Gas-cooled Fast Reactor Research Articles

Influence of the Oxygen Partial Pressure on the Oxidation of Inconel 617 Alloy at High Temperature

Ni-based superalloy Inconel 617 (IN617) is one of the main candidate structural materials for high temperature components (heat exchanger) of the gas-cooled fast reactor (GFR), a possible candidate for generation IV nuclear reactor. The material in operating conditions will be exposed to impure He at a temperature of around 850 °C. The impurities are expected to be oxidizing (such as O2, H2O) but since no feedback experience is available for this type of reactor, the level of impurities is completely unknown. Hence, an attempt has been made to understand the influence of oxygen partial pressure on oxide composition and on the oxidation mechanisms of IN617 at 850 °C. To achieve this, oxidation tests were performed at 3 different range of partial pressure: 10−5, 0.2 and 200 mbar. Tests were performed from 1 h to 28 days and the obtained oxide layers were characterized using MEB, EDX, XPS, XRD and GD-OES. The oxide layers were mainly composed of chromia containing TiO2 and thickening with time. Aluminium oxide formed internally. Other oxides were detected in the scale, such as NiO, CoO, MoO3 and MnO2, except for the lowest oxygen partial pressure experiments, where a selective oxidation took place. The scale-growth mechanism was cationic for low and medium oxygen partial pressure conditions. A growth following a transient oxidation mechanism was observed for high oxygen partial pressure.

The thermal hydraulics in a rod bundle representative of the start-up core of the ALLEGRO Gas cooled Fast Reactor—Experimental and numerical approaches

The ALLEGRO project, led by the CEA in the European framework, is devoted to the conception and the building of the first Gas cooled Fast Reactor (GFR) demonstrator. In order to fulfil this ambitious challenge, the ESTHAIR program has the objective to validate some thermal-hydraulic aspects, mainly the correlations on the pressure loss and on the heat exchange coefficients within the core sub-assemblies, between the fuel pin and the helium coolant. This program is supported by: • an experimental air test section, which is designed to produce data for pressure drop and heat transfer coefficients for the various ALLEGRO fuel assembly concepts; • a numerical simulation program based on a simplified modelling of the ESTHAIR sub-assembly. The experimental pressure loss and heat transfer coefficients are compared to the existing correlations and to the calculated results. The study concludes in the good reproducibility of the experimental results by the numerical simulation, which will allow the selection of the most appropriate correlations.

Adhesion of YSZ suspension plasma-sprayed coating on smooth and thin substrates

The design concept of the gas-cooled fast reactor which is a 4th generation nuclear reactor, requires protective coatings able to operate at 850 °C and protect the underlying structure in case of sudden increase of the functional temperature up to 1250 °C and depressurization from 0.70 MPa to atmospheric pressure. The parts to be covered are made of 1-mm thick materials resistant to heat and erosion and exhibiting high mechanical properties at high temperatures, such as the Haynes® 230 nickel-based alloy. In this study, the use of suspension plasma spraying to manufacture zirconia coatings is explored. The spraying conditions were optimized for the elaboration of coatings on stainless steel AISI 304L substrates and then adapted for Haynes 230 substrates. A special attention was paid to coating adhesion that was investigated by using a Vickers indentation cracking method.

High temperature interfacial reactions of TiC, ZrC, TiN, and ZrN with palladium

SiC as a coating layer of nuclear particle fuels has shown susceptibility to the attack of fission products like palladium. Other high temperature capable ceramics such as TiC, ZrC, TiN, and ZrN, which were identified as candidates for gas-cooled fast reactor systems, can also be examined as potential replacement materials for SiC. The resistance to Pd attack of TiC, ZrC, TiN, and ZrN was examined by diffusion couples annealed at 1400 °C for 10 h. Scanning electron microscopy, energy dispersive X-ray spectroscopy, and electron backscatter diffraction were employed to characterize the interfacial reactions of the diffusion couples. The thickness of the reaction zone of the diffusion couples is dependent upon the microstructure of the ceramics. The results indicate the best resistance to Pd attack of TiN and ZrN, followed by TiC and ZrC in sequence. All the four ceramics showed better resistance to Pd attack than SiC.

FAST Code System: Review of Recent Developments and Near-Future Plans

The FAST code system is currently being developed and used at the Paul Scherrer Institut for static and transient analysis of the main Generation 4 fast-spectrum reactor concepts: sodium-, helium-, and gas-cooled fast reactors. The code system includes the ERANOS code system for static neutronics calculations, as well as coupled TRACE/PARCS/FRED for neutron kinetics, thermal hydraulic, and fuel transient analysis. The paper presents the status of the recent developments in neutronics (new 3D procedure for equilibrium cycle simulation and new transient cross section generation procedure), in thermal hydraulics and chemistry (equations-of-state for new coolants, two-phase flow models for sodium, and new model for oxide layer buildup in heavy-metal flow), and in fuel behavior (new model for the dispersed gas-cooled fast reactor fuel). Near-future plans for the further development of FAST are outlined.

The role sol–gel process for nuclear fuels-an overview

The paper reviews the sol–gel methods used for the preparation of nuclear fuel materials in the form of microspheres. It also discusses how these microspheres can be fabricated into nuclear fuels for reactors such as High Temperature Gas Cooled Reactors and Fast Reactors. The performance of these microsphere-based fuels is reviewed. More recent applications, such as the transmutation of minor actinides, (Np, Am and Cm) and hydrogen production, are also briefly covered.

An optimized Line Sampling method for the estimation of the failure probability of nuclear passive systems

The quantitative reliability assessment of a thermal–hydraulic (T–H) passive safety system of a nuclear power plant can be obtained by (i) Monte Carlo (MC) sampling the uncertainties of the system model and parameters, (ii) computing, for each sample, the system response by a mechanistic T–H code and (iii) comparing the system response with pre-established safety thresholds, which define the success or failure of the safety function. The computational effort involved can be prohibitive because of the large number of (typically long) T–H code simulations that must be performed (one for each sample) for the statistical estimation of the probability of success or failure. In this work, Line Sampling (LS) is adopted for efficient MC sampling. In the LS method, an “important direction” pointing towards the failure domain of interest is determined and a number of conditional one-dimensional problems are solved along such direction; this allows for a significant reduction of the variance of the failure probability estimator, with respect, for example, to standard random sampling. Two issues are still open with respect to LS: first, the method relies on the determination of the “important direction”, which requires additional runs of the T–H code; second, although the method has been shown to improve the computational efficiency by reducing the variance of the failure probability estimator, no evidence has been given yet that accurate and precise failure probability estimates can be obtained with a number of samples reduced to below a few hundreds, which may be required in case of long-running models. The work presented in this paper addresses the first issue by (i) quantitatively comparing the efficiency of the methods proposed in the literature to determine the LS important direction; (ii) employing artificial neural network (ANN) regression models as fast-running surrogates of the original, long-running T–H code to reduce the computational cost associated to the determination of the LS “important direction” and (iii) proposing a new technique for identifying the LS “important direction”, based on the genetic algorithm (GA) minimization of the variance of the LS failure probability estimator. In addition, this work addresses the second issue by assessing the performance of the LS method in estimating small failure probabilities with a reduced (e.g., lower than one hundred) number of samples. The issues are investigated within two case studies: the first one deals with the estimation of the failure probability of a nonlinear structural system subject to creep and fatigue damages [1,2]; the second one regards a passive decay heat removal system in a gas-cooled fast reactor (GFR) of literature, [3].

Heavy-gas injection in the Generation IV gas-cooled fast reactor for improved decay heat removal under depressurized conditions

Decay heat removal is a key safety and design issue for the Generation IV gas (helium)-cooled fast reactor. This paper investigates the natural convection capability of the dedicated DHR loops under depressurized conditions while injecting a heavy gas into the system. Investigated is a loss-of-coolant accident using the TRACE code. The goal of the study is to improve fuel/cladding temperature behavior during LOCA transients with the enhancement of passive safety by operation in natural convection only, while accepting 10 bar back-up pressure in the guard containment. The paper investigates the cooling capabilities of different heavy gases. Furthermore, different injection locations and mass flow rates have been tested, in order to address possible core-overcooling problems resulting from rapid depressurization of the gas reservoir. It has been shown that, among the gases investigated, CO 2 is the best choice from the thermal-hydraulics viewpoint, being able to cool the core satisfactorily for a broad range of injection rates. N 2 can be envisaged as an alternative solution in case of chemical problems with CO 2. Supplementary studies carried out for the CO 2 and N 2 injection cases include that of the sensitivity to the number of available DHR loops and to the LOCA break-size. The effect of the resulting neutron spectrum changes on the shutdown-reactivity margin has also been investigated.

High temperature oxidation of SiC under helium with low-pressure oxygen. Part 2: CVD β-SiC

In the frame of the Generation IV International Forum, Gas-cooled Fast Reactor (GFR) is one system studied by CEA (France). Helium pressurized at 7 MPa is the coolant and the nominal temperature of use is about 1300 K. The cladding materials currently considered is a SiC/SiC composite with a β-SiC coating. In case of accident, reactor temperatures can reach 1900–2300 K. A previous study was carried out to determine the physico-chemical behavior of another polytype, α-SiC, for comparison on the position of the active to passive transition and of the mass loss rates under active conditions to simulate a typical accident. Experimental oxidation tests at high temperature (1400–2300 K) on massive β-SiC samples processed by Chemical Vapor Deposition (CVD) coupled to mass variation, SEM, XPS, AFM and roughness analyses enabled to determine the transition between passive and active oxidation regimes, and to study the resistance to oxidation of such material in some conditions that might be encountered in case of accident (high temperature increase up to 2300 K). Finally, the experimental results have shown that the transition from passive to active regime occurs at higher temperature for β-SiC than for α-SiC and that the mass loss rate of β-SiC is lower than the one measured for α-SiC on the common temperature range investigated (up to 2100 K).

High temperature oxidation of SiC under helium with low-pressure oxygen—Part 1: Sintered α-SiC

Gas-cooled Fast Reactor (GFR) is one system of the Generation IV International Forum systems studied by CEA (France). The considered coolant, helium, is pressurized at 7 MPa and the nominal working temperature is about 1300 K. In case of accident, reactor temperatures can reach 1900–2300 K. Several studies are carried out to determine the physico-chemical behavior of structural and cladding materials undergoing various environmental constraints. The cladding materials currently considered are refractory carbides, and particularly silicon carbide, SiC. Experimental tests at high temperature (1400–2100 K) on massive α-SiC samples coupled to mass variation, SEM and roughness analyses enabled to determine the transition between passive and active oxidation regimes, and to study the resistance to corrosion of such material in some conditions encountered in case of accident (high temperature increase up to 2300 K). The transition between passive and active oxidation obtained experimentally appears at 1300 K for an oxygen partial pressure of 0.2 Pa and at 1600 K for 100 Pa. The experimental mass loss rate obtained for conditions encountered during an accident – active oxidation – have shown that a maximal oxygen partial pressure of 10 Pa in helium is well suited to avoid a huge degradation of the α-SiC.

Comparison of MCNPX-C90 and TRIPOLI-4-D for fuel depletion calculations of a Gas-cooled Fast Reactor

The Gas-cooled Fast Reactor is one of the reactor concepts selected by the Generation IV International Forum for the next generation of innovative nuclear energy systems. Several fuel design concepts are being investigated. Burnup depletion of mixed fuel of uranium and plutonium, cooled with gas in a fast neutron energy spectrum must be simulated. Various codes are being developed and/or adapted to improve the quality of the results, and also to reduce the computing time required for the simulations. The main objective of this work is to compare the fuel depletion results obtained with MCNPX-CINDER90 code and the new TRIPOLI-4-Depletion 1 TRIPOLI® is a registered trademark of CEA. 1 code (developed by the Commissariat à l’Énergie Atomique) of a fuel design concept for the Gas-cooled Fast Reactor. Calculations were made for an equivalent homogeneous model of fuel rods in a hexagonal mesh assembly. Total reflection conditions were applied on the six lateral faces and the two axial faces of the assembly. The materials used in the fuel assembly are: carbide of uranium and plutonium as fuel, silicon carbide as cladding, and helium gas as coolant. JEFF libraries of effective cross sections were used in both codes. Two methods of burnup step calculations were performed with TRIPOLI-4-D, the Euler and the CSADA, and their results were compared with the MCNPX-CINDER90 CSADA method. A period of 300 days of irradiation time was considered, which was divided into 12 steps. Results of the infinite multiplication factor as function of the irradiation time, and the evolution of the isotope concentrations for a selected group of nuclides were compared. The main conclusion is that very similar results were obtained for the three types of depletion calculations which were compared: (1) MCNPX-C90 CSADA; (2) TRIPOLI-4-D CSADA, and (3) TRIPOLI-4-D EULER. The best calculation time was obtained with the TRIPOLI-4-D EULER method, which needed approximately half the time than the other two. In summary, it is sufficiently good to use the first order approximation (Euler method) for the time dependent fuel depletion simulation of the fast neutronic spectrum GFR fuel studied in this work.

TRACE qualification via analysis of the EIR gas-loop experiments with smooth rods

Several gas-loop experimental programs were carried out during the 1970-1980s at the Eidgenossisches Institut fur Reaktorforschung (EIR, now PSI), Switzerland. In the present work, a wide range of thermalhydraulics tests for smooth rods, which form part of the experiments conducted at the time, have been reanalyzed with the aim of qualifying the TRACE code. The latter constitutes the thermal-hydraulics module of PSI's FAST code system, currently being applied to the multi-physics analysis of advanced fast reactor systems, including the Generation IV Gas-Cooled Fast Reactor (GFR). It is shown, from the present analysis, that the built-in TRACE heat transfer and friction correlations are indeed quite suitable for gas cooling under prototypic GFR conditions. (C) 2010 Elsevier Ltd. All rights reserved.

Quantitative functional failure analysis of a thermal–hydraulic passive system by means of bootstrapped Artificial Neural Networks

The estimation of the functional failure probability of a thermal–hydraulic (T–H) passive system can be done by Monte Carlo (MC) sampling of the epistemic uncertainties affecting the system model and the numerical values of its parameters, followed by the computation of the system response by a mechanistic T–H code, for each sample. The computational effort associated to this approach can be prohibitive because a large number of lengthy T–H code simulations must be performed (one for each sample) for accurate quantification of the functional failure probability and the related statistics. In this paper, the computational burden is reduced by replacing the long-running, original T–H code by a fast-running, empirical regression model: in particular, an Artificial Neural Network (ANN) model is considered. It is constructed on the basis of a limited-size set of data representing examples of the input/output nonlinear relationships underlying the original T–H code; once the model is built, it is used for performing, in an acceptable computational time, the numerous system response calculations needed for an accurate failure probability estimation, uncertainty propagation and sensitivity analysis. The empirical approximation of the system response provided by the ANN model introduces an additional source of (model) uncertainty, which needs to be evaluated and accounted for. A bootstrapped ensemble of ANN regression models is here built for quantifying, in terms of confidence intervals, the (model) uncertainties associated with the estimates provided by the ANNs. For demonstration purposes, an application to the functional failure analysis of an emergency passive decay heat removal system in a simple steady-state model of a Gas-cooled Fast Reactor (GFR) is presented. The functional failure probability of the system is estimated together with global Sobol sensitivity indices. The bootstrapped ANN regression model built with low computational time on few (e.g., 100) data examples is shown capable of providing reliable (very near to the true values of the quantities of interest) and robust (the confidence intervals are satisfactorily narrow around the true values of the quantities of interest) point estimates.

High Temperature Oxidation of SiC under Helium with Low Oxygen Partial Pressure

The Gas-cooled Fast Reactor (GFR) is one system of the Generation IV International Forum systems studied by CEA (France). The coolant considered is helium pressurized at 7 MPa and the nominal working temperature is about 1300 K. In case of accident, the reactor temperature can reach 1900 to 2300 K. Several studies are carried out to determine the physico-chemical behavior of structural materials undergoing various environmental constraints. The cladding material currently considered is SiC. Numerical calculations are carried out to determine the transition between passive and active oxidation regimes. Then an experimental study is performed to fix the position of the transition and to study the oxidation resistance of the two polymorphs α- and β-SiC at high temperature (1300 to 2300 K), coupled to a material characterization based on mass variation, SEM, XPS, AFM and roughness analyses.

A comparative study on recycling spent fuels in gas-cooled fast reactors

This study evaluates advanced Gas-cooled Fast Reactor (GFR) fuel cycle scenarios which are based on recycling spent nuclear fuel for the sustainability of nuclear energy. A 600 MWth GFR was used for the fuel cycle analysis, and the equilibrium core was searched with different fuel-to-matrix volume ratios such as 70/30 and 60/40. Two fuel cycle scenarios, i.e., a one-tier case combining a Light Water Reactor (LWR) and a GFR, and a two-tier case using an LWR, a Very High Temperature Reactor (VHTR), and a GFR, were evaluated for mass flow and fuel cycle cost, and the results were compared to those of LWR once-through fuel cycle. The mass flow calculations showed that the natural uranium consumption can be reduced by more than 57% and 27% for the one-tier and two-tier cycles, respectively, when compared to the once-through fuel cycle. The transuranics (TRU) which pose a long-term problem in a high-level waste repository, can be significantly reduced in the multiple recycle operation of these options, resulting in more than 110 and 220 times reduction of TRU inventory to be geologically disposed for the one-tier and two-tier fuel cycles, respectively. The fuel cycle costs were estimated to be 9.4 and 8.6 USD/MWh for the one-tier fuel cycle when the GFR fuel-to-matrix volume ratio was 70/30 and 60/40, respectively. However the fuel cycle cost is reduced to 7.3 and 7.1 USD/MWh for the two-tier fuel cycle, which is even smaller than that of the once-through fuel cycle. In conclusion the GFR can provide alternative fuel cycle options to the once-through and other fast reactor fuel cycle options, by increasing the natural uranium utilization and reducing the fuel cycle cost.

Nuclear fuel recycling: National and regional options for the US nuclear energy system

A system dynamics model is applied to study the impact of spent fuel recycling on the US nuclear energy system and the implications of a bilateral hypothetical collaboration between the US and Brazil. In particular, the impact of different options for advanced fuel cycle facilities needed for the US nuclear energy market alone and under a Brazil partnership is studied. Different recycling technology options are considered: (1) thermal recycling of transuranics (TRUs) in Light Water Reactors (LWRs) using Combined Non-Fertile and UO2 Fuel (CONFU); (2) recycling of TRU in fertile-free metallic fuel in fast Actinide Burner Reactors (ABRs); and (3) fast recycling of TRU with UO2 in self-sustaining Gas-cooled Fast Reactors (GFRs). These recycling options are truly advanced and will require significant development, but they allow the evolution of nuclear energy systems with minimum transuranic elements in the nuclear waste, hence reduce the long term environmental burden of the waste. Case studies for different advanced technology introduction dates are examined under a prescribed rate of growth in demand for nuclear electricity. The timing of introduction of recycling is important for proper technology development. In the model, the rate of deployment is restrained by the industrial capacity to build TRU separation plants, as well as the desire for high utilization factor of the deployed facilities over their life time. It is notable that all recycling options result in only modest reductions in total uranium consumption at the end of the century. Due to its unity fissile conversion ratio, early introduction of the self-sustaining GFR recycling scheme leads to the largest reduction in uranium consumption and enrichment requirements, about 20%. On the other hand, the CONFU recycling scheme keeps the TRU inventory in the entire system well below other schemes, and guarantees equilibrium between the generation and consumption of TRU without investments in fast reactors. The three schemes reduce the TRU sent to the repository for disposal by two orders of magnitude, but the ABR and the GFR schemes require the introduction of the more costly fast reactors. The rate of deployment of fast reactors becomes limited by the availability of TRU from LWRs in the last quarter of the century. Moreover, an assessment of the impact on the US nuclear energy system of a major commitment to supply the nuclear fuel cycle needs of other countries, with Brazil as the example, is made. The partnership with Brazil implies that the US recycling facilities would be used at higher utilization factor, and more TRU will be available to start fast reactors in the US. Economic analysis for the US nuclear energy market indicates that the CONFU technology is more attractive at current uranium price level, and that fast recycling becomes as attractive as thermal recycling at much higher uranium prices.

GFR equilibrium cycle analysis with the EQL3D procedure

Advanced fast reactors of the fourth generation should be capable to breed their own fuel from 238U feed and to recycle the actinides from their own spent fuel. This recycling or virtually the closure of fuel cycle can converge to an equilibrium fuel cycle and has impact on the safety-related parameters. The goals of this study are: (i) to apply an equilibrium cycle procedure EQL3D to the Gas cooled Fast Reactor (GFR), (ii) to simulate and confirm the GFR neutronics capability for closed fuel cycle, and (iii) to evaluate the safety-related parameters of the equilibrium cycle. Equilibrium cycle method for considering the homogeneous recycling of actinides is a known approach. However, in EQL3D the equilibrium method is newly applied for hexagonal-z 3D core geometry and 33 energy-groups neutron-flux calculation. This geometry enables to characterize the equilibrium cycle for complex reloading patterns within a multi-batch cycle. Two GFR geometries were studied, the first based on an international neutronics benchmark with a simple set-up and the second based on more advanced core design. For the advanced design, three reloading patterns within a multi-batch cycle with four different feeds were compared. The GFR neutronics capability for closed cycle was proved. The negative impact of the fuel cycle closure on safety-related parameters was confirmed and quantified. The GFR core with closed fuel cycle could serve after prospective optimization as a sustainable and clean energy source.

Comparison of bootstrapped artificial neural networks and quadratic response surfaces for the estimation of the functional failure probability of a thermal–hydraulic passive system

In this work, bootstrapped artificial neural network (ANN) and quadratic response surface (RS) empirical regression models are used as fast-running surrogates of a thermal–hydraulic (T–H) system code to reduce the computational burden associated with estimation of functional failure probability of a T–H passive system. The ANN and quadratic RS models are built on a few data representative of the input/output nonlinear relationships underlying the T–H code. Once built, these models are used for performing, in reasonable computational time, the numerous system response calculations required for failure probability estimation. A bootstrap of the regression models is implemented for quantifying, in terms of confidence intervals, the uncertainties associated with the estimates provided by ANNs and RSs. The alternative empirical models are compared on a case study of an emergency passive decay heat removal system of a gas-cooled fast reactor (GFR).

Stability of Ti–Zr–N coatings under Xe-ion irradiation

Titanium nitride is a technological important refractory material combining the physical properties of ceramics and metal. In addition to its wide use as protective hard coating for cutting tools or metallization layer in microelectronics, it is one of the inert matrixes proposed to surround the fuel in future gas cooled fast reactor (GFR) systems. The application of low-dose inert gas ions in materials is a powerful method for studying material properties changes due to irradiation damage. In this work, the influence of Xe-ion irradiation on the structure, electrical and mechanical properties of ternary TiN-based coating is investigated. Ti x Zr 1 − x N ( x = 0, 0.25, 0.50, 0.75, 1) thin films (∼ 300 nm thickness) were deposited on (100) Si substrate using unbalanced reactive magnetron co-sputtering. Irradiation with Xe ( q = + 18) ions were carried out at fluence of 8 · 10 14 ions cm − 2 and energy of 360 keV. The morphology, phase and element compositions of the as-grown and irradiated samples were studied using SEM, XRD and RBS. It is found that the simultaneous sputtering of metal targets in Ar/N 2 atmosphere leads to formation of ternary solid solution nitrides with Na–Cl structure with enhanced hardness compared to binary counterpart while the nanohardness decreases for binary nitrides after ion irradiation and increase is observed for ternary nitrides. A detailed analysis of elastic strains and stresses changes after ion irradiation is also performed.

Evaluation of system codes for analyzing naturally circulating gas loop

Steady-state natural circulation data obtained in a 7 m-tall experimental loop with carbon dioxide and nitrogen are presented in this paper. The loop was originally designed to encompass operating range of a prototype gas-cooled fast reactor passive decay heat removal system, but the results and conclusions are applicable to any natural circulation loop operating in regimes having buoyancy and acceleration parameters within the ranges validated in this loop. Natural circulation steady-state data are compared to numerical predictions by two system analysis codes: GAMMA and RELAP5-3D. GAMMA is a computational tool for predicting various transients which can potentially occur in a gas-cooled reactor. The code has a capability of analyzing multi-dimensional multi-component mixtures and includes models for friction, heat transfer, chemical reaction, and multi-component molecular diffusion. Natural circulation data with two gases show that the loop operates in the deteriorated turbulent heat transfer (DTHT) regime which exhibits substantially reduced heat transfer coefficients compared to the forced turbulent flow. The GAMMA code with an original heat transfer package predicted conservative results in terms of peak wall temperature. However, the estimated peak location did not successfully match the data. Even though GAMMA's original heat transfer package included mixed-convection regime, which is a part of the DTHT regime, the results showed that the original heat transfer package could not reproduce the data with sufficient accuracy. After implementing a recently developed correlation and corresponding heat transfer regime map into GAMMA to cover the whole range of the DTHT regime, we obtained better agreement with the data. RELAP5-3D results are discussed in parallel.