TY - JOUR

T1 - Modeling flow in naturally fractured reservoirs

T2 - effect of fracture aperture distribution on dominant sub-network for flow

AU - Gong, J.

AU - Rossen, W. R.

PY - 2017/2/1

Y1 - 2017/2/1

N2 - Fracture network connectivity and aperture (or conductivity) distribution are two crucial features controlling flow behavior of naturally fractured reservoirs. The effect of connectivity on flow properties is well documented. In this paper, however, we focus here on the influence of fracture aperture distribution. We model a two-dimensional fractured reservoir in which the matrix is impermeable and the fractures are well connected. The fractures obey a power-law length distribution, as observed in natural fracture networks. For the aperture distribution, since the information from subsurface fracture networks is limited, we test a number of cases: log-normal distributions (from narrow to broad), power-law distributions (from narrow to broad), and one case where the aperture is proportional to the fracture length. We find that even a well-connected fracture network can behave like a much sparser network when the aperture distribution is broad enough (α ≤ 2 for power-law aperture distributions and σ ≥ 0.4 for log-normal aperture distributions). Specifically, most fractures can be eliminated leaving the remaining dominant sub-network with 90% of the permeability of the original fracture network. We determine how broad the aperture distribution must be to approach this behavior and the dependence of the dominant sub-network on the parameters of the aperture distribution. We also explore whether one can identify the dominant sub-network without doing flow calculations.

AB - Fracture network connectivity and aperture (or conductivity) distribution are two crucial features controlling flow behavior of naturally fractured reservoirs. The effect of connectivity on flow properties is well documented. In this paper, however, we focus here on the influence of fracture aperture distribution. We model a two-dimensional fractured reservoir in which the matrix is impermeable and the fractures are well connected. The fractures obey a power-law length distribution, as observed in natural fracture networks. For the aperture distribution, since the information from subsurface fracture networks is limited, we test a number of cases: log-normal distributions (from narrow to broad), power-law distributions (from narrow to broad), and one case where the aperture is proportional to the fracture length. We find that even a well-connected fracture network can behave like a much sparser network when the aperture distribution is broad enough (α ≤ 2 for power-law aperture distributions and σ ≥ 0.4 for log-normal aperture distributions). Specifically, most fractures can be eliminated leaving the remaining dominant sub-network with 90% of the permeability of the original fracture network. We determine how broad the aperture distribution must be to approach this behavior and the dependence of the dominant sub-network on the parameters of the aperture distribution. We also explore whether one can identify the dominant sub-network without doing flow calculations.

KW - Effective permeability

KW - Naturally fractured reservoir

KW - Non-uniform flow

KW - Percolation

KW - Waterflood

UR - http://www.scopus.com/inward/record.url?scp=85006340162&partnerID=8YFLogxK

UR - http://resolver.tudelft.nl/uuid:db0f7ec7-6a75-4f5b-a1f1-4eedfa092548

U2 - 10.1007/s12182-016-0132-3

DO - 10.1007/s12182-016-0132-3

M3 - Article

AN - SCOPUS:85006340162

SN - 1672-5107

VL - 14

SP - 138

EP - 154

JO - Petroleum Science

JF - Petroleum Science

IS - 1

ER -