The Barua Field, Venezuela: Comparison of the Results and Process of Deterministic versus Stochastic Reservoir Characterization

The onshore Barua Field, near the southeast edge of Lake Maracaibo, is a highly faulted structure producing from the Eocene Pauji and Misoa intervals. This mature field has undergone significant depletion from approximately 30 wells since 1958. A classical, deterministic reservoir characterization study followed by reservoir simulation history matching was conducted. Prior to completing the history match, a stochastic reservoir characterization, including multiple realizations, was performed. Barua’s Eocene interval of near shore littoral bar environments can be modeled using a variety of geostatistical techniques. A pixel-based method was selected over object modeling due to the long continuous nature of the reservoir sands and shales, and the absence of any definitive local genetic depositional facies geometries. Understanding the petrofacies and lithofacies and their relationship in the depositional environment was key to creating a representative stochastic reservoir characterization. This modeling process is described. The process of integrating seismic, well logs, depositional concepts, and facies data is also described. The comparative reservoir simulation results between the stochastic and deterministic approaches, and the resulting impact from an engineering perspective on the history matching process provide insight into the different methodologies. A discussion of the differences and benefits is provided. Introduction An integrated (geophysical, geological, petrophysical, and engineering) study was conducted to (1) develop a geologicalpetrophysical model covering the field area and formations of interest, (2) evaluate past reservoir performance, (3) predict future performance under various operating plans, and (4) prepare appropriate recommendations for field development and operations. Initially, the study plan called for a deterministic approach to reservoir characterization. Subsequently, the operator elected to develop the reservoir description using stochastic methods. Flow simulations were carried out using both the deterministic and stochastic reservoir descriptions. General Information and Field History The Barua field is located onshore near the southeast edge of Lake Maracaibo. It is south of the Mene Grande field and west of the Mototan field, Fig.1. Production is from the Eocene Pauji and Misoa formations at depths ranging from about 3,100 m to more than 4,200 m. Figure 2 is an example log showing sequences and parasequences within the two formations. Figure 3 is a generalized facies cross section, again illustrating the sequences and parasequences. At least 24 sands have been observed to contain by hydrocarbons in some area of the field. The field contained numerous faults, some with throws of several hundred feet. Figure 4 is a structure map on the top Misoa. A 3-D seismic survey covered the area. Production was initiated in 1958, but development was relatively slow and production low until the mid-1980s. Most wells were completed in multiple zones. Production logs in a few wells generally indicated uneven profiles with some zones not contributing. RFT pressures in a number of wells drilled from 1993 to 1998 indicated large differentials, both areally and vertically throughout the field. The reservoir fluid was highly undersaturated initially, but pressures are now below the bubblepoint pressure in some zones in certain areas. The main production mechanism has been fluid and rock expansion. Most wells are on some form of artificial lift. Reservoir Description Data In addition to 3-D seismic data and normal open hole logging suites at the times the wells were drilled, the following items were available for use in developing the geologicalpetrophysical model. Detailed core descriptions including standard facies classifications in four wells Thin section descriptions Routine core analyses Special core analyses Image logs on four wells (one cored) Dipmeters on seven wells SPE 56656 The Barua Field, Venezuela: Comparison of the Results and Process of Deterministic versus Stochastic Reservoir Characterization A. Carnes, SPE, J. Yarus, Smedvig Technologies, P. Cordova, SPE, M. Delgado, H. Rodriguez, PDVSA, P.J. Black, R.L. Brown,SPE, K. Kramer, K. Yang, and L.D. Green, Smedvig Technologies 2 A. Carnes, et al. [SPE 56656] Deterministic Model The structure at top Misoa (Fig. 4) was based upon seismic, well control, and available dipmeter data. The fault interpretation was almost entirely from the seismic. The structural surface was exported to commercial mapping software where depth matching at the wells was ensured. Zone thicknesses were deduced from well control and seismic and contoured using the mapping software. Zone thickness grids were then added or subtracted from the top Misoa grids to develop the structural framework. Nine lithofacies were identified in the core descriptions, four of which were interpreted as being of reservoir quality by oil staining as well as porosity and permeability ranges and fluid contents. After adjustment of core depths to log depths, a correlation was developed which related certain log responses to the lithofacies from core descriptions. The correlation consistently separated reservoir from non-reservoir facies. The prediction also was quite good in the four wells having image logs. After investigating other methods of determining net rock (porosity cutoff, etc.), the facies predictor was considered to be the best approach. A correlation also was developed for predicting permeability from log data. It involved a two-step process. The first was to correlate log responses to the Flow Zone Indicator (FZI)1 which was derived from core data. The second was to calculate permeability using log porosity and the corresponding value of FZI. Figure 5 is a graph of calculated permeability vs core permeability on one of the cored wells. The corresponding depth plot is presented on Fig. 6. The next step in preparing the deterministic model was to apply the facies discriminator to the 0.25-ft log analysis results in each well/zone and calculate net thickness and average values of porosity, permeability, water saturation, and hydrocarbon thickness. Finally, well values of the various parameters were input to the computer mapping program. The contouring algorithm was constrained to broadly honor the conceptual geological model through the use of pseudo well control points as well as the estimated oil-water contact where appropriate. Stochastic Model The stochastic model began with much the same input data as the deterministic model. In general, the work flow used to model heterogeneous reservoirs consists of five steps: construct the structural framework, construct the stratigraphic framework, model the facies, model the petrophysical properties, and upscale the results for flow simulation. Structural Framework. Twenty-four maximum flooding surfaces defining the parasequences along with faults and fault polygons were used to define the structural framework. Five surfaces were added to provide greater resolution where thick shales were present in some of the main reservoir intervals. All surfaces honored the well tops. The zonation is illustrated on Fig. 2. Stratigraphic Framework. The stochastic model used a grid with a fine scale resolution. Each parasequence was discretized into 10 layers using proportional gridding. Proportional gridding was used to emulate the stratigraphic thinning and thickening associated with typical Type II sequences. Thus, the internal fine grid structure consisted of approximately 290 layers with a horizontal resolution of 100m x 100m, and a vertical resolution varying between <0 .33m to 1.5m. The final fine scale model consisted of 8,000,000 cells. Facies Modeling. Facies modeling is important because it provides information about the vertical and lateral relationships between the various reservoir and non-reservoir rocks, ultimately constraining the distribution of the petrophysical properties. Each sedimentological facies has variable directions, geometries, and petrophysical properties that must be considered. Lithofacies derived from petrophysical analyses do not guarantee a geometrical distribution pattern that can be understood in geological terms. Thus, it is preferable to model the sedimentological facies patterns to the extent that they can be defined. The sedimentological facies in the Barua Field are well understood. The reservoir zones consist of shallow marine deposits in stacked Type II parasequences. The sands appear continuous in each progradational parasequence; and local variation in sand thickness is gradual, giving rise to broad sand bars or ridges. Widespread thin shales between and within the reservoir sands indicate a uniform and periodic sedimentological process. The seismic attribute, instantaneous frequency was generally correlated with the depositional facies. In fact, this attribute along with net sand isopachs were used in the deterministic model to develop a set of generalized facies distribution maps (Fig. 7). Integrating such attribute maps into the model is one way to honor the depositional facies pattern and constrain the ensuing petrophysical model. The attribute was extracted for each parasequence conformable to the defined flooding surfaces (horizon slices, Fig. 8). Instantaneous frequency was used as an external trend or pseudo-facies. Petrophysical Modeling. Variogram models were constructed for each zone using the petrophysical data. In general, the data indicated a W-NW anisotropic trend with the maximum direction of continuity 1.5 times that of the minor direction. Variograms were modeled with a either a spherical or exponential function and small nugget (less than 1% of the sill). Figure 9 depicts a set of typical variograms for x, y, and z directions. Figure 10a depicts one layer of the fine scale porosity distribution. Porosity values and the other petrophysical variables were distributed using a Sequential Gaussian Simulation algorithm constrained by instantaneous frequency. Note the similarity in the distribution patterns between Figs. 10a and 8. Ten porosity realizations were made for one of the primary reservoir intervals to evaluate the variation in stochastic modeling by the approach selected in this study. Observation of the multiple realizations shows that the general pattern of petrophysical distribution remained the same and variations are localized. A typical example from another realization is shown in Fig. 10b. [SPE 56656] [THE BARUA FIELD, VENEZUELA:] 3 To further evaluate the variation in volumetrics, a part of the field was selected for pore volume calculation. The differences from realization to realization are very small, usually less than 1% (Fig. 11). Upscaling is described in the Flow Simulations section. Comparison of Two Models. Net thickness, porosity, and permeability maps from one of the zones were selected to illustrate the differences in reservoir description obtained using the two different models. The maps shown here are the upscaled versions for the flow simulations, but they clearly demonstrate the results of the two different approaches. Figures 12 and 13 are the respective deterministic and stochastic net/gross maps for the same zone. Figure 14 through 17 are the corresponding maps for porosity and permeability. The contrast in results from using the two different approaches is obvious. Permeability probably is the most dramatic. One has to believe that the stochastically derived permeability distribution is more realistic than one which looks like a smooth surface constructed by a mapping program. Flow Simulations The same 34 x 45 areal grid system was used for both simulation models (Fig. 18). Grid lines conform to faults where considered feasible. The stochastic reservoir description was upscaled using the same layering as in the deterministic model. Net thickness (net/gross ratio) and porosity upscaling are straightforward. Permeability upscaling has several options. A commonly used option for horizontal permeability is the combination of harmonic and arithmetic averaging. When this was applied, there were a sufficient number of low values in the stochastic model grid to cause the upscaled value to be lower than the zone average value at the well, sometimes much lower. As a result, arithmetic averaging only was used in the upscaling. The resulting values were more in line with those at the wells. A history run with each upscaled permeability further demonstrated the need for the higher values. Vertical permeability was assigned rather than computed, because the shales between the sands were considered to be sealing, unless indicated to be otherwise during the matching process. Reservoir volumes for the two models are very similar. Unfortunately, it can not be concluded which reservoir description provided the better starting point for the history match. The main reason is that the degree of communication among layers at faults is the single biggest factor affecting field and well performance. History matching using the deterministic model was abandoned, but a satisfactory history match ultimately was obtained using the stochastic reservoir characterization. Conclusions 1) Reservoir characterization of the Barua field was an evolutionary process. It is very unlikely that the same procedure would be followed again. If the study were to be started now, it seems probable that the stochastic approach to reservoir characterization would be followed. An added benefit is that the flow simulation grid and the grid data arrays can be generated with the stochastic modeling software in formats competition with commercial simulators. 2) In stochastic reservoir characterization, consequences of using the selected permeability upscaling method(s) should be investigated thoroughly. In our model, the upscaling scheme used was rather simple. More sophisticated methods are available, some of which may prove valueable. Acknowledgements Many colleagues of the authors participated in various aspects of the study. Their contributions are acknowledged and appreciated. Thanks also are due Helga Ehrhardt of Smedvig who typed the manuscript. References: 1. Amaefule, V.O., Altunbay, M., Tiab, D., Kersey, D,G., and Keelan, D.K: “Enhanced Reservoir Description UsingCore and Log Data to Identify Hydrandic (Flow) Units and Predict Permeability in Uncored Intervals/Wells.” SPE 26436 (1993). 2. Van Wagoner, J.C., Mitchum, R.M., Campioun, k.m., and Rahmanian, V.D.: Siliciclastic Sequence Stratigraphy in Well Logs, Cores, and Outcrops, AAPG Methods in Exploration Series, No. 7, American Association of Petroleum Geologists, Tulsa (1996) 55. 3. Deutsch, C.V. and Meehan, D.N.: “Geostatistical Techniques Improve Reservoir Management,” Petroleum Engineer International, (March 1996) 21. 4. Goovaerts. P: Geostatistics for Natural Resource Evaluation, Oxford University Press,New York (1997) 483. 5. Deutsch, C.V. and Journal, A.G.: GSLIB Geostatistical Software Library and User’s Guide, Oxford University Press, New York (1998) 369 SI Metric Conversion Factors ft x 3.048* E-01 = m * conversion factor is exact.