Annals of Nuclear Energy 38 (2011) 2213–2217

Contents lists available at ScienceDirect

Annals of Nuclear Energy

journal homepage: www.elsevier .com/locate /anucene

BWR fuel rod behavior evaluation for preconditioning power ramps with FEMAXI-V

Hector Hernandez Lopez ⇑, Marco A. LucateroInstituto Nacional de Investigaciones Nucleares, Carr. Mexico-Toluca s/n (km 36.5), La marquesa, Ocoyoacac, Mexico 52750, Mexico

a r t i c l e i n f o a b s t r a c t

Article history:Received 14 January 2011Received in revised form 8 June 2011Accepted 10 June 2011Available online 8 July 2011

Keywords:BWRFuel rodThermomechanical behaviorPCIPreconditioning procedures

0306-4549/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.anucene.2011.06.009

⇑ Corresponding author. Tel.: +52 55 53297200x24E-mail addresses: hector.hernandez@inin.gob.mx (

o@inin.gob.mx (M.A. Lucatero).

The licensing authorities around the world usually set a limit value to the operation LHGR, as a functionof burnup. Such limit provides a bound state to a steady state operation, but also prevents against somethermal and mechanical phenomena that could occur during overpowered transients. In particular, insome countries, the PCI limit is set based on experimental ramp tests and directly related to the LHGRlimit value. Thus, to avoid violating the PCI limit, fuel conditioning procedures are still required for bothbarrier and non-barrier fuel. Simulation of the power ramp procedures to be performed by the reactoroperator during startup or power increase maneuvers is advisable as a preventive measure of possibleoverpower consequences on the fuel thermomechanical behavior.

The thermomechanical behavior of BWR fuel rod is analyzed for fuel preconditioning procedures. Fivedifferent preconditioning computations were performed with the FEMAXI-V code, each with three differ-ent ascending linear power rate ramps. The starting point of the ramps was taken from data of the Unit 1from the Laguna Verde Nuclear Power Plant, located in MEXICO. The top limit of the ramps was thethreshold linear power at which failure by PCI could occur, as a function of burnup.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Because of safety and economic reasons, setting adequate ther-mal and mechanical limits for the operation of a nuclear powerreactor depends on several aspects, as reactor type, fuel rod com-position, power generated, etc. Particularly, power increase duringreactor startup, or because higher demand of electrical power tothe network, makes mandatory to consider the changes in the ther-momechanical properties of the specific fuel rod type used in a fuelassembly. For example, in a BWR, the density change of the coolantand moderator along a fuel pin causes different thermomechanicalstresses and oxidation levels of the fuel rod cladding at the differ-ent axial sections of the fuel pin.

Although the number of failed rods is still low, in comparison tothe number of fuel elements in the core of the operating power nu-clear reactors, the recent advantages in competitiveness of nuclearenergy can be challenged by public opinion, and thus forcing theregulatory entities to restrict the use of the new core managementstrategies, particularly on power peaking and linear heat genera-tion rate (LHGR) operation limits.

An increasing emphasis in economic revenue has leaded theutilities to apply for or perform power up-rates, pursue longeroperating cycles, and introduce innovative fuel reload patterns.These current more aggressive operation strategies have as result

ll rights reserved.

61; fax: +52 55 53297301.H.H. Lopez), marcoa.lucater-

improved plant capacity factors, leading the nuclear industry toreach its lowest electricity production costs in many countries. Inthe last few years, however, the BWR fuel failure rate has pre-sented a new and noticeable increase (Yang et al., 2004). The causeis considered to be a combination of very diverse areas, as waterchemistry, new cladding materials and manufacturing procedures,and higher fuel duty.

During normal steady state operation of a nuclear power reac-tor, the gap and fuel thermal conductivities are the main physicalproperties dominating the thermal behavior of a fuel rod. On theother hand, during transient events, the heat capacity of the fuelis the ruling physical property of the thermal behavior. In bothcases, if a fast power increase occurs, thermal expansion of the fuelpellet could lead to pellet–cladding interaction, which is a primarytype of defect that could lead to further clad degradation, andeventually cause clad failure. If the power ramp rate to whichthe fuel rod is subjected is appropriately limited, the dimensionalchanges of the fuel pellet and cladding may be moderate, and thuscreep, and relaxation can alleviate the consequences of PCI mech-anism. Appropriate slopes for the power ramps can be outlinedfrom the results of thermomechanical fuel behavior computercodes.

Fuel conditioning is the physical mechanism that includes allthe local thermomechanical phenomena that help limiting the con-sequences of power transients in fuel elements. Fuel densificationand stress relaxation are examples of the physical phenomenaoccurring during fuel conditioning that reduces the contact pres-sure between the fuel pellet and cladding and reopens the gap. Fuel

Table 1Fuel rod specifications and test conditions.

Parameter Value

CladdingO.d. (cm) 1.0262I.d. (cm) 0.8941Material Zircaloy 2

Fuel pelletD (cm) 0.8763Height/diameter ratio 1% TD (UO2) 96.5

RodPlenum volume (cm3) 1.08Fill gas initial pressure (MPa) 1.013Active length (cm) 381

Test system conditionsCoolant inlet temperature (K) 560Reactor pressure (MPa) 7.14Coolant mass flux (kg/cm2 s) 0.166

Fig. 1. Axial distribution enrichment of 235U in fuel rod.

2214 H.H. Lopez, M.A. Lucatero / Annals of Nuclear Energy 38 (2011) 2213–2217

conditioning may take from hours to days. Once equilibrium be-tween cladding creep and pellet swelling is reached, a new steadystate condition at a higher power level is established. Fuel de-con-ditioning, contrary to fuel conditioning, is the phenomena thataggravate PCI, such as fuel swelling, by increasing the contact pres-sure and reducing the gap size.

The term conditioning power level is defined as the rod powerlevel at a typical reference stress, when cladding creep and pelletswelling equilibrate each other (Ito et al., 1983). This conditioningpower level is also known as the conditioning LHGR, which is thelimit where neither conditioning nor de-conditioning occurs(NEA, 2003). That is, contact pressure between pellet and claddingis moderate and constant. If fuel conditioning occurs, then the rodpower level needs to increase to reach the conditioning LHGR. Onthe contrary, if fuel de-conditioning occurs, the rod power levelneeds to decrease to reach the conditioning LHGR. Once a newsteady state is reached, the conditioning LHGR asymptotically ap-proaches the current value of the fuel rod LHGR.

One measure normally taken in a nuclear power reactor opera-tion to avoid the failure mechanism due to PCI is to establish a pro-cedure to limit the number and types of sudden power increasesthat could reach the levels at which clad failure by PCI occurs.Many countries still require using such procedures. This is so be-cause it is necessary to moderate the consequences of the fuel con-ditioning and de-conditioning phenomena described above. Theoperational procedures used to reduce the probability of such typeof clad failure are known as fuel preconditioning operations. Thepreconditioning is a controlled and constant power increase thatfollows a previously set ascending ramp. This process is consideredat nodal level, and not the average power of the whole fuel rod.

The initial point of the ramp is a reference power level, and theending point is the nominal power at which reactor operation isdesired. Preconditioning rules are normally applied during reactorstartup or after a control rod blade pattern change. By following anappropriate preconditioning power ramp the possibility of fueldamage is greatly reduced, and it also helps the fuel to betterassimilate further and faster power changes, below the precondi-tioned envelope. However, even at this controlled conditions, it isnecessary to perform thermomechanical analysis of the fuel rodsto ensure that clad failure by PCI will not occur during the precon-ditioning action, or to determine the linear heat generation rate va-lue at which the failure could occur.

In this paper, the thermomechanical behavior of BWR fuel rod isanalyzed for fuel preconditioning procedures. Five different pre-conditioning computations were performed with the FEMAXI-Vcode (Suzuki, 2000; Suzuki and Saitou, 2001), each with three dif-ferent ascending linear power rate ramps. The starting point of theramps was taken from data of the Unit 1 from the Laguna VerdeNuclear Power Plant. The top limit of the ramps was the thresholdlinear power at which failure by PCI could occur, as a function ofburnup.

2. Fuel rods description

The fuel rod has natural uranium at both top and bottom ex-tremes. While in the middle part, the fuel rod has two regions with235U enrichment of 4.90 w/0 and 4.40 w/0, respectively. The total ac-tive length of fuel rod was 381.0 cm. Table 1 presents the designdimensions for fuel rod. Also, shown the test conditions typicalof BWRs. Fig. 1 shows the axial distribution of the 235U enrichmentfor fuel rod.

The FEMAXI-V geometry model used consisted of 10 axialnodes, 10 radial segments in the fuel pellet, one for the gap, andtwo segments for the cladding. Although the maximum numberof axial nodes allowed by FEMAXI-V is 12, only 10 were used

because this is the same number of axial nodes used in theRODBURN (Uchida and Saito, 1993; RODBURN, 1999) code forthe computation of the power distribution. The top and bottomnodes represented the natural uranium areas of the actual fuelrods. The middle nodes were all assumed to have the average235U enrichment corresponding to the fuel rod.

3. Computation procedure

For the computations, it was firstly assumed that the fuelassemblies containing the fuel rod were the hottest assemblies inthe core (the hot channel). The considered reactor operating condi-tions corresponded to the nominal steady-state operation, see Ta-ble 1.

Fuel preconditioning is burnup dependent, since the thresholdlinear power for possible clad failure also changes as a functionof burnup. A threshold power is thus previously set, as the

Table 2Threshold power as function of burnup for the fuel rod.

Nodal exposure(MWD/TU)

Threshold power(kW/ft)

0–3700 133701–5000 11.55001–6400 106401–7800 9.57801–65,915 9.5

H.H. Lopez, M.A. Lucatero / Annals of Nuclear Energy 38 (2011) 2213–2217 2215

maximum power at which a fuel element can (conservatively)operate. In this work, the starting point of the linear power in-crease ramps were set to 1 kW/ft below the preconditioningthreshold power. The calculations are performed for the fuel pinnode with the highest power peaking factor. The axial power peak-ing factor values used was 1.81 for the fuel rod. The values are highbecause they also consider some Anticipated Operational Occur-rences. The threshold powers for fuel rod considered in this workare shown in Table 2, for five burnup ranges. These values can beconsidered typical of fuel 10 � 10. Three rates for the power ramps

Fig. 2. Gap width behavior appl

Fig. 3. LHGR at contact fo

were considered at each burnup interval: 0.11 kW/ft, 0.22 kW/ft,and 0.33 kW/ft. No Xenon (Xe) redistribution was considered inthe computations. Each axial node in FEMAXI-V had a differentpower peaking factor, to realistically represent the BWR core axialpower distribution.

4. Results

For the fuel rod, the initial power average of 12.0 kW/ft and ini-tial average exposure between 0 and 3700 MWD/TU for the fuelpellet, shows that when reaching 13 kW/ft of power preconditionalthreshold with the power ramp, it does not have pellet–claddingcontact. For the initial power average of 10.5 kW/ft and averageinitial exposures between 3701 and 5000 MWD/TU of the fuel pel-let, it shows that when reaching the power preconditional thresh-old of 11.5 kW/ft with the power ramp, it does not have pellet–cladding contact.

For the initial power average of 9.0 kW/ft and initial exposureaverage between 5001 and 6400 MWd/TU of the fuel pellet, itshows that when reaching the power preconditional threshold of

ied for three power ramps.

r three power ramps.

Fig. 4. Maximum temperature at contact for power ramps.

Fig. 5. Pressure applied for pellet at contact for power ramps.

2216 H.H. Lopez, M.A. Lucatero / Annals of Nuclear Energy 38 (2011) 2213–2217

10.0 kW/ft with the power ramp, it does not have pellet–claddingcontact. For the initial power average of 8.5 kW/ft and initial expo-sure average between 6400 and 40,000 MWd/TU of the fuel pellet,it shows that when reaching the power preconditional threshold of9.5 kW/ft with the power ramp, it does not have pellet–claddingcontact. In other words, the fuel rod, not having contact in eithercase, is observed in Fig. 2.

Since contact between pellet and cladding did not appear whenthe limits of LHGR were reached, thresholds were determined bythe manufacturer, and the power ramp was applied until reachingcontact between pellet and the cladding, observing that contact oc-cur when power has been increased 35% over the established limit,Fig. 3.

When establishing the LHGR where the pellet–cladding interac-tion appears, the maximum temperatures that appear in the fuelcenterline were calculated and it was determined that in the threepower ramps for the initial exposure average between 0 and3800 MWD/TU, the maximum temperature was approximately1980 �C, Fig. 4.

Fig. 5 shows the behavior of pressure applied for the pellet on thecladding at time the pellet–cladding interaction appears under theconditions established by the power ramps, determining the initialexposure average between 4000 and 7000 MWD/TU and both powerranges thresholds 10 and 11.5 kW/ft for the ramp of 0.33 kW/ft/h,and thus, the pressures greater than 6 MPa are reached. While forthe initial exposure average between 7000 and 8000 MWD/TU andpower threshold of 9.5 kW/ft, pressures greater than 8 MPa arereached, and they exceed the reactor operating pressure.

5. Conclusions

The results show that fuel rod does not present pellet–claddinginteraction for the preconditioning procedures, using the three dif-ferent power ramp slopes considered in this study. However, onceit reaches the pellet–cladding interaction, seen pressures over therange of reactor operating pressure, especially for initial averageexposure at 7000–8000, when have been applied the power rampof 0.33 kW/ft/h.

H.H. Lopez, M.A. Lucatero / Annals of Nuclear Energy 38 (2011) 2213–2217 2217

These results show an ample operational LHGR margin in steadystate reactor condition to accommodate sudden power excursionsthat must be limited to have power ramps below 0.33 kW/ft/h. How-ever, further investigations must be done in transient conditions toevaluate how much LHGR operational limits can be relaxed due tothe new core management strategies to deliver more electricity.

References

Ito, K., Ichikawa, M., Okubo, T., Iwano, Y., 1983. FEMAXI-III a computer code for fuelrod performance analysis. Nuclear Engineering and Design 76, 3–11.

Suzuki, M., 2000. Analysis of high burnup fuel behavior in Halden reactor byFEMAXI-V Code. Nuclear Engineering and Design 201, 99–106.

NEA/Committee on the Safety of Nuclear Installations, 2003. Fuel Safety Criteria inNEA Countries. NEA/CSNI/R(2003)10.

RODBURN Ver. 1.2. 1999. Input manual [in English].Suzuki, M., Saitou, H., 2001. Light water reactor fuel analysis code FEMAXI-V. JAERI-

Data/Code.Uchida, M., Saito, H., 1993. RODBURN: A Code for Calculating Power Distribution in

Fuel Rods, JAERI-M. pp. 93–108 (in Japanese).Yang, R., Ozer, O., Edsinger, K., Cheng, B., Deshon, J., 2004. An integrated approach to

maximizing fuel reliability. In: 2004 International Meeting on LWR FuelPerformance. Orlando, FL, September 19–22, 2004.