US2642943A - Oil recovery process - Google Patents

Description

Patented June 23, 1953 OIL RECOVERY PROCESS Robert L. Smith, Western Springs, and Kenneth M. Watson, Madison, Wis., assignors, by mesne assignments, to Sinclair Oil and Gas Company, Tulsa, Okla, a corporation of Maine No Drawing. Application May 20, 1949, Serial No. 94,506

3 Claims.

This invention relates to improvements in the production of oil and gas from oil-bearing formations, and more particularly to an improved method for recovering gas and oil from partially depleted oil-bearing formations by thermal means.

It is well known that the recovery of oil from producing fields is in general an incomplete process. The production of oil from subterranean oil-bearing formations requires available energy and its efficient expenditure. Where natural sources of energy such as high formation gas pressure or natural water drive have existed and have been coupled with efficient production methods, recoveries have been said to run in rare instances as high as 70% to 80% of the original oil content. In general, however, it is a matter of common knowledge that recoveries of oil have rarely exceeded 25% to 40% of the original oil content, and that vast quantities of oil remain in the produced fields of this countary and the World, even though some have been produced steadily for 25 to 40 years. For example, the Nowata County field of Oklahoma is a relatively shallow field underlain by a portion of the Bartlesville sand averaging about 30 feet in thickness at a depth approximating 600 to 800 feet. This field has been in production for about 40 years, and yet it is estimated that only 44% to 46% of the original oil content has been recovered, say about 17% by primary methods including vacuum, 17% to 19% by gas and air pressuring, and say by water flooding where this has been applied.

Since the very early days of production of oil from subterranean formations, thermal methods for improving recoveries by lowering viscosity and through distillation and cracking have been proposed. Despite the theoretical soundness of some of these proposals these methods have proved characteristically impractical because of the high heat consumption required to improve recovery, for these methods have depended upon heating the entire underlying oil-bearing stratum to the temperature at which the desired improvement in flow would develop, and often have required combustion of substantial proportions of the recoverable oil to Warrant consideration at all.

In contrast to these methods, we have now discovered a method of recovery embodying the advantages of the old thermal schemes but which is characterized by propagation of a high temperature zone coupled with an approximately coincident internal combustion wave front within the oil-bearing formation. The high temperature zone is expanded in a relatively horizontal direction within the formation in such a manner as to sweep out residual oil by force of a heat transmissive gas flow. The combustion wave front is continuously or intermittently generated within the formation and maintained in an approximately coincident relation to the high temperature zone or heat transfer wave so as to continuously or intermittently revivify the heat content of the expanding hot zone.

Thus we propagate a high temperature zone within the formation by forcing hot combustion gases into the formation from an inlet well under elevated pressure for a period of time sufiicient to heat up the surrounding sand structure and produce gas and/or oil flow from one or more nearby outlet wells. The hot gases preferably derive from bottom-hole burning, that is, from the burning of fuel at high rates of heat release at formation level in the input well. When a high temperature zone surrounding the inlet well for a considerable distance has been established, fuel burning is discontinued, and the movement of the heat zone horizontally out into the formation is continued by cycling cold gases through the inlet well.

We have found that a small percentage of asphaltic or carbonaceous material usually remains in the porous structure following passage of the hot gases. For example, purging of an oil sand having a porosity of about 20% and an oil and connate water content of about 2.25% each with inert hot gases at temperatures ranging up to about 1350 F. leaves about 0.5% residual carbon on the sand. This material constitutes a non-recoverable residue, but we have found that it can be ignited and is burned at elevated temperature, say upwards of 450 F. to 500 F., by flow of air or oxygen-containing gas so as to regenerate or revivify the heat transfer zone in compensation for the dissipation of its heat values resulting during its progress out into the formation.

We have found that this residual carbon represents a valuable fuel for improving thermal recovery if it is burned in such a way as to add its heat values to the heat values put into the formation by burning start-up fuel. For we have found that arelatively high temperature, say above about 700 F., is needed to effect practical improvements in oil recoveries after ordinary methods of secondary recovery as by water drive or gas repressuring have reached a point of diminishing return. Yet the dissipation of heat within the vast underground volume of the formation as the ring of the heat transfer zone is expanded from the input well is large even with closely situated output wells. We have found, however, that we can continuously or intermittently revivify the degenerating heat transfer wave by burning the residual carbon left in the sand in a relatively narrow band or moving combustion wave that approximately coincides with the moving heat transfer wave in point of position and rate of movement. For we have discovered that the rate of movement of the combustion wave relative to that of the heat transfer wave primarily depends upon the oxygen content of the driving gases and the residual carbon content of the sands, and that we can obtain the desired coincidence in rate by limiting the oxygen content to a small proportion. It is important that the relative movement of the peak hot zone and the combustion wave be controlled because otherwise the heat values of each will be independently given up to the structure. Thus if the combustion Wave lags behind the heat transfer wave, the heat generated primarily goes into elevating the temperature of that portion of the formation which has been already swept, and the temperature peak is reduced. If the combustion wave ranges too far ahead of the heat transfer wave, the peak again falls off and finally combustion is lost in the too rich and too cold environment of the outer fringe.

In one method of operating according to our invention, therefore, we include in the cold cycle gas drive up to about 5% to 6% of oxygen. The

heat transfer wave is continuously projected out into the formation so as to sweep out the oil and is continuously revivified by the recuperative capacities of the cycling gases in picking up preheat from the swept out portions of the formation and then igniting and burning the residual carbon in the sands in the approximate zone of peak temperature.

We have also discovered that we can intermittently revivify the heat transfer wave by cycling 1 produced gas from outlet Wells or other gas of low oxygen content to advance the heat zone until heat losses reduce the frontal temperature to a point somewhat above the ignition temperature of the carbonaceous residue in the sands. At this point we revivify the heat front by introducing sufficient air or other oxygen-containing gas with the cycling gases to reinitiate combustion or raise the combustion level of residual carbonaceous matter in the zone of high temperature so as to replenish the heat content of the frontal area. Large bands of unburned carbon may be left behind, if desired, to avoid overheating the structure in this cyclical method of operation. Thus by skipping bands of residue, control is bad on the average per cent residue burned, giving an additional measure of control on the rate of the combustion wave.

Thus, according to our invention, the cold cycling gases entering the inlet well are heated as they move through the zone of previously heated sand, leaving cooler sands behind. Heat is thereby recovered and the heat front is simultaneously advanced, for the regenerated heat is transmitted to the cold sand in front of the previous hot zone. In this manner the high temperature Wave is continuously moved forward, and a heat wave is created within the formation by the heat transmissive flow of the hot gases followed by the heat regenerative flow of the cold cycle gases which give up the regenerated heat to the structure in advance of the moving wave front. Since the non-recoverable carbon residue is left behind the advancing front, it may be burned in a relatively narrow band, forming a combustion wave behind the moving heat wave so that a minimum of recoverable oil is consumed. As the frontal zone progresses in distance from the inlet well, the flow rate may be increased. The flow rate requirement may be reduced by a limited use of water injection.

We have found that good recovery is effected at operating temperatures within the range of approx mately 700 F. to 1000 F. Combustion of the residual matter will occur as low as 450 F. to 500 F. but it is Weak, and even at 600 F., the burning band tends to be too broad in area and slow moving for good revivification. In addition, mild cracking begins at about 700 F., reducing the gravity somewhat. Accordingly, we prefer to maintain a temperature above about 700 F., but under the fusion or sintering temperature of the structure, which is ordinarily somewhat above 2000 F. to 2500 F. Ordinarily, incipient sintering below the fusion temperature does not adversely affect permeability. Oil recovery improves with temperature. For example, the percentage of oil produced from Nowata oil sands of about 20% porosity and 2.25 weight per cent oil content approximates 50% of the oil present at 900 F. under an inert gas flow of about 275 cubic feet, per barrel volume of sand per hour while at 1350 F. the recovery approximates Similarly, the desorption of connate water increases with temperature and rate of flow, although the percentage produced is of a significantly lower order under the conditions described.

We have also found surprisingly enough by physiochemical analysis of typical oil sands that, although most of the material is crystalline quartz, as high as 10% or more is porous, high area adsorptive clay, and that substantially all of the oil and water is associated with this portion. It may be that the primary function of the elevated temperature, 1. e. upwards of about 700 F., in recovering oil from exhausted oil sands is to break an adsorption complex between a relatively adsorptive matrix for the crystalline sand, such as kaolin, and water and oil rather than merely to promote flow by cracking and distillation.

Of course, the before-mentioned methods of operating require the presence of sulficient residual carbonaceous matter in the sands to provide fuel for revivification, which is usually the case (or depend upon building up a sufficiently large body of heated sand in the formation). In those other cases, e. g., fields of good permeability bearing high gravity crudes, we intermittently introduce fuel gas along with the oxygen-containing gas. The mixture does not ignite until it contacts the zone of high temperature, since the sands between the frontal zone and input well are cooled by prior cold cycling. Thus this method of producing lean formations requires the preceding cycling of cold gas to produce a sizable area of cooled sand, otherwise burn-back of the incoming mixture to the well will occur with resulting heating of the entire structure and prohibitive expenditure of heat values as well as excessive pressure drop.

As a result of applying the methods of our invention, a moving region of high oil saturation is set up ahead of the moving heat front by reason of the greater susceptibility to flow induced by the thermal effects of cracking, vaporization, and viscosity reduction assisted by the drag effect of the flowing gas. By cooling and condensation, the ases and oil give up heat to the cooler sands in front of the hot zone so that the progression of the high temperature front is maintained. As the region of oil saturation approaches an outlet well, oil under gas pressure flows or is pumped. from the Well.

Since the strata overlying and underlying oilbearing formations are impervious while the formation is relatively permeable, there are no heat losses by convection. Losses by conduction are usually small because of the low thermal conduc-.

tivities of shale and other materials commonly found as bounding strata. Moreover, some of these heat losses will be returned to the formation and picked up again by the cold gas flow. However, suficient gas drive must be maintained to prevent immobilization of the front, or conduc tion losses may become prohibitive.

It should be observed that the travel of the hot zone radially outward does not in itself ordinarily result in a temperature decrease. Although the vertical area of the zone broadens, the heated mass remains the same. The sharpness and height of the temperature profile then remains the same if the sand is uniformly permeable and if the actual heat losses are exactly compensated for by combustion of deposit. The most important effect in causing lowering of temperature and broadening of the mass of the heated zone is non-uniform permeability. That is, both Waves move rapidly through a highly permeable streak, hence heating it up before the adjacent and less permeable streaks but losing heat to them by con' duction, thus lowering the temperature level and increasing the mass of the heated zone. For this reason it is important to keep both waves as close together as possible.

In practicing our invention the high temperature front is propagated within the formation by burning fuel, conveniently natural gas or crude oil, with air at high temperature either on the surface, or preferably within the hole. The resuiting combustion gases under elevated pressure are forced into the porous oil-bearing stratum for a length of time sufficient to raise the temperature of a large body of sand surrounding the well to a temperature below the fusion temperature of the structure but well above the ignition temperature of the residual carbon. At the same time, the formation is partially rte-pressured with resulting now of cold oil and gas from one or more outlet Wells.

During the fuel burning period, there is normally a large excess of oxygen in the combustion continuous and uniform combustion. However, a jet-type, spark ignited combustion system with a high velocity, turbulent air and gas flow in which a relatively large volume of diluent air gas is introduced progressively into the burning zone provides perhaps the most convenient and reliable bottom-hole system. Recycle gas may be utilized as diluent after the temperature has been built up, and Where gas compressors are employed, the compressor exhaust may be utilized as recycle or diluent gas.

The inlet pressures will vary according to the distance between producing and input Wells, the thickness and permeability of the oil-bearing stratum, and the oil and water content of the formation. Similarly, the quantity of gas introduced will be affected by the desired pressure. temperature of the input gases, and the con ditions and heat capacity of the oil-bearing stratum. The pressure for example .Will ordinarily exceed 60 p. s. i. g., but because of problemsof reservoir control, is maintained at a moderate figure, and, of course, is ultimately limited by the over-burden. To minimize plugging or cementing of the porosity of the formation, after-scrubbers or other filtering devices should be employed on compressors in order to remove iron rust or other troublesome carry over. Similarly, it is desirable to incorporate similar devices in the liquid knock-out system employed for handling recycle gas The gas produced is recovered and after recompression to compensate for pressure drop is recycled. Maintaining superatmospheric pressure on producing Wells markedly reduces power costs for recompression in recycling operations. Depending upon its fuel content, the produced gas may be burned as fuel or utilized as flue gas recycle. Naturally, readily liqueriable or other valuable components may be recovered as by ab sorption prior to utilization as recycle.

Pressure drop is desirably kept low and may be controlled by keeping the volume gas rate low and by operating at a high pressure level. Use of water as the principal heat carrying medium serves to keep the volume rate and pressure drop low except in the heated zone, but may require emulsion breaking to avoid plugging troubles. The rate of oil production ordinarily should be high to minimize the proportion of labor and capital expense, although this consideration con= gases because of the use of dilution air in the burner system, which insures clean combustion and prevents the formation of soot that might clog the structure, and additionally assists in heating up the structure by burning carbonaceous residue and some oil.

The use of a bottom-hole burner is advantageous because it eliminates the need for expensive fiicts with most other considerations except that of heat losses to the bounding strata.

When a high temperature, say about 1500 F. but advantageously in the range of about 1000" F. to 2000 F., has been established over the area surrounding the inlet Well, we discontinue the input of hot combustion gases, but continue gas flow by increasing recycle, or admixture of air with combustion gases produced from outlet wells. The cold gases are heated by moving through the zone of previously heated sand, at the same time cooling the sand in back of the previous hot zone. The heated gas and oil vapors at the same time heat the sands in front of the previous hot zone, thus maintaining a heat transfer Wave. With oxygen-containing gas drive, the combustion wave is established by burning residual carbon under the temperature prevailing in the heat transfer wave, and the combustion. Wave is maintained in the zone of peak temperature by regulating rate through control of oxygen content and the rate of gas flow. For example, We consider a total gas input of about 500,000 standard cubic feet per day per well and an oxygen content of from about 1.33 to about 5.0 per cent particularly desirable. With nonoxygen containing gas drive, the temperature in the peak zone is intermittently raised by introducing cold air into the circulating gas stream as it falls off.

Residual carbonaceous matter in the sands burns in a relatively narrow zone determined by depletion of the oxygen at the front and the combustible at the back. Thus this burning zone also travels forward as a wave, and the combustion raises the temperature of and furnishes additional heat for the moving hot zone. Although the flowing air-gas mixture is preheated by the sand, and the sand is preheated by the combustion products, the problem is usually one of effectively controlling the relative rates of movement so that the heat values of the carbon burning wave are effectively added to the heat values of the heat transfer wave. Where exceptionally high carbon residues occur, however, it may be necessary to provide some means of control to I prevent the temperature from rising too high, The loss of some heat to overlying and underlying strata of course is helpful, but again the principal means of control is varying the ratio of heat carrying medium to oxygen in the stream. Liquid water intermittently slugged into the formation may be utilized to supplement the recirculated flue gases for this purpose. The peak temperatures may also be limited by deliberately varying the rates of travel of the waves, and by reducing the rates of wave travel.

Gas analysis of the produced gas, as by the Great method, for oxygen and carbon dioxide content provides a means for determining the state and progress of the front within the formation. During the combustion cycle oxygen content decreases and combustion products increase, while during the cold gas cycle the reverse is true. It is helpful to observe pressure differentials between inlet and outlet flow, which are affected by the temperature, the permeability, whether fusion is occurring, whether an oil and water block is building up, and the position of the heat transfer point and the revivifying combustion point. Control timing is also assisted by the use of deep-well thermometers or temperature recording devices.

For optimum recovery in an economical process, we consider that the rate of travel of the heat transfer hot wave should be almost coincident with that of the revivifying combustion hot wave, in order to minimize losses to sand and bounding strata. If the rate of travel of the combustion wave is appreciably either faster or slower than that of the heat transfer wave, the whole mass of sand between the front of one and the back of the other will be heated to high temperature with consequent losses to sensible heat that may be prohibitive, or so large a body of sand may be heated that the temperature may be too low. Rate of travel of the heat transfer wave is proportional to the ratio, mass velocity of total gas divided by sand density, and to the ratio, specific heat of gas to specific heat of sand. Rate of travel of the combustion wave is proportional to the mass velocity of oxygen consumed (or oxygen supplied less oxygen out), and is inversely proportional to the carbonaceous residue of the sand (at a given temperature level, reactivity, etc.) which is burned.

In applying our invention to large scale recovery operations, it is advantageous to utilize a logically spaced pattern of input and outlet wells.

In many of the fields which have been extensively gas pressured or water flooded. wells have been drilled in 5-spot or 9-spot patterns which will be suitable for applying our invention to further recovery. It may be necessary, however, to drill a new input well or to pull the old casing and replace it with pressure tight piping. The holes may be tightly cemented with a high temperature resistant cement at the top of the formation where necessary to confine the combustion and recycle gases within the formation stratum. Generally an ultimate 1:1 ratio of input and outlet wells will be found advantageous. In planning recovery patterns, it is particularly advantageous to utilize water dams built up in the structure by water injection through selected surrounding wells in order to segregate the working area and retain high pressure.

One of the major problems in secondary recover through gas pressuring is non-uniformity in the porous structure of the formation both vertically and horizontally, so that channeling occurs and control of oil movement within the formation is made diflicult; Non-uniform permeability creates the further problem with our invention of tending to promote broadening of individual hot zones and lowering of temperature. Core analysis is a helpful guide to methods for overcoming channeling by indicating the desirability of well packing at levels where fissures or excessively permeable strata appear. For this purpose, high temperature resistant cements may be used in place of the usual rubber packing devices of low temperature pressuring. The cement is poured into place and solidified. The casing is perforated as by shooting at the desired input levels. In order to prevent horizontal channeling through a relatively narrow streak of relatively high permeability so as to defeat uniform and economic recovery, higher back pressures on the more productive wells may be maintained to increase the resistance to flow in those directions. In addition, well blocking may be resorted to, as by water injection in or mechanical plugging of peripheral wells or wells of quick producing tendency.

The principles underlying our invention will be illustrated in the following examples, which, however, are intended to be merely illustrative and not limiting with respect to conditions and means utilized. Thus no attempt at recycling was made in the pilot scale testing, but air was utilized as the pressuring medium, and control was exercised through variation in the flow rate.

EXAMPLE I We tested the soundness of our concept of developing both a moving high temperature wave and a combustion wave in oil-bearing sand by means of heat transmissive and heat regenerative drive in a trial in which 40 inches of sand having an oil, content of 10% were packed in a 4- inch steel pipe. The test unit consisted of a 4- foot section of 4-inch steel pipe insulated with magnesia, which was maintained in a vertical position with an oil drain-off located at the bottom. A burner was used to preheat the simulated oil sands and was located at the upper end of the pipe, and an arrangement of thermocouples and an automatic temperature recorder were used to follow the movement of the heat zone through the simulated oil sands.

The pipe was equipped with a 1%; inch 0. D. stainless steel thermowell permitting the insertion of a bundle of eight thermocouples, staggered l inch apart, into the well. This thermowell entered the bottom of the pipe through the bottom oil take-off and extended. upward through the center of the pipe to within 1 /2 inches of the tip of the burner which was positioned 3 inches above the top of the sand bed. A fast-acting temperature indicator having a F. to 1600 F. range was used to record the temperature. In order to determine the flame temperature of the burner, a thermocouple was inserted into the thermowell to measure the flame temperature 1 /2 inches below the burner tip. A portable potentiometer was used to measure this temperature,

The pipe was; charged with enough gravel to make a 5-inch layer in the bottom. A 25-mesh stainless steel screen was placed over the gravel, followed by a layer of glass wool. The oil sand charged to the unit was prepared by mixing thoroughly 30.3 lbs. of builders sand, 3.5 lbs. of Santa Barbara 50% reduced crude and 0.15 lb. of coke ground to 100 mesh and finer. The coke was added to the oil sand mixture to simulate the entrapped oil and carbon in a natural oil sand formation. The amount of coke added was equivalent to 0.5 of 1% based on sand. This entire mixture was packed firmly into the 4-inch pipe by tamping during filling. The total depth or the sand bed was 40.5 inches and the top of the bed was located 3 inches below the tip of the gas burner. A wet test meter was used to measure outlet gases,

To start the test the burner was lighted and adjusted visually to give a hot flame and then inserted into the unit. Prior to inserting the lighted burner, the sand bed was air blown to determine whether or not any oil could be removed by airblowing alone. No oil could be removed by airblowing. After the lighted burner had been placed in position above the sand, the flue gas vent at the top of the pipe was closed thus causing all flue gases from the burner to pass through the sand bed. The burner was then adjusted by the aid of the flame temperature thermocouple to give its maximum temperature. The burner was kept on for 38 minutes, until the upper 3 inches of the sand bed was 1000 F. or over. During this period the pressure drop through the sand bed was 20 p. s. 1. After this 38-minute preperiod, the gas was turned off and air alone was passed into the oil sand bed. The amount of fed to the sand was 53 cubic feet per hour. With this amount of air passing into the sand bed, it was determined through the use of a bundie of eight thermocouples, that a flame front or peak temperature zone was bein maintained. Th re was a definite peak temperature in the bed, but it was noticed that as the wave progressed down through the bed, the peak temperatures kept dropping so that 1% hours after turning off the gas the hottest temperature in the sand bed was 800 F. At this time, it was decided to increase the air rate to 132 cubic feet per hour thereby increasing the pressure drop across the bed to p. s. i. The peak temperature responded immediately to this increase in air rate and lined out at about 1000 F. The remainder of the bed was burned off with this increased air rate. Examination of the temperature data obtained indicated that the flame front or combustion wave traveled through the sand bed at the rate of 15 inches per hour. The pressure drop through the sand at the end of the test was 22 p. s. i.

The oil recovered from the unit was found to weigh 3.6 lbs. compared to 3.5 lbs. charged. The increase in weight was caused by the removal of 10 some water from the sand used to make the oil sand mixture.

The temperature data obtained during the test indicated that the high temperature zone was 10- calized, and it appeared that it was confined to a narrow band less than an inch in width.

Upon dumping the sand from the pipe, it was found that there was no oil-bearing sand left. Most of the sand had been burned clean. The sand that had not been burned clean was cemented together, adhering in a cylindrical crust to the inside of the 4-inch pipe. It was observed that this cylindrical crust was thickest at the point where the peak flame temperature was the lowest. At the point where the temperature peak was only 800 F. the cemented crust was so thick that the burned out core through the center of the pipe was only about 1.5 inches in diameter. However, lower down in the sand bed this crust thinned out due to the increased flame front temperature caused by the increased air rate.

EXAMPLE II Following the completion of the test conducted in the 4-inch pipe, another test was made in a unit which more nearly simulated the conditions which would exist in the field. For this test, a rectangular box made of 16 gauge sheet metal was utilized. The inside dimensions of this box were 31 inches wide by 43 inches long by 24 inches deep. In order to reinforce the box, 2- inch by 2-inch angles were used to stiffen the sides and bottom. In mounting the box, one of its narrow edges was raised 6 inches above the other to give the box an 8 slope. A l -inch pipe was placed in the box at its higher edge in a vertical position. This pipe extended to the bottom of the box and served as the burner pipe, through which the gas flame used to ignite the bed was inserted. The burner pipe was perforated with .-inch holes for 15 inches along its lower end. The burner tip was so positioned that 4 inches of its nozzle was within the cover of the box. A second l -inch pipe was placed near the middle of the lower edge of the cover plate in a vertical position with its lower end touching the bottom of the box. This pipe was perforated with /;-inoh holes along 9 inches of its lower end, and the perforated area was covered with wire mesh to keep sand out of the pipe. To remove the oil which drained into the pipe, a inch pipe extending to the bottom of the larger pipe was used. The upper end of the larger pipe was connected to a valve so that back pressure could be imposed in order to cause the oil to flow up through the /4-inch pipe to a receiver.

The material charged to the box consisted of 1600 lbs. of builders sand mixed thoroughly with i 177 lbs. of oil plus 8 lbs. of finely ground coke.

The'oil was Santa Barbara reduced crude, and constituted 10% by weight, while the coke charge was 1 of 1% based on the sand. The oil sand was tamped firmly while being loaded into the box. After being thoroughly tamped into the box, the depth of the sand mixture was 19 inches. On top' of this sand a -inch layer of fine Olmsted earth was spread and the remainder of the 24-inch depth of the box was filled in with a mixture of half Olmsted earth and half builders sand. The box was slightly overfilled so that when the -inch thick cover plate was put into position, it had to be pulled down by clamps in order to be able to weld 'it'to the box. This procedure was used'in order to compress the sand so as to remove any voids within the box. The

4-in hie e e l ad four nc he metal fins extending across the inner width of the box welded to its under side. These fins were spaced about 8 inches apart and served to prevent lay-passing of the gases through any voids forming above the oil sand mixture. After. the box h s be n e tirel welded, hut. i was. ested with n e- 1- i p e s re nd al leak e W l d. shu

In der to ab e fo l t e cours f; the hi h emperatu zon throu h. h and thr e thermowells were used. One passed horizontally hr ugh the m d le of. h an bed al n the ens axis of the ex- AnQ her passed throu h th m ddle. O the sand. b hor zontal y throu h ts h as The. oth r well. nt ed the box ve i a to the en n pass d hr u h the. middl e the. e Be id s hese. h rmewe ls. there ere. tw me hrou h. h ch. t e. flam t rn.- n iet r an tempera ure he nd near the. u ne R pe eeuld e. ees red- 1. ord r. to. follow the high temperature zone along the long axis of the box a bundle of eight thermocouples spaced 1-inch apart. was used. This. bundle. of thermocouples. could be positioned. anywhere in the Well so asto explore thetemperature through-e. out the length of the bed. In the short horizontal and the vertical. thermowells. therewere bundles of five. thermocouples. spaced" 2v inches a tlocatedconcentrically in the. burner; pipe and entered through the. bottom, of the box; This thermowell extended to within 1. inch of. thetip ofthe burner. The thermowell for measurement of the sand temperature adjacent the burner pipe hada bundle of three thermocouplesspaced 2 inchesapart. This. thermowell paralleled the burner pipe and was located 2 inchesfrom it.

During the first attempt to initiate combustion in the sand bed insufiicient preheat was given to the sand surrounding the burner. tube and thus combustion wasnot started. Theqpreheat time was then extended to nine hours. During the reh at neri d h -ay r .p ess re :drop throu h he-b d Wasabeuta ps. 1- Aft r th pr h at perio d ,;the;air andgas. to theburner-Were cut f. 41 a l l ed.- e. he-bex through the .burner Pine a a sed. hr ugh-a r tome en o h tth e1 7.- a e; deems; ey end: b rn ng pe iod could e. et mine u in he; fir t. p rtionof: t end. urnin riedtbe ir ate as. kep hi h nei e te n edeee a .1:.- s iplie eu e on ross h v en -$1.2.- he yer. ur ns es ea erportionof the test a 5 p s, i. d-ifierentialwas'maine ame eree he. end. ed. he ave a a r. rate fer he 38- our u ning ri as. I a ut ub cre tn .-he r- During the initial. part of the sand burning period the peak temperature of the hot zone was about 1l00 F., butas it advanced away from the burner pipe and its area increased, the peak temperaturesfdropped. It was determined that increase of air flow was necessary to maintain or raise. the frontal temperature, but. unfortunately the. air rate through the bed could notv be appreciably increased "because, the box was made of light gauge steel. Thus the air rate through each unitarea of the hot wave decreasedduring the course of the test. The peak temperature in the bed at thetime the test was stopped was 695 F. V

The flue gases from the box-wereanalyzed by Orsat. The CO2 content at the'start of the sand burning period,was;13;5% and decreased gradually as the frontal temperature decreased;

lhe flame temperature thermowell was When it was noticed that the peak'frontal temperatures were continuously dropping a small amount, 6 cubic feet per hour of propane gas were fed to the sand bed along with the air. This procedure did not have any decided eifect on the frontal temperatures or CO2 content of the outlet gases. It was noticed, however, that this inclusion of propane with the. air fed to the box caused some localized heating of the burned off sand around the inlet pipe.

The total elapsed time for the test including the time required for preheating was 47 hours. During this time 126.5 lbs. of oil were recovered of the 1.77 lbs. placed on the sand, representing a recovery of 71%. The recovered oil contained about 5% water as shown by distillation. The gravity of the oil decreased from its original value of 21.8 API to 207 due to the increase in water content caused by the removal of water from the sandy used in the mixture. The odor of the recovered oil was substantially the same as that of the original oil. A, comparison of distillation, data obtained from the original and the. re.- cover d. oilfollowsz.

5%. 1110 in sample.

Upon completion the cover plate was-removed from the box and its contents were examined. It was found that the dry sand-Olmsted earth mixture placed on top of the oil sand bed had not absorbed any oil, and there was no indication that any-gases had by-passed through this dry mixture. There was no evidence of channeling through the oil sand bed. The sand that had been burned clean represented about 35% of the volume of the oil'sand bed and formed an egg-shaped volume. The large end of the eggshape surrounded the inlet or burner pipe, while the small end of the egg pointedtoward and almost touched the-outlet pipe.

EXAMPLE III Testswere made upon coregsamples from typical areas of the Nowata Field'to determine the effectiveness of purging with air and nitrogen within thetemperature range of, 80 F. to 1350. F. at atmospheric pressure. The sample, 442 grams (320 cc.) was ground to l4-mesh and cemented in a quartz tube. The purge gas was passedthrough the tubeafter preliminary drying and preheating. The eilluent gases were led through a; pair" of ice-water traps and calcium sulfate driers' in parallel. The gaseswere col- 13 450 F. to 500 F., judged by oxygen disappearance. Cracking begins at about 700 F., and the recovered oil had an API gravity of 275 to 29.0 and a U. 0. P. characterization factor 11.? to 11.9

"14 Core analysis of the area selected indicates a porosity of approximately 20% and a permeability of approximately 40 millidarcys. The oil content approximates 300 barrels per acre-foot and g i t i 5 the water content, which is entirely connate original oi l 3 2321551 35 1 030 53; 205031 8% re water approximates the same There is Very littl a r i covery increased. Wlbh temperature-rise rate, but e g S ecovemble P the p q decreased with increasing purge rate. The air For t purposes of muStFatmn' an Input Wen rates, it should be noted, however, were high for 10 typlcal 7'mch plpe 1S firmed and 1s field rates; e. g., about 300 cubic feet per minute, {earned and Shot Wlth mtroglycerme to expand but were as low as practicable in the laboratory. W hole Volume Wlthm t Sand to a to 34001? Total oil recovery did not appear to be affect d diameter cavity. In this area wells have been by the nature of the purge gas, air or nitrogen, fi m l patterns for gas pressuring, S0 at approximately Zhourly space velocity and 5. 1 that old Wells may be utilized. The well spacing "F. per minute temperature-rise rate. The iiow approximates 200 feet between input and outletdata are tabulated below in Table 13, nd gas wells and 330 feet between like wells. For test analyses by the mass spectrometer of th efpurposes all surroundin wells except one which fluent of run 0 are reported in Table 0'. is 230 f distant are pp d o plu ed.

Table B Charge Purge Recovered Materials Wt. Percent 011 Recovery, Temperature Based on origi- T t 1 R r Pressure 011 H 0 oo t t e fifgi sand (442 gm) 3. 3. 1. fi r tg. Rise-Rate (Gm.) Gin (cm?) by pentane ex- Range, 3 R/Mm) (Ave) traction (Ave) 1 CoreA N, .235 .020 4.20 3.00 53 2 CoreA(4l5gm.).. Air 1.133 .103 0 .97 3.37 0 OoreB .320 .027 4.40 3. 88 0.90 52 d .427 .013 4.50 5.10 0.39 53 .425 .025 0 .54 5.01 0 .912 .030 5. 82 0.27 10. 03 09 9.020 .103 1.00 12.30 23. 50 27 .327 .020 4.04 0.32 9.30 .490 .019 3.11 0.93 13.10 54 .597 .025 3.00 0. 03 3.47 52 1 Actual recovery. Some oil was lost (leakage).

Table C Cumulative M01 Percent Time of Tempera- Air Purge ture, F.

(HR) H1 N2 0: 00 CO3 01 O2= C2 C3= C3 C4 C4:

1 No samples taken.

EXAMPLE IV 60 To introduce combustion gas intothe forma As an example of an application in the field, our invention may be illustrated by application to the Delaware-Childers field of Nowata County, Oklahoma. This field is underlain by the Bartlesville sand at a depth averaging from 600 to 800 feet. The oil-bearing sand averages to feet in thickness and is overlain by impervious cap rock. Although the sand is somewhat lenticular it represents a formation of average or better uniformity. This field has been produced by primary methods, including vacuum, for over 40 years and has been subjected to major gas and air-pressuring projects and, to a minor. extent, to water flooding.

tion a high-pressure burner with an elongated combustion chamber is lowered through the well casing to formation level. A suitable burner is disclosed in application Serial No. 97,142 filed June 4, 19 19. The burner comprises an elongated combustion chamber to which fuel gas and primary air are separately introduced through concentric piping by means of a mixing plate at its head. A turbulent tangential motion is imparted to the flame by passage of the fuel gas and air through angularly directed ducts in the mixing plate. The mixture is ignited by a spark plug centrally located in the mixing plate. The burner is constructed of stainless steel or other 1 heat resistant metal, advantageously in two sections, the combustion, and an upper section containing an inner pipe for gas and a high tension cable leading to the spark plug. For a heat release rate approximating 500,000 to 600,000 B. t. u. per hour at 40 to 70 p. s. i. g., the dimensions of the burner advantageously are about 2 in diameter by 24" in length of each pipe section. The inner gas line and outer air pipe are extended to the surface with ordinary piping, and primary air and fuel gas are supplied in approximately theoretical proportions for perfect combustion. Secondary air to dilute the combustion gases and control the flame temperature is admitted through the oil-well casing around the burner tubing and combustion chamber.

The burner is operated at a flame temperature of about 1500 F. in the hole and under an inlet pressure of about 360 p. s. i. g. Fuel is burned at a rate in sufficient quantity for a heat release approximating 500,000 13. t. u. per hour. Suflicient air is provided to control the flame temperature and provide excess oxygen in'the fuel gases to accelerate the temperature rise in the formation through internal combustion. A back pressure of about 300 p. s. i. g. is maintained on the outlet well, and produced gas is recycled to the inlet well after repressuring.

The data on an illustrative example of operation are tabulated below;

Table D Nature of Field:

Depth to top of oil sands .i.

Thickness of oil sands.

Well spacing Average permeability of oil sands, millidarcys- 40 Residual oil after other recovery methods, wt.

percent of oil sands Residual carbonaceous solids, wt. percent of oil sands 0.5 Conditions of operation:

Total gas input, std. cu. ft.lday/well Base pressure level, output gas, ii /in.

Input pressure sufficient to maintain gas rate, ii /in.

approximately 360 02 content of total gas to input well, percent 3.33 5.0 Average production, bbl./day/input well. 3. 2 8. 12

At the start of operations, fuel is burned with air, and later recycle gas is also included to give temperatures of combustion of about 1200 F. to 2000 F., until about 180,000,000 cubic feet per well of total gas have been put in. The burner is extinguished and gas drive is now continued without burning extraneous fuel. Under these conditions with an oxygen content of below approximately 1.33%, the heat transfer wave is in front, and the rate is constant. With above 1.33% oxygen, the combustion wave is in front, travelling at a rate proportional to the oxygen content. oxygen, the fire in front will go out because the temperature is too low to sustain combustion and the oxygen-containin gas will simply be cycled through.

Alternatively, the gas drive may be continued with inert or substantially oxygen-free gases until the temperature of the front falls to a figure affecting recovery efficiency and approaching an effective in situ combustion temperature for the residual carbonaceous matter, say about 700 F. example. Air or other oxygen-containing gas is then added to the cycling cold gases so as to provide a sufficient quantity to ignite and support combustion of the residual carbonaceous material under the high temperature condition existing. The cycling of oxygen-containing gas in the gas flow is continued until a high temperature has been re-established. Again the temperaturemay be followed by the means of heat 16 transfer calculations or by deep well thermometers or thermal recording devices situated within the formation.

The use of heat transfer calculations to predict the temperature of the heat transfer wave front as it is moved out into the formation depends upon taking into account the factors mentioned above in columns '7 and 8, including the total amount of combustible material available for combustion in the high temperature region and takin into account the temperature level of the initial high temperature zone established at the input well, checked by data obtained through analysis of the gas recovered from the output well. During the combustion cycle, the presence of excess oxygen and carbon monoxide in the effluent gases is indicative of the completion of the combustion reaction and the correctness of the ratios of oxygen to fuel. The best indication of the temperature of the heat wave is furnished by the extent of decomposition of hydrocarbon gases occurring during the cycle gas or cold gas drive, the period of intermittent operation, Which can be followed by means of a recording calorimeter employed to determine the calorific value or the B. t. u. content of the produced gas. As a result of passage through the high temperature Wave front, the average molecular weight of the cycling hydrocarbon gases is reduced, with the tendency to form hydrogen increasing at conditions of extreme severity. Comparison of the ratios of light to heavy hydrocarbon followed by changes in calorific content forms a relatively sensitive indication of changes in the peak temperature level, which is not seriously confused by moderate changes in the width of the hot zone,

7 provides an additional means for checking the But with above about 5% to 6% of i methods of analysis and calculation but requires additional drilling into some intermediate portion of the formation before the high temperature wave has reached the test area.

We claim:

1. The method of recovering gas and oil from oil-bearing formations penetrated by at least one input well and at least one adjacent output well which comprises establishing a zone of high temperature at about 700 to 2500 F. within the formation surrounding the input well by means of hot gas iiow, moving the high temperature zone as a wave front by means of cold as flow outwardly from the input well and through the formation in a relatively horizontal direction, revivifying the high temperature wave front by establishing a combustion wave front by introduction of free oxygen in an average amount approximating 1.33 to 6.0% into the formation and controlling the rate of movement of the combustion wave front relative to the rate of movement of the high temperature wave front in a manner obtaining approximate coincidence between the combustion wave front and the high temperature wave front by regulating the velocity of free oxygen introduced to the formation during any period of operation and thus controllin the rate of combustion in the front by which its rate of movement is determined, while recovering gas and oil from the outlet Well.

2. The method of recovering gas and oil from oil-bearing formations penetrated by at least one input well and at least one adjacent output well which comprises establishing a zone of high temr 17 perature at about 700 to 2500 F. within the formation surrounding the input well by means of hot gas flow, moving the high temperature zone as a Wave front by means of cold gas flow outwardly from the input well and through the formation in a relatively horizontal direction, continuously revivifying the high temperature Wave front by burning residual carbonaceous matter left in the formation so as to form a combustion Wave front within the zone of high temperature by including free oxygen in thelcold gas drive in an average amount approximating 1.33 to 6.0% and controlling the rate of movement of the combustion wave front relative to the rate of movelment of the high temperature wave front in velocity of free oxygen introduced to the forma- 1 tion during any period of operation and thus controlling the rate of combustion in the front by which its rate of movement is determined, While recovering gas and oil from the outlet well.

3. The method of recovering gas and oil from oil-bearing formations penetrated by at least one input well and at least one adjacent output well which comprises establishing a zone of high temperature at about 700 to 2500 F. Within the formation surrounding the input well by means of hot gas flow which is substantially free of uncombined oxygen outwardly from the input well 18 and through the formation in a relatively horizontal direction, intermittently revivifying the high temperature wave front by burning residual.

carbonaceous matter left in the formation so as to form a carbonaceous wave front in the zone of high temperature by cyclically introducing a free oxygen containing cold gas flow to the formation, controlling the rate of movement of the combustion wave front relative to the rate of movement oftbe high temperature wave front in a manner obtaining approximate coincidence between the combustion wave front and the high temperature wave front by regulating the mass velocity of free oxygen introduced to the formation during this period of operation and thus controlling the rate of combustion in the front by which its rate of movement is determined and by limiting the amount of free oxygen introduced to theformation to an average amount within the range approximating 1.33 to 6.0% of the total References Cited in the file of this patent UNITED STATES PATENTS Number Name Date 2,382,471 Frey Aug. 14, 1945 2,390,770 Barton et al Dec. 11, 1945

Claims (1)

1. THE METHOD OF RECOVERING GAS AND OIL FROM OIL-BEARING FORMATIONS PENETRATED BY AT LEAST ONE INPUT WELL AND AT LEAST ONE ADJACENT OUTPUT WELL WHICH COMPRISES ESTABLISHING A ZONE OF HIGH TEMPERATURE AT ABOUT 700* TO 2500* F. WITHIN THE FORMATION SURROUNDING THE INPUT WELL BY MEANS OF HOT GAS FLOW, MOVING THE HIGH TEMPERATURE ZONE AS A WAVE FRONT BY MEANS OF COLD GAS FLOW OUTWARDLY FROM THE INPUT WELL AND THROUGH THE FORMATION IN A RELATIVE HORIZONTAL DIRECTION, REVIVIFYING THE HIGH TEMPERATURE WAVE FRONT BY ESTABLISHING A COMBUSTION WAVE FRONT BY INTRODUCTION OF FREE OXYGEN IN AN AVERAGE AMOUNT APPROXIMATING 1.33 TO 6.0% INTO THE FORMATION AND CONTROLLING THE RATE OF MOVEMENT OF THE COMBUSTION WAVE FRONT RELATIVE TO THE RATE OF MOVEMENT OF THE HIGH TEMPERATURE WAVE FRONT IN A MANNER OBTAINING APPROXIMATE COINCIDENCE BETWEEN THE COMBUSTION WAVE FRONT AND THE HIGH TEMPERATURE WAVE FRONT REGULATING THE MASS VELOCITY OF FREE OXYGEN INTRODUCED TO THE FORMATION DURING ANY PERIOD OF OPERATION AND THUS CONTROLLING THE RATE OF COMBUSTION IN THE FRONT BY WHICH ITS RATE OF MOVEMENT IS DETERMINED, WHILE RECOVERING GAS AND OIL FROM THE OUTLET WELL.