Forschungs- & Entwicklungsinformationsdienst der Gemeinschaft - CORDIS

FP7

ERCOSAM Berichtzusammenfassung

Project ID: 249691
Gefördert unter: FP7-EURATOM-FISSION
Land: Switzerland

Final Report Summary - ERCOSAM (Containment thermal-hydraulics of current and future LWRs for severe accident management)

Executive Summary:
The EURATOM co-financed ERCOSAM project and the ROSATOM co-financed SAMARA project have been conducted in parallel during the period 2010-2014, in the framework of a cooperation agreement, to investigate the containment thermal-hydraulics of current and future LWRs for severe accident management. The projects had two main objectives. The first was to establish, for a severe accident sequence chosen from existing plant calculations and representative of a LOCA in a LWR, the strength of hydrogen stratification that can be established during part of the transient period, starting from the late blowdown until the end of hydrogen release from the reactor vessel into the containment. The second was to determine whether this stratification, once established, can be broken down by the operation of SAM devices: sprays, coolers and heat release by PARs. The test scenarios in the experimental facilities, i.e. TOSQAN (IRSN, France), SPOT (JSC Afrikantov OKB, Russia), MISTRA (CEA, France), PANDA (PSI, Switzerland), and in the HYMIX (IBRAE-RAN, Russia) conceptual facility have been determined based on the scaled scenarios from real nuclear power plants. Analytical activities, for the definition and the analysis of the tests have been conducted using, the lumped parameter codes ASTEC, COCOSYS, TONUS_LP, KUPOL-MT; the CFD-type codes GOTHIC and GASFLOW; the commercial CFD codes FLUENT and CFX; and the open source CFD code, OpenFOAM.
The phenomenology addressed in TOSQAN, MISTRA, SPOT, PANDA tests and in the HYMIX analytical tests include pressurization, gas stratification build-up, stratification break-up, gas mixing, condensation, re-evaporation, de-pressurization under the effect of spray, cooler or heater activation. These phenomena are expected to take place in the containment of LWR/HWR during a postulated severe accident with core degradation and release of hydrogen.
The cooler and spray activation led both to containment de-pressurization. The de-pressurization rate was faster with the spray due to the high spray water mass injection flow rates and the drop-steam direct contact heat/mass transfer. The effect of spray activation on the gas species evolution was overwhelming, whereas the effect of the cooler was limited to the region below the cooler and immediately above the component. Due to the strong mixing effect of the spray, the analyses of these tests resulted less complicated than for the cooler tests, for which the fidelity of the simulation depends on the correct representation of various processes.
The heater tests showed that the convection induced by the heat release has a mixing effect on the region above the heater inlet. The mixing below the heater inlet is controlled by slow processes (diffusion and thermal effects), and affected by the specific geometry of the containment. Moreover, the modeling of thermal radiation, although it was not critical for the prediction of the gas species evolution in the tests addressing only the thermal effects of the PARs, was important for accurately reproducing other variables, and should certainly be considered in full-scope simulations.
The two step approach (pre- and post-test analysis) resulted in a substantial and generally valuable progress in the modelling of phenomena associated with spray, cooler and heater operation, and therefore in the confidence to apply several codes to containment safety analysis.
The ERCOSAM-SAMARA projects have provided an unique contribution toward the understanding of LWR containment phenomenology, under severe accident conditions and in increasing the confidence in the use of LP and CFD codes for the analysis of hydrogen distribution and the effect of severe accident mitigating systems in real nuclear power.

Project Context and Objectives:
The ERCOSAM-SAMARA projects had two main objectives. The first was to establish, for a severe accident sequence chosen from existing plant calculations and representative of a LOCA in a LWR, the strength of hydrogen stratification that can be established during part of the transient period, starting from the late blowdown until the end of hydrogen release from the reactor vessel into the containment. The second was to determine whether this stratification, once established, can be broken down by the operation of SAM devices: sprays, coolers and heat release by PARs.

Based on the collection of existing LOCA calculations for various nuclear power plant (NPPs) designs, the reference scenario which has been selected is a Small Break LOCA in a PWR with dry containment, which generally constitutes the highest contribution to the core damage frequency [1]. The considered NPPs characteristics have been used to define the containment of a generic NPP of about 1000 thermal MW power. The generic containment has been determined by scaling down from PWR 1300 configuration, considering the ratio between the generic and the PWR 1300 containment volumes. Validation of the generic containment, scaled down from real plant, was performed through new calculations of the selected scenarios with the corresponding scaled source terms on the generic containment. Comparisons of the generic calculation results with the reference plants show similar conditions in terms of predicted pressure and temperature and also of global and local steam and hydrogen quantities such as mass and concentration.
The source terms (steam, hydrogen) have been averaged to obtain constant mass and enthalpy flow rates during the steam and hydrogen injection phases.
The approach to define the tests in the various facilities, where hydrogen was replaced by helium, was to follow the simplified accident scenarios with constant steam and helium injection mass flow rates, by considering four distinct and consecutive phases. Phase 1 addressed the blowdown, characterized by steam release from the break during the postulated LOCA and the containment pressurization. Phase 2 simulated the phase of the accident following the blowdown involving core damage leading to the release of hydrogen and steam into the containment. Phases 3 simulated the phases of the accident when no more steam and hydrogen was released from the reactor coolant system into the containment. In Phase 4, a component (spray, cooler or heater) was activated to study its effect on the gas species (helium, steam, air) evolution in the containment.
The test initial and boundary conditions in Phases 1, 2 and 3 were determined, with the target to reach at the end of Phase 3 the total containment pressure and the helium volume fraction in the upper dome similar to those obtained in the generic containment analysis: ~2.5 bar; ~12% helium molar fraction. Due to the difference in the steam/helium release characteristics in the experimental facilities (through injection pipes in a large volume, which result in a given gas composition at the pipe exit) and the releases expected from the steam generator compartment (large opening in the steam generator compartment, which allow a mixing of the release fluid with the compartment gas mixture) of the generic containment, refined calculations applied to the facility geometries and preliminary tests (shake-down tests which aim to develop the appropriate preconditioning protocols) were carried out to identify the appropriate test initial conditions (air/steam mixture temperature, pressure and composition) at the start of Phase 1.
In TOSQAN, three tests have been preformed (T114, T115, T116) addressing the effect of spray activation, and a full cone spray nozzle has been used. In MISTRA tests have been performed which address the effect of activation of a hollow cone spray nozzle (MERCO_1), a cooler (MERCO_2), one heat source (MERCO_3), two heat sources (MERCO_4). In addition a test (MERCO_0) has been performed without SAM activation and with an extended Phase 3, to observe the effect of diffusion and natural circulation on the gas mixture composition evolution. In PANDA, one test has been performed with activation of a hollow cone spray nozzle (PE1), one test with a full cone nozzle (PE2), two tests with a cooler (PE3 and PE5) and one test with one heat source (PE4). In SPOT two tests have been performed with activation of a cooler (S1 and S2). The HYMIX analytical tests addressed the effect of activation of spray (K1) and cooler (K2).
Despite the fact that several aspects of the tests performed in each facility are different, an effort has been made to define the Phase 4 for selected tests with comparable initial and boundary conditions. The remaining tests have been defined to address parameters which provide a broader understanding on the basic phenomenology and in some cases the effect of component designs. For instance the same hollow cone spray nozzle and spray injection flow rate has been used in MERCO_1 and PE1 tests, so that it was possible to investigate the effect on the gas species evolution of spray activation for two types of multi-compartments (PANDA: two large/empty interconnected volumes; MISTRA: one large volume with inner compartment). The PE2 test was performed using a full cone nozzle, and spray water injection conditions as for the PE1 test. Comparing these two PANDA tests allowed for study of the effect of a spray induced flow pattern, which can be more uniformly distributed in the case of the hollow cone nozzle. The T114, T115, T116 tests have been performed with a full cone nozzle and for different initial and boundary conditions and allowed the observation of the phenomenology evolution in a smaller scale facility. The K1 spray analytical test has been defined considering geometrical similarities with the PE2 test, for example, HYMIX configured as an empty containment compartment, and leaving an empty volume between the spray nozzle exit and the top of the containment. Moreover, full cone spray nozzles have been considered in K1 on a ring in the containment periphery. The analysis of spray tests (e.g. MERCO_1, PE2, etc.) allowed for assess to the code and refinement of the models, while the code to code K1 analysis allowed for insight on the code predictions at the HYMIX scale (e.g. intermediate scale toward a LWR real containment).
The Phase 4 for the tests performed with a heater, MERCO_3, MERCO_4 and PE4, have been defined considering heater power curves resembling the energy which could be released by a PAR. The metallic cases (box type geometry) for the heater elements used in MISTRA and PANDA allowed similar flow paths, for example a horizontal opening at the inlet (vertical flow path into the heater) and vertical opening at the exit (horizontal flow path at the exit of the heater). In the PE4 test, the heater element was installed at an intermediate elevation near the interconnecting pipe elevation. The MERCO_3 test was performed with a single heater installed in the periphery of the facility, in the region of the internal compartment exit elevation. The MERCO_4 test was performed with two heater elements installed at the same elevation and radial location as in MERCO_3 but a different circumferential angle. By comparing the three tests, it was possible to observe the induced heat source convective flow for different facility geometries (PE4, MERCO_3) and moreover the flow pattern created by two heat sources (MERCO_4).
The PE3, PE5, MERCO_2, S1, S2 tests and the K2 analytical test have been performed by activating a heat sink in phase 4, for example cooler devices or, in the case of MERCO_2, a condenser. The initial vessel wall and fluid condition for PE5 and S2 tests were defined in such a way as to avoid wall condensation and therefore to observe specifically the effect from the cooler operation. These two tests permitted observation of the effect of cooler design for two different facility scales. The combined cooler and wall condensation effects have been investigated in PE3 test.

The controlled wall condensation has been investigated in MERCO_2 with the middle condenser acting as heat sink. The S1 and S2 tests have been defined with the cooler and the source release in the middle region of the containment as for PE3 and P5 tests. The S1 test has been defined with the same scenarios as PE3. In the S2 test the cooler has been activated during the helium release phase to observe the cooler effect on the helium stratification build-up during the helium release and on the overall mixing (after the helium release phase has been completed).
The K2 analytical test has been defined with the design of the coolers the same as those used in the S1 and S2 tests and considering the volume to cooler-power scaling ratio. Moreover the source (steam, hydrogen) release in K2 analytical test has been inside a room/compartment and the test allows observing hydrogen distribution in a multi-compartment including accumulation in dead end volumes.
All the Organizations participating in the ERCOSAM-SAMARA projects contributed, via code simulations, to the definition and analysis of TOSQAN, SPOT, MISTRA and PANDA experiments and the HYMIX analytical tests. For the simulations, the following lumped parameter codes have been used: ASTEC, COCOSYS, TONUS_LP, KUPOL-MT; the CFD-type codes GOTHIC and GASFLOW; the commercial CFD codes FLUENT and CFX; and the open source CFD code, OpenFOAM. Some tests have been analyzed using the same computational tool by various Organizations. The analytical activities associated with the test definition included planning calculations, and the analyses of the tests included pre-test and post-test calculations.
Moreover, the preliminary calculations (before the actual planning calculations), performed to verify that the gas mixture conditions in PANDA, MISTRA, TOSQAN and SPOT facilities would be representative of the conditions in the generic containment, revealed that due to the facility simplified geometries (which cannot represent the mixing in the break room) a strong steam stratification was to be expected during Phase 1 which would then remain for the following Phases 2 and 3, and this would have caused a major distortion in the mixture stratification in the facilities with respect to the generic containment, for which approximately uniform steam/air mixture was expected over the containment height. The approach to overcome to this distortion was to define for the scenario an additional phase, called Phase 0, which would create a more uniform steam/air mixture composition in the facilities, and therefore be more representative of the generic containment conditions.
Planning calculations have been performed for all phases of the tests to study the effects of various parameters on the gas distributions, and the interaction of the components (cooler, heater, spray) with the stratified ambient conditions. Additionally these studies provided indications on the optimal positions for the key measurements.
The pre-test and post-test analyses have been carried out to assess the code capability to simulate the tests. For the pre-test analysis, the test specifications (nominal values) have been used, while the actual test conditions were used for the post-test analysis.



Project Results:
The facilities used in the project are different in their main dimensions and compartmentalization. The TOSQAN (IRSN, France) facility [2] is a single compartmental vessel with a volume of 7 m3, with a diameter of 1.5 m and height of 4.8 m.
The MISTRA (CEA, France) facility [3] has a multi-compartment geometry and a total volume of 97.6 m3, an inner diameter of 4.25 m and a height of 7.38 m.
PANDA (PSI, Switzerland) is a multi-compartment facility [4], [5] (Figure 4) with a total volume (six vessels) of approximately 515 m3 and an overall height of 25 m and vessels with inner diameter of 4 m.
The SPOT (“JSC Afrikantov OKB”, Russia) facility [6] (Figure 5) is a single compartment vessel with a total volume of 59 m3, internal diameter of 3.2 m and height 9 m.
A conceptual facility [7] named HYMIX (IBRAE-RAN, Russia) has been defined with a volume of 3181 m , inner diameter of 14 m and height 23 m. The HYMIX analytical tests addressed the effect of activation of spray (K1) and cooler (K2).

Results and their significance
The TOSQAN [8], MISTRA [9-12], SPOT [13-15], PANDA [16-20] tests have been carried out according to test protocols developed through preliminary tests (shake-down tests) and considering the findings obtained with the planning calculations. It should be pointed out that the test protocols to reach the target conditions were defined considering the facility characteristics: auxiliary systems and location of injection and venting lines and therefore are different in the various facilities. Moreover, the facility designs such as compartment type, volume/surface ratio, and presence of localized heat sinks, had their own effect on the evolution of test phenomenology. Protocols have been defined also for the HYMIX analytical tests [21].
The global phenomena taking place in Phases 1 (release of steam), 2 (release of steam and helium) and 3 (stabilization) are: containment pressurization (TOSQAN: 2.2- 2.8 bar [22]; MISTRA: 2.56 bar [23-27]; SPOT: 2.5 bar [28]; PANDA: 2.5 bar [29-33]), and gas mixture stratification build-up (helium in the upper layer of the vessel, e.g. TOSQAN: up to 100%; MISTRA: 9.5%; SPOT: 12%; PANDA: 12-14%). Moreover, thermal stratification was observed in the vessel wall and gas space. The thermal stratification was related to the injection conditions, (such as injection exit elevation), to the facility characteristics (presence of concentrated heat sinks, inter-compartment flow transport) and to the condensation and re-evaporation phenomena, which were taking place in some of the tests due to the specified initial and boundary conditions.
The global phenomena which took place in Phase 4 (activation of spray, cooler or heater), were containment depressurization (cooler and spray), as well as gas atmosphere mixing (cooler, spray, heat source).
For instance, with respect to the tests performed with spray activation, the PE1 (with hollow cone nozzle) and PE2 (with full cone nozzle) tests had similar depressurization rates; the containment pressure decreased from 2.5 bar to 1.5 bar in about 2000 sec. Nonetheless in the PE1 test, the phenomenology associated with gas species evolution was more complex, due to the initial higher gas/wall thermal stratification in the vessel. This further affected the condensation and re-evaporation phenomena as well as the inter-compartment flow transport. As result, in the PE1 test the helium-rich layer was broken more quickly than in PE2. In MERCO-1 test (hollow cone), the pressure decreased from 2.56 bar to 1.5 bar in about 1500 sec and in a shorter time the helium-rich layer was broken. In the T115 test the pressure decreases to 1.7 bar in about 5000 sec. In the T114 and T116 tests the re-evaporation of sump prevailed on the condensation of steam on the injected spray droplets and as one consequence the facility pressure increased.
In the tests performed with activation of a cooler, the containment pressure decreased more slowly than in the spray tests. In the PE3 test the pressure decreased from 2.5 bar to 1.47 bar in 7710 sec and in PE5 from 2.5 bar to 1.35 bar in 7262 sec. In MERCO_2, the pressure decreased from 2.56 bar to 2.2 bar in 2500 sec. In the S1 test the pressure decreased from 2.5 bar to 1.2 bar in 9380 sec and in the S2 test from 2.5 bar to 1.2 bar in 9304 sec. The evolution of gas species, in particular the cooler’s capability to destabilize a stratified hydrogen atmosphere, is related to the release conditions (radial and vertical release locations) for the given scenarios, the cooler design (exposed cooler tubes, or tubes inside a case with closed faces), the cooler position within the containment and with respect to the release location. The cooler used in PE3 and PE5 tests was installed in the middle region of the PANDA containment, and the cooler design included cooling pipes inside a frame with open sides (leaving a possibility for the helium/steam/air mixture to leave the cooler flowing through the cooler pipes). The convection induced by the cooler was confined within 1 m above the cooler and therefore the helium rich-layer was only partially eroded.
In the MERCO-2 test, with activation of the condenser installed in the medium elevation of the MISTRA containment, a stratified atmosphere consisting of two main zones persisted until the end of the test. A helium-rich layer remained in the free volume above the internal MISTRA compartment. Below the helium-rich layer, in the annular space between the condenser and the compartment and above the annular ring, remained a stable air-rich mixture.
In the S1 and S2 tests, the convection induced by the cooler activation also led to a partial erosion of the helium-rich layer located in the upper region of the facility. Moreover the activation of a cooler in the S2 test during the helium-release phase did not produce qualitative differences in the helium-rich layer break-up.
In the tests performed with activation of a heater, the energy released to the gas had a minor effect on the gas pressure evolution. In the PE4, MERCO_3 and MERCO_4 tests the heat loss from the facility walls prevailed over the heat source by the heater/s and the pressure slightly decreased during the heater activation. The convection induced by the heater in PE4 led to complete homogenization of the gas mixture located above the heat source inlet elevation. A gas mixture with a lower helium content slightly diffused downward and flowed through the top of the IP towards Vessel 2. Nonetheless the heater effect on gas mixing in the region below the heater was rather limited. In MERCO-3 and MERCO-4, with the heater located in the upper compartment region, the helium homogenization extended to the elevation of the ring plate, and the facility compartments allowed the creation of flow patterns leading to the transport of helium below the heater elevation.
The analysis of the TOSQAN, MISTRA, SPOT and PANDA tests with the advanced computational tools allowed insight on the actual code capabilities and on the modeling needs to simulate the tests. The two step (pre- [34-43] and post- [44-57], test) approach, led in general to improved simulation results in the post test analysis, with respect to the pre-test analysis. Indeed, the major discrepancies of the code simulations from the test results, identified with the pre-test analysis, were eliminated by using a more complete set of test initial and boundary conditions (actual test conditions instead of nominal values). Therefore the pre-test analyses cannot be considered as representative of the code capabilities to perform blind simulations. They have been very useful, however, because they contributed to a closer look at all the available experimental data and revealed some technical issues, which could be properly addressed in the post-test analysis.
With respect to the Phases 1, 2 and 3 of the experiments with wall condensation, the codes in general predicted the pressurization rate well in the post-test analysis. The substantial improvement with respect to the pre-test analyses was mainly due to the correct calculation of the condensation rate and the use of the correct steam/helium sources. In the post-test analyses the detailed initial distributions of gas concentration and wall temperature , the actual heat losses and nominal values for the steam and helium injection rate could be used, and the correct setting of initial and boundary conditions was crucial for the success of the simulations. In particular, the wall temperature and the initial gas concentrations had a major effect on the pressurization rate. A difference of few degrees (~2-3oC) in wall temperature and a few percent in molar fractions had a major effect on the containment pressurization rate.
Various sensitivity studies, such as: improving mesh resolution near the wall, modifications in the modelling of wall heat transfer, refined modeling of wall condensation (including spurious effects) also contributed to improve the simulation results.
For tests without wall condensation, wall heat transfer played a dominant role, and the effect of radiative heat transfer, although debated, in some simulations has been shown to be an important one. This effect could be especially large in regions where the fluid is nearly stagnant, such as the region across the density interface during steam/helium injection. Indeed, most simulations in this region predicted too high of temperatures due to compression heating effects, not compensated by adequate wall transfer.
The overall gas species distribution in Phases 1, 2 and 3 is predicted reasonably well over the facility height, with the exceptions of the peak helium concentration in PANDA, which could not be accurately predicted by nearly all codes, and the helium distribution in the region of the steam/helium exit elevation, where some simulations predicted too large mixing, Due to the fact that in some of the tests, (PANDA/SPOT cooler test, PANDA heater test) the release exit elevation is in the same region of the heater and cooler, the discrepancy in predicting the gas species composition in that region had a major influence in predicting Phase 4.
Due to the complexity of the phenomenology, among the effects of the three type of component investigated (cooler, spray, heater), the cooler operation resulted to be the most difficult to be predicted with the computational tools. Although, in accordance with the scoping calculations (confirmed by the experimental results) stratification was eroded very slowly in most calculations, many details of the phenomena observed in the tests could not be reproduced. In particular, some of the CFD simulations for some tests predicted full mixing of helium, while in the experiments a stratified atmosphere persisted until the end of the cooler operation. Moreover, the rate of the erosion of stratification was over-estimated in other tests, and this prevented the correct prediction of the evolution of the helium concentration above the cooler. In general, the cooler activation does not have an overwhelming effect on the evolution of containment phenomenology, and therefore it is reasonable to presume that the overall accident scenario depends on complex interaction between several variables and, to a certain extent, on the conditions before the cooler is activated (history effect). Parametric studies for some tests revealed that actually this dependence could be stronger than expected, and therefore the code simulations for phase 4 could strongly depend on the correct simulation of the transient conditions before the cooler activation. This sensitivity to the correct initial conditions may pose a severe challenge to the codes when these are used to assess the cooler effect on the hydrogen distribution in a real containment.
Contrarily to the cooler, the spray activation has an overwhelming influence on the phenomena evolution. Therefore the correct prediction of the spray activation is less dependent on the containment conditions in the early phase of the accident scenarios (Phases 1 to 3).
In general, the de-pressurization rate in most of the transient stage was controlled by the steam condensed on the droplets, which is much larger that the vaporization from droplets, sump or liquid film on the wall. Since the heat/mass transfer models used by the codes are well established the initial depressurization was well captured by all codes. In the final period of the spray operation, however, pressure reduction due to condensation on droplets is partly compensated by the re-evaporation from the sump and heat transfer from the still hot walls. Under these conditions, the modelling of wall heat transfer and wall-to-liquid heat transfer on the sump becomes crucial for an accurate estimate of the final pressure. Therefore, in the post-test analysis, a large effort was dedicated to the proper representation of these two processes. In particular, in some simulations, enhanced turbulence due to the spray operation was assumed to justify large heat transfer rates and empiric solutions were proposed to represent sump water re-evaporation. Two-phase turbulence effects and radiative heat transfer to droplets, as well as mechanistic modelling of the sump could not be properly addressed. Nevertheless, this activity resulted in a substantial and generally valuable progress in the modelling of phenomena associated with spray operation, and therefore in the confidence to apply several codes to containment safety analysis.
The gas species distribution below the spray nozzle elevation is generally well predicted, despite there some specific effects in the transient state which are not well captured. For the space above the spray nozzle, the simulations are generally much less accurate, with the prediction depending on a variety of modelling features and detail of the mesh. Although complete parametric studies could not be performed, for nearly all simulations the spatial distribution of the spray water injection (in simulations with CFD codes the spray angle) resulted to be a key parameter in the timing of the complete mixing.
For the heater tests (PANDA/MISTRA), nearly all simulations showed a good capability to reproduce the rather fast mixing above the heater inlet, and the very slow downward progression of the stratification front below this level. In this respect it is interesting to note that the heater effect is rather weak, and the quality of the simulations should be rated comparing them with the calculation of a transient without the component. This was only possible using the data in the MISTRA facility, where the “unofficial” test MERCO-0 (without heater) served as reference for observing the effect of the heater on the transient. For this test, the few simulations addressing this issue were reasonably successful in predicting the genuine mixing effect of the heat source(s). Another important finding (which confirmed results from previous projects) was that it is important to model the heat radiation from the heater to the wall structure and, for the case of MISTRA tests, flow transport behind the condenser driven by the heat loss sink, which played an important role in the helium distribution below the heater inlet.
The two benchmarks based on the HYMIX K1 and K2 analytical tests allowed code-to-code comparison at large scale (one order of magnitude larger than the facilities) and in a “blind-type” mode: not influenced by the knowledge of the experimental results. Since both LP and CFD codes were used for the analysis of K1 and K2, it has been possible to identify general differences on these two types of codes. For instance, in the calculation of the containment pressurization, the CFD codes tend to provide higher pressures because of higher calculated gas temperature due to gas compression and lower wall condensation. With respect to hydrogen stratification build up, the LP codes predicted a more diffuse hydrogen distribution and therefore lower hydrogen concentration in the upper dome. The hydrogen concentration break-up by spray activation resulted to be much faster with the CFD and GOTHIC codes than that predicted using the LP codes.
Despite the fact that the cooler model was calibrated/verified against the SPOT data, the simulation of K2 (HYMIX, cooler activation) led to poor agreement between the code simulations, with respect to both hydrogen stratification break-up and hydrogen inter-compartment transport in dead-end volumes, where formation of flammable mixtures was predicted by several codes. This disagreement shows that the validation of the codes using smaller scale and simplified geometries is not sufficient to demonstrate the applicability of the codes to the real containment, because scalability of modelling choices and mesh remains to be investigated.

Conclusions and recommendations
The experimental and analytical investigations carried out within the EURATOM-ROSATOM ERCOSAM-SAMARA projects have contributed to the advancement of knowledge on issues that have relevance for reactor containment safety analysis.
The phenomenology addressed in TOSQAN, MISTRA, SPOT, PANDA tests and in the HYMIX analytical tests include pressurization, gas stratification build-up, stratification break-up, gas mixing, condensation, re-evaporation, de-pressurization under the effect of spray, cooler or heater activation. These phenomena are expected to take place in the containment of LWR/HWR during a postulated severe accident with core degradation and release of hydrogen.
The cooler and spray activation led both to containment de-pressurization. The de-pressurization rate was faster with the spray due to the high spray water mass injection flow rates and the drop-steam direct contact heat/mass transfer. The effect of spray activation on the gas species evolution was overwhelming, whereas the effect of the cooler was limited to the region below the cooler and immediately above the component. Due to the strong mixing effect of the spray, the analyses of these tests resulted less complicated than for the cooler tests, for which the fidelity of the simulation depends on the correct representation of various processes.
The heater tests, showed that the convection induced by the heat release has a mixing effect on the region above the heater inlet. The mixing below the heater inlet is controlled by slow processes (diffusion and thermal effects), and affected by the specific geometry of the containment. Moreover, the modeling of thermal radiation, although it was not critical for the prediction of the gas species evolution in the tests addressing only the thermal effects of the PARs, was important for accurately reproducing other variables, and should certainly be considered in full-scope simulations.
The experimental database obtained with the projects has already been used by the project Organizations to assess various computational tools and to advance modeling capabilities and simulation approaches. The two step approach (pre- and post-test analysis) resulted in a substantial and generally valuable progress in the modelling of phenomena associated with spray, cooler and heater operation, and therefore in the confidence to apply several codes to containment safety analysis.
A detailed description of the experiment phenomenology and of the test analysis has been reported by the project Organizations in the various project reports. Some of the project findings have been already published in various conferences and a special session devoted to ERCOSAM-SAMARA is planned in ICAPP 2015.
As follow-up of the ERCOSAM-SAMARA activities, the following general recommendations are outlined.
• The ERCOSAM-SAMARA approach to scale down from a generic scenario to experiments in various facilities and use the models validated with the facility data to analyze scenarios at large scale (HYMIX type) should be followed also in future projects intended to assess and validate advanced computational tools for nuclear safety analysis.
• The ERCOSAM-SAMARA experimental database should be further exploited and the project findings should be disseminated through scientific publications in conference proceedings and scientific journals
• The investigations on hydrogen risk should be continued, addressing more complex configurations and accident scenarios, more prototypical flow obstructions and compartments as well as different SAM component designs.
• The analytical activities associated with the tests should also aim for the improvement of numerical and physical models, especially for representing condensation/re-evaporation, turbulence, convective wall heat transfer in nearly stagnant regions, radiative heat transfer and the effects of spray on turbulence.
• Analytical tests (of type HYMIX) for code-to-code comparisons have to be carried out as blind type exercises for assessing the scalability of code simulations to scale approaching real LWR containments.

References
(ERCOSAM-SAMARA reports)

[1] S. Benteboula, A. Bentaib, J. Malet, A. Bleyer, “Scaling down analysis real NPP calculations to experimental facilities”. IRSN, D1.1 (2011-09), 2011.
[2] E. Porcheron, P. Lemaitre, A. Nuboer, “TOSQAN facility description”. P3.2 (2011-04), 2011.
[3] J. Brinster, I. Tkatschenko, J. L. Widloecher, “MISTRA facility description”. CEA, P3.3 (2011-03), 2011.
[4] N. Erkan, D. Paladino, G. Mignot, R. Zboray, R. Kapulla, M. Fehlmann, C. Wellauer, W. Bissels, “PANDA facility description”, PSI, P3.1 (2011-02), 2011.
[5] G. Mignot, R. Kapulla, D. Paladino, S. Paranjape, R. Zboray, M. Fehlmann,
W. Bissels, “Addenda to the facility description report of PANDA facility for ERCOSAM/SAMARA project’, P3.1B (2012-14/1), 2012
[6] M. Kamnev, A. Khizbullin, “SPOT facility description”, JSC “Afrikantov OKBM”, D5.11 (2012-09), 2012.
[7] A. Zaytsev, T. Popova, A. Philippov, A. Sorokin, T. Yudina, “HYMIX concept facility description”. IBRAE RAN, D5.5 (2013-21), 2013.
[8] C.Ledier, P. Lemaitre, A. Nuboer, E. Porcheron, “TOSQAN Test Protocols for 114 – 115 – 116 Tests”. IRSN, P3.5 (2012-07), 2012.
[9] D. Abdo, J. Brinster, F. Dabbenne, J.L. Windloecher, ”Test Protocol of MISTRA MERCO-0 Test, Phases I to III without SAM Activation”. CEA, P3.4_MERCO_0 (2012-6a), 2012
[10] D. Abdo, J. Brinster, F. Dabbene, J.L. Widloecher, “Test Protocol of MISTRA MERCO-1 Test: hollow cone spray test”. CEA, P3.4_MERCO_1 (2012-6b), 2012
[11] D. Abdo, J. Brinster, F. Dabbene, J.L. Widloecher, “Test Protocol of MISTRA MERCO-2 Test: middle condenser cooling test”. CEA, P3.4_MERCO_2 (2012-6c), 2012.
[12] D. Abdo, J. Brinster, F. Dabbene, O. Norvez, J.L. Windloecher, “Test Protocol of MISTRA MERCO_3 and MERCO_4 tests: Heater tests”. CEA, P3.4_MERCO_3&4 (2012-6c), 2012.
[13] M. Kamnev, A. Khizbullin, “SPOT test protocols”, JSC “Afrikantov OKBM”, D5.12 (2013-15), 2013.
[14] M. Kamnev, O. Tyurikov, “Test protocol for Test S1 (draft version)”. P3.13a (2013-01), JSC “Afrikantov OKBM”, P3.13a (2013-01), 2013.
[15] M. Kamnev, O. Tyurikov, “Test protocol for Test S2 (draft version)”. JSC “Afrikantov OKBM”, P3.13b (2013-02), 2013.
[16] G. Mignot, R. Kapulla, R. Zboray, M Fehlmann, C. Wellauer, W. Bissels, D. Paladino “Test protocol for PANDA Test PE1, hollow cone spray test“. PSI, P3.6A (2012-02), 2012.
[17] G. Mignot, R. Kapulla, R. Zboray, M Fehlmann, C. Wellauer, W. Bissels, D. Paladino “Test protocol for PANDA Test PE2, full cone spray test“. PSI, P3.6B (2012-10), 2012.
[18] G. Mignot, R. Kapulla, R. Zboray, M Fehlmann, C. Wellauer, W. Bissels, D. Paladino, “Test protocol for PANDA Test PE3, cooler test”. PSI, P3.6C (2012-11), 2012.
[19] G. Mignot, R. Kapulla, R. Zboray, M Fehlmann, C. Wellauer, W. Bissels, D. Paladino “Test protocol for PANDA Test PE4, heater test”. PSI, P3.6D (2012-01), 2012.
[20] G. Mignot, R. Kapulla, S. Paranjape, R. Zboray, M. Fehlmann, W. Bissels, D. Paladino, “Test protocol for test PE5, cooler test”. PSI, P3.6E (2012-16), 2012.
[21] A. Zaytsev, A. Lukyanov, T. Popova, “HYMIX benchmarking test protocol”. SSC RF-IPPE, D5.6 (2012-18), 2012.
[22] C. Ledier, P. Lemaitre, A. Nuboer, E. Porcheron, “TOSQAN test report, ERCOSAM tests 114, -115, -116”. IRSN P3.8 (2013-09), 2013.
[23] D. Abdo, J. Brinster, R. Tomassian, J. D. Wildloecher, “Quick Look Report for MISTRA MERCO_0-Steam (01/12/2011) and MERCO_0 (07/02/2012) and 13/4/2012) Tests”. CEA, P3.9A (2012-12), 2012.
[24] D. Abdo, J. Brinster, O. Norvez, J. D. Wildloecher, “Quick Look Report for MISTRA MERCO_1 Tests”. CEA, P3.9B (2012-12b), 2012.
[25] D. Abdo, J. Brinster, F. Dabbene, O. Norvez, J. D. Wildloecher, “Quick Look Report for MISTRA MERCO_2 Tests”. CEA, P3.9C (2012-12c), 2012.
[26] D. Abdo, J. Brinster, F. Dabbene, O. Norvez, JL. Widloecher, ”Quick-look report
for MISTRA MERCO_3 and MERCO_4 tests”. P3.9D (2013-18), 2013.
[27] D. Abdo, J. Brinster, F. Dabbene, O. Norvez, JL. Widloecher, “ Characterization
of Heater devices representatives, of a PAR before implementation inside the MISTRA facility”. CEA (2013-17), 2013
[28] M. Kamnev, “SPOT data analysis report”. JSC “Afrikantov OKBM”. P3.14 (2014-01), 2014.
[29] G. Mignot, R. Kapulla, S. Paranjape, R. Zboray, M. Fehlmann, W. Bissels, D. Paladino, “Test Report for Test PE1 – Hollow cone spray test”. PSI, P3.7A (2012-17), 2012.
[30] S. Paranjape, G. Mignot, R. Kapulla, R. Zboray, M. Fehlmann, W. Bissels, D. Paladino, “Test report for test PE2 full cone spray test”. PSI, P3.7B (2013-03), 2013.
[31] G. Mignot, S. Paranjape, R. Kapulla, R. Zboray, M. Fehlmann, W. Bissels, D. Paladino, “Test report for test PE3 cooler test with wall condensation”. PSI, P3.7C (2013-04), 2013.
[32] G. Mignot, R. Kapulla, D. Paladino, S. Paranjape, R. Zboray, M. Fehlmann, W. Bissels, “Test Report For Test PE4 Heat Source Test”. PSI, P3.7D (2012-13), 2012.
[33] R. Zboray, S. Paranjape, G. Mignot, R. Kapulla, M. Fehlmann, W. Bissels, D. Paladino, “Test report for test PE5 cooler test with wall condensation”. PSI, P3.7E (2013-05), 2013.
[34] N. Drobyshevsky, A. filippov, S. Grigoreyev, O. Tarasov, T. Yudina, “IBRAE RAN
Pre-test analysis report”. IBRAE RAN, P2.13 (2013-06) 2013.
[35] D. C. Visser, N. B. Sicarna, E. M. J. Komen, “ERCOSAM-WP2 pre-test report”. NRG, P2.5 (2013-07), 2013
[36] S. Benz, P. Royl, Z. Xu, T. Jordan, “Pre- and planning test calculations with GASFLOW”. KIT, P2.5 (2013-08), 2013.
[37] S. Kelm, W. Jahn, M. Klauck, L. Gotz, R. Gehr, „JUELICH Pre-Test Analysis
Report“. JUELICH, P2.17 (2013-10), 2013.
[38] R. Liang, ”AECL Planning and Pre-Test Analysis Report for the ERCOSAM-SAMARA projects”. AECL, P2.6 (2013-11), 2011.
[39] M. Andreani, “Planning and pre-test calculations for the ERCOSAM tests”. PSI, P2.1 (2013-12), 2013.
[40] F. Dabbene, “MERCO_0, MERCO_3 and MERCO_4 pre-test simulations”. CEA P2.2 (2013-13), 2013.
[41] M. Kamnev, O. Tyurikov, “JSC “Afrikantov OKBM” planning and pre-test analysis report”. JSC “Afrikantov OKBM”, P2.19 (2013-14), 2013.
[42] A. Zaytsev, A. Lukyanov, T. Popova, “Pre-test analysis report”. SSC RF-IPPE P2.14 (2013-19), 2013.
[43] T. Yudina, “IBRAE Report “HYMIX benchmarking tests”. IBRAE RAN, P4.1a (2013-16), 2013.
[44] R. Liang, “AECL Post-Test Analysis Report for the ERCOSAM-SAMARA Projects – Part I”. P2.12 (2014-02), 2014.
[45] A. Filippov, V. Tomashchik, T. Yudina, “HYMIX benchmarking tests: code-to-code comparison”. IBRAE RAN, P4.1 (2014-14), 2014.
[46] C. Boyd, “ERCOSAM WP2 post-test report”. US NRC P2.22 (2014-04), 2014.
[47] M. Andreani, M. Sharabi, “Post-test calculations for the ERCOSAM tests”. PSI P2.7 (2014-05), 2014.
[48] A.Bleyer, “ERCOSAM project: TOSQAN, PANDA and HYMIX post-tests analysis using ASTEC/CPA code”. IRSN, P2.8A (2014-06), 2014.
[49] J. Malet, “ERCOSAM project: CFD calculations of stratification and spray tests”. IRSN, P2.8B, 2014
[50] F. Dabbene, “MERCO_0, MERCO_3 and MERCO_4 post-test simulations”. CEA, P2.9 (2014-07), 2014.
[51] Z. Xu, “PANDA spray post-test simulations”. KIT, P2.10 (2014-08), 2014
[52] D. Visser, “ERCOSAM WP2 post-test report NRG”. NRG, P2.11 (2014-09), 2014.
[53] A.S. Filippov, S.Y. Grigoryev, O.V. Tarasov, T. A. Judina, “IBRAE RAN post-test Analysis report”. IBRAE RAN, P2.15 (2014-10), 2014.
[54] S. Kelm, M. Klauck, L. Goetz, R. Gehr, W. Jahn, “JUELICH Post-test analysis
report“. JUELICH, P2.18 (2014-12), 2014.
[55] M. Kamnev, O. Tyurikov, “ JSC “Afrikantov OKBM” post-test analysis report, Part I”. JSC “Afrikantov OKBM”, P2.20a (2014-13), 2014
[56] M. Kamnev, O. Tyurikov, A. Khizbullin, “JSC “Afrikantov OKBM” post-test analysis report, Part II”. “ JSC “Afrikantov OKBM, P2.20b (2014-13), 2014
[57] M. Kamnev, O. Tyurikov, A. Khizbullin, “SPOT code to data comparison”. “JSC “Afrikantov OKBM, P4.2 (2014-14), 2014
[58] J. Malet et al., “ERCOSAM WP4 Synthesis of tests and calculations activities”. IRSN, D4.1 (2014-29), 2014.


Potential Impact:
The ERCOSAM-SAMARA projects allowed a step forward in understanding, LWR containment phenomenology under Severe Accident conditions and in assessment and validation various advanced computational tools.
Most of the computational tools used in the project are in general used also for the analysis of real nuclear power plants and therefore the knowledge gained within the projects has the potential to be transferred to the analysis of real nuclear power plants.
A detailed description of the experiment phenomenology and of the test analysis has been reported by the project Organizations in the various project reports.

As dissemination activities, project findings have been already published in various conferences:

•D. Paladino, S. Guentay, M. Andreani, I. Tkatschenko, J. Brinster, F. Dabbene, S. Kelm, H.-J. Allelein, D.C. Visser, S. Benz, T. Jordan, Z. Liang, E. Porcheron, J. Malet, A. Bentaib, A. Kiselev, T.Yudina, A. Filippov, A. Khizbullin, M. Kamnev, A. Zaytsev0, A. Loukianov, “The EURATOM-ROSATOM ERCOSAM-SAMARA Projects on Containment Thermal-hydraulics of current and future LWRs for Severe Accident Management”. Proceedings of ICAPP ’12, Chicago, USA, June 24-28, 2012
•Z. Liang, M. Andreani, “Numerical Study on Interaction of Local Air Cooler with Stratified”. Proceedings of ICAPP ’12, Chicago, USA, June 24-28, 2012
•Z. Liang, M. Andreani, “Planning Calculations of Spray Tests for the ERCOSAM-SAMARA Project”. 24th Nuclear Simulation Symposium, Ottawa, Ontario, Canada, 2012 October 14-16.
•S. Paranjape, R. Kapulla, G. Mignot, D. Paladino, R. Zboray, “Light Water Reactor Hollow Cone Containment Spray Performance Tests In The Presence of Light Non-condensable gas”. The 15th International Topical Meeting on Nuclear Reactor Thermal Hydraulics (NURETH-15), May 12-16, 2013 Pisa (Italy)
•S. Paranjape, R. Kapulla, G. Mignot, D. Paladino, “Effect of thermal stratification of thermal stratification on full cone spray performance in reactor containment for a scaled scenarios”. ICONE 22, July 7-11, 2014, Prague, Czech Republic.
•A.S. Filippov, S.Y. Grigoryev, O.V. Tarasov, T.A. Yudina, “CFD Simulation Of PANDA and MISTRA Cooler Tests Of ERCOSAM-SAMARA Project”. ICONE 22, July 7-11, 2014, Prague, Czech Republic.
•A. Filippov, S.Grigoryev, N. Drobyshevsky, A. Shyukin, T. Iudina, “CMFD simulation of ERCOSAM PANDA spray tests PE1 and PE2”. CFD4NRS-5, Zurich 9-11 September 2014
•Grigoryev S.Yu., Filippov A.S., Shyukin A.A., Development and validation of condensation model for CFD investigations of NPP containment safety during a severe accident. Bull. Russ. Acad. Sci. Energetics, 4, P. 123-141, 2014 (in Russian)

A special session devoted to ERCOSAM-SAMARA is planned in ICAPP 2015:

•D. Paladino, A. Kiselev, “Main outcomes from the EUROATOM-ROSATOM ERCOSAM SAMARA parallel projects for hydrogen safety of LWR”. Abs in: International Congress on Advances in Nuclear Power Plants (ICAPP 2015) May 3-6, Nice, France.
•A.S. Filippov, S.Y. Grigoryev, T.A. Yudina, A.E. Kisselev, O.V. Tarasov, “Complete CFD analysis of ERCOSAM-SAMARA exercises: a step toward advancing modeling of LWR containment under severe accident conditions”. Abs in: International Congress on Advances in Nuclear Power Plants (ICAPP 2015) May 3-6, Nice, France.
•S. Benteboula, J. Malet, A. Bleyer, A. Bentaib, D. Paladino, S. Guentay, M. Andreani, I. Tkatschenko, J. Brister, F. Dabbene, S. Kelm, H.-J. Allelein, D.C. Visser, S. Benz, T. Jordan, Z. Liang, A. Kiselev, T. Yudina, A. Filippov, A. Khizbullin, M., Kamnev, A. Zaytsev, A. Loukianov, “EUROATOM-ROSATOM-ERCOSAM-SAMARA PROJECTS Scaling from Nuclear Power Plant to experiments”. Abs in: International Congress on Advances in Nuclear Power Plants (ICAPP 2015) May 3-6, Nice, France.
•M. Andreani, M. Sharabi, A. Bleyer, J. Malet, F. Dabbene, D. Visser, S. Benz, Z. Xu, Z. Liang, S. Kelm, A. Filippov, A. Zaytsev, M. Kamnev, O. Tyurikov, C. Boyd, “Modelling of stratification and mixing of a gas mixture under the conditions of a severe accident with intervention of mitigation measures”. Abs in: International Congress on Advances in Nuclear Power Plants (ICAPP 2015) May 3-6, Nice, France.
•F. Dabbene, J. Brinster, E. Porcheron, G. Mignot, S. Paranjape, R. Kapulla, et al. “Experiemtnal activities on stratification and mixing of a gas mixture under the conditions of a severe accident with intervention of mitigation measures performed in the ERCOSAM-SAMARA project”. Abs in: International Congress on Advances in Nuclear Power Plants (ICAPP 2015) May 3-6, Nice, France.
•J. Malet, E. Porcheron, A. Bentaib, M. Andreani, G. Mignot, D. Paladino, S. Guentay, F. Dabbene, J. Brinster, D. Visser, Z. Xu, Z. Liang, S. Kelm, A. Kiselev, T. Yudina, A. Filippov, A. Zaytsev, M. Kamnev, C. Boyd, “Analysis of stratification and mixing of a gas mixture under severe accident conditions with intervention of mitigation measures”. Abs in: International Congress on Advances in Nuclear Power Plants (ICAPP 2015) May 3-6, Nice, France.
•M. Kamnev, O.Tyurikov, A. Khizbullin, A. Kiselev, T.Yudina, D. Paladino, M. Andreani, “Overview of SPOT experimental and analytical activities with KUPOL”. Abs in: International Congress on Advances in Nuclear Power Plants (ICAPP 2015) May 3-6, Nice, France.
•T.Yudina, A. Filippov, A. Kiselev, A. Bleyer, A. Bentaib, J. Malet, M. Andreani, D. Paladino, S. Guentay, R. Gehr, S. Kelm, H.-J. Allelein, S. Benz, T. Jordan, Z. Liang, T.Popova, “HYMIX benchmarking tests: code-to-code comparison”. Abs in: International Congress on Advances in Nuclear Power Plants (ICAPP 2015) May 3-6, Nice, France.

The ERCOSAM-SAMARA experimental database should be further exploited and the project findings will be disseminated further through scientific publications in conference proceedings and scientific journals.



List of Websites:
http://ercosam.web.psi.ch

Verwandte Informationen

Dokumente und Veröffentlichungen

Kontakt

Domenico Paladino, (Group Leader, Deputy Head Laboratory Leader)
Tel.: +41 56 310 2677
Fax: +41 56 310 2199
E-Mail-Adresse

Fachgebiete

Nuclear Fission
Datensatznummer: 183447 / Zuletzt geändert am: 2016-06-07
Informationsquelle: SESAM