Vol. No. 12

December 1999

Frontiers of Technology

Reservoir Engineering: Augmented Recovery

In 1888, roustabout James Dinsmoor was working on several Third Venango sand wells on the William Hill property in Venango County, Pennsylvania. On an adjoining property, a Third Venango sand well was being deepened by its operator to tap the Speechley sand to obtain natural gas for use on the lease. The operator found a considerable amount of gas in the Speechley but did not have pipe available. So the well was shut in temporarily to save the gas.

Dinsmoor observed that the three Third Venango sand wells on the William Hill property immediately experienced an improvement in oil production and that this increase was maintained until the nearby Speechley well was completed by its operator. Following the well’s completion, Dinsmoor noticed that the Venango wells returned to their previous production levels. This accidental repressuring represents the first known application of augmented recovery in the U.S.¹

Additional instances of augmented-recovery methods occurred in the late 1870s or early 1880s when other early well completions began to lose formation pressure and production started to wane. Attempts by early operators to augment production included the application of gas (vacuum) pumps to faltering wells. This practice established an increased pressure differential between the reservoir and wellbore, increased the rate of fluid production slightly, and extended the wells’ producing lifespan.²

During the ensuing period of approximately 119 years, a variety of augmented- recovery methods have been developed and introduced. These methods include waterflooding, gas injection, chemical flooding, and thermal recovery. Some methods, such as waterflooding and gas injection, use materials native to the reservoir to either replace or augment natural drive forces. However, they do not alter any of the fundamental factors that act to retain the oil within the reservoir. Other techniques, such as chemical flooding and thermal recovery, use more dramatic methods to overcome forces, such as surface tension and viscosity, that inhibit the flow of oil from the formation.

Waterflooding

As early as 1880, observers of oil producing operations concluded that water could be an effective method for driving oil flow within the formation. Most of these observations resulted from the accidental intercommunication of natural forces under favorable circumstances that resulted in increased production.

One of the first documented waterfloods was in Pennsylvania’s Bradford field. The accidental flooding of the field is thought to have begun in 1905, six years after the field began production. Flooding continued for 15 years. During this time, production rates trended upward. Most operators credited the production increase to the accidental flooding, and, even though it was illegal, operators in both the U.S. and Canada instituted intentional field waterfloods. While the unintentional flooding of fields is well-documented, data on intentional waterfloods by operators prior to 1921 (the date waterflooding was legalized) are sketchy. Still, evidence from that time suggests that intentional waterflooding occurred as early as 1875.

The earliest waterfloods were called “circle floods” because of the growth pattern of the water-invaded zone. As nearby wells were watered out, they too became injection wells in order to continue the extension of the area of waterflooding. The rate of advance of the water, diminishing with time and cumulative injection, prompted one operator to convert a series of wells simultaneously to form a line drive, a technique that increased oil production rates even more.

The first five-spot-pattern waterflood was attempted in 1924 on a tract in the southern part of the Bradford field. Frank Haskell is credited with the idea, but Arthur Yahn receives credit for the technique’s first successful deployment. Haskell’s attempt failed to produce a speedy response because of the 500-ft (152-m) distance between like wells. Yahn’s 190-ft (58-m) distance between wells produced a much quicker response. Initially, only surface water entered the wells, but, in late 1929, a pressure plant was installed to increase the rate of injection. Also, the five-spot pattern required that wells be reworked to achieve replacement of prior withdrawals in reasonable time frames, and later five-spot well spacing varied, depending on formation permeability. The technique gained widespread acceptance by 1937.

However, operators were slow to extend waterflood activities outside Pennsylvania due to the economic conditions of 1929 and 1930. However, in 1931, the Carter Oil Co. initiated a pilot flood in the shallow Bartlesville sand of Oklahoma. Soon, others followed; and all enjoyed favorable results. In early 1936, waterflooding operations were extended to the shallow sands of the Fry Pool in Brown County, Texas. However, results were marginal, and it wasn’t until Magnolia Petroleum Co. initiated the West Burkburnett flood in 1944 that outstanding results were achieved. Operations soon followed in other states between 1944 and 1949.

During this expansion period, engineers became aware of the advantages of pressure control by reinjecting produced water in natural-waterdrive fields. In 1936, the East Texas field was the site of initial experiments involving reservoir-pressure control as a result of the disposal of produced water in natural-waterdrive fields. Earlier analytical studies of the reservoir by Ralph J. Schilthuis and William Hurst led to the conclusion that, as the reservoir pressure declined, salt water contained in the aquifer of the Woodbine sand expanded and encroached into the oil reservoir. Depending on the rate of production, the water sustained an equilibrium level of reservoir pressure that was interdependent with the rate of production.³ After several years of observation, the program was declared a success and was expanded to other fields. The field’s natural waterdrive resulted in a recovery factor that has, to date, exceeded 70%.

Gas Injection

Interest in gas injection continued during the days following the turn of the century. In gas injection, compressors are used to force the air or natural gas through injection wells drilled for that purpose or through old wells taken off production and used as “key wells.” Gas injection has four objectives: to maintain or to restore formation pressure, to act as a drive mechanism, or to place the gas in storage until it is needed.⁴

In August 1911, I.L. Dunn successfully demonstrated that repressuring a reservoir by injecting gas could increase oil production. Dunn based his experiments on an idea he had gained in Ohio in 1903 when gas, at a pressure of 45 psi, was forced into an oil well producing from a 500-ft (152-m)-deep sand. According to Dunn, “After 10 days the gas pressure was released and the well began to pump much oil, which continued until the gas had worked out again.”

To demonstrate his technique, Dunn conducted a series of experiments in which 150,000 ft3 of free air was compressed and forced into one well daily at a pressure of 40 psi. Within a week, the production of surrounding wells increased, after which the compressed-air method was extended to other parts of the property.⁵ As a rule, the application of air resulted in a three- to four-fold increase in the production rate, and the use of air proved more economical than natural gas.

In 1927, Marland Oil Co.’s Seal Beach field in California was the site of the first use of higher injection pressures. The higher pressure was necessary because of the hydrostatic pressure exerted by high-head edge water. Pressures as high as 1,800 psi were required to force the gas into the sand; however, after the gas was flowing into the formation, the pressure never exceeded 1,500 psi. A year later, 173 million ft³ of gas had been injected into the Bixby sand. As a result, production increases as high as 50% were obtained in wells upstructure from the injection wells, with little increase downstructure.

According to historians, annual U.S. production resulting from gas-injection projects reached a peak in 1935, maintained a constant level until 1945, and then increased rapidly to 1952. Annual production from gas-injection projects is estimated to have reached 212 million bbl in 1955.

Early Field Successes

Early gas- and water-injection projects were designed and implemented with empirical methods. Fundamental scientific understanding of fluid flow began in the 1930s, and breakthrough understanding of fluid displacement can be attributed to M.C. Leverett and his collaborators. This understanding established the foundation for engineered design and prediction of waterflood performance, including the method for layered reservoirs. It also set the stage for large-scale applications in some of the biggest fields of the period.

Magnolia Petroleum Co.’s West Burkburnett field in Texas is an example of the success of a large waterflood secondary-recovery operation. Using a five-spot water-injection program, the company’s flood recovered 9 million bbl of oil between 1944 and 1953, or 1.4 times that recovered by primary methods.

In Illinois, an initial five-spot program by Adams Oil & Gas Co.-Felmont Corp. in 1943 in the Patoka field, Marion County, resulted in a much more rapid response in the oil production rate than expected. The field, which produced 2.8 million bbl of oil by primary production, produced an additional 6.4 million bbl of waterflood oil by August 1960.

In the Bradford field of Pennsylvania, results from an air-injection project were quickly realized. The daily injection of air at an average of 68,600 ft3/well at 300 psi increased per-well production from 0.25 to 12 BOPD in less than two months. Annual production from the 22-well project increased from 3,474 bbl in 1925 to 18,524 bbl in 1927, and total production from the field is estimated to have increased by 25% from the air injection alone.

Enhancing Displacement Efficiency Miscible-Gas Injection

The first use of miscible-gas injection by the petroleum industry occurred in the 1950s as a result of a search for a miscibility process that would recover oil effectively during secondary and tertiary production. The injection fluids used include liquefied petroleum gases (LPGs), such as propane, high-pressure methane, methane enriched with hydrocarbons, and high-pressure nitrogen and carbon dioxide (alone or followed by water). All of these are effective in displacing trapped reservoir oil, but applications are dependent on the field and a variety of economic considerations, such as the commercial marketability of these products individually.

The injection of products other than air or water into formations to encourage oil to flow to the wellbore actually began as early as 1927. At that time, the Midwest Refining Co. injected surplus liquefied-gas products into the secondary-gas-cap area of the Salt Creek First and Second Wall Creek reservoirs. While they planned to improve gas-drive operations, they were not aware of the enriched-gas-drive mechanism that was used later.

Block 31 Field

Started in 1949, the oldest active operation involving miscible-gas injection is Atlantic Richfield’s (Arco) Block 31 field in Crane County, Texas. It began as a miscible-hydrocarbon-gas injection and is still in operation as a mixed hydrocarbon/nitrogen-injection project.

“The Block 31 field was discovered in 1947,” states Ben Caudle, a former Arco employee and now a professor of petroleum engineering at the U. of Texas, “and the light crude in the field had a high shrinkage factor. Our forecasted recovery rate was determined to be ±10% at best, so we began looking for a method of getting the oil out in quantities that would make the economics work.”

Barney Wharton, at Arco’s research center, had an idea. “Barney suggested that in order to handle the shrinkage, we should inject readily available natural gas into the formation at high pressures. The gas would then evaporate the light-ends liquids before we recovered the oil at the surface,” Caudle recalls.

The Texas Railroad Commission, Texas’ oil and gas governing body, was contacted to obtain approval for the new, untried recovery method for the reservoir. They approved the procedure and also exempted the field from the mandatory shut-in period in effect at the time. “In those days, we were on production allowables,” says Caudle. “Wells couldn’t be produced more than 9, 10, maybe 11 days a month because of the abundance of available crude,” Caudle says.

“Next, we conducted experiments in the laboratory to determine the most appropriate method of applying our idea. Nobody was more surprised than we were when, halfway through our experiment, it dawned on us that the injected gas was becoming miscible due to the high pressure. The multiple contacts of the gas in the pore spaces was drawing the gas and oil closer together until it formed a miscible slug that drove the oil ahead of it,” explains Caudle. “We put our method into operation and, as a result, Block 31 became the industry’s first miscible-gas oil-recovery operation.” Later on, it was determined that nitrogen gas could be substituted for the natural gas. As a result of the miscible-gas-injection project, the field’s recovery factor has reached approximately 70%. The field is still in operation.

During the 1970s, the U.S. industry switched largely to carbon dioxide gas that was available near major west Texas fields. It achieves miscible displacement at low pressures, has a greater viscosity under pressure than many other gases, and is less costly than LPGs or methane.⁶ Hydrocarbon gases continue to be used widely in Alaska and elsewhere in the world.

Chemical Flooding

During 1936–1937, alcohol was injected into oil wells to displace capillary-held water from the near-wellbore region, where it offered the most severe restriction to oil flow. Later, when oil production from the well was resumed, the operator anticipated he could remove the alcohol surrounding the wellbore and realize an increase in oil productivity.⁷ His activities marked one of the earliest uses of a chemical to displace oil in a formation.

Natural-drive fluids, like water and gas, leave a large supply of oil behind in the reservoir under the best of conditions. Because they are immiscible with formation oil, the oil resists being displaced from the rock pores. Also, these fluids have densities and mobilities that are incompatible with oil. Chemical flooding adds chemicals to the water in order to overcome these problems. Three types of chemical floods are used: polymer, micellar/polymer, and micellar/alkaline floods.

Polymer flooding is a type of chemical flood in which long, chainlike, high-weight molecules are used to increase the viscosity of injected water. This improves the mobility ratio of injected water to reservoir oil, resulting in a more effective displacement process. Micellar and alkaline floods are two methods used that result in improved microscopic oil displacement by reduction of oil/water interfacial tensions. In alkaline floods, the injected material reacts with naturally occurring acidic oil components to form a surfactant. Micellar/polymer flooding is a two-part recovery technique in which a surfactant/ water solution is injected to reduce oil/water interfacial tensions, resulting in improved microscopic oil-displacement efficiency. This is followed by polymer-thickened water to push the oil and surfactant slug toward the producing wells.

The popularity of chemical flooding reached a peak during the 1970s when research projects abounded, and a large number of field tests were conducted in the 1980s. Chemical-flood field tests reached a peak of 206 in 1986.⁸ However, the oil-price collapse that same year interrupted the popularity of this recovery method when it presented operators with a fundamental dilemma linked to economics. Since the cost of the materials is generally linked to the cost of petroleum, a vicious cost spiral ensued as oil prices collapsed. Since the mid-1980s, oil prices have seen some growth; however, this improvement has not been sufficient to reignite the interest seen in the 1970s and 1980s. Proponents of chemical floods are seeking ways of breaking this economic linkage by generating surfactants from nonhydrocarbon feedstocks.

Thermal Recovery Steamflooding

Of the two main methods of thermal recovery—steam injection and in-situ combustion—the injection of hot fluids, such as steam, into the reservoir is the oldest and most controllable method. Flooding with heated water, steam, or superheated steam has been around almost as long as conventional waterflooding. In fact, the idea for using heated fluid supplied from the surface can be traced back to a proposal by B.W. Lindsly in 1928.⁹

To date, most thermal-recovery work has been accomplished with steam. It has found significant application in many parts of the world, including the U.S., Venezuela, Canada, Germany, Russia, China, and Indonesia (the site of the largest steamflood).

Hot-fluid-injection theory is simple; heated water or steam is generated on the surface and introduced into the formation through injection wells. The heat serves to lower the oil viscosity in the formation, allowing it to flow more easily to the producers. Steam is preferred because it is much more efficient at delivering thermal energy to the reservoir because of the latent heat of vaporization. Although steamflooding can be effective in both light and heavy oils, it is used predominantly in heavy oils.

In-Situ Combustion

Purposeful underground combustion started in Russia around 1933 (unintended combustion had occurred previously during some air-injection projects). This early project was conducted in a pressure-depleted reservoir containing 36°API oil. Since then, in-situ combustion projects have been implemented in many locations including Romania, Canada, the U.S., and India

In-situ combustion generates heat in a reservoir through the introduction of air into the reservoir, after which a fire is ignited in the formation near an injection well. The fire and airflow move simultaneously toward the production wells. This forward-combustion method uses the injected air and vaporized formation water as the heat carrier (dry combustion) or may combine air and water injection (wet combustion) to increase process efficiency. A seldom-used variant—reverse combustion—allows the fire to move from the production well toward the injection well(s), and the fire “flows” counter to the flow of injected air.

The first known attempts to apply an in-situ combustion process in the U.S. occurred in 1952, when Magnolia and Sinclair engineers, each group working independently only 300 miles apart in Oklahoma, initiated movements of combustion fronts in pattern-type experiments after several years of laboratory testing.10 Reports of these two experiments provided the impetus for additional research in other laboratories. Later on in the 1950s, General Petroleum Corp. and Magnolia Oil Co. generated a cooperative underground combustion-field-test effort supported by ¹⁰ other oil companies in the South Belridge field in California.

Simultaneously, thermal-recovery techniques also were beginning to be applied in other areas of the world. One of these areas was Venezuela.

“The use of steam (cyclic steaming and steamflooding) began in the oil fields of Venezuela in the late 1950s and was in routine use in eastern and western Venezuelan oil fields by the mid-1960s,” says Tom Reid, a former Phillips Petroleum Co. engineer who now works for the U.S. Dept. of Energy. “I took steam to the Morichal field in eastern Venezuela in 1964 because our management was convinced that the economics of steaming high-sulfur, heavy oil, which had been successfully substantiated by operators in California, could be used in Phillips’ high-sulfur, heavy-crude operations in Venezuela,” he explains.

“During the 1960s, cyclic-steam, steam-drive, cyclic-hot-water and hot-water drive raised production to over 100,000 B/D until the latter part of that decade when the company cut production back to 25,000 B/D to match the needs of its refinery in England that was processing the heavy crude into asphalt for paving roads,” he continues.

“During these steam operations, we learned a lot,” Reid says. “We received a couple of surprises when the first well was steamed. One of these surprises was the production of hydrogen that occurred when the high-temperature steam reached the reservoir sands. We traced this produced hydrogen in the offset producers, and it indicated the direction of the steam front. Also, the cyclic-hot-water and hot-water drive we used produced excellent responses in these virgin reservoirs.”

Reid also comments on the use of fireflooding in the Morichal field. “It wasn’t successful,” he states candidly. “Counterflow combustion wasn’t possible because of the occurrence of spontaneous ignition, and conventional direct-drive fireflooding failed because of the low structure (flatness) of the reservoirs. The nitrogen and carbon dioxide produced by the fireflood ‘gassed out’ distant producers, thereby reducing the overall field production.”

The Future

Because of the wide dissimilarity in rock and fluid properties, combined primary and secondary reservoir recovery factors of 30 to 40% of original oil in place are considered good. This leaves 60 to 70% of the original oil in place as the target for augmented-recovery methods. Since the injection of any fluid into a reservoir is expensive, either low-cost injectants or high increased recovery factors are needed to ensure good economics.

Looking at the future of augmented recovery, it is readily apparent that water is the cheapest fluid and that improvements in waterflood technology are still occurring as a result of better control of fluids in the formation. Also, thermal techniques continue to offer distinct advantages in that the elevated temperatures reduce viscosity and heat flow is advantageous to sweep efficiency. On the other hand, it is also apparent that chemical injectants can be very expensive; therefore, their application is more limited. Finally, there is a need for carbon dioxide sequestration to help reduce the CO2 released to the atmosphere from the combustion of hydrocarbons. This could lead to a double dividend from an environmental benefit and from improved oil production.

Based on these observations, it is important that reservoir engineers plan the augmented phases of oil production at the same time they plan the primary phase. This will ensure that future recovery factors of original oil in place are maximized during the life of the field.

References

Lewis, James A.: “Chapter 13: Fluid Injection,” History of Petroleum Engineering, API, New York City (1961) 849–873.
Lewis, James A.: “Chapter 13: Fluid Injection,” History of Petroleum Engineering, API, New York City (1961) 849–873.
Lewis, James A.: “Chapter 13: Fluid Injection,” History of Petroleum Engineering, API, New York City (1961) 849–873.
Miller, H.C.: Function of Natural Gas in the Production of Oil, API, New York City (1929) 144.
Lewis, James A.: “Chapter 13: Fluid Injection,” History of Petroleum Engineering, API, New York City (1961) 855.
Fundamentals of Petroleum, third edition, Petroleum Extension Service, U. of Texas, (1986) 203.
Lewis, James A.: “Chapter 13: Fluid Injection,” History of Petroleum Engineering, API, New York City (1961) 895.
“EOR Production Up Slightly,” Oil & Gas J., Pennwell Publishing Co., Tulsa, OK (20 April 1998) 49–56.
Lindsly, B.W.: “Oil Recovery by Use of Heated Gas,” Oil & Gas J., Pennwell Publishing Co., Tulsa, OK (20 Dec. 1928) 26–32.
Nelson, T.W. and McNiel, J.S., Jr.: “Past, Present and Future Development in Oil Recovery by Thermal Methods,” Petroleum Engineer, Petroleum Engineer Publishing Co., Dallas, TX (Feb. 1959) Part 1.