Short-Term Probabilistic Hazard Assessment in Regions of Induced Seismicity

2020 ◽  
Vol 110 (5) ◽  
pp. 2441-2453 ◽  
Author(s):  
Ganyu Teng ◽  
Jack W. Baker
Keyword(s):  
Hydraulic Fracturing ◽  
Induced Seismicity ◽  
Aftershock Sequence ◽  
Earthquake Occurrence ◽  
Induced Earthquakes ◽  
Short Term ◽  
Wastewater Disposal ◽  
West Texas ◽  
Hazard Level ◽  
Natural Earthquakes

ABSTRACT This project introduces short-term hazard assessment frameworks for regions with induced seismicity. The short-term hazard is the hazard induced during the injection for hydraulic-fracturing-induced earthquakes. For wastewater-disposal-induced earthquakes, it is the hazard within a few days after an observed earthquake. In West Texas, hydraulic-fracturing-induced earthquakes cluster around the injection activities, and the earthquake occurrence varies greatly in time and space. We develop a method to estimate the hazard level at the production site during the injection, based on past injection and earthquake records. The results suggest that the injection volume has a negligible effect on short-term earthquake occurrence in this case, because injection volumes per well fall within a relatively narrow range, whereas the regional variations in seismic productivity of wells and b-values are important. The framework could be easily modified for implementation in other regions with hydraulic-fracturing-induced earthquakes. We then compare the framework with wastewater-disposal-induced earthquakes in Oklahoma–Kansas and natural earthquakes in California. We found that drivers of short-term seismic hazard differ for the three cases. In West Texas, clustered earthquakes dominate seismic hazards near production sites. However, for Oklahoma–Kansas and California, the short-term earthquake occurrence after an observed mainshock could be well described by the mainshock–aftershock sequence. For Stillwater in Oklahoma, aftershocks contribute less to the hazard than San Francisco in California, due to the high Poissonian mainshock rate. For the rate of exceeding a modified Mercalli intensity of 3 within 7 days after an M 4 earthquake, the aftershock sequence from natural earthquakes contributed 85% of the hazard level, whereas the aftershock contribution was only 60% for induced earthquakes in Oklahoma. Although different models were implemented for hazard calculations in regions with hydraulic fracturing versus wastewater injection, injection activities could be drivers of short-term hazard in both cases.

What Governs the Spatial and Temporal Distribution of Aftershocks in Mining-Induced Seismicity: Insight into the Influence of Coseismic Static Stress Changes on Seismicity in Kiruna Mine, Sweden

2020 ◽  
Author(s):  
Maria Kozłowska ◽  
Beata Orlecka-Sikora ◽  
Savka Dineva ◽  
Łukasz Rudziński ◽  
Mirjana Boskovic
Keyword(s):  
Induced Seismicity ◽  
Temporal Distribution ◽  
Aftershock Sequence ◽  
State Model ◽  
Induced Earthquakes ◽  
Good Tool ◽  
Coulomb Stress Changes ◽  
Mining Induced Seismicity ◽  
Stress Changes ◽  
Natural Earthquakes

ABSTRACT Strong mining-induced earthquakes are often followed by aftershocks, similar to natural earthquakes. Although the magnitudes of such in-mine aftershocks are not high, they may pose a threat to mining infrastructure, production, and primarily, people working underground. The existing post-earthquake mining procedures usually do not consider any aspects of the physics of the mainshock. This work aims to estimate the rate and distribution of aftershocks following mining-induced seismic events by applying the rate-and-state model of fault friction, which is commonly used in natural earthquake studies. It was found that both the pre-mainshock level of seismicity and the coseismic stress change following the mainshock rupture have strong effects on the aftershock sequence. For mining-induced seismicity, however, we need to additionally account for the constantly changing stress state caused by the ongoing exploitation. Here, we attempt to model the aftershock sequence, its rate, and distribution of two M≈2 events in iron ore Kiruna mine, Sweden. We could appropriately estimate the aftershock sequence for one of the events because both the modeled rate and distribution of aftershocks matched the observed activity; however, the model underestimated the rate of aftershocks for the other event. The results of modeling showed that aftershocks following mining events occur in the areas of pre-mainshock activity influenced by the positive coulomb stress changes, according to the model’s assumptions. However, we also noted that some additional process not incorporated in the rate-and-state model may influence the aftershock sequence. Nevertheless, this type of modeling is a good tool for evaluating the risk areas in mines following a strong seismic event.

scholarly journals Statistical physics models for aftershocks and induced seismicity

2018 ◽  
Vol 377 (2136) ◽  
pp. 20170397 ◽  
Author(s):  
Molly Luginbuhl ◽  
John B. Rundle ◽  
Donald L. Turcotte
Keyword(s):  
Induced Seismicity ◽  
Statistical Physics ◽  
Aftershock Sequence ◽  
Induced Earthquakes ◽  
Aftershock Activity ◽  
Gas Field ◽  
Natural Time ◽  
Parkfield Earthquake ◽  
Groningen Gas Field

A standard approach to quantifying the seismic hazard is the relative intensity (RI) method. It is assumed that the rate of seismicity is constant in time and the rate of occurrence of small earthquakes is extrapolated to large earthquakes using Gutenberg–Richter scaling. We introduce nowcasting to extend RI forecasting to time-dependent seismicity, for example, during an aftershock sequence. Nowcasting uses ‘natural time’; in seismicity natural time is the event count of small earthquakes. The event count for small earthquakes is extrapolated to larger earthquakes using Gutenberg–Richter scaling. We first review the concepts of natural time and nowcasting and then illustrate seismic nowcasting with three examples. We first consider the aftershock sequence of the 2004 Parkfield earthquake on the San Andreas fault in California. Some earthquakes have higher rates of aftershock activity than other earthquakes of the same magnitude. Our approach allows the determination of the rate in real time during the aftershock sequence. We also consider two examples of induced earthquakes. Large injections of waste water from petroleum extraction have generated high rates of induced seismicity in Oklahoma. The extraction of natural gas from the Groningen gas field in The Netherlands has also generated very damaging earthquakes. In order to reduce the seismic activity, rates of injection and withdrawal have been reduced in these two cases. We show how nowcasting can be used to assess the success of these efforts. This article is part of the theme issue ‘Statistical physics of fracture and earthquakes’.

Comprehensive Characterization and Mitigation of Hydraulic Fracturing-Induced Seismicity in Fox Creek, Alberta

SPE Journal ◽  
2021 ◽  
pp. 1-12
Author(s):  
Gang Hui ◽  
Shengnan Chen ◽  
Zhangxin Chen ◽  
Fei Gu ◽  
Mathab Ghoroori ◽  
...  
Keyword(s):  
Spatial Distribution ◽  
Hydraulic Fracturing ◽  
Pore Pressure ◽  
Hydraulic Fracture ◽  
Induced Seismicity ◽  
Mitigation Strategy ◽  
Moment Magnitude ◽  
Induced Earthquakes ◽  
Data Set ◽  
Comprehensive Characterization

Summary The relationships among formation properties, fracturing operations, and induced earthquakes nucleated at distinctive moments and positions remain unclear. In this study, a complete data set on formations, seismicity, and fracturing treatments is collected in Fox Creek, Alberta, Canada. The data set is then used to characterize the induced seismicity and evaluate its susceptibility toward fracturing stimulations via integration of geology, geomechanics, and hydrology. Five mechanisms are identified to account for spatiotemporal activation of the nearby faults in Fox Creek, where all major events [with a moment magnitude (Mw) greater than 2.5] are caused by the increase in pore pressure and poroelastic stress during the fracturing operation. In addition, an integrated geological index (IGI) and a combined geomechanical index (CGI) are first proposed to indicate seismicity susceptibility, which is consistent with the spatial distribution of induced earthquakes. Finally, mitigation strategy results suggest that enlarging a hydraulic fracture-fault distance and decreasing a fracturing job size can reduce the risk of potential seismic activities.

Complex Shear-Wave Anisotropy from Induced Earthquakes in West Texas

2020 ◽  
Vol 110 (5) ◽  
pp. 2242-2251 ◽  
Author(s):  
Regan Robinson ◽  
Aibing Li ◽  
Alexandros Savvaidis ◽  
Hongru Hu
Keyword(s):  
Shear Wave ◽  
Induced Seismicity ◽  
High Rate ◽  
Normal Faults ◽  
Permian Basin ◽  
Induced Earthquakes ◽  
Large Delay ◽  
Delaware Basin ◽  
West Texas ◽  
Shallow Earthquakes

ABSTRACT We have analyzed shear-wave splitting (SWS) data from local earthquakes in the Permian basin in west Texas to understand crustal stress change and induced seismicity. Two SWS parameters, the fast polarization direction and the delay time, are computed using a semiautomatic algorithm. Most measurements are determined in the Delaware basin and the Snyder area. In both regions, SWS fast directions are mostly consistent with local SHmax at stations that are relatively far from the earthquake clusters. Varying fast directions at one station are related to different ray paths and are probably caused by heterogeneity. In the Snyder area, most northeast–southwest fast directions are from the events in the northern part of the cluster, whereas the northwest–southeast fast directions are mostly from the southern part. The northeast–southwest and northwest–southeast fast directions could be attributed to the northeast-trending normal faults and the northwest-trending strike-slip faults, respectively. SWS results in the Delaware basin have two unique features. First, most shallow earthquakes less than 4 km deep produce relatively large delay times. This observation implies that the upper crust of the Delaware basin is highly fractured, as indicated by the increasing number of induced earthquakes. Second, diverse fast directions are observed at the stations in the high-seismicity region, likely caused by the presence of multiple sets of cracks with different orientations. This situation is possible in the crust with high pore pressure, which is expected in the Delaware basin due to extensive wastewater injection and hydraulic fracturing. We propose that the diversity of SWS fast directions could be a typical phenomenon in regions with a high rate of induced seismicity.

Factors Influencing the Probability of Hydraulic Fracturing-Induced Seismicity in Oklahoma

2020 ◽  
Vol 110 (5) ◽  
pp. 2272-2282 ◽  
Author(s):  
Rosamiel Ries ◽  
Michael R. Brudzinski ◽  
Robert J. Skoumal ◽  
Brian S. Currie
Keyword(s):  
Hydraulic Fracturing ◽  
Induced Seismicity ◽  
Lower Probability ◽  
Pressure Gradients ◽  
Volume Number ◽  
Wastewater Disposal ◽  
The Past ◽  
Well Depth ◽  
Depth Relationship ◽  
Available Information

ABSTRACT Injection-induced seismicity became an important issue over the past decade, and although much of the rise in seismicity is attributed to wastewater disposal, a growing number of cases have identified hydraulic fracturing (HF) as the cause. A recent study identified regions in Oklahoma where ≥75% of seismicity from 2010 to 2016 correlated with nearly 300 HF wells. To identify factors associated with increased probability of induced seismicity, we gathered publicly available information about the HF operations in these regions including: injected volume, number of wells on a pad, injected fluid (gel vs. slickwater), vertical depth of the well, proximity of the well to basement rock, and the formation into which the injection occurred. To determine the statistical strength of the trends, we applied logistic regression, bootstrapping, and odds ratios. We see no trend with total injected volume in our Oklahoma dataset, in contrast to strong trends observed in Alberta and Texas, but we note those regions have many more multiwell pads leading to larger cumulative volumes within a localized area. We found a ∼50% lower probability of seismicity with the use of gel compared to slickwater. We found that HF wells targeting older formations had a higher probability of seismicity; however, these wells also tend to be deeper, and we found the trend with well depth to be stronger than the trend with age of formation. When isolated to the Woodford formation, well depth produced the strongest relationship, increasing from ∼5% to ∼50% probability from 1.5 to 5.5 km. However, no trend was seen in the proximity to basement parameter. Based on previously measured pore pressure gradients, we interpret the strong absolute depth relationship to be a result of the increasing formation overpressure measured in deeper portions of the basin that lower the stress change needed to induce seismicity.

A risk-based approach for managing hydraulic fracturing–induced seismicity

Science ◽  
2021 ◽  
Vol 372 (6541) ◽  
pp. 504-507
Author(s):  
Ryan Schultz ◽  
Gregory C. Beroza ◽  
William L. Ellsworth
Keyword(s):  
Hydraulic Fracturing ◽  
Induced Seismicity ◽  
Red Light ◽  
Maximum Magnitude ◽  
Traffic Light ◽  
Induced Earthquakes ◽  
Eagle Ford Shale ◽  
Eagle Ford ◽  
Sensitivity Tests ◽  
Spatially Varying

Risks from induced earthquakes are a growing concern that needs effective management. For hydraulic fracturing of the Eagle Ford shale in southern Texas, we developed a risk-informed strategy for choosing red-light thresholds that require immediate well shut-in. We used a combination of datasets to simulate spatially heterogeneous nuisance and damage impacts. Simulated impacts are greater in the northeast of the play and smaller in the southwest. This heterogeneity is driven by concentrations of population density. Spatially varying red-light thresholds normalized on these impacts [moment magnitude (Mw) 2.0 to 5.0] are fairer and safer than a single threshold applied over a broad area. Sensitivity tests indicate that the forecast maximum magnitude is the most influential parameter. Our method provides a guideline for traffic light protocols and managing induced seismicity risks.

Induced Seismicity in the Delaware Basin, West Texas, is Caused by Hydraulic Fracturing and Wastewater Disposal

2020 ◽  
Vol 110 (5) ◽  
pp. 2225-2241 ◽  
Author(s):  
Alexandros Savvaidis ◽  
Anthony Lomax ◽  
Caroline Breton
Keyword(s):  
Hydraulic Fracturing ◽  
Oil And Gas ◽  
Daily Basis ◽  
Fort Worth ◽  
Wastewater Disposal ◽  
Delaware Basin ◽  
West Texas ◽  
Oil And Gas Operations ◽  
Epicentral Location ◽  
Different Sources

ABSTRACT Most current seismicity in the southern U.S. midcontinent is related to oil and gas operations (O&G Ops). In Texas, although recorded earthquakes are of low-to-moderate magnitude, the rate of seismicity has been increasing since 2009. Because of the newly developed Texas Seismological Network, in most parts of Texas, recent seismicity is reported on a daily basis with a magnitude of completeness of ML 1.5. Also, funded research has allowed the collection of O&G Op information that can be associated with seismicity. Although in the Dallas–Fort Worth area, recent seismicity has been associated mostly with saltwater disposal (SWD), in the South Delaware Basin, West Texas, both hydraulic fracturing (HF) and SWD have been found to be causal factors. We have begun to establish an O&G Op database using four different sources—IHS, FracFocus, B3, and the Railroad Commission of Texas—with which we can associate recent seismicity to HF and SWD. Our approach is based on time and epicentral location of seismic events and time, location of HF, and SWD. Most seismicity occurs in areas of dense HF and SWD-well activity overlapping in time, making association of seismicity with a specific well type impossible. However, through examination of clustered seismicity in space and time, along with isolated clusters of spatiotemporal association between seismicity and O&G Ops, we are able to show that a causation between HF and seismicity may be favored over causation with SWD wells in areas of spatially isolated earthquake clusters (Toyah South, Reeves West, Jeff Davis Northeast, and Jeff Davis East). Causality between SWD and seismicity may be inferred for isolated cases in Reeves South and Grisham West.

Seismicity Declustering and Hazard Analysis of the Oklahoma–Kansas Region

2019 ◽  
Vol 109 (6) ◽  
pp. 2356-2366 ◽  
Author(s):  
Ganyu Teng ◽  
Jack W. Baker
Keyword(s):  
Induced Seismicity ◽  
Hazard Analysis ◽  
Aftershock Sequence ◽  
Eastern United States ◽  
Etas Model ◽  
Epidemic Type Aftershock Sequence ◽  
Window Method ◽  
Seismicity Declustering ◽  
Hazard Level ◽  
Seismicity Catalog

Abstract This study is an evaluation of the suitability of several declustering method for induced seismicity and their impacts on hazard analysis of the Oklahoma–Kansas region. We considered the methods proposed by Gardner and Knopoff (1974), Reasenberg (1985), Zaliapin and Ben‐Zion (2013), and the stochastic declustering method (Zhuang et al., 2002) based on the epidemic‐type aftershock sequence (ETAS) model (Ogata, 1988, 1998). The results show that the choice of declustering method has a significant impact on the declustered catalog and the resulting hazard analysis of the Oklahoma–Kansas region. The Gardner and Knopoff method, which is currently implemented in the U.S. Geological Survey one‐year seismic‐hazard forecast for the central and eastern United States, has unexpected features when used for this induced seismicity catalog. It removes 80% of earthquakes and fails to reflect the changes in background rates that have occurred in the past few years. This results in a slight increase in the hazard level from 2016 to 2017, despite a decrease in seismic activities in 2017. The Gardner and Knopoff method also frequently identifies aftershocks with much stronger shaking intensities than their associated mainshocks. These features are mostly due to the window method implemented in the Gardner and Knopoff method. Compared with the Gardner and Knopoff method, the other three methods are able to capture the changing hazard level in the region. However, the ETAS model potentially overestimates the foreshock effect and generates negligible probabilities of large earthquakes being mainshocks. The Reasenberg and Zaliapin and Ben‐Zion methods have similar performance on catalog declustering and hazard analysis. Compared with the ETAS method, these two methods are easier to implement and faster to generate the declustered catalog. The results from this study suggest that both Reasenberg and Zaliapin and Ben‐Zion declustering methods are suitable for declustering and hazard analysis for induced seismicity in the Oklahoma–Kansas region.

A Long-Lived Swarm of Hydraulic Fracturing-Induced Seismicity Provides Evidence for Aseismic Slip

2020 ◽  
Vol 110 (5) ◽  
pp. 2205-2215 ◽  
Author(s):  
Thomas S. Eyre ◽  
Megan Zecevic ◽  
Rebecca O. Salvage ◽  
David W. Eaton
Keyword(s):  
Hydraulic Fracturing ◽  
Pore Pressure ◽  
Induced Seismicity ◽  
Aftershock Sequence ◽  
Fluid Migration ◽  
Aseismic Slip ◽  
Seismicity Rate ◽  
Slow Earthquakes ◽  
Seismic Swarms ◽  
Event Distribution

ABSTRACT Seismic swarms are defined as an increase in seismicity that does not show a clear mainshock–aftershock sequence. Typically, swarms are primarily associated with either fluid migration or slow earthquakes (aseismic slip). In this study, we analyze a swarm induced by hydraulic fracturing (HF) that persisted for an unusually long duration of more than 10 months. Swarms ascribed to fluid injection are usually characterized by an expanding seismicity front; in this case, however, characteristics such as a relatively steady seismicity rate over time and lack of hypocenter migration cannot be readily explained by a fluid-diffusion model. Here, we show that a different model for HF-induced seismicity, wherein an unstable region of a fault is loaded by proximal, pore-pressure-driven aseismic slip, better explains our observations. According to this model, the steady seismicity rate can be explained by a steady slip velocity, while the spatial stationarity of the event distribution is due to lateral confinement of the creeping region of the fault with increased pore pressure. Our results may have important implications for other induced or natural seismic swarms, which could be similarly explained by aseismic loading of asperities driven by fluid overpressure rather than the often-attributed fluid-migration model.

Sign in / Sign up

Export Citation Format

Share Document