CA1320007C - Method for determining residual oil saturation of a gas-saturated reservoir - Google Patents

Abstract

METHOD FOR DETERMINING RESIDUAL OIL
SATURATION OF A GAS-SATURATED RESERVOIR

Abstract A method for determining in situ residual oil saturation of a gas-saturated reservoir by injecting at least two non-reactive tracers into an injector well and analyzing the samples from a production well. The tracers have different partition coefficients and are chromatographically retarded to different extents during their passage through the formation.

Description

:~2~7 A~ cation for Patent Title: METHOD FOR DETERMINING RESIDUAL OIL
SATURATION OF A GAS-SATURATED RESERVOIR
Inventors: Joseph S. Tang and Bradford C. EIarker :~pecification Field of the Invention:
The present invention relates to a method of using tracers to determine the in situ residual oil saturation between two locations in subterxanean gas-saturated oil reservoirs with gas being the ~obile phase. More specifically, -the present inverltion relates to the : determination o~ the relative concentrations of oil and gas within subtèrranean reservoirs: by measuring over a period o~ tlme -the chromatographic separation of -tracers I
having distinctly different Henry's law constants or : K values in the oil and gas phase fluids in the reservoir., Background of:the:Invention: :
Typical oil reservoir:formations are made up of rock containing tiny, interconnected pore spaces which are saturated with oil, water, and gas. KnowIedge of the concentrations of these fluids in the formation is critical for the effic~ient p~oduction of the oil. When the formation is first drilled, it is necessary to know : -the or~ginal oil saturation in order to plan the : 20 exploitation of the ~ield. Later in the life of the field, the amount of oiI remaining in the ormation will 13 2 ~
often dictate the most efficient secondary and tertiary recovery operations.
Several methods are currently used to determine Eluid saturations of a formation. One techni~ue involves coring, i.e., direct sampling of the formation rock and fluids wherein a small portion of rock satura-ted with fluids is removed and brought to -the surface where its fluid content can be analyzed. Coring, however, is susceptible to several shortcomings. First, the small sample may not be representative of the formation as a whole since it only investigates the immedia-te vicinity of the wellbore. Second, the coring process itselE may change the fluid satur~tions of the samples. Finally, coring can usually only be done in newly drilled wells.
Another method of de-termining fluid saturations involves logging techniques. This method, too, suffers from the shortcoming of investigatin~ a limited area which is in the immediate vicinity of the wellbore. In addition, logging techniques are often unable to differentiate between properties of the rock and those of its fluids. U~fortunately, coring and log~ing are not suitable for low-pressure, low-porosity, gas-filled carbonate reservoirs.
Another approach involves material balance ~5 calculations based on production history. ~Iowever, this approach is susceptible to error since it requires a knowledge of the ini-tial fluid saturation of the formation ~y some other independent means.
More modern methods for determining fluid saturations involve the injection and production of tracers. The technigues are based on chromatographic theory~
Typically, two tracers having different ~artition coefficients are used. The tracers are chromatographically retarded to different e~tents as they 3~ pass through the formation. The degree -to which the two tracers are differentially re-taxded can be used to determine the formation fluid saturations.

22658/16/~ MF

~ 3 ~ 3~ r~

Most tracer techni~ues for the determination of fluid saturations involve using a single well. A fundamental problem wi-th single well testing is that only a very limited portion of the formation, the area immediately surrounding the wellbore is investigated. ~part from this fundamen-tal problem, single well testing which a-ttempts to take advantage of chromatographic principles also suffers from an additional shortcoming -- the "mirror image"
effect. The mirror image effect occurs where two or more tracers having different partition coefficients are injected into a formation. The tracers will separate as they are injected into the formation, and thP degree of separation will be a function of the oil sat~lration.
However, when the tracers are withdrawn from the formation by means of the same well, the separation will disappear.
When the tracers moved away from the well, one tracer moved faster than the other due -to the difference in partition coefficients and the residual oil saturation.
When the well was placed on production, the faster moving tracer again moves further than -the other and the two tracers arrive at the wellbore at approximately the same time.
Several schemes have been devised to avoid this problem. In one techni~ue, the well is shut in after the injection of the tracers for an extended period of time.
This allows the tracers to drift, i.e. to move in the formation under -the in1uence of forces unrelated to the injec-tion or withdrawal of fluids a-t the well. When the well is pu-t on production, the tracers are somewhat separated and a determination of fluid saturations becomes more feasible. The problem ~ith this technique is that it is difficult to determine the -time necessar~ for the tracers to drift. Furthermore, e~tended residence time in the formation creates other problems, such as gravita-tional separation of the tracers.
Another way of getting around the "mirror image"
effect is to inject a non-reactive tracer along with a 22658/16/~

~L32~37 tracer precursor. The injec-tion is followed by a shut~in period during which -the precursor is allowed to react to Eorm a tracer. The precursor and corresponding tracer have different par-titioning coefficien-ts. During -the in~ection phase, the precursor and non-reactive tracer move away from the well a-t certain velocities determined by their partition coefficients. During the production phase, the non-reactive tracer moves back toward the wellbore at the same rate but the newly formed tracer, because it has a different parti-tion coefficient from that of its precursor, moves at a rate different from -that of its precursor. The resul-t is a separation at the wellbore of the two tracers. The problem with this method is that it depends for its success on chemical reactions which are influenced by various factors, such as formation temperature. In addition, this method is not applicable to gas-filled reservoirs because o~ a lack of suitable tracers which can react in the gas phase or the oil phase to generate the necessary tracers.
The mirror image problem can be completely circum~ented by injecting a carrier fluid containing at least two non-reactive tracers having different partition coefficients between the fluid phases into one location in the formation and produciny f~om another. Typically, one well is used to inject the carrier ~luid bearing the tracers while another well is used to p~od~ce formation fluids. Because different injection and production locations are used, it is unnecessary to rely on fluid drif-t for the separation of the tracers. Nor is it necessary -to use tracer precursors and rely on chemical reactions to produce tracers with different partitioning coefficients. Instead, non-reacti~e tracers can be used which are chromatographically separated as they pass through the formation, and this chromatographic separation is a func-tion of the saturation of the immobile phase.
The basic idea of chromatographic separation was disclosed by Dr. Claude Cooke in U.S. Patent 3,590,923.

22658/16/l-1-l/MF

-5- ~3~ 7 Cooke injected fluid containing at least two tracers of differen-t partition coefficients. The tracers were chromatographically retarded in their passage -through the formation to dif~erent ex-tents. The breakthrou~h of -the tracers was detected in another location, and inferences were drawn about the relative proportion of forma-tion fluids.
While the Coo]~e method was superior to any of those previously used, it suffered from a number of serious drawbacks. First, little guidance was given on the selection of appropriate tracers. Second, the Cooke method used only tracer breakthrough quantities -to calculate residual oil saturation. Because of dispersion, stratification, streamline effects and the detection properties o~ various properties, it was usually difficult to determine the precise time of breakthrough with great accuracy. Even when breakthrough was determined wi-th considerable accuracy, -the effect of using only the breakthrough was that only the residual oil saturation of the most permeabl~ layer was determined. The saturation of other layers in the formation were not determined by this techni~ue. Thus, there still exists a need in the industry for a method to accurately determine the residual oil sa-turation of a formation.
Summary of the Invention:
The present invention relates to an improved process in which residual oil satuxations o a hydrocarbon-containing, ~as-saturated formation are determined by injecting a fluid containing at least two properly selected non-reactive tracers into the formation.
The tracers have different partition coefficients and are`
chromatographically retarded in their passage through the formation to differing extents. The presence and amounts of the tracers are detected over extended periods of time at another location. The complete results are analyzed using chromatographic theory or reservoir simulation methods to determine the relative proportions of formation 22658/16/1-1-l/MF

:~ 3 ~ 7 fluids for various por-tions of the formation be-tween the injection and production locations.
The presen-t invention re~uires selecting appropria-te tracers. One of these tracers is selec-ted from the group consisting of halocarbons, halo~hydrocarbons and tritiated or carbon 14 tagge~ hydrocarbons. This first tracer is oil partitioning with a Henry's law constant in -the appropriate range. The second tracer is preferably an oil non-partitioning tracer, such as sulfur hexafluoride, -tritium gas or radioactive isotopes of inert gases with a Henry's law constant at reservoir con~itions which is hi~h~
relative to the selected oil partitioning tracer.
However, this second -tracer can be selected from the same group of oil partitioning tracers as the first tracer so long as it has a Henry's law constant different from the ¦
first tracer.
The tracers are injected into the formation through an injector well. Production is from a well in communication with the injection well. Samples are taken of -the produced fluids over an extended period o~ time and are anal~zed for the presence and amount of the tracers.
The residual oil saturations of the various layers of -the formation are calculated accord:ing to chromatographic theory using the breakthrough ~uantities and the production function of the tracers.

Brief Description of the Drawings:
Fig. 1 is a plot of inverse Henry's law constants with retention pore volume for a slimtube test 1.
Fig. 2 is a plot of inverse Henry's law constants with retention pore volume for a slimtube test 2.
Fig. 3 shows tracers partition profile (concentration as a func-tion o time).
Fig. 4 shows tracers partition profile ~concentration as a function of time) with four "landmarks" indicated.
Fig. 5 shows plots of inverse Henry's law constants versus production time for the four "landmarks" of Fig. 4.

2265~/16/1-1-1/MF

~3 ~
Fig. 6 is a plo-t of "Sor" versus cumulative recovery.

Description of the Invention:
Selectlon of Tracers Tracers useful in this invention include oil partitioning tracers, such as halocarbons, halo-hydrocarbons, triti.a-ted or carbon 14 tagged hydrocarbons, as well as oil non-partitioning tracers, such as sulfur hexafluoride, tritium gas and radioactive isotopes of inert gases with a ~Ienry's law constant at reservoir conditions which is high relative to the oil partitioning tracer used.
Tritiated hydrocarbons are ordinary hydrocarbons with at least one hydrogen replaced by tritium. Likewise, carbon 14 tagged hydrocarbons are ordinary hydrocarbons with at least one carbon replaced by carbon 14. These txitiated or carbon 14 tagged hydrocarbons, which can be saturated or unsaturated, may contain up to 5 carbons.
Selection of the appropriate halo hydrocarbon, halocarbon, and tritiated or carbon 14 taggecl hydrocarbon is based on ~0 the Henry's law constant (or K value) at reservoir conditions and on the detection limit.
Henry's law constan-ts can be determined in the lab using pre-e~uilibrated reservoir gas and oil at reservoir temperature and pressure by a static partitioning test or by a dynamic slim-tube test. In -the static partitioning test, a 0.001~ to 2% concentration of halo-hydrocarbon or halocarbon is introduced into the pre-e~uilibrated gas which is then eguilibrated with the oil for at least 4 hours in a rocking cell. The halo-hydrocarbon or halocarbon concentration in the gas and liquid phases can then be determined by: (a) direc-t measurement; ~b) measuring the halo-hydrocarbon or halocarbon concentration in the gas phase and determining the concentra-tion in the oil phase by head-space analysis; or (c) measuring the halo-hydrocarbon or halocarbon concentration in the gas ~ 3 ~
phase before and af-ter equilibrium and de-termining its concentration in the oil phase by material balance.
In the static partitioning tes-t, a small dosage of tritiated hydrocarbon or carbon 14 -tagged h~drocarhon, typically less than 0.001 m Curie, is introduced into -the pre-eguilibrated gas which is then equilibrated with the oil for at least 4 hours in a rocking cell. The tritiated hydrocarbon or carbon 14 tagged hydrocarbon concen-tration in the gas and liquid phases can then be determined by first separating the components by preparative gas chromatography or other means with or without added inert carriers and measuring the activities of the individual tri-tiated hydrocarbons or carbon 14 tagged hydrocarbons.
Alternatively, the tritiated or carbon 14 tagged hydrocarbon concentration in the gas phase before and after equilibrium can be measured, and its activity in the oil phase can be determined by material balance.
Henry's law constant "Hi" is defined as:

Hi = xi (atm~ (1) where "yi" is the mole fraction of -tracer "i" in the gas phase, "xi" is the mole fraction of tracer "i" in the oil phase, and "P" is the reservoir pressure, atm. Henr~'s law constant can be calculate~ as:
Hi - ~ Ai* )( RTZ(T,P) ) Vo ~2) Ai~-Ai* Vg Vm where Ai, Ai* are the gas chromatograph peak area counts for halo-hydrocarbon or halocarbon "i" or the decompositions per cc of gas at a reference T, P per minute for tritiated or carbon 14 -tagged hydrocarbon "i"
before and after e~uilibrium respectively, "Vol' and 'IVg are the volumes of oil and gas in the rocking cell for partitioning test, IlVm" is the molar volume of the oil, llZll is the compressibility factor, I'Rl' is the gas constant, and "T" is the temperature, in degrees Kelvin.

-~32~
A slimtube with 0.1 -to 0.5" ID and 40-120 ft. lenyth is suitable for the dynamic test. Th~ tube is packed with glassbeads and maintained at reservoir pressure by using a back pressure regulator. The test is conducted in an oven at reservoir temperature. The tube is first saturated with oil and then gas-flooded to residual oil saturation with the pre-equilibrated reservoir ~as. ~lterna-tively, the tube can also be first saturated with formation water and then oil-flushed followed by a gas-flood to connate water and residual oil saturations. A known but small slug containing the halocarbons, halo-hydrocarbons or tritiated or carbon 14 tagged hydrocarbons to be tes-ted is injected into the tube, and the tracer concentrations in the effluent gas are continuously monitored through a sampling valve by a gas-chromatograph with an electron capture detector in the case of the halo-h~drocarbons or haloearbons and b~ a gas proportional counter or li~uid scintillation counter in the case of the tritiated or carbon 14 tagged hydrocarbons. Component separation prior to coun-ting is not required sinee various tritiated or carbon 14 tagged hydrocarbons have been separated in the slimtube.
The Menry's law constants for individual tracers can be ~alculate~ from the produeed peak pore volumes (retention ~olumes or Vpi) aceording to the following equation:

Hi = (RVm) ( Vpi - Sg where "So" is the oil saturation and "Sg" is the gas saturation~
In order to obtain the responses of the partitioning tracers in a reasonabl~ short time and get sufficient separation of peaks, the best range of "Vpi" is 1.5 Sg to 5.0 Sg with corresponding "Hi" values of:
_ 1 RTZ So Hl 5 ~ Vm ) ( Sg ) for Vpi = 6 Sg (4) to ~ 32~7 Hi = 2 RTZ ( Sg ) for Vpi = 1.5 Sg (5) Using a non-partitioning tracer, such as -tritium gas or any raclioactive isotopes of light inert gases, in the tracer slug is preferred. The non-partitioning tracer is produced at Vpi = Sg, corresponding to an infinite Henry's law constant. If, however, a non-partitioning tracer cannot be identified, it can be substituked with a second partitioning tracer having a Henry's law constant value different from the Henry's law constant of -the first tracer. If both tracers are partitioning, the tracers must have a Henry's law constant such that the retention volume of one tracer (Vpi~ is at least 1.5 times larger than that of the other tracer.
Henry's law constant can also be estimated from the vapor pressure (Poi) of tracer "i" at reservoir tempera-ture or from equations of state. The halo-hydrocarbon and halocarbon vapor pressure which is always lower than the Henry's law constant can be improved through khe use of Regular solution theory which can par-tially account for the non-ideal behavior of halo-hydrocarbon in oil.
Detection Limits of Halo-Hydrocarbons and Halocarbons Linearity of the response and detection limit can be determined by measuring the ga~ chromatograph signals of a series of samples prepared by successive dilu-tion of a mi~ture of halo-hydrocarbons with reservoir gas. The detection limit for a chemical is adversely affected by possible interferences from the reservoir ~as, as well as by the noise level of the detector. For any design to be prac-tical, the detection limit of the chemical should be lower than 1 ppm unless the sweep volume of the test zone is ex-tremely small. In order to be detectable at the 1 ppm detection level, khe halv-hydrocarbon and halocarbons should have at least ~ chlorine atoms or 1 bromine atom or 1 iodine atom.

22Ç58/16/1-1-1/MF

~ 3 2 ~ ~ ~ J

Detection Limits of Tritiated and Carbon 14 Ta~ge~
EIydrocarbons A good separa-tion o:E individual tritia-ted and carbon 14 -tagged hydrocarbons is essential for -the success of the test. The most common means o~ separation is preparative gas chromatography and, less commonly, low temperature flashing. ~periments have to be carried out to chec~ the separation of the selected tritia-ted and carbon 1~ tagged hydrocarbons.
Description of the Method:
A mixture of at leas-t two tracers made up preferably of a partitioning tracer, such as halo-hyclrocarbons, halocarbons, or carbon 1~ tagged or tritia-ted hydrocarbons and a non-partitioning -tracer, typically sulfur hexafluoride, tritium gas or other radioactive isotopes of inert gases, is injected into the formation with the injection gas above the dew point. The mi~ture can be injected as a slug, a spike, o:r continuously at low concentration.
Tracers can be injected separately from cylinders each of which contains a single tracer at or about its own vapor pressure. Pumps are adjusted so that the tracers are delivered to the injection line at the desired proportions. The tracers are then vaporized and mixed thorou~hly with the injection ~as whi~h is subse~uently warmed to the reservoir temperature by an in~line heater.
Alternatively, the tracers can be premixed in a cylinder and discharged to the injection line by a single pump. In this mode of injection, however, a vapor-liquid phase e~uilibrium calculation has to be carried out to study the effect of differential liberation of individual tracers during discharging. A fast injection rate may help minimize the impac-t of differential liberation on the process. Tracer injection time typically ranges from less than an hour to continuous injection.
The production rates and injection rates should remain as steady as possible throughout the -test.
Production rates do not have to be egual to the injection 22658/16/1-1-1/~`

~ 3 2 Q ~ ~ ~

ra-tes as long as the reservoir pressure is not significantly changed by the unbalanced injection and production. As a limiting case, the producers can be used as observation-sampling wells where small but representative samples are obtained wi-ch time. This zero production rate has the merit o~ giving a non-disturbed flow pattern which provides a reliable means o~
determining the permeability and permeability thickness distribu-tion as well as sweep volume.
Produced gas samples are collected at a predetermined freguency and analyzed for the tracers. If all the tracers follow the various flow paths in similar proportions, and the residual oil saturations are identical along the various ~low paths, the residual oil saturation can he determined by the chromatographic theory shown below. If, however, either of these criteria is not met, simulation must be used.
This method can be extended to the simultaneous determination of three-phase saturations ~y including one or more high vapor pressure oil-non partitioning/water-participating and/or oil-partitioning/water-partitioning tracers.
Data_I_terpretation 1. Chromatographi~ Th~ory According to chromatographic theory:

Qi = Qs [1 t RTZ So ] (63 ~iVm Sg where "Qi" is the breakthrough volume for tracer "i" and "Qs" is the gas volume in the flow path (sweep volume).
A plot o~ "l/Hi" versus "Qi" ~or the various tracers yields a straight line. The slope of this line is used to determine -the residual oil/gas saturation ratio, i.e., So/Sg. In order to solve for "So" and "Sg", at least one o~ the following needs to be known.
(i.~ any saturation (i.e., Sw, So or Sg)i (ii) any combin~tion o~ two saturations; or (iii) porosity and sweep volume.

22658/16/1-1-l/MF

-13 ~ 3~ 7 Usually the breakthrough volume (Qi) is difficult -to obtain. Instead one can use extrapolated break-through volumes, peak volumes or half height volumes (i.e., volume corresponding to half of the peak height) in the "1/Hi"
plo-t. It is usually mos-t convenient and most accurate to use half height volume.

2. Simula-tion Reservoir simulators capable of modeling sweep volume can be employed to interpret -the results by matching the entire tracer production profiles.
Example 1: Slimtube Test 1 Gas N2 Oil Tetradecane Water 110,000 ppm brine Tracers* CH4, F23, F22 and F12 Sg 0.63 So 0.~1 Sw 0.06 P 300 psi Example_2~_ Sl tube Test 2 Gas Pre-equilibrated reservoir gas Oil Pre-equilibrated reservoir oil Water 110,000 ppm reservoir brine Tracers* F23, F22 and F12 S~ 0.42 So 0.36 Sw 0.22 P 600 psi *NOTE: Fl2 = dichloro-difluoro-methane F22 = chloro-difluoro-methane F23 = trifluoro-methane (fluroform) 22658/16/1-1-1/~

14- ~32~

By combinin~ equations ~2) and ~3~, it can be shown that:

Vpi = Sg + ~ Ai i Ai* } V~So (7) Therefore, as shown in Figure 1 (Slimtube Test 1), and Figure 2 (Slimtube Test 2), a plot of halowhydrocarbon peak volume Vpi vs. reciprocal EIenry's law constant (Ai - Ai*)/Ai* measured independently in a static test yields a straight line with the slope given by:

~ (8) Vo From the above relationship, the residual oil saturations are determined to be 31% and 36% which are in excellent agreement with the experimental values of 32% and 36%, respectively.

Example 3 A freon interwell test was carried out to determine residual oil saturation in a gas-filled carbonate reservoir located in Central Alberta, Canada. The reservoir, which has an average porosity of 17%, was first produced by primary depletion and then by gas-cycling since 1953. The upper part o~ the reservoir has already been gas-flooded down to residual oil saturation. The reservoir pressure and temperature were 600 psi and 60C
respec-tively at the time of the tracer test. The tracer test was conducted in a 2-SpQt (i.e., one injector and one producer) unconfined pattern. The wells were 152 m apart and an interval of 5 m was perforated for the test. ~he test was run at steady state with the injection and production rates maintained at 40,000 SCM/day and 7~,000 SCM/day, respectively, throughout the test.
To satisfy the vapor pressure and detection limit re~uirement, SF6, F13B1 and F12 were selected for the test. Some of the parame-ters for these tracers are listed in Table 1.

22658/16/l-1-1/ME

~32~7 T~BLE 1 Parameters of the Tracers Formula Sulfur Hexafluoride ~olecular Weight 146.07 v.p. @ 70F 310 psig l/HEi 0.561 Detection Limit (lab) 0.2 ppb F13Bl Formula Bromo~trifluoro-methane Molecular Weight 148.93 v.p. @ 70F 190 psig l/HEi 2.13 Detection Limit (lab) 2 ppb Formula Dichloro-difluoro-methane Molecular Weight 120.93 v.p. @ 70F 70.19 psig 1/~Ei 4.45 Detection Limi-t ~lab) 6 ppb A slug of freons composed of 5.6 Kg of SF6, 17.4 Kg oE
F13B1 and 33.6 Kg of F12 was injected with dry gas at the specified injection rate and gas samples were collected from the pro~ucer 4 times ~ day using an automatic sampler. The samples were subsequently analyzed. From the production profile o~ the most volatile component SF6, there appear to be 3 layers (or flow paths) contributing approximately 20%, 30% and 50% to the flow.
The production curves for SF6, F13B1 and F12 are plotted in Figure 3 for comparison. It is noted that the production curves for F13Bl and F1~ are "delayed"
fingerprints of the SF6 profile, every detail of which is preserved in the production curves of the two higher boiling tracers. Because of fre~uent sampling, the hreakthrough times were found exactly to be 4.4 days, 5.1 days and 6.1 days for SF6, F13B1 and F12, respectfully, as would be expected from their Henry's law constan-ts. The 2265~/16/1~ MF

-16 ~ 3 ~ 7 cumulative recoveries for the three tracers were transposed to a recovery-tracer profile (i.e., "landmark") cross-plot indicated in Figure 4. I-t can be shown in the cross-plot tha-t recovery correlates well with -the con-tour of the tracer produc-tion profile, i.e., the "landmark."
For instance, a cumulative recovery of 18% corresponds to the peak positions (i.e., peak C) of -the three tracer curves. Therefore, when determininy residual oil saturation by the chromatographic -theory technique either "landmark," e.g., breakthrough time and peak -times, or equal recovery time can be used for comparison.
For constant injection and production rates, Equation (6) can be written as:
ti = to*[1 ~ So/(HEi*Sg~] (9) where "ti" and "to" are the production times for the partitioning tracer "i" and the non-partitioning tracer at a given cumulative recovery or "landmark," and "HEi" is the effective Henry's law constant defined as:
HEi = Hi*Vm/(RTZ) (10) Therefore, a plot of "l/HEi" versus "ti" for the three tracers would yield a straight line from the slope and intercept of which the oil to gas saturation ratio (So/sg) can be determined. "So" can -then be solved explicitly if connate water satura-tion is known. To demonstrate the chromatographic method, "l/HEil' is plotted against "ti" in Figure 5 at five cumula-tive recoveries of 0%, 2%, 9%, 18%
and 22.5% which correspond to breakthrough, peak A, peak B, peak C and late production respectively. It is found that, as predicted by Equation (9), -the three tracer points fall on a straight line for each of the five recovexies with the corresponding residual oil saturations determined to be 9%, 7%, 15%l 19% and 20% at an estimated connate water saturation of 8%. Residual oil saturations measured at various cumulative recoveries from ?2658/l6/1-1-1/MF

-17~

breakthrough -to 40% are plotted in Figure 6 with arrows indicating the peak posi-tions. The variation oE residual oil saturation with recovery is due to the overlapping of three layers with different residual satura-tions, i.e., 7%
for peak A, 15% for peak B and 20% for peak C. Residual oil saturation reaches a constan-t level of 20% in late production where layer C domina-tes the tracer production.
All the above observations, i.e., 1. Recovery correlates with the contour of the production profile ('!landmar~"), 2. The "HEi" vs. "ti" plot yields a straight line at various recoveries, 3. Residual oil saturation reaches a constan-t value at late production, demonstrate that chromatographic theory and the "landmark"
comparison technique in -this case accurately estimate residual oil saturation ~rom the tracer profiles, and thus simulation is no-t necessary.
A dual-porosity mixing cell model was used to check the validity of -the "landmark" comparison techni~ueO The simulation results indica-te that only under a pseudo-single porosity situation, i.e., no non-flowing fraction or extremely fast or extremely slow mass transfer between flowing and non-~lowing fractions, will -the above phenomena be ohserved. For a real double-porosity system wi-th an intermediate ma~s transfer rate, the peak tends to give the residual oil saturation in the flowing fraction and, as the mass transfer rate increases, the pore-average xesidual oil saturation, whereas the late produc-tion tends to give the non-flowing oil saturation. For such a system, irregularities may be encountered in using the chromatographic technique to calculate ~'Sor". Also, the non-flowing oil saturation cannot be accurately measured even if a dual porosity simulator is used for profile matching. From the simulation results in this case, it would appear that this carbona-te reservoir behaves as a single porosit~ reservoir.
The prînciple of the invention, a detailed description of one specific application of the principle, -18- ~32~

and the best mode in which it ls contemplated to apply that principle have been described. I-t is to be understood that the foregoing is illustrative only and that other means and -techniques can be employed without departing from the true scope of the invention defined in the follow.ing claims.

22658/16/l-1-l/ME

Claims (7)

1. A method for determining in situ residual oil saturation of a gas-saturated saturated reservoir comprising:
selecting a first oil partitioning tracer from the group consisting of halocarbons, halo-hydrocarbons, and tritiated or carbon 14 tagged hydrocarbons wherein said first tracer has a Henry's law constant;
selecting a second tracer which is either oil partitioning or oil non partitioning Prom the group consisting of halocarbons, halo-hydrocarbons, tritiated or carbon 14 tagged hydrocarbons, sulfur hexafluoride, tritium gas and radioactive isotopes of insert gases wherein said second tracer has a Henry's law constant different from said first tracer;
injecting into an injector well a mixture comprising said tracers;
producing said mixture from a production well in communication with said injection well;
collecting produced gas samples and analyzing said samples for the presence of said tracers, and calculating residual oil saturation.

2. The method of claim 1, wherein the first tracer has a Henry's law constant at reservoir conditions of between (1/5)x and 2x, wherein x is (RTZ/Vm)(So/Sg), and wherein "R" is the gas constant, "T" is the temperature in degrees Kelvin, "Z" is the compressibility factor, "Vm" is the molar volume of oil, "So" is the oil saturation, and "Sg" is the gas saturation.

3. The method of claim 1, wherein the second tracer is oil non-partitioning and is selected from the group consisting of sulfur hexafluoride, tritium gas and radioactive isotopes of inert gas.

4. The method of claim 1, wherein the second tracer is oil partitioning and is selected from the group consisting of halocarbons, halo-hydrocarbons, and tritiated or carbon 14 tagged hydrocarbons and has a Henry's law constant at reservoir conditions such that the retention volume of one tracer is at least 1.5 times larger than that of the other tracer.

5. The method of claim 1, wherein residual oil saturation is calculated according to chromatographic theory, wherein:
breakthrough volume is determined by extrapolation of volumes selected from the group consisting of breakthrough, peak and half-height volumes; and So/Sg is determined by the slope of a straight line yielded by a plot of reciprocal Henry's law constant versus breakdown volumes, wherein "So" is the oil saturation and "Sg" is the gas saturation.

6. The method of claim 1, wherein residual oil saturation is calculated using reservoir simulators capable of modeling weep volume.

7. The method of claim 1, wherein the injection rate and the production rate are controlled so as to maintain a substantially constant reservoir pressure.