Accounting for bias and uncertainty in facies estimations from deterministic inversions
John Pendrel, Henk J. Schouten, R. Bornard · 2016
Bayesian inference procedures can be used as interpretation tools for seismic inversions. The results are facies and their probabilities of occurrence derived from the native outcomes of inversions or their derivatives. Although deterministic inversions produce a single outcome, they have uncertainties associated with them. Further, due to inappropriate low frequency models or thin bedding, biases in the inversion properties can arise. We use a phenomenological approach to model these effects and separately correct for them in the subsequent Bayesian inference. The results are facies interpretations and pay maps which account for bias and uncertainties and will provide greater confidence in reservoir volume estimates. Presentation Date: Tuesday, October 18, 2016 Start Time: 3:45:00 PM Location: 156 Presentation Type: ORAL