Integrating Multiple Log Measurements for Uncertainty Reduction in Reservoir Evaluation
Palavras-chave:
Reservoir Characterization, Uncertainty Analysis, Unsupervised Clustering, Data AnalyticsResumo
The methodology presented in this work aims at reducing uncertainties in the petrophysical evaluation using
machine learning, statistics, and physics-driven methods to increase the level of confidence in advanced workflows
in complex formations.
The proposed method saves days in repetitive processes when compared to traditional workflows and quanti-
fies the level of confidence associated with the answers considering multiple uncertainty sources.
This methodology enables the petrophysicist to provide an analysis containing one final porosity, water sat-
uration, and permeability estimation with minimized uncertainty considering the multiple measurements available
for the well and the multiple sources of uncertainties in a reproducible and fast manner.
