ML19122A309

ANS 2019 Grant S.S. Prediction of Reflood in Rod Bundle

ML19122A309
Person / Time
Issue date:	02/19/2019
From:	Stephen Bajorek, Beck F, Cheung F, Garrett G, Grant G, Hoxie C, Jin Y, Kirk Tien NRC/RES/DSA/CDB, Pennsylvania State Univ, University Park, PA
To:
Tien K
References
Download: ML19122A309 (4)

Text

STEADY STATE PREDICTIONS OF REFLOOD CHARACTERISTICS IN A ROD BUNDLE Grant R. Garrett, Yue Jin, Faith R. Beck, Fan-Bill Cheung, Stephen M. Bajorekx, Kirk Tienx, Chris L. Hoxiex Nuclear Engineering Department, 127 Reber Building, University Park, PA 16803, grg5094@psu.edu, yuj118@psu.edu, fxc4@psu.edu Mechanical Engineering Department, 127 Reber Building, University Park, PA 16803, frb115@psu.edu xOce of Nuclear Regulatory Research, United States Nuclear Regulatory Commission, Washington, D.C. 20555, stephen.bajorek@nrc.gov, chris.hoxie@nrc.gov, kirk.tien@nrc.gov INTRODUCTION Accurate predictions of two-phase "ow phenomena ob-served in the Penn State University (PSU)/United States Nu-clear Regulatory Commission (NRC) Rod Bundle Heat Trans-fer (RBHT) facility were made by performing a "rst order analysis from fundamental thermal hydraulic and heat transfer relations. The calculations were made from a developed script for Engineering Equation Solver (EES). The calculations per-formed are part of an ongoing eort in model development and code validation for re"ood phenomenon. In this paper, results from "rst order calculations are compared to experimental data from the PSU/NRC RBHT facility. The facility is designed to represent a section of a PWR core. Further details regarding the RBHT facility are covered in the following section.

The calculations made by the "rst order analysis com-pared well with experimental data. Assumptions made during the analysis, and uncertainty in the experiments contribute most to the dierences between calculated results and experi-mental measurements.

NRC/PSU RBHT FACILITY This section brie"y covers the details of the PSU/NRC RBHT facility. Further details can be found in the Rod Bun-dle Heat Transfer Test Facility Description (Hochreiter, 2012).

Figure 1 provides an isometric view of the RBHT facility and "gure 2 provides a P& ID of the RBHT facility. The facility, which contains 49 vertical, 3.66 m (12 ft) long test rods (four unheated corner rods and 45 heated rods) with Inconel 600 cladding in a 7x7 geometry, having the rod diameters, rod pitches and spacer grids comparable to those in commercial PWRs, was speci"cally designed to obtain fundamental "ow and heat transfer data during re"ood transients. The 45 non-corner rods are electronically heated by a heating coil inside each of the heated rods.

The experiments are performed by introducing power to the heated rods and heating up the test section with no liquid phase water present. The system pressure is controlled by a PID controller. When the desired initial conditions are met, water, at a controlled temperature and test section inlet velocity is pumped through the test section. The experiments are performed until the rod bundle completely quenches or a speci"ed time after a steady state condition is achieved.

Fig. 1. Isometric View of RBHT Facility (Hochreiter, 2012)

Fig. 2. RBHT Facility P& ID (Hochreiter, 2012)

EES SCRIPT DEVELOPMENT This section reviews the "rst order approach used to de-velop the Engineering Equation Solver (EES) script used to make steady state RBHT predictions. The script developed in EES is used to predict the steady state quench height in the bundle and the vapor temperature at the exit of the bundle for various RBHT experiments. Physically, liquid phase water will enter the bottom of the test section through the lower plenum. As the coolant rises through the bundle it will remove power from the heated rods and heat up. Once the bulk coolant temperature reaches its saturation value, nucleate boiling will

occur. As the coolant continues to "ow through the facility more boiling will occur until nearly all of the liquid in the coolant is vaporized and the vapor is superheated.

To predict the steady state quench location, it was as-sumed that all of the power from the heated rods of the RBHT facility was transferred to the liquid "ow until all of the liq-uid phase is vaporized. Since the power pro"le of the RBHT facility is known, the power required to completely vaporize all of the liquid in the coolant can match with the integrated power of the RBHT facility to a certain height. It is expected that the steady state quench location should be near this lo-cation because when there is little to no liquid in the coolant, the rods in the facility can no longer be quenched. To make this prediction, the power to heat the coolant from its facility inlet value to its saturation value is calculated and added to a calculated power required to completely vaporize all the liquid in the coolant once it has reached its saturation temperature.

Equation 1 is implemented into the EES script and used to determine the power, qsl required to raise the temperature of the coolant from its temperature, Tinlet at the test section inlet to its saturation temperature, Tsat. Equation 2 is implemented into the same EES used to determine the power, qlv required to completely vaporize the liquid in the coolant after it has reached its bulk saturation temperature. In equations 1 and 2, m is the mass "ow rate of the coolant, -cp is the speci"c heat, and hlv is the latent heat of vaporization. Equations 1 and 2 implemented and solved in EES, with "uid properties taken from the STEAM_IAPWS database.

qsl = m -cp (Tsat Tinlet)

(1) qlv = mhlv (2)

To predict the steady state vapor temperature at the exit of the RBHT test section, an energy balance approach, similar to that used in the prediction of the steady state quench front location was used. Since the total bundle power, qbundle is known and the power to completely vaporize the coolant is calculated in the prediction of the steady state quench front location, the excess power, qexcess in the bundle that remains after all liquid in the coolant has vaporized can be determined by subtracting these two values. For the prediction of the steady state vapor temperature at the exit of the bundle, it is assumed that all of the excess power remaining in the bundle after vaporizing all of the liquid in the coolant is transferred to the vapor. Equations 3 and 4 provide the equations for the excess power in the bundle after completely vaporizing all the liquid and the vapor temperature, Tv at the exit of the test section respectively. In equation 4, cpv is the speci"c heat of the vapor, found in the STEAM_IAPWS database.

qexcess = qbundle qsl qlv (3)

Tv = Tsat + qexcess mcpv (4)

Using the EES script developed according to the approach described in this section, predictions were made for the steady state quench front location and the steady state vapor temper-ature at the exit of the test section for multiple experiments.

The experimental conditions and comparison of results are described below.

RESULTS AND ANALYSIS This section provides a comparison of the experimental results and predictions made using a "rst order approach via a developed EES script. Two experiments are used in the comparison of results, EXP 8095 and EXP 8100. Table I provides the test conditions for EXP 8095 and EXP 8100.

TABLE I. RBHT Experiments 8095 and 8100 Conditions Parameter EXP 8095 EXP 8100

Pressure, kPa (psia) 275 (40) 275 (40)

Inlet Coolant Velocities, cm/s (time after re"ood, s) 7.62 (0-15) 5.08 (15-30) 2.54(30-45) 1.32 (45-1812) 7.62 (0-15) 5.08 (15-30) 2.54(30-45) 1.32 (45-550) 1.22 (550-1920)

Peak Power, kW/ft 0.4 0.4 Initial Peak Bundle Temperature, K (F) 1033 (1400) 1033 (1400)

Test Section Inlet Subcooling, K (F) 24 (43) 24 (43)

Figures 3 and 4 provide a comparison of the predicted and experimentally measured steady state quench front height for EXPs 8095 and 8100 respectively. Also included in these "gures are experimental quench locations preceding the steady state quench front to show its progression through time as it approaches its steady state height.

Figures 3 and 4 show that the steady state predictions of the quench front location agrees well with the experimental data for both EXP 8095 and EXP 8100. Experimentally, the quench front will oscillate slightly due to oscillating bound-ary conditions. For example, experimentally the pressure is not perfectly controlled, and pressure oscillations will occur that will cause "ashing and condensing. This will result in oscillations in the quench front. Similarly, oscillations in the test section inlet velocity occur that will subsequently result in quench front location oscillations.

500 1000 Temperature (K) 0 50 100 150 Rod Bundle Axial Elevation (in) t = 250 s 500 1000 Temperature (K) 0 50 100 150 t = 750 s 500 1000 Temperature (K) 0 50 100 150 t = 1250 s 500 1000 Temperature (K) 0 50 100 150 t = 1650 s 500 1000 Temperature (K) 0 50 100 150 t = 1795 s Fig. 3. EXP 8095 Quench Front Location 500 1000 Temperature (K) 0 50 100 150 Rod Bundle Axial Elevation (in) t = 250 s 500 1000 Temperature (K) 0 50 100 150 t = 750 s 500 1000 Temperature (K) 0 50 100 150 t = 1250 s 500 1000 Temperature (K) 0 50 100 150 t = 1750 s 500 1000 Temperature (K) 0 50 100 150 t = 1900 s Fig. 4. EXP 8100 Quench Front Location Figure 5 provides a comparison of the predicted and exper-imentally measured steady state vapor temperatures at the exit of the test section for EXPs 8095 and 8100. Figure 5 shows that the prediction of the vapor temperature at the exit of the bundle agrees well, but on average is slightly higher than the experimental data. For the predictions, it is assumed that the test section is perfectly insulated. Although insulation is used in the RBHT facility, some fraction of energy is still conducted through the insulation and not transferred to the coolant. The vapor temperature at the exit of the bundle is sensitive to any power that is conducted through the insulation. Also, experi-mentally liquid that condenses in the upper plenum has been shown to enter the test section at the exit of the test section and act as a heat sink to the vapor. Also, liquid droplets can contact the vapor temperature probes and cause large, nearly instantaneous drops in the measured vapor temperature. Since a small change in power causes a much larger change in the temperature of vapor, as compared to liquid phase water, the combination of the eects explained are likely the cause of the over-prediction of the vapor temperature at the exit of the test section.

The discrepancy in predicted vs experimentally measured vapor temperature at the exit of the test section is larger for EXP 8095 than for EXP 8100. For EXP 8095, the steady state quench front is at a location closer to the steam temperature measurement probes. This causes more liquid droplets to be in the "ow at the measurement probe locations, and droplets will impinge on the steam temperature probes at a higher average frequency than for EXP 8100. Droplet impingement on the steam temperature probe can be seen in "gure 5 by observing the large drops in temperature that approach the saturation temperature of the "uid. High enough droplet impingement on the temperature probes prohibit them from reaching a steady state temperature themselves and registering an accurate read-ing for the steam temperature. This can be seen in "gure 5 for the experimentally measured vapor temperature for EXP 8095 by examining regions that are between droplet impingements.

These regions are steadily increasing, but are not able to reach a steady state temperature before another droplet impinges on the surface of the measurement probe. Since the average frequency of droplet impingement on the vapor measurement probes is less for EXP 8100, the temperature probe itself is

able to reach a steady state temperature for certain times. Ad-ditionally, liquid droplets in the "ow will cool the vapor due to interfacial heat transfer. The vapor at the exit of the test section will experience more interfacial heat transfer for EXP 8095 than for EXP 8100 because the interfacial heat transfer area will be larger for EXP 8095 (more droplets).

0 20 40 60 80 100 Experimental Steady State Time (seconds) 400 420 440 460 480 500 520 540 560 580 600 Temperature (K)

EXP 8095 experimental vapor temperature EXP 8095 predicted vapor temperature EXP 8100 experimental vapor temperature EXP 8100 predicted vapor temperature Fig. 5. EXP 8095 and EXP 8100 Steady State Vapor Tempera-tures at exit of Test Section CONCLUSIONS A "rst order study was used to predict steady state be-havior in the NRC/PSU RBHT facility. The RBHT facility is designed to model a section of a PWR core and was speci"-

cally designed to obtain fundamental "ow and heat transfer data during re"ood transients. An EES script was developed that predicted the steady state quench front location and va-por temperature at the exit of the RBHT test section using an energy balance approach.

For the prediction of the steady state quench front loca-tion, an energy balance was performed to determine the power required to completely vaporize all of the liquid in the "ow, assuming all power from the bundle was transferred to the liquid phase. Since the power pro"le of the RBHT facility is known, the prediction of the steady state quench location can be made by "nding the height in the bundle that corresponded to the same total integrated power from the test section inlet as the calculated power to completely vaporize all of the liquid in the coolant. The predicted results from using this approach agreed well with the experimental data. The experimental steady state quench front locations oscillated about the pre-dicted values for both EXP 8095 and EXP 8100. Oscillating experimental values are caused by not perfectly controlled boundary conditions.

To predict the steady state vapor temperature at the exit of the RBHT test section, it was assumed that the excess power in the bundle after vaporizing all of the liquid in the coolant was completely transferred to the vapor in the coolant. The experi-mental steady state vapor temperature at the exit of the bundle oscillated below the predicted values for both EXPs 8095 and 8100 and were on average a few degrees lower than the pre-dicted values. Experimentally, a fraction of the power from the bundle is conducted through the insulation and not trans-ferred to the coolant. This fraction of energy is not accounted for in the predictions. Also, unintended liquid entering the system through the upper plenum acts as a heat sink for the superheated vapor. Additionally, liquid droplets can contact the vapor temperature measurement probes. These factors are not accounted for in the steady state predictions made.

Oscillating experimental boundary conditions contribute to the oscillations seen in the experimental vapor temperature measurements. Overall the prediction of the steady state va-por temperature at the exit of the bundle agrees well with the experimental data.

ACKNOWLEDGMENTS The work performed at the Pennsylvania State University was supported by the U.S. Nuclear Regulatory Commission under Contract Number: NRC-HQ-60-16-T-0002.

REFERENCES Hochreiter, et. al., (2012), RBHT Re"ood Heat Transfer Ex-periments Data and Analysis, The Pennsylvania State Uni-versity, U. S. Nuclear Regulatory Commission, NUREG/CR-6980.