ML19340B046
| ML19340B046 | |
| Person / Time | |
|---|---|
| Site: | Dresden  |
| Issue date: | 02/22/1974 |
| From: | US ATOMIC ENERGY COMMISSION (AEC) |
| To: | |
| Shared Package | |
| ML19340B045 | List: |
| References | |
| NUDOCS 8010170755 | |
| Download: ML19340B046 (6) | |
Text
_
UNITED STATES ATOMIC ENERGY C0btfISSION SAFETY EVALUATION BY W E DIRECTORATE OF LICENSING DOCKET NO. 50-10 EFFECTS OF FUEL DENSIFICATION Introduction On November 2,1972 the Staff requested from Commonwealth Edison infomation on the effects of fuel densification on the operation of the Dresden-1 nuclear power plant.
On February 27, 1973 the Staff received from Commonwealth Edison the reply j
to this inquiry.
In addition, the Staff also began discussions of a generic nature with the fuel supplier, Gulf United Nuclear Fuels Corporation (GUNFC).
Tnese discussions dealt with the models for fuel thermal perfomance and cladding collapse described in the GUNFC topical report GU-5300(1) and later in G' -5300 Revision 1(2). The results of the Staff evaluation of these two J
GUNFC topical reports are presented in the Staff report ' Technical Report on Densification of Gulf United Nuclear Fuels Corporation Fuels for Light Water Reactors" dated November 21, 1973.
The phenomenon of fuel densification and its possible effects on reactor fuel are described in detail in references (3) and (4). Briefly, these effects are:
-(1) The size of the gap between the fuel pellet and the cladding will increase due to a decrease in the diameter of the fuel pellet. Tnis tends to decrease the gap conductance and increase the stored energy of the fuel.
I (2) Axial gaps due to pellet column shrinkage and pellet cladding mechanical inter.ction will fom. This causes an increase in j
the neutron fl1x due to the increased moderation.
Tnis increase in neutron flux is called a power spike and.must be considered in detemining the operating power of the fuel rod.
1 Another effect of the axial gap is that, since the cladding is not supported by the fuel column, it might collapse due to creep caused by the extemal-coolant pressure and temperature.
(3) The linear heat generation rate will increase due to a decrease I
in the fuel column length.
All these effects were considered by Commonwealth Edison in predicting +he behavior of Core IX of Dresden-1 under steady state, transient and accident ccnditions. Tne results of the Staff evaluation of these analyses follows.
801017e 7 g5-
(-
- 4 l
?cnsification Effects Cap Conductance The gap conductance for Core IX of presden-1 is calculated with the HOTROD 2
fuel rod themal perfomance codet ; with appropriate modifications as specified by the Staff in reference (4). The Staff, by means of independent calculations using the Staff's fuel themal performance code, GAPCON-4 i
11ER'ML-1(6), has checked these results and agrees that the minimum gap 2
conductance is greater than 300 BTU /hr ft ep, Cladding Collapse 4
CJNFC calculates that Core IX fuel, even if it densifies to such an extent that axial gaps are fomed, will not collapse at any time during its life 4
in Dresden-1. These calculations were done by using the GUNFC computer codes CREBUCK and COLLAPSE described in reference 2.
Tne Staff independently verified these results with its om cladding collapse computer code, BUCKLE.
~
Rod Drop Accident The rod drop accident and its analysis has certain characteristics that should be considered in the evaluation of potential densification effects on the accident. These are:
i (1) The analysis of the lower power rod drop accidents assumes
~
adiabatic fuel pins, i.e., no heat transfer out of the pins during the transient. Tnese conditions mean that densifi-cation effects related to transient heat transfer are not relevant. The low power operating region is of primary interest since at higher power the rod worths are less and feedback more prompt resulting in greater margin to the 280 cal /gm accident limit criteria.
4 (2) Tne transient parameters of importance to the analysis (such as dropped rod reactivity, scram reactivity, Doppler feed-back, delayed neutron fraction and gross power distributions) are not significantly affected by potential densification phenomena.
For these reasons the only effect of importance in the analysis of the low power rod drop accident is the possible increased local = power density peaking
=
in the fom of power spikes.. In limiting accidents of this class the axial power peaks in the upper region of the core. Thus, a-required modification of the analysis must involve an increase in peak fuel enthalpy by the power i
i
=
3-sp:ke cciculated for the upper several feet of the core.
For a pre-
._ensification result near the allowable limit, consideration of the power spike, which is about 4 percent, would result in an increase of 11 cal /gm.
This would result in only a minor reduction of mximum allowed rod worth (frca 1.64 to 1.585 delta k for the limiting cold critical conditicn) to rect the 280 cal /gm limit.
At higher power in'tial conditions, the increase in initial stored energy and transient energy resulting from the decreased gap conductivity and power spikes would increase the final peak enthalpy by less than 25 cal /gm. This increase, coupled with the much larger margin to the 280 cal /gm limit for the higher power accidents, would still not exceed the limit.
Power Spike Abdel Tne effects of densification on power density distributions have been calcu-lated by Gulf United using models in general confonnance with those discussed in Section 4 of the Staff Densification Report (Reference 4). These calcu-lations take into account the peaking due to a given gap, the probability distribution of peaking due to the distribution of gaps, and the convolution of the peaking probability with the design radial power distribution. The model has been described in GU-5300, " Evaluation of Densification -
Dresden Unit 1 Fuel Supplied by Gulf United," Feb. 23, 1973, and m modifi-cations to the model described in the answer to question 34 in the June 12, 1973 Response to Request for Additional Information.
It is the modified model that is discussed in this evaluation since it contains the presently required mximum gap size fonnulation. The calculation uses models and n=erical data the same as the power spike model described in Appendix E of Reference 3, except that:
(1) The calculational model uses probabilities rather than Fbnte Carlo techniques. Both methods provide similar results.
(2)
Initial density is that applicable to the reactor being con-sidered, e.g., 93.7% of theoretical density for Dresden 1.
(3) Maximum gap size is calculated with a clad irradiation growth factor of 0.3 percent (rather than 0.4) as appropriate to one cycle of operation with Gulf United fuel.
(4) Tne gap size distribution is based on that recomended by the Staff in Section 4.2 of Reference 3, but is even more con-servative in that it predicts a larger percent of gaps which are large size gaps.
i I
i I
,+
,--a e
,~
. t (5) The convolution _alculation uses an expectation value of 0.355 rather than 0.255 to provide a 95 percent confidence level. This comes from the use of a discrete probability analysis rather than a continuous approximation and is in agreement with a suggestion which had been independently made by our Erookhaven National Laboratory (BNL) consultants.
i (6) The gap peaking factor'for a single gap of various sizes and positions differs numerically from that given.in Appendix E and is calculated for conditions appropriate to the reactor being considered. This may include, as for Dresden calcu-lations, the use of zero peaking effects across inter-assembly water channels so that a full latticeis not necessarily included when an assembly corner cr edge pins are involved in power spike probability calculations.
The use of individually calcu-lated gap effects for a given reactor is a satisfactory procedure, but requires evaluation on a case-by-case basis.
Thus, the overall model presented is a satisfactory modification of the acceptable procedures indicated in the Staff densification report. Specific calculations using this model for Dresden 1 are in general agreement with extrapolations of calculations which have been done by our consultants, Brookhaven National Laboratory, in the course of doing check calculations of power spikes for a range of PWRs and BhRs (Reference 5).
Steady State Operation and Transients During steady state operation the power spi.kes created by fuel densification are postulated to influence the ninimum critical heat flux ratio (MGFR) and the calculated cladding strain. The licensee has detemined that the MGFR will be greater than 1.5 at 125% design power with the inclusion of a power spike penalty accounted for by increasing the local power peaking fattor. A peak centerline temperature of 4860*F was calculated at 125%
power including densification effects.
In addition, the maximum clad strains calculated using HOTROD do not exceed 0.7%.
Based on our comparison i
of HOTROD and GAPCON, the Staff considers these analyses to be suitably conservative and acceptable.
+
Those operating transients reevaluated are:
(1)
Control rod run out during startup and at full power.
(2) Sudden isolation valve closure.
(3) Cold water injection.
(4)
Fuel loading accident.
J
- ..., O. futI center
- ina terg:-rutura vc; -
i-
.1...:,,cin, cons n'.ering densifict.tacn effects.
'ax s wf:... ;. 8 w;te
. licasce that the consequenccs of these transients takr ' t: accc a:.
Cc'r of fuel densification wcuid r.ot result in 1cc.ccr 2:e theiri
.ia (i.e.,
- JEF Icss than 1.5) than in previous analyses.
b::i-of-Ccolant Accident 1:. February,1973 GU'GC supplied the Staff with calculations of the effect cr. Dresden-1 of a hypothetical loss of coolant accident.
Lese calculations icere done with the M1\\T code in the form of a parametric study of peak clad tcngerature versus gap conductance, all other conditions remaining,the same.
Tnese calculations show that at a gap conductance of 300 PTU/hr fte F the
- mai clad tc:merature will be 2111 F.
Since the calculated rinimun gap ccn-Tulce calculated for Dresden-1, Core IX is greater than 300 i.TU/hr ft#F,
.':esden-1, Core IX meets the Interin Acceptance Criteria lir.it of peak clad te.rgerature less than 2300 F.
Conclusions Eased on our evaluation of the information submitted bv Cc=.cnuealth Edison fcr Dresden-1 and our gencric evaluation cf infomatic' subr.itted by Gulf n
United Nuclear Fuels Corporation we have ccncluded that no cl.2:ges in cperating limitations are necessary as a result of fuel densification censiderations.
mq o
3" D
b.
o ee e
2
s i
-6 References 1.
" Gulf Light Water Reacter Fuel Rod Thermal-Mechanical Analysis Methods," GU-5300, June, 1973.
f i
2.
" Gulf Light Water Reacter Fuel Rod Thermal-Mechanical Analysis Methods," GU-5300, Revision 1, October, 1973.
}
3.
" Technical Report on Densification of Light Water Reactor Fuels,'~
Regulatory Staff, U.S. Atomic Energy Commission, November 14 1972.
4.
" Technical Report on Densification of Gulf United Nuclear Fuels Corporation Fuels for Light Water Reactors," Regulatory Staff, U.S. Atomic Energy Commission, November 21, 1973.
i 5.
Brookhaven National Laboratory, Interin Reports; Power."eaking Due to Fuel Densification in a BWR; December 19, 1972 and February 20, 1973.
Spike Factors in a BWR; April 24, 1973, April 26, 1973, May 8, 1973, June 5, 1973, August 13, 1973.
Peaking Factors in Pressurized Water Reactors with Fuel Densification, December,1972.
6.
Hann, C. R., et.al., "GAPCON THERMAL-1: Computer Program for Calculating the Gap Conductance in Oxide Fuel Pins," BNWL-1778.
4
- ]
i i
t i
f T
_, _. _ _