EAGE/SEG Research Workshop 2017

The Containment and Monitoring Institute (CaMI) of CMC Research Institutes, Inc. and the University of Calgary have constructed a Field Research Station for research into monitoring technologies for containment and conformance of subsurface fluids, particularly CO2. The field research station is a platform for development and performance validation of technologies intended for measurement, monitoring and verification of CO2 storage. Outcomes of the research will also have adjacent applications such as assessing and monitoring cap rock integrity in oil sands during production, assessing fugitive emissions in shale gas production and monitoring CO2-enhanced oil recovery programs. The project will undertake applied research into monitoring technologies, graduate student and industry professional training, as well as public outreach and engagement activities.

Repeated site survey data acquired before and after an underground blowout that occured when exploration well 2/4-14 was drilled in 1989 are used to map the migration of gas from the deep reservoir into shallow sediments. We use 4D difference sections to map the extent and vertical position of gas being trapped in various overburden sand layers. A dedicated sand tank experiment is conducted to improve our understanding of gas flow through the uppermost layer situated 100 m below seabed. This laboratory flow tests are compared to videfilms acquired close to the well.

The Containment and Monitoring Institute (CaMI) has built a Field Research Station (FRS) near Calgary, Alberta, Canada, designed around small injections (up to 1000 tonnes per year) of CO2 at depths of approximately 300 m and 500 m. Baseline borehole geophysical surveys have been completed at the CaMI site as part of a larger monitoring effort aimed at detecting and monitoring the small (1-2 ton) gas phase CO2 plume as well as any vertical migration. Crosswell seismic and EM data, as well as borehole-to-surface EM data, were recorded and have sufficient data quality to proceed with time-lapse monitoring. Initial processing for quality control has been accomplished. Seismic processing to date includes initial travel time picking and indicates that data is sufficient for velocity inversion. EM processing includes comparison of acquired data (amplitude and phase) with a layered model based on resistivity logs with repeatability of ~2%. The use of fiberglass casing in one monitoring well supports the use of EM methods. Joint inversion of the seismic and EM measurements (i.e. P-wave velocity and electrical conductivity) is planned to improve the estimate of CO2 saturation in the subsurface.

Whenever fluids are injected or extracted from a reservoir, pore pressure and stress field change both in the reservoir and in the overburden. According to geomechanics, such variation in stress or pore pressure can produce a reservoir compaction, trigger pre-existing faults or even generate new ones. Al these events will translate into surface deformation, which turns to be a valuable information to better understand the subsurface phenomena. In the last few years, the analysis of multi-temporal SAR data sets represents an important layer of information, for reservoir monitoring and management, thanks to the possibility to provide surface deformation data with millimetre precision over large areas. In order to help decision makers in analyzing different scenarios and planning specific actions it is important to update ground displacement measurements every new satellite acquisition.

CO2CRC Otway project was the first Australian of CO2 geosequestration project. The project site is located 240 km away from Melbourne, Victoria. During the Stage 1 of this project ~66 thousand tons of supercritical CO2/CH4 gas mixture was injected into a depleted gas reservoir at approximately 2 km depth in 2008-2010 to prove that gas can be safely transported and stored in a geological formation. The ongoing Stage 2C of the project is focusing seismic monitoring capabilities and, also, on proving that the injected gas plume plume will stabilize over the period of time. A very limited amount of the same gas mixture (15 000 tonnes) was injected into a saline aquifer at ~1500 m. In order to monitor the gas injection a comprehensive 4D seismic program was rolled out. In this presentation we outline the monitoring program and show the time-lapse seismic results obtained after the first four monitor surveys, acquired at 5 000 t, 10 000 t, 15 000 t of the injection and, also, 9 months after completion of the injection.

The UK GeoEnergy Test Bed (GTB) is a field laboratory founded by the University of Nottingham and The British Geological Survey (BGS) for subsurface research into fluid migration through natural pathways. The GTB was initiated with funding from the founding partners and has received capital investment from the UK Government’s Treasury as part of the Energy Research Accelerator (ERA) project. The GTB will act as a research hub to catalyse scientific collaboration between researchers from academic and industrial backgrounds. The aim of the GTB is to advance geoscientific research to support the global need for secure, sustainable and safe energy. Geophysical, geochemical, petrochemical and hydrogeological data have been collected during site characterization and used to build a conceptual geological model. Permission is being sought from the relevant authorities to inject a small amount of food-grade CO2. Surface and downhole technologies and techniques will be used to monitor site behaviour before, during and after fluid injection. The sensing capability at the GTB has been designed by the BGS to enable interrogation of the subsurface in 4D and to a level of detail commensurate with energy processes in the subsurface.

Full Waveform Inversion (FWI) and Controlled Source Electro-Magnetic (CSEM) inversion can provide quantitative estimates of useful subsurface properties. In the context of CO2 monitoring, FWI has the capability to characterize velocity changes induced by CO2 injection, while CSEM can be useful to determine the CO2 saturation. Quantifying the uncertainty of the reconstructed parameters is however challenging. In this study, we assess the reliability of the results obtained using CSEM and FWI when monitoring CO2 storage at the Sleipner pilot in the North Sea. The uncertainty evaluation methodology is described in a first stage. In a second stage, two CO2 monitoring examples are presented, one synthetic case using CSEM and one real data case using FWI. For the synthetic CSEM case, the inversion of the EM data has a clear effect on the uncertainty, reducing the covariance of the probability distribution for possible conductivity models. For the real, rather large-scale FWI case, computationally efficient strategies are considered. The covariance is not reduced to the same extent as for the CSEM case, but an improvement is observed. In both cases, the quantification approach provides a means to assess the quality of the derived models.

Background studies are made prior to the beginning of the injection to ensure the security of the geological sequestration of CO2. In the CaMI.FRS site near Brooks, Alberta, numerous wells give the information about the lithology, the porosity, the permeability, the velocities (and others parameters) of the medium. In addition to logs data, seismic surveys were conducted in order to characterize the subsurface. After acquiring well logs information and baseline seismic surveys, we applied numerical simulations in order to characterize the feasibility of the time-lapse seismic monitoring. Indeed, once the injection begins, seismic survey will be made at regular intervals to monitor the CO2 injection. Fluid simulations allow us to work on synthetic models, but yet are close to what it is expected in the reality. We use Gassmann fluid substitution to obtain the elastic parameters (VP, VS and density) at different injection times (1 year after the beginning of the injection and 1 year after the end of the injection), for a 300m depth CO2 reservoir. In those 3D models, synthetic data are generated then processed. This work give us a good approximation of the feasibility of a time-lapse seismic monitoring, considering the conditions of CaMI.FRS project.

An ultra-high-resolution mini 3D survey was acquired at the Otway site to characterise structures and discontinuities in a depth range of 0-100 m for the controlled CO2 fault plane release experiment. The survey demonstrated that very high spatial data density is essential for imaging and characterisation of shallow structures. Nevertheless, significant processing efforts were required to precisely image shallow fault geometry and the extent of fault tips. The main fault F1 and the accompanied discontinuities were clearly imaged with a PSDM routine after a 3D velocity model was derived from 3D tomography and then further improved through an iterative process. Fault F1 now appears more complex and additional studies are needed. Further assessment of the fault F1 and its associated discontinuities will be conducted using three component seismic data.

Time-lapse seismic in a desert environment is a major geophysical challenge due to the complex and changing nature of the near surface that inhibits both imaging and data repeatability. In this study, the problem is further compounded by the injection of CO2 into a stiff carbonate reservoir which results in small changes in reservoir acoustic impedance. Therefore, a highly repeatable acquisition system was required to detect these weak 4D signals. A hybrid system using deep buried receivers and surface vibroseis sources was found to provide the optimum solution. Along with careful survey design and acquisition, a specialized data processing workflow was developed to overcome the challenges faced by single source, single receiver data. In particular, the stacking of neighboring shots in a process known as supergrouping is essential for the reduction of 4D noise. Using this system, highly repeatable data has been achieved with mean NRMS of 6% between dry season surveys separated by more than one year. This level of repeatability allows small variations that may be related to CO2 injection to be observed.