water experiments

Nuclear Engineering and Design 245 (2012) 113–124

Contents lists available at SciVerse ScienceDirect

Nuclear Engineering and Design journal homepage: www.elsevier.com/locate/nucengdes

Counter-current flow limitation in a model of the hot leg of a PWR—Comparison between air/water and steam/water experiments Christophe Vallée a,∗ , Tobias Seidel a , Dirk Lucas a , Matthias Beyer a , Horst-Michael Prasser b , Heiko Pietruske a , Peter Schütz a , Helmar Carl a a b

Helmholtz-Zentrum Dresden Rossendorf e.V. (HZDR), Institute of Safety Research, D-01314 Dresden, Germany ETH Zürich, Dept. of Mechanical and Process Engineering, Institute of Energy Technology CH-8092 Zürich, Switzerland

a r t i c l e

i n f o

Article history: Received 27 July 2011 Received in revised form 31 December 2011 Accepted 3 January 2012

a b s t r a c t In order to investigate the two-phase flow behaviour in a complex reactor-typical geometry and to supply suitable data for computational fluid dynamics (CFD) code validation, a model of the hot leg of a pressurised water reactor was built at Helmholtz-Zentrum Dresden Rossendorf (HZDR). The hot leg model is devoted to optical measurement techniques, therefore, a flat test section design was chosen and equipped with large windows. In order to enable the operation at high pressures, the test section is installed in the pressure chamber of the TOPFLOW (Transient twO Phase FLOW) test facility of HZDR, which is used to perform the experiments under pressure equilibrium with the inside atmosphere. Counter-current flow limitation (CCFL) experiments were performed, simulating the reflux-condenser cooling mode appearing in small break loss-of-coolant-accident (LOCA) scenarios. The fluids used were air and water at room temperature and pressures of up to 3.0 bar, as well as steam and water at pressures of up to 50 bar and the corresponding saturation temperature of 264 ◦ C. One selected 50 bar experiment is presented in detail: the observed behaviour is analysed and illustrated by typical high-speed camera images of the flow. Furthermore, the flooding characteristics obtained from the different experimental runs are presented in terms of the Wallis parameter and Kutateladze number, which are commonly used in the literature. However, a discrepancy was first observed between the air/water and steam/water series. Further investigations show that the steam was probably wet due to heat losses and liquid entrainment from the heater circuit. Consequently, a correction of the steam measurements was required. The amount of parasitic water was evaluated indirectly over the shift of the zero liquid penetration noticed in the CCFL diagram. Finally, the experimental results confirm that the Wallis similarity is appropriate to scale flooding in the hot leg of a pressurised water reactor over a wide range of pressure and temperature conditions. © 2012 Elsevier B.V. All rights reserved.

1. Introduction In the event of a loss-of-coolant-accident (LOCA) in a pressurised water reactor (PWR), emergency strategies have to be mapped out in order to enable a safe removal of the decay heat from the reactor core, also if the operational control of the facility is degraded by additional component breakdown. During a small break LOCA with partial failure of the high pressure emergency core cooling systems (e.g. 2 of 4) and of the reactor coolant pumps, a natural circulation starts in the primary circuit. This passive mechanism allows heat removal, also if steam is generated in the reactor core due to the depressurisation of the primary circuit. But if, furthermore, the water level in the reactor pressure vessel (RPV) falls below

∗ Corresponding author. Tel.: +49 0 351 260 3227; fax: +49 0 351 260 2818. E-mail addresses: [email protected] (C. Vallée), [email protected] (H.-M. Prasser). 0029-5493/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.nucengdes.2012.01.001

the hot leg nozzle, only steam will flow to the steam generator (SG). Therefore, the natural circulation breaks down and switches to the reflux condenser mode. In the reflux condenser mode, the steam coming from the RPV condenses in vertical U-tubes of the steam generator. In each half of the steam generator, the condensate flows down the tube in which it has been formed. Therefore, about one half of the condensate flows as usual over the pump to the downcomer, whereas the other half flows over the hot leg back to the upper plenum. In the hot leg, the condensate has to flow in counter-current to the steam. The horizontal stratified counter-current flow of condensate and steam is only stable for a certain range of flow rates. If the steam flow rate exceeds this limit, for example due to the unavailability of the steam generator of another loop, the condensate can be clogged in the hot leg. This is the beginning of the counter-current flow limitation (CCFL): the liquid is carried over by the steam and partially entrained in opposite direction to the steam generator. As a consequence, the hot leg and steam generator are flooded, which

114

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

Nomenclature A D g H j J* K ˙ m p t W

flow cross-section [m2 ] diameter, depth [m] gravitational acceleration [m/s2 ] channel height [m] superficial velocity [m/s] Wallis parameter [–] Kutateladze number [–] mass flow rate [kg/s] pressure [bar] time [s] width [m]

Greek letters  difference [–] density [kg/m3 ]   surface tension [N/m] Indexes cond corr d G h L S ZLP

condensate corrected discharge gaseous phase hydraulic liquid phase steam zero liquid penetration

further decreases the water level in the RPV and, therefore, deteriorates the core cooling. In case of an additional increase in the steam flow, the condensate could be completely blocked and the cooling of the reactor core from the hot leg would be impossible. Detailed examples of such LOCA scenarios leading to the reflux condenser mode as well as to CCFL are described by Jeong (2002). For the validation and optimisation of accident management strategies, such transient scenarios are reproduced in dedicated facilities or rather simulated numerically. The use of onedimensional system codes is state of the art, but these codes are principally based on empirical correlations more than on first principle fluid dynamics models. In particular the CCFL conditions are dominated by 3D effects, which cannot be resolved by the system codes and, consequently, require the use of a computational fluid dynamics (CFD) approach. However, the actual CFD codes applied to two-phase flows do not meet the high level of confidence needed in the field of nuclear reactor safety. Moreover, the developed models and closure laws embedded in CFD must be validated to allow reliable simulations. Therefore, high-resolution experimental data at reactor typical boundary conditions is needed for comparison with CFD calculations. In order to support these tasks, counter-current flow limitation experiments were performed in a model of the hot leg of a PWR. This “hot leg model” is installed in the pressure chamber of the TOPFLOW (Transient twO Phase FLOW) test facility of Helmholtz-Zentrum Dresden Rossendorf (HZDR). The experiments were performed before and around the onset of flooding with air and water as well as with steam and water at pressures of up to 50 bar. The measured global parameters like water levels and pressure drop will be analysed in order to characterise the flow. Furthermore, the comparison with detailed visual observations will be used to explain this behaviour. The flooding characteristics of the hot leg model will be determined and plotted in terms of different nondimensional parameters. Finally, the proposed non-dimensional

parameter, which succeeds to regress all the experimental series, will be discussed with the results of previous investigations.

2. Previous investigations of counter-current flow limitation in hot leg typical geometries The first detailed investigations on counter-current flow limitation in a hot leg typical geometry (i.e. a horizontal conduit connected to a riser) date back to the late seventies. Richter et al. (1978) performed air/water experiments in a scaled down model of the hot leg of a PWR. The test section was made of acrylic glass in order to allow visual observation of the two-phase flow. They proposed to correlate the obtained flooding data with the nondimensional superficial velocity introduced by Wallis (1969) for vertical counter-current flows in pipes. Krolewski (1980) established the experimental flooding characteristics of five different hot leg geometries with air and water. It is shown that the characteristic of the CCFL depends significantly on the angle of the riser as well as on the inlet and outlet geometry. Later on, Ohnuki (1986) performed counter-current flow limitation experiments in a horizontal pipe connected to an inclined riser with air/water and saturated steam/water, both under atmospheric pressure conditions. From his results, Ohnuki concluded that the flooding characteristics is independent from the fluid combination. Furthermore, he varied the most important geometrical aspects of the hot leg: the conduit diameter, the length of the straight pipes and the angle of the riser. As a result of his investigations, Ohnuki proposed an empirical correlation to predict the onset of flooding by using the Wallis parameter, in which the y-intercept constant is a function of the length to diameter ratio of the horizontal pipe as well as of the length of the inclined riser. Furthermore, steam/water CCFL experiments under increased pressure conditions were performed in the Upper Plenum Test Facility (UPTF), which simulates the primary circuit of a PWR at full scale. The experiments related by Weiss and Hertlein (1988) simulate the reflux condenser mode after a small break LOCA. These were conducted at pressures of 3 and 15 bar and saturation conditions. A comparison of the results with the correlations of Richter et al. and Ohnuki confirmed that the Wallis parameter allows a proper geometrical scaling of the effects of counter-current flow limitation. Moreover, reflux condenser experiments were performed in the German integral test facility PKL at a pressure of 40 bar (Schmidt and Limprecht, 1991). The power of the reactor core simulator was increased stepwise to reach CCFL in the hot leg or steam generator. As a result, the distribution of the coolant in the primary circuit was measured in function of the core power. However, the flooding characteristics was not determined. More recently, Kim and No (2002) have merged in one database the experimental results obtained by eight different research groups, which were published between 1986 and 1999. The database includes cold air/water as well as steam/water experiments. By the regression through a total of 356 data points, Kim and No proposed a flooding correlation as function of the length to diameter ratio of the horizontal part of the hot leg. The prediction error of the correlation was evaluated against the considered database to 8.7%. Lately, Minami et al. (2008) performed experiments in a model of the hot leg of a pressurised water reactor with rectangular crosssection. The test section is made of acrylic glass and the fluids used are air and water at atmospheric pressure and room temperature. The study of Minami et al. focuses on the flow patterns observed in the hot leg and the results are compared with the flooding characteristic of the test section.

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

This non-exhaustive review of the literature shows that previous investigations cover many aspects of the CCFL in hot leg typical geometries. However, the goal of most of the previous experiments was the development and validation of one-dimensional system codes. Therefore, the available data focuses mainly on macroscopic effects, which do not allow a detailed validation of the CFD codes. Furthermore, to the knowledge of the authors, no experiments were performed in one facility with the fluid combinations air/water and steam/water over a wide range of pressure and temperature conditions. However, Damerell and Simons (1993) or Jeong (2002) indicated that the reflux condenser mode could appear at primary system pressures of up to 80 bar. Therefore, the CCFL experiments performed in the hot leg model of HZDR provide improved comparison possibilities between air/water data at low pressure and room temperature on the one hand and steam/water data at pressures of up to 50 bar on the other hand. Furthermore, the detailed visualisation of the two-phase flow with a high-speed camera over large windows also at reactor typical boundary conditions delivers unique possibilities for the development and validation of CFD for future applications in nuclear reactor safety.

3. The hot leg model of the TOPFLOW test facility The test section of the hot leg model is schematically shown in Fig. 1. The main components consist of the test section itself, the RPV simulator located at the lower end of the horizontal channel and the SG separator connected to the SG inlet chamber. The test section reproduces the hot leg of a pressurised water reactor from the German Konvoi type at a scale of 1:3. In order to provide optimal observation possibilities, the test section is not composed of pipes like in the original power plant, it is a 50 mm thick channel representing a cut through the vertical mid-plane of the hot leg and of the steam generator inlet chamber. Consequently, the test section is composed of a horizontal rectangular channel, a bend that connects it to an upward inclined and expanded channel, and a quarter of a circle representing the steam generator inlet chamber. The horizontal part of the test section is 2.12 m long and has a rectangular cross section of 0.05 m × 0.25 m. The SG separator and RPV simulator are identical vessels with 0.8 m × 0.5 m × 1.55 m (D × W × H) cubic shape. In order to visualise the flow over large windows at high pressures, the hot leg model is operated under pressure equilibrium. This condition is realised in the pressure vessel of the TOPFLOW test facility of HZDR (Fig. 2), in which the test section is installed. For steam/water experiments, a special heat exchanger condenses the exhaust vapour from the test section directly in the pressure vessel. As shown in Fig. 2, the cold end of this condenser is permanently connected with the inside atmosphere of the vessel in order to guarantee full pressure equilibrium at all times (cf. Prasser et al., 2006; Vallée et al., 2009). The vessel can be pressurised up to 5 MPa either with air for cold experiments or with nitrogen for steam experiments. Thanks to this method, the test section does not have to support overpressures and can be designed with thin materials. Furthermore, this makes it possible to equip the test section with large glass side walls for visual observations. This was realised in the bent region of the hot leg and of the steam generator inlet chamber (cf. Fig. 1). The flow behaviour was recorded by a high-speed video camera at frame rates of 60–100 Hz and a shutter speed of 1/500–1/1000 s during 40–180 s. Furthermore, global parameters were measured via a data acquisition system running at 1 Hz, which was synchronised with the camera. A vortex meter was used to measure the injected water mass flow rate with a precision of 0.4%. The injected air mass flow rate was measured and controlled using thermal mass flow meters, the steam flow rate over the pressure drop through an ISA

115

nozzle. The accuracy of the flow meters was evaluated to about 1%. The temperatures of the fluids were measured by calibrated thermocouples at various positions in the facility with an accuracy better than 1.0 K. Furthermore, the water levels in both separators were determined by the measurement of the differential pressure between the top and the bottom of the vessels with differential pressure transducers. The pressure drop over the test section was measured by a differential pressure transducer placed between the SG and RPV separators. These global parameters give important input or output values for the comparison with simulations. 4. Counter-current flow limitation experiments 4.1. Experimental procedure and test matrix During the experiments, a constant water flow rate was injected at the bottom of the SG separator (see Fig. 1), from where it can flow through the test section to the RPV simulator. The gas was injected into the RPV simulator from the top and flowed through the test section in counter-current to the water flow to the SG separator. The increase in the water level in the RPV simulator was used to determine the water flow rate streaming over the test section (discharge flow), which indicates the intensity of the counter-current flow limitation. Therefore, special attention was paid to the water level in this separator, which was always kept below the inlet nozzle of the hot leg. In case the level approached the height of the lower edge of the hot leg, the experiment was interrupted and the water was drained from the RPV simulator. In order to investigate the onset of flooding as well as the recovery of stable counter-current flow conditions after CCFL, two types of transient experiments were performed: • flooding experiments: the gas flow rate is increased stepwise with small increments to reach CCFL conditions and, if possible, zero liquid penetration (zero discharge flow). • deflooding experiments: in the beginning, a high gas flow rate is injected in order to establish counter-current flow limitation. During the run, the gas flow rate is decreased stepwise until the CCFL breaks down and a stable counter-current flow is achieved. The boundary conditions varied during the CCFL experiments are detailed in Table 1. The test matrix includes 2 pressure levels with the fluid combination air/water and 3 with steam and saturated water. The mass flow rate was varied between 0.1 and 0.9 kg/s for water, between 0.23 and 0.41 kg/s for air and between 0.3 and 1.2 kg/s for steam. 4.2. Typical counter-current flow limitation experiment As an example, one of the experiments was chosen to describe the observed phenomena and to explain the methodology used to analyse the measured data. This run was performed under the following boundary conditions: a system pressure of 50.0 bar, a temperature of about 262 ◦ C and a water flow rate of 0.72 kg/s. An analysis of the evolution of the global parameters over time (Fig. 3) allows the characterisation of the flow behaviour. Especially the water levels measured in the separators and the pressure difference between them give an indication of the flow regime. Furthermore, the increase in the water level in the RPV simulator allows the determination of the water flow rate streaming over the test section (discharge flow). This indicates the onset of flooding and beyond it the intensity of the counter-current flow limitation. Therefore, linear interpolation lines were added to the measured water level in Fig. 3. Based on the trend indicated by these lines, the experiment was divided into 4 regions:

116

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

Fig. 1. CAD side view of the hot leg model test section (dimensions in mm).

Fig. 2. Simplified scheme of the experimental apparatus.

I. At steam flow rates lower than 0.82 kg/s (t < 65 s), the water level in the SG separator is constant and the slope of the water level increase in the RPV simulator corresponds to an average water flow rate of 0.71 kg/s. The discrepancy with the direct measurement of the injected water flow rate is within the measuring accuracy. This behaviour indicates a stable countercurrent flow, confirmed by the camera images (Fig. 4a), which

is characterised by a constant and very low pressure drop over the test section (<0.2 kPa). II. At t = 65 s, the steam flow rate is increased to about 0.94 kg/s (±0.02). Immediately, the pressure difference between the separators increases, indicating the beginning of the countercurrent flow limitation. With a delay of about 5 s, the slope of the water level in the RPV simulator decreases to a discharge

Table 1 Test matrix of the counter-current flow limitation experiments. Gas [–]

Pressure [bar]

Temperature [◦ C]

Air Air Steam Steam Steam

1.5 3.0 15.0 30.0 50.0

18–24 18–24 197 232 262

Mass flow rates [kg/s] Water 0.1–0.9 0.1–0.9 0.3–0.9 0.3–0.6 0.3–0.7

Number of runs [–] Gas 0.23–0.41 0.23–0.41 0.3–0.7 0.35–0.85 0.6–1.2

Flooding

Deflooding

5 7 3 2 4

1 8 3 2 2

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

117

flow to a minimum of 0.08 kg/s and the pressure drop over the test section increases to values of over 4 kPa. The camera pictures (Fig. 4c) reveal a highly mixed two-phase flow: large slugs are observed which flow up the riser and transport water into the SG separator, where the water accumulates. IV. After t = 150 s, the steam flow rate stabilises to 0.95 kg/s (±0.005) for about 40 s, leading to a discharge water flow rate of 0.14 kg/s. The further increase in the water level in the SG separator occurring during this time seems not to have a significant impact on the discharge water flow. However, the high-speed camera observation shows the additional formation of large slugs in the steam generator inlet chamber (Fig. 4d). All the slugs developing in the riser and in the SG inlet chamber obstruct the steam flow, increasing consequently the droplet entrainment significantly.

Fig. 3. Variation of the steam mass flow rate (top diagram, red curve), of the pressure drop over the test section (top diagram, green curve), of the water level in the RPV simulator (bottom diagram, blue curve) and in the SG separator (bottom diagram, ˙ L = 0.72 kg/s and p = 50 bar. (For purple curve) during the CCFL experiment at m interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)

water flow rate of 0.23 kg/s. Consequently, the water level in the SG separator increases significantly. Furthermore, the pressure difference between the separators becomes unstable and fluctuates between 1 and 2.5 kPa due to the slugs generated in the hot leg (Fig. 4b). III. For 105 < t < 150 s, the steam flow rate is slowly increased to values of up to 0.99 kg/s. This further reduces the discharge water

The slight decrease of the steam flow rate at the end of the experiment leads to a decreasing CCFL intensity after t = 150 s, which is similar to the processes observed during deflooding experiments. The flow conditions described here are similar to those observed during the air/water experiments performed in the same test facility and reported by Deendarlianto et al. (2011). 5. Flooding characteristics of the hot leg model 5.1. Data processing method 5.1.1. Principle The plot of the gas flow rate versus the discharge water flow rate during CCFL leads to the flooding characteristics. In order to automate the arrangement of the flooding diagram, a data treatment routine was developed. The discharge water flow rate was determined from the time derivative of the water level in the RPV simulator. However, the derivation amplifies each slight fluctuation

˙ L = 0.72 kg/s and p = 50.0 bar. Fig. 4. Flow behaviour during steam/water counter-current flow at m

118

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

3.2

3.2

Δp > 0.5 kPa Δp < 0.5 kPa

3.0

3.1

2.6

unfiltered low-pass filter selected points

time

2.4

region III

jG [m/s]

jG [m/s]

2.8

2.2 2.0 0.00

region IV

3.0

2.9

region II 0.02

0.04

0.06

0.08

0.10

2.8 0.000

jL [m/s] Fig. 5. Steam vs. discharge liquid superficial velocity during the CCFL experiment at 50.0 bar with and a water flow rate of 0.72 kg/s.

of the measured signal and induces too large oscillations of the superficial velocity (see “unfiltered” curve in Fig. 5) for further analysis. In order to damp these fluctuations, the measured water level was first treated with a low-pass filter of Gaussian type with a time constant of 6.0 s. Fig. 5 shows the result of the time derivation after filtering, which presents a significant reduction of the oscillations in comparison with the raw measurement data. As the last procedure, selection criteria were applied to the filtered curve in order to consider only the quasi-stationary points obtained after the onset of flooding. Consequently, to be selected for the flooding characteristics, each point (i.e. time step) had to fulfil the following 3 criteria simultaneously: 1. The pressure drop measured over the test section was used to distinguish between stable counter-current flow and CCFL (cf. remarks in the phenomenological analysis of Section 4.2). In fact, it was considered that only points with a pressure drop higher than 0.5 kPa correspond to counter-current flow limitation and can, therefore, be included in the flooding diagram. This threshold value was chosen on the basis of general observations. As an example of its representativeness, Fig. 3 shows that around the onset of flooding the pressure drop rises rapidly from values lower than 0.2 kPa to 1 kPa. 2. In order to evaluate the quasi-stationarity of each point, the time derivative of the superficial velocities was calculated. This should not exceed 0.0005 m/s2 for the discharge water flow (after filtering) in order to be considered for the flooding characteristics. 3. The second procedure was applied equally to the gas superficial velocity with a limit fixed at 0.01 m/s2 . 4. Besides the stable CCFL points, also inflexion points can satisfy the two aforementioned conditions. Therefore, points are only considered to be stable if they additionally are immediately preceded or followed in time by at least another point fulfilling the first 3 criteria. As an example, the result of the data processing method for the steam/water experiment presented in Section 4.2 is shown in Fig. 5 (series “selected points”). This shows already the trend of the flooding characteristics with a relatively restrained scatter compared to the raw measurement data. 5.1.2. Validation of the processing method In order to validate the data processing method presented in the previous section, the outcome shown in Fig. 5 was compared with the results from the linear regression lines of the RPV water level shown in Fig. 3. The slope of the regression lines was used to calculate the averaged value of the discharge water flow rate over the chosen period of time. However, the corresponding steam flow rate is not evident to define. In fact, due to the difficulties to operate the

0.005

0.010

0.015

0.020

0.025

0.030

jL [m/s] Fig. 6. Comparison between the points selected with the data processing method and the linear regression lines shown in Fig. 3.

test facility during these highly transient steam/water experiments (the steam flow rate was nearly doubled within a few minutes), the steam flow rate could not be stabilised perfectly after each stepwise increase (cf. Fig. 3). Furthermore, because the steam flow rate is the main input parameter of the considered thermal–hydraulic system, the reduction of the measured values to a time average is not significant enough. Therefore, it was decided to bound the average with error bars ranging from the minimum to the maximum measured steam flow rate in the considered region (as defined in Section 4.2). The comparison between both methods is shown in Fig. 6. The points delivered by the data processing method accumulate around the quasi-stationary equilibrium states. In Fig. 6, 3 accumulation zones can be identified, which correspond to the regions defined in Section 4.2, as shown by the comparison with the error bars. This agreement makes clear that the 2 methods are comparable. However, the data processing method allows to reduce the scatter of the flooding characteristics because a couple of actual flow rates are associated at each time step. In fact, using this method, the slight variations of the steam flow rate during the experiments lead to a scan of the flooding characteristics and allows to recognise more clearly its slope. However, some differences can be seen between the results of both methods, in particular in region II. These differences are due to the slightly transient character of the flow in this region caused by the varying steam flow rate. This explains why the data processing method has selected only 2 points for the flooding characteristics. Consequently, the probability that the selected time steps correspond to the averaged value is low. On the contrary, the agreement between both methods is nearly perfect in region IV due to the very stable steam flow rate. This indicates that in spite of the selection method used to treat the data, point accumulation zones should be considered as the most reliable. Altogether, the comparison shows that the developed data processing method allows to determine precisely the flooding characteristics, in spite of the measured flow rate fluctuations. 5.2. Flooding characteristics in the Wallis parameter diagram For a meaningful comparison of experimental data, the nondimensional superficial velocity Ji∗ (or Wallis parameter) is commonly used to plot the flooding diagram. As shown in Vallée et al. (2011), for the phase i this is defined for channels with rectangular cross-section as follows:



Ji∗

= ji

1 i gH L − G

(1)

where j is the superficial velocity and  the density of the fluid (indicated by L for the liquid and G for the gas), g the gravitational

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

0.7

119

2.4

0.5

Air/water : Air/water: 1.5 bar 3.0 bar Steam Stea m/water: 15 bar 30 bar 50 bar

2.0

1/2

KG

J*G

1/2

[-]

0.6

[-]

2.2

Air/water : Air/water: 1.5 bar 3.0 bar Steam Stea m/water: 15 bar 30 bar 50 bar

1.8

1.6

0.4 0.0

1.4

0.1

0.2

0.3

0.4

0.0

J*L1/2 [-]

0.2

0.4

0.6 1/2

KL

0.8

1.0

1.2

[-]

Fig. 7. Flooding characteristics of the hot leg model plotted in terms of the Wallis parameter.

Fig. 8. Flooding characteristics of the hot leg model plotted in terms of the Kutateladze number.

acceleration and H the height of the channel. The Wallis parameter is convenient for all types of comparisons because this is a nondimensional parameter. Furthermore, due to the density ratio term, it takes into account the effect of the pressure on the fluid densities. As indicated in the review of the literature (Section 2), the Wallis parameter was used in the past by many researchers to correlate CCFL data obtained in hot leg typical geometries. For all the experiments (see details in Table 1), the points belonging to the flooding curve according to the method exposed in Section 5.1 were plotted in terms of the Wallis parameter in Fig. 7. A zoom on the region of interest was chosen (i.e. the origin of the vertical axis is not zero) in order to show the results in more details. This diagram first shows the typical decreasing trend of flooding curves, indicating that during CCFL an increase of the gas flow rate further decreases the discharge water flow rate. Furthermore, Fig. 7 reveals a slight segregation between the air/water experiments on one hand and the steam/water experiments on the other hand: mainly due to a higher zero liquid penetration point (interception of the flooding curve with the ordinate axis), the flooding of the steam/water flows was obtained at higher non-dimensional gas superficial velocities.

to the Wallis diagram. These results show that the surface tension obviously does not explain the observed discrepancy.

5.3. Flooding characteristics in the Kutateladze number diagram The Kutateladze number is commonly used to correlate flooding experiments in large vertical pipes (cf. Levy, 1999) or through horizontal perforated plates (cf. Hawighorst et al., 1984; No et al., 2005). This non-dimensional number is defined as:

 Ki = ji

i2 g(L − G )

1/4 (2)

The Kutateladze number includes the surface tension  and therefore one essential physical property of the fluids varied indirectly with the temperature in our experiments. Furthermore, according to Kim and No (2002), this is used together with the Wallis parameter as a second possibility to predict counter-current flow limitation in the hot leg in the one dimensional system code RELAP5. However, Glaeser (1992) has shown in a theoretical derivation of the classic flooding correlations that amongst the two non-dimensional parameters, only the Wallis parameter is applicable to horizontal counter-current flows. This result is supported by an analysis of the own air/water experiments (cf. Vallée et al., 2011). As shown in Fig. 8, the Kutateladze number also fails to correlate our flooding data, in particular the air/water and the steam/water experimental series. Furthermore, the steam/water series at 15, 30 and 50 bar tend to separate in the Kutateladze diagram compared

5.4. Consideration of the effects of subcooling and wet steam As shown in the previous sections, the classical Wallis parameter and Kutateladze number both fail to properly correlate our flooding data. Discrepancies were found especially between the experimental series conducted with air/water on the one hand and steam/water on the other hand. One of the main uncertainties arising whilst conducting steam/water experiments is the steam quality, which can differ from unity due to the presence of moisture formed for instance by condensation as a consequence of heat or pressure losses in the system. Since these undesirable effects may explain the observed discrepancies, this issue is discussed in the following sections. 5.4.1. Direct effect of slight subcooling During steam/water experiments, the inevitable heat losses in the system always lead to a slight water subcooling, measured at about 2 K as shown in Table 1. Nevertheless, this value underestimates the water temperature at the interface, the thermocouple being located at the bottom of the RPV separator, where thermal stratification in the liquid is likely. Furthermore, at this level of subcooling, no noticeable effect on the CCFL behaviour is being expected, as demonstrated in the experimental study by Wan (1986). He investigated CCFL in a 51 mm diameter horizontal pipe connected over a 90◦ elbow to a vertical one, which reproduces the geometry of the coolant inlet and outlet lines of a CANDU reactor. The experiments were performed at atmospheric pressure with steam and slightly subcooled water. In order to investigate the influence of condensation effects, the water subcooling was varied during the steam experiments between 0 and 6 K. The results show that for low liquid Wallis parameters (JL∗ ≤ 0.35), the CCFL characteristics were not influenced by the subcooling (cf. Figs. 5 and 6 of his paper). 5.4.2. Effect of wet steam: qualitative considerations The steam quality, and in particular condensation caused by heat losses in the test facility, could affect the effective gas flow rate available for flooding. In fact, the real steam flow rate through the test section may be overestimated because the flow meter is located far upstream of the hot leg model. This explanation matches the fact that the gaseous Wallis parameter needed to reach flooding with steam/water was found to be higher than that with air and water. Unfortunately, no specific instrumentation was available during the experiments which could be used to measure the

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

quality of the steam injected into the test section. Furthermore, no dedicated test was performed to estimate directly or indirectly the arising amount of liquid. Nevertheless, in Figs. 7 and 8 it is remarkable that zero liquid penetration was clearly reached during the air/water experiments (especially at 1.5 bar), but apparently not during the steam/water experiments. However, this observation becomes surprising when noticing that the flooding characteristics of the steam experiments present obviously two different parts: ∗1/2

1/2

- for JL > 0.05 or KL > 0.2 (in Figs. 7 and 8, respectively), the CCFL characteristics describe a line almost parallel to that of the air/water experiments; - for smaller parameters, the steam/water flooding points present a nearly vertical trend. The nearly vertical evolution of the flooding characteristics at low liquid discharge flow rates is particularly pronounced for the 50 bar experiments (cf. diagrams). However, this trend is abnormal in a CCFL diagram plotted in terms of the square root of nondimensional parameters. In fact, according to the knowledge of the authors, such behaviour has never been described in the relatively abundant CCFL literature. Consequently, we will suppose hereafter that the nearly vertical part of the flooding characteristics is caused by a steam quality lower than unity. In that case, the liquid transported by the steam into the test section is likely to accumulate in the RPV simulator: due to the large cross-section, this is a place of low gas velocities and, furthermore, the lowest point of the overall experimental apparatus. If this hypothesis is right, the RPV simulator is filled on the one hand by the discharge water, and on the other hand by the deposition of mist carried by the steam. This means that the method used to measure the discharge water flow, which is based on the water level increase in the RPV simulator, possibly includes a disturbance due to wet steam. In that case, the water level in the RPV simulator further increases also after the zero liquid penetration has been reached due to the parasitic liquid contained in the steam. According to this analysis, the nearly vertical part of the steam/water flooding characteristics in Figs. 7 and 8 should correspond to the zero liquid penetration. However, it has to be mentioned that the plot of the square root of the non-dimensional parameters leads to a distortion of the diagram which enlarges distances close to the zero liquid penetration. This effect emphasises graphically the problem already at small liquid amounts. 5.4.3. Quantification of the parasitic amount of liquid Following the reasoning of previous section, the flooding characteristics of the steam/water experiments have to be corrected in order to take into account the effects of mist carried by the steam flow. A determination of the parasitic amount of liquid in retrospect is only possible over the abscissa of the near vertical part of the CCFL characteristics. As a result, this parasitic liquid flow rate is extrapolated to the partial water delivery conditions. This appears to be reasonable since the slug flow regime established after the onset of flooding was observed up to zero liquid penetration (cf. Section 4.2). In order to improve the statistics, the followed approach supposes that this amount of liquid only depends on the system pressure. Consequently, for the determination of the parasitic amount of water, the flooding points were plotted in terms of the mass flow rate separately for each of the three pressure levels. As shown exemplarily for the experiments performed at 30 bar in Fig. 9, the two regions of this CCFL diagram have been delimited by a line. The coordinates of the line were chosen manually to isolate above the points belonging to the zero liquid penetration.

0.90 0.85

zero liquid penetration

0.80

mG [kg/s]

120

30 bar Limit Selected Average

counter-current flow limitation

0.75 0.70 0.65 0.60

33.7 g/s 0.55 0.0 0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

mL [kg/s] Fig. 9. Selection of the flooding points belonging to the zero liquid penetration (example for the 30 bar experiments).

Finally, the parasitic liquid flow rate was calculated from the average abscissa of the selected points. At 30 bar, 33.7 g/s of water was obtained (cf. Fig. 9), which corresponds to a vapour quality of 96.0% with the steam flow rate of about 0.8 kg/s measured during zero liquid penetration. Furthermore, this represents a liquid mass fraction of 4.2% or a volume fraction of only 0.8‰. 5.4.4. Plausibility check for steam condensation In order to further support or invalidate the hypothesis of parasitic water entrainment in the steam flow, it was attempted to identify its origin. Therefore, a plausibility check concerning heat losses and the resulting steam condensation was performed. Indirectly it is possible to evaluate approximately the steam condensation amount over three different ways: 1. During other experimental series (e.g. co-current flow experiments) performed at the same pressure and temperature conditions, steam flow rates down to 0.035 or 0.075 kg/s were injected in the test section. The high-speed camera observations indicate that steam was still flowing through the riser of the hot leg. Consequently, the condensation amount must be significantly lower than 0.075 kg/s. 2. The heat flux released to the atmosphere over the air/air heat exchanger used to cool the inner atmosphere of the pressure chamber (see Fig. 2) was measured during the experiments. This is a good gauge for the heat losses of all the components inside of the pressure chamber. The measured heat flux was lower than 40 kW, which represents at 50 bar about 0.024 kg/s of steam condensation in the pressure chamber. This is in agreement with the maximised condensation amount evaluated in the previous point. However, this value probably overestimates the parasitic amount of liquid formed in the chamber as only the steam condensing upstream the location of CCFL (situated in the test section, likely around the riser) disturbs the measurement. 3. The condensation rate in the steam line was evaluated with a theoretical model. The empirical correlations recommended by the VDI-Wärmeatlas (1994) were used to calculate the heat flux through the insulation as well as the natural convection around the line. At each pressure level, steam saturation temperature was assumed at the outer wall of this DN100 pipe and the room temperature was supposed to be 20 ◦ C. The temperature dependency of the heat conduction of the insulation material was modelled with a polynomial function of the second degree according to the material properties for rock wool indicated in the VDI-Wärmeatlas (1994). Furthermore, the outer diameter of the steam line was set to 260 mm after measurements performed in the test facility and its total length (i.e. from the TOPFLOW

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

Condensate amount relative to the 50 bar case [%]

100%

80%

60%

40%

Zero liquid penetration Cooling power of heat exchanger Heat losses in steam line

20%

0% 0

5

10

15

20

25

30

35

40

45

50

55

Pressure [bar] Fig. 10. Evolution of the relative amount of condensate in function of the system pressure (base 100 at 50 bar).

separator outlet to the inlet of the pressure vessel) was evaluated to 44 m according to isometric drawings. Because of the complicated pipe routing, the natural convection was calculated with the correlations for horizontal pipes on one hand and for vertical pipes on the other hand. The temperature of the outer surface of the insulation material constitutes the interface between both models. Consequently, this was calculated iteratively in order to get convergence between the heat fluxes through the insulation and the heat losses due to natural convection. The correlations for natural convection around horizontal and vertical pipes lead to differences of less than 1%, therefore, the average values were taken. Heat losses of up to 4.4 kW were calculated, representing a maximum condensation rate of 2.6 g/s. Although the insulation conditions of the line are for sure worse compared to the model assumptions, the calculation results are about one order of magnitude lower than the heat losses measured in the pressure vessel. Consequently, the condensation effects along the steam line are probably of second order. For the different possibilities mentioned previously, the variation of the evaluated amount of condensate in function of the pressure level is shown in Table 2. Absolute values calculated for the heat losses are smaller than the amounts of condensate that have to be assumed to correct the zero liquid penetration. Furthermore, the relative trends are given in Fig. 10 normalised to the values obtained for 50 bar. This comparison points out that the amount of condensate due to heat losses in the pressure chamber (from measurements of the cooling power) or in the steam line (from calculations) depends strongly on the pressure. However, the evolution of the parasitic amount of water measured from the zero ˙ ZLP , cf. previous section) is significantly less liquid penetration (m sensitive to the pressure. Consequently, besides the absolute values, these different trends reveal that steam condensation due to heat losses is probably not the main reason for the observed shift of the zero liquid penetration. 5.4.5. Plausibility check on liquid entrainment from the steam/water separator Another parasitic source of liquid water in the steam line could be liquid entrainment from the separator of the heater circuit of the TOPFLOW test facility. In fact, the steam flow rate needed to reach CCFL, and a fortiori zero liquid penetration, is relatively high and a perfect separation of the two-phase flow in the separator cannot be guarantied. During previous experimental series conducted at the TOPFLOW test facility, liquid entrainment has already been observed with help of conductivity probes installed in the steam line above the separator. Unfortunately, this instrumentation has

121

been removed in the meantime and was not available during the hot leg experiments in order to support this hypothesis. Therefore, again a dedicated plausibility check was performed. As shown in Table 3, the steam superficial velocity in the outlet pipe of the TOPFLOW separator is higher than 10 m/s at zero liquid penetration. In order to have an idea of the flow pattern reached in this section at such gas velocities and high void fractions, the flow regime transition model for vertical pipes proposed by Taitel et al. (1980) was used. According to their analysis, annular flow can only exist if the gas flow rate is sufficient to raise the droplets entrained in the gas core. Consequently, if the developed model predicts annular flow, the steam flow is likely to entrain liquid water out of the separator. The criteria proposed by Taitel et al. (1980) to describe the transition to annular flow is only a function of the fluid properties: jG = 3.1

(g) √ G

1/4

(3)

The flow transition velocity was calculated accordingly in function of the pressure level as shown in Table 3. A comparison with the flow conditions at zero liquid penetration reveals that the superficial velocity is far above the transition, confirming that the steam injected into the hot leg model is probably wet. Moreover, the liquid entrainment can explain why the parasitic amount of water measured over the zero liquid penetration is almost independent of the pressure. In fact, the superficial velocity at zero liquid penetration decreases with an increase in pressure (cf. Table 3). As a result, the amount of entrained water is expected to decrease as well. The dependency of the heat losses to the system pressure being inverse, the superposition of both effects can lead to a mutual compensation. 5.5. Correction of the flooding characteristics Finally, the parasitic water accumulating in the RPV simulator during zero liquid penetration is probably due to both steam condensation and liquid entrainment from the TOPFLOW separator. Consequently, for a proper correction of the flooding characteristics, the single contribution of each effect should be determined separately. In fact, liquid entrainment only affects the discharge liquid flow, whilst steam condensation also reduces the steam flow rate. Furthermore, the condensation rate probably only depends on the pressure level, but the amount of entrained liquid should additionally be a function of the steam flow rate. As these effects cannot be quantified in retrospect from the available data, only two limit cases are investigated here. For the correction of the flooding characteristics, independently ˙ ZLP was subof its origin, the obtained parasitic amount of water m ˙ L as follows: tracted from the discharge water flow m ˙ L (t) − m ˙ ZLP (p) ˙ L,corr (t, p) = m m

(4)

This is the sole needed correction whilst assuming that the parasitic water originates just from liquid entrainment. However, if on the contrary the parasitic water is only due to steam condensation, its amount should be subtracted from the measured steam flow ˙ G as well, leading to: rate m ˙ G (t) − m ˙ ZLP (p) ˙ G,corr (t, p) = m m

(5)

The flooding characteristics obtained after application of these corrections for the limit cases are presented in terms of the Wallis parameter in Fig. 11. On one hand, assuming liquid entrainment only (Fig. 11a), a small discrepancy between the air/water and steam/water experiments is observed. On the other hand, in case of pure steam condensation (Fig. 11b), a good agreement between all the experimental series is obtained. As the parasitic water is

122

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

Table 2 Comparison of the amount of condensate determined by different methods. ˙ ZLP [g/s] m

Pressure [bar]

15 30 50

˙ S [g/s] (1) min. m

32.7 33.7 38.0

(2) Cooling heat exchanger

35 75 75

(3) Heat losses in steam line

Power [kW]

˙ cond [g/s] m

Power [kW]

˙ cond [g/s] m

16.5 31 36

8.5 17 22

2.78 3.59 4.36

1.43 2.00 2.66

Table 3 Evolution of the steam flow rate at zero liquid penetration in function of the pressure. Pressure level [bar]

˙ S @ zero liquid penetration [kg/s] m

jS @ at separator outlet [m/s]

jS @ at transition to annular flow [m/s]

15 30 50

0.67 0.80 1.05

24.8 15.0 11.6

4.76 3.14 2.22

probably due to a combination of both effects, the real CCFL characteristics of the steam experiments should be in between the investigated limit cases. This means that a slight difference may remain between the air and steam experiments. However, considering the scatter of the data as well as the remaining uncertainties, one can speak of a reasonable agreement between all the measured CCFL series. Consequently, this result shows that the Wallis similarity is appropriate to scale flooding in the hot leg of a PWR over a large range of pressure and temperature conditions. In particular, no significant discrepancy could be observed between the air/water and steam/water series, although the fluid properties vary noteworthy. This confirms the results of Ohnuki (1986) obtained in smaller scale pipes and over a more limited range of boundary conditions (atmospheric pressure).

in Siemens/KWU (1987). The experiments were performed with steam and saturated water at pressures of 3 and 15 bar in order to check the applicability of the Wallis similarity to the original NPP geometry and boundary conditions. Furthermore, investigations were focused on the influence of the several metres long pipe of the Hutze (ECC nozzle placed at the bottom of the hot leg) with respect to the flooding behaviour. As a result, it was recommended to calculate the Wallis parameter based on the flow path in the region of the Hutze. This concerns in particular the flow cross-section A for the calculation of the superficial velocity and the hydraulic diameter Dh as length scale. Later on, this approach was supported by the small scale experiments of Ohnuki et al. (1988). The resulting data points (calculated with A = 0.3974 m2 and Dh = 0.639 m) are compared with the own flooding characteristics in Fig. 12a. For clarity reasons, only the experiments performed at the same pressure levels as those of UPTF were plotted: the air/water tests at 3 bar and the steam/water tests at 15 bar (denoted as “S/15 bar” in Fig. 12, the series (a) and (b) corresponding to the limit cases indicated in Fig. 11). Fig. 12a shows an approximate agreement between both experimental works: at UPTF the zero liquid penetration was obtained at higher gas fluxes and the slope of the flooding characteristics is steeper. Consequently, in a second comparison presented in Fig. 12b, the UPTF data points were recalculated with the geometrical parameters in the parts of the hot leg without Hutze: A = 0.4418 m2 and Dh = 0.750 m. In that case, the overall agreement with the own results is better, in particular for the zero liquid penetration. This may indicate that flooding does not mainly occur along the Hutze. However, close inspection reveals that the slope is again too steep, especially close to zero

5.6. Comparison with the results obtained at UPTF As discussed in Section 2, most of the past CCFL experiments published in the literature were performed with air and water at atmospheric pressure and room temperature. In a previous paper, the own air/water experiments were compared with similar experimental data and empirical correlations for pipes (cf. Vallée et al., 2011). Consequently, in this study the comparison with literature is focussed on the steam/water experiments. With respect to geometry, scale and boundary conditions, the only relevant data was obtained at the UPTF. In UPTF, the counter-current flow limitation in a hot leg was investigated at the original power plant scale during the dedicated test series no. 11 reported by Weiss and Hertlein (1988) and

0.7

Air/water : Air/water: 1.5 bar 3.0 bar Steam Steam/water: 15 bar 30 bar 50 bar

0.6

[-]

0.6

J*G

J*G

1/2

1/2

[-]

0.7

0.5

0.5

0.4

0.4 0.0

0.1

0.2

J*L

1/2

0.3

[-]

(a) Only liquid entrainment

0.4

0.0

0.1

0.2

J*L1/2

0.3

0.4

[-]

(b) Only steam condensation

Fig. 11. Flooding characteristics of the hot leg model plotted in terms of the Wallis parameter after correction of the steam quality effects.

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

0.75

0.70

0.70

0.65

air/3.0 bar S/15 bar (a)

[-] 1/2

0.60

J*G

J*G

TOPFLOW:

0.60

1/2

[-]

0.65

123

0.55

0.55

S/15 bar (b) UPTF:

0.50

0.50

3 bar 0.45

0.45

15 bar

0.40

0.40 0.0

0.1

0.2

J*L

1/2

0.3

0.4

[-]

(a) With consideration of the Hutze (A = 0.3974 m² / Dh = 0.639 m)

0.0

0.1

0.2 1/2 J*L

0.3

0.4

[-]

(b) Without consideration of the Hutze (A = 0.4418 m² / Dh = 0.750 m)

Fig. 12. Comparison of the present data with the CCFL characteristics of UPTF (Weiss and Hertlein, 1988; Siemens/KWU, 1987).

liquid penetration. Since none of the geometrical considerations give satisfying agreement, the observed differences may also be due to the rectangular cross-section of the TOPFLOW hot leg model. Nevertheless, with respect to the uncertainty of the data, the differences between both experimental series are acceptable and precise explanations would require dedicated investigations. 6. Summary and conclusions The two-phase flow behaviour in a flat model of the hot leg of a pressurised water reactor was investigated during counter-current flow limitation. Experiments were performed with air and water at low pressure and room temperature as well as with steam and saturated water at pressures of up to 50 bar. The operation of the test section in a pressure chamber allowed detailed optical observations of the two-phase flow at these reactor typical boundary conditions. As an example, one steam/water experiment was presented and analysed with help of the recorded high-speed camera pictures. Commonly, the macroscopic effects of CCFL are represented in a flooding diagram using the non-dimensional superficial velocities as coordinates. Therefore, a numerical data treatment method was developed to plot the flooding characteristics based on the discharge water flow accumulating in the RPV simulator. A comparison between the air/water and steam/water flooding curves first revealed a difference when plotted in terms of the classical Wallis parameter or Kutateladze number. Further investigations show that the steam was probably wet, which requires a correction of the steam measurements. The amount of parasitic water was evaluated indirectly over the zero liquid penetration noticed in the CCFL diagram. Finally, the experimental results confirm that the Wallis similarity is appropriate to scale flooding in the hot leg of a PWR irrespective of the gas (air or steam) and for pressures ranging from 1.5 to 50 bar and temperatures of 18 to 264 ◦ C. However, the assumptions made for data correction unfortunately could not be conclusively confirmed from the available data. Furthermore, it was not possible to determine the exact origin of the parasitic water due to lack of specific instrumentation. Plausibility checks show that the amount of liquid cannot only be explained by heat losses and is probably also due to liquid entrainment from the separator of the TOPFLOW heater circuit. Consequently, uncertainties remain in the adapted method to correct the results, which should be clarified in a second experimental campaign.

Acknowledgements This work was carried out within the framework of a research project funded by the German Federal Ministry of Economics and Technology, project number 150 1329. The authors would like to thank the TOPFLOW team for their work on the test facility and the preparation of the experiments, by name Klaus Lindner, Heiko Rußig, Marko Tamme and Steffen Weichelt.

References Damerell, P.S., Simons, J.W., 1993. Reactor safety issues resolved by the 2D/3D program. International Agreement Report. U.S. Nuclear Regulatory Commission Report No. NUREG/IA-0127, Washington, USA. Deendarlianto, Vallée, C., Lucas, D., Beyer, M., Pietruske, H., Carl, H., 2011. Experimental study on the air/water counter-current flow limitation in a model of the hot leg of a pressurised water reactor. Nuclear Engineering and Design 241 (8), 3359–3372. Glaeser, H., 1992. Downcomer and tie plate countercurrent flow in the Upper Plenum Test Facility (UPTF). Nuclear Engineering and Design 133 (2), 259–283. Hawighorst, A., Kröning, H., Mayinger, F., 1984. Fluid dynamic effects in the fuel element top nozzle area during refilling and reflooding. Nuclear Science and Engineering 88 (3), 376–385. Jeong, H.Y., 2002. Prediction of counter-current flow limitation at hot leg pipe during a small-break LOCA. Annals of Nuclear Energy 29 (5), 571–583. Kim, H.Y., No, H.C., 2002. Assessment of RELAP5/MOD3.2.2(against flooding database in horizontal-to-inclined pipes. Annals of Nuclear Energy 29, 835–850. Krolewski, S.M., 1980. Flooding limits in a simulated nuclear reactor hot leg. In: Submission for a Bachelor of Science. Massachusetts Institute of Technology, Cambridge (MA), USA. Levy, S., 1999. Two-Phase Flow in Complex Systems. John Wiley & Sons, New York, USA. Minami, N., Nishiwaki, D., Kataoka, H., Tomiyama, A., Hosokawa, S., Murase, M., 2008. Experiments on air-water countercurrent flow in a rectangular duct simulating PWR hot leg. In: Proceedings of the 16th International Conference on Nuclear Engineering, Orlando (FL), USA, May 11–15 (Article No. ICONE16-48113). No, H.C., Lee, K.-W., Song, C.-H., 2005. An experimental study on air-water countercurrent flow limitation in the upper plenum with a multi-hole plate. Nuclear Engineering and Technology 37 (6), 557–564. Ohnuki, A., 1986. Experimental study of counter-current two-phase flow in horizontal tube connected to inclined riser. Journal of Nuclear Science and Technology 23 (3), 219–232. Ohnuki, A., Adachi, H., Murao, Y., 1988. Scale effects on countercurrent gas-liquid flow in a horizontal tube connected to an inclined riser. Nuclear Engineering and Design 107 (3), 283–294. Prasser, H.-M., Beyer, M., Carl, H., Manera, A., Pietruske, H., Schütz, P., Weiß, F.-P., 2006. The multipurpose thermalhydraulic test facility TOPFLOW: an overview on experimental capabilities, instrumentation and results. Kerntechnik 71 (4), 163–173. Richter, H.J., Wallis, G.B., Carter, K.H., Murphy, S.L., 1978. Deentrainment and counter-current air-water flow in a model PWR hot leg. U.S. Nuclear Regulatory Commission Report No. NRC-0193-9, Hanover (NH), USA. Schmidt, H., Limprecht, H., 1991. Versuch PKL III B 3.4. Siemens/KWU Report No. KWU E312/91/16, Erlangen, Germany.

124

C. Vallée et al. / Nuclear Engineering and Design 245 (2012) 113–124

Siemens/KWU, 1987. UPTF-Test no. 11: Countercurrent flow in PWR hot leg test. Siemens/KWU Quick Look Report R515/87/08, Erlangen, Germany. Taitel, Y., Bornea, D., Dukler, A.E., 1980. Modelling flow pattern transitions for steady upward gas-liquid flow in vertical tubes. American Institute of Chemical Engineers (AIChE) Journal 26 (3), 345–354. Vallée, C., Seidel, T., Lucas, D., Beyer, M., Prasser, H.-M., Pietruske, H., Schütz, P., Carl, H., 2009. Influence of the fluid properties on co-current two-phase flows in a horizontal channel connected to a riser. In: Proceedings of the 7th World Conference on Experimental Heat Transfer, Fluid Mechanics and Thermodynamics (ExHFT-7), Krakow, Poland, 28 June–03 July, 2009, pp. 443–452. Vallée, C., Seidel, T., Lucas, D., Tomiyama, A., Murase, M., 2011. Comparison of counter-current flow limitation experiments performed in two different

models of the hot leg of a pressurised water reactor with rectangular crosssection. Journal of Engineering for Gas Turbines and Power – Transactions of the ASME 133 (5) (Article 052917). VDI-Wärmeatlas, 1994. Berechnungsblätter für den Wärmeübergang. Verein Deutscher Ingenieure, VDI Gesellschaft Verfahrenstechnik und Chemieingenieurwesen, Siebte erweiterte Auflage, VDI-Verlag, Düsseldorf. Wallis, G.B., 1969. One-Dimensional Two-Phase Flow. Mc Graw-Hill, New York, USA. Wan, P.T., 1986. Countercurrent steam-water flow in an upright 90◦ elbow. In: Proceeding of the Eighth International Heat Transfer Conference, vol. 5, San Francisco (CA), USA, August 17–22, pp. 2313–2318. Weiss, P.A., Hertlein, R.J., 1988. UPTF test results: first three separate effect tests. Nuclear Engineering and Design 108 (1–2), 249–263.