On "Tc, 137 CS and 90Sr in the Kara

International Conference on Environmental Radioactivity in the Arctic, NvJ970001 8

90 On "Tc, 137CS and Sr in the Kara Sea

Henning Dahlgaard Rise National Laboratory, Roskilde, Denmark

Abstract Technetium-99 in the Arctic seas originates mainly from European reprocessing plants whereas 1} Cs and ^Sr have many sources. It appears that for' 'Cs, re-mobilisation from the Irish Sea ofsedimented activity from earlier discharges and the outflow of Baltic -water contaminated with Chernobyl activity, are more important sources to the present contamination of the Kara Sea than new European discharges. As opposed to 137Cs and "Tc, KSr is correlated with low salinity waters. It is argued, that this is due to runoff of global atmospheric fallout ^Srfrom land.

Sources Man made radionuclides have been studied in waters of the North-East Atlantic and the Arctic for 4 decades (Aarkrogef ai, 1987; Livingston, 1988; Dahlgaard, 1994; Dahlgaard et al., 1995; Kershaw & Baxter, 1995). Global fallout from atmospheric nuclear test explosions contaminated the world oceans during the 1950s and 1960s, with a distinct peak in 1962-1965 and a characteristic "'^ ratio of-1.5. Later, European nuclear fuel reprocessing plants, especially Sellafield in the United Kingdom gradually increased the 137Cs/"0Sr ratio in the North East Atlantic, including the Arctic Ocean and the East Greenland Current. The European discharges resulted in an increase of the 99TC/IJ7CS ratio by an order of magnitude in the East Greenland Current Polar Water (DahJgaard, 1994). Finally, the Chernobyl Nuclear Power plant accident in April/May 1986 provided significant injections of 137Cs and I34Cs, especially in the Baltic Sea.

Transport from West-Europe To quantify the transport from point sources in Europe to the Arctic, Transfer Factors have been used. Having estimated a representative transport time, t years, a Transfer Factor (TF) is calculated as the quotient between observed concentrations in the environment and an average discharge rate t years earlier. The unit for the transfer factor, Bq m"3 / PBq yr'1, is equal to 10'15 yr m'3; ie. a TF value of 1 indicates that 10'15 annual discharges are present per m3. Transfer factors from Sellafield to different sea areas are shown in Figure 1. The estimated transit time from Sellafield to the south- eastern Barents Sea and the Kara Sea is 5-6 years. The transit time from La Hague is estimated as 3 - 4 years, and from the Baltic outlet through the Danish Straits 2-3 years. Further transit times are given in Dahlgaard (1995). Discharges of 137Cs from Sellafield (UK) and La Hague (France) are shown in Figure 2 with an example of resulting concentrations in the Kara Sea applying the transfer factor concept. In practice, the picture is much more complicated. Due to complex hydrographic transport patterns, each individual discharge will be dispersed to a variable degree when arriving at the sampling points downstream from the discharge. Therefore, distinct peaks in the discharge will be smoothed over a certain period of time, and water samples will contain fractions of discharges that are several years older than the given central transport time. This "tail" effect is presently very important in the Arctic as the discharges from Sellafield has decreased 2 orders of magnitude for 137Cs and one order of magnitude for "Tc.

Kara Sea Figure 3 and Table 1 shows radionuclide concentrations measured at Rise on Kara Sea surface waters from the Russian-Norwegian sampling cruises 1992 and 1993. Applying a Transfer

91 40°W 30°W 20°W 10°W Figure 1. Major surface currents and transfer faaors from Sellafield to different sea areas (from Dahlgaard, 1995).

Table 1: measured radionuclide concentrations (Bq m"3) in Kara Sea surface waters from the Russian-Norwegian sampling cruises 1992 and 1993. Year °N °E Salinity %o "Tc 137Cs "Sr 1992 71.30 58.00 33.0 0.197 9.8 3.4 1992 74.30 62.00 31.3 0.156 12.1 3.7 1992 75.30 72.00 8.7 0.064 4.5 5.6 1992 74.30 70.00 10.1 0.063 4.8 6.0 1992 74.30 66.00 27.5 0.172 10.5 4.5 1992 73.30 68.00 20.6 0.135 9.6 6.1 1992 72.30 62.00 31.0 0.150 9.8 3.4 1992 71.30 64.00 30.6 0.155 8.9 3.2 1992 70.00 56.00 30.6 0.157 8.5 3.2 1992 70.23 31.31 33.9 0.144 7.1 2.7 1993 74.22 58.41 14.4 0.132 5.7 5.7 1993 74.26 58.37 14.7 0.110 5.8 5.4 1993 74.19 58.53 14.2 0.083 5.2 6.0 1993 74.13 58.55 15.4 0.097 5.5 5.3 1993 72.31 55.30 18.1 0.127 9.2 5.4 1993 72.33 55.22 22.1 0.124 6.7 5.0 1993 72.30 55.36 17.8 0.101 6.2 5.1 1993 72.40 58.10 31.8 0.154 7.9 2.8 1993 72.21 57.50 31.7 0.142 7.3 2.5

92 •6000

65 70 75 ' ' ' ' 80 85 90 Year in Kara & E.-Barents

->*«- Estm. Cs-137 | | Cs-137 discharge Figure 2. Releases of 137Cs to coastal waters from Sellafield (UK) and La Hague (France), and estimated concentrations in the Kara Sea assuming a Transfer Factor of 5 Bq m"3 / PBq yr"1.

Tc-99, Bq m-3, Kara 1993, Surface water Tc-99. Bq m-3, Kara 1992

'3 0.11 '

.-• -0.16-0.17~O.OS

/ ?x c-o.u <~ 10.10

71-N J

•V-.-N v

STE S6-E to-E 71TE TS-E Figure 3. Sampling sites and level ofw Tc in the Kara Sea 1992 and 1993.

Kara 1992 -1993, Cs-137/Sr-90

Figure 4. Cs / Sr ratios in Kara Sea surface water samples 1992 and 1993.

93 factor of 10 Bq m"3 / PBq yr"' and the above mentioned 5 and 3 year transit times from Sellafield and La Hague, respectively, would give 0.10 Bq wTc 1992-1993. As the average value observed is 0.13 Bq m'\ uew transports from Sellafield and La Hague may explain most of the observed activity, whereas the "tail" effect can account for the rest. Doing the same for I37Cs only explains 0.25 out of the observed 5 - 12 Bq m'3 as direct transport. As the decrease in the discharge of 137Cs has been an order of magnitude sharper than for 99Tc, more recirculation of old activity could be expected. There are, however, two additional sources of 137Cs: remobililisation of sedimented I37Cs in the Irish Sea has been estimated at approximately 260 TBq yr"1 (Hunt & Kershaw, 1990), and after the Chernobyl accident in 1986, the efflux of lj7Cs from the Baltic through the Danish Straits has been around 100 TBq yr'1. Under the same assumptions as above, this would lead to additionally 2.6 and 1 Bq l37Cs m'3 in 1992-1993. The global fallout background in the Atlantic water flowing towards the Arctic is 2.8 Bq I37Cs m"3 (Dahlgaard et al., 1995), Le. we can now explain 6 - 7 out of the observed 5 - 12 Bq m'. The remaining activity could be explained with the "tail" effect including recirculation of Arctic water holding higher levels of I37Cs than the Atlantic water (Dahlgaard, 1994). Furthermore, applying transfer factors from Sellafield may underestimate transport from La Hague and the Baltic. The concentration of ^Sr is less influenced by the European discharges, and the !37Cs / 90Sr was low in the Arctic seas before Sellafield increased the level (Livingston, 1988), probably because of excess ^Sr runoff with river water. This effect, caused by the larger mobility of '"Sr as compared to ly7Cs, is clearly visible in the present material. It appears furthermore, that the observations for 1993, where the fresh water originates in the fjords of Novaya Zemlya, fits with the data from 1992, where the fresh water originates from the large rivers Ob and Yenisey (Figure 4). Both nuclides relate clearly to salinity. Extrapolating linear regression lines to zero salinity gives 3.4 Bq I37Cs m"3 and 7.7 Bq 90Sr m'3. The observations could thus be explained if these values represented concentrations in the fresh water runoff from Ob/Yenisey as well as from Novaya Zemlya. These numbers can be compared with a resent dataset on drinking water from Greenland measured at Risa showing 1.1 ± 1.1 Bq lj7Cs m'3 and 3.7 ± 1.1 Bq 90Sr m"3. These observations strongly supports the hypothesis, that the excess ^Sr connected to loiv salinity water in the Kara Sea originates mainly from the above mentioned natural process: runoff of 90Sr from land. The Greenland data furthermore suggests, that the Kara Sea excess 90Sr is mainly runoff of global fallout. How much of the observed radionuclides could come from the nuclear material dumped in the Kara Sea? As the present marine discharges from Sellafield and La Hague show a considerable degree of 137Cs depletion, a significant 137Cs contamination from the dumped material would presumably result in inhomogeneous 137Cs / 99Tc ratios. Also, '"Sr might be expected to be more soluble than 137Cs leading to low l37Cs / ^Sr ratios close to the dumped material. This does not seem to be the case. Based on the present data, the dumped activity in the Kara Sea seems thus not 90 to increase the level of "Tc, 137CS and Sr measurably in seawater. As there is a political need to show an effect of dumping, it is therefore suggested to intensify the studies in the Kara Sea. Otherwise, we may have to conclude, that dumping in the Arctic marine environment is an acceptable option, which is not in agreement with the current environmental policy.

References Aataxj, A., Boelskifte, S., Dahlgaard, H., Duniec, S., Hallstadius, L, Holm, E. & Smith, J.N. (1987). Technetium-99 and cesium-134 as long distance tracers in Arctic waters. Estuar. Cstl. ShetfSci. 24,637-647. Dahlgaard, H. (1994). Sources of 13'Cs, ^Sr and 59Tc in the East Greenland Current. J. Environ. Radioactivity 25,37-55. Dahlgaard, H. (1995). Transfer of European coastal pollution to the Arctic: Radioactive tracers. Marine Pollution Bulletin (in press). Dahlgaard, H., Chen, Q., Hermann, J., Nies, H., Ibbett, R.D. & Kershaw, P.J. (1995). On the background level of "Tc, '"Sr and i37Cs in the North Atlantic. Journal of Marine Systems (in press). Hunt, G.J. & P.J. Kershaw (1990). Remobilisation of artificial radionuclides from the sediment of the Irish Sea. J. Radio). Prof., 10, 147-151 Kershaw, P. &. Baxter, A. (1994). The transfer of reprocessing wastes from NW Europe to the Arctic. Deep-Sea Res. (in press). Livingston, H.D. (1988). The use of Cs and Sr isotopes as tracers in the Arctic Mediterranean Seas. Phil. Trans. Roy. Soc. Lond., Ser. A 325, 161-176.