Analysis of the turbine deblading in an HTGR with the CATHARE code

The direct coupling of a Gas Cooled Reactor (GCR) with a closed gas-turbine cycle leads to a specific dynamic plant behaviour. This behaviour is described and illustrated through computer analyses performed at CEA with the computer code CATHARE. This analysis requires a 1D code able to simulate the whole reactor, including the core, the vessel, the piping and the components (turbine, compressors, heat exchangers).This paper is devoted to deblading accidents. The problems and solutions encountered in various types of gas-turbines are presented: aero engines, steam turbines (EDF-Porcheville steam turbine accident feedback) and finally the feedback from previous High Temperature Gas Reactor experiments (EVO helium loop, HHT project and other HTR projects) are displayed. From this literature survey, some recommendations are drawn for a future High Temperature Reactor. It is shown that for safety reasons in case of deblading, a horizontal shaft aligned with the reactor vessel is recommended for the turbomachinery.This paper presents simulations of different scenarii performed with CATHARE code:

(a) Turbine deblading with and without reactor trip. In these calculations, a pessimistic assumption has been made: all turbine blades break off.

(b) Total flow blockage. The flow area is entirely blocked by the turbine blades.

(c) Partial flow blockage. Previous conclusions consider the worst cases of deblading and total flow blockage. An intermediate case based on a partial deblading has also been performed.

CATHARE results illustrate that the loss of turbine blades is accompanied by abrupt changes in the Power Conversion System and reactor flow conditions: large axial pressure drop, reverse flow through the core and high rate de-pressurization.     

Feasibility of the electrochemical way in molten fluorides for separating thorium and lanthanides and extracting lanthanides from the solvent

An alternative way of reprocessing nuclear fuel by hydrometallurgy could be using treatment with molten salts, particularly fluoride melts. Moreover, one of the six concepts chosen for GEN IV nuclear reactors (Technology Roadmap - http://gif.inel.gov/roadmap/) is the molten salt reactor (MSR). The originality of the concept is the use of molten salts as liquid fuel and coolant. During the running of the reactor, fission products, particularly lanthanides, accumulate in the melt and have to be eliminated to optimise reactor operation. This study concerns the feasibility of the separation actinides-lanthanides-solvent by selectively electrodepositing the elements to be separated on an inert (Mo, Ta) or a reactive (Ni) cathodic substrate in molten fluoride media. The main results of this work lead to the conclusions that:

•

The solvents to be used for efficient separation must be fluoride media containing lithium as cation.

•

Inert substrates are suitable for actinide/lanthanide separation; nickel substrate is more suitable for the extraction of lanthanides from the solvent, owing to the depolarisation occurring in the cathodic process through alloy formation.

     

LWR fuel rod behavior during severe accidents

Chemical interactions between UO₂ fuel and Zircaloy cladding up to 2350°C are described. UO₂/Zircaloy single effects tests have been performed with short LWR fuel rod segments in inert gas and under oxidizing conditions. The reaction kinetics of molten Zircaloy cladding with solid UO₂ fuel has been investigated with UO₂ crucibles containing molten Zircaloy. The UO₂/Zircaloy reactions obey parabolic rate laws. The oxygen uptake by solid Zircaloy due to chemical interaction with UO₂ occurs nearly as quickly as that from the reaction with steam or oxygen.To study the competing effects of the external and internal cladding oxidation under realistic boundary conditions and the influence of the uncontrolled temperature escalation due to the exothermic steam/Zircaloy reaction on the maximum cladding temperature, single rod and bundle experiments have been performed. Electrically heated fuel rod simulators, including absorber rod material (Ag, In, Cd alloy), guide tubes and grid spacers are used. The maximum measured cladding temperature during the temperature escalation was about 2200°C. The failure temperature of the absorber rods and the extent of bundle damage depends on the guide tube material (Zircaloy or stainless steel) and varies between 1200 and 1350°C. The molten materials and liquid reaction products can relocate and form large coherent lumps on solidification, which may result in complete blockage of the fuel rod bundle cross section. In the future, 7 × 7 bundle experiments of 2 m overall length will be performed in the new CORA facility to study, in addition, the influence of quenching on fuel rod integrity.     

Effects of UO2/Zircaloy-4 Mole Ratios on Reaction of UO2 Fuel with Molten Zircaloy-4

Dissolution of UO₂ crucibles by molten Zircaloy-4 (Zry) was investigated in the temperature range of 2,223-2,373 K and for specimens having UO₂/Zry mole ratios between 7 and 18.2. The uranium concentration in the Zry melt rapidly increased during initial reaction time and approached saturated values, depending on reaction temperature and UO₂/Zry mole ratio. Kinetics of uranium concentration increase in the melt was analyzed based on a natural convection mass transfer model that takes into account the change of contact surface area/melt volume ratio with reaction time. The saturated uranium concentration in the Zry melt was inversely proportional to the U0₂/Zry mole ratio. An empirical correlation of saturated uranium concentration in the Zry melt was obtained as a function of UO₂/Zry mole ratios and reaction temperature. This study of the empirical correlation was intended to estimate maximum UO₂ fuel dissolution by molten Zry cladding during severe fuel damage accidents for three different reactor type fuels.     

Degraded core reflood: Present understanding and impact on LWRs

The severe fuel damage (SFD) research at Forschungszentrum Karlsruhe started with the single rod experiments [Hagen, S., Malauschek, H., Peck, S.O., Wallenfels, K.P., 1983. Temperature Escalation in PWR Fuel Rod Simulator Bundles due to the Zircaloy/steam Reaction Test ESBU-1. Test Results Report, FZK-3508] already before the TMI-2 accident revealed that the knowledge on accident initiation and progression was scarce. The general focus on primary circuit research was the understanding of the relevant processes and their interactions, and of possibilities for countermeasures. However, even today, the most prominent countermeasure, the flooding of a degraded core, is not yet completely understood, namely the influence of adverse effects such as enhanced core degradation and hydrogen spikes under the most probable scenarios.Our work shows that the design basis accident procedures of core reflood can be extended to peak core temperatures of approx. 2200 K, if sufficient reflood mass flow rate can be supplied. Assuming an initial core heat-up rate of 0.5 K/s, this gives only approx. 12 min additional time with respect to design basis accidents (DBA) cases, but moreover a strategy for a successful core reflood is seen to be feasible before a large in-core pool is formed, that might be uncoolable.     

Use of Thorium-Plutonium MOX in the inner pins of CANDU fuel bundles

This research is focused on using Thorium-Plutonium MOX fuel in the inner fuel pins of the CANDU fuel bundles for plutonium incineration and reduction of uranium demand and to reduce coolant void reactivity. The delayed neutron fraction and the power distribution amongst the fuel elements of the fuel bundle have been considered as main safety parameters.The 700 MWe Advanced CANDU Reactor (ACR-700) was selected as a case study. The inner eight UO₂ fuel pins of the ACR-700 fuel bundle are replaced by Thorium-Plutonium MOX fuel pins in the proposed design with 3% reactor grade PuO₂. This amount represents 23.4 w/o of the fuel in the bundle. The outer two fuel rings (35 pins) enrichment is reduced from 2.1 w/o U-235 to 2 w/o U-235. The simulation using MCNP6 showed that about 27% reduction of uranium demand can be achieved. The proposed fuel bundle eliminate the use of burnable poisons in the central pin that was used for negative coolant void reactivity and more reduction in the coolant void reactivity was achieved (about 3.5 mk less than the reference fuel bundle). The power distribution throughout the fuel bundle is more flat in the proposed fuel bundle. Use of this fuel bundle reduces the delayed neutron fraction from 540 pcm in the reference case to 480 pcm in the proposed case.     

Characteristics of corium debris bed generated in large-scale fuel-coolant interaction experiments

The formation of corium debris as the result of fuel-coolant interaction (energetic or not) has been studied experimentally in the FARO and KROTOS facilities operated at JRC-Ispra between 1991 and 1999. Experiments were performed with 3–177 kg of UO₂–ZrO₂ and UO₂–ZrO₂–Zr melts, quenched in water at depth between 1 and 2 m, and pressure between 0.1 and 5.0 MPa. The effect of various parameters such as melt composition, system pressure, water depth and subcooling on the quenching processes, debris characteristics and thermal load on bottom head were investigated, thus, giving a large palette of data for realistic reactor situations.Available data related to debris coolability aspects in particular are:

• Geometrical configuration of the collected debris.

• Partition between loose and agglomerated (“cake”) debris.

• Particle size distribution with and without energetic interaction.

These data are synthesised in the present contribution.     

European Validation of the Integral Code ASTEC (EVITA): First experience in validation and plant sequence calculations

The main objective of the European Validation of the Integral Code ASTEC (EVITA) project is to distribute the severe accident integral code ASTEC to European partners in order to apply the validation strategy issued from the VASA project (4th EC FWP). Partners evaluate the code capability through validation on reference experiments and plant applications accounting for severe accident management measures, and compare results with reference codes.The basis version V0 of ASTEC (Accident Source Term Evaluation Code)—commonly developed and basically validated by GRS and IRSN—was made available in late 2000 for the EVITA partners on their individual platforms. Users’ training was performed by IRSN and GRS. The code portability on different computers was checked to be correct. A “hot line” assistance was installed continuously available for EVITA code users. The actual version V1 has been released to the EVITA partners end of June 2002. It allows to simulate the front-end phase by two new modules:

• for reactor coolant system 2-phase simplified thermal hydraulics (5-equation approach) during both front-end and core degradation phases,

• for core degradation, based on structure and main models of ICARE2 (IRSN) reference mechanistic code for core degradation and on other simplified models.

Next priorities are clearly identified: code consolidation in order to increase the robustness, extension of all plant applications beyond the vessel lower head failure and coupling with fission product modules, and continuous improvements of users’ tools.As EVITA has very successfully made the first step into the intention to provide end-users (like utilities, vendors and licensing authorities) with a well validated European integral code for the simulation of severe accidents in NPPs, the EVITA partners strongly recommend to continue validation, benchmarking and application of ASTEC. This work will continue in Severe Accident Research Network (SARNET) in the 6th Framework Programme where ASTEC plays a key role as the reference European integral code.     

Current perspective on fast reactor containment

The concept of “containment” is to provide a series of physical barriers between the radioactive products of the fission process and the public. All nuclear reactors have several such barriers and LMFBRs have more than most. These barriers are, successively:

1. fuel, which retains fission products;

2. fuel cladding, which encloses the fuel;

3. sodium coolant, which absorbs fission products released through fuel caldding;

4. primary coolant boundary, which has energy absorption and leakage control capabilities;

5. containment building, hereafter referred to as containment, which provides the final engineered barrier for control of radioactive releases;

6. exclusion distance, which provides space for natural attenuation of radioactive releases before reaching the public.

These barriers, along with the design approaches and features which protect their integrity under normal and accident conditions, assure that the public is adequately protected from the potential hazards of radioactivity residing in the core. It is only in the case of hypothesized core disruptive accidents (HCDAs) that these successive barriers can be sufficiently threatened as to pose a significant threat to the public. These HCDAs involve an extremely low probability sequence of successive failures resulting in core cooling imbalances which lead to fuel overheating. Under such conditions, the fuel and cladding barriers can be lost and energy sources can be generated which threaten the primary coolant boundary and containment. This paper addresses current perspectives on containment of HCDAs with emphasis on the approach and programs in the US.     

Determination of Heats of Formation for Solid and Liquid Lithium Metasilicates by Vaporization Method

In-reactor experiments were performed in Nuclear Safety Research Reactor of Japan Atomic Energy Research Institute to study the failure behavior of stainless steel clad fuel rods under a simulated reactivity initiated accident (RIA) condition. A single test fuel rod with stainless steel cladding was contained in a capsule filled with water at room temperature and atmospheric pressure and irradiated by pulsing power simulating an RIA. It was revealed through the experiments that the failure mechanism of the stainless steel clad fuel rod was cladding melting, which was different from oxygen-induced embrittlement observed in the Zircaloy clad fuel rod in the same test condition, and the failure threshold energy was determined to be about 240cal/g·UO₂ (–1,000 kJ/kg·UO₂), which was about 20 cal/g·UO₂ (–85 kJ/kg·UO₂) lower than that of the Zircaloy clad fuel rod. It was also found that the mechanical energy was generated by explosive vaporization of coolant due to molten fuel-coolant interaction as a consequence of the fuel rod failure accompanying fuel pellet fragmentation at an energy deposition of nearly 380 cal/g·UO₂ (–1,600 kJ/kg·UO₂) or more.     

Cladding tube deformation and core emergency cooling in a loss of coolant accident of a pressurized water reactor

The paper summarizes the dominant effects which finally ensure the core coolability of a pressurized water reactor in a loss-of-coolant accident (LOCA).The main results are summarized as follows:

• — The cooling effect of the two-phase mixture which is intensified during reflooding increases temperature differences on the cladding tube circumference and thus limits the mean circumferential burst strains to values of about 50%.

• — An unidirected flow through the fuel rod bundle during the refill and reflooding phases causes maximum cooling channel blockage of about 70%.

• — The coolability of deformed fuel elements can be maintained up to flow blockages of about 90%.

All effects investigated indicate that in a LOCA no impairment of core coolability and public safety has to be expected.     

Qualification of the LF-eddy current technique for the inspection of stainless steel cladding and applications on the reactor pressure vessel

As part of the re-inspection of the reactor pressure vessel of the nuclear power plant, the low-frequency-eddy current technique was implemented during the 1995 outage. Since then, this inspection technique and the testing equipment have seen steady further development. Therefore, optimization of the entire testing system, including qualification based on the 1995 results, was conducted. The eddy current testing system was designed as a ten-channel test system with sensors having separate transmitter and receiver coils. The first qualification of the testing technique and sensors was performed using a single-channel system; a second qualification was then carried out using the new testing electronics. The sensor design allows for a simultaneous detection of surface and subsurface flaws. This assumes that testing is performed simultaneously using four frequencies. Data analysis and evaluation are performed using a digital multi-frequency regression analysis technique The detection limits determined using this technique led to the definition of the following recording limits for testing in which the required signal-to-noise ratio of 6 dB was reliably observed.

• Detection of surface connected longitudinal and transverse flaws:

• notch, 3 mm deep and 10 mm long, for weave bead cladding;

• notch, 2 mm deep and 20 mm long, for strip weld cladding.

• Detection of embedded planar longitudinal and transverse flaws:

• ligament of 7 mm for 8 mm clad thickness and 3 mm;

• ligament for 4 mm clad thickness, notch starting at the carbon steel base material with a length of 20 mm.

• Detection of embedded volumetric longitudinal and transverse flaws:

• 3 mm diameter side-drilled hole (SDH) for 8 mm clad thickness; ligament, 4 mm. For 4 mm clad thickness: diameter, 2 mm SDH; ligament, 2 mm. All SDHs are 55 mm deep.

Molten fuel-coolant interaction during a reactivity initiated accident experiment

The results of a reactivity-initiated accident experiment, designated RIA-ST-4, are discussed and analyzed with regard to molten fuel-coolant interaction (MFCI). In this experiment, extensive amounts of molten UO₂ fuel and zircaloy cladding were produced and fragmented upon mixing with the coolant. Coolant pressurization up to 35 MPa and coolant overheating in excess of 940 K occurred after fuel rod failure. The initial coolant conditions were similar to those in boiling water reactors during a hot startup (that is, coolant pressure of 6.45 MPa, coolant temperature of 538 K, and coolant flow rate of 85 cm³/s). It is concluded that the high coolant pressure recorded in the RIA-ST-4 experiment was caused by an MFCI and was not due to gas release from the test rod at failure, Zr/water reaction, of UO₂ fuel vapor pressure. The high coolant temperature indicated the presence of superheated steam, which may have formed during the expansion of the working fluid back to the initial coolant pressure; yet, the thermal-to-mechanical energy conversion ratio is estimated to be only about 0.3%.     

The 2006 CHF look-up table

CHF look-up tables are used widely for the prediction of the critical heat flux (CHF). The CHF look-up table is basically a normalized data bank for a vertical 8 mm water-cooled tube. The 2006 CHF look-up table is based on a database containing more than 30,000 data points and provides CHF values at 24 pressures, 20 mass fluxes, and 23 qualities, covering the full range of conditions of practical interest. In addition, the 2006 CHF look-up table addresses several concerns with respect to previous CHF look-up tables raised in the literature. The major improvements of the 2006 CHF look-up table are:

• An enhanced quality of the database (improved screening procedures, removal of clearly identified outliers and duplicate data).

• An increased number of data in the database (an addition of 33 recent data sets).

• A significantly improved prediction of CHF in the subcooled region and the limiting quality region.

• An increased number of pressure and mass flux intervals (thus increasing the CHF entries by 20% compared to the 1995 CHF look-up table).

• An improved smoothness of the look-up table (the smoothness was quantified by a smoothness index).

A discussion of the impact of these changes on the prediction accuracy and table smoothness is presented. The 2006 CHF look-up table is characterized by a significant improvement in accuracy and smoothness.     

SEMER: a simple code for the economic evaluation of nuclear and fossil energy-based power production systems

The code system, SEMER, was recently developed to evaluate the economic impact of various nuclear reactors and associated innovations. Models for nearly all fossil energy-based systems were also included to provide a basis for cost comparisons.Essentially, SEMER includes three types of model libraries: the global model, for a rapid estimation of various nuclear and fossil energy-based systems, the detailed models, for the finer cost evaluation of individual components and circuits in a PWR type of reactor and the fuel cycle models, for PWRS, HTRs and FBRs, allowing the cost estimations related to all the steps in the nuclear fuel cycle, including reprocessing and disposal.This paper summarises our on-going investigations on new developments in, and on the validation of, the SEMER system.Details of the modelling principles, and the results of validation carried out in the context of an EDF/CEA Joint Protocol Agreement, are also presented.First results of this validation are highly encouraging:

• Relative errors for the total kWh or overnight and investment costs are less than 5% for large PWR systems operating in France or other countries.

• These errors are less than 3% for small-sized compact PWRs and they are of the order of 4–7% for HTRs (as compared to IAEA estimations).

• For fossil energy-based power plants, the relative error, even with slightly different cost breakdown between SEMER and that of existing installations, is from 5 to 20%.

• Similarly, errors on the nuclear fuel cycle costs are about 1–4%, compared to published reference values.

CEA and AREVA R&D on HTR fuel fabrication and presentation of the CAPRI experimental manufacturing line

In the framework of the French V/HTR fuel development and qualification program, the Commissariat à l’Energie Atomique (CEA) and AREVA are conducting R&D projects covering the mastering of UO₂ coated particle and fuel compact fabrication technology. To fulfill this task, a review of past knowledge, of existing technologies and a preliminary laboratory-scale work program have been conducted with the aim of retrieving the know-how on HTR coated particle and compact manufacture:

• The different stages of UO₂ kernel fabrication GSP process have been reviewed, reproduced and improved.

• The experimental conditions for the chemical vapor deposition of coatings have been defined on dummy kernels and development of innovative characterization methods has been carried out.

• Former CERCA compacting process has been reviewed and updated.

In parallel, an experimental manufacturing line for coated particles, named GAIA, and a compacting line based on former CERCA compacting experience have been designed, constructed and are in operation since early 2005 at CEA Cadarache and CERCA Romans, respectively. These two facilities constitute the CAPRI line (CEA and AREVA PRoduction Integrated line).The major objectives of the CAPRI line are:

• to recover and validate past knowledge,

• to produce representative HTR TRISO fuel meeting industrial standards,

• to permit the optimization of reference fabrication processes for kernels and coatings defined previously at a laboratory-scale and the investigation of alternative and innovative fuel design (UCO kernel, ZrC coating),

• to test alternative compact process options and

• to fabricate and characterize fuel required for irradiation and qualification purpose.

This paper presents the status of progress of R&D conducted on HTR fuel particles and compact manufacture by early 2005 and the potential of the laboratory-scale HTR fuel CAPRI line.