Transient behavior is likely to dominate over most of the duration of a foam injection field project. Due to the lack of experimental data, little is presently known about foam flow transient behavior. Foam flow does not follow established models such as the Buckley-Leverett theory and no general predictive model has been derived. Therefore, experimental data are badly needed. This study investigates transient foam flow in a simple linear porous medium.
Foam is injected at a constant volume rate into a one-dimensional sandpack of 1-inch diameter and 24-inch length, initially saturated with distilled water. The system is placed in a CAT-Scanner. Data are accumulated at low temperature and pressure including the pressure field and saturations obtained by scanning. The pressure field and saturations obtained by scanning. The liquid saturation can be obtained at each point of a cross section of the pack with less than 1% error over the range of interest. Calculations from the numbers obtained from the scanner show that the best special resolution available is a volume element (voxel) of 0.5x0.5x5 mm. Hence a relationship can be found between saturation and pore volume injected for each location in the sandpack. The pressure profiles show that the pressure drop along the sandpack is not evenly distributed and varies with time. The pressure gradient is much greater between the injection end and the foam front than it is ahead of that front. Moreover, the pressure gradient keeps changing as the foam advances in the sandpack. It is this behavior that differs from Buckley-Leverett theory. The cat-scan results demonstrate that the foam displacement is not piston-like. Gas channeling appears near the front, and eventually the foam blocks all these channels. The foam flows through the sandpack, con tinuously breaking and reforming. It takes two or more pore volumes of foam injection to reach residual liquid pore volumes of foam injection to reach residual liquid saturation.
The saturation versus pore volumes relationships for a given section of the sandpack have been matched with exponential formulas. In general the matches are satisfactory, however the exponents and constants in the formulas vary as a function of the position of the section considered. Graphing the data as saturation versus equivalent pore volumes injected reduces the range of these variations but does not eliminate them. We are unable to explain this behavior to date.
Techniques of measurement of saturations by X-ray CAT-scanning as well as interpretation of CT data will be presented.
Gas injection is one of the many enhanced recovery techniques. But gas injection is characterized by high gravity forces and high mobility ratios since gas has lower density and viscosity than reservoir oils. Thus gas projects are prone to gravity segregation, channeling and fingering, which cause early breakthrough with much of the oil left behind the displacing front. The most widely used and most successful enhanced oil recovery methods are the various thermal recovery techniques. Cyclic steam injection and steam drives are typically applied to heavy oil reservoirs with good results. Because of their higher ultimate recovery, steam drives are becoming the dominant thermal recovery technique.
Small volume hydraulic fracture tests, defined as microfrac tests, are conducted to determine the in situ principal stresses in oil formations. Interpretation difficulties arise when the tests are conducted in relatively shallow, unconsolidated sands which are typical of the oil sand deposits in Alberta, Canada. Interpretation methods for such tests are outlined and their use is discussed. Three microfrac case histories in the Athabasca oil sands deposit are presented and analyzed. Recommendations are given for performing microfrac tests in these formations. performing microfrac tests in these formations
The past fifteen years have witnessed the progressive development of the large progressive development of the large hydrocarbon reserves, 1350 x 10 to the 9th bbl (214 x 10 to the 9th m), located in the oil sands deposits of Alberta. Two major surface mining operations are currently in operation in the Athabasca deposit, but less than 10% of the area is shallow enough to allow surface mining operations. Therefore, other recovery strategies are being investigated or used for the remaining oil sands. Esso Resources is operating a full scale in situ production project in the Cold Lake deposit production project in the Cold Lake deposit involving steam stimulation techniques. Due to high bitumen viscosities (e.g. 3 x 10 to the 6th cp (3.0 x 10 to the 3rd Pa.s)) under reservoir conditions, the bitumen will not readily flow through the sand matrix. Thermal methods, such as cyclic steam stimulation and steam flooding, are used to improve the bitumen mobility. In the most common technique, cyclic steam stimulation, the payzone is fractured to improve the areal payzone is fractured to improve the areal conformance of the steam.
Field performance and research have shown that the in situ stress state is the first order control on hydraulic fracture behaviour. For oil sands, shear failure may occur prior to tensile fracturing depending on the relationship between the in situ stresses and the formation pore pressure. Rational design of well patterns, pore pressure. Rational design of well patterns, spacing, and operating procedures requires knowledge of the in situ stresses along with geomechanical properties of the oil sands. The small volume hydraulic fracture test, the microfrac test, is the only method available for measuring in situ stress at reservoir depth. Although hydraulic fracturing stress measurements have been practiced for many years, it has only been recently that the test method has been standardized. Furthermore, the analysis of hydraulic fracturing tests for determining in situ stresses remains the subject of varying interpretation as shown by the papers of the 1988 International Workshops on Hydraulic Fracturing Stress Measurements.
The objective of this study is to measure the phase behavior and physical properties of a Canadian phase behavior and physical properties of a Canadian heavy crude oil (11 degrees API) saturated with carbon dioxide as a function of pressure and temperature. These data are used to evaluate the feasibility of using carbon dioxide as a pretreatment for steam floods or as a steam additive.
Experimental measurements were carried out at 69.8 degrees F (21 degrees C), the reservoir temperature, and at 284 degrees F (140 degrees C). The saturation pressures were between 400 psi (2.8 MPa) and 1,500 psi (10.3 MPa). At each pressure level the solubility of CO in the oil, the pressure level the solubility of CO in the oil, the density, and the viscosity of the saturated mixture were measured.
The results showed an increase in the solubility of CO in the oil with increasing pressure and decreasing temperature, while the densities of the mixtures did not change to a large extent with changing compositions. The most dramatic effect observed was the decrease in viscosity with CO dissolution in the oil: A 53-fold reduction occurred at 69.8 degrees F (21 degrees C) and a 2-fold reduction at 284 degrees F (140 degrees C).
SARA analyses of various samples were carried out. It was found that at 69.8 degrees F (21 degrees C) some redistribution of asphaltenes and resins occurred at pressures above the liquifaction pressure of CO (850 pressures above the liquifaction pressure of CO (850 psi or 5.9 MPa). psi or 5.9 MPa). These results indicate that, from a phase behavior point of view, enhanced oil recovery by CO injection in the heavy oil reservoir is feasible, and that viscosity reduction would play a major role in improving productivity at lower temperatures. The asphaltenes/resins balance should be taken into account when selecting the injection pressure.
Conventional use of CO for improving oil recovery has been mostly for miscible flood applications where the displacement of crude oil from the pore space within the reservoirs is achieved by the solvent action of CO which prevents formation of an interface between the driving and the driven fluid. The displacement process is very effective and potentially 100% of the oil in place could be recovered from the swept zone.
However, in heavy oil and bituminous reservoirs such miscibility cannot be obtained and it will not be a factor in increasing oil recovery. Rather, the enhancement of oil recovery will be the result of the high solubility of CO in low API gravity oils which in turn would result in:
(1) A substantial viscosity reduction - 10-fold or more can be easily achieved.
(2) Considerable oil swellin (10-30%) depending on the saturation pressure, reservoir temperature, and the crude oils composition.
(3) Improved reservoir energy after CO treatment since CO gas will come out of solution and provide a driving force to sweep the oil out of the reservoir.
(4) The lowering of interfacial tension between the oil and the aqueous phases.
Also, due to the chemical reactions between carbonic acid (CO + H O mixture) and the carbonate portions of tie reservoir rock, increased portions of tie reservoir rock, increased permeabilities can be obtained by CO injection. permeabilities can be obtained by CO injection. This can be used as a preconditioning process prior to steam injection to improve steam injectivity.
Long-term production response to matrix acidizing has been analyzed in the Stevens Sand in the North Coles Levee field in the San Joaquin Valley. Twenty-nine acid jobs on production wells, using either sequential hydrofluoric acid (SHF) or regular hydrofluoric acid (HF), have been performed over the last 15 years. SHF treatments were considerably more effective than regular HF treatments over the entire period of response to the stimulations.
The North Coles Levee field is located in the southern San Joaquin Valley, approximately 15 miles west of Bakersfield, California (Fig. 1). Production is primarily from four sandstone reservoirs in primarily from four sandstone reservoirs in the Stevens formation, a turbidite of Miocene age. Average depth is approximately 9000 ft (2740 m).
Since discovery in 1939, the field has been successively exploited through primary depletion, dry gas reinjection, gas cap blowdown, waterflooding, and CO injection. The waterflood continues to date. Cumulative production is 161 MMBO (25.6 x (10)6 m), out of 483 MMBO (76.8 x (10)6 m) originally in place. Current production is 1150 BOPD (153 m/d) from 75 producing wells.
Since 1973, 21 SHF and eight regular HF acid jobs have been performed on producers in the North Coles Levee field producers in the North Coles Levee field (Fig. 2). The acid jobs were conducted in response to (1) rapidly declining gross production rates, (2) anomalously poor well production rates, (2) anomalously poor well performance as compared to offsets, and (3) performance as compared to offsets, and (3) damaging events occurring during drilling, completion, and workovers. Over the period of incremental oil production response to acidizing (ranging from 1 to 54 months), significant differences between the effectiveness of the two acid types became apparent.
This paper describes the mineralogy of the Stevens Sand, compares the SHF and HF processes, and summarizes oil and water processes, and summarizes oil and water production responses to both acid types. production responses to both acid types. MINERALOGY OF THE STEVENS SAND
The Stevens Sand is a poorly sorted arkosic sandstone, with abundant clay matrix infilling pore spaces between the coarser grains. The Stevens was deposited as part of a deep-sea fan complex. Numerous turbidite sequences are overlain to produce a reservoir with thicknesses as great as 650 ft (195 m) (Fig. 3). Porosity is fairly constant at about 16%, but permeability varies considerably and is inversely related to the presence of the clay matrix. Liquid permeability is generally 6 md or less. permeability is generally 6 md or less. Feldspars, rather than quartz, are the dominant minerals in the Stevens Sand at North Coles Levee (Table 1). Feldspars in the Stevens Sand are the primary source for the largely authigenic clay matrix. Although x-ray diffraction indicates that clays comprise an average of 7.2% of rock weight, thin section analysis suggests that in some wells clay content may be as high as 19%. The clay fraction is composed of chlorite, mixed layer illite-smectite, illite, and kaolinite in descending order of prevalence (Table 2).
Texaco operates large steam floods in the San Joaquin Valley in which Individual wellhead mass flow rates and steam quality measurements are difficult to obtain. Uneven and unpredictable phase splitting at lateral and dead end tees is the problem. In one instance, Texaco has solved this problem by utilizing a horizontal separator to separate 10,500 BPD of 70% quality cogeneration steam.
The separator is able to deliver saturated water and dry steam. These two single phase flows can be divided into any number of flow streams and then measured using simple orifice plate, differential pressure, technology.
The steam flood distribution piping consists of the separator, two single phase trunk lines running the length of the field, and small headers located along the trunk lines. The purpose of these headers is to measure, control, and combine the water and steam for each well, using orifice meters and adjustable chokes. Because each injector is controlled separately, the Injector mass flow rates and steam qualities are independent of one another. Not only does this solve the phase splitting problem, but this flexibility phase splitting problem, but this flexibility Is helpful when some Injectors need higher or lower qualities than others. For Instance, lower qualities may be warranted for displacement patterns experiencing steam breakthrough, or for occasions where patterns requiring different qualities need to be served by the same trunk line. A 36 megawatt cogeneration facility feeds the horizontal separator which is 30 feet long, 5 feet in diameter, and 2.5 inches thick. Saturated water is taken from the bottom and dry steam is taken off the top. Both these flow streams are divided in order to serve two properties. For one property, the water and steam are measured and combined right away before feeding an existing distribution system. For the other, on which a new flood is being started, the two trunk lines run the length of the field to serve the injection headers. The hardware for each injector consists of an orifice meter and choke to control the steam, an orifice meter and choke to control the water, and a fixer to combine the two flows. All this hardware is located at the header so a choke can be set while monitoring the output from the corresponding orifice meter. A spare port on one of the headers is used, not for an injector, but to serve an existing stimulation header. In this way, cogeneration steam is used to provide all the displacement and stimulation steam requirements.
One of the problems associated with this type of operation is the chemical treatment of the dry steam line. The feedwater contains a high concentration of bicarbonate. Therefore, a neutralizing amine is added to combat the high ph caused by the formation of carbonic acid. Steam samples are collected at various points along the line and are monitored for iron concentration and ph. Also, corrosion probes and corrosion coupons are used to verify data. The goal is to hold corrosion to 0.00625 inches/year (1/8 inch over 20 years).
Consider the case of a sand stone reservoir that has been under a water or steam flooding program, that was not efficient. A lot of effort is currently being done to evaluate and recover the oil that was left in the ground. As the flood fluid has now been introduced to the Reservoir element, it becomes very difficult to find the value of the water saturation as the salinity of the flood water is different from the original formation water, and involves a lot of coring and other expensive operations to come up with the value of Sw; water saturation.
The technique suggested in this paper finds the value of the saturation using logs, both old and new, that are available for that field the old being the original resistivity logs and the new being the logs of the new well drilled and recorded after the flood. This technique agrees very well with the saturation obtained by other methods including other logs such as the EPT log, and the core analysis results, within a tolerance of + or - 7%.
In many fields that use water or steam flooding, the problem of fingering and bypassing some of the oil zones frequently happen. As a result, programs are made to recover the oil that is left behind in these secondary recovered reservoirs. The question becomes; " How much oil is left behind ? ". The problem now is that the pay zone is partly flooded, that leaves the old oil zone split by a new water zone or any other shape that might form due to the new presence of the flood water. The new water zone, however, does not have the same salinity as the flood water or the original formation water, but rather some salinity inbetween, that makes the saturation evaluation from resistivity curves a complicated problem. (Rw will vary over the zone and from one well to another for the same zone, depending on the position of the well from the flood profile )
This fact was realized by the industry and rather expensive methods are followed, such as extensive coring, Nuclear Magnetism tests, and EPT logs, etc. just to determine the oil saturation over those unswept zones.
This paper offers a solution to this problem that could save substantial expenses to answer that particular question. The results of this study particular question. The results of this study were compared to other saturation values obtained by core analysis, EPT and GST logs. The results were agreeable, and will be discussed later.
The real data tested is a case in the Wilmington field, CA. This field is a typical case of a long producer that underwent waterflooding and the producer that underwent waterflooding and the chances of bypassed oil are very much there. Fig. (1) shows what the problem would look like on a resistivity log, where a new water zone has formed within or above the oil zone. That water zone has the characteristic that it's SP curve full deflection does not reach the SSP value of the clean thick adjacent wet zone that was not mixed with the flood water. Yet, the resistivity log shows low resistance indicating water in the formation. This indicates that those two waters are different; the newer being fresher.
The solution suggested in this paper will utilize the conventional Archie's (Pickett) crossplot, with some important modifications to suit the new situation. From there we will arrive at a new defined 100% water line ( Ro-Line ) which contains the new water regardless of it's mixture, and that will be our base line for evaluating the saturations of the adjacent oil zones, ( in this case could be above or below it, as shown in the figures attached of wells #1B and 2B).
An offshore polymer flood in the Dos Cuadras field near Santa Barbara, California was implemented early in 1986. The polymer selection process included laboratory screening and a month long field injectivity test with three prospective polymer suppliers. Polymer injectivity, required polymer concentrations Polymer injectivity, required polymer concentrations to meet design viscosities, the need for additional polymer inversion chemicals and core plugging polymer inversion chemicals and core plugging behavior of the polymer solution were critical parameters in this evaluation. Full-scale equipment parameters in this evaluation. Full-scale equipment design was aided by observing the performance of skids supplied by each vendor for polymer inversion and injection. Following the selection of a polymer product, shear tests were conducted to improve product, shear tests were conducted to improve polymer solution injectivity through Berea core polymer solution injectivity through Berea core plugs. Although these tests were sufficient to plugs. Although these tests were sufficient to ensure injection into the EP zone, additional shearing of the polymer solution was required to improve injectivity into the FP zone. A filtration test, which was found to be more sensitive than the core plug test to the plugging behavior of polymer solutions, was used to compare properties of injected solutions for the FP zone. Hall plots have been found to be useful for understanding the response of the injection wells to the various injectants.
Lease OCS P-0241 has three offshore platforms, "A", "B" and "C", which stand in 190 feet (58 meters) of water. Peripheral waterflooding began in 1971 into the EP and FP zones. Dual well completions are used for the injection wells with one string in each zone. The EP zone has a nominal surface injection pressure limitation of 200 psi (1380 kPa) at a depth pressure limitation of 200 psi (1380 kPa) at a depth of 1150 feet (350 meters) subsea while the FP zone has a nominal limitation of 400 psi (2760 kPa) at a depth of 1700 feet (520 meters) subsea. The average thickness of the EP and FP zones are 151 feet (46 meters) and 87 feet (27 meters), respectively.
Laboratory work showed the polymer flooding process could improve and accelerate oil recovery. The poor mobility ratio, low reservoir temperature and poor mobility ratio, low reservoir temperature and existing injection system were all favorable for polymer flooding. Conversely, the saline brine, polymer flooding. Conversely, the saline brine, the restricted wellhead pressure and the offshore location of the field were judged to be negative factors for implementing a polymer flood. Laboratory testing suggested that a three slug tapered design be used consisting of 5.5 cp, 3.5 cp and 2.2 cp (5.5 mPa s, 3.5 mPa s and 2.2 mPa s, respectively). Based on both laboratory data and the existing peripheral waterflood, a polymer slug size of 0.5 pore volume was selected.
Polymer selection for the offshore polymer flood Polymer selection for the offshore polymer flood was made by comparing three emulsion polyacrylamides during a 30-day injectivity test. All three polymers were chosen for field testing based on polymers were chosen for field testing based on laboratory studies. The chosen polymer was selected based on lower overall cost and ease of injection even though some samples failed the core plug test for injectivity. Polymer injectivity was a critical parameter because of the severe pressure limitations of the shallow reservoir. No additional chemicals were required to assist the inversion of the chosen polymer, as were needed for the other candidates. This feature requires one less pump per skid and substantially reduces the possibility per skid and substantially reduces the possibility of injecting uninverted polymer into the well, which can result in major formation damage.
Later studies showed the chosen polymer passed all core plug tests when mild shearing was performed on the concentrate stream prior to dilution.
As previously reported, a scaled physical model was developed to study the oil recovery efficiency of steam injection into horizontal wells. Experimental performance evaluations were made with two types of synthetic oils with injection of steam at different qualities. In this paper, the results of these laboratory studies are compared to predicted performance using a three-dimensional, three-phase thermal performance using a three-dimensional, three-phase thermal simulator.
Previously, most history matching studies have relied heavily Previously, most history matching studies have relied heavily upon adjustments to the relative permeability curves to obtain agreement between calculated and observed production data. In this study, to match production data we devised a new technique in which at elevated temperatures oil viscosity is lowered and water viscosity is increased so as to converge to a common value at steam temperature. Conceptually, this technique implies that at steam temperature oil and water tend to form as a combined emulsion phase. The presence of an emulsion phase is evidenced by experimental studies and thermal oil production in the field.
We successfully used the proposed technique to match the oil and water production data obtained from laboratory experiments. The efficacy of the technique demonstrated by comparing observations to three separate prediction runs; (1) base case using standard relative permeability and viscosity data (2) results computed with only relative permeability adjusted, and (3) results with only viscosity adjusted. The history matching studies include oil production from vertical wells in a five-spot pattern and horizontal wells in a trench pattern.
Steam stimulation and steam drives in reservoirs containing crudes with gravity 11 degrees API and higher have been very successful. However, economic recovery of heavier crudes, bitumens and tar sands, has still not been generally realm The concept of horizontal well technology for exploiting bitumen reservoirs offers exciting potential. Results are presented for two oils, one whose viscosity falls in the first presented for two oils, one whose viscosity falls in the first category and one in the latter.
The scaled physical model referred to this paper was constructed under a Small Business Incentive Grant from the Department of Energy to study the effectiveness of horizontal wells in recovering extra heavy oil. A 3-D thermal simulator called ASTRO (All Purpose Simulation of Thermal Recovery Operations) available at Union Oil Science and Technology Division was used to perform history matching studies of the experimental results. Both theoretical and experimental studies demonstrate the improved recovery gained from horizontal wells. Scaled physical models of the steamflooding processes are of great value in designing a steamflood and processes are of great value in designing a steamflood and obtaining insight into the process mechanics.
A three-dimensional thermal simulator is used for further studies of thermal processes. The purpose of the simulation study is to interpret the experimental results and confirm the suitability of a horizontal well for steamflooding.
An unique feature of this work is its effort to duplicate the results of a laboratory steam drive with a thermal recovery simulator. History matching is usually done by manually adjusting data through a trial-and-error procedure.
The use of surface calculated pressure buildup (SCPBU) tests has proven to be an accurate and cost effective technique for acquiring pressure transient data in the Kuparuk River Unit on the North Slope of Alaska. The technique involves measurement of the surface tubing pressure and use of a computerized acoustic device to measure the depth of the gas/liquid interface in the tubing of a shut-in production well. The bottom-hole pressure is then calculated by applying the gradients of the respective fluid columns. Conventional pressure transient analysis is then undertaken which provides useful information regarding provides useful information regarding reservoir and completion characteristics.
The equipment used was developed for performing SCPBU's in rod-pumped wells by performing SCPBU's in rod-pumped wells by monitoring the casing pressure and fluid level. However, with minor adaptations, it is successfully being employed at Kuparuk to measure the downhole pressure buildup via the tubing string of producing wells. To avoid the potential of natural gas accumulation in the wellhouses which protect wellheads in the arctic, the acoustic signals are generated by explosion signals from a nitrogen bottle secured in the wellhouse rather than through implosion to the atmosphere. As these acoustic signals are reflected by downhole jewelry such as gas lift mandrels (GLM), additional attention is required to ensure that the computer does not mistake a GLM for the fluid level as the fluid level rises or recedes in the tubing. This is currently being accomplished by frequent on-site monitoring of the fluid level's location in relation to downhole equipment, and adjustment of the "mute time" after which the computer will recognize an acoustic reflection as the true fluid level. After determining the measured depth to the FL, the computer then uses Input well deviation data to determine the true vertical height of the gas and liquid columns in the wellbore. The bottom-hole pressure is then calculated by applying the gradients of the respective fluid columns, with conventional pressure transient analysis techniques applied to the resulting pressure transient data.
It should be noted that using the subject technique presents an additional unknown in many pressure transient calculations in that the flowing bottom-hole pressure (FBHP) must be estimated instead of measured.In most situations at Kuparuk, proven hydraulic correlations allow the FBPH to be assumed within a range suitable for analysis of most pressure transient parameters, such as skin, pressure transient parameters, such as skin, permeability, and flow efficiency. permeability, and flow efficiency. In the past year, over 50 SCPBU's were completed on wells in the Kuparuk River Unit. Of these tests, over ten were measured simultaneously with downhole pressure measurement devices. These concurrent tests provide a means to evaluate the accuracy of provide a means to evaluate the accuracy of the surface calculated measurement technique through direct comparison with the current industry standard of downhole measured pressure transient testing. Results of these pressure transient testing. Results of these concurrent tests indicate close agreement after the dynamics of phase separation in the tubing have taken place. In Kuparuk wells, the first 4-8 hours of shut-in time is a period marked by the dynamic migration of period marked by the dynamic migration of natural gas from the live oil present in the wellbore and reservoir, resulting in a constantly changing fluid gradient in the wellbore. Due to the numerous complicating factors involved, a correlation to accurately model the changing gaseous-liquid gradient has not yet been developed for Kuparuk.
Decisions concerning future development options at the Kuparuk River Field are complex and interrelated. Various development strategies are being evaluated. To help address these issues, a facilities model has been added to a full field reservoir simulator. Facility limits are used with a well ranking system to shut in marginal production when equipment capacity constraints are exceeded. The model simulates field operating strategies which maximize production. Operational strategies are modelled using (1) a wellbore hydraulic routine that enables bottom hole pressures to change (reflecting the benefit of various gas lift rates), (2) a gas lift allocation correlation that was determined from field data, and (3) a gas management routine which optimizes oil production while honoring gas capacity constraints and field gas production while honoring gas capacity constraints and field gas lift needs. These gas management and hydraulic features are essential to properly evaluate future projects in a gas capacity constrained environment and can be applied to other reservoir simulation tools. An example of how the reservoir simulator is being used to address a key issue in the large scale facility expansions evaluation is included.
The Kuparuk River Reservoir contains approximately five billion barrels of original oil in place. The reservoir is located on the Alaskan North Slope, approximately forty miles west of the Prudhoe Bay Field. The field produces from two physically independent sands of the Kuparuk River Formation, a lower cretaceous, shallow marine sandstones The productive sands are Informally named the A and C Sand members. The C Sand overlies the A Sand and is highly permeable, having an average permeability thickness of 5000 md-ft (1.5 mu m2-m). The A Sand is considerably less productive than the C Sand, having an average permeability thickness of 1000 md-ft (0.3 mu m2-m). The two sands do not communicate within the reservoir; however, they are hydraulically coupled through the wellbores.
Many development issues have faced the Kuparuk River Field Unit Owners since startup in 1981. The field has experienced a variety of development processes including primary production, waterflood, a water alternating immiscible gas injection (immiscible WAG) project, and a pilot scale miscible WAG project. With these various recovery mechanisms, the field has grown in complexity. An uncertain economic climate further complicates development decisions. The Kuparuk River Field has reached a point in field development where decisions concerning future strategies are usually complex and interrelated. Various development strategies are simultaneously being investigated: infill drilling, further peripheral development, and enhanced oil recovery projects. Some of these strategies require facility additions with long equipment lead times, which necessitates early identification of field development plans and reliable rate projections. In addition, future operating strategies centering around a maturing waterflood are also uncertain. Gas lift gas requirements for wells with steadily increasing water cuts are still being investigated. Factoring these possibilities and uncertainties into the evaluation of development options at Kuparuk required enhanced simulation tools.
To address the complex nature of these decisions, a Kuparuk-specific facilities model and well management logic for a gas capacity constrained system have been added to a full field reservoir simulator. A modelling scheme specific to the Kuparuk River Field was needed because of the unique facility configuration. Equipment limits are used with a well ranking system to shut in marginal production when facility limits are exceeded. To properly simulate field operations in a compression capacity constrained environment, additional features were developed for the well management package. A wellbore hydraulics routine that enables flowing bottom hole pressures to change as a function of gas lift rates, fluid production characteristics, and reservoir pressure was implemented.