ML20054H721
| ML20054H721 | |
| Person / Time | |
|---|---|
| Site: | Clinch River |
| Issue date: | 06/21/1982 |
| From: | Longenecker J ENERGY, DEPT. OF |
| To: | Check P Office of Nuclear Reactor Regulation |
| References | |
| HQ:S:82:052, HQ:S:82:52, NUDOCS 8206240363 | |
| Download: ML20054H721 (37) | |
Text
_
Department of Energy Washington, D.C. 20545 Docket No. 50-537 HQ:S:82:052
)eg Mr. Paul S. Check, Director CRBR 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
Reference:
Letter, P. S. Check to J. R. Longenecker, "CRBRP Request for Additional Information," dated April 30, 1982 This letter formally responds to your request for additional information contained in the reference letter.
Enclosed are responses to Questions CS 760.15, 17, 18, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 78, 80, 81, 83, 84, 85, 87, 88, 91,101, and 104; which will also be incorporated into the PSAR Amendment 69; scheduled for submittal in July.
Sincerely, AS.
J[hnR.Longen ker Acting Director, Office of the Clinch River Breeder Reactor Plant Project Office of Nuclear Energy Enclosures cc: Service List 0/
Standard Distribution Licensing Distribution
~
8206240363 820621 PDR ADOCK 05000537 A PDR
I;Gge - t Sus-@WWD LW,2GJ Ut@
Ouestion CS76015 The reactivity ef fect associated with the f uel rod bowing can be either positive or negative depending upon the assembly support structure and the radial temperature gradient. In the CRBRP, the assembl ies are hel d at the top and bottm of the core. Depending upon the radial temperature gradient, fuel assembl ies can bow either toward the center of the core, resulting in a posi-tive reactivity contribution, or away fran the core, thus causing a negative reactivity contribution. This ef f ect can resul t in a substantial net positive reactivity, it shoul d be eval uated f or dif f erent power-to-flow conditions and al so f or dif f erent inter-assembly gap sizes (nominal gaps to reduced gaps).
For CRBRP, the project-computed bowing reactivity contribution for the hmo-geneous core was as high as +654. For the heterogeneous core, this val ue shout d be smal I er.
Pl ease provide a discussion or Inf ormation that woul d conf irm that these concerns have been covered.
Reponse The informatIon requested Is provided In PSAR SectIon 4.2.2.4.3.3.
QCS760.15-1 Anend. 69 July 1982 82-@398
f P gs - 2 (82-0358) [8,22] 489 Ouestion CS760.17 It is clear throughout Section 15.2, Reactivity insertion Design Events, that transient fuel cladding mechanical interaction is ackowledged only for rapid reactivity insertion events; that is, events occurring during a time span of one second or less. This is an unwarranted assumption for any reactivity insertion event, slow or f ast, particularly for fuel pins in which the fuel-cladding gap is expected to be closed at steady-state. Please provide the justification for this assumption.
Resconse The Information requested by this question has been provided in the response to Question CS490.31.
l QCS760.17-1 Amend. 69 July 1982
Page - 3 (82-0358) [8,22] #89 Ouestion CS760.18 Has it been considered whether or not ang of the corrosion phenomena changes character at elevated temperatures (1600 F and beyond to coolant saturation) so as to cause wholesale or catastrophic cladding failure, or become in itself a cladding failure mechanism? Could this occur under accident conditions where it might be postulated that control over oxygen or other impurities in the sodium would be lost?
Resoonse The condition of severe sodium overtemperature for any appreciable length of time is generally characterized by events which are caused by undercooling (LOF) scenarios. These events are rather short and generally are over in less than five minutes.
It is not expected that there will be a change in character of the corrosion mechanism. Extrapolating the NSMH corrosion rate equation (1 ppm Oxygen) yields the following:
at 1600 F: R = 8.7 mpy at 1650 F: R = 13.1 mpy at 1700 F: R = 18.5 mpy As a worst-case condition assume hat oxygen level reaches 10 ppm and the 6
temperature rises to 17g0 F/15 min. This would result in a theoretical clad wall wastage of 5 x 10 mil. Catastrophic failure would not be expected at such a small value. In addition, since the undercooling events are characterized by reduced flow, thus the corrosion rate would also be reduced from full-flow conditions (i.e., the sodium can only saturate with corrosion product; mass transport would occur by thermal convection only).
Under accident conditions of the type postulated, it is not expected that control of oxygen in sodium wou:d be lost. In addition, because there are low flow conditions in the reactor fuel assemblies, if there were to be postulated an Ingress of oxygen, it would not reach the assemblies in as rapid a time, or have as much influence on corrosion, as under full flow conditions.
l l
l l
l t
i l
QCS760.18-1
! Amend. 69 1
July 1982
rep ., m us.o. Lt:,22] 59 Ouestion CS760.59 The flow allocation to the various assemblies is constrained by fuel life coolant boliIng The PSAR states on page 4.4-11 that assembly mixed mean outlet temperatures and radial temperature gradients at the assembly exit are among the constraints used in the flow orificing. However, r.o f urther information is provided as to how those two constraints are applied in your analysis.
Resoonse As mentioned at the end of PSAR Section 4.4.3.5.1, the philosophy followed in CRBRP was first to allocate the flow to satisfy the most restrictive of the lifetime and transient contraints (limited flow), then to orifice the core on the basis of the limited flows and identify whether excess flow exists. If this is Indeed the case, "the excess flow is allocated among the f uel assemblies to minimize and equalize the assemblies exit temperatures and temperature gradients".
As discussed in Section 4.4.2.5.3, the minimum amount of orificed core flow i necessary to satisfy the limiting constraints was 93.07% of the total reacter flow. Since: 1) the amount (less than 1% of total reactor flow) of excess flow allocation; and 2) maximum gradients were within the UlS capability, the l available excess flow was distributed roughly evenly among the various orificing zones to provide extra design margin.
To summarize, satisfying the lifetime and transient contraints is a requiroment of the orificing process while optimizing other conditions such 3s exit temperature and temperature gradients is a desired feature. Tnus, only the excess flow,' eventually available af ter the lifetime /translent constraints are satisfied, is used in the optimization process.
i QCS760.59-1 Amend. 69 May 1982
,m u - .__ , __ __ _
Page 4 W82-0320 [8,22] 59 Ouestion CS760.60 The fuel life and coolant boiling constraints are quantified by defining equivalent limiting temperatures which shall not be exceeded. The constraints are defined for Slant Expected Operating Conditions (PEOC). The uncertainty factors are at the 2-sigma level of confidence. Assembly lifetime /burnup goals are achieved when both the cladding Inelastic strain and cladding CDF are within established limits of 0.2% for the ductility strain limit and 0.7 for the CDF during steady state operation. Strain Equivalent Limiting Temperature (SELT) is defined as the end-of-life temperature which, if maintained throughout life, would cause for the particular assembly an EOL cumulative strain of 0.25. A Damage Equivalent Limiting Temperature (DELT) is defined similarly as the equivalent EOL temperature corresponding to a CDF of 0.7 for fuel assemblies and 0.5 for blanket assemblies.
The calculation of the Transient Equivlant Limiting Temperature (TELT) is performed in three steps:
A. to provide an ade:,uate margin to bolling, a temperature of 1550 F is defined as the maximum coolant temperature allowable during the natural circulation translent at a 3-sigma level of confidence assuming Thermal-Hydraulic Design Value (THDV) conditions.
B. this limiting temperature is translated into a temperature Tg which is defined as:
- the maximum steady state coolant temperature corresponding to a 1550 F transient maximum coolant temperature for PEOC
- at the 2-sigma confidence level C. finally, T is y translated into the Transient Equivalent Limiting Temperature TELT by multiplying the difference between Tp and the Inlet by the ratio of the coolant temperature rises at EOL and temperatureTl9ewhenthemaximumtransienttemperatureoccurs, the time in l consider-Ing also the axial position where this temperature is reached, and adding to this the inlet temperature and correcting for the ID temperatures needed. The assemption for this correction is that the temperature difference Tp-T, wo;ld increase / decrease with time In the same manner as the temperature Ifference T Cool
-T In*
It is understood that the orificing is an iterative process whereby a flow distribution is assumed which yields SELTs, DELTs and TELTs. These numbers In turn provide for a new flow distrubution which yields new values for these l
temperatures, etc.
The design basis requires no f uel centerline melting at 115% overpower conditions.
Why has this criterion not been used for the flow orificing?
i l
QCS760.60-1
' Amend. 69 July 1982
Page 5 W82-0320 [8,22] 59 Resoonse Fuel centerline temperatures are only a weak function of the cladding temperature; therefore the no fuel centerline melting would have been a
" flow-insensitive" criterion if used for orif icing. Rather, detailed ad hoc analyses were perf ormed to guarantee that the no melting criterion is satisfied, as discussed in Section 4.4.3.3.6.
It should also be clarified that the orificing is only partially an iterative process, and the question statement "It is understood..." is only partially true. A flow distribution which yields SELTs, DELTs and TELTs is actually assumed. However, once the limiting temperatures are determined, the corresponding flow Is calculated by the OCTOPUS code and this represents the limiting flow adopted in the orificing process. Thus, there is no Iteration process on the temperatures constraints; once they are determined the orificing configuration follows through. However, final verification that all design constraints are Indeed satisfied is performed following calculation of the detailed performance prediction parameters reported in Section 4.4.3.3.
Thus, orificing constraints are quidelines not limits. Guideline values have margins to any limits and provide guidance in core orificing (i.e.,
establishing the optimum flow allocations in core). Subsequent detailed structural analyses (PSAR Section 4.2) determine the design adequacy of core components, using as input the design data from Section 4.4.3.3.
QCS760.60-2 Amend. 69 July 1982
Pcgm 6 W82-032b [8,223 59 6
i Ouestion CS760.61 By selecting the natural circulation transient as the worst transient, cladding temperatures are of greater concern than f uel temperatures. Does this translent selection preclude f uel centerline melting from any consideration?
Resoonse The core transient design considers hath the overpower and undercooling type events. Maximum f uel temperatures and maximum cladding temperatures are both of primo consideration for their respective event types and are accounted for in core design. Core orificing has significantly more ef fect on limiting maximum cladding temperatures than f uel temperatures and thus the undercooling events play a major role in optimization of the core flow allocations.
Natural circulation is the most limiting (with regard to maximum cladding temperature) of the undercooling events. Other events, including TOP, are
- accounted for in the CDF allowances f or transients.
l Fuel conterline melting Is also Considered by designing to the steady-state 15% overpower margin (with uncertaintles) to f uel melting criterion as
, di scussed i n Section 4.4.3.3.6. Fuel melting, if it would occur, is f actored into denign evaluations.
l l
QCS760.61-1 Amend. 69 l May 1982 1
-_ .c -t:"A _, __ _ _ . -- -- - - - - -- - -
~ _ - - - -
Pagn 7 tS2-0320 [8,22] 59 Ouestion CS760.62 ,
What is the basis for excluding TOP events from consideration't Resnonse As discussed in the response to Question CS760.61, TOP events are included in the design, i
i I
QCS760.62-1 ",
Amend. 69
, May 1982
Pcg2 8 W82-0320 [8,22] 59 Ouestion CS760.63 The PEOC do not represent the " worst conditions". The coolant temperature const.aint is defined for THDV conditions.
Why are THDV conditions not used as a basis for flow orificing since they encompass the worst conditions?
l Resoonse A distinction must be clearly made between plant conditions used in performance predictions and in orificing.
When predicting the CRBR T&H performance, use of Thermal-Hydraulic Design Values (THDV) conditions yields more conservative results (lower core flow, higher inlet temperature, hence higher temperatures) than use of Plant Expected Operating Conditions (PEOC). THDVs are used for all safety analyses for both permanent and reptaceabie components. THDVs are also used for ati duty cycle events (transient and steady-state) for all permanent components.
For replaceable components, THDVs are used for all transients except upset events; upset transients and steady-state analyses are based on PE0V conditions.
In orificing, the purpose is not to predict the actual temperature distribution, but to optimize the distribution of the flow among the core assemblies to satisfy specified constraints. The output of the orsficing process is not temperatures, but relative flow allocation ratios among the verlous orificing zones. Thus, when a set of plant conditions is consistently used in expressing quantitatively the various constraints, the resulting flow orifIcIng is Independent of the plant conditions and applles equally to THDV and FEOC.
l .
l QCS760.63-1 Amend. 69 July 1982
Pago 9 W82-0320 [8,22] 59 Question CS760.64 At EOL conditions, the fuel temperatures are the lowest because of the low power level in the f uel, and therefore cladding temperatures are the lowest, too. The opposite is true for the blanket assemblies, which have the highest power level at EOL. The transient constraints come from natural circulation transients and not TOP events. TOP events, however, could lead to higher fuel temperatures at EOL or eariy in life, and higher blanket temperatures at EOL in blanket assembiles. Please address this and provide a discussion which justifies your position.
Resoonsa The justification basis for the selection of limiting conditions is provided in the response to Questions CS760.61 and CS760.66.
QCS760.64-1 Amend. 69 July 1982
Page 10 W82-0320 [8,22] 59 OuestIon CS760 65 Since CRSR is a high-temperature design compared with the design of the conceptual Design Study (CDS), transients needs to be evaluated in greater detall. Please explain why no apparent correlation is established between the unlikely event CDF margins of 0.3 and g.S for f uel and blanket assemblies, respectively, and TELT, Tg or the 1550 temperature I imit.
Response
The requested explanation is provided in the responses to Questions CS760.61 and CS760.66.
1 QCS760.65-1 Amend. 69 July 1982 t e
Pcg310 W82-0320 [8,22] 59 Ouestion CS760.66 Because of the degradation in material / fuel properties with burnup/ fluence, the damage caused by the same transient is dependent on the time in life when the transient occur. This means that there is a correlalon between CDF and the time of the transient. The use of the TELT constraint rather than a
" transient DELT" ignores this f act since the ef fect of time in life when the transient temperatures has its maximum, is only accounted for by an adjustment of tanpr ature rises as described above. Please provide quantitative justi f i n ' on f or thi s concl usion.
Resoonse The transient core design calculations to date have conservatively used the maximum.46T that can be calculated at any time in the design life. These ~~
JkT max's are then superimposed (at the times they occur) on the steady-state ----
temperature histories.
The final design of the CRBRP (to be reported in the FSAR) wil I account for both overpower and undercooling transients throughout the lifetime of the assembly in a more realistic manner than the CDF allowances used in the PSAR.
The purpose of the TELT used in PSAR analyses was to assure satisfaction, with a large margin, of the no-bolling criterion during the worst undercooling transient.
QCS760.66-1 Amend. 69 May 1982
=2-ILL2D
Pcge 12 W82-0320 [8,22] 59 Ouestion CS760.67 How was the time determin' , at which the maximum translent temperature occurs?
Resnonse Each assembly was followed throughout life to determine the time at which the maximum temperature occurs (since the flow and inlet temperature are fixed, the lifetime temperature profile follows the assembly / rod power profile). All the blanket assemblies reached their maximum power / temperature at end-of-life; most of the fuel assemblies attained their maximum at beginning-of-life, except a few whose maximum occurred at the end of the first or beginning of the second year of residence.
l l
I QCS760.67-1 Amend. 69 July 1982
Page 13 W82-0320 L8,22] 59 Ouestion CS760-6B On page 4.4-12, it is stated that the CDF l imit f or steady-state and translant operation is "by def ini tion" 1.0. This statement is correct only if one has a totally accurate description of the f uel perf ormance and its Interaction with the cladding f rom BOL to EOL conditions. Since there is no code which describes nature perf ectly, the CDF l imit is not 1.0 but can be greater or smaller, depending on the conservati sm in the f uel model s.
It is the understanding that the LIFE code, f or exampl e, attempts to provice a best estimate perf ormance characterization and not a conservative estimate of temperatures and stresses. This would mean then, that by necessity, the LIFE code predictions have uncertainties which in addition to modelling approximations come f rom swelling and creep uncertainties, uncertainties in thermal properties, fission gas release f ractions, etc. Theref ore, a CDF value of 1.0 is not a limit and it is not clear if the DELT constraint provides a conservative estimate. Please provide a discussion of the logic used and justif ication f or these concl usions.
Resoonse The design level CDF is computed using cladding properties which are combined at their worst possible levels and cladding temperatures and plenum pressures at thei r l evel s of uncertainty. Thus, the value taken by the design level CDF at any time is the maximum expected value and the limiting value of 1.0 is Intended to precl ude cl adding breach.
In the design procedure, LIFE is used as a source of information only for f uel-cl adding contact pressure. In this regard, LIFE has been calibrated and validated using data f rom operating f uel pins. Thus, the cl adding def ormation and corresponding contact pressures represent state-of-the-art estimates of these phenomena.
QCS760.68-1 Amend. 69 July 1982
P;ge 14 W82-0320 [8,22] 59 Duestion CS760.69 The equivalent limiting temperature (ELT) concept assumes that the actual temperature / pressure history can be simulated by a single temperature value which if maintained throughout life, causes the sa:ne damage / strain as the actual temperature / pressure history, at EOL conditions. This temperature value is then used as a constraint in the orificing analysis and flow rates are determined such that this temperature value is not exceeded. Proof is needed that this temperature which has to be maintained, is a good figure of merit for the fuel performance. Because of the change in operating conditions, the temperature limit can only be met in some form of an average fashion. It needs to be seen how this temperature averaging has to be done.
Non-linearities in the models seem to prohibit the use of ELTs since for example, the damage caused by temperatures fluctuating about a mean, can be greater than the damage caused by the mean temperature. Please address the areas of concern discussed above.
Resoonse The actual temperature / pressure history was analyzed for all assemblies. For orificing, three (or more in a few cases) different flows were utilized and corresponding CDF's calculated. A curve of CDF versus end-of-life temperature was constructed, as discussed in Section 4.4.2.5.2. The end-of-life temper-ature is only a convenient parameter of representation; any other time in life might have been used, except end-of-life was felt to be the most represent-ative time since most of the steady state damage occurs at end-of-life.
The DELT is obtained from the CDF versus temperature curve as the temperature corresponding to the Iimiting CDF.
l Thus, the definition of the ELT as "a single temperature value which if maintained throughout life, causes the same damage / strain as the actual temperature / pressure history, at EOL conditions" is simply a way of physically expressing a correlation parameter devised to put on the same basis an array of assemblies each having different lifetime power profiles and uncertaintles.
No averaging is performed and no damage analysis is performed using the ELT as the mean temperature throughout life. Actual power / temperature / pressure histories throughout lifetime are analyzed for each assembly. The only source of error exists in Interpolating the CDF/EOL temperature curve around the limiting CDF. However, if the actual CDF's are calculated close enough to the limiting value, the interpolating error is negligible (also as shown in Figures 4.4-10 through 4.4-14, the correlation is practically linear on a semi-log scele). The three flows used in determining the temperatures /
pressure histories were in fact selected to yield values of CDF very close to the limit. In a few instances, when it was found that the initial three guesses yielded CDF's too far from the limiting value, more than three flows were used.
Again, subsequent detailed structural analyses are performed to predict limiting fuel and blanket rod performance. In this manner, feedback is obtained which provides proof that this is a " good figure of merit for the l fuel performance".
QCS760.69-1 Amend. 69 July 1982
Pcge 15 W82-0320 [8,22] 59 Ouestion CS760.70 The multiple constraint approach using ELTs rather than temperature limits, introduces another concern and this is the stability of the solution.
Quantitative models are employed to describe cumulative strain and CDF and transient description. How much of a margin exists to account for ;
inaccuracles in the models? Are the HCFs the only data to account for uncertaintles?
Resnonse The accuracy of the solution is discussed in the response to Question ,
CS760.69. As f ar as uncertainties are concerned, HCF's as reported in Section l 4.4.3.2 only cover the thermal-hydraulic aspects of these evaluations.
Structural / material type uncertainties were considered in the subsequent detailed CDF evaluations which are covered in Section 4.2.1.
QCS760.70-1 Amend. 69 July 1982
Pcge 16 W82-0320 [8,22] 59 l
1 Ouestion CS760.71 in determining the flow allocations, have slight flow redistributions been taken into account which comes from partial load operation? Has any consideration been given to the deformation of bundles toward the EOL and how this af fects the hydraulics of the system?
Resoonse See the response to question CS760.86 for a discussion of how the ef fects of bypass flow redistribution at partial load operation are f actored into the design.
Flow redistribution at low flow conditions has conservatively not been accounted f or in determining the flow allocations. However, for final design, transient analyses will be conducted accounting for flow redistribution ef fects through use of the COBRA-WC code, thus they will be indirectly f actored into the orificing process. The FSAR will report the results of these studies.
Ef f ect of bundle def ormation throughout lif e has been eccounted f or by appropriate uncertainty f actors which are part of the set of HCF's; as discusse'd in Section 4.4.2.5, hot channel f actors are an integral part and key input to the orificing process, it is pointed out in response to question CS 760.78 that there is no evidence of premature cladding f ailure due to pressure of distortion and that experimental data as well as theoretical analyses Indicate that locate hot spots are within the range of current uncertainty factors. The FOTA's and WBA-45/46 experiments in FFTF will provide f urther data to update the magnitude of the " distortion" hot channel factors.
Therefore, the effect of bundle distortion on orificing has been and will be accounted f or through uncertainty factors rather than by explicitly considering a "def ormed geometry" in the flow allocation process.
QCS 760.71-1 Amend. 69 July 1982
Page 17 W82-0320 [8,22] 59 Ouestion CS760.72 Since the reacter woul d not be scrammed in case of a " gas lenker", has any consideration been given to the possibility that sodium might enter into the f uel pin thus experiencing a dif f erent temperature environment which in turn coul d require dif f erent constraints on cl ad tanperatures?
Resnonse The ef f ect of sodium entry on cladding temperature has not been considered.
Since the large dif f erence in temperature between f uel and cladding is a major unf avorable ef f ect on f uel rod design, substituting a " sodium gap" to a " gas gap" will only hel p the rod temperature environment.
in PSAR Soction 4.2.1.1.3.8 (pa ge 4.2-34), however, an example is given for a case of a CRBRP f uel element f alling late in lif e with a sizable cladding breach such that unif orm reaction of the f uel with sodium is assumed to occur and go to equilIbriun. This case is somewhat similar to that which might occur in the case of a " gas leaker" where a f uel/ cladding gap remains and sodium enters the f uel rod to fill, entirely or partially, the annulus. Under these conditions, the unif orm vol ume expansion of the f uel due to the f uel-sodium reaction was calculated to be nJ5% giV/V or about 1.67% diametral v' increase, which would have very little ef fect on f uel perf ormance; this will be conf inned in the Run Beyond Clad Breach program.
l l
[
QCS760.72-1 Amend. 69 July 1982 m - _ _ _ _ _ _ - _ _ _
p:gil W82-0298 (8,22) l43 Ouestion OCS 760.73:
Part An in the uncertainty analyses presented in section 4.4.3.2 of the CRBR PSAR, the rationale used to determine 2 and 3 uncertainty factors for thermal and hydraulic data is discussed. The discussion does not include a quantitative justification for non-statistical f actors nor does it provide Information about the methods used to determine statistical factors. Please Indicate for the data presented in Tables 4.4-18A through 4.4-31 which of the uncertainty f actors are determined statistically, which are based on engineering judgement, and which are based on manuf acturing tolerances. Also, for the tolerance based f actors, please provide quantitative data, i.e.,
actual tolerances. For the statistical f actors, please provide a detailed description of the methods i.e., the procedures used to arrive at specific uncertainty f actors for experimental and calculated data. For engineering judgement based f actors, please provide a discussion of the logic used to arrive at them.
Part B: In addition to uncertainties in material property data, design tolerances, and similar data there are uncertainties associated with the numerical methods (including models) used in the various computer codes. Are uncertainties in numerical methods (including models) included in the uncertainty factors presented in Tables 4.4-18A through 4.4-31? If uncertainties in numerical methods are included in the overall uncertainties, please provide a detailed mathematical description of the methods used to determine these uncertainties. If numerical method uncertainties are not accounted for, please explain why they are not.
Resnonse:
The Information requested is provided in the response to question QCS 490.36.
QCS 760.73-1 Amend. 69 May 1982
p:go 2 W82-0298 (8,22) 43 Ouestion OCS 760.74 According to the CRBR PSAR (Section 4.4.2.5) the procedure used to determine assembly orificing f or the heterogeneous core is based on a 3 loop naturag circulation transient with an imposed maximum coolant temperature of 1550 F.
Using this method, minimum required flows are calculated and used to detennine flows for 12 orificing zones. The above procedure resulted in a minimum core flow of 93.07% out of a maximum allowed core flow of 94% of total primary system f l ow. What would the result have been If, instead of using PEOC, THDV at 3a had been used to define the temperature T I N
Resoonse:
The information requested is provided in the response to question QCS 490.37.
QCS 760.74-1 Amend. 69 w
Page 18 W82-0320 [8,22] 59 Ouestion CS760 25 in section 4.4.2.6 of the CRBR PSAR there is a discussion of reactor coolant flow distribution at low flow conditions. It is stated there that a system of three computer codes (DEMO, COBRA-WC, and FORE-2M) was used to assess the ef f ect of all natural circulation cooling on the maximum coolant temperatures in CRBR. Please provide a detailed description of the geonetry modeled by each of the codes and of the data coup!Ing between them, i.e., output used as input, for the calculations discussed in the above section. The geonetry model Inf ormation shoul d incl ude the number and type of assembl ies modeled, the number of f uel or blanket rods in each assembly that are modeled explicitly, the LIM model, and the Uls model. Al so, pl ease prov ide detai l ed results, i.e., temperature distributions and flow rates as a f unction of time, for the ca!culations used to arrive at the conclusions presented. No experimental evidence of natural circulation cooling f or the CRBR heterogeneous core geometry is presented in this section. Are there any experimental data? If not, what type of experiments are planned to demonstrate the concl usions presented?
Resoonse The inf ormation requested is provided in the response to Question CS490.38.
QCS760.75-1 hnend. 69 July 1982
p293'$ b$$c@@@@ d@,$$)~42 ~
5 Ouestion DCS 760.76:
In section 4.4.2.8.5 there is a discussion of f uel-cladding gap ef fects on peak cladding temperatures reached during an undercooling transient. The discussion concludes that under LOF conditions with scram it is conservative to overestimate heat transfer to the cladding early in the transient, i.e., a higher peak cladding temperature will be calculated. Please provide quantitative justification, i.e., transient temperature results, for this conclusion.
Resoonse The inf ormation requested is provided in the response to Question QCS 490.39.
QCS 760.76-1 Amend. 69 May 1982
Page 19 W82-0320 [8,22] 59 Ouestion CS760.78 Has thermal and Irradiation induced distortion of core structures (i.e., fuel rods, ducts, LIM, and UlS) been accounted for in the hot channel f actors? Has it been accounted for in pressure drop calculations? If there have been calculations and/or experiments to determine thermal and Irradiation induced distortion of core structures, please provide details of them.
Resoonse Thermal and irradiation induced distortions have been accounted for in the l CRBRP design and analyses. Hydraulic calculations are performed using hot component dimensions, thus accounting for the effects of linear thermal i expansion on assembly pressure drop and flow rate. Thermal distortion of f uel rods has been calculated and is less than 0.001 inch which is a negligible effect.
Irradiation Induced distortion includes thermal creep, Irradiation creep and swelling, all of which are time and temperature dependent. Consequently, their effects only appear after significant Irradiation time and have no ef f ect near beginning-of-lif e. Furthermore, Irradiation induced distortions are limited to components near the core, i.e., the f uel column region of rods and ducts. The upper and lower lengths of the fuel rods, orifices and upper and Iower internals do not receive suffIclent fluence for irradiation creep or swelIing, and thermal creep ef fects are hydraulIcalIy instgnifIcant.
Calculations for fuel and blanket assemblies with 316 cladding and ducts showed the flow areas were reduced at end-of-life, resulting in a 1% f uel assembly and 1.8% blanket assembly total flow reduction. Locally, uniform everage rod spacing and uniform flow distribution are maintained until bundle-duct clearances are eliminated. Following loss of clearance the side channel areas reduce more than inboard areas due to fewer rod-duct support points than rod-rod support points, resulting in increased coolant flow and lower temperatures in inboard channels. Since inboard channels are always the hot channels, the net result is a negligible change in hot channel temperature.
Therefore, rod and duct distortions have minimal effects on hot channel temperatures and are adequately covered by the current hot channel / spot factors. Conf Irmatory data are expected f rom the FFTF FOTA's, BOTA's and the
- driver fuel surveillance program.
l l
! QCS760.78-1 Amend. 69 July 1982 rwmno
P:ge 23 W82-0320 [8,22] 59 Ouestion CS760.BD The sodium property data (G. H. Gol den and J. V. Tokar, "Thermophysical Properties of Sodium," ANL 7323, August 1967) ref erenced in Section 4.4 of the CRBR PSAR is not the most recent sodium property data available. Al so, the af orementioned data deviates signif icantly from that reported by A. Padilla (A. Padilla, "High Temperature Thermodynamic Properties of Sodium6 "
HEDL-T)E-77-27, UC-79b, February 1978) at temperatures above 1200 K (1700 F).
Why hasn't the newer sodium data been used for the calculations presented in the PSAR? Will the newer data be used in f uture CRBR analyses?
Resoonse As stated, the Padilla data deviates signif icantly from the Gol den & Tokar data only at temperatures above 1200 K (1700 F) which are well above any conditions considered in Section 4.4, hence, the temperatures of concern are in the range where the two sets of properties agree. The Project is required to use properties, data and correlations reported in the Nuclear Systems Material s Handbook (NSNH). The NSNH adopts the Golden and Tokar properties.
Checks were made of the dif f erence between Golden & Toker and Padilla val ues in the range of temperatures of concern to Section 4.4. The deviation for density and enthal py was less than 0.1%; f or specific heat ranged f rom 0.06%
(at 600 K) to 0.4% (at 1120 K); and viscosity is not reported by Padilla.
These deviations are an order of magnitude less than the coolant uncertainties used in design and thus are not significant.
l l
[
l l
l l
i QCS760.80-1 l Amend. 69
- July 1982 l- eawwaw,
. Question CS760.81 On page 4.4-13, SET is defined as the EOL temperature which, if maintained constant throughout lifetime, would cause the same EOL strain as the actual temperature / pressure history. SELT is defined as the SET which would cause, for the particular assembly relative behavior of cladding temperature and pressure throughout lifetime, an EOL cumulativo strain of 0.2%.
Table 4.4-2 shows that the SETS are greater than the SELTs for assembiles 11, 14, 36, 68 and 101 which seems to be in violation of the criteria set forth for the flow orificing. Furthermore, for several of those assemblies, the EOL strain is welI above the 0.2% limit (page 4.4-13).
Please provide justification why the temperature constraints can be violated and why the EOL strain is above the 0.2% limit.
Resoonse The SET was calculated for a temperature / pressure history corresponding to an assumed flow. The SELT corresponds to the SET temperature which must not be exceeded to keep the EOL strain below 0.2%. When the SELT is higher than the SET, it simply means that the assumed flow if higher than the flow necessary to stay within the 0.2% strain limit. Vice versa, when SET >SELT, the assumed flow is less than the limiting flow; this was the initial assemblies ill, 14, 36, 68 and 101 at the first step in the orificing process. Consequently, as Indicated in Table 4.4-2, for these five assemblies the EOL strain corresponding to the assumed flow exceeded 0.2%. The flow allocated to these assemblies was, therefore, increased (with respect to the assumed flow) to reduce both the SET and the EOL strain below their limiting values. At the completion of the orificing process, both the CDF limitation (which was more restrictive than the strain limitation) and the EOL 0.2% strain limitation were met for.all assemblles.
QCS760.81-1 Amend. 69 July 1982 82-0338
Pcg3 26 W82-0320 [8,22] 59 '
Ouestion CS760.83 It is stated that flow redistribution during a natural circulation transient is important. The core thermal head becomes important relative to the form and f riction loss across the core below 5% of f ull flow. Furthermore, it is stated that flow redistribution significantly reduces maximum core temperatures. Reference is made to publications by Agrawal using the SSC-L code and Khatib-Rabar using the CURL code to support this claim.
Both SSC and CURL use coarse core representations. The total number of assembly groupings is typically less than 20. Flow redistribution effects can therefore not be represented in an adequate manner.
The inference that the hot assemblies take away sodium from the colder assemblies in an amount suf ficient to avoid bolling in either assembly type, still needs to be supported. The flow redistribution is a " redistribution" of flow and not an overall increase in sodium flow. The very nature of a redistribution Implies that whatever the increase in flow allocation, it has to be balanced b,y a flow r eduction somewhere else. More refined analysis tools are necessary to study the natural convection flow redistribution, than the coarse models in SSC. Moreover,-SSC results also showed the occurrence of flow reversal, a f act the Project denies will happen in CRBR. Flow reversal is preceeded by flow stagnation. In the presence of a power source, any flow stagnation in coolant flow leads to very high coolant temperatures. Please provide additional information and explanation to support the claim that flow redistribution significantly reduces maximum core temperatures during low flows. Also, explain why flow reversal will not occur !n CRBR.
Resoonse This question is similar to Question CS490.38 and thus, the same Information would apply. As noted in the response to Question CS490.38, the CRBRP natural circulation analyses for t% PSAR have not taken credit for inter- and Intra-assembly flow and heat redistribution. However, sufficient work has been perf ormed to substantiate that this is a conservative approach with l regard to predicting maximum core temperatures. This is exemplified by the comparison of predictions to prototypic FFTF natural circulation test data as l provided in response to Question CS490.38.
Prediction of detailed core-wide Inter- and intra-assembly flow and heat redistribution phenomena will be part of the FSAR analyses. As described in the report supplied to NRC, " Verification of Natural Circulation in Clinch l River Breeder Reactor Plant - An Update", (Ltr. HQ: S:82:042 dated June 1982) a
! verified system of three computer codes (DEMO, Q58RA-WC and FORE-2M) will be l used for the predictions, and these codes do have a refined modeling of core Inter- and intra-assembly flow and heat redistribution effects.
l l
QCS760.83-1 Amend. 69 July 1982
P:ga 27 W82-0320 [8,22] 59 Flow reversal and/or flow stagnation is not anticipated to occur in the core for the highest temperature natural circulation case that has been addressed in the PSAR studies (I.e., scram from fulI power and flow with maximum decay heat loads plust all uncertainties included in the worst case direction). j Signficant flow reversal may only occur for postulated cases where natural circulation is initiated f rom Isothermal primary loap conditions. This situation has been precluded by design in CRBRP because of redundant power supplies and equipment qualification. However, even if it would occur, the maximum core temperatures attained should be less than those presented for the i heterogeneous core in (R8RP-ARD-0308.
I
)
QCS760.83-2 Amend. 69 July 1982
P ge 27 W82-0320 [8,22] 59 Ouestion CS760 84 From the presentation in the PSAR, it is not clear if the code package COBRA-DEMO-FORE-2M has been used only once or on a routine basis. The Project needs to clarify what data were obtained f rom this code package.
Resoonse This question is similar to Question CS490.38 and thus, the same inf ormation would apply. As noted in the response to Question CS490.38 and summarized here, the key points are that the CRBRP natural circulation analysis f or the PSAR have conservatively neglected the ef f ects of Inter- and intra-assembly flow and heat redistribution which would be predicted with the system of three codes (COBRA-WC, DEMO and FORE-2M). Thus, they have not been used f or the CRBRP PS AR predictions. This system of codes has been extensively verif ied, however, and will be used f or FSAR natural circulation analyses. Part of the verif ication ef f ort incl uded pre-test predictions by CRBRP of the prototypic FFTF natural circulation experiments; the post-test comparisons of the experimental data and the predictions showed good agreement f or both a high power /high flow central and a low power / low flow peripheral fuel assembly.
QCS760.84-1 Amend. 69 July 1982
P;ge 28 W82-0320 L8,22J 59 Ouestion CS760 81 It needs to be clarified what code package has been used to calculate the TELTs. The calculation of the maximum clad temperature during a natural convection transient is very sensitive to the model selection, since by def inition, the maximum temperature is the highest temperature assumed anywhere in the core. By using a coarse core representation in the transient analysis, one cannot expect to obtain better than regional averages. The Project needs to explain how much sensitivity analysis accompanied these TELT cal cul ations? The TELT calculations are very important since the TELTs severely constrain the bl anket perf ormance.
Resoonse TELT calculations were based on resul ts f rom FORE-2M analyses of the most l imiting f uel, inner bl anket and radial bl anket assembl ies. No coarse representation or regional averages were involved. FORE-2M provides a very detailed representation of the considered pin. The TELT eval uation adopteg in the CRBRP PSAR is quite conservative, for three main reasons: 1) the 1550 F maximum transient temperature is consi derably bel ow the boll Ing l imit; 2) as discussed in Section 4.4.2.5.2, all the other core assemblies of the same type were assumed to have the same transient behavior as the worst three assembi les analyzed and guidelines derived f rom these three assemblies were applied to all assembl ies; and 3) as stated in the response to Question CS760.79, current transient predictions are extranely conservative due to the numerous conserva-tive assunptions util ized in these eval uations.
It is to be expected that this conservative approach has led to some over-cool ing of the bl anket assembl ies; f inal design studies will better eval uate the above conservatisms and the results will be reported in the FSAR.
QCS760.85-1 Amend. 69 July 1982
P ga 33 W82-0320 [8,22] 59 Ouestion CS760.87 Explain the discrepancy between the 3-cr and 3- cr+ 15% overpower values f or the fuel and blanket assemblies. Should not the latter be 115% of the 3-4 values?
The CRBR PSAR, Figure 4.4-32 Indicates this ratio to be 1.12 and not 1.15.
The peak values of heat generation in the f uel and blanket pins are used to ascertain that there is no fuel melting (or blanket melting) for the 3-coupled with 115% overpower case.
Resoonso The (3e) uncertainty f actors include a " power level measurement and control system dead band" direct f actor equal to 1.03. The 15% overpower f actor is defined as 115% of the nominal rated power. Theref ore, the 15% overpower f actor is also inclusive of the 1.03 f actor and the ratio between 3 and 3*r +
overpower must be 1.12 as indicated in Figure 4.4-32 and not 1.15 so as to not account f or the 1.03 f actor twice.
I 1.
h
/
QCS760.87-1 Amend. 69 July 1982
Page 34 W82-0320 [8,22] 59_ _ _ _ _ _ _ _ _ _ _ _ _ _ _
OuestIon CS760.88 Provide details of the assumptions made on fuel performance parameters (such as restructuring, the gap conductance, the maximum inner-surface cladding temperature, etc.) in ascertaining that there would be no melting for the 3-plus 1155 power case.
Response
No assumptions "per se" were made in the power-to-melt analyses. The LIFE code with its models for fuel restructurir.g, gap conductance, etc. was used a priori in these analyses.
l
- ~
~
>+
s ;~
QCS760.88-1 Amend. 69 July 1982
%M. _. _ - _ _ _ _ _
PCge - 1 (82-0358) L8,22J #91 Ouestion CS760 91 The guar'd vessel which is described in Section 5.2.1.4, performs an Important Initiative f unction for pipe leaks or breaks below the level of the core.
Guard piping around the inlet and outlet pipes are necessary to f ulfill this f unction, particularly for the primary inlet pipes, and are shown schematl-cally in Figure 5.2-1 of the PSAR but they are not mentioned in the text.
\'
Responie' Sect $n 5. 1.4 has 'been amended to describe the reactor guard vessel pipes.
~
~
t 4
i i
a F.
'l 7
h t
44 I
l i
QCS760.91-1 Amend. 69 I- 3 - July 1982 f
, ._ , - - - r p-.- -. - e ,
p::ga 1 W82-0358 (8,5) 42 Riser Elastomer Seals The balance of the seals on the riser assembly operate at temperatures below 125 F.
Uoner Internals Structure Jackina Mechanism The UlS Jacking mechanism utIIIzes metal buf fered seals in the 400 F areas.
These seals are part of the mechanical assemblies. The seals will be removed with components at the appropriate maintenance period. Elastomer seals are located in the cooler regions, have a service life of five years, and will be replaced using hands-on maintenance.
Llauld Level Monitor Ports Pluas Four of these components, operating at 400 F, are located on the reactor vessel head and provide receptacles for holding the liquid level monitors.
Three small port plugs are attached to the top surfaces of the closure head rotating plugs by partial penetration welds, two on the Intermediate and one on the large rotating plug. One large port plug is bolted to the top surface of the large rotating plug and is sealed to the plug by double metal "0" rings. The seals remain attached to the port plug during Installation and removal. Because the port plug remains stationary relative to the head assembly, the metal "0" rings beneath the plug flange are not expected to require maintenance.
5.2.1.4 Guard vessel The guard vessel provides for the retention of the primary sodium coolant in the event of a leak in the portion of the primary coolant boundary which it surrounds. The guard vessel geometry assures reactor vessel outlet nozzle submergence after such a leak which will maintain continuity in operating primary coolant loops to provide core cooling. The guard vessel also prcvides a unifor.m annulus for in-service Inspection of the reactor vessel, with l clearances that preclude contact with the reactor vessel and piping under l
accident conditions. Insulation for the reactor vessel and a heating system l
for the reactor vessel to be used prior to sodium fill and during prolonged shutdown are also mounted upon the guard vessel.
l The Reactor Guard Vessel has pipes which surround the primary inlet and outlet piping to an elevation which matches the top flange of the guard vessel, 788' 0" (shown in Figure 5.2-1). The piping is welded to nozzles in the guard vessel shell. The piping elbows are formed from five mitered sections of piping. The mitered sections are welded together to form a 54" radius on the outlet pipes and a 36" radius on the inlet pipes. The piping is made from 1" '
plate which was rolled and welded with one longitudinal seam per section. the seams have been staggered on alternate sides of the mitered elbows and straight sections. The inside diameter of the outlet piping is 53" and the
! Inlet piping is 41". The outlet piping has a 4'5.7" straight section above the elbow. The Inlet piping has three 9'5" shell courses and one l'6.5" shell j course above the elbow. All sections are welded together with circumferential i seams. The inlet and outlot piping has a plate flange at the top with a 60" ;
0.D. on the outlet and 48" 0.D. on the inlet. The flanges have been drilled l and tapped with eighteen, 3/4" diameter holes equally spaced to provide for !
l i
l 5.2-4a l Amend. 69
" Y '982
'~
p:ga 2 h82-0358 (8,5) 42 insulation support. The inlet piping has been provided with a seismic restraint which allows thermal growth by providing a support with a concentric gap around the piping but limits lateral pipe movement during a seismic event.
The seismic restraint is attached to the guard vessel shell. All welds and NDT have been made in accordance with the ASME Code and Guard Vessel Equipment Specification.
The maximum and minimum widths of the radial gap between the guard vessel and the reactor vessel have been conservatively calculated, taking into account all relevent f actors such as tolerances on the diameters of the two vessels, permissible out-of-roundness of the two vessels, possible deviations from straightness due to manuf acture and subsequent operation, thermal expansion, Initial deviations in the alignment of the two vessels, etc. The transporter for the television camera will be designed to accommodate itself to this maximum possible range of gaps as it moves in the space between the two vessels.
5.2-4b Amend. 69 July 1982
P:ge - 4 ( 82-0358) [ 8,22] # 91 Ouestion CS760 1Q1 The 6" continuous drain nozzle in the steam drum presumably leads to the line to the topping heater. Is this flow cut of f by closing the steam drum drain i sol ation val ve? What system (s) would cause this valve to close and under what circumstances?
Resoonse Flow to the topping heater is cut of f by closing the steam drum drain i sol ati on val ves. The steam drum drain valves are closed by SGAHRS Initle-tion, SWRPRS Initiation, or low steam drum pressure. The drum drain isolation valve actuating signals are Identified in PSAR Table 5.5-5. The drum drain control logic is shown as part of the SWRPRS control l ogi c di agram, PS AR Figure 7.5-6, sheet 3 of 6.
l l
1 i
l QCS760.101-1 Amend. 69 July 1982
P;ge - 5 (82-0358) [ 8, 2 2 ] # 91 Ouestion CS760101 in Section 5.5.3.1.5.2, page 5.5-18n it is impl ied that the pressure rel ief valves in the evaporator lines are not designed to operate at the high temperatures that those at the superheater outlet must endure. Is It possible, under some accident conditions, for evaporator to dry out and actually superheat its inventory? Would such high temperatures damage those rel ief val ves and lead to f urther problems?
Resoonse The event which produces the highest steam temperature at the evaporator outlet is the loss of of f site power (U-18). Upon loss of power, the recirculation pump trips, initiating natural circulation in the recirculation loop. The sodium pump switches to pony power and continues to circulate liquid sodium on the shell side of evaporator, heating the water / steam in the evaporator. The evaporator does not dry out at the inlet, but the outlet steam can be superheated to temperatures of 800 to 900 F for a short time.
This thermal transient is included in the equipment specification for these val ves. Analyses have been made to show that the valves will not be damaged by such transients.
{
QCS760.104-1 Amend. 69 July 1982 1 -