Radiat Environ Biophys (2003) 42:169–174 DOI 10.1007/s00411-003-0212-9

ORIGINAL PAPER

P. Jacob · Y. G
ksu · V. Taranenko · R. Meckbach · N. G. Bougrov · M. O. Degteva · M. I. Vorobiova

On an evaluation of external dose values in the Techa River Dosimetry System (TRDS) 2000 Received: 18 June 2003 / Accepted: 4 September 2003 / Published online: 25 October 2003
Springer-Verlag 2003

Abstract Absorbed doses were determined by thermoluminescence (TL) measurements for bricks from a height of 6 m from the south-western wall of the former mill in Metlino that faced the Techa river. Measurements of the internal beta-radiation and alpha-radiation in the brick samples and of radionuclide activities in soil samples from the Techa river valley were performed. The absorbed dose in bricks due to the natural radiation was derived and subtracted from the total dose in order to obtain the absorbed dose in the bricks caused by anthropogenic sources. The results were combined with results from two previous studies. The absorbed dose in the bricks due to the radiation field after relocation of the Metlino population in 1956 was derived from dose rates in air measured in front of the sampling locations in 1996/ 1997. Based on these dose rates the dose to bricks was calculated by means of conversion factors from the literature. The absorbed dose accumulated in the bricks in the period 1949–1956 was nearly 80% of the total dose that had been determined by TL measurements. Previously derived conversion factors were applied to obtain an estimate of the gamma dose in air at the former shore of the Techa river. An uncertainty and sensitivity analysis was performed with the program package Crystal Ball. Care was taken to treat statistical and systematic uncertainties separately and to take parameter correlations into account. The resulting distribution for the gamma dose accumulated in the period 1949–1956 at the Techa river shore has a median value of 32 Gy with a 95% confidence interval of 21–45 Gy. This study confirms the corresponding value of 26.6 Gy that is used in the Techa River Dosimetry System (TRDS) 2000. P. Jacob ()) · Y. Gksu · V. Taranenko · R. Meckbach GSF–National Research Center for Environment and Health, Institute of Radiation Protection, 85764 Neuherberg, Germany e-mail: [email protected] Tel.: +49-89-31874008 Fax: +49-89-31873363 N. G. Bougrov · M. O. Degteva · M. I. Vorobiova Urals Research Center for Radiation Medicine, 454076 Chelyabinsk, Russia

Introduction In a recent summary of Southern Urals radiation studies, Kellerer [1] indicated the possibility that the Techa River Dosimetry System (TRDS) 2000 [2] might underestimate the external exposures of members of the Extended Techa River Cohort (ETRC). The external exposure assessment in the TRDS-2000 is based on data published before the year 2000 [3]. Recently, a new evaluation of the liquid releases of the Mayak Production Association (MPA) was published by Mokrov [4, 5] who suggested short-lived radionuclides such as 95Nb and 95Zn were inadequately described in the TRDS-2000 which in turn would lead to an underestimation of external doses. Although Mokrov’s re-analysis of the MPA liquid releases gives valuable information on the time-dependence and radionuclide composition of the releases that might be useful in future developments of the TRDS, his dose estimate based on re-evaluated radionuclide activities has a serious flaw. Mokrov converted the estimated radioactive contamination along the Techa river system to the gamma dose in air by assuming an infinite plane source geometry. It is, however, more appropriate to use a dose-reduction factor for a river shoreline geometry as has been done in the TRDS-2000 [6]. The two different geometries result in conversion factors that differ by a factor of 0.2 [7] which explains an important part of the discrepancy between Mokrov’s results and TRDS-2000. Nevertheless, an evaluation of the external dose estimates in TRDS-2000 by applying an independent method is necessary in order to increase the credibility of future risk estimates derived from the ETRC. The village with the highest gamma dose rate levels was Metlino which was located at the Techa river, 7 km downstream of the release point of liquid waste from the MPA. The exposure of the inhabitants of Metlino started in 1949, and seven years later they were relocated and were, therefore, no longer directly exposed to the high radiation in the town. Subsequently, a dam was built further downstream of the Techa river. The backwater

170

formed the so-called Reservoir no. 10 that flooded parts of Metlino. In the past, luminescence methods have been applied to assess absorbed dose in bricks from buildings in the Techa river valley [8, 9]. In a first independent evaluation of external dose values of TRDS-2000 [10], absorbed doses in bricks from a building close to the former Techa river bed were measured with thermoluminescence (TL) methods, and were compared with calculated doses based on the TRDS-2000 value for the external exposure at the Techa river shore in Metlino. TRDS-2000 was confirmed in so far as measured and calculated results for the absorbed dose in bricks were not significantly different. It is the purpose of the present work, to extend the data base of the study of Taranenko et al. [10] by adding further measurement results, to improve the statistical analysis by taking into account systematic errors of TL measurements, to perform a sensitivity analysis in order to identify main sources of uncertainty, and to calculate the absorbed dose in air accumulated at the Techa river shore during the period 1949–1956. This allows us to more directly estimate the confidence intervals for the evaluation of external dose values in the TRDS-2000 system for Metlino. Also, details of the luminescence measurements are described more explicitly.

Fig. 1 South-western wall of Metlino mill. Arrows indicate locations where brick samples have been extracted at heights of 6 and 4m where Nij is the TL signal of the i-th aliquot due to the j-th irradiation. For each sample and for each irradiation, average values:

Materials and methods Luminescence measurements of absorbed dose in bricks The south-western wall of Metlino mill faced the former Techa riverbed location at a distance of 10 m (Fig. 1 in [11]). During several missions in 1996 and 1997 bricks were sampled from this wall. The uppermost sampling locations (nos. 29, 32 and 34) were taken at a height of 6 m above the present water level in Reservoir no. 10 (Fig. 1). There are three reasons why the present work focusses on brick samples from these highest locations: first, it has previously been shown that the exposure period of interest, i.e. during 1949–1956, provided the major contribution to the absorbed dose in bricks from the uppermost part of the south-western wall of the mill [11]. Second, absorbed doses in bricks from the uppermost part show the lowest variability due to inhomogeneities of the activity distributions on the ground level. Third, the uppermost level of the wall has the smallest risk of surface contamination due to water spray from Reservoir no.10. Conventional methods were applied to extract the quartz fraction from the samples [12]. The quartz samples were etched by hydrofluoric acid (HF) in order to eliminate the contribution from the a radiation of natural radionuclides. TL measurements were performed and conventional tests for the suitability of the quartz samples were applied. The absorbed dose, DTL, accumulated in the bricks in the period from their production up to the TL measurement was determined by the multi-aliquot-regeneration technique [13, 14]. From each sample, six aliquots were taken and the TL signals Ni0; i=1,...,6 recorded with a RISØ TL/OSL-12 reader. Each of the aliquots were heated three times and subsequently exposed to the built-in 90Sr source with doses Dj; j=1,...,3. The 90Sr source was calibrated with respect to the 60Co source in the Secondary Standard Dosimetry Laboratory at GSF. Three dose estimates Dij were obtained for each of the aliquots according to: Dij ¼ Dj
Ni0
Nij1

ð1Þ

Dj ¼

6 X

Dij =6

ð2Þ

i¼1

were calculated. All samples passed the suitability test that the largest and the smallest of the three dose estimates differed by less than 3%. For each of the samples, an estimate of the absorbed dose in the brick, DTL, was obtained by: DTL ¼

3 X

Dj =3

ð3Þ

j¼1

DTL measurements were performed for depth layers of 10€5 mm and 20€5 mm. The stochastic uncertainty of DTL for each of the samples is described by a normal distribution with a standard deviation that is equal to the standard deviation of the mean value of the results for the six aliquots. The systematic uncertainty is assumed to be determined by the calibration error with a rectangular distribution with €5% of the measurement value. The absorbed dose, DX, caused by anthropogenic sources in the bricks is obtained from DTL by subtracting the natural background dose, DBG, which is determined by the product of the age, A, of the brick with the sum of dose rates due to cosmic, beta, and gamma radiation: DBG ¼ A
ðD_ cos þ D_ b þ D_ g Þ

ð4Þ

The best estimate of the age A of the bricks from the mill has been assessed with TL methods in earlier work to be 129 years [11]. The measurements had been performed in 1996, therefore the production year was assessed to be 1867. Historical records about statistical data from the Urals region mention the mill in 1873 [15] so that the bricks have definitely been produced before 1873. For the present analysis, the production year of the bricks was assumed to be in the period 1861–1873. The TL measurements for the present work were performed in 2002, that means the best estimate for the age of the bricks was 135 years.

171 ˙ cos [12]. A generic value of 0.18 mGy year has been used for D ˙ b is the absorbed beta dose rate in quartz measured using thin D ˙ g is the absorbed dose rate in layer Al2O3:C beta dosimeters [16]. D the surface layer of the wall due to the gamma radiation from natural radionuclides in each of the bricks and in the soil. It has been assessed according to: 1

surface brick ˙ g ¼ 1=2ð0:85
D ˙ soil ˙ infinite D þD Þ g g

ð5Þ

˙ gsoil surface, the gamma dose rate at the surface of the For D ground, a generic value of 0.89 mGy year1 has been assumed, based on measurements of the content of natural radionuclides in soil samples from the Techa river valley. The factor 0.85 was introduced to take into account the attenuation of the radiation in air ˙ ginfinite brick is the gamma dose rate in an up to the sampling height. D infinite brick medium due to uranium, thorium and potassium contents. The content of uranium and thorium was determined in each of the bricks with conventional thick-sample a counting procedures [17]. Their contribution was subtracted from the measured beta dose rate (see above) in order to assess the 40K content. Then infinite medium dose-rate conversion factors [18] ˙ ginfinite brick. The factor 1/2 is introduced were used to calculate D because the surface brick layer is exposed from one half space to the radiation from the radionuclides in the ground and from the other half space to the radiation from radionuclides in the wall. The wall of the mill was sufficiently thick (37 cm) to allow this approximation. The stochastic uncertainty of DBG is estimated from ˙ b and D ˙ ginfinite brick. the variability of the sample specific results for D The stochastic uncertainty of the dose estimate DX was calculated from the stochastic uncertainties of DTL and DBG by error propagation assuming that the two sources of uncertainty are independent of each other. Resulting uncertainties were small and turned out to be not representative for the variability of absorbed doses in the bricks from the different sampling positions (see Results section). Therefore, the measured values were combined with published values for other bricks from the same wall in order to estimate the stochastic uncertainty of DX from the variability of these measured values. The systematic uncertainties have been described by rectangular ˙ cos (€14%) [12], D ˙ b (€5%), and D ˙g distributions for A (€6 years), D (€15%). It has been assumed that the different sources of systematic error are independent of each other. The complete uncertainty distribution of the absorbed dose DX was calculated from the statistical and systematic uncertainty with the program package Crystal Ball assuming that both sources of uncertainty are independent. A supplementary calculation was done with results for bricks sampled from a height of 4 m. Corresponding sampling locations are indicated in Fig. 1. Absorbed dose in air accumulated at the Techa river shore during 1949–1956 The present dose rate in air at a height of 6 m in front of the southwestern wall of the Metlino mill was measured during summer time with the Automess 6150 and Gamma Sonde AD-18 [10]. Subtracting the natural background dose rate of 0.1 mGy h1, a value of 1.2 mGy h1 (95%CI: 1.0–1.5) was obtained for the present dose rate in air, D_ air res , due to the anthropogenic radionuclides in the reservoir geometry. The absorbed dose, Dbrk res , accumulated in the bricks during the presence of Reservoir no.10, i.e. since 1957 until the sampling in 1996 and 1997, was calculated according to: Dbrk res

¼ Cres
kres


Z1997

D_ air res
expð0:693t=Teff Þ
dt

ð6Þ

1957

In this calculation the following results of Taranenko et al. [10] were used: – Compared to the dose-rate in summertime, dose rates averaged over the whole year are reduced due to the snow cover during

wintertime by a factor kres with a symmetrical triangular distribution with minimal and maximal values of 0.80 and 0.93 (throughout the paper, doses in air relate to summertime and doses in bricks to an average over all seasons). – The dose-rate decreased in the considered period exponentially according to the half-life Teff with a symmetrical triangular distribution with a mode of 21.2 years (taken from the data survey on measurements of radionuclide concentrations in the water of Reservoir no. 10 [10] and a maximal value 30.1 years (radioactive decay only). – The conversion factor Cres from the dose in air to the dose in bricks has a distribution with a median value of 0.45 and a 95% confidence interval of 0.28–0.63. The results for Dbrk res were subtracted from DX in order to obtain the absorbed dose, Dbrk riv , in bricks due to anthropogenic sources in the period 1949–1956: brk Dbrk riv ¼ DX  Dres

ð7Þ

As a basic quantity, the TRDS-2000 uses the integrated gamma dose-rate during summertime in air at a height of 1 m above the shore of the Techa river, Dair riv . In the present work an independent estimation of Dair riv was obtained by: brk Dair riv ¼ Driv =ðCriv
kriv Þ

ð8Þ

Based on the work of Taranenko et al. [10], the following assumptions were made in the calculation: – The dose in air at the shore is obtained by dividing the dose in the brick by a conversion factor Criv that has a distribution with a median value of 0.10 and a 95% confidence interval of 0.086– 0.15. – The integral over summer dose-rates at the shore is obtained by dividing the dose in air by a snow cover shielding factor, kriv, with a symmetrical triangular distribution with minimal and maximal values of 0.68 and 0.89. The conversion factor Criv has been calculated in [10] by photon transport calculations considering radioactive contamination of the shore (extending up to 1 m on both sides of the river) and of floodplains (extending up to the south-western wall of the mill on one side and up to 18 m on the opposite side of the river). The ratio of the activities per unit area at the shore and at the floodplains was derived from the ratio R of the gamma dose-rate in air at the shore and at a distance of 10 m from the river as it was measured (with large uncertainty) in the period 1952–1954 [10]. The ratio b of the activities per unit area on the two floodplains was introduced as an uncertain input parameter. The same calculation as outlined in Eqs. 6, 7 and 8 was performed with slightly changed parameter values [10] for bricks sampled from a height of 4 m. These supplementary calculations were combined with the results for 6 m by forming a weighted sum for the gamma dose-rate in air at the Metlino shore of the Techa river during summer time integrated over the period 1949–1956: h .
2 4m air .
4m air 2 i 6m air Dair Driv 6m sair þ Driv sriv = riv ¼ riv h .
i . 
 2 2 1 6m sair ð9Þ þ 1 4m sair riv riv where the upper left indices ‘6 m’ and ‘4 m’ indicate from which air sampling height of the bricks Dair riv was derived, and sriv indicates the standard deviation expressing the complete uncertainty. Correlations of parameters in the two calculations for bricks from 6 m height and from 4 m height were taken into account. The uncertainty and sensitivity analyses were performed with the program package Crystal Ball.

172

Results and discussion Luminescence measurements of absorbed dose in bricks The background dose rates in the etched quartz samples are about 3 mGy year1 (Table 1). The internal beta radiation contributes about 65% to the background doserate. The magnitudes of the statistical uncertainty and the systematic uncertainty are about the same. For both, the uncertainty of the gamma dose rate contributes most to the total uncertainty. Values of DTL at a depth 10€5 mm in the three brick samples from a height of 6 m of the south-western wall of the mill ranged between 3.24 Gy at location no. 29 to 4.44 Gy at location no. 34 (Table 2). The natural radiation contributes only about 0.4 Gy to the dose in the quartz samples. The absorbed dose due to anthropogenic sources at a depth of 10€5 mm was found to be by 6–18% larger than at a depth of 20€5 mm. This is in agreement with an attenuation in this depth range of about 15%, as it is expected for gamma-ray with energies larger 600 keV that were emitted by radionuclides on the ground in front of a wall [12]. Since the systematic errors are correlated for the measurement of the samples, the statistical errors would at a first glance be expected to describe the variability of the absorbed dose at a given depth. However, the variability of the measured doses is considerably higher. It indicates that the variability of the absorbed doses in the aliquots from the single samples is smaller than the variability of the absorbed doses in the quartz samples from different bricks. It is not clear whether these differences can be explained by the exposure conditions. In order to get a more reliable estimation of the statistical uncertainty of the absorbed dose DX at a depth of 10€5 mm, results of in total 8 measurements were compiled (Table 3). The mean value of the measurements

was 3.06 Gy with a standard deviation (of the mean) of 0.24 Gy. The distribution of the values was asymmetric, i.e., the mean value was larger than the median. Therefore, in the further steps of the calculation the statistical distribution of DX was assumed to be lognormal. According to Table 2, a best estimate for DX of 3.06 Gy corresponds to a systematic uncertainty with a standard deviation of 0.10 Gy. The calculation of the complete distribution of DX (Fig. 2) resulted in a median value of 3.05 Gy with a 95% confidence interval of 2.57– 3.60 Gy, and 83% of the complete uncertainty was found to be due to the statistical uncertainty. The two values from Bougrov et al. [11] in Table 3 were obtained with the pre-dose method. There is a potential of dose underestimation in the application of the pre-dose method to absorbed doses exceeding 1 Gy [13]. Therefore, an alternative calculation was performed in which only the six values of Wieser et al. [19] and of the present work were used. The resulting distribution of DX is nearly the same as in the case when all eight measurements were taken into account. The reason for a comparable uncertainty range is that two effects com-

Table 3 Absorbed dose DX caused by anthropogenic sources at a depth of 10€5 mm in bricks that were sampled at a height of 6 m from the south-western wall of Metlino mill Location no.

Absorbed dose DX (Gy)

29 32 34

Bougrov et al. [11]

Wieser et al. [19]

Present work

– 1.96 3.63

2.92a 2.61 2.84b

2.82 3.58 4.04

a Average b

of 2 values obtained by measurement repetition. A second value (3.01 Gy) with a large uncertainty was given by Wieser et al. [19] obtained by the additive dose method and is not included in the present analysis.

Table 1 Annual beta, gamma and total background dose rates in brick samples. Systematic p uncertainty has been expressed by the standard deviation that is related to the half width of the uniform distribution by a factor of 1/ 3 Location no. 29 32 34

˙ b (mGy year1) D

˙ g (mGy year1) D

˙ cos+D ˙ b+D ˙ g (mGy year1) D

Best estimate

sstat

ssyst

Best estimate

sstat

ssyst

Best estimate

sstat

ssyst

1.96 1.94 1.88

0.05 0.06 0.05

0.06 0.06 0.05

0.97 0.91 0.94

0.12 0.10 0.10

0.08 0.08 0.08

3.11 3.03 3.00

0.13 0.12 0.12

0.10 0.10 0.10

Table 2 Absorbed dose DTL in bricks that were sampled at a height of 6 m from the south-western wall of Metlino mill, and contributions, DBG, due to natural background radiation and, DX, due to anthropogenic sources Location no./depth (mm) 29/10 29/20 32/10 32/20 34/10 34/20

D

TL

(Gy)

D

BG

(Gy)

D

X

(Gy)

Best estimate

sstat

ssyst

Best estimate

sstat

ssyst

Best estimate

sstat

ssyst

3.24 3.06 3.99 3.33 4.44 3.80

0.05 0.03 0.03 0.03 0.06 0.03

0.09 0.09 0.12 0.10 0.13 0.11

0.42

0.02

0.02

0.41

0.02

0.02

0.41

0.02

0.02

2.82 2.64 3.58 2.92 4.04 3.40

0.05 0.03 0.04 0.03 0.06 0.03

0.10 0.09 0.12 0.10 0.13 0.11

173

values obtained in the GSF laboratory again yielded a similar result. Absorbed dose in air accumulated at the Techa river shore during 1949–1956

Fig. 2 Distribution of absorbed dose DX due to anthropogenic sources in a depth of 10€5 mm at a height of 6 m of the southwestern wall of the mill at Metlino. The dotted line represents the distribution according to statistical uncertainty (lognormal with a standard deviation of 0.24 Gy), the broken line represents the distribution according to the systematic uncertainty (uniform with a standard deviation of 0.10 Gy), and the solid curve represents the resulting complete distribution as calculated with the program package Crystal Ball

pensate each other. On one hand the uncertainty of the mean is larger due to the smaller number of measurements, on the other hand it is smaller because the results are more consistent. For a sampling height of 4 m, five TL measurements of DX were compiled (Table 4). The resulting contribution taken into account the complete uncertainty has a median value of 3.24 Gy with a 95% confidence interval of 2.60– 4.01 Gy, and 90% of the complete uncertainty was found to be due to the statistical uncertainty. Restriction to the

Table 4 Absorbed dose DX due to anthropogenic sources at a depth of 10€5 mm in bricks sampled at a height of 4 m from the southwestern wall of Metlino mill Location no.

27 28 33

Absorbed dose DX (Gy) Bougrov et al. [11]

Wieser et al. [19]

Present work

4.38 2.55

3.70 2.70

2.98 -

Table 5 Distribution of absorbed doses due to anthropogenic sources at a depth of 10€5 mm in 8 bricks that were sampled at a height of 6 m from the south-western wall of Metlino mill, and absorbed doses in air at the Techa river shore, accumulated during 1949–1956 (without shielding due to snow cover during winter time)

Quantity

DX Dbrk res Dbrk riv Dair riv a

For a sampling height of 6 m, the radiation contribution to the absorbed dose in the analysed bricks in the period 1957–1997 was found to be about 0.35 Gy (Table 5). The result has a large uncertainty. However, since the contribution to the dose DX is small (12%), the large uncertainty does not significantly influence the result for the absorbed dose in the bricks accumulated during the period 1949–1956. According to the calculations based on eight measurements of absorbed doses in bricks at a sampling height of 6 m, the absorbed in air at the shore of the Techa river (neglecting the shielding of the snow cover during winter time), Dair riv , has a best estimate of 32 Gy with a 95% confidence interval of 21–45 Gy. More than 50% of the uncertainty is contributed by the parameters R and b, i.e., by uncertainties of the radionuclide distribution (not the amount of the contamination which is not an input parameter of the calculation) during the period 1949– 1956. A significant decrease in the uncertainty of the result could therefore only be achieved by an improvement of the knowledge on the radionuclide distribution. The statistical uncertainty of the absorbed dose in the bricks contributes 23% to the total uncertainty. Increasing the precision of the absorbed dose determination in the bricks, e.g. by increasing the number of TL measurements, is the second candidate to reduce the uncertainty of the final result. Results based on TL measurements for bricks from a sampling height of 4 m, and for restricted sets of measurements essentially give the same result as discussed above. However, the uncertainties are considerably higher for a sampling height of 4 m (Fig. 3). Combining the results for both sampling heights leads to a result that is close to the calculation for 6 m sampling height showing that taking into account the results for 4 m sampling height does not improve the precision of the final result. In TRDS-2000, a value of 26.6 Gy is used for Dair riv which is not significantly different from the results derived from the TL measurements of bricks (Fig. 3). The TRDS-2000 value is lower by a factor of 1.2 than the best

Absorbed dose (Gy) Median

95% CI

3.05 0.35 2.69 32.4

2.57–3.60 0.20–0.60 2.16–3.27 21.1–44.5

Parameter contribution to uncertaintya

sstat (83%), ssyst (17%) Cres (60%), Teff (27%) sstat (74%) R (36%), sstat (23%), b (19%)

Only parameters contributing at least 15% to the uncertainty.

174

Fig. 3 Absorbed dose in air at the shore of the Techa river in Metlino accumulated in 1949–1956 (without shielding by snow cover during winter time) as derived from luminescence measurements of the absorbed dose in bricks that were sampled at heights of 6 m and 4 m from the south-western wall of the mill. Lines indicate the median value and boxes indicate the 95% confidence interval of the estimated distribution. The corresponding TRDS2000 value is shown for comparison

estimate in the evaluation exercise. According to the evaluation, Dair riv lies with 95% confidence in a range of 1.65–0.78 times the TRDS-2000 value. There is no indication for a considerable underestimation of the absorbed dose in air accumulated in Metlino at the shore of the Techa river in the period 1949–1956. It should however be noted that a major source of uncertainty of the external dose estimates in the TRDS-2000 are assumptions on the times spent by the cohort members at and in contaminated water bodies, and assumptions on radiation levels at more distant locations. These two factors could not be evaluated in the present study. Acknowledgement This work was supported by the European Commission under contract no. FIGH-CT1999–00007 and International Science and Technology Center Project no. 509.

References 1. Kellerer AM (2002) The Southern Urals radiation studies. Radiat Environ Biophys 41:307–316 2. Degteva MO, Shagina NB, Tolstykh EI, Vorobiova MI, Napier BA, Anspaugh LR (2002) Studies on the extended Techa river populations: dosimetry. Radiat Environ Biophys 41:41–44

3. Mokrov Y, Glagolenko Y, Napier B (2000) Reconstruction of radionuclide contamination of the Techa River caused by liquid waste discharge from radiochemical production at the Mayak Production Association. Health Phys 79:15–23 4. Mokrov YG (2002) A reconsideration of the external dose assessment for the Techa river population. Radiat Environ Biophys 41:303–302 5. Mokrov YG (2003) Reconstruction of the radionuclide spectrum of liquid radioactive waste released into the Techa river in 1949–1951. Radiat Environ Biophys 42:7–15 6. Degteva MO, Vorobiova MI, Kozheurov VP, Tolstykh EI, Anspaugh LR, Napier BA (2000) Dose reconstruction system for the exposed population living along the Techa river. Health Phys 78:542–554 7. Eckerman KF, Ryman JC (1993) External exposures to radionuclides in air, water, and soil. Federal Guidance Report No. 12. EPA-402-R-93-081 8. Gksu HY, Heide LM, Bougrov NG, Dalheimer AR, Meckbach R, Jacob P (1996) Depth-dose distribution in bricks and by Monte Carlo calculation for external g-dose reconstruction. Appl Radiat Isot 47:433–440 9. Gksu HY, Degteva MO, Bougrov NG et al. (2002) First international intercomparison of luminescence techniques using samples from the Techa river valley. Health Phys 82:94–101 10. Taranenko V, Meckbach R, Degteva MO, Bougrov NG, Gksu Y, Vorobiova MI, Jacob P (2003) Verification of external exposure assessment for the upper Techa riverside with luminescence measurements and Monte Carlo photon transport modeling. Radiat Environ Biophys 42:17–26 11. Bougrov NG, Gksu HY, Haskell E, Meckbach R, Jacob P, Degteva MO (1998) Issues in the reconstruction of environmental doses on the basis of thermoluminescence measurements in the Techa riverside. Health Phys 75:574–583 12. ICRU (2002) Report 68. Retrospective assessment of exposures to ionising radiation. International Commission on Radiation Units and Measurements. Journal of the ICRU 2 No 2 13. Bailiff IK, Bøtter-Jensen L, Correcher V, Delgado A, Gksu HY, Jungner H, Petrov SA (2000) Absorbed dose evaluations in retrospective dosimetry: methodological developments using quartz. Radiat Meas 32:609–613 14. Gksu HY, Schwenk P (2000) Thermoluminescence dating of terrazzo from the monastery church of Tegernsee (Bavaria, Germany) using the 210C TL peak of quartz. Radiat Environ Biophys 39:301–308 15. Choupin N (1873) Geographical and statistical dictionary of the Perm region (in Russian). Publishing House Popovoi, Perm 16. Gksu HY, Bulur E, Wahl W (1999) Beta dosimetry using thinlayer a-Al2O3:C TL detectors. Radiat Prot Dosim 84:451–455 17. Aitken MJ (1985) Thermoluminescence dating. Academic Press, London 18. Nambi KSV, Aitken MJ (1986) Annual dose conversion factors for TL and ESR dating. Archaeometry 28:202–205 19. Wieser A, Aragno D, Baiankine S et al. (2000) Dose reconstruction for workers of Mayak and for the Techa river population. GSF-Report 19/00, GSF–National Research Center for Environment and Health, Neuherberg, Germany