High resolution monitoring, real time visualization and reliable modeling of highly controlled, intermediate and up-scalable size pilot injection tests of underground storage of CO2

Final Report Summary - TRUST (High resolution monitoring, real time visualization and reliable modeling of highly controlled, intermediate and up-scalable size pilot injection tests of underground storage of CO2)

Executive Summary:
The key objective of the TRUST project aimed at conducting and analyzing highly controlled CO2 injection experiments at the Heletz site in Israel. This site is located at the edges of a depleted oil reservoir (the Heletz sands) at a depth of 1610-1630 m. Above the reservoir there is a 40 meter thicj caprock. Static pressure is of 143 bar and the reservoir temperature is of 64 C. This site was developed with two deep wells (1640 m), H18A and H18B. H18A was instrumented for CO2 injection, water injection, water abstraction and monitoring and H18B was instrumented for water abstraction and monitoring. Monitoring technologies included: shall and deep tracer injection (only at H18A); downhole pressure and temperature monitoring, downhole fluid sampling and Distributed temperature sensing (DTS). Above the ground, the site is dotted with a versatile CO2 injection system and water injection system. Around the site we laid out a seismic monitoring system. The site was extensively characterized for its geology and hydraulic properties, prior to the undertaking of CO2 injection experiments. Three CO2 injection experiments were carried out: 1) two-push pull experiments aimed at determining field values of two key trapping processes (dissolution trapping and residual trapping); 2) detecting the injection of CO2 in the reservoir via seismic monitoring. Extensive modeling took place both for the design and planning and for the interpretation of the experiment data. We were able to predict the pressure build-up subsequent to the CO2 injection as well as determine the field scale dissolution and residual saturation. The site licensing and injection permit required a comprehensive environmental impact statement and a number of hazard and risk evaluation (HAZOP) statements. The experiments were conducted without any major HSE (Health – Safety and Environment) incident. We also conducted a number of investigations on large scale simulations of CO2 storage as well as on how optimize the CO2 storage, considering a number of schemes (pure CO2 phase, alternating CO2 and water injection and injection of trapped CO2). We also investigated hydro-mechanical effects, such as the impact col CO2 injection. New monitoring approaches and technologies were developed: 1) pressure tomography as a surrogate of seismic monitoring; 2) a new generation of tracers (Kinetic Interface Sensitive) capable of providing information on the mixing of the stored CO2 in the formation. We addressed the legal aspects of the storage as well as formulated a template for site licensing. In TRUST we undertook capacity building activities by means of three advanced courses (in 2013, 2015 and 2017). Societal aspects were addressed by creating a dialogue with a key NGO (Bellona), operatives from SKB in Sweden and other experts. We gathered information from other sites (Hontomin though in a limited extent, The Baltic storage projects and one EOR project in Brazil). We developed new monitoring technologies driven by optic fiber (FBG temperature sensors), which can now be extended to conduct pressure, strain and seismic measurements in a way that seems to be far more robust and resilient than the currently existing technologies. Analytical solutions estimating the fate of CO2 in simple geometries were developed and could be used for model verification. We extended existing software so it can now simulate the coupling between the injection well and the reservoir, including phase changes in the CO2. We provide an evaluation of Monitoring – Measurement and Validation technologies (MMV), based on our experience at Heletz and suggest to switch to optic fiber driven monitoring and measurement for resilience, robustness and upgradability. We also developed sealing solutions for leakage mitigation and tested them at laboratory scale and in shallow well. We developed a software platform for the online and real-time visualization of monitoring data in any device with browsing capabilities. We also conducted a cross-model validation, using a number of key simulation models over bench-marks tests of various complexities and found out that the models provide comparable results.

Not all of the stated objective could be achieved, mainly due to the fact EWRE (bearing the budget for the experiment) was denied the overhead for equipment and consumables as agreed in the grant agreement one year after the beginning of the project, there were numerous failures in the downhole fluid sampling system, which required resources shifting to cover the costs of repair and a cost of the CO2 that was far higher than expected due to the regional geo-political conditions. However, despite these limitations and in a wider retrospective of TRUST, the Heletz site is so far the only site funded by EU in which CO2 is stored in a way that is consistent with geological storage (deep conductive formation sealed by a caprock, supercritical CO2).

Project Context and Objectives:
The disposal or storage of greenhouse gases emissions (essentially CO2) is viewed by many as a critical mean to stabilize or revert the trend of global warming and climate change, until the world stops relying on the production of energy from fossil fuels. The principle is relatively simple; the emitted CO2 is captured from the sources (essentially, fossil fuel fired power plants, the cement industry and the steel and other metals industry). It is then transported and stored in in a geological layer that is sufficiently deep (in order to allow a high density of the stored CO2) and sufficiently sealed from above. This in order to prevent upward migration and leakage, as the stored CO2 will be lighter than the host formation fluid. Candidate geological layers could be depleted oil and gas reservoirs, abandoned coalmines and saline formations. Saline formations are abundant and well distributed around the world and they present by far the largest storage potential. They also have no value due to their high salinity. Technologies for CO2 capture and transport have been developed and substantial effort has been invested in order to optimize the cost and efficiency of the process. The question of the geological storage remains problematic because deep formations, particularly saline formations, are poorly characterized, with regard to their properties, the quality of the sealing layers and the expected behavior of the stored CO2. This is why pilot demonstrations are critical to bridge the knowledge and understanding gaps. Over the year number of pilot CO2 storage sites have been developed, mostly in the USA but also in Australia, Japan and Canada. In Europe, the EU has funded the development of three sites: 1) Ketzin (Germany); 2) Hontomin (Spain), and 3) Heletz (Israel). Pilot sites are critical to demonstrate safety and manageability of the geological storage of CO2, to monitor the storage process and the CO2 spreading in the reservoir, to detect possible leakage and to create datasets for model validation. Models are essential tools in the design and planning of CO2 storage. So far, in Europe, industrial scale storage of CO2 takes place in Norway (Sleipner and Snohvit). The key reason for such a small number of pilot sites is the strong opposition of the public of opinion against CO2 storage. The key reason for lack of deployment of industrial CO2 storage is the lack of business model / incentive in addition of the public opposition.

The Heletz site was initially developed in the frame of the EU-FP7 project MUSTANG. In this context, TRUST ambition was to achieve a sizeable injection of CO2 under highly control conditions of measurement, in order to provide a high degree of understanding of the CO2 injection process and of fate of the stored CO2. In simple words, the key objective of TRUST was to develop a site for pilot injection of CO2 (going through all the technical and regulatory phases) and conduct CO2 injection experiments capable of improving our understanding of the behavior of the stored CO2 and how to manage and optimize the CO2 storage.

Accordingly, we structured the project in a number of work-packages (sub-projects) as follows:

WP02: Sites and Field tests

Prepare the Heletz site for the CO2 injection experiments to be conducted in TRUST. This included:

• The design, planning and installation of a versatile CO2 injection facility allowing the testing of CO2 injection in different conditions (super-critical conditions, cold conditions, dissolved within water), including a comprehensive monitoring system (pressure, temperature and massflow rate).
• The instrumentation of two wells: one well was designed for CO2 injection monitoring and one well for monitoring, with an extensive set of monitoring technologies, including downhole pressure and temperature, downhole fluid sampling, optic fiber distributed temperature sensing (DTS), tracers injection. The wells were also instrumented for fluid abstraction in order to allow conducting complex experiments of CO2 injection aimed at providing a deep understanding of the key trapping mechanisms of the stored CO2.
• Gather data and information from other sites: Hontomin (Spain), Reconcavo (Brazil) and Swedish/Baltic sites.

WP03: Monitoring Measurement and Validation (MMV) technologies and real-time reporting and visualization

This work-package was aimed at developing cost-effective and multiple purpose monitoring technologies and implement them at Heletz for the purpose of obtaining high quality data on the storage process. This included:

• Design and field test of a new integrated monitoring device incorporating the following technologies: 1) downhole fluid sampling; 2) Electrical resistivity tomography (ERT); 3) Distributed temperature sensing via optic fiber (DTS); 4) High precision vertical distribution of temperature via FBG (Fiber Bragg Grating); 5) downhole pressure and temperature monitoring. A specific deep well was planned to be drilled.

• Design and implement a new integrated active and passive seismic monitoring system, allowing the detection of relatively small amounts of CO2 in the reservoir.

• Develop and laboratory testing of a new generation of tracers, KIS (Kinetic Interface Sensitive) capable of providing improved information on the area of the water-CO2 interface in the reservoir and thus information on the mixing.

• Determine of permeability changes as a proxy for the extent of the CO2 plume, via pressure tomography.

• Integrate surface and downhole measurements and monitoring results in an internet platform with online visualization tools. This to allow the online and real-time visualization of the monitoring in any device with internet browsing capability.

WP04: Modeling

Modeling is an essential tool both for the diagnostic and prognostic purposes. But in order for the models to acquire reliability they need to be validated. Validation is achieved when a model reproduces within satisfactory limits the behavior of the system (history matching). In TRUST a substantial modeling, effort was planned. It included:

• Model and modeling approach development, evaluation and validation against comprehensive sets of field data collected from the injection experiments to be conducted at Heletz.

• Development of approaches and operational models for long-term and large scale (industrial) predictions of CO2 injection and storage. Highly sophisticated models are computationally intensive. For large scale simulations, the use of simpler yet reliable models could be more effective. Here our objective was to identify and evaluate such kind of simplified models.

• Design simulations of injection experiments, including testing of different injection scenarios. Prior to conducting them in the field, injection experiments need to be carefully design and planned in order to determine the parameters that need to be measured and the measurement frequency.

• Interpretation and model matching of the injection experiments. Once the experiments are completed, model are applied in order to reproduce the measurements. The reliability of a model is determined by its ability to reproduce the measured data within satisfactory error bounds.

• Providing a field data case for model cross-validation and providing a first approach for such cross-validation. Assess the consistency of different model for a given type of problems. This would provide a good understanding of the differences between different modeling platforms.

WP05: Strategies for storage management

Using the knowledge gained in the field and in modeling field results was used to delineate the appropriate ways to inject CO2 in a given reservoir. This included:

• Defining optimal injection strategies and modes of injection in order to maximize the trapping while minimizing the reservoir pressure build-up and minimizing the energy usage and other major costs of the operation.

• Planning and implementing injection strategies, first for the experiments to be carried out within the frame of TRUST, including testing of different modes of injection, evaluating the trade-offs of injecting dissolved CO2, assessing the effectiveness of different injection geometries and envisaging testing injection of CO2 micro-bubbles in brine.

• Extend the analyses, by means of modeling, to other conditions and site characteristics, essential industrial scale storage.

• Analyze results and suggest recommendations on best practices for injection, from the outlet of the supply line to the reservoir and the related pressure management.

WP06: Leakage detection and mitigation

• Develop a new technology for remediation based on reactive fluid injection (self-carbonation solute).

• Determine the optimal fluid formulations and uses according to the local hydro-thermo-chemical properties of the reservoir.

• Conduct a controlled experiment of CO2 leakage in a poorly plugged or unplugged well at Heletz site.

• Monitor the start of the leaking and its development with time.

• Perform the remediation fluid injection.

• Demonstrate that it is an operational technology for the mitigating of leaky wells.

WP07: Risk Assessment: Procedures, Protocols for Certification and Licensing

• Develop and implement a site-specific risk management procedure at the Heletz site, aimed at achieving a comprehensive risk assessment process, controlling and monitoring, communication of information, handling and minimizing residual risks (especially downhole, near the wellbore, and within the caprock) and detecting and handling possible failures of the seal or near-well completion.

• Compare the risk management tools applied for Heletz site with the approaches and findings in Hontomin.

• Use the risk management findings as input for further applications such as the development of guidelines, protocols for site licensing and certification, liability issues.

WP08: Communication, public engagement and liabilities

• Promoting education, training and capacity building in the field of geological storage of CO2.

• Construct in at least one of sites (the Heletz site) facilities for the technical training and demonstration.

• Promoting education and competence building by organizing formal educational courses for the international forum and related to this, develop course material for teaching.

• Promoting education and competence building by affiliating of PhD students, Post-Doctoral scientists and Master students to the R&D work of TRUST.

• Actively disseminating to the scientific/technical community (industries, regulators) via the end-users group (BAP) of the project.

WP10: Extrapolation

Extrapolate the results of the project to the industrial scale application as well as to develop generic approaches for such upscaling.

• Identification of typical settings and characteristics of industrial scale operations.

• Upscaling of the model simulations and the results of the field injection experiments to industrial scale.

• Upscaling of the monitoring strategies and network design for large scale;

• Preparation of protocols for site management;

• Formulation of best practices and recommendations and exploitation of results.

Project Results:
The S/T results were achieved from the work conducted in work-packages WP02, WP03, WP04, WP05, WP06, WP07 and WP10.

WP02: Sites and Field tests

The hydraulic characterization of the Heletz reservoir was completed, including the conduction of fully monitored pumping tests and chemical analysis of the pumped water. The interpretation of the pump provided an assessment of the values of the hydraulic conductivity of the formation (see Figures for WP02).

The wells instrumentation was completed. The injection well (H18A) was instrumented for CO2 injection, water injection, injection of tracers, water abstraction, chemicals and CO2 mixing at depth, downhole pressure and temperature monitoring at two different vertical horizons, distributed temperature sensing via optic fiber (DTS) and downhole fluid sampling. The monitoring well was instrumented for water abstraction, water injection, downhole pressure and temperature monitoring, distributed temperature sensing via optic fiber (DTS) and downhole fluid sampling (see Figures for WP02).

A versatile CO2 injection system allowing the injection of CO2 as a free phase and or the injection of CO2 for mixing with water, including a CO2 storage tank, a booster pump, a low-flow CO2 pump (up to 250 Kg/hour) and a high flow CO2 pump (up to 4000 Kg/hour). A heat exchanger providing the ability to set temperature of the CO2 to be injected, fueled by gas. A monitoring system (pressure and temperature sensors, Coriolis mass flowmeter was laid out between the CO2 tank and the well-head for a full control of the CO2 conditioning prior to injection. The CO2 injection kit was instrumented for safety thus making (see Figures for WP02).

A compressed air distribution pipe network for supplying compressed air from an air compressor to the wells.

A control room to which all the communications line from the wells and the CO2 injection kit are gathered and connected to a data acquisition system. The control room also hosts the TiT fluid connection panel to which the downhole fluid samples are carried for analysis. The TiT panel allows the collection of high-pressure and low-pressure samples. The TiT panel was modified so it includes three additional functionalities: 1) installation of a high pressure electrical conductivity probe (for detecting the arrival of the sample under stable conditions; 2) a high pressure pH probe (for determining the amount of dissolved CO2 in the high pressure sample) and 3) a device for measuring the partial pressure of gases in the high-pressure sample. The control room has also a quadrupole mass spectrometer for the determination of gas composition in a gas sample (obtained from the depressurized gas samples) and devices for the low pressure measurement of pH, alkalinity and electrical conductivity. The control room also hosts the data server for the seismic monitoring. (see Figures for WP02).

A system for the conditioning of water prior to its injection to the formation. In order to avoid clogging and or any unwanted chemical reaction, we only inject formation water (either abstracted and previously stored or directly from the monitoring well). Prior to its injection the produced water is filtered with a sand filter in order to remove solid particles, then treated with UV for bacteria removal and electrolyzed (to form chlorine for the chloride present in the water) as a downhole biocide. These operations are necessary to ensure the injectivity of the well (see Figures for WP02).

Drilling of three shallow wells (~250 m) to be used in the seismic monitoring setup ((see Figures for WP02).
A number of hydraulic and thermal experiments were conducted, using airlift as the abstraction system. The airlift experiments allowed controlling the water abstraction via airlift, which is usually extremely noisy. The thermal tests were used to characterize the heat exchange process between the well and its environment and the vertical heterogeneity of the reservoir sand layers.

Planning and execution of the CO2 injection experimental sequence

The objectives of the CO2 injection experiments were to assess at field scale key parameters of trapping (dissolution trapping and residual trapping) and conduct a seismic monitoring of small amounts of injected CO2. The former required the creation of a residual zone of the CO2 in the formation, with CO2 saturated formation water. We carried out three CO2 injection experiments:

Residual Trapping Experiment I: in this experiment, the reference tests used to determine the residual saturation were mainly hydraulic withdrawal tests, carried out before and after creating the residually trapped zone. To create the residually trapped zone, 100 tons of CO2 was first injected to the target reservoir, followed by fluid withdrawal until CO2 was at residual saturation.

Residual Trapping Experiment II: Unlike in experiment I, in this experiment the residually trapped CO2 zone was created by CO2 injection, followed by the injection of CO2 saturated water, to push away the mobile CO2, leaving the residually trapped CO2 behind. In this test, the main reference test carried out before and after creating the residually trapped zone was injection and recovery of partitioning tracers Krypton and Xenon.

CO2 injection and monitoring with geophysical methods: The main objective of this experiment was to see how a smaller amount of CO2 can be detected with geophysical methods. If successful, the analysis can also give information concerning the overall spreading pattern of CO2 in the formation and thereby provide support to analyses of previous experiments (1) and (2). This experiment and the monitoring has been reported in Deliverable.
Overall, extensive sets of data were collected during the various stages of the experiments indicated in the tables above, that allow analyzing the reservoir behavior, especially the residual trapping under CO2 injection and withdrawal. The two residual trapping experiment carried out at Heletz site will provide unique basis for determining CO2 residual trapping in-situ. To the best of our knowledge, attempts to determine CO2 residual trapping in-situ has until now only been carried out at Otway, Australia. The experiments at Heletz will provide important new information concerning the CO2 trapping behavior. In particular, two distinctly different approaches are used in the two experiments, both in terms of how the residually trapped zone was generated and what where the reference tests for determining the degree of residual trapping.

Capacity building and exchange related to programs on Swedish and Baltic sites – Sweden

We carried out model simulations to determine the dynamic storage capacity of selected regions of the Baltic Sea, as part of the Bastor project. In Sweden, there has been a working group consisting of a wide presentation of industries and also academies, working on a plan to address the government’s mission of zero CO2 emissions, and to look for a road-map to meet the goal of zero CO2 emissions, with a particular focus on process industry. The work has continued to establish a full-chain CCS demonstration project in Sweden, involving capture, transport and storage (led by Swedish Energy (Svensk Energi)). Parallel to the above, work is ongoing to establish a research motivated pilot-scale injection site in Sweden, with the objective to demonstrate the relevant storage and sealing properties of the key formations in the Baltic Sea (Dalders Monocline) in a pilot scale setting on the island of Gotland. None of this work has been carried out within TRUST. However, there is a well-functioning exchange of information between these projects and TRUST (mediated through presentations in the project progress meetings), working in both directions. In particular, this exchange 1) supports the work in TRUST by providing insight to the issues and challenges of industrial scale CCS projects and 2) the experience gained in TRUST and its predecessor MUSTANG provide information helpful in establishing a pilot/demo site in Sweden /Baltic.

Data gathering and distribution at Hontomin (CSIC)
Characterization tests at Hontomin included a high pressure injection test to characterize the (thermo-) hydro-mechanic response of the reservoir and cap rock; brine push-pull tests where conservative tracers will be injected to characterize mixing, porosity structure and reactivity of the rock; and short hydraulic tests to monitor possible changes in permeability.
Interpretation of these tests has been performed in the framework of this task. The most relevant outcome derives from the different behavior observed in the injection and recovery stages, which suggests that the medium transmissivity is changing with the overpressure, probably because of the elastic aperture of pre-existing fractures.

Data gathering and synthesis of EOR experiments for the Reconcavo field, Brazil

Petrobras started injecting high-pressure CO2 in 2009 into the Miranga onshore field (Reconcavo basin, Bahia state) to test technologies that might contribute to future development projects for the Santos Basin Pre-Salt cluster. The carbon dioxide produced by the pre-salt fields is now reinjected in a Water-Alternating-Gas (WAG) mode into the reservoirs themselves to boost the oil recovery factor following an Enhanced-Oil-Recovery (EOR) strategy.
The Miranga field was selected for the tests on account of its geological characteristics and existing logistics available at the site. The technique to be used in Miranga is based on injecting CO2 under high pressure, using it like a solvent that changes the properties of the oil. The Miranga field project foresees the geological sequestration and removal of 370 tons of CO2 from the atmosphere per day. In addition to the environmental gains, the EOR process considerably increases the recovery percentage of the oil nestled in the reservoir. This project is used as a proving ground for new technologies that might be applied in other fields to be developed in Brazil, particularly in pre-salt discoveries, since the reservoirs there have shown the presence of natural CO2 associated to the oil.
Increasing oil recovery by injecting CO2 is not new to Petrobras. The procedure was first used 28 years ago, in Bahia. When it started experimenting with the concept, Petrobras applied a high-pressure CO2 injection technique at the Araçás field, in the Reconcavo Basin. In 1991, it deployed a low-pressure CO2 injection project in the Buracica field, in the same basin. The project was highly successful and resulted in the partial maintenance of the field's oil production for nearly 20 years.

Work-package 3: MMV technologies and real-time reporting and visualization

Comparative study of the different Measurement, Monitoring and Validation systems

We have prepared a full review of the different Measurements Monitoring and Verification (MMV) technologies and real-time reporting and visualization, on the Heletz and Hontomín sites and summarized in Deliverable D3.1. We present in this deliverable the different TRUST project sites and the common and specific characteristics of their monitoring systems. We have detailed the specific borehole monitoring program and instruments developed in Heletz and Hontomin sites, such as the downhole fluid sampling system (TiT), the high precision pressure and temperature sensors and the fiber optic distributed temperature sensing (DTS) in Heletz and the hydrophone and electrical resistivity sensors as well as the TiT setting in Hontomin. We also explained and described the different surface monitoring technologies developed in both site, such pressure and temperature sensors along the pipeline and CO2 sensors near the injection kit. We also add an example of EOR monitoring system in the Buracica field.
RTS downhole monitoring

The plan was to drill a new monitoring well and to test an integrated multi-parameter monitoring string using a "behind the casing" well completion approach. The monitoring parameters included: 1) downhole pressure and temperature measurement, downhole fluid sampling, optic fiber distributed temperature sensing (DTS), vertical distribution of temperature via optic fiber FBG (Fiber Bragg Grating) sensing and Electric resistivity tomography. Most of the design and preparation work of the single components were prepared. However, due to severe restrictions set by EU on EWRE budget one year after the beginning of the project, it was not feasible to finance the drilling and the completion of this well and this task had to be cancelled.

Passive and active near-surface seismic monitoring

The seismic program of the TRUST Project had as main task to perform integrated active and passive seismic measurements by using permanent installations of borehole receivers for long term monitoring of the noise before, during and after injection of CO2 and to develop and build the tools and methodology to make the above possible. The development, construction and preparation of the various instruments and techniques has been successful, in as much as 3-component, wide-band digital receivers were built and deployed in 3 boreholes, approximately 250m deep each, installed in arrays of five units per borehole. In addition to the initial plan, sub-surface receivers were also built and deployed as buried arrays along three lines. This component was bound to replace the construction of semi-permanent buried sources, the use of a more extensive receiver array being seen as a major advantage, while a surface source could be operated periodically at multiple locations, instead of fixed buried sources. Nonetheless, a prototype of a buried source has also been designed as a part of the TRUST project. The seismic investigations concept developed within the TRUST Project has been intended to provide the experimental frame for monitoring small injections of carbon dioxide, as needed for leakage prevention and leakage mitigation. The aim is to identify, map and characterize potential and actual migration paths in depth, within a volume centered on the injection well. The concept is based on permanent and semi-permanent seismic receivers and active sources placed in the shallow sub-surface and boreholes and includes continuous data acquisition, online detection, selection and processing of relevant seismic events (natural micro-seismic events and artificial source-generated events), as well as communication and data transfer to remote users.

The monitoring system developed and built by Vibrometric consists of buried semi-permanent sources, semi-permanent receiver arrays deployed in borehole and under the ground surface. The system is completed by a server for data collection, on-site processing and interpretation, which can produce processed data ready for being displayed and interpreted in nearly real-time. 4-component, wide-band digital receivers were built by Vibrometric and were installed semi-permanently in shallow observation boreholes. These receivers are also suitable for deep observation wells. Variations of the design allow multi-level strings to be installed as part of the injection well completion, in deep observation wells drilled at the injection site in the immediate vicinity of the injection well, or in several shallower wells drilled in a regular pattern around the injection site.

Novel technological solutions were found, tested and implemented for the construction of these digital receivers. Each borehole receiver unit contains six 14-Hz geophones oriented as 3 orthogonal components in a double differential configuration, and a low frequency (2 Hz) vertical geophone. Four programmable amplifiers and four 24-bit A/D converters are placed immediately adjacent to the geophones. This construction allows very low signals to be measured. Data ports for communication with the acquisition server and power supply modules complete the configuration. The receiver has connectors at both ends, permitting various configurations to be implemented. All fittings and couplings are sealed with multiple level O-rings, to assure long-term water tightness at both low and high pressure.

The seismic monitoring system is meant to contribute to the understanding of the structural site geometry, to the monitoring of the expansion of the CO2 plume with time and will lead to a cohesive interpretation of the seismic results over scales. With the deployment of seismic receivers proposed for Heletz, the monitoring objectives are addressed by combining two complementary (and partly overlapping) approaches to seismic subsurface imaging: 1) Sparse surveys yielding high temporal resolution and 2) Easily deployed high resolution surveys adaptable to focused studies.

Seismic equipment installation at Heletz

Surface and borehole receivers developed and built by Vibrometric were deployed at Heletz, in a semi-permanent installation (the instruments were kept stationary for the duration of the experiment and in the end of the project were retrieved). Lead-in cables conveying the digital signal and the necessary electric power were partly buried between the well heads and the acquisition stations. The intention has been to burry all cables, as required by a monitoring experiment, which in principle could have lasted for several years. Due to unforeseen project delays beyond our control, the installation was however adapted to a short term exercise, and limited installation permitting.

The layout of the seismic monitoring network was planned assuming that the CO2 plume will migrate to the west (up-dip direction). It consists of three buried lines: Line 1 on the eastern side of the site, connecting the other two shallow boreholes intended for the installation of 4 component borehole receives. Line 2 in the E-W direction passing in the immediate vicinity of the injection site and along one of the three shallow boreholes in which groups of five 4-component receivers were installed (Hole B), and two N-S lines. Line 3 parallel with Line 1, to the West. Each line length was approximately 600 m. The installation depths of the 4 component borehole receivers were planned to be 20m, 70m, 120m, 170m, and 220m, while actual depths were 2.2m 52.2m 102.2m 152.2m and 202.2m in Hole B and 4.2m 54.2m 104.2m 154.2m and 204.2m in Hole C.

For long term use, the cables, digitizing hubs and sensors were planned to be installed in PVC tubes and junction boxes, to be protected from rodents. The network was composed of six module lines, each connecting 20 sensors, to a total of 120 sensors distributed as 5- level x 4 component receivers (one line per each of the three boreholes) and 20 buried vertical sensors on each of the three subsurface line modules.

Semi-permanent shallow hole seismic monitoring network

The shallow wells used for installation of the down-hole digital receivers were cased and cemented from bottom to top. The outer diameter of the casing is 5 inches – grade 18 #/ft - inner diameter 4.276 inch - drift diameter 4.151 inch. The casing collars are located in 1m x 1m x 1m concrete-lined pits, covered by a metal lid. The casing collar is at a depth of 0.5 m below the lid. The digital bus cables for six data line modules (bus cables) were deployed, but permitting difficulties were encountered when attempting to bury them in pipes as planned. The cables were buried only under the roads and areas where active agricultural traffic was expected. Three of the bus cables went uninterruptedly from the acquisition server to each of the three boreholes. For the remaining three bus cables serving the sub-surface receivers, connecting PVC pits were set every 100m and A/D hubs placed in these pits. Due to the same permitting difficulties, the pits were not buried completely, although with a long term semi-permanent installation they would be.

Semi-permanent near surface seismic monitoring network

60 vertical receivers were prepared for being installed in the sub-surface at Heletz along perpendicular lines with an interval along the line of approximately 25m. The receivers were buried in Auger boreholes.
However, only 36 receivers were installed in the sub-surface, because of the disagreement of the land owner. The installation was done by drilling an auger borehole, inserting a plastic pipe, filling and compacting the ground around the pipe, lowering the receiver into the pipe and filling the pipe with sand. The geophone was then connected to the data line. The rest of the geophones were only placed on surface and connected as all the others. As mentioned previously, the cables were not buried in PVC pipes as originally planned, the connection T-piece remaining over the ground surface. The digitizing hubs were also spread along the lines, connected and placed in plastic connection pits.

Seismic source

The VIBSIST class of seismic sources have been developed by Vibrometric and are time-distributed devices which accumulate energy in time by emitting a non-repeatable sweep, later correlated with a pilot time-function to obtain a regular seismogram. Similarly to regular vibrators, this characteristic allows high energy to be pumped into the ground by using a moderate instantaneous power. Also common with regular vibrators is the computerized command and feedback control of the sweep parameters. Unlike with regular vibrators, the VIBSIST sweep consists of a coded series of wavelets, each containing the whole frequency band. A fully reconstructed record is obtained in uneven terrain and with complex near surface ground conditions. This makes this version of time-distributed source more robust and it has been regarded as a natural step towards extending the use of time-distributed sources from oil and gas exploration towards shallower, higher resolution applications. To reduce the overall mass, the ground coupling force in the VIBSIST case is realised dynamically. In the VIBSIST-3000 case, at a reference wavelet frequency of 100 Hz, 3000 J are injected into the ground in ~ 2.5 ms, generating a peak force in excess of 200,000 N.

Recording station

A dedicated recording station was developed and instrumented by Vibrometric. This was installed at Heletz, for continuous data acquisition.

Data acquisition and preliminary evaluation

Because of the stretched schedule of the TRUST Project, the seismic fieldwork at Heletz could be started only in November 2017. After nearly two weeks of site preparations, the data collection was started on December 3rd, with the baseline active source measurements along the three lines deployed, as shown in Figure 3.13. Site coordinates of all source and receiver positions on surface are listed in deliverable D3.3. The measurements continued with passive listening before and during the injection of CO2. A second listening session was performed in January 2018. Details of the seismic data acquisition program are given in deliverable D3.3.

3D seismic synthetic seismic data modeling for Heletz

Combine passive and active seismic data sets

The low-frequency information in the seismic data is an important factor that determines the success of the full waveform inversion of seismic data. In the complex situation of actual seismic exploration, active source data is usually lacking low-frequency information due to various factors such as source signal, receiver, acquisition cost, and so on. The full-wave inversion result of seismic data is also affected by the lack of low-frequency information. Currently, the main solution is to perform low-frequency compensation through mathematical methods. However, this method lacks a physical basis and may not be able to obtain real and effective low-frequency information in a complex actual acquisition environment. Given the above problem, we choose to use passive source data to compensate for the low-frequency information of the active source data. This is because passive source data is derived by processing seismic signals generated by actual noise and other subsurface hypocenters. It usually contains a wealth of low-frequency information and has a solid physical basis. A synthetic data set is used here to test our method. We use the seismic interference method to process the synthetic passive source data to generate a virtual seismic shot gather. Then match the virtual shot gather with the simulated active shot gather. The main purpose of matching is to adjust the spectrum and energy information between the two to find a suitable intersection. Figure 3.14 shows the results of our test. We could find from the matched power spectrum and the compensated results that the method succeeds in reconstructing the low-frequency information recorded by using the passive source data. Applying this method to the full waveform inversion, the result of low-frequency compensation will inevitably achieve better results than the missing low-frequency original data. It can effectively improve the stability of the full waveform inversion and its results when applying that to the time-lapse monitoring of CO2 injection.

Multi-source time-lapse full waveform inversion: A synthetic example.

Full waveform inversion of multi-source time-lapse seismic data can directly reconstruct the change of velocity and other parameters of underground media, providing another effective method for time-lapse seismic monitoring and interpretation. For conventional reflection time-lapse seismic processing method, it is a useful supplement. However, the full waveform inversion of time-lapse seismic data is not only limited by the difficulty of conventional full waveform inversion, such as surface conditions, noise, and bad initial velocity modeling, but also by the acquisition equipment, acquisition time, time-varying noise caused by near surface changes.Based on the previous research, we develop a workflow to combine passive and active data sets to improve time-lapse full waveform inversion.

First, the full waveform inversion is performed using the passive source data of the baseline. Because the passive source data contains low-frequency information, it can effectively avoid the phenomenon of "cycle-skipping" and increase the stability of inversion. Although the dominant frequency is relatively low, the inversion result can provide the inversion of the active source data a more accurate initial model.

Using the inversion result of the baseline passive source data as the initial model, the active source data of the baseline survey is inverted to obtain the final model of the baseline survey.

Taking the final inversion result of the baseline survey as the initial model, the double-difference inversion method was used to invert the active source data of the repeated survey to obtain the final model of the repeated survey.

Due to the lack of real time-lapse multi-source seismic data, we use synthetic data to verify the above workflow. The model is based on the Ketzin pilot site for the carbon dioxide geological sequestration. We performed the full-wave inversion on the active and passive sources time-lapse seismic data. We use the smoothed version of the real velocity model as the initial model for inversion. As a comparison, We also performed a conventional double-difference time-lapse full-waveform inversion of the active source data. The result is shown in Figure 3.20. For the joint inversion of active and passive sources, we have performed high-pass filtering on the active source data to remove the low-frequency information below 25 Hz to simulate the situation where the active source data is missing at low frequencies. We also took the velocities from the middle of the model for further comparison (more like a logging data). By contrast, we can find that by using passive source data we effectively compensate for the lack of low frequency of active source data. The stability of inversion is increased, and the inversion results are also improved.

Reactive tracers

We have conducted a wide range of KIS tracer laboratory experiments using the setup. The experiments used two categories of KIS tracers, externally and in-house produced.

The two-phase flow experimental data was matched by performing a parametric sweep on the two most uncertain parameters in the experiment — permeability and injection rate. In the case of parameter “permeability”, both, the porous media (glass beads) and the porous frits permeabilities were determined with a falling head permeameter. It was found that the low permeability of the porous frits impacts significantly the overall permeability of the system. Difficulties quantifying the exact effect of the porous frits resulted in permeability being approximated numerically. The variability in the injection rates due to the peristaltic pump was shown to be in the order of approximately +/- 0.05 ml/min. Since the numerical simulation is particularly sensitive to these parameters, a parameter sweep was performed to best fit the experimental data. A prerequisite for the correct matching of the experimental data is to know the capillary pressure saturation-specific IFA relationships. Multiple studies have shown the ability to match drainage and imbibition data onto bi-quadratic surfaces. Projecting this surface onto a 2D plane for Sw-awn, we attain a curve that shows an increase in awn up until some wetting saturation S_w (a_wn^max) , followed by a decrease in the specific IFA with a continued increase in wetting saturation.The experimental results were fitted by running a parameter sweep on parameter a0 from the polynomial. This parameter controls the maximum specific IFA, a_wn^max while maintaining the general shape of the S_w-a_wn curve. For capturing all experimental data points, we present the entire range of a_wn^max numerical solutions required to fit the most off-set data points.

Fluid pressure tomography

Monitoring techniques are crucial for being able to immediately react to potential CO2 leakage from the geological reservoir and for formulating remediation strategies. For this purpose, geophysical methods are extensively used. However, a major pitfall of most geophysical methods is that they merely provide indirect information on flow properties. This yields uncertainties especially for the estimation of CO2 saturation. This ETH contribution to the TRUST project introduces an alternative approach, fluid pressure tomography, which has potential to overcome this by direct linkage of the observed signals and the inversion procedure to the flow regime.

Due to the complications of CO2 properties, CO2 sequestration involves more complexities than single-phase flow. We have developed a single-phase proxy, which can significantly reduce the computing burden of full multiphase simulation and accelerate the inversion procedure. In this proxy, CO2 and brine are assumed as a phase mixture, neglecting the secondary processes, such as thermal and chemical processes. Disposal of CO2 in the brine-rich formations alters the mixed-phase flow properties. The mixed-phase specific storage increases greatly with increased CO2 saturation, since CO2 is much more compressible than brine. In contrast, variations in the mixed-phase conductivity are relatively small. Similar to the ratio of the mixed-phase conductivity and specific storage, mixed-phase diffusivity can change by up to two orders of magnitude, which can be recognized by fluid pressure tomography. Implementation of pressure tomography involves brine or CO2 injections as sources, and pressure measurements in different locations as receivers. Pressure transients at the observations are utilized for travel-time based inversion, which yields the structural information of the subsurface. Plume development is inferred by comparing and clustering the inverted diffusivity tomograms acquired at different times. The CO2 saturation of the identified plume is then derived by calibrating the measured pressures based on the single-phase proxy. A synthetic homogeneous case has been used for demonstrating the feasibility of the method.

Applying pressure tomography not only to the storage formation, but also to the above aquifer, can detect potential CO2 leakage occurred at different times. A no-leakage case is simulated as a reference to be compared with various leaky cases. It is demonstrated that pressure responses and hydraulic travel times in storage formation and the aquifer above provide a first insight in the leakage type. Comparison of the diffusivity tomograms in both storage formation and above aquifer among no-leakage and leaky cases can localize the leakage. Furthermore, the influence of data noise and well distance is examined. Results indicate that the noise has an impact on the inversion results and leakage detectability. Increase of well distance also weakens the detectability of CO2 leakage, since it reduces the inversion resolution.
In the last part of our contribution, pressure tomography is conducted in heterogeneous formations, in comparison with crosswell seismic tomography under comparable conditions. Hydraulic travel times show much larger relative spread than seismic tomography, which allows pressure tomography to better resolve the more complicated geometries. Moreover, from the inverted tomograms, these two approaches show different capability for resolving the aquifer structure and the CO2 plume. Pressure tomography delineates the structure of the initial CO2-free formation better than seismic tomography, because it directly relates to formation permeability. For the post-injection periods, however, seismic tomography can always depict the main part of the plume, while pressure tomography is more influenced by the heterogeneity of the aquifer. Joint clustering of pressure and seismic tomography results combines the advantages of these two approaches. The plume shape is better identified, and also the estimation error of plume CO2 saturation is reduced.

This contribution reveals the theoretical potential of the new concept of “time-lapse pressure tomography” to further adapt the single-phase hydraulic tomography to a two-phase flow system in a time-lapse manner. Based on the single-phase proxy, the tomographical inversion and calibration of the flow properties become rapid and computationally efficient. With the promising theoretical results, the methodology introduced here is ready for field applications future.

From the previous work, we have demonstrated that pressure tomography (PT) can be used for timely characterizing a CO2 plume in deep saline aquifers, as well as detecting potential CO2 leakage during early-stage injection. Inversion is based on a single-phase proxy, in which CO2 and brine are assumed as a phase mixture. Not only the plume geometry, but also the CO2 saturation can be identified by the proxy. For testing our methodology, we compare PT to crosswell seismic tomography (ST) under same conditions. We implemented the tomographic experiments (multilevel fluid injections and P-wave pulses) in several heterogeneous models, and then inverted the hydraulic and seismic traveltimes, respectively. Results show that PT can clearly depict the aquifer structure prior to CO2 injection, as it directly relates to permeability. However, after CO2 injection, ST performs better than PT for tracking the CO2 plume in more heterogeneous aquifers. We then jointly clustered the inverted diffusivity and velocity differences from PT and ST. Clustering results indicate that PT can complement ST by improving the shape identification and reducing estimation errors of CO2 saturation.

Integration of MMV results at Heletz

The data handling system built by Vibrometric and tested at Heletz is functional for real-time acquisition, monitoring, online processing, data integration and visualization of seismic data. All the data is integrated into Database Management Systems (DBMS) using MSACCESS 2016. All the data gathered for each experiment is organized in tables of different natures (times series of downhole pressure and temperature management, time series of measurements above the ground, results of the downhole sampling etc.). The integration work is explained in detail in Deliverable D3.5. DMBS files will be placed in the public domain, once the partners have completed the relevant publications. We estimate that the time needed for the partners to complete the publication to three years from the end of the project. After this time the datasets will be downloadable from the TRUST site.

Datasets including all the measurements from the experiments were prepared using the MS ACCESS 2016 database management system. It comprises tables of time series of downhole pressure and temperature, time series of above the ground pressure, temperature and massflow rate, results of the gas compositions, high pressure and low pressure pH alkalinity and electrical conductivity, results of laboratory analysis, results of the partial pressure of gases and fluid sampling.

Integrated visualization platform

EWRE has developed a web-based software platform for online and real-time visualization of monitoring data. It was presented in report for period #3 and in deliverable D3.6. The system is fully functional. It has undergone a number of improvements: 1) improving the reliability of the online data collection process for the real-time data; 2) the online data (historic data) has been organized per experiment; 3) the user can set a time window and retrieve results; 4) data can be downloaded from the project site (from the restricted area).

In his case the user can select time range via the selection brush (shows all the history) on the bottom, and more precisely via the date and time picker on the top. In addition, user can get the data for all active sensors in csv files. EZ measures are all in on file as their timing is the same.

A key advantage of this system is that it is independent of the operating system and works in all devices with browser capability (smartphones, tables, laptops and computers).

Work-package 4: Modeling

Experiments planning simulations and validation of the simulation models against experimental field data

We carefully planned the CO2 injection experiments at Heletz via modeling, using as input the findings of the geological characterization, the results of the hydraulic tests and the site-specific two-phase brine CO2 capillary pressure and relative permeability data. Protocols for each experiment were outlined. They describe along the timeline of the experiment, the activities that needed to take place and the non-automatic monitoring and measurement activities.

Modeling of Heletz injection experiments

We conducted an extensive analysis and modeling of the two residual trapping experiments (denoted as RTE I and II) carried out at Heletz.

We modeled the experiments with ‘full-physics’ TOUGH2 (USA-LBNL Simulator) approach, matching the entire test sequence and thereby getting understanding of the residual trapping behavior of the site. First we carried out the modeling of RTE I. A simplified analytical model was first used for guidance, followed by ‘full-physics’ modeling with the TOUGH2 simulator, where all the data (temperature, pressure, flow rates, two-phase flow behavior etc.) were matched. Comprehensive calibration procedure led to a best estimate of the test behavior, suggesting an in-situ residual saturation of 0.1 including a hysteretic behavior in the relative permeability functions, similar properties in the two reservoir layers and preference of the CO2 to enter the upper layer.

The ‘full-physics’ model calibrated with RTE I was then used to model the later RTE II experiment. Without any further calibration, the results showed an excellent agreement for the early parts of the experiment (prior to establishing the residually trapped zone) and relatively good agreement even with the later parts, with residually trapped CO2 in the formation. In particular, the amount of the Krypton tracer partitioned into CO2 was well captured with the earlier calibrated model, without any further adjustments, indicating a similar estimate of residually trapped CO2 than from RTE I.

Supporting modeling of the behavior the residual trapping experiments - modeling of the behavior of the injection well and detailed analysis of the hydraulic test response

We first present a coupled wellbore-reservoir model – by means of T2WELL simulator - for the period of CO2 and CO2/water self-release during the creation of the residually trapped zone. This is especially needed to provide information concerning the CO2 lost during the self-release period, but also to get overall supporting information concerning the system behavior during the experiment. The observed geysering behavior can be well matched by CO2 exsolution in the well and reduced relative permeabilities in the formation, due to this exsolution. The model provides valuable supporting information concerning the overall behavior of the test and will later be incorporated with the overall reservoir model of RTE I.

We present an in-depth analysis of the details of the hydraulic response of all the hydraulic tests carried out in the Heletz injection well, both as part of the site characterization program and as part of RTE I. The results show that while the response of the two tests with no CO2 in the system show a similar behaviour, the response from the test with residual CO2 in the system is different. The storage coefficient in the well is greater, delaying the response, but the skin permeability has lower values. This behaviour can be explained by a small amount of residual CO2 around the well, which is consistent with the other model analyses.

Altogether, the modeling carried out in Deliverables D4.3 and D4.6 will provide a good understanding of the system performance during CO2 injection and the resulting residual trapping. The analysis work and summarizing the results will continue beyond the time of final reporting for TRUST and the resulting peer-reviewed publications will be uploaded to the TRUST website.

Modeling the injection well

We further developed the modeling capabilities as follows. An existing simulation tool (TOUGH2-T2Well-ECO2N) for the combined flow along the well and in the formation of multi-phase CO2 and water has been modified in order to enable phase change of CO2. The enhanced software uses the equation of state module (termed ECO2M within TOUGH2), which includes all combinations of an aqueous phase and one or two CO2 phases. This modification is necessary in order to design injection scenarios, match measurement results of experiments during injection as well as during the release of CO2 from the well after the completion of the injection. The new module has been used for these goals and work is underway to interpret the results of the first injection experiment at Heletz.

A sequence of simulations has been conducted in order to examine the possibility of a “cold injection” scenario for a well whose parameters are similar to Heletz’. Our main finding is that it is necessary to inject at the very beginning warm CO2 at increasing pressure until the whole well is filled with supercritical CO2. Then, after approximately two hours, it is feasible to reduce the temperature and pressure at the wellhead and to continue injecting at cold temperature. The sequence ends with wellhead pressure of 100 bar and 60 C. Our main result is that once the CO2 at well head first reaches the pressure / temperature point along the gas-liquid transition line it continues its trajectory in the phase space along the transition line and with time the fluid in the part of the well below the well-head follow . Yet, the simulation stops converging soon afterwards and the time scales of the resulting flow do not match the time scales of the real process. It seems that the gas-liquid transition curve acts as a phase-space “sink” during the CO2 release process whereas during injection it acts as a “source”. Further plans include studying the release process and matching with the results from Heletz.

Modeling methodology and modeling of the long term and industrial scale behavior

We presented the main approaches for the modeling of CO2 storage in large domains by including simpler analytical and semi-analytical solutions to vertical equilibrium models and 3D full-physics simulations. We addressed issues of geological heterogeneity and upscaling and/or by simpler or more effective methods. The modeling approaches were applied to the work reported in WP10. Another approach presents the use of multi-level Monte Carlo approaches to treat heterogeneous systems with CO2 injection).

We have investigated the effect of injection strategies on CO2 spreading and trapping by means of simulations with PFLOTRAN model, by addressing large industrial scale domains. The focus was efficiency of CO2 spreading and trapping in a deep saline water containing geological formation under different injection strategies. A series of numerical simulations of injection and spreading of CO2 into a reservoir with the characteristics of the Heletz site (but with larger dimensions) and the time and space distribution of pressure, saturation and CO2 concentration was monitored. The following strategies have been studied: (1) Effect of temperature of injected CO2; (2) Effect of intermittent injection of CO2: (3) Effect of brine injection, with (a) Injection of water (brine) between periods of CO2 injection and (b) with injection of water (brine) in the upper portion of the reservoir between periods of CO2 injection, or simultaneously with CO2 injection. The model domain consists of a 10km×6km×50m rectangular box. The injection point is located in the cross section in the middle of the 10km×50m face. The domain is slightly inclined with respect to the larger horizontal direction. The time scale of the simulations are of the order of 10-30 years. The injection rates in the various simulations are of the order of 1,000,000 ton/year.

The results from large scale simulations indicated the following, when modelling

The effect of temperature: in the range 30—67o C, temperature has no observed effect on the spreading.
The low temperature is felt only very close to the injection well.
The effect of simultaneous injection of water in the upper part of the conducting layer and CO2 in the lower part can be summarized as follows:
Injected water will rise and spread out on the “ceiling” of the reservoir above the CO2 only if injected water is lighter (e.g. warmer or less saline than the water close to the ceiling of the reservoir).
Alteration of the injection temperatures does not provide any advantage in terms of the dissolved CO2 mass, CO2 spreading or formation pressure. There is some advantage in injecting warmer CO2.
Injection strategies

The design simulations for analyzing the effect of injection strategies on CO2 residual and dissolution trapping (Rasmusson et al., 2016) show the following: Residual and solubility trapping are two important processes providing trapping, and their effectiveness ultimately determines the feasibility of geological storage. By means of numerical modeling, a systematic analysis was conducted with regard to the factors potentially affecting trapping, to guide the planned injection experiments at the Heletz test injection site. The effect of enhanced-trapping injection strategies along with the role of geological heterogeneity and the choice of trapping model (TM) were evaluated. The results showed that adding chase-fluid stages to a conventional CO2 injection enhanced the trapping. Taking into account the geological heterogeneity decreased trapping, as this retarded the buoyant migration, resulting in less imbibition and residual trapping. The choice of TM was significant, with the simplified Land TM producing the highest trapping, and the Aissaoui TM the lowest. The results stress the importance of using an appropriate TM as well as heterogeneity model for the site in question for any predictive modeling of CO2 sequestration, as different assumptions may lead to significant discrepancies in the predicted trapping.

Large-scale simulation methodologies

The goal has been two-fold, firstly to develop models and address the system behavior under geological heterogeneity, second to tackle the issues related to problem scale as modelling of the CO2 storage systems can become prohibitively complex when large systems are considered. The work starts from a Monte Carlo analysis of heterogeneous 2D domains with a focus on the sensitivity of two CO2 storage performance measurements, namely, the injectivity index (Iinj) and storage efficiency coefficient (E), on parameters characterizing heterogeneity. It is found that E and Iinj are determined by two different parameter groups which both include correlation length (λ) and standard deviation (σ) of the permeability. Next, the issue of upscaling is addressed by modelling a heterogeneous system with multi-modal heterogeneity and an upscaling scheme of the constitutive relationships is proposed to enable the numerical simulation to be done using a coarser geological mesh built for a larger domain. Finally, in order to better address stochastically heterogeneous systems, a new method for model simulations and uncertainty analysis based on a Gaussian processes emulator is introduced. Instead of conventional point estimates this Bayesian approach can efficiently approximate cumulative distribution functions for the selected outputs which are CO2 breakthrough time and its total mass. After focusing on reservoir behaviour in small domains and modelling the heterogeneity effects in them, the work moves to predictive modelling of large scale CO2 storage systems. To maximize the confidence in the model predictions, a set of different modelling approaches of varying complexity is employed, including a semi-analytical model, a sharp-interface vertical equilibrium (VE) model and a TOUGH2MP / ECO2N model. Based on this approach, the CO2 storage potential of two large scale sites is modelled, namely the South Scania site, Sweden and the Dalders Monocline in the Baltic Sea basin. The methodologies developed and demonstrated in this work enable improved analyses of CO2 geological storage at both small and large scales, including better approaches to address medium heterogeneity. Finally, recommendations for future work are also discussed.

Geomechanical modeling of cold CO2 injection

An investigation of the in situ stresses of the Heletz region was undertaken, and the minimum and maximum horizontal stresses have now been characterized and found to be consistent with world stress map estimates and local observations. A 3D geometric model of the Heletz site has been constructed based on the Isopach maps generated in the MUSTANG project. A set of mechanical and fluid properties has been compiled in order to populate the material properties of the geomechanical simulations. A study of the mesh refinement requirements for the fracture interaction simulations was conducted, including additional validations of the fracture growth code. As planned, the CSMP and ICGT libraries have been extended in order to include a fully coupled thermo-hydro-mechanical (THM) model suitable for the geomechanical analysis of Heletz and other sites. The pre-existing THM model in CSMP was sequential and did not couple the interactions between these processes. The developed THM model has been coupled to the fracture growth model that pre-existed in ICGT. The new superior model now combines thermal, mechanical, and flow effects on growth. The fracture growth model was validated in the context of hydraulic fracturing in the context of poroelastic effects and thermal effects.

Simulations specific to the Heletz site show that the thermal properties of the limestone layer that delimits the boundary between reservoir and upper shale are a key influencing factor on the temperature shock experienced by the upper shale. Preliminary simulations show that fracturing due to cold injection within the Heletz site is unlikely, and that the predictions depend strongly on the thermo-mechanical properties of both limestone and caprock shale. Future characterization of these properties is necessary.

We further refined and validated the methodology geomechanical modelling approaches for large-scale systems. This work advances our understanding of the geomechanical processes that affect fractures during CO2 injection. The injection of cold CO2 is modelled using a two-stage coupled thermo-poro-elastic model. Non-isothermal flow is considered within the fractures and rock matrix, and the two flow domains are coupled through a mass transfer term. The numerical model has been developed using standard finite elements, with spatial discretisation achieved using the Galerkin method, and temporal discretisation using finite differences. A full-scale field case geometric model, based on the Goldeneye reservoir, has been developed, and is used for simulations. In the present approach, fractures are modelled as surface discontinuities in the three-dimensional matrix. Flow through the fractures is separated from the flow through the matrix; thus, five interacting conceptual models are considered: (i) mechanical deformation, (ii) flow through the matrix, (iii) flow through the fracture, (iv) heat transfer through the matrix, and (v) heat transfer through the fracture.

A specific case is investigated, assuming that there is a pre-existing set of faults. In situ faults are modelled discretely as discontinuous surfaces in a three-dimensional matrix, with four geological layers. The faults are assumed to be initially of low permeability, with the same permeability as the caprock. However, simulations show that their apertures (and as a result their permeability) vary due to the thermo-poro-elastic effects of the injection of the relatively cold CO2. The change in the apertures is mainly due to thermal loading. Four cases are simulated: (i) injection temperature Tinj = 20oC and Young’s modulus E = 20 GPa, (ii) injection temperature Tinj = 40oC and Young’s modulus E = 20 GPa, (iii) injection temperature Tinj = 20oC and Young’s modulus E = 10 GPa, and (iv) an isothermal case with Young’s modulus E = 20 GPa. The simulations are performed for 160 years. An example output of the simulations is presented in Figure 4.2. Apertures are indicative of the extent of the deformation of the fracture surface during injection.

Results show that the reservoir layer undergoes contraction due to the cooling, which significantly increases the fault’s aperture in the reservoir section. As the far-field stress is kept constant, the contraction of the reservoir layer applies additional compression on the surrounding layers, including the caprock. Thus, the fault’s aperture reduces in sections located in the caprock.

Results also show that the thermo-elastic stresses dominate the reservoir deformation and the fault’s aperture distribution within the reservoir, while the aperture in the caprock is controlled by poro-elastic deformations. Matrix contractions due to the injection of cold fluid in the reservoir layer induces compression on the caprock layer, which reduces the aperture of the faults within the caprock. Results show that injecting warmer fluid, or having a softer matrix (lower Young’s modulus) reduces the thermo-elastic effects. The possible fracture growth is evaluated by computing stress intensity factors at the fault tips. Shearing of the existing fractures due to thermo-poro-elastic deformations is also investigated, and it is shown that the fractures are susceptible to shear under combined mode II and III due to the thermo-poro-elastic deformations. A sensitivity analysis is performed to study the effect of the reservoir stiffness as well as the injection temperature. Softer reservoir rocks reduces the compaction due to the matrix contraction. Higher injection temperature (lower temperature difference) also applies lower contraction into the reservoir layer, and reduces the compaction. The shearing of the existing fractures is also studied using the Stress Intensity Factors (SIFs). Results show that a combination of low friction and high in situ shear stress on the fracture surfaces may trigger fracture propagation under combined mode II and III in the reservoir layer or at the interface of the reservoir-caprock. No tensile growth is observed for any of the modelled cases.

Code and modeling approach inter-comparison against benchmark case

Completion of this task has required a significant effort of UU, CSIC, IMPCOL, UGOE, EWRE and IIT. At different stages of the comparison we used TOUGH2, VE-model, PFLOTRAN T2WELL-ECO2M, CSMP, CODE-BRIGHT, DuMux and eWoms. The main conclusions are that we indeed performed the code cross-comparison of six numerical simulators to evaluate their capabilities to model non-isothermal multi-phase flow in deformable porous media. We compared the relevant impacts arising during CO2 injection (e.g. CO2 plume shape, fluid pressure and temperature evolution, deformation, etc.) in three benchmark test examples, which display a gradual increase in their physical complexity. The first benchmark exercise proposes CO2 injection through a vertical well in a homogeneous, isotropic, horizontal deep saline aquifer without considering mechanical, chemical and thermal effects. The second benchmark test extends the scope of the first test by incorporating non-isothermal effects resulting from cold CO2 injection. The third example addresses the geomechanical effects induced by CO2 injection in order to assess caprock integrity. An industrial scale injection rate of 30 kg/s, which corresponds to approximately 1 Mt/year, is simulated. Due to the symmetry of the problem, the domain is solved either using an axisymmetric two-dimensional model, a quarter of the domain, or a slice (“piece of cake”) of the reservoir, depending on the computational resources and the specifications of each software. Simulation results show that all codes can solve the highly coupled non-linear system of partial differential equations and that they are in fairly good agreement with each other. Yet, differences arise for some results. The sources of discrepancies come from differences in the equations of state of brine and CO2, in the numerical methods (i.e. finite elements or finite volumes), or implementation. By conducting a series of result inter-comparison sessions, the participating groups were able to improve the quality of results and identify errors.

Work-package 5: Strategies for storage management

Comparison of different CO2 injection modes at field scale

We suggest optimal ways of dimensioning the WAG (water-alternating-gas) injection scenarios. At Heletz, the injection well is instrumented with a "Chemical Injection Line" (CIL), a 3/8 inch tubing that connects to the main injection tubing in the well at a depth of 1000 meters. This line was designed to allow the injection of CO2 at low-flow rates for the purpose of mixing with water flowing in the tubing and saturating it with CO2 at conditions of pressure and temperature that are close to the reservoir conditions and of tracers to be mixed with CO2 (such as SF6). Soon we understood that this setting allows testing and validating the storage of trapped CO2 (dissolved in the water) in way that is almost optimal (exploiting the higher CO2 solubility at higher pressures). In view of the pressure and temperature conditions at this depth, there is no risk of hydrate formation and therefore it was possible to inject the CO2 with a relatively small pressure increase and no heating (cold liquid CO2). This represented a field validation of the injection of trapped CO2, with a relatively small energy footprint and using a setting that is simple and effective.

Task tsk_5.2: Evaluation of additional modes of operation by means of model simulation

We analyzed the effectiveness of injecting dissolved CO2 in order to handle potential situations which may complicate the storage. In fact, storing supercritical CO2 requires a low permeability caprock to reduce CO2 migration and high entry pressure to prevent buoyancy driven upwards escape of CO2. Also, pressure should be limited in order to avoid caprock failure or induced seismicity. Injection of dissolved CO2 shows some advantages since it provokes a smaller overpressure with respect to the injection of supercritical CO2. Moreover, CO2-rich brine will tend to sink to the deep of the reservoir, because of its density, which means that the risk of escape is reduced.

Task tsk_5.3: Laboratory analysis and analytical modeling on combination of dissolution and capillary effects and their influence on enhanced trapping

We continued the development of combined theoretical and experimental models of capillary and dissolution trapping. These are very useful for the assessment of their combined impact on CO2 propagation and the concomitant risks of leakage. Two key aspects have been concentrated upon in this area. First, we have used the large aspect ratio of propagating CO2 currents to develop so-called vertical equilibrium models. These leverage the nearly hydrostatic pressure in the CO2 and brine phases, along with empirical capillary pressure and relative permeability curves, to model the multiphase flow of CO2. An important recent contribution is the incorporation of trapping models to link the saturation of the propagating current with the residual saturation of the remaining trapped CO2. Classical models of spreading which incorporate trapping suggest that, for propagation along horizontal strata, currents may slow but not stop even in the presence of significant capillary trapping. We find that this result remains, even when the vertical saturation distribution within the currents is fully resolved, in the limit when the trapping model relating initial to residual saturation is linear. In contrast, when a nonlinear trapping model is used, such as the popular Land’s trapping model, which is more difficult to implement but also more accurate, currents may be arrested in a finite distance. These results have important implications for our understanding of the likely propagation distance of injected CO2, and have been published, with considerable acclaim from the referees, in the Journal of Fluid Mechanics. This study is complemented by a more detailed study on the geometry and maximal trapping efficiency in pendant capillary bridges between cylindrical barriers. In addition we have analysed the capillary rise of fluids within specified pore-size distributions.

Additional research has been on the slumping of stratified gravity currents in porous media, motivated particularly by the slumping of CO2-saturated brine beneath the CO2 plume. The slumping rates of these currents can in turn play the dominant role in determining dissolution rates. In addition, due to mixing during dissolution from the CO2 plume, currents will become density stratified in their interior. Such currents need to be well understood because they pose significant challenges when modelling their behaviour over long time and space scales. We have developed an analytical technique for modelling their behaviour which simultaneously captures the rate at which they advance, and the evolving density structure within these currents. These theoretical ideas have resulted in a publication in the Journal of Fluid Mechanics, and will be considered in specific test cases for CO2 sequestration scenarios.

We have put considerable effort into understanding in detail the data obtained from Sleipner in the North Sea, where approximately one million tons of CO2 have been sequestered each year since 1996. With insight and use of aspects of inverse theory, we have managed to get good agreement between our model, which relies heavily on the topographic variation of the confining reservoir, and the observed results. These ideas can be easily (and importantly) transferred to other potential future sites. We have also developed a theory, tested in the laboratory, of how the current behaves and spreads when it is initially put into, or reaches, a trapping horizon.

Conclusions of the benefits of different storage management options

We reviewed possible options for the storage management, including the integration of CO2 storage in other non-conventional sources of energy (EGS, Shale gas and or oil). These sources of energy require the injection of very large amounts of water, often under saturated in CO2. It could be possible envisaging the injection of trapped CO2 instead and or in addition of free-phase CO2. The advantages of the latter reside in the drastic reduction of the uncertainties.

Work-package 6: Leakage detection and mitigation

The remediation test is in the shallow well LAV 2 (at Lavallette, CEEL site, France) where a major hydraulically active fracture has been identified at depth of 54.25m.

For application to the CEEL site (low temperature), we focused on the solgel solution. This consisted in testing the sealing potential in porous and fractured media. We examine optimal sealing in terms of reduction of the medium permeability and irreversibility by means of laboratory experiments. We decided to work with ETS40 based sol, which we can acquire in large volume at reasonable price. We compared the properties of ETS40 with regard to those of TEOS based sol yielding similar hydrolysis and condensation behavior but different hydrolysis and condensation rates. ETS40 based sol has a lower hydrolysis rate (presence of longer chains) and a faster condensation rate than TEOS based sol. The different tests run to evaluate the drying process at different conditions also pointed to ETS40 as a better option. In addition, Silimin B75 (end product) was also tested but then was discarded because, based on the experiments performed, it appears to be not adapted for our application.

We first performed preliminary tests in order to (i) evaluate the T40 based sol, ethanol and water in terms of gelation and solidification within the porous structure of different rocks (mainly carbonates), (ii) determine the influence of synthesis time and temperature in the consolidation process and (iii) verify the injectivity.

The first objective of the experiments were to evaluate the drying process at different conditions that can influence the solgel synthesis. Some P and T conditions studied are representative of the test sites (Lavallette and reservoir conditions). The results are:

Drying under room conditions with open flasks resulted in gel formation in 2-3 days and crack formation in 4-5 days. In contrast, drying under room conditions with closed bottles led to gel formation in 8-14 days and gel solidification in more than 250 days. Note that when gel is formed in closed bottles a water/alcohol layer remains in the gel surface.
An increase in temperature (from room to 40°C), for experiments dried under room conditions and using closed bottles, accelerated gel formation from 8-14 days to 4 days.
An increase in P (from atmospheric to 4 bar), for experiments dried under the same T (room T) and in a closed system, did not affect gel formation.
The addition of water for experiments dried under room and Lavalette conditions (P = 4 bar and room T) and using closed bottles, slowed down gel formation from 8-14 days to 14-37 days.
Drying at CO2 supercritical P-T conditions accelerated gel formation and solidification in comparison to atmospheric/Lavalette P-T conditions (gel solidification in less than 10 days).

The second objective of the experiments was to obtain a solgel suitable for injection. It consisted in preparing different ETS40 synthesis by changing the water/Si and alcohol/Si ratios, in order to achieve a good viscosity and gel formation in an appropriate time (three weeks maximum). In addition, the amount of water and alcohol used in the synthesis should be minimized in order to favor solidification and avoid very important shrinking processes when solidification takes place.

Taking as a reference the alcohol/Si ratio equal to 3, different water/Si ratios were tested in the solgel synthesis (from 4 to 1, 1.5 and 2). This change was observed to be directly related to the gel formation time. The results showed that using a water/Si ratio of 1, 1.5 and 2 the gel formation time varied in 15, 8 and 5 days, respectively. Note that higher amount of water used shorter gel formation time obtained, which pointed that water is crucial in the first hydrolysis process.

The third objective was to achieve the decrease of a carbonate rock permeability by injection of ETS40 based sol. The rock sample used is a carbonate from Mallorca with circa 100 mD of permeability. We observed that with one ETS40 injection the permeability decrease is only around 70mD.

Design and planning of the mitigation test CEEL (CNRS, EWRE)

Downhole equipment

The injection system comprises two main components:

A dual packer system is installed in order to isolate the fracture level. Hereafter this part of the system is called the external tubing. The two packers limit a screened zone) through which the solgel will be injected. The positioning of the packer is optimized using a camera that is fitted inside the tubing. The dual parker system is attached to PVC tubing that run up to the surface (Figure 6.3).
The second part is the injection system itself. It is made of a single packer that can slide into the tubing of the dual packer system and be positioned in front of the screened zone. Hereafter this part of the system is called the internal tubing. It is equipped with a specific recirculation system made of two tubes connected to a pump and the solgel tank at the surface. A pressure-gauged valve is installed at the deeper location of the two tubes forming the recirculation loop (Figure 6.4). The solgel is injected when the pressure in the recirculation loop reach a pressure value higher than the fixed opening valve pressure (e.g. 1 MPa) and stop if the pressure is decreased to a value bellow the valve operation pressure. This unique two-part system allows 1) performing the pumping test before and after the solgel injection by removing the internal tubing, 2) inject the solgel using the internal tubing and 3) inject flushing fluid and install a camera to image the fracture by removing the internal tubing.

Surface equipment

The schematic representation of the surface equipment is given in Figure 6.5. It comprises the two tanks required for preparing the solgel mixture as well as the tank for the water flushing and for storing the wastes. The injection is performed using a specific pneumatic pump (PGP35-5 Geopro) that can produce a flow rate up to 45 L/min (at ≤ 0.7MPa). The solgel fluid is produced at surface by mixing two components that are specifically formulated in order to 1) to optimize the viscosity of product and control the time of solidification and 2) minimize the increase of temperature of the reaction. The fracture displays a large aperture and a large transmissivity (because of the incipient diagenesis occurring during the life of the fracture) that is probably much larger than is expected in the frame of mechanically produced fracture during CO2 storage operations. Accordingly, we cannot exclude the possibility that more than one injection will be required to fully seal the fracture at the Lavalette site. All together the duration of the injection experiment and verification of the product efficiency and chemical stability with CO2-enriched water could range between 2 to 4 months depending on the number of injection required to seal the fracture. The chemical stability of the sealing product when injected then aged in the well at Lavalette will be tested using CO2 enriched water eventually, i.e. after it will have been verified that permeability was reduced sufficiently according to our expectations. If necessary a second injection will be performed before making the test of durability (not leakage) with CO2. It is worth noticing that the fracture display a large aperture and large permeability (because of the incipient diagenesis occurring during the life of the fracture) probably much larger than is expected in the frame of mechanically produced fracture during CO2 storage operations.

Work-package 7: Risk Assessment: Procedures, Protocols for Certification and Licensing

We developed a risk management plan for the operation of the Heletz site, which included:

An environmental risk assessment (related to onsite handling of CO2, Nitrogen, SO2 and tracers).
A hazard and operability study (HAZOP) for the operation of the CO2 injection system.
A hazard and operability study (HAZOP) for the operation of other high-pressure facilities in the site (water production by airlift, operation of the downhole fluid system and depressurization of the injection well).
Certification with key regulating authorities in Israel (The Israel Water Authority, The ministry for the protection of the environment, the Fire Department and the Home-front command of the Israel Defense Forces).
Preparation of safety drills, prior to the start of any CO2 injection activity and or any activity related to high-pressure environments.
HSE (Health – Safety – Environment) document regulating the behavior of the personnel onsite.

During the whole period one HSE event, without consequences, occurred.

A draft of a certification procedure was formulated. This document provides guidelines and a certification scheme for the CO2 storage. In a first part, the document provides some general information about CO2.

We propose Different types of certifications, depending on the stage of the project:

Approval in Principle (AIP);
Design Approval (DA);
Certificate of Conformity Certificate for CO2 Injection and Storage Phase;
Certificate of Conformity Certificate for CO2 Injection, Storage and Closure Certificate ;
Certificate of Conformity Certificate for CO2 Injection, Storage Monitoring and Post-Closure Certificate;
Certificate of Compliance to the legislations or Modules selected.

The document provides guidelines for the selection of the site, the injection and the storage of CO2. The scheme takes into account the different phases of the CO2 storage, from the selection of the site to the final closure and monitoring. For each phase, we foresee a certification, from the Approval in Principle to the Post closure certificate. The certification scheme is compatible with the regulations of all countries in the world, including the most severe ones. This scheme may be applied, among others, to the Heletz field.

Work-package 10: Extrapolation

Identification of typical settings of industrial scale applications and relating these characteristics to the scales analyzed in the present field studies

Preparing the input tables with key parameters of typical industrial scale CO2 storage reservoirs. The parameters in the ‘typical CO2 storage settings’ database were thoroughly checked and updated. They constitute the input for CO2 injection TRUST code inter-comparison work.
Development of efficient, integrative core-based methodologies. These are useful because drilling costs, at large depths are high and deep wells are not readily available within most pilot projects. The techniques were compared with respect to their suitability to determine the relevant parameters, and the effect of heterogeneity, sample size, as well as the measuring bias of the analytical technique were discussed.
We conducted further laboratory and numerical analysis of Heletz sandstone and caprock from the drilled injection (H18A) and monitoring (H18B) wells. The aim is to characterize the flow, transport and mineralogical parameters of the Heletz pilot CO2, thus constituting an important addition is the numerical investigation of reactive properties of Heletz reservoir.
Update of the literature review of integrative approaches for reservoir characterization.
Appraisal of the global CO2 storage opportunities.

Simulation of industrial scale injection and storage for both CO2 storage and EOR

Large scale simulations of CO2 storage were conducted also as part of WP04. When injecting CO2 as a pure phase, the most important conclusions are:

It is very hard to guarantee taking advantage of the pore space. Gravity forces dominate far from the well, ensuring that CO2 will occupy only the top portion of the aquifer. Brine abstraction is needed to use a significant portion of pore space for CO2 storage.

From the point of view of stability, the worst situation takes place during early times assuming that injection is made through vertical wells and the aquifer is extensive.

Dissolution is an important mechanism controlling the extent of the CO2 plume but it is only relevant for moderately high aquifer permeability. Still, if permeability is very high, the late time rate of dissolution becomes close to the injection rate.

Pressure build-up can be very large depending on the aquifer permeability. The main pressure control mechanism is leakage from the aquifer across the caprock, but may have to be complemented by pumping.

Pressure buildup control through pumping is required if leakage is concentrated.

Contrary to current views, mineralization will rarely be a significant trapping mechanism.
Dissolved CO2 injection should be considered as an alternative to single phase injection be very profitable. It consists of pumping brine, cooling it to increase solubility and injecting it in another borehole sufficiently far, to prevent recirculation. CO2, or flue gas, or mixtures (we have only analyzed the CO2 case) are also injected and mixed downhole. Selling the thermal energy extracted from the pumped (and hot) brine, could substantially improve the economic feasibility of the process. This may require several dipoles (pairs of pumping and injection wells).

While our results imply that CO2 storage at industrial scale is technically and economically feasible, a number of issues remain open:

Induced seismicity: even if leakage is significant, a large pressure buildup will be produced, which may provoke induced seismicity. Seismic risk may also favor dissolved injection storage because the pressure is much smaller in this case.

Horizontal wells: We have concluded that they are not needed for pure phase injection. However, this may not be the case for dissolved injection.

Compartmentalization: We have assumed extensive aquifers, but sedimentary formations are often compartmentalized. The issue is not as severe as it sounds. With proper pressure control, large amounts of CO2 can be stored even in small compartments. The dissolved injection concept is a compartment concept.
Flue-gas: A most exciting concept would be to extend the dissolved injection concept to direct dissolution of the most soluble components of flue gas, which eliminates capture costs. We have only performed scoping calculations of solubility, which suggest that the concept is feasible, but much more work is needed.

We configured the HPC (High Performance Computing) code PFLOTRAN for regional scale modeling. In the frame of TRUST this code was essentially applied simulation of CO2 injection in the Heletz structure. We have continued the development of the EWRE-VASP software platform as a pre and post processor of PFLOTRAN. The open-source simulator Dumux was used together with the TORIS database derived typical geological settings of CO2 storage reservoirs.

We also developed a set of modeling approaches to address the behavior during industrial scale/long term CO2 injection and the behavior after the stop of the injection. The set tools consist of (i) analytical and semi-analytical models for preliminary estimation of the pressure increase. Two-phase solutions are needed for estimating the near-field pressure increases while single-phase models can be used for estimating the far-field responses; (ii) Vertical equilibrium based models to get first estimates of the CO2 plume spreading and (iii) Full-physics modeling approaches, with massively parallel TOUGH2 for obtaining the most detailed analyses. These models have been applied in a successive order from (i) to (iii) to simulate industrial scale injection in two sites in the Baltic Sea region (Dalders Monocline and South Scania).

In continuation to previous work UGOE, CSIC, UU, IMPCOL and EWRE have collaborated to perform a code cross-comparison of six numerical simulators evaluating their capabilities to model non-isothermal multi-phase flow in deformable porous media. We compare the effects of interests arising during CO2 injection (e.g. CO2 plume shape, fluid pressure and temperature evolution, deformation, etc.) in three benchmark test examples, which present a gradual increase in their physical complexity. The first benchmark exercise proposes CO2 injection through a vertical well in a homogeneous, isotropic, horizontal deep saline aquifer without considering mechanical, chemical and thermal effects. The second benchmark test extends the scope of the first test by incorporating non-isothermal effects resulting from cold CO2 injection. The third example addresses the geomechanical effects induced by CO2 injection in order to assess caprock integrity. An industrial scale injection rate of 30 kg/s, which corresponds to approximately 1 Mt/year, is simulated. Due to the symmetry of the problem, the domain is solved either using an axisymmetric two-dimensional model, a quarter of the domain, or a slice (“piece of cake”) of the reservoir, depending on the computational resources and the specifications of each software. Simulation results show that all codes can solve the highly coupled non-linear system of partial differential equations and that they are in fairly good agreement with each other. Yet, differences arise for some results. The sources of discrepancies come from differences in the equations of state of brine and CO2, in the numerical methods (i.e. finite elements or finite volumes), or implementation. By conducting a series of result inter-comparison sessions, the participating groups were able to improve the quality of results and identify errors.

Evaluation of the monitoring strategies and development of optimized monitoring network for large scale operations

At Heletz, we found out that while downhole fluid sampling provides critical information during the experimental phases of a project, it might not be needed during the operational phase (industrial storage). The key reason is that fluid sampling systems can be a liability in the sense that a failure in their functioning may present a CO2 leakage path.

Pressure and temperature monitoring are key in all phases of a CO2 storage project. We found out the optic fiber distributed temperature sensing (DTS) was robust and provided continuous information, while the pressure/temperature sensors installed in the monitoring well failed. Today it is also possible to measure pressure via optic fiber sensing and the precision of the measurements is steadily increasing (via FBG or Fiber Bragg Grating). This approach, optic fiber driven DTSPS (distributed temperature, strain and pressure sensing) may become the way to monitor the CO2 reservoir, due to the high resilience of these systems.

Preparation of protocols for site licensing and management, check lists and best practices

We have conducted a literature survey on the existing regulation related to the geological storage of CO2 and found out that recommendations and protocols for CO2 injection and storage projects prepared under the US-EPA Class VI Rule present a comprehensive platform for site licensing. The U.S. Environmental Protection Agency (EPA) Federal Requirements Under the Underground Injection Control (UIC) Program for Carbon Dioxide Geologic Sequestration (GS) wells are codified in the U.S. Code of Federal Regulations and are referred to as the Class VI Rule. This Rule was issued on December 10, 2010, and it establishes a new class of injection wells (Class VI) and sets minimum federal technical criteria for Class VI injection wells for the purposes of protecting underground sources of drinking water (USDWs).

We compared the activities carried out for the licensing of Heletz as an experimental injection facility and found out that they are compatible with the class VI regulations.

Work on TRUST impact was carried out in the frame of work-packages WP08 and WP09.

Work-package 8: Communication, public engagement and liabilities

TRUST website design, implementation, and social media

The TRUST website has been created and reviewed at the early stage of the project (www.trust-co2.org). The website is first an internal organisation tool to assist the consortium partners in sharing the findings of their work as well as useful information about internal meetings.

It also hosts the software allowing online and real-time visualization (in the partner's area).

The website is also an external tool to disseminate and promote the results of TRUST by gathering and integrating the results of the work packages in a consistent way. In order to build a better knowledge about CCS within the different stakeholder’s communities.

The website is an evolving tool as is updated as needed.

We held discussions within the consortium regarding the effectiveness of social media in European CCS research projects, such as TRUST. For this reason, an analysis of available social media was undertaken and shared with the consortium members. The reflection paper considers the pros and cons of the following tools: facebook, twitter, linkedin and viadeo, as well as its application to the TRUST project.

After having analyzed and discussed with the consortium about the use of social media, MERI created a twitter account and a LinkedIn group. The twitter account and LinkedIn group are managed by MERI and as of 16 October 2015 the @TrustCo2 twitter account has 70 followers.

Systematic review of current and past CCS projects to apply lessons learned to TRUST sites (MERI and BV)

We gathered information and extracted lessons learned from past and current CCS projects. It brought together the main findings from the literature on social science research around CCS, from reports from European funded projects, findings from national research projects and from surveys on public perception and acceptance as well as reports from different CCS demonstration projects and initiatives. The results of the main studies and surveys on public perception on CCS at EU level and compared to other technologies have been considered to extract key findings. Second, we have looked at European research projects on governance and CCS, mainly ACCSEPT, nearCO2 and SiteChar to depict main lessons learned. In addition, the ongoing R&Dialogue project which covers low carbon technologies has also been included in the report. Third, the following national CCS projects and initiatives reviewed include: Barendretch (Netherlands); Beeskow and Ketzin (Germany), Laq (France), Otway (Australia), Compostilla project (Spain) and FutureGen (US). In the conclusions, the report identifies lessons learnt with regards to communication and engagement around CCS activities.

Originally this task involved also undertaking semi-structured interviews in the area of Hontomín in Spain to assess the work undertaken in the field of communication and stakeholder involvement regarding CCS. Nevertheless, due to political and economic drawbacks in the region and in Spain, CIUDEN has stopped most territorial development actions regarding the CCS project in Hontomín. The new action plan is based on austerity and has adjusted the budget, investing only in the projects related to wine tourism and the energy museum. Contacts with representatives at CIUDEN have been established without success. Given this limitation, it is foreseen that some effort of this task could be transferred to the organization of a module on social aspects in the next training courses organized under WP9.

Public outreach (MERI)

First we worked at compiling information on social and regulatory processes in different countries and particularly, in Europe. Contacts have been established with some key persons representing different interests, such as NGOs, academics and members of the civil society. Additionally, a key player from the nuclear industry has been contacted to share their experience regarding the local stakeholder involvement in Sweden in the development of the spent fuel repository. These contacts have been invited to take part in a round table organized as part of the next project meeting which took in Göttingen (October 2014).

We Prof. Derek Taylor, professor at the Geo-Energy Centre of the University of Nottingham and previously Head of Unit for nuclear energy policy at the European Commission, to the meeting in Göttingen (20-21 October 2014). Prof. Taylor offered a keynote speech entitled “Why I support CCS?” From that meeting, Prof. Taylor is regularly informed of the progress of the TRUST project and invited to project meetings.

For the project meeting held in Uppsala (19-20 May 2015), we contacted Bellona. Ms. Sirin Engen, policy advisor and policy manager at Bellona Norway, attended the meeting and gave a presentation entitled “CCS communication and policy challenges”. In addition, we also invited Ms. Saida Laâruchi Engstrom, Vice President of SKB, and the radioactive waste management agency in Sweden. Ms. Engstrom gave a presentation entitled “stakeholder dialogue in nuclear waste management in Sweden”. The consortium learnt about the challenges associated with communication and engagement in CCS policy-making from an NGO perspective as well as the challenges regarding the siting process of a spent fuel repository.

During the second period and after following the situation with CIUDEN, communication has been initiated and maintained with CIUDEN representatives from February 2015. The representative of TRUST, Ms. Martell (MERI), was invited to be a member of the Spanish National Council established as part of the R&Dialogue FP7 project. She attended two meetings (on 6/2/2015 and 3/6/2015) of the Spanish R&Dialogue project in Madrid and contributed to both the discussion paper and the vision paper developed within the project. However, the limitations of the current structure of CIUDEN on social aspects make it difficult to create opportunities for interaction between TRUST and CIUDEN regarding the site in Hontomín, beyond the involvement in the R&Dialogue project.

Investigation of the liabilities problem for site certification and licensing

This task was attributed to a law firm to be subcontracted by EWRE. The purpose of this task is to explore the legal aspects related to CO2 storage in Israel. EWRE issued a tender to three Israeli Law firms with proven expertise and experience in environmental legislation and regulation.

The tender was sent to the following offices: Yuval Levy & Co from Ness-Ziona (www.yuvalaw.co.il) Levinson Environmental Law firm from Haifa (www.environment.co.il) and Gross, Kleinhendler, Hodak, Halevy, Greenberg and Co. (GKH, www.gkhlaw.com). Only two offices responded to the tender (Levinson and GKH). Both proposals were of high quality but GKH was selected due to their extensive experience in legal issues related to deep drilling in relation to oil and gas exploration.

The GKH law firm of Tel-Aviv, subcontracted by EWRE, prepared an overview of the environmental legal implications in relation to CCS activities in Israel, including relevant liability issues. TRUST is a pioneer project initiative in Israel and currently there is no particular legislation specifically addressing CCS activities in this country. However, it is important to analyze the existing legal framework since most of the Israeli environmental legislation addresses the uses of the land and natural resources and the permits that are required in order to carry out all activities with potential environmental impact. We point out the possible applicability of the Water Act, the Land Reclamation Bill, the Environment Protection Act (emission and transfers to the environment – reporting and registration obligations), the Mining Ordinance, the Planning and Zoning Act, the Oil Act, and the Land Act. Given that the Israeli Government is most likely to adopt international standards to regulate the activity, in addition to the current relevant Israeli legislation, the present memorandum also discusses the European Union Directive on Geological Storage of Carbon Dioxide, which provides the most comprehensive guidelines for the implementation and operation of CCS activities. Additionally, we also address liability issues arising from CCS activities both under the Israeli legal framework and under the EU Directive on Environmental Liability. We recommend a close legal and technical review during the licensing and operation of the TRUST Project in order to minimize the risks of obstructions and delays during the implementation, as well as the risks of liability.

Potential Impact:
Work on TRUST impact was carried out in the frame of work-packages WP08 and WP09.

Work-package 8: Communication, public engagement and liabilities

TRUST website design, implementation, and social media

The TRUST website has been created and reviewed at the early stage of the project (www.trust-co2.org). The website is first an internal organisation tool to assist the consortium partners in sharing the findings of their work as well as useful information about internal meetings.

It also hosts the software allowing online and real-time visualization (in the partner's area).

The website is also an external tool to disseminate and promote the results of TRUST by gathering and integrating the results of the work packages in a consistent way. In order to build a better knowledge about CCS within the different stakeholder’s communities.

The website is an evolving tool as is updated as needed.

We held discussions within the consortium regarding the effectiveness of social media in European CCS research projects, such as TRUST. For this reason, an analysis of available social media was undertaken and shared with the consortium members. The reflection paper considers the pros and cons of the following tools: facebook, twitter, linkedin and viadeo, as well as its application to the TRUST project.

After having analyzed and discussed with the consortium about the use of social media, MERI created a twitter account and a LinkedIn group. The twitter account and LinkedIn group are managed by MERI and as of 16 October 2015 the @TrustCo2 twitter account has 70 followers.

Systematic review of current and past CCS projects to apply lessons learned to TRUST sites (MERI and BV)

We gathered information and extracted lessons learned from past and current CCS projects. It brought together the main findings from the literature on social science research around CCS, from reports from European funded projects, findings from national research projects and from surveys on public perception and acceptance as well as reports from different CCS demonstration projects and initiatives. The results of the main studies and surveys on public perception on CCS at EU level and compared to other technologies have been considered to extract key findings. Second, we have looked at European research projects on governance and CCS, mainly ACCSEPT, nearCO2 and SiteChar to depict main lessons learned. In addition, the ongoing R&Dialogue project which covers low carbon technologies has also been included in the report. Third, the following national CCS projects and initiatives reviewed include: Barendretch (Netherlands); Beeskow and Ketzin (Germany), Laq (France), Otway (Australia), Compostilla project (Spain) and FutureGen (US). In the conclusions, the report identifies lessons learnt with regards to communication and engagement around CCS activities.

Originally this task involved also undertaking semi-structured interviews in the area of Hontomín in Spain to assess the work undertaken in the field of communication and stakeholder involvement regarding CCS. Nevertheless, due to political and economic drawbacks in the region and in Spain, CIUDEN has stopped most territorial development actions regarding the CCS project in Hontomín. The new action plan is based on austerity and has adjusted the budget, investing only in the projects related to wine tourism and the energy museum. Contacts with representatives at CIUDEN have been established without success. Given this limitation, it is foreseen that some effort of this task could be transferred to the organization of a module on social aspects in the next training courses organized under WP9.

Public outreach (MERI)

First we worked at compiling information on social and regulatory processes in different countries and particularly, in Europe. Contacts have been established with some key persons representing different interests, such as NGOs, academics and members of the civil society. Additionally, a key player from the nuclear industry has been contacted to share their experience regarding the local stakeholder involvement in Sweden in the development of the spent fuel repository. These contacts have been invited to take part in a round table organized as part of the next project meeting which took in Göttingen (October 2014).

We Prof. Derek Taylor, professor at the Geo-Energy Centre of the University of Nottingham and previously Head of Unit for nuclear energy policy at the European Commission, to the meeting in Göttingen (20-21 October 2014). Prof. Taylor offered a keynote speech entitled “Why I support CCS?” From that meeting, Prof. Taylor is regularly informed of the progress of the TRUST project and invited to project meetings.

For the project meeting held in Uppsala (19-20 May 2015), we contacted Bellona. Ms. Sirin Engen, policy advisor and policy manager at Bellona Norway, attended the meeting and gave a presentation entitled “CCS communication and policy challenges”. In addition, we also invited Ms. Saida Laâruchi Engstrom, Vice President of SKB, and the radioactive waste management agency in Sweden. Ms. Engstrom gave a presentation entitled “stakeholder dialogue in nuclear waste management in Sweden”. The consortium learnt about the challenges associated with communication and engagement in CCS policy-making from an NGO perspective as well as the challenges regarding the siting process of a spent fuel repository.

During the second period and after following the situation with CIUDEN, communication has been initiated and maintained with CIUDEN representatives from February 2015. The representative of TRUST, Ms. Martell (MERI), was invited to be a member of the Spanish National Council established as part of the R&Dialogue FP7 project. She attended two meetings (on 6/2/2015 and 3/6/2015) of the Spanish R&Dialogue project in Madrid and contributed to both the discussion paper and the vision paper developed within the project. However, the limitations of the current structure of CIUDEN on social aspects make it difficult to create opportunities for interaction between TRUST and CIUDEN regarding the site in Hontomín, beyond the involvement in the R&Dialogue project.

Investigation of the liabilities problem for site certification and licensing

This task was attributed to a law firm to be subcontracted by EWRE. The purpose of this task is to explore the legal aspects related to CO2 storage in Israel. EWRE issued a tender to three Israeli Law firms with proven expertise and experience in environmental legislation and regulation.

The tender was sent to the following offices: Yuval Levy & Co from Ness-Ziona (www.yuvalaw.co.il) Levinson Environmental Law firm from Haifa (www.environment.co.il) and Gross, Kleinhendler, Hodak, Halevy, Greenberg and Co. (GKH, www.gkhlaw.com). Only two offices responded to the tender (Levinson and GKH). Both proposals were of high quality but GKH was selected due to their extensive experience in legal issues related to deep drilling in relation to oil and gas exploration.

The GKH law firm of Tel-Aviv, subcontracted by EWRE, prepared an overview of the environmental legal implications in relation to CCS activities in Israel, including relevant liability issues. TRUST is a pioneer project initiative in Israel and currently there is no particular legislation specifically addressing CCS activities in this country. However, it is important to analyze the existing legal framework since most of the Israeli environmental legislation addresses the uses of the land and natural resources and the permits that are required in order to carry out all activities with potential environmental impact. We point out the possible applicability of the Water Act, the Land Reclamation Bill, the Environment Protection Act (emission and transfers to the environment – reporting and registration obligations), the Mining Ordinance, the Planning and Zoning Act, the Oil Act, and the Land Act. Given that the Israeli Government is most likely to adopt international standards to regulate the activity, in addition to the current relevant Israeli legislation, the present memorandum also discusses the European Union Directive on Geological Storage of Carbon Dioxide, which provides the most comprehensive guidelines for the implementation and operation of CCS activities. Additionally, we also address liability issues arising from CCS activities both under the Israeli legal framework and under the EU Directive on Environmental Liability. We recommend a close legal and technical review during the licensing and operation of the TRUST Project in order to minimize the risks of obstructions and delays during the implementation, as well as the risks of liability.

Work-package 9: Training and capacity building

Progress towards the objectives

Establish facilities for the technical training and demonstration at the Heletz site and facilitate

The Heletz site is now operational and CO2 injection has already taken place. During the injection, personnel from UU contributed to the monitoring and measurement work. We have established very detailed protocols for the operation of the various monitoring technologies (Pressure and temperature from above the ground and downhole transmitters, the Distributed Temperature Sensing, via optical fiber, downhole fluid sampling, the high-pressure pH of fluid samples, the partial pressure of CO2 in fluid samples, gas composition of depressurized samples. We suggest conducting a training workshop at Heletz during the last year of the project.

Organize formal educational courses for the international forum

A formal training course (Training Course on Geological Storage of CO2) was organized in Göttingen on 9-12th October, 2013. This course was in collaboration with two other EU projects (Mustang and Panacea) and had its focus on the technical and scientific considerations for CO2 injection, the exploration of critical process in laboratory studies and by field techniques, as well as the related numerical modeling. The proceeding are to be found on Mustang and TRUST web sites. The course included both theoretical aspects of CCS and a two-day workshop on the PFLOTRAN simulator (delivered by key developers of the software from the USA LANL).

The second training course took place at the CNRS premises, Montpellier, France, 19-21 October, 2015. The course objectives were: To provide participants with the current state of the art and knowledge concerning the scientific and technical aspects of planning and implementing a project of CO2 sequestration in deep, brine- containing geological formations. The course was organized by three participants: CNRS which hosted the event, headed by Philippe Pezard. IIT, headed by Jacob Bear which organized the scientific content of the course and MERI, headed by Meritxell Martell who was responsible for the publication, registration evaluation and secretariat organization. The lectures focused on three subgroups - general background, processes and modeling, and characterization and monitoring. They were given by 13 lecturers (from EWRE, IIT, UU, MERI, CNRS, UGOE, CSIC and IMPCOL and UNOT). A field trip to the near-by injection experiment Maguelone site was conducted on the last day. The course was attended by 29 students from Spain, Germany, UK, France, Norway, Czech Republic, Belgium and Israel. An on-line evaluation survey was distributed after the course and the majority of the 21 students who responded were satisfied from the course and think that they would be working in the field of CCS.

The Third TRUST Training course (3rd Advanced Course on CO2 Sequestration in Deep Geological Formations) was organized in Barcelona, Spain on 13-14th September 2017, with partner CSIC as the main organizer. The details of the training course are reported in Deliverable D9.4 and the training courses are also discussed in the overview Deliverable D9.5 that summarizes all capacity building activities.

Affiliate PhD students, Post-Doctoral scientists and Master Students to the R&D work

Overall, 11 PhD Theses have been completed fully/partly financed by the TRUST project or otherwise in affiliation to the TRUST project. Additional 10 PhD projects are underway. These PhD projects have been/are being completed at Uppsala University (Sweden), Göttingen University (Germany), CSIC/UPC (Spain), Cambridge University (UK), Imperial College (UK) and ETH (Switzerland). In addition, 12 MSc and other comparable student projects have been prepared at Göttingen University and Imperial College.
Finally, 13 Post-doctoral fellows have been affiliated to the project, at Uppsala University, CSIC, Cambridge University and Imperial College.

Active dissemination to the scientific/technical community (EWRE, IIT, MERI)

There is a continuous dissemination of TRUST results by means of scientific publications, and presentations at key conferences (Trondheim, GHGT). Invited or contributed presentations specifically presenting the results of Heletz injection experiments during the present reporting period include the following:

• Bensabat J. Presentation of the TRUST project at the EU-Australia workshop held in Melbourne and Sydney (Australia), March 2013.

• Niemi, A., Bensabat, J. et al, (2018) Study of CO2 residual trapping in-situ – results from push-pull experiments at Heletz (Israel) pilot CO2 injection site. IEAGHG GHGT-14 Conference, 21-16th October, Melbourne, Australia.

• Niemi, A. and Bensabat, J. (2018) Progress at Heletz CO2 injection site and first results of the residual trapping experiments. Invited talk at Venice Open Forum, CO2GeoNet. 24-26 April 2018.

• Niemi, A, Bensabat, J. et al. (2017) Progress at Heletz, Israel site since 2016 and first results for evaluating the in-situ residual trapping test IEAGHG (International Energy Agency Greenhouse Gas RD Program) Monitoring Network Meeting. Traverse City, Michigan, USA. 12-14.6.2017.

We also have been invited to organize the following sessions in major international conferences:

• Organizing the Mini-symposium ‘Fundamentals of Geological Storage of CO2’ in Interpore 10th Annual Meeting and Jubilee, May 14-17th, 2018, New Orleans, USA. https://events.interpore.org/event/2/

• Co-organizing the session ‘Confrontations of models with field data: applications to CO2 storage, geothermal production and managed aquifer recharge’ in Conference Computational Methods in Water Resources XXII, June 3-7th, 2018. Saint Malo, France. http://cmwrconference.org/

Examples of direct dissemination to relevant industry includes

• Niemi, A, Bensabat, J. et al. (2018) Overview - our recent activities in R&D in Geological CO2 Storage in EU funded projects and Heletz pilot injection site. Invited presentation for Total, France. Jan 23rd 2018.

• Bensabat J., Niemi a. and J. Bear (2018): The Heletz experimental CO2 injection site and prospects for strategic storage of natural gas. The Israel Ministry of Energy.

List of Websites:
www.trust-co2.org

Final Report Summary - TRUST (High resolution monitoring, real time visualization and reliable modeling of highly controlled, intermediate and up-scalable size pilot injection tests of underground storage of CO2)

Verwandte Dokumente

Diese Seite teilen

Herunterladen