ML20012C277

SER for BWR Coolant Chemistry Loop to Be Installed & Operated in Mitr

ML20012C277
Person / Time
Site:	MIT Nuclear Research Reactor
Issue date:	03/09/1990
From:	Ames M, Baeza J, Driscoll M MASSACHUSETTS INSTITUTE OF TECHNOLOGY, CAMBRIDGE
To:
Shared Package
ML20012C270	List: ML20012C269 ML20012C277
References
MITNRL-031, MITNRL-31, NUDOCS 9003200430
Download: ML20012C277 (43)

Text

._

i SAFETY EVALUATION PIPORT FOR i

THE B'a'R COOLANT CHEMISTRY LOOP (BCCL)

MITNRL-031 t o be i

installed and operated in the MITR ltarch 9, 1989 Prepared by:

M. R. Ames J. L. Baeza M. J. Driscoll l,

O. K. Harling G. E. Kohse J. O. Outwat e r for Review by the illTR Saf eguards Committee

_m 9003200430 900309 ADOCK 0500g O

PDR P

e.

'O 1.

INTRODUCTION 1.a. Foreword This is the saf ety evaluation report for an in-pile loop facility designed to simulate core coolant chemistry of a boiling water reactor (BVR).

The loop will be used to carry out research into the effects of radiolysis and chemical additions (hydrogen, for example) on coolant composition (oxygen and hydrogen peroxide concentration, for example),

as part of a program to evaluate the ef fect of the coolant environment on the corrosion of ma t e ri als.

Loop construction and operation are funded by the Electric Power Research Institute and the Empire State Electric Energy Research Corporation.

The objective of this report is to present a summary description of the design and operating procedures of the BVR Coolant Chemistry Loop (BCCL) in sufficient detail, and with supporting analyses, to demon-strate that it can be operated safely within the envelope of applicable MIT Reactor Technical Specifications.

Considerable reliance upon, and reference to, the SER for the companion PWR loop will be made (1),

because of the many f eatures shared in common (in pile containment thim-ble, in-core heater bath).

As will be seen, the major difference addressed in this report is the need f or steady-state in-core boiling in the BCCL.

l l

l (1) HITNRL-020 (2/13/87), and supplements thereto, in their most recent revisions, PCCL (4/19/88) and Large Circulating pump (10/24/88).

~

2-1.b. General Description of BWR Loop 1.b.1. Introduction The design of the BWR loop has evolved through the operation of two full-scale laboratory mockups, by Oliviera (1) and by Baeza (2).

i Many of the components will be used as-is in the in-pile version, now under construction.

Figure 1.1 is a schematic of the major f eatures of the BWR Coolant Chemistry Loop (BCCL) which will be operated in the MIT Reactor.

As can be seen, it is a once-tnrough system high-purity feedwater is pumped from out-of-pile through the in pile Zircaloy core section, where approximately 10 weight percent boils; the ef fluent ~ is then separated in an outlet plenum and taken out-of-pile for analysis, and ultimate re-use.

It is similar in many respects to the by-now-familiar PWR loop shown in Fig.

1.2.

In particular, the in-core section is physically identical.

The major dif ference elsewhere is the substi-tution of titanium for stainless steel and Inconel as the principal con-struction material.

1. b.2. Design detaf is Figures (1.3) through (1.8) show additional details of the BCCL; Table 1.1 summarizes system parameters of interest.

The overall layout is sketched in Fig. l. 3.

Starting on the rights cold feedwater is l

(1) C.

Oliveira, " Design and Proof-of-Principle Testing of an In-Pile Loop to simulate BWR Coolant Chemistry," Nuclear Engineer's Thesis, MIT Nuclear Engineering Department, Decembe r 1987.

(2) J. Baeta, " Refinement of an In-Pile Loop for BWR Chemistry Studies,"

S.M. Thesis, MIT Nuclear Engineering Department, January 1989.

3 N-18 n

DETECTORSO I )

(

I:

ELECTRIC FEEDWATER HEATER e-0UTLET PLENUW/

,']

REGENERADVE HEAT e

STEAM SEPARATOR EXCHANGER am,

9 Q *- MITR-il SECONDARY

+ COOLANT CHEMICAL ECP at Ha a INACTION HEAT o

.'~

m

,g PONTg EXQ4 ANGER h

O HIGH PRESSURE PUWP

==

PLENUW

_.y N

M, RECOMBINER/

y s

HEATED DEMINERAUZER IN-CORE SECDON CHARQ1NG TANK (Hg COVER GAS)

FIGURE 1.1: SCHEMATIC OF IN-PILE BWR COOLANT CHEMISTRY SIMULABON LOOP 1

\\

.r.

I MAGNEDRIVE l

PUMP i

l O

m DISCHARGE TO l

1 L_q V

' SAMPLING SYSTEM POD I

CONTROLLED I

t

+= CHEMISTRY

?q i

- 1 COOLANT e

3) I ADDITIONS L_____

s l

/

CORE TANK BOUNDARY COPPER i

l SHOT I

GD l

A 'O i

LEAD BATH

_IN Ti T1JBE l

ALUMINUM i

THIMBLE

(

20 kW HEATER

=

J l0 STAINLESS i

STEEL INLET

[

MITR-Il PLENUM /1 1

REACTOR l

7 ZlRCALOY-2 CORE l

h HEATED SECTlON w

3 lN LEAD BATH l

l i

l i

I i ' DUMMY FUEL v

ELEMENT NOT TO SCALE l

FIGURE 1.2: SCHEMATIC OF LOOP TO SIMULATE P_VULCOOLANT CHEMISTRY

______-r-_____y------,i--m

--vvy--*me=-+a-

+

5-TABLE 1.1 SELECTED BCCL DESIGN AND OPERATING PARAMETERS _

Outlet Plenum Volume (water + steam) 1200 cc Plenum Pressure and Temperature 1000 pois, 545'T Downcomer Plenum Volume 55 ce Feedwater Flow Rate 3000 cc/ min Core Section (Zircaloy U-Tube) Inlet T 530'T Length of In-core U-Tube 48 in Maximum Core Ef fluent Steam Quality 15 w/o In-Pile Electric Heater Rating 20 kW Teodwater Electric Heater Rating 20 kW Maximum H Content in Teodwater 2 cc/kg In-core Dose Rates, neutron 4 x 10' R/hr =

• 1.6 w/g (Mean, in H 0) gammas 4 x 10' R/hr = = 1.6 w/g Operating Temperature of Hotwell Tank (max) 100'T l

Rating of Heat Exchangers Regenerative 36 kV Non-regenerative 7.5 kW Void /Reflood Reactivity of In-Core U-Tube 0.0421 Af=5.4e l

l l

l i

i i

l i

IN-PILE:

OUT-OF-PILE LEGEND I

_L g FLIIWEEEER suter K E I T

[:":] I N TunesE i

m

_L 1 OUTLET

I TI N

XING TEE PLENUM

.Q h

1

-m gl l

anor-carstaar l

j n;

mEastavuut rLoAT i

.f N

f I

LEVEL rnEssunt siasic m

i ETER g

- I

_ CHARGNdG

~

hh HEIT88ETER 4 L 81HX i v

-t i

O

=

gU-

~

h nacx-cutssunc 0

VAVE h 2 m w

=

CDRE

/

NN HKI g

%T g,

1 seest M RIC seITR-II oo o vatvt

-K ATERS gay w filT TO SCALE I

FIGURE I.3: SCilEMATIC OF BCCL CONFIGURATION: OVERALL l_AYOUT

1 1200 psi pumped by a charging pump rated at at a rate of approxi-mately 3 liters per minute through a regenerative heat exc ha ng e r, and electrically heated feedwater heater to the in-pile thimble.

The same total mass of water, as steam and liquid at = 300'C, is then brought back out of the MITR core tank, where the two ef fluent streams are com-bined, lose heat in the regenerative and non-regene ra tive heat ex-changers, and are returned to the hotwell tank.

The similarity to the PVR system has already been noted, but there are differences the main ones are summarized in Table 1.2.

These dif-forences and other details of the BCCL are elaborated upon in the para-graphs which follow.

1.b.3. Containment thimble Of perhaps greatest interest are the dif ferences in the contain-ment thimble, since it is the actual interf ace between the MITR core and coolant and the subj ect experiment.

The material of construction (alu-minum) and all shapes / dimensions are the same as the PVR thimble except (see Fig. 1.4):

(a) the PVR shot port has been eliminated and replaced by a circumfer-l ential (metal) gasketed, bolted flange.

(b) a second elliptical thimble section has been added to the bottom to i

increase the dose rate in the loops' ex-core "downcomer plenum."

l The streaming snout terminates several inches above the fueled region of the MITR core, is beveled at its bottom, and positioned such l

l-l that " shadowing" of adjacent core fuel assemblies, from the eme rgenc y cooling spray, is minimized:

see Figs.

1. 5 a nd 1. 6.

Thus no interf er-1 ence in either normal core coolant flow or emergency core spray reflood should take place.

l

I t

TABLE 1.2

SUMMARY

OF SIMILARITIES AND DTTTERENCES BrNEEN BVR AND PVR LOOPS 111!E Similarities Differences Aluminum Same !!11 gas, pres-Flange in transition containment sure relief system, region, additional short thimble main structural com-elliptical section for ponents.

enhanced streaming to plenum, no shot port.

In-core Identical same heater.

Zircaloy-2 U-tube in heater bath Ti sheath, place of Zircaloy-4 Thimble

• Larger exit plenum in internals BWR, with steam / water interface and level measurement system.

• No recirculating pump in BWR loop.

Insulating shot (quarts gravel) in place of conducting shot (copper).

Local sample extrac-tion tape for out-of-pile analysis.

Makeup / letdown Substant' ally different l

system BWR flow rate

• 600 x higher, and = 10 w/o of effluent is steam.

BWR has heat ex-changers, feedwater heater.

Charging, dis-Similar tank construc-Vater is returned to charge tanks tion.

charging tank and re-cycled.

Operating con-

• same temperature Substantially differenti ditions/ pro-Steady-state boiling cedures in-core Substantially lower pressure (1000 psi vs. 2200 psi).

• Shorter runs.

• Attended runs.

Handling facili-Identical ties and equip-ment

)

i l

1 t

t 12

-)

4-8 l

+

DIMENSIONSi INCHES

,s

-s 4+

4-BED OF QUART 2 PEBBLES

/

(SHIELDING AND INSULATIDN) 48

/

CONTAINED IN INTERNAL CANNISTER W

149 v

101 i

DUTLET PLENUM 36 y

p DQVNCOMER PLENUM 17 hil qs v

t v

N 9-3/8 a N

-FLANGE

\\

^

36 GAMMA, NEUTRON STRCAHING CHANNEL IN-CDRE ELLIPTICAL SECTION P

v v

FIGURE 1.4: BCCL THIMBLE

-to-SC ALE - lili DIMENSIONS - INCHES GAMMA POD 1-11/ 3 2

~ 3/8 IN-CORE ELLIPTICAL 1/2 SECTION u

il il 1-1/2 GRID o

il i

N0ZZLE x -

s I

DUMMv FUEL ELEME',~

(actual placement overlaps; shown here for clartty)

FIGURE 1.5: BCCL IN CORE TANK: RELATIVE PLACEMENTS

11 i

SCALEi lil FUEL ELEMENT 3

CUTLINE GANMA POD j

i S

N0ZZLE OUTLINE

\\

'\\,

^

CORT CDRE

\\

POSITION POSITION m

I' 31 FUEL THIM3LE FUEL

{

4-INCH SECTION l

w

{

IN-CDRE ELLIPTICAL

/

SECTION

\\

f CORE DUMMY-1 POSITIDN ELEMENT v

I P

EGURE 14LBCC L IN CORE LANKETOE VIEW- - -

1.b.4 Cheaf stry control f eatures Figure 1.7 shows the cheatstry control system for the loop.

Of main interest here is that for some tests a small concentration

(= 0.2 ppa = 2 cc/kg) of hydrogen gas (in a hydrogen-helium mixture to give a total hotwell tank overpressure of 15 psia) will be main-

=

tained.

A systes for addition of ppb ppm levels of trace elements to the feedwater is also provided, but will probably not be used in the first round of tests.

In place of the deposition monitor of the PWR loop, we instead have two Na! detectors to measure N-16 activity in the steam and water phases extracted f rom in pile.

s Finally, Fig. 1.8 shows the special chemical sample extraction features built into the BCCL.

Here small side streams (roughly equiva-lent to the PWR loop let-down stream) are cooled in situ (by cold-sinking them to the thimble wall) and then led out of pile for analy-sis. Table 1.3 summarizes the initial campaign planned for the loopt a

i succession of short runs (on the order of hours) to study the radiolysis chemistry of BWR coolant particularly H 02 2 concent rations, since this constituent is thought to aggravate stress corrosion cracking.

'l

L

.j};I

• i'

i,!

-[!l'

\\fli:

,ti o

s l

nen ca s

y, em e

ve m

ae t

sc m

ac

.ca s

ta I

L SL a

.o a

e I

O s

n A

i s

s T

t eq 2

a E

m e

s D

a c

\\' v e.o e

s e

sa c.

s n

N u;

1=.*

a u

s%

O i

I t

n.

T ew AM w

s ms RE e

a UT s

u a

S j

s[

um G Y s

I ne FS ur rl NL e'

OO s

CR P

t LT R _.

D CN t _

l CO l

A e.

R W p

C I

I D

e n

s FY E

e o

s 2t o

OR E _

N M#

m_s t

t CT f

]

c v.

s S

s A

1 Wsmt 0

s I

r n

TI e

0 i

g v

e AM o

t.

_ s s

n l

. m.,

E u

Ml u

i e

m L (o m

t=

c k

a u

m,

i u

i l

s e

C t

l m

gT a= a l

s C

i a

C:

1 c

i m.

S0 t

uy

, u, n

rB i

l m

st sl

, i e

7:

l t A.

up e

u, u

s nn I

e eG uD e

n cK s

I A

)f n

v R

r 6 a

i 4

I

((

L G

)

g l

\\)

l r

6 s

i a

{(

a tv

- (

tr cA R l

l te -

inwE u

A a-ieT N

n t

ttA e

h" d

s ae a

sat w

e irs w

l 4 AC t

lI a

t e

ic

(

MeM s

i tt t

i e

1

<l!

i

'i!i!ll'

\\l

!ll!j

!lllll

I "EASURE"E'JTS ON-L!NE ION CHROMATOGRAPH H

Ha, 0:, ;H, a l

(

STEAM i

g a

f V

LXTRACTED SAMPLE H0 2 2 AA NAA OUTLET g PLENUM NYLON TUBING EFFLUENT 1 150'F ECP PROBE -%, I g

/ }'

/

GAMMA DETECTOR '-

4 6

r 00WNCOMER PLENUM l

ALUMINUM BLOCKS COLO-5 INKED

%4 TO Al THIMBLE '4ALL (IMMERSED IN MITR COOLANT)

I I

/>l t/

ECP : ROBE QUARTE-L:NED METIL CA?!LLARY

'ic : to Scale

~

FROM CORE FIGURE 1.8: FEATURES RELATED TO CHEMICAL ANALYSIS

1 i

TABLE 1.3 i

PROPOSED BWR LOOP TEST HATRIX 4

WATER CHEMISTRY

• i A

B C

D let 2-wk 2nd 3rd 4th I

sequence sequence sequence sequence of runs of runs of runs of runs ppm H2 FULL HWC al/4 Hj al/2 H 2 (Loop feedwater) Normal WC (-0.15 ppa) to give to give 1

10-20 to give specified specified Runs +

ppb H2 approx.

interme-interse-

-230 mV diste ECP diate ECP ECP 1

Downconer exit HO 2

core effluent HO 2

(a) Full Flow (transit time)

X X

X X

(b) 1/2 Flow X

X X

X (c) 1/4 Flow X

X X

X II Bypass (i.e. core at very low power & flow)

(a) Full Flow X

X X

X (b) 1/2 Flow X

X X

X (c) 1/4 Flow X

X X

X III Rerun I (core) but at (a) Lower heat flux X

X X

X (b) Lower dose rate (MITR power)

X X

X X

IV Downcone r (a) 1/2 Dose Rate X

X X

X (b) 1/4 Dose Rate X

X X

X HEASUREMENTS:

ECP, H0, H,

0, N-16 in steam and water, for all 22 2

2 runs; also pH, conductivity and cations (by AA) an1 anions (by Ion Chromatography)

NOTE:

The two-week-long BWR experimental campaigns are planned ta run in alternation with approximately one-month-long PWR loop runs.

Thus, this set of BWR tests could be completed in aboa:

six months under such a scenario.

)

i

. l I

i 2.

SAFETY-RELATED ASPECTS OF DESIGN AND OPERATION i

2.a. Conventional Considerations i

2.a.l. Introduction Accident scenarios for the Bk1 loop parallel those of its Pkt counterpart in most respects, dif fering primarily in details which have no significant impact on ultimate safety.

In the subsections which follow, four such categories are evaluated (1) Loss of coolant flow (2) In-thimble leakage (3) Nydrogen release (4) Radiation exposure s

By inf erence, items not addressed here are judged not to be of safety significance - with the exception of reactivity-related aspects which raise sufficient questions to deserve attention in a separate sec-tion of their own (See Section 2.b).

2.a.2. Loss of coolant flow As wi t h the Pkt loop, this is the type of incident having the highest probability of occurrence.

The end result is the same the in-core lead heater bath temperature increases until the redundant over-temperature trip cuts off electric heater power, following which gamma heat is safely rejected by passive means (radiatirm and conduction to t he thimble wall, and thus to the MlTR-II coolant).

The intervening sequence differs somewhat in detail, however.

In the Pkt loop loss of flow ref ers to the canned-rotor circulat-ing pump, since its out-of-pile charging / pressurization pump cannot pro-vide enough coolant by feed and bleed to remove significant heat by forced circulation.

In the BWR loop, on the other hand, the chargin.;.

t

.o

)

O i

pressurisation pump supplies all of the coolant flow.

So long as this pump operates, even if the loop is fully depressurized, some 15 kW of y

heat removal capacity will be available.

If interrupted for more than I

roughly one minute, however, the loop will boil dry, and overtemperature will follow shortly.

While the pump is accessible, the short time interval before restart is required limits the nature of corrective actione to those which can be automated, such as restoration or substi-tution of power following an outage.

Since the BWR loop is made up of insulated components inside the

(

in-pile thimble, its internal heat rejection capacity is limited - un-like the PWR loop, whose " steam generator" section can easily and

^

paesively reject 15 kW at 600*F.

Thus for all heat loads above about s

5 kW, the BWR loop relies upon the regenerative and (especially) the l

non-regenerative heat exchangers in its out-of-pile auxiliary systems, for which the ultimate heat sink (shared with the out-of-pile test tank) is a ~ coolant system linked to the MITR-II secondary cooling system.

l Failure of the circulating pump in this heat sink loop will therefore require shutting down the loop.

Since several minutes' warning will be available, this can easily be done by manually shutting off power to the in-pile electric heater using the switch provided in the reactor control room.

Although the final line of defense is automatic heater cutoff l

l, triggered by lead bath overtemperature, it is worth noting that the first line of defense - loop operator action - will be enhanced for the BWR loop, since it will be staf fed by a crew ef attendants during most operations. Unlike the PVR loop, which is designed to operate at steady state unattended for month-long runs, the BVR loop will be run in a

n 1

t o

t. '

succession of short runs during which a number of on-the-spot chetnistry analyses will be carried out.

Thus it is quite likely that corrective 4tr. ion can be initiated to interrupt potential accident scenarios before any automatic protective systems are required to operate.

'2.a.3.

In-thimble leakage The concern here is whether a rapid large leak could overpres-L surize the aluminum containment thimble.

It is concluded that this will not occur for several reasons:

.i (a) During normal operation the BWR loop contains approximately the same inventory of liquid water (600 g) as the PWR loop; its maximum inventory, when the outlet plenum is flooded (no steam / water interface) is 1000 g.

The BWR loop, however, operates at a much lower pressure (1000 psia vs. 2200 psia).

(b) The cold' thimble wall can rapidly quench any st'eam it ' con-tacts:

the entire loop inventory, if need be, within a few seconds (see appended estimates).

1 (c) As with the PWR loop, ultimate ' protection is provided by the thimble relief valve (set to lif t at 20 psia) and rupture disk (designed to break at 65 psia).

The large area lid on the top of the thimble would also leak through its gasketed seal at several hundred pounds overpressure.

Note, that we cannot rely upon the attemperating effect of the.

copper shot bed used in the PWR loop.

The qua r t /. pebbles have a much longer time constant for heat absorption ( = 0. 5 sec), and may be at i higher initial bed-average temperature (because of the insulattad dip between the liner and thimble).

Nevertheless, the other features of the BWR design more than compensate for this dif f erence.

Following the initial transient of a large leakade incident, the BWR loop response will differ itgnificantly from that of its PWR counterpart.

The nor:nal feedsater charging rate for the BWR loop is approximately. 3000 cc/ min, about 600 times the feed and bleed rate far the PWR loop.

Hence the BWR thimble internal void space can flood a

f 4-within several minutes after initiation of leakage.

No additional

[

safety problems would ensus, but the consegnt:cos vould la undesirable (e.g., shorting out of the in pile electric rester).

Thus consideration is being given to augmentation of the alarm syne - indication of this incident (currently high humidity), by addition of an internal leak tape and a low loop pressure alarm, perhaps including an automatic heater shutoff.

Note also that the thimble PRV/ burst-disk system will be designed to accommodate thimble overflow at the maximum charging rate.

3 2.a.4.

Hydrogen Combustion

' Hydrogen has not traditionally been added to the coolant in BWR's as it has in PWR's.

Recently, however, hydrogen water chemistry has been ' implemented in an attempt to suppress oxygen, and thereby I

reduce the susceptibility to stress corrosion cracking of materials exposed to BWR coolant.

It is in fact the evaluation of the efficacy of-this fix (and confirmation of the scientific basis for it in terms of aqueous radiochemistry) which is the central focus of BWR loop design and operations.

We see no problems not already addressed in the PWR loop SER, for the following reasons:

(a) The nominal required concentration of H2 in BWR coolant is

= 2 cc/kg, an order of magnitude less than that in PWR cool-l ant.

(b) The makeup water system for the BWR loop is quitG similar to that for the PWR loop, including the hydrogen cover gas /

sparging/recombiner subsystems.

(c) The same in-containment limit on total hydrogen inventory per loop (.i 20 SCF) will be observed.

(Note that in the future the PWR and BWR and/or IASCC loops may be operated simultaneously).

___._.___._...=......;..

-o.

The most significant dif f erence between the BWR and PWR loops. is that the former has a makeup / discharge rate some 600 times the latter.

This means that a large leak in the charging or letdown systems outside of the MlTR-II core tank would vent hydrogen into the containment volume at a correspondingly higher rate.

Even so that release would amount to onir 360 cc of H2 (STP) per hour - a negligible amount which would be readily dissipated, and considerably less hazardous than, for example, a butane-filled cigarette lighter (compared on the basis of total energy release).

2. a. $., Radiation Exposure In this category dif ferences between the BWR and PWR loops will be negligible.

Specific issues are as follows:

The single la rges t dose contributor is the in-core section of the loop and thimble.

Here the BWR and PWR assemblies are virtually identical (with the only, and trivial, dif ference being the substitution of Zircaloy-2 for Zircaloy-4 for the heater bath U-tube).

Since BWR.-

loop runs are to be two weeks in duration (compared to five weeks for the PVR), end-of-run dose rates should be considerably less.

Handling doses for the remainder of the loop's components should be less because of the following considerations:

(a) The BWR loop will be all-titanium construction (to avoid catalytic decomposition of peroxide formed during radioly-sis).

(b) Crud deposition studies are not currently planned; hence the in-core section will not be disassembled for segmentation, decontamination and assay of Zircaloy U-tube deposits.

(c) copper shot is replaced by qua rt z (SiO ),

which has much 2

less radio-activation; a nd the quartz shot bed will not be drained af ter each run.

3,~

21-

+

Here again the main dif f erence comes from the higher 1stdown rate in the BWR loop.

The increased flow rate and shorter transit time means that N-16 decay out of pile will comprise a more significant source. (except when the version of the PWR loop emp1.oying an ex-tank circulating pump is used).

Even so, dose rates are not significant.

Extensive calculations have been made in the course of designing a sys-tem to measure N-16 in the steam and water ef fluents* ~ and while count rates are more than adequate for measurement purposes, the corresponding addition to local exposure dose rate is less than =10 mr/hr at 1 meter.

By compari, son the background dose rate at this location is approxi-mately 15 mr/hr.

In addition, the subject discharge lines are heavily shielded in the detector assembly to reduce undesirable background from the MITR-III this coincidentally reduces the contribution of loop N-16 to passersby.

Finally, since neither boron or lithium are added to BWR cool-ant, unlike the PWR situation we will not have (a small amount of) tritium in the ef fluent.

2. a. 5.

Section Summary In this section, a number of saf et y concerns for which there is a one-to-one parallel with comparable PWR loop features have been re-viewed.

Although dif f ering in detail, none of the issues addressed -

loss of flow, in-thimble leakage, hydrogen escape, and radiation expo-sure - appear to dif fer significantly as regards consequences.

We con-clude that the BWR loop is at least as safe as its PWR counterpart.

• Enhanced N-16 carryover in steam is a consequence of hydrogen water chemistry (thought to be in the form of ammonia, NH ).

Hence measure-3 ment of this phenomenon is a major subtask in the planned research program.

i

~

i i.

In the following section the major dif ference between the two loops will be addressed:

boiling in core - a low probability transient for the PWR loop, but a steady-state feature for the BWR loop.

2.b In-Pile Boiling 2.b.1.

Introduction The obvious and major difference between the PWR and BWR loops is the intent to operate without steady-state boiling in the former, and with steady-state boiling in the latter.

While MITR-II technical specifications preclude boiling in core, it is clear [and has been formally so-interpreted (1)] that this refers to boiling in fueled assemblies, and not to boiling inside experiments.

Moreover the total void /reflood reactivity dif ference for the PCCL and BCCL in-core coolant U-tubes is only about 0.042% Ak/k (5.4 cents), as experimentally determined (2).

This is well below technical specifica-tion limits for either "non-secured" ak/k f 0.5% or " movable" ak/k 1 0.2% experiments.

Thus boiling is not a reactor safety issue per se.

It is, ra the r, an issue related to the effect of boili ng on ease of reactor operations in two respects:

(a) reactivity instrumentation noise (b) spurious scrams due to intermittent short period indica-tions.

Table 2.1 outlines the range of reactivity oscillations which boiling in the BCCL might encompass, during steady state operating con-ditions:

background information essential to unde rs t anding these two issues.

(1) J.

Bernard memo to MITRSC Subcommittee for the Coolant Loop dated B/31/88, " Conduct of Experiments Involving In-Core Boiling."

(2) K. Kwok memo dated 1/28/88, "PCCL" Reactivity Measurement Results -

Revision I."

1

~ ;..

t TABLE 2.1

. PARAMETER ENVELOPE FOR BCCL BOILING EFFECTS Feedwater Flow ' Rate *:

3000 cc/ min = 50 cc/sec Steam Quality:

10 wt:

Steam / Water Specific Volurae Ratio:

20/1 (SAT @ 1000 psia)

Volume of In-Pile Zircaloy U-Tube:

50 cc Thus one has:

2 ce steam /1-cc H O exiting core 2

I second core transit time Limiting bubble frequencies:

(a) 100 ce steam, then 50 cc H O every second for step reactivity of 2

12.7 ( (5.4 / peak to valley = 0.042% &/k).

h- ~ -

(b)

= 100 t 10 one ce steam bubles exiting the core each second with a core-average steam content of 20 2 2 cc, to give noise of 2 0.2 f at = 100 Hertz ( 20.0016% 3/k).

+

WWW

= 0. 2 p t

• NOTE: All. values are approximate full flow, full power quantities.

I

l l

2.b.2.

Possibility of causing reactor scram The nature of the false scram concern is explained in Fig. 2.1.

As shown there, step reactivity insertions of any magnitude involve ex-tremely short transient periods (in a purely mathematical sense) before reaching their asymptotic value.

Thus an instantaneously responding period scram circuit would conceivably trip the reactor during virtually any and all power maneuvers.

The actual circuit, however, has its own built-in time lag to avoid just such occurrences.

The question, then, is whether loop boiling transients can somehow excite detectable periods of just the right type to dupe the period scram system.

While this would not compromise reactor safety, it is clear that spurious chal-lenges to safety shutdown systems are undesirable, and if too f requent, would ef fectively prevent meaningful operation of the BWR loop for the intended purpose.

2.b.3.

Reactor noise With respect to the issue of steady state boiling noise, it is more difficult to establish a specific criterion for judging it to be innocuous.

Clearly both amplitude and frequency are of interest.

Table 2.1 defines one such scenario:

2 0. 2 c 0 100 Hz.

Higher frequen-cles cannot be ruled out, but they would in deneral be associated with smaller amplitudes.

Amplitude will surely be le ss than in a BWR reac-tor, where neutron flux variation related to boiling noise is = 5%.

While bo t h extremes discussed above appear tole ra ble, experi-mental verification is considered desirable.

Hence a boiling experiment will be carried out using the PWR loop prior to BWR loop operations, and the results reported to the MITRSC.

I i

l

.,4

=*

-i

,^

d

__ _ _ All.0WABLE REACTIVITY FOR FIXED EXPERIMENT REACTIVITY TOTAL REACTIVITY 1

INSERTION I-

% RAMP INSERTION (BECOMES " STEP" IN LIMIT OF VERY SHORT DURATION) o z_ TIME h

MITR CORE D, n/cm2 ASYMPTOTIC EXPONENTIAL PERIOD sec DOMINATES @(t)

PROMPT EFFECT AND TRANSIENT PERIODS

- INITIAL VALUE (ASSLHED STEADY STATE)

+ TIME A

INSTRUMENT RESPONSE e

ASYMPTOTIC

PERIOD,

'~

PERIOD

. SECONDS

'f,,,

MITR-tI

--P

'- - - PERIOD SCRAM:

REACTOR TRI?S ON ALL ?ERIODS SHORTER

% MATHEu.ATICALLY \\

THAN TH!S EXACT o/

. !!ME

-. J The issue is whether the instrumentation circuit responds fast enough to pick up the inevitable verv short t rans ien t Comnonents.

FIGURE 2.1: SCHEMATIC SEQUENCES SHOWING PERIOD SCRAM DILEMMA

t A final note is of interest.

From a purely neutronic viewpoint,

. operation of the BCCL will be quite analogous to a classical. rod oscil-lator experiment (1,2) in which a neutron absorber is used to measure a o

reactor's transfer function by subjecting it to a variable frequency reactivity perturbation (10-3 -101 cycles /sec).

This is of ten done one f requency at a time, but even mo re relevant he re are experiments in which random sequences are used, capable of exciting many frequencies simultaneously (3).

(1)

A..E. Profio, " Experimental Reactor Physics," John Wiley & Sons, NY (1976).

(2) J. M. Harrer, " Nuclear Reactor Control Engineering," D. Van Nostrand Co., NY (1963).

(3) T.

W. Kerlin, " Methods for Frequency Response Measurements in Power Reactors," in Dynamics of Nuclear Systems, D.

L.

Hetrick, ed.,

University of Arizona Press, Tucson (1972).

3.

SUMMARY

AND CONCLUSIONS

[

The project staff has completed its safety evaluation of the Boiling Water Reactor Coolant Chemistry Loop (BCCL), as docume nt ed in this report.

As summarized in Table 3.1, in mos t saf ety-related aspects, the

[

BWR loop dif fers only moderately in degree compared to its PWR counter-part.

Moreover, differences of ten favor the BWR e.g., its operating.

pressure is approximately 1200 psia. lower.

1 The major noteworthy dif ference between the two loops is the pres-ence of steady-state boiling in the in-core Zircaloy U-tube of - the B[/)C(, L While measurements of the void /reflood reactivity change show that. it -is well within allevable limits (i.e.

only 0.042% ak/k), the uncertain frequency of the boiling noise and precisely how this will interact wi t h -

MITR-II instrumentation has led us to recommend a proof-by-test approach to this aspect.

Accordingly a low power boiling experiment is currently being. planned for execution using the PWR loop.

We anticipate that this will confi rm the benign nature of BCCL ope rations.

A report on this test will be prepared and circulated prior to requesting final-permission for in-pile operations of the BWR loop, currently scheduled for the week beginning April 3,1989.

In this re port we have not gone into detail on the saf ety inst ru-

' mentation and control systems because they are virtually identical to those. on the PWR loop.

Fundamentally, all high-severity scenarios, if' h

not checked by intervening actions or events, lead to overtemperature of the in-pile lead heater bath - which then triggers (through redundant detection and control circuits) electric heater shutoff.

The loop then

. need only rej ect approximately 6 kW of gamma heat - which can be accom-1

, l l

l l

TABLE 3.1

SUMMARY

OF SAFETY REALTED ASPECTS OF BVR LOOP OPERATIONS Aspect Comments Hydrogen release External inventory:

same as PWR.

Internal inventory (dissolved in coolant) less than PWR, Discharged He is recycled, not vented.

Consensus as safe or safer than PWR.

Leak inside thimble Loop pressure is much lower than PWR.

Quartz shot has slower heat transfer, hence quenching must rely upon condensation by cold thimble wall.

Consensus: Thimbls rupture disk should protect against overpressure damage to thimble.

Loss of coolant or Loop will boil dry, overtemperature will shut coolant flow off electric heater, conduction and radiation will safely dissipate gansna heating.

Since there is no internal heat sink analogous to. steam generator tube section in PWR, wet heat rejection will not be effective.

Consensus:

Ultimate-passive safety mode is exactly same as PWR proven effective by demonstration.

Note:

This mode may be used as the normal dry layup mode between runs if the loop is left in-core.

Steady-state and/or A total void /reflood Ap 5.4 / and a steady

=

transient boiling state ap variation of = 2 0. 2 p. mus t be accom in-core modated (values to be established by in-pile-tests with PWR' loop).

Max step Ap will not cause period scram.

Boiling noise shown tolerable in experiments using PWR loop.

Consensus:

Operation will not lead to chal-1enges of MITR-II safety system.

l l

plished by passive conduction and radiation without exceeding roughly 1700'r in the heater bath and its contents.

Baesa has drafted a preliminary list of incident re sponse ac-tions, reproduced here as Table 3.2.

More comprehensive operating and emergency procedures, paralleling - those draf ted for the PWR loo p, will be prepared prior to in pile operations.

As a result of our re vie w, it is concluded that there are no - un-reviewed safety questions involved in operating the BCCL in the-MITR-II.

TABLE 3.2

SUMMARY

'OF MAIN BCCL SAFETY EVENTS EXPERIMENTER AUTOMATIC EVENT SYMPTOMS ACTIONS ACTIONS Charging pump failure /

--Reduction or no flow readings

--Shut off heaters

--H5gh T heater Loss of coolant in any-of 3 flowmeters and charging pump shut off

--Low loop pressure.

--Eventual high-lead bath tem-perature Backpressure regulator

--Low loop pressure

--Shut off heaters.

failure Pla c e loop in standby mode.

• Loop leak (small)

--Humidity increase

--Normal shutdown procedure Thimble leak (severe)

--Helium pressure decrease

--Shut off heaters

--Humidity Increase and open heater y

power breaker

--Notify operators and consider re-questing reactor shutdown NRHX cold side failure

--Loep temperature increase

--Shut off heaters.

--Place loop in stand-by mode

• Float level meter

--No output signal

--Normal shutdown failure

--Turbine flowmeter over-proce; dure speed or orifice flow-meter differential pressure decrease

• In st andby mode cooling water is circulated through the loop by the main charging pump at' low temperature ~

to remove nuclear heat (elect ric heat ers of f ).

e L

+

L t...-

m

.a :.

-L

c,. -

l-APPENDIX A Condensation on Thimble Wall Steady state condensation relations are of the form (1):

1/4 1

Gr Pr g

Nu

=-1.13 where Ja

. = Jak'ob numbe r = AHi/ AHv, ratio of enthalpy rise of liquid water (between wall-and saturation temperatures) to

' enthalpy of evaporation.

Pr

= Prandtl number for liquid water at Tyttg.

2 3 2 G et,

= Grashof number go t /N, also for ' liquid water at

=

- Tyttg (this form valid because SPAT E Ap = p)

Tygtg = 1/2 (TWALL + TSAT)

For TWALL = 100'F and TSAT at 1000 psia, the above expressions Gr = 1.66 x 10 15 Nu = 7800-Pr = 1.I h

= 520 BTU /hr ft2 7

Ja = 0.73 and for a 3.5-inch ID, 6 ft long thimble section the heat removal accounts for condensation at a rate of 0.54 lb/sec.

This estimate should be taken as a rough approximation _ only, because of several approximations:

(1)

G.

J. Geankoplis, Transport Processes:

Momentum, Heat, and Mass.

Allyn and Bacon, Inc., Boston (1983).

t 32 (a) It applies in _ the laminar regime; turbulence will lead to sub-i stantial increases.-

(b) It applies to-the steady state; condensation will be much more rapid in the initial transient period.

(c) On the other hand, the retarding ef f ect of non-condensible gas (here a 2 atm. of He) is neglected.

Nevertheless, the presence of a fairly rapidly acting and ef fective heat sink is evident.

m----

1 Safety Review #-0-89-20 BWR Coolant Chemistrv Imp (BCCL) 1.

Dewrintion of Chance An in-pile loop designed to simulate the primary coolant system of a boiling water reactor (BWR) is to be installed in the MITR-Il core. The facility is described in detail in

)

[1], " Safety Evaluation Report (SER) for the BWR Coolant Chemistry Loop (BCCL),"

Report No. MITNRL 031, March 9,1989, prepared for review by the MIT Reactor Safeguards Committee and attached hereto.

2.

BCCL Safety Evaluation Reoort The BCCL SER provides both a description of the BCCL and an evaluation of the safety of its design and operation. Aspects of the BCCL which are virtually identical to those of a previously installed loop, the PCCL, are not discussed here in detail. This includes the safety instrumentation and control systems. Both the BCCL and PCCL can be placed in a safe condition by deenergizing the elecuic heaters. (No.te: The PCCL is a

' Pressurized Coolant Corrosion Loop' used to investigate the formation, transmrt, and deposition of radioactive crud in Pressurized Water Reactors. The PCCL :Tas been operationalin the MITR-II since August 1989. It has functioned without posing a problem to the reactor.)

The BCCL SER concludes that the 'most noteworthy difference between the two loops (BCCL and PCCL) is the presence of steady-state boiling in the in-core Zircaloy U-tube of the BCCL.' The Reactor Staff agrees with this finding. In addition, the Reactor Staff has concludod that the potential for the BCCL's shadowing the reactor fuel during use of the emergency core cooling system (ECCS) is another major difference. Accordingly, this safety review examines these two issues (boiling and ECCS) in some detail. Other

~ issues such as the presence of a high pressure system in-core and the use of H2 gas are given less consideration because, for these issues, the BCCL parameters fall well within the envelo ye of conditions established by the PCCL. Refer to the PCCL SER, the PCCL SER Supp ement, and SR #0-86-9 for details [2-3].

3.

Comoliance of Boiline Coolant Chemistry Looo (BCCL) with MITR-Il Tech'nical Soecifications

~

~

A description of the BCCL is given in [1] and is therefore not te 3eated here except to note that the in core portions of the BCCL and PCCL are identical. ALso, the BCCL will be inserted in the A-Ring, as was the case with the PCCL.

3.1 Technical Specification Desien Criteria MITR Technical Specification #5.2.2 specifies that in-core experiments must comply with four criteria. These are:

a.

They shall be positively secured in the core to prevent movement during reactor operauon.

b.

Materials of construction shall be radiation resistant and compatible with those used in the reactor core and primary coolant system.

c.

Sufficient cooling shall be provided to ensure structural integrity of the assembly and to preclude any boiling of the primary coolant.

SR#-0 89-20 DEC 201989

2 d.

The size of the irradiation thimble shall be less than sixteen square inches in cross section.

The proposed BCCL design meets these criteria as noted below:

3.1.1 I ack of Movement During Ooeration The BCCL, in the same manner as the PCCL and all previous in core sample assemblies, is secured via the upper grid plate.

3.1.2 Materials of Construction The exterior of the BCCL is aluminum and hence is compatible with the MITR-II's fuel, core vessel, and water chemistry.

3.1.3 Heat Removal Energy is generated by means of a 20 kW electric heater and a lead bath (gamma-ray attenuation). In this respect, the BCCL is identical to the PCCL. Energy removal is different in that the PCCL uses a copper shot bed in the riser section of the loop. The BCCL rejects heat outside the reactor tank. The BCCL is protected by redundant sensors that will shut off the electric heater if the temperature bemmes excessive. Should that cutoff fall, an alarm will sound in the control room and the operator can deenergize the system. This arrangement satisfies Technical Specification #5.2.2(c).

3.1.4 Cross-Sectional Area The BCCL occupies only one fuel element position. Its cross-sectional area is less than sixteen square inches.

3.2 Reactivity of the BCCL Exoeriment MITR-II Technical Specification #6.1.1 limits the reactivity worth of experiments to the following values:

l Single Experiment Worth Total Wortir Movable 0.2% AK/K 0.5% AK/K Non Secured 0.5% AK/K 1.0% AK/K 1-Total of the above 1.5% AK/K Secured 1.8% AK/K The three types of experiments are defined in Section 1 of the Technical Specifications as follows:

Secured Experiment A secured experiment is an experiment or experimental facility held firmly in place by a mechanical device or by gravity, such that the restraining forces are SR# 0-89-20 DEC 201989

3 substantially greater than those to which the experiment might be subjected by hydraulic, pneumatic, buoyant, or other forces which are normal to the operating environment of the experiment or by forces which can arise as a result of credible malfunctions.

Movable Experiment A movable experiment is one where it is intended that the entire experiment may be moved in or near the core or into or out of the reactor while the reactor is operating.

Non Secumd Experiments Experiments where it is intended that the experiment should not move while the reactor is operating, but is held in place with less restraint than secured experiment.

No reactivity measurements of the BCCL have as yet been made. However, given that the BCCL's in-core section is identical to that of the PCCL, the figures obtained for the PCCL are germane, The principal difference between the two is that soluble boron is not used in the BCCL coolant. This difference is minimal because of the small amount of boron present. (ND.tc: For documentation of PCCL reactivity measurements, refer to MITR core configuration #87, dated 01/19/88. It should be noted that the more important reactivity numbers were measured on several occasions to be certain of accuracy.)

3.2.1 BCCL Water and Contained Chemicals For the PCCL, these were classed as 'non secured.' However, for the BCCL a change of phase is intended. Hence, the water and contained chemicals are classed as

' movable.' The limit is therefore 0.2% AK/K or 254 mbeta. Measurements for the PCCL L

put the reactivity worth of the coolant at +54 mbeta.

1 3.2.2 Floodinn/Refloodine Flooding /reflooding scenarios were examined for the accessible volume of the BCCL. This includes the void volume in the thimble and the coolant channel annulus I

between the thimble and the dummy element. For the PCCL, the effect of flooding these L

spaces was classed as 'non-secured' because it was not intended that these spaces be flooded or voided during normal operation. The same classification applies to the BCCL.

L The measured reactivity (PCCL data) was +181 mbeta. Hence, the flooding /reflooding scenarios meet the 'non secured' limit which corresponds to 0.5% AK/K or 636 mbeta.

L 3.2.3 BCCL Comoonents The BCCL components (lead, loop, heater, and fixtures) are classified as a secured l

experiment, because they are mechanically held in position by the loop tubing and other structural components. Their complete ejection from the thimble followed by flooding must not exceed 1.8% AK/K. The measured reactivity worth for the PCCL was -407 mbeta. The absolute value of this figure is well below the limit of 1.8% AK/K or 2290 mbeta.

SR#-0-89-20 DEC 201989 1

I

4 3.2.4 Erneriment Interactions As noted at the outset of section 3.2 of this report, Technical Specification #6.1.1 imposes limits on the total amounts of reactivity associated with both movable and non-i secured experiments. Initially, the BCCL will be the only in-core experiment. Ilowever, at some future time, two loo? experiments might be run simultaneously. It is noted here that the MITR II core coulc. contain three such experiments and not exceed either the movable or the non-secured limit.

3.3 Pn ssure Effects Technical S xcification #6.1 requires that experiments be designed to withstand twice the anticipate < pressure. The loop portion of the BCCL will operate at approximately 1000 psi. Surrounding the actualloop is an elliptical thimble made of 6061 aluminum. It would be pressurizec, in the event of failure of the BCCL internals. The thimble is protected by a pressure relief valve set at 15 psi and a burst disk set at 65 psi. The maximum expected pressure is therefore 65 psi. The thimble will therefore be hydrostatically tested to 150 psi. (&lt: The same approach was used for the PCCL following discussions with the U.S. Nuclear Regulatory Commission.)

3.4 In-Core Boiling The BCCL experiment will involve boiling of the loop coolant which will, of course, be physically within the core. The MITR-II's technical specifications were carefully examined to determine if there was any prohibition of such an activity. It was concluded that boiling within an experiment (as opposed to boiling of the primary coolant -

itself) is permitted. The relevant technical specifications were:

3.4.1 Icchnical Soecification #2.2 - Limiting Safety System Settings The objective of this specification is: "To assure that automatic protective action will prevent incipient boiling in the reactor core and will prevent conditions from exceeding a safety limit." The specification itself establishes limiting conditions for flow, power, core tank Icvel, and temperature to "... prevent incipient boiling... which is initiated prior to the initiation of flow instability."

As to the applicability of this specification, the specification section itself contains no prohibition agamst boiling in an experimental facility. Its purpose is to restrict certain reactor parameters so that there will be no problem with flow in the fuel element channels.

Accordmgly, it is concluded that this specification does not pertain to in core experiments that are isolated from the core.

3.4.2 Technical Soecification #5.2 - Reactor Core Paragraoh 2a of this specification states that the " Design of in core sample assemblies shal conform to the following criteria," one of which is that " sufficient coohng shall be provided to ensure structuralintegrity of the assembly and to preclude any boiling of the pnmary coolant."

As to the applicability of this specification, it clearly requires that the experiment be designed so that there wH1 be no boiling on the outer surface of the experiment thimble and tube. The type of boiling, nucleate or bulk, is not mentioned. Presumably both are prohibited. There is no prohibition of boiling within the thirnble.

SR#-0-89-20 DEC 201989

i

,6 i

5 -

i e

3.4.3 Technical Snecificarian n1 -General Exneriment criteria

~

Paragraph 2b of the specification states that, "The outside surface temperature of a '

submerged experiment or capsule shall not cause nucleate boiling of the reactor coolant during opemtion of the reactor."

As to the applicability of this specification, it clearly requires that the experimer.: be designed so that there will be no nucleate boiling of the coolant. There is no prohibition of boiling within the thimble.

t 3.5 Ernlosion Haran!

Technical Specification #6.1.3(b) limits explosive materials to the equivalent of 25 mg of TNT. The BCCL will have a hydrogen gas concentration of 2 cc/kg. Assuming this concentration is in all of the BCCL coolant that is within the reactor's biological shield, the total amount of hydrogen present will be about 2.5 cc at standard temperature and pressure. This corresponds to less than a milligram equivalent of TNT.

4.

Saferv Annivsis In the previous section of this safety review, the compliance of the BCCL experiment with the MITR-II's technical specifications was shown. In this section.

specific safety issues are examined with the objective of demonstrating that an 'unreviewed safety question' or URSQ, does not exist. For the record, an URSQ is defined in 10 CFR l

50.59(2) to exist if:

A proposed change, test, or experiment shall be deemed to involve an unreviewed safety question (i) if the probability of occurrence or the consequences of an accident or malfunction of equipment important to safety previously evaluated in the safety analysis report may be increased; or L

(ii) t'f a possibility for an accident or malfunction of a different type than any evaluated previously in the safety analysis report may be created; or (iii) if the margin of safety as defined in the basis for any technical specification is reduced.

Certain of the analyses given here closely follow those for the PCCL [2].

4.1 Temnerature Effects The in-pile loop assembly will be heated both by a 0-20 kW heater and by a combination of gamma and fast neutron radiation. The normal combined heat load will be less than 20 kW. The radiation heating is estimated to be 9.6 kW at 5 MW [2]. Hence, 29.6 kW would be the maximum heat load potentially available under malfunction conditions. This is not much more than the hottest running fuel plate in the MIT Reactor.

Hence, the thimble is easily cooled by the flow of primary coolant through the 0.050 inch thick channel between the thimble and the dummy fuel element that surrounds it.

The potential for a Zircaloy-water reaction was analyzed in the PCCL SER [2]

where it was shown that cooling by conduction and radiation should prevent temperatures l

in the thimble from exceeding 1845 F for the maximum radiation heating i.e. 9.6 kW.

This was based on a very conservative extrapolation of temperatures measured in a test mock-up of the PCCL assembly operating out of core at heater powers in the range of 2470-4510 watts. The 1845 F temperature is significantly below the 2200 F post-LOCA limit on Zircaloy temperature imposed for PWR units by NRC (4].

SR#-0-89-20 DEC 201989 l

g-4

.h Assurance that total heating will not exceed the 9.6 kW which might result from gamma and fast neutron heating with the reactor at full power is achieved by redundant

- heater shut-offs that are activated by high lead bath temperatures. The sensors and relays that interrupt power to the heaters are completely independent, thus avoiding compromise by a single failure.

Elevated lead bath temperatures are not a threat to the aluminum thimble, because there is no contact between the thimble and the titanium can holding the lead except at occasional small points of contact with high spots on the weld bead stiffener on the outer surface of the titanium can and at the su? port ring which is at the top of the titanium can extension about 12 inches above the lead ath.

In view of the above active and passive safety features, it is not credible that temperature effects within the thimble can affect the fuel, core structure, or other components important to safety and, hence, there is no unreviewed safety question in this regard.

4.2 Hydrogen 12ak and Combucion This issue was thoroughly reviewed as part of the PCCL SER[3]. The concentration of hydrogen gas in the PCCL coolant is approximately 20 cc/kg. That in the BCCL's coolant will be 2 cc/kg, a factor of ten less. Hence, the BCCL's use of hydrogen is well within the envelope of previously established conditions.

The hydrogen gas used for the BCCL will be stored and handled using the same procedure as is now fo lowed for the PCCL. He only difference is that the total inventory of hydrogen gas present in the containment will be doubled. Specifically, a storage limit of 20 SCF per loop not to exceed a total of 80 SCF will be observed. This does not present a hazard because discharge of 80 cubic feet of hydrogen into the containment building volume of 200,000 ft3 will result in a concentration far below the lower explosive limit of 4.1% Moreover, there is a fan mounted near the hydrogen station that ensures rapid dispersion of the gas.' The maximum amount stored in any one container will not exceed 20 SCF.

4.3 Loss of Loco Pumoing Power or loss of Flow -

As noted earlier, the heat sources for the BCCL are the same as those for the existing PCCL. Namely, a 0-20 kW electric heater and a lead bath for gamma ray attenuation. He heat sinks are quite different. The PCCL rejects heat to a copper shot bed located within the reactor vessel. The BCCL uses a non regenerative heat exchanger located outside of the reactor vessel.

He BCCL SER makes several arguments to show that under most circumstances, a loss of flow would be more easily rectified for the BCCL than for the PCCL, The principal reasons are the external location of the circulating pump and the likelihood of operating in a continuously, attended mode. However, these arguments are not permane because for some of the operatmg cycle, the BCCL may well be unattended. He important consideration is that, in the absence of any human interaction following a loss of BCCL flow, the in-core lead heater bath temperature will increase until the redundant over-temperature trips cut off electric heater power, following which gamma heat will be safely rejected by passive means, radiation and conduction to the thimble wall, and thus to the MITR-II coolant.

I SR#-0-89-20 DEC 201989

7.

7 4.4-In-Thimble 12nknge Section 2.A.3 of the BCCL SER addressed the issue of whether a rapid,large leak could result in an overpressurization of the thimble. The MITR Staff agrees with the analysis given and with the conclusion that the relief valve and burst disks installed on the thimble are adequate.

4.5 1.eg! Bath Can Leak Section 3.6 of SR #0 86-9 (PCCL Safety Review [3]) addressed the issue of large and small leaks of lead from the titanium can and concluded that there was no credible-mechanism by which the loop could adversely affect MITR Safety. That conclusion applies here because the can construction is identical.

4.6 Electrical System Mnifunction The electric power supply for the BCCL is essentially identical to that for the PCCL Section 3.8 of SR #0-86-9 (PCCL Safety Review [3]) examined the possibility of an electrical short and concluded that the loop was properly designed to prevent such an occurrence. That same analysis applies to the BCCL and is therefore not repeated here.

4.7 Effect of Boiling on Reactor Ooeration The concern here was that the presence of boiling in the BCCL might cause reactivity fluctuations which, while within the magnitude allowed by the technical specifications, would cause ope, rational problems such as excess movement of the regulating rod. Following discussions with the experimenter, an experiment was proposed to resolve this issue. Known as the 'In-Pile Boihng Experiment,' it involved performing boiling tests in the PCCL while at low power. Safety aspects of the proposed experiment were documented in SR #0-88 6 and approval for the experiment was given by the Special

- Subcommittee of the MITRSC on 09/15/88 and by the full MITRSC on 12/20/88.

The 'In-Pile Boiling Experiment' was conducted on 6/19/89 and it was found that the presence of boiling in the PCCL in core section did not have any discernible effect on MITR II operation. Additionalinformation is given in Appendix A to this safety review which provides the ex 3erimenter's documentation of the test results. There is no credible mechanism by which tle BCCL can adversely affect MITR safety on this issue.

4.8 Emergency Core Coolina System (ECCS)

The concern here was that the BCCL might shadow some of the fuel elements from the water sprayed onto the top of the core by the ECCS system in the event that the reactor core shoulc, not be covered by water. This shadowing could occur either as the result of the BCCL's riser section or as the result of the gamma ray pod.

Tests were made using a mock-up of the core top, primary coolant flow guide, and BCCL.

It was found that sprayed ECCS water splashed randomly from the experimental facilities and from the interior surfaces of the flow guide so that any shadowing effect was minimized, and each fuel sosition received at least 40% of the average flow per element.

Five tests were performec and the results averaged.. The final set of figures is given in Table I. The tests were conducted according to PM 6.1.1 which is used to verify compliance of the installed ECCS system with the MITR Technical Specification criteria on an annual basis. Technical Specification #3.6 requires that a total of 10 gpm be delivered via the ECCS system. Section 6.1 of the MITR II's SAR further requires that each fuel SR#-0-89-20 DEC 201989 v

e,

.g.

'~ O :

I.abit.1 BCCL Test of Emergency Core Cooling Sprav Distribution -

Position Grams H Q

% of Average 2

' A1 (not fueled) 146.5 35.5 2

591.3 143.4 3 (not fueled) 939.9 227.8 B1 356.1 86.3 2

184.9 44.8 3

268.4 65.1 4

-359.8 87.2 5

349.5 84.7 6

558.6 135.4 7

533.3 129.2 8

419.7 101.7 9

506.2 122.7 C1 450.0 109.1 2~

194.5

47.1 3

174.7 42.3 4

239.8 58.1 5

319.5 77.4 6

261.6.-

63.4 7

284.7 69.0-8 503.7 122.0 9

374.4 90.7 10 463.8 112.4 11-513.6 124.5 12 425.5 103.1 13 644.2 156.1 14' 568.0 137.7 15 508.8 123.3 4

SR#-0 89-20 DEC 201989

_______________i._.______._______________________________ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

w

. C' element nceive at least 20% of the average flow per element. As Table I shows, there was significant variation in the flow distribution. However, all fueled element positions received at least 42% of the average, which is well above the 20% required. Accordingly, the installation of the BCCL will not interfere with the proper functiomng of the emergency core cooling system. (Holc: The mock up tests were performed for the BCCL alone.

Should the PCCL and BCCL be run simultaneously, a new set of tests would be required.)

5.

Conclusion It is concluded that failures or accidents originating with the BCCL loop cannot interact with the reactor fuel, core structure, or other components important to safety, except through reactivity effects. In this case loop failures or accidents will not cause reactivity changes exceeding those authorized by the Technical Specifications. For

. equipment important to safety, (i) the probability of an accident or malfunctico is not increased (ii) the possibility for an accident or malfunction of a different type than that previously evaluated in the SAR is not created, and (iii) no margin of safety in any technical specification is reduced. Consequently, the BCCL experiment does not involve an umeviewed safety question.

References 1.

Safety Evaluation Report for the BWR Coolant Chemistry Loop (BCCL), Report No. MITNRL-031, March 9,1989.

2.

Safety Evaluation Report for the PWR Coolant Chemistry Loop (PCCL), Report No. MITNRL-020, February 13,1987 plus supplement dated March 22,1988.

3.

SR #0-86 9, "PWR Coolant Chemistry Loop," April 21,1988.

4.

.10 CFR 50.46(b)(1)

SR#-0 89-20 DEC 201989

, o -.

~t, q

A1

_o Apnendix A to SR #0-89-20 Results ofIn Pile Boiling Experiment Filed as part of Q/A program #M 89 2 are copies of the notebook entries which describe the thermal /hydntulle parameters established during a test on 6/19/89 to determine if reactivity noise generated by controlled boiling in an in core loop poses a problem for reactor operation. The experiment was carried out using the PCCL loop with the circulating pump bypassed. The charging pump was used to provide forced once-through flow at a low flowrate and pressure was adjusted using the outlet line back pressure regulator. Note that the PCCL and BCCL in core sections are identical except that the PCCL tubing is Zircaloy 4 while the BCCL tubing is Zircaloy 2.

Two reactor criticalities were established at low power. During the first of these the coolant was solid and the loop outlet temperature was maintained at ap3roximately 515 F.

Once a baseline reactivity was determined under these conditions, the reactor was shut down and boiling was established in the loop. This was accomplished by lowering pressure to -1000 psia (th,e planned operating point for the BCCL). Constant loop outlet temperature with increasing power indicated that boiling was occurring in the in core tubing. The conditions achieved at the two operating points were as follows:

Flowrate Pressure Electric Power Loop Inlet T Loop Outlet T 1

(lb/h)

(psia)

(kW)

( F)

(F)

Solid 6.3 1515 2.8 107 500-530 Boiling -

6.3 1005 1035 3.4 120 543 Note that TSAT at 1020 psia is 547 F. This agrees well with the measured loop outlet temperature. Using the solid run to calibrate the energy losses from the lead bath through the thimble walls, and energy balance gives a loop outlet quality of 50 60%.

No effect of the boiling was observed by the operators at reactor power levels up to 10 kW. The planned maximum quality in the BCCL is 10-15%. The higher flowrates to be used in the actual loop should result in less spatial and time vadation of bubble sizes.

.-