Observations of a potential size effect in experimental determination of the hydraulic properties of fractures

TLDR

Recent experiments on how normal stress affects hydraulic conductivity of single fractures have used rock specimens with cross‑sectional areas ranging from 0.02 m² to over 1.0 m², while field cores are typically no larger than 0.15 m in diameter. The study aims to determine the optimum specimen size that yields reliable fracture conductivity measurements under stress. This was investigated by comparing hydraulic conductivity at maximum attainable stresses (10–20 MPa) in ultralarge cores (0.95 m diameter) to smaller specimens. The results show that minimum hydraulic conductivity values increase with specimen size, revealing a size effect that can make laboratory measurements nonconservative for typical core sizes and suggesting that fracture permeabilities measured in small cores may be significantly lower than those found in the field.

Abstract

In several recent investigations, experimental studies on the effect of normal stress on the hydraulic conductivity of a single fracture were made by using three rock specimens ranging in cross‐sectional area from 0.02 m 2 to over 1.0 m 2 . At the maximum stress levels that could be attained (10–20 MPa), minimum values of the fracture hydraulic conductivity were not the same for each rock specimen. These minimum values increased with specimen size, an indication that the determination of fracture conductivity may be significantly influenced by a size effect. The implications of these results are important. Cores collected in the field are normally not larger than 0.15 m in diameter. However, the results of this work suggest that when a core of this size is used for laboratory investigations, the results may be nonconservative in that fracture permeabilities will be significantly lower than those that will be found in the field. If the results with an ultralarge Core (0.95 m in diameter) are more indicative of field conductivities for a fracture under stress, then further work is needed to determine optimum specimen size so that reliable results on flow in fractures under stress will be available.