The Barents Sea is an epicontinental marine basin, located on a passive margin of the Eastern Europe Platform.

In the interior shelf plate a seismic activity is very low, and for the XX Century the world seismic monitoring network have recorded here earthquakes with magnitudes less than 4.0 (Assinovskaya, 1994; Bungum, Lindholm, 1996). Before 1923 a network capability was rather low, and so the upper limit magnitude of events, which have 90% probability of detection, was estimated as m_b > 5.1. Since 1950-es the threshold magnitude (TM) decreased to m_b > 4.3.

After 1983, when the Barents Sea seismic network was designed by the Kola Science Centre (KSC) of Russian Academy of Sciences , TM declined to 3,9. Then, after 1992 a digital facilities were installed, and data exchange between KSC and Norwegian Seismic Arrays (NORSAR) was established, resulting in a new approach in both the sensitivity (TM > 3.0) and event location precision (error less than 5-10 km). In the Stokman gas field the recent seismic network allows to detect and locate any event with M>TM=2.5; along a Northern part of a future pipeline from the field to the Kola Peninsula TM is assumed as 2,0; and near coast and on land at Vidyaevo, where port terminal and plant for LNG would be situated, TM=1.5.

The area, involved in the Stokman Project (development of gas deposit, construction of underwater pipeline, coastal port and plant for LNG), is characterized by different levels of seismicity: the gas field and northernmost part of pipeline are situated within an aseismatic province, whereas the southern part of pipeline and coastal facilities will be situated within the Finnmark-Murmansk seismogenic zone (FMSZ) with moderate activity. As estimated by B.Assinovskaya (1994), the seismotectonic potential in FMSZ is higher in compare with the shelf interior in 1.3-1.5 times. Earthquakes are mainly locating close to the marginal Trollfjord-Komagelv (TK) and Karpinsky (K) shear-faults.

The low depth hypocenters are dominant, and this has been interpreted as a sign of control by a sublateral fluid saturated waveguide, revealed in depth of 8-12 km by seismotomography and superdeep drilling (Isanina et al., 2002; Yudakhin et al., 2003). Of a history some evidences on strong earthquakes at Kola Town are well known, the highest levels were noted in 1772 and 1873 when macro effects indicated on local magnitude of 5.0 and strike in the epicenter up to 7±0,5 balls according to Russian MSK-64 scale. Regional seismic network in 1953-2005 was recorded some variations in seismic energy release: periods (1-5 years) with rather high activity (up to 17 earthquakes per year within FMSZ) are divided by quiet periods, when events with M > 2.5 did not occurred of 1 up to 7 years. The last "blow" of activity in FMSZ was recorded in 1987-1993, and nowadays the zone is slipping under geodynamic regime of "anomaly calm". Assessment of probability of strong strikes based on cumulative frequency-magnitude graph for time span 1963-2005 shows that earthquakes of 6 balls could be expected once in 79 years, and strikes up to 7.5 balls - once in 745 years.

Technogenic induced seismicity, related to NAVY operations or sinking of missile parts, can generate ground motions equal to earthquakes of 3-5 balls. Taking into consideration a lithology of upper sediment layer and a sea bottom topography at the Stokman field and along underwater pipeline, we can assess that even so low seismic events would be able to provoke sliding of large mud masses. As assumed, the induced seismicity, related to a subsidence trough formation, would be potentially much more hazardous. There are two types of technogenic induced earthquakes with high risk or catastrophic consequences: 1) technogenic rock strikes within recovered hydrocarbon reservoirs, low deep and M=3.0-3.5; 2) induced strikes out of reservoirs (upper, below or beside) at depth of 10 km and deeper, with M=4.0-6.0 (Khaustov, Redina, 2006).

As the reference case for the latter, the Lacq earthquake in France can be mentioned: after 10 years of recovery with annual output of 5 bin. cub. m strike at depth 3-5 km was happened with M=4.2, which reflected at the earth surface by ground motions up to 7.5 balls (Lahaie, Grasso, 1999). In the Stokman field after 10 years of recovering with annual release of 71 bin. cub. m of gas the subsidence trough have to reach up to 35 km in width and 10-15 meters in depth (Kozlov, 2005). If assume that a trough draw will be performed in one stage and any fault would be sheared totally in length and deep (up to bottom of productive reservoir), then a probable ground acceleration in the epicenter would exceed 1089 cm/sek , what is equal to 11 balls earthquake. Evidently, to prevent a catastrophic event we have to control drawing and provide sharing of stress and motion in time and space.

According to the national concept "Geodynamic safety in development of hydrocarbon reserves in Russia" (2000) the aimed Block for seismic monitoring (BSM) has to be established as obligatory unit in the future system for geoecological control of the Project.

The main requirement for BSM: to provide a guaranty for detection and correct location of low strikes with M=2.0 in upper crust, responsible for mud slides at the sea bottom, as well as to allow permanent monitoring of detail 3D pattern of microseism sources within and nearby the subsidence trough. So, first of all we have to improve and extent the Barents Sea regional seismic network, because nowadays a gap between the nearest stations and Stocman field exceeds 600 km, and they can not provide a comprehensive monitoring of light events in the Stokman field and northernmost section of pipe line.

Due to get a reliable level of sensibility and precision new seismic arrays, similar to "Apatity" one (Vinogradov, 2005), have to be installed at nearest islands - Novaya Zemlya (280 km from the field) and Kolguev (400 km) (fig.l). Beside that, based on experience of monitoring of the pipeline "Blue Sream" in the Black Sea (Levchenko, 2005), we can recommend to install a special network of bottom seismographs in the field and along the pipeline in the Barents Sea, because any coastal stations/arrays are unable to reveal and detect very low seismic signals of geodynamic process into loose sediments at the sea bottom . The optimal points for bottom seismographs installation are intersections of active faults and outer rim of the subsidence through (73° 00' N. and 44°10* E., 73° 05* N. and 43°35' E., 73° 15' N. and 43°50* E.). In the southernmost part of pipeline optimal points are closed to TK and К faults (70° 00' N. and 35°00' E., 69°40' N. and 34°00' K, 69°30' N. and 33°30* E., 69°25' N. and 33°25* E.).

References
Bungum H., Lindholm C. Seismo- and neotectonics in Finmark, Kola and the Southern Barents Sea, part 2: Seismological analysis and seismotectonics. Tectonophysics, v. 270,1996. -P. 15-28.

Assinovskaya B.A. Seismicity of the Barents Sea. Moscow, 1994. -128 p.

Vinogradov Yu. A. Seismo-acustic complex "Apatity" - the modern tool for monitoring of Environment // Physical Acustics. Waves propagation and diffraction. Geoacustics. Proc. Of XVI session of Russian Acustic Society. Vol.1. Moscow: GEOS, 2005. -P. 358-362.

Isanina E.V., Verba M.L., Ivanovo N.M. et al. Deep structure and seismogeological boundaries in Pechenga district of the Baltic Shield and in adjacent part of the Barents Sea shelf// Geology of Ore Deposits, vol. 42, No. 5, 2000. - P.476-487.

Kozlov S.A. Concept grounds for a engineering-and-geological study of the West-Arctic hydrocarbon province // Electronic Magazine "Oil and gas business", 2005: http: //www. ogbus.ru/authors/Kozlov/Kozlov_4. pdf
National Concept "Geodynamic safety in development of hydrocarbon reserves in Russia", Moscow, 2000.

Khaustov A.P. and Redina MM. Environment protection at oil recovery. Moscow: Delo, 2006. - 552 p.
Yudakhin F.N., Shchnkin Yu,K. and Makarov V.I. Deep structure and recent geodynamic processes in lithosphere of the Eastern European Platform. Ekaterinburg, 2003. - 299 p.

Lahaie, F.; Grasso, J. R. Correction to loading rate impact on fracturing pattern: lessons from hydrocarbon recovery, Lacq gas field, France //J. Geophys. Res. - Solid Earth, 1999, V. 104, №10.

OIL AND GAS OF ARCTIC SHELF 2006 (PROCEEDINGS OF INTERNATIONAL CONFERENCE)