Energy Procedia 37 ( 2013 ) 4749 – 4756

1876-6102 © 2013 The Authors. Published by Elsevier Ltd.Selection and/or peer-review under responsibility of GHGTdoi: 10.1016/j.egypro.2013.06.384

GHGT-11

Induced seismicity; observations, risks and mitigation measures at CO2 storage sites

Andy Nicola,b, Matt Gerstenbergera,b, Chris Bromleyc, Rachel Carneb, Lauriane Chardotc, Susan Ellisb, Charles Jenkinsa,d, Tony Siggins a,e, Paul Viskovicb

aCO2CRC, Australia, bGNS Science, Lower Hutt 5040, New Zealand,

cGNS Science, Wairakei 3377, New Zealand, dCSIRO Canberra Australia, eCSIRO Melbourne Australia

Abstract

Seismicity produced by human activities (i.e. induced seismicity) has been widely reported over the last 40 years. To date few induced earthquakes have been recorded at CO2 storage sites, however, the volumes of injected CO2 and the number of operational sites are small. A review of induced seismicity from different types of fluid injection and extraction sites confirms that these events are typically magnitude (M) and in many cases have no reported earthquakes. Although the size (and associated risks) of induced earthquakes at CO2 storage sites is most likely to be small, these events could decrease seal integrity or raise public concerns, while rare larger events (>M5) could also have ramifications for CCS beyond a single site. These risks can be reduced by careful site selection and development of site-specific risk reduction and mitigation programmes. Forecasts of induced seismicity using physical and statistical models and real-time monitoring will be key planning and decision making tools. The utility of monitoring and mitigation programmes will be maximized by establishing prior to injection, site performance and management guidelines for acceptable levels of induced seismicity, and agreed control measures. Further improvements to risk management practices, understanding induced seismicity processes and stakeholder confidence may be achieved by; a) increasing the number of publically available induced earthquake catalogues for development and testing of physical and statistical models, b) undertaking more systematic studies of individual sites populated by well constrained sub-surface geomechanical information and seismicity data complete down to small magnitudes (e.g., M-3), c) enhancing the physical reality of numerical dynamic models, d) studying the scaling effects of seismicity associated with moving from pilot projects to full commercial implementation of CO2 storage, e) developing standard risk management procedures and guidelines for induced seismicity at CCS sites and, f) filling induced seismicity knowledge gaps in the CCS community by collaborating with seismologists and modellers working in other industries. © 2013 The Authors. Published by Elsevier Ltd. Selection and/or peer-review under responsibility of GHGT

Available online at www.sciencedirect.com

© 2013 The Authors. Published by Elsevier Ltd.Selection and/or peer-review under responsibility of GHGT

4750 Andy Nicol et al. / Energy Procedia 37 ( 2013 ) 4749 – 4756

Keywords Induced seismicity; earthquake magnitude; risk management;

1. Introduction

Seismicity induced by anthropogenic activities including water injection into the sub-surface and hydrocarbon extraction has been widely reported over the last 40 years [1-32]. Although these earthquakes are mainly of small to moderate magnitude (e.g., M4.5) and are unlikely to cause significant damage, they could promote leakage of CO2 from the storage container [32] or raise public concerns [16, 33]. To date few induced earthquakes have been recorded at CO2 sub-surface storage sites, however, the volumes of injected CO2 are typically small, the number of sites low and the attendant seismometer networks limited in spatial extent or temporal coverage. These knowledge gaps mean that questions remain about the magnitude of risks associated with induced seismic and point to the need for further work. As with many parts of the CCS system, understanding the physical processes resulting in induced seismicity is crucial for assessing and reducing the associated risk and increasing the likelihood of successful project completion.

Induced seismicity occurs when anthropogenic activities cause a new or pre-existing fault to move (Fig

1). A large part of the understanding of induced seismicity has been developed though empirical case studies and analytical modelling where particularly active or energetic sequences have been analysed using standard earthquake catalogue parameters such as earthquake magnitude, location and time. Although useful for understanding the processes and risks, these datasets are biased towards sites where the rates and/or magnitudes of induced seismicity are high. Systematic case studies of seismicity induced by fluid injection across multiple sites are rare and studies of catalogue data have been largely limited to individual sites with variable quality data [e.g., 2-4,10,17,19]. To improve understanding of the overall behavior of induced earthquakes, reviews and analyses of published data have been compiled by numerous authors [e.g., 6,8,12,15-16,25,27]. More recently research has been undertaken to develop computational models that specifically target the induced seismicity process and aim to understand the underlying physics that is driving it including recent models developed specifically for CO2 injection and storage [e.g., 14]. These studies have improved our understanding of the mechanisms that produce induced seismicity, however, it is presently not possible to forecast these earthquakes with sufficient precision for detailed risk assessment.

With the aim of improving understanding of induced seismicity at CO2 storage sites and reducing the

associated risks the relevant literature from a broad spectrum of industries has been reviewed. With this knowledge we can begin to estimate what the potential is for induced seismicity in CO2 injection and storage projects. The ultimate goal of this contribution is to identify important gaps in understanding of induced seismicity, to highlight where future research may have the greatest impact for reducing uncertainties associated with such events, and to recommend procedures for reducing the risk to CCS projects.

Andy Nicol et al. / Energy Procedia 37 ( 2013 ) 4749 – 4756 4751

Fig. 1. Processes inducing seismic activity at CO2 injection sites (from reference 24).

2. Observations

Induced seismicity data from commercial (millions of tonnes) and experimental (<100 000 tonnes) scale CO2 storage sites are sparse [7,24,28]. Few events have been reported at Sleipner (Norway), In Salah (Algeria), Otway Basin (Australia) and Ketzin (Germany) projects [6], while small numbers (<50 per year) of low magnitude (M-3 to -1) earthquakes occurred at Weyburn during CO2 injection [28]. The absence or low numbers of induced earthquakes and the observation of no events exceeding M1 could be interpreted to suggest that they pose little risk. However, the number of CO2 storage sites where detailed seismicity information are available is small and data from other types of fluid injection and extraction operations have been considered to increase the size of the database. Seismicity associated with fluid injection and extraction share many common features, while differences in the compressibility, viscosity, density and relative permeability of water and supercritical CO2 do not appear to produce contrasting induced seismicity populations [29]. Given these observations there is value in using induced seismicity from water injection and petroleum production sites as analogues for induced seismicity at future CO2 storage sites.

4752 Andy Nicol et al. / Energy Procedia 37 ( 2013 ) 4749 – 4756

4 Author name / Energy Procedia 00 (2011) 000 000

The spatial distribution of induced earthquakes for sites of fluid/gas extraction and injection often

follows generally understood patterns, even if the scale and details of the pattern cannot be predicted ahead of time. Fault slip and rock failure typically occur in seconds and produce induced earthquakes that range in magnitude from M-3 to M>5 [e.g., 6,8,15,25,27]. These earthquakes are most often small to

too small to be felt or recorded by seismometer networks. Due in part to the dominance of small earthquakes and a lack of local seismometer networks, induced seismicity has not been reported at a large number of sub-surface injection or extraction sites. Where observed induced seismicity begins almost immediately after the onset of sub-surface operations and continues for their duration. Induced events typically commence adjacent to the injection/extraction well and migrate outwards from the well with time and increasing volume of displaced fluid [e.g., 2,18,20]. Following the injection/extraction period, the rate of earthquakes typically follows an exponential decay as is seen in tectonic aftershock sequences [e.g., 4,11,21]. In some cases, the induced seismicity continues for years or tens of years after injection has ceased [e.g., 10,18].

extraction of large

volumes of hydrocarbons or water, however, the anthropogenic origin of many of these events is debated in the literature [26]. This debate is fuelled by the recognition that larger extraction-induced earthquakes can occur kilometres from the reservoir, many years after the beginning of the extraction operation and in areas of existing tectonic activity [e.g., 8,13,15,23,26]. Although the likelihood of earthquakes being induced by CO2 injection will probably be low, it is not zero and the possibility of these events should be considered for management of seismicity and risk.

Fig. 2. Variations in induced earthquake magnitude and frequency with changes to injection rate and wellhead pressure for injection operations at Paradox Valley, Colorado, USA during the period 1996-2004 (figure from reference 21 and seismicity data original presented by reference 1). Over 4,000 induced earthquakes were recorded by the 15-station Paradox Valley Seismic Network, which was installed in 1995 prior to the onset of injection.

Andy Nicol et al. / Energy Procedia 37 ( 2013 ) 4749 – 4756 4753

While the mechanics of shear failure and the earthquake process are well understood [e.g., 15,25,34,35], the key factors that control the magnitude range, seismicity rates and spatial and temporal distributions of events appear to vary between sites. At some sites earthquakes are strongly influenced by the rate and volume of fluid injected and corresponding changes in reservoir pressure [e.g. 4,10,11,21,22](Fig 2). Induced seismicity is also influenced by the mechanical properties of the injection interval. The highest rates of induced seismicity and largest earthquake magnitudes are often, for example, observed within crystalline basement perhaps because these rocks are more brittle than sedimentary strata and/or they have low permeabilities which promote rapid pressure increases [6,32]. The presence of relatively weak pre-existing faults may also locally enhance induced seismicity [5,25,27,31], with these faults often accruing slip during earthquakes at stresses well below that required to fracture intact reservoir rock.

3. Risks and mitigation measures

One of the principal reasons for studying induced seismicity arising from injection of CO2 is to assess

and manage the associated risks. For the vast majority of existing fluid injection or extraction projects induced seismicity has not significantly disrupted operations. Despite the shallow hypocentral depths (e.g., <5 km) the small to moderate magnitudes common for induced earthquakes mean that they rarely cause damage to infrastructure or private property, especially in regions of modern earthquake-resistant construction [e.g., 25,26,30,36]. With increases in the number of sites and storage volumes of injected CO2, maintaining the present record of no adverse induced seismicity events for CCS projects requires that operators select sites judiciously and employ risk management to minimise the impact of these risks. Experience from Enhanced Geothermal Systems suggests that an incomplete operational framework (e.g., incomplete legislation, poorly designed operational procedures and unresolved liability) and a lack of public support and acceptance may pose important risks to CCS projects [see 33]. Another factor that may be important is the role of earthquakes in the leakage of CO2 through the seal [32]. The earthquake magnitudes required to generate unacceptably high risks varies between sites and for different risk factors. Close to populated areas events as small as M2 may be felt and raise stakeholder concern, while events of at least M3-4 could cause minor damage to infrastructure and injury.

The risks associated with induced seismicity at CCS sites can be reduced and mitigated using a

systematic and structured risk management programme. Experiences such as those at the Basel geothermal experiment [e.g., 9] support the notion that a risk assessment of induced seismicity, including forecasting of reservoir response to the predicted pressure changes, installation of a microseismic array and consideration of mitigation measures, should be carried out prior to the commencement of injection. While precise forecasts of the expected induced seismicity may never be possible, a thorough risk management procedure will require estimates of the expected magnitude, number, location and timing of potential induced earthquakes. Such forecasts should utilise site specific observations (e.g., pre-injection stress conditions, rock strengths, locations and frictional properties of pre-existing faults) together with physical and statistical models optimised for each site. Combining forecasts with real-time monitoring of induced seismicity is necessary to maintain an accurate picture of the seismicity and to allow for mitigation of the associated risks as they evolve. The design of monitoring networks for induced seismicity could vary between sites depending on a range of factors including; desired event observational magnitude range, site location, size and reservoir depth, levels of background seismicity and ambient noise, and cultural site constraints (e.g., existing infrastructure and financial priorities). To optimise the utility of monitoring and mitigation programmes site performance and management guidelines for induced seismicity should be established prior to injection, as has been proposed for

4754 Andy Nicol et al. / Energy Procedia 37 ( 2013 ) 4749 – 4756

Enhanced Geothermal System sites [e.g., 33,37]. Guidelines include setting the acceptable level (i.e. magnitude range and productivity) and impacts of seismicity and outlining the control measurements to be implemented if original expectations are exceeded. Such measures, in combination with public education to ensure that the perceived and actual risks are comparable, and reservoir pressure management plans, could be expected to significantly reduce the risks of induced seismicity to successful completion of CO2 storage projects. 4. Future focus

While there is acceptance that injection of CO2 has the potential to induce earthquakes, there is not

agreement about the level of risk to CCS projects posed by these earthquakes. What is clear from experience to date is that the seismic response to CO2 injection is likely to vary dramatically between sites. At many sites the earthquakes will be too small to be felt or recorded instrumentally and induced seismicity may pose a low risk to CCS projects. At a few sites, however, earthquakes triggered by the injection of CO2 have the potential to negatively impact on the successful completion of CCS projects. Being able to identify which CCS sites could present an unacceptable induced seismicity risk and developing management strategies to address these risks will be important for the successful sub-surface storage of CO2. Six main areas of future work are recommended to improve our understanding of induced seismicity processes, of the associated risks they pose to CO2 storage projects and the risk reduction measures that may be employed. These main areas are summarized below.

1) Induced Seismicity Catalogue Database: These data are critical for the development and

validation of statistical and physical models. Creation of a central database for global injection induced seismicity observations and encouragement of data sharing will greatly facilitate model development and improve induced seismicity forecasting. To ensure that such data are collected it is recommended that microseismic networks (including some down-hole instruments) are operated at all CCS sites prior to, during and following injection.

2) Systematic and detailed site studies: To improve understanding of induced earthquake processes and develop predictive models, a greater number of systematic site studies rich in data (i.e. seismicity catalogues, reservoir properties and injection histories) are required. As more information becomes available, future studies will undoubtedly begin to reduce the uncertainties in the forecast induced seismicity parameters, such as the largest possible magnitude or the expected rate of events for a site. Uncertainty reduction is important for robust quantitative risk assessment.

3) Physical modelling: The development of physical models is at an early stage but shows promise for informing induced seismicity predictions. In order to provide improved predictions these models must incorporate well quantified system parameters (e.g., porosities, permeabilities, pore pressures, rock strength, temperatures and fluid compositions) and more accurately replicate system processes. For both validation and testing of the codes and models it is important that codes continue to be benchmarked against each other and tested, in a rigorous and predictive sense, against observations of induced seismicity.

4) Scaling from pilot projects to commercial-scale injection: CCS projects to date have mainly consisted of pilot projects. The volumes of injected CO2 for these pilot projects are probably orders of magnitude less than what will be required to achieve the ultimate goal of reducing the amount of CO2 in the atmosphere to slow global warming and climate change. It is currently not well understood how the processes will scale to the much larger and potentially more complex reservoirs that will be required.

5) Risk management protocol: A CCS specific protocol or guideline for risk management specific to induced seismicity does not yet exist. Such guidelines have been developed for other areas of CCS

Andy Nicol et al. / Energy Procedia 37 ( 2013 ) 4749 – 4756 4755

[e.g., 38] and for induced seismicity in other industries [e.g., 33]. A CCS specific protocol should describe the major steps necessary to manage the induced seismicity risk through to guidelines for monitoring during injection and beyond. Such a protocol would educate stakeholders and project managers and ease the risk management process.

6) Induced Seismicity Collaboration: Induced seismicity is not a CCS specific problem and a number of industries are currently devoting time and resources towards improving our understanding of this topic. The CCS community will benefit from interaction and integration with the other industries (e.g., Petroleum and Enhanced Geothermal) and interest groups (e.g., seismological community). These groups are more advanced in their understanding of induced seismicity and are either working on, or could be focused on, issues of direct relevance to CCS.

Acknowledgements

The contents of this paper are derived from an Induced Seismicity report commissioned and funded by the IEA GHG.

References

[1] , Block L. Deep-Injection and Closely Monitored Induced Seismicity at Paradox Valley, Colorado. Bull Seism Soc Am 2005; 95, 664-683. [2] Baisch S, Weidler R, Vörös R, Wyborn D, de Graaf L. Induced seismicity during the stimulation of a geothermal HFR reservoir in the Cooper Basin, Australia. Bull Seism Soc Am 2006; 96, 2242-2256. [3] Baisch S, Voros R, Weidler R, Wyborn D. Investigation of Fault Mechanisms during Geothermal Reservoir Stimulation Experiments in the Cooper Basin, Australia. Bull Seism Soc Am 2009; 99, 148-158. [4] Charléty J, Cuenot N, Dorbath L, Dorbath C, Haessler H, Frogneux M. Large earthquakes during hydraulic stimulations at the geothermal site of Soultz-sous-Forêts. Int Journ Rock Mech &Mining Sci 2007; 44, 1091-1105. [5] Doser DI, Baker MR, Luo M, Marroquin P, Ballesteros L, Kingwell J, Diaz HL, Kaip G. The not so simple relationship between seismicity and oil production in the Permian Basin, West Texas. Pure & Applied Geophys 1992; 139, 481-506. [6] Evans KF, Zappone A, Kraft T, Deichmann N, Moia F. A survey of the induced seismic responses to fluid injection in geothermal and CO2 reservoirs in Europe. Geothermics 2012; 41, 30-54. http://dx.doi.org/10.1016/j.geothermics.2011.08.002. [7] Gerstenberger M, Nicol A, Bromley C, Carne R, Chardot L, Ellis S, Jenkins C, Siggins T, Viskovic, P. Induced seismicity and its implications for CO2 storage risk. IEAGHG Report 2012. [8] Grasso J-R. Mechanics of Seismic Instabilities Induced by the Recovery of Hydrocarbons. Pure & Applied Geophys 1992; 139, 507-534. [9] Häring MO, Schanz U, Ladner F, Dyer BC. Characterization of the Basel 1 enhanced geothermal system. Geothermics 2008, 37, 469-495. [10] Healy J.H, Rubey WW, Griggs DT, Raleigh CB. The Denver Earthquakes 1968; Science, 161, 1301-1310. [11] Ladner F, Häring MO. Hydraulic Characteristics of the Basel 1 Enhanced Geothermal System. Geothermal Resources Council Transactions 2009; 33, 199-203. [12] Majer EL, Baria R, Stark M, Oates S, Bommer J, Smith B, Asanuma H. Induced seismicity associated with Enhanced Geothermal Systems. Geothermics 2007; 36, 185-222. [13] McGarr A. On a possible connection between three major earthquakes in California and oil production. Bull Seism Soc Am 1991; 81, 948-970. [14] . Modelling induced seismicity due to fluid injection. Abstract S42B-05 presented at 2011 Fall Meeting, AGU, San Francisco, Calif. 5-9 Dec.

4756 Andy Nicol et al. / Energy Procedia 37 ( 2013 ) 4749 – 4756

[15] Nicholson C, Wesson RL. Triggered Earthquakes and Deep Well Activities. Pure & Applied Geophys 1992; 139, 561-578. [16] Nicol A, Carne R, Christopersen A, Gerstenberger M. Induced seismicity and its implications for CO2 sequestration risk. Energy Procedia 4 2011; 3699-3706. [17] Raleigh CB, Healy JH, Bredehoeft JD. An Experiment in Earthquake Control at Rangely, Colorado. Science 1976; 191, 1230-7. [18] Seeber L, Armbruster JG, Kim W.-Y. A Fluid-Injection-Triggered Earthquake Sequence in Ashtabula, Ohio: Implications for Seismogenesis in Stable Continental Regions. Bull Seism Soc Am 2004; 94, 76-87. [19] Segall P. Earthquakes triggered by fluid extraction. Geology 1989; 17, 942-6. [20] Shapiro SA, Dinske C. Scaling of seismicity induced by nonlinear fluid-rock interaction. J Geophys Res 2009; 114, B09307. [21] Shapiro SA, Dinske C, Kummerow J. Probability of a given-magnitude earthquake induced by a fluid injection. Geophys Res Lett 2007; 34, L22314. [22] Shapiro SA, Kruger OS, Dinske C, Langenbruch C. Magnitudes of induced earthquakes and geometric scales of fluid-stimulated rock volumes. Geophysics 2011; 76, Issue 6 WC55 (2011); http://dx.doi.org/10.1190/geo2010-0349.1 [23] Simpson DW, Leith W. The 1976 and 1984 Gazli, USSR, Earthquakes Were they induced? Bull Seism Soc Am 1985, 75, 1465-8. [24] Sminchak J, Gupta N. Aspects of induced seismic activity and deep-well sequestration of carbon dioxide. Environmental Geosciences 2003; 10, 81-9. [25] Suckale J. Induced seismicity in hydrocarbon fields, Advances in Geophysics 2009; 51, 55 106. [26] Suckale J. Moderate-to-large seismicity induced by hydrocarbon production. The Leading Edge 2010; 29, 310-9. [27] Van Eijs RMHE, Mulders FMM, Nepveu M, Kenter CJ, Scheffers BC. Correlation between hydrocarbon reservoir properties and induced seismicity in the Netherlands. Environmental Geology 2006; 84, 99 111. [28] Verdon JP, Kendall J-M, White DJ, Angus DA, Fisher QJ Urbancic T. Passive seismic monitoring of carbon dioxide storage at Weyburn. The Leading Edge 2010; 29, 200-6. [29] Verdon, JP, Kendall J-M, Maxwell SC. A comparison of passive seismic monitoring of fracture stimulation from water and CO2 injection. Geophysics 2010; 75, No. 3 P. MA1 MA7, doi:10.1190/1.3377789. [30] Zoback MD. Managing the seismic risk posed by wastewater disposal, Earth Magazine 2012; April. [31] Zoback MD, Zink JC. Production-induced normal faulting in the Valhall and Ekofisk oil fields. Pure & Applied Geophys 2002; 159, 403-420. [32] Zoback MD, Gorelick SM. Earthquake triggering and large-scale geologic storage of carbon dioxide. PNAS 2012; 1-5, www.pnas.org/cgi/doi/10.1073/pnas.1202473109. [33] Majer E, Baria R, Stark M. Protocol for induced seismicity associated with enhanced geothermal systems. Report produced in Task D Annex I (9 April 2008), International Energy Agency-Geothermal Implementing Agreement (incorporating comments by: C. Bromley, W. Cumming, A. Jelacic and L. Rybach). Available at: http://www.iea-gia.org/documents/ Protocol for Induced Seismicity 2008; EGS-GIADoc25Feb09.pdf [Accessed February 9, 2011]. [34] Hubbert MK, Rubey W.R. Role of fluid pressure in mechanics of overthrust faulting: 1. Mechanics of fluid-filled porous solids and its application to overthrust faulting. Bull Seism Soc Am 1959; 70, 115-166. [35] Zoback MD. Reservoir Geomechanics. Cambridge University Press, 2007. [36] Van Eck T, Goutbeek F, Haak H, Dost B. Seismic hazard due to small magnitude, shallow-source, induced earthquakes in The Netherlands. Environmental Geology 2006; 87, 105 121. [37] Bommer J, Oates S, Cepeda JM, Lindholm C, Bird J, Torres R, Marroquin G., Rivas J. Control of hazard due to seismicity induced by a hot fractured rock geothermal project. Engineering Geology2006; 83, 287-306. [38] DNV (Det Norske Veritas). CO2WELLS Guidelines for the risk management of existing wells at CO2 geological storage sites (No. 2011-0448) 2011; Hovik, Norway.