Second EAGE Sustainable Earth Sciences (SES) Conference and Exhibition

In this paper, we demonstrate the importance of predicting the effects of fracture networks on flow, using a case study from the In Salah CO2 storage site in Algeria. We show how fracture permeability is closely controlled by the stress regime determining the conductive fracture network, and that the anisotropy of the conductive network is reflected in surfaces deformation imaged by InSAR. Our results demonstrate that fracture network prediction combined with present day stress analysis can be used to successfully predict CO2 movement in the sub-surface.

Inducing hydraulic fractures is vital in developing enhanced geothermal systems. To the same extent, simulation of such fluid-driven fractures is a complicated process to model even for simple geometries. This difficulty is subjected to the moving boundary conditions due to propagation of crack, non-linear governing equations of fluid flow within the fractures and the high gradient of mechanical deformation near the fracture tip. This paper presents an XFEM framework developed for numerical modeling of hydraulically induced fractures in a porous media interacting with the flow of fluid. This model takes the fully coupled hydro-mechanical processes into account and also, uses the concept of cohesive crack which is more suitable to describe the nonlinear behavior of the quasi-brittle material in the fracture process zone.

Many uncertainties exist in geochemical modeling. Mineral reactive surface area is one of the uncertain parameters. QEMSCAN analyses are performed on sandstone samples from a Dutch CO2 natural analogue to determine reactive surface areas. Geochemical modeling is performed using QEMSCAN surface areas and surface areas which are conventionally used in modeling. The model predicts that the QEMSCAN reactive surface areas result in higher reaction rates and faster equilibration of the sandstones with CO2. The fact that reactions are predicted to occur in a sandstone which has been in contact with CO2 for geological times suggests that the reservoir mineralogy is not yet in equilibrium with CO2. This would indicate that, even if mineral reactive surface areas are determined in detail, geochemical models strongly overestimate reaction kinetics. Other uncertain parameters which can affect kinetics, like mineral nucleation and ion diffusion, should be evaluated in order to be able to better predict long-term CO2-mineral reactions and storage integrity.

The goal of this work is to simulate CO2 storage in a deep saline aquifer case that is currently considered for field injection. Results of RTM are obtained using both TOUGHREACT and the in-house Shell reservoir simulator MoReS, which was recently coupled with the geochemical software PHREEQC. The obtained results are very similar when the same geochemical database and input parameters are used. Small changes in porosity, up to a maximum of 0.003, are computed as a result of dissolution and precipitation reactions. Benchmarking of simulation software is important for quality control and confidence in obtained results.

Renewables suffer from fluctuating energy generation, which means that they generate energy overspills during beneficial weather conditions which, due to lack of large scale storage options, cannot be used at a later point but is wasted. A viable option would be to convert the energy into hydrogen and to store it in existing porous subsurface formations. To assess the influence of hydrogen rock interaction during underground storage of natural gas hydrogen mixtures in porous subsurface formations, a geochemical gibbs free energy simulation was done. The simulation was conducted for the mineral assembly of two existing porous subsurface formations. The model showed that hydrogen increases the pH, and therefore changes the mineral composition of the reservoir rock by dissolving dolomite and precipitating calcite and talc. Additional investigations included crystalline mine rals which were not affected and clay minerals which could not be assessed as reliable data for these minerals is not yet available. Finally the stability of sulphides was assessed to rule out the possible generation of H2S. It was found that under subsurface conditions, sulphides will stay stable

This paper presents the outline of the CO2-DISSOLVED project whose objective is to assess the technical-economic feasibility of a novel CCS concept integrating geothermal energy recovery, aqueous dissolution of CO2 and injection via a doublet system, and an innovative post-combustion CO2 capture technology. Compared to the use of a supercritical phase, this approach offers substantial benefits in terms of storage safety, due to lower brine displacement risks, lower CO2 escape risks, and the potential for more rapid mineralization. However, the solubility of CO2 in brine will be a limiting factor to the amount of CO2 that can be injected. Consequently, and as another contributing novel factor, this proposal targets low to medium range CO2 emitters (ca. 10-100 kt/yr), that could be compatible with a single doublet installation. Since it is intended to be a local solution, the costs related to CO2 transport would then be dramatically reduced, provided that the local underground geology is favorable. Finally, this project adds the potential for energy and/or revenue generation through geothermal heat recovery. This constitutes an interesting way of valorization of the injection operations, demonstrating that an actual synergy between CO2 storage and geothermal activities may exist.

Use of the subsurface has become steadily more intensive and globalized. This is due to wider application of conventional uses and to the emergence of relatively new, unconventional uses. This raises the question of whether we are managing the subsurface space in a sustainable way that also enables us to understand and mitigate the risks that may emerge. Depending on the purpose of subsurface space use, typical risk challenges may include e.g. impacts to groundwater, oil and gas resources. Expanding uses of the subsurface motivates additional effort to mitigate well-known and relatively new risks. Estimating risk levels in complex systems can be a daunting task if the strategy is to construct simulation models of all known physical processes combined with uncertainties in the system, which for subsurface projects, often dominate the system description. An alternative approach described here isolates the main risk drivers in a high-level probabilistic format known as a Bayesian (Belief) Network (BN). The BN approach accommodates more general relationships between uncertain variables than event or fault trees and allows expression of probabilities to consistently influence the top-level risk indicators. A BN risk model will typically be more compact and legible than its fault/event tree equivalent.

The CO2 storage atlas of the Norwegian part of the North Sea and Norwegian Sea has been prepared by the Norwegian Petroleum Directorate at the request of the Ministry of Petroleum and Energy. The main objectives have been to identify the safe and effective areas for long-term storage of CO2 and to avoid possible negative interference with ongoing and future petroleum activity. 27 geological formations have been individually assessed, and grouped into saline aquifers. The assessed aquifers have been ranked according to guidelines which have been developed for this study. The evaluation of geological volumes suitable for injecting and storing CO2 can be viewed as a step-wise approximation as shown in the maturation pyramid. A modeling study will be presented to understand the timing and extent of long distance CO2 migration

Successful CO2 storage in deep saline aquifers relies on economic efficiency, sufficient capacity and long-term security of the storage formation. Unfortunately, these three criteria of CO2 storage are generally in conflict, and often difficult to guarantee when there is a lack of geological characteristics of the storage site. We overcome these challenges by developing: 1) multiwell CO2 injection strategies using a multi-criterion optimization to handle conflicting objectives; 2) CO2 injection management that is robust against model uncertainty. PUNQ-S3 model was modified as a leaky storage to study injection strategies associated with the risks of CO2 leakage under geological uncertainty. Based on our numerical results, the NSGA-II with the ASGI technique can effectively obtain a set of efficient-frontier injection strategies. For the uncertainty assessment, the impact of the model uncertainty to the outcomes is significant. Therefore, our findings suggest using the mixture distribution of the objective-function values, as opposed to the traditional Gaussian distribution to cover model uncertainty

A rock physics diagnostics study is performed for the Tubåen formation in the Snøhvit field with the aim of characterizing the transport properties for the different sand intervals. This is combined with co2 injection laboratory experiments for selected plugs from the sand intervals to understand the impact the CO2 injection have on the microstructure of the rock framework and its physical properties. The rock physics diagnostic indicates that parts of the Tubåen formation are dominated by stiff load bearing contact cement, other parts by pore-filling grains that can be related to poor sorting, and that don’t contribute to the stiffness of the rock. Some of these fine particles may become dislodged due to CO2 injection, which permanently changes the load bearing framework of the formation. The laboratory experiments and SEM images indicate that the CO2 injection cause a re-arrangement of the rock framework and of the movable fines that change the rock framework. The importance of fines migration is likely to depend on the initial microstructure of the rock. The results of the ongoing laboratory experiments will help to better understand the effect on the different sand intervals of the Tubåen formation.

In applications such as geothermal energy, underground storage, mining, hydraulic fracturing, it becomes current practice to implement local seismic networks to monitor induced seismicity and to help mitigating the associated risks. In such contexts, it is crucial to guarantee that the network is able to detect a seismic event of predefined magnitude in a specific area. We propose a method to estimate this detection capability for existing kilometric-scale seismic networks which did not record any seismicity in the target zone yet. However, the network should be running for a time period long enough to record several local events listed in a reference catalogue. These earthquakes are used to calibrate an amplitude-magnitude relationship, knowing that the amplitudes are at the basis of the detection of seismic event candidates. This observation-based approach can take into account uncertainties in the magnitude estimate. The procedure was applied on the seismic network deployed over Bruchsal (Germany) geothermal field. Since mid-2010, no seismicity in the reservoir has been recorded by the network despite its good working order. The proposed technique suggests that there is 95% probability that no seismic event with ML ≥ 0.7 occurred below the network down to the reservoir depth at 2400 m.

The paper focuses on the petrophysical evaluation of wells for the development of a geothermal district heating plant in Tønder,Denmark. The evaluation was carried out by analysis and interpretation of available data. The data used included completion reports, well logs, conventional core and side wall core data. The results from interpretation of the data include total porosity, permeability, clay volume and effective porosity for three wells. An unusual finding was the inverse relationship between porosity and permeability in the targeted layer of the reservoir. This inverse relationship, maybe attributed to the precipitation of salt and the presence of caverns or fractures.

A 100 sqkm seiscmic survey was acquired to search for faults in the crystalline basement of the Westerzgebirge (Saxony). The aim is to explore the crystalline basement for deep geothermal reservoirs over 5 km in depth . The subsurface is formed by large granitic intrusions which are partly covered by metaphorphc piles and the area is located in a main fault zone. The seismic image show a detailed structure of the granitic body and a complex fault system, which in this type was not predicted by the prelimenary study. The aim of the project, which is funded by the German Ministry of Environment, Nature Conservation and Nuclear Safety, is to adapt the seismic method for the exploration of deep geothermal reservoirs in the crystalline basement.

Renewable energy resources are intermittent and need buffer storage to the bridge time-gap between production and demand peaks. North German Basin has a very large capacity for CAES in saline aquifers and cavities. Replacement of brine by gas in aquifer cause strong changes in electrical resistivity. Previously we have shown the applicability of 2D electric resistivity tomography (ERT) for monitoring sequestered CO2 in these reservoirs. Towards more real world problems, this paper aims at studying resolution of 3D ERT in boreholes for CAES mapping in deep saline aquifers. We used for this 3D surveys 3-9 vertical arrays of borehole electrodes installed surrounding the targets (caprock, aquifer, CAES plume, aquitard). We applied non classical electrode configurations of pseudo (optimized) and real 3D measurements in different boreholes. The results show that the applied 3D technique can generally map the dome structure targets with varying resolution, smearing and artefacts. However, resistivity amplitudes are less correctly recovered. The resolution increases with increasing number of borehole electrode arrays, increasing of layer thicknesses and the a priori fixing of boundaries in the inversion. Optimized 2D arrays have better resolution than the other arrays and thus needs to be developed to real 3D arrays.

We analyse microseismic data from a pilot installation in the commercial sized CO2 injection project at Krechba, Algeria. Over 5000 microseismic events were detected from August 2009 to June 2011 using a master event waveform cross-correlation method. Most events can be directly related to the injection at well KB502. The period with the highest microseismic activity correlates with a period where the maximum applied pressure has most probably exceeded the minimum principal stress in the injection interval. Using S-P wave traveltime differences and event azimuth we can separate various distinct event clusters. Although an accurate event location is not possible due to the limited receiver network, the events most probably come from the injection reservoir interval. Moment magnitudes of the events have been determined and range between MW -1 and MW 0 with a few larger magnitude events up to MW 1. Differences in b-value between event clusters may give some insight about different types of fracturing and may be related to the in-situ stress regime.

Issues related to Carbon Capture and Storage (CCS) includes technical, economic, and public acceptance challenges. Only though field scale demonstration projects can industry address these challenges in preparation for commercial deployment. The capture component takes place at the James M. Barry Electric Generating Plant in Bucks, Alabama utilizing capture technology licensed by Mitsubishi Heavy Industries America. CO2 captured at the plant is being transported by pipeline for underground storage in a deep, saline geologic formation within the Citronelle Dome located in Citronelle, Alabama. A 12-mile pipeline was constructed to transport CO2 to the injection site located on the flank of Citronelle Dome. Operations started in 2012, up to 550 metric tons of CO2 per day, the equivalent emissions from 25 MW of the plant’s capacity are being captured. To date over 160 metric tons of CO2 have been captured and over 65,000 metric tons have been injected for storage. The injection target is the lower Cretaceous Paluxy Formation (a sandstone saline reservoir) which occurs at a depth of 2,865 meters (9,400 feet). Transportation and injection operations will continue for one to two years. Subsurface monitoring will be deployed through 2017 to track plume movement and monitor for leakage.

To optimize the design and to model the working cycle of an underground thermal energy storage facility (UTES) a geothermal reservoir simulator was created on the basis of the Complex Systems Modelling Platform C++ (CSMP++) software library (Matthhäi, et al., 2001) following two different approaches: a Boussinesq approximation and a full Pressure-Temperature-Enthalpy scheme (based on Coumou et.al. 2009). A series of benchmark tests were performed to assess efficiency and accuracy of both schemes, given the standard operating temperature ranges. The simulator is then used to determine optimal well locations and operating schedules for a UTES construction project.

We performed a series of tests for passive seismic tomography with data from Paralana, a new Enhanced Geothermal System (EGS) located in South Australia. An injection well was drilled at Paralana in 2009 into a sedimentary basin down to a high heat-producing basement at ~4000 m depth. The first main hydraulic stimulation of the well took place in July 2011 in order to create/enhance a geothermal reservoir. Induced seismicity was monitored by a network of 20 stations (from surface to 1800 m depth) and more than 7000 microearthquakes were detected during the five days of injection. The synthetic tests indicate that small velocity heterogeneities into the reservoir can be resolved at Paralana if high-precision relative arrival times are used, such as provided by waveform-cross correlation. Preliminary results indicate a low P-wave velocity anomaly at the base of the well.

Mitigations of anthropogenic GHG demand developments of renewable energy resources. These energy resources are intermittent and need buffer storage to bridge the time-gap between production and demand peaks. North German Basin has a very large capacity for CAES in porous saltwater aquifers and salt cavities (natural and artificial). Replacement of brine by compressed gas in saline formations cause strong changes in electrical resistivity and density, and therefore justify the application of geoelectrics, electromagnetics and gravity etc. In this study we study the applicability of these geophysical techniques in mapping CAES reservoirs (pore and cavern) in the underground of NW Germany. Our constrained techniques of electric resistivity tomography in boreholes are able to highly resolve the unusual problem of super resistive air caverns within the extreme resistive salt rocks. For gravity techniques, we could determine the lowest detectable mass deficit and offset of double caverns (2-fold their depth) in the underground resulting from pore and salt cavern CAES at some study sites.

Keynote Speach: Rien Herber

During the three last decades, geomechanics has been, in a large part, driven by the evolution in the domain of energy. After its contribution in the conventional oil and gas resources and in the nuclear waste repository, new domains such unconventional geothermal energy and CO2 underground storage have stepped in since around 10 years and more recently, unconventional gas and oil resources continue to contribute to the evolution of geomechanics. In this paper, we have focused on the Enhanced Geothermal Systems and on the CO2 underground storage, highlighting the new challenges for each of these two domains and the common ones: geomechanical role of the faults, stress fields and behavior of the near well.

Reservoir monitoring improves our understanding of reservoir behavior and helps achieve more effective reservoir management and prediction of future performance with obvious economic benefits. Volumetric changes in reservoirs due to fluid extraction and injection can induce either subsidence or uplift which can trigger fault reactivation and threaten well integrity (as well as reservoir structural integrity). Depending on the depth of the reservoir and the characteristic of the cap rock, deformation may also be detectable at the surface. Surface deformation monitoring can provide valuable constraints on the dynamic behavior of a reservoir enabling the evaluation of volumetric changes in the reservoir through time, allowing the calibration of geo-mechanical models. Whatever the surveying technique, the detection of millimeter level surface deformation is required to monitor small surface displacement rates, which could impact risk evaluation and environmental impact assessment in oil&gas operations, as well as in geothermal plants.

The P18-4 compartment (operated by TAQA Energy B.V.) would be injected by CO2 at a minimum temperature of 12 degrees C. At these temperatures, the CO2 phase will either be a gas or a liquid. As the initial temperature of the reservoir is 120 oC, the CO2 will eventually be the gaseous or (at higher pressures), the supercritical phase. An outstanding issue was the simulation of the fate and transport of the injected liquid CO2, with emphasis on the temperature effects. In this presentation, we will present results of our latest research efforts

The project CauCasCCS is concerned with fundamental research on geological CO2 capture and storage. The study aims to investigate the possible alteration of rocks exposed to natural CO2-spring water. Such rocks represent natural analogues for reservoir cap rocks and storage.

The target siliciclastic aquifer investigated by the Longyearbyen CO2 Lab as a possible test-scale CO2 storage unit is a dual-permeability reservoir characterized by fractured, tight lithologies. By integrating borehole and outcrop data, the reservoir section has been subdivided in intervals defined by 5 litho-structural units (LSUs), each one characterized by different lithologies and fracture sets interpreted to represent pseudo-geomechanical units. Due to their contrasting features, these LSUs are believed to have a crucial influence on subsurface fluid migration. Our results indicate that fractured shale intervals control lateral fluid flow (predominance of low-angle fracture) whereas sandy and coarser intervals seem to control vertical fluid flow (predominance of high-angle fractures), locally enhancing the contribution of the matrix porosity. Horizontal and vertical high permeability conduits can be found at the LSUs’ interfaces, along the chilled margins of igneous sills and dykes, and along the damage zone of mesoscopic faults, due to the localized enhanced fracturing (fracture corridors). A large database containing structural data on fractures has been acquired and analyzed in order to extrapolate calibrated parameters for numerical modeling and flow simulations. These in turn allow reservoir volumetric calculations, assessment of seal integrity and forecasting of vertical/lateral connectivity of the reservoir.