HOUSTON – Hampson-Russell Software & Services has released LithoSi, an addition to its recently revamped HRS-9 suite for reservoir characterization.
LithoSi is designed to predict facies type and fluid content from seismic inversion attributes, while quantifying the uncertainty in the prediction.
The LithoSi workflow is based on modeling multi-variate probability density functions from log data or from inverted elastic attributes at well locations. The probability functions are then mapped to elastic property volumes from seismic inversion using a supervised Bayesian classification. The outputs are a series of litho-probability cubes and a cube of the most probable facies.
Example output of the LithoSi process, showing the distribution of sand within a geologic formation, color coded with the probability of its presence. The ability to differentiate facies type along with its probability of occurrence yields more refined risk analyses.