Upcoming Presentation

Dr. Barrie Wells will give a presentation on "Modelling Risk and Uncertainty using Multipoint Statistics" at the "What Makes A Geoscience Model Useful?" conference at the Geological Society in London on Tuesday 22nd September.

Abstract:

Modelling uncertainty typically involves creating multiple equally likely model scenarios, or realisations, which can be used to model the probability density of outcomes. The development of software that enables construction of realisations from Sequential Gaussian Simulation (SGS) was therefore a significant breakthrough. As Multiple Point Statistics (MPS) was promoted as a further development of geostatistical modelling, potentially replacing SGS, the question is asked: how can this be applied to risk and uncertainty modelling? The answer, that it cannot, is not based on questions of meeting stationarity requirements but on its ability to generate equally likely realisations: as presently implemented in software, it cannot do this and MPS is therefore not useful as a tool for creating models designed to investigate risk. Recent work on flume tank experiments may be able to change this and make MPS models useful for risk analysis. Clearly, it is not feasible to run a flume tank experiment for each experiments realisation but, by using mass balance considerations and parameterising the sediment input to the flume tank experiment, Duller and Straub, at University of Liverpool, UK, and Tulane University, USA, have shown how key reservoir architecture can be parameterised to allow equally likely scenarios to be generated for subsequent input to Monte Carlo models, and hence to generate models of probability density for the required outcomes. Whilst this is limited to a subset of sedimentary processes, it is nevertheless a breakthrough in moving risk and uncertainty analysis forward in line with developments in deterministic modelling.

PETROG v6.0.5 now available for download

A number of new features, enhancements and fixes are available in this update for PETROG v6 users. >>