Rebecca Harrington

RebeccaHarringtonPhD(UCLA)2008,Postdoc(KIT)2008-2013,McGill2013–2016RuhrUniversitätBochum

2017-Present

TheRoleofFluidsinTriggering“Unconven?onal“Earthquakes

Sliponafaultduringanearthquake

✭Depth

into theearth

Surface of the earth

Distance along the fault plane 100 km

Sliponanearthquakefaultsecond2.0












P/%,&,QI"&-J&)*J-(R%S-*&);&)*&%&;");R-+(%RT&

P/%,&,QI"&-J&)*J-(R%S-*&);&)*&%&;");R-+(%RT&

!"&1<K(%S-*@A6&

#$&1;)U"6&

V("LK"*$Q&1'U6&

HI"$,(%E&%RIE),K

<"&

H)U"&%*<&<K(%S-*&;$%E"&%;&&&&J-(&,"$,-*)$&"%(,/LK%M";&&

M0 !!3, M0 ! fc

"3

V%KE,&(KI,K(";&W),/&%&<K(%S-*X&&&&0K(%S-*&("E%,"<&,-&J$&#QX&

&

! =LV

! = fc!1

?)R"&1;"$6&D(-K

*<&R-S

-*&

Howdosourceparametersscalewithsize?

KanamoriandBrodsky,2004

€

M0 ∝τ3

DatasuggestsduraSonscalesas

0 2 4 6 8 10 12 14 16!4000

!2000

0

2000

4000

disp

. (nm

)

time (sec)

10!1 100 101 102

100

105

frequency (Hz)

!0 (n

ms)

dataBrune fitconfidence int.

Z,-&/"EI&K*<"(;,%*<&)J&"%(,/LK%M";&#"/%["&)*&,/"&;%R"&W%Q&("+%(<E";;&-J&;)U"&-(&W/"("&,/"Q&/%II"*N&&

!to understand earthquake scaling relationships (key to estimating hazard)&&

'-W&);&"*"(+Q&I%(SS-*"<&)*&"%(,/LK%M";T&&

P/Q&;/-KE<&W"&$%("T&&&

1464 H Kanamori and E E Brodsky

Figure 18. Illustration of simple stress release patterns during faulting. (a) ——: simple case ofimmediate stress drop. - - - -: general case without slip-weakening. (b) Slip-weakening model:hatched and cross-hatched areas indicate the fracture energy and frictional energy loss, respectively.(c) The energy budget: hatched, cross-hatched and dotted areas indicate the fracture energy, thermal(frictional) energy and radiated energy, in that order. All the figures are shown for unit area of thefault plane.

An earthquake is viewed as a stress release process on a fault surface S . The solid linesin figure 18(a) show the simplest case. At the initiation of an earthquake, the initial (beforean earthquake) shear stress on the fault plane σ0 drops to a constant dynamic friction σf , andstays there, i.e. σf = σ1. If the condition for instability is satisfied (Brace and Byerlee (1966),Scholz (2002), also section 6.1.1), rapid fault slip motion begins and eventually stops. At theend, the stress on the fault plane is σ1 (final stress) and the average slip (offset) is D. Thedifference "σs = σ0 − σ1 is the static stress drop. During this process, the potential energy(strain energy plus gravitational energy) of the system, W0, drops to W1 = W0 − "W where"W is the strain energy drop, and the seismic wave is radiated carrying an energy ER. Fromequation (3.14),

"W = σ̄DS, (4.32)

where σ̄ = (σ0 + σ1)/2 is the average stress during faulting (section 3.1.4). Graphically, "W

(for unit area) is given by the trapezoidal area shown in figure 18(c).The variation of stress during faulting can be more complex than shown by the solid lines

in figure 18(a). For example, the stress may increase to the yield stress σY in the beginning ofthe slip motion (curve (1) in figure 18(a)) because of loading caused by the advancing rupture(figure 15(e)), or of a specific friction law such as the rate- and state-dependent friction law(Dieterich 1979) (figure 17). In fact, some seismological inversion studies have shown this

Gutenberg-Richter Log(N) = a - bM

Do small earthquakes behave like large earthquakes? Can a breakdown in source scaling give information about seismic rupture? &

&+*%*@$,#B"@C*%+/"$0*%:*@,"$*+%1*@+"$*<%:$/1%+*(+1/=$@1+D%

For large populations, source parameter scaling relationships roughly hold.

My focus: How does the faulting environment relate to the

source parameter scaling? Are earthquakes on some faults different than others?

•  Physical factors may account for some of the range in observations.

–  Fault geometry may be a factor. Faults become more smooth with increasing slip.

cumulative slip fault maturity/geometry

–  Fault frictional properties may also change rupture speed. (Fluids!)

1464 H Kanamori and E E Brodsky

Figure 18. Illustration of simple stress release patterns during faulting. (a) ——: simple case ofimmediate stress drop. - - - -: general case without slip-weakening. (b) Slip-weakening model:hatched and cross-hatched areas indicate the fracture energy and frictional energy loss, respectively.(c) The energy budget: hatched, cross-hatched and dotted areas indicate the fracture energy, thermal(frictional) energy and radiated energy, in that order. All the figures are shown for unit area of thefault plane.

An earthquake is viewed as a stress release process on a fault surface S . The solid linesin figure 18(a) show the simplest case. At the initiation of an earthquake, the initial (beforean earthquake) shear stress on the fault plane σ0 drops to a constant dynamic friction σf , andstays there, i.e. σf = σ1. If the condition for instability is satisfied (Brace and Byerlee (1966),Scholz (2002), also section 6.1.1), rapid fault slip motion begins and eventually stops. At theend, the stress on the fault plane is σ1 (final stress) and the average slip (offset) is D. Thedifference "σs = σ0 − σ1 is the static stress drop. During this process, the potential energy(strain energy plus gravitational energy) of the system, W0, drops to W1 = W0 − "W where"W is the strain energy drop, and the seismic wave is radiated carrying an energy ER. Fromequation (3.14),

"W = σ̄DS, (4.32)

where σ̄ = (σ0 + σ1)/2 is the average stress during faulting (section 3.1.4). Graphically, "W

(for unit area) is given by the trapezoidal area shown in figure 18(c).The variation of stress during faulting can be more complex than shown by the solid lines

in figure 18(a). For example, the stress may increase to the yield stress σY in the beginning ofthe slip motion (curve (1) in figure 18(a)) because of loading caused by the advancing rupture(figure 15(e)), or of a specific friction law such as the rate- and state-dependent friction law(Dieterich 1979) (figure 17). In fact, some seismological inversion studies have shown this

Y()\E"@<K$SE"&,(%*;)S-*&U-*"&&1A]&@&B]&MR&<"I,/6&

H");R-+"*)$&U-*"&1]&F&A]&MR&<"I,/6&

H/%EE-W&;");R-+"*)$1T6&U-*"&18&F&]&MR&<"I,/6&

^2*$-*["*S-*%E_&H");R)$&H)+*%E;&&

E,"<F('=%,#*%,$@'+(?/'%G/'*%H4I6JI%C1K%('%7@$CL*9<M%N@9(:/$'(@X&&H%*&5*<("%;&V%KE,&

HowfaultfricSonalproperSesinfluencerupture

DifferencesinwaveformsandscalingproperSescouldberelatedtochangesinfricSonbasedontremorlocaSons.

Tectonictremorwaveforms:“earthquakes”(low-frequencyevents,orLFEs)thathappendeepinthefaultzone

E,"<F('=%,#*%,$@'+(?/'%G/'*%H4I6JI%C1K%('%7@$CL*9<M%N@9(:/$'(@X&&H%*&5*<("%;&V%KE,&

H");R-+"*)$&c-*"&17@A]&MR6&>*,(%@IE%,"&"%(,/LK%M";&)*&,/"&3/%(E"[-)d&;");R)$&U-*"&

Theshallowseismogeniczone(<5km)

Roleoffluids?Usinginducedseismicity

O/@9X&%&R%+*),K<"@I("<)$,%#E"&("E%S-*;/)I&#",W""*&`K)<&`-W&%*<&"%(,/LK%M"&-$$K(("*$"&

V%KE,&-()"*,%S-*;&

H,(";;&<(-If;");R)$&"*"(+Q&

PQ+/9",*%+,$*++*+R%7/$*%S$*++"$*%0#@'=*+R%%% .-(-"E%;S$&

R-<"E)*+&

3-R#)*"&-#;"([%S-*;fR-<"E)*+&,-&$-*;,(%)*&R"$/%*);R;&J-(&W/)$/&`K)<;&%E,"(&,/"&;,(";;&;,%,"&%*<&$-*,(-E&"%(,/LK%M"&*K$E"%S-*&&

E*(+1(0%/Q+*$)@?/'+% T/<*9('=%>*g"$S-*&[-EKR";:&

+"-E-+)$%E&)*J-(R%S-*:&W"EE&I(";;K(";&

7/$*%S$*++"$*%0#@'=*+R%%%

E,@?0%,$(==*$('=% UF'@1(0%,$(==*$('=%

What can we learn from earthquake triggering near injection sites?

E,@?0%,$(==*$('=%UF'@1(0%,$(==*$('=%UF'@1(0%,$(==*$('=%UF'@1(0%,$(==*$('=%


E,@?0%,$(==*$('=%UF'@1(0%,$(==*$('=%UF'@1(0%,$(==*$('=%UF'@1(0%,$(==*$('=%

?()++"()*+&;,(";;&);&I(-I-(S-*%E&,-&["E-$),Q&


Earthquake triggering in Canada

Triggering in Canada (determined by statistical "-value) following distant earthquakes from 2004-2014 (Wang et al., 2015). Red/pink indicates triggering.

TriggeringinAlberta

Wangetal.,2015

Triggeringstresses~0.1kPa,

frequencydependent

30'

%&$% 1:7 1%

119o W

121o W

121o

123o W

123oW

125o

61 oN61 oN

60 oN60 oN

59 oN59 oN

58 oN58 oN

57 oN57 oN

56 oN56 oN

55 oN55 oN

NBC1

NBC2NBC3

NBC4

NBC5

NBC6

40' 40'

20' 20'

40' 40'

20' 20'

MOON

PYRD

TULN

NOWN

CARC

HKNB

WCNBELNBSRNB

SVNB

125o W

126o W

1o

127o W

12oW

128o

64o W o W

65oo W

66oW

67o

64 oNoN

65 oNoN

66 oNoN

44 oNoN

45 ooN

46 oNN

47 oN

30'

30'

(a) (b) (c)

Focal mechanism

Breakout

��Ý:��Ý:

��Ý: ��Ý:��Ý:

��Ý:

��Ý

��Ý

��Ý

��Ý

��Ý

LookforremotedynamictriggeringinthreesedimentarybasinswhereinjecSonoccurs•  Directtriggeringinthewaveforms

•  MulS-staSonmatchedfilteràenhancedcatalog

Roleoffluids?TriggeringnearinjecSonsites

30'

%&$% 1:7 1%

119o W

121o W

121o

123o W

123oW

125o

61 oN61 oN

60 oN60 oN

59 oN59 oN

58 oN58 oN

57 oN57 oN

56 oN56 oN

55 oN55 oN

NBC1

NBC2NBC3

NBC4

NBC5

NBC6

40' 40'

20' 20'

40' 40'

20' 20'

MOON

PYRD

TULN

NOWN

CARC

HKNB

WCNBELNBSRNB

SVNB

125o W

126o W

1o

127o W

12oW

128o

64o W o W

65oo W

66oW

67o

64 oNoN

65 oNoN

66 oNoN

44 oNoN

45 ooN

46 oNN

47 oN

30'

30'

(a) (b) (c)

Focal mechanism

Breakout

��Ý:��Ý:

��Ý: ��Ý:��Ý:

��Ý:

��Ý

��Ý

��Ý

��Ý

��Ý



ImmediateTriggering


Immediate triggering: Mainshock surface wave train (BCAB)

30'

%&$% 1:7 1%

119o W

121o W

121o

123o W

123oW

125o

61 oN61 oN

60 oN60 oN

59 oN59 oN

58 oN58 oN

57 oN57 oN

56 oN56 oN

55 oN55 oN

NBC1

NBC2NBC3

NBC4

NBC5

NBC6

40' 40'

20' 20'

40' 40'

20' 20'

MOON

PYRD

TULN

NOWN

CARC

HKNB

WCNBELNBSRNB

SVNB

125o W

126o W

1o

127o W

12oW

128o

64o W o W

65oo W

66oW

67o

64 oNoN

65 oNoN

66 oNoN

44 oNoN

45 ooN

46 oNN

47 oN

30'

30'

(a) (b) (c)

Focal mechanism

Breakout

��Ý:��Ý:

��Ý: ��Ý:��Ý:

��Ý:

��Ý

��Ý

��Ý

��Ý

��Ý



CatalogedTriggering


N@,@9/=*<%,$(==*$('=%

N@,@9/=*<%,$(==*$('=%

Cataloged triggering: Statistical implications of delayed triggering

p = e!! !"

" !P =1! p

.(-#%#)E),Q&-J&$/%*+";&)*&;");R)$),Q&(%,"&

N@,@9/=*<%,$(==*$('=%

N@,@9/=*<%,$(==*$('=%

?()++"()*+&%*<&E%$M&-J&I%(S$E"&["E-$),Q&<"I"*<"*$"&&

H,(";;";X&8N7&F&A&M.%&H,(";;";X&8N7&F&7&M.%& H,(";;";X&8NA&F&8N7&M.%&

?()++"()*+&%*<&J("LK"*$Q&<"I"*<"*$"&

!118˚30' !118˚00' !117˚30' !117˚00'

54˚00'

54˚30'

CRANE network RAVEN network TD network CN networkHypoDD relocations Catalogue locations

M 1 M 2 M 3 M 4Wells Cities

Fox CreekJan 23, 2015

June 13, 2015

!120˚ !112˚

50˚

52˚

54˚

56˚

58˚

CRANE network RAVEN network TD network CN networkHypoDD relocations Catalogue locations

M 1 M 2 M 3 M 4Wells Cities

B.C.

U.S.A.

Alberta

Fox Creek

Edmonton

7$/S*$?*+%/:%*@$,#B"@C*+X&&&/Q<(%KE)$&J(%$,K()*+&)*<K$"<&"["*,;&

54˚00'

54˚30'

Fox Creek

June 13, 2015

CRANE network RAVEN network TD network CN networkCRANE network RAVEN network TD network CN networkCRANE network RAVEN network TD network CN networkHypoDD relocations Catalogue locationsHypoDD relocations Catalogue locationsHypoDD relocations Catalogue locations

M 1 M 2 M 3 M 4M 1 M 2 M 3 M 4M 1 M 2 M 3 M 4

2;"&;I"$,(%E&(%S-&R",/-<;&,-&";SR%,"&;,%S$&;,(";;&<(-I&

Proper?esofearthquakes:hydraulicfracturinginducedevents

Stressdropvaluesdon’tdifferfromnaturalearthquakes

Clercetal.,2016

Stressdropsdon’tchangewithdepthordistanceformthewell

Stressdropsmaynotbeagoodmetricfordiscrimina?ngbetweeninducedandnaturalearthquakes

Clercetal.,2016

PoroelasScmodelof(fracking)inducedearthquakesinFoxCreek,AB

Dengetal.,2016

CoulombstresschangesontheesSmatedfaultplane

InducedearthquakesandΔCFS:poroelasScresponseseemstodominateporepressurechanges.

Dengetal.,2016

Summary•  Naturalearthquaketriggeringoccurswithsmall(~0.1kPa)triggeringstressesthatappeartobefrequencydependent.

•  Humaninducedearthquakesdonotappeartobephysicallydifferentthannaturalearthquakes(stressdrops,b-values,…)

•  NewporoelasScmodelingandtriggeringobservaSonssuggeststresstransferoftherockmatrixmaybemoreimportantfornucleaSngearthquakes.

•  DoobservaSonsandmodelingresultshelpdisSnguishmechanismsforimmediatevs.delayedtriggering?

&&

X/$C('=%/'%"'<*$+,@'<('=%#/W%Y"(<+%('Y"*'0*%$"S,"$*%@,%@99%<*S,#+%('%,#*%:@"9,%G/'*Z%,$*1/$%H<**SKM%*@$,#B"@C*+%H(',*$1*<(@,*%@'<%+#@99/WK[%!*9@,*%,/%=*/9/=(0@9%0#@$@0,*$(+?0+%,/%"'<*$+,@'<%

1*0#@'(0@9%S$/0*++[%%%!"#$%&'()&*+)%&,-'.%)$/".$%01\%

X/$C('=%/'%"'<*$+,@'<('=%#/W%

!*+"9,%/:%@==$*++()*9F%0"$(/"+%0/W+%

Rebecca Harrington

Documents