On the Representative elementary volume concept

Real porous media heterogeneities scales differ by the order of magnitude and have significant influence on filtration flows. Laboratory measurements of rock properties are carried out using core samples of limited size, which may lead to inaccuracies in full-scale reservoir simulation. In the paper, the correctness evaluation criterion for this procedure is investigated using representative elementary volume (REV) concept. It is shown that REV must contain significant number of medium features both for stochastic and deterministic permeability distribution cases. Study of the REV for a two-phase system has revealed that the single-phase case consideration is sufficient for REV sizes estimation. Using the modern volume tomography methods, one can obtain 3D parameters distribution, which allows to use the criteria suggested and to complement the results of commonly used laboratory measurements.