1. SEISMIC NETWORK DESIGN FOR THE MONITORING OF NATURAL GAS RESERVOIR OF “SANT'ANDREA” (TREVISO, NORTHERN ITALY)
S. Carannante1, E. D'Alema1, S. Lovati1, M. Massa1, P. Augliera1, G. Franceschina1.
1-Istituto Nazionale di Geofisica e Vulcanologia, Milano.
ABSTRACT
The natural gas reservoir of “Sant’Andrea” (Treviso, Northen Italy), falls within the hydrocarbon exploitation concession named “Casa Tonetto”, conferred by the Italian Ministry of Economic Development (MiSE-DGS-UNMIG) to Apennine Energy S.p.A. on July The concession involves an area of 4.5 km^2, in the municipalities of Nervesa della Battaglia, Spresiano and Susegana. The exploration activities carried out by Apennine Energy in this area allowed to identify a methane gas field for which it was decided to proceed with production activities, using the well named “Sant’Andrea 1 dir ST” (SA1dST). This work presents the results obtained by the INGV team in order to assess the feasibility of a monitoring micro seismic network able to guarantee an adequate level of detection of seismic events, occurring in the neighborhood of the reservoir. In particular, considered the depth and the areal extension of the deposit, we defined a monitoring crustal volume of 8.0 x 8.0 x 5.0 km^3, centered on the position of the well SA1dST. Because of the limited mineral potential of the reservoir (about 84 MSm^3), we considered the possibility to detect and locate seismic events with local magnitude, ML, of at least 1.0 unit, within the above defined crustal volume. In order to establish the detection threshold of the network, we used point source simulations of earthquakes characterized by different values of magnitude and distance. For each installation site, the power spectrum of the simulated earthquake was compared with the observed power spectrum of ambient seismic noise. Data recorded by a seismic station installed in the area were employed to calibrate the simulation parameters. Point source simulations were carried out for seismic sources placed in 625 equally-spaced points of 3 regular grids, located within the monitored volume at 2.0, 3.5 and 5.0 km depth. The levels at 2.0 and 5.0 km coincide with the depth of the “Sant’Andrea” gas field and with the bottom of the monitored volume, respectively. A number of microseismic networks, with variable inter-station distances, and composed of 5 or 6 stations installed within the surface projection of the monitored volume, were hypothesized. In particular, by requiring the detection by at least 4 stations of the network, a configuration obtained by using 6 stations, with inter-station distances of the order of 4 km, allows to reach a ML localization threshold of 0.7 and 0.9 units at the depth of the reservoir and at the bottom of the monitored volume, respectively.
Monitoring Area
Crustal velocity model (Anselmi et al., 2011)
Detection Domain (DD) (cfr. MiSE-DGS-UNMIG (2014)
Seismicity of the area
(ISIDe Working Group – INGV, 2010)
Blue box - Detection Domain (DD): crustal volume within which some type of induced seismicity is expected (8.0 x 8.0 x 5.0 km^3);
Red triangle - CTNS station installed in the inner area of the Sant’Andrea 1 dir ST well; continuous recordings are transmitted in real time to the INGV data acquisition center of Milan;
Blue triangles - 4 sites selected for seismic noise measurements;
Yellow triangles: hypothetical stations adopted to estimate the localization thresholds;
Red closed line: surface projection of the reservoir.
Section N135° of Anselmi et al. (2011), shown in terms of Vp absolute values.
Ambient Noise Evaluation
Microseismic Network Design
Continuous recordings are processed using the PQLX software that evaluates, in real time, the probability density function (PDF) of the power spectral density (PSD) for stationary random seismic data (McNamara and Buland, 2004). By comparing the observed PDF of ambient seismic noise with the reference curves NHNM (New High Noise Model) and NLNM (New Low Noise Model) obtained by Peterson (1993), it results that, for each period up to 18 s, the observed mean curves are about 10 dB lower than the NHNM (upper gray line).
Two hypothetical networks were designed:
A) CTNS, GRV1, NRV1, SLP1, SPR1;
B) CTNS, CT01, CT02, CT03, CT04, CT05;
Detection Threshold (DT): minimum magnitude at which an earthquake can be recorded by at least one station;
Localization Threshold (LT): minimum magnitude at which an earthquake can be localized with 3 or 4 stations.
The simulation parameters were calibrated using data from a seismic sequence (max ML = 3.0) recorded by CTNS. Seismic point sources were placed in 625 equally­ spaced points of 3 regular grids (2.0, 3.5 and 5.0 km depth). DT and LT were estimated by comparing, at each site, the PSD of the simulated events with the observed PSD of noise.
DETECTION THRESHOLD, expressed in units of magnitude ML, for events localized at 2.0 km depth, for the networks A and B, respectively.
Probability density function (PDF) of the power spectral density (PSD), recorded at CTNS. Left panels: overall PDF; right panels: temporal variation of the PDF.
LT of the network B at all the examined depths, obtained by requesting the detection of a seismic event by at least 4 stations of the network.
PDF of the power spectral density (PSD) for EW components recorded at the temporary stations GRV1, NRV1, SLP1 and SPR1.
LEFT COLUMN - LT of the network A: A1) LT obtained by requesting the detection of a seismic event by at least 3 stations; A2) LT obtained by requesting the detection of a seismic event by at least 4 stations.
RIGHT COLUMN - LT of the network B: B1) LT obtained by requesting the detection of a seismic event by at least 3 stations; B2) LT obtained by requesting the detection of a seismic event by at least 4 stations.
CONCLUSIONS
References
1) The monitoring area is characterized by elevated levels of anthropic noise, with mean values of PSD comparable with the NHNM curves
of Peterson (1993).
2) By considering the 90° percentile curves of the measured PDF of noise, and a network composed of 5 surface stations placed within the surface projection of DD, we obtain, at the reservoir depth (2 km):
- ML location thresholds in the range [ ] when the event detection is required by at least 3 stations of the network.
- ML location thresholds in the range [ ] when the event detection is required by at least 4 stations of the network.
3) With a network composed of 6 surface stations, we obtain a GAP reduction of all the events located within DD and, at the reservoir depth (2 km):
- ML location thresholds in the range [ ] when the event detection is required by at least 3 stations of the network.
- ML location thresholds in the range [ ] when the event detection is required by at least 4 stations of the network.
4) By requiring the event detection by at least 4 stations, we obtain, at the bottom of DD (5 km depth), ML location thresholds in the ranges
[ ] and [ ] in the cases 2) and 3), respectively.
- Anselmi M., Govoni A., De Gori P., Chiarabba C. (2011). Seismicity and velocity structures along the south-Alpine thrust front of the Venetian Alps (NE-Italy), Tectonophysics, 513, 37–48.
- ISIDe Working Group (INGV 2010), Italian Seismological Instrumental and parametric Database: (last accessed: 03/2016).
- McNamara D.E., Buland R.P. (2004). Ambient Noise Levels in the Continental United States, Bull. Seism. Soc. Am., 94, 1517­-1527.
- MiSE-DGS-UNMIG (2014). Ministero dello Sviluppo Economico – Direzione Generale per le Risorse Minerarie ed Energetiche - Gruppo di lavoro CIRM, Indirizzi e linee guida per il monitoraggio della sismicità, delle deformazioni del suolo e delle pressioni di poro nell’ambito delle attività antropiche, Roma, 24/11/2014,
- Peterson J. (1993). Observation and modeling of seismic background noise, U.S.G.S. Tech. Rept., 93­-322, 1-­95.
- PQLX: A Software Tool to Evaluate Seismic Station Performance, (last accessed: 10/2015).