ML20151R727

SER for PWR Coolant Chemistry Loop (Pccl) MITNRL-020

ML20151R727
Person / Time
Site:	MIT Nuclear Research Reactor
Issue date:	02/13/1987
From:	Ames M, Driscoll M, Harling O MASSACHUSETTS INSTITUTE OF TECHNOLOGY, CAMBRIDGE
To:
Shared Package
ML20151R699	List: ML20151R696 ML20151R727 ML20151R734
References
NUDOCS 8804280073
Download: ML20151R727 (74)

Text

-

8 .. q . A m M

P SAFETY EVALUATION REPORT FOR THE PWR COOLANT CHEMISTRY. LOOP (PCCL)

MITNRL-020 to be '

installed and operated in the MITR

?

5 February 13, 1987 Prepared by:

M. R. Ames M. J. Driscoll

0. K. Harling G. E. Kohse K. S. Kwok D. D. Lanning J.H. Wicks For Review by the MITR Safeguards Coanittee 8804280073 880421 PDR ADOCK 05000020 P DCD

s . ,

TABLE OF CONTENTS P, age

1. I NT R O DUC T I O N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I a ) Fo r e wo rd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I b ) Ge n e r a l De s c r i p t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I

1) In pile loop design....................................... 1 2 ) Ex pe r im e n t a l p ro t o c o l . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2. DE TA I LS OF T HE PCC L LOO P DES I CN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 a) Design Spec ifications , Tolerances and Safe ty Margins . . . . . . . . . 9 b) Loop Materials - Compatibility with MITR-II Core and Coolant. 9 c ) S t r uc tur al Sup po r t s and Lo ad ing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 d ) Po we r Pe a k i n g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 e ) In s t r umen t a t i o n and Co n t r o l . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 f) Quality Assurance Requirements............................... 15

3. OPE RAT IO NAL AND E XP E R IM E NTAL PR0C E DUREE . . . . . . . . . . . . . . . . . . . . . . . . 17 a ) Lo o p O p e r a t i o n s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 b ) Po s t -Ope r a t i o n Ra n d 1 i n g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 1 ) Loo p r emov a 1. . . . . . . . . . . . . . . . . . . . . . . . . . ................... 21 2 ) Lo o p d i s a s s e m b l y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

4. MAXIMUM EFFECTS OF REACTIVITY, PRESSURE AND TEMPERATURE. . . . . . . . . 25 a ) Re a c t i v i t y E f f e c t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 1 ) Re ac t iv i t y e f f ec t o f B-10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 5

2) Water flooding / voiding in c i d e n t s . . . . . . . . . . . . . . . . . . . . . . . . . . 27 b ) Pr e s s u re E f f e c t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 c ) Tem pe r a t u r e E f f e c t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

5. RADIAT ION LE VE LS AND ALA RA CO NS IDE RAT I0h S . . . . . . . . . . . . . . . . . . . . . . . 32

6. PC C L SAF ETY E VA LUAT IO N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 a) Sa fe ty and Ope ra t ional Envelo pe s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 b ) Mal f unc tio n Se quenc e a and Conse quenc e s . . . . . . . . . . . . . . . . . . . . . . . 36 l 1 ) Lo o p 1 e a k . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 2 ) Lo s s o f pum p i n g po we r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3 ) Hyd rog e n l e a kage a nd c omb u s ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4 ) Th i m b l e 1 e a k . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 5 ) Le ad b a t h c a n 1 e a k . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

7. W A S TE RA NDL I N G AN D D I S P0 S A L . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

8. F UT URE W 0 R K . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 9 7

f t

.- , t I

9. R

SUMMARY

AND CONCLUSIONS......................................... 51 a) Summary 51

1) Reactivity changes 51

-2). Thermal-hydraulic effects 51

3) Chemical ef fects 53

4) Radiolytic . decomposition' 53

5) Experimental scrams 54

6) Prototype testing and proof testing 54

7) Radioactive releases, waste handling and disposal, radiation levels, ALARA considerations, and related procedures 54 b) Conclusions 54 APPENDICES Appendix 1: Safety Related Experiments and Calculations a) Radiative dissipation of gamma and neutron heating of loop components b) Reactivity worth of B-10 c) Thimble stresses Appendix 2: Extracts from the MITR-II Technical Specificetions ,

1 9

J 4

r I i

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

-r--,,

1. INTRODUCTION a) Foreword This is the safety evaluation report for en in pile loop facility designed to : simulate the primary coolant system of a pressurized water reactor (PWR). The loop will be used to carry out research into the effects of coolant chemistry on the traasport and deposition of corrosion product radionuclides as part of a program to develop means for the l reduction of maintenance doses in the nuclear utility industry. Loop constructic,n and operation are funded by the Electric Power Research Institute and the Empire State Electric Energy Research Corporation; loop conceptual design has been funded in-house and by a research project sup-ported by regional utilities (Boston Edison, PSE&G, and Duke Power) under the Electric Utility Program of the MIT Energy Laboratery. This program also calls for the design and operation of a loop simulating BWR condi-tions; however, the BWR loop will be covered by a separate submission.

The objective of this report is to present a summary description of the design and operating procedure of the PWR Coolant Chemistry Loop (PCCL) in suf ficient detail, and with supporting analyses, to demonstrate that it can be operatet safely within the envelope of applicable MIT Reactor Technical Specifications. To this end, a number of topics will be emphasized, including the ef fect of the PCCL on core reactivity, energy dissipation following loss of normal energy removal capability, the consequences of leakage, and hydrogen handling.

b) General Description

1) In-pile loop design The guiding design philosophy in the development of the PCCL concept has been to simulate all bnportant PWR primary coolant system

(

• 1 a

parameters (e.g., pressure, temperature, velocity, materials, surf ace area ratios, etc.) as closely as possible, as suaaarized in Tatle 1.1, but at a greatly reduced scales on the order of 105 smaller. Despite being quite small on a macroscopic basis, good simulation of a PWR coolant flow unit cell (steam generator tube--inter-fuel pin cha.mel) is achieved on approximately a one-third scale. The resulting design is depicted schematically in Fig. 1.1. It is a relatively simple layout consisting mainly of 0.25 inch ID tubing, containing less than 0.1 liter of coolant circulated at approximately 1-2 gpm with a canned rotor pump. An electric heater supplies 10-20 kW of energy to the Zircaloy in pile segment. The system is externally pressurized using a positive displacement diaphragm-type pump plus backpressure valve--a practice proven in many years of out-of pile autoclave experiments operated at ,

MIT, Westinghouse, and General Electric.

Features which are particularly significant from a safety viewpoint are as follows:

The entire loop is encapsulated by an aluminum thimble of 2 inches diameter in-core (where it is housed in a dummy fuel element), increasing to 4 inches above the core, topped by a pod

, containing the pump. The thimble atmosphere is helium gas at

< 50 psi, and the thimble is protected from overpressurization by a pair of 100 psi relief valves. Relief valves also protect the loop itself against > 2,500 psia (see Section 2.a).

Energy is added to the in-core section of the loop, a Zircaloy U-tube, by electric resistance heaters immersed in a small lead bath surrounding the U-tube. Calculations and confirmatory experiments have shown that when electrtic heat is shut off,

r.

, TABLE 1.1: COMPARISON OF PWR AND PCCL PARAMETERS i

Parameter Representative MIT PCCL*

PWR

1. Designed to Match Virtually Exactly Pressure, psia 2,240 hence same T.

H2 O density.

Temperature: high/ low, 'F 605/547 viscosity Flow velocity in core, ft/a 14 hence same fluid shear Thermal neutron flux, n/cm2 -s 2x10 13 hence real-time activation 2

Core heat flux, Btu /h-ft 180,000 hence same film AT 2

Steam generator heat flux, Btu /h-ft 67,000 Purification rate, sys. vols/m 10-3 hence, same exogenous sink

. strength

2. Other Comparisons of Significance Steam generator:

Tube velocity, ft/s 21 15 5 5 Reynolds Number 7.4x10 2.4x10 Nusselt Number 1,100 450 Length / diameter ratio 940 550 Axial T gradient 'F/in 0.08 0.38 Area ratio SG/ core 2.7 3.3 Core:

5 Reynolds Number 5x10 2.4x10 6 Nusselt Number 770 450 Length / diameter ratio 270 170 Axial I gradient, *F/in 0.4 1.2 2

Fast neutron fluv. n/cm -s 2x101 " 1x101 "

Loop transit tist s 15 5

• At maximum power (2C4vj maximum flow (2 gpm); operating conditions may be changed for specific experiments.

( - t 4_

CHARGING

, 2 SAMPLING f

m l

Ul -

. j i f D L CANNED 30 PSIA R0"OR THIMBLE He PUHP HELIUM SYSTEM 7au .. . . . :. . . . . . ::.

J vt**  : .... ..::: ... .

..::i.::

i:!:i:!.:i.:i.:.

. . . :i:i.:i.

.s p .:.:. ..:. .:...... .:.. .:. .. .

g . .... .::.

at . .. . .

zsHG13cm ..

3 -

.:::.r:.:.:.:.:.:.p .

PLI:NA w '

s l

ALUF INUM a T T/

THDIBLE ELECTRIC HEATER h

ZIRCALOY

' FUEL -

CLAI DING SECTION HITR CORE TI- AN Pb-llATH 9 U

MITR CORE TANK BOUNDARY

=

FIGURE 1.1: a) Schematic of Pb'R coolant chemistry loop: Portion inside

'!ITR core tank.

_ W i

dI "E

"V R k L U T n A I a

"/

" I ' M t

Y D H T l E T e T I r

LA R o CB I YR RTE l T H p GMg A

H c

ANR CD R LOU ST T LCT N l IS D I I

A D PR AV C .

M COE LP T. e 1 FME T

RI N

O OH V' e N

d i

t s

u o

' TT IA G E 1

MLI N G n YR o OT G XE i

(  % TNE R IN A

DN I t

( V nJL V P /B NM r

}V S

] ED GC P o

g DE g

p i RR D  :

y I Y p g E H o S N j1 V. o G E .

E G

D ) 7 y NRG GEAK I

N M l y

9

\'

R R T R - .

r D ( A AO TA T r

g'

, HVT C S A

t s

D i L m F

I e

I h

' c

/'

[ IdT

/

' bR hT N

I TD EK GN RA AT l

DLT IEH U

Q VG I Q"

l t

n a

o o

H IES M/,'

SI CE c DN SG LL "

/ PO EM IA R D DR W

' O U P P

7 '

DT TS [F "

^ f

/ $$> n 4" o

/ _ E c

/ V L IL i

t

' AA

'TI a

' A

, VT- i m H ]b e L/

U f E FL EL h c

/G N S J

IE LT E N

E N S

'I  ! S E

R G G )

L/N A'

M

'GN D

R D D Y

X E

L P

b

'R V I Y H

M A  :

/EH / [D bPIL 1 T

T

- N, I J R q [S .

/ 5, #

G EO 1

/ / l $

0 0

I N S I

E RT UA ]F E R

'/ -

I j c e

0 GKv RC A

H y SL KSU CEG ARE b 'F U G

I C BPR F 2 .

LD n .-

e .< .

i i

. L gamma heating by thc MITR at-full power can be safely dissipated  !

by radiation from the wall of the titanium bath can to the cold aluminum thimble wall (see Appendix 1.a).  !

Energy (1 20 kW) is normally rejected by conduction through a bed of copper shot to the thimble wall, and thence to the pool of MITR j coolant in the region well above the core (see Section 3.a).

As in a full-scale PWR, water decomposition is suppressed by main- .

t tenance of a small amount [1 50 cc (STP)/kg) of H2 dissolved in i

the coolant. A catalytic recombiner on the PCCL makeup tank 1 atmosphere provides assurance that a combustible mixture will not exist within the tank, and the discharge tank is vented to the MITR off gas system through a flame arrestor. Total in-

- 3 containment H2 inventory is limited by use of a small, low pres- i sure, transfer flask as the only source of this combustible gas (see Section 6.b).

Total in-core H2 O inventory in the loop is i 100 cc, hence void /

reflood reactivity is well within MITR experiment limits. Up to 2,000 ppm of boron (as boric acid) may be added during experi-  !

ments, but the total boron inventory is inconsequential, and, in

! any event, 98% B-ll is used (to reduce tritium production; Li-7 is used for LiOH treatment for similar reasons) (see Section 4). ,

In-core materials have been selected (and screened using test irradiations) to insure that even unshielded dose rates during j loop handling could not exceed several R/h--a value easily reduced by two orders of magnitude using a shielded transfer flask / storage [

container (see Sections 2.b and 3.b).

t

\

l

._c . . . _ _ - . - . . _ _ . _ , . _ , _ . _ . . _ _ . . _ _ ._m__..- _

-e . i 7

• The loop support arrangement is conservatively designed to pre-clude dropping an unsupported loop into the core, and to insure that excessive weight is not borne by the MITR core grid (see Section 2.c).

The above considerations, and others, are described in the various sections of the body of the report, where supporting calculations and references are also documented.

2) Experimental protocol An appreciation of the type of experiments to be conducted using the PCCL is essential to the understanding of its various design features, and the safety implications of loop operations.

Research contracted with EPRI and ESEERCO for the first several years of operation is devoted to measurement of the effects of coordi-nated LiOH/H 3 B03 treatment (i.e., effective operating pH) on the produc-tion, transport, and deposition of corrosion product radionuclides on ex-core surfaces in a PWR environment. Thus, the experimental procedures are relatively straightforward:

• operate the PCCL for approximately one month out-of pile to pre-condition the corrosion film on all loop surfaces;

• move the loop into the MITR core tank for another one- to two-month run under steady state conditions (temperature, heat flux, flow rate) in the presence of neutron and gamma irradiation;

• remove and disassemble the loop to assay the amount and spatial distribution of important radionuclides such as Co-60 and Co-58 on loop surfaces (amounts measured in microcuries are to be expected);

( . 1

~

e repeat.otherwise identical runs, varying only -the LiOH/H 503 3

ratio, and compare results to activity transport models; use the  ;

L results as the basis of recommendations to PWR operators for a [

regimen of coolant chemistry control which will reduce exposure doses. (Improvements by as much as a factor of 10 can be antici-pated, based upon the current level of understanding.)

Section 3 of this report provides appropriate-detail on the proposed operational and experimental procedures, and Section 7 discusses the sub-sequent disposal of radioactive waste products.

l 1

l-i ,

i I

4

2. DETAILS OF THE PCCL 140P DESIGN

-a) Design Specifications, Tolerances and Safety Margins Figure 1.1 above, indicates the layout of the loop components which will be within the MITR-II core tank. Specifications for the major

components are summarized in Table 2.1. Specifications most relevant to the safety of the loop and the reactor are the thimble and loop maximum pressure specifications and the heater shut-of f mechanisms. Dimensional tolerances are standard machine shop practice except in the case of the thimble-dummy element mating surfaces. For this fit , the dummy element will be measured af ter fabrication and the critical thimble dimensions specified to allow 0.005 in , clearance , +0.000-0.005.

b) Loop Materials - Compatibility with MITR-II Core and Coolant Several types of materials issues must be considered in evaluat-ing PCCL safety compatibility with reactor primary coolant, reactiv-ity, and activation. Reactivity issues are dealt with in Sections 2.d and 4, and activation is covered in Sections 5 and 7. This section deals with the compatibility of loop materialu with the reactor primary cool-ant.

As discussed above, the loop components are encapsulated in an aluminum thimble , which will constitute the major surf ace in contact with the reactor coolant. 'the material used will be a certified reactor grade alumintan and is thus within the envelope of materials approved for use in the core tank as specified in the MITR-II Technical Specifications Sec-

tion 5.3. Apart from the aluminum, small amounts of gasketing material will be used. A Viton 0-ring will be used to seal the upper flange at the top of the thimble (above core tank water level), and a pure lead gasket or other metal-to-metal seal will be used to seal a 2" port at the a

< . s l

TABLE 2.1: PWR COOLANT CHEMISTRY LOOP DESIGN SPECIFICATIONS IN-CORE SYSTEMS Pumpt Capacity (GPM) 1-2 Design Canned-rotor Temperature (F/C) 603/317 Normal Operating Pressure (PSI / Bars) 2200/152 Maximum Design Pressure (PSI / Bars) 3000/207 Loop Differential Pressure (PSID/ Bars) 15-25/1-1.7 Material Inconel/ Stainless Steel Power Supply 220 or 440 VAC, 100 watts Variable Speed Hestert Power (variable) 0-20 kW Power Distribution Linear Length (heated section)(in/cm) 24/61 Diameter (in/cm) 0.316/0.803 Sheath Material Carbon Steel Voltage (VAC) 220 Shutoff systems - Automatic shutoff initiated by one of two independent thermo-couple temperature signals.

- Manual shutoff by experi-menter/ reactor operator under loop operating procedures.

- Passive shut-off by melting of aluminum link in power line if other systems fail and bath temperature reaches

• 1100 T Thimblet Material 6061 Aluminus Wall Thickness (in/mm) 0.125/3.2 Design Pressure (PSI /*ars) 30/2.1 Maximum Pressure- 100 PSI (relief valve)/

Loop Leak Accidt <500 PSI (no relief)

Proof Pressure 150 PS1

. .. .- 3.

.. -Heated Section' (Simulated Fuel Pin)

Material Zircaloy 2 or 4 Diameter OD (in/ min) 0.312/7.9  :

ID (in/ min) 0.26/6.6 Con figura tion "U" Tube t

. Heated .1Angth (approx)(in/Cm) 50/127 Lead Bath Container:

Material TI 6Al-4V Wall Thickness (in/ sun) 0.032/0.79 i Simulated Steam Generator Tube:

Shot-Bed Heat Transfer Medium Copper Shot Tubing Inconel Diameter OD (in/mm) 0.312/7.9 ID (in/aun) 0.26/6.6 Out-of-Core System: .

6 Charging / Pressurization Pump:

Metering Pump Positive Displacement Flow Rate (cc/ min) <1000  ;

Maximum Pressure (PSI / Bars) 3000/207

, Back-Pressure Valve Gas or Spring Loaded i

Check Valves Dual Ball-Type to prevent back

, flow and depressurization  ;

k i

a 4

bottom of the snor bed heat exchanger section (approximately 18" above the top of the MITR-II core). Note that all screws or bolts on the thimble, which could potentially fall into the core tank, will be cap-tured or held so that they cannot fall if inadvertently dropped. Small amounts of gasketing material are permitted under the existing approval, and it is not expected that they will produce any significant impact on the primary coolant. Inconel tubing for feed /bieed flow, and instrumen-

, tation and heater and pump power wiring will be fed through the top flange of the thimble and through the core tank wall. These feedthroughs will be above the core tank water level and will be isolated f rom the core tank environment in polyethylene tubing or stainless steel conduit (as is currently done with other experimental and operational facili-ties). These components will be subject to splash and humidity from the primary coolant, but will not be continuously exposed. Again, no signi-ficant impact on the primary coolant is expected, and the proposed PCCL falls within the envelope of previously approved procedures from the standpoint of coolant compatibility.

It is recognized that the materials which will contact the primary l

coolant as described above are subject to certification requirements.

Procedures for procurement and quality assurance of such materials are n

described in Section 2.f. The consequences of a thimble leak followed by re-release of primary coolant to the core are discussed in Section 6.b.4.

c) Structural Supports and Loading Figure 2.1 shows the thimble support system. It consists of a f stainless steel bridge bolted to the core tank wall and capable of sup-porting the full weight of two loops (in air) with appropriate safety i i margin. Each loop will be attached to this support using two spring-a I

[ . . ,

% ,96' c,oM 0 \

at#K#

\nol eT# ,9Y" h

S 1

(

G

\ -

\ \ . ~

9, 1

u r

bO  %%* ,o*

\@I

%. ,A vW o -

• p -

.\

<d #

\

FIGURE 2.1: Loop mounting bridge position showing the configuration for operating the loops in the A-3 and B-3 element posi-tions.

e . .

-13 loaded bolts designed to accommodate the thermal expansion of the loop at the operating condition. These supports will be designed to carry up to the full 200 lb. weight of the loop in water. To prevent vibration of the thimble in the flowing primary coolant, a small fraction of the loop weight could be carried on the lower grid plate, or an upward force could be exerted by the hold-down tabs against the upper grid plate as in the case of the ICSA's.

The loop thimble will be secured agalnst ejection from the core by locking of the upper grid plate over a latch tab on the thimble. This is equivalent to the fuel element securement. (Note also that the clearance between the top of the loop thimble and the bottom of the core tank lid will be only several inches, and complete ejection of the loop from the core is therefore impossible even if the locking system fails.)

Provisions against dropping the loop and/or shield structures into the core tank are discussed in Section 3.b.

d) Power Peaking in the MITR-II Core Computer calculations of the effect of the loop experiment under various operating and accident scenarios are being carried out by Opera-tions staff. Experience with previous experiments such as the FCE sug-gests that the PCCL will meet the Technical Specification requirements in this regard. A report on the results of the calculations will be made available to MITRSC members ac soon as it is completed.

I

e) Instrumentation and Control The instrumentation and control requirements for the PCCL are rather simple. Redundant thermocouples on the loop's inlet and outlet plena are used to generate a loop average temperature signal, and power to the heater is varied to maintain constant T-average: much the same approach as used on actual PWR units. Control and safety instrumentation both come in two categoriest essential and supplementary. The former category consists of redundant thermocouples which measure loop hot and cold plenum temperatures and heater bath temperature. The PCCL will be cooled down and depressurized unless one thermocouple of each pair is functional. Other instrumentation (such as the flow meter and humidity detector) is used to provide supplementary information, and loop opera-tion may continue should items in this category become inoperable.

The principal safety system consists of redundant thermocouples measuring the temperature of the lead heater bath; an overtemperature signal is programmed to cause the interruption of power to the heater, since this is indicative of a serious accident in progress (see Sec-tion 6.b). All severe accidents ultimately lead down this path, and very little damage is done if heater power is cut off.

Most of the other instrumentation on the loop itself is diagnostic in function (e.g., to measure flow or pressure), or for the purpose of logging data pertinent to the interpretation of experimental results (e.g., H2 and 02 concentration, pH and conductivity). Readings (and for certain signals, alarms) from this instrumentation can help loop opera-tors identify the specific nature of an incident which disables the loop, but heater bath overtemeprature alone is sufficient to satisfy all safety protection needs.

, A parallel set of cafety-related diagnostic instrumentation is pro-vided on the thimble gas space: over ' and under pressure alarms to indi-cate . loop or thimble leakage, respectively, and a humidity gauge which will respond to water ingress in either scenario. The humidity detector (solid state moisture detector) will be mounted out-of-core on a thimble vent streme obtained through a capillary bypassing the thimble pressure relief valves. This allows for periodic testing and replacement of the detector without interrupting loop operations, which is important given the relatively low reliability of such detectors. The long response time of this system and the reliability question preclude automatic loop shut-down from this signal . Considerable time is available for deliberate i action by the operator subsequent to most malfunctions; those rare sequences that are more serious will be interdicted by the heater bath overtemperature protection syste:n.

. All alarms for the PCCL which require reactor operator response will be brought to a comon panel in the control room. Breakers for emergency shut-down of the loop heaters and pumps will also be provided in the con-trol room. (See Section 3 for typical response sequences.) It should be emphasized again, however, that operator response is not required for reactor safety, and in most cases is aimed only at minimizing conse-quences to the loop equipment.

f) Quality Assurance Requirements The PCCL will be designed , construc ted and installed in conform-ance with the MITR quality assurance (QA) program. A copy of the rele-vant parts of this program is appended to this report. Under this program, a QA file will be maintained in the Reuctor Operations of fice inco rpo rating : material certifications, design and construction

drawings, safety experiment and proof testing documentation and procedure approvals.

It should be noted that a hierarchy of PCCL systems with varying impact on reactor operation and salety can be established. Materials which contact the primary coolant or which are exposed to significant neutron flux are the most critical and must be the most stringently controlled. Materials within the MITR-II core tank but not in the above two categories will have somewhat less stringent certification require-ments. Safety-related equipment such as pressuro relief valves and the overtemperature protection system (including thermocouples, relays and the fusible link) will be subject to testing and calibration require-ments.

Authorization of personnel to carry out critical loop operations, and to carry out or approve quality assurance procedures, will consist of c a letter signed by one of the project co principal investigator; and by the Director of Reactor Operations.

1

_17

3. OPERATIONAL AND EXPERIMENTAL PROCEDURES In this section the general procedures for system operation and the conduct of experiments are outlined. Since the PCCL has been designed to achieve and maintain an invariant steady state status for up to several months non-stop, these procedures are fairly simple. Except for a few hours during startup and shutdown, the loop is under automatic control, and those responsible for its operation need only make occasional (less than daily) small adjustments in power level, pump speed (i.e., flow rate), and feed-and-bleed rate. Even less frequent changes will be made in thimble helium pressure and makeup tank hydrogen overpressure. Most data of significance is measured by built-in instrumentation and logged by computer, and makeup / discharge tank samples need only be drawn for supplementary analyses on a weekly basis. The most significant data from the point of view of the experimentalist will be recorded by the on-line deposition monitor, and measured by post-mortem dissection and gamma scanning of the loop's tubing.

In what follows, the procedures are outlined in narrative form; detailed step-by-step checklists will be prepared for use by the loop operators and the MITR-II operating staff, a) Loop Operations Each run will begin with rea9sembly of the loop, usually using new in-core Zircaloy and out-of core Inconel tubing. The loop will then be installed in an out-of pile tank for its preconditioning run. The objective of this phase is to establish a significant corrosion product film on all internal surfaces. Operation will be virtually identical to subsequent in pile operation (i.e., at full pressure and temperature, hooked up to the same control and auxiliary systems as will be used to

support its in pile operation), except that the thimble helium space will be evacuated to permit running at very low power, simulating "isothermal" operation. Full power operational tests will also De undertaken before transfer of the loop to its in-core position.

Since the PCCL will then be moved into the MITR-1I core tank with as little perturbation s possible, the preconditioning run--about one month long--is an excellent "shakedown cruise," which should help insure that all systems are operating flawlessly before each in pile run in initi-ated. Transfer from out-of pile will be effected with the loop discon-nected (and sealed off) from the feed and bleed train, in a cooled-down and depressurized state. The insertion peccedure will include the use of mechanical stops to prevent the application of "missile forces" to core structures if the thimble is dropped. The thimble helium pressure of 30 psia will be maintained during transfer operations, since out-of pile operations at power serve to verify that air was not left in the thimble prior to helium back-fill (see Section 6.b.3). Thimble outside diameter measurements will be made at this point ot ensure that shot-bed ratchet-ing is not occurring. After it is emplaced and secured in the core tank (see Section 2.c), all fluid, power and instrumentation linee will be reconnected, and the MITR-II button up/startup can then proceed as normal.

In parallel with MITR-II startup the PCCL is then pressurized cold using its feed-and-bleed system. Next the PCCL's in pile heater is turned on and power is ramped up, at rates set in the operating proce-dures, to its normal full power rating (10-20 kW) as specified for the particular experimental run which is scheduled. During heatup, thimble helium is vented to keep its pressure at about 3025 psia.

, At this point loop heater power / loop temperature control is put on automatic. As the MITR-II itself comes up to power, gama heating will gradually assume about 10% of the PCCL heat load, and electric power will autceatically be reduced to keep the loop at steady state with respect to coolant tempe ra t ure . From this point on the objec tive is to hold the established conditions until the PCCL is shut down for removal (in 1-2 months). Over wekends, when the MITR-II is shut down, the FCCL control system will compensate for the reduction in gamma heating.

Of more interest as regards a safety analysis is the operator response in the event of an accident which disrupts normal loop opera-tion. Sections 4 and 6.b discuss such sequences in more detail. He re we confine the discussion to operator action. Table 3.1 sumarizes the responses appropriate to a variety of incidents. Note that in each case the loop control system is programmed to execute a suf ficient action (usually cut-of f of heater power) to put the loop in a safe state, and l

hence operator action is of the nature of confirmation and backup.

Heater bath temperature will indicate and alarm in the control room. In the event of heater control (automatic power cutof f) failure, operator l control is important , but a passive ("fusible link") heater shut-of f is provided to avoid serious overtemperature incidents without any interven-tion.

b) Post-Operation Handling Following completion of a PCCL irradiation the loop will be removed f rcrn the reactor and transferred to a shielded test stand where it can be disassembled for analysis . This section outlines the procedures to be f ollowed for removal, disassembly ard analysis, with i

l l

TABLE 3.1: INCIDENT RESPONSES Event _ Symptom / Alarm Automatic Control Operator Response Loop rupture / - High humidity Cut-off power to Open circuit breakers loss of coolant - Low loop pressure heater for heaters Relief valve open - High heater bath T Turn off charging and circulating pump Severe thimble leak - Low He P Cut off power Open circuit breaker for

- High humidity to heater heater; execute cooldown Failure of makeup / - Low loop pressure Cut off power to Open circuit breakers for pressurization pump heater heaters and pump Heater control failure - High Pb bath T (Heater control Open circuit breaker for (overheat) - High loop T failure postulated) heater but T alarms work

• e Failure of circulating - Low loop flow Cut off power to Open circuit breakers for c$

pump - High Pb bath T heater

• heaters and pump Computer Failure - Loss of all signals Fall-safe feature: Execute cooldown and control heater power cutoff capability Loss of all Cambridge - MITR control room Heater goes off; sys- Open circuit breakers for electric power procedures relied tem is designed to entire system; consider upon prevent heater re- restart in conjunction with start when power is experimenter.

restored NOTES: Cutting off power to the pump, where indicated, is to protect the pump against damage and is not a required safety procedure per se.

Emergency electric power is not required in the event of a local or global power failure.

Relief valves are all spring-loaded, and energy removal in extremis is by radiation.

• In the event of failure of both automatic and manual shut-off of the electric heaters, the fusible link in the heater power supply line will melt and cut off heater power. .

5 O

emphasis on the measures used to prevent possible damage to the reactor and to minimize irradiation dose to personnel.

1) Loop removal Loop removal will take place on or af ter Monday, allowing about sixty hours of decay time from Friday evening shutdown. (See Section 5 for an estimate of the radioactive inventory at this time.)

We heater power vill be ramped down and loop depressurization vill follow when the temperature has dropped suf ficiently. With the loop cold and depressurized and the core tank lid removed , the power and instrumentation leads will be disconnected, freed frcxn the core tank wall feedthrough and secured to the thimble head . The feed and bleed and helium pressurization lines will be disconnected at the thimble feedthrough and capped.

A lifting harness will then be fastened to the eyebolts on the thimble head, and the harness placed on top of the thimble lid so that it may be reached later. The lifting harness has three lines, with two of them adjustable, so that the loop hangs vertically, he nuts dich hold the thimble down to its support bridge are now removed. At this point a special core tank lid is to be lifted into place by the overhead reactor crane. This lid is larger than the maximum opening to the core tank, and fits over the studs protruding from the top of the core tank. A single pott in the lid is positioned directly over the thimble to be removed so that in lif ting the loop from the reactor nothing may be dropped into the core tank. Two three foot long alignment rods are then screwed into the pins on the support bridge so that in lif ting the loop from the core (and more importantly when put t ing it in) the loop does not rotate or tip, which might cause it to wedge into the dummy fuel element , he rods (and

all other hardware over the core tank) will be secured to prevent their falling into the reactor. The loop lif ting wires will then be threaded through a shielded cask (similar to the cask used by Operations for con-trol blade removal) which is then positioned on the lid to allow the thimble to be drawn through.

Loop removal will be performed using_the 3-ton hoist. Once the pump pod at the top of the thimble has cleared the top of the cask, additional shielding with a 4 in diameter central hole (to fit closely around the 4 in thimble section) will be attached to the top of the cask. This structure also serves as a stop by supporting the pump "pod" .at the top of the thimble if the thimble is dropped. When the lower thinble section is in place in the cask, a-support frame for the upper thimble will be bolted to the cask.

Having secured the loop in this frame the operatoe will detach the loop from the lifting hoist and connect cask lif ting embles to the large reactor crane. The cask containing the loop will then be moved over and down to the reactor floor (just to the right of the hot cells, against the containment building wall).

The thimble lid may now be removed and all the electrical and fluid lines to the loop disconnected. The pump will be drained through a tee fitting and the water will be collected for analysis. Once the pump is empty it should be completely disconnected and removed from the thi; i head. The open end of the tee should be capped and a line connected to the open end of the loop. The water in the loop may then be pumped out through the tee using helium pressure. Both lines may then be removed and the loop ends capped.

o A small hand hoist fixed to the wall above the loop is then attached to the loop itself. At this point all operations in the line of the radiation beam have been completed and so the copper shot which has been serving as shielding may be drained through the shot drainage port at the bottom of the heat transfer section of the thimble. Care will be taken to contain the shot. The loop is now ready to be removed from the thimble. Since the in-core section of the loop will not be immediately surrounded by shie* ding during this procedure all non-essential personnel should be away from the area and the operator will be shielded by a con-crete wall. The operation may be observed through a video camera and monitor. The loop is to be removed from the thimble to a position over a disassembly t ack next to the loop stand. The hoist will move by pivoting 4

its bracket between two well-defined stopo. Cables will run from the 1

bracket through the stops and down to the operator position where they 1

will be secured. When the loop is over the disassembly rack, it is lowered into a plastic sock (for contamination control) with the in-co re section going into a transportation cask. Once within the cask the in-core section may be disconnected from the rest of the loop at its SWAGELOK8 fitting and heater connection; this will be done using tools designed to avoid personnel exposure in the radiation beam from the in-core sections. The cask containing the activated in-core section is then rolled out fran under the rest of the loop and a lid may be placed over the to p .

2) Loop disassembly Disassembly and sectioning of the in-core section of the loop is to be carried out in the reactor floor hot cell. The transport cask containing the activated section will be lif ted to the top of the hot

cell and positioned over an access port. A small power supply will be connected to the heater and a lowering bracket attached to the top of the Zircatoy tubes. By opening a sliding door in the bottom of the cask the in-core loop section is then lowered down into a holder in the hot cell.

The lead in the titanium can is then melted by the heater and the Zircaloy lifted out. Sectioning and characterization is then carried out using the hot cell remote manipulators and tools prepared for this purpose.

Activities in the loop water and the out-of-core loop sections are expected to be easily manageable, because the activities involved are small. Contamination control is critical to the quality of experimental data acquired, and such control can be achieved through careful application of standard decontamination procedures. Analysis of the tubing will be performed within the exclusion area, avoiding the removal of this activity.

O o e

~

4. MAXIMUM EFFECTS OF REACTIVITY, PRESSURE AND TEMPERATURE In this section a conservative set of maximum ef fects will be esti-maced . In many cases a highly improbable sequence of events would be required to create a situation leading to the ef fect in question, ber our interest here is more in establishing an envelope of limits rather than examining a variety of more plausible scenarios.

a) Reactivity Effects As with all other in-core experiments, it is essential that the PCCL meet the Technical Specifications (Section 6.1, see Appendix 2) with respect to its potential ef fect on MITR core reactivity during postulated accident scenarios. In the case of the PCCL two phenomena are of pseticular concern: the presence of B-10 (in the form of di6 solved boric acid) in the simuisted PWR coolant contained within the PCCL circuit; and the reactivity worth of loop and thimble water contents during void /reflood incidents.

. 1) Reactivity ef fect of B-10 Two hypothet i cal scenarios of increasing severity are postu-3 lated, involving:

i a) sudden voiding of the in pile water and its containad maxi-mum (2000 ppm) boron content

, b) a non-mechanistic incident in which all in-pile boron is first concentrated in pile and then ejected i

To minimize tritium production and reactivity effects, boron enriched in B-11 will be used in experiments; the B-10 content is 2 w/o

! instead of the natural value of 20 w/o. Strict administrative controls will be imposed to prevent inadvertent substitution of natural boric i

acid. In addition, the tritium concentration of the PCCL coolant inven-tory will be a'sayed weekly; among other things, it will serve as a posi-tive indication of B-10 concentration via the 10B (n,8Be)3 T reaction.

We then have for our three scenarios:

Case H2 O Involved Total B B-10 (a) 40 g 8.0x10-2 g 1.6x10~3 g (b) 500 g 1.0 g 2.0x10 2 g Appendix 1.b develops an estimate for the reactivity worth of B-10 in the MITR-II core:

$3l = 0.998%

3 B-10 Thus, we have:

Case Scenario Max. % ik /k (a) Eject all ir- ,re B-10 1.6 x 10-3 (b) Eject all PCCL B-10 2.0 x 10-2 As can be seen, cases (a) and (b) fall well within the allowable limit for a movable in pile experiment, namely ak/k < 0.2%.,

Since we will assay PCCL discharge tank water weekly for boron content and since the feed / bleed rate is only 5 kg H2 0/wk, it is not conceivable that hide-out in pile of all of the charging tank inventory would go undetected.

..~ Also, note that the ef fect of voiding boron and water are opposite in signi since boric acid is highly soluble, realistic scenarios which would decouple these compensatory effects are hard to imagine.

2) Water floodins/ voiding incidents There are three scenarios of interest here

-(c) sudden flooding (preceded by. undetected voiding) of all in-core PCCL inventory (40 cc = 40 g under cold condi-tions)

(d) loop rupture and drainage into the in pile thimble

-(1 75 ce = 75 g, again cold) or thimble rupture (e) undetected voiding of the volume between the thimble and the dummy element, followed by sudden reflooding

] (1 100 cc = 100 g) l For small water voids, we have:I I

SE. k/i I < 2 milli 8/g EHO 8 -

2 112 0 Thus, we find for 7 = .00786, $

%h/g 1 1.57x10 3 Case Scenario %Ak/k (c) PCCL tube flooding 0.062*

(d) Loop drainage into thimble 0.12*

(e) Dummy / thimble channel reflood 0.16 I

Personal communication, J. Bernard, October 6,1986, and his memorandum to 0. Harling dated October 7, 1986. See also MITR-II Start-up Report for Core IV.

• Note that all values exclude the opposing effect of any B-10 dissolved in the water involved.

A.1 these cases are within the allowable reactivity restriction for a movable MITR-II experiment , namely Ak/k < 0.2%.

Note that measurements show that flooding a 1.75 inch ID ICSA would add 0.982% Ak/k, for an ef fective large void coaf ficient of =1.51 milli 8/g. I This figure is less than the small void coef ficient of 2.0 milli 8/g used above.

The following additional qualifiers to the preceding analyses should be noted:

1) The potential events (sudden, undetected flooding) can only occur in their most severe version when the PCCL is in a cold startup/

shutdown mode ; otherwise, the in-pile heaters (and energy stored in their molten Icad bath at > 600'F) would immediately boil any water in the (low pressure) thi=ble's in-pile gas space. Th e r e -

fore, it will be normal practice (but not an absolute require-ment) that the PCCL power level exceed 10 kW whenever the MITR is critical. Thi s is not a severe imposition, since the anticipated experimental protocol involves non-stop steady state runs of a month or more in duration with loop insertion /removs1 over week-ends.

2) The helium atmosphere in the PCCL thimble will be maintained at 2 atm (30 psia), so that out-leakage of hellun would take prece-dence over in-leakage of MITR cooling water (maximum pressure at the bottom of the core--atmospheric plus hydrostatic of

= 22 psia) if there is thimble fa il ure ,

I Memo from L. Clark to MITRSC dated July 5,1978.

i i

. 3) The thimble will be tested for leakage (visual indication of gas bubbles) at 50 psia internal pressure at the end of its out-of-pile preconditioning run, prior to being transferred into the MITR core tank. Post-manufacture and periodic testing with a helium leak detector will also be performed.

4) A humidity detector is installed in the thimble's vent line to provide indication of the presence of small amounts of water ingress. An alarm will be sounded in the MITR control room to warn of a potential problem.

5) Most importantly, the titanium can housing the in pile heater bath fits snugly into the aluminum thimble at its top (except for small grooves to admit helium to this region). Thus, rapid drainage of either PCCL or MITR water into this region from above is unlikely. Hence, only sudden massive failure of the conserva-tively designed lower thimble is a plausible maximum accident initiator.

Based upon the very conservative analyses documented above, there does not appear to be any way in which installation and operation of the PCCL can exceed MI7d-II Technical Specifications with respect to reac-tivity effects. A confirmatory measurement of the reactivity difference with and without water in the FCCL's Zircaloy in pile tube will be made as part of its initial checkout.

b) Pressure Effects .

In its normal operating state, the PCCL consists of roughly 1/2 liter of hot (*600*F) water under = 2,200 psia pressure inside a loop comprised mainly of 0.25 inch ID tubing, all surrounded by.an aluminum 3

thimble with approximately 1 ft free volume containing helium gas at

" 30 psia.

~

The high pressure part of the system is, because of its configura-tion (spall di.ameter), small volume, and encapsulation, less hazsedous than common corrosion test autoclaves ; and a small amount of helium at two atmospheres is also not a safety problem. Thus , normal operation does not create any situations of concern, and one nust turn to low-probability accident scenarios to postulate consequences of potential interest: the most severe being an instantaneous loop rupture or D3CA

( lo s s-o f-c ool an t acc iden t) .

A particularly simple, but quite realistic, calculation of an upper limit on the pressure attainable in the thimble following instan-taneous release of the entire loop water inventory can be made when one recognizes that the thermal balance is dominated by the large mass of copper shot (=200 lbs), which completely overwhelms the small amount of water involved (=1 lb) and the even smaller mass of helium present 9

3

(= 3 ft at STP) and even the = 3 lbs of rolten lead at = 700*F in the in-pile heater bath.

The average temperature of the shot bed remains virtually unaltered P.hroughout at = 350*F. Enough stored energy is available to vaporize all wa te r no t immediately flashed to steam upon loop depressurization. Thus, we need only compute the pressure of the helium-steam mixture at 350*F.

At 350*F saturated steam has a vapor pressure of 135 psia and a specific volume of 3.342 f t3/lb; hence, compression of 1 lb into 1 f t 3 (loop-free volume) would require a pressure of approximately 135 x 3/342 = 451 psi; to this , add the initial (and final) helium pressure of 30 psi = (at 350*F) to obtain a post-blowdown pressure of 481 psia in the thimble.

., . l This value is conservative because it ignores condensation of steam on the cold (=100*F) aluminum walls of the thimble. Even so, it is a quite tolerable pressure. Moreover, in the interests of design conserva-tism, the loop is protected by a relief valve which will vent the thimble at < 100 psi, and a back-up rupture disk which will burst at = 500 psi.

Note that the thimble and all of its contents are designed to with-stand a full vacuum, hence cold conditicas in which steam condenses after expelling the helium fill gas will not lead to additional problems. In fact, when operated in its "isothermal" pre-conditioning mode out-of-pile, the loop gas space is evacuated to increase the thermal resistance of the shot bed, and permit high temperature operation at "zero" (actu-ally very low) power.

c c) Temperature Effects The design philosophy underlying our approach to PCCL safety,dur-Ang extreme accident scenarios has been to provide'for sufficient passive (i.e., radiation) cooling to prevent material temperatures from ever exceeding recognizably safe values: for example, conforming to the same

< 2,100*F post-LOCA limit on Zircaloy temperature as is imposed on actual PWR units by the NRC. Appendix 1.a gives an estimate of the gamma and fast neutron heating of the in core section, with experimental data and calculations to show that the total radiation heating can be dissipated at maximum temperatures < 1500'F.

The combination of active and passive (aluminum fusible link enposed to the lead bath) shut-off mechanisms for the electric heaters is expected to prevent heater power being applied at lead bath temperatures above = 1500*F. The nuclear heating analysis is therefore sufficient to show that Zircaloy-water reactions will not occur.

5. RADIATION LEVELS AND ALARA CONSIDERATIONS The PCCL components which will be exposed to significant neutron flux in-core are Zircaloy tubing (Zircaloy-2 or Zircaloy-4), pure lead for the conductive bath, electric heaters consisting of Nichrome heating elements with magnesium oxide insulation and carbon steel sheaths, a titanium-aluminum-vanadium alloy can containing the lead bath and the lower section of the aluminum thimble. Stainless steel sheated thermo-couples (chromal/alumel) in the lead bath and heaters will also be exposed to neutron flux. Table 3.1 sives an inventory of the elements present in-core in the proposed PCCL.

Apart from the Zircaloy tubing, which is necessary for the simula-tion of the PWR flow loop, all in-core component materials were chosen to minimize activation within the constraints imposed by the functional requirements of a given component. In several cases, the activation of the components is due largely to impurity elements, leading to some uncertainty in prediction of the activity levels. Activation experiments have been performed in an equivalent core position to that proposed for the PCCL to assist in estimating post-irradiation activity levels.

Table 5.2 gives estimated dose rates after a one month irradiation and 60 hours6.944444e-4 days <br />0.0167 hours <br />9.920635e-5 weeks <br />2.283e-5 months <br />' decay.

The activities present lead to a total unshielded gamma dose rate of less than 10 R/h after a typical one-month irradiation and decay from reactor shutdown at 1800 hours0.0208 days <br />0.5 hours <br />0.00298 weeks <br />6.849e-4 months <br /> on Friday to Monday morning loop remo'ral.

This dose level is within the range of experience of MITR operators and radiation protection, and the loop handling, shielding and procedures discussed in Section 3.b will keep personnel doses small for transfer and disassembly operations. A small amount of the a-emitter polonium-210

TABLE 5.1: IN-CORE MATERIALS INVENTORY FOR THE PCCL Component Material Composition (wt.%) Wt. (g)

U-tube Zircaloy-4 1.45Sn, 0.21Fe, 0.10Cr, 110 bal. Zr and impurities Heater 1018 Steel 0.200, 0.90Mn, bal. Fe and 85 impurities Inconel 600 76Ni , 0. 2 5Cu, 8Fe , 15. 5Cr , 35

0. 25Si, 0.51ht Magnesium oxide 60.3Mg, 39.70 (high purity) 180 Conduction bath Lead 99.9 Pb (<5 ppm Bi) 6000 Containment tube Titanium 99.5 Ti, bal. impurities 240 Thimble 6061 Aluminum 0.8Si, 0.7Fe, 0.4Cu, 0.15Mn, 670 1.2Mg, 0.35Cr, 0.25Zn, 0.15Ti TABLE 5.2: ESTIKATED UNSHIELDED GAMMA DOSE RATES FOR PCCL IN-CORE

y. COHPONENTS AFTER 20 FULL POWER DAYS AND 60 HOURS' DECAY Component Principal Activities Y Dose Rate (R/h at 1M)

U-tube 95Zr, 51Cr 1.5 51 Heater 58Co, 60Co, 597 ,, Cr 0.3 Conduction bath 203Pb, 203Hg 4.0 46 Sc 0.2 Containment tube 24 59p ,, SI Cr Thimble Na, 58 Co, 60Co, 1.0 7.0 NOTE: The dose rates expected are strongly dependent on impurity content in several components. It will cherefore be necessary to recheck the activation data with the actual loop materials when these be-come available.

will also be generated in the lead bath, principally from bismuth impur-ity. Polonium production in reagent grade lead has been estimated from experiments at = 1 pCL/kg lead for a one-month irradiation. The plug which isolates the lead bath from the upper thimble is' expected to be an effective barrier (by condensation and adsorption) to polonium, and all lead bath disassembly operations will be performed in a hot cell.

Personnel exposure to polonium-210 is therefore not expected, but care must be taken in handling the Zircaloy tubing section, which may take up some polonium. Component handling proccdures used will call for routine wipe tests and alpha counting.

It should be noted that the out-of-core sections and the loop water are expected to have only pCi levels of activity. Doses from water chem-istry analysis and post-irradiation inspection and gamma-counting of the Inconel sections will be insignificant once rhese sections are disassem-bled from the in-core section in the shielded test stand. Tritium pro-duction in the loop will be minimized by the use of lithium hydroxide and boric acid enriched to 99.9% Li7 and 98% Bll, respectively. Since the flow rate through the bleed capillary is slow (=30 cc/h), virtually no 16 N activity will be transported outside the reactor core tank during loop operation, and no effect on core top radiation levels is expected.

Personnel exposure during loop operations will be limited. Routine oper-ation will require experimenters to be present on the reactor top and loop platform for approxiuately one hour daily, resulting doses far below the allewable limits for each worker.

Procedures for all loop operations will be developed and implemented with shielding, ventilation and appropriate controls to insure that radi-I ation exposure to all personnel is as low as reasonably achievable.

6. PCCL SAFETY EVALUATION a) Safety and Operational Envelopes To facilitate the review of the safety considerations for the PCCL, it is helpful to make specific reference to the MITR-II Technical Specifications that where written to provide an envelope within which the MITR Sa feguards Committee and MITR staf f can approve experiments. Pages 6-1 to 6-7 from Section 6 and Section 5.3 of the Technical Specifications are provided as Appendix 2. The PCCL design and operation as described in this Safety Evaluation Report is to be in conformance with these s peci fica tions .

In considering the effect of normal PCCL operations on MITR opera-tion, the following points should be noted:

--The gamma energy deposited by the MITR in the PCCL would otherwise be deposited elsewhere in the MITR structure; hence there is no increase in total heat load from this phenomenon--merely a redistribution (which is usually beneficial).

--In incidents severe enough to warrant scramming of the MITR, gamma heating of the PCCL is reduced proportionally. Hence the PCCL does not aggravate the consequences for its host reactor.

--Accordingly, only the energy added by the electric heater (PCCL pumping energy is negligible) need be considered in assessing the impact of the PCCL during steady state, tran-sient or accident scenarios.

Furthermore, investigation has confirmed that insertion and opera-tion of the PCCL will not interfere with proper operation of the MITR-II

o l cooling system under normal conditions. In particular:

-- The transition between the 2 in. diameter lower thimble and the 4 in. upper thimble is. gradual, and al2 in. above the core, so as not to perturb in-core flow patterns.

Pre-operational tests will attempt -to ; verify that core flows are not significantly altered.

-- Total PCCL power input, 10-20 kW, is comparable to the reactor decay heat by Saturday morning following Friday evening shut-down (=25 kW). Provision for additional decay heat removal will therefore be necessary.

-- Mechanical interference with control rod drives is ruled out because the PCCL is firmly captured, both in the core grid plate and at the PCCL support bridge across the reactor top, in a position with clearance all around.

.. -- Interference with the emergency core cooling sprays (ECCS) will be prevented by relocating one spray nozzle to ensure that no areas of the core are shadowed by the PCCL loop or loops. Operations staff is investigating the necessary physi-cal and procedural changes.

b) Malfunction Sequences and Consequences As noted in the preceding sections of this report, design features have been incorporated in the PCCL to either preclude, limit, or mitigate the consequences of severe malfunctions or misoperation. Never-theless, there are plausible sequences which cannot be ruled out and l

l which therefore merit detailed discussion and analysis here.

t 1

I l

l l

- In this category are the following :

(1) loop leak (2) loss of PCCL pump (3) leak in hydrogen cover gas system (4) thimble leak (5) lead bath can leak Each of these scenarios is examined in some detail in the subsec-tions below.

1) Loop leak The most likely site of leakage in the PCCL is at one of the several miAGELOKs fittings, and leakage is likely to be at a slow rate.

Some of the hot (600*F), high pressure (2200 psia) water will flash to 3 steam immediately and the remainder will be vaporized if the water con-tacts hot metal in the tnimble shot bed or the in-core heater lead bath.

t No undue safety consequences will result for several reasons:

(1) If called upon, the thimble's pressure relief valve will relieve pressure at < 100 psia and the back-up rupture disk will burst at - 500 psia. (Failure to relieve could result in ultimate thimble pressure < 500 psia , see Sec-tion 4). The thimble will withstand 500 psia with ade-quate sa fety margin.

(2) Steam will condense (thereby reducing temperature and pressure) on the cold thimble wall .

(3) Leakage will normally be detected by the humidity detec-i tor, located on a gas bleed line from the thimble's gas space.

(4) The reactivity consequences of either losing or adding the maximum amount of water and boron in the in-core thimble are acceptable (see Section 4).

Although loop leakage is no t serious from a reactor safety view-point, continued operation of the experiment is inadvisable. Hence, fol-lowing positive confirmation of significant leakage, an orderly shutdown of the PCCL will be carried out. Electric power to the in-core heater will be cut of f and the temperature allowed to fall to its equilibrium value--which is expec ted to be about 250*F when only gamma heating has to be removed.

At this point, the loop can be depressurized and allowed to boil <

dry. Af ter this, gamma heating is dissipated by radiation to the thimble wall. This mode is maintained until the next regularly scheduled weekly MITR shutdown, at which time the PCCL is removed . Boiling dry is prefer-able to continued, wet operation, to avoid either flooding the thimble

" solid or damaging the loop's canned rotor pump by cavitation.

Small, steady leaks cannot cause any damage of consequence in the upper thimble because the loop is surrounded by a shot bed which will absorb the energy in water or steam jets; similarly, a cast ceramic insulator surrounds the inlet and outlet plena. Leakage into the lead heater bath is more serious since it will result in instantaneous flash-ing into steam. While the energy is absorbed by the lead, it will churn and spatter the bath. To allow for this , the upper foot-long section o f the titanium can housing the lead bath is empty, and a baf fle/ plug is used to cap this container. Note that water injection into molten lead is a well studied and benign process which forms the basis of liquid-metal-type MHD generators now under development for solar and nuclear applications.I I.L Blumenau et al. , "Liquid Metal MHD Power Conversion Systems with l Conventional and Nuclear Heat Sources ," 24th Symposium on Engineering i Aspects of Magnetohydrodynamics (SEAM), June 24-27, 1986.

l l

.e ,

l

2) Loss of pumping power Without circulation, the coolant in the PCCL will overheat, even after the coincident cut-off of the heater power by the loop control system. The loop pressure will rise to 2,500 psia, at which point the relief valve will lift and depressurize the system, following which it will boil dry. Loop removal can again be postponed until the next MITR shutdown.

3) Hydrogen leakage and combusion It is standard practice on all PWR units to maintain a small concentration of dissolved hydrogen (< 50 cc @ STP/kg H 2 O) in the primary coolant to suppress water dissociation. Since the PCCL will, in general,

, be operated under representative PWR conditions, it will be necessary to

, adhere to this practice. To accomplish this, a small-volume, low-pressure, hydrogen flask is used to provide H2 cover gas for the makeup water storage tank. This will maintain 50 cc/kg of H2 in the < 30 kg of makeup water (and in the < 1 kg of water in the loop circuit it serves);

hence, a total water-borne inventory of approximately 1,500 cc (STP) will be present--a virtually negligible amount.

The maximum H2 inservice, however, will occur when the makeup tank is in a near-empty status, in which case its volume (30,000 cc - ft 3) will, at 3 atm, contain some 3 ft 3 (STP) of H2 . Accordingly, we will restrict the hydrogen inventory of the transfer flask to < 10 SCF by design and administrative controls. These controls will include locking the transfer flask in place, restricting access to the parent bottle to authorized loop operators, and maintaining hydrogen inventory records.

Thus, the maximum instaataneous H2 combustion incident, which would occur outside the reactor biological shield, could conceivably involve a

maximum of 10 SCF; the heat of combustion of H2 is 343 Btu /ft 3 Henc e ,

the maximum energy release is = 3,400 Btu .(a kWh)--about the same as would be generated in combustion of one cup of fuel oil: a tolerable amount as long as not suddenly released in a confined space.

Various safety features are provided either to help preclude or to

. mitigate the consequences of even such small-scale events:

. The charging water makeup tank has a hydrogen-oxygen recombiner circuit attached since we need to reduce feedwater 02 to

< 1 ppb. Hence, except during initial fill operations, there will not be enough oxygen to support significant, confined com-bustion. During filling , the gas space will be << 1 f t3, and partially evacuated before the first addition of hydrogen ~.

. The discharge tank is vented through a flame-arrestor to the MITR ventilation system. The rate of water (hence hydrogen) discharge is extremely low: < 30 cc/h H 20, thus < 50 cc H2 Per day.

. A small fan powered by a spark-proof electric motor will be mounted near the hydrogen transfer flask and the makeup / discharge tanks to prevent the local accumulation of a combustible gas mix-ture.

. A helium atmosphere is maintained in the thimble gas space.

In view of the above considerations and precautions, the combustion of hydrogen in the systems associated with the feedwater and H2 supply tanks , and all other connected components out side the reactor biological shield, is not considered a credible tmzard to either personnel or equip-ment. The amount of hydrogen in the water circulating in the loop (i.e.

the normal hydrogen inventory within the biological shield) is equivalent i

to approximately 20 mg of TNT, based on relative heats of combustion, i

1

- -- , v,

. In reviewing possible failure or accident scenarios, we considered another scenario which postulates a water leak in the high pressure PCCL located inside the thimble which is located in the core tank. If this leak goes undetected, water from the makeup tank and eventually hydrogen gas might be pumped into the thimble. If at the same time oxygen were-available f rom s ome sou rc e , e .g . , if the thimble had not been purged of air and filled with helium, as would normally be done, a combustible gas mixture could result .

The above scenario postulates simultaneous occurrence of a loop leak, failure of several sensors and experimenter inattentiveness, as well as the availability of oxygen in a space normally purged with heli-um. This scenario is very unlikely. However, since the energy released during an optimal rapid burn of a significant fraction of the hydrogen inventory of the supply tank could be significant, it will be necessary to show that it does not compromise thimble integrity, or that by suit-able design and instrumentation of the PCL system this scenario can be made inc red ible .

Our approach to designing the PCCL system so that a significant fraction of the inventory from the hydrogen flask cannot enter the thimble is as follows :

a. PCCL experimental thimbles will not be inserted into the core tank until a preconditioning run has taken place outside the core tank. This will insure that air has been properly purged from the thimble because shot bed conductivity and loop temperatures will be sensitive to relatively small amounts of air in the helium cover gas.

b.- A high quality instrumented level indicator will be installed in the makeup or charging tank. This will assure _ that the water.

level in the tank never approaches the bottom of the tank. It is only af ter this tank has been emptied of water that significant amounts of hydrogen could be pumped into the thimble by the charging pump.

c. The amount of water in the charging tank, even at the end of an experiment, will be more than the in-core thimble volume. This assures that a loop leak can conceivably only result in filling the thimble with water and not with gaseous hydrogen.

d. A float operated check valve can be installed in the bottom of it the charging tank to prevent pumping hydrogen when the tank is em pt y . This will be added if the other measures cited here are not deemed adequate.

Note that these measures are independent of the instrumentation and procedures which would normally detect a loop leak, which must precede or accompany hydrogen ingress to the thimble. The humidity detector and the loop temperature would respond rapidly to loop leakage.

Finally, a burst disk will be installed in the top of the in-core thimble to guard against overpressurization from whatever source.

4) Thimble leak The only potentially troublesome leakage event would involve a sudden large rupture in the thimble below the water level in the MITR pool. Since the helium gas in the thimble will be maintained at 30 psia, small leaks will involve egress of helium rather than ingress of water.

A large rupture is highly unlikely--no credible mechanistic sequence has been identified leading to this event. If it did occur , however , the

=

p .. .

consequent loss of helium would permit flooding by water. Initially the hot metal in the shot bed and lead bath would cause some of the water to flash to steam. Eventually cooldown would progress to the point where flooding of the thimble would persist. As discussed in Section 4, flooding of the fill space between the heater bath and the thimble does not add a consequential amount of reactivity. Similarly, the volume between the thimble and dummy fuel element is insufficient to permit the addition of unacceptable reactivity upon reflooding af ter helium or steam is expelled into this gap.

Upon detection of significant thimble leakage, PCCL heater power will be cut of f and the loop operated in a cooled-down mode, but at its normal pressure, until its removal at the next regular MITR shutdown.

Another possible consequence of thimble leakage is related to the exposure of the primary coolant water to the loop internals, with subse-

< quent re-release of some of this water to the MITR coolant. Since inter-nal and external pressures are equalized, no large driving force exists for this process. 'Ihe solubility product of Cu(OH)2 is approximately 10-20,5; thus in pure water only about 20 pg Cu/ liter would be expected to solubilize . A completely flooded PCCL shot bed would contain less than ten liters of water. Hence considerably less than one milligram of copper should be introduced into the MITR coolant if thimble in-leakage were to drain back out--during removal of the damaged loop, for example .

(The extremely small solubility of copper in water is attested to by its extensive use as a durable anti-fouling sheathing for wooden ships in the 19th century.) Diffusive release would be many orders of magnitude smaller. Note that this small amount of copper is the only contaminant of concern in a back leakage incident. No long-lived radionuclide

production would result, corrosion of core tank structures would not be significantly accelerated, and the MITR ion exchange purification system would be capable of removing the small quantities of material involved.

Of greater concern would be escape of the copper shot itself. This is guarded against by the monolithic structure of the loop thimble, its conservative design and low operating pressure. The port used to drain the shot f rom the thimble is securely and redundantly bolted in place, and the outer seal plate is backed up by an internal plug which is threaded into the drain hole, and capable by itself of retaining the shot bed. The diameter of the shot has been selected to be too large to pass down between MITR-II fuel plates, and the high settling velocity of the

- shot (=2 ft/s) makes extensive entrainment unlikely. The shot could, however, pass through the circulation valves and get under the core,

, resulting in blocked fuel channels. Care will be taken in all shot han-dling procedures to insure that none escapes to become an uncontrolled

, source of in-containment debris.

5) Lead bath can leak Titanium can failure will permit contact of molten lead with the cold aluminum thimble wall, which will cause virtually instantaneous freezing of the lead. Laboratory tests are planned which are anticipated to show that the freezing process seals the leak and that the resulting thermal shock will not cause the aluminum to fail. However, the thermal short circuit will create a hot spot on the thimble wall, and local boll-ing cannot be ruled out at the present time. Again, experiments are planned to explore the consequences of such an event. Note that if this i

boiling occurred, it would not be on the surface of a fuel plate.

l i

. A91arge can failure would be indicated by a large decrease in loop average temperature -(with the heater on), and would be cause for loop shutdown. If operation continued, for a short time, and maximum loop power (20 kW) were shunted through the thimble wall, a lower thimble 2

average heat flux of approximately 68,000 Btu /h-ft would result. This is a quite tolerable result considering that heat transfer coefficients in excess of 104 Btu /h-ft2_.F are to be anticipated.

Small leaks will be difficult to detect. If the leak eventually has significant effect on the heat losses in the lead bath, this would prob-ably be detected by thermocouples or heater power level and the experiment would be shut down if necessary. Another indication might be an increase in reactivity noise if local boiling were to occur.

It should be noted that the lead bath can will be subject to vacuum leak testing and pre-operational hot testing. It is unlikely that leaks 4

will develop in a can which has passed this testing, since no significant stresses are applied, and the liquid lead and helium environments are benign. A thermocouple may be installed to monitor the temperature of the thimble bottom and provide an indication of small lead leaks.

The overall conclusion of the preceding revit;w is that there are no credible mechanisms by which the PCCL can adversely affect MITR safety or otherwise create a situation hazardous to operators or experimenters.

The most serious outcome projected is the loss of valuable experimental data, since large changes in the internal chemical environment of the PCCL would probably invalidate any subsequent analysis of corrosion prod-uct characteristics.

Some concluding remarks are also in order on the inverse question as to whether there are any normal transient or accident situations of the i

MITR'itself which could induce a hazardous response by the PCCL. In

.brief, none have been id ent i fied . The MITR interacts significantly with the PCCL only by 1) provision of approximately 10% of PCCL energy input via gamma heating, and 2) serving as a heat sink for the PCCL. Since electrical power to the PCCL in-core heat can be cut off in an emergency, thereby greatly reducing the need for's heat sink, a loss of MITR capa-bilities suf ficient to severely inconvenience the PCCL would unquestiona-bly be a far greater threat to the MITR it self.

6 I

7. WASTE HANDLING AND DISPOSAL Two major types of radioactive waste will be produced by the PCCL experiments: loop coolant water contaminated by activated corrosion pro-ducts, and activated in core components. As with other aspects of the loop operation, no new issues are raised in this regard.

Approximately 22 liters of water will accumulate in the discharge tank during a one month run. The activity level of this water is expected to be very low, since most of the corrosion products in the effluent (a few precent of those generated in the loop itself) should deposit on the bleed capillary surfaces before reaching the tank. The activity levels will be sufficieltly low to permit disposal of the dis-charge water to the drain after counting. Although particulates are expected to deposit on the capillary surfaces, tests will be run on ini-tial discharges to determine if filtration is necessary. The approxi-mately 0.5 liters of water present in the loop at the end of a run will be collected, counted and disposed to the drain after verification that it is within acceptable limits.

In core loop components will be stored for decay and eventually 210 shipped as solid waste. The P0 produced in the lead bath has a half-life of 138 days and can therefore be decayed to insignificant levels before disposal.

Note that the PCCL is designed for 1:1 simulation of an actual PWR, hence coolant activity levels will be comparable. Table 7.1 gives repre-sentative radionuclide concentrations expected after one month of irradi-ation. Using values for a typical PWR given by Benedict et al.,1 I

M. Benedict, T.H. Pigford and H.W. Levi, Nuclear Chemical Engineering, Second Edition, McGraw-Hill, Mcw York (1981) p. 396.

7 scaling by water volume and taking into account the use of II B and Li, tritium concentrations are expec ted to be = 5 pCi/1.

TABLE 7.1: REPRESENTATIVE ACTIVITY VALUES ESTIMATED FOR A ONE-MONTH. PCCL IRRADIATION (from typical PWR values)

"Steam Generator" toop Water 2

Activity (pCi/cm ). Activity (uCi/ liter) 58 3.0 no data Co 60 Co .2 0.02 597 ,

51 Cr <<60Co <<60Co 54 Mn 4

L r

b 4

i

, ,-. s - -n---,- , , - - . _ _ . . . - - , . , - , , , , , , , _ . , , . , - - . , . . , . . - - - , _ , , . ,,

.,,,,,,-,,n-nr, ,g, , , - --,- - -

8. FUTURE WORK

~ This section summarizes the requirements which remain to be ful-filled before loop installation in the reactor. It also gives a complete tabulation of the safety-related experiments and pre-operational tests which will be carried out , see Table 8.1. Note that the results of these experiments and tests will be reported by PCCL personnel, certified by senior PCCL personnel (Professors M. J. Driscoli and O. K. Harling) and incorporated into the QA file.

The following items are required before experiment irradiation:

1) Standard operating procedures

2) Abnormal operating procedures

3) Calibration procedures and test schedule for safety-related instrumentation

4) An Irradiation Request Form (Part I)

5) A final signed safety review by the MITR staff

6) Pre-operational test procedures and resulta

7) Material certifications (see Section 2.f on QA procedures)

Development of the required procedures and checklists for operating the loop will be done in conjunc tion with MITR sta f f.

-5 0- ,

TABLE 8.11 SAFETY-RELATEJ EXPERIMENTS AND PRE-OPERATIONAL TESTS TO BE PERFORMED BEFORE PCCL IRRADIATION l

a) Safety-related experiments Shot bed conductivity measurements Radiative heat transfer measurement

• Shot bed average temperature determination Liquid lead compatibility experiment Thimble cooling / coolant AT measurement Lead bath can leak / break simulations b) Proof testing and pre-operational tests

.s Circulation pump low T/P performance testing Loop pressure proof test (3000 psi) i Thimble vacuum / pressure proof test l

Pressure relief valve proof test l

Out-of-core conditioning, "shakedown" run Grid plate clear 2nce test

• Low reactor power reactivity tests (empty dummy fuel element, unfilled loop, cold loop, hot loop)

Loop disassembly rehearsal

l s

9.

SUMMARY

AND CONCLUSIONS a) Summary This safety evaluation report pertains to the operation and research use of a small pressurized water loop in the MITR-II core, core-tank and containment building. The preceding sections and the appendices which follow document the design features, as they relate to safety, of an in pile facility which is intended to closely simulate the primary system of a PWR. Figure 9.1 is a schematic illustration and sum-mary of the safety features of the loop.

Although designed to operate at PWR temperatures and pressures, the small size, the design of the facility and the control instrumentation limit the potential hazards. In this SER we have addressed issues asso-ciated with 1) reactivity changes, 2) thermal-hydraulic effecte,

3) chemical ef fects, 4) radiolytic decomposition, 5) experiment scrams,

6) prototype testing and proof testing, 7) radioactive releases, waste handling and disposal, as well as other concerns such as operational and experimental procedures, radiation levels and ALARA considerations. Our conclusions in these areas are summarized below:

1) Reactivity changes The loop is designed so that the mayimum reactivity changes which are conceivable are less than the 0.20% ak/k allowed for an experiment which would be movable during reactor operation.

2) Thermal-hydraulic ef fects The most serious concern overall is assurance of the heat input to the loop: 10-15 kW of electric heater power and 2-4 kW of gamma heat. Radiation to the thimble wall is suf-ficient to remove the latter in a purely passive mode of

FLAME ARRESTOR RUPTURE O REDUNDANT .

DISK l' THIMBLE MOISTURE DETECTOR CORE TANK I At S BOUNDARY q

\

" HITR VENTILATION k  : LIMITED CLEARANCE g PREVENTS EJECTION SUPPORT BEARS UP TO -

100% OF REDUNDANT LOOP RELIEF VALVES V:., . < . .. e: / p

~ ~

n p.:k.':.'.

. .'.-;'. ::.k NATURAL CONDUCTION THROUGH .

CONVECTION COPPER SHQY TO COOLANT I POOL __

DISCHARC E TANK

= REDUNDANT THERMDCOUPLES AUTOMATIC 7 HEATER SHUT-OFF FLAME ARRESTOR

/

LOCK-DOVN ,

TABS BELOV m - ]

CORE GRID  :: FUSIBLE LINK HEATER PLATE SHUT-OFF TCHELT)=1200 F p H

/ MITR VENTILATION

\ RADIANT REJECTION OF GAMMA HEATING EXPLUSION -

V PROOF FAN n

SMALL FLD0DABLE VOLUMES TO LIMIT HYDROGEN / HYDROC EN REACTIVITY ADDITIONS OXYGEN I RECOMBINER v HARGING (j TANK h

HINIMUM VOLUME &

PRES Y OGEN FIGURE 9,1

SUMMARY

OF PCCL SAFETY FEATURES

l_....

L operation; and redundant means for interruption of electric power are provided--an active overtemperature control system based upon thermocouples which monitor heater bath tempera-ture, and a passive fusible link which melts to sever the electrical connection to the resistance heaters in the lead bath.

3) Chemical effects A moderate amount of hydrogen gas (< 10 sc f) will be used inside the reactor containment. The systems outlined in the body of the report assure safe use of hydrogen through two basic approaches avoiding the contact of hydrogen with oxygen or high dilution levels to prevent accumulation of an explosive mixture in the event of hydrogen leakage.

A scenario which involves the potential for hydrogen entry into the in-pile thimble has been analyzed and is discussed in 6.b.c. The operational, instrumental and design features which we have incorporated into the PCCL are deemed adequate to assure that hydrogen and oxyFen can never be present in siga.i fic an t amounts in the thimble when it is in-core.

4) Radiolytic decomposition No significant unresolved issues have been identified. Basi-cally the PCCL operates in the mode and with materials identi-cal to a PWR.

f

)

5) Experiment scrams This is addressed in the body of this report, and the protec-tion of the reactor systems as well as the PCCL can be assured.

6) Prototype testing and proof testing A considerable amount of testing of this type is planned to assure the safe conduct of PCCL experiments. Section 8 out-lines the testing which will be performed in this category prior to routine in pile operation.

7) Radioactive releases, waste handling and disposal, radiation levels, ALARA considerations and related operational proce-dures These are discussed in various sections of the SER. For the most part the PCCL breaks no new ground in this area. Pre-vious experiments reviewed and authorized for operation in the MITR-I and MITR-II, such as the Fatigue Cracking Experiment, have involved comparable or higher levels of activities, radi- ,

ation fields and radioactive material handling procedures.

b) Conclusion The overall conclusion of this SER is that operation and experi-mentation with the PCCL in pile loop, having the design features outlined above, can be made to fully satisfy the MITR-II Technical Specifica-tions. Furthermore, we feel that no significant health or safety hazards will result from these activities.

i

Y APPENDICES f

Appendix 1: Safety Related Experiments and Calculations a) Radiative dissipation of gamma and neutron heating of loop components.

b) Reactivity worth of B-10.

c) Thimble stresses.

Appendix 2: Extracts from the MITR-II Technical Specifications

+

APPENDIX 1.a: RADIATIVE DISSIPATION OF GAMMA AND NEUTRON HEATING OF LOOP COMPONENTS As discussed above, one of the priuary safety features of the loop is the passive rejection of nuclear heating of itecore components at a temperature below the zirconium-steam reactio n t- mperature of 2100'F.

An experiment to verify calculations of the radiative heat transfer rate between the lead-filled titanium tube and the cooled aluminum thimble has been completed. A simulation at approximately full scale with respect to radial dimensions and half the length of the actual in-core section was used. A one kilowatt electric heater was used at varying power levels, and the temperature difference from the titanium tube wati to the aluminum tube (which was maintained at constant temperature by a water bath) was measured.

The experimental data was fit to the expression: i c(T2 -Tt ")

1 1 1 4 _t c1A1 A2 C2 where: Q = total heat transfer rate o = Stefan-Boltzmann constant c = emissivity A = surface area 1 - Al thimble, 2 - Ti can

O The denominator of this expression r,an be estimated from the experimen-tal data, and should be approxiercely constant since the change of emis-sivity with temperature is small over the range encountered. The cal-culated values, however, varied by a factor of five, indicating that convection was contributing significantly to the heat ;ransfer. An approximate value for the denominator at T2 = 2100*F was nevertheless obtained by extrapolation, and the heat rejection at that temperature was estimated to be 8.5 kW.

At full power, the cors average Y-heating rate is $1 w/g. Wood 2

and Brown have shown that neutron heating adjacent to a thermal neutron fission spectrum is one tenth the gamma heating. Thus, usJng a t)tal nuclear heating rate of 1.1 W/g, and with 6.5 kg total mass in the tita-nium tube and contents, the total nuclear heating is $7.2 kW. The anal-ysis above indicates that this will result in temperatures below th/

zirconium-steam reaction temperature, although the margin is not large.

When the loop is constructed, tests of the actual passi+1e heat rejection will be performed. If necessary, the maximum temperature attained in passive cooling mode can be reduced by reducing the mass of lead in the conductive bath and/or by treating the titanium and aluminum surfaces to increase their emissisity. The experimental results described above indicate that adequate safety margins should be attainable.

I P.J. Wood and M.J. Driscoll, "Assessment of Thorium Blankets for Fsst Breeder Reactors," MITNE-148, July 1973.

2 G.J. Brown and M. J. Driscoll, "Evaluation of High Performance LMFBR Blanket Configurations," MITNE-150, May 1974.

e, .

f

(

. APPENDIX 1.b: ESTIMATION OF BORON WORTH ',

The (unknown) worth of' Boron-10 can be estimated f rom the (known) worth of Uranium-235. Perturbation theory yields the simple ratio:

(Ao/g) B-10 ,

I a, B-10 , 235- l A-1)

(ap/g) U-235 (n-1) o,, U-235 10 For cross-sections, the Maxwellian-averaged thermal values can be used, since over 80% of the neutron absorption in both B-10 and U-235 are in l the thermal region. Cross-section valuss are a follows:

a n -

B-10 3,400 0-1 U-235 588 2.07 I s

& Finally, we have for the MITR-II core: ,

bE < 10 milli S/gl (worst-case value - A-ring, 445 g element)80-235 ,

Thus, Eq. (A.1) gives:

• 8 ..

bE g

= 1270 milli S/g

. .B-10 i -- -

and if S U-235 - 0.00786 If. s ak/k' , ]p( = 9.98x10-3 = 0.998%

8 8 8

. _B-10 . .B-10 . .B-10 F

I Personal communication, J. Bernard, October 6, 1986, and his memorandum s to 0. Harling dated October 7, 1986.

ll i

1 e1 . . , , . - - , . ,,. , an. . . . . . , , , , ,,,e - -,, ...--,~..-, - ,, ,,-,,, , ,

e n= , e

. ,,o o.

.4 APPENDIX 1.c: THIMBLE STRESSES The thimble is designed to withstand the stresses expected under both normal and accident conditions. The cases of interest which have been identified are:

1) internal pressurization - normal helium pressure

- shot bed hydrostatic pressure

- loop leak pressurization accident

- hydrogen deflagration accident

2) external pressure on evacuated loop

3) thermal stresses The thimble consists of three sections: an approximately elliptical section 2.50 in. x 1.40 in. x 0.125 in, wall, 3 ft. long; a cylindrical section 4.0 in. OD x 0.250 in call (possibly with 0.125 in. deep grooves for added heat transfer), 8 ft. long, and a cylindrical section 8.0 in. OD x 0.250 in, wall. All these components are constructed from

. 6061-T6 aluminum tubing with minimum yield strength of 40,000 psi.

1 For the case of internal pressurization in a long, thin-walled cylinder, the hoop stress is given by:

o Ef l

vhere P = pressure r = tube radius t

• tube thickness.

.a For the cylindrical sections, the pressure at which the hoop stress is one-half the yield stress can readily be derived: P = 1250 psi (equal for both sections if a minimum thickness of 0.125 in. is assumed for the 4 in. section, as would be the case for the grooved version). For the elliptical section, maximum stress due to internal pressure occurs at the minimum radius point where hoop stress ist o=q where b = semi major axis, and the bending moment is given by:I M= b2 _

where Ix = moment of inertia of a quadrant of the ellipse about the x-axis 1 = moment of inertia of a quadrant of the ellipse about the 7

y-axis S = are length of a quadrant of the ellipse Using these relationships, an internal pressure of =850 psi can be shown to produce a maximum stress of one-half the yield stress. It is evident from this analysis that the normal helium operating pressure of 20-30 psig can easily be withstood, and the =20 psi additional pressure due to the weight of the shot bed is also inconsequential. (Note that a simple I Baumeister, Avallone and Baumeister, eds. , Marks' Standard Handbook for Mechanical Engineers (8th Ed.), McGraw-Hill, New York (1978) pp. 5-51.

4- .

analysis indicates that shot-bed ratchetting, which could exert signifi-cant forces on the thimble, is unlikely to occur. Dimensional monitor-ing of the thimble before and after its out-of pile runs will be used to verify the absence of ratchetting forces.) Tne loop leak pressurization accident was shown above to result in a maximum pressure of less than 400 psi if the two thimble pressure relief valves fail to open. Pres-sures in a hydrogen deflagration accident are much more difficult to as-sess, but this accident has been shown to be extremely unlikely, and a burst disk set at 500 psi or lower will be provided to gaurd against thimble failure in this scenario. Considerable margin to yield is pro-vided, and the ductility of the aluminum (min. = 10% elongation) gives further assurance that thimble leakage is a remote p asibility even in extreme pressurization accidents.

The stresses produced on the thimble by evacuation (plus submer-gence to about 12 ft. in water) are negligible, and the ability of the thimble to withstand this loading will be amply demonstrated during the conditioning run, which will be made with the thimble at vacuum. The worst case for thermal stresses will occur on the in-core thimble sec-tion during a loss-of-coolant accideat. Us'.ng a conservative value of 10 kW total heat transfer to this 150 in.2 area gives AT = 3*F;, and a corresponding thermal stress of "300 psi, which is again negligible.

/

t

's e

APPENDIX 2 EXTRACTS FROM THE MITR-II TECHNICAL SPECIFICATIONS l

l l

l l

^ '

- - 5.3 Primary Coolant System . ,

Applicability This specification applies.to the design of the primary coolant .

system.. ,

3 Objective -

To assure compatibility of the primary coolant system with the safety analysis. ,

Specification T e, reactor coolant system shall consist of a. reactor vessel, a single cooling loop containing three heat exchangers, and appropriate pumps and valves. All materials, including those of the reactor vessel, in contact with primary coolant (H 2O), shall be aluminum alloys or stainless steel, except small non-corrosive components such as gaskets, 4

filters and valve diaphragas. The reactor vessel shall be designed in It shall g

accordance with the AS12 Code for Unfired Pressure Vessels.

be designed for a working pressure of 24 psig and 150*F. Heat exchangers shall' be designed for 75 paig and a temperature of 150'F. The connecting

- piping shall be designed to withstand a 60 psig hydro test. .

Basis

~

l The reactor coolant system has been described and analyzed in the i Safety Analysis Report as a single loop system containing two heat -

exchangers. Additiona'l analy. sis based on the use of three heat exchangers, has been described in the flRC staff's Safety Evaluation of Amendmant flo,14 to these Technical Specifications. 11aterials of construction, being l

primarily stainless steel, are chemicalTy compatiMe with the H 2O coolant.

The stainless steel pumps are heavy-walled members in areas of low stress and should not be susceptible to chemical attack or stress corrosion failures. N i

failure of the gaskets and velve be'le.i,, although undesirable, would Amendment No. 14 g

• q

• _4 i

not result in catastrophic failure of the primary system; hence, strict 1 .

~"-

material limitations are not required for technical specifications. The

• design, temperature, and pressure of the reactor vescel and other primary system components provide adequate margins over operating temperatures and pressures, and it is believed prudent to retain these margins in order to further reduce the probability of a primary system failure. The reactor vessel .was designed to Section VIII,1968 edition, of the ASME Code for Unfired Pressure Vessels. Subsequent design changes should be made in accordance with the most recent edition of this code.

Since the safety analysis is based on the reactor coolant system as presently designed and with th'c present margins, it is considered l -

j necessary to retain this design and these margins or to redo the analysis.

k-t i

r i

• i, -

e I

i i t

i

.I l

l 5-6 l

l:

l I . m

4 W

6. EXPERIMENTS 6.1 General Exneriment Criteria Applicability This specification applies to experiments using the reactor.

Objective To assure that experiments in the reactor do not affect the safety I

of the reactor.

Specifications All experiments within the reactor sha'11 conform to the following

. conditions.

. 1. Reactivity Effects The reactivity worth of experiments shall not exceed the values indicated in the following table:

Single Experiment Worth Total Worth Movabic O.2% AK/K 0.5% aK/K Non-secured 0.5% AK/K 1.0% AK/K Total of the above ,

1.5% AK/K Secured 1.8% AK/K

2. Thermal-llydraulic Effects

a. All experimental capsules shall be designed against failure from internal and external heating at a reactor power level or procons variabic corresponding to the Limiting Safety System Setting associated with that power level or process variabic.

l 6-1 l

t , ,

v

, b. The outside surface temperature of a submerged experiment or capsule shall not cause nucleate boiling of the reactor coolant durir.g operation of the reactor.

c. The insertion of an experiment into the core should not cause a coolant flow redistribution that could negate the safety consideration implicit in the Safety Limits.

3. Chemient Effects

a. Metastable or other materials that could react to create a rapid pressure rise shall be encapsulated. The capsule shall be prototype tested under experimental conditions to demonstrate that it can contain without failure an energy release equivalent to at least twice the material to be irradiated or at least twice the pressure that could be s

expected from any reaction of these materials. These tests must also include effcces of any fragments which may be generated. If a change in experimental conditions could result in a greater potential for failure than design experimental conditions, the capsule shall also be tested under these changed conditions. In addition, the quantity of material should be limited such that if the maximum l

calculated energy release should occur, significant damage to the reactor core will not result, assuming the material is not encapsulated.

1

b. No explosive materials (defined to include all materials that would concticute Class A Class B .'ad Class C cxplo-siven as described in Title 49, Parts 172 and 173 of the 6-2 o

d

s. Code of Federal Regulations) shall be placed in the reactor core or within the primary biological shield, which, if completely detonated, could cause any rearrangement or damage to the reactor core. Proposed quantities of explo-sive materials greater than the equivalent of 25 milligrams of TNT shall require a documented safety analysis and approval by the MIT Reactor Safeguards Committee. Capsule design for explosive material shall be prototype tested to demonstrate that it can contain at least twice the pressure produced inside the capsule as a result of detonation of the material or the pressure produced by the de*onation of twice the amount of material.

c. Corrosive materials that could af fect or react with another material present in the reactor system will be doubly encap-sulated. If the materi'al can adversely affect the reactor core or any of its component parts or auxiliary systems or the building containment to cause loss of function of the affected component or system, means shall be provided to monitor the integrity of the material container.

4. Radiolytic Deconnosition

a. Compounds subject to radiolytic decomposition shall be irradiated in containers which can withstand the maximum gas pressure produced as a result of the decomposition under irradiation including the effect of any temperaturc

, 6-3

,e- .

v rise. This pressure will be determined by previous expe-rience or by costing' as described in Specification 6.1, subsection 6.

b. Consideration shall also be given to any pressure buildup resulting from the decomposition of the sample container, such as might occur with a polyethylene vial.

c. . Compounds rubject to radiolytic decomposition may be i

irradiated in a capsule that is vented, provided that the vented release is less than 1.0% of the limits of 10 CFR 20 at any point of possible exposure.

' ~

5. Experimont Scrans Provisions have been nade in the Reactor Safety System for the addition of experiment scrams. These scracs may be added for the protection

-. of the experimental equipment and/or reactor coeponents in the event of some malfunction. If malfunction of the experiment can adversely affect the reactor core or any of its component parts or auxiliary systems or the building containment to cause loss of function of the affected component or system, the experiment scrams shall be redundant so as to satisfy the singic mode f ailure protection discussed in Section 7 of the SAR.

6. Prototype Testing l

Materials whose properties (coeposition, heating, radiolytic l decomposition, etc.) are unecrtain must be prototype tested. These tests will be designed to give a stepwise approach to final operating conditiens.

The tents may either be stepuise time or flux irradiations with proper instrumentation to determine temperature, pressure and radioactivity for cach step an required.

6-4

o ,.

o

7. Radioactive Releases

- a. Experiments shall be designed so that operation or mal-function is not predicted to result in exposures or releases of radioactivity in excess of the. limits of 10 CFR 20 to either onsite or offsite personnel.

b. The total radioactive materials inventory of an experiment

.or credibly coupled experiments shall be limited such that i

the dose in unrestricted areas resulting from release of this inventory at its calculated maximum value shall not exceed that of the Design Basis Accident (Section 5.3.1 of the SAR).

Bases .

Accidents resulting from the step insertion of reactivity have been discussed in the SAR. It was determined that following a step increase of 1.8% AK/K, fuel plcte temperatures vould be below the clad molting ten-perature and significant core damage would not result. The 0.2% AK/K limit for movable experir.ents corresponds to a 20-second period, one which can be easily controlled by the reactor operator with little effect on reactor power. The limiting value for a single non-secured experiment, 0.5% 6K/K is set conservatively below the prompt critical value for reactivity insertion and below the minimum shutdown margin. The sum of the magnitudes of the static reactivity worths of all non-secured experi-cents, 1.6? SK/K, does not exceed the minimum shutdown margin. The total worth of all oevabic and non-secured experiments will not reduce the mini-num shutdown margin as the shutdown margin is determined with all movabic experiments in the most positive reactive state.

6-5

a. 4 v

Specifications 2, 3, 4, 5 and 6 are intended to minimize the prob-ability of experiment failures. Experiment capsules should be designed to withstand expected temperatures, pressures, chemical and radio-chemical effcces. The requirement for testing containers at twice the pressure or with twice the amount of explosive or metastable material to be irradiated provides a f actor of 2 safety margin as allowance for experimental uncertainties. Table 6.1-1 gives a summary of the require-ments for specimen irradiations for ease of review and classification of the specifications.

The radiological consequences of experiment malfunctions must be considered as' stated in Specification 7. Consistent with the Commission's

- regulations, predicted onsite personnel exposures or offsite concentrations resulting from these malfunctions must not be in excess of those permitted b'y 10 CFR Part 20. ,

6-6 8

• I I Table 6.1-1 .

Sur: ary of Requirements for Specimen Irradiations b

.-4

. SS Requirement 24 $

ve

$ 8 c u e v4.

en ,A4 o.

e 24 a a .

3a. 8-u 3,

-a 8 u e x o c: c Consideration of reactivity effects, induced activity, heating and temperature distribution X X X X X Estimation of pressure buildup X X X Single encapsulation , X, ,X X i Deuble encapsulation X w

Container pressure test X X Other d a, ~d b c

a. Amounts above the equivalent of 25 mg TNT require safety analysis and approval of MITESC.

b. If corrosion can cause loss of functions of the reactor core or any of its component parts or auxiliary systens or the building containment, integrity of container must be monitored during irradiation. ,

c. Container may be vented if' release is less than 1.0% of 10 CFR 20.

d. Amounts limited such that reaction will not damage reactor core.

1

- - - , - - -