DORT/TORT Analysis of the Hatch Unit-1 Jet Pump Riser Brace Pad Neutron Dosimetry Measurements with Comparisons to Predictions Made with RAMA (NUREG/CR-6887)
On this page:
Download complete document
Manuscript Completed: April 2004
Date Published: November 2005
J.F. Carew, K. Hu, A. Aronson,
A.N. Mallen and M. Todosow
Brookhaven National Laboratory
Upton, NY 11973-5000
With contributions by
DJB. Jones and K.E. Watkins
TransWare Enterprises, Inc.
W.E. Norris, NRC Project Manager
Division of Engineering Technology
U.S. Nuclear Regulatory Commission
Washington, DC 20555-0001
NRC Job Code Y6883
Regulatory Guide 1.190 for determining pressure vessel fast neutron fluence requires that the vessel fluence calculational methodology be evaluated using dosimetry measurement benchmarks. To insure the viability of the underwater welding that has been proposed for repairing Boiling Water Reactor (BWR) highly irradiated stainless and high nickel alloy vessel internals, an accurate and well benchmarked calculational method is required for determining the BWR thermal fluence. The recent Electric Power Research Institute (EPRI)/ Nuclear Regulatory Commission (NRC) measurement program performed at Hatch-1 provides both a fast and thermal neutron dosimetry data-base for benchmarking BWR fluence calculation methodologies.
Under the NRC "BWR Fluence" Program (JCN-Y-6391), the neutron dosimetry measurements performed at Hatch-1 are being used as a data base for assessing the accuracy of calculational methodologies used to predict neutron fluence accumulated by BWR internal components and the vessel. Results from both the well established DORT/TORT computer codes and the new state-of-the-art RAMA Fluence Methodology were to be compared in order to reduce the uncertainty in fluence estimates, and in the future, to permit a better evaluation of the feasibility of underwater welding techniques of highly irradiated components.
Detailed calculations of the Hatch-1 jet pump riser brace pad thermal and fast neutron dosimetry measurements have been performed by Brookhaven National Laboratory (BNL) using the DORT/TORT discrete ordinates transport methodology and by Transware Enterprises Inc. (TWE) using the RAMA three-dimensional fluence methodology. The calculations for both code models were performed using nuclear data primarily based on the BUGLE-96 nuclear data library. The calculations were performed using a detailed description of the Hatch-1 core/internals/vessel material and geometrical configuration. The core neutron source includes the effects of the pin-wise power distribution on the core periphery and the effects of plutonium buildup on the magnitude and energy dependence of the neutron source.
Comparisons of the fluences calculated by DORT/TORT (performed by BNL), and fluences and activations calculated by RAMA (performed by TWE), and the Hatch-1 measurements have been performed to assess the accuracy of the methodologies for predicting the fast and thermal neutron fluence of BWR internal components and the vessel. Measurement-to-Calculation (M/C) fluence comparisons were also performed. The DORT/TORT methodology was found to predict the fast and thermal fluence measurements to within ∼5% and ∼15%, respectively. The DORT/TORT fluence measurement predictions by BNL are considered to be within the combined accuracy of the calculations and measurements. The DORT/TORT thermal fluence calculations showed an average measurement to calculation ratio of 1.17. The DORT/TORT calculation under-predicted the thermal fluence measurement by ∼15%.
The RAMA methodology by TWE was found to predict the fast fluence measurements to within ∼7%. The thermal fluence measurements were predicted by RAMA with a measurement to calculation ratio of ∼0.62 with a standard deviation of 0.2%. The RAMA methodology performs a direct three-dimensional solution of the transport equation and calculates fluence predictions. The fast predictions are in good agreement with measurements. However, the RAMA code over-predicted the measured thermal fluence by ∼60%. The agreement between calculated and measured fast and thermal activations was similar to that of the fluence in both direction and magnitude. The large difference between the RAMA calculated thermal fluence and the measured thermal fluence at the pressure vessel wall warrants further investigation.
The DORT/TORT-to-RAMA differences for1the fast fluence measurements are ∼4%, and are consistent with the accuracy of the two calculational methods. The large over-prediction (∼60%) in the RAMA thermal fluence, compared to the measured fluence, prevents any meaningful comparison with the DORT/TORT synthesis approach at BNL. The DORT/TORT approach under-predicted the thermal fluence measurements by an average of ∼15%.