Title:
|
Incorporation of the stress sensitivity of the permeability of natural fractures and breccia zones into production simulation models : an example from the Clair Field
|
Fracture permeability in some reservoirs may be stress sensitive; which can affect fluid flow
during production and should be taken into account in reservoir management. However, the
petroleum industry rarely incorporates the impact of stress sensitive fracture permeability into
production simulation models.
This study investigates the potential impact of stress sensitive fracture permeability in the Clair
Field using a unique approach that integrates data and workflows from geology, petrophysics,
geophysics, geomechanics and numerical simulation. This is first tested on a synthetic block
model and is later implemented on the Clair Field reservoir model to confirm the impact of
changes in fracture permeability during production on productivity index (PI). This is one of the
first studies where the stress path parameters are mechanically justified to determine the
appropriate simulation method to be used.
Despite offering extensive reservoir and production data for this study, the Clair Field is an
incredibly complex and heterogeneous reservoir. The 27 year appraisal period alone
demonstrates the high level of complexity, particularly in terms of its fracture system, much of
which remains to be understood. In this respect, this study is particularly ambitious.
Analogue outcrops have been studied, increasing the understanding of the controls on fracture
distribution in the Clair Field. Fracture spacing Is shown to have no impact on the stress
sensitive permeability, whereas fracture apertures, which are difficult to obtain, may have a
significant impact.
Evaluation of fracture representation in synthetic block models reveal explicit and implicit
models may produce very different simulated results, particularly when describing large scale
stress sensitive fracture zones that connect wells. This demonstrates the danger of poor
representation of the fracture system in the simulation model. Despite increased
computational expense, more accurate fracture descriptions are likely to generate more
precise simulation outcomes.
A significant impact on PI was observed in the synthetic block models, while little variation was
observed in the Clair full field model because during production the reservoir experiences very
little pressure drawdown. Non-unique and highly heterogeneous full field simulation outputs,
however, are difficult to interpret. Furthermore, owing to the poor understanding of relative permeability and insufficient knowledge regarding the fracture distribution render the results
of this study largely inconclusive. While PI may be used to demonstrate stress sensitivity, time-dependent
or pressure-dependent changes in PI may also be accredited to other factors. These
should be ruled out or confirmed to obtain a more realistic expectation of what drives the PI
changes during production. Therefore, an appropriate decision of whether to incorporate
stress sensitive permeability to the simulation can be made.
Although much attention has been given to the case study in this work, the same methodology
and recommendations are applicable to other fractured reservoirs. It is both difficult and
impractical to design a general study whereby findings can be used to make generalizations for
all reservoirs due to the natural variability of oil and gas reservoirs worldwide. Synthetic block
models demonstrate that homogeneous and small scale simulations cannot necessarily be
used to make inferences to the behaviour of full field heterogeneous reservoir models.
However, with appropriate fracture parameter input, both these simulation and pre-simulation
studies of how stress changes are likely to impact the permeability can provide
good approximations of the likely reservoir stress sensitivity. This highlights the importance of
real data in this study.
|