The objective of the project is to produce a software package for modelling Hydrocarbon Naturally Fractured Reservoirs, incorporating all the improvements achieved by new conceptual and stochastic models, able to produce input data compatible with the major reservoir simulators.

For the past 3 years we have been studying and developing new techniques and mathematical models to describe fractured reservoirs, in order to improve the confidence in simulation input data. The work has encompassed detailed geological/geophysical and petrophysical studies and an intensive research on fractal analysis, in order to characterize and evaluate, with the assitance of a black oil simulator, the performance of a fractured reservoir.

The methodology behind this research was tested with a Middle-East fractured resevoir, where geological/geophysical, petrophysical and production data was available. The success of alternative realizations for fluid predictions was measured and quantified.

Specific software to perform fractal analysis and stochastic simulation has been developed. Several realizations of the reservoirs fault pattern were obtained, and their impact on fluid flow was tested.

A numerical simulation model was prepared, wherein each block was assigned using interwell properties derived from a deterministic approach. Other important parameters governing fluid flow such as relative permeabilities, pressure and mobility ratios, were also input into the model. Several realizations of structural discontinuities were generated using a stochastic approach.

The selection of the most representative realization was made according to the similitude between the fractal dimension of the stochastic fault pattern and the fractal dimension of the original seismic fault pattern of the field, as well as cross-checking both areal fault density maps.

Although, the software that was developed to do fractal analysis of the fault patterns is completed it still needs further work for the integration of the software modules.

The software can be subdivided into two main groups. The first group has to do with the calculation of the fractal dimension of the fault patterns, while the second one consists of an algorithm that enables the user to apply correction effectsto the number of faults to be generated.

A statistical approach has been developed for the generation of bidimensional spatial distributions of fracture networks based on the analysis of their geometrical parameters and using different types of statistical distributions such as exponential, normal and lognormal.

An important achievement obtained during this project was to show that fractal geometry analysis, applied to a fractured reservoir, yields an effective method for selecting the most likely stochastic realizations of several fault patterns, by using the fractal dimension.

The Project spanned different critical areas from Structural Analysis (Phase I), Geology and Reservoir Engineering (Phase II), Reservoir Description (Phase III) to testing programme with a commercial simulator (Phase IV).

Phase I has been completed. The available data was screened and organized, geological and geophysical information was processed, tectonic interpretation of the fault pattern was performed, and geological computer modelling was finalized.

In Phase II, important enhancements have been achieved and a statistical approach was developed for modelling the fracture networks. This approach aims at the generation of a synthetic bidimensional spatial distribution of fractures, based on statistical information of the geometric parameters and using different statistical distributions.

Phase III related to Reservoir Description, was focused on the analysis and integration of data from different sources such as core analysis, logs and well test interpretation, in order to provide clues for the identification of fractured zones and characterization of the petrophysical properties.

Phase IV of the Project consisted on the preparation of input data for the initialization of simulation models, in order to test the developed methodologies. The simulation was developed using single-well models, cross sectional models and sector models, so that improvements in the description of dual-porosity systems, specially the fracture properties and the matrix blocks/fractures interaction, were obtained.

Different simulation scenarios using different spatial distributions of faults were performed and a detailed sensitivity analysis is presented and discussed.

As a consequence of the application of this methodology, more reliable Reservoir Description Models are built, and useful performance predictions are made, creating conditions for the improvement of simulation results.

The project has fulfilled the main objectices which were in the originally proposed plan. However, for two or three items, time constraints did not allow a deeper analysis and, consequently, they can be worked out in future research programmes. These items are related to the following subjects :

i) Transference of scales - further work is required in order to build a robust Mathematical Model able to transfer reservoir properties ans fracture geometric characteristics recorded in one scale (e.g. seismic) to other scales (e.g simulator grid blocks).

ii) Attempt to characterize reservoir heterogeneity by a multiple correlation model - further work is required to tackle the multiple correlations between well logging, well testing, geological, petrophysical and seismic parameters in order to quantify the degree of reservoir heterogeneity. A critical point at this stage is to gather all the required information on a particular field but quite often, this task was not possible, which is the major reason behind the shortfalls of Project Phase IV (Quantification of Reservoir Heterogeneity).