ML20079F833
| ML20079F833 | |
| Person / Time | |
|---|---|
| Site: | Clinch River |
| Issue date: | 06/01/1982 |
| From: | Longenecker J ENERGY, DEPT. OF |
| To: | Check P Office of Nuclear Reactor Regulation |
| References | |
| HQ:S:82:035, HQ:S:82:35, NUDOCS 8206080225 | |
| Download: ML20079F833 (30) | |
Text
_ - _ _ _
Department of Energy Washington, D.C. 20545 Docket No. 50-537 HQ:S:82:035 JUN 01 1982 Mr. Paul S. Check, Director CRBRP Program Office Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C.
20555
Dear Mr. Check:
RESPONSES TO REQUEST FOR ADDITIONAL INFORMATION - CORE PERFORMANCE
Reference:
Letter, P. S. Check to J. R. Longenecker, "CRBRP Request for Additional Information," dated March 23, 1982 This letter formally responds to your request for additional information contained in the reference letter.
Enclosed are responses to Questions CS 491.1 through 4, 6, 8 through 17, and 19 through 22.
Responses to questions CS 491.5, 7, and 18 will be provided by June 4.
These responses will also be incorporated into the PSAR Amendment 69; scheduled for submittal later in June.
Sincerely, W
. Dvgn JqpnR.Longenpdonm,entalManager cker Licensing & Env Coordination Office of Nuclear Energy Enclosures cc: Service List Standard Distribution Licensing Distribution gDN 8206080225 820601 PDR ADOCK 05000537 A
,,Page 1 (82-0315) [8,22] #57 Ouestion CS491.1 In Section 3.1, " Conf ormance with General Design Criteria," there is no design criterion comparable to 10CFR50, Part A, Criterion 28, " Reactivity Limits".
Why is this general design criterion not a part of Section 3.17 Are appropriate Ilmits on the potential enount and rate of reactivity increase discussed in this criterion going to be quantitatively specified?
Resnonse 10CFR50, Part A, Criterion 28, " Reactivity Limits" is not a part of Section 3.1 since the CRBRP Design Criterion was established prior to the present 10CFR50, Part A.
However, as discussed in Question Response CS421.2, the present CRBRP design f ully meets Criterion 28 of 10CFR50, Part A.
In Section 3.1, Criterion 23, " Protection System Requirements f or Reactivity Control Mal functions", Criterion 24, " Reactivity Control System Redundancy and Capability", and Criterion 25, " Combined Reactivity control Systems Capabliity" serve as suf ficient criteria f or limiting the amount and rate of reactivity increase.
l l
QCS491.1-1 Amend. 69
9 Pcg2 2 (82-0315) L8,22] #57 Ouestion Cs491.2 Please explain why Criterion 29, " Protection against Anticipated Operational l
Occurrences", of 10CFR50, Part A is not a criterion in Section 3.1 of The l
PSAR.
_Resoonse Likw Criterion 28, of 10CFR50, Part A, discutsed in Question Response CS491.1, t
CHBRP Criterion 29 was establisned prior to the present 10CFR50, Part A.
Question Response Cd421.2 discusses how the CKBHP design tully neets this i
criteria.
1 4
QCS491.2-1 Amena. 69 Y
1
P:ge - 1 L8,22] #32 s
s s
t-Questfon OCS 491.3 (4.2.2.1)
The specific speed of response requirements does not seem to be presented in Secti on 4.2.2. t.3.
Where is It presented?
Resoonse The PSAR section l isted shoul d be 4.2.3.1.3.
The specific speed of response requirements are given in PSAR Figures 4.2-114 and 4.2-119.
i
\\
l t
. _5 '
t s.
\\,
N i
\\
\\.
1
?-
s QCS491.3-1 Amend. 69 May 1982 w__--__-_________
P;go 3 s82-0315) L8,22] #57 OuestIon CS491.4 (4.2.3.3)
Why aren't there sufficient calculational uncertainties listed.To enable one to judge she fragility of the PCRS and SURS scram conclusions?
l Resoonse l
1 It is not possible to identify and combine statistic.sl uncertainties in the PCRS scram analysis for the constituents which produce the results presented in rigures 4.2-114 and 4.2-119.
PSAR Section 4.2.3.3 has been updated to include more recent results, and where possible, calculational uncertainties have oeen estimated.
PSAR dection 4.2.3.3 will be updated to include :nore recent results whicn will consist of ths DYNALSS Code verification (comparison between predictions and test results) predicted CnBRP SCRS scram performance and required CKBHP SCRS scram performance. The DYNALSS code is used to predict SLRS scram performance.
Where possible, calculational uncertalntles will be included in the SLRS scram performance predictions.
This information will be incorpoaratefinto the PdAR in CY 1982.
/
/
l I
QCS491.4-1 Amena. 69 N w.w
P ga - 2 [8,22] #32 OUESTION-DCS 491.6 (4.2.3.4)
Are the PCA position Indicators and dampers also being tested?
Which of the tests mentioned in Section 4.2.3.4.1 have been completed and documented?
RESPONSE
The PCA position Indication systems and the PCRDM dashpots were included in the system level tests. PSAR Section 4.2.3.4.1 has been revised to reflect this information.
The following tests of Section 4.2.3.4.1.1 (Primary Control Rod System) have been completed:
A.
Component Tests 1.
Dynamic Seismic Friction Test - Test report not issued 2.
Control Assembly Hydraulic (Flow) Test - Test report not issued 3.
Control Assembly Pin Compaction Test - Test report not issued 4.
Control Assembly Rotational Test Joint
. Documentation complete 5.
B C Data Test - See response to Question 490.28 4
6.
Friction and Wear Tests - Documentation complete 7.
Control Assembly Analytical Methods - Test report not issued B.
System Level Tests - Test report not issued C.
PCA irradiation Test to begin in FFTF Core 2 The following tests of Section 4.2.3.4.1.2 (Secondary Control Rod System) have been completed except as noted:
A.
Latch - Tests - Documentation Completel B.
Damper Tests - Test report not issued C.
Position Indication Tests - Test report not issued D.
SCA Status Flow Tests - Test report not issued E.
Prototype Tests - Test report not issued NOTE: While SCRS Prototype tests P-1, P-2, and EL-4 have been completed, SCRS Prototype tests P-3 and P-4 tests are in progress.
QCS491.6-1 Amend. 69 May 1982
F.
Coll Cord Tests - Test report not issued, G.
Latet Seal Tests - Test report not issued 2
H.
Noseplace Flow Tests - Documentation canplete 1.
Argon Control System - Test report not issued 1.
Documentation includes, "CRBRP-GEFR-00524, "SCRS Latch Assembly Scram Cycling Test Final Evaluation Report," May 1980 and CRBRP-GEFR-0054, "SCRS Latch Assembly Real Time Test Numers One and Two Report," June 1981.
2.
CRBRP-GEFR-00487, " Development of Noseplace Orif ice For the Secondary Control Rod System," October 10, 1979.
QCS491.6-2 r,. - m
\\
Paga - 5 L8.22Jf 66 6.
Friction and Wear Tests The base technology materials test program being conducted at ETEC and ARD provides data for the material couples selected for f abrication of the primary control rod system.
l 7.
Control Assembiv Analvtical Methods Provides an analytical model calibrated with test results for predicting primary control assembly thermal-hydraulics performance.
lifetime characteristics and scram dynamics behavior.
B)
Svstem Level Tests: A series of Primary Control Rod System Prototype Tests have been performed to verify that the Primary Control Rod System perf ormance is consistent with its design equirements under design basis operating conditions. The Control Rod Prive Mechanism was evaluated in a CRDM Accelerated Unlatching Life Test. This test program vertfled the unlatch perf ormance characteristics of a prototype primary control rod drive mechanism over twice the design lifetime travel and scrams. The Accelerated Lif e Test involved testing of a f ull size prototype primary control rod system in sodlum, sodium vapor. and argon gas environments that simulate operations in the Clinch River Breeder Reactor Plant.
Phase I testing in this series canpleted 1/2 of the PCRS lifetime scrams.
1.3 of the leadscrew travel requirement, and about 5 times the PCA travel requirements. Phase I; of this series will extend total test scrams and tervel beyond CFBRP lifetime requirements.
During System Level Tests of the Primary Control Rod System, each subsystem was also tested, including the position Indication system and the dashpot.
Four prototype systems were tested and the results show PCRS perf ormance including Position Indicator accuracy and dashpot performance were within their design requirements.
t C)
PCA Irradiation Test: A PCA Irradiatloe test is scheduled to be Inserted in the FFTF for 600 FPD. The Intent of this test Is to provide near-prototypic f rradiation performance data on the PCA absorber assembly to support the PCA lifetime evaluations. The test assembly will contain 37 deconaary control,C and will function as an integral part of the FFTF pina vi ein iched B Assembly Bank. The parameters of the test assembly have been selected to provide temperature and cladding strain. Data from this test are expected to be available in 1985.
D)
Other Tests: See Appendix C for Reliability Test Program.
4.2.3.4.1.2 Secondarv Control Rod Svstem The SCRS testing progran consists of the following major testing activities:
A) Latch Tests: Component development tests of the scram latch configuration for the secondary control rod system verlflod the design of this component.
B) Damner Tests: Component develpment tests of the damper configuration for the secondary control rod system verified the design of this component.
4.2-305 Amend. 69 rn arm >
Page - 3 [8,22] #32'~~~
Ouestion CS491.8 (4.3.2.2)
The applicant Indicates that power distribution limits are derived f rom maximum allowable peak heat generation rates for nominal and anticipated operational conditions, which combined with the rod mechanical and thermal parameters, assure that Inciplent fuel melting does not occur in the fuel pellet with peak power. What are these speelfic, quantitative, power l
distribution limits? What are the maximum (quantitative) allowable peak heat generation rates (linear power) for nominal and anticipated oprestional conditions? What clad and coolant temperatures correspond to these maximum peak heat generation rates?
Resoonse The peak calculated linear power discussed in PSAR section 4.3.2.2 is 12.4 KW/ft in the fuel at the beginning of cycle one and 16.5 KW/ft in the Inner blankets at the end of cycle four after 550 effective full power days of Irradiation.
Uncertaintles include calculation to experiment ratio differences in isotopic fission and capture rates determined from analysis of experiments performed in ZPPR-7, methods /modeling uncertaintles, and CRBRP engineering tolerances (fissile and pellet heavy metal content tolerances, reactor power normalization, control rod banking tolerance,...). The corresponding maximum linear power with 3cruncertainty and 115% reactor power, is 15.7 KW/ft in the fuel at beginning of cycle one (15.9 KW/ft in the ref ueled assemblies at the beginning of cycle two) and 20.0 KW/ft in the inner j
blanket at the end of cycle four.
The core design has been based on limits of 16 KW/ft in the fuel and 20 KW/ft in the blankets.
However, there are no maximum allowable peak heat generation rates per se; rather the limiting criterion is to prevent incipient fuel melting at 115% overpower, thermal-hydraulle design (THDV) conditions, and accounting for 3 uncertaintles. The highest power and temperature (peak and hot) rods in the fuel and blanket are analyzed with the LIFE code (as reported in PSAR section 4.4.3.3.6) to guarantee that no incipent melting occurs at the aforementioned conditions. Cladding and coolant temperatures are calculated by the NICER code for each rod and input as boundary conditions to LIFE Maximum cladding ID temperatures are provided in Figures 4.4-45 and -46.
Design, extrapolation and experimental /modeling uncertainties, at the 3 cr level of confidence as reported in 4.4.3.2.2, are f actored into this analysis.
Note that the peak clad and coolant temperatures do not necessarily correspond to the maximum peak heat operation rates because of orificing for the various constraints. Actual peak temperatures for all assemblies are provided in PSAR Section 4.4.
QCS491.8-1 Amend. 69 May 1982
P;ge - 5 [8,22] #32 I
Ouestion OCS491.9 What type of Instrumentation is planned to allow detection of flux (power) tilts in the core at operational levels?
Resoonse Three separate types of instrumentation will Indicate the presence of flux tilts.
Firstly, two redundant control rod position Indication systems wilI provide evicence of any deviation of single control rods from their correct banked locations, which could result in flux tilts.
Secondly, a core exit temperature measurement system wilI monitor the sodium temperature at the exit of almost all fuel assemblies, in this way, the changes in thermal power distribution arising from a flux tilt will be Indicated to the operators.
Finally, two separate flux monitoring systems located on the periphery of the Reactor vessel and each containing three equidistantly located detectors wilI provide evidence of overall tilts in Reactor flux.
In conjunction with the rod position Indicators and the flux monitoring Indicators, the plant computer and plant annunciator system will be used to Inform the operator of flux tilts.
QCS491.9-1 Amend. 69 May 1982
~ ~ " " ^ '
Pag 3 T 4 [8,223 #32 Ouestion CS491.10 (4.3.2.2)
In Section 4.3.2.2.9c clarify the expression "36 equivalent uncertaintles".
Resoonse in the radial blanket power uncertainties evaluation discussed in section 4.3.2.2.9C of the PSAR, the experimental component of the power uncertainty is based on the pro-Engineering Mockup Critical (ZPPR-7) Isotopic fission rate data. This data base was very limited for the radial blanket and therefore it does not provide a good statistical basis for developing the uncertainty.
Consequently, two times the maximum range of observed calculation to axperiment ratio (C/E) variations for each isotopic reaction rate in the redi rl blanket was used to develop the uncertainty. This approach is suf ficiently conservative for preliminary design to be equivalent to t3r urrertainty limits. The radial blanket end of life peak linear power of 16.6 kW/f t at the end of cycle 4 and 18.0 kW/f t at the end of cycle 8 (Including 3 /
uncertainty and 15% overpower margin) is not particularly close to the 20 KW/ft peak in the Inner blankets.
A more extensive data base is being obtained from the Engineering Mockup Cr!tical (ZPPR-11) so that conventional statistics can be utilized to evaluate the 3r uncertianty envelope for final design.
QCS491.10-1 Amend. 69 May 1982
P2ge - 7 [8,22] #32
. Question CS491.11 (4.3.2.3)
"The Doppler reactivity constant has been computed for temperatures above 2100 degress K (e.g., 3000, 4000 and 5000 degrees K).
What assumptions were made, or how uncertain are.these high temperature coef ficients when your basic 30-group library probably only gives temperature dependent cross sections to 2100 degrees K7" l
"Can you argue that any safety considerations only very weakly depend on accurate high temperature Doppler coeffIclents?"
Resoonse The high temperature coef fIclents have one sigma uncertainties of i 7% for temperature dependence, i 10% for Doppler constant, and i 12% for the combined Doppler feedback reactivity at elevated temperature.
A discussion of the Doppler reactivity constant is presented in revised Segtlon 4.3.2.3. The temperature dependent cross sections are given to 2100% with extrapolation limited to 40000K in the proposed PSAR Figure 4.3-27a.
The basic ref erence on the subject of sensitivity of energy release during an HCDA to the uncertainty in the Doppler coefficient is the paper by Nicholson and Jackson.(1) They identified a relationship between the Doppler coef ficient, A = T -pr, and the core peak temperature of:
( T, ) _ ( R,
} **
- t T, /
~
, at /
for ramp rates as dif ferent as 20$/see and 100$/sec. The peak temperature is a usef ul parameter for comparison, since it bears a close relationship to the energy of Isentropic expansion at slug impact on the head.
(1)
R. B. Nicholson and J. F. Jackson, "A Sensitivity Study for Fast-Reacter Disassembly Calculations", ANL-7952, January 1974.
QCS491.11-1 Amend. 69 May 1982
P:ge 7 (82-0298) [8,22] #65 A calculation perf ormed on a molten pool configuration derived from the CRBR BOC-1 heterogeneous core (ramp = 30$/sec) shows a relationship of 3
, _ Y-gi f
f A, kT (A./
2 i i The higher exponent is explained by the greater peak to average temperature distribution of the heterogeneous core relative to the homogeneous design used by Nicholson and Jackson.
In the calculation perf ormed for CRBR the reactivity peaks before a core average temperature of 34000K is attained. By the time 36000K average is reached, the rate of increase to feedback frm material motion exceeds that of the Doppler contribution. Af ter the power peaks the Doppler contribetion quickly stabilizes whereas the motion contribution continues to accelerate and soon exceeds the Doppler in absolute magnitude. The relationship between Doppler coef ficient and core average temperature is even more loosely coupled as
/T)
[A, W 6 2
i
( Ti /
(Az./
l l
Because of the low specific power in the internal blanket assemblies of the I
heterogeneous core, the average temperature responds more sluggishly than the peak.
By the time peak temperatures of 50000K are reached the internal pressures are disrupting static structures and accelerating f uel mass away from peak power locations. These peaks are associated with core average temperatures of 40000K in the heterogeneous core. By this time the magnitude of the Doppler coef ficient has become secondary to the details of material relocation.
In ennetusion, both general and specific calculations support the view that Doppler uncertainties in the range of temperatures above 40000K are not significant. Doppler plays its most significant role below 35000K before f uel begins to move.
i QCS491.11-2 Amend. 69 May 1982
P2ge 1 (82-0298) [8,22] #65 buildup, 2) the mid-term row 6 ref ueling, and 3) control rod bank withdrawal offects. Equivalent Doppler constants at the beginning-of-cycle one and at the end-of-cycle f our in a sodlum-volded environment are shown In Table 4.3-17.
The eifect of the removal of sodlum is to harden the neutron energy spectrum and substantially reduce the magnitude of the Doppler constants.
Table 4.3-18 presents a t pleal nodal-average Doppler distribution in the
/
f uel, inrcr, and radial blankets at the beginning-of-cycle one.
Each region contains a total of seven axial nodes; five equal-volume nodes in the 36-Inch high "f uel" region and one node each in the upper and lower blankets (extension). The row 1 and row 2 radial blanket Doppler constants have been combined additively into a single region.
This combination results in a slightly conservative (less negative) feedback reactivity due to temperature dif f erences in the two rows of radial blankets.
Figures 4.3-27u and b show the distribution of Doppler constant by assembly in the 36-inch active f uel and Inner blankets at the beginning of cycle one and the end of cycle f our, respectively. The values in Figures 4.3-27a and b are condensed f rom three-dimensional (VFNTURE) first-order perturbation theory calculations which were used to develop nodal feedback coef ficient input to SAS analyses (see Chapter 15).
The temperature dependence of the Doppler constant is discussed in Ref erence 5.
The Doppler contribution of the fissile material is a small positive of f ect, and general ly fof lows a T-372 dependence.
However, the U-238 contribution is strongly negative and overrides the small positive contr:bt!on f rm. the f IssiIe nuclIdes. Calcuiattons of the temperature depend:nce f cr a series of U-238 resonances result in a Doppler temperature relationship of T-T.
Sel f-shielding ef fects will tend te decrease the absolute value of the temperature exponent, but f or a f ast reactor having a f ertile/ fissile content similar to CRBRP, the overall Doppler constant has approximately a T-T variation.
The temperature dependence of the Doppler coef ficient is in general
?
characterized by the expression dk/dT = AT-n, where 1he theoretical limits of n are 0.5 and 1.5 and the expected value f or CRBRP and other LWBR's is close to 1.0 (see Ref erence 6). Analysis of the SEFOR results (see Ref erence 7)
Indicates a value of n between 0.9 and 1.0; however, the temperature depenaence of tne SEFOR Doppler feedback In the range of the measurements is insensitive to the value of n f or n between 0.8 and 1.2.
For extrapolation to higher temperatures, calculations Indicate a value f or SEFOR of n = 1.0.
Yarlations in n of 20% over the f uel temperature range of the SEFOR measure-ment: lead to 10% variations in the Doppler feedback extrapolated f ran LMFBR operating conditions to HCDA conditions (see Figure 4.3-27a).
The one cr uncertalMy in Doppler feedback reactivity due to the uncertainty in n was estimated (see Ref erence 7a) to be 75 for temperatures characteristic of an extreme (HCDA) accident.
This uncertainty is independent of and uncorrelated with the uncertelnty in the value of the Doppler reactivity constant.
When estimating the total uncertainty in Doppler feedback reactivity, in accident conditions, these uncertainties should be combined stati sti cal ly.
4.3-35 Amend. 69 May 1982
Page 2 (82-0298) [8,22] #65 The i 7% (le ) uncertainty in temperature dependence when statistically combined with the i 10% (17) uncertainty in Doppler constant (see next paragraph) yields a total uncertainty in Doppler feedback reactivity at elevated temperatures of i 12 (IT). This additional component on the Doppler feedback uncertainty would only apply to the highly unilkely condition of an HCOA type of event. For static or operational and design transient evaluations, the Doppler feedback uncertainty is characterized by the i 10%
(1 tr ) uncertainty In the CRBRP Doppler constant.
Doppler Uncertainty:
The uncertainty in the CRBRP Doppler Constant has been developed from the analysis of the SEFOR Core I and ll experiments. The Southwest Experimental Fast Oxide Reactor (SEFOR) was constructed specifically to determine the LWBR Core Doppler feedback through a series of power coef ficient ((/Wth) and sub-and super-prompt transient energy coef ficient (//Wth.sec) measurements.
SEFOR Core II had a material composition, resultant neutron energy spectrum and f uel temperature that was reasonably characteristic of that in CRBRP. The SEFOR experiments are described in, for example, Reference 8.
The SEFOR Core 11 Doppler constant derived from these measurements (T dk/dT = -0.0060) is in good agreement with the value of -0.0062 calculated by% (to equivalent)
GE in Reference 9.
GE estimated the SEFOR Doppler constant uncertainty as 19 In Referer.ce 9.
The principal contributions themselves, are estimated uncertainties in the f uel temperature-power relationships (fuel to coolant thermal conductance and f uel specific heat) required to extract the Doppler constant, -T dk/dT, from the measured power and energy coef ficients (p/ Wit and p /Wth.sec, respectively). Additional uncertainties in the extrapolation of the SEFOR power and energy coef ficients to LWBR power reactors are l
attributable to ef fects which are significantly d!f ferent between the two l
reactors (uncertainties in f uel thermal properties, delayed neutron data, and the Ilke are highly correlated between SEFOR and power reactors so that these uncertainties largely cancel in the normalization). The net extrapolated uncertainty in LWBR power or energy coef ficient was determined to be i 11%
(1ir) in Reference 9.
This extrapolation accounted for dif ferences in the SEFOR and LWBR core composition and spectrum, fuel thermal property
(
differences, and spatial temperature and importance weighting uncertainties.
The neglect of spatial temperature and importance weighting (that is, the use of region Doppler constants with average f uel temperatures) tends to (ce s-veMyely) underestimate l
4.3-36 Amend. 69 May 1982
, Page 3 (82-0298) [8,22] #65 l
REFEREN&S FOR SECTION 4.3 1.
R. L. Childs, F. R. Mynett, L. S. Abbott, " Analyses of the TSF First-Fission Stored-Fuel and Ex-Vessel Low-Level Flux Monitor Experiments f or the Clinch River Breeder Reactor", ORNL-M5057, March 1976 (Chapters 3, l
4 and 5).
2.
J. B. Bul lock, M. V. Mathis, J. T. Michalczo, " Inverse Kinetics Rod Drop Measurements With a Mockup of the Cilnch River Breeder Reactor Shields",
ORNL-W4828, March 1976 (Chapters 5 and 6).
3.
J. W. Allen, " Development and Application of a Three-Point Inverse Kinetics Rod Drop Technique f or Suberiticality Determination",
ORNL-W4758, November 1975.
4.
J. W. Allen, J. C. Robinson and N. J. Ackerman, Jr., " Statistical Errors in Suberifical Reactivity inf erred f rom Inverse Kinetics Rod-Drop Measurenents Using the Three-Point Method", ORNL-W4101, July 1973.
5.
Harry H. Hummel and David Okrent, " Reactivity Coef ficients in Large Fast Power Reactors", Monograph Series on Nuclear Science and Technology, American Nuclear Society,1970.
6.
P. Greebl er, B. A. Hutchins, and J. R. Sueoke, Calculation of Doooler l
CoeffIclent and Other Safety Parameters for a large Fast Oxide Reactor, GEAP-3636, General Electric Company,1%1.
l 7.
D. D. Freement et al., "SEFOR: Verification of the Doppler Transient Shutdown Capabil Ity in LMFBR's", Proc. ANS National Toolcal Meeting on New Develooments in Reactor Physics and Shielding, Conf.-720901, Book 2, Klarnesha Lake, NY,1972.
7a.
P. Greebler, et al., " Status of Reactivity Feedback and Stability", Fast Breeder Reactor - Status of Saf ety Technology Development, DOE / TIC-11209, Department of Energy, Reactor Research and Technology, August 1980.
8.
a
- 9. Kussmaul and S. L. Derby, " Experimental Progran Results In SEFOR Core iI", GEAP-13838, June 1972.
9.
D. D. Freeman, "SEFOR Experimental Results and Appilcation to LWBRs",
GEAP-13929, January 1973.
- 10. R
- g. Kalser and J. M. Gasidio, "On the Ef f ect of Core Conf Iguratton on Doppler Measurenents in ZPPR Assembly 2", New Develooments in Reactor Phystes and Shielding, Conf.-720901, September 12-15, 1972,
- p. 1006.
4.3 -87 Amend. 69
% WXd
.)
.~
INSERT C DOPPLAR PEEDSACK REACTivlTY ethT = Af l
a = la 2J l
\\ /, '*
44(T) = t(T).hfam Ki n=0A I ak" = h(12 Kl.kM53 El
/
/n = 17
/
0 SE FOR CALCULAT4083 p
NPtsE RICAL IDI
/
u
+
4massiaseso f
/
UteCERTAmeTY, f
jfl
..ia
,/
\\
/
\\
g
/
I
g tavsRAGsPust
/
l TsweaATuasi
/
i l
i
/
I 1
i l/
14 1
/
I I
I
/
I l
/I I
pust
/
l j MSLTweG ronacT 1
/
o i
i l
l l
l 1
i i
l l
I u.
/
l
.I e
I 2
3 4
AVEAAGE PuSL TluPE R ATURE (K s it's 27c Figure 4.3.28*.
Extrapolation of Doppler Feedback to Elevated T,emperatures O
1 P ge - 8 [8,22] #32 Ouestion 005491.12 (4.3.2.3):
Uncertainty in the Doppler constant has been based principally on the analysis of the SEFOR Core I and il experiments perf ormed by GE.
Is the extrapolation from SEFOR to CRBR simply from one reactor to another or between reacters and methods?
If the PSAR method for calculating the Doppler constant is dif ferent from that used by GE to calculate the SEFOR Doppler constant, please provide a comparison of the two methods.
Also provide justification as to why the uncertainty of the GE method should be accepted as that for the CRBR method.
The PSAR also indictes that Ref. 9 provides data for extrapolation from SEFOR to LWBR power reacters accounting for dif ferences in core composition, core-spectrum, etc. Provide justification that the SEFOR to LMFBR power reactors (1973 type) extrapolation should be identical to the SEFOR to CRBR extrapolation.
Resoonse The CRBRP Doppler constant uncertainty is based in large part on the SEFOR measurements and analyses perf ormed by General Electric, Hanf ord Engineering Development Laboratory, and others. The HEDL analysis, documented in HEDL-TME-73-42, " Analysis of the Doppler Constants of Cores I and li of SEFOR," by R. A. Harris (May,1973), was perf ormed with the same neutronics codes and basic cross section data as that employed in the CRBRP nuclear design. The results f or SEFOR-Core ll are:
Experimental Doppler Constant
-0.0063 (T dk/dT)
(with ENDF/B celayed neutron data consistent with that in the CRBRP design)
GE Calculated Doppler Constant
-0.0063 HEDL Calculated Doppler Constant
-0.0062.
The close agreement between the HEDL and GE calculations, as well as the good agreement with the measured value, demonstrates that there is no significant dif ference in GE and Westinghouse Doppler calculation methods.
QCS491.12-1 Amend. 69 May 1982
......s.,
P:ge - 8 [8,22] #32 The SEFOR Doppler uncertainty was assessed by GE based on the potential for uncertainties in the Interpretation of the experimental SEFOR parmeters (fuel temperature determination, separation of non-Doppler feedback components, and oth ers). A recent survey article by Paul Greebler, " Reactivity Feedback and Stability; A Status Report 'n Saf ety Lines of Assurance," DOE / TIC-11209 (1980), provides an excelltat summary of the SEFOR Doppler constant uncertainty and extrapolation assessments. The Doppler uncertainty, extrapolated to a large LWBR, Is 110% (la). Most of the extrapolatton f actors considered by Dr. Greebler apply to the SEFOR-to-CRBRP extrapolation.
The principal additional extrapolation f actors in CRBRP are associated with blanket ef fects in the heterogeneous core configuration. Doppler calculational capability in a heterogeneous core arrangement has been investigated using small-sample Doppler measurements in ZPPR. Argonne National Laboratory has produced a core U238 Doppler map for ZPPR-11B, the clean beginning of life CRBEP Engineering Mockup Critical (references AHL-ROP-103 ). The Integral measured ZPPR-118 fuel Doppler constant is 0.00331. The Westinghouse calculated value is -0.00327. The good agreement between calculated and measured ZPPR-11B Doppler constant suggests that there is no significant additional Doppler constant error in the heterogeneous core arrangement.
f QCS491.12-2 Amend. 69 Mar / 1982 l
Prg2 - 10 [8,22] #32 Ouestion CS491.13 (4.3.2.3)
It appears that Expansion and Bowing Reactivity Coef ficients are computed by Integrating core movements over axial and radial worth curves. Has the reactivity of the small core reconfiguration been accounted for?
As the core heats up the structural and f uel materials expand increasing the size of the core. Does the mass of sodlum necessarily increase In the expanded core? That is, is it possible that sodium expands enough with higher temperature that Its mass in the core stays the same or even decreases?
Resoonse The uniform radial expansion reactivity feedback coef ficient was determined from eigenvalue dif ference calculations with reference and expanded-core models in hexagonal geometry. The reactivity associated with the f undamental mode flux change has theref ore been acocunted for explicitly in the expanded core eigenval ue calculation. The radial bowing reactivity coef ficients in Tables 4.3-24 and 25 were determined by moving axial segments or assembly rows radially through material worth gradients in a first-order perturbation sense in RZ geometry.
A uniform radial t <pansion coef ficient can also be synthesized from these bowing coef f t:lents by summing the reactivity worths f or a uniform radial movement of cach node. This synthesized coefficient agrees very well with the directly calculated radial expansion coefficient.
The uniform axial exp ston reactivity feedback coef ficient is also calculated frcm f!r:t crd:r porturbation theory in RZ geometry. The axial expansion reect!vity coef ficient is the dif ference between the average material worth in the 36-inch core and the worth at the core / axial blanket boundaries.
As the core heats up, the structural and f uel materials expand, increasing the volume of the core. The volume of sodlum increases in this heat-up.
For t
I example, the sodium volume Increase in the transition from ref ueling temperature (4000F) to hot f ull power conditions Is less than 15. The sodium density decrease in this same ref ueling temperature to hot f ull power transition is aobut 6%, so that there is a Tet reduction in sodium mass of 5-6%.
QCS491.13-1 Amend. 69
Page - 11 [8,22] #32 I
Cuestion CS491.14 (4.3.2.4)
There appears to be insuf ficient control rod neutronics. Provide descriptions of calculations and data that characterize control rod burnup, management, and flux and power distributions.
I Resoonse Control rod B10 depletion in the partly inserted Row 7 Corner primary control rod bank amounts to 5-6 atom % in a 275 ef fective f ull power day equilibrium c cle. Accumulated B10 burnup is determined by time and space Integrating the B 0 capture rate from RZ dif fusion calculations in the 2DB code with the Row 7 corner primary control red ring inserted at beginning-of-cycle depth for the first half of the cycle and end-of-cycle depth for the remainder of the cycle.
Because of the strong spatial self-shielding in the fully enriched CRBRP control rods, this depletion only accounts f or a 3% loss in reactivity worth which is approximately 0.2% Ak for the R7C bank. During the course of an equilibrium burnup cycle, the core excess reactivity is depleted nearly 2%Ak so that the primary shutdown margin, in fact, increases substantially more than the loss in control rod worth.
From Table 4.3-29, the primary control rod worth margin at the beginning-of-cycle 4, the second of the two-cycle equilibrium batch burnup eles, is larger than that at the beginning of cycle 3 such that control rod Bg depletion from cycle 3 does not preclude use of these same assemblies in cycle 4.
Control rod management is addressed in Sections 4.3.2.5 and 4.3.2.6 of the PSAR.
Calculations to characterize CRBRP control assembly flux and power distributions are based on multigroup (55 energy groups, 42 neutron and 13 gamma) two-dimensional dif fusion theory computer calculations using the ODTillW computer progran. A series of triangular mesh calculations was perf ormed for each Individual assembly in the reactor core for beginning of cycle 1 (BOC1), B003, and end of cycle (EOC4) conditions. The conditions were modeled at the radial midplane with separate cases run to model the full-in conditions f or each type (primary or secondary) and location (R4, R70, or R7F) of assembly. Output flux distributions from the DOTTillW runs were input to the HEDPIN computer program to generate 37 (PCA) and 31 (SCA) pin radial flux and r6 war distributions using a polynomial equation fit to the individual group flux distributions in each assembly.
Two dimensional, R-Z, dif fusion theory calculations were performed at BOC1 conditions to define the variation of nuclear heating as a f unction of control absorber assembly position for the row seven corner (R7C) primary control assembly for a total of seven Insertion positions ranging from the control absorber assmbly fully inserted to a fully withdrawn condition. Axial power QCS491.14-1 Amend. 69 N MYi92
Pcge - 7 [8,22] #32
~~
and flux distributions were generated f or time-in-lif e conditions and specified insertions with the output data files using a series of Interpolation routines and cycle-to-cycle f actors. Normalized axial distributions at a selected insertion position are applicable to each assembly type and absolute values for each type of assembly are predicted using the f ully inserted flux and power values for each assembly type developed in the detailed triangular mesh radial midplane analyses.
l l
QCS491.14-2 l
May 1982
P:ge 2^8 [8,22] #32 ~ ' ~
Question CS491.15 (4.3.2.7)
In order to establish the criticality of the hot-full power CRBRP, why not perform a direct K calculation at hot-full power conditions?
Resoonse l
l All CRBRP criticality (k hot-full-powercoregeomINy)calculationshavebeenperformedin models with major heavy metal cross sections which have been corrected for hot-full-power pellet temperatures.
(
l QCS491.15-1 Amend. 69 Y
POge - 13 [8,22] #32 Ouestf on OCS491.16 (4.3.2.7):
Why were the minimum critical configuration calculations performed with only P cross sections?
o
Response
The minimum critical configuration calculations In 4.3.2.7 were performed wi1h transport-corrected Po cro ions. The correction takes the form of an adjustment to the transporks se and n-group scattering cross sections using the cosine of the average scatttering angle. This transport correction gives results equivalent to a P1 approximation, 1
1 l
l QCS491.16-1 Amend. 69 C3798@
Pcge - 15 [8,22] #32 Questfon CS491.17 (4.3.2.9)
Section 4.3.2.9 and Table 4.3-35 report neutron flux and fluence data at locations within the core and in structural components outside the core, i.e.,
core barrel and reactor vessel. Please respond to the following question regarding the ex-core calculations:
1.
What was the calculational method used?
2.
What was the geometrica! model used and what modeling approximations were made?
3.
What was the neutron source used in the calculations? What was the basis for this source and what approximations were made in incorporating it into ex-core calculations?'
4.
What procedure was used to insure that the ex-core neutron fluences, over plant life, are conservative and representative of the worst points on the core barrel and pressure vessel?
5.
What is the accuracy of the calculated fluxes?
6.
What are the limiting flux (fluence) values for the core barrel and pressure vessel?
Resoonse 1.
Calculations to def ine the in-vessel radiation environment at f ull power conditions use a two-dimensional RX modeling of the CRBRP reactor system.
The radiation environment was developed using the discrete ordinates transport /dif fusion theory computer progran DOTillW.
Multigroup (42 neutron energy groups) noturon flux distributions were calculated in dif fusion theory based on a core region fission source distribution defined by nuclear analysis methods defined in Section 4.3.2.2 of the PSAR.
Neutron cross sections were generated with the computer progran XSRES/WIDX. The cross sections, the FTR 300-S ENDF/B-IV library, are a 42-energy group set f or P'O scattering, and transport enc rected f or P1 scattering by an extended transport approximation.
2.
Cylindrical (RZ) modeling of tne CRBRP is used. The reactor internals RZ model represents a cylindrical description of the regions internal to and including the reactor vessel in the radial direction, and from the core support structure plate to the upper internals structure / sodium pool region in the axial direction. The RX mode!!ng defines the spatial dependence of the reactor internals Irradiation environment based on homogenized material regions. The modeling approximation uses a model of the array of hexagonal assemblies (fuel, inner blanket, radial blanket, and radial shield) as developed from conservation of mass and volume.
QCS491.17-1 Amend. 69 May 1980
P;ge - 16 [8,22] #32 3.
The fission source distributions within the core were obtained as output from the nuclear analysis methods defined in Section 4.3.2.2 of the PSAR.
Cylindrical (RZ) modeling of the fissile relgons in DOTillW were identical to those of the fission regions used in the nuclear analysis. The nuclear analysis model is expanded f or ex-core regions to adequately represent the reactor system.
4 The permanent structures radiation environments are based on an equilibrium cycle average flux distribution, (e.g., BDC3-EOC4).
Cylindrical (RZ) maximum flux levels by component (e.g., core barrel), and by region (e.g., lower ring, lower ring-middle cylinder circumferential weld, middle cylinder, middle cylinder longitudinal weld), are defined radlally and axially relative to the core axis and core midplane, respectively. Detailed final design analyses will consider azimuthal flux variations resulting from hexagonal-to-cylindrical interf aces, and neutron flux streaming f actors for actual component design.
5.
Analysis has defined an 8%-12% (17) neutron flux uncertainty of the fixed radial shield-core barrel region.
Flux uncertainties are to be incorporated into the CRBRP radiation environment predictions at the time of final design analysis.
6.
The core barrel material fluence limits are defined in PSAR Table 4.2-53 basemgtal (SS304) and the reactor vesseg2 n/cd, respectively,
( Amend. 54). Fluence limits weldment (SS308) are 2.1 x 10((n/cm 2 and 1.4 x 10 i
QCS491.17-2 Amend. 69 May 1980
~ ~ ~ ~ ~ ~ ~ ~ '
Piga ~ 11 [8,223 132 Qugstion CS491.19 (4.3.3)
The applicants method of calculating the CRBR and applying biases derived from critical assembly investigations Is similar to that performed for the FFTF.
Since startup measurements on the FFTF have been completed, what investigations have been performed to di.'ver which methods and calculations did not stand up well for the FFTF and hen;e may be suspect for the CRBRf Does the 30-group neutron cross section library (your basic starting point) have a reference? Canthis exact library be obtained in order to reproduce any of your calculations?
Resoonse The CRBRP nuclear design has been based on an extensive Engineering Mockup Critical experiments program in which the critical mockup exhibited a high degree of similitude with the CRBRP, much closer in many respe:ts than FFTF.
This similitude includes important characteristics like the heteogeneous core configuration, plutonium enrichment and composition control rod pattern and worths, boundary conditions and others.
Consequently, bias f actors and uncertainties f rom core Integral physics characteristics in ZPPR are accurately extrapolated to CRBRP.
Startup and low power physics characterization measurements have been perf ormed in FFTF. These measurements include Initial criticality and control rod worth determinations, Isotopic fission rate characterization measurements, and temperature and power coefficient determinations. The results of these measurements are currently being evaluated primarily by the FFTF staf f.
An ef fort is currently underway, however, to calculate the FFTF control rod worth measurements with CRBRP design tools.
The 30-group cross section library, used in the PSAR nuclear analysis, is based on ENDF/B-lll data which has been processed at HEDL with the ET0X code.
The library Is essentially the same as FTR Set 300. The exact 30-group CRBRP library, consisting of Infinitely dilute cross section and self-shielding f actors in the Bondarenko f ormat, can be obtained f rom CRBRP.
QCS491.19-1 Amend. 69 May 1980
Pcge - 18 [8,22] #32 Ouestion CS491.20 (4.3.3.7) in this section you state that ZPPR-4 control rod bank worth C/E values range f rom 0.95 to 1.04.
Why do these numbers dif fer from those presented in Table 4.3-40.
Resoonse The ZPPR-4 progran consisted of four distinct phases (critical configurations) simulating a clean-beginning-of-life core and a burned-snd-of-life containing plutonium in the radial blankets, both with and without inserted control rod banks. The calculation to experiment (OBE) ratios for control rod worths in alI phases of ZPPR-4 range from 0.95 to 1.04 (PSAR reference 29 in Chapter 4.3).
The data In PSAR Table 4.3-40 summarizes the average C/E, from ali four phases, for each control rod bank. The range of variations is dampened by the averaging process in the Table 4.3-40 values.
QCS491.20-1 Amend. 69
t'Igo - 15 L6,22J 732 Ouestion 0CS491.21 Section 15.1.2 descrites qualitative core limits f or normal operations, transients, and accidents.
Are specific, quantitative, design limits going to l
be specified?
If not, please justify why qualitative limits are pref erable to l
quantitative I imits.
I Resoonse This subject was addressed at the February 25, 1982 meeting with NRC.
The qualitative Iimits of PSAR Table 15.1.2-1 have been translated into specific acceptance guldelines f or preliminary saf ety evaluation for each event ciassi f Ication, as shown in PSAR Tabl e 15.1.2.2.
The gu t dei ines of PSAR Tabl e 15.1.2-2 are derived based upon the design limits and methodology and Insure that the qualitative limits of PSAR Table 15.1.2-1 are preserved.
Detailed calculations of mechanical damage to meet the selected umbrella transients have been reported in Chapter 4.
DetalIed calcuiations of mechanical damage w11I be perf ormed f or the f Inal saf ety revlew (FSAR).
If other spectf Ic core design limits are required, they will be specified and discussed with NRC prior to the FSAR submittal.
QCS491.21 -1 Amend. 69 D K'f*2
Question CS491.22 (15.1.2)
No sodium boiling is used as a limit for extremely unlikely faults in Table 15.1.2-2.
This limit does not appear to have a specific value (temperature) as it depends on the coolant pressure.
If this criterion i
results in a variable quantitative temperature limit for the various events considered, why is the corresponding design limiting coolant temperature (and its basis) not specified for each event?
Response
The no sodium boiling limit does result in a variable quantitative temperature limit for the various events considered. A corresponding design limiting temperature is not specified for each event because of the following:
(1) For those events where the maximum coolant temperature stays significantly below the boiling temperature (regardless of potential variations in pressure) it is not necessary to calculate specific temperature limits to have confidence that boiling is avoided.
(2) In cases where the hot channel coolant temperature approaches the expected saturation temperature of the coolant, the time-dependent pressure is examined to arrive at an estimate of the applicable saturation temperature. Due to the continuous variation of temperatures and pressures during a transient,
- ka mavimum temperature in the hot channel coolant may not be associated with the highest likelihood of boiling. The minimum difference between the saturation temperature and the hot channel coolant temperature is presented as a margin to boiling. Thus, the calculated margin is equivalent to using a unique requirement for each event.
_ _ _ _ _ _ _ _ _ _ - - - - - - _. - _ _