工业下水总排口γ连续监测系统刻度因子的计算方法研究

【作者】 李航；

【导师】 肖雪夫；

【作者基本信息】 中国原子能科学研究院 ， 辐射防护与环境保护， 2006， 硕士

【摘要】 原子能院工业下水连续监控系统自上世纪90年代建成试运行至今，曾监测到放射性异常工业下水排放40余次，截获并处理含放射性异常的工业下水约4800m³，有效地杜绝了含放射性异常的工业下水直接进入环境，保证了工业下水的受控排放，为保护原子能院周边环境起到了积极有效的作用。但该系统刻度工作比较困难，而且由于探测器及监测环境的影响，对核素分析也十分困难，严重限制和制约了该系统的推广应用。作者认为，蒙特卡罗库最小二乘法（MCLLS）是解决上述问题的最好方法之一，而探测器响应函数作为MCLLS方法的基础，是本论文的主要研究内容。本文对建立探测器响应函数的两种方法——解析法和Monte Carlo法——进行了详细比较。解析方法对物理过程处理不准确，且只能用于简单的几何结构，而Monte Carlo法不仅能对物理过程进行详细地描述，而且能处理任意三维几何结构。因此本论文采用Monte Carlo法建立探测器的响应函数。采用Monte Carlo法计算时，探测器的计算模型决定了探测器响应函数的准确度。本研究使用ORTEC Maestro软件的增益稳定功能测试了²⁴¹Am、⁵⁷Co和¹¹³Sn的稳谱效果，选用¹¹³Sn作为增益稳定的参考源，取得了较好的结果。使用MCNP-4C建立了NaI（T1）探测器的探测效率计算模型，利用模型计算不同几何的探测效率，计算结果与实验测量结果比较，最大相对偏差为11％，绝大部分＜±5％。在工业下水总排口连续监控系统4号水池内使用¹³⁷Cs和⁶⁰Co点源实验，将点源置于水池径向距离分别为20.5cm、25.5cm和30.5cm的垂线上不同位置测量。将实验获得的能量刻度参数和FWHM（半高宽）刻度参数加入MCNP的输入卡内，使用MCNP对整个系统进行计算，计算谱和实验谱基本吻合。将实验测量的点源全谱探测效率和计算所得的全谱探测效率分别对整个池水进行加权积分，获取水池的刻度因子，实验刻度因子和计算刻度因子在±5％以内吻合。使用分析纯的KC1配制了浓度为80.15 Bq／1的⁴⁰K液体体源进行实验，计算值和实验值吻合效果不理想，本研究对此进行了分析。通过与王里玉研究员合作，完成了蒙特卡罗库最小二乘法解谱软件的编制，为将来工业下水监控系统进行核素分析奠定了基础。更多还原

【Abstract】 In 1990’s, A system used in continuously monitoring and controlingγ-radioactivity in Industrial Drainage was built in CIAE. From then on, it hadmonitored more than 40 times abnormal releases of induatrial drainage containingradioactive water, which was about 4800 m³ water in all. The system doesefficiently protect the abnormal release of Industrial Drainage fromenvironmental pollution, ensure the controllable release of Industrial Drainage,and protect the ambient environment of CIAE. Because of the difficulty incalibration and nuclide analysis, the system can not be widely applied in nuclearfacilities, it’s original function is restricted. In this research, the Monte CarloLibrary Least Squares methods（MCLLS） is thought to be the best way to solvesuch problem. The detector response functions（DRFs）, which are the fundamentalin MCLLS, became the major content in this research.There are two methods to estabilish the DRFs, Numeric method and MonteCarlo method, which are compared in detail in this research. Numeric method iseffective with simple geometry and it can’t deal with physical mechanismcorrectly, but Monte Carlo method can not only deal with physical mechanism indetail, but also deal with arbitrary 3-Dimension geometry, so Monte Carlomethod is chosen to estabilish DRFs in this research.The DRFs accuracy is depended on the equivalent calculation model of thedetector. In this research, the sources of ²⁴¹Am, ⁵⁷Co and ¹¹³Sn were used to testthe Gain Stabilization function in Ortec Maestro. As a result, the ¹¹³Sn presentsbetter result than other two sources. MCNP-4C version was used to estabilish thedetector calculation mode, By virtue of the comparison between the calculationdetection efficiency and the experimental detection efficiency, the calculationmodel was iteratively corrected. Eventually, the calculation model wasestabilished. The relative errors between the calculation results and the experiment results are mostly smaller than 5%, only a few greater than 5%.¹³⁷Cs and ⁶⁰Co point sources were used to verify whether the calculationmodel is right. According to the energy calibration and FWHM（Full Width atHalf Maximum） calibration results, MCNP calculation results got a good matchwith experiment results. The whole spectrum detection efficiency of experimentand calculation were used to calculate the whole pool detection efficiencyindividually, the relative error between the experimental result and calculationalresult was small than 5%.The volume source was made by dissolved 5000 gram analytical reagent ofKCl in 1 m³ water, which was 80.15 Bq/l. But the calculational result got a badmatch with the experimental result. There are some comments on this problem.The spectrum analysis software based on Monte Carlo Library Least Squareshas been compiled by cooperated with Professor Liyu Wang, which is thefundamental of nuclide analysis in the system of monitoring and controlingγ-radioactivity in Industrial Drainage.更多还原