Estimating Uncertainty in Reserves from Deterministic Seismic Inversion

We present a methodology for deriving uncertainty estimates of rock and fluid distributions for qualitative and quantitative interpretation of deterministic seismic inversion data. The method integrates rock physics modelling with geostatistical forward modelling to provide a basis for interpretation of elastic properties from seismic inversion. The workflow allows for the inclusion of uncertainties due to seismic noise, limited seismic resolution, overlap in elastic properties of different rocks and fluid fills, time-to-depth conversion for net pay calculations and scatter in the relationships between elastic properties and reservoir properties. The sparsity of control, well log data is also handled. <br><br>The initial output of the process is a volume of probability for each rock/fluid type categorised. These volumes can be used for qualitative interpretation of possible distributions of the rock/fluid types and connectivity of geobodies. In addition, by combining the probability volumes with the elastic property volumes a range of estimates of net pay can be produced that are constrained to the known values at well locations. This provides uncertainty estimates of the distribution and volume of hydrocarbons in place and can therefore be used for risk management.<br>