Injection-Induced Earthquakes William L. Ellsworth …users.clas.ufl.edu/prwaylen/GEO2200 Readings/Readings/Fracking... · Injection-Induced Earthquakes William L. Ellsworth Background:

DOI 101126science1225942 (2013)341 Science

William L EllsworthInjection-Induced Earthquakes

This copy is for your personal non-commercial use only

clicking herecolleagues clients or customers by you can order high-quality copies for yourIf you wish to distribute this article to others

herefollowing the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

) July 18 2013 wwwsciencemagorg (this information is current as of

The following resources related to this article are available online at

httpwwwsciencemagorgcontent34161421225942fullhtmlversion of this article at

including high-resolution figures can be found in the onlineUpdated information and services

httpwwwsciencemagorgcontentsuppl2013071034161421225942DC1html can be found at Supporting Online Material

httpwwwsciencemagorgcontent34161421225942fullhtmlrelatedfound at

can berelated to this article A list of selected additional articles on the Science Web sites

httpwwwsciencemagorgcontent34161421225942fullhtmlref-list-1 20 of which can be accessed freecites 34 articlesThis article

CopyrightAmerican Association for the Advancement of Science 1200 New York Avenue NW Washington DC 20005 (print ISSN 0036-8075 online ISSN 1095-9203) is published weekly except the last week in December by theScience

Injection-Induced EarthquakesWilliam L Ellsworth

Background Human-induced earthquakes have become an important topic of political and scientifi c discussion owing to the concern that these events may be responsible for widespread damage and an overall increase in seismicity It has long been known that impoundment of reservoirs surface and underground mining withdrawal of fl uids and gas from the subsurface and injection of fl uids into underground formations are capable of inducing earthquakes In particular earthquakes caused by injection have become a focal point as new drilling and well-completion technologies enable the extraction of oil and gas from previously unproductive formations

Advances Microearthquakes (that is those with magnitudes below 2) are routinely produced as part of the hydraulic fracturing (or ldquofrackingrdquo) process used to stimulate the production of oil but the process as currently practiced appears to pose a low risk of inducing destructive earthquakes More than 100000 wells have been subjected to fracking in recent years and the largest induced earthquake was magnitude 36 which is too small to pose a serious risk Yet wastewater disposal by injection into deep wells poses a higher risk because this practice can induce larger earthquakes For example several of the largest earthquakes in the US midcontinent in 2011 and 2012 may have been triggered by nearby disposal wells The largest of these was a magnitude 56 event in central Oklahoma that destroyed 14 homes and injured two people The mechanism responsible for inducing these events appears to be the well-understood process of weakening a preexisting fault by elevating the fl uid pressure However only a small fraction of the more than 30000 wastewater disposal wells appears to be problematicmdashtypically those that dispose of very large volumes of water andor communicate pressure perturbations directly into basement faults

Outlook Injection-induced earthquakes such as those that struck in 2011 clearly contribute to the seismic hazard Quantifying their contribution presents diffi cult challenges that will require new research into the physics of induced earthquakes and the potential for inducing large-magnitude events The petroleum industry needs clear requirements for operation regulators must have a solid scientifi c basis for those requirements and the public needs assurance that the regulations are suffi cient and are being followed The current regulatory frameworks for wastewater disposal wells were designed to protect potable water sources from contamination and do not address seis-

mic safety One consequence is that both the quantity and timeliness of information on injection volumes and pres-sures reported to regulatory agencies are far from ideal for managing earthquake risk from injection activities In addition seismic monitoring capabilities in many of the areas in which wastewater injection activities have increased are not capable of detecting small earthquake activity that may presage larger seismic events

ARTICLE OUTLINE

Mechanics of Induced Earthquakes

Earthquakes Induced by Hydraulic Fracturing

Earthquakes Induced by Deep Injection

Lessons from Three Case Studies of Deep

High-Volume Injection

Other Causes of Induced Earthquakes

Hazard and Risk of Induced Earthquakes

Unknown Knowns

Reducing the Risk of Injection-Induced

Earthquakes

ADDITIONAL RESOURCES

The following resources provide an introduction to earthquake hazards and risk the science of induced earthquakes and strategies for managing the risk

C Nicholson R L Wesson ldquoEarthquake hazard associated with deep well injection A report to the US Environmental Protection Agencyrdquo US Geol

Surv Bull 1951 (1990) httppubsusgsgovbul1951reportpdf

Committee on Induced Seismicity Potential in Energy Technologies Induced Seismicity Potential in Energy

Technologies (National Research Council Washing-ton DC 2012) httpdelsnaseduReportInduced-Seismicity-Potential-Energy-Technologies13355

S Horton Disposal of hydrofracking waste fl uid by injection into subsurface aquifers triggers earth-quake swarm in central Arkansas with potential for damaging earthquake Seismol Res Lett 83 250ndash260 (2012) doi101785gssrl832250

Tutorial material on probabilistic seismic hazard analysis (PSHA) wwwopenshaorgsitesopenshaorgfi lesPSHA_Primer_v2_0pdf

M D Zoback Managing the seismic risk posed by wastewater disposal Earth Magazine 57 38ndash43 (2012)

Earthquake Science Center US Geological Survey Menlo Park CA 94025 USAE-mail ellsworthusgsgov

Cumulative number of earthquakesM gt 3

1970 1980 1990 2000 2010

Earthquakes with magnitude (M) ge 3 in the US midcontinent 1967ndash2012 After decades of a steady earthquake rate (average of 21 eventsyear) activity increased starting in 2001 and peaked at 188 earthquakes in 2011 Human-induced earthquakes are suspected to be partially responsible for the increase

12 JULY 2013 VOL 341 SCIENCE wwwsciencemagorg 142

REVIEW SUMMARY

READ THE FULL ARTICLE ONLINE

httpdxdoiorg101126science1225942

Cite this article as W L Ellsworth Science 341 1225942 (2013) DOI 101126science1225942

Published by AAAS

Earthquakes in unusual locations have become an important topic of discussion in both NorthAmerica and Europe owing to the concern that industrial activity could cause damagingearthquakes It has long been understood that earthquakes can be induced by impoundment ofreservoirs surface and underground mining withdrawal of fluids and gas from the subsurfaceand injection of fluids into underground formations Injection-induced earthquakes have inparticular become a focus of discussion as the application of hydraulic fracturing to tight shaleformations is enabling the production of oil and gas from previously unproductive formationsEarthquakes can be induced as part of the process to stimulate the production from tight shaleformations or by disposal of wastewater associated with stimulation and production Here Ireview recent seismic activity that may be associated with industrial activity with a focus onthe disposal of wastewater by injection in deep wells assess the scientific understanding of inducedearthquakes and discuss the key scientific challenges to be met for assessing this hazard

Earthquakes are expected within tectoni-cally active regions such as along plateboundaries or within distributed zones of

deformation Recent seismic activity across thecoterminous United States for example concen-

trates along the plate boundaries of the WestCoast and within the intermountain West (Fig 1)Within such actively deforming zones elasticstrain energy accumulates in the crust sometimesfor centuries before being released in earth-quakes The potential for earthquakes also existswithin continental interiors despite very low de-formation rates (1) This is because shear stresslevels within the interior of plates or near plate

boundaries are commonly found to be near thestrength limit of the crust (2) Under these con-ditions small perturbations that effect fault sta-bility can and do trigger earthquakes (3ndash6) Forexample the injection of water under high pres-sure into impermeable basement rocks beneathBasel Switzerland to develop an enhanced geother-mal system beneath the city induced four momentmagnitude (Mw) 3 earthquakes in 2006 and 2007(7) (earthquake magnitudes measured using otherscales are denoted byM ) These small earthquakesled to the abandonment of the project loss of theinvestment and ongoing litigation over compen-sation for damage The extraction of natural gasfrom shallow deposits in the Netherlands alsocauses earthquakes (8) A recent M 34 event nearLoppersum damaged scores of homes in the arearesulting in large losses for the property owners (9)

Within the central and eastern United Statesthe earthquake count has increased dramaticallyover the past few years (Fig 2) More than 300earthquakes with M ge 3 occurred in the 3 yearsfrom 2010 through 2012 compared with an av-erage rate of 21 eventsyear observed from 1967to 2000 States experiencing elevated levels ofseismic activity included Arkansas Colorado NewMexico Ohio Oklahoma Texas and Virginia Thegreatest rise in activity occurred in 2011 when 188M ge 3 earthquakes occurred Although earthquake

REVIEW

Earthquake Science Center US Geological Survey MenloPark CA 94025 USA

E-mail ellsworthusgsgov

-80-90

Longitude

-100-110-120

Fig 1 Seismicity of the coterminous United States and surrounding re-gions 2009ndash2012 Black dots denote seismic events Only earthquakes with M ge 3are shown larger symbols denote events with M ge 4 Background colors give the

probability of peak ground acceleration with a 2 probability of exceedance in50 years from the US National Seismic Hazard Map (1) Red ge 1g orange 03to 1g yellow 01 to 03g light green 003 to 01g darker green 003 to 01g

wwwsciencemagorg SCIENCE VOL 341 12 JULY 2013 1225942-1

detection improved for M lt 3 as the USArraytransportable seismograph array began to passthrough the region starting in 2008 (10) a recentreport on seismicity in the central and easternUnited States found that the probability of missingM ge 3 earthquakes in the region has been nearzero for decades (11) Consequently the increasedearthquake count represents a temporal change inearthquake rate Because the hazard of damagingground shaking is fundamentally related to the rateof earthquake occurrence (1) regions where the rateincreased may be more hazardous than forecast bythe 2008 version of the US National SeismicHazard Map (Fig 1) (1) Understanding why seis-micity increased and how this increase affects thehazard have become a priority for the earthquake-research community

A number of these recent earthquakes occurredin areas where specific types of nearby industrialactivities raise the possibility that these events wereinduced by human activity Here I will use the termldquoinducedrdquo to include both earthquakes triggeredby anthropogenic causes that primarily releasetectonic stress and those that primarily releasestresses created by the industrial activity (4) Un-derstanding which earthquakes may have beeninduced and if so how are challenging problemsto solve in the current data-poor environment

Several examples since 2011 highlight thedifficulty in determining whether earthquakeswere induced by human activity The Mw 40earthquake on 31 December 2011 in YoungstownOhio appears to have been induced by injectionof wastewater in a deep Underground InjectionControl (UIC) class II well (12) The Mw 4727 February 2011 central Arkansas earthquakehas also been linked to deep injection of waste-water (13) The Mw 44 11 September 2011 earth-quake near Snyder Texas occurred in an oil fieldwhere injection for secondary recovery has beeninducing earthquakes for years (14) TheMw 4810 October 2011 earthquake near Fashing Texasoccurred in a region where long-term productionof gas has been linked to earthquake activity (15)For others such as the Mw 57 6 November 2011central Oklahoma earthquake (16) or theMw 4917 May 2012 east Texas earthquake (17) whereactive wastewater-injection wells are located neartheir respective epicenters the question of naturalversus induced remains an active topic of research

The potential association between deep waste-water disposal wells and earthquakes has receivedconsiderable attention due to the association ofthis activity with the development of tight shaleformations for gas and petroleum by hydraulicfracturing or ldquofrackingrdquo (5) Wells used in the USpetroleum industry to inject fluids are regulatedas UIC class II wells Approximately 110000 ofthese wells are used for enhanced oil recoveryIn addition 30000 class II wells in the UnitedStates are used for wastewater disposal Of thesewells most have no detected seismicity withintens of kilometers although a few are correlatedwith seismicity (18) However this can be saidwith confidence only for earthquakesMw ge 3 as

smaller earthquakes are not routinely reportedin the central and eastern United States So it ispossible that smaller earthquakes could be morecommon in the vicinity of these wells In Califor-nia where the completeness threshold is belowMw 2 the majority of the 2300 active wastewater-injection wells are located in regions of low seis-micity As with elsewhere in the United States asmall fraction of the California wastewater wellscoincide with earthquakes which raises the ques-tion of what factors distinguish those seismicallyactive wells from the majority of wells if theearthquakes and injection activities are related

Mechanics of Induced EarthquakesEarthquakes release stored elastic strain energywhen a fault slips A fault will remain locked aslong as the applied shear stress is less than thestrength of the contact The failure condition toinitiate rupture is usually expressed in terms ofthe effective stress tcrit = m(sn ndash P) + to where thecritical shear stress tcrit equals the product of thecoefficient of friction m and the effective normalstress given by the difference between the ap-plied normal stress sn and the pore pressure P(3 19 20) For almost all rock types m lies be-

tween 06 and 10 and the cohesive strength ofthe sliding surface to is negligible under typicalcrustal conditions Increasing the shear stress re-ducing the normal stress andor elevating the porepressure can bring the fault to failure triggeringthe nucleation of the earthquake (Fig 3) Once ini-tiated sliding resistance drops and seismic wavesradiate away driven by the imbalance between theelastic stress stored in the surrounding rock massand the frictional resistance of the dynamicallyweakened sliding surface Rupture will continue topropagate as long as the wave-mediated stress atthe rupture front exceeds the static strength andmay extend into regions where the ambient stressesare below the failure threshold

Rocks fail in tension when the pore pressureexceeds the sum of the least principal stress s3and the tensile strength of the rock forming anopening-mode fracture that propagates in theplane normal to s3 The industrial process of hy-draulic fracturing commonly involves both tensileand shear failure Depending on the local stressstate hydraulically conductive fractures may beinduced to fail in shear before P = s3 A successfulldquofrac jobrdquomay create a fracture network dominatedby pathways created by shear failure (21)

Longitude

1967 through 2012

1970 1980 1990 2000 2010

105 100 95 90 85

Fig 2 Cumulative count of earthquakes with M ge 3 in the central and eastern United States1967ndash2012 The dashed line corresponds to the long-term rate of 212 earthquakesyear (Inset)Distribution of epicenters in the region considered here

12 JULY 2013 VOL 341 SCIENCE wwwsciencemagorg1225942-2

REVIEW

Earthquakes are known to be induced by awide range of human activities (3ndash5) that modifythe stress andor pore pressure (Fig 3) At presentwith the use of seismological methods it is notpossible to discriminate between man-made andnatural tectonic earthquakes Induced earthquakessometimes occur at the source of the stress or pres-sure perturbation at other times these events takeplace deep below and kilometers away from thesource When removed from the source inducedearthquakes typically release stored tectonic stresson preexisting faults as do natural earthquakesSometimes induced events occur shortly after theindustrial activity begins but in other cases theyhappen long after it has been under way or evenceased Factors that should enhance the probabilityof a particular stress or pore-pressure perturbationinducing earthquakes include the magnitude ofthe perturbation its spatial extent ambient stresscondition close to the failure condition and thepresence of faults well oriented for failure in thetectonic stress field Hydraulic connection betweenthe injection zone and faults in the basement mayalso favor inducing earthquakes as the tectonicshear stress increases with depth in the brittle crust(2) In addition the larger the fault the larger themagnitude of earthquakes it can host

Methods for anticipating the time of failurehave long been the ldquoholy grailrdquo of seismology(22) Though short-term prediction remains anelusive goal it has been proposed that criticallyloaded faults have enhanced triggering suscep-tibility to dynamic stresses from distant earth-quakes (23) Specifically some but not all of thesites where fluid-injectionndashinduced earthquakesare suspected of contributing to the recent in-crease in seismicity in the midcontinent (Fig 2)experienced increased rates of microearthquakes

in the days immediately after three recent greatearthquakes (23)

Earthquakes Induced by Hydraulic FracturingThe industrial process of hydraulic fracturinginvolves the controlled injection of fluid underpressure to create tensile fractures thereby in-creasing the permeability of rock formations Ithas been used for well over half a century tostimulate the recovery of hydrocarbons Formany decades the primary application was toimprove the output of aging oil and gas reser-voirs Beginning in the late 1990s technologiesfor extracting natural gas and oil from tightshale formations led to the development of newnatural gas fields in many parts of the centraland eastern United States western Canada andEurope Global development of oil and gas fromshale will undoubtedly continue as the resourcepotential is high in many parts of the world

Extracting hydrocarbons from shale requiresthe creation of a network of open fractures con-nected to the borehole Horizontal drill holes ex-tending up to several kilometers within the shaleformation undergo a staged series of hydraulicfractures commonly pressurizing a limited sectionof the cased well at a time to stimulate the flowof gas or oil into the well Each stage involves thehigh-pressure injection of water into the formationFracking intentionally induces numerous micro-earthquakes the vast majority with Mw lt 1

Several cases have recently been reported inwhich earthquakes large enough to be felt but toosmall to cause structural damage were associateddirectly with fracking These cases are notable be-cause of the public concern that they raised de-spite maximum magnitudes far too small to causestructural damage Investigation of a sequence of

felt events with maximum M 29 in south centralOklahoma revealed a clear temporal correlationbetween fracking operations in a nearby well andthe seismic activity (24) Available data were insuf-ficient to definitely rule out a natural cause due tothe occurrence of some natural seismicity in thegeneral area In April and May 2012 a series ofinduced earthquakes with maximum M 23 oc-curred near Blackpool United Kingdom (25)during fracking to develop a shale gas reservoir

One of the major shale plays in the UnitedStatesmdashthe Marcellus Shale of the AppalachianBasin in Pennsylvania West Virginia Ohio andNew Yorkmdashlies within a region characterizedby low levels of natural seismic activity (Fig 1)The regional seismographic network operated byLamont Doherty Earth Observatory (LDEO) sys-tematically catalogs all earthquakes with M ge 2in Pennsylvania (Fig 4) Although thousands ofhydraulic fractures were done in Pennsylvaniasince major development of the field began in2005 only six earthquakes with M ge 2 were de-tected by the LDEO network within the footprintof the Marcellus Shale the largest of which wasjust M 23 The largest earthquake in the regionsince the development of shale gas happenedacross the Ohio border in Youngstown where itwas induced by injection (12) much of the fluidapparently coming from wells in Pennsylvania

Beginning in 2009 an unusual sequence ofearthquakes was noted in the Horn River Basin ofBritish Columbia including 21 events with Mw 30and larger Only the largest at Mw 36 was reportedas felt by workers in this remote area where it did nodamage (26) The investigation into the cause of theseevents by the BC Oil and Gas Commission (26)concluded that the events ldquowere caused by fluid in-jection during hydraulic fracturing in proximity ofpre-existing faultsrdquo Two of the hydrofrac treatmentswere recorded by dense seismometer deploymentsat the surface Precise hypocentral locations showedthat the induced earthquakes occurred on previouslyunknown faults located outside of the stimulationinterval that were well oriented for failure in the am-bient stress field Apparently fracture pressure wasquickly communicated through hydraulically conduc-tive pathways and induced slip on critically stressedfaults via reduction of the effective normal stress

Earthquakes Induced by Deep InjectionThere has been a growing realization that the prin-cipal seismic hazard from injection-induced earth-quakes comes from those associated with disposal ofwastewater into deep strata or basement formations(5) Before 2011 the Mw 48 event on 9 August1967 near Denver Colorado was the largest eventwidely accepted in the scientific community ashaving been induced by wastewater injection (5)The hazard landscape of what is possible has shifteddue to the role that wastewater injection into adepleted oil field may have played in the Mw 576 November 2011 central Oklahoma earthquake(16) although a consensus on its origin has not yetbeen reached (27) This earthquake damaged homesand unreinforced masonry buildings in the epicentral

Changes in solid stressdue to fluid extraction or injection

(poro-thermoelastic effects changes in gravitational loading)

Permeable reservoir

aquifer

Volume andor mass change

Permeable reservoiraquifer

Direct fluid pressure effects of injection

(fluid pressurediffusion)

Increase in pore pressure along fault (requires high-permeability pathway)

Change in loading conditions on fault (no direct hydrologic connection required)

Fig 3 Schematic diagram of mechanisms for inducing earthquakes Earthquakes may be in-duced by increasing the pore pressure acting on a fault (left) or by changing the shear and normalstress acting on the fault (right) See (4)

REVIEW

area and was felt as far as 1000 km away inChicago Illinois

The November 2011 central Oklahoma earth-quake sequence initiated very close to a pair ofwastewater-injection wells where disposal operationbegan 18 years earlier (16) No unusual seismicitywas detected in this historically quiet region whereonly a few events of M lt 2 were noted until aMw 41 earthquake occurred near the wells in early2010 Aftershocks of this event continued spo-radically through 2010 and into mid-2011 Thisdecaying sequence was shattered by a Mw 50earthquake on 5 November 2011 followed 20 hourslater by theMw 57 mainshock With the initiatingpoint of the November sequence within 15 km ofthe injection wells and some earthquake hypocen-ters at the same depth as injection the potential fora causal connection between injection and the earth-quakes is clear The long delay between the start ofinjection and the earthquakes however deviatesfrom the pattern seen in other documented casesof injection-induced seismicity such as the 2011Youngstown Ohio earthquake where there was atmost a few months of delay before induced seis-micity began In the Oklahoma case years of injec-tion may have been needed to raise the pore pressureabove the preproduction level in this depletedoil field before fault strength was exceeded (16)

Much of the concern about earthquakes andfracking centers on the injection of wastewatercomposed of flowback fluids and coproducedformation brine in deep wells and not on frack-ing itself Wastewater disposal appears to haveinduced both the 2011 central Arkansas earthquake(13) and the 2011 Youngstown Ohio earthquake(12) as mentioned above Unprecedented levelsof seismicity have also been seen in the BarnettShale in north central Texas where commercialdevelopment of shale gas was pioneered Sincedevelopment began in late 1998 nineearthquakes ofM ge 3 occurred comparedwith none in the preceding 25 years Anotable sequence occurred in the DallasndashFortWorth area fromOctober 2008 throughMay 2009 A detailed investigation of thissequence concluded that the earthquakeswere most probably caused by disposal ofshale gas wastewater in a UIC class IIdisposal well at the DallasFort WorthInternational Airport (28) although aswith the Oklahoma earthquake not allinvestigators agree that the case is proven(29) Because routine earthquake report-ing in the region is incomplete for eventsof M lt 3 the passage of the USArrayTransportable Array through the regionover an 18 month period in 2009ndash2011made it possible to improve magnitudecompleteness toM 15 and location accu-racy by several fold Epicenters for themost reliable locations were clustered ineight groups all within 3 km of high-rate(gt25000 m3month) wastewater-injectionwells (18) These results suggest thatthe injection rate as well as the total

volume of injection may be a predictor of seis-mic potential

Lessons from Three Case Studies of DeepHigh-Volume InjectionConclusions about the cause of many of the recentearthquakes suspected of being induced by injec-tion are complicated by incomplete information onthe hydrogeology the initial state of stress and porepressure the pumping history of the well(s) andwhere pressure changes are being communicated atdepth Routine earthquake locations with uncertain-ties of 5 to 10 km and a high magnitude-detectionthreshold are of limited use Three particularly well-documented cases of injection-induced seismicityfrom Colorado illustrate what can be learned whenmore is known about the pre-injection stress stateand seismicity as well as the injection history

Rocky Mountain ArsenalIn 1961 a deep injection well was drilled at theRocky Mountain Arsenal (RMA) northeast ofDenver Colorado to dispose of hazardous chem-icals produced at this defense plant (30) Withinseveral months of the start of routine injection inthe 36-km-deep well in March 1962 residents ofthe northeastern Denver area began to report earth-quakes and events registered on two nearby seis-mic stations Between the start of injection andits termination in February 1966 a total of 13 earth-quakes with body wave magnitudes (mb) 4 andlarger occurred The following year the three largestof the Denver earthquakes occurred including theMw 48 event on 9 August 1967 that caused minorstructural damage near the epicenter By this timethe earthquakes had migrated as far as 10 km fromthe injection point (31) Hydrologic modeling showedthat the migrating seismicity would track a criticalpressure front of 32 MPa (32) Although declining

earthquake activity continued for the next twodecades including a mb 43 earthquake on 2 April1981 The RMA earthquakes demonstrate how thediffusion of pore pressure within an ancient faultsystem can initiate earthquakes many kilometersfrom the injection point delayed by months oreven years after injection ceased

RangelyThe insights gained from RMA led to the sug-gestion that earthquakes could be controlled bymodulating the fluid pressure in the fault ac-cording to the effective-stress relation (19) In 1969the US Geological Survey (USGS) began an ex-periment to test the effective-stress hypothesis inthe Rangely oil field in northwestern Colorado(20) Water injected into the reservoir under highpressure had been used to enhance oil productionat Rangely since 1957 The operator Chevron OilCompany gave USGS permission to regulate thefluid pressure in a portion of the field that wasknown to be seismically active Laboratory mea-surements of the coefficient of friction on coresamples of the reservoir rocks and in situ deter-mination of the state of stress led to the predictionthat a critical fluid pressure of 257 MPa would berequired to induce earthquakes Two cycles of fluidinjection and withdrawal were conducted between1969 and 1973 When the pressure in a monitoringwell exceeded the target pressure earthquake ac-tivity increased when pressure was below thethreshold earthquake activity decreased In partic-ular the earthquake activity ceased within 1 day ofthe start of backflow in May 1973 providing strongevidence that the rate of seismicity could be con-trolled by adjusting the pore pressure at the depthwhere earthquakes initiate if stress conditions andthe strength of the faulted rock mass were knownThe rapid response of seismicity at the onset of

backflow also emphasized the importanceof understanding the geohydrology and inparticular the importance of hydraulicallyconductive faults and fractures for trans-mitting pore pressure within the system

Paradox ValleyAn ongoing fluid-injection project has beenunder way since 1996 in Paradox Valley insouthwestern Colorado where the salineshallow water table is being suppressedby pumping to prevent salt from enteringthe Dolores River as it crossed the valleyand eventually the Colorado River furtherdownstream (33) In its natural state theDolores River picks up salt from thegroundwater as it crosses Paradox ValleyAfter extensive study of alternatives theUS Bureau of Reclamation determinedthat high-pressure injection of brine intoa deep disposal well (UIC class V) pro-vided the best method for reducing thesalinity of the Dolores River Injection oc-curs in a tight but highly fractured dolo-mitic limestone with a fracture-dominatedporosity of less than 6 located 43 km

PENNSYLVANIAPENNSYLVANIAPENNSYLVANIAPENNSYLVANIAYLVAYLVAPENNSYLVANIA

NEW YORK

WEST VIRGINIA DELAWARE

NEW YORK

NEW YORKNEW YORKNEW YORKNEW YORKNEW YORK

OHIOOHIOOHIOOHIOOHIO

WEST VIRGINIAWEST VIRGINIAWEST VIRGINIAWEST VIRGINIAWEST VIRGINIA REREREREDELAWARDEL WARDEL WARDEL WARDELAWARDELAWARDELAWARDELAWARRRRRDELAWARE

42deg30N

40degN

80degW 75degW77deg30W

bullbullbullbullbullbullbullbullbullbull

bullbullbullbullbullbull bull

bullbull

bullbullbull bullbullbullbullbullbull bull bullbullbull bull

bullbullbull

bullbull

bullbull bullbullbullbull

bullbullbullbullbull bull

bullbullbull

bullbull

bullbullbull bullbullbullbullbull

bullbullbullbullbull

bullbullbull

bullbullbullbullbull bullbullbullbull

bullbullbull

bullbullbullbull

PENNSYLVANIA

NEW YORK

WEST VIRGINIA

MARYLAND

DELAWARE

VIRGINIA

Fig 4 Seismicity of Pennsylvania and surrounding regions1970ndash2012 Shading indicates areas underlain by deposits of theMarcellus Shale Blue dots earthquakes before 2005 red dots after2005 Seismicity was determined by the Lamont Doherty Earth Ob-servatory (45)

REVIEW

below land surface To date more than 7 times 106 m3

of brine have been injected One operational ob-jective based on both the RMA and Rangelyexperiences was the need to minimize the mag-nitude of earthquakes induced by injection

A local seismic network was established in1985 to determine background levels of seis-micity before the drilling of the well and initialinjection tests Between 1985 and June 1996only three tectonic earthquakes were detectedwithin 15 km of the well and just 12 within 35 km(33) However hundreds of earthquakes were in-duced during injection tests conducted between1991 and 1995 Most of these earthquakes wereconcentrated within 1 km of the injection pointalthough a few were located 3 to 4 km from thissite All events were below M 3 The occurrence

of induced earthquakes is not notable here asinjection required a bottom hole pressure in ex-cess of the hydraulic fracture pressure of 70 MPa

High injection pressure was needed to keeppace with the disposal requirements consequent-ly induced earthquakes were expected when dis-posal operations went into production in 1996Continuous monitoring of injection pressures andvolumes along with seismicity is being conductedto insure the safe operation of the project Duringthe first few years of operations several of the in-duced earthquakes exceeded M 3 necessitatingchanges in injection procedures in an attempt tolimit the maximum magnitude The dimension ofthe activated zone also grew with earthquakes asfar as 8 km from the injection point appearingwithin a year and events to beyond 12 km several

years later (Fig 5) Because seismicity rapidlyabated after each injection test it was hypothe-sized that occasional shutdowns of 20 days wouldallow the fluid pressure to equilibrate reducing thepotential for larger events (33) By itself thisprocedure proved inadequate as a M 43 eventwas induced in May 2000

After this earthquake a new procedure wasintroduced in 2000 that involved periodic 20-dayshutdowns and a 33 reduction in the injec-tion volume which initially reduced the requiredbottom hole pressure to 78 MPa Over the fol-lowing decade the pressure required to inject thatvolume steadily increased to more than 84 MPain 2012 drawing the revised strategy into ques-tion as a steadily increasing injection pressure isnot sustainable in the long term On 24 January

108deg45W

109degW

108deg45W

38deg15N 38deg15N

0 63 KilometersN

Paradox

Injection well

Dolores River

Paradox Valley

Fig 5 Seismicity near Paradox Valley Colorado The US Bureau of Reclamationextracts saline groundwater from shallow wells where the Dolores River crossesParadox Valley to prevent its entry into the Colorado River system Since 1996 the

brine has been disposed of by injection into a 43-km-deep UIC class V wellInjection has induced more than 1500 earthquakes with M ge 1 including the Mw39 earthquake on 25 January 2013 which was located 8 km northwest of the well

REVIEW

2013 a Mw 39 earthquake occurred 8 km north-west of the well in a previously active clustercausing strong shaking in the town of ParadoxColorado (Fig 5) As a consequence injectionwas halted for 12 weeks before restarting at areduced rate The Paradox Valley experience il-lustrates how long-term high-volume injectioncan lead to the continued expansion of the seis-mically activated region and the triggering oflarge-magnitude events many kilometers fromthe injection well more than 15 years after ob-servation of the initial seismic response This casestudy also illustrates the challenges for manag-ing the risk once seismicity has been induced

Other Causes of Induced EarthquakesAccording to the effective-stress model describedabove earthquakes can be induced by either re-ducing the effective normal stress or raising theshear stress (3ndash5) It has been known for decadesthat large reservoirs can induce earthquakes eitherfrom the effect of the elastic load of the reservoiror by diffusion of elevated pore pressure (34)Well-known examples include the deadly 1967M 63 earthquake in Koyna India (35) Yetestablishing a causal connection can be difficultwhen natural seismicity occurs nearby For ex-ample the debate about the role of the Zipingpureservoir in triggering theMw 79 2005WenchuanChina earthquake may never be resolved (36 37)What is clear however is that deep reservoirsin tectonically active zones carry a real risk ofinducing damaging earthquakes

Earthquakes throughout the world are alsorecognized to be associated with mining petro-leum and gas production and geothermal energyextraction Withdrawal of large volumes of fluid orgas from a reservoir or creation of a void space ina mine may modify the state of stress sufficientlyto induce earthquakes that relax the stress pertur-bations (4) Production may also release tectonicstress The long-term pumping of groundwater mayhave induced the deadly Mw 51 earthquake inLorca Spain on 11 May 2011 (38) Pore-pressurechanges alone can also induce seismicity such asby waterflooding for secondary recovery of oil orto maintain the fluid level in a geothermal reser-voir or when a mine is abandoned and allowed toflood (3 4) The physical connection between op-erational parameters such as injected volume andthe seismic response can be complex In the SaltonSea Geothermal Field for example the seismicityrate positively correlates with the net volume ofproduced fluid (extraction minus injection) ratherthan net injection as would be expected if seis-micity rate simply tracked pore pressure (39) Thisunderscores the importance of geomechanical mod-eling for transferring understandings developedin one setting to others

Hazard and Risk of Induced EarthquakesThe hazard from earthquakes depends on proximityto potential earthquake sources their magnitudesand rates of occurrence and is usually expressedin probabilistic terms (1 40) The US National

Seismic Hazard Map for example gives the ex-ceedance probabilities for a variety of ground-motion measures from which the seismic designprovisions in the building codes are derived (Fig 1)(1) Our understanding of the hazard will evolveas new information becomes available about theunderlying earthquake sources which are ideallyderived from a combination of fault-based infor-mation and historical seismicity Accounting for thehazard of induced earthquakes however presentssome formidable challenges

In the current US map (Fig 1) for examplethe estimated hazard in most parts of the centraland eastern regions of the country derives exclu-sively from historical seismicity How should in-creases in the earthquake rate since 2009 (Fig 2)be incorporated in the model Should identifiedor suspected induced earthquakes be treated thesame as or differently than natural events Inparticular do induced earthquakes follow thesame magnitude-frequency distribution modelsas natural earthquakes This issue has particularimportance as the high end of the magnitude dis-tribution where events are infrequent contributesdisproportionately to both the hazard and riskAlthough injection-induced earthquakes have doneonly minor damage in the United States to date(5) the 2011 central Oklahoma earthquake wasthe same magnitude as the 1986 San SalvadorEl Salvador tectonic earthquake that killed morethan 1500 people injured more than 10000 andleft 100000 homeless (41) Losses on this scaleare unlikely in North America and northern Europewhere a catastrophic building collapse in aMw 57earthquake is unlikely but the same cannot be saidfor large portions of the world where nonductileconcrete frame or unreinforced masonry buildingsare prevalent The earthquake that killed nine andcaused serious damaged Lorca Spain was evensmaller at Mw 51 (40) The heavy losses in thispossibly induced earthquake resulted from the ex-posure of many fragile buildings to strong shakingfrom this very shallow-focus earthquake (42) Thisevent should serve as a reminder that risk is theproduct of the hazard exposure and vulnerability

Unknown KnownsIgnorance of the things that we understand weshould know but do not leaves us vulnerableto unintended consequences of our actions Theeffective-stress model provides straightforward guid-ance for avoiding induced earthquakes but re-quires knowledge that we rarely possess of thestress state and pore pressure acting on the faultQuantitative predictions from the model dependon knowing initial stress and pore-pressure con-ditions and how perturbations to those conditionsdue to injection will affect the surroundings Forexample pore-pressure changes in a fault kilome-ters from the injection point depend on the hy-drologic characteristics of connecting pathwaysthat will in all likelihood be poorly known Theseismic response might not take place immediatelyand decades may elapse before a damaging eventoccurs as illustrated by the recent Paradox Valley

earthquake and possibly the central Oklahomaearthquake as well Simply injecting water by grav-ity feed (pouring it down the well with no surfacepressure) sounds safe enough But if the deep aquifersystem was originally underpressured and the faultswere in frictional equilibrium with the stress (2)this apparently benign type of injection can bringfaults to failure by raising the water table and hencethe pore pressure acting on the faults

The fact that the great majority of UIC class IIinjection wells in the United States appear to beaseismic at least for earthquakesMw gt 3 suggeststhat ambient conditions in geologic formationscommonly approved for disposal are far enoughremoved from failure that injection can be donewith low risk provided that the pressure perturba-tion remains confined within the intended forma-tion The largest injection-induced events have allinvolved faulting that is considerably deeper thanthe injection interval (13 16 30 43) suggestingthat transmission of increased pressure into thebasement elevates the potential for inducing earth-quakes Consequently detection of seismicity inthe vicinity of the well or changes in seismicity inthe neighborhood should prompt reevaluationof the hazard

License and operational requirements for UICclass II wells in the United States are regulatedunder the Safe Drinking Water Act by the USEnvironmental Protection Agency or by dele-gation of authority to state agencies The lawrsquosprovisions are primarily directed toward protec-tion of potable aquifers by requiring injection intoformations deep below and geologically isolatedfrom drinking water sources As such the lawfocuses on well integrity protection of imperme-able barriers above the injection zone and settingoperational injection pressure limits to avoid hy-draulically fracturing the well Diffusion of porepressure into basement faults or injection pres-sure that would raise critically stressed faults tofailure is not considered in US federal regula-tions From a scientific standpoint measuring theinitial stress state and pore pressure tracking ofinjection history and careful seismic monitoringwould be of great value At present little more isrequired by regulation than an estimate of thefracture pressure (not to be exceeded) and monthlyreporting of total injection volume and averageinjection pressure In most cases this informationis not sufficient to apply the effective-stress mod-el or gain an understanding of the hazard posedby injection activity

Reducing the Risk ofInjection-Induced EarthquakesHow can the risk of inducing damaging earth-quakes through human activity be minimizedin an information-poor environment Long-termand high-volume injection in deep wells clearlycarries some risk (18) even though most wellsare apparently aseismic (5) In contrast earth-quakes induced during hydraulic fracturing havelower risk because of their much smaller magni-tudes The largest fracking-induced earthquakes

REVIEW

(24 26) have all been below the damage thresh-old for modern building codes

One approach for managing the risk ofinjection-induced earthquakes involves settingseismic activity thresholds that prompt a reduc-tion in injection rate or pressure or if seismicactivity increases further suspension of injec-tion (44) Such ldquotraffic-lightrdquo systems have beenused selectively going back to at least the RMAwell pump tests in 1966ndash1967 The traffic-lightsystem used in in Basel Switzerland (7) did notstop the four Mw 3 earthquakes from happeningbut might have prevented larger events The de-cision to stop injection in the Youngstown Ohiowell based on the seismicity (12) and made theday before theMw 40 event resulted in seismicitynear the well declining within a month All ofthese examples feature better seismic monitoringcapabilities than currently exist in much of theUnited States or most of the rest of the worldLowering the magnitude-detection threshold inregions where injection wells are concentrated tobelow Mw 2 would certainly help as a traffic-lightsystem using the current US detection thresholdofMw 3 in many of these areas would have limitedvalue Improvements in the collection and time-liness of reporting of injection data to regulatoryagencies would provide much-needed informationon hydrologic conditions potentially associated withinduced seismicity In particular daily reporting ofvolumes peak and mean injection pressures wouldbe a step in the right direction as would measure-ment of the pre-injection formation pressure

Ultimately better knowledge of the stress andpressure conditions at depth the hydrogeolgicframework including the presence and geometry offaults and the location and mechanisms of naturalseismicity at a few sites will be needed to developa predictive understanding of the hazard posedby induced earthquakes Industry regulatory agen-cies and the public are all aware that earthquakescan be induced by fluid injection Industry needsclear requirements under which to operate regu-lators must have a firm scientific foundation forthose requirements and the public needs as-surance that the regulations are adequate and arebeing observed

References and Notes1 M D Petersen A D Frankel S C Harmsen

C S Mueller K M Haller R L Wheeler R L WessonY Zeng O S Boyd D M Perkins N Luco E H FieldC J Wills K S Rukstales ldquoDocumentation for the 2008update of the United States National Seismic Hazard MaprdquoUS Geol Surv Open-File Rep 2008ndash1128 (2008)

2 J Townend M D Zoback How faulting keeps the cruststrong Geology 28 399ndash402 (2000) doi 1011300091-7613(2000)28lt399HFKTCSgt20CO2

3 C Nicholson R L Wesson ldquoEarthquake hazard associatedwith deep well injection A report to the US EnvironmentalProtection Agencyrdquo US Geol Surv Bull 1951 (1990)httppubsusgsgovbul1951reportpdf

4 A McGarr D Simpson L Seeber ldquoCase histories ofinduced and triggered seismicityrdquo InternationalHandbook of Earthquake and Engineering Seismology(Academic Press Waltham MA 2002) vol 8 chap 40

5 Committee on Induced Seismicity Potential in EnergyTechnologies Induced Seismicity Potential in EnergyTechnologies (National Research Council Washington

DC 2012) httpdelsnaseduReportInduced-Seismicity-Potential-Energy-Technologies13355

6 K F Evans A Zappone T Kraft N Deichmann F MoiaA survey of the induced seismic responses to fluidinjection in geothermal and CO2 reservoirs in EuropeGeothermics 41 30ndash54 (2012) doi 101016jgeothermics201108002

7 N Deichmann D Giardini Earthquakes induced by thestimulation of an enhanced geothermal system belowBasel (Switzerland) Seismol Res Lett 80784ndash798 (2009) doi 101785gssrl805784

8 T van Eck F Goutbeek H Haak B Dost Seismic hazarddue to small-magnitude shallow-source inducedearthquakes in The Netherlands Eng Geol 87 105ndash121(2006) doi 101016jenggeo200606005

9 J Tagliabue ldquoParts of low country are now quakecountryrdquo New York Times 26 March 2013 p A6

10 F L Vernon L Astiz Seismicity in the mid-continentalUS as recorded by the Earthscope USArray TransportableArray Geol Soc Am Abstr Programs 44 511 (2012)

11 ldquoTechnical report Central and eastern United Statesseismic source characterization for nuclear facilitiesrdquoAppendix B (Electric Power Research Institute Palo AltoCA US Department of Energy and US NuclearRegulatory Commission Washington DC 2012)

12 W-Y Kim Induced seismicity associated with fluidinjection into a deep well in Youngstown OhioJ Geophys Res 101002jgrb50247 (2013)

13 S Horton Disposal of hydrofracking waste fluid byinjection into subsurface aquifers triggers earthquakeswarm in central Arkansas with potential for damagingearthquake Seismol Res Lett 83 250ndash260 (2012)doi 101785gssrl832250

14 S D Davis W D Pennington Induced seismicdeformation in the Cogdell oil field of west TexasBull Seismol Soc Am 79 1477ndash1495 (1989)

15 W D Pennington S D Davis The evolution of seismicbarriers and asperities caused by the depressuring offault planes in oil and gas fields of south TexasBull Seismol Soc Am 76 939ndash948 (1986)

16 K M Kerenan H M Savage G A Abers E S CochranPotentially induced earthquakes in Oklahoma USA Linksbetween wastewater injection and the 2011 Mw 57earthquake sequence Geology 41 699ndash702 (2013)doi 101130G340451

17 W A Brown C Frohlich Investigating the cause of the17 May 2012 M 48 earthquake near Timpson east Texas(abstr) Seismol Res Lett 84 374 (2013)

18 C Frohlich Two-year survey comparing earthquakeactivity and injection-well locations in the BarnettShale Texas Proc Natl Acad Sci USA 10913934ndash13938 (2012) doi 101073pnas1207728109pmid 22869701

19 M K Hubbert W W Rubey Role of fluid pressurein mechanics of overthrust faulting Geol SocAm Bull 70 115ndash206 (1959) doi 1011300016-7606(1959)70[115ROFPIM]20CO2

20 C B Raleigh J H Healy J D BredehoeftAn experiment in earthquake control at RangelyColorado Science 191 1230ndash1237 (1976)doi 101126science19142331230 pmid 17737698

21 C Barton D Moos K Tezuka Geomechanicalwellbore imaging Implications for reservoir fracturepermeability AAPG Bull 93 1551ndash1569 (2009)doi 10130606180909030

22 W H Bakun et al Implications for prediction andhazard assessment from the 2004 Parkfield earthquakeNature 437 969ndash974 (2005) doi 101038nature04067 pmid 16222291

23 N J van der Elst H M Savage K M KeranenG A Abers Enhanced remote earthquake triggering atfluid-injection sites in the midwestern United StatesScience 341 164ndash167 (2013)

24 A Holland Earthquakes triggered by hydraulic fracturingin south-central Oklahoma Bull Seismol Soc Am 1031784ndash1792 (2013)

25 C A Green P Styles ldquoPreese Hall shale gas fracturingReview and recommendations for induced seismicitymitigationrdquo (2012) wwwgovukgovernmentuploads

systemuploadsattachment_datafile157455075-preese-hall-shale-gas-fracturing-reviewpdf

26 ldquoInvestigation of observed seismicity in the Horn RiverBasinrdquo (BC Oil and Gas Commission Victoria BritishColumbia Canada 2012) wwwbcogccanode8046downloaddocumentID=1270

27 G R Keller A Holland ldquoOklahoma GeologicalSurvey evaluation of the Prague earthquake sequenceof 2011rdquo (Oklahoma Geological Survey Norman OK2013) wwwogsoueduearthquakesOGS_PragueStatement201303pdf

28 C Frohlich C Hayward B Stump E PotterThe DallasndashFort Worth earthquake sequenceOctober 2008 through May 2009 Bull Seismol Soc Am101 327ndash340 (2011) doi 1017850120100131

29 E Janskaacute L Eisner Ongoing seismicity in the Dallas-FortWorth area Leading Edge 31 1462ndash1468 (2012)doi 101190tle311214621

30 J H Healy W W Rubey D T Griggs C B RaleighThe Denver earthquakes Science 161 1301ndash1310(1968) doi 101126science16138481301pmid 17831340

31 R B Herrmann S-K Park The Denver earthquakes of1967-1968 Bull Seismol Soc Am 71 731ndash745(1981)

32 P A Hsieh J S Bredehoeft A reservoir analysis ofthe Denver earthquakes A case of induced seismicityJ Geophys Res 86 903ndash920 (1981) doi 101029JB086iB02p00903

33 J Ake K Mahrer Deep-injection and closely monitoredinduced seismicity at Paradox Valley ColoradoBull Seismol Soc Am 95 664ndash683 (2005)doi 1017850120040072

34 D W Simpson W S Leith Two types of reservoir-inducedseismicity Bull Seismol Soc Am 78 2025ndash2040 (1988)

35 H Gupta A review of recent studies of triggeredearthquakes by artificial water reservoirs with specialemphasis on earthquakes in Koyna India Earth Sci Rev58 279ndash310 (2002) doi 101016S0012-8252(02)00063-6

36 S Ge M Liu N Lu J W Godt N Luo Did theZipingpu Reservoir trigger the 2008 Wenchuanearthquake Geophys Res Lett 36 L20315 (2009)doi 1010292009GL040349

37 K Deng et al Evidence that the 2008 Mw 79 Wenchuanearthquake could not have been induced by the ZipingpuReservoir Bull Seismol Soc Am 100 2805ndash2814(2010) doi 1017850120090222

38 P J Gonzaacutelez K F Tiampo M Palano F CannavoacuteJ Fernaacutendez The 2011 Lorca earthquake slip distributioncontrolled by groundwater crustal unloading Nat Geosci5 821ndash825 (2012) doi 101038ngeo1610

39 E E Brodsky L J Lajoie Anthropogenic seismicity ratesand operational parameters at the Salton Sea GeothermalField Science 101126science1239213 (2013)doi 101126science1239213

40 R K McGuire ldquoSeismic hazard and risk analysisrdquoEarthquake Engineering Research Institute MonographMNO-10 (2004)

41 D H Harlow R A White The San Salvador earthquakeof 10 October 1986 and its historical context BullSeismol Soc Am 83 1143ndash1154 (1993)

42 J-P Avouac Earthquakes Human-induced shakingNat Geosci 5 763ndash764 (2012) doi 101038ngeo1609

43 J L Rubinstein W L Ellsworth The 2001 -present triggered seismicity sequence in the Ratonbasin of southern ColoradoNorthern New MexicoSeismol Res Lett 84 374 (2013)

44 M D Zoback Managing the seismic risk posed bywastewater disposal Earth Magazine 57 38ndash43 (2012)

45 wwwldeocolumbiaeduLCSNindexphp

Acknowledgments This paper is a contribution of theUSGSrsquos John Wesley Powell Center Working Group onUnderstanding Fluid Injection Induced Seismicity I thankL Block C Frohlich S Hickman W Leith A McGarrJ Rubinstein and an anonymous reviewer for insightfulcomments during the development of this paper

101126science1225942

REVIEW

contentscivol341issue6142pdf1225942pdf

Injection-Induced Earthquakes




Lessons from Three Case Studies of Deep High-Volume Injection

Rocky Mountain Arsenal

Rangely

Paradox Valley



Unknown Knowns

Reducing the Risk of Injection-Induced Earthquakes

References and Notes

Background Human-induced earthquakes have become an important topic of political and scientifi c discussion owing to the concern that these events may be responsible for widespread damage and an overall increase in seismicity It has long been known that impoundment of reservoirs surface and underground mining withdrawal of fl uids and gas from the subsurface and injection of fl uids into underground formations are capable of inducing earthquakes In particular earthquakes caused by injection have become a focal point as new drilling and well-completion technologies enable the extraction of oil and gas from previously unproductive formations

Advances Microearthquakes (that is those with magnitudes below 2) are routinely produced as part of the hydraulic fracturing (or ldquofrackingrdquo) process used to stimulate the production of oil but the process as currently practiced appears to pose a low risk of inducing destructive earthquakes More than 100000 wells have been subjected to fracking in recent years and the largest induced earthquake was magnitude 36 which is too small to pose a serious risk Yet wastewater disposal by injection into deep wells poses a higher risk because this practice can induce larger earthquakes For example several of the largest earthquakes in the US midcontinent in 2011 and 2012 may have been triggered by nearby disposal wells The largest of these was a magnitude 56 event in central Oklahoma that destroyed 14 homes and injured two people The mechanism responsible for inducing these events appears to be the well-understood process of weakening a preexisting fault by elevating the fl uid pressure However only a small fraction of the more than 30000 wastewater disposal wells appears to be problematicmdashtypically those that dispose of very large volumes of water andor communicate pressure perturbations directly into basement faults

Outlook Injection-induced earthquakes such as those that struck in 2011 clearly contribute to the seismic hazard Quantifying their contribution presents diffi cult challenges that will require new research into the physics of induced earthquakes and the potential for inducing large-magnitude events The petroleum industry needs clear requirements for operation regulators must have a solid scientifi c basis for those requirements and the public needs assurance that the regulations are suffi cient and are being followed The current regulatory frameworks for wastewater disposal wells were designed to protect potable water sources from contamination and do not address seis-

mic safety One consequence is that both the quantity and timeliness of information on injection volumes and pres-sures reported to regulatory agencies are far from ideal for managing earthquake risk from injection activities In addition seismic monitoring capabilities in many of the areas in which wastewater injection activities have increased are not capable of detecting small earthquake activity that may presage larger seismic events

ARTICLE OUTLINE

Lessons from Three Case Studies of Deep

High-Volume Injection

Unknown Knowns

Reducing the Risk of Injection-Induced

Earthquakes

ADDITIONAL RESOURCES

The following resources provide an introduction to earthquake hazards and risk the science of induced earthquakes and strategies for managing the risk

C Nicholson R L Wesson ldquoEarthquake hazard associated with deep well injection A report to the US Environmental Protection Agencyrdquo US Geol

Surv Bull 1951 (1990) httppubsusgsgovbul1951reportpdf

Committee on Induced Seismicity Potential in Energy Technologies Induced Seismicity Potential in Energy

Technologies (National Research Council Washing-ton DC 2012) httpdelsnaseduReportInduced-Seismicity-Potential-Energy-Technologies13355

S Horton Disposal of hydrofracking waste fl uid by injection into subsurface aquifers triggers earth-quake swarm in central Arkansas with potential for damaging earthquake Seismol Res Lett 83 250ndash260 (2012) doi101785gssrl832250

Tutorial material on probabilistic seismic hazard analysis (PSHA) wwwopenshaorgsitesopenshaorgfi lesPSHA_Primer_v2_0pdf

M D Zoback Managing the seismic risk posed by wastewater disposal Earth Magazine 57 38ndash43 (2012)

Earthquake Science Center US Geological Survey Menlo Park CA 94025 USAE-mail ellsworthusgsgov

1970 1980 1990 2000 2010

Earthquakes with magnitude (M) ge 3 in the US midcontinent 1967ndash2012 After decades of a steady earthquake rate (average of 21 eventsyear) activity increased starting in 2001 and peaked at 188 earthquakes in 2011 Human-induced earthquakes are suspected to be partially responsible for the increase

12 JULY 2013 VOL 341 SCIENCE wwwsciencemagorg 142

REVIEW SUMMARY

READ THE FULL ARTICLE ONLINE

httpdxdoiorg101126science1225942

Cite this article as W L Ellsworth Science 341 1225942 (2013) DOI 101126science1225942

Published by AAAS

REVIEW

-80-90

Longitude

-100-110-120

Longitude

1967 through 2012

1970 1980 1990 2000 2010

105 100 95 90 85

REVIEW

Permeable reservoir

aquifer

REVIEW

NEW YORK

42deg30N

40degN

bullbull

bullbullbull

bullbull

bullbullbull

bullbull

bullbullbull

bullbullbullbull

PENNSYLVANIA

NEW YORK

WEST VIRGINIA

MARYLAND

DELAWARE

VIRGINIA

REVIEW

108deg45W

109degW

108deg45W

38deg15N 38deg15N

0 63 KilometersN

Paradox

Injection well

Dolores River

Paradox Valley

REVIEW

101126science1225942

REVIEW







Rangely

Paradox Valley



Unknown Knowns



REVIEW

-80-90

Longitude

-100-110-120

Longitude

1967 through 2012

1970 1980 1990 2000 2010

105 100 95 90 85

REVIEW

Permeable reservoir

aquifer

REVIEW

NEW YORK

42deg30N

40degN

bullbull

bullbullbull

bullbull

bullbullbull

bullbull

bullbullbull

bullbullbullbull

PENNSYLVANIA

NEW YORK

WEST VIRGINIA

MARYLAND

DELAWARE

VIRGINIA

REVIEW

108deg45W

109degW

108deg45W

38deg15N 38deg15N

0 63 KilometersN

Paradox

Injection well

Dolores River

Paradox Valley

REVIEW

101126science1225942

REVIEW







Rangely

Paradox Valley



Unknown Knowns



Longitude

1967 through 2012

1970 1980 1990 2000 2010

105 100 95 90 85

REVIEW

Permeable reservoir

aquifer

REVIEW

NEW YORK

42deg30N

40degN

bullbull

bullbullbull

bullbull

bullbullbull

bullbull

bullbullbull

bullbullbullbull

PENNSYLVANIA

NEW YORK

WEST VIRGINIA

MARYLAND

DELAWARE

VIRGINIA

REVIEW

108deg45W

109degW

108deg45W

38deg15N 38deg15N

0 63 KilometersN

Paradox

Injection well

Dolores River

Paradox Valley

REVIEW

101126science1225942

REVIEW







Rangely

Paradox Valley



Unknown Knowns



Permeable reservoir

aquifer

REVIEW

NEW YORK

42deg30N

40degN

bullbull

bullbullbull

bullbull

bullbullbull

bullbull

bullbullbull

bullbullbullbull

PENNSYLVANIA

NEW YORK

WEST VIRGINIA

MARYLAND

DELAWARE

VIRGINIA

REVIEW

108deg45W

109degW

108deg45W

38deg15N 38deg15N

0 63 KilometersN

Paradox

Injection well

Dolores River

Paradox Valley

REVIEW

101126science1225942

REVIEW







Rangely

Paradox Valley



Unknown Knowns



NEW YORK

42deg30N

40degN

bullbull

bullbullbull

bullbull

bullbullbull

bullbull

bullbullbull

bullbullbullbull

PENNSYLVANIA

NEW YORK

WEST VIRGINIA

MARYLAND

DELAWARE

VIRGINIA

REVIEW

108deg45W

109degW

108deg45W

38deg15N 38deg15N

0 63 KilometersN

Paradox

Injection well

Dolores River

Paradox Valley

REVIEW

101126science1225942

REVIEW







Rangely

Paradox Valley



Unknown Knowns



108deg45W

109degW

108deg45W

38deg15N 38deg15N

0 63 KilometersN

Paradox

Injection well

Dolores River

Paradox Valley

REVIEW

101126science1225942

REVIEW







Rangely

Paradox Valley



Unknown Knowns



REVIEW

101126science1225942

REVIEW







Rangely

Paradox Valley



Unknown Knowns



101126science1225942

REVIEW







Rangely

Paradox Valley



Unknown Knowns



Documents

Injection-Induced Earthquakes William L. Ellsworth …users.clas.ufl.edu/prwaylen/GEO2200 Readings/Readings/Fracking... · Injection-Induced Earthquakes William L. Ellsworth Background: