원전 종사자 삼중수소수 내부피폭 선량평가의 불확도 평가

Abstract

방사성 핵종 섭취로 인해 개인이 받는 선량은 장기의 흡수선량을 직접측정 할 수 없기 때문에 생체시료의 측정과 가상의 개인에 대해 주어진 선량평가에 관한 모델과 인자들을 통해 예측 및 계산된다. 하지만 이러한 모델과 인자들은 사실적 묘사의 한계, 개인 간의 다양성, 불완전한 측정 및 평가 등에 의해 필연적으로 큰 불확도를 내포하게 된다. 특히 증기형태로 흡입되는 삼중수소수(tritated water; HTO)의 경우, 중수로 원전에서 대량으로 발생하며 중수로 원전 종사자 방사선 피폭의 약 30%를 차지하는 주요 내부 피폭원이므로 평가된 선량에 비해 지나치게 과한 피폭을 피하기 위해 불확도 정량화가 필수적이다. 이에 본 연구에서는 원전 종사자의 삼중수소수 내부피폭 선량에 대한 불확도 평가를 수행하여 그 대표성을 재고하였으며 민감도 분석을 통해 인자별 불확도에 대한 기여도를 분석하였다. 삼중수소수 내부피폭 선량 불확도 평가를 위해 기저 불확도와 선택 불확도를 정의하였고 불확도 인자를 그 성격에 따라 구분하여 선정하였다. 본 연구에서의 기저 불확도 인자는 생리역동학 모델 전이율, 전신 질량, 전신 수분량, 방사선 가중치 및 소변시료 내 삼중수소 측정이 포함되었고 선택 불확도 인자로는 소변 제출시간, 체내 붕괴 수 계산법, 선원 조직 및 성별이 분석되었다. 기저 불확도 인자는 실험 데이터를 기반으로 근사(Fitting)와 조율(Tuning)을 통해 그 확률분포를 도출하였으며 몬테칼로 기법을 이용하여 삼중수소 내부피폭 선량의 분포로 합성하였다. 그 결과 선량의 최소·최대 값이 15배 가량 차이가 나는 매우 넓은 범위에 분포되어있는 것을 확인하였다. 또한 선량 분포의 중간 값으로 개인의 선량을 평가한다면 90% 신뢰도로 실제 선량은 약 2배의 범위 내에 존재할 것으로 판단된다. 현재 원전에서의 삼중수소 감시 방법으로 산출된 단일 선량 값은 모든 기저 불확도를 고려한 선량 분포에서 8% 내지 12%에 위치하는 다소 낮은 값으로 나타났으며 이는 선량 분포 중간 값보다 약 50% 작은 값이다. 하지만 방사선 가중치의 불확도를 배제하는 경우 기존의 단일 선량 값이 63% 내지 80%로 나타나 현재 중수로 원전 종사자 삼중수소 내부피폭 선량의 보수성을 확인할 수 있었다. 한편 민감도 분석을 통해 삼중수소 내부피폭 선량 불확도는 방사선 가중치의 불확도에 가장 큰 영향을 받는 것을 확인하였고 이는 불확도 저감화를 위해서는 방사선 가중치에 초첨을 맞추어야 함을 시사한다. 특히 RBE 표준선원에 대한 보다 더 명확한 정의가 필요함을 확인하였다. 작업 감시의 경우 체외 배설율, 즉 생물학적 반감기에도 크게 영향을 받는 것으로 나타났으며, 이는 개인별 삼중수소 교유 유효반감기를 적용하여 삼중수소 내부피폭 선량 불확도를 효과적으로 감소시킬 수 있음을 나타낸다. 삼중수소 내부피폭 선량의 선택 불확도 인자에 따라서는 선량이 10% 가량 차이가 나는 것을 확인하였지만 이는 내부피폭 선량에 필연적으로 내포되는 기저 불확도의 크기를 고려할 때 큰 의미를 가지는 차이는 아닌 것으로 판단된다. 본 연구의 결과는 삼중수소 내부피폭 선량 불확도의 연구목적을 위해 산출된 선량 분포로서 정량적 자료로 직접 적용하기에는 무리가 있다. 하지만 본 연구에서는 삼중수소에 대한 최신 자료를 기반으로 불확도를 평가하였고 원전에서의 삼중수소 감시 방법과 실질적인 비교가 수행되었으며 민감도를 분석을 통해 인자별 선량에 대한 기여도를 분석한 데에 의의가 있다. 본 연구의 결과와 논의가 원전 종사자의 삼중수소 선량에 대한 신뢰성 재고와 불확도 감소를 위한 노력에 있어서 합리적인 방향을 제시하는 근거자료로 활용될 것으로 기대된다.| Because absorbed doses of organs cannot directly measured, internal dose caused by radionuclide intakes is predicted and calculated using bioassay measurement and dosimetric models and parameters for hypothesis individual. These models and parameters, however, necessarily have a large uncertainty from various reasons like a simplification of reality, inter-individual variability and lack of knowledge. In particular, tritiated water(HTO) as vapour form mainly produced in heavy water reactor(HWR) accounts for 30% of total amount of exposure for HWR workers, thus, the quantification of uncertainty in internal dose for HTO is necessary to avoid much larger exposure than estimated dose with reference values. In this study, therefore, the uncertainty in internal dose for HTO was assessed to reconsider the representability with sensitivity analysis of uncertainty factors to the dose. To assess the uncertainty, we selected factors contributing the uncertainty and assigned to 'underlying uncertainty' or 'selective uncertainty' defined in this study according to its characteristics. Probability distributions of underlying uncertainty parameters, including the transfer rates of biokinetic model, total-body mass, total-body water, radiation weighting factor and measurement of urine sample, were obtained by 'fitting' and 'tuning' with experimental data and propagated using monte-carlo technique. As a result, the calculated internal doses widely distributed in a range that was a factor of 15 from the highest to lowest values. Additionally, we could find the internal dose for HTO, considering the central value of dose distribution, can be regarded as known to within a factor of 2 with 90% confidence. The single dose value estimated on the basis of the current monitoring method, not considering uncertainty, falls at the 8th-12th percentiles and approximately 50% lower than the median value calculated. When the radiation weighting factor was fixed as unity, however, the value is at the 63th-80th percentiles and appears to be appropriately conservative. The results of sensitivity analysis show that radiation weighting factor of tritium is of paramount importance when estimating internal dose for HTO. Therefore, in order to reduce the uncertainty related to the tritium dose, our focus should be toward radiobiological effects of tritium. In particular, more narrow definition of RBE standard source is needed. For operational monitoring, the excretion rate(eg. biological half-time) also has a large sensitivity on committed effective dose. Adopting individual effective half-time, therefore, can be considered as one of effective methods to reduce the uncertainty. Meanwhile, the comparative analysis according to selective uncertainty factors(eg. including submission time of urine sample, method for estimating the number of disintegration in body, source region and sex) shows that subjective selection of a dosimetrist can change the median value of dose distribution up to 10%. However, the effect is considered to be insignificant when compared to the magnitude of underlying uncertainty. The results in this study are inappropriate to be directly applied as quantitative data because the dose distributions are just for study purpose. However, this study has great significance in that it includes recent biokinetic data of tritium, practical comparison to current monitoring method of tritium in NPP, and sensitivity analysis for estimating contribution of each parameters. The results and discussion of this study are expected to be useful to assess the reliability of tritium internal dose and reasonability of efforts to reduce the uncertainty.; Because absorbed doses of organs cannot directly measured, internal dose caused by radionuclide intakes is predicted and calculated using bioassay measurement and dosimetric models and parameters for hypothesis individual. These models and parameters, however, necessarily have a large uncertainty from various reasons like a simplification of reality, inter-individual variability and lack of knowledge. In particular, tritiated water(HTO) as vapour form mainly produced in heavy water reactor(HWR) accounts for 30% of total amount of exposure for HWR workers, thus, the quantification of uncertainty in internal dose for HTO is necessary to avoid much larger exposure than estimated dose with reference values. In this study, therefore, the uncertainty in internal dose for HTO was assessed to reconsider the representability with sensitivity analysis of uncertainty factors to the dose. To assess the uncertainty, we selected factors contributing the uncertainty and assigned to 'underlying uncertainty' or 'selective uncertainty' defined in this study according to its characteristics. Probability distributions of underlying uncertainty parameters, including the transfer rates of biokinetic model, total-body mass, total-body water, radiation weighting factor and measurement of urine sample, were obtained by 'fitting' and 'tuning' with experimental data and propagated using monte-carlo technique. As a result, the calculated internal doses widely distributed in a range that was a factor of 15 from the highest to lowest values. Additionally, we could find the internal dose for HTO, considering the central value of dose distribution, can be regarded as known to within a factor of 2 with 90% confidence. The single dose value estimated on the basis of the current monitoring method, not considering uncertainty, falls at the 8th-12th percentiles and approximately 50% lower than the median value calculated. When the radiation weighting factor was fixed as unity, however, the value is at the 63th-80th percentiles and appears to be appropriately conservative. The results of sensitivity analysis show that radiation weighting factor of tritium is of paramount importance when estimating internal dose for HTO. Therefore, in order to reduce the uncertainty related to the tritium dose, our focus should be toward radiobiological effects of tritium. In particular, more narrow definition of RBE standard source is needed. For operational monitoring, the excretion rate(eg. biological half-time) also has a large sensitivity on committed effective dose. Adopting individual effective half-time, therefore, can be considered as one of effective methods to reduce the uncertainty. Meanwhile, the comparative analysis according to selective uncertainty factors(eg. including submission time of urine sample, method for estimating the number of disintegration in body, source region and sex) shows that subjective selection of a dosimetrist can change the median value of dose distribution up to 10%. However, the effect is considered to be insignificant when compared to the magnitude of underlying uncertainty. The results in this study are inappropriate to be directly applied as quantitative data because the dose distributions are just for study purpose. However, this study has great significance in that it includes recent biokinetic data of tritium, practical comparison to current monitoring method of tritium in NPP, and sensitivity analysis for estimating contribution of each parameters. The results and discussion of this study are expected to be useful to assess the reliability of tritium internal dose and reasonability of efforts to reduce the uncertainty.