Application of the automated optimization framework at the BEST project field site. Top figures show the locations of one injection well, one extraction/observation well and one passive relief/observation well. Fresh water injected into a brine-bearing sandstone formation. Injection occurs into two isolated sand layers of the formation. The extraction and passive relief wells are used to control the pressure in the region and along the hypothetical fault shown in the top figures. The left bottom shows that the time-dependent optimized extraction rates (blue) that were calculated based on the updated model and the hypothetical ‘true’ data, to keep the pressure buildup below a critical value, in response to a fixed injection rate (red). The bottom right shows the cross-sectional view of salinity changes between the injector and the passive relief well.
Brine Extraction and Storage Test (BEST):
Field Demonstration at Plant Smith Generating Station Assessment of Opportunities for Optimal Reservoir Pressure Control, Plume Management and Produced Water Strategies
The overall objective of the EPRI-led BEST project to be performed at the Lansing Smith electric generating station near Panama City, Florida, is to help develop cost-effective pressure control, plume management and produced water strategies that can be used to improve reservoir storage efficiency and capacity, and demonstrate safe, reliable containment of CO2 in deep geologic formations with CO2 permanence of 99% or better. In addition, operational experience gained from implementing the field demonstration at a power plant site will provide realistic and practical learnings that can be incorporated into future updates of the United States Department of Energy (DOE) best practice manuals.
In collaboration with EPRI, ARI, and Southern Company, LBNL will evaluate cost effective CO2 injection, brine production and brine treatment strategies for the control of subsurface pressures and plume movement.
Preliminary simulation and optimization studies were conducted in Phase I of the project to: (1) design optimized water production strategies that meet the pressure management goals while minimizing the volume of extracted brine; (2) query the reservoir response to the proposed injection and extraction schemes; and (3) to demonstrate that detectable monitoring signals are generated for measuring pressure buildup, differential pressure plume evolution, and injected fluid plume migration.
Currently in Phase II, the project involves a field demonstration to implement and test the strategies at the host site Plant Smith near Panama City, Florida, operated by Gulf Power Company. No structural risks for leakage such as faults and fracture zones were identified during the screening effort at Plant Smith. For the purpose of demonstration, the pressure management scenarios included a purely hypothetical and critically stressed fault near the planned injection area. The Phase II field demonstration site includes a recently drilled injection well (TIW-INJ), a monitoring and extraction well (TIW-5), and an existing injection well (TIW-1) that is repurposed as monitoring and passive relief well. The project plans to inject low-salinity water and extracted brine through TIW-INJ into isolated sandstone layers of the Lower Tuscaloosa Massive sand at about 1500 m depth by testing and applying “active” extraction and “passive” pressure relief strategies for pressure control. Injection is planned to start by mid-2021.
This field demonstration project will use advanced simulation and optimization tools to implement an adaptive pressure management scheme for the control of subsurface pressures produced by injected waste water that is tailored to the management goals and geologic conditions at the field site. During the field demonstration, the monitoring program has the following principal objectives: (1) To track the position of the pressure front and low-salinity plume created by injected wastewater with sufficient spatial and temporal resolution such that adaptive pressure management strategies can be demonstrated; and (2) to validate predictions of pressure, fluid movement, and differential pressure plumes in the reservoir. This will be accomplished using monitoring methods over a range of spatial scales and at a number of time steps. Continuous and time-lapse borehole measurements of fluid pressure, flow rate, temperature, and/or conductivity will be used to provide high-resolution, ground-truth measurements at the three project wells. In addition to borehole-based measurements, the monitoring plan relies heavily on time-lapse geophysical methods, which provide broad spatial coverage and are sensitive to subsurface pressure and fluid salinity changes.
LBNLs PROJECT TEAM MEMBERS
3D view of the differential pressure distribution (in MPa) (top) and the injected fluid plume in terms of salinity (× 106 parts per million, ppm) (bottom) in the Lower Tuscaloosa injection layer, at the end of the selected 18-month pressure management base case scenario. Both graphs show the location of a hypothetical fault as a scenario for pressure management. Bottom graph shows a cross-sectional view of the layer-cake permeability values in the back. The injected fluid reaches the passive relief well (TIW-1) at about 12 months and then starts migrating down into the deeper Lower Tuscaloosa/Lower Cretaceous layers. Permeability values are given in m2 (~ 1012 Darcy).
Inversion of the crosswell vertical-electric field component for the BEST site’s plume model after 180 days and 365 days of freshwater injection. (a) Actual plume at 180 days. (b) Crosswell data inversion for plume at 180 days. (c) Actual plume at 365 days. (d) Crosswell data inversion for plume at 365 days. Both inversions employ variable lower parameter bounds derived from geometrical plume approximations.