Mt. Bieber et al., MEASUREMENT AND OVERALL CHARACTERIZATION OF PERMEABILITY ANISOTROPY BY TRACER INJECTION, Revue de l'Institut francais du petrole, 51(3), 1996, pp. 333-347
Citations number
7
Categorie Soggetti
Energy & Fuels","Engineering, Chemical","Engineering, Petroleum
Reservoir rocks can exhibit very strong permeability (K) anisotropy. T
he classical anisotropy measurement methods, which consist of taking s
everal plugs with differently oriented axes from a single core, or of
taking measurements on samples with a particular shape, do not general
ly allow the permeability anisotropy to be fully defined. We have deve
loped a simple, overall method to measure and characterize this anisot
ropy. If, at a point in a porous, permeable, infinite medium, totally
saturated with a relatively incompressible fluid, a second fluid is in
jected that is perfectly miscible with the first, and has the same den
sity, the interlace between these two fluids (i.e., the invasion front
) describes a surface such that, at a given moment, the distance from
the injection point to the surface is proportional to root K in the di
rection under consideration. To provide an overall quantification of t
he permeability of a medium, it suffices to describe the geometrical c
haracteristics of an invasion front during a miscible displacement due
to a pinpoint injection, and to measure a single absolute value of pe
rmeability. The proposed method consists of injecting a salt solution
(e.g., KI) that absorbs X rays into a rock that has been previously sa
turated with brine; the resulting invasion front can be followed easil
y using X ray tomography. The method's validation is based on experime
ntal verification that there is no disturbance due to ionic diffusion,
that the results are insensitive to injection parameters, and there a
re no edge effects. The method has been applied to four rocks that are
often studied in the laboratory, and whose permeability anisotropy is
known from classical measurements. An excellent quantitative concorda
nce is observed between CT scan results and conventional results, as l
ong as the intrinsic heterogeneity of natural porous media as it affec
ts permeability is taken into account. After smoothing raw data using
a polynomial approximation, the experimental data are inverted in term
s of the permeability tensor, using a method analogous to the one deve
loped at Institut Francais du Petrole (IFP) for inverting the elastici
ty tenser. We will show an example based on a real reservoir case befo
re concluding with a discussion of the applicability of this method to
other scales of observation.