Page 374 - Numerical Analysis and Modelling in Geomechanics

RESERVOIR COMPACTION, SUBSIDENCE AND WELL DAMAGE 355
Field-scale model
Contours of the pore pressure field for four of the eighteen years of the
simulation, 1978, 1980, 1989, and 1995, are superposed on the deformed mesh in
Figure 11.9a to d. For clarity, the deformations were exaggerated by a factor of
75. The contours in the center of the reservoir were due to decreases in pore
pressures from production. The pore pressure field varied considerably in the
horizontal direction due to the detailed representation of individual wells and
hydraulic fractures, as well as vertically due to the gravitational effect. Two
darker areas above and to the left and right of the reservoir were due to water
injection. Vertical shortening of elements within the reservoir illustrates
reservoir compaction, which produces ground subsidence. Moreover, rock has
been dragged towards the center of the reservoir, which illustrates the mechanism
of shearing of rock. Local well deformations can also be observed. Elements
laterally shorten around production wells due to the local reduction in pore
pressure, while elements laterally expand around water injection wells. These
deformations can cause shearing of wells within their area of influence, an effect
referred to as “well-to-well interactions.” 38
A field program to monitor surface subsidence using surface monuments has
been in place for several years in the Belridge field. Computed surface
subsidence is compared with data collected from a surface monument array for
1991 and 1995 in Figure 11.10. In 1995, the subsidence reached a maximum of
approximately 10 feet (3 metres) in this area of the field. The agreement between
computed and measured subsidence as a function of time verifies the accuracy of
the magnitude of the pore pressures, as well as the transient nature of simulated
field operations. However, subsidence is a large-scale effect produced by
deformations on the reservoir scale. Local effects, such as deformations around
individual wells, are more difficult to capture accurately on such a scale with the
resolution of a field-scale model. Changing the friction factor for the contact
surfaces did not affect the magnitude of the subsidence substantially. Also, there
is considerable spatial variation in subsidence within Section 33, as production
and injection programs vary over the field. 39,40
Relative slip between the Al and Dl shale horizons and between the
Tularediatomite unconformity is shown in Figure 11.11 for the 1995 simulation.
For the shown curves the sliding friction factor was assumed to be 0.2. Relative
slip is observed to be highly heterogeneous along each interface, due to local
variations in the pore pressure field. In addition, relative slip is observed to
increase as the depth to the interface increases, which is due to the increase in
shearing deformations with depth. Relative slip is observed to reach a maximum
of about 11 inches in 1995.
The transient nature of the relative slip is illustrated by comparing plots of slip
on the Tulare-diatomite unconformity computed in the 1987 and 1995
simulations, as plotted in Figure 11.12. Both the magnitude and the character of
the slip change with time.