EAGE GET 2022

Depleted and saline reservoir types are the main candidates to consider for any CCS-scenario. Both kinds of reservoir have fundamentally different risk profiles and degrees of uncertainty and each of them has a role to play in successful CO2 sequestration scenarios.

Depleted reservoirs provide a relatively rapid path to storage with significantly less uncertainty on containment and injectivity efficiency. However, the large number of well penetrations that reduce the performance uncertainty also increase the risk of potential leakage points.

In contrast, deep saline reservoirs will take longer to bring to project maturity but have an order of magnitude more storage potential. Additionally, they will almost certainly support super-critical injection in dense phase making the project design much simpler to manage and monitor.

Successful long-term CO2 storage projects are likely to employ a combination of both types of reservoirs through a phased life-cycle where the use of a depleted reservoir enables early injection whilst an adjacent saline reservoir is brought “on stream” later on.

The implementation of successful CO2 storage projects requires capabilities with strong adjacency to the oil and gas sector, supplemented with those specific to CO2 storage, requiring flexible and adaptable approaches to the upskilling of key staff.

Decarbonization is a real target these days that will would require, initially, carbon storage projects for further achievements. Even though there are several Carbon Capture & Storage (CCS) standards, guidelines or other recommendations on the published literature, it is paramount to collect and discuss how operator are progressing (visualization phase) on these projects with real milestones and decision process gateways. This project´s reality is most of the time hidden. This paper focus on a practical and fit for purpose integrated CCS visualization workflow through agile methodologies. Traditionally, each discipline (G&G, geomechanics, etc…) used to work on a separate workflow which results may be assembled later on. Presented workflow helped to reduce expected results uncertainties by incremental iterations (“sprints”), that progressively corrected areas of improvements (adaptative fit for purpose), reduced uncertainties (efficiency) and focus deliverables (visualization modelling). This workflow execution on real data sets provided several paramount lessons learned for future projects. These insights proved that data can be converted to useful knowledge (CCS candidate technical feasibility visualization).

Storage capacity in European countries is computed from an extensive database gathered in Hystories Project. This database was built by expanding existing databases such as CO2STOP/ESTMAP.

The 1082 traps of the Hystories database were modeled using a synthetic approach. The synthetic approach considers simplified geometrical parameters such as area, thickness, porosity and depth. For each trap, a simplified petrophysical properties distributions for permeability and the porosity of the storage formations were used to assess the porous volume. A database of 747 synthetic geological models was established where sufficient publicly data is available. The proposed volumetric approach extends the approach developed for CO2 storage to Hydrogen storage. It uses the hydrogen, brine and hydrocarbon properties computed at the storage conditions in terms of densities and viscosities. The approach also estimated the uncertainties of the different parameters from the database.

The approach was validated against published dynamic capacities for hydrogen storage and is extended to all the traps in the Hystories database.

For most European countries, the storage capacity provided by the porous media covers the expected demand as established within Hystories project. The different type of storages may require different lead time to qualify for hydrogen storage.

Monitoring an underground gas-storage is complex and challenging. Figuratively speaking, it is a jigsaw, in which not all pieces are available, and not completely clear which motif will be created. At the same time, it is necessary to explain the motif. This process is already challenging for experts, but even poses great problems for a large number of local stakeholders to understand the process. The research cooperation Epe aims to make the ground movement at the cavern storage Epe (NW-Germany) explainable and transparent. For this purpose, the research cooperation was founded by the city of Gronau, the citizens’ initiative Kavernenfeld Epe e.V., EFTAS GmbH, Münster and the Research Center of Post-Mining of the Technical University Georg Agricola, Bochum. The operators in the cavern field also support with field data and knowledge. Radar remote sensing (InSAR) is used to analyse the ground movement, which might be caused by the cavern operation. The InSAR result will be merged with the available public geodata and mine surveying data. A public participation process will accompany all work. For the first time, the cooperation leads to a direct exchange of various participants in a mining project for establishing a common understanding of the process.

In order to reduce greenhouse gas emissions, decarbonized electricity should become an increasingly important part of the world’s energy mix. However, a major issue in covering the energy demand with electricity is its storage, to even out grid load. One solution is to store electricity in the form of hydrogen; this can be performed on land, in dedicated reinforced containers, or in hewn out lined caverns (usually in salt formations). Another option which is studied in this paper, is to store hydrogen offshore, in depleted oil and gas reservoirs. In this work, the mechanical risks of fatigue, failure and fracture in and around a storage reservoir from the North Sea, subjected to hydrogen injection and depletion cycles, are investigated, through finite element analysis. The work focuses on the mechanical integrity aspects, neglecting longer-term geochemical interactions. Our study identifies that for sufficiently large reservoir pressure amplitude cycles, tensile failure risks exist during the injection phase and shear failure risks during the subsequent injection and depletion phases, at specific locations where stress concentration occurs. We demonstrate that the amplitude of the loading cycles has a large influence on the evolution of the stress state around the reservoir and on plastic deformation accumulation.

The Central Cordillera of Colombia volcano-tectonic province has suitable geological conditions for geothermal renewable and sustainable energy sources. The use of 3D geologic modeling and simulations are an effective tool for better understanding the subsurface geology and have a better estimation of potential resource assessment. This work aims to characterize and simulate the geothermal and hydrothermal processes in a volcano-hosted system, using a 3D crustal-basin modelling software, to estimate the in-place geothermal energy potential in a high-enthalpy deep geothermal system, establishing spatial relationships between heat sources, flow patterns, lithologies and fault networks. The results confirm the large convective-conductive geothermal-hydrothermal system activity in the metamorphic Cajamarca complex and the andesitic-dacitic lavas within the upflow zone, and the hydrothermal flow pattern in the outflow zone structurally controlled by the fault network. Furthermore, the simulated heat energy in-place results served as input data for a reservoir pre-feasibility energy recovery simulation with closed-loop systems, to investigate the ultimate recoverable heat energy potential exploitable as an EGS.

Conventional time-lapse seismic data analysis for CO2 monitoring goes through a full processing workflow, which includes complex procedures such as velocity model building, seismic imaging, and full waveform inversion. These procedures are very time consuming, often taking months to years depending on the survey area, requiring extensive manual interactions from domain specialists which can introduce human bias into the results. To bypass these computationally expensive and manually intensive processes, we introduce a novel deep learning algorithm for the subsurface property estimation directly from pre-stack monitoring seismic data. With this algorithm, the nonlinear mapping between the depth domain property contrast and the time domain seismic response is established by a deep learning neural network trained on a set of processed baseline and the corresponding monitoring dataset. The trained network then predicts the subsurface property changes caused by the injected supercritical CO2 plumes directly from any new monitoring dataset without conventional velocity model building and imaging procedures. This neural network features a novel multi-branch design with enhanced feature extraction and a customed binary cross-entropy loss function handling the imbalanced training labels. It is applicable to surface seismic data, vertical seismic profile, cross-well data, or other types of geophysical data.

The estimation of CO2 storage resources is a key tasks to understand the viability of a CCS project. There are different methodologies to calculate the CO2 resources. All these methods apply some storage coefficients. These coefficients are difficult to estimate and generates high uncertainty becoming critical parameters. This paper will focus in understand the impact of these critical parameters in the CO2 storage resources estimation using the U.S. Geological Survey methodology. Three critical parameters have been selected to perform a sensitivity analysis: the buoyant trapping storage efficiencies, the residual trapping storage efficiencies and the CO2 density. Using a simulation digital tool based on Monte Carlo algorithm, the P90 and P10 values were analyzed for each critical parameter and compare with the mean CO2 storage volume of a base case generated from a static model. Depending on which coefficient values have been chosen, the deviation of the storage resources could be up to 33% having a potential impact in the CCS project. The use of simulation digital tools to analyze these critical coefficients become necessary, being able to generate thousands of realizations allowing to decrease the uncertainty of these risky parameters and having more control in the storage resources estimation.

A straightforward numerical modeling approach is developed aiming to facilitate the CO2 geochemical impact evaluation for CCS in depleted gas reservoirs. This method is applied to a sandstone reservoir for which extensive experimental analyses on core samples are made to determine the formation bulk mineralogy, bulk chemical composition, and stoichiometry of site-specific minerals.

First, 0D simulations are used to fine-tune formation water composition and define the system’s geochemical stability without CO2. Second, CO2 is added into a static batch cell to identify the prevailing CO2-brine-rock geochemical reactions. In the last stage, a 2D radial reactive transport model (RTM) is set up in accordance with the project’s requirements. Based on the RTM results, evaluations regarding the mineral dissolution/precipitation paths, porosity and permeability changes, dry-out effects and salt precipitation, impact on CO2 injectivity and quantitative evaluations on CO2 permanent trapping are performed.

The results show the occurrence of carbonate dissolution in the entire area contacted by the CO2 and halite precipitation within the near wellbore area during the injection phase. The associated petrophysical changes within the reservoir are nevertheless minor and are not considered threatening to the injectivity or fluid migration. However, for containment evaluation, a separate caprock geochemical study is required.