Using the theory of impulse testing and principle of superposition, Nolte et al  developed a method which allows the identification of radial flow and thus the determination of reservoir transmissibility and reservoir pressure. The exhibition of the radial flow is ensured by conducting a specialized calibration test called mini-fall off test. Benelkadi and Tiab  proposed a new procedure for determining reservoir permeability and the average reservoir pressure in homogeneous reservoirs. In this paper, the procedure is extended to naturally fractured reservoirs.
A new method for the determination of reservoir transmissibility using the after closure radial flow analysis of calibration tests was developed based on the pressure derivative. The primary objective of computing the pressure derivative with respect to the radial flow time function is to simplify and facilitate the identification of radial flow and the characteristic trough of a naturally fractured reservoir. The proposed method does not require a-priori the value of reservoir pressure. Only one log-log plot is used to determine the reservoir permeability, average pressure, storativity ratio, and interporosity flow coefficient.
The technique is presented as a step-by-step procedure, with appropriate equations and graphs. The log-log diagnostic plot of pressure and the pressure derivative is employed here as the main tool to identify the radial flow regime and determine properties of a naturally fractured reservoir (i.e. permeability, interporosity, storativity ratio, and average reservoir pressure). The application of the proposed method is demonstrated on real field data from calibration tests.
The main conclusion of this study is that small mini-fracture treatments can be used as an effective tool to identify the presence of natural fractures and determine reservoir properties.
The mini-frac injection test has permitted the determination of the reservoir description in homogeneous reservoirs where fluid leakoff is dependent on the matrix permeability, fluid viscosity, and reservoir fluid compressibility. Applying this type of test to naturally fractured reservoirs introduces new factors that are difficult to measure, e.g. fluid leakoff dominated by the natural fractures that vary with stress or net pressure. This study allows the identification of naturally fractured reservoirs from after closure tests and the estimation of their respective reservoir parameters.
Naturally Fractured Reservoirs
A naturally fractured reservoir is also referred to as dual-porosity system due to one of its main characteristics: the presence of two distinct types of porous media (matrix and fracture), which present different fluid storage and conductivity characteristics.
Because of the complexity in the geometry of naturally fractured reservoirs, different mathematical approaches have been developed for diverse geometric shapes in an effort to simulate the effect of matrix block shapes in the transition period. One of the most popular approaches was proposed by Warren and Root . They introduced two parameters that they referred to as the storativity ratio (?) and the interporosity flow coefficient (?) to characterize naturally fractured reservoirs.
Mini-frac Injection Test
In the last two decades, mini fracture injection tests -also called calibration treatments or injection tests- have been developed to diagnose features including interpretation of near wellbore tortuosity and perforation friction, fracture height growth or confinement, pressure-dependent leak-off, fracture closure, and more recently transmissibility and permeability.
Elnaga, Aly Abou (Chevron San Jorge S.R.L.) | Almanza, Edgar A. (Halliburton Energy Services Group) | Batocchio, Marcelo Pavanelli (Halliburton Argentina S.A.) | Folse, Kent C. (Halliburton Energy Services Group) | Schoener-Scott, Martin F. (Halliburton Co.)
Chevron San Jorge S.R.L. operates in the Loma Negra area and El Trapial field located in the Neuquén Basin, Argentina.
El Trapial wells are characterized by stratified, shallow- to medium-depth reservoirs with permeabilities of 35md to 85md and porosities of 18 to 30%. The wells are completed in oil reservoirs that have been perforated using conventional methods, fracture stimulated to increase production, and later, completed with electro-submersible pumps (ESPs).
To effectively meet the operator's needs for a method that would help optimize well productivity, and at the same time, be cost effective without compromising the results of the operation, an improvement over traditional tubing-conveyed perforating (TCP) was required. A propellant-assisted (PA) perforating method that could optimize well productivity while maintaining stringent health, safety and environmental standards was proposed. The propellant-assisted perforating method uses standard perforating components and procedures, thus providing the same safety features available to the industry today when conventional TCP operations are used. The propellant is an oxidizer that creates carbon dioxide gas at extremely high peak pressures in the millisecond time regime to overcome in-situ stresses and create perforation breakdown and mild fracturing near the wellbore.
This paper will focus on two case histories describing the methods and improved results obtained from the application of the propellant perforating technique.
The information in this paper will verify that the technique was capable of satisfying the operational challenges in these reservoir types. Instrumental in the success of the propellant-assisted perforating methodology was 1) the proper screening of candidate wells, 2) good pre-job design modeling, and 3) adherence to industry accepted best practices with constant communication.
El Trapial Field was discovered in 1991. It is part of a near-shore, shallow marine environment. It is situated in the Neuquen Basin (Fig. 1) in the Northeast quadrant of the Neuquen Province of Argentina. The main reservoir in the El Trapial field is the Lower Troncoso where 57% of the total reserves reside. It is composed of sandstone in eolian faces, of a medium-grained clast-supported texture, scarce matrix and dolomite cement, with good to very good porosity and permeability.
New well logs were run, and when they identified increasing water cut in the production, Chevron decided to try a new technology to perforate these wells rather than the conventional tubing-conveyed perforating (TCP) techniques previously used followed by conventional hydraulic fracturing.) This new technology chosen included propellant sleeves with the TCP guns, which was capable of increasing the exposure area (fissures) created by the gas (CO2) in the propellant sleeves.
Loma Negra field complex on the Rio Negro Norte Block in southeastern Neuquen has hosted several discoveries since opening of Loma Negra field in 1997. Loma Negra, with reserves of at least 240 million bbl. could be one of Argentina's top producing fields of this decade. The formation is a sandstone called Punta Rosada with permeabilities in excess of 75md and porosities in excess of 15% and bottom hole pressures in the vicinity of 4000psi.
Description of Propellant-Assisted Perforating
Propellant-assisted perforating techniques have been in use for many years as a method to stimulate completions. Results have been somewhat mixed due to a lack of understanding of the physics involved during the propellant burn event. Advances in the propellant technology have continued to evolve, and these have led to improvements in techniques for well candidate selection, and subsequently, stimulation results. The propellant-assisted perforating technique is used in many different applications with positive results, as described by Boscan ET al.1
The propellant-assisted perforating process combines perforating and perforation breakdown with propellant in a single tool and operation.
Reed, Skip (Halliburton Energy Services Group) | Torne, Juan Pablo (Halliburton) | Pacheco, Erick (Halliburton Co.) | Casillas, Augusto Zenon Lara (Halliburton) | Palacios, Cesar (Pemex E&P) | Riano, Juan Manuel (Pemex) | Corral Aguirre, Alejandro (Pemex E&P) | Casares Vazquez, Martin de Jesus (Pemex) | Castellanos, Jose | Arteaga, Rafael
The Veracruz basin, located in eastern Mexico, has sometimes been very difficult to evaluate with openhole wireline tools because over-pressured gas formations present a problem when trying to reach a balanced borehole condition. One of the fields was planned for development using nearly horizontal wells to maximize production, but acquiring openhole logs in these wells has proven to be difficult.
Well A-21 is the first highly horizontal well to be completed using only LWD resistivity data in the nearly horizontal section. To assist in the evaluation after the well was completely drilled and cased, a pulsed neutron tool was logged across the entire well using a tractor device to reach the total depth of the well. An evaluation model for the A field was then developed using the CHI (cased hole interpretation)modeling program, by using the openhole resistivity and porosity data acquired in the vertical A-1 well, combined with the cased hole pulsed neutron data acquired across the same interval. Pseudo resistivity and porosity logs were then created for the A-21 well using only the pulsed neutron data acquired across the nearly horizontal section of the well, based on the Chi Model developed for the field. The pseudo resistivity was then compared to the LWD resistivity data acquired in the A-21 well. Next, the interpretation was completed by defining the optimum perforating interval for the reservoir conditions and the mechanical condition of the well. After evaluating the
interpretation, the On-Balance perforating technique using coiled tubing was decided upon as the optimum technique to perforate the A-21 well, to minimize reservoir damage. This paper will present the procedures used to evaluate and complete the A-21 and the A-31 wells, as well as a comparison of the CHI Modeling pseudo resistivity with the LWD resistivity measurement.
The use of pulsed neutron logs to acquire pseudo openhole data is shown to be a valid alternative when drilling conditions do not permit normal data acquisition in openhole. The integration of the data obtained, along with applied reservoir geomechanics, for perforating design and production planning is shown to be a valid alternative to maximize production, and to prevent sanding and completion problems while reducing costs.
Highly deviated wells have traditionally required the use of either logging-while-drilling (LWD) or tool pusher techniques to acquire the information needed to perform formation volumetric evaluation. This evaluation is used to determine the best intervals to be completed. A new technique is currently available that can be used in development fields to optimize drilling and completion cost while minimizing the risk of getting stuck with a set of LWD tools. Sticking LWD tools downhole can lead to a complicated recovery of the drilling
string. In addition, because they contain chemical radioactive sources, LWD porosity tools left in the well present further complications.
The innovative technique presented in this paper includes an application of pulsed neutron log data to acquire pseudo openhole data after the completion of the well. This acquisition can even be accomplished without a drilling rig, which further minimizes both the risk and the cost of logging the openhole section. This paper presents the theory behind the technique and the process required to reach a result. It also includes techniques currently used to acquire the information in cased hole, such as wireline logging and perforating using either coiled tubing or the wireline tractor tool.
Los Molles Formation in the Neuquén basin (Fig 1), west-central Argentina, contains a series of deep-water submarine channel complexes deposited in an elongate basin, during the Pliensbachian-Toarcian ages. These events developed during the final stages of the rifting phase. In the studied area, these submarine meandering canyons cut the shelf and gentle slope in a SSE-NNW direction, and represent the transfer zone of the system, where erosional features are more frequent and lithological distribution is more complex. The canyons are approximately 3 Km wide and several hundred meters thick and host a number of migrating channel-fill units inside, lying on erosional surfaces that cut into adjacent interchannel facies.
Applying neural network techniques in the three wells that penetrate this deep-marine strata, allowed the identification of five main lithofacies: muddy-matrix conglomerates, sandy-matrix conglomerates, coarse-grained sandstones, fine-grained sandstones and mudstones. Furthermore, the use of 3D seismic attributes was crucial to obtain the distribution of these facies within the canyons. For this purpose, techniques based on neural network and representative supervised "seed points?? next to each lithology around the well, were applied.
This work resulted in a seismic volume with the distribution of four seismic facies along the system in a very heterogeneous way. The fine grained facies clearly located in an overbank position; the sandstones and conglomerates show a distribution constrained inside the canyons, and is also easy to see how the net-to-gross relationship increases towards the distal positions of the system.
The techniques applied, greatly improve the success of prediction of potential reservoirs locations.
Exploring deep-water reservoirs is a challenging task, where an accurate characterization of the stratigraphy and facies distribution, happen to be the main target towards a robust geological facies model.
Los Molles Formation consists of a NNW-SSE elongate submarine system where a number of meandering canyons of strong erosional profiles cut into the shelf and slope, during Pliensbachian-Toarcian ages. Three main depositional provinces were detected: A proximal erosional zone, a channelized gentle slope with erosional and depositional features, and a downslope depositional channel mouth zone where the sandy fractions were pressumably deposited in a "collapsing?? way.
The discoveries of gas accumulations in nearby areas, triggered a detailed exploration of this unit and the need of understanding the facies distribution along the system. The development of models of deep-water systems for predictive exploration, has come to rely on all the available data like seismic, cores and logs.
Peru has a great background in the oil industry; the first well in South America was drilled in Zorritos, located on the northwest coast of Peru, in 1863. Originally, the wells were completed in the shallower pays with cemented casing and using perforated liners in the productive areas. Production was driven by the natural energy of the reservoirs.
Around 4600 wells have been drilled In Block X and, currently, there are 2021 producing wells, with an average production of 6.4 bls./day per well. Hydraulic fracturing is required to produce them economically.
These works have been done since 1953, and continue up to the present applying available technical innovations. Initially, stimulations were carried out through hydraulic fracturing using diverse multifrac techniques, with crude oil as base fluid. In the 90's, oil was substituted for water as a base fluid for stimulations, to generate better fractures and increase recoveries.
In spite of the maturity of Block X and the high wells density, new alternatives were searched in order to continue operating economically, considering:
In this paper a historical perspective as well as currently applied technologies for stimulation are reviewed, and related field results are presented.
Block X operated for Petrobras Energía Perú is located in Talara Basin, on the northwest coast of Peru. It is composed of 17 main fields (Fig. 1). Approximately 4600 wells have already been drilled currently, and there are 2021 producing wells, with an average production of 6.4 bopd per well.
The reservoirs of Talara Basin are characterized by their low permeability ranging from 0.1 to 60 md. For that reason, hydraulic fracturing is required to be done to produce them economically. The producing formations are sandstones characterized by their heterogeneity and presence of shale.
Drilling began in Block X in 1910, using percussion drilling equipment. South America's first well in Zorritos, was drilled to a final depth of 78 ft. (24 m). From 1926 on, changes in technology to rotary rigs have enabled to drill deeper and faster.
Initially, wells were completed with a surface casing of around 500 ft. depth, and several sections of perforated liners (Fig.2). Driving mechanism for these wells was natural flow.
From 1951 on, the wells have been completed cementing both surface (300 ft.) and production casing (5 ½?? ó 4 ½??), this one up to the surface.
Currently, completions are performed with 5 to 6 consecutive stages stimulation jobs, through casing; For workover jobs 3 stages hydraulic fractures through 2 7/8?? ó 3 ½?? tubing are needed (Fig. 3).
Evolution of Stimulation Jobs
SANDOIL TREATING technique, consisting in pumping a mixture of crude oil and sand, started to be used in 1953. Both fluids and proppant agent volumes were low, with figures ranging from 200 to 400 gals. and ½ lb/gal sand concentration, as well as a 2 to 4 gpm flow rate were typical.
In 1956, viscous crude oil (VISOFRAC) started to be used as frac fluid in some fracturing jobs, which allowed increasing the volumes of treatments.
From 1957, PERFPAC technique started to be used, which performs fracturing in stages by using nylon balls as divergent agent for temporally isolating previously fractured zones, as well as leading stimulation towards untreated formations. This technique was used in large pays, avoiding the necessity of multiple stage fractures with higher costs.
Frac-pack is a pervasively used completion technique in wells targeting high permeability, poorly consolidated and depleted sandstone formations located in Bachaquero, Tía Juana and Lagunillas fields in West Venezuela. This technique combines stimulation and sand production control in a single treatment by placing a short and wide fracture which bypasses the near-wellbore damage, while gravel-packing the zone of interest.
This paper describes a novel and economical frac-and-pack technique which consists of pumping a sand plug with the downhole tool set for circulation to isolate a bottom set of perforations, followed by conventional frac-and-pack. When this procedure is followed, the fracture is forced to propagate along the upper intervals. This novel technique is particularly useful for wells with water-producing zones near the bottom of the target zone, because it induces selective growth of the fracture along the upper intervals and mitigates the risk of growing the fracture into the water-producing zone.
A case study of a frac-and-pack performed in a Lagunillas field well with a water contact 40 ft below the target zone is reviewed. The intervention rendered an increase in well production rate from 27 to 173 net barrels per day with a reduction in water cut from 25% to 9%. In contrast, two wells in the same field and with very similar characteristics which were frac-and-packed conventionally rendered 100% and 63% water cuts, respectively.
Another application of this technique refers to frac-and-pack of wells with long perforated intervals where early wellbore screen-out may occur due to proppant bridging of the annular volume between the screen and the casing. Conventional frac-and-pack of twenty wells in these fields with perforated intervals exceeding 90 ft rendered a 40% early wellbore screen-out rate. The early wellbore screen-out rate was reduced to 12% in a sample of twenty eight wells with the new technique. The average production rate increased from 2 to 135 BOPD, whereas the average estimated after-treatment production was 130 BOPD, for which this technique was considered successful. A shortcoming of the technique for this application is the fact that the bottom of the perforated interval is not fractured. High-end frac pack techniques that overcome this issue such as use of shunt tubes were found to render higher normalized oil production rates.
In mature fields, operators are often seeking ways to increase the hydrocarbon recovery, with the help of reputable service companies. Well stimulation continues to be, by far, the preferred method of achieving such goal. Operators and service companies are continually screening out technologies which will deliver the highest benefit/cost ratio for a particular stimulation well treatment, maintaining focus on operational and health, safety and environment excellencies .
This paper addresses the rebirth of a past hydraulic fracturing technique, born in the 50's, and how it is being successfully applied on onshore mature fields in Brazil: batch fracturing. It is effective due to several technological advancements on proppant density, becoming lighter than conventional frac sand and yet with sufficient mechanical properties to withstand bottom-hole environments. Batch fracturing is now contributing to equally efficient, and more economical well stimulation treatments, providing good economical returns to operating companies.
Batch-Fracturing had limited success in the past. This was due to the available frac fluid and proppant technologies at that time. It is desirable that proppants have low settling when carried by a fracturing fluid, from the time they are added into such fluid, until the end of the pumping process. Batch fracturing applications are on the rise, due to the new families of ultra lightweight proppants, with specific gravities ranging from 1.05 to 1.75. In batch fracturing, the proppant is added to the carrier fluid prepared in standard oilfield mixing tanks,
eliminating the need of specialized mixing equipment such as blenders. Less sophisticated equipment on location implies in lower operational and logistical costs. The carrier fluid ("frac fluid??) does not need to yield high levels of viscosity, and, by consequence, does not have a high load of chemicals (gelling agents, cross-linkers, related breakers…). With batch fracs it is possible to perform common but effective types of fracturing treatments, such as "skin-by-pass?? (a fracture that by-passes the damaged zone), and "partial mono layer' fracturing, both exemplified in this paper, through case histories.
Today, most of the producing oil and gas fields are considered mature. Although continually being redefined, a field is considered mature when its current level of hydrocarbon production has passed its past production peak. Associated with the reservoir's production depletion, there are other hydrocarbon recovery issues inducing operators to continually seek ways to overcome these natural effects. They look, with their subcontracted service companies, for cost effective techniques and technologies able to increase production and
Everyday, it is more frequent and common to adequately know the dynamical behavior of stress and strain sensitive formations and to improve their geomechanical characterization, as well. For this purpose, a coupled simulator of fluid flow and rock deformation was used in this study to investigate the impact of the elastic constants of the rocks during a well pressure test. The results were compared with those obtained for a reservoir with average properties in which the geomechanical effects were neglected.
The numerical experiments were focused on simulating drawdown tests for four different scenarios as follows: a standard case without geomechanical effects, two cases with isotropic initial state of stress and two cases with anisotropic stress. For interpretation and comparison purposes, we utilized the TDS technique to determine the reservoir parameters of the simulated pressure data. The results allow us to visualize that Young's modulus causes a pronounced effect on the pressure test during pseudosteady-state flow which leads to a wrong estimation of reservoir drainage area; a range of values were established above which its variation does not cause important effects. On the other hand, it was found that variations of Poisson's ratio are not significant since they fit in a small range of the values customary reported in the literature.
Since conventional reservoir engineering has overlooked geomechanics effects, research on reservoir geomechanics has lately increased. Recent investigations2-5 have concluded that an adequate characterization of stress-dependent reservoirs can help to achieve a more realistic reservoir forecast and, then, appropriate decisions regarding production optimization can be made. For these goals, coupled models to simulate both fluid flow and rock deformation behavior have been developed.
In this paper, a study for evaluation the effect of the rock's elastic constant on pressure transient tests using a coupled simulator is presented. The results permitted to establish a range of high influence of the elastic constant values during well pressure tests. The study also allowed us to discard critical values for the elastic constants which generate high numerical instability and possess low practical application.
2. Simulation Background
In this study a numerical coupled simulator developed by Alcalde and Wills1 was used. The numerical model consists of a fully implicit 3D finite difference solution in cylindrical coordinates with irregular grid using lattice grid points. The model considers the following assumptions: (i) the rock is deformable under an elastic, nonlinear behavior, (ii) monophasic and isothermal flow, (iii) Because of the stress dependence, the fluid and geomechanical properties are subject to temporal and spatial variations. The numerical solution is achieved by a Picard-Gauss-Seidel iteration type2.
The governing equations for the fluid flow in porous media coupled with geomechanical deformation are determined by four expressions which are developed detailly in Ref. 1.
Evaluating the complex clastic reservoirs in El Tordillo field of the San Jorge Basin in Argentina using conventional logs is greatly affected by variations in formation water salinity, texture, and lithology (primarily the amount of volcanic tuff material) combined with extreme changes in rock permeability and wide variations in oil viscosity. All of these factors affect most of the conventional logs responses in such a way that traditional log analysis methods may fail to provide proper results and consequently may not achieve appropriate forecasts. Such failures have been proven by the high degree of mismatch between conventional log analysis and test/production results.
To address uncertainties in conventional log evaluations, operators may resort to excessive well testing for reservoir characterization and production verification. However, well testing is known to be costly (considering the rig time and the frac jobs used). On the other hand, if the operator does not proceed with well testing, productive zones can be bypassed. That is why an effective log evaluation would be the optimum, cost effective method when it demonstrates agreement with production results.
The latest acquisition and interpretation techniques of the magnetic resonance imaging logs (MRIL) have demonstrated promising results in the complex shaly-sand reservoirs of El Tordillo field. The magnetic resonance imaging (MRI) technology run on wireline provides an in-situ evaluation of the reservoir properties vital for producibility predictions. The ability of MRI to estimate the type of fluid in terms of viscosity influences the selection and elimination of zones to be tested based on their mobilities. MRI also helps in properly preparing for the testing procedure by pre-identifying the type of hydrocarbon in a zone and by identifying its reservoir quality in terms of permeability and porosity. Thus, the MRI serves well when determining the need for well testing and in enhancing the effectiveness of any particular well test.
This article discusses the effectiveness of applying the MRI logging technology to the characterization of the shaly-sands of El Tordillo field in both the Comodoro Rivadavia and Mina El Carmen formations. It also argues for the value of MRI technology in minimizing the cost (and associated risk) of well testing by identifying best candidate zones for testing and by providing necessary "prior-to-testing?? information on the fluid type (i.e., water, light oil, heavy oil or gas) present in a zone and rock properties, such as porosity and permeability.
El Tordillo field was discovered in 1932. It was operated by YPF (the Argentinean national oil company) from 1932 until 1991 when the Consortium El Tordillo, in which Tecpetrol S.A. is the operator, assumed operations. The field is situated on the north flank of the San Jorge Basin in Chubut Province, Argentina. It is approximately 50 kilometers from the town of Comodoro Rivadavia and 1,500 kilometers south of Buenos Aires (Figure 1). More than 1,200 wells have been drilled in the field, and production is spread over approximately 57 square kilometers.
Three Cretaceous units cluster the numerous fluvial sand bodies that structurally and stratigraphically trapped hydrocarbons in the field. These formations are El Trebol, Comodoro Rivadavia, and Mina El Carmen (Figure 2). The latter two formations are the focus of this work.
The fluvial reservoir units are sandstones with high lithic and pyroclastic (tuffaceous material).1 The quality of the reservoirs generally improves up-section as pyroclastic content decreases, but hydrocarbon accumulations throughout the producing interval are highly compartmentalized as a result of faulting and the discontinuous nature of fluvial sandstone reservoirs. Relevant geological, geophysical, and engineering observations from studies performed within the field can be found in publications by Muruaga, et al. 2 and Taboada, et al. 3
This paper describes three wells distributed in different production blocks within the field in which MRI logs were run in addition to the conventional data usually adopted as the minimum acquisition of open hole logs: Caliper, SP, GR, Induction and Sonic.
Upscaling reservoir properties for reservoir simulation is one of the most important steps in the workflow for building reservoir models. Upscaling allows taking high-resolution geostatistical models (107-108 grid blocks) to coarse scale models (104-105 grid blocks), manageable for reservoir simulation, while retaining the geological realism and thus effectively representing fluid transport in the reservoir 1,2. This work presents a study of the effectiveness of different available techniques for permeability upscaling and the implementation of a new technique for upscaling of relative permeability curves based on the numerical solution of a two-phase system and the Kyte and Berry method3.
The reference fine scale model considered in this study is a conceptual fluvial reservoir based on the Stanford V model4. The reference fine scale isotropic and locally heterogeneous permeability distribution was upscaled to different upscaling ratios by means of analytical (static) and numerical single-phase (pressure solver, dynamic) techniques. Two-phase flow simulations were performed on the reference fine grid and upscaled models using a comercial black-oil simulator. Arithmetic, harmonic, and geometric averages were defined for static upscaling of the permeability distribution. The dynamic upscaling process considered one-phase and two-phase upscaling. One-phase upscaling considered upscaling of the permeability distribution and two-phase upscaling considered upscaling of the permeability distribution and relative permeability curves.
Flow simulation results for waterflooding in the coarse scale model indicated relevant discrepancies with the fine grid results. Compared to fine-scale, flow results of the single-phase upscaling process indicated that the coarsest upscaled models did not match the water breakthrough times, water cut values, or well pressures from the reference model. The finer upscaled models reproduced the reference results more accurately than the coarser models. The two-phase dynamic upscaling technique implemented in this work resulted in the best match with the flow simulation results of the fine grid model. Results show that the most accurate upscaling scheme should be defined using the two-phase dynamic upscaling technique on the model with the smallest upscaling ratio.
A geological model generated using geostatistical techniques often can contain detailed geologic information in multiple directions and at different scales. The detailed geologic information can be comprised of varying degrees of heterogeneity, anisotropy, or different length scales. As much detail as possible is desired to make an accurate and precise geologic model. Such an accurate and precise geologic model is capable of characterizing reservoirs accurately in terms of compartmentalization, connectivity, and structure. In terms of computer memory and storage, however, more detail means models of larger sizes, on the order of 107 to 108 cells. Although an accurate and well-characterized reservoir is desired, the complexity, and thus the size of the model, can introduce significant computational problems when performing reservoir simulations. An effective way of upscaling is requiered, which reduces the
CPU demand and run time while preserving the geology, especially the important flow features such high permeability zones.