Analytical Study of Thermo-Hydrodynamic Behavior of Reflood-Phase during LOCA

The objective of this study is the establishment of the thermo-hydrodynamic model of the reactor core during reflood phase of LOCA.

Based on the quench model proposed by the author, and assuming a reflood model including a flow model and a set of the thermo-hydrodynamic correlations, a reflood analysis code named “REFLA-1D” was developed.

Considerably close agreement between PWR-FLECHT tests and the results calculated by REFLA-1D code for the critical Weber number We_c= 1 was obtained for fuel clad temperature histories and the quench time and the quench temperature except for the quenching from the top of the fuel rod. It was found that the errors of calculated quench time and temperature are within ±20% under the following conditions: (1) pressure 4.5–1.5 kg/cm²·a or core inlet velocity 15–4.8 cm/s, (2) inlet subcooling more than 30°C. In the transition flow region, the calculated tendency of the temperature histories is different from the measured. This reflood model appears to be reasonable but some modifications on the low flooding quench model and the transition flow are necessary.     

Film Boiling Heat Transfer during Reflood Phase in Postulated PWR Loss-of-Coolant Accident

A single heater rod PWR reflood heat transfer experiments and analyses of the PWR- Full Length Emergency Core Heat Transfer (PWR-FLECHT) Group I data were carried out. The objectives of the experiments and the analyses were to evaluate film boiling heat transfer coefficients in the core during reflood phase of a postulated loss-of-coolant accident in pressurized water reactors, and to provide necessary information on heat transfer correlations for development of a safety analysis computer code.

The results of these experiments showed that the film boiling heat transfer coefficients are strongly dependent upon the local subcooling at the quench front. It was found that when the subcooling at the quench front was zero, the saturated film boiling heat transfer coefficients could be expressed by a correlation similar to the Bromley correlation by introducing a representative length which is defined as the distance between the quench front and the elevation at which the coefficients are evaluated. When the subcooling at the quench front is not zero, the subcooled film boiling heat transfer coefficients could be expressed by a simple correlation. This correlation predicted that experimental results within the error band of ±20%.     

Investigation of Break Location Effects on Thermal-Hydraulics during Intermediate Break Loss-of-Coolant Accident Experiments at ROSA-III

The rig of safety assessment (ROSA)-III facility is a volumetrically scaled (1/424) boiling water reactor (BWR/6) system with an electrically heated core designed for integral loss-of-coolant accident (LOCA) and emergency core cooling system (ECCS) tests. Break location effects on thermal-hydraulics during intermediate LOCAs were investigated by using four experiments at the ROSA-III, the 15 and 25% main recirculation pump suction line break (MRPS-B) experiments, the 21% single-ended jet pump drive line break (JPD-B) experiment and the 15% main steam line break (MSL-B) experiment. Water injection from the high pressure core spray (HPCS) was not used in any of the experiments. Failure of ECCS actuation by the high containment pressure was also assumed in the tests.

In the MRPS-B experiments, the discharge flow turned from low quality fluid to high quality fluid when the downcomer water level dropped to the main recirculation line outlet elevation, which suppressed coolant loss from the vessel and the core. In the JPD-B experiment, the jet pump drive nozzle was covered with low quality fluid and low quality fluid discharge continued even after the downcomer water level reached the jet pump suction elevation. Low quality fluid discharge ceased after the ADS actuation. It suggestes that the JPD-B LOCA has the possibility of causing larger and more severe core dryout and cladding temperature excursion than the MRPS-B LOCA. The MSL-B LOCA was characterized by mixture level swell in the downcomer and the core. The core mixture level swell resulted in the much later core dryout initiation than that in the MRPS-B LOCA, however, ECCS actuation was also delayed because of slow downcomer water level drop.     

Experimental Study of Two-Dimensional Thermal-Hydraulic Behavior in Core during Reflood Phase of PWR LOCA

Two-dimensional effects on the core cooling behavior during the reflood phase of a PWR-LOCA were experimentally studied by performing four tests with various radial core power profiles under the same total power and initial core stored energy conditions using the Slab Core Test Facility (SCTF). The heat transfer was enhanced and the cladding temperature was reduced for the higher and average power bundles in the steep radial power profile test especially at the upper elevation. The effect of radial power profile on the cladding temperature was quantitatively evaluated. For all tests with different radial power profiles, the collapsed water level in the upper plenum became higher in the hot leg side and the quench in the upper half of the core was delayed in the bundles corresponding to the outer bundles of a PWR core. The delay of the quench is considered to be caused by a flow stagnation trend in those bundles because the pressure in the outer bundles became higher than the pressure in the inner bundles due to the nonuniform water accumulation in the upper plenum.     

Analysis of Precursory Cooling in Quenching Phenomena

It has been noted that precursory cooling plays an important role in quenching phenomena. In this work, a new model is presented by assuming that the heat transfer in precursory cooling is mainly due to film boiling, which persists in a finite length. Then the quench velocity and temperature profile are obtained based on the three-region model for one- dimensional axial heat conduction in a heated tube. We applied our model to several existing experimental results and obtained a correlation to predict the effective length of precursory cooling region. It turns out that the correlation takes an identical form for both falling-film rewetting and bottom flooding.     

Transient Analysis of Boiling Transition Phenomena Using Liquid Film Flow Model

An evaluation has been made for the covariances of neutron cross sections of ⁵²Cr, ⁵⁶Fe, ⁵⁸Ni and ⁶⁰Ni contained in JENDL-3.2. Reactions considered were the threshold reactions such as (n, 2n), (n, nα), (n, np), (n, p), (n, d), (n, t) and (n, α), the radiative capture reaction above the resonance region, and the inelastic scattering to discrete and continuum levels. Evaluation guidelines and procedures were established during the work.

A generalized least-squares fitting code GMA was used in estimating covariances for reactions of which JENDL-3.2 cross sections had been evaluated by taking account of many measured data. For cross sections that had been evaluated by nuclear reaction model calculations, the KALMAN code, which yields covariances of cross sections and of associated model parameters on the basis of the Bayesian statistics, was used in conjunction with reaction model codes EGNASH and CASTHY.

The evaluated uncertainties of a few percent to 30% in the cross sections look reasonable, and the correlation matrices show understandable trends. Even though there is no strict way to confirm the validity of the evaluated covariances, tools and procedures adopted in the present work are appropriate for producing covariance files based on JENDL-3.2. The covariances obtained will be compiled into JENDL in the near future. Meanwhile, new sets of optical model and level density parameters were proposed as one of byproducts obtained from the KALMAN calculations.     

Transient Cooling Process of Fuel Rod in Reactivity Initiated Accident

Out-of-pile experiments were performed with Zircaloy-4 rods in subcooled water environment to study the basic phenomena occurring in the transient cooling process undergone by a fuel rod during a reactivity-initiated accident (RIA) affecting a light water reactor (LWR). The experimental results show that the cooling process of the fuel rod during an RIA can be divided into three phases separated by the quenching temperature T_q and the rewetting temperature T_q .

It is also noted from the experimental results that with increasing degree of subcooling, T_q tends to rise to levels far exceeding the maximum liquid superheat temperature of water; T_q , on the other hand, is little affected by the cooling water temperature, and remains close to that of the maximum superheat temperature.

Numerical calculations indicate conclusively that radial heat transfer to coolant water is the dominant factor that governs the transient cooling process in an RIA affecting the cold start-up of a BWR, rather than the axial heat conduction through rod which is considered to be the basic mechanism of cooling that governs the reflooding process during a LOCA.     

Application of REFLA Core Model to Safety Evaluation Code

In the analysis of the core thermal-hydraulic behavior during the reflood phase of a PWR-LOCA, current safety evaluation codes like WREM code system are usually limited to use in narrow region where the employed empirical correlations are validated. In order to make a safety evaluation code more flexibly applicable, the empirical correlations in TOODEE2 code in the WREM code system was replaced with the core model built in the REFLA code. By changing the multiplication factor for the calculated heat transfer coefficient for the region above the quench front, the predicted clad surface temperatures were compared with those measured in Cylindrical Core Test Facility (CCTF) tests.

It was found that the multiplication factor 0.9 gives always a conservative prediction against CCTF data.