Validation of Doppler Reactivity Feedback in SPERT-III E-Core With the Best-Estimate Transient Code TRAC Toshiba Version

Abstract We have developed a best-estimate transient code TRAC Toshiba version, which is Toshiba’s advanced version of the TRAC code for BWR. This code is expected to simulate thermal hydraulic and neutron kinetics phenomena three-dimensionally with high accuracy for BWR transient events. For the validation of the code, reactivity accident tests in cold-startup of The Special Power Excursion Reactor Test III (SPERT-III) E-core were calculated and compared with experiment results. Two lattice code “NEUPHYS” and “TGBLA” were used for the calculation. “NEUPHYS” developed by Nuclear Fuel Industries, Ltd. is a neutron transport calculation code with Method of Characteristics. “TGBLA” is a current licensing code of Toshiba. For the confirmation of properly simulation of initial core condition of the tests, static nuclear characteristics of the core were calculated. keff, spatial power peaking, ∧/βeff, worths of control rod and transient rod were compared with experiment results, and good agreements were confirmed. In the transient tests, estimated C/E-1 of peak power without simulation error in transient rod worth were −2%±4% for “NEUPHYS”/TRAC and 1%±4% for “TGBLA”/TRAC in high reactivity insertion tests more than 1$ excluding test 20 which is considered outlier case (±4% were evaluated from uncertainty of experiment). In high reactivity insertion transient tests, temperature of UO2 pellet behaves adiabatically by the time of power peak. Therefore, temperatures of UO2 pellet at the power peak are expected to contain small amount simulation error, it is expected that doppler reactivity simulation error are dominant in aforementioned simulation errors.