Incorporating Information About Edge Effects When Simulating Lithofacies

Some years ago, earth scientists came to realize that knowing more about the geology of an orebody or an oil reservoir makes it easier to make the appropriate decisions concerning mine planning or reservoir exploitation. Geostatistical techniques for simulating lithofacies—that is, the geometry of the geology—were developed as a result of this. These methods should be able to produce geological images that respect not only the anisotropies of the different lithofacies but also their spatial layout relative to one another. While indicator variograms ensure that anisotropies are respected, another tool needs to be incorporated in the simulation technique to reflect the relative spatial layout of the different lithofacies. We propose to use the concept of edge effects that define the position of one lithofacies relative to another. Simple tests using direct and cross indicator variograms confirm the presence or absence of edge effects. We investigate if and how edge effect information can be incorporated in the different indicator simulation techniques—sequential indicator simulations, simulated annealing, the truncated Gaussian method and plurigaussian simulations. Results show that the choice of simulation method must be guided by the edge effect characteristics of the experimental lithologic data.