Conclusion• With a 2D modeling approach calibrated for the Basel EGS, we observe that considering

earthquake interactions in the modelling can lead to a larger number of expected seismic

events (24% more) if compared to a pressure-induced seismicity only. The increase of

the rate is true particularly after the end of the injection activity, in accordance with the

simultaneous increase of the Coulomb Index.

• The use of a more sophisticated 3D modeling approach allows to generalize the effect of

earthquake interactions during an injection-induced seismicity. The model confirms that for

a case with distributed seismicity (e.g. Basel), the static stress transfer has a large effect

on the number of events

• We conclude that implementing a model for estimating the static stress changes due to

mutual event interactions increases significantly the understanding of the process.

REFERENCES1) Catalli, F., Meier, M.-A., Wiemer, S., (2013). The role of Coulomb stress changes for injection‐induced seismicity: The Basel enhanced

geothermal system. Geophys. Res. Lett., 40, 72-77.2) Catalli, F., Rinaldi, A. P., Gischig, V., Nespoli, M., Wiemer, S., (2016). The importance of earthquake interactions for injection-induced

seismicity: Retrospective modeling of the Basel Enhanced Geothermal System. Geophys. Res. Lett., 43, 4992–4999.3) Kraft, T. Deichmann, N., (2014). High-precision relocation and focal mechanism of the injection-induced seismicity at the Basel EGS,

Geothermics, 52, 59–73.4) Rinaldi, A. P., Nespoli, M., (2016) TOUGH2-seed: A coupled fluid flow and mechanical-stochastic approach to model injection-induced

seismicity. Comput. Geosci., in press.

This research is funded by a Swiss National Science Foundation (SNSF) Ambizione Energy grant (PZENP2_160555).

TOUGH2-seed: a 3D modeling approach (Rinaldi & Nespoli, 2016)

-APPLICATION TO BASEL

-SYNTHETIC CASE

Time (days)

0

10

20

30

40

Pres

sure

cha

nges

(MPa

)

0

20

40

60

80

Num

ber o

f eve

nts/

12h

0 4 8 12 0

5

10

15

20

25

Pressure (MPa)

-1 -0.5 0 0.5 1Easting (km)

-1

-0.5

0

0.5

1

Nor

thin

g (k

m)

10 days

-1 -0.5 0 0.5 1Easting (km)

-1

-0.5

0

0.5

1

Nor

thin

g (k

m)

10 days - XY

-16

-15

-14

-13

log10 (κ)

-1 -0.5 0 0.5 1Easting (km)

4.6

4.2

3.8

3.4

Dep

th (k

m)

10 days - XZ

5.0

-16

-15

-14

-13

log10 (κ)

Time (days)

0

200

400

600

800

Cum

ulat

ive

# ev

ents

0 4 8 12

1000

No stress transferCoulombSpring-Block

(a)

Time (days)

0

200

400

600

800

Cum

ulat

ive

# ev

ents

0 4 8 12

1000

(b)

-1 -0.5 0 0.5 1Easting (km)

-1

-0.5

0

0.5

1

Nor

thin

g (k

m)

Coulomb model - 10 days - XYseeds on preferential direction

Mw < 1.51.5 < Mw < 2.5Mw > 2.5

-1 -0.5 0 0.5 1Easting (km)

-1

-0.5

0

0.5

1

Nor

thin

g (k

m)

Coulomb model - 10 days - XYuniformely distributed seeds

(c) (d)

Uniformely distributedseed

Seeds onpreferentialdirection

12Time (days)

0

100

200

300

Num

ber o

f eve

nts/

12 h

0

10

20

30

Pres

sure

cha

nges

(MPa

)

PressureSimulatedMeasured

84012Time (days)

0

10

20

30

Pres

sure

cha

nges

(MPa

)

MeasuredSimulated

840

Stochastic seed model

ComputesE�ective stress

Checksseed reactivation

Computespermeability changes

InitializationSeeds distribution

Initial stress �eld

only first time step

Seismicitycatalog

TOUGH2/EOS3

ComputesMultiphase,

multicomponent �uid �ow

ChangesPermeability

ComputesPressure

TOUGH2out�le

inj.

rate

(kg/

sec)

10

20

30

40

50

time (days)4 102 6 8 12

North

=N155°E

Hmax= 1 N144°E

Hmin= 3

n

2D modeling approach (Catalli et al., 2016)

0

50

100

150

200

250

300

350

0 100 200 3000

50

100

150

-100 0 1000

50

100

150

200

250

Dip Strike Rake

# Se

eds

# Se

eds

# Se

eds

#SeedsSmoothed distribution from Kraft and Deichmann (2014)scaled to number of seeds

20 40 60 80

CI

(%)

0

20

40

60

80

100

Time (days)102 4 6 80

# ev

ents

0

200

400

600

800

1000

1200 1stdSTRESS TRANSFER1stdPRESSURE ONLYOBSERVATIONS

Time (days)102 4 6 80

Det

erm

inin

g tr

igge

r (%

)

0

20

40

60

80

100

R (km)1.00.2 0.4 0.6 0.8

1stdEQ. Interaction trigger1stdpressure trigger

Det

erm

inin

g tr

igge

r (%

)

0

20

40

60

80

100

Time (days)102 4 6 8

Time (days)10

Num

ber o

f eve

nts

0

100

200

300 PRESSURE ONLY

Pres

sure

cha

nges

(MPa

)0

5

10

15

20

25

30

35

2 4 6 80

STRESS TRANSFER

Time (days)10

Num

ber o

f eve

nts

0

100

200

300

Pres

sure

cha

nges

(MPa

)

0

5

10

15

20

25

30

35

2 4 6 80

Pres

sure

cha

nge

[Mpa

]

30

0

10

20

Mag

nitu

de M

L

1

2

3

4

Time [days]2 12108640

1 reduction 2 shut-in3 bleed-off

Distance from casing shoe [m]0 200 400 600 800 1000

Cum

ulat

ive Δ

CFS

[MPa

]

0.5

0

-0.5

-1

-1.5-2

1

CI [

%]

85

80

75

70

65

90

CI [

%]

80

70

60

50

40

90

Cum

ulat

ive Δ

CFS

[MPa

]

0.5

0

-0.5

-1

-1.5-2

1

Event occurance0 40 80 120 160

0 100 200 300 400

−1

0

1

Time from 1st event [days]

mL=1.5mL=2.5mL>3.0

CFS

[MPa

]Δ

0

40

80

Num

ber o

f cas

es

ΔCFS (MPa)

−10 −1

−0.1

−0.0

1−0

.001

0

0.01 0.1 1 10

0.00

1

ΔCFS > 0−0.8

0.4

−0.4

0

0.4 CI = 75%

X [km]

−0.1 MPa−1 MPa0.1 MPa

1 MPa

mL > 3.0casing shoe

Y [k

m]

0-0.4ΔCFS < 0

0.4X [km]

0-0.4

Data from Basel show that the

earthquake interactions may have

played a role in the distribution

of the seismic cloud. The amount

of events with positive Coulomb

Stress increases with time and far

from casing shoe. The percentage

of events with positive CFS changes

reaches about 90%.

The case of Basel EGS (Catalli et al., 2013)IntroductionWe explore the role of earthquake interactions

during an injection induced seismic sequence. We

first propose a 2D model based on data from Basel

and that considers both a transient pressure and

the static Coulomb stress redistribution due to

event interactions as triggering mechanisms for

induced seismicity. Then, we generalized the model

for injection induced seismicity to a full 3D fluid

flow and geomechanical-stochastic configuration,

including stress transfer and permeability changes.

We applied the 3D model to synthetic cases and

compare two different model of static stress

transfer. For each approach, we produce a number

of stochastic seismic catalogues that allow a

probabilistic analysis of the problem.

Modeling of earthquake interactions for induced seismicityRinaldi A. P.(1), Catalli F.(2), Nespoli M.(3), Gischig V.(4), Wiemer S.(1)

(1) Swiss Seismological Service, ETH Zürich, Switzerland; (2) Deutsches GeoForschung- Zentrum, GFZ – Potsdam, Germany; (3) Dip. di Fisica e Astronomia, Università di Bologna, Italy; (4) Swiss Competence Center for Energy Research, ETH Zürich, Switzerland