The STAT7 Code for Statistical Propagation of Uncertainties in Steady State Thermal Hydraulics Analysis of Plate Fueled Reactors

Abstract

60439. For information about Argonne and its pioneering science and technology programs, see www.anl.gov.

Verification and Validation of the STAT7 Code

“…The STAT7 code first generates uniformly distributed random numbers from the user-inputted random number seeds. From these uniformly distributed numbers, the final set of normally distributed random numbers are generated to be used in the TH calculation for uncertainty propagation [1]. The Chi-squared test is used to assess the "goodness-of-fit" of a given set of numbers to a known distribution.…”

“…As mentioned in the previous section, STAT7 approximates the normal distribution [Section 4.1.1-(a)] based on the uniform distribution [Section 4.1.1-(b)] bijectivity [1]. Thus, the randomness of the source uniform distribution and the target normal distribution is expected to be identical.…”

“…In this test case, core power, core flow rate, and coolant outlet temperature are set as the Whittle and Forgan experimental condition (the highlighted case from Table 3.4) [1]. In STAT7, OFI temperature is calculated using the Whittle and Forgan correlation with the bubble detachment parameter η as formulated in Equation ( 2) [36].…”

“…Figure 4.12 illustrates the node configuration. As all the boundary conditions are aligned with the experimental conditions from Whittle and Forgan [1], all 16 test cases should present OFI conditions at the outlet (see Table 3.4). However, because both PLTEMP/ANL and STAT7 OFI calculation capabilities impose bubble detachment parameter η =32.5 by default to account for uncertainties from the experiment [1,3,9], calculated OFI temperature conditions may differ from the Whittle and Forgan experimental results [1].…”

“…As all the boundary conditions are aligned with the experimental conditions from Whittle and Forgan [1], all 16 test cases should present OFI conditions at the outlet (see Table 3.4). However, because both PLTEMP/ANL and STAT7 OFI calculation capabilities impose bubble detachment parameter η =32.5 by default to account for uncertainties from the experiment [1,3,9], calculated OFI temperature conditions may differ from the Whittle and Forgan experimental results [1]. While STAT7 supports both OFI temperature calculation and OFI power search capabilities, PLTEMP/ANL only supports the power search capability and does not directly output the OFI temperatures.…”