Estimation of Seismic Hazard on a Prospective Npp Site in Southern Finland
Total Page:16
File Type:pdf, Size:1020Kb
ESTIMATION OF SEISMIC HAZARD ON A PROSPECTIVE NPP SITE IN SOUTHERN FINLAND Pentti Varpasuo Fortum Nuclear Services Ltd Finland 1 Introduction The purpose of the present work is the estimation of seismic hazard in the territory of the prospective nuclear power plant site in Southern Finland. Because there are no registered strong motion acceleration recordings of earthquakes in Finland, the earthquake recordings from Saguenay and Newcastle regions from Canada and Australia were taken as sources of initial data because of their geological and tectonical similarity to Fennoscandia. The probabilistic seismic hazard assessment consists of three parts: 1) source effects, 2) path effects, 3) site effects. The site effects phase of the study is not relevant to this study because the prospective target sites are located on solid bedrock.Theoretical bases of determination of seismic hazard, questions of seismicity of a southern part of Finland, initial data on earthquakes and techniques of their processing, are considered below. 2 Regional seismicity in Fennoscandia Finland is situated on the Baltic shield, which is the one of the seismically quietest areas in the world. According to the fault plane solutions of earthquakes the push from the North Atlantic Ridge in the NWSE direction seems to be the major stress related to the seismicity of Finland. Other factors of the stress field, such as glacial rebound and local seismotectonics, are more local. Earthquake recurrence rates in Fennoscandia are very low if compared with plate boundary regions worldwide. Nonetheless, Fennoscandia is an active seismic region, albeit at low earthquake recurrence rates and with relatively low magnitudes.The earthquake catalogue for Northern Europe (FENCAT), maintained by the Institute of Seismology of the University of Helsinki, was used in this study [i]. This catalogue encompasses the whole of Fennoscandia and adjacent areas inside a window of about 5580 oN and 10 oW 45 oE. The catalogue includes all documented earthquakes in the region since 1375. Although instrumental earthquake observations started in Finland in the 1920's, local short period recordings started in 1956 [ii]. The events in Finland and in Fennoscandia have been predominantly instrumentally located since the mid 1960’s [i]. The instrumental magnitudes are based on the Richter's classical local magnitude scale, ML, modified for the Fennoscandian region. The uncertainty of macroseismic magnitudes is assumed to be 10% at best [i]. Figure 21 shows that southern Finland, which is the target area of this study, is characterised by relatively low seismicity. The most active belts of seismicity close to it are the Swedish coast from the Bothnian Sea to the Bothnian Bay, western Lapland and the northern Bothnian BayKuusamo region. Southern Finland and areas south and east of it are characterised generally by a lower seismic activity. However, two NW SE oriented belts of relatively high seismic activity run through the region [iii]. The northern zone of higher activity runs from the southern Bothnian Bay towards Ladoga (BL). The other 1(34) active belt (ÅPP) runs from the Åland archipelago to southeastern Estonia, where it extends to from Paldis to Pskov. The zones are distinguished from their surroundings particularly by the occurrence of relatively large earthquakes (Figure 22). However, the period of pronounced seismic activity from 1920 to 1941 brings out the same seismic belts in southern Finland, but elsewhere as well. E 0 : n o ti a t n e 70 N s re p e th f o r e rd o B Lapland SWEDEN Kuusamo 65 N BB Oulu järvi FINLAND RUSSIA Bothnian Sea 60 N 200 km ESTONIA M = 4.56.1 M = 3.54.4 M = 2.53.4 M = 1.52.4 20 E 25 E 30 E 55 N Figure 21. Distribution of the earthquake epicenters in northern Europe since 1375 according to FENCAT. The outlines of the seismic belts applied in this study are presented in Figure 22. Both of the seismically active belts of southern Finland are characterised by long NWSE oriented fracture zones [iii]and [iv]. The more prominent seismotectonic zone in eastern coast of Sweden is associated with considerable (300400 m) vertical displacements along the eastern coast of Sweden [vii] and [v]. This zone, over 200 km west of Olkiluoto, separates the inland Precambrian bedrock from the offshore Phanerozoic sedimentary rocks. 2(34) ))) ))) 333 )) )3)3)33 )))) ))) ))) ))))) )))333 ))))) ))) 333 333 333 333 333 )) ))) 333 333 3333 ))) 333 ))) 333 ))) White )))333 33) ) ) 333 ))) 333 Sea ))) ))) 333 33333 3 3 3 3 333 ))) 333 333 )) 333 ))3)33 3 3 3 3 3 333 333 333 ))) 3333 33 333 33 333 33 ))) 333 )))33 )) 333 333 3 33 Kuusamo SWEDEN ))) ))) 333 333 3 33 333 )) 333 ))) 333 333 ) ) )3 3 3 )))))3 3 3 )))3 3 333 )))3 3 3 3 )) ))) 333 )))) ) ) ))) 333 333 333 65 N ))) ))) 333 333 333 )3)3)333 )) ))) ))) ) ))33)3 ) ) ) ) ))) ))) Oulu ))) ))333 ))) järvi ))))) ))) ))) ))) ))) )))) CFQZ 333 333 )) B 333 333 ) )) ))) ))) )) 333 333 ) ))) ) ) 333 ))) )))3)3)3 )))) ) ) 333 ))) FINLAND NLAND ))))) 333 ))) 333 ))333 ))) 33 RUSSIA 333 333 ))) ))))) )))333 ))) 3 33 ) ) ) Bothnian )))333 ))) 3)3)3) ))) ))) Sea 33 ))) ))) Olkiluoto 333 %% SFQZ 33333 333 333 L ))) 333 ))) 60 N 333 %% 333 )))333 3 3 333 %% Å Hästholmen P 333 ))) ESTONIA 35 E ))) P ))) ))) )) 20 E 25 E 30 E Figure 22 Belts))) of higher seismic activity (shaded areas). Historical events (13751964) with magnitude M ≥ 3.5 and instrumentally located (19651995) events with magnitude M ≥ 3.0 are shown by open and filled black circles, respectively. Crosses denote earthquakes during the period of increased seismic activity 19201941. BL Southern Bothnian Bay Ladoga Zone, Å PP = Åland Archipelago PaldisPskov Zone, CFQZ= Central Finland Quiet Zone, SFQZ = Southern Finland Quiet Zone. The study utilises a similar size of subset of FENCAT for both of the nuclear power plant sites. The seismic characteristics of Finland and examinations of attenuation (Chapter 5 ) suggest that areas within a distance of 500 km from the nuclear power plant sites are considered large enough to include all significant seismic events.Within 500 km of the sites, most of the earthquakes are small (M<5.0) and well below the threshold of engineering concern. Only one event, which occurred in 1894 (M=5.1) in central Sweden can be classified as a moderate (M≥5.0) earthquake. The highest intensity at the Olkiluoto site is only IV (MSK64 intensity scale) [ii and vii]. MSK 64 intensity IV is partially described as: felt indoors by many people, outdoors by few, some awakened, vibration is like that due to the passing of a heavily loaded truck; Windows, doors and dishes rattle; Noticeable in standing motor cars [vi]. In the Loviisa site the maximum observed intensity is higher: VI. Within a distance of 500 km from Olkiluoto, a clear majority (72%) of the seismic events has occurred in Sweden [vii]. The areas in southern Sweden (Telemark Vänern region) and the eastern coast of Sweden are well known belts of higher seismicity (see e.g. [i]). These belts are outside the Loviisa data set. Both of the data sets include a small part of the northern Bothnian BayKuusamo seismic belt in the north. 3(34) 3 Theoretical bases of Probabilistic Seismic hazard Assessment (PSHA). In the execution of the given work the seismic hazard assessment methodology described in reference viii and ix was used. The theory, on which the methodology of references viii and ix is based, has been developed for a number of years by Cornell x (1968) and Cornell xi (1974) Merz and Cornell xii (1973). The PSHA theory as presented in reference ix can be submitted in the basic form as the Theorem of total probability Equation 31 P [A] = òò P [A | S and R] fs (S) fr (R) dsdr. In Equation 31 P indicates the specified probability, A is the event whose probability is sought, S and R are continuos independent random variables that have an effect on the value of A. In other words the probability that A occurs can be determined by multiplying the conditional probability of A, when s and r are given, times the independent probabilities of s and r and integrating over all investigated of s and r. Variables s and r represent earthquake size in used measure and distance from the site of interest. Random size and random location of the events are taken into account in the method. At an estimation of continuous probability of logarithm of ground acceleration exceeding some value i at the given station, the normal distribution assumption of reference xiii The average size of continuous distribution of logarithm of ground acceleration is defined as Equation 32 MI (S, R) = C1 + C2 * S + C3 * ln (R + h), In Equation 32 C1 , C2, C3 are constant factors and S is the size of earthquake and R is epicentral distance in km. h is depth of a seismic source in km. This equation is called the attenuation equation of ground motion. The standard deviation of logarithm of ground acceleration sI is generally constant and independent from S and R. Applying normal distribution and Equation 32 , we have Equation 33 P [A | S and R] = P [I > i | S and R] = F* ((iC1 C2 * S C3 * ln (R + h)) / sI) In Equation 33 F* cumulative standard normal distribution. Characteristic ground motion variable is generally assumed to be lognormally distributed. Thus the logarithms of this variable is normally distributed. This normality assumption is fundamental to the methodology. The number Nm of earthquakes having magnitude greater than m, occurring in the source area is assumed to be of the form xiv Equation 34 Log10 Nm = a b m, In Equation 34 a and b are constants describing the seismicity of a particular source area.