The Yufutsu gas-condensate field in Hokkaido, Japan, is an unusual reservoir. The fluid in the reservoir is heavier at the top, and the C7+ fraction decreases with depth despite pressure communication. Another unusual aspect of the reservoir is the decrease of the GOR from an initial value of around 1350 vol/vol to around 950 vol/vol in the first five years of production. In this work, we present the results of compositional simulation. The simulation is based on the initialization from a model that takes into account molecular, pressure, and thermal diffusion. The GOR predicted from the compositional simulation is in agreement with measured GOR.
The Yufutsu field in Hokkaido, Japan, is a large gas-condensate reservoir. The reservoir is naturally fractured with a very tight matrix of granite/conglomerate rock of negligible porosity. The fractures provide both the storage and conductivity. The gas condensate is rich and has a high wax content. The field has an areal extent of 4 km × 8 km with a maximum hydrocarbon column of 1 km. Top of the reservoir is at 3800 m sub-sea level (mSSL). No distinct gas-water contact (GWC) or gas-oil contact (GOC) has been found. The Yufutsu field was discovered in 1989. A total of twelve wells have been drilled. Production started in February 1996. The initial reservoir pressure is about 550 bar and the temperature is 150° C at 4500 mSSL.
Until early 2000, there was only one producer in the field (well MY1). Fig. 1 depicts the measured GOR produced from the well. This figure reveals that the GOR decreases gradually from around 1350 vol/vol to less than 1000 vol/vol. Under the same size bean, gas production has a decreasing trend while the condensate rate stays constant. The GOR from another well (MY2) which started producing in early 2000 shows a similar trend.
The GOR decrease is believed to be due to the heavier fluid at the top of the formation at initial conditions (prior to depletion). In a previous work,1,2 we have studied compositional variation in the Yufutsu field. A new numerical algorithm based on a diffusion model from irreversible thermodynamics3,4 was developed to predict composition and pressure in the entire reservoir using a single PVT sample as a reference point. The results from the model were in good agreement with compositional data from different wells. It was demonstrated that thermal diffusion is the main phenomenon affecting compositional variation in the Yufutsu field. Due to thermal diffusion, a heavy fluid floats on the top of a light fluid. The model results also indicate the existence of liquid (in the near critical region) on the top of the vapor column in some upper part of the Yufutsu field. The predictions from the model are in agreement with pressure data from the shut-in wells tubing which show a high pressuregradient region between two low pressure-gradient regions. At the bottom of the tubing, the density is less than 400 kg/m3; it gradually increases to over 500 kg/m3 in the middle; then there is a sharp density decrease to about 350 kg/m3. A density of 500 kg/m3 corresponds to a liquid state while a density of 400 kg/m3 and less corresponds to a vapor state.
The GOR decrease in the Yufutsu field is believed to be related to the state of initial fluid distribution: the liquid dropout at the top is higher than that in the bottom (see Fig. 2). Constantvolume depletion (CVD) data show that the retrograde-liquid dropout decreases substantially with depth. Data from some of the wells in the formation also show that the heavy fraction (C7+) decreases with depth.1,2 On the other hand, methane mole fraction increases with depth. The vertical compositional variation of heptane-plus in the Yufutsu formation is similar to that reported by Temeng et al. from the Ghawar Khuff reservoirs.5
Barenblatt, G.I. (Lawrence Berkeley National Laboratory and University of California) | Patzek, T.W. (Lawrence Berkeley National Laboratory and University of California) | Silin, D.B. (Lawrence Berkeley National Laboratory )
Forced oil-water displacement and spontaneous countercurrent imbibition are crucial mechanisms of secondary oil recovery. The classical mathematical models of these phenomena are based on the fundamental assumption that in both these unsteady flows a local phase equilibrium is reached in the vicinity of every point. Thus, the water and oil flows are locally redistributed over their flow paths similarly to steady flows. This assumption allowed the investigators to further assume that the relative phase permeabilities and the capillary pressure are universal functions of the local water saturation, which can be obtained from steady-state flow experiments. The last assumption leads to a mathematical model consisting of a closed system of equations for fluid flow properties (velocity, pressure) and water saturation. This model is currently used as a basis for predictions of water-oil displacement with numerical simulations.
However, at the water front in the water-oil displacement, as well as in capillary imbibition, the characteristic times of both processes are comparable with the times of redistribution of flow paths between oil and water. Therefore, the non-equilibrium effects should be taken into account. We present here a refined and extended mathematical model for the non-equilibrium two-phase (e.g., water-oil) flows. The basic problem formulation as well as the more specific equations are given, and the results of comparison with experiments are presented and discussed.
The problem of simultaneous flow of immiscible fluids in porous media, and, in particular, the problem of water-oil displacement, both forced and spontaneous, is fundamental to the modern simulations of transport in porous media. This problem is also important for engineering applications, especially in the mathematical simulation of the development of oil deposits.
The classical model of simultaneous flow of immiscible fluids in porous media was constructed in late thirties-early forties by the distinguished American scientists and engineers M. Muskat and M.C. Leverett, and their associates. The model was based on the assumption of the local equilibrium, according to which the relative phase permeabilities and the capillary pressure can be expressed through the universal functions of the local saturation.
The Muskat-Leverett theory was in the past and is nowadays of fundamental importance for the engineering practice of the development of oil deposits. Moreover, this theory leads to new mathematical problems involving specific instructive partial differential equations. It is interesting to note that some of these equations were independently introduced later as simplified model equations of gas dynamics.
Gradually, however, it was recognized that the classical Muskat-Leverett model is not quite adequate, especially for many practically important flows. In particular, it seems to be inadequate for the capillary countercurrent imbibition of a porous block initially filled with oil, one of the basic processes involved in oil recovery, and for the even more important problem of flow near the water-oil displacement front. The usual argument in favor of the local equilibrium is based on the assumption that a representative sampling volume of the water-oil saturated porous medium has the size not too much exceeding the size of the porous channels. In fact, it happens that it is not always the case and the non-equilibrium effects are of importance.
A model, which made it possible to take into account the non-equilibrium effects, was proposed and developed by the first author and his colleagues. This model was gradually corrected and modified. It was confirmed by laboratory and numerical experiments. In its turn, this model leads to non-traditional mathematical problems.
One proposed method for the chemical control of produced water is by using a "relative permeability modifier" (RPM) which is usually based on a water soluble polymer. A considerable amount of literature on such materials has appeared although there is still not a complete consensus on the actual mechanism through which RPMs operate within a porous medium. One mechanism that has been suggested is based on the existence of an adsorbed polymer layer where this layer, within a water-wet porous medium, reduces water flow more significantly than oil. In this paper, we propose further developments of this adsorption mechanism of relative permeability modification. The present study is based on two complementary approaches as follows: (i) a series of detailed coreflood experiments has been performed to assess the impact that an adsorbed polymer layer has on the fluid flow pathways. Experimental results on tracer pulses, end point relative permeabilities/saturations and pressure profiles have been utilised in our analysis; (ii) a pore-scale network model has been developed which incorporates the proposed mechanism of RPM operation. The predictions from this model have then been compared with the experimental results. It has been possible to deduce that, not only is water flow affected more significantly than oil, but that the manner in which this occurs is different for each phase. The contributions to both the water and oil conductivity of relative permeability effects (due to saturation changes) and actual "pore blocking" is deduced.
As oilfields around the world become more mature, there is an increasing focus on the large quantities of water that are being produced and on how this water should be managed, both from an economic and environmental perspective. Relative permeability modifiers (RPMs) may provide a simple and effective method of controlling water production without risking damage to oil producing zones. This type of treatment is applied by "bullheading" it into all of the near wellbore formation. This is of particular interest in situations where complete diagnosis of the source of the water is prohibitively costly or the productive oil zone cannot be isolated due to insufficient diagnosis or well geometry. Some success has been reported when RPMs have been applied in the field .
It is accepted within the industry that obtaining a clear understanding of the mechanisms by which RPM treatments act will aid future treatment design and thus enhance the probability of success. The mechanism by which such polymer chemistries have a disproportionate effect on water and oil permeabilities has therefore been the subject of research for some time and various mechanisms have been proposed to explain this phenomenon [2-7]. One mechanism that has been proposed is where an adsorbed polymer layer restricts water flow more significantly than oil flow due to its hydrophilic (oleophobic) nature . Effectively, the adsorbed layer thickness for water, dw, is reduced in the presence of oil, to do, where do<
Three methods are proposed for quickly evaluating the history match of a numerical simulation to actual reservoir performance. All of the methods rely on computing a set of deviation values, each of which is defined to be a calculated simulator result minus the corresponding surveillance measurement value.
For any particular type of surveillance data, such as rates, watercuts, or gas-oil ratios, the deviation values can be grouped by well, by area, or combining all measurements in the database. The first two proposed methods rely on simple graphical presentations of each group of deviation values to show how well the simulation results match the surveillance data. Plotting together the results from more than one simulation run allows a quick comparison of the match for each run, which is useful during the history match process.
The third method converts each deviation value to a quantity called Match Factor, which is a relative measure of the confidence that the simulator actually reproduced the particular reservoir performance at the time the surveillance measurement was made. Weighted-average Match Factors can reveal the degree of match by well, by area, and by data type.
These techniques are especially valuable when matching reservoirs with a large volume of surveillance data. They can help focus the history matching process by identifying areas less well matched. They can identify when the history matching process is not significantly improving the match and can stop.
One of the more challenging aspects of history matching large numerical simulators is assessing how well the simulator results match observed field behavior. Unfortunately, there are few objective measures of the degree of match readily available.
The traditional approach is to plot the observed data values versus time, along with the corresponding simulator output, and visually assess how well the simulator reproduced the measured values1. The quality of this evaluation can vary, depending on the experience and judgment of the simulation engineer. It can also be very time consuming, especially for fields with many wells and years of surveillance data.
This paper describes a two-stage approach, developed by BP Kuwait for Kuwait Oil Company, and validated on large models of giant reservoirs in Kuwait. In the first stage, observed surveillance data values are directly compared to the corresponding predicted values extracted from simulator output. Each pair of observed and predicted values defines a deviation value. Groups of deviation values are presented in two graphical ways, showing the overall degree of match for the type of data in that group for that run. Plotting together results from two or more simulation runs can be used to quickly compare the matches for the runs.
In the second stage of the analysis, each deviation is converted to a Match Factor value, which represents the confidence level that the simulated result matches the actual field behavior represented by the observed surveillance value. This conversion takes into account the inherent uncertainty of the field measurement technique, and the limits of the simulator calculation. Plotting on a map the Match Factors averaged by well quickly shows where a model is better matched compared with other areas. This helps guide where changes in the reservoir description should be made in subsequent history match runs.
Averaging the Match Factors for all values in one data type quickly provides a measure of the overall degree of match for that data type for that simulation run. Comparing this average to that from other runs, shows whether the history match is improving or deteriorating. If the average Match Factor value has stabilized, the useful end of the history match process may have been reached (unless a significant alteration in the reservoir description is made).
This paper reports on a mathematical model to simulate hydrodynamics and fluid-mineral reactions in the fracture within a permeable media. Fluid convection, diffusion and precipitation / dissolution (PD) reaction inside a finite space are solved as a simplified representation of natural fracture mineralization. The problem involves mass transfer within the fluid accompanied by chemical reaction at the fracture surface. Mass-conservation equations for components in the fluid are solved in this problem, and these are coupled with chemical reaction at the fracture surface. The intent of this model is to show the time evolution of fracture aperture shrinkage patterns caused by PD reactions. We present the aperture distribution along the fracture with various boundary conditions. Partially cemented fractures are created if cementation fails to completely fill the fracture or if subsequent dissolution leaches out some of the mineral.
A fracture has been recognized as a fluid conduit that has high permeability relative to that of the surrounding rock matrix in a fractured reservoir. To characterize the reservoirs, core and well log data can be investigated. However, only a few fractures can be observed by the direct measurements; determining which particular fractures are controlling fluid flow and by which mechanisms the wells in fractured media produce are difficult tasks. Often, the most reliable way of characterizing the response of a fractured reservoir is through the analysis of reservoir production behavior, such as well productivity and breakthrough data.
The reservoir quality of sedimentary rocks is closely related to diagenesis, a process involving post-depositional alteration of previously deposited sediments. Rock properties such as porosity, permeability, pore-size distribution, reservoir heterogeneity and spatial correlation can be the product of diagenetic modification of original properties. One of the most important mechanisms of diagenetic processes is chemical reaction between minerals and migrating fluids.
Fluid-mineral reactions are dynamic processes that involve many effects: complex fluid flow, pore space changes, surface chemistry, and the mineral composition. We have developed a simple precipitation / dissolution (PD) model for the flow in a finite parallel plate that represents a single fracture. The idea behind the model is to approximate the description of a variable fracture aperture, surface reaction, and diffusion by a set of parameters and simple rules governing the alteration of fractures. Fracture aperture can either grow or shrink as chemical reaction proceeds at the fracture surface. In this study, we solve mass conservation equations for the components of aqueous and solid phases simultaneously and show the fracture aperture size distribution with time. Calcite cementation is as the form of precipitation considered.
Characterization of Fractured Reservoir
The spatial variations of fracture properties, such as aperture size and orientation, are complicated and irregular so that the characterization of a fractured reservoir is more difficult than that of an unfractured reservoir. One way to approach the characterization is to start from a local characterization of a single fracture and proceed to fracture systems. Parameters for the characterization of fractures include fracture property distributions, fracture density and the size and the shape of matrix blocks.
Especially in low permeability reservoirs, natural fracture permeability is an important issue. In crystalline rocks, the system permeability is almost entirely the result of the fracture network even though the matrix contains most of the reservoir fluid.
In general, the two rock surfaces that bound a fracture are rough. The degree of roughness can be a function of the fracture aperture and the fluid properties within the fractures. Fractures can be partially or fully-filled by mineral precipitation. The nature of a fracture is reservoir-specific, depending on mineral composition, tectonic stress history, diagenesis and petrophysical properties. One purpose of this work is to put some of these connections into a quantitative understanding.
Physically and mathematically rigorous improved models, considering equilibrium and nonequilibrium diffusion gas transport in the liquid phase and resistance of the gas-liquid interface to gas dissolution in liquids, are developed. This allows the accurate determination of the gas diffusion coefficient from experimental measurements of the gas pressure decline in a closed tank by dissolution of gas in the liquid. The short- and long-time solutions of these models are derived analytically under various conditions. These solutions are reformulated for direct determination of the best estimate of the diffusion coefficient by regression of the resultant analytical expressions to experimental data. Procedures are presented and demonstrated for accurate estimation of the gas diffusivity coefficient by conforming the present models to experimental data.
The results developed in this paper can be utilized for determination of the gas diffusivity and the rate of dissolution of the injection gases used for secondary recovery and the rate of separation of light gases from reservoir oil and brine. The present analytic expressions can be facilitated to establish the significance of the equilibrium vs. nonequilibrium conditions under in situ conditions for gas injection, including carbondioxide, nitrogen, and methane. The gas diffusivity in drilling muds and completion fluids can also be determined using the present analytical interpretation methods.
Gas diffusivity is an important parameter determining the rate of dissolution of the injection gases used for secondary recovery and the rate of separation of light gases from the reservoir oil and brine. The frequently used equilibrium assumption in reservoir simulation may lead to significant errors in the prediction of oil recovery by miscible flooding and the miscibility can be optimized for best recovery by developing proper gas injection strategies.
Laboratory measurement of gas diffusivity in liquids is usually accomplished via the measurement of the pressure of gas in contact with certain liquids, such as oil, brine, drilling muds, and completion fluids, in a closed PVT-cell (see Fig. 1) during gas dissolution in the liquid phase. The accuracy of the available models, including by Riazi1, Sachs2,3, Zhang et al.4, are limited by their inherent simplifying assumptions involved in their analytic treatise used for interpretation of the experimental data. Zhang et al.4 have shown that there is no consensus amongst the available simplified models used for diffusivity measurement.
The purpose of this paper is to show how to estimate geomechanical parameters for improved recovery and coalbed methane production processes using an integrated flow model. An integrated flow model combines a petrophysical model with a traditional flow simulator. The usefulness of reservoir geophysical information from an integrated flow model is discussed for the following scenarios: forecasting the reservoir geophysical response of CO2 injection in a mature oil field; estimating subsidence during depletion of an oil reservoir with a gas cap; and predicting the change in geomechanical properties during the life of a coalbed methane reservoir.
The purpose of this paper is to show how to estimate geomechanical parameters for improved recovery and coalbed methane production processes using an integrated flow model. An integrated flow model combines a petrophysical model with a traditional flow simulator. The integrated flow model was originally devised to assist in the design and analysis of timelapse seismology 1-3 because the petrophysical model can calculate such reservoir geophysical attributes as acoustic impedance, reflection coefficient, shear velocity, and compressional velocity. We have found that integrated flow models have other important uses.
Using the integrated flow model, we can readily calculate such geomechanical properties as Poisson's ratio, Young's modulus, and uniaxial compaction. These properties are calculated from a minimal input data set and are provided throughout the life of the reservoir. They give us insight into the behavior of the structure of the reservoir and the impact of structural changes on fluid flow.
The usefulness of reservoir geophysical information from an integrated flow model is discussed for the following scenarios: forecasting the reservoir geophysical response of CO2 injection using advanced well technology in a mature oil field; estimating subsidence during depletion of an oil reservoir with a gas cap; and estimating the change in geomechanical properties during the life of a coalbed methane reservoir. The petrophysical algorithm used in the integrated flow model is described first.
A prototype integrated flow model (IFLO) based on a widely used petrophysical model has been developed and applied to a range of reservoir systems3,4. The petrophysical model must be able to calculate reservoir geophysical attributes that can be compared with seismic velocity and impedance measurements. The algorithm for calculating seismic velocities is a rock physics model5. We refer to the algorithm used in the integrated flow model as a petrophysical algorithm because of its dependence on rock physics properties and petroleum fluid properties.
Bulk density for a porous rock with porosity f is given by ?B=(1-)?m+f?f. Rock matrix density ?m and initial porosity are user-specified input data. Porosity depends on fluid pressure P and porosity compressibility cf =(1/f) (?f / ?P)T at constant temperature T. Oil, water and gas densities (? o, ?w, ? g) and saturations (S o,Sw,Sg) are needed to calculate fluid density ?f. Phase densities are obtained from the fluid properties model in the traditional flow simulator, and saturation distributions are obtained as solutions of the flow equations.
Analytical solutions for the initial stage of one-dimensional countercurrent flow of water and oil in porous media are presented. Expressions are obtained for the time dependence of the water saturation profile and the oil recovered during spontaneous countercurrent imbibition in rod-like, cylindrical and spherical cores for which water is the wetting liquid. Some of the analytical solutions are found to be in good agreement with existing numerical solutions and available experimental data for oil recovery from cores with strong water wettability.
Capillary-driven fluid flow is often important in two-phase flow in fractured porous media and in layered media where individual layers are thin. In such cases, the parameters in the flow equations are complicated functions of saturation due to high nonlinearity arising from a realistic shape of the capillary-pressure curve. The common approach to the problem solution is the use of numerical techniques.
Analytical solutions to fluid flow problems are desirable, because they allow a better understanding of the underlying physics and verification of numerical models. For capillary-driven flow, only a handful of authors have proposed analytical solutions of various degrees of complexity and with certain restrictive assumptions.
Yortsos and Fokas  obtained an analytical solution for a one-dimensional flow with account of capillary pressure; the relative permeabilities and capillary pressure were, however, severely restricted in functional form. Chen  proposed combined analytical-numerical techniques for analysis of radial one-dimensional flow. His work is based on the use of certain asymptotic conditions; it has a strong numerical component.
McWhorter and Sunada  reported quasi-analytical solutions for one-dimensional linear and radial flow. Their work includes both countercurrent and cocurrent flow. These authors limited their solution to an infinite acting medium and assumed that the volume flux at the inlet is of the form At-1/2 where A is constant and t is time.
Pavone et al.  also solved the one-dimensional and two-dimensional (gravity drainage) problem analytically; several assumptions were made by these authors to provide a closed-form solution. The assumptions included (i) infinite gas mobility, (ii) linear liquid-phase relative permeability, and (iii) capillary-pressure dependence on saturation in the form of logarithmic function. As a result of these assumptions, the flow equations became linear.
In this paper, we provide approximate analytical solutions for the initial stage of linear, cylindrical and spherical countercurrent flow of water and oil in a porous medium. We solve the flow equations without restricting the functional form of the relative permeabilities and the capillary pressure. We only assume that the imbibing and the displaced liquids are incompressible and that the porous medium is water-wet. These two assumptions have been made in the work of all authors referred to above.
The "Diffusion" Coefficient
The flow of water and oil in a porous medium is described by a diffusion-type equation in which the quantity
plays the role of diffusion coefficient [McWhorter and Sunada, 1990; Pooladi-Darvish and Firoozabadi, 2000]. In this expression Sw is the water saturation, k (m2) is the absolute permeability of the medium, krw and kro are the relative permeabilities to water and oil, respectively, µw (Pa.s) and µo (Pa.s) are the viscosities of water and oil, respectively, f is the fractional porosity of the medium, Pc=po-pw is the capillary pressure (positive when water is the wetting liquid), and pw (Pa) and po (Pa) are the water and the oil pressures, respectively.
A methodology for constructing an analogue 3D model for clastic reservoirs in an estuary-shoreface depositional environment using outcrop information is discussed. Such analogue models provide valuable information related to reservoir architecture and rock properties that can be used to model sedimentary structures in the subsurface. A new approach for upscaling high-resolution models using nonuniform coarsened grid is introduced. Simulation results for a viscous dominated flow process show that nonuniform grids yield better results compared to uniform grids.
Hydrocarbon accumulations in estuary-shoreface type depositional environment are found at numerous locations worldwide. Complex sub-tidal and inter-tidal estuarine channels and shoreface deposits make these reservoirs extremely heterogeneous and difficult to model based on the information available at a few wells. Developing a detailed, high-resolution analogue model based on extensive outcrop data provides valuable information pertaining to spatial variations in reservoir architecture and rock properties. Information available on analogue models can be used to generate high resolution, stochastic models that are constrained to information such as well data, seismic attribute maps, etc. In order to utilize these equi-probable stochastic models for assessing production performance of the reservoir, a robust technique for upscaling the high-resolution models is necessary.
This paper addresses two important aspects of reservoir characterization: (1) Construction of an analogue model for clastic reservoirs in an estuary-shoreface depositional environment; and, (2) Development and validation of a new approach for upscaling high-resolution reservoir models using a nonuniform coarsened grid.
Analogue models are an important first step towards generating high-resolution stochastic models for reservoirs with complex depositional systems. Outcrops provide valuable information related to reservoir heterogeneity that can be used to model sedimentary structures in the subsurface. The proposal is to utilize outcrops of the upper Cretaceous Virgelle Member of the Milk River Formation in southern Alberta, Canada for developing the analogue model. This was a progradational depositional system representing environments from marine offshore through shoreface and foreshore1. The presence of sub-tidal and inter-tidal estuarine channels is a typical feature of such deposits. A consistent methodology for assembling outcrop data available on multiple 2-D sections, into 3D gridded models is presented. The approach integrates permeability and porosity measurements, stratigraphic columns, multiple vertical sections, and facies information to form a deterministic model for the clastic reservoir.
Modern geostatistical reservoir modeling techniques, such as stochastic methods, are routinely used to generate multiple, equally probable models, each optimally constrained to the available data. These models quantify the uncertainty associated with sparsity of information. Unfortunately, these models are often too large and the resulting flow models are extremely cpu demanding. Upscaling methods try to reduce the size of these detailed models without loss of accuracy. The proposed algorithm first identifies regions of high connectivity using streamline simulation. The nonuniform coarse-scale grid is then constructed preserving areas with high connectivity. This method preserves the geological and flow characteristics better than uniform upscaling techniques.
Using outcrops for modeling
The evaluation of critical factors associated with estuary-shoreface clastic reservoirs can be improved using data from analogue outcrops where large and small-scale heterogeneities can be studied in more detail. Excellent outcrop conditions allow continuous tracing of sedimentary units over large areas leading to a better understanding of reservoir architecture.
The Lagunillas 07 reservoir is located Lake Maracaibo in Venezuela. The reservoir contains the Laguna, Lagunillas and La Rosa formations. The sands are of the Miocene age, poorly consolidated, well sorted and fine-grained. The oil had an 18° API and a viscosity of 21 cp at initial conditions. Oil production began in 1926 and more than 1,000 wells have been drilled. Flank waterflooding, at an average water injection rate of 100,000 stb per day, was introduced for pressure maintenance purposes in 1984. By December 1999, cumulative water injected was 558 million barrels and 36.7 % of the initial oil in place had been produced.
This paper evaluates the impact of the flank waterflooding project in the Lagunillas 07 reservoir. The net oil recovery due to the flank waterflooding is estimated. The real time water front movement is determined with a front-tracking technique that uses fluid production data. The movement shows water fingering due to reservoir heterogeneity. A good match is obtained when the water front movement is cross-correlated with the petrophysical properties of the reservoir. Finally, the impact of the flank waterflooding on recovery factor is highlighted.
The paper concludes that a net oil recovery of 17 million barrels can be attributed to the flank waterflooding from 1984 to 1999. A net oil recovery of 160 million barrels is expected if the flank waterflooding is continued until 2019. The flank waterflooding is therefore considered a success. A future development plan is proposed for the reservoir.
The Lagunillas 07 reservoir is made up of an arch structure with an average strike of 300° and a southwest dip of 3° to 3.5°. The average sand thickness is 86 ft. The reservoir consists of 3 units with cross flow and good vertical communication. The units are the Laguna, the Lagunillas Inferior and the La Rosa. The Laguna and the Lagunillas Inferior units consist of fluvio-deltaic sediments. There is sand continuity in both units. The Laguna unit pinches out to the northeast and is more complex with thinner, less continuous sands. The Lagunillas Inferior unit contains the best oil sands and accounts for about 90% of all initial oil in place. The La Rosa unit is mainly marine and is more heterogeneous with less oil sands.1
The first well was drilled in 1926. Early production data was of poor quality. Subsidence data were not taken until 1940 although surface subsidence due to rock compaction had occurred earlier. Rock property data were obtained from 2 core analysis studies. Average porosity was 30% and connate water saturation was 16%. The reservoir temperature was 152°F.
Field Production and Injection History
Oil production began in 1926 and reached a daily peak of 115,000 stock tank barrels (stb) per day in 1937. It fell to an average of 63,000 stb per day between 1938 and 1958 despite the installation of artificial lift to improve productivity due to pressure depletion. After 1958, the production rate became market dependent. It steadily declined to a low level of 16,000 stb per day in 1972. New infill wells were drilled and the oil production rate increased to 49,000 stb per day. The average reservoir pressure declined by more than 50% between 1926 and 1980. This decline prompted a study to determine the best pressure maintenance method.2 A flank waterflood was recommended by the study. The flank waterflooding was initiated in 1984. Figure 1 shows the daily oil production and the watercut. The watercut was in the 15% to 20% range between 1944 and 1984. It increased to 55% in 1998 as a result of the flank waterflooding.