OR/17/033 Appendix 3 - Publication summaries
| Baptie, B. 2017. Earthquake seismology 2016/2017 - BGS seismic monitoring and information service. British Geological Survey Internal Report, OR/17/033. |
Operational Seismic Monitoring During Hydraulic Fracturing
Baptie, B. 2017
This report was produced by the British Geological Survey (BGS) at the request of the Oil and Gas Authority (OGA) to provide an overview of the requirements for local seismic monitoring required for unconventional oil and gas activities in UK in order to comply with existing regulations.
Baseline Seismic Monitoring
Baptie, B, and Horleston, A. 2017
In this report we discuss some of the guiding principles for baseline seismic monitoring using a network of seismic sensors. These include; the design and installation of a network of sensors to ensure reliable detection and location of seismic activity in the area of interest; duration of monitoring and its dependence on background earthquake activity rates.
Unconventional Oil and Gas Development: Understanding and Monitoring Induced Seismic Activity
Baptie, B, Segou, M, Ellen, R, and Monaghan, A. 2016
Scotland is characterised by low levels of earthquake activity and the risk of damaging earthquakes is low. The largest recorded earthquake in Scotland had a magnitude of 5.2 ML with only two other earthquakes of 5 ML or greater in the last 400 years. Most earthquake activity in Scotland is north of the Highland Boundary Fault, on the west side of mainland Scotland, with less activity in northern and eastern Scotland. Earthquake activity in the Midland Valley of Scotland is also lower and in the 1970’s to 1990’s was mostly induced by coal-mining. On average there are eight earthquakes with a magnitude of 2 ML or above in Scotland every year.
Catalogues of earthquake activity in Scotland are incomplete at magnitudes below 2 ML and for higher magnitudes prior to 1970. This is due to the detection capability of the seismometer networks. This limits identification of areas that might present an elevated seismic hazard for Unconventional Oil and Gas (UOG) operations. Limited information on the stress in the Earth’s Crust mean that it is not possible to identify areas where faults are more likely to be reactivated.
Hydraulic fracturing to recover hydrocarbons is generally accompanied by earthquakes with magnitudes of less than 2 ML that are too small to be felt. In the United States, the large number of hydraulic fracturing operations (1.8 million) and the small number of felt earthquakes directly linked to them (3) suggests that the probability of induced earthquakes that can be felt is small. In western Canada, the increase in earthquakes over the last ten years corresponds to the increase in hydraulic fracturing, suggesting an increase in induced earthquakes. There have also been a number of induced earthquakes with magnitudes larger than 3 in Canada, including a magnitude 4.4, which is the largest earthquake linked to hydraulic fracturing in the world. However, as in the US, the probability of induced earthquakes that can be felt appears small given the large number of hydraulically fractured wells (>12 000).
In the UK, regulatory measures for the mitigation of induced seismicity (DECC, 2013) include: avoiding faults during hydraulic fracturing; assessing baseline earthquake activity; monitoring seismic activity during and after fracturing; and a ‘traffic light’ system to control injection. These are similar to regulatory measures that are in place in the US and Canada. In the UK, the magnitude limit for hydraulic fracturing operations (0.5 ML) is considerably lower than California (2.7 ML) and Illinois, Alberta and British Columbia (4.0 ML) and improved monitoring of seismicity will be required to implement the UK limit.
British Standards define limits for ground vibrations caused by blasting and quarrying above which cosmetic damage could take place. Modelling of ground motions for a range of earthquake magnitudes suggests that those with magnitudes of 3 or less are unlikely to exceed the limits for cosmetic damage except at distances less than a few kilometres.
Improved understanding of the hazard from induced earthquakes and the successful implementation of mitigation measures requires additional data from a number of sources:
- Improved monitoring and higher quality earthquake catalogues. Data should be openly available to maintain public confidence.
- Geological and geophysical data to map sub-surface faults in high resolution, measurements of the stress field and hydrological properties of the sub-surface.
- Industrial data from hydraulic fracturing operations.
Unconventional Gas Exploration and Extraction: Baseline Characterisation of Seismicity
Baptie, B, Jordan, C, Mosca, I, Cigna, F, Burke, S, McCloskey, J, Nic Bhloscaidh, M, Bean, C, and Möllhoff, M. 2016
This assessment of the potential risk of seismic activity induced by UGEE operations has examined international experience of such induced activity, natural seismic activity in the island of Ireland, methodologies for monitoring distortion of the surface and of background and induced seismic activity, and developed techniques for predicting induced seismicity. There is general consensus that UGEE operations can result in low magnitude seismic activity from the hydraulic fracturing process but that these events are unlikely to cause damage or even be felt. Larger events could occur if slip on existing faults is initiated, but again this is considered to be high unlikely in Ireland where the available data indicates the rate of natural seismicity to be extremely low. A greater risk is perceived through injection of high volumes of wastewater that might result from UGEE operations and so any such proposals should be examined in detail in the context of the local site geology. Modelling techniques developed by this project offer potential to predict earthquake activity, including fracture lengths, but better baseline data on the geological structure of the study areas and background seismicity is required to provide input parameters for the models. Using conservative assumptions, the modelling demonstrated that fracture lengths from hydraulic fracturing are relatively short and extremely unlikely to exceed 500 m; as a consequence, pollution of aquifers would not occur by movement of pollutants along fracture paths as long as the separation between the fracture zone and the aquifer exceeds this distance. Detailed seismic monitoring would be required during any UGEE operations and linked to a traffic light system implemented to control operations should seismic activity occur.
Local magnitude discrepancies for near‐event receivers: implications for the UK. Traffic‐Light Scheme.
Butcher, A, Luckett, R, Verdon, J P, Kendall, J‐M, Baptie, B, and Wookey, J. 2017.
Local seismic magnitudes provide a practical and efficient scale for the implementation of regulation designed to manage the risk of induced seismicity, such as Traffic‐Light Schemes (TLS). We demonstrate that significant magnitude discrepancies (up to a unit higher) occur between seismic events recorded on nearby stations (<5 km) compared with those at greater distances. This is due to the influence of sedimentary layers, which are generally lower in velocity and more attenuating than the underlying crystalline basement rocks, and requires a change in the attenuation term of the ML scale. This has a significant impact on the United Kingdom’s (UK) hydraulic fracturing TLS, whose red light is set at ML 0.5. Because the nominal detectability of the UK network is ML 2, this scheme will require the deployment of monitoring stations in close proximity to well sites. Using data collected from mining events near New Ollerton, Nottinghamshire, we illustrate the effects that proximity has on travel path velocities and attenuation, then perform a damped least‐squares inversion to determine appropriate constants within the ML scale. We show that the attenuation term needs to increase from 0.00183 to 0.0514 and demonstrate that this higher value is representative of a ray path within a slower more attenuating sedimentary layer compared with the continental crust. We therefore recommend that the magnitude scale ML=log(A)+1.17log(r)+0.0514r−3.0 should be used when local monitoring networks are within 5 km of the event epicenters.
The problem with magnitudes calculated using nearby stations
Luckett, R, and Butcher, A. 2016
In April 2011, fracking near Blackpool caused a 2.4 ML earthquake at a shallow depth. This was felt by local people and there was considerable public concern. The British Geological Survey (BGS) installed temporary seismic stations close to the epicentre and recorded several subsequent, smaller events. There was, however, some ambiguity over the magnitude of these later events. The magnitudes calculated for the temporary stations were too high for unfelt events that were not, in general, recorded on the national network. A single induced earthquake was recorded both by the temporary stations and by a few stations of the UK national network. The local magnitude calculated from amplitudes recorded on the more distant stations was 1.2 ML but the very nearby stations recorded amplitudes corresponding to a magnitude of 2.3 ML. In subsequent studies, this one event was used to scale amplitudes from the nearby stations to magnitudes that were probably similar to the magnitudes that would have been calculated using distant stations — a most unsatisfactory solution. The regulatory approach adopted in the UK to manage the risk of induced seismicity is a ‘traffic light’ monitoring scheme, with a remedial action level, or ‘red light’, set at 0.5 ML. As the UK national network has at a nominal detection level of ML > 2, the installation of local seismic stations is critical for the operation of this scheme. However, the suitability of the current UK local magnitude scale is questionable, given that it was not calibrated using very near-receiver events. In fact, the evidence of magnitude discrepancies demonstrated near Blackpool and elsewhere suggests that the scale is not suitable. The single event recorded on both nearby and distant stations at Blackpool is not sufficient to base any further work on. However, analysis of the BGS catalogue shows that this affect has been observed on several other occasions. In particular, over 500 small earthquakes were recorded by a network installed within a few kilometres of the New Ollerton coal mine in 2014. Those events that were also recorded by stations of the UK national network had magnitudes calculated using the local network much larger than those calculated at more distant stations. We use this data to analyse amplitudes recorded very close to earthquakes and test various ideas. We then discuss possible alternatives to the current UK ML scale that might allow near event seismic data to be used to calculate robust magnitudes.
SISMIKO: emergency network deployment and data sharing for the 2016 central Italy seismic sequence
Moretti, M, et al. 2016
At 01:36 UTC (03:36 local time) on August 24th 2016, an earthquake Mw 6.0 struck an extensive sector of the central Apennines (coordinates: latitude 42.70°N, longitude 13.23°E, 8.0 km depth). The earthquake caused about 300 casualties and severe damage to the historical buildings and economic activity in an area located near the borders of the Umbria, Lazio, Abruzzo and Marche regions. The Istituto Nazionale di Geof- isica e Vulcanologia (INGV) located in few minutes the hypocenter near Accumoli, a small town in the province of Rieti. In the hours after the quake, dozens of events were recorded by the National Seismic Network (Rete Sismica Nazionale, RSN) of the INGV, many of which had a ML > 3.0. The density and coverage of the RSN in the epicentral area meant the epicenter and magnitude of the main event and subse- quent shocks that followed it in the early hours of the seismic sequence were well constrained.
However, in order to better constrain the localizations of the aftershock hypocenters, especially the depths, a denser seismic monitoring network was needed.
Just after the mainshock, SISMIKO, the coordinating body of the emergency seismic network at INGV, was activated in order to install a temporary seismic network integrated with the existing permanent network in the epicentral area. From August the 24th to the 30th, SISMIKO deployed eighteen seismic stations, generally six components (equipped with both velocimeter and accelerometer), with thirteen of the seismic station transmitting in real-time to the INGV seismic monitoring room in Rome. The design and geometry of the temporary network was decided in consolation with other groups who were deploying seismic stations in the region, namely EMERSITO (a group studying site-effects), and the emergency Italian strong motion network (RAN) managed by the National Civil Protection Department (DPC). Further 25 BB temporary seismic stations were deployed by colleagues of the British Geological Survey (BGS) and the School of Geosciences, University of Edinburgh in collaboration with INGV.
All data acquired from SISMIKO stations, are quickly available at the European Integrated Data Archive (EIDA). The data acquired by the SISMIKO stations were included in the preliminary analysis that was performed by the Bollettino Sismico Italiano (BSI), the Centro Nazionale Terremoti (CNT) staff working in Ancona, and the INGV-MI, described below.
Seismic hazard assessment for Pavúa, Mozambique. ‘Commercial in Confidence’
Mosca, I, Ellen, R, and Sargeant, S. 2016
This report presents a probabilistic seismic hazard assessment (PSHA) for Pavúa Hydropower Project. This was completed as a desk study without any fieldwork.
Earthquake science in DRR policy and practice in Nepal
Oven, K, Milledge, D, Densmore, A, Jones, H, Sargeant, S, and Datta, A. 2016
Nepal is a geologically active country with a long history of destructive earthquakes — most recently in the 2015 Gorkha earthquake sequence. There have been substantial advances in the scientific understanding of earthquake hazard in Nepal, but it is not clear how that understanding has informed, or could inform, national and international investment in earthquake disaster risk reduction (DRR) activities, and to what effect. This paper aims to understand the role that earthquake science plays in DRR policy and practice in Nepal by seeking answers to the following. What earthquake science is used by DRR stakeholders in Nepal, and for what purpose? To what extent is earthquake DRR policy and practice in line with current scientific knowledge? Where and how is scientific knowledge seen as particularly useful for policy and practice, and where is it seen to be less useful and why? What are the drivers of and constraints on the production and use of earthquake science? Are there opportunities to better produce or broker scientific knowledge for policy and practice? What effects could better use of earthquake science deliver, and to whom?
Reflections on recent recommendations on the use of science in disaster risk reduction using case studies from Bangladesh and the Western United States
Sargeant, S L, and Lindquist, E. 2016
The valuable role that science has to play in disaster preparedness and risk reduction is widely recognized and was highlighted during the development of the successor to the Hyogo Framework for Action for disaster risk reduction that was adopted in March 2015. However, there are many factors that limit how effectively science can inform both disaster risk reduction policy and practice. Understanding these factors and taking steps to overcome them require a broad view, and a comparative approach can be instructive. We focus on two projects that were independently completed by the authors: earthquake risk management in Bangladesh and flooding and wildfires management in the United States. We use each case to reflect on the implications of recent recommendations made by the Science and Technology Advisory Group (STAG) of the United Nations Office for Disaster Risk Reduction that attempt to increase the integration of science in disaster risk reduction policy making. We then use the STAG recommendations as a framework for integrating our independent case study findings. Despite the differences in the geographic contexts and hazards being considered, these examples broadly support the STAG recommendations. However, the fine details of the way in which science is used in decision making need to be given careful consideration if science is to fully support disaster risk reduction. Although our collective observations suggest that science is an important part of the disaster risk reduction (DRR) process, suggesting that it is ‘key to post-2015 DRR efforts’ as the STAG recommendations do, may perhaps overstate the role that science is able to play.
Prospective Earthquake Forecasts at the Himalayan Front after the 25 April 2015 M 7.8 Gorkha Mainshock
Segou, M, and Parsons, T. 2016
When a major earthquake strikes, the resulting devastation can be compounded or even exceeded by the subsequent cascade of triggered seismicity. As the Nepalese recover from the 25 April 2015 shock, knowledge of what comes next is essential. We calculate the redistribution of crustal stresses and implied earthquake probabilities for different periods, from daily to 30 years into the future. An initial forecast was completed before an M 7.3 earthquake struck on 12 May 2015 that enables a preliminary assessment; postforecast seismicity has so far occurred within a zone of fivefold probability gain. Evaluation of the forecast performance, using two months of seismic data, reveals that stress-based approaches present improved skill in higher-magnitude triggered seismicity. Our results suggest that considering the total stress field, rather than only the coseismic one, improves the spatial performance of the model based on the estimation of a wide range of potential triggered faults following a mainshock.
Frequency-magnitude Distribution for Natural and Mining-induced Seismicity in UK
Segou, M, and Baptie, B. 2016
Over the last 30 years mining-induced seismicity in United Kingdom has been monitored by the British Geological Survey. About 4000 events with local magnitudes between -0.7 and 3.0 have been reported in coal mining reporting areas. The magnitude-frequency distribution follows a Gutenberg-Richter relation with a higher b-value of about 1.35 and 1.1 in Scotland and England, respectively. The above indicates the absence of large magnitude events for this type of induced seismicity. Within the instrumental seismicity period about 45% of seismicity corresponds to coal mining events. Reliable baseline for natural seismicity is essential to both discriminating from induced seismicity, such as hydraulic fracturing, and providing stake-holders/decision makers real-time earthquake probabilities during operational phase. The magnitude-frequency distribution suggests a b-value of 0.85 with standard deviation of 0.12 for the entire catalog of naturally triggered events with some variability met at different regions. Analytically b-values of about 0.95, 0.70 and 0.9 are reported in Scotland, Wales and England with standard deviations of 0.1, 0.07 and 0.14, respectively. The 1984 Ml=5.4 Llyn Peninsula remains the larger earthquake reported the last half-century and it is characterized by a low p-value indicating a slow aftershock decay rate. In the UK a moderate sized event with magnitude larger than 5, occurs on average every 15 years followed by few (<10) felt aftershocks. In this low-seismicity environment only local population at the vicinity of mining fields has experienced light (EMS 5/6) ground shaking in the past.
Earthquake hazard assessment in Kazakhstan
Silacheva, N, and Mosca, I. 2016
Engineers, architects and planners generally require something more specific to design their structures and infrastructure (including emergency plans) to be resilient to earthquakes. In particular, they need some estimate of likely ground motion or shaking at a particular place like a hospital or a school. This is usually expressed as maximum ground acceleration (peak ground acceleration, or PGA), which is what produces the forces that destroy buildings. Hazard assessments bring together knowledge from all the modern earthquake science techniques described in the previous sections to develop a picture of the likely distribution, size, character and frequency of occurrence of earthquakes. This picture can be very localised — to a particular active fault, or a place of particular interest — or it can be quite general, across a wide region. Its aim is to make a statement on the nature of the threat, not what can be done about it; nonetheless, it is the necessary first step for developing disaster-risk-reduction policies and strategies, and provides information needed by engineers, architects and planners.
Environmental Baseline Monitoring Project. Phase II, Final Report
Ward, R S, Smedley, P S, Allen, G, Baptie, B J, Daraktchieva, Z, Horleston, A, Jones, D G, Jordan, C J, Lewis, A, Lowry, D, Purvis, R M, and Rivett, M O. 2017
This report is submitted in compliance with the conditions set out in the grant awarded to the British Geological Survey (BGS), for the period April 2016–March 2017, to support the jointly-funded project ‘Science-based environmental baseline monitoring’. It presents the results of monitoring and/or measurement and preliminary interpretation of these data to characterise the baseline environmental conditions in the Vale of Pickering, North Yorkshire and for air quality, the Fylde in Lancashire ahead of any shale gas development. The two areas where the monitoring is taking place have seen, during the project, planning applications approved for the exploration for shale gas and hydraulic fracturing. It is widely recognised that there is a need for good environmental baseline data and establishment of effective monitoring protocols ahead of any shale gas/oil development. This monitoring will enable future changes that may occur as a result of industrial activity to be identified and differentiated from other natural and man-made changes that are influencing the baseline. Continued monitoring will then enable any deviations from the baseline, should they occur, to be identified and investigated independently to determine the possible causes, sources and significance to the environment and public health. The absence of such data in the United States has undermined public confidence, led to major controversy and inability to identify and effectively deal with impact/contamination where it has occurred. A key aim of this work is to avoid a similar situation and the independent monitoring being carried out as part of this project provides an opportunity to develop robust environmental baseline for the two study areas and monitoring procedures, and share experience that is applicable to the wider UK situation. This work is internationally unique and comprises an inter-disciplinary researcher-led programme that is developing, testing and implementing monitoring methodologies to enable future environmental changes to be detected at a local scale (individual site) as well as across a wider area, e.g. ‘shale gas play’ where cumulative impacts may be significant. The monitoring includes: water quality (groundwater and surface water), seismicity, ground motion, soil gas, atmospheric composition (greenhouse gases and air quality) and radon in air. Recent scientific and other commissioned studies have highlighted that credible and transparent monitoring is key to gaining public acceptance and providing the evidence base to demonstrate the industry’s impact on the environment and importantly on public health. As a result, BGS and its partners initiated in early 2015, a co-ordinated programme of environmental monitoring in Lancashire that was then extended to the Vale of Pickering in North Yorkshire after the Secretary of State for Energy and Climate Change (BEIS) awarded a grant to the British Geological Survey (BGS). The current duration of the grant award is to 31st March 2018. It has so far enabled baseline environmental monitoring for a period of more than 12 months. With hydraulic fracturing of shale gas likely to take place during late 2017/early 2018, the current funding will allow the environmental monitoring to continue during the transition from baseline to monitoring during shale gas operations. This report presents the monitoring results to April 2017 and a preliminary interpretation. A full interpretation is not presented in this report as monitoring is continuing and it is expected that there will be at least six months of additional baseline data before hydraulic fracturing takes place. This represents up to 50% more data for some components of the montoring, and when included in the analysis will significantly improve the characterisation and interpretation of the baseline. In addition to this report, the BGS web site contains further information on the project, near real-time data for some components of the monitoring and links to other projects outputs, e.g. reports and videos (www.bgs.ac.uk/research/groundwater/shaleGas/monitoring/home.html).
Micro-seismic source location with a single seismometer channel using coda wave interferometry
Zhao, Y, Curtis, A, and Baptie, B. 2016
Finding relative locations of seismic events is essential for discriminating earthquake fault and auxiliary planes from the sequences of aftershocks or foreshocks, studying earthquake interaction and recurrence, and monitoring stress state and induced (micro-)seismicity. Conventional methods, such as joint hypocenter determination and double-difference location, usually require a large number of seismic stations and good event-station azimuthal coverage to obtain reliable results. However, such requirements are not always fulfilled. To this end, a source location method based on coda wave interferometry (CWI) is developed, which uses the scattered waves in the coda of seismograms to estimate the differences between two seismic states, in this case to estimate the distance between pairs of earthquake locations.
Those are then used jointly to determine the relative location of a cluster of events in a probabilistic framework. The purpose of this study is to test the performance of this novel approach on induced micro- seismicities, where it was applied to a micro-seismic dataset of mining induced events recorded in England. We find that source separation estimates are highly consistent and the earthquake location results agree to within estimated uncertainties when using different individual seismometer channels. We also discuss three issues that arose during the implementation for this dataset and provide solutions that can be used in future applications.
