CO2 storage associated with CO2 enhanced oil recovery: A statistical analysis of historical operations

Abstract

This work analyzes a database of 31 existing CO2 enhanced oil recovery (EOR) projects that was compiled for the estimation of oil reserves to better understand the CO2 retention, incremental oil recovery, and net CO2 utilization for these oil fields. The measured data begin at the start date of the CO2 flood and extend through the year 2007. Cumulative CO2 retention (in the formation), incremental oil recovery factors, and net CO2 utilization factors were calculated for each of the sites. To express all site data on a common dimensionless scale, the data were extrapolated to 300% cumulative hydrocarbon pore volume (HCPV) by fitting nonlinear functions. Summary statistics were then calculated from 0% to 300% HCPV. Across all 31 sites, the 10th, 50th (median), and 90th percentile values for the three factors at 300% HCPV were: CO2 retention: 23.1%, 48.3%, and 61.8% retained; incremental oil recovery: 5.3%, 12.2%, and 21.5% of OOIP (original oil in place); and net CO2 utilization: 4.8, 8.7, and 10.5 Mscf/STB (stock-tank barrel). This work employs a novel approach that incorporates nonlinear functions to quantify uncertainty in the estimated values as a function of HCPV and to describe the shape of the CO2 retention or incremental oil recovery response with a handful of parameters, providing insight into the behavior of the reservoir across the entire timeline of the CO2 flood. These nonlinear curve fits are focused on statistical inference – i.e., what is the likely outcome and uncertainty ranges for CO2 retention, incremental oil recovery, and net CO2 utilization given the historical data from the 31 CO2 EOR sites? However, the approach described in this work also provides useful information for prediction – i.e., given a set of inputs from another site with similar geology, what are plausible ranges in outcomes for each of these factors? Consequently, the results of this work can be used to estimate the potential range of expected performance for similar candidate oil fields that are not currently undergoing CO2 injection, including estimates of the associated CO2 storage potential of these candidate fields. The results of this work allow estimation of CO2 storage capacity in CO2-EOR operations with various degrees of confidence. The sites in the dataset reflect water – alternating gas CO2 floods – all within the continental United States and heavily dominated by the West Texas carbonate floods. Other floods outside of this region, where the data were available, are also included in this study (i.e., the Rocky Mountain region and the State of Oklahoma).