SUB SURFACE REPRESENTATIONS. PATENTS, PROBLEM AND POTENTIAL

Since geocellular modeling was first introduced into the geoscience armoury in the late 80’s it has
become a ubiquitous tool. Once the domain of those large companies with supercomputers or
high performance workstations, the software andcomputing power have both advanced so that
geocellular models are now routinely built on laptops in airport lounges.

In the late 1980’s when reservoir characterization was emerging as an investigative technique. The
software that was used was designed to treat the reservoir in an abstract way. There was not
necessarily any requirement to construct a model of your reservoir, rather a region that was in some
way had similar characteristics to regions within your reservoir. It is the departure from this
experimental approach to the understanding of the subsurface that, I believe has caused most of the
problems with characterization. In some ways it is now more important to have a model that looks like your reservoir rather than a model which
behaves like your reservoir and from which, inferences and predictions may be made.

Early software particularly coming out of BP Research was based around constructing objects
that would represent the heterogeneous aspects of the reservoir. Shale barriers, fluvial channels and
fractures were all modeled as objects like trees and signs in computer games. The driving forces
for the construction of geocellular models came from two sources. Firstly, from the software
company Stratamodel, which was probably the first geological modeling software, who created
the idea of dividing the reservoir into cells. Secondly from the emergence of geostatistical
techniques in the petroleum industry which could take advantage of the compute power of the super
workstations of the day and required regular – or vaguely regular 3D grids to operate. By the early
1990s the battle had been won and object modeling was not to reappear in any serious form
for another 10-12 years.

By :

Stephen Tyson, C.Math
Paradigm, Kuala Lumpur, Malaysia.
stevet@pdgm.com