ML20054J776

Forwards Responses to NRC 820430 & 0514 Requests for Addl Info.Responses Will Be Incorporated Into PSAR Amend 69, Scheduled for Submittal in Late Jul 1982

ML20054J776
Person / Time
Site:	Clinch River
Issue date:	06/25/1982
From:	Longenecker J ENERGY, DEPT. OF
To:	Check P Office of Nuclear Reactor Regulation
References
HQ:S:82:058, HQ:S:82:58, NUDOCS 8206290545
Download: ML20054J776 (33)

Text

.

Department of Energy Washington, D.C. 20545 Docket No. 50-537 HQ:S:82:058 JUN 2 51982 i

Mr. Paul S. Check, Director CRBR Program Office Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Comission 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 and May 14, 1982 This letter formally responds to your request for additional information contained in the reference letters.

Enclosed are responses to C astions CS760.8,12,13,14,16,19, 20, 22, 25, 26, 31, 33, 34, 42, 53, 54, 56, 92, 96, 111, 119, 124, 125, 128, 133, and 156; which will also be incorporated into the PSAR Amendment 69; scheduled for submittal later in July.

Sincerely, JothR.Longene r Actlng Director, Office of the Clinch River Breeder Reactor Plant Project Office of Nuclear Energy Enclosures cc: Service List Standard Distribution Licensing Distribution DoD /

8206290545 820625 PDR ADOCK 05000537 A PDR

Pigo 1 (82-0374) [8,22] #99 Question CS760.8 PSAR page 4.2-15 and Table 15.1.2-2 on PSAR page 15.1-53, list the acceptance criterion f or the " Extremely Unlikely Fault" of " Postulated Accident". The aim is to preserve coolable geometry, and the acceptance criterion is listed as coolant saturation; that is, no boiling. This appears to be a criterion that may not be adequate of itself for ensuring coolable geometry in the event of a reactivity insertion accident. While calculated cladding temperatures have a high degree of uncertainty, nevertheless gross f uel expulsion is known to occur in TREAT overpower tests at peak cladding temperatures as much as 200oK or more below coolant saturation, and in many cases 100oK or more below the cladding design temperature guideline. This behavior has now been confirmed in a slow transient (HEDL's W-2 test). The phenomena involved in these overpower events are exceptionally complex, but recent calculations Indicate a f uel enthalpy limit with appropriate allowance f or uncertainty may be an appropriate acceptance criterion to ensure coolable geometry for over-power conditions. A criterion of that sort is now used f or the same purpose in LWR reactors.

The applicant is requested to comment on the adequacy of the no-boiling criterion as an acceptance criterion to ensure coolable geometry in the light of the TREAT test experience.

Resoonse The areas addressed in this question regarding the adequacy of the no-bolling criterion are similar to those expressed in Question CS490.23, and reference is made to the response already provided, it was shown in that response that there was considerable margin f or preserving coolable geometry in the applicable design basis accident, although the event involved a conservatively postulated set of conditions including a large overpower transient, delayed shutdown and release of fission gas f rom all the pins in the af fected assembly. The emphasis of Question CS760.8 is on the revelance of observa-tions from TREAT and other tests to the no-sodium-bolling guideline. In general, the tests impose f ar more severe conditions on the f uel than would be applicable f or the design basis events. Care should be taken in applying results f rom unterminated transient tests which exceed conditions typical of a design base to the terminated conditions in a design base. Undercooling tests have consistently shown that bolling precedes major damage to the cladding and the f uel . Overpower tests have demonstrated the satisf actory perf ormance of CRBRP f uel within the design envelope. Where pin disruption has occurred in overpower bef ore the coolant reaches saturation, the f uel linear power rating has f ar exceeded the predicted levels in CRBRP design basis events. For example, the test which is noted in the question, W-2, was an unterminated HCDA oriented test, and the peak linear power rating at the time of major cladding f ailure was 41.2 kw/ft (1351.5 W/cm). This should be compared with a value of approximately 16 kw/ft (525 W/cm) maximum, corresponding to 15% over-power at which a trip would have occurred in CRBRP f or a similar reactivity insertion rate.

QCS760.8-1 Amend. 69 July 1982

Paga - 1 (82-0358) [8,22] #89 Question CS760.12 Since small ramp rates take a long time to develop, justify the assumption that the hot fuel rod is limiting, since the longer thermal Inertia of the blanket rods does not apply here.

Resnonse For small ramp rates that take a long time to develop, the worst overpower condition that can exist would be operating at just under the 15% overpower trip level. If the power exceeds this value, core shutdown would be initiated through the plant protection system. Thehighestmaximumcladdingtemgerature during sustained ogeration at 15% overpower would be 1496 F (3r)/<1300 F (nominal) and 1418 F (3a)/<1300 F (nominal) for the fuel and blanket assemblies, respectively. With regard to fuel temperatures, all fuel and blanket rods have been designed to operate at this condition with no molten f uel as described in Section 4.4.3.3.6. After the overpower condition is terminated through scram, higher temperatures Ln be attained for the blanket rod cladding in the subsequent shutdown heat removal phase due (in part) to the large thermal inertia of the blanket rods alluded to in the question.

This mechanism is described in Section 15.1.4.1 where a maximum blanket rod cladding temperature of 1587 F (3r)/<1300 F (nominal) is reported (attained

~47 seconds af ter shutdown).

QCS760.12-1 Amend. 69 July 1982

Prga - 2 (82-0358) [8,22] #89 Question CS760.13 Section 15.2.2.2 analyzes a 604 radial movement (stick slip) Incident. The analysis does not distinguish between primary or secondary scram. (Only one temperature curve is given). Provide analysis for this transient, IIsting the appropriate primary and secondary trip functions.

Re'sponse For a 60$ step reactivity insertion the power increases in almost step f ashion '

from 100% to over 200% as shown by Figure 15.2.3.3-3. Both the primary and secondary high power trip signals are significantly below the increased power level and thus, both trips would occur simultaneously. The table below summarizes results for the highest cladding temperature hot rod in FA-52 considering both primary and secondary scram (each separately).

MAXIMUM TE WERATURES (3-)

REACTOR POWER CLADDING FUEL COOLANT SHUTDOWN SYSTEM AT TRIP A B C A B C A B C Primary 115% 1491 0.63 1.6 4576 0.53 1.3 1417 0.63 1.6 Secondary 1225 1544 0.83 2.0 4752 0.63 1.7 1467 0.83 2.1 A - maximum 3 e hot spot temperature attalned, F.

B - time to reach maximum temperature, sec.

t C - length of time temperature is above initial steady state value, sec.

It should be noted that occurrence of a 60( step reactivity insertion combined with failure of the primary scram would be less probable than an extremely unlikely category event in which case the primary shutdown of a Safe Shutdown Earthquake (Section 15.2.3.3) would envelope the consequential core damage.

QCS760.13-1  ?

Amend. 69

_ _ _ _ _ _ _ - - - _ er0rl DE8 -

P;ge - 3 (82-0358) [8,22] #89 Question CS760.14 Reactivity insertions during startup must be more closely assessed. Under these conditions many of the PPS subsystems are bypassed and the PCS musT ou relied upon to mitigate the transient. In light of this, discuss the ef fects of PCS maloperation or operator error under these conditions.

For the cold sodium Insertion event:

a. The transient is analysed using instantaneous core inlet temperature and flow rate changes. Shouldn't this be analyzed with more realistic (i.e., ramp type) changes in these conditions?

b. Although the loop transient time is 60s, the actual core inlet temperature will rise slowly. Therefore, shouldn't the transient be analyzed longer than 60s? (Especielly for secondary PPS trip).

c. With a minimum Doppler coef ficient, can you use B001 values for sodium density feedback coef ficient?

d. What are the mechanical ef fects on pins due to cold sodium insertion?

Resoonse Although specific subsystems may be bypassed, suf ficient protection in both the primary and secondary scrm systms stili exist to ensure that damage Iimits are not exceeded. Speci fIcally, post tive f Iux to delayed flux, positive modified nuclear rate, flux to pressure (autmatically reinstated above 15% power), flux to total flow, and startup flux subsystes would all trip for unacceptable positive reactivity insertions occurring during startup caused by PCS maloperation or operator error.

With regard to the cold sodium insertion event, the following are the responses to the item-by-Itm questions:

a. The controlling mechanism of this analysis is the positive reactivity from Doppler feedback which occurs when the fuel is initially cooled by the lower sodium inlet temperature. Since it was not known "a priori" that this event would be benign, it was assumed that condi-tions which would result in the most rapid power increase would be limiting. Thus, a step change in the inlet coolant tmperature was conservatively used in the analysis. A slower change of inlet temperature would not increase the transient power level but would decrease the rate at which the power increases.

b. It is estimated that the reactor inlet temperature would increase at a rate of less than 200F in 300 seconds which would not cause any significant increase in core structural damage. This i ncrr,re of i inlet tmporature would attenuate the equil ibrium power atiai ned f rom that shown in the analyses (Figure 15.2.3.1-1). If the analyses were carried out until exact equilibrium is reached, the maximum tempera-tures would still be less than those shown for full power steady state operati on.

QCS760.14-1 Amend. 69 July 1982 P2-2 7.* . -

Pcge - 4 (82-0358) [8,22] #89 I

c. Yes, but the transient would be less severe because less positive Doppler f eedback would result (see item a).

d. The worst case hot rod cooldown rates shown by Figure 15.2.3.1-3 are very similar to those experienced during a normal scram and insignificant damage is incurred. This is due to cladding being very thin which mitigates thermal shock damage.

1 QCS760.14-2 Amend. 69 July 1982 82-0374 _ _ .., -- _. . _ _ . . _ _ . . _ _ -. - ~ _ _. -- --- -

P ge - 6 (82-0358) [8,22] #89 Questfon CS760.16 On page 15.2-2a of the PSAR, the first sentence of the fourth paragraph states, "The first two of the above restrictiens are obvious". Subsequently, a third restriction is alluded to . It is not at all clear what the ref erenced restrictions are. Please clarify.

Resoonse The treatment of Fuel-Cladding Mechanical Interaction (FCMI) was described in response to Question CS490.10 and at the May 12, 1982 meeti ng with NRC.

Revisions to Section 15.2 of the PSAR will be made to clarify the treatment of FCMI by August 15, 1982.

QCS760.16-1 Amend. 69 July 1982 E2-C_.______________________

Page - 9 (82-0358) [8,22] #89

.Qybstfon CS760.19 in Section 15.2, present or ref erence the f uel and cladding temperature histories f or the worst f uel and blanket rod during the U2b event. Also, present or ref erence a synopsis of this event and its consequences, as was done for the other overpower transients in this section.

Resnonse The U2b event is an enveloping condition f or various types of accidents that can be postulated which insert positive reactivity to the core (either in steps or ramps). This event is described in PSAR Appendix B, Section B.I.2.2.2. It should be kept in mind that this event assumes the f ailure of the rod block at 103% power in either the manual or automatic modes of the PCS. The evaluation results in an analysis of core temperatures at 15%

sustained overpower for a period of 300 seconds. If the power increase would be any larger than this magnitude the reactor would automatically be shut down due to scram from an overpower trip signal of the plant protection system.

Equilibrium temperatures Indicative of steady-state operation at 115% power are reached within the 300 second hold even f or the large diameter blanket size rods.

As indicated in response to Question CS760.12, the maximum cladding tempera-ture during the sustained operation at 15% overpower would be 14960F (3a)/<13000F (nominal) and 14180F (35)/<13000F (nominal) for the f uel and blanket assemblies, respectively. With regard to f uel temperatures, all fuel and blanket rods have been designed to operate at this condition with no molten f uel as described in Section 4.4.3.3.1. Af ter the overpower condition is terminated through scram, higher temperatures can be attained f or the blanket rod cladding in the subsequent undercooling phase due to stored heat and decay heat ef fects as described in Section 15.1.4.1. In th i s post-shutdown period the maximum blanket hot rod cladding temperature of 15870F (3a)/<13000F (nominal) can be attained af ter-47 seconds af ter shutdown.

i QCS760.19-1 Amend. 69 July 1982

kagS-10(82-0358)[8,22]#89 I Question CS760.20 The design guideline for cladding temperature for anticipated events is listed in Tabl e 15.1.2-2 as 15000F. However, Table 15.2-2 lists temperature criteria

. presumably cladding) of 14500F and 14000F (blanket rod reactivity insertion).

Please explain the relation between the limits listed in these two tables.

Please also provide in detail the basis for these two sets of guidelines.

Response

Table 15.1.2-2 does provide design guidelines which includes the 15000F cladding temperature f or both f uel and blanket f or anticipated events. The basis for the guidelines was explained in response to Question CS490.21 and addressed at the February 25, 1982 meeting with NRC.

There is no Table 15.2-2 in the PS AR. Originally, in earlier revisions to the PSAR, there were such conservative cladding temperature limits as those described. However, analyses consistent with those now appearing in Section 4.2 of the present PSAR showed the limits to be overly conservative and thus, is the basis f or the 15000F guideline now appearing in Table 15.1.2-2.

QCS760.20-1 Amend. 69 July 1982

l Paga - 12 (82-0358) [8,22] #89 I

Question CS760.22 Discuss how the changes to a heterogeneous core af fect re ctivity ef fects for the operating basis (OBE) and saf e shutdown (SSE) earthquakes.

Resnonse The change to a heterogeneous core has no significant ef fect on reactivity changes f or the operating basis (OBE) and safe shutdown (SSE) earthquakes.

This is because parameters which control the response, the radial reactivity worths and mechanical dimensions, are very similar. Uncertainties in the core mechanical response to the earthquake and core motion reactivity worth f actors will mask any reactivity dif ferences between the homogeneous and heterogeneous core designs.

l t

l i

I QCS760.22-1 l Amend. 69 July 1982

_ - . - - _ _ _ _ _ . ~ -

_ _7, Pags - 12 (82-0358) [8,22] #89 Questfon CS760.25 Provide the basis for the assembly power distribution numbers.

Response

The CRBRP power distribution calculations and uncertainties are discussed in depth in Section 4.3.2.2 of the CRBRP PSAR.

QCS760.25-1 Amend. 69 July 1982

---

.---c--gve--w-- - - - w v -

y-- w---y--7 gwym.y.--.,y. ---.wg .r --v., %w-1,-- ym,mm--g--- y ww , -w>-. ----,pme-- wm-7---,m.----w- y. - 1 y-TNT

~

Page - 16 (82-0358) [8,22] #89 I

• l l

Ouestion CS760.26 What is the ef fective bypass inertia (L/A) ef f.?

Response

The total reactor flow during a flow coastdown from full flow is governed by the inertia in the pumps (the principal contributor) as well as the fluid inertia. The contribution of the fluid inertia in the reactor is small. For example, the L/A for the flow path between the pump and the reactor vessel inlet is 94.4 ft.-l whil e th at f or the f uel region i s <3 f t.-l . The effective L/A for the " bypass" region would be even smaller because of the lower L/A value and the small fraction of total flow associated with the bypass region.

Flow redistribution between the four parallel flow paths in the DEMO model (fuel, inner blanket, radial blanket and " bypass") occurs when the total flow derivative is very small and inertial ef fects are negligible. Thus, the ef fective bypass inertia is ignored (i.e., set to zero) and the computailon of the redistribution of flow between the four parallel flow paths within the reactor is accomplished by balancing the flows between the four paths such that each path has an identical pressure drop. The pressure drop f or the bypass is the sum of gravitational and form pressure losses.

l l

i i

l l

QCS760.26-1 Amend. 69 July 1982

Pag 2 9 (82-0374) [8,22] #97 Duestion CS760.31 The most notable change in the new natural circulation analysis (CRBRP-ARD-0308, Feb.1982) is the change in pump coastdown t!me f rom 55 to 120 seconds. This change is attributed to a substantial decrease in low speed f rictional torque seen in prototype tests. Since the previous estimates based on design requirements have been replaced by the new best estimate torque, will all future pumps be required to conform to this behavior (i.e., will any pump with a higher frictional torque be rejected)?

Resoonse The coastdown requirements specified for the CRBRP primary sodium pump / pump drive motors were stated in Ref erence CS760.31-1. These requirements were established in 1975 and were based on the results of analyses which examined the ef fects of various coastdowns on the plant's natural circulation capability as welI as transient ef fects on structures such as the UlS. These requirements provided a basis f or the design of the pump and pump drive system. Since the design of the pumps as welI as the drive motors (and associated bearing and seal assemblies) is now completed, analyses are based on the actual coastdown characteristics shown by tests of the pump / pump drive prototype.

The availability of prototype pump water test data has made it possible to develop and refine predicted pump characteristics for opertion in sodium.

Correlations for pump head-flow and pumping torque characteristics were developed f rom water test head-flow and ef ficiency data. A val ue f or pump Inertia was obtained by using water test coastdown speed vs. time data, and approximation for loss torque at high pump speeds and the equation of motion f or the pump. A trial-and-error procedure was then used to select a set of coef ficients for frictional torque correlations at low pump speeds that most closely matched the measured results of water test coastdown runs. This method ensures that inaccuracles in the pumping torque correlation due to departures f rom similarity at low speeds wilI be absorbed in the frictional torque correiatton.

The prototype pump coastdown characteristics developed from water test results represent an accurate model of sodium pump performance in CRBRP. Coastdown tests conducted during the Maximum isothermal System Tests (MIST) for FFTF primary and secondary main coolant pumps showed only slight variations in pump-to-pump coastdowns. For exampie, at 30 seconds af ter trip, the respective pump speeds in primary loops 1, 2 and 3 were 155 RPM,156 RPM and 155 RPM. The greatest dif ferences between measured pump speeds occurred below 4% speed immediately prior to the pumps stopping. The variation in coastdown time for the primary pumps was approximately 155, while for the secondary pumps it was 12%. Since a similar correspondence in coastdown performance can be expected for CRBRP sodium pumps, variations in f rictional torque large QCS760.31-1 Amend. 69 July 1982

Pcge 10 (82-0374) [8,22] #97 enough to warrant rejection are unlikely to occur. If, however, pump coast-down characteristics in the plant pumps result in pump coastdowns that occur earlier than used in the analyses, the Project will demonstrate adequate natural circulation conditions still exist or provide modifications to the pump coastdown characteristics to achieve acceptable natural circulation conditions.

Reference CS760.31-1 R. R. Lowrie and W. J. Severson, "A Preliminary Evaluation of the CRBRP Natural Circulation Decay Heat Removal Capability",

CRBRP-ARD-0132, November,1977.

QCS760.31-2 Amend. 69 July 1982

pig 2 10 (82-0374) [8,22] #97 Question CS760.33 in the revised natural circulation report (CRBRP-ARD-0308, February, 82) by Severson, et al., it is stated that no credit is "taken for inter- and intra-assembiy fiow and heat redistribution." This Is consistent wIth the DEMO-REY 4 which has a fixed flow fraction to each group of assemblies. However, it is also stated that "the code calculates flow redistribution between the four regions." This appears to be a major change f rom the previous conservative approach.

What would the hot-spot temperature be if the fixed flow fraction were maintained throughout the transient?

Response

A four region reactor model which provides for flow redistribution between the f uel assemblies, inner-blanket assembiIes, outer-blanket assembiles and bypass channel has been added to DEMO since the publication of DEMO-REV4 (WARD-D-0005, REV 4, January,1976) . Flow distribution between the four flow paths is computed by equating the pressure drop for each path. Flow redistri-bution is a more physically accurate approximation to a two dimensional reactor flow model than the fixed flow fraction model. Thi s f low redi stribu-tion model was used in the DEMO analysis (which generated the total reactor flow rate vs. time) reported in CRBRP-ARD-0308.

The DEMO analysis employing the four region reactor model which accounts for redistribution between the above mentioned regions results In slightly lower reactor flows than that which would be computed by the earlier fixed flow f raction model . The reason is that the f uel assembly flow which establishes the plenum to plenum AP in the fixed flow fraction version of DEMO will be higher in the redistribution model thus increasing the dynamic pressure losses and at the same time reducing the thermal head. Thus, the redistribution model produces a conservatively low total reactor flow.

. The highest core temperatures at the hottest locations of the hottest rods for l the f uel, inner-blanket and radial blanket assemblies were then computed usf rg l FORE-2M based on fixed fractions (equal to their initial fractions) of the i total reactor flow. In this hot channel analysis, no credit was taken for j inter- and intra-assembly flow and heat redistribution.

No analysis of hot-spot temperatures using a forced flow fraction has been performed which could be directly compared with the valves reported in CRBRP-ARD-0308, however, because of the increased flow associated with using a fixed flow fraction the hot-spot temperatures would be lower for such a case and therefore the valves reported in CRBRP-ARD-0308 can be considered bounding for this aspect.

QCS760.33-1 Amend. 69 July 1982

Pcga 11 (82-0374) [8,22] #97 Question CS760.34 There is a built-in time delay in reverting f rom perf ect mixing in the upper plenum to strati fied f low. What is the basis for the specific delay and what is the ef f ect if no delay is added?

Resnonse The DEMO upper plenum uses two distinct modes of mixing, the f ully mixed and stratified mode, to simulate the mixing in the reactor upper plenum. At the initiation of a transient, perfect mixing was assumed in the outlet plenum.

The transition from perfect mixing to stratified flow mode depends on jet height. A Jet height of 20 f t. was assumed in the analysis reported in CRBRP- ARD-0308. The stratified flow portioin of the model begins with the filling of the lower region of the outlet plenum by the cold sodium existing from the top of the chimney. During this time and until the time the hot / cold sodium interf ace reaches the bottom of the outlet nozzle, the reactor vessel outlet temperature is set equal to the hot outlet plenum temperature at the start of the stratification. This time period is the delay referred to in the above question. When the hot / cold sodium interf ace rises above the bottom of the outlet nozzle, the outlet nozzle temperature starts to decrease as a f unction of the cold flow area in the outlet nozzle. The cold f low area is that portion of the outlet nozzle covered by the cold fluid. The remaining outlet nozzle area is assumed to be covered by the hot plenum fluid. The outlet nozzle temperature is calculated by assuming perfect mixing of the hot and cold sodium. The resulting temperature is, therefore, given by:

TVO = TCH + (THL - TCH) EXP (-(t - ft )/r) (1) where T is the vessel outlet temperature VO T is the mixed mean temperature of fluid entering the upper plenum fh$mtheUlSchimney T

HL is the hot sodium temperature in the upper plenum t is the transient time t fis the time when the stratified cold sodium rises to the bottom of the nozzle r is a convective time constant relating to the rate of change of the hot / cold sodium interface from the bottom to the top of the outlet nozzle.

, When the cold fluid rises to the top of the outlet nozzle, the flow through I the nozzle and its temperature will be the same as the cold sodium. The level of the hot / cold interface is assumed to be constant and the temperature above the interf ace is hot and below the interf ace is cold.

It is clear from this model description that a delay between switching f rom f ully mixed to stratified mode of calculation is required by virture of the physical dimension of the plenum and the fluid transport time. It should be pointed out that the model assumptions used are quite conservative. Credit QCS760.34-1 Amend. 69 July 1982

us usa, Pago 12 (82-0374) [8,22] #97 for heat transfer to the cold fluid from both the hot plenum metal and the plenum sodium was not accounted f or in this stratified mode plenum calculation. In addition, the values of tf nel reported in the CRBRP-ARD-0308 report were so afnd morev used in the fve hsis conservat an the actual data. This was done deliberately to allow the cold sodium entering the plenum to leave the outlet nozzle earlier. Since the sodium exiting the reactor plenum rises abruptly into the vertical run of the primary hot leg pipe, this would result in a lower overall primary loop thermal head. Thi s lower thermal head in turn, would result in a more conservative estimate of the primary loop flow.

'the ef fect of neglecting the delay totally would mean that the cold sodium entering the plenum will appear instantaneously at the reactor vessel outlet nozzle. This is not only overly conservative, It is physically impossible.

The two mode model described above is both adequate and conservative f or the analysis of a natural circulation transient event.

i l

QCS760.34-2 Amend. 69 July 1992

Prg3 - 1 [82,0357] 8,22 #92 l Duestion Cs760.42 in Section 15.4.1.1.5 (page 15.4-12), it is stated that "... cladding- def ects (in the f uel zone region) of 0.1% of the f uel rods...would result in an end-of-life plutonium concentration of 0.1 ppm in the primary system sodium."

Please provide the reference from which these data were taken.

Resnonse The design basis limit for plutonium release to the primary coolant is 100 ppb (0.1 ppm). This has been related to a continuous 0.1% failure rate of the i fuel rods for 30 years based on a Pu escape rate coefficient of 9 7/ x 1014 atoms /cm2 sec if each of the f ailures correspond to a 0.03 Inch hole.

~

5 QCS760.42-1 ',

July 1982

o. .--.

Pcg3 - 6 [82,0357] 8,22 #92 l

Ouestion CS760.53 This section on extremely unlikely events treats only the first case on the core component pot. Have all of the other cases been treated and what are the j results?

Resoonse PSAR Section 15.7.3.1 discusses the event of leakage of the sodium from a core component pot (CCP) suspended in the EVTM, resulting in overheating of a contai ned f uel assembly. A single ever.t is considered, however, extensions of it beyond its expected termination are considered. The event would result in partial melting of the cladding and f uel assembly ducts but the columns of f uel pel lets would stay intact. Hypothesized extensions of this event discussed in this section of the PSAR are the collapse of the pellets to the bottom of the core assembly and the redistribution of the pellets outside the f uel assembly duct in the bottom of the CCP. In all cases, the core assembly materials would be contained within the CCP. The only material which would escape f rom the CCP would be fission products which are volatile bef ore the maximum temperature is reached by the f uel. While this temperature is high and the release would include many elements, most would be deposited on the colder Interior walls of the EVTM.

The release of materials outside the EVTM pressure boundary is through EVTM elastomer seals which are more than 6 ft below the bottom of 1ae suspended CCP. The highest seal temperature f or the event (including the hypothesized extensions) would be 2600F. The worst case release would be of elements which are volatile above this temperature and this is the enveloping release f or all cases described in PSAR Section 15.7.3.1. Is it the same release as for the unlikely event described in PSAR Section 15.5.2.3, single f uel assembly cladding f ailure and subsequent fission gas release during refueling (in the EVTM) . That event is referenced in PSAR Section 15.7.3.1.3 for the of fsite exposure f rom the extremely unlikely CCP leak event. (It should be noted that the treatment of the isotopes which are volatile above 2000F is described in the response to NRC Question 001.212 (15.5.2.3.2). The 2000F temperature considered there and in PSAR Section 15.5.2.3 is equivalent to the 2600F real temperature in PSAR Section 15.7.3.1, since the only isotope with a melting point between the two temperatures, Iodine (see PSAR Table 15.7.3.1-3), is included among the elanents considered).

QCS760.53-1 Amend. 69 July 1982

P gn 4 WB2-0320 [8,22] 59 Ouestion CS760.54 Regarding Section 15.7.1.3 on sodium leaks. Sometimes (as in the' Phenix Reactor), the IHS springs a leak as a result of the strains which occur with shutdown and startup. If the primary loop pumps come up before the secondary pump loops, then it may be possible for contaminated primary sodium to be driven into the secondary loop. Please discuss this possibility with an undetected IHX leak.

Resoonse -

i The IHTS is designed to insure the pressure in the IHT5 is always higher than the PHTS by at least 10 psid. lHTS design also includes a low IHTS/PHTS Ap alarm on the Main Control Panel (MCP) to alert the operator of a problem.

It is possible to bring up the Primary Heat Transport System (PHTS) pumps before the Intermediate Heat Transport System (lHTS) pumps and achieve a pressure in the PHTS higher than the lHTS, however, CRBRP operating procedures would have to be ignored or violated and the low IHTS/PHTS Ap alarm would have to f all to alarm or be ignored when it is received.

Assuming normal lHTS pressure when the PHTS pumps are started, the PHTS flow would have' to be increased to >85% flow bef ore the PHTS pressure would exceed IHTS pressure.

Therefore, if one assumes an undetscted leak in the IHX combined with several operator errors and alarm f ailures, it is possible to get primary sodium into IHTS system, although it is considered a very unilkely event.

Assuming the PHTS to' IHTS leak did occur, the IHTS boundary would prevent any release of radioactivity to the atmosphere and the health and safety of the public would not be endangered from the event.

i '

i l

QCS760.54-1 ,

Amend. 69 July 1982

Pcgs 5 WB2-0320 [8,22] 59 Ouestion CS760.56 The PS AR cl aims, in Section 15.7.1.6 regarding NaK spills in the EVST system, that the NaK will be non-radioactive. It is possible, however, that some radioactivity could get into this NaK by such sources as the 1% failed fuel or carryover sodlum f rom the f uel transf er. The cover gas could also become contaminated by leakage of fission gas f rom f ailed spent f uel rods.

a. Does CRBR have Instrumentation in the EVST to detect radioactivity in the cover gas and in the NaK? .

b. Does the EYST have any Instrumentation for local detection of activity, temperature, or local bolling within a possibly partially blocked subassembly?

Resoonse The NaK will be non-radioactive because it is kept separate from the EVST sodlum which is the primary coolant. EYST heat is transferred from sodium to NaK in a heat exchanger (see PSAR Section 9.1.3.1). The EVST sodlum will contain radioactivity from carryover of primary sodium during ref ueling and perhaps f rom f uel assembly fission gas releases in the EYST. The activity will not be transferred to the NaK because of the sodium and NaK separation.

The NaK is kept at a higher pressure than the sodium at the heat exchanger to prevent contamination of the NaK in the event of a leak (see PSAR Section 9.1.3.1.3). The plant design provides the capability to detect NaK leaks into the EVST sodlum so that NaK levels and pressures could not decrease to the point where radioactivity could leak from the EYST sodium into the NaK coolant.

The plant design provides the capability to detect radioactivity in the EVST cover gas and sodium coolant. The cover gas activity is continuously monitored by maintaining a flow of cover gas through a radiation detector.

The sodium activity is monitored by sampling the sodium periodically for laboratory analysis of its radioactivity concentration.

Individual core assemblies in the EVST are not instrumented because the conditions are not severe enough to require it. Core assembly power levels (20 kW maximum power per assembly versus several MW in the reactor) and cle'dding temperatures (p6000F versus ae11000F in the reactor) are both low compared to reactor operating conditions, and the cooling method is such that the of fect of partial bl.ockage of a subassembly is minimized. The flow of EVST sodium coolant is outside the core component pots (CCPs) in which core assembl ies are stored. Heat romoval by this coolant is f rom the CCP walls.

There is no significant mixing between this sodium and the sodium in a CCP.

Core assembly decay heat is transferred to the CCP walls from the f uel rods by conduction and by convection of the sodlum in the CCP. Since the driving f orce f or the convection f low is the f uel temperature, any partial blockage causing a temperature rise would be self-correcting by increasing the flow through the assembly rather than diverting flow through a lower resistance asserbly as would occur with forced flow cooling.

QCS760.56-1 Amend. 69 July 1982

Pag 3 - 2 (82-0358) [8,22] #91 Ouestion CS760.92 Frequent changes in power level (e.g., following plant trip) can entall swings of up to several hundred degrees in the coolant temperatures. Thus, large thermal stresses may appear in the reactor vessel, coolant piping, or other components, which may eventually threaten the system integrlty. What are the methods and models presently used to determine these temperature swings? How are they f actored in to provide assurance that they conservatively cover the entire duty cycle of the plant?

Resoonse The basis for NSSS component structural evaluations is the plant Design Duty Cycle. The transients specified for the structural evaluation of plant components are generally the results of the DEMO code output. The DEMO code is described in Appendix A.21 of the PSAR.

The plant Design Duty Cycle transient events were selected to be representa-tive of operating conditions, which are considered To occur during plant operation and which are suf ficiently severe or frequent to be of possible signi fIcance to components. These transients are based on a conservative envelope of plant operation and were developed primarily for use in component stress analyses. The events, as well as their associated frequencies, are based on LWR, FFTF, and fossil plant experience; system and component reliability estimates; and engineering judgment. The description of the transient events that are used in CRBRP component analyses, and their assigned f requencies are presented in Appendix B of the PSAR.

The analysis of each Design Duty Cycle event is based on conservatively biased paraneters for each system and/or component using the DEMO computer code. The analysis of each event was perf ormed such that the rates of change and total range of temperature change were conservatively computed for each run of piping or component. The rated power Initial conditions included hot and cold leg temperatures biased upward 200F to account for temperature measurement and cointrol uncertainties. For some transients an alternate set of initial conditions were used which employed hot to cold leg temperature dif ferences of 3000F f or the primary system and 3400F for the intermediate systems (with a total temperature dif ference of 3900F between the PHTS hot leg and lHTS cold leg). In addition, some of the plants sensible heat (piping heat capacity, for example) was neglected, energy delivery rates f rom reactor to SGS were maximized by conservatively high pony motor flows and reactor upper plenum stratification was included or not included to assure conservatism. Other plant parameters such as system pressure drop, pump Inertia, pump loss torques, decay heat, pony motor speed, PPS actuation time, rod reactivity and delay and valve stroke time and capacity are individually biased in the most conservative direction. These conservatively developed duty cycle transients are then included in all NSSS component / piping histograms as discussed below.

All normal events (and frequencies) are applied to each component in the system at thei r specified f requency. Upset events are grouped into a smaller set of umbrella events (typically 10 to 13). Less severe transients are combined with more severe transients by increasing the event frequency of the umbrella event, such that the f requency of the umbrella event equals the sum of the f requency of that event and the frequency of each event umbrellaed QCS760,92-1 Amend. 69 July 1982

us uso, Paga - 3 (82-0358) [8,22] #91 under it. Emergency events are incorporated by determining the most significant event and applying it five times (evenly spaced in time) plus two consecutive occurrences of the most severe event or combination of events.

All events that are defined as a f aulted event f or a component are included in that component evaluation.

These duty cycle transients are in general applied approximately evenly in time over the thirty year life of the CRBRP, divided into ten three year peri ods. Worst case sequencing is assumed within these periods consistent with physical possibility. This combination of events results in a conservative histogram for component evaluation.

In conclusion, due to the combined use of:

o design thermal transients based on worst case plant conditions f or the component under evaluation for that event; o conservative estimates of events and their associated f requencies; o conservative umbrellaing techniques; and o applying a worst case histogram.

the components have been evaluated against temperature swings that wil l i conservatively cover the transients that are expected to occur in the plant.

l k

QCS760.92-2 Amend. 69 July 1982

Paga - 2 (82-@)M@) L8,f#2.] ##8 Ouestion CS760.96 The recirculation pump is described as single speed, yet will experience varying mass flow rates at different power levels and will go through varying speeds while coasting down after a trip. How does the pump head vary with flow rate and speed, i.e., what does the homologeous pump curve look like?

Resoonse The pump curves are attached as Figures QCS760.96-1, 2. The curves were established by actual vendor pump testing. The recirculation pump is a single speed unit with a Design Speed of 1794 rpm, a Design Flow of 5920 GPM, and a Design Dynamic Head of 397 f t. This design point is marked on the attached curve. The pump conditions given in the " Coast Down Curve" are measured on the suction side of the pump.

l l

l l

l I

! QCS760.96-1 Amend. 69 July 1982 82-0358

P;gs - 6 (82-0358) L8,22] i91 Figure QCS760.96-1 PUMP CHARACTERISTICS CURVE g Y urs*

.m ,

EFE 5 a:: see .

se

,e Mo4D re 6e goe .

I .

t w y3.. , nt.

g ee se

o. W ,y ge no n G 8000 toes go'e e yeee gg$so edeo Pese Soes 9880 65&Le# $ Pd4 NIWETS Figure QCS760.96-2 COAST DOWF CURVE s.

Pump CONDITIONS TEMPEJt4 TURK 568

• F PREssultE 1906 Psi A CAPhCITY M2p GPM

(.

RPM,'P M 3 .

(a 100 o) 1 -

I N

O A A G $ lo IL I4 Tims (sac)

QCS760.96-2 Amend. 69 July 1982

. _ .. .. ~. - . _ _ _ _ _ - - . - - ~ . - . _ - - - - ._-. . .. . . . _ - - . - _ - - - _ - ...

Page 6 (82-0358) [8,22J #93 .

t cuamtIon cs760111  !

What is the flow area through the SGMRS vent valves when they are f ully open't l I

Response .

The flow area of a f ully open SGMRS vent valve is 9.01 square Inches.

l l i i l

i

i i  ;

1 i

1 h

4 ., ,

s

, , s i

f

- 1

- ~

I l '

j i

QCS760.111-1 Amend. 69 i -

' July 1982 1

' M , , _ , - , _ _ , , . _ - , , , - _ , , . - , _ .

._,_,..}._,____,,.

page 3 t:2-0358 (8,22) 94 Ouestion CS760.119 in Section 5.6.2.3.2 of the PSAR, !t is stated that the DHRS is not designed to provide heat removal by natural circulation. Since the overflow concept requires pumping in order to function within its design objectives please provide a discussion of the following:

a. How is the DHRS diverse f or electrical power (onsite and of fsite failure)?

1

b. Other potential common mode f allures.

Response

DHRS is not designed, nor intended to be, diverse for electrical onsite and offsite failures. The diversity provided in the plant f or onsite and of fsite electrical f ailure is the natural circulation capability through the the PHTS/lHTS/SGS/SGAHRS. DHRS provides dlyersity for those f alIures whIch could dirrupt heat removal through the IHTS and steam generator syster.

QCS760.119-1 Amend. 69 July 1982

p:g3 5 WB2-0358 (8,22) 94 Question CS760.124 in reviewing pump coastdowns how were of f acts of extended coastdown considered? How were dif ferences between " Identical" pumps considered in your analysis?

Rannonse it is assumed that the above question relates to a natural circulation event.

Extended coastdowns f or main coolant pumps beyond those presently used in the DEMO plant simulation code enhance the natural circulation decay heat removal capability of CRBRP. The critical period f or the natural circulation decay heat removal mode occurs shortly af ter the primary and intermediate pumps have stopped. At that time, thermal driving headc necessary to promote adequate flows are required to prevent resulting core temperatures f rom exceeding acceptable limits. Maximum core temperatures reached during the natural circulation transient are largely dependent upon the decay power. Extended pump coastdowns allow time f or reductions in both decay heat and reactor sensible heat and consequently provide greater margins to bolling in the core.

The analysis of the natural circulation event used pump coastdown characterls-tics developed from prototype pump water tests.

Dif f erences between " Identical" pumps were not considered in analysis of the natural circulation event. Dif ferences In pump-to-pump perf ormance during plant operation are not expected to be significant enough to justify inclusion of separate models f or Individual pumps. Further discussion of this point is provided in the response to Question CS760.31.

QCS760.124-1 Amend. 69 July 1982

us uso, paga 6 t]2-0358 (8,22) 94 Question CS760.121 in reviewing the progression of uncertainties how were the following items considered:

o Pressure drop core pump piping IHX valves o Flow Coastdown pump Inertia pump f riction dif ferences between " identical" pumps o S trati f ication upper plenum piping o intra-assembly heat and flow redistribution o inter-assembly flow redistribution o Heat losses to outside o Bypass flow o Decay heat.

EB122010 The Individual data sources f or the current natural circulation assessment are discussed in the response to Question CS760.28.

In developing the transient response of the CRBRP, each of the Design Duty Cycle events, has a set of parameters individually chosen at their limits and a series of models individually incorporated or deleted to the DEMO computer code that are appropriate f or that duty cycle event. The Individual uncer-tainties requested are discussed in the following table.

I QCS760.125-1 Amend. 69 July 1982

....s.,

page 7 [22-0358 (8,22) 94 Parameter Consideration o Pressure drop core Maximum or minimum used as required to assure conservatism.

piping included with lHX in analysis.

pump Head flow characteristic assumed at the minimum for all analyses and locked rotor resistance assumed to be at the maximum.

lHX Piping and IHX pressure drops combined and chosen as maximum or minimum as required to assure conservatism.

Val ves Maximum or minimum used as required to assure conservatlsm.

o Flow coastdown pump Inertia Chosen consistent with the maximum or minimum specified requirement for the pumps.

pump f riction Chosen consistent with the maximum or minimum spect fled f or the pumps.

di f f erences between None - See response to Question CS760.124.

" Identical pumps" o Strati f ication upper plenum Fully mixed or stratified model used to provide the most severe transient.

- piping See response to Question CS760.28.

o intra-assembly No credit taken.

hest and flow redi stribution o Inter-assembly No credit taken, flow redistribution o Heat losses to No credit taken, outside o Bypass flow No uncertainty applied.

o Decay heat Maximum or minimum chosen to provide most severe transient.

QCS760.125-2 Amend. 69 July 1982

ucuna paga 5 WB2-0358 (8,22) 94 Question CS760.128 During the descent f rom 10% power, what are the safeguards to prevent unacceptably high usage of feedwater from the protected water storage tank?

Resnonse The Auxillary Feedwater System (AFWS) is not used during normal descent from 10% power; therefore, no water is drawn f rom the protected water storage tank (PWST). The AFWS is only operated when the Steam Generator Auxillary Heat Removal System (SGAHRS) Is Initiated. The PWST water use is discussed in PSAR Section 5.6.1.3.9.

QCS760.128-1 Amend. 69 i July 1982

page 8 WB2-0358 (8,22) 94 Question CS760.133 Very little basis is given for the assumed frequency of events. Please categorize the f requency as to source (in order of pref erence),

a. Commercial reactor experience

b. Test reactor experience

c. Other data

d. Engineering judgment Resoonse This response is prepared assuming the question refers to Table 5.7-1,

" Preliminary Summary at Heat Transport System Design Transient".

The f requency for the overall plant duty cycle events was initially determined f rom a review of available commercial reactor experience and specific meetings with commerical reactor vendors. The selection of specific duty cycle events and the allocation of frequencies to the specific events was developed based on engineering judgment and an understanding of the design dif ferences between an LWR and an LMFBR. The structural evaluation of the ef fects of each Individual duty cycle event on each reactor plant component was analyzed by grouping the duty cycle events for each component into a single transient event (umbrella) which is conservatively representative of the group with the f requency of the entire group. Since the individual transients have dif ferent ef fects on dif ferent components, the umbrella transients and the transients grouped under that umbrella are developed dif ferently for each component.

Dif ferent f requencies are therefore assigned to each umbrella transient for each component. The selection of umbrella transients, and the groupings under each umbrella transient, was based on preliminary analysis of the ef fects (temperature, pressure, and resultant stresses) of each duty cycle transient on each component. This engineering ef fort resulted in the frequencies shown in Table 5.7-1 of h.e PSAR for each major component of the Heat Transport System.

QCS760.133-1 Amend. 69 July 1982

Page - 4 (8,22) #95 o

Question CS760.156 Discuss the leak test method used following replacement of the equipment hatch. How were the permissible leak rates determined?

Resoonse The leak test method to be used f or periodic testing of the equipment hatch af ter completion of each ref ueling will be local pneumatic pressurization of the dual compressible hatch seals utilizing the in-place test connection.

Determination of the actual leak rate will be by measuring the pressure decay for a prescribed time duration, in the case of anticipated actual replacement of the equipment hatch, special Installation checks such as dye and chalk tests for alignment verification coupled with pneumatic pressurization of the dual seals will be perf ormed.

Permissible leak rates to be finalized in early 1983 will be consistent with the acceptance criteria for type B tests as delineated in 10CFR50, Appendix J.

It should be noted that the equipment hatch is always closed during all Reactor Plant operations and is only opened for ref ueling and/or maintenance activities.

0 QCS760.156-1 Amend. 69 July 1982 82-0374 _ . _ _