Summary
Much of the oil in Saudi Arabia is stored in giant and supergiant
multireservoir fields. The Arab-D limestone is the most important of these and
the most prolific. The large volumes, excellent porosity, and high productivity
of these reservoirs do not mask the fact that these carbonates have complex
pore systems. The problems associated with heterogeneous carbonate reservoirs
pose significant and longstanding modeling complications that are not yet fully
addressed by the industry. One important difficulty is the accurate modeling of
the substantial transition zones present above the free-water levels (FWLs). In
these giant fields, these transition zones hold large amounts of oil and are
important commercial objectives. Commerciality requires accurate assessment of
saturations and rock properties. Standard J-function methods are
inadequate to model the well-log observed saturation-height behavior in the
transition zones. It is necessary to characterize and account for the pore
system variations and scale when modeling the saturation behavior of large rock
volumes. The reservoir properties of geocells and wellbores must be reconciled
with the measurements on core plugs. The measurements performed on these tiny
pieces of rock need to be upscaled to represent the reservoir bulk
properties.
Upscaling of core-plug-scale, laboratory measured porosimetry data and
transport properties has been a general and persistent problem since the
beginning of reservoir simulation. This critical step has been handled, over
the years, using a wide variety of numerical computational schemes,
approximations, and empirical methods. In this paper, we take the different and
very specific approach of upscaling the capillary pressure data for the Arab-D
limestone. We base the approach on the availability of a large amount of
mercury (Hg) -injection data and statistical analysis thereof, obtained by
fitting hundreds of individual core plugs to Thomeer functions.
For the Arab-D limestone, and similar carbonates, we derive a closed-form
analytic expression for the upscaled capillary pressure function, which has
significant implications for improving transition-zone hydrocarbon-volume
estimates for this important petroleum system. The analytic expression also
offers major efficiencies compared with other methods used by petroleum
engineers, provided that the pore systems are adequately investigated and
statistically characterized. A key result of the upscaled formalism is that
reservoir cells, consisting of a large variation of pore systems, will start to
fill with hydrocarbons much closer to the FWL than when using saturation-height
functions based on simple averaged pore system parameter values. Therefore,
transition zones for upscaled reservoir elements (and well-log volumes) are
thicker than simple calculations based on data from single core plugs would
indicate. The accurate upscaling of pore-system architecture is a major step
toward the full understanding of the fluid-rock interactions of giant-field
transition zones in the Middle East and is an industry technical milestone.
© 2010. Society of Petroleum Engineers
View full textPDF
(
980 KB
)
History
- Original manuscript received:
8 October 2009
- Revised manuscript received:
25 April 2010
- Manuscript approved:
24 May 2010
- Published online:
23 December 2010
- Version of record:
21 February 2011
View Cart
