NC experiments have also been performed in ITF that simulate the BWR system performance. Relevant NC data have also been recorded from the operation of BWR NPP and used for benchmarking system code performance. The flow map for the operation of BWR systems shows the core power as a function of the core flow rate. A parabola like curve in the considered plane is derived from experiments and confirmed from code applications. A steep power increase up to about 50% of nominal core power can be observed from the mentioned diagram when core flow rate achieves roughly 30% of the its nominal value (i. e. value at 100% core power). This implies that the BWR systems can operate at 50% power in NC. However, in these conditions the system is prone to instabilities, identified in the literature as density wave oscillations (DWO). A wide literature exists related to the DWO that can be considered as a NC phenomenon. A state-of-the-art report on this topic has been recently issued by OECD/CSNI, Ref. [14].

Lesson learned from the application of system codes to BWR related NC experimental situations are summarized in the following, again including references where details for the analyses can be found.

(a) The curve core power versus core flow rate in NC conditions predictable by system codes is close to the experimental values in relevant ITF, Ref. [15]. The code is also capable of predicting the same curve related to the BWR operation. The capability in predicting the NC flow map in BWR is not affected by the scaling of the system.

(b) The code has been successfully used in predicting the NC measured between core and a heat exchanger installed in a pool outside the main vessel in a configuration that is typical for the SBWR equipped with the IC (isolation condenser), Ref. [16].

(c) The code has capability to predict DWO occurring recorded during the NC in typical BWR conditions, Ref. [17]. However, the prediction is largely affected by the user choices.

3.4.2. Final remarks

The system codes have been widely applied to the prediction of NC in situations relevant to PWR and BWR conditions. The scaling problem has been addressed by demonstrating that the accuracy of the prediction is not affected by the geometric dimensions of the involved systems. No major deficiencies have been detected. It can be concluded that codes are suitable in predicting NC phenomena in conditions relevant to the present generation reactors. An exception is constituted by the predictive capability of DWO. In this case the user may substantially affect the predictive capabilities.

The above conclusions can be extended to a number of systems that are part of the design of advanced reactors like the PRHR and the CMT in the case of the AP-600 and the IC in the case of the SBWR, in relation to which a suitable experimental database exists. However, in this last case, more rigorous procedures may be established where:

— Precision requirements are established (e. g. core flow rate in NC must be predicted with an error of 3%);

— Measurements are shown to comply with the fixed error-threshold;

— Accuracy of the predictions, adopting well established input decks, is shown to lie within established error bands.

— Needs for experiments and for development of new models are derived from deficiencies found from the above process.

J. Meseth

Siemens AG, Unternehmensbereich KWU, Germany

Abstract. In some of the passive safety systems of Siemens’ SWR 1000 boiling water reactor (i. e. the emergency condensers and containment cooling condensers), natural circulation is the main effect on both the primary and secondary sides by which optimum system efficiency is achieved. Other passive safety systems of the SWR 1000 require natural circulation on the secondary side only (condensation of steam discharged by the safety and relief valves; cooling of the RPV by flooding from the outside in case of core melt), while still other systems require stratification to be effective (i. e. the passive pressure pulse transmitters and steam-driven scram tanks). Complex natural circulation and stratification can take place simultaneously if fluids with different densities are enclosed in a single volume (in a core melt accident, for example, the nitrogen, steam and hydrogen in the containment). Related problems and the solutions thereto planned for the SWR 1000 are reported from the designer’s viewpoint.

1. INTRODUCTION

In recent years the Power Generation Group (KWU) of Siemens AG, in conjunction with German electric utilities and with the support of European partners, has been developing the SWR 1000, a medium-capacity boiling water reactor (BWR) with an electrical generating capacity of approximately 1000 MW (Figure 1 presents the conceptual arrangement of the containment). This reactor is evolutionary in its design for normal plant operation. The main difference to the existing Siemens BWR design is the increased water inventory in the reactor pressure vessel (RPV) which allows full depressurization without high-pressure coolant makeup. As a result, no high-pressure flooding system is required for the RPV.

However, this evolutionary development has been supplemented by an innovative approach which involves replacing the active safety systems in part with passive features. If all essential safety functions can be fulfilled by an active system and additionally by a diverse passive system, the probability of loss of both systems due to failure is much lower than with only an active system. The core damage frequency over the plant operational period is thus some 5E-9 per annum. As the passive safety systems are able to become effective without actuation from outside sources, operating personnel are in principle not necessary in order to return the plant to a safe condition. Human error can be more or less ruled out, as operating personnel need not actively intervene after an accident for a period of several days.

The passive safety systems utilize basic laws of physics such as gravity, for example, enabling these systems to function without electrical power supply or actuation by instrumentation and control (I&C) systems. In many of these passive systems and devices, natural circulation is essential under accident conditions, while stratification is essential while in the standby mode (i. e. under normal plant operating conditions). Other passive systems function on the basis of both natural circulation and stratification, while still others require neither natural circulation nor stratification to be effective.

Natural circulation is very effective if phase transitions take place as in an RPV with natural circulation. Here, evaporation occurs in the core which is located at a quite low elevation within the plant design. However, the opposite is also possible, i. e. condensation in an apparatus located at a quite high elevation in the design.

The driving pressures of these passive safety systems are comparably high, i. e. several kPa per meter of elevational difference. It is only in single-phase flows that driving pressures remain low, at only several Pa per meter of elevational difference. If a choice has to be made between single-phase flow and phase-transition flow, the latter is the preferable condition.

In the following, the main passive systems and devices of the SWR 1000 are described with respect to the various conditions in which natural circulation or stratification can take place.

TAEA participated to the OECD/NEA International Standard Problem No:42 (ISP-42) which is hosted by the Paul Scherrer Institut (PSI), Switzerland. The ISP-42 test was performed in the PANDA test facility, at the PSI, as a sequence of Phases A through F, representing typical passive safety system operating modes covering certain specific phenomena. The configuration used for ISP-42 was corresponding to the European Simplified Boiling Water Reactor containment and passive decay heat removal system at about 1:40 volumetric and power scale, and full scale for time and thermodynamic state.

1 n

< X ^—4

Mean Deviation^— > abs n 1

250000

225000

200000

cT 175000 £

5 150000 gj 125000

Ї 100000

(6

03

x 75000 50000 25000 0

0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5

Axial Position (m)

FIG. 4. Heat flux distribution along the condenser tube (Pn=4 bar, Rev=77000-86000 andRev

= 45000for Xi=52 %) [4].

250000

200000

СМ

Е

^ 150000

х

3

100000 га ш X

50000 0

0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5

Axial Position (m)

Axial Position (m)

FIG. 6. Effect of system pressure (pure steam) [4]. TABLE I. COMPARISON OF CORRELATIONS

Both the experimental results and prediction of the RELAP5/mod3.2.2 code reveals the fact that the system behaviour during Phase-A is highly affected by the performance of Passive Containment Cooling System (PCCS) heat exchangers. The objective of Phase-A is to investigate the start-up of passive cooling system when steam is injected into a cold vessel (dry-well) filled with air and to observe the resulting gas mixing and associated system behaviour. This simulation demonstrates the importance of both pure steam condensation and steam condensation in the presence of air for natural circulation in the system which in turn governs the realistic system behaviour. The system transient has been developed into two distinct parts: first, system heat-up and pressurization period (~3800 s) due to evaporation in the reactor pressure vessel with constant heat input from the heaters in the core and weak heat removal rate from PCC heat exchangers as the result of high air mass fraction; second, system pressure stabilization period (from 3800 s to the end of analysis) during which PCC heat exchangers become active as the result of venting of air from PCC tubes. The results of RELAP5 predictions for PCC-1 heat exchanger are presented in Figs. 7 and 8 for time 1500 s and 5000 s, respectively. These two distinct times are selected to demonstrate the PCC heat exchanger performance during aforementioned two distinct parts of the transient. In these figures, only the results of 5 tubes (out of 20 tubes) which were lumped to single pipe consisting of 10 control volumes were shown. Two parameters are essential with respect to PCC heat exchanger performance; local heat flux and air mass fraction. As given in Section

4.1, the system pressure is also an important parameter for the rate of condensation. However, the effect of the system pressure is expected to be small in these two figures since the pressure difference is small, i. e. 0.7 bar.

As could be seen in these figures, the local heat flux is affected by the presence of air inside the PCC heat exchanger tubes, as expected. Since the local air mass fraction is about 0.94 (almost pure air) and constant throughout the length of the condenser tubes as predicted at t=1500 s (Fig. 7), the local heat flux values are suppressed to about 0.2 % of the local heat flux values predicted at t=5000 s during which condenser tubes are full of almost pure steam down to 1.3 m (about % of total length). The maximum air mass fraction at the bottom of the condenser tubes is less than 0.3 at 5000 s. It is to be noted that some amount of air is accumulated in bottom part of tubes and lower drum of PCC-1 after 3800 s due to terminated vent flow from PCC lower drum to the wet-well tanks. The accumulation of air at the bottom of tubes shorten the active condensation length to about % of total lenght, as seen in Fig. 8, and this reveals the fact that percent of shortening of active condensation length is also the function of system pressure and differential pressure developed between PCC lower drum and wet-well tanks. In other words, system pressure and differential pressures developed between components in a system operating under natural circulation could highly affect the rate of vent of air from PCC tubes and in turn the effective condensation length.

6. CONCLUSIONS

In this paper, activities of the TAEA concerning condensation in the presence of NC gas were given. In the experimental analysis, it was observed that the overall agreement between the analytical analysis and the experimental data obtained for heat flux or heat transfer coefficients is reasonably good. For example, the heat flux significantly decreases as the inlet air mass fraction increases. Moreover, it could be promulgated that the effect of superheating of inlet stream possesses no considerable effect on the heat flux. Another conclusion emerging from experimental studies is that the local heat flux values for pure steam and mixture runs come closer towards the bottom of the condenser tube.

PCC-1,t= 5000 s
P= 2.8 bars, Re^^l 0,000

The correlations obtained from UCB database show that the mixture side Reynolds number is also a strong parameter affecting heat transfer coefficient. However, it should be noted that the given correlations stay behind the real phenomenon occurring inside the tube. Because, neither interfacial waviness nor the suction effect is taken into account. At the same time, these correlations depend on the flow regimes of either phase. If turbulent — turbulent flow regime is in question, these correlations fail. Therefore, the studies have been concentrated on an analytical solution in which a film-wise condensation of a down-flow steam/NC gas mixture in a vertical tube is considered. In this analytical model, the mixture side is treated as turbulent flow. The effect of Prandtl number, interfacial shear stress, interfacial waviness, entrainment and deposition and especially the suction have been covered in our model. The two-fluid formulation constitutes the main routine. The interfacial temperature is estimated using the stagnant film theory. Moreover, for the mixture side, the turbulence model is developed in order to elaborate suction effect, which is one of the primary reasons of the enhancement of the mixture side heat transfer coefficient. Finally, it should be stated that the diffusion layer theory is superimposed into the model to define the closure relations.

The condensation is an important heat transfer mode for natural circulation in innovative systems nuclear reactor systems like the Simplified Boiling Water Reactor design. The realistic prediction of local heat flux in heat exchangers of passive containment cooling system is essential and due to this reason physical models in computer codes for condensation and effect of NC gases on condensation should be assessed. Though very preliminary, ISP-42 study on PANDA reveals us the fact that the realistic prediction of the performance and behaviour of PCC heat exchangers could affect the overall system behaviour and the rate of condensation heat transfer is the function of air mass fraction at the inlet of

The first experiment series included five Small Break LOCA tests with break in one hot leg of PACTEL. Three break sizes (2, 4 and 6 mm or 0.5, 2 and 4.4%, respectively) were used. Three tests included secondary system depressurization as an accident management measure. The operators of the loop also depressurized the secondary system by opening a steam generator relief valve. The core power in the tests was 80 kW corresponding to 1.8% of the scaled thermal power of the reference reactor. In all tests, only PSIS provided ECC water to the core. The initial temperature of the water in the CMT was 40°C. The primary pressure used in the tests was lower than the nominal operation pressure of PACTEL. Maximum operation pressure of the passive accumulators determined the upper limit of the experiment pressure (3.8 MPa).

Munther [4] and Munther et al. [5, 6, 7] have published the results of these experiments. The main results of the first test series can be summarised as the following:

• three flow modes of CMT operations were present in the experiments: recirculation mode with water circulating through the PSIS lines, oscillating phase with two-phase flow in to the tank and injection phase with flow of steam in to the CMT.

• the ECC water injection stopped totally several times during injection, due to rapid vapour condensation in the CMT.

• flow reversed in the broken cold leg leading to flow of cold water from the downcomer towards the break

•

Although some problems with condensation in the CMT occurred no core heat-up was detected. It should be mentioned here that the CMT used in the test was very large, about twice as large as the scaled volume of four accumulators of the reference plant. On the other hand, the volume of the CMT is about the same as the volume of the rest of the loop. All the tests in this first series were terminated before the CMT was totally empty.

The results presented in the following chapters were gained within the project "European BWR-R&D Cluster for Innovative Passive Safety Systems (INNO-IPSS)" supported by the EU within the 4th framework programme.

2.1. The Emergency condenser

The design of the Emergency Condenser, as already used in the Gundremmingen Unit A Power Plant, is shown in Fig. 2. The tubes are about 10 m long. The tube inner diameter is

38.7 mm and the wall thickness 2.9 mm. Because the wall thickness resulted in a thermal conductivity of about two thirds of the total thermal conductivity a second bundle was designed with an inner diameter of 44.3 mm and the minimum acceptable wall thickness of 2 mm. Tests with the second bundle were performed with the bundle oriented in a vertical position, under an angle of 40.9° and in a horizontal position. The orientation under an angel of 40.9° should decrease the height between a flooded bundle (with zero energy transfer) and an empty bundle (with maximum energy transfer to the pool); in addition the mixing within the pool should be enhanced.

The tests have been performed under the same thermal conditions as in a real BWR; the pressures chosen were 7, 5, 3 and 1 MPa. The tests were performed in the NOKO facility, see Appendix A. The test procedure started with a flooded bundle, then the water level in the pressure vessel was decreased stepwise until the bundle was empty.

In Fig. 3 and 4 the experimental results are given. It should be added that the accuracy of the results ranged between less than 10 per cent for high powers and several ten per cent for power levels below 1 MW; the conditions for a flooded bundle are — on the other hand — relatively accurate.

In principle, the dependence of the transferred power on the water level in the pressure vessel is similar for both bundles. The influence of the orientation of the bundle is only minor, see Fig. 4.

Post-test calculations with ATHLET are shown in Fig. 3 for the first bundle and in Fig. 5 for the second bundle; the agreement is good. The distortion at about 5 m is due to the vertical part of the bundle.

It was shown that despite the good agreement for the (integral) transferred power the local deviations in the heat transfer during condensation as calculated by the ATHLET and CATHARE, see Fig. 6, are large.

FIG. 5. NOKO-2 power levels from tests and calculations for the bundle in vertical position

—I——— 1——— 1——- 1——— 1——— 1——— 1——— 1——— 1

0123456789 10

Tube length [m]

FIG. 6. Comparison between ATHLET and CATHARE calculation residts.

Therefore, for orientation tests were performed with a single tube (dimensions of the first bundle) instrumented with needle probes for the identification of the film thickness. To better study the film thickness a test tube with a tube of the second bundle and movable film probes are underway.

It was a result of the single tube tests that non-condensables were accumulated in front of the condensed water rather than been distributed along the tube.

When considering natural circulation systems for severe accident mitigation, one should note that the implementation of the natural circulation systems intended to cope with DBAs can essentially reduce the probability of severe accidents. Furthermore, many systems designed to cope with DBAs can be also used for severe accident mitigation purposes. The level of knowledge of natural circulation phenomena under severe accident conditions (i. e. with a degraded core, reactor pressure vessel and other plant equipment) is much lower than the level of knowledge for conditions where the plant damage is within the design limits. Keeping in mind the above considerations, in new reactor concepts the priority in the implementation of the natural circulation systems is given to the systems intended to prevent severe accidents. As for the mitigation systems, their implementation is mainly aimed to ensure that the integrity of the containment is maintained, because it is the last barrier against the release of radioactivity to the environment.

Safety systems for the mitigation of severe accident are designed to meet level 4 of the defence in depth strategy against significant releases of radioactivity to the environment. From this point of view, the main goals of severe accident mitigation are as follows:

(a) Prevention of H2-detonation,

(b) Energy removal from the containment,

(c) Corium retention inside the RPV by in-vessel and/or ex-vessel cooling, or

(d) Corium retention inside the containment by spreading and cooling of the molten core.

According to the above goals, systems for severe accident mitigation can be classified by function: H2-control, heat removal and corium/debris retention. Below the passive systems for severe accident mitigation in new designs are discussed in the context of their use of natural circulation as a design principle.

When a station blackout accident occurs, the decay heat of AC600/1000 reactor core can be removed to the atmosphere as the ultimate heat sink through natural circulation flows established by the primary coolant loop, the secondary side cycle of SG and air cycle
respectively. Nuclear Power Institute of China (NPIC) has built a passive emergency residual heat removal (ERHR) experiment facility in order to research the natural circulation flow characteristics for AC600/1000 ERHR system and to verify the computer code ERHRAC. The

facility parameters are as the following:

Heating power: 80-400 kW

Design pressure in the secondary side of SG: 8.6 MPa

Elevation difference between air cooler and SG: 7.2 m

Equivalent diameter of chimney: 0.7 m

Height of chimney: 12 m

The core make-up water tanks are used in the design of AC600/1000 so as to eliminate the high-pressure safety injection pumps and increase reliability of the safety system. The function characteristics and coolant flow transient of core make-up water tank, accumulator and automatic depressurization system can be tested using NPIC core make-up water tank experiment facility to simulate a small LOCA. The facility parameters are as the following:

Design pressure:

Design temperature:

Core make-up tank volume: Accumulator volume:

RPV volume:

Pressurizer volume:

Break sizes:

The safety function “fuel cooling during transients and accidents” is ensured by provision of sufficient coolant inventory, by coolant injection, sufficient heat transfer, by circulation of the coolant, and by provision of an ultimate heat sink. Depending on the type of transient/accident, a subset of these function or all of them may be required. Various passive systems and components are proposed for WWER-1000/V-392 and WWER-640/V-407 reactor concepts to fulfill these functions.

For WWER-640/V-407 reactor, steam generator passive heat removal system (SG-PHRS) which does not require the electricity supply is designed to remove the decay heat in case of non-LOCA events and to support the emergency core cooling function in case of LOCAs. Reactor coolant system and passive heat removal equipment layouts provide heat removal from the core following reactor shutdown via steam generator to the tanks of chemically demineralized water outside the containment and further to the atmosphere by natural circulation as it is shown in Figure 1. Reactor power that can be removed from the core by coolant natural circulation is about 10% of the nominal value, which guarantees a reliable residual heat removal. Thus, in case of non-LOCAs the decay heat is removed by coolant natural circulation to steam generator boiler water. The steam generated comes into the passive heat removal system where steam is condensed on the internal surface of the tubes that are cooled on the outside surface by the water stored in the demineralized water tank outside the containment. The water inventory in this tank is sufficient for the long-term heat removal (at least 24 hours) and can be replenished if necessary from an external source.

Containment passive heat removal system (C-PHRS) of V-407 reactor removes heat from the containment in case of a LOCA and is designed to fulfill the following functions: (1) emergency isolation of service lines penetrating the containment and not pertaining to systems intended to cope with the accident; (2) condensation of the steam from the containment atmosphere; (3) retention of radioactive products released into the containment; (4) fixing of the iodine released into the containment atmosphere. The steam from the containment atmosphere condenses on the internal steel wall of the double-containment being cooled from the outside surface by the water stored in the tank. So, the system operates due to natural circulation of the containment atmosphere and water storage tank. The design basis of this system is to condense the amount of steam equivalent to decay heat release during 24 hours after reactor trip without the water storage tank replenishment.

The emergency core cooling system of V-407 reactor comprises three automatically initiated subsystems: (1) hydroaccumulators with nitrogen under pressure, which are the traditional ECCS accumulators being used at operating WWER-1000 reactors, (2) elevated hydrotanks open to the containment, and (3) equipment for deliberate emergency depressurization of the primary circuit. All these subsystems are based on the principle of passive operation providing for long-term residual heat removal in case of a loss-of-coolant accident accompanied by the plant blackout (i. e. AC power supply is not needed for ECCS operation). In the first stage of the accident, primary pressure is decreased due to loss of coolant and operation of the passive heat removal system. The further cooling down and pressure decrease are realized via steam generator PHRS and containment PHRS. When the pressure difference between primary circuit and containment has decreased to 0.6 MPa, the passive valves of the emergency depressurization system open connecting reactor inlet and outlet with the fuel pond space. When the reactor and containment pressure difference has decreased below 0.3 MPa, the ECCS hydrotanks begin to flood the reactor. This sequence results in creating of so called emergency pool where the reactor coolant system is submerged to and in connection of this emergency pool with the spent fuel pond. The natural circulation along the flow path shown in Figure 2 (reactor inlet plenum — core — reactor outlet plenum — “hot” depressurization pipe — fuel pond — “cold” depressurization pipe — reactor inlet plenum) provides the long-term heat removal from the core in case of a LOCA combined with loss of all electric power. The water in the emergency pool and spent fuel pool reaches the saturation point in about 10 hours. The steam generated will condense on the internal surface of the steel inner containment wall, and condensate flows back into the emergency pool. This configuration ensures also the heat removal from reactor vessel bottom to keep the corium inside the reactor in case of postulated core melt event.

The passive residual heat removal system (PHRS) is included in the V-392 design to remove heat from the reactor plant. The design basis of this system is that in case of station blackout, including loss of emergency power supply, the removal of residual heat should be provided without damage of the fuel and of the reactor coolant system boundary for a long time period. The PHRS consist of four independent trains; each of them is connected to the respective loop of the reactor plant via the secondary side of the steam generator. Each train has pipes for steam and condensate, valves and modular air-cooled heat exchanger installed outside of the containment as it is shown in Figure 3. The steam that is generated in the steam generator due to the heat released in the core condenses in the air-cooled heat exchanger, and condensate is returned back to the steam generator. The motion of the cooling media (steam, condensate and air) takes place in natural circulation.

FIG. 3. V-392. Passive heat removal system

The passive system for reactor flooding during LOCA in V-392 design comprises two groups of hydroaccumulators as it is shown in Figure 4. First group (so called first stage accumulators) consists of four traditional ECCS accumulators being used at operating WWER-1000 reactors; these accumulators are pressurized by nitrogen to 6 MPa and connected in pairs to the upper and lower plenums through special nozzles in the reactor pressure vessel. Second stage accumulators are 8 tanks connected to the reactor coolant system through the check valves and special spring-type valves. These valves are kept closed by the primary pressure; when the primary pressure drops below 1,5 MPa, the spring open the valve. Such a connection configuration and valve design ensures continuity of hydrostatic head irrespective of the primary pressure change during an accident. Installation of hydraulic profiling of the outlet route ensures a step-wise limitation of the water flow rate from the tank when the water level in the tank is decreasing. The water inventory in the second stage accumulators (about 1000 t) ensures the core cooling for 24 hours during a LOCA even if all active ECCS mechanisms are inoperable. Joint operation of the second stage accumulators and SPOT gives a possibility to increase the period indicated.

For two-phase natural circulation, scaling laws are provided by Ishii-Kataoka (1984) which is widely applied. The PUMA facility simulating the SBWR has been designed based on this philosophy. The power to volume scaling philosophy proposed by Zuber (1980) is also applicable for two-phase systems. The integral test facility being set-up to simulate the Advanced Heavy Water Reactor (AHWR) has been designed based on this philosophy. Prior studies by Nayak et al. (1998) has shown that this philosophy is well-suited for pressure-tube — type reactors where it is possible to use full-size components for most parts of the loop. It requires 1:1 scaling for elevation, pressure, temperature and velocity with the same fluid used in the prototype and the model. Such constraints do not exist for the scaling philosophy proposed by Ishii and Kataoka. [3]

(1) (2)

A 2 і Ai

FIG. 1. Schematic of non-unform diameter natural circulation loop.

The above equations can be non-dimensionalised using the following substitutions.

A A.
z H	D D.
W a a =—- ; в W
ai	and di
; S	т ■

Where tr, Ar and Dr are respectively the reference values of time, flow rate and hydraulic diameter defined as

Vtp.
(6)
r
W

It is easy to see that the reference time is nothing but the loop circulation time. Dr and Ar are respectively the length average hydraulic diameter and flow area of the loop and the total circulation length, Lt=ELi. In case of negligible local pressure losses, X(Leff)i becomes equal to the total circulation length, Lt of the loop. The non-dimensional equations can be expressed as:

	Grm Re3
(7)
дв a дв = Vt dT+ah dS=V
heater
(8a) (8b) (8c)
дв + a дв = stPcLt в дт ac ds Ac
cooler
дв + a дв = 0 дт ap дs
pipes

Where T=ESi/ai, (leff)i=(Leff)i/Lt, Grm=DI3pI2PgATr/p2, and St=Nu/RessPr. Assuming fully developed forced flow correlations are valid, the friction factor, fi can be expressed as

_ p _ pa = Rif = Re (d / a )

Where p=64 and b=1 for laminar flow and assuming Blassius correlation to be valid for turbulent flow p and b are respectively 0.316 and 0.25. In a non-uniform diameter loop, it is possible that some pipe sections are in turbulent flow (Re>4000) and some in laminar flow (Re<2000) and still others in transition flow (2000<Re<4000). However, if we assume the entire length of the loop to be under either laminar or turbulent flow conditions, then equation (7) can be expressed as

(10)

(11)

1.1 The test facility

The test facility SUCOS-3D consists of a tank (1298 x 580 x 275) whose outer walls are 20 mm thick and made of Plexiglas (Fig. 3). The ratio between lengths in the test facility and EPR is 1:20. A 30 mm thick copper plate which is heated by electric heat conductors simulates the core melt; it is isolated from the ground, from the walls, and from outside with Teflon. The heating power is scaled according to the volume of the sump as (1/20)3. Plexiglas structures replace the structures of concrete in the prototype. The heat exchangers are slab heat exchang­ers made of copper; their cooling tubes have a meander form and use water as a cooling fluid. The horizontal heat exchanger is divided in four separate sections: two small and two large ones. The vertical heat exchangers consist of 8 sections: 4 inner and 4 outer sections respec­tively. Additional coolers in an upright position are present in the test facility. The space above the heated copper plate is called pool, see Fig. 2. The space above the horizontal heat ex­changers is called horizontal side area. The space between the vertical heat exchangers is called vertical side area. The vertical channel without heat exchangers connecting the pool and the horizontal side area is called chimney.

Several experiments were performed in SUCOS-3D varying the value of the power input to the copper plate, the arrangement of the heat exchangers, the inlet temperature of the secondary fluid in the heat exchangers, and the level of water. Only temperatures were measured using thermocouples in two planes near the mid section. The experiments are characterized by a so called pool temperature. This is the mean temperature in the pool area which is measured by six thermocouples below the tilted roof of the pool.

In order to perform the numerical simulation and interpretation with FLUTAN, a SUCOS-3D experiment had to be chosen which is consistent with the one already simulated for SUCOS — 2D. Therefore, one was chosen in which the horizontal and the outer vertical coolers were in operation, while the internal vertical coolers were not active. The electric heat supply of the heated copper plate amounted to 1,240 W. The inlet temperature of the coolant on the secon­dary side of the heat exchangers was set to 20°C and the flow rate was 20 or 40 g/s. The measured pool temperature was 32.6 °С.