ML20245F591
| ML20245F591 | |
| Person / Time | |
|---|---|
| Site: | Brunswick  |
| Issue date: | 08/04/1989 |
| From: | Loflin L CAROLINA POWER & LIGHT CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NLS-89-224, TAC-71112, TAC-71113, NUDOCS 8908140427 | |
| Download: ML20245F591 (9) | |
Text
_
6 it D g.
g Carolina Power & Lipy Company AUG 4 1989 SERIAi: NLS-89-224 i
United States Nuclear Regulatory Commission I
ATTENTION: Document Control Desk Washington, DC 20555 l
BRUNSWICK STEAM ELECTRIC PIANT, UNIT NOS, 1 AND 2 DOCKET NOS. 50-325,6 50-324/ LICENSE NOS. DPR-71 & DPR-62 H
RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION I
REACTOR CAVITY NEUTRON DOSIMETRY PROGRAM (TAC NOS. 71112-AND 71113)
Gentleme-n:
On June - 21, 1989, the NRC requested additional'information regarding CP&L's l
submittals dated October 26, 1988 and March 30, 1989. contains the Company's response to this request.
Please refer any questions regarding this submittal to Mr. William R. Murray j
at'(919) 546-4661.
1 Yours very truly,
/2 - ) _f Leo ard I.
flin 4
Mana r Nuclear Licersing Section LIL/ MAT
~ Enclosure cc:
Mr. S. D. Ebneter Mr. W. H. Ruland Mr. E. G. Tourigny 8908140427 890004 7
PDR ADOCK 05000324 y
.P PNU r
e l
411 Fayetteville Street
- P. O. Box 1551
- Raleign. N C. 27602 h
j, L
ENCLOSURE 1 BRUNSWICK sID.M ELECTRIC PLANT, UNITS 1 AND 2 NRC DOCKETS 50-325 & 50-324 l
OPERATING LICENSES DPR-71 & DPR-62 RESP 0MSE TO REQUEST FOR ADDITIONAL INFORMATION REACTOR CAVITY NEUTRON DOSIMETRY PROGRAM (TAC NOS. 71112 AND 71113)
The following information is offered in reply to the NRC request for additional information regarding the Brunswick Pressure-Temperature Amendment i
request.
Specifically, the questions. pertain to the reactor cavity dosimetry l
program carried out during the sixth fuel cycle at Brunswick Unit 2.
The results of that measurement program were documented in WCAP-10903.
The ten questions posed in the attachment can be conveniently divided into two broad categories. Questions 3 through 7 deal primarily with the analytical aspects of calculating the pressure vessel exposure from first principals, while questions 1, 2, 8, 9, and 10 deal with the evaluation of the cavity l
dosimetry itself and the subsequent use of that data to determine the actual l
exposure of the inner vessel wall.
In this case, it is convenient to address these two categories separately.
l Prior to answering the individual questions that have been posed, a general statement regarding the overall intent of the calculation in the context of the dosimetry program is in order. The two-dimensional calculation discussed in WCAP-10903 was not intended to be a stand-alone predictor of vessel exposure but, rather, to be a tool for the interpretation and extrapolation of the cavity measurements.
The absolute magnitude of the vessel exposure was to be set by the measurements, not the calculation and, therefore, the burden on t1.c discrete ordinates evaluation was merely to provide cavity spectra for dosimetry interpretation and radici slopes through the vessel wall for extrapolation. The intent of the measurements was not to " benchmark" this particular analytical model.
The following are replies to the specific analytical questions.
Question 3 There was no dis ussion how the void affected the neutron source (neutron leakage) which was taken into account in the calculation of the middle-plane two-dimensional neutron transport.
Such leakage is axially asymmetric.
Response
The void fraction, which varies both azimuthally and axially, has a significant impact on neutron leakage from the core and, hence, on the absolute magnitude of the tautrr n radiation levels at the pressure vessel.
However, the one-dimer.sional studies described in WCAP-10903 indicated that over a void fraction range of 0 - 90 percent, the interpretation of dosimetry data and the extrapolation through the El-1
ll vessel wall were not significantly impacted.
That is, over the void:
fraction' range examined, the derived flux at the pressure vessel inner radius varied by plus.or minus 1 percent.
In other words, the slope
.through the'. vessel wall as well as the neutron spectrum in the reactor cavity are relatively insensitive to coolant void fractions. The void fraction studies performed with the one-dimensional transport analysis may be summarized as follows:
Range of Evaluation 0% to 90% vo 4 fraction Maximum Change in 2.0%
Dosimetry Evaluation Maximum Change in-0.2%
Vessel Slope Question 4 The outer assembly power (source) distribution was defined on an assembly by l
assembly basis. This method is known (from PWR analyses) to yield l
non-conservative results by about'10-15 percent which is the level of the indicated uncertainty.
Is there any justification for thic practice?
Response
The definition of the power distribution on an assembly basis rather than fuel pin basis removes gradients that actually exist, thus, increasing the neutron leakage from the reactor core.
Experience has shown that this treatment of core power distribution leads to results that are conser-rative rather than non-conservative.
In any event, as with the case.of the coolant void fraction, the power distribution gradients in the peripheral fuel assemblies have little impact on the relat'.ve_ slope through the vessel wall or on the neutron energy spectrum in the cavity. Therefore, since pin powers were not readily available and.the calculation of the absolute magnitude of the neutron field was not of prime importance, the approach described in WCAP-10903 was taken j
for this' analysis.
Question 5 In the two-dimensional analyses the nominal (rather than the as-built) design dimensions were used. Given the high sensitivity of this calculation to the J
amount of water between source and detector, v% was this factor ignored in the assessment of the uncertainty?
Respon$e The uncertainty in the projected vessel exposure levels depends on the data evaluation in the cavity and on the additional uncertainty in the relative slope through the vessel wall.
The void fraction studies, discussed above, demonstrated that these parameters are not sensitive to El-2
the total amount of water between the source and cavity.
This has been borne out by numerous evaluations of PWR cavity dosimetry where amounts of water can vary significantly as a function of azimuthal angle. These PWR studies have shown that the neutron environment in the reactor-cavity is governed almost totally by the thickness of the reactor vessel and the local cavity geometry and not by the reactor internals configuration.
Question 6 l
Staff estimates indicate that BWR peripheral assembly power calculations have i
significantly higher uncertainty than the inner assemblies.
Why'.ias this effect ignored?
Response
The uncertainties in the core power distributions are reflected in an uncertainty in the absolute magnitude of the calculated vessel exposures.
They have no significant bearing on either the i
interpretation of cavity dosimetry or on the extrapolation of dosimetry results to the vessel inner radius, i
Ouestion 7 Toward the end of cycle the upyir part of the core may experience significant plutonium burning, thus, affecting the neutrons / fission ratio as well as the neutron upectrum hardness. Was this factor accounted for?
Response
i Again, the burn-in of plutonium isotopes impacts the absolute magnitude of calculated fluence values. Cavity measurement programs at PWR plants have demonstrated that the cavity spectra and slope through the vessel wall are the same for out-in loading patterns and extreme low leakage loading patterns; i.e.,
they are insensitive to the amount of plutonium fissioning in the reactar core. An example of this behavior is represented by a comparison of 16 sets of multiple foil data obtained from Westinghouse 4-loop reactor plants. The data base contair.s eight data points from fuel cycles with fresh fuel on the periphery of the
]
ccTe and eight additional points with thrice burned fuel on the 1
periphery.
The following reference data set represents the average j
measured reaction rates af:er normalization to the Fe-54 (n.p) reaction.
Cu-63 (n.a) 3.87E-19 6.8% 1 sigma I
Ti-46 (n,p) 5.23E-18 7.4% 1 sigma j
Fe-54 (n,p) 3.06E 17 i
Ni-58 (n,p) 4.14E-17 10.5% 1 sigma U-238 (n,f) 2.07E-16 8.3% 1 sigma Np-237 (n.f) 3.52E-15 7.7% 1 sigma i
El-3 l
s l-This data set demonstrates that the neutron spectrum, as manifested in l
reaction' rate ratios relative to the Fe-54 (n,p) reaction, is impacted neither by the implementation of low leakage fuel management nor by the azimuthal position of the sensor set in the cavity.
In fact, the sensor behavior is quite similar to that observed in the 4-loop surveillance capsule data base summarized belov l
Cu-63'(n,a) 5.81E-17 5.1% 1 sigma L
Fe-54 (n,p) 5.81E-15 Ni-58 (n,p) 7.60E-15 4.7% 1 sigma l
U-238 (n,f) 3.35E-14 6.7% 1 sigma Np-237 (n,f) 3.04E-13 6.7% 1 sigma
'The questions discussed above raise several valid points regarding the
()
. complexities of BWR fluence analysis. Many of the factors brought up in these questions vary not only azimuthally and radially, but also with time during the fuel cycle. Additional complexity is introduced for the analyst by the presence of jet pump structures in the regions adjacent to the pressure vessel wall. All of these factors lead to large uncertainties in the calculated l
, values of exposure for the pressure vessel. These potential large uncertainties are precisely why the cavity measurements were performed and are so valuable. 'In essence, the cavity dosimetry results depend primarily on the vessel thickness and the relative neutron spectrum in the cavity.
- Thus, uncertainties in void fraction, power distribution, and reactor internals geometry can be effectively removed from the evaluation, resulting in a
. reduced overall unce.rtainty in the vessel exposure projections.
The following are replites to the questions regarding the cavity dosimetry evaluation and extrapolation.
Question 1 Is there any benchmarking of the solid state track recorders used for fluence measurement?
Response
The solid state track recorders (STTRs) used in Westinghouse' cavity dosimetry programs have been benchmarked by irradiation in calibrLted fields at the National Bureau of Standards. The results of these benchmarking evaluations indicate that the STTRs can be used to accurately. measure fission rates even with the low mass deposits that are utilized in reactor cavity irradiations. A summary of a series of NBS benchmarking comparisons is provided in Table 1.
i El-4 l
l l
l l
Ouestien_Z There was no discussion of neutron streaming in the cavity. How was it l
established that the measured values were not affected by neutron streaming or other cavity effects?
l
Response
The reactor cavity at the Brunswick Unit 2 reactor is extremely " clean" in the sense that there is little if any extraneous material present; thus, the area between the reactor vessel and the sacrificial shield is essentially an open annulus over the axial extent of the reactor vessel beltline.
The area between the carbon steal vessel and the sacrificial shield is filled with 4.5 inches of reflective insulation (3% stainless steel by volume) and a 10.5-inch air gap. This configuration very much resembles " narrow cavity" PWR configurations. A series of measurements at PWR plants with similar cavity geometries has demonstrated that the measured axial distributions in the reactor cavity can be directly correlated with the axial distribution of neutron leakage from the core.
This statement applies, cf course, to the axial zone in the cavity that is coincident with the core height. The fact that this correlation can be shown indicates that streaming effects within approximately plus or minus 5 feet of the core midplane are not significantly distorting the axial shape of the neutron field in these relatively tight reactor cavity configurations and, therefore, can be ignored in the dosimetry evaluations. A comparison of measured axial flux profiles in the reactor cavity and core axial power distributions is presented in Figure 1.
The agreement between the two sets of data support the conclusion expressed above.
Question 8 Beyond the calculated slope of the neutron flux through the pressure vessel, were any other cavity effects taken into account in the extrapolation?
Response
i In determining the relative neutron spectrum in the cavity as well as the radial slope through the vessel wall, the reflective insulation, aic gap, and sacrificial shield were included in the analytical model.
In 1
addition, a one-dimensional parametric study was carried out to ascertain the impact of varying concentrations of iron in the sacrificial shield on the interpretation and extrapolation of cavity dosimetry results. This study indicated that over a concentration range of 0 - 30 percent iron in the sacrificial shield, the derived fast neutron flux at the pressure vessel inner radius varied by plus or minus 1.5 percent.
El-5
J 1
~
'Ouestion 9 An uncertainty of 5-7 percent is estimated for the projection of the neutron flux'within the pressure vessel. This is said to be based on the PCA l
bsnchmark. However, the kind of cavity and the extraneous materials present could affect the extrapolation. Has the similarity of Brunswick-PCA cavity i
been established.
Response
The PCA benchmark was used to establish an uncertainty in the calculated slope through the carbon steel vessel wall. As noted in the many I
write-ups on the PCA, calculated slopes tended to be too steep due to defects in iron inelastic scattering cross-sections currently in the ENDFB-IV data base. An examination of the measured and predicted slopes for the PCA 12/13 configuration indicated that calculated attenuation across.the 8.5-inch PCA vessel mockup exceeded measurement by 5-7 percent. This analytical defect was included as an extrapolation uncertainty.
In addition to the PCA data, there have been three occasions evaluated by Westinghouse where surveillance capsule dosimetry has been removed coincident with reactor cavity dosimetry.
In all y
cases, the cavity sensors and the surveillance dosimetry experienced the same exposure history. That is, all sets were irradiated for a single L
fuel cycle.
In these three cases, vessel thickness ranged from 8.5 to l
9.5 inches.
In all three cases, the derivation of vessel exposure from-capsule dosimetry versus cavity dosimetry supported the 5-7 percent l
slope defect observed in the PCA comparisons.
It should be noted that
~
since the vessel thickness at Brunswick Unit 2 is considerably less than 8.5 inches, the 5-7 percent extrapolation uncertainty may be somewhat overstated.
Furthermore, since the calculated slopes through the vessel wall are demonstrably too steep, the normalization of cavity results to
- that slope will produce slightly conservative results at the pressure vessel inner radius.
Question 10 j
The PCA was an order of magnitude more compact arrangement than the Brunswick reactor. How does this affect the applicability of the PCA benchmarking?
Response
The PCA benchmark was used only to ascertain uncertainties in the calculated slope through the steel block, not to assess uncertainties in the absolute magnitude of neutron radiation levels within the block.
In that sense, the 8.5-inch PCA vessel mockup spans most of the vessel thicknesses of operating reactors.
El-6 I
4
L l
TABLE 1 NBS Benchmark Irradiations and Mass Assignments l
Deposit Mass (ng)
Mass (ng)
Mass Ratio Label Isotope Source HEDL NBS NBS/HEDL
- 876 238 Dil. 12 8.01 7.50 0.936 g
- 880 238 Dil. 12 2.88 2.71 0.941 U
- 894 238 Dil. 12 4.34 4.19 0.965*
g
- 874 238 Dil. 12 3.27 2.82*
0.862*
U
- 871 238 Dil. 12 35.2 29.4*
0.835*
g j
- 1104 238 Dil. 13 1.53 1.57 0.994 g
- 1114 238 Dil. 13 1.64 1.59 0.970 U
- 1112 238 Dil. 13 2.06 2.13 1.034 U
.,2288 237 Dil. 8 3.63 3.38 0.931 Np
- 129C 237 Dil. 8 0.667 0.626 0.939 Np
- 1033 237 Dil. 7 0.315 0.306 0.971 Np
- 1043 237 Dil. 7 0.302 0.265 0.877**
Np W-21' 7 237 W-7 4.88 4.59 0.941 Np Average (0.953+0.040)***
From damaged dosimetry set.
This exposure resulted in a small number of tracks and a correspondingly high statistical uncertainty on the resultant total. This value is less than 2a from the average of all of the mass ratios (excluding #874 and #871). Note that the mass ratios for deposits #874 and #871 are more than 2o and 30 from the average, respective, since these measurements resulted in higher track counts with less statistical uncertainty.
l
- This tverage does not include the ratics for deposits #894, #874, and #871. The I
cause for the 4.7% discrepancy between the radiometric and NBS mass scales is being
]
investigated.
If the average of the two masses is used, an additional 2.4%
j uncertainty in the mass value results.
I (435CRS)
n_ -
.. ?. v,,
j 3:ri.:LI ::_; Ei^
e
- r. : :.
. r _. :.._.
FIGURE 1 a
..s. w.w.-o,o IMr-CO.
- w. _ _on, m_ _e,n__y n n_ e_ _r n_ _o m.
vurgwv
~ WITE CAVITY GRADIENT MEASUREMENTS a
l
{.:. :
.I 2.
..._..__._._..________.u______...
t i
2 t
b.*
i l
i 5
d J
l I
{
z.
p ---
CYCLE 2 3
- y j
X" g;
f r
.r s 1
E..
.s B
v,i
/
o t
.w-t q
- g l
l t
. s 3
1 n.
w a
r t
_-..r..
s, l
_-+
/
j
..L..g.
t t
- -- m_\\..
4
.. l t.
I i ':
~:
l I __.
3
_j 1
g z.
.o r
2
. -. 6 I
CYCLE.1 2~
.... _ _. _ _. _. _. _... ~ _.
2;
,.=
3
- . 5 g
. ! = t'o '
__.. _ y _ _= _,
g.
q.. s_ _-
=
i 5
- ..;:==
,,ni ; a _ _
- ~ ~
=- + M M E-MM-2 W=+ +N
-2=
Qty ;fgii+ ~
-z=-2
- -=iMkn~=t4~~
-~
E E4-I N,:
.em
...__ m.;..
y
._ - _.. _._..w=-----.__._..-
.. __ : = _=.
w
=. ___;.:. -
.___t._._._..-_-_-_--.r--.
6 1,
t-i s.
._. n_ ;3. 4. _. _w.,
,. w
_m..m...
+
.._.__.__-._m
+
i g,- - 3 m. --. -e.
m'
= =. : = r __.+----
=_.y,__
_ _37== = n-5 54.. t. : \\ l
..... -..r.__
- n. _ _: 4..;; _.__..__._._.s,,_-
, 'a;gp.=_. J, ~.'=N.e3tPW._ ^^ '". '.T ' _ :.
- ~L_ : ;'= :
n r-o
.. _.. _ _. ~... -.
. y.m... a..
m 3
m
__ _ _=...__ _;;._:. :
....]...._.._
.... t... _.. _.
.. ~..-.
- -~
. :==:- -.-
. i.
a...
r
.-.4
_..i
_.a
._. g l
.__7_._.
..4
_._y.._..
.._..4..__-..
}
. _. y _
1
-e
-5
-4
-3
-z
-1 0
1 z
3 4
5 6
l Distance From Core Midplane (ft)
_ _ _ _ _ _ _ _ _ _. _ _ _ _ _ _ _ _