Accounting for Proportion Data in Stochastic Simulation of Fluvial Deposits

@inproceedings{ViseurRM1999,
 abstract = { In the frame of stochastic simulation procedure applied to reservoir characterization, the weil data represents "hard" data. This data gives minute parts of the distribution of the geological features encountered into the subsurface of the earth. It is a punctual and geometrical data. This has led us to develop within a stochastic procedure a direct approach [13] for fitting the simulated objects (channels) to the weil data. This approach is also based on heurisitics stemming from geological knowledge. It can potentially incorporate deterministic data, derived from weil-test (e.g. weil correlations) or seismic interpretations (e.g. object boundaries). As opposed to weil data, proportion data is often considered to be "soft" data to be taken into account by a stochastic procedure. This kind of data may be computed from weil observations and weil-tests [9] or inferred from seismic analysis. Vertical proportion curves or areal maps of lithofacies proportions may be seen as rellections of object distribution trends or as probabilities of lithofacies presence. This data should control the global number of objects to simulate and the il' preferential locations. Concerning the approach we describe in this paper, it seems more suit able to simulate in one go the geological features (channels) that intersect the weil bores and the other ones, while being controlled by the proportion data. Indeed, honouring weil data independantly from this proportion information can generate discrepant simulations as compared to the proportion maps and curves. Before explaining such an approach, we will use terms that must be defined. Indeed, as Iluvial deposits have a hierachical architecture, several orders of geological bodies may be simulated, ranged over the small-scale to the macroscale features. Our trend is more to simulate objects that are correlated to major events during the channel life : namely the avulsions[l]. In this way, we will first generate objects that will be referred to as avulsion surfaces, in the following, and second, one channel-beU body pel' constructed surface. },
 author = { Viseur, Sophie },
 booktitle = { 19th gOcad Meeting },
 month = { "june" },
 publisher = { ASGA },
 title = { Accounting for Proportion Data in Stochastic Simulation of Fluvial Deposits },
 year = { 1999 }
}