EXPERIMENTAL STUDY ON CRITICAL HEAT FLUX BEHAVIOR IN SINGLE FUEL PIN WITH AND WITHOUT WIRE SPACER

Dan Tri LE*, Noriaki IN ABA** and Minoru TAKAHASHI**

*Department of Nuclear Engineering, Tokyo Institute of Technology Nl-18, 2-12-1 Ookayama, Meguro-ku, Tokyo, 152-8550, Japan [email protected]. ac .j p

**Research Laboratory for Nuclear Reactors, Tokyo Institute of Technology Nl-18, 2-12-1 Ookayama, Meguro-ku, Tokyo, 152-8550, Japan inaba@2phase .nr.titech. ac.j p [email protected]

Abstract: Light reactor could have fast neutron spectrum with high conversion ratio nearly equal unity by using tight lattice fuel assembly with wire spacer. Besides, the critical heat flux (CHF) data base for grid spacer has a great quantity of studies but it is not in case of wire type. In order to investigate coolability in tight lattice core in water reactor (BWR), an experiment of critical heat flux (CHF) was conducted using single pin flow channels with a heater pin with and without wire. We determined the critical heat flux for this system by varying the inlet temperature from 333 to 373 K and mass fluxes from 200 to 700 kg/m1 2s and the pressure was atmospheric pressure. The result show the CHF data in two- flow condition base on inlet and outlet condition in both case of heater pin with and without wire. The CHF values were higher with wire than without wire due to the effect of wire and spiral flow.

Keywords: Critical Heat Flux, Tight Lattice, Two-phase Flow, Wire Spacer.

1. INTRODUCTION

Nowadays, the fourth generation reactors are the focus of most study in the nuclear engineering. The fuel used for the fourth generation reactors are Plutonium and natural Uranium instead of the U. Most fourth generation reactors are metal-cooled fast reactor. It is known that the liquid metal-cooled fast reactors (LMFR) that have the breeding ratio higher than unity meet the requirement of sustainability. However, the LMFR poses some unique design problems compared to the light water reactors which have been commercialized and operated for long years. On the other hand, tight lattice core could have a high conversion ratio of nearly equal to unity even in light water reactors (LWR), particularly in boiling water reactors (BWR). The use of wire spacers is more suitable for the tight lattice core than the use of grid spacer. However, the tight lattice core has smaller volume compared with normal core. From thermal hydraulic point of view, the coolability or heat removability is one of key issues for the feasibility of the tight lattice core with wire spacers. The most crucible feature of coolability is the CHF of the fuel rod in the BWRs. There have been numerous studies on CHF so far. Nevertheless, the studies on CHF of tight lattice core with wire spacer are still few. In case of thermal-hydraulics for nuclear reactor, there have been a lot of studies on the CHF data base for grid spacer, but it is not in case of wire type. Thus, the critical heat flux or burnout phenomenon in tight lattice core is one of the most important studies for such kind of reactor. The CHF experiment was performed by Cheng & Muller [1] who reported that the CHF in hexagonal rod bundle with wire spacer was higher than that with grid spacer in low quality condition. The study was performed by using Freon-12 because of low latent heat and low critical pressure. The results were transferred from Freon-12 to water condition by using the fluid-to-fluid scaling law of Courtaud, et al [2], In 2009, the study of Diller, et al [3] showed that the advantages of fuel assembly with wire spacer type over grid spacers are their significant reduction in pressure drop and increase in CHF. This study was performed by simulating the hexagonal fuel assembly of pressurized water reactor (PWR) core. Their studies contributed to the evaluation of wire spacer compared with grid spacer and for the upgrade or developing a new type of light water reactor. Besides, the fuel bundle with wire spacer has advantages not only in thermal-hydraulic design of fuel but also in nuclear design of fuel. In 2009, Olander [4] performed the study about nuclear fuels, and showed that by using the hydride fuel with wire spacer, the fuel assembly could have good fuel performance and could reduce the core size. By reviewing of some studies for wire spacer, it can be seen that wire spacer has some interesting advantage, particularly in thermal-hydraulic design. Reviewing of some studies related to CHF for wire spacer, it can be seen that the studies on CHF for the fuel assembly with wire spacer are still not complete. Therefore, the main objectives of this study are to investigate the characteristics of CHF in single fuel pin, in particular the effect of wire spacer on CHF.

2. EXPERIMENTAL APPARATUS AND PROCEDURE 2.1. Experimental Apparatus The test loop used in this study is shown in Fig.l. The test loop consists of the storage tank, the water circulation pump, the orifice flow meter, the preheater, the test section and the condenser. The water flows from the water tank through the circulation pump and the pre-heater and enters the test section. Steam after going through the test section is condensed prior to flow back to the water tank. There is some consideration about the position of the orifice flow meter in the circulation loop after a preliminary test. The orifice flow meter was at first located in the downstream of the pre-heater. In the case of inlet temperature near the saturation temperature, a small amount of steam bubbles flowed out of the pre-heater into the flow meter. Therefore, the location of the orifice flow meter was moved to the upstream of the pre-heater so that steam bubbles did not affect the measurement of water flow rate.

Fig.l.Experiment apparatus 2.2. Test section The test section with the oriented vertically is shown in Fig.2, and the cross section of the flow channel is shown in Figs.3a and 3b. The main part of the test section are the heater pin, the copper electrode, the glass tube, the thermocouple and the wire spacer in case of experiment for heater pin with wire spacer. The heater pin was made of a thin stainless steel tube with an outer diameter, D, of 8 mm and a length of 420 mm. It was connected to the copper electrode at both ends by silver soldering. The electrodes were inserted into the stainless steel tube by 10 mm for the soldering and connection. Therefore, the length of the heater pin, L, that was used for experiment and data analysis was 400 mm. On the other hand, the length to diameter ratio, of around 50 was large enough to minimize the effect of heated length on CHF [3]. The wire spacer was made of a Teflon tube in which a stainless steel wire was inserted so that the spacer was electrically insulated to the heater pin (Fig.3c). The axial pitch of wire, is the axial distance over which the wire completely wraps around the heated rod (Fig.4a).The axial pitch of the wire spacer, H, was set with two different values 100 and 200 mm. For this two different values of the axial pitch, the value of HID less than 50 was the upper limit of the wire correlation for both CHF and pressure drop [3]. Gap size, S, is the distance from the outer diameter of heater pin to the inner diameter of the glass tube as show in Fig.4 (b). For the rod diameter of 8 mm, three different values of gap size were chosen: 1.1, 1.5 and 2.0 mm. By changing the rod diameter, wire diameter and glass inner diameter, we could change the size of the gap. The heater pin was directly Joule-heated by a direct current electrically, which provided an uniform heat flux on the heater pin surface. The maximum power and current of the power in this experiment were 15kW and 500A, respectively.

Cy electrode Outlet hermocouples Wire wrai T1 T2 Glass tub

L=400 mm

Cu electrode

Fig.2. CHF test section.

Thermocouples were setup to detect the CHF occurrence through the rapidly increase of surface temperature. The thermocouples were mounted of the heater pin surface at the axial locations of 10 mm and 20 mm upstream from the downstream end of the heated length, being marked as T1 and T2, respectively (Fig.5b). The reach to the CHF was detected by a rapid and sudden increase of the surface temperature which was measured by the thermocouples. Figure 5a shows the technique about the setup of the thermocouples. Figure 5a is detail each of T1 or T2 in Fig. 5b. Since the heater pin was the stainless steel tube with direct heating by the current, the three point junction technique (Fig.5a) for compensation of voltage induced by the current between two points junctions were used to setup the thermocouples. C and A mean the wires of Type K thermocouples: Chromel and Alumel, respectively. The voltage induced by the current between C and A was eliminated by adjusting the resistance of the variable resistor under low power input condition where temperature did not increase appreciably. The glass tube used in this study is Pyrex glass and crystal glass. On the outside surface of the glass tube, we made the position indications to recognize the location of CHF. There are two types of burn out or CHF phenomena, that is, in the annular flow and low flow rate, the burnout or CHF occurs in the downstream region in the channel. The burnout occurs when a liquid film on the heated surface is lost. The other type of burnout occurs in the upstream region of the channel at high heat flux with high flow rate and very high heat flux. Therefore, by using thermocouples, glass tube with length measurement and the camera, the positions where burn out or CHF occur were detected.

Fig.3. Cross-section of the test section: (a) heater pin with wire; (b) heater pin without wire; (c) wire structure

Fig.4. Axial pitch of wire, H, and Gap size, S , ----- Heater pin Thermocouple 10 mm T1

10 mm T2 A

Heater pin

(a) (b) Fig.5. Setup of the thermocouples; a-Three point junctions; b- Thermocouple position

2.3. Experimental conditions The experimental conditions are shown in Tables 1,2 and 3.

Table 1. Experimental conditions for heater pin with and without wire Parameter With wire Without wire Outer diameter of heater pin, D 8 mm 8 mm Heated length, L 400 mm 400 mm Inner diameter of glass tube, d 12 mm 12 mm Gap size, 8 2.0 mm 2.0 mm

Wire diameter 1.79 mm -

Wire axial pitch, H 200 mm - Mass flux 205-410kg/m2s 205-410kg/m2s Flow rate 0.012-0.024kg/s 0.012-0.024 kg/s Hydraulic diameter of flow 0.0035 m 0.004 m channel

Table 2. Experimental conditions forthree I different values of P/D. Experiment 1 Experiment 2 Experiment 3 Gap size, 8 2.0 mm 1.5 mm 1.1 mm Wire diameter 1.79 mm 1.4 mm 1.06 mm Inner diameter of glass tube 12 mm 11 mm 10.2 mm Hydraulic diameter 3.52 mm 2.7 mm 2.0 mm Mass flow rate 0.018 kg/s 0.018 kg/s 0.018 kg/s

Table 3. Experimental conditions for two different values of wire pitch Experiment 1 Experiment 2 Gap size, 8 1.5 1.5 Wire diameter 1.4 mm 1.4 mm Wire axial pitch, H 200 mm 100 mm 2.4. Experimental procedure and measurement item Water from storage tank was circulated into the test section by using the circulation pump. The preheater was used to heat up water before staring the experiment. Water flow rate was kept constant during the test. It was necessary to keep high flow resistance in the test section because if the flow resistance was low, there was the possibility of the instability in the measured flow rate due to the instability of pressure drop of two-phase flow in the test section. If the instability of the measured flow rate occurred, the experimental data were judged to be unreliable. When the outlet water temperature of the preheater or the inlet water temperature of the test section reached the desired value, the experiment was started by increasing power of the heater pin. The electric power was increased step by step by remote controlling the electric DC source, and the surface temperature of the heater pin was recorded by using the thermocouples with the sampling frequency of 100 Hz. As the flow rate slightly increased with the increase of the power, it was adjusted again at each time when the power was increased. The power was increased until the surface temperature of the heater pin suddenly and rapidly increased because of the reach of heat flux to CHF. When the heat flux reached the CHF, the power source is automatically shut down just after the detection of the sudden rise of the temperature. Figure 6 shows an example of the behaviors of the measured surface and inlet temperatures, voltage and current during the experiment. It can be seen that the inlet temperature was kept stable during the experiment. The surface temperature increased due to the increase of power step by step. It was judged that the bum out took place or the heat flux reached the CHF when the surface temperature of the heater pin rapidly increased.

Time (s) Fig.6. Temperature behavior during the experiment

The CHF was obtained from the measured voltage and current at the time of shutdown as follows:

Q = VI (1)

The CHF value is obtained by

G chf t f c H F s (2) where Q c h f ^s the electric power when CHF occurs, qCHF is the critical heat flux (CHF) and S is the heated surface area of the heater pin. The local quality is defined as K~hf X = (3) hfg where hf is the specific enthalpy of saturated liquid, hfg is the latent heat of vaporization, and hz is the local enthalpy that could be expressed by hz=hm+Ah (4) where hin is the inlet enthalpy, and Ah is the increase of enthalpy from the inlet of the test section to the measurement position of surface temperature on the heater pin with thermocouples. The inlet enthalpy and the increasing enthalpy are calculated as K = h f -{Tsat-Tm)cp (5) and

A h — — — (6) W where Tsat is the saturation temperature of liquid, Tin is the inlet temperature, cp is the specific heat capacity, and W is the water flow rate. The experimental data of the CHF was obtained in the range of mass flux and inlet temperature mentioned above with three different values of 8 and two different values of wire pitch, H. The measurements were made under the boiling two-phase flow condition. The accuracy of the heat flux value was high enough since it was determined from measured voltage and current. In order to keep the uncertainty of the results small, only the experimental data obtained under the condition which met the following requirements were chosen: The flow rate was stable during the experiment. The inlet temperature fluctuated in the range less than 1°C. 3

3. RESULTS AND DISCUSSION 3.1. Effect of wire spacer Figure 7 shows the comparison of the CHF values between single pin with wire and without wire. It was found that the CHF values in both types of heater pin with wire and without wire decreased with the decrease of the equilibrium steam quality. However, the CHF values in case of heater pin with wire were higher than that in case of heater pin without wire. It could be explained by the enhancement of bubble removal from the heated surface due to the effect of wire and spiral flow. In the study by Cheng& Muller [1], the results were obtained in smaller range of the quality from -0.1 to 0, and the CHF values of wire type still higher than CHF values of without wire type even if the quality close to zero. Besides, the different of CHF values between heater pin with and without wire was larger in lower qualities compared with the different of CHF in higher qualities. Figure 8 shows the comparison of the CHF values between three different values of gap size. It can be seen that at the same local quality and same total amount of water supply for the channel (same water flow rate), the CHF values were enhanced with the decrease of gap size. In more detail, the CHF values were higher in case of smaller value of gap size. It can be explained by the fact that with the same total amount of water supply for the channel or the same water flow rate, the channel which has the smallest flow area has the highest velocity. On the other hand, the coefficient depends on the velocity. Therefore, with the higher velocity, the heat transfer coefficient enhances leading to the enhancement of CHF values. Figure 9 illustrates the CHF values of two different values of wire pitch, H. It can be seen that the CHF values were nearly the same between two different values of H. It can be explained that, there were two flow channels with two different values of wire pitch. However, the flow area and the water flow rate did not change with the change of wire pitch so the flow velocity was the same. Since the velocity was the same so the heat transfer coefficient did not enhance in this case. Therefore, it is clear that the axial pitch did not have a large influence on the CHF.

3.2. Axial position of CHF The relation between the CHF values and the heated length to the position of arriving at CHF is shown in Fig. 10. It is found that at the same mass flow rate, the CHF occurred mostly in the downstream of the flow channel. Fig. 10 also shows the comparison of CHF position among three different values of gap size. In case of gap size of 1.5 and 2.0 mm, the CHF positions were mainly at the downstream of flow channel. Therefore, the CHF in this case was principally caused by liquid film dry-out. On the other hand, with the smallest flow area, gap size of 1.1 mm, there was the possibility of CHF occurring at the upstream of flow channel because of the high heat flux and high flow rate. According to the results, the CHF positions changed slightly with different values of the flow area or gap size. In more detail, the CHF position tended to slightly move from downstream to the upstream of the flow channel due to the decrease of gap size

3.3. Effect of mass flux The experiment in different mass flux value is needed for clearly understand the CHF behavior in tight lattice core. For that reason, the CHF experiments for heater pin with wire spacer were carried out under different mass flux condition. Figure 11 .a and 11 .b show the CHF values base on different mass flux. In both cases of gap size of 1.5 mm and 2.0 mm, it is clearly seen that the CHF values increased with the increase of mass flux. With the increase of mass flux, the bubble detachment from the heated surface was enhanced. Therefore, the CHF values increased with the mass flux.

1200 ------1------1------1------1------1------1------1------1------• Wire A Without wire 1000 • • • •

LU/ • oo o o -A & * A , • w 4H A A, • so o o A O A

400 G=400 kg/m2s p=0.1MPa 200 ____ .____ i____ .____ i____ .____ .____ .____ .____

Inlet quality Fig.7. Critical heat flux base on inlet condition Fig.8. Critical heat flux of three different values of gap size.

Fig.9. Critical heat flux of two different values of wire pitch,H.

Fig.10. Critical heat flux base on the heated length. (a) (b) Fig.ll. Critical heat flux base on different mass flux

3.5. Comparison with prediction methods Doerffer et al.[5] proposed the CHF prediction methods for annuli with inside heating. The prediction methods were based on the 1986 CHF look-up table of Groveneveld et al. [6]. Base on their explanation, the critical quality, gap size and pressure are three main factors that have the strongest influence on the CHF ratio between annuli channel and round tube channel. Thus, the authors developed a correlation based on three main factors as mentioned above. In this correlation, the CHF values in the 1986 look-up table of Groveneveld et al. [6] were multiplied with three correlation factors representing for critical quality, gap size and pressure. The CHF correlation for concentric annuli is

Q CHF An,cone = QCHFd=8^ p (7) where qcwv d=8 is the CHF values for a tube with the diameter of 8 mm from the 1986 CHF look-up table. However, in this study, we used the [5] correlation (Eq.(7)) with the CHF values derived from the 2006 look-up table [7] for higher accuracy. kx, k& kp are three factors representing the effect of quality, gap size and pressure, respectively. The three factors are calculated by

kx =0.859-16.179x15 +15.6x2 -7.195*2 lnx (8) ks = 0.2872 + 1.209<52 -1.156£2 5 +0.2873£3 (9) II o (10) 73 where x is the quality(x > 0.025),8 is the gap size (8< 4.26 mm), p is the pressure (p < 3.30 MPa). Fig. 12 shows the comparison between CHF results for heater pin with the wire spacer and the calculated values of Eq. (7) using the 2006 look-up table [7]. It can be seen that the experimental values of CHF for heater pin with wire spacer were higher than the calculated values of CHF for heater pin without wire. Therefore, it is evident that the CHF values were enhanced by the effect of the wire spacer. Exit quality xout Fig.12. Comparison of experimental results of wire type with the calculated results of non-wire type using concentric annuli correlation of Doerffer et al.[5] for the 2006 CHF look-up table of Groveneveld et al.[7]

Fig. 13 compares the experimental results obtained in the flow channel with a wire spacer and without a wire spacer with the 2006 CHF look-up table of [7] which is calculated Eq. (7) for annuli channel. The ratio of the measured CHF to the calculated CHF is obtained at the same pressure, mass flux and exit vapor quality. As shown in Fig. 13, the CHF values in case of heater pin without wire spacer become closer to the calculated values in high quality region. On the other hand, the different values between CHF values of heater pin with a wire spacer and the calculated CHF values without a wire spacer have an upward trend with the increase of the exit vapor quality. • No Wire Spacer _ 3 A Wire Spacer P-H — LUT-2006 u (annuli, no wire spacer) A P-H A u

£ 1 p = 0.1 MPa o 8= 1.5 mm G = 658 kg/m2s

0 0.1 0.2 Exit quality xout Fig.13. CHF ratio of experimental results to calculated results obtained from Eq.(7)

The correlation shown in Eq.(7) is the concentric annuli correlation. On the other hand, in this study, the heater pin was located at the center of the flow channel, so that the comparison between experimental data and the calculated values for the rod-centered case is needed. In the study of [5], the authors also proposed the CHF correlation for the rod-centered approach. According to this correlation, the vapor quality in annular channel is calculated by x' = x + Ax (11) where x is the quality that given from Eq. (3). The calculation for Ax is given by Ax = 0.658- 0 .3 3 /428Gal08

y< 'in- in(A x) = q_cHFd=%( PXj,x') (13)

Exit quality xout Fig.14. Comparison of experimental results with the calculated results using CHF correlation for the rod-centered approach of Doerffer et al.[5] for the 2006 CHF look-up table of Groveneveld et al.[7]

Following Fig.14, the calculated values were obtained by CHF correlation for the rod-centered approach [5] (Eq.(13)) were closer to the experimental results, compared with the concentric annuli correlation [5] (Eq.(7)).

4. CONCLUSIONS

According to the experiment results, the CHF values were higher in case of wire type so the coolability or heat removability was optimized due to the effect of wire and spiral flow. On the other hand, the effect of wire on CHF was mainly depending on the local quality. However, the change of wire pitch, H did not have a large influence on the CHF. Through the results, it could be seen that the CHF in boiling two-phase flow was mostly base on the dry-out phenomena. Nevertheless, in case of high velocity and high heat flux, the burn-out at high heat flux appeared. Base on the experiment data which obtained with different values of gap size, the CHF values were increased with the decrease of gap size. Therefore, the flow channel with smaller value of gap size has a higher coolability compare with the flow channel with large gap size value. However, it is necessary to consider about the pressure drop for the safety purpose. Besides, CHF values were also depending on the change of mass flux. Directly comparison with the CHF look-up tables of [7] had a big different with the test results because of the different in geometric condition. In case of the CHF look-up tables [7] the CHF was performed in a round tube which was different with the CHF in annuli flow as shown in our experiment. The experimental values of single pin without wire spacer agreed well with the predicted method by using the rod-centered approach method of [5]. It also shown the predicted values of CHF in case of channel without wire spacer were lower than the experimental data of channel with wire spacer. That agreed with the experimental results. All in all, through this paper, the effect of wire spacer on CHF phenomena in tight lattice core was investigated. However, the bundle experiment is needed for clearly understanding about such kind of phenomena. Therefore, the bundle experiment for this kind of study is constructing and starting to operate.

ACKNOWLEDGMENTS

The study was supported by Hitachi Nuclear Scholarship project. The authors would like to give the special thanks to Mr. Shoji Matsui for his assistance in the technical work.

REFERENCES

[1] .Cheng, X., Muller, U., 1998. Critical heat flux and turbulent mixing in hexagonal tight rod bundles. International Journal of Multiphase Flow 24, 1245-1263. [2] .Courtaud, M., Deruaz, R., D’Aillon, L.G., 1988. The French thermal-hydraulic program addressing the requirements of the future pressurized water reactors. Nucl. Technol. 80, 73-82. [3] .Diller, P., Todreas, N., Hejzlar, P., 2009. Thermal-hydraulic analysis for wire-wrapped PWR cores. Nuclear Engineering and Design 239, 1461-1470. [4] .01ander, D., 2009. Nuclear fuels - Present and future. Journal of Nuclear Materials 389, 1-22. [5] .Doerffer, S., Groeneveld, D.C.,Cheng, S.C, Rudzinski, K.F., 1994. A comparison of critical heat flux in tubes and annuli. Nuclear Engineering and Design 149 (1-3), 167-175. [6] .Groeneveld, D.C.,Cheng, S.C,Doan T., 1986. AECL-UO critical heat flux look-up table. Heat Transfer Eng 7 (1-2), 46-62. [7] .Groeneveld, D.C., Shan, J.Q., Vasic, A.Z., Leung. L.K.H., Durmayaz. A., Yang. J., Cheng, S.C., Tanase, A., 2007. The 2006 CHF look-up table. Nuclear Engineering and Design 237(15-17), 1909-1922.

NOMENCLATURE

cp specific heat d outer diameter of heater pin D inner diameter of glass tube Dhe heated equivalent diameter D),y hydraulic diameter dwire wire diameter G mass flux H axial pitch of wire spacer hf specific enthalpy of saturated liquid hfg latent heat of vaporization hin inlet enthalpy Ah increase in enthalpy hz local enthalpy I current L heated length P pitch p pressure Q chf power when CHF occurs qcHF critical heat flux Tsat saturation temperature of liquid Tin inlet temperature S heated surface area V voltage W mass flow rate x quality Ax additional term of quality

Greek letters 8 gap size

Abbreviations An annulus BWR boiling water reactor CHF critical heat flux cone concentric annulus LMFR liquid metal-cooled fast reactors LUT look-up table LWR light water reactor