ML25350C333

Task 1 - Realistic Estimates of Fuel Dispersal from High Burnup PWR Cores

ML25350C333
Person / Time
Issue date:	07/07/2025
From:	James Corson, Joseph Staudenmeier NRC/RES/DSA
To:
References
Download: ML25350C333 (0)

Text

REALISTIC ESTIMATES OF FUEL DISPERSAL FROM HIGH BURNUP PWR CORES James Corson RES/DSA/FSCB Joseph Staudenmeier RES/DSA/CRAB I July 7, 2025 1

INTRODUCTION The NRC staff is currently reviewing the Electric Power Research Institute (EPRI) Alternative Licensing Strategy (ALS) topical reports, which utilize a combination of probabilistic fracture mechanics and leak-before-break (LBB) to assert that large break LOCAs are so unlikely to occur that the effects of Fuel Fragmentation, Relocation, and Dispersal (FFRD) do not need to be considered, consistent with how certain phenomena are treated by the exception within GDC 4 for LBB. The NRC staff is considering an alternative risk-informed regulatory approach which would benefit from a reasonable technical evaluation of the impact of FFRD on the LOCA event progression and outcomes. This information may be able to support regulatory findings that FFRD does not need to be explicitly analyzed up to a certain point based on the fact that the impact would not be significant. This technical basis could be used in similar risk-informed approaches for regulatory findings on other approaches to disposition FFRD for LOCA analyses.

The Office of Nuclear Regulatory Research (RES) is supporting this alternative risk-informed effort by performing a series of analyses intended to address certain technical issues associated with fuel fragmentation, relocation and dispersal (FFRD) of high burnup/extended increased enrichment (HBU/IE) nuclear fuel due to large-break loss of coolant accidents (LOCAs) per a request from the Office of Nuclear Reactor Regulation (NRR). Task 1 requests that RES develop a range or upper bound on realistic quantities of FFRD that would be expected for higher burnup cores during a large-break LOCA.

This report is organized as follows:

Section 2 briefly describes the computational methods used in this study.

Section 3 presents the results of the TRACE and FAST calculations.

Section 4 describes the conclusions and recommendations of the analytical work.

2 DESCRIPTION OF MODELING APPROACH This analysis used the Workflow for Holistic Accident Multiphysics (WHAM) methodology described in Ref. [1], together with a high burnup, increased enrichment core design developed by a utility for a 4-loop Westinghouse plant [2]. The following paragraphs briefly describe the methodology and the core design, but interested readers should refer to Refs. [1, 2] for further information.

Overview of WHAM WHAM combines NRC-sponsored neutronics, thermal hydraulics, and fuel performance codes to simulate the steady-state and transient behavior of the complete reactor system. Note that the codes are coupled through input and output; that is, the output of one code often provides the boundary conditions for another code, as summarized below and represented graphically in Figure 1. The workflow is facilitated using several python scripts that read output and write input files. The coupling is only one-way. This means that fuel rod deformation calculated by FAST during the transient is not seen by TRACE. This may have some impact on the results, particularly following fuel rod ballooning and burst, though it is not expected to change the conclusions reached in this study regarding estimated quantities of dispersed fuel.

Figure 1: WHAM workflow; blue boxes represent assembly-level calculations, while orange boxes represent pin-by-pin calculations (reproduced from Ref. [1])

Fuel Loading Pattern Characteristics For this study, we performed calculations for a 3,626 MWt Westinghouse 4-loop plant using a core design described in Ref. [2]. The 24-month equilibrium cycle alternates between 84 and 85 Westinghouse VANTAGE+ feed assemblies with enrichments of 5.95, 6.2, or 6.6 weight percent U-235. Our calculations used the 85-feed loading pattern, which is shown below. The loading

pattern contains significant amounts of IFBA and WABA to counteract the excess reactivity and to limit the soluble boron concentration at beginning of cycle. The IFBA rods include annular blankets to provide additional rod void volume to limit end-of-life rod internal pressures.

However, as we will show later, the IFBA rods still tend to have higher end-of-cycle pressures than the non-IFBA rods, which impacts the fuel rod ballooning and burst behavior.

Figure 2: Quarter-core representation of the fuel loading pattern, with feed assemblies in red, once-burned assemblies in green, and twice-burned assemblies in blue; enrichments in the top-left corner of each assembly; number of IFBA and WABA rods in the top middle and top right; and beginning-and end-of-cycle assembly-average burnup at the bottom (from Ref. [1])

3 RESULTS TRACE Results The TRACE calculations used a detailed Vogtle Class Westinghouse 4 loop plant model with a cartesian core VESSEL component with bundle level resolution placed within cylindrical VESSEL component. The model also used a CONTAN model to get containment pressure feedback. The model is shown in Figure 3. Two different power shapes using the piecewise linear power shape input option were used in these calculations. One was a based on a chopped cosine with an axial peaking factor of 1.5002 and the other was a power shape from a PARCS core depletion calculation and had power peaks near the bottom and top of the core.

The power shapes are shown in Figure 4. Calculations were performed with and without the

loss of offsite power (LOOP). A loss of offsite power delays the start of the pumped injection by 30 seconds but does not significantly affect the results because the accumulator flow is larger than the pumped flow and is large enough to refill the lower plenum and achieve bottom of core recovery (BOCREC) by approximately 30 seconds. Much of the water is lost through the break before the vessel pressure comes into quasi-equilibrium with the containment pressure.

Figure 3 Hydraulic component view of TRACE model used in FFRD calculations.

The locations of the cladding ruptures are concentrated near the peaks in the axial power profiles so the location of the ruptures (shown in Figure 8) in the PARCS power shape case are higher in the core than for the chopped cosine case because the peak in the power shape is at a higher elevation in the core. Only a small difference between the PCT curves between the delayed (LOOP) pumped ECCS injection and un-delayed ECCS injection results is seen in the

chopped cosine power shape case shown in Figure 5. It should be noted that the PCT in the plots is the maximum cladding surface temperature in the core at a given time. The location of the PCT can move in time. This is because the accumulator injection always comes in without delay and is enough to fill the lower plenum and downcomer and the time of point of bottom of core recovery (BOCREC) is approximately the same for both cases.

Figure 4. Power shapes used in the FFRD calculations.

Figure 5 PCT for delayed (LOOP) and un-delayed pumped ECCS injection.

The power shape and peaking have a bigger impact. Figure 6 shows the results for the chopped cosine shape and the PARCS shape. It should be noted that the highly peaked chopped cosine is a bounding power shape used in ECCS licensing calculations and is not realistic for normal steady-state operations.

Figure 6 PCT for the chopped cosine and PARCS power shapes.

The number of trains of pumped ECCS injection has a significant impact on the temperatures in the long term as is shown in Figure 7 for the case of the PARCS power shape. Two trains of pumped ECCS results in a significantly lower PCT and will lead to fewer cladding failures. This will be seen in the FAST results in the section on FAST results that follows.

Figure 7 Effect of the number of pumped ECCS trains on the PCT.

FAST Results Using the thermal hydraulic boundary conditions from TRACE, we simulated the behavior of every rod in the core for several combinations of axial power profile (either chopped cosine with a peaking factor of 1.5 or the end-of-cycle profile calculated by PARCS), ECCS availability (one

or two trains), and offsite power availability. In addition, we looked at three different fuel dispersal models:

Model C from RIL 2021-13: All fuel disperses through the burst opening from contiguous nodes with burnup > 55 GWd/MTU and hoop strain > 3%.

Model A from RIL 2021-13: Fuel disperses through the burst opening from contiguous nodes with burnup > 55 GWd/MTU and hoop strain > 3%. For this case, the amount of fuel that disperses from a given node increases linearly from 0% at 55 GWd/MTU to 100% at 80 GWd/MTU.

Model A, but accounting for the impact of spacer grids: This case is the same as the model A case, but the amount of fuel susceptible to dispersal is limited to the grid span containing the burst. Note that the distance between grid spans in the upper half of the assembly is less than 12 inches for the fuel assembly design used in these calculations.

Table 1 shows the results of the FAST calculations for the four sensitivity parameters (number of ECCS trains, axial power profile, offsite power availability, and dispersal model). From the results, it is clear that offsite power availability has only a minor impact on peak cladding temperature, rod ballooning and burst, and fuel dispersal, as explained in the previous section on the TRACE results.

On the other hand, the number of ECCS trains, the axial power profile, and the dispersal model significantly impact the amount of fuel dispersal. Interestingly, the chopped cosine distribution results in a lower percentage of burst rods (both total and second cycle only), but it results in more fuel dispersal. The lower burst percentage may be due to the quench front reaching the peak power node in the cases with the chopped cosine power shape earlier than it reaches the top peak in the best-estimate end-of-cycle power shape calculated by PARCS.1 At the same time, the higher peaking factor in the chopped cosine case results in a higher PCT, so there are competing phenomena impacting the burst behavior. With that said, the dispersed fuel mass using RIL model A is more than a factor of two greater for the chopped cosine cases because the bursts happen lower in the bundle, where the burnup is higher, as illustrated in Figure 8 (burst node) and Figure 9 (axial burnup profile). For example, Rod 72/15/16 bursts in all cases; however, it only disperses fuel in the chopped cosine cases because the burnup at the burst location is slightly above the dispersal threshold, whereas the burnup at the burst location is below the dispersal threshold for the PARCS axial power case.

1 Recall that the PARCS power shape is flatter than the chopped cosine shape and features peaks near both the bottom and top of the fuel rod. The top peak has a lower magnitude than the bottom peak.

Table 1: FAST Results PARCS

Power, Offsite Power Available, 2 ECCS Trains PARCS

Power, Offsite Power Available PARCS

Power, Offsite Power Unavailable Chopped

Cosine, Offsite Power Available Chopped

Cosine, Offsite Power Unavailable FAST Peak Cladding Temperature (oC) 774 816 834 852 863 Burst Rods

(% of total) 25 49 55 44 49 Second Cycle Burst Rods (% of total) 24 50 58 38 41 Dispersed Mass (RIL Model C) (kg UO2) 1300 2000 2100 3400 3700 Dispersed Mass (RIL Model A) (kg UO2) 700 940 980 2300 2500 Dispersed Mass (RIL Model A, single grid span) (kg UO2) 380 530 540 1400 1500

Figure 8: Number of high burnup rod bursts (burst node burnup > 55 GWd/MTU) as a function of elevation for the PARCS (top) and chopped cosine (bottom) power profiles; offsite power and one ECCS train are available in both cases2 2 Note that the bursts in node 4 for the PARCS power case occur at the lower peak in the axial power profile, early in the transient, during the stored energy heatup phase.

Figure 9: End-of-cycle axial burnup profiles for three sample rods, identified by assembly number / row / column Note that the results are heavily influenced by the loading pattern developed by Southern Nuclear Operating Company. The loading pattern has a significant number of fuel assemblies in the interior of the core that are well above the fuel fragmentation threshold at the end of life.

Compare this to existing fuel loading patterns, which typically have few (if any) high burnup assemblies in the core interior. (For example, see Ref. [3].) This means that a large number of assemblies are potentially susceptible to fuel dispersal.

The high burnup assemblies in the core interior also have high linear heat rates throughout the second two-year operating cycle. This leads to significant fission gas release and very high predicted end-of-cycle rod internal pressures, which are shown in Figure 10 for both burst and non-burst rods. As noted, the high rod internal pressures also contribute to the high predicted burst fraction, though of course there are other important parameters that influence burst (e.g.,

end-of-cycle linear heat generation rates and flow patterns during the LOCA, which impact the cladding temperature response). A core loading pattern featuring gadolinia instead of IFBA would likely have lower end-of-cycle rod internal pressures and a potentially lower cladding burst fraction during the LOCA. Furthermore, utilities and vendors may come up with core loading patterns that better manage rod internal pressure concerns, leading to lower predicted dispersed fuel mass during a LBLOCA. On the other hand, modern cladding designs may be able to handle the high rod internal pressures predicting here without experiencing liftoff during normal operations, such that future loading patterns could result in predicted rod internal pressures and fuel dispersal masses similar to those in this study. Vendors may also choose other strategies that have not been considered here that could significantly impact the rod internal pressures, which would in turn impact the amount of fuel dispersal.

Figure 10: End-of-cycle rod internal pressures calculated by FAST for rods predicted to burst (top) and to remain intact (bottom) in the PARCS power, offsite power available case The fuel dispersal calculations are also heavily influenced by the cladding ballooning and burst models in FAST. Figure 11 compares the burst temperatures calculated by FAST to the Chapman correlation and the historical data on which it is based (Ref. [4]; see Figure 12 below).

For rods with burst stresses below approximately 12 kpsi (corresponding to a rod internal pressure of approximately 10 MPa), FAST is underpredicting the burst temperature when compared to the Chapman correlation. These lower pressure rods tend to be first cycle rods or lower power (and thus, lower burnup) second cycle rods, which are much less susceptible to fine fragmentation and dispersal. The FAST predictions tend to overpredict the burst temperature for rods with a burst stress greater than 12 kpsi, which tend to be the rods more susceptible to fuel dispersal. However, the predicted burst temperatures for almost all of these lower pressure rods are below 700ºC, which is the minimum burst temperature in the historical data on which the Chapman correlation is based. Unfortunately, there are very few data points representing failure stresses above 10 kpsi and low ramp rates similar to the average heatup rate at failure predicted by FAST, so it is difficult to assess whether or not the low temperature (<

700ºC) failure predictions are accurate.

It must also be noted that there is considerable scatter in the historical data on which the Chapman correlation is based [4], and in cladding burst testing in general. Furthermore, FAST accounts for irradiation damage and hydrogen uptake in its mechanical models, which influence ballooning and burst behavior, and that the temperature profiles calculated based on the TRACE thermal hydraulics are more complicated than the linear temperature ramps used for the historical tests describes in NUREG-0630 [4] or in more recent in-pile and out-of-pile testing at Halden, Studsvik, and ORNL [5]. This makes it difficult to assess whether the lower burst temperature predicted by FAST compared to the historical data is a result of code biases or whether it is due to parameters (e.g., hydrogen and irradiation effects) that were not included in the historical tests. FAST does tend to predict earlier burst for the small number of LOCA tests in the FAST assessment database, but this may be due in part to challenges with predicting the pressure and temperature in the non-prototypical fuel rod plena in these tests [6].

Figure 11: Cladding burst temperatures predicted by FAST for the PARCS power, offsite power available case, compared to the Chapman correlation

Figure 12: Comparison of the Chapman correlation to historical burst test data (reproduced from Fig. 3 of NUREG-0630 [4])

4 CONCLUSIONS The TRACE and FAST calculations demonstrate that the assumption of loss of offsite power at the start of the LOCA has little impact on the peak cladding temperature or dispersed fuel mass.

On the other hand, the axial power profile has a more pronounced impact: the chopped cosine power profile results in a higher PCT and more fuel dispersal than the best-estimate end-of-cycle power profile calculated by PARCS. In the chopped cosine power profile cases, the bursts tend to occur lower in the bundle, where burnup is slightly higher and the distance between grid spans is greater than higher in the bundle, where most of the bursts occur in the PARCS power profile cases. ECCS availability also has a significant impact: the case with two trains of ECCS available results in about half the cladding bursts and about 75% of the fuel dispersal of the case with a single train available.

However, the parameter explored in this study that has the greatest impact on the dispersed fuel mass is the fuel dispersal model itself. The most conservative model from RIL 2021-13 results in over 2% (> 1,300 kg) of dispersed fuel in these cases, whereas the less conservative model used here (i.e., model A within a single grid span) in a much smaller dispersed fuel mass (as low as 0.37%, or 380 kg) for the PARCS power profile cases.

The latter calculations (i.e., with the less conservative dispersal model, PARCS power profile, two trains of ECCS, and onsite power available) provide a best estimate for the amount of dispersed fuel for this particular loading pattern at end-of-cycle. Note that the dispersed mass would likely increase with the time in cycle, as the fuel accumulates more burnup, so in that sense it is conservative to only consider the end of cycle.

There are several phenomenological uncertainties that we have not accounted for here. For one thing, the available experimental data suggest that the least conservative dispersal model used here (i.e., model A within a single grid span) may still be conservative; indeed, the model tends to slightly overpredict the dispersed mass from the SCIP-III tests, and more recent tests in the SATS facility at ORNL show even less fuel dispersal during the tests than was observed in SCIP-III. There is also considerable debate about the representativeness of the single-rod tests conducted with ~30-cm refabricated rod segments to full-length rods in a bundle geometry; about the applicability of thermal-hydraulic boundary conditions used in the Halden, ORNL, and SCIP tests to predicted thermal hydraulic behavior during a LBLOCA; and about the differences in the results produced at Studsvik and at ORNL.

Furthermore, there are uncertainties in the fuel rod ballooning and burst models used in FAST.

As noted, FAST is predicting burst temperatures below 700ºC for many high burnup rods. The historical data does not include any tests with such low burst temperatures. At the same time, the historical tests were performed on as-fabricated cladding tubes, not on high burnup rods that have accumulated irradiation damage and absorbed hydrogen due to water-side corrosion during normal operations.

Perhaps more importantly, the core design used in this study includes a large number of rods with very high rod internal, due in part to the very high IFBA loading. The high rod internal pressures lead to a significant number of fuel rods ballooning and bursting during the accident.

Furthermore, the loading pattern includes a large number of rods in the interior of the core with burnup above the fine fragmentation threshold. Together, these features of the loading pattern would result in significantly more fuel dispersal than would be expected for existing fuel loading patterns with lower discharge burnups. This may or may not be representative of future high burnup core loading patterns.

Note that all of these issues are under active investigation: researchers at ORNL and Studsvik are discussing possible mechanisms that could contribute to different fine fragmentation behavior, INL plans to conduct several LOCA tests in TREAT to determine how a more prototypic test could influence fragmentation and dispersal, the vendors are working on optimizing their high burnup core loading patterns, and NRC is sponsoring work at PNNL to evaluate the high temperature mechanical models in FAST. RES staff will keep NRR staff informed of the latest developments in these areas.

5 References

[1] A. Bielen, J. Corson and J. Staudenmeier, "NRC's methodology to estimate fuel dispersal during a large break loss of coolant accident," Nuclear Engineering and Design, vol. 426, p.

113377, 2024.

[2] N. Capps, A. Wysocki, A. Godfrey, B. Collins, R. Sweet, N. Brown, S. Lee, N. Szewczyk and S. Hoxie-Key, "Full Core LOCA Safety Analysis for a PWR Containing High Burnup Fuel (ORNL/TM-2020/1700)," Oak Ridge National Laboratory, Oak Ridge, TN, 2020.

[3] P. Raynaud and I. Porter, "Predictions of Fuel Dispersal During a LOCA," in Proceedings of the Water Reactor Fuel Performance Meeting, Sendai, Japan, 2014.

[4] D. Powers and R. Meyer, "NUREG-0630, Cladding Swelling and Rupture Models for LOCA Analysis," U.S. Nuclear Regulatory Commission, Washington, DC, 1980.

[5] M. Bales, A. Chung, J. Corson and L. Kyriazidis, "RIL 2021-13, Interpretation of Research on Fuel Fragmentation, Relocation, and Dispersal at High Burnup," U.S. Nuclear Regulatory Commission, Washington, DC, 2021.

[6] K. Geelhood, D. Colameco, W. Luscher, L. Kyriazidis, C. Goodson, J. Corson and J.

Whitman, "PNNL-35703, FAST-1.2.1: Integral Assessment," Pacific Northwest National Laboratory, Richland, WA, 2024.