AAPG Annual Convention and Exhibition

Datapages, Inc.Print this page

Constructing a Geologic Model to Simulate and Optimize the Commercial Scale Injection and Storage of CO2 at Citronelle Field, Mobile County, Alabama


At Citronelle Field, the oil-producing Lower Cretaceous Donovan sandstones (Rodessa Fm.) are overlain by approximately 3,658 vertical meters of water-bearing, predominately clastic sediments. Numerous potential CO2 storage horizons are present, as well as thick, regional-scale, potential confining units. Four-way structural closure from Citronelle Dome is present at all reservoir horizons, making Citronelle Field an attractive target for potential injection and storage of commercial scale volumes of anthropogenic CO2. The project objective is to model the complete earth volume from surface to the base of the oil-producing sands, for an area of 145 km2 centered on Citronelle Field. The purpose of the geologic model is three-fold: 1) to adequately characterize the relevant reservoir heterogeneity controlling CO2 injectivity, flow and storage capacity; 2) to quantify each storage horizon in terms of CO2 volume stored and areal expanse; 3) to optimize potential commercial-scale CO2 injection and storage within the many possible storage horizons, using reservoir simulation to track and manage the CO2 plume and pressure field. Citronelle Field contains 400+ wells on 40-ac spacing. Most wells have only 1960's vintage SP-resistivity logs and no porosity logs. Three new wells were drilled by the USDOE-NETL sponsored Southeast Carbon Sequestration Partnership anthropogenic CO2 injection test on the pads of existing abandoned wells. The old-new well pairs offer a unique opportunity to apply a neural network approach to predict porosity from legacy resistivity logs, using the porosity data from these new wells to train the neural network. A geostatistics approach extrapolates porosity throughout the study area from a subset of individual well points, and is used to generate ‘low’ and ‘high’ areal heterogeneity cases. Permeability is predicted from porosity based on available core data. Vertical heterogeneity is represented by approximately 375 reservoir layers for the 3658 m sedimentary column. Implementation of the massive model in the GEM-GHG reservoir simulator and 3-D visualization of the model is in progress. Planned use of the geologic model and reservoir simulation is to explore strategies of well siting and design, CO2 injection, and pressure management to maximize CO2 storage capacity, while minimizing the operations footprint of a large-scale commercial CO2 storage site.