| Title | Real-Time Induced Seismicity Monitoring During Wellbore Stimulation at Paralana-2 South Australia |
|---|---|
| Authors | Michael Hasting, Julie Albaric, Volker Oye, Peter Reid, Mathieu Messeiller, Ella Llanos, Peter Malin, Eylon Shalev, Mathew Hogg, Mark Alvarez, Alex Miller, Christina Walter, Carolin Boese, Nora Voss |
| Year | 2011 |
| Conference | Australian Geothermal Energy Conference |
| Keywords | |
| Abstract | In 2009 the Paralana JV, drilled the Paralana-2 (P2) Enhanced Geothermal System (EGS) borehole east of the Flinders Range in South Australia. Drilling started on June 30th and reached a total depth of 4,003m (G.L AHD) on Nov 9th. A 7- inch casing was set and cemented to a depth of 3,725m and P2 was officially completed on the 9th Dec 2009. On Jan 2nd 2011 a six meter zone was perforated between 3,679 and 3,685 mRT. A stimulation of P2 was carried out on Jan 3rd by injecting approximately 14,668 litres of fluid at pressure of up to 8.7 kpsi and various rates up to 2 bpm. During the stimulation ~125 micro-earthquakes (MEQ) were triggered in the formation. Most of the MEQ events occurred in an area about 100 m wide and 220 m deep at an average depth of 3,850 m. The largest event, a MW 1.4, occurred after the shut-in. Between 11th and 15th of July 2011, the main fracture stimulation was carried out with ~3 M litres injected at pressures up to 9 kpsi and rates up to 10 bpm. Over 11,000 MEQ were detected by the seismic monitoring network. This network consisted of 12 surface and 8 borehole stations with sensor depths of 40 m, 200 m and 1,800 m. Four accelerometers were also installed to record ground motions near key facilities in the case of a larger seismic event. MEQ were automatically triggered and located in near-real-time with the software MIMO provided by NORSAR. A traffic light system was in operation and none of the detected events came close to the threshold value. More than ½ of the detected events could be processed and located reliably in the full automatic mode. Selected MEQ events were also manually picked on site in order to improve the location accuracy. A total of 875 MEQ events were picked, located and plotted on site to give the operator, Petratherm, a sense of the fracture created while post processing yielded another 1,025 events. After a data download in mid August an additional 750 events were located from this data set. As such over 2,600 events were hand-picked and located to form the final picture of the stimulation fracture. Results show that fracturing occurred in three swarms. The 1st swarm occurs near the well and deepened with time from 3.7 km to over 4.1 km. The 2nd swarm occurred a few days in and shows as a circular patch extending a few hundred meters east of the 1st one. The 3rd swarm occurred after shut-in and extends downwards to the NNW and reaches 4.4 km depth. Petratherm believes that there is a primary NE/SW structure that takes most of the fluid. Then, two NNE/SSW structures are highlighted after day 5 to 6 and continue growing after shutin. The first fracture appears to have a sygmoidal pattern. The two later structures appear to act as boundaries to the East and West and are subparallel to the major faults that define the graben in which P2 was drilled. They appear to deepen towards the north. A later shallower structure is highlighted to the SE of the well. Overall, it appears that at least 4 structures have been enhanced and stimulated. The well head pressure after the minifrac and after the main frac shows a value of about 3940 psi. This shows that the injection of fluids in P2 has connected into a naturally fractured network, with in situ fluid. While drilling P2, overpressured brines were intersected at depths between 3680 and 3860m. We believe that this zone of fracture permeability has been connected to and enhanced. The MEQ cloud shows a complex fracture network of at least 4 structures that can be interpreted as conjugated faults/fractures. |