The Kuparuk River oil field has had a “miscible” water-alternatinggas (MWAG) enhanced-oil-recovery (EOR) pilot in place since 1988. This original pilot was implemented on two drillsites previously on waterflood. The pilot was expanded to a third drillsite in 1993 to test the MWAG process in an area previously fl ooded by an “immiscible” water-alternating-gas (IWAG) recovery technique. Large-scale application of MWAG at Kuparuk is in progress, wit h a project that will more than triple current MWAG injection. Full implementation is slated for mid-1996. When completed, the Kuparuk large-scale EOR (LSEOR) project will be among the largest of its kind in the world. This paper covers the basic aspects of moving the EOR process at Kuparuk from pilot to large-scale application. A number of hurdles had to be overcome, including scaleup of field and simulation results in a consist ent manner and the need to use existing infrastructure to ensure economic viability.
The Kuparuk River field is on Alaska’s North Slope, approximately 260 miles north of the Arctic Circle and 30 miles west of the Prudhoe Bay field (Fig. 1). On the basis of remaining reserves and a current production rate of nearly 300,000 BOPD, Kuparu k is the second largest field in North America. From the time of first production in 1981, Kuparuk has operated under a variety of recovery mechanisms, including primary depletion, waterflood, lean-gas reinjection, and both IWAG and MWAG injection. Reservoir development decisions are often complex and interrelated, requiring the integration of reservoir recovery mechanisms and surface facility facility constraints. constraint s.1-4 This was especially true for the Kuparuk LSEOR project. Kuparuk Development History. Discovered in 1969, the decision to develop the Kuparuk River oil field was delayed until 1979 because of economic uncertainties and resource constraints tied to the startup of the Prudhoe Bay oil field. By late 1981, the Kuparuk River field began producing 80,000 BOPD from 40 wells on five drillsites. Production exceeded expectations. Fluids were processed by a single central production facility (CPF-1), with the oil being transported 27 miles to the Trans Alaska Pipeline System (TAPS) through a 16-in. common-carrier pipeline. Gas not burned as fuel was reinjected into an area of the reservoir designated for gas storage. The field did not have an original gas cap. Because of the promising results, a plan was in place in 1982 to develop the entire field with three CPF’s. Estimates at that time were to spend more than U.S. $8 billion in capital to recover approximately 1.3 bi llion bbl of oil, with an estimated peak production rate of 250,000 BOPD. The second CPF (CPF-2) was commissioned in 1984. To accommodate the increased oil production, the 16-in. pipeline from Kuparuk to TAPS was replaced with the current 24-in. Kuparuk pipeline. Fieldwide expansion of a waterflood pilot ini tiated in 1983 was made possible with the 1985 startup of the 300,000-BWPD capacity seawater treatment plant (STP). Concurrent with STP startup, an IWAG pilot was initiated on three thr ee drillsites. IWAG IWAG injection was intended to provide a means of both storing gas and improving recovery.5,6 To date, a total of 19 drillsites drillsi tes have been subjected to IWAG, IWAG, with expansion to two additional drillsites currently under way. Copyright 1997 Society of Petroleum Engineers Original SPE manuscript received for review 11 March 1996. Revised manuscript received 11 March 1997. Paper peer approved 12 March 1997. Paper (SPE 35698) first presented at the 1996 SPE Western Regional Meeting held in Anchorage, 22–24 May.
SPE Reservoir Engineering, May 1997
The third and final CPF (CPF-3) was commissioned in 1986 to develop the northern portion of the field. An extensive fieldwide refracturing program was initiated in 1988 to increase withdrawal from the less prolific A-sand.7,8 An EOR pilot was initiated in 1988 on two drillsites,9,10 and subsequently expanded to one of t he IWAG IWAG drillsites in 1993 to test the MWAG after IWAG process. Largescale expansion of EOR at Kuparuk is in progress. Development capital expended to date has been nearly U.S. $5 billion to drill and equip more than 700 wells (approximately half of which are injectors) at 42 drillsites. Peak production reached 353,00 0 BOPD on 15 December 1991. With cumulative production topping 1 billion bbl on 26 May 1993, Kuparuk became only the 14th field in North America to reach this milestone. Ultimate recovery is now expected to top 2 billion bbl of oil at a total capital cost of around U.S. $7 billion. A simplified schematic depicting major Kuparuk facilities and pre-LSEOR recovery mechanisms is shown in Fig. 2. Geologic Setting. The Kuparuk River formation resides at a nominal depth of 6,000 ft; is a laterally extensive, slightly dipping anticline (0.8 ) covering over 200 sq miles; and comprises two primary producing horizons of Early Cretaceous sandstone. The trapping mechanism is a combination of structural closure to the north and east and stratigraphic closure to the south and west. The geologic history and reservoir stratigraphy have been discussed in detail elsewhere. 11-14 The lower horizon (A-sand) is subdivided into six imbricated intervals (A-1 through A-6) of interbedded sandstone, siltstone, and mudstone, which together contain roughly two-thirds of the approximately 6 billion bbl of original oil in place (OOIP). The upper horizon (C-sand) is hydraulically separated from the lower horizon by the Lower Cretaceous unconformity and is subdivided into four intervals (C-1 through C-4) of poorly sorted, glauconitic, sideritecemented, intensely bioturbated sandstones. While net pay for the C-sand can reach 80 ft, the average is just over 30 ft. Ave Average rage net pay for the A-sand is typically less than 30 ft. Production from both horizons are hydraulically commingled at the wellbore. The Kuparuk River formation is heavily faulted, with regional fluid movement influenced predominantly by the north-south fault system.9 Virtually all Kuparuk faults are normally ori ented and behave as barriers to flow in some cases and as fluid conduits in o thers. Throws of up to 300 ft are observed in some of the larger faults; however, fault throws of 30 to 8 0 ft are most common. 11 Faulting plays a key role in reservoir management and in determining infill-well placement. Even fault throws of as little as 10 ft are capable of disrupting the lateral continuity of otherwise floodable intervals.12 A map depicting major faults within Kuparuk is shown in Fig. 3. °
Rock and Fluid Properties. The reservoir was undersaturated at the time of discovery, with an average reservoir pressure of 3,100 psi, an average bubblepoint pressure of 2,800 psi, and a solutiongas/oil ratio (GOR) of 500 scf/STB.9 Oil gravity varies from 18 to 27 API throughout the field, but is 24 API on average. Table 1 contains a boiling-point breakdown of oil currently being shipped from the field. No discernible differences have been detected between Aand C-sand fluid samples. Asphaltene content (pentane insolubl es) ranges from 3 to 17 wt% of the C 7+ fraction. Laboratory experiments with the Kuparuk fluid system have identified the formation of a third hydrocarbon phase, high in asphaltene content, when miscible injectant (MI) is mixed with reservoir oil. Onset of third-phase formation was found to be d irectly related to the asphaltene content of the reservoir oil.15 The A-sands, with an average permeability-thickness of 1,000 mdft, are considerably less prolific than the C-sands, which have an aver°
°
83
Fig. 1—Kuparuk field locator map.
age permeability thickness of 5,000 md-ft.1,2 The more-permeable C-1 interval often behaves as a thief zone when present. Both sands have an average porosity in the 23 to 24% range. Wettability is intermediate. Laboratory derived values for S orw and S orm for the A-sand are approximately 42% and 5 to 13%, respectively. The laboratoryderived S orw for the C-sand is approximately 342% for a small number of pore volumes (PV’s) throughput. A single-well tracer test indicated that the S orw after many PV’s is approximately 27 3%.9 The tracer-test results were dominated by a thief zone. These two independent values of S orw are reasonable estimates for S orw in a few of the subunits of the C-sand at field conditions, which range from low to high throughput. An S orm value has not yet been measured in the C-sand. The high S orw values provide a significant target for EOR. Laboratory corefloods show that S orm increases with core dispersivity and with asphaltene content (Table 2).
In the mid-1980’s, the suitability of a miscible WAG process was reviewed for Kuparuk. The process appeared to be applicable in terms of the following.
1. The reservoir pressure of 3,000 psi (normal range required is between 1,500 and 3,000 psi 16,17). 2. The presence of a number of thin, low-permeability sands with vertical flow barriers, which are needed to minimize gravity segregation in horizontal floods.16,17 3. A high residual oil saturation of 28 to 42% (better than the preferred value of 25%17). 4. The required MI could be manufactured from approximately 70 to 80% lean gas and 20 to 30% enriching fluid, depending on the enriching fluid composition. 5. Lack of a gas cap. However, the oil properties were not ideal. The oil viscosity of 2 cp was not in the preferred range of less than 1 cp (although up to 5 cp is considered acceptable). Formation of a second liquid phase at high MI concentrations was a concern. The presence of highpermeability thief zones in the C-1 sand was also a concern. It was determined that the prudent development strategy was a phased approach. The initial stage was to manufacture some lowcost MI and evaluate the required MI quality and process performance on a number of patterns within two drillsites. Later, the volume of indigenous MI and number of patterns could be expanded. The third phase would be to import a much larger volume of enriching fluid from another field.
An EOR pilot project was implemented at Kuparuk in late 1988 at two drillsites (1Y and 2Z) that had previously been on waterflood. The pilot consisted of 11 patterns covering an area of 3,300 acres with an OOIP of 210 MMSTB. The EOR flood on 1Y and 2Z was bisected by an area set aside to evaluate waterflood performance on 80-acre well spacing. Fig. 4 depicts the original EOR and 80-acre waterflood pilot areas. Early performance of the original EOR pilot area has been discussed in detail elsewhere.9,10 The EOR response from the original project area, after accounting for the effects of subsequent infill drilling and well work, is clearly evident ( Fig. 5). This performance has been corroborated by extensive one-dimensional, 2D, and 3D reservoir simulation studies10 and is backed by 10 years’ worth of laboratory data involving more than 75 experiments ranging from static pressure/volume/temperature to sophisticated flow measurements.
Fig. 2—Kuparuk field area map. 84
SPE Reservoir Engineering, May 1997
TABLE 1—KUPARUK CRUDE DISTILLATION Crude Yield (vol %)
Component Propane Isobutane Normal butane Pentane To boiling point of 175 F Boiling-point range of 175 to 350 F Boiling-point range of 350 to 450 F Boiling-point range of 450 to 650 F Boiling-point range of 650 to 1,000 F At boiling points of 1,000+ F
0.2 0.1 0.5 3.0 11.6 8.1 21.6 27.9 27.0
°
° ° °
°
°
TABLE 2—KUPARUK COREFLOOD RESULTS FOR A-SAND AT 2 HPV INJECTED
Fig. 3—Kuparuk fault map.
Table 3 lists the types of experiments that have been performed. The effect of the extensive faulting on field performance has been addressed first by a gas tracer injection project in the plot area. 9 Direct communication between injectors and producers was confirmed across low throw faults. When the fault throw was on the order of the sand thickness, the faults act as sealing faults. These effects were included in performance analysis by defining the patt ern boundaries on the basis of the presence of sealing faults. Juxtapositioning of sands across faults was included in the 3D reservoir simulation studies. An example of site-specific simulation results is presented in Fig. 6. In 1992, the 80-acre waterflood area separating the 1Y and 2Z EOR pilot areas was converted to MWAG injection. Drilling of 11 infill wells at 1Y was initiated in late 1993 and drilling of two infill wells at 2Z in early 1995. This brings the total number of SSEOR patterns on 1Y and 2Z to 28. Areal SSEOR coverage stands at 5,800 acres, with a corresponding OOIP of 390 MMSTB. The MI for this pilot was made by blending separator off-gas with indigenous enriching fluids in the form of gas-scrubber and lowtemperature separator liquids. At CPF-1, the indigenous enriching fluids were made up of natural gas l iquids (NGL’s) from the CPF-1 NGL plant and scrubber liquids from the CPF-1 artificial-lift gascompression system. At CPF-2, the indigenous enriching fluids were scrubber liquids from the CPF-2 artifi cial-lift gas compression system. The injection history for the SSEOR pilot is shown in Fig. 7. This project was able to make approximately 30 MMscf/D of MI from the produced fluids. With recycle (bootstrapping), which is discussed later, the injection rate rose to approximately 50 MMscf/D. The MI injection rate dropped in 1995 owing to lengthy equipment downtime associated with startup of the SSEOR expansion project described below. Incremental EOR recovery from the SSEOR pilot to date has been estimated to be 15 to 20 MMSTB, which represents 3.5 t o 5% of the OOIP. The ultimate incremental EOR recovery is expected to be 30 to 40 MMSTB, which corresponds t o 7 to 10% of the OOIP.
The SSEOR pilot-expansion project, once fully operational, increased the volume of MI that could be manufactured from Kuparuk fluids by approximately 14 MMscf/D. At CPF-1, the increased MI capacity came from additional scrubber liquids in the artificial-lift gas-compresSPE Reservoir Engineering, May 1997
Asphaltene Content (%)
Peclet Number
S orm
3.8 12.0 3.8 12.0
648 648 64 64
5.3 7.8 10.5 13.0
(%)
Component Recovery (%) C10 to C22 C36+ 98 98 91 93
87 81 74 68
sion system and naphtha from the Kuparuk River Unit crude-oil topping plant. At CPF-2, the increased MI capacity came from additional scrubber liquids in the artificial-lift gas-compression system and NGL’s from the fuel-gas conditioning skid. With these plant modifications, all low-cost MI liquids available at CPF-1 and CPF-2 were extracted. Although early performance from the SSEOR pilot was very encouraging, it did not correspond directly to the reservoir mechanisms associated with the majority of future expansion opportunities. As noted previously, the original SSEOR pilot was initiated following waterflood. Future expansions will be almost exclusively into areas previously on IWAG. Taking advantage of the existing IWAG high-pressure-gas distribution infrastructure is key to the economic viability of expanding Kuparuk EOR. A number of mechanistic, fully compositional simulation studies were undertaken to address the impact of prior IWAG injection on the MWAG process. These studies incorporated compositionally consistent relative permeability treatment and the effects of hysteresis on gas trapping. Of particular concern was t he change in in-situ oil composition following lean-gas injection, miscible gas dilution by prior lean-gas injection, and the potential formation o f preferential gas paths that could l ead to excessive cycling of the miscible gas and thus lower incremental EOR recoveries. Simulation results indicate that IWAG will negatively impact incremental MWAG recoveries, but that the magnitude is minimal and is not sufficient to alter the decision whether to proceed with the project. The reduction in MWAG recovery is primarily the result of a reduction in target oil saturation attributed to t he IWAG recovery process. The combined recovery from MWAG and IWAG was found to be great er than tha t of MWAG alone by an av erage of 1% OOI P. However, because IWAG recoveries on a stan dalone basis average 2% OOIP, the stand-alone MWAG incremental is reduced by the difference, or 1% OOIP. The incremental average IWAG recovery of 2% OOIP is consistent with the 1 to 3% OOIP reported elsewhere for Kuparuk. 6 Although simulation results indicate that MWAG/IWAG interaction should not be a concern, there were no known precedents for IWAG preceding MWAG. As such, there were neither field data nor experience to rely on for assessing the impact of IWAG on the MWAG process. Given the magnitude of investment associated with a major EOR expansion at Kuparuk, a field test was initiated in 1993. The existing pilot was expanded to an adjacent mature IWAG drillsite (1A). This expansion increased the total number of EOR patterns in the field to 40 and t he total areal EOR coverage to 85
Fig. 4—SSEOR area map.
8,700 acres, with a corresponding OOIP of 575 MMSTB. As shown in Fig. 8, early field results from 1A show that the miscible process has worked effectively following IWAG.
The SSEOR and SSEOR pilot-expansion project s manufacture all the low-cost MI possible from the Kuparuk CPF-1 and CPF-2 enriching fluids. Therefore, any major expansion of EOR at Kuparuk requires solvent from an external source.9 Without solvent import, future EOR expansions would be limited to between two and five additional drillsites. With solvent import, the number grows to 18 or more. Project Overview. As depicted in Fig. 2, a high-pressure-gas distribution system was install ed as part of the Kuparuk I WAG project. This existing infrastructure forms the facility backbone of the LSEOR project, serving to enhance the economic viability of the project while minimizing capital investment risk. Reliance on existing infrastructure has always played a key rol e in North Slope development. Had it not been for the infrastructure constructed as a result of the Prudhoe Bay d iscovery, Kuparuk itself would not have been economically viable.14 Current plans are to expand EOR into all 18 drillsites currently on IWAG. This expansion will be staged, with 9 to 12 drillsites initially converted to MI injection. The remaining six to nine drillsites will be converted as the initial drillsites mature. The significant volume of MI required for this EOR expansion will be provided by b lending NGL’s from Prudhoe Bay with Kuparuk’s produced gas. Prudhoe Bay NGL’s are the only North Slope enriching fluid that meets Kuparuk’s requirements of immediate availability, reservoir compatibility, and facility compatibility. 86
While other solvent sources may become available in the future, Prudhoe Bay NGL’s are the only substanti al source of North Slope enriching fluid currently available for use at Kuparuk. Although miscibility at Kuparuk is controlled by the combined condensing/ vaporizing mechanism,18 the reservoir’s strongly condensing character makes it well suited to an NGL-based solvent. Because the Kuparuk reservoir has very little indigenous CO 2, introducing CO2 into the system could potentially trigger significant corrosion mitigation costs. Prudhoe Bay NGL’s do not pose a CO2 concern. Facility modifications associated with the LSEOR project are depicted in Fig. 9 and include the following. 1. Tie-ins at Prudhoe to allow a portion of the Prudhoe Bay NGL stream to be diverted into the ori ginal 16-in. pipeline between Kuparuk and TAPS (the “Oliktok” pipeline).
Fig. 5—SSEOR production history. SPE Reservoir Engineering, May 1997
TABLE 3—KUPARUK EOR EXPERIMENTAL DATA LIST Data Types
Number of Experiments
Utility
Saturation pressures, saturation temperatures, densities
12
Black-oil endpoint; correct component volatility.
Injectant properties
14
Proper phase behavior after injection; ensure single phase when injected.
3 6 21 3
Examine condensing mechanism. Examine vaporization mechanism. Tune to specific conditions; viscosities. Extension of single contact utility.
13 2 2
Low dispersion displacement; multicontact miscibility. Higher dispersion displacement; multicontact miscibility. Visual observation of miscible process.
Single-cell solvent/oil contacts Swelling Titration Single contacts Multiple contacts Solvent displacements Slim Tube Corefloods Micromodel
2. Modification of the Oliktok pipeline for conversion from gas to NGL service. 3. Installation of three pipeline pumps at CPF-1 to boost the pressure of the incoming NGL’s for transport to CPF-2 vi a a new 8-in. NGL line and for entry into the new enriching fluid collection drums to be installed at both CPF-1 and CPF-2. 4. Rewheeling the SSEOR expansion-pilot injection pump at CPF-1 from 220 to 360 gal/min. 5. Replacing the existing 180-gal/min SSEOR injection pump at CPF-2 with a new 800-gal/min injection pump. 6. Linking the CPF-1 and CPF-3 power systems to accommodate the increased electrical demand at CPF-1, especially during winter months when power consumption peaks. While typical North Slope modular development can require 3 years from the start of final design engineering to plant startup,19 the LSEOR project is expected to be fully operational within 18 months of final engineering owing to the use of truckable modules. Unlike sea-lift designs, which have a limited 6-week North Slope delivery window, truckable designs can be transported to the North Slope virtually year-round. This has enabled startup of the CPF-1 portion of the project within 10 months of final engineering; well in advance of full project startup.
approximately 230 patterns on 21 dri llsites. Areal coverage will be increased from 8,700 acres to approximately 68,000 acres, with the corresponding OOIP coverage increasing from 575 to 2,900 MMSTB. Incremental reservoir recovery resulting from the LSEOR project is expected to top 200 MMSTB. These factors easily place the Kuparuk LSEOR project among the largest hydrocarbonmiscible gasfloods in the world.20 Cumulative NGL-import volume over the life of the project will be approximately 100 MMSTB, one-third of which is expected to be recaptured and sold as part of the Kuparuk oi l stream. This recaptured NGL volume is above and beyond the 200 MMSTB incremental reservoir recovery noted earlier. While the number of factors that go in making a go/no-go decision on a project of this magnitude are well b eyond what can be condensed into a singl e paper, some of the more salient aspects are described next. Each, taken in combination with the positive pilot performance observed in the field, contributes to the confidence needed to proceed with project funding and execution. Recovery Estimates. The recovery estimates for the LSEOR project are based on (1) injection of 1 Tcf of MI, (2) execution of the planned infill-drilling program, (3) adequate well injectivity, and
The LSEOR project will more t han triple Kuparuk’s MI injection capacity, raising the annual average injection rate from 65 to 220 MMscf/D. Coupled with infill drilling, the number of patterns receiving MI will increase from 40 patterns on three drillsites to
Fig. 7—SSEOR MI injection history.
Fig. 6—DS-2Z simulation vs. performance. SPE Reservoir Engineering, May 1997
Fig. 8—Drillsite 1A EOR production. 87
Fig. 9—LSEOR facility schematic.
In Fig. 10, diminishing returns are evident as slug size increases. Results from 3D modeling of drillsite 2Z on the basis of the Todd-Longstaff 21 approach have been discussed elsewhere. 10 In that work, the mixing parameter, , was adjusted to fit historical performance. It was implicitly assumed that the value remained const ant for a variety of slug sizes. The predictions were consistent with those derived from finely gridded compositional simulations. A 30% HPV slug size will be targeted at the main sand in each pattern. Because of different Asand and C-sand injectivities and volumes, not all sands will receive a 30% slug. As such, the incremental recovery from individual drillsites is expected to range from 4 to 10% OOIP. EOR Process Efficiency. Table 4 provides a comparison of the process characteristics between MI made from Prudhoe Bay NGL’s and MI made entirely from enriching fluids indigenous to Kuparuk. The actual MI will be a combination of these two sources.
Fig. 10—LSEOR “type” curves.
(4) expansion into all targeted drillsites. Simulation studies indicate that, while the rate of injection affects recovery timing, tot al recovery is not impacted provided these conditions are met. To ensure reasonably that estimated recoveries are achieved wit hin the expected remaining life of the Kuparuk River field, the scope of the LSEOR project was designed so that i njection of the 1 Tcf of MI will be completed within 15 to 20 years of project startup. The 1 Tcf of MI will be generated from a combination of imported Prudhoe Bay NGL’s and enriching fluids indigenous to Kuparuk, including the solvent bootstrap volumes discussed later. Reservoir modeling and field performance indicate that the optimum MI slug size is 20 to 30% on a hydrocarbon PV (HPV) basis.
Third-Hydrocarbon-Phase Formation. The formation of a third heavy-liquid phase was of initial concern to Kuparuk EOR development. Onset of third-phase formation was observed to be directly related to the compositions generated during the multicontact miscible mechanism of the condensing/vaporizing process. Simulation results with Michelsen’s three-phase algorithm22 indicate that a third phase forms near the critical point and that all three phases are rich in solvent range components. The impact of third-phase formation is no longer viewed as being a major cause for concern because (1) all three phases are light, with reasonably similar physical properties; (2) the heaviest, most dissimilar phase only occurs in very low saturations, typically less than 10%; (3) the three phases appear for only a small range of compositions; and (4) there is no experimental (slim tube, coreflood, and micro-model) or field evidence indicating that there is an impact on the process or project performance. Additional details regarding these observations are re-
TABLE 4—EOR PROCESS EFFICIENCY MI Enriching Fluid (EF)
Enrichment at MMP of 2,900 psi, STB of EF/Mcf of MI Gross solvent utilization, STB of EOR oil per Mcf of MI injected Gross solvent utilization, STB of EOR oil per total STB of EF injected Net solvent utilization, STB of EOR oil per net STB of EF injected Net solvent utilization, STB of EOR oil per STB of trapped saleable EF †
Prudhoe NGL’s
Kuparuk Indigenous
0.170 0.22 1.3 1.8 2.7
0.215 0.20 0.93 1.5 4.4
†Assumes 50% of Kuparuk indigenous EF and 100% of Prudhoe Bay NGL’s are saleable liquids before injection.
88
SPE Reservoir Engineering, May 1997
TABLE 5—MMP CORRELATION vs. EXPERIMENTAL RESULTS Slim-Tube Experiment
Experimental MMP (psi)
Correlation MMP (psi)
1 2 3 4 5 6 7 8 9 10 11 12 13
2,825 3,500 3,200 2,825 2,825 2,825 2,825 2,825 2,825 2,150 3,600 3,850 3,050
3,039 3,538 3,273 2,675 2,999 2,665 2,819 2,799 2,702 2,540 3,431 3,848 2,801
TABLE 6—TYPICAL LSEOR LEAN-GAS AND NGL COMPOSITIONS Component
Lean Gas (mol%)
Prudhoe NGL’s (mol%)
CO2 C1 C2 C3 i -C4 n -C4 i -C5 n -C5 C6 C7 C8+
1.0 85.0 8.0 4.0 0.5 1.0 — 0.5 — — —
— — — 0.1 4.1 24.5 14.4 19.8 19.2 12.5 5.4
ported by Godbole et al.15 The most significant effect might well be a slight increase in S orm already present in the laboratory data. Injectivity. Injectivity calculations accounting for three-phase flow and gas trapping indicate that water injectivity should be noticeably reduced as a result of gas injection. However, this effect has not been observed in the field data. Overall, water injectivit y does not appear to have been affected, while gas injectivity is 15 to 20% greater than
that of water. This is in contrast to several CO 2 WAG injection proj ects in west Texas that have seen substantial losses in water in jectivities following gas injection.23 Minimum Miscibility Criteria. To ensure that the resultant mixture of enriching fluids and separator off-gas will achieve the desired miscible behavior in the reservoir, a simplified minimum miscibility pressure (MMP) correlation that can be implemented in the field is needed. In developing this correlation, the primary objectives were to (1) focus on the NGL solvent to be used in the LSEOR project, (2) base the correlation strictly on experimental data, (3) minimize the number of adjustable parameters, and (4) maintain consistency with the equation-of-state (EOS) model. Too many parameters can limit the use of a correlation to the immediate vicini ty of the data points from which it is derived and can present a level of complexity that is problematic from a facility-controls standpoint. As shown in Table 3, a total of 13 slim-tube experiments has been conducted. These experiments cover a wide range of solvents, from very lean MI to relatively heavy (high molecular weight) MI. Because the empirical MMP correlation was developed over a wide range of MI compositions, the correlation will robustly handle seasonal variations in the enriching fluid and gas compositions. Data from these experiments exhibit a clear trend whereby the MMP increases slightly for heavier oils and decreases significantly with increasing MI molecular weight. The following equation represents the MMP correlation that will be used in controlling blending levels in the field. p M , min
ab
y T M i
c ci
C 7.
. . . . . . . . . . . . . . . . . . . . (1)
The summation is over all single carbon number components in the MI, with i-C4, n-C4, i-C5, and n-C5 being treated separately (i.e., they are not combined together). The critical temperature for CO 2 is taken to be 435 R rather than 548 R on the basis of the fact that CO 2 is known to be much more effective in developing miscibility17 than a T c of 548 R would indicate. The dependence of MMP on oil composition is reflected by a 100-psi increase for every 10-unit increase in M C . 7 As shown in Table 5, the correlation has an average prediction error of 137 psi, which is withi n the 200 psi uncertainty in the experimental data. Further, there is essentially no error for th e experiment with MI made fr om NGL’s (Slim-tube Series 7), which is the closest analog to the LSEOR injectant. For the t ypical LSEOR lean-gas and NGL compositions presented in Table 6, a blending ratio of 18.7 mol% NGL’s renders the desired MMP o f 2,900 psi for an oil having an M C of 285. °
°
°
7
Fig. 11—Bootstrap process. SPE Reservoir Engineering, May 1997
89
Fig. 12—Kuparuk EOR enriching fluid.
Infill Drilling. Infill drilling is directly related to EOR recovery at Kuparuk in at least two different ways. First, infill drilling provides improved reservoir sweep through underperforming patterns and increased reservoir contacting in areas of partial or total structural isolation (see Fig. 3). The more reservoir oil that can be contacted wi th MI, the higher the EOR recovery potential. Second, infill drilling provides for increased injectivity on a given drillsite, allowing the target slug size to be injected in a shorter period of time. This allows for more rapid drillsite expansion and reduces field life associated risks. Up to 66 new wells are expected to be drill ed within the Kuparuk LSEOR area. These wells include 13 peripheral wells, 19 pattern infill wells, and 34 structural infill wells. The incremental EOR recoveries cited earlier do not include the non-EOR incremental benefits associated with the infill-drilling program. Solvent Bootstrap. “Bootstrap” refers to the process by which in jected fluids are reproduced and become available for reuse. Fig. 11 is a schematic representing the process. During the first pass through the reservoir, 40 to 45% of an injected barrel of NGL-range components is trapped in the reservoir and 55 to 60% returns to the surface. Of the returned volume, approximately 40% is stabilized in the oil and sold, while the other 60% is available for reinjection (determined by compositional reservoir and surface-facility simulation). For LSEOR, this process, on average, will complete two to three cycles. In this way, the 100 million bbl of imported NGL’s will be injected approximately 1.4 times, or the equivalent of 140 million bbl. The NGL-import volume is directly influenced by the desired MI enrichment level, pump capacity constraints, and the degree of bootstrap. The higher the bootstrap, the lower the required NGL import and vice versa. Fig. 12 depicts the anticipated NGL-import volumes along with the combined indigenous/bootstrap enrichingfluid rate profiles.
Fig. 13—LSEOR dimensionless recovery curves.
3. Multiply each pattern’s dimensionless EOR and MI production by the pattern’s OOIP. 4. Sum all pattern recoveries. Because the LSEOR project scope includes approximately 230 patterns of multiple zones, the scaleup tool treats each of the 18 dril lsites as two patterns—one each for the A-sand and the C-sand. Even with reliable dimensionless recovery curves, the scaleup process will produce unreliable performance forecasts unless each drillsite’s MWAG conversion timing and injection rate are accurately predicted. The A-sand and C-sand MWAG injectivities (% HPV/ year) for each drillsite have been calculated on the basis of the drillsite’s historical gas and water injectivities and estimated future conditions, such as number of injection wells, WAG ratio, reservoir pressure, MI composition, and MI injection wellh ead pressure. The drillsite conversion timing, or project development scenario, is determined primarily by facility constraints. For example, the field MI injection rate will be constrained by either a contractual NGL-import rate limit or by the injection-pump capacity, which prohibits simultaneous MWAG conversion of all 18 LSEOR drillsites. On the other hand, the MI distribution piping configuration requires th at specific drillsites convert t o MWAG simultaneously because the high-pressure gas trunk line to each group of drillsites can carry either lean or enriched gas, but not both. Another potential facility impact on the LSEOR project is production-well proration based on high GOR or water/oil ratio (WOR). While production and injection prorat ion are common in the Kuparuk field, these constraints were not incorporated in to the scaleup process. On the basis of reservoir simulation of Kuparuk’s MWAG process following IWAG, it is expected that the LSEOR project wi ll maintain or even reduce the average GOR and WOR of the affected wells.
Reservoir Surveillance. In terms of operating the LSEOR project in the field, there are certain data that should be collected to answer the following two questions: “How well is the process proceeding?” Reservoir Scaleup. Laboratory studies, reservoir simulations, and the and “How can the process be improved?” The information required SSEOR pilot project provided confidence in the EOR performance of to address these questions can be broken down into t hree categories: a typical Kuparuk MWAG pattern. The next challenge was to scale up (1) placement of MI, (2) location and amount of MI producti on; and individual pattern performance to predict the composite injection re- (3) maintenance of optimal miscibility. The first consideration, inquirements and production benefits associated with 230 MWAG pat- jectant placement, involves metering the total amount of injected terns in various stages of maturity over a 15- to 20-year project life. The miscible gas and the split of MI among the dominant zones. This is cornerstone of the reservoir scaleup process is the dimensionless recov- one of the primary criteria for MI scheduling and pattern abandonery curve. By use of the MWAG process evaluation techniques disment. The second issue, MI production, is slightly more complex. cussed in this paper, unique curves describing EOR and MI recovery Gas compositions and oil gravities are enormously useful for deter(% OOIP) as functions of total MWAG injection (% HVP) were estab- mining when EOR response has been observed and how much relished for several ultimate MI slug sizes. Two such EOR curves are turned MI and lean gas make up the produced gas. Single-zone proshown in Fig. 13. Multiple dimensionless curve sets are needed to scale duction tests are used t o determine from which zone water, gas, and up patterns with various WAG ratios. For simplification, the LSEOR oil production originates. Finally, miscibility is ensured by mainanalysis assumed a constant WAG ratio of 1:1. taining reservoir pressure, which is estimated through pressure fallConceptually, the scaleup process involves the following steps off at injectors and pressure buildup at producers. for each time increment. The initial focus will be directed toward optimal placement of MI 1. Predict the MI and water volumes injected in each MWAG by avoiding areas where MI thief zones exist. When these are depattern. tected, MI will be diverted from these patterns, and remedial action 2. Find the corresponding EOR and MI production for each pat- will be taken, if warranted. Once the first-pass optimization is under tern from the appropriate dimensionless curves. way, further optimization can proceed. One example is the determina-
90
SPE Reservoir Engineering, May 1997
7. One-third of the imported NGL’s is expected to be recaptured and sold with the Kuparuk oil stream. 8. Infill drilling within the LSEOR project area will provide for improved reservoir contacting and the ability to reach targeted slug sizes for a given drillsite in a shorter period of time. 9. With less than half of the Kuparuk OOIP subject to the EOR recovery process, the potential for future EOR expansions at Kuparuk is significant.
Injectant Enrichment Level (mol%) Fig. 14—Typical Kuparuk slim-tube results; MMEminimum miscibility enrichment.
tion of pattern-abandonment criteria on the basis of field data and compositional reservoir simulations. Produced-gas samples and cumulative MI injection will be the basis for determining when (or if) MI should be replaced by lean-gas injection. Compositional simulations will be used to develop guidelines on the threshold amount of returned MI in a pattern before the pattern should be abandoned or placed on lean-gas injection to recover the remaining enriching fluids.
The opinions expressed in this paper are the authors’ and do not necessarily represent the views of the Kuparuk River Unit owners. We thank Arco Alaska Inc. and BP Exploration (Alaska) Inc. for permission to publish this paper and all others who contributed to the successful evolution of this project. Special thanks go t o Gary Youngren for his hands-on guidance on reservoir simulation and phase behavior issues and to Andy Spence for directing a highly successful laboratory program.
a regression constant1.70103 b regression constant2.66104 c regression constant1.6 M C molecular weight of the C 7+ fraction of the oil 7 p M ,min minimum miscibility pressure, m/Lt2, psi S orw residual oil to waterflood, % total PV S orm residual oil to miscible gas, % total PV yi mole fraction of Component i in the MI T ci critical temperature of Component i, T, R °
Future Opportunities. Even with the LSEOR project, less than half of Kuparuk’s OOIP will be under EOR. Access to the remaining OOIP presents a formidable challenge given the lack of facilities infrastructure. However, a project is under way to use the NGL in jection pump that will be removed from service as a result of the LSEOR project. High-pressure-gas piping will be extended from drillsite 1Q to CPF-3, where two drill sites will be placed on IWAG and a third placed on MWAG. With two-thirds of the injected NGL’s remaining trapped in the reservoir following LSEOR, the merits of conducting a lean-gas chase project will certainly be investigated. As shown in Fig. 14, Kuparuk’s EOR process exhibits a gradual transition between miscible and immiscible behavior. With the LSEOR project being highly leveraged by NGL-volume requirements, opportunities to optimize the blend ratio will be continually scrutinized. As field experience is gained, operating parameters (such as WAG ratios and conversion timing) will also be refined. And finally, should Kuparuk development expand to include other potential horizons and/or nearby satellite accumulations, opportunities for expanding EOR will likely be more than just a passing consideration.
The Kuparuk LSEOR project is among the largest of its type in the world. Incremental oil recovery from the project is expected to top 200 MMSTB. The enriched-gas EOR recovery process at Kuparuk has been confirmed both in the laboratory and in the field with two successful pilot programs. The following are key aspects of the Kuparuk LSEOR project. 1. Maximizing the utilization of existing infrastructure enhances economic viability and minimizes capital investment risk. 2. Significant EOR expansion at Kuparuk requires more solvent than is available from indigenous sources. 3. Approximately 100 MMSTB of Prudhoe Bay NGL’s will be imported during the life of the project. 4. MI injection capacity will be increased by more than a factor of three. 5. EOR coverage at Kuparuk will be increased by 18 drillsites through a staged expansion program. 6. The strongly condensing character of the Kuparuk fluid system is well suited to an NGL-based solvent. SPE Reservoir Engineering, May 1997
1. Weiss, J.L. et al.: “Large-Scale Facility Expansion Evaluations at the Kuparuk River Field,” paper SPE 20046 paper at t he 1990 SPE California Regional Meeting,Ventura, California, 4–6 April. 2. Starley, G.P. et al.: “Application of Full-Field Simulation to Development Planning and Reservoir Management at t he Kuparuk River Field,” paper SPE 20045 presented at the 1990 SPE Cali fornia Regional Meeting,Ventura, California, 4–6 April. 3. Stoisits, R.F. et al.: “Application of Nonlinear Adaptive Modeling for Rigorous Representation of Production Facilities in Reservoir Simulation,” paper SPE 24898 presented at the 1992 SPE Annual Technical Conference and Exhibition, Washington, DC, 4–7 October. 4. Stoisits, R.F., Scherer, P.W., and Schmidt, S.E.: “Gas Optimization at the Kuparuk River Field,” paper SPE 28467 presented at the 1994 SPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, 25–28 September. 5. Champion, J.H. and Shelden, J.B.: “An Immiscible WAG Injection Project in the Kuparuk River Unit,” paper SPE 16719 presented at the 1987 SPE Annual Technical Conference and Exhibition, Dallas, 27–30 September. 6. Ma, T.D. and Youngren, G.K.: “Performance of Immiscible Water-Alternating-Gas (IWAG) Injection at Kuparuk River Unit, North Slope, Alaska,” paper SPE 28602 presented at the 1994 SPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, 25–28 September. 7. Pospisil, G. et al.: “Results of a Large-Scale Refracture Stimulation Program, Kuparuk River,” paper SPE 24857 presented at the 1992 SPE Annual Technical Conference and Exhibition, Washington, DC, 4–7 October. 8. Pearson, C.M. et al.: “Optimal Fracture Stimulation of a Moderate Permeability Reservoir, Kuparuk River Unit,” paper SPE 20707 presented at the 1990 SPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, 23–26 September. 9. Reinbold, E.W. et al.: “Early Performance and Evaluation of the Kuparuk Hydrocarbon Miscible Flood,” paper SPE 24930 presented at the 1992 SPE Annual Technical C onference and Exhibition, Washington, DC, 4–7 October. 10. Hallam, R.J., Ma, T.D., and Reinbold, E.W.: “Performance Evaluation and Optimization of the Kuparuk Hydrocarbon Miscible Water-Alternating-Gas Flood,” New Developments in Improved Oil Recovery, H. J. de Haan (ed.), Publication No. 84, Geological Soc. (1995) 153–64. 11. Masterson, W.D. and Paris, C.E.: “Depositional History and Reservoir Description of the Kuparuk River Formation, North Slope, Alaska,” Alaskan North Slope Geology, Pacific Section SEPM, Bakersfield, California, and AGS, Anchorage, AK (1987) 1, 95. 91
12. Gaynor, G.C. and Scheihing, M.H.: “Shelf Depositional Environments and Reservoir Characteristics of the Kuparuk River Formation (Lower Cretaceous), Kuparuk Field, North Slope, Alaska,” Giant Oil and Gas Fields, SEPM Core Workshop No. 12 (1988) 1, 333. 13. Carman, G.J. and Hardwick, P.: “Geology and Regional Setting of Kuparuk Oil Field, Alaska,” AAPG Bulletin (June 1983) 67, 1014. 14. Masterson, W.D. and Eggert, J.T.: “Kuparuk River Field—U.S.A.: North Slope, Alaska,” Stratigraphic Traps III (1992) 257. 15. Godbole, S.P., Thele, K.J., and Reinbold, E.W.: “EOS Modeling and Experimental Observations of Three-Hydrocarbon-Phase Equilibria,” paper SPE 24936 presented at the 1992 SPE Annual Technical Conference and Exhibition, Washington, DC, 4–7 October 1992. 16. Herbeck, E.F, Heintz, R.C., and Hastings, J.J.: Fundamentals of Tertiary Oil Recovery, Energy Publication (1982). 17. Stalkup, F.I. Jr.: Miscible Displacement, SPE Monograph Series, SPE, Richardson, Texas (1984) 8. 18. Zick, A.A.: “A Combined Condensing/Vaporizing Mechanism in the Displacement of Oil by Enriched Gases,” paper SPE 15493 presented at the 1986 SPE Annual Technical Conference and Exhibition, New Orleans, 5–8 Oct ober. 19. Bass, J.W.: “Modular Route for Prudhoe Bay,” CEP (November 1982) 70. 20. Moritis, G.: “Production Report: OGJ Biennial EOR Survey,” Oil & Gas J. (26 September 1994) 92, No. 39, 51. 21. Todd, M.R. and Longstaff, W.J.: “The Development, Testing, and Application of a Numerical Simulator for Predicting Miscible Flood Performance,” JPT (July 1972) 874. 22. Michelsen, M.J.: “The Isothermal Flash Problem Parts 1 and 2,” Fluid Phase Equilibria (1982) 9. 23. Schneider, F.N. and Owens, W.W.: “Relative Permeability Studies of Gas-Water Flow Following Solvent Injection in Carbonate Rocks,” SPEJ (February 1976) 23.
Hoolahan
acre 4.046 873 E01 ha 141.5/(131.5API) g/cm3 API bbl 1.589 873 E01 m3 cp 1.0* E03 Pas ft 3.048* E01 m ft3 2.831 685 E02 m3 C (F32)/1.8 F gal 3.785 412 E03 m3 in. 2.54* E00 cm mile 1.609 344* E00 km psi 6.894 757 E00 kPa sq mile 2.589 988 E00 km2 *Conversion factor is exact.
McDuffie
Hallam
SPERE
Circle 1 on Reader Service Card 92
SPE Reservoir Engineering, May 1997
