Uncertainty Analysis for GIS Estimates of the CO₂ Storage Resource in the Oriskany Sandstone

This study focuses on the geologic CO₂ sequestration resource in deep saline-filled formations, a class of repositories believed to make up the bulk of the storage resource. The goals of this research are (1) to better understand the sources of uncertainty in deep saline-filled formation (DSF) resource estimates by developing a sequestration resource model, (2) to employ kriging and cokriging to estimate input parameters, and (3) using this model to probabilistically quantify the sequestration resource for the Oriskany sandstone in Pennsylvania. The geologic framework of the model is based on data provided by the various government agencies. Due to the fact that this dataset includes information from wells drilled by different operators there exists a need to properly estimate values for parameters that aren't available for all wells.

The results of statistical studies of reservoir and structural properties of the Oriskany sandstone suggest the best-fit distribution for average formation depth. Regression models allow for the prediction of (1) porosity as a function of depth and (2) formation temperature and pressure as functions of depth. The equation of state developed by Span and Wagner is used for calculation of density as a function of temperature and pressure. Storage resource estimates are developed using a Monte Carlo simulation. Four parameters are treated as uncertain: average formation depth and regression model parameter estimates for porosity, pressure and temperature. The results of the simulation for storage resource show that there is a large variation in storage resource estimates and that a point estimate using mean values of input parameters does not result in the same capacity as the mean of the simulated storage resource distribution. Results indicate that for the baseline storage scenario, which assumes an efficiency factor of 2%, the Pennsylvania part of the Oriskany formation can hold 0.14 gigatonnes of CO₂. In this scenario mass of CO₂ varies from a 5^th percentile of 0.05 gigatonnes to a 95^th percentile of 0.5 gigatonnes. Sensitivity analysis indicates that the two variables that contribute the most to the uncertainty in estimates are porosity and temperature.
 

AAPG Search and Discovery Article #90142 © 2012 AAPG Annual Convention and Exhibition, April 22-25, 2012, Long Beach, California