The in situ hydromechanical
behaviors
within a fracturing fracture (induced by hydraulic fracturing) without
a proppant support was usually estimated using artificially splitting
fractures (induced by Brazilian splitting). This causes a deviation
of fracture permeability from the actual one, but it is not fully
understood how strong it could be. We investigate this issue by series
of laboratory gas flow tests under effective confining pressure varying
from 1 to 40 MPa and inlet gas pressure varying from 0.1 to 2.3 MPa
and by a numerical simulation based on reproduction of fracture geometry
and matching the experimental data. The obtained results show that
(1) fracturing fracture surfaces are relatively smooth with fewer
asperities and are prone to be in contact in response to the loading;
(2) the permeability of fracturing fractures is lower than that of
splitting fractures, and the error between them grows with the increasing
loading, implying the actual permeability evaluated with splitting
fracture is somewhat overestimated; (3) laminar and transitional flows
dominate in both fractures, but a nonlinear effect is stronger in
splitting fractures; (4) the onset of nonlinearities occurs at a relatively
large Reynolds number (>10) because of the larger amount of inertial
losses in fracturing fractures; and (5) the tortuous fracture propagation
path is dependent on rock heterogeneity, and such dependence is stronger
in the Brazilian splitting test.