Resumo da Dissertação da Tese de Doutoramento do amigo Herlander Lima

Inverse Modelling of Petroleum Reservoir

Integration of Dynamic Data into Stochastic Model and Upscaling
abstract

Reduction of the uncertainty related to petrophysical properties is still a challenge in reservoir characterisation. An accurate estimation of the spatial distribution of petrophysical properties (e.g. permeability) translates into higher success rates in infill drilling, and fewer wells required to drain the reservoir.
On the other hand, reservoir simulation of very large fine grid model (greater than 1 million blocks) is impracticable using traditional simulators, such as those based on finite difference. Owing to this difficulty, upscaling is justifiable for very fine-scale reservoir models and/or when many runs are necessary and the computational performance is very important.
Considering the above mentioned problems this thesis was developed with the following two purposes:

i) enhance the characterization of subsurface reservoirs by combining reservoir static and dynamic data and models to improve the prediction of petrophysical properties. This goal is fulfilled by formulating, implementing, and demonstrating the applicability of an emerging algorithm, the integration of static and dynamic (production-related observations) data by inverse conditional simulation in which stochastic images of permeability are simulated and perturbed through Direct Sequential Simulation and Cosimulation and flow simulations are obtained from a traditional black oil simulator. This inversion approach leads to better and less uncertain results because of the larger amount of data and constraints available to inform the predictions; and

ii) reduce the number of reservoir parameters through an upscaling technique based on hydraulic zonation of porous media in order to build a coarse (upscaled) grid which preserve the characteristics of grid simulation. The upscaling technique leads to consistent results and can be a useful tool for reservoir simulation purpose.

Key-words: Stochastic modelling; Deterministic modelling; Inverse conditional simulation; Data integration; Reservoir characterisation; Upscaling.