Second Annual Meeting

Transcription

1 SHale gas Exploration and Exploitation induced Risks Detection, location and migration of AEs induced by hydraulic fracturing: the Aspö mine (Sweden) experiment Simone Cesca Second Annual Meeting Blackpool June 57, 2017 This project has received funding from the European Union s Horizon 2020 research and innovation programme under grant agreement No

2 Previous results (Zang et al., 2017) HF3 HF2 HF1 HF1 HF2 HF3! HF1 49 HF2 HF

3 Typical waveforms of seismic AE event Detections with triggered mode Sampling = 1Mhz Waveform duration ~ miliseconds Accelerometers (ACC) [50 Hz 25 khz] Acoustic emission sensors (AE) [1Khz 100 Khz] Bandpass filter [3Khz 70 Khz] AE Magnitude (MAE) (Eisenbläter and Spies, 2000) M AE 20 log10 (A) A0 = r = r0 = α = A A0 r (r ro ) e r0 Pwave amplitude normalized sourcereceiver distance (m) 20 m (scaling parameter) 0.02/m (damping) Zang et al., (2016)

4 e in p re p a ra tio n fo r th e c o n s tru c tio n l. Hydraulic Fracture Experiments fo r h e le ft s is ts o f t rig h t).! 280 m level AE11 AE08 AE09 AE10 AE07 AE06 HF2 AE03 AE04 AE05 AE01 AE m level Fig u re ' 4 :' S e n s o r s ' e m p lo y e d ' in ' n e a r?f ie ld.' P R E L I M I N A R Y ' P I C T U R E ' ' C O O R D I N A T E S ' O F ' G E O P H O N E ' N E T W O R K ' O F ' N E A R?F I E L D ' S E N S O R S.' M s e n s o r s ' A E 1?A E 2,' b o r e h o le ' M 2 ' is ' e q u ip p e d ' w it h ' s e n s o r s ' A E 3?A E 5,' a n d

5 Automated full waveform detection (LASSIE) Characteristic function sensitive to changes in energy: square amplitude and stacking (Heimann et al., in preparation) Spatial grid search: 90 x 90 x 15 m (dx = dy = dz = 2 m) Earth model: homogeneous full space model (vp = 5810 m/s, vs =3400 m/s) Threshold detections for the Amplitude of the Characteristic function

6 Acoustic emission detections (HF2) a) AE strong event Acf = 2090 b) Electronic noise Acf = 1589 c) Anthropogenic noise Acf = 1959 AE01 AE02 AE03 AE04 AE05 AE06 AE07 AE08 AE09 AE10 AE11 08:51: d) AE weak event 2 ms Acf = :33: e) Other signals 2 ms Acf = :50: ms f) Long period noise Acf = 592 AE01 AE02 AE03 AE04 AE05 AE06 AE07 AE08 AE09 AE10 AE11 08:14: ms 08:50: ms 08:43: ms

7 Acoustic emission detections (HF2) Pressure in Mpa / Flow in lpm a) Acf det Acf det Acf 500 Pressure in Mpa / Flow in lpm b) AE events = 4158 False detections = 4172 Electronic noise = Other signals = 215 Anthropogenic noise = 58 Long period noise = det Triggered 102 det.

8 Acoustic emission locations (LOKI) LOKI: Automated locations by waveform analysis coherence with and without master event (Grigoli et al., 2013; 2014; 2016) Grid locations: 15 x 15 x 10 m => 15 x 15 x 15 m mapview a) 15 m 15 m master event view direction in b) view direction in c) b) sideview 1 c) sideview 2 (dx = dy = dz = 10 cm) Coherence matrix for master event

9 Acoustic emission locations (LOKI) The largest 258 AE events located a) preliminary locations relocations using a master event view direction in ii) i) view direction in i) ii) 4158 AE events relocated b) view direction in ii) i) view direction in i) ii)

10 Pressure in Mpa / Flow in lpm

11 Frequencymagnitude distribution Mc = 1.1 b = 2.38 MAE,min = 0.77 MAE,max = 2.79 Mc calculated by Wiemer and Wyss, 2000

12 Frequencymagnitude distribution a) b) PT F RF1 RF2 RF3 RF4 RF5 Applications of the algorithm from Hainzl, (2016) to estimate the ratedependent completeness magnitude Mc (t).

13 The fracture growth b) a) view direction in d) view direction in b) c) d)

14 The fracture growth HF2 Frac c e a b Refrac 1 Refrac 3 Refrac 4 Refrac 5

15 RF11 a) RF12 RF13 RF14 Refrac 1 RF14 Pressure in Mpa / Flow in lpm RF11 RF12 RF13 X (m) RF51 b) X (m) RF52 X (m) X (m) RF53 Refrac 5 RF52 RF53 Pressure in Mpa / Flow in lpm RF51 X (m) X (m) X (m) Figure 11

16 The fracture growth Maximum magnitude for the worst case scenario b) a) Worst case scenario: Rupture length = 10 m Worst case scenario: Size of seismicity cloud in terms of d char considering: Source model of Madariaga (1976) Stress drop Δσ = 3 MPa MW= 0.08 dchar = square root of the planar area covered by the seismicity cloud = km MW= 3.07

17 Conclusions Robust and accurate AE locations obtained applying waveform stacking techniques. Thanks to the implementation of these novel techniques, we are able to detect ~40 times more events than in the previous triggerbased catalogue. AE events cover the depth interval between 404 and 414 m. A general upward migrations is observed; however, at the end of the experiment, after the injection, AE activity appears located in the area of the initial fracture phase, suggesting a crackclosing effect. We estimate a high bvalue (2.38) and a magnitude of completeness of MAE 1.1 (~Mw 4.3) (G. Kwiatek, pers. Comm.). MAE,max ~ Mw 3.5, compatible with upper estimation of Mmax = 3.1 following Dahm et al. (2008); estimation after McGarr (2014) would exceed by 3 order of magnitudes For further information: LópezComino, J. A., S. Cesca, S. Heimann, F. Grigoli, C. Milkereit, T. Dahm and A. Zang (2017). Characterization of hydraulic fractures growth during the Äspö Hard Rock Laboratory experiment (Sweden). Journal of Rock Mechanics and Rock Engineering, under review.

18 SHale gas Exploration and Exploitation induced Risks Detection, location and migration of AEs induced by hydraulic fracturing: the Aspö mine (Sweden) experiment Simone Cesca Second Annual Meeting Blackpool June 57, 2017 This project has received funding from the European Union s Horizon 2020 research and innovation programme under grant agreement No

19 a) b) c) d) HF1 HF2 HF3 Frac Refrac 1 Refrac 3 Refrac 4 Refrac 5

20 Discussion and conclusions a) b) c)

21 Frequencymagnitude distribution Relative magnitude in db (Eisenbläter and Spies, 2000; Spies and Eisenbläter 2001a,b; Spies et al., 2005 ) r (r ro ) e r0 M AE0 20 log10 (A) A A0 M AE1 20 log10 (A) r A A1 e (r ro ) r0 r = sourcereceiver distance (m) r0 = 20 m (scaling parameter) α = 0.02/m (damping) A0 = Pwave amplitude normalized (Zang et a., 2017) A1 = Maximum amplitude normalized Absolute magnitude (Gutenberg Richter distribution) (Köhler et al., 2009; Maghsoudi et al., 2013, 2014) M AE log10 (A) A A1 r (r ro ) e r0 MAE in the range 0 to 5 Mw in the range 8 to 3 (Eisenbläter and Spies, 2000)

22 Frequencymagnitude distribution

23 Acoustic emission locations (LOKI) LOKI: Automated locations by waveform analysis coherence with and without master event (Grigoli et al., 2013; 2014; 2016) Grid locations: 15 x 15 x 10 m => 15 x 15 x 15 m (dx = dy = dz = 10 cm) mapview a) 15 m 15 m master event view direction in b) view direction in c) b) sideview 1 c) sideview 2 Zang et al., (2017)

24 Acoustic emission locations (LOKI) LOKI: Automated locations by waveform analysis coherence with and without master event (Grigoli et al., 2013; 2014; 2016) Grid locations: 15 x 15 x 10 m => 15 x 15 x 15 m (dx = dy = dz = 10 cm) mapview a) 15 m 15 m master event view direction in b) view direction in c) b) sideview 1 c) sideview 2 Acf > AE detection Acf > AE detections Total detections Zang et a., AE 102 AE detections