- Home
- Conferences
- Conference Proceedings
- Conferences
79th EAGE Conference and Exhibition 2017 - Workshops
- Conference date: 12 Jun 2017 - 15 Jun 2017
- Location: Paris, France
- ISBN: 978-94-6282-219-1
- Published: 12 June 2017
21 - 40 of 144 results
-
-
How widespread is induced seismicity in Canada and the USA?
Authors M. van der Baan and F. CalixtoThe seismicity rate in Oklahoma in the last 5-8 years is correlated to increased large-scale hydrocarbon production. Contrary to Oklahoma, analysis of oil and gas production versus seismicity rates in 6 other States in the USA and 3 provinces in Canada finds no State/Province-wide correlation between increased seismicity and hydrocarbon production, despite 8-16 fold increases in production in some States, including North Dakota (Bakken formation) and Pennsylvania, West Virginia (Marcellus shale). However, in various areas, seismicity rates have increased locally.
-
-
-
Case study: Lacq pilot CO2 storage in old gas field - microseismic monitoring
Authors J. Barnavol and X. Payreor reservoir seal integrity and supply all seismological metrics and attributes necessary for seismic risk management and population information. During the monitoring (2009–2015), 2637 events have been detected and 717 were located inside the surveillance perimeter. Magnitude of events ranged between -2.6 and +1.1 in two distinct clusters: one in the vicinity of the injection point (reservoir perimeter) and another one close to the “Meillon/Saint-Faust” fault complex, located 2km north of the injection site (local perimeter) (figure below). The seismic monitoring on this CO2 storage pilot shows a strong influence of the injection in the reservoir perimeter (≈1 km around the injection point). At a larger scale (local perimeter), this study did not enable to assess reliably eventual seismicity rate changes during injection operations. Nevertheless, this “multi-scale” seismic network design perfectly achieves the goals assigned in terms of risk management, population information and injection mapping.
-
-
-
Dynamics of Fault Activation by Hydraulic Fracturing in Overpressured Shales
By D. EatonFluid-injection processes can induce earthquakes by increasing pore pressure and/or shear stress on faults. Natural processes, including transformation of organic material (kerogen) into hydrocarbon, can similarly cause fluid overpressure. Here we document examples where earthquakes induced by hydraulic fracturing are strongly clustered within areas characterized by pore-pressure gradient in excess of 15 kPa/m. By contrast, induced earthquakes are virtually absent in the same formations elsewhere. Monte Carlo analysis indicates that there is negligible probability that this spatial correlation developed by chance. A detailed analysis was undertaken within a region in Alberta, Canada where uniquely comprehensive data characterize dynamic interactions between seismicity and well completions. Seismicity is strongly clustered in space and time, exhibiting spatially varying persistence and activation threshold. The largest event (ML 4.4) can be reconciled with a previously postulated upper bound on magnitude, only if the cumulative effect of multiple treatment stages is considered. Induced seismicity from hydraulic fracturing reveals contrasting signatures of fault activation by stress effects and fluid diffusion. Patterns of seismicity indicate that stress changes during operations can activate fault slip to an offset distance of > 1 km, whereas pressurization by hydraulic fracturing into a fault yields episodic seismicity that can persist for months.
-
-
-
Effective stress drop of fluid induced seismicity
Authors T. Fischer and S. HainzlIn this paper we test how the effective stress drop is comparable to the static stress drop of a single earthquake rupturing the same fault portion. To this purpose, we compare the spatiotemporal evolution of the seismic moment release and analyze the uncertainties of the resulting stress drop estimates. We show that the effective stress drop is only comparable to earthquake stress drops in specific cases. In particular, the effective stress drop values significantly underestimate the earthquake stress drops in the presence of aseismic deformation. Furthermore, the values are only scale-independent if pre-stress and post-stress conditions are uniform in space. Our analysis of data from injection-induced seismicity, natural earthquake swarms and aftershock sequences shows that in most cases the effective stress drop estimate is rather stable during the cluster evolution. Slightly increasing estimates for injection-induced seismicity are indicative for the local forcing of the system, while overall low effective stress drop values hint to the important role of aseismic deformations. While normal values up to 1MPa are found for seismicity associating geothermal reservoirs stimulation, anomalous small effective stress drop occur in case of fracking tight sands and shales, which may indicate aseismic deformation during these treatments.
-
-
-
Scaling of induced seismicity, implications for the role of geological setting on seismic hazard
Authors G. Viegas, A. Baig and T. UrbancicIn this study we present conflicting stress drop estimates of injection-induced events in two regions of the Western Canadian Sedimentary Basin. Horn River Basin events show lower stress drops than Duvernay Basin events by a factor of 10 to 20. We propose that the observed stress drop differences are caused by different regional stress characteristics, assuming the seismic events are generated during similar injection programs. A potential difference the Fox Creek region is characterized by the presence of reefs in the Leduc formation that cross-cut the Duvernay shale formations which form drapes over the reef-off reef facies (Stoakes, 1980). We suggest that differences in stress drops reflect differences in the regional stress state with events occurring in more stressed regions having higher stress drops, that is, being able to release larger quantities of stored elastic strain. Higher stress drop earthquakes have a significant role in seismic hazard as they generate higher frequency strong ground motions which can potentially cause more damages.
-
-
-
Mechanisms driving earthquake faulting during a case of injection-induced seismicity
Authors M. Diez and R. del PotroInduced seismicity is currently one of the main geomechanical and environmental challenges faced by the underground industry. Current efforts focus mainly on seismic monitoring and risk assessments while progress on unraveling the mechanisms that drive induced seismicity have been addressed to a lesser extent. Here we explore stability conditions for unstable slip, and potential dynamic weakening mechanisms to explain earthquake faulting in a case of injection-induced seismicity. Injection of natural gas into the Castor Underground Gas Storage, offshore Spain, which generated a ~2 bar pressure increase, induced a seismic swarm that culminated in a series of Mw~4 earthquakes, two weeks after shut in. We focus our attention on frictional weakening and on the dynamic weakening effect of shear heating-induced thermal pressurization to explain Mw >3 earthquake faulting during the Castor sequence. These new mechanisms that we explore may help improve our understanding of cases of injection-induced seismicity, in regions of low natural seismicity, where the external forcing, or amplitude of the stress perturbation is relatively small.
-
-
-
Verification of Network Design for Induced Seismicity
More LessInduced seismicity monitoring for hydrocarbon or geothermal energy extraction is usually designed to meet political or environmental goals and limitation (e.g. UK limitations on magnitude of completeness thresholds). Therefore operators seek monitoring network designs to meet or exceed these goal or limitations. Generally, before starting any microseismic monitoring the geometry of array has to be designed to achieve optimal performance of the network and to fulfil all demands for obtaining of the required data quality. Design of the monitoring array should follow several rules. For surface monitoring networks, a proper detection and location of events requires station spacing approximately twice as large as the expected depth of the seismic events. In addition, seismic noise in some area may significantly decrease the monitoring network performance. Last but not least the monitoring network performance is dependent on the assumed velocity and attenuation models. All of these factors significantly affect the network performance and it is a challenge to verify that the proposed performance is going to be real. Additionally, network performance in seismically quiet areas is particularly challenging as there is no seismicity to benchmark it on before the start of the operations (e.g. Gaucher, 2016).
-
-
-
Examining the capability of statistical models to mitigate induced seismicity during hydraulic fracturing of shale gas reservoirs
Authors J. Verdon and J. KendallIn this paper we test the ability of statistical methods to estimate the expected size of the largest event during stimulation, applying these approaches to two datasets collected during hydraulic stimulation of a North American Devonian Shale. We apply these methods in a prospective manner: using the microseismicity recorded during the early phases of a stimulation stage to make forecasts about what will happen as the stage continues. We do so to put ourselves in the shoes of an operator or regulator, where decisions must be taken based on data as it is acquired, rather that a post hoc analysis once a stimulation stage has been completed. We find that the proposed methods are able to provide a reasonable forecast of the largest event to occur during each stage. This means that these methods can be used as the basis of a mitigation strategy. Applying such a strategy to our case studies, we find that the majority of stages would have been allowed to continue as planned, while that the need for mitigation would have been identified for all of the stages that ended up inducing larger events.
-
-
-
Microseismic Geomechanical Evaluation of Fault Slip Associated with Hydraulic Fracturing
By S. MaxwellThe paper describes applying a coupled hydraulic-mechanical model and the resulting seismicity catalogue to explore repeated fault activation and seismicity patterns associated with multi-stage hydraulic fracturing. The model is also used to explore seismic risk mitigation by changing the viscosity of the fracturing fluid.
-
-
-
Integration of geomechanical modeling with induced seismic source mechanisms to assess deformation and stress changes
Authors D. Angus, G. Viegas, T. Urbancic and A. BaigThe recording of induced seismicity plays a significant role in monitoring geo-industrial activities in terms of providing improved understanding of the failure mechanisms and as well as quantitative tool for risk assessment. The induced seismicity is due to local perturbations to the in situ stress field from such industrial activities. Significant work is being directed at developing fracture models that enable the assessment of local deformation and stress evolution due to stimulation design. Passive seismic monitoring provides an additional data source to study the stress field evolution by imaging the in situ response of the rock mass over and characterizing the failure mechanisms and high-order inelastic response of the rock mass. We integrate microseismic data with a geomechanical model to quantify deformation and stress field evolution. Our approach utilizes moment tensor solutions to represent localized discrete rupture zones. The geomechanical algorithm evaluates the Green’s functions for each rupture, and subsequently calculates the co- and post-seismic deformation using linear superposition. We perform this work flow on a passive seismic dataset to validate the technique as well as provide insight into the observed seismicity response due to a large (~M5) event.
-
-
-
Production Induced Seismicity in the Netherlands - From quick-scan to advanced models
More LessApproximately one out of six producing onshore gas fields (both the Groningen Field and small fields in the north-western part of the country) in the Netherlands experiences production-induced seismicity. Poro-elasticity and related differential compaction were referred to as the prime mechanisms causing this seismicity. The observed onset of induced seismicity in the Netherlands occurred after a considerable pressure drop in the gas fields. A large range of methods has been applied to study the background of Dutch seismicity, ranging from quick scans to advanced 3D geomechanical modelling studies. We have shown that both simplified 2D and full-field 3D geomechanical models can be used to model the onset of reactivation and identify faults which are prone to be reactivated. One of the approaches we used was inclusion of dynamic rupture modelling in traditional geomechanics workflows. In dynamic rupture modelling, traditional static friction laws are displaced by dynamic slip evolution. This enabled us to include realistic nucleation and propagation of the seismic events, as e.g. extend of the rupture area and slip displacements could be determined. We present an overview of the different methods that have been used to better understand induced seismicity in the Dutch onshore gas fields, including our dynamic rupture modelling method and an outlook to what this could mean in the future.
-
-
-
Numerical modelling of production-induced stress changes and fault reactivation in Rotliegend gas fields of the North German Basin
Authors C. Haug, A. Henk and J. NüchterProduction-induced seismicity is an increasing challenge to the E&P industry. Related poroelastic stress changes in reservoirs with complex geometries, their interference with tectonic stresses and their interaction with faults cannot be sufficiently explained by analytical models. Here, we develop generic numerical models to study production-induced stress changes and fault reactivation in and near compartmentalized gas reservoirs. The models are inspired by features of Rotliegend gas fields of the Northern German Basin but do not describe a specific reservoir. In a linear-elastic model series I, stress changes during pore pressure drawdown are investigated for different model parameters. Field properties leading to an increased tendency of fault reactivation are, among others, a high Biot-Willis coefficient, a locally reduced overburden load and a large reservoir thickness. In model series II a contact surface pair simulating a simplified fault is incorporated and the mechanical response of the contact surface to different schemes for absolute stress developments are simulated. For high friction coefficients the contact surface stays stable with production while for µ<0.6 the contact surface slips after a stage of poroelastic stress increase. Modelling results provide insight into the mechanisms that control production-induced poroelastic stress changes in compartmentalized reservoirs with complex geometries.
-
-
-
Pore-scale Processes in Amott Spontaneous Imbibition Tests
Authors M. Rücker, W.B. Bartels, M.A. Boone, T. Bultreys, H. Mahani, S. Berg, A. Georgiadis, S.M. Hassanizadeh and V. CnuddeWe observed the redistribution of the oil phase in the pore space of the rock in real-time in water-wet and mixed-wet (by ageing in crude oil) carbonate samples. During the imbibition of the water phase both, pore filling events with connection to the surrounding brine as well as snap-off events connected through water films only were detected. The distribution of the oil in different pore sizes as well as the different event types help to identify the wettability state of the system and understand how pore scale processes lead to the oil production at the larger scale.
-
-
-
Differential imaging of porous plate capillary drainage in laminated sandstone rock using X-ray micro-tomography
Authors Q. Lin, B. Bijeljic, H. Rieke and M. BluntThe experimental determination of representative capillary pressure curves as a function of saturation is of utmost importance for the determination of the initial reservoir fluid distribution and subsequent flow properties under production. We design an experimental procedure to image porous plate capillary drainage using X-ray micro-tomography based on differential imaging for a laminated sandstone micro core (4.86 mm in diameter). The pore structure, including the sub-resolution micro pores, was characterised and quantified using both the initial dry scan and the scan fully saturated with Potassium Iodide (KI) doped brine (30 wt%). During the porous plate capillary drainage, nitrogen (N2) was injected at a constant pressure and the capillary pressure was controlled by the pressure drop through the core sample. A full range of capillary pressure curve against saturation from 0 to 1.17 MPa is provided from the image analysis and is compared with the Special Core Analysis (SCAL) for the original core (35 mm in diameter). We are also able to discern that brine remained predominantly within the sub-resolution micro-pores, such as regions of fine lamination. Moreover, brine covering the rock grain surface and in the corners of the macro-pores can also be visualised.
-
-
-
Estimation of source time functions and yields of explosions directly from seismograms using the cube-root scaling law
More LessI estimate the source time functions and yields of explosions directly from seismograms. The method requires seismograms at a single receiver for two events of different size at the same source location and eliminates the path effect between source and receiver by finding a ratio filter that shapes the seismogram of the smaller event to the seismogram of the larger. If the noise is small, the convolution of the filter with the source time function of the smaller event yields the source time function of the larger event. The two source time functions are also related by the well-known scaling law in which the injected volume is proportional to the yield and the time constant is proportional to the cube-root of the yield. These two independent equations are solved for the two source time functions by a trial-and-error method that gives the ratio of the yields. The two seismograms are then deconvolved to recover two estimates of the Green’s function. Applying the method to seismograms from the 2009 and 2013 North Korean underground nuclear tests gives yields of 6 and 16 kt, respectively. This method has applications in seismic exploration on land using a dynamite source.
-
-
-
Passive seismic imaging of structure discontinuities around the active fault using scattered earthquake waveforms
More LessFor seismically active fault with relative dense seismic network, seismic tomography using first arrival times from earthquakes is routinely applied to image fault zone structures. To characterize structure discontinuities around the active fault, seismic scattering imaging using scattered waves can be applied. This technique has been extensively used with active seismic sources, especially in oil/gas industry. However, scattering imaging using waveforms from earthquakes is rare. Here we present the imaging results using scattered SH waves from earthquakes around the SAFOD, California. Near vertical reflectors are clearly imaged around the San Andreas fault (SAF), similar to results obtained using scattered P-P waves. However, for strike-slip focal mechanisms for earthquakes along the SAF, first arrival polarities for P waves are different at different regions. In comparison, SH waves have the same polarities. Because wavelength for P waves is longer than S waves, making the resolution of the imaging result is higher when using scattered S waves. For this study, we found coherent phases (scattered SH waves) after direct SH waves. Overall, the imaging results using different types of waves are bery similar, thus supporting the reliability and usefulness of using passive seismic events to image structure discontinuities.
-
-
-
A unified inversion scheme for diffractions and passive events
Authors B. Schwarz, A. Bauer and D. GajewskiWe present a unified inversion scheme that can likewise be applied to passive events and different active-source diffraction data configurations. For the active case, we specifically target the weak but highly illuminating diffracted background through automated adaptive subtraction of the dominant reflected contributions, while only making use of the available near-offset channel. Based on local stacking and coherence evaluation, the recorded diffracted events are treated as passive source wavefields, which are characterized in terms of local properties of wavefronts emerging at the registration surface. In an industrial field data example offshore Israel, we show that the diffraction-based wavefront inversion of only the near-offset channel leads to results, which are in reasonable agreement with available geological interpretations. By application to an academic low-fold dataset recorded near Santorini, we demonstrate that the inversion scheme offers the opportunity to construct laterally resolved depth-velocity models even in the absence of large-offset recordings. In addition, we suggest a simple fully data-driven strategy to globally link the independently performed local coherence measurements that share the same origin in depth. Through this global characterization of events, we motivate to perform event-consistent statistics, which allow to estimate mean scattering or passive source locations and conveniently asses location uncertainties.
-
-
-
Wave field inversion of ambient seismic noise
Authors S.A.L. de Ridder and J.R. MaddisonWe formulate a full wave field inversion for ambient seismic noise recorded by large and dense seismograph arrays. Full wave field inversion exploits the constraints on the gradients of the wave field that array data inherently possess (the sum is greater than its parts). Consequently, we can relax the spatial and temporal constraints on the wave field source functions in the seismological inverse problem. The result is that we become insensitive to the noise sources in, and changing character of, the ambient seismic field. In principle the formulation holds equally for ambient noise wave fields and for wave fields excited by controlled sources. We support the theory with examples in one dimension in the time domain, and in two dimensions in the frequency domain. The latter being of interest to invert surface wave ambient noise for phase velocity maps. We include checker board tests for geometries mimicking USArray and Ocean Bottom Cables.
-
-
-
-
-
Imaging seismic anisotropy in a shale gas reservoir by combining microseismic and 3D surface reflection seismic data
Authors W. Gajek, J. Verdon, M. Malinowski and J. TrojanowskiA strong VTI fabric can dominate the influence of weaker azimuthal anisotropy on seismic wave propagation, causing ambiguities, making it challenging to invert geophysical observations for fracture orientations and densities. We employ SWS technique for a microseismic dataset collected in a vertical borehole during a hydraulic stimulation of a shale gas target in northern Poland to image the fracture strike and density masked by a strong VTI signature. In order to overcome VTI fabric influence and enhance the inversion stability we integrate SWS data with parameters obtained from the surface 3D seismic survey. We succesfully image a pre-existing vertical fracture set not aligned with the maximum horizontal stress. The obtained results are consistent with fracture strike and crack density interpreted from well logs data.
-