Direct Vessel Injection Line Break Research Articles

Uncertainty and sensitivity analysis of the DVI line break loss of coolant accident for small modular reactor

The ACP100 is a small modular reactor with integrated passive design, and the best estimate plus uncertainty method is utilized to conduct a realistic analysis of the loss of coolant accident caused by a double-ended direct vessel injection line break of the ACP100. The thermal-hydraulic system code ARSAC is used, and the figure of merit for reactor safety is determined to be the minimum collapsed core level. Subsequently, important phenomena which may affect the core level are identified, and uncertainty distributions of the inputs are quantified, followed by the uncertainty propagation and sensitivity analysis. The tolerance limit of the minimum collapsed core level is obtained, and the key parameters during different accident phases are identified. According to the results, parameters related to the flow resistance of the injected coolant and some constitutive models of the ARSAC can be optimized to further deepen the reliability and safety design of the ACP100.

RELAP5 Foresight Thermal-Hydraulic Analysis of Hypothesis Passive Safety Injection System under LOCA for an Existing NPP in China

Qinshan nuclear power plant is the first large Chinese-designed nuclear power station based on pressurized water reactor, and the second generation main stream active safety injection system is adopted for Qinshan nuclear power plant. In this paper, a novel passive safety injection system (PSIS) has been proposed for ocean-based Qinshan Phase One nuclear power plant to replace the original active one. The PSIS contains high-pressure, medium-pressure, and lower-pressure safety injection systems, and a two-stage automatic depressurization system. To evaluate the system performance, small-break LOCA has been investigated using RELAP5. Various break sizes and locations including 2-inch, 10-inch cold leg break, and double-ended direct vessel injection line break were analyzed. Key safety parameters such as safe injection mass flow rates, coolant level of the core, system pressure, and fuel cladding temperature were monitored during the accident process. The results illustrate that the performance of the safety injection system can guarantee the effective core cooling and submerged under different LOCA even with only half of the safety injection system, which can fulfill the single failure criteria. The thermal-hydraulic analysis for the Qinshan passive safety injection system is significant to master the related technologies for Chinese engineer and develop the Chinese-designed third-generation nuclear power plants, and the PSIS can guarantee the reactor submerged under LOCA even plus the station block out accident.

Integral Effect Test and MARS-KS Calculation with Uncertainty Propagation Analysis for Direct Vessel Injection Line Break Intermediate-Break Loss-of-Coolant Accident

To investigate thermal-hydraulic phenomena during an intermediate-break loss-of-coolant accident (IBLOCA) and evaluate the effect of a direct vessel injection (DVI) line break, an integral effect test using the Advanced Thermal-hydraulic Test Loop for Accident Simulation (ATLAS) test facility was conducted as the B3.2 test item of the international cooperation project Organisation for Economic Co-operation and Development (OECD)–ATLAS Project Phase 2 (ATLAS-2) (OECD-ATLAS2). The initial and boundary conditions for the test were determined referring to the Advanced Power Reactor 1400 MW(electric) (APR1400) as a prototype with three-level scaling methodology. A single-failure criterion was applied to the operation of the safety injection pump (SIP), and four safety injection tanks (SITs) were available to cool down the reactor coolant system. In the test result, as the break nozzle was located at the DVI line, the clearance of the upper downcomer could make an effective flow path of the steam toward the break and quench the reactor core. Maximum cladding temperature was measured before clearance of the upper downcomer. Coolant inventory in the reactor pressure vessel was maintained due to the safety injection without any further core heatup. So, it was proved that the current design of the safety systems in APR1400 had a sufficient long-term cooling capability with a single SIP during a DVI line break IBLOCA. The ATLAS test data were utilized to evaluate the prediction capability of the thermal-hydraulic system code Multi-dimensional Analysis of Reactor Safety KINS Standard (MARS-KS) for a DVI line break IBLOCA scenario. The calculation result with the uncertainty propagation analysis using the PArallel computing Platform IntegRated for Uncertainty and Sensitivity analysis (PAPIRUS) toolkit proved that major phenomena such as uncovery of the core or intermittent injection of the SIT flow could be reasonably predicted by the MARS-KS code.

Analysis of the upflow conversion modification and influence on selected LOCA accidents in a PWR plant

The baffle jetting phenomenon, intensified by pressure difference across baffle plates, may cause excessive fuel rod vibrations and possible damage. The downward flow direction through the baffle-barrel region opposite to the upward flow through the core in pressurized water reactors affects increase of pressure gradient. Gaps between the baffle plates widen with reactor operating time and water jets become more pronounced. During the 2015 outage the NPP Krško has undergone the upflow conversion (UFC) modification in order to minimize baffle jetting and prevent fuel rod failure. The modification consisted in altering reactor vessel internals in such a way that the coolant downflow path in the baffle-barrel region was converted to an upflow path.The impact of the UFC modification on the nuclear power plant behaviour was analyzed in the paper. The double ended guillotine break of one direct vessel injection (DVI) line was selected as a representative scenario because the DVI line break is exhibiting the highest peak cladding temperature among the small break LOCA cases. The thermal-hydraulic analysis was carried out with the RELAP5/Mod3.3 code for core average channel conditions, while additional FRAPTRAN runs were performed to evaluate peak cladding temperatures for high power and high burnup fuel assemblies. Distribution of pressure and velocity fields in the reactor pressure vessel baffle-barrel region and change in bypass flow rates due to modification were determined by the ANSYS-Fluent calculation. The accident analysis has been performed for both the downflow configuration (pre-UFC) and the upflow configuration (post-UFC) to assess the influence of the modification. Additional cases with different reactor coolant pump trip times, higher core decay heat and reduced capacity of safety systems were also analyzed to evaluate the impact on reactor safety.

Reflux condensation behavior in SBLOCA tests of ATLAS facility

The behavior of the reflux condensation heat transfer in a hot side steam generator (SG) U-tubes during a cold leg (CL) pipe and a direct vessel injection (DVI) line break in small break loss-of-coolant accident (SBLOCA) tests of the ATLAS facility was investigated including MARS code calculations. Among the SBLOCA tests, a 6″-CL pipe and 50%-DVI line break SBLOCA test were selected to investigate the behavior of the reflux condensation. A reflux condensation heat transfer seemed to occur from the time the SG U-tubes were half-empty to near the loop seal clearing (LSC). It was found that a transition regime existed between the reflux condensation heat transfer and reverse heat transfer. The remaining reflux condensate in SG U-tubes owing to the counter-current flow limit (CCFL) phenomenon and a separating effect of liquid carry-over and/or entrainment with steam moisturizing seemed to affect the thermal-hydraulic behavior of the transition regime. It was also found that the steam flowrate of the loop pipings and SG U-tubes seemed to have a strong effect on the duration time of the transition regime, e.g., a larger steam flowrate results in a longer duration. From a comparison of the reflux condensation behavior between the ATLAS tests and MARS code calculations, overall qualitative agreements were found between the two cases. The largest discrepancies were found in the SG inlet plenum water level between the two cases, and the authors suggest that the combination effects of the remaining reflux condensate in SG U-tubes and a separating effect of liquid carry-over and/or entrainment with steam moisturizing are the reasons for the discrepancies in the water levels of the SG inlet plenum, which seems to be difficult to simulate in the current version of the MARS-KS code.

Risk reduction due to modification of normal residual heat removal system of AP1000® reactor to meet European Utility Requirements

As part of AP1000® design adaptation to European Utility Requirements, Westinghouse has carried out several modifications of the standard AP1000 design. One of these modifications has been the physical separation of the normal residual heat removal system (RNS) into two independent trains which improves the plant defense in depth. The AP1000 Probabilistic Risk Assessment, assumes in the standard design that in the event of direct vessel injection line break (DVILB) the RNS cannot inject efficiently, and therefore it is necessary to depressurize the Reactor Coolant System until atmospheric conditions in order to allow injection through the IRWST and to achieve safe conditions. This work presents the analysis performed for DVILB event tree taking into account the RNS modification and the effectiveness of RNS injection. With this purpose, different possibilities of DVILB event tree have been simulated with TRACE V5.0 patch 2 thermal–hydraulic code. The results show that some sequences that were classified as damage with the standard RNS design became success with the consideration of RNS improvement and therefore the contribution of DVILB to the total core damage frequency can be minimized.

Transient analysis of ”2 inch Direct Vessel Injection line break” in SPES-2 facility by using TRACE code

In the past few decades a lot of theoretical and experimental researches have been done to understand the physical phenomena characterizing nuclear accidents. In particular, after the Three Miles Island accident, several reactors have been designed to handle successfully LOCA events. This paper presents a comparison between experimental and numerical results obtained for the “2 inch Direct Vessel Injection line break” in SPES-2. This facility is an integral test facility built in Piacenza at the SIET laboratories and simulating the primary circuit, the relevant parts of the secondary circuits and the passive safety systems typical of the AP600 nuclear power plant. The numerical analysis here presented was performed by using TRACE and CATHARE thermal-hydraulic codes with the purpose of evaluating their prediction capability. The main results show that the TRACE model well predicts the overall behaviour of the plant during the transient, in particular it is able to simulate the principal thermal-hydraulic phenomena related to all passive safety systems. The performance of the presented CATHARE noding has suggested some possible improvements of the model.

The characteristics of break flows of SBLOCA tests in ATLAS facility

The characteristics of break flows of small break loss of coolant accident (SBLOCA) tests in the ATLAS (Advanced thermal–hydraulic Test Loop for Accident Simulation) facility were investigated. For an investigation of the characteristics of break flows for SBLOCA tests, cold leg (CL) pipe break and direct vessel injection (DVI) line break tests were compared with respect to several kinds of viewpoints, e.g., pressurizer (PZR) pressure vs. accumulated break mass; comparison of accumulated break masses; occurrence of choking at the break nozzle; comparison of upstream conditions; and comparison of the discharge coefficient. In the comparison of PZR pressure vs. accumulated break mass, the overall trends were similar between CL pipe and DVI line breaks. In the comparison of accumulated break masses, the accumulated break mass of the CL pipe break was larger than that of the DVI line break for four different break sizes. In the comparison of the occurrence of choking at the break nozzle, it was found that there should be choking at the break nozzle in the CL pipe and DVI line SBLOCA tests. For a comparison of the upstream conditions, upstream fluid temperatures showed effects of downcomer wall temperature in a DVI line break. For a comparison of the Cd values, the geometry effect, e.g., the L/D ratio, was discussed, especially regarding the two smallest break tests.

An analytical investigation of cold leg small break accidents of the ATLAS facility

A previous parametric study of the direct vessel injection (DVI) line breaks was re-evaluated to see its applicability to that of the cold leg (CL) pipe breaks in advanced thermal-hydraulic test loop for accident simulation (ATLAS). Evaluation results of the tests and analyses for the major parameters, e.g., the pressurizer (PZR) pressure, downcomer water level, collapsed core water level, and clad temperature, were compared for four different CL pipe break scenarios. The overall trends of the major parameters showed reasonable behaviors between the tests and analyses. The clad temperature showed conservative behaviors in the analyses using the suggested options. The suggested counter-current flow limit (CCFL) options for the fuel alignment plate (FAP) and cross-over legs (COLs) can be applicable to any small-break loss-of-coolant accident (SBLOCA) scenario for the CL pipe and DVI line breaks in the ATLAS tests.

An analytical investigation of direct vessel injection line break accidents of the ATLAS facility

For parametric evaluations of direct vessel injection (DVI) line break scenarios, the pressurizer (PZR) pressure, core collapsed water level, and peak cladding temperature were investigated between analyses and tests. The PZR pressure was mainly dependent upon the break flow model, e.g., discharge coefficient of the Henry-Fauske critical model. The core collapsed water level and peak cladding temperature were mainly dependent upon the counter-current flow limit (CCFL) option of the fuel alignment plate (FAP). The CCFL option of the cross-over leg (COL) affected the PZR pressure due to loop seal clearings, and seemed to have little effect on the core collapsed water level. Proper Cd values and applicable CCFL options were summarized. Cd values seemed to be dependent upon the sizes of the DVI line break. Although there is little difference for the CCFL options of the COL, the Ku-option was a preference for COLs’ CCFL option. The CCFL options of the FAP appeared to be sensitive to the core collapsed water level and peak cladding temperature. The Ku-option of the FAP tended to negatively exaggerate the core behavior and showed excessively conservative results, especially on the peak cladding temperature. For smaller breaks, e.g., ⩽25%, NA- and Wa-options will be applicable for the FAP. However, for larger breaks, e.g., ⩾50%, the Wa-option of the FAP was the preferred option. From comparisons between tests and analyses for four different DVI line break scenarios with respect to the PZR pressure, the downcomer water level, core water level, and clad temperature, selected Cd, and CCFL options showed conservatively good results.

AP1000® SBLOCA simulations with TRACE code

The AP1000® is an advanced pressurized water reactor (PWR) design developed by Westinghouse which implements passive safety systems to provide core cooling in case of accident. The performing of such systems must be evaluated through the performance of experiments and simulations with a variety of thermal–hydraulic codes. This paper presents the results which has been obtained for different SBLOCA break sizes with the best estimate TRACE V5.0 patch 2 thermal–hydraulic code and their comparison with those obtained by Westinghouse with NOTRUMP code. The main results show that TRACE code predicts a similar trend in all sequences with some differences that are expected to be an issue of the more conservative models and hypothesis assumed in the SBLOCA licensing analysis performed with NOTRUMP. Some particular characteristics of this reactor are also shown in this paper such as the importance of core makeup tank (CMT) and passive residual heat removal (PRHR) system for core cooling in case of small break sizes and the behavior of the plant in case of availability of half of the total passive safety injection systems which is the case of the double-ended direct vessel injection line break (DEDVI).

Comparison of integral thermal-hydraulic behaviors of a DVI line break SBLOCA with an equivalent cold leg break

An APR1400 (Advanced Power Reactor, 1400MWe) adopts a DVI (Direct Vessel Injection) method for ECC (Emergency Core Cooling) water delivery rather than a conventional CLI (Cold Leg Injection) method as an advanced safety feature. Several break scenarios of one DVI nozzle are taken into account in the safety analysis as one of the spectra of small break LOCAs. An 8.5-in. double-ended break of one DVI nozzle was simulated with ATLAS, and a counterpart test for the DVI break was performed at the cold leg with an equivalent break size for comparison. It was found from the counterpart tests and the analysis that the DVI break is more limiting than the corresponding cold leg break. This comparison will contribute to enhancing the comprehensive understanding of the thermal hydraulic behavior during transients. A constructed integral effect database can also be used to validate the existing conservative safety analysis methodology and to develop a best-estimate safety analysis methodology for small-break LOCAs. A post-test calculation was performed with a best-estimate safety analysis code, MARS-KS, to examine its prediction capability and to identify any code deficiencies for thermal hydraulic phenomena occurring during the transient.

ANALYSIS OF THE ISP-50 DIRECT VESSEL INJECTION SBLOCA IN THE ATLAS FACILITY WITH THE RELAP5/MOD3.3 CODE

The pressurized water reactor APR1400 adopts DVI (Direct Vessel Injection) for the emergency cooling water in the upper downcomer annulus. The International Standard Problem number 50 (ISP-50) was launched with the aim to investigate thermal hydraulic phenomena during a 50% DVI line break scenario with best estimate codes making use of the experimental data available from the ATLAS facility located at KAERI. The present work describes the calculation results obtained for the ISP-50 using the RELAP5/MOD3.3 system code. The work aims at validation and assessment of the code to reproduce the observed phenomena and investigate about its limitations to predict complicated mixing phenomena between the subcooled emergency cooling water and the two-phase flow in the downcomer. The obtained results show that the overall trends of the main test variables are well reproduced by the calculations. In particular, the pressure in the primary system show excellent agreement with the experiment. The loop seal clearance phenomenon was observed in the calculation and it was found to have an important influence on the transient progression. Moreover, the collapsed water levels in the core are accurately reproduced in the simulations. However, the drop in the downcomer level before the activation of the DVI from safety injection tanks was underestimated due to multi-dimensional phenomena in the downcomer that are not properly captured by one-dimensional simulations.

Counterpart Tests of SBLOCA on the Effect of Break Location with the ATLAS: DVI Nozzle versus Cold-Leg Breaks

Results from two integral effect tests were compared to discuss the effect of break location between the direct vessel injection (DVI) nozzle and cold leg during the small-break loss of coolant accident (SBLOCA) scenario. One is the SB-DVI-09 test for a 6-inch (50% break area of a DVI nozzle) DVI line break and the other is the SB-CL-06 test for an equivalent break size of cold leg. Both counterpart tests were performed with the same control logic and initial/boundary conditions except for different break locations of the DVI line and cold leg. Experimental results showed that the maximum heater surface temperature increased more with the broken DVI nozzle (SB-DVI-09) than with the broken cold leg (SB-CL-06) due to the delayed and simultaneous occurrence of the loop seal clearing and the momentary decrease in the collapsed water level in the core region.

Effects of Break Size on Direct Vessel Injection Line Break Accidents of the ATLAS

The direct vessel injection (DVI)-adopted power plant APR1400 considers a DVI line break among the analyzed small-break loss-of-coolant accidents in safety analysis. The first-ever integral effects test database for various DVI line break sizes from 5% to 100% was established with the Korea Atomic Energy Research Institute’s Advanced Thermal-Hydraulic Test Loop for Accident Simulation (ATLAS) test facility. This database enhances our physical understanding of the major thermal-hydraulic behaviors of the APR1400 during DVI line break accidents, and it can also be used to examine the prediction capabilities and identify any deficiencies in the existing best-estimate safety analysis codes. Effects of the break size were experimentally investigated, and the best-estimated MARS code was assessed against the experimental database. On the whole, the prediction of the MARS code shows a good agreement with the measured data. However, the code predicted a higher core level than the data just before a loop seal clearing occurs, and it also produced a more rapid decrease in the downcomer water level than the data. These disagreements are the expected consequence of uncertainties in predicting countercurrent flow or condensation phenomena in a downcomer region. The present integral effects test data will be used to support the present conservative safety analysis methodology and to develop a new best-estimate safety analysis methodology on the DVI line break accidents of the APR1400.

Counterpart Tests of SBLOCA on the Effect of Break Location with the ATLAS: DVI Nozzle versus Cold-Leg Breaks

Results from two integral effect tests were compared to discuss the effect of break location between the direct vessel injection (DVI) nozzle and cold leg during the small-break loss of coolant accident (SBLOCA) scenario. One is the SB-DVI-09 test for a 6-inch (50% break area of a DVI nozzle) DVI line break and the other is the SB-CL-06 test for an equivalent break size of cold leg. Both counterpart tests were performed with the same control logic and initial/boundary conditions except for different break locations of the DVI line and cold leg. Experimental results showed that the maximum heater surface temperature increased more with the broken DVI nozzle (SB-DVI-09) than with the broken cold leg (SB-CL-06) due to the delayed and simultaneous occurrence of the loop seal clearing and the momentary decrease in the collapsed water level in the core region.

AN EXPERIMENTAL STUDY WITH SNUF AND VALIDATION OF THE MARS CODE FOR A DVI LINE BREAK LOCA IN THE APR1400

In order to analyze thermal hydraulic phenomena during a DVI (Direct Vessel Injection) line break LOCA (Loss-of-Coolant Accident) in the APR1400 (Advanced Power Reactor 1400 MWe), we performed experimental studies with the SNUF (Seoul National University Facility), a reduced-height and reduce-pressure integral test loop with a scaled down APR1400. We performed experiments dealing with eight test cases under varied tests. As a result of the experiment, the primary system pressure, the coolant temperature, and the occurrence time of the downcomer seal clearing were affected significantly by the thermal power in the core and the SI flow rate. The break area played a dominant role in the vent of the steam. For our analytical investigation, we used the MARS code for simulation of the experiments to validate the calculation capability of the code. The results of the analysis showed good and sufficient agreement with the results of the experiment. However, the analysis revealed a weak capability in predicting the bypass flow of the SI water toward the broken DVI line, and it was insufficient to simulate the streamline contraction in the broken side. We, hence, need to improve the MARS code.

EXPERIMENTAL SIMULATION OF A DIRECT VESSEL INJECTION LINE BREAK OF THE APR1400 WITH THE ATLAS

The first-ever integral effect test for simulating a guillotine break of a DVI (Direct Vessel Injection) line of the APR1400 was carried out with the ATLAS (Advanced Thermal-hydraulic Test Loop for Accident Simulation) from the same prototypic pressure and temperature conditions as those of the APR1400. The major thermal hydraulic behaviors during a DVI line break accident were identified and investigated experimentally. A method for estimating the break flow based on a balance between the change in RCS inventory and the injection flow is proposed to overcome a direct break low measurement deficiency. A post-test calculation was performed with a best-estimate safety analysis code MARS 3.1 to examine its prediction capability and to identify any code deficiencies for the thermal hydraulic phenomena occurring during the DVI line break accidents. On the whole, the prediction of the MARS code shows a good agreement with the measured data. However, the code predicted a higher core level than did the data just before a loop seal clearing occurs, leading to no increase in the peak cladding temperature. The code also produced a more rapid decrease in the downcomer water level than was predicted by the data. These observable disagreements are thought to be caused by uncertainties in predicting countercurrent flow or condensation phenomena in a downcomer region. The present integral effect test data will be used to support the present conservative safety analysis methodology and to develop a new best-estimate safety analysis methodology for DVI line break accidents of the APR1400.

An integral loop test and MARS code analysis for a DVI line break LOCA in the APR1400

The transient thermal–hydraulic phenomena of a DVI (Direct Vessel Injection) line break LOCA (Loss-of-Coolant Accident) in pressurized water reactor, APR1400, were investigated. In order to understand the phenomena during the LOCA transient, a reduced-height and reduced-pressure integral loop test facility, the SNUF (Seoul National University Facility), was constructed with scaling down the prototype. For the appropriate test conditions in the experiment with the SNUF, the energy scaling method was suggested with scaling the coolant mass inventory and the thermal power for the reduced-pressure condition. According to the conditions determined by the method, the experimental study was performed with the SNUF. The experimental results showed that the phenomenon of the downcomer seal clearing played a dominant role in the reduction of the system pressure and the recovery of the coolant level in the core. That phenomenon occurred when the steam incoming from cold legs penetrates the coolant in the upper downcomer toward the broken DVI line. The experimental results were compared with the prototype analysis to estimate the energy scaling method, so that the experiment reasonably simulated the phenomena in the prototype. For the analytical investigation, the experiment was simulated with MARS code to validate the calculation capability of the code, especially for the downcomer seal clearing, which showed good agreement with the results of experiment.

Commissioning of the ATLAS thermal-hydraulic integral test facility

KAERI recently constructed a new thermal-hydraulic integral test facility for advanced pressurized water reactors (PWRs) – ATLAS. The ATLAS facility has the following characteristics: (a) 1/2-height&length, 1/288-volume, and full pressure simulation of APR1400, (b) maintaining a geometrical similarity with APR1400 including 2(hot legs) × 4(cold legs) reactor coolant loops, direct vessel injection (DVI) of emergency core cooling water, integrated annular downcomer, etc., (c) incorporation of specific design characteristics of OPR1000 such as cold leg injection and low-pressure safety injection pumps, (d) maximum 10% of the scaled nominal core power. The ATLAS will mainly be used to simulate various accident and transient scenarios for evolutionary PWRs, OPR1000 and APR1400: the simulation capability of broad scenarios including the reflood phase of a large-break loss-of-coolant accident (LOCA), small-break LOCA scenarios including DVI line breaks, a steam generator tube rupture, a main steam line break, a feed line break, a mid-loop operation, etc. The ATLAS is now in operation after an extensive series of commissioning tests in 2006.