SUMMARY
Large-magnitude fluid-injection induced seismic events are a potential risk for geothermal energy developments worldwide. One potential risk mitigation measure is the application of cyclic ...injection schemes. After validation at small (laboratory) and meso (mine) scale, the concept has now been applied for the first time at field scale at the Pohang Enhanced Geothermal System (EGS) site in Korea.
From 7 August until 14 August 2017 a total of 1756 m³ of surface water was injected into Pohang well PX-1 at flow rates between 1 and 10 l s–1, with a maximum wellhead pressure (WHP) of 22.8 MPa, according to a site-specific cyclic soft stimulation schedule and traffic light system. A total of 52 induced microearthquakes were detected in real-time during and shortly after the injection, the largest of Mw 1.9. After that event a total of 1771 m³ of water was produced back from the well over roughly 1 month, during which time no larger-magnitude seismic event was observed. The hydraulic data set exhibits pressure-dependent injectivity increase with fracture opening between 15 and 17 MPa WHP, but no significant permanent transmissivity increase was observed.
The maximum magnitude of the induced seismicity during the stimulation period was below the target threshold of Mw 2.0 and additional knowledge about the stimulated reservoir was gained. Additionally, the technical feasibility of cyclic injection at field scale was evaluated. The major factors that limited the maximum earthquake magnitude are believed to be: limiting the injected net fluid volume, flowback after the occurrence of the largest induced seismic event, using a cyclic injection scheme, the application of a traffic light system, and including a prioriinformation from previous investigations and operations in the treatment design.
SUMMARY
We investigate attenuation (Q−1) of sediments of 2.5–3.5 km thickness underneath the city of Basel, Switzerland. We use recordings of 195 induced events that were obtained during and after ...the stimulation of a reservoir for a Deep Heat Mining Project in 2006 and 2007. The data set is ideally suited to estimate Q as all events are confined to a small source volume and were recorded by a dense surface network as well as six borehole sensors at various depths. The deepest borehole sensor is positioned at a depth of 2.7 km inside the crystalline basement at a mean hypocentral distance of 1.8 km. This allows us to measure Q for frequencies between 10 and 130 Hz. We apply two different methods to estimate Q. First, we use a standard spectral ratio technique to obtain Q, and as a second measure we estimate Q in the time domain, by convolving signals recorded by the deepest sensor with a Q operator and then comparing the convolved signals to recordings at the shallower stations. Both methods deliver comparable values for Q. We also observe similar attenuation for P‐ and S‐ waves (QP∼QS). As expected, Q increases with depth, but with values around 30–50, it is low even for the consolidated Permian and Mesozoic sediments between 500 and 2700 m.
•Coupled hydro-mechanical simulation and characterization of the cyclic soft stimulation in August 2017 at the Pohang EGS reservoir (second stimulation in well PX-1) is performed.•Pressure history ...match was achieved by splitting the simulated treatment into phases.•The permeability is increased through opening of existing fault.•The modeled extent of direct overpressure of 0.01 MPa at the end of the stimulation is approx. 180 m from the injection point.
In this study, we investigate numerically the hydro-mechanical behavior of fractured crystalline rock due to one of the five hydraulic stimulations at the Pohang Enhanced Geothermal site in South Korea. We use the commercial code FracMan (Golder Associates) that enables studying hydro-mechanical coupled processes in fractured media in three dimensions combining the finite element method with a discrete fracture network. The software is used to simulate fluid pressure perturbation at fractures during hydraulic stimulation. Our numerical simulation shows that pressure history matching can be obtained by partitioning the treatment into separate phases. This results in adjusted stress-aperture relationships. The evolution of aperture adjustment implies that the stimulation mechanism could be a combination of hydraulic fracturing and shearing. The simulated extent of the 0.01 MPa overpressure contour at the end of the treatment equals to ∼180 m around the injection point.
We investigate attenuation (Q-1) of sediments of 2.5-3.5km thickness underneath the city of Basel, Switzerland. We use recordings of 195 induced events that were obtained during and after the ...stimulation of a reservoir for a Deep Heat Mining Project in 2006 and 2007. The data set is ideally suited to estimate Q as all events are confined to a small source volume and were recorded by a dense surface network as well as six borehole sensors at various depths. The deepest borehole sensor is positioned at a depth of 2.7km inside the crystalline basement at a mean hypocentral distance of 1.8km. This allows us to measure Q for frequencies between 10 and 130 Hz. We apply two different methods to estimate Q. First, we use a standard spectral ratio technique to obtain Q, and as a second measure we estimate Q in the time domain, by convolving signals recorded by the deepest sensor with a Q operator and then comparing the convolved signals to recordings at the shallower stations. Both methods deliver comparable values for Q. We also observe similar attenuation for P- and S- waves (QP similar to QS). As expected, Q increases with depth, but with values around 30-50, it is low even for the consolidated Permian and Mesozoic sediments between 500 and 2700 m.
The history of reservoir stimulation to extract geothermal energy from low permeability rock (i.e. so-called petrothermal or engineered geothermal systems, EGS) highlights the difficulty of creating ...fluid pathways between boreholes, while keeping induced seismicity at an acceptable level. The worldwide research community sees great value in addressing many of the unresolved problems in down-scaled in-situ hydraulic stimulation experiments. Here, we present the rationale, concepts and initial results of stimulation experiments in two underground laboratories in the crystalline rocks of the Swiss Alps. A first experiment series at the 10 m scale was completed in 2017 at the Grimsel Test Site, GTS. Observations of permeability enhancement and induced seismicity show great variability between stimulation experiments in a small rock mass body. Monitoring data give detailed insights into the complexity of fault stimulation induced by highly heterogeneous pressure propagation, the formation of new fractures and stress redistribution. Future experiments at the Bedretto Underground Laboratory for Geoenergies, BULG, are planned to be at the 100 m scale, closer to conditions of actual EGS projects, and a step closer towards combining fundamental process-oriented research with testing techniques proposed by industry partners. Thus, effective and safe hydraulic stimulation approaches can be developed and tested, which should ultimately lead to an improved acceptance of EGS.
•Problems in EGS and induced seismicity targeted in scaled underground experiments.•Stimulation experiments show large variability within a 10 m scale rock volume.•Stimulations with acceptable induced seismicity are tested on the 100 m scale.
The increased interest in subsurface development (e.g., unconventional
hydrocarbon, engineered geothermal systems (EGSs), waste disposal) and the
associated (triggered or induced) seismicity calls ...for a better
understanding of the hydro-seismo-mechanical coupling in fractured rock
masses. Being able to bridge the knowledge gap between laboratory and
reservoir scales, controllable meso-scale in situ experiments are deemed
indispensable. In an effort to access and instrument rock masses of
hectometer size, the Bedretto Underground Laboratory for Geosciences and
Geoenergies (“BedrettoLab”) was established in 2018 in the existing
Bedretto Tunnel (Ticino, Switzerland), with an average overburden of 1000 m.
In this paper, we introduce the BedrettoLab, its general setting and
current status. Combined geological, geomechanical and geophysical methods
were employed in a hectometer-scale rock mass explored by several boreholes
to characterize the in situ conditions and internal structures of the rock
volume. The rock volume features three distinct units, with the middle fault
zone sandwiched by two relatively intact units. The middle fault zone unit
appears to be a representative feature of the site, as similar structures
repeat every several hundreds of meters along the tunnel. The lithological
variations across the characterization boreholes manifest the complexity and
heterogeneity of the rock volume and are accompanied by compartmentalized
hydrostructures and significant stress rotations. With this complexity, the
characterized rock volume is considered characteristic of the heterogeneity
that is typically encountered in subsurface exploration and development. The
BedrettoLab can adequately serve as a test-bed that allows for in-depth
study of the hydro-seismo-mechanical response of fractured crystalline rock
masses.
PDF
