ML20042F029

Fuel Follower Control Rod Safety Analysis for Armed Forces Radiobiology Research Inst Triga Reactor Facility.

ML20042F029
Person / Time
Site:	Armed Forces Radiobiology Research Institute
Issue date:	04/30/1990
From:	Forsbacka M, Maria Moore ARMED FORCES RADIOBIOLOGICAL RESEARCH INSTITUTE
To:
Shared Package
ML20042F028	List: ML20042F027 ML20042F029 ML20042F031
References
NUDOCS 9005070125
Download: ML20042F029 (15)

Text

_ ~. _ _ . _ _ . . .

)

ATTACHMr.NT 2 '

i FUEL FOLLOWER CONTROL ROD SAFETY ANALY SIS FOR THE AFRRI TRIGA REACTOR FACILITY t l

?

i J

let Lt Matt Forsbacka, USAF Reactor Executive Omcer Reviewed and approved:

l MARK MOORE Reactor Facility Director .

1 1

900507012D 900430 PDR ADOCK 05000170 P PDC 1

. . . . .. - - . . - ...1

I

]

INTRODUCTION Operational requirements of the Armed Forces Radiobiology Research Institute ]

(AFRRI) TRIGA reactor facility necessitate the implementation of fuel follower '

control rods Fuel follower control rods are like the standard TRIGA l contrcl rods a(FFCR's).

s described in section 4.10.1 of the AFRRI TRIGA Safety Analysis Report (SAR) except that they have a fuel filled follower rather than an air or  ;

aluminum follower. The primary purpose of the FFCR's is to oEset the long-term effects of fuel burnup.  !

The Code of Federal modifications Regulations of a portion (CFR)d(Title (s) of a license facility, 10, Part 50.59) as described in requires the facilitythat SAR, '

be documented with a written safety analysis. Such documentation ensures that all safety issues associated with the change are reviewed. Based on the analysis in '

this report, the authors have determined that implementing FFCR's will allow the standard control rods to_ function in their intended purpose and will restore core reactivity economically. FFCR's have been implemented in approximately a dozen TRIGA reactors and have been used for over 20 years without reported failure, t The proposed modifications require minor administrative changes to the technical i and the SAR. These specifications of theare administrative changes current specifiedreactor license (R-84)B of this report.

in Appendix "

This report has been submitted to the AFRRI Radiation Facility Safety Committee to ensure that all safety questions have been reviewed before submission to the -

USNRC, as required under 10 CFR. 50.59.

l GENERAL DESCRIPTION OF FUEL FOLLOWER CONTROL RODS The current AFRRI TRIGA standard control rods were installed in 1964. The I standard approximatelycontrol rod consists 1.25 in in diameter and 31 inof a sealed long. The upperaluminum 15.25 in o tube (0.065; tube contains a compacted borated graphite rod or other boron compounds), which functionsneutron as 4

a(Babsorber C with 25-percent or poison. The free boron lower end of the tube contains a 15.25-in long and 1.125-in diameter solid aluminum rod called the aluminum follower. The follower functions as a mechanical guide for the control rod as it in withdrawn from or inserted to the.

reactor core.

The proposed FFCR's difier from the current standard control rods in the '

following respects:

• The aluminum cladding is replaced by smooth stainless-steel (SS304) cladding

, with a wall thickness of 0.020 in. The inner and outer diameters are 1.085 in and 1.125 in, respectively.

• The length of the control rod is increased to 37.85 in; the absorber and fuel follower section are both 15 in long. '

• The outer diameter of the absorber section and the fuel follower are both 1.085 in.

The absorber 'or poison material of the proposed FFCR's is, however, identical to the standard control rods presently installed.

1

The fuel contained in the FFCR consists of a fuel-moderator element in which i zirconium hydride is homogeneously mixed with partially enriched uranium. The FFCR fuel element contains 12 pgcent uranium by weight and has a nominal enrichment of 20 pergtU-this in theis 83%U isotope.

about 31.4 grams of of the 'pe.U FFCR fuelofelement loading contains a standard AFRRI TRIGA fuel element.

I SAFETY- ANALYSIS OF FFCR IMPLEMENTATION The two principal safety issues that must be addressed are the maximum excess <

reactivity limit of 85.00 set by Technical Specification 3.1.3.(a) and the maximum R fuel temperature of 600*C set by Technical Specification 2.2, With regard to the maximum excess reactivity limit, Reactor Operat!ng Procedure VII, Reactor Core

~

Loading and Unkading (Appendix D), ensures that the $5,00 limit on excess .

reactivity is not breached.

A thermalhydraulic analysis of the FFCR fuel element to determine the maximum i fuel temperature uses the following model:

• The neutron mean free path for neutrons of all energies is smaller than the diameter of the TRIGA fuel rods, so the reactor must be treated as a heterogenous reactor. Thus, the active volume of the core is taken to be the volume of fuel contained within the reactor core.

• The ratioisof1.21.

uranium power in a fuel element witn 12- wt-% uranium versus 8.5 -wt-%,

This is determined by General Atomico design calculations.

• The reactor is operating at a steady-state power level of 1.0 MW and the heat flux across the fuel element is- described by Fourier's law of thermal conduction:8 q"(r) = -kVT(r) (1)  !

t where q"(r) = heat flux at position r  !

k = thermal conductivity

• T(r) = tempe ature at position r.

For steady-state heat transport, the heat production rate and the rate of energy loss due to heat transport are equal. This can be generally expressed y 88 q"'(r) =- Vaq"(r) (2)  ;

where q'"(r) = vclumetric heat rate (heat production rate) at position r.

Substituting equation (1) into equation (2) yields the time-independent equation of thermal conduction:

q"'(r) = -VakVT(r) (3) .

Equation 3) is, thus, the second-order ordinary differential equation that must be s(olved to determine the maximum temperature attained in the fu portion of the FFCR.

2

i 1-l

Using this model to determine the maximum fuel' temperature divides the analysis 1 i into two separate tasks: determining the power density in the FFCR in a D-rmg 1 i- grid position and solving equation (3) for the given power density. ,

\

l Power Density in FFCR Fuel Element J The anticipated fuel loading for the AFRRI TRIGA reactor core with FFCR's

~

installed will consist of 77 standard TRIGA fuel elements and the three FFCR fuel elements. - Presuming that the control rods are fully withdrawn to achieve a power level of 1.0 MW, the total active fuel volume .will be 30,597.9 cm8 . Thus,- the average power density at'1.0 MW will be 32.7 :W/cm3 The maximum fuel temperature .ls the important parameter, so only the radial_

variation of tht, core centerline power density is considered. To determine the maximum power density in the D-ring location of the IFCR fuel element, the .

. following calculations are made: t i

For the AFRRI TRIGA, the radial and axial, peak.to-average power ration are 1.55 and 1.30, respecth>ely.a Thus the maximum power density (heat rate) will be l q"', = (1.55)(1.30)q"', (4)  ;

= 65.9 W/cm To determine q'"D rig ##l"ti"*

• 9'" , it is u8eful t c mpute -a scaling  ;

, factor from the gross variation of the=rmal neutron flux in the-radial direction (thermal flux and power density are directly- proportional). The normalized radial- flux distribution for the AFRRI TRIGN core is best represented'by a ',

Bessel function of the first kind of order zero:

2.405 r '

f therm * #o( ) (5) where R*= 21.78 cm, the extrapolated core radius f r = 11.99 cm, radial position of D ring element 4 and scaling factor = J, (1.3240) = 0 6074 3 The power density for the D-ring ie thus computed to be 1

( q'"D rig = (0.6074) q"', = 40.0.' W/cm 8

(6) l Because the FFCR fuel element differs from the standard fuel element in L concentration of uranium, the power density in a FFCR fuel element is greater -

than the power density in a standard fuel element by a factor of 1.21. i

[

Taking the above scaling factor into account, the power density of a FFRC I fuel element is found to be q"'Frca = (1.21)q"' (7)

F = 48.4 W/cm ,

3 )

Maximum Temperature in FFCR Fuel Element-Equation (3) takes the following form for cylindrical geometry with axial and azimuthal symmetry:

1d dT- ,

- [-kr-] + q' ' ' =0 (8) j r dr -dr I whleh has a general solution of the form 2

r T(r)-=-q + 011n(r) + O2 (8)  !

The boundary conditions required to solve for the' constants of integration in equation (9) are as follows: '

dT

- = 0 at r = 0 and T = T, at r = 0 (10) I dr  !

Parameters of interest are represented in figure 1. The solution to equation' (8) takes the form 2 2 qR q,,'R - 1 R+c 1 n T, - Tf= +

[ - In( ) '+ ] (11) 4k g 2 k, R h(R+c) where T = maximum centeriine fuel temperature T*= coolant temperature R = 1.38 cm (radius of FFCR fuel element) c = 0.051 cm (cladding thickness) it = 0.18 W cm.'C '(thermal conductivity of. UZrH)4 k = 0.138 -C thermal conductivity of SS304)2 '

h* = 1.339 /c:cs ,.

coefficient of wat,er) (free convective heat transfer l

i u l M; (

• \

8 k.

T. ,

i

\

l'igure L Cross-section of FFCR fuel element.

4 -

l& ;l.?~Ql$Y$f& &yk$$'ll/? D

s 1

It should be noted that the free convective heat transfer coefficient, h, was an experimentally derived quantity. The method by which h was determined is presented in Appendix A. Solving equation (11) using a volumetric heat rate of 48.4 W/cm3 and a bulk water temperature of 48.6'C (the conditions at which h was determined) yields a maximum fuel temperature of 212.4'C. The maximum temperature achieved in the FFCR is nearly 180*C less than the normal temperature of 390'C in a standard fuel element in the B-ring during a 1.0 MW steady-state . power operation. I l

Fuel Ternperature.in Pulse Mode Operation  ;

The fully withdrawn FFCR tends to run cooler than the standard TRIGA fuel elements in the D-ring. Solving equation (8) by applying the boundary conditions of the annular design (with respect to heat production) of the standard TRIGA fuel element with' heat flow out of the outer ,urface, the maximum calculated temperature is 255.8'C. Thus, during pulse operations, we can expect the FFCR to attain a lower peak temperature than the standard TRIGA fuel elements in the-same fuel ring. The maximum temperatures measured in the instrumented TRIGA fuel element in the B-ring for $3.50 and $2.00 pulses are 532.5'C and 267.0'C ,

, respectively, so it is expected that the FFCR temperature behavior will-be well within technical specifications limits.

L CONCLUSION FFCR's are a standard design offered as a stock item by General Atomics and

!' have been used in several TRIGA reactors for over 20 years. FFCR's are currently implemented in approximately a dozen TRIGA reactors.- There has been no reported evidence of fuel failure as a result of FFCR utilization.

The neutronic characteristics for the TRIGA core will remain unchanged! as the prompt negative temperature coefficient of the-12 wt-% fuel followers is virtually the same as that of the standard 8.5 wt-% TRIGA fuel element. Because the maximum temperature achieved in a FFCR fuel element at 1.0 MW is nearly 790*C less than the AFRRI TRIGA technical specification limit of 1000*C, the FFCR's can be safely implemented with no danger of damage to the FFCR' faci l element cladding, t

4 a

k I

s 1

-5

u 1

APPENDIX A. DETERMINATION OF FREE CONVECTIVE HEAT TRANSFER COEFFICIENT i

Introduction l

We can measure the bulk water temperature within the AFRRI TRIGA core to I determine the average free convective heat transfer coefficient of the cooling water. ,

This experiment involves inserting a temperature-measuring probe between the B-  !'

and C-ring fuel daraents while the reactor is operating at a steady-state power level of 1.0 MW and measuring the water temperature at various axial positions.

Once the bulk water temperature has been determined, Newton's law of cooling can be used to calculate the average free convective heat transfer coefficient.

Experirnental Apparatus and Procedure The equipment used in this experiment consistsLof two approximwly 18-ft lengths  ;

of chromal-alumel thermocouple wires fused together at one end, encased in a 16- '

ft-long, 3/8-in-diameter aluminum (Al) tube, and the thermocouple display readout un- the AFRRI computerized reactor contre ansole (Figure A-1). .;

Thermocouple Junction l A1 Tube '!

Z: C Figure A-1. Experimental apparatus.

The potential difference generated at the thermocouple junction as the water-is heated by the reactor is amplified and displayed by the thermocouple circuitry in the AFRRI computerized reactor control console. The thermocouple is initially .

inserted into the core to correspond to position I. The thermocouple resides in each region for several minutes to allow it to attain thermal equilibrium. Once -  :

thermal equilibrium is attained, ten temperature readings. are taken at 10-sec intervals. After each temperature measurement, the thermocouple is withdrawn to [

the next position and the temperature measuring procedure is repeated. -!

Figure A-2 shows that the temperature is measured in five axial positions: (I)3 ,

in below midpoint (14 in of thermocouple wire inserted into the core); (II)

Midpoint in axial dimension; III) Halfway between midpoint and bottom of graphite slug; (IV) At top of fuel region; (V) 1.5 in above top of fuel region.

m m E  : w i

m 2  : 2 n

i Hgo m =

..+. . . . ...

Figure A-2. Axial measuring points. i 6

-- -_- ._ L

. Safety Considerations There are two safety considerations associated with this experiment: radiation streaming and an- unintentional positive' change iri reactivity if _the thermocouple .

wires are rapidly withdrawn from the reactor core while it is 'at power. Radiation '

streaming is avoided by flooding the aluminum tube with water and bending the tube so- that it is at an angle not normal to the top of the core. The thermocouple wire will displace only 0.043 8in of water when it is fully inserted in the core, so using the void coefficient of reactivity,-the thermocouple wire represents a negative reactivity insertion of only 0.001 cents. If we were to make the conservative estimate that the thermocouple wire had the same neutron-absorbing properties of a control rod, the maximum negative reactivity would be only 0.01 cents. Thus there is no possibility of a reactivity accident associated with the apparatus used in this experiment.-

Data Table A-1 summarizes the data gathered during a 1.0 MW steady-state run of the  !

AFRRI TRIGA reactor. The variation in the temperature measurements is most likely due to variance in the radial position of the temperature probe in the t channel.

Table A-1. Bulk Water Temperature at Each Axial Position in the AFRRI TRIGA Reactor Core Axial Position Inlet temp (*C) Measured core bulk water temp ('C)

I 22 72.9 II 24 65.0 III 25 48.6 IV 26 51.6 V 27 59.7 Analysis / Conclusion i

The purpose of this experiment is to determine the bulk water temperature within

! the core shroud; thus, it is the lowest measured value of the ' water temperature 4

that is sought. Figure A-3 illustrates the temperature variation within a cooling C"""**

i am as '

e T /~

I U

- \

1 b E i

=

v w:p i

,::c:c Et Figure A-3. Temperature variation within a cooling channel.

Table A-1 shows that the measured value that most closely represents the bulk i water temperature within -the core shroud is 48.6*C.

4 7

i 1

g -

1 The free convective heat transfer coefficient, h,-is found by solving equation (8) for J boundary conditions given by a standard TRIGA fuel element. Equation (A-1) gives the solution .In terms of h. 1

' 2 I r 1 1-1 l(Tg-T)- qrg lr r(- )2 - 21n( ) - 11 l f

1 1 4k g ( rg rg -) 1 r +c'l  ;

h=( )I _ In( , o)l (A-1)

[ r, + c, I qrg (r k, r, I I I (-- ) 2 - 11 1 l 2 L rg ) J i where T = measured ~ fuel l temperature at ~1.0' MW '

T,i = measured bulk coolant temperature in the core r = fuel outer radius,1.816 cm '

r' = fuel inner radius, 0.229 cm c = cladding thickness,. 0.051 cm  !

k = thermal conductivity of fuel, 0.18 W/em *C -

k = thermal conductivity of clad, 0.138 W/cm *C q'" =- volumetric heat rate.

The measured fuel temperature in the B-ring at.1.0 MW steady-state power level 8

is 390*C, and the calculated volumetric heat rate is 65.9 W[cm . Using .the +

measuged value of the bulk coolant- temperature of '48.6'C yields a value of 1.339 W/cm *C for the free convective heat transfer coefficient. ,

i 1.

i 2

e 4

8 ,

- APPENDIX B. REQUESTED - ADMINISTRATIVE CHANGES Changes to the AFRRI TRIGA technical specifications as effected by the implementation of FFCR's are as follows:

Section 1.0 - Definitions Pare 2. Section 1.9: Replace the definition for a fuel element in its entirety to accommod.-te fuel follower control rods as follows: w 1.9 FUEL ELEMENT A fuel element is a single TRIGA fuel rod, or the fuel portion 01~ a  !

fuel follower control rod. i Section 2.0 - Safety Limits and Limiting Safety System Settines No changes.

Section 3.0 - Limiting Conditions for - Ooerations s

No changes.

Section -4.0 - Surveillance Recutrements Pare 22. Section 4.2.5: Replace the Specifications in its entirety as follows to  ;

clarify the requirement for fuel element surveillance and to accommodate the- i fuel follower -control rods: q All-the fuel elements present in the reactor core, to include fuel follower control rods, shall be inspected for damage or deterioration, and measured for length and bow at -intervals separated by not more ,

than 500 pulses of insertion greater than $2.00 or annually (not to I exceed 15 months), whichevej occurs first. Fuel elements in long-term storage need not be measured until just prior to being returned to core; however fuel elements routinely moved to temporary storage shall be measured every 500 pulses of insertion greater than 82.00 or.-

annually (not to exceed 15 months), whichever occurs first.

Section 5.0 - Design Features Pares 25. Section 5.2.1: Replace the Applicability to accommodate fuel I follower control rods with the following-  !

Apolicability 1

These specifications apply to the fuel elements, t'o include fuel follower control rods, used in the reactor core.

j Pare 26. Section 5.2.1: Replace part to accommodate fuel follower control ds ro(a) withofthe thefollowing:

Specifications in itr entirety

a. Uranium content: Maximum of 9.0 weight percent enriched to less than 20% uranium-235. In the fuel follower, the maximum uranium 9

content wl!! be a maximum of 12.0 weight percent enriched to less-than 20% uranium 235.

Page 26. Section 5.2.1: Add the following paragraph Lto the Basis to accommodate fuel follower control rods:

The power density of a 12.0 weight percent fuel follower element of the

< same diameter as a control rod, which is- smaller than the standard TRIGA element, will produce the same power density in the local area as the standard 8.5 weight percent TRIGA ' elements- due to its-increr. sed hydraulic diameter.

Pare 27. Section 5.2.3: Replace part to accommodate fuel follower controlds ro(a) withof the-

. the Specifications' following: - in its entirety

a. The standard control rods shall. have scram capability, and shall:

contain borated graphite, B4C powder, or boron and its compounds in solid form as a poison in ' aluminum .or stainless-steel cladding.

These rods may have. an aluminum, air, orl fuel follower. If fuel followed,, the fuel region.will conform to the Specifications of 5.2.1.

Section 6.0 - Administrative Controls No changes.

The following. modification paragraph of 4.10.1 of the Safety Analysis Report is required to accommodate the implementation of the fuel follower control rods:

4.10.1 Standard Control Rods and Guide Tubes The shim rod (SHIM), safe rod (SAFE), and regulating rod (REG) cons;itute the three standard control rods and are located in core positions D 1. D-7, D-13 respectively (Figure 4-3).

A standard control rod (Figure 4-7) consists of a sealed stainless steel tube (0.020-inch thick) approximately 37.85 inches long and 1.125 inchu la diameter.' The upper half of the tube contains a 15 inch long compacted borated graphite rod (25 percent free boron or boron compounds) as the neutron absorber, or poison. The lower end of the tube (the follower) contains a solid 15 inch long 1.085 inch diameter rod of UZrH fuel which is is 12 w-% in uranium and has a nomina! enrichment of ?O %. The control rod ' guide tubes are attached to.

the lower grid plate and provide space for inserting and withdrawing the control rods and pass through the upper and lower grid plates.

Figure 4-7 will be modified to reflect the implementation of FFCR's.

0 10 c

APPENDIXL C. RESULTS OF RRFSC REVIEW i

The FFCR safety analysis and proposed technical specification changes were reviewed by the full Reactor and Radiation Facility Safety Committee (RRFSC) on March 27,1990. The result of this review was unanimous approval for the required technical specificatien changes and installation of FFCR's.

y e

t J

t 4

4 w

~

11

n 1

1

~

APPENDIX -

D. PROCEDURE

FOR. REACTOR CORE LOADING AND i UNLOADING i Revised: March 1990 PROCEDURE VII REACTOR CORE LOAi)ING AND UNLOADING General: Loading and unloading of- the reactor core shall be performed under the supervision of the Reactor Facility Director or the Reactor Operations Supervisor. 1 These procedures are superseded during CET Operations (see Procedure I, Tab B) i and during annual shutdown maintenance (see the current Annual Shutdown  !

Checklist).  !

Specific: 1

1. Setup

a. Ensure that at least one nuclear instrumentation channel is' operational.

b. Ensure-that the source is in core. .

c. Ensure that an operator monitors the reactor console during all fuel movements. -

d. Check new fuel -lements before insertion into the core; this includes i cleaning,-visua' hispection, and length and bow measurements. l

e. If irradiated fuel elements are to be removed unshielded from the pool, l from the Safety and Health obtain a Special Department (SHD); doWork Permit not remove fue (SWP) l elements with a power history (greater than l' KW) in the previous 2 weeks from the reactor pool.

2. Core loading

a. After each step of fuel movement perform the following:

Record detector readings.

Withdraw control. rods 50%; record readings.

Withdraw control rods 100%; record readings.

Calculate 1/M.

Plot Predict1/Mcritical versusloading.

number of elements (and total mass of U-235). ,

Insert ALL rods; continue to next step. '

b. L ad elements in the following . order:

Load the B-ring thermocouple element.

Load the C-ring thermocouple element.

Install temperature measurement system (to measure fuel temperature).'  ;

(4) Install any other thermxouple elements. '

(5) Complete loadirg of B- and C-ring elements (total of 18 standard elemente plus 7 FFCR's).

(6) Load D-ring (total .of 33 standard elements plus 3 FFCR's)

(7) Load the following E-ring elements in order:

16, 17, 18, 20, 6, 8, 9, 10 (8) Complete the E-ring by loading (total of elements the following 41 elements in order: plus 3 FFCR's).

15, 21,11, 5,14, 22, 4,12,13, _1 (total of 57 standard elements plus 3 FFCR's)

(9) Load the following 7-ring elements in two elements per step until - '

criticality is achieved using the following loading order:

22, 23, 24, 21, 20, 25, 26, 27, 28,'29, 30,' 1, 2, 3, 4, 5, 19, 18, 17, 16, 15, 14, 13, 6, 12, 7, 11, 8, 10, 9.

I 12

=

_ _ - - . . . . - . . .. .= M

l ,

(10) Load core to- $2,00 excess reactivity by loading two elements per step - ,

l - using the loading order in instruction 9.

(12) Estimate' control rod worth using rod drop techniques.

• l (13) Measure control rod worth; calcula'e value of remaining ' elements as >

! they are added.

L (14) Load the core to achieve 'a K-excess that will allow calibration of the s'

' TRANS rod' based- on the last available worth curve of the TRANS ~

rod (approximately $4.00). ,

. Calibrate the TRANS rod. ,

s  !

, Estimate the shutdown margin. .

Estimate K-excess with a fully loaded core (must not exceed $5.00).

l- Load core to fully operational load using loading order ~ 1n instruction 9, and recalibrate all control rods. ,

. 3. Core Unloading: .

a. Unload' the reactor core starting with the F-ring and ending with _ tht- B rmg. .

b. Remove the fuel elements individually from the . reactor core, identify them by serial number, and place them in the fuel storage racks or a shipping c88h.'

c. If elements are to be loaded into a shipping cash, clean _ the cask :

1 completely and check for radiological contamination before' placing the~

cask in or near the pool. Load cask in accordance with procedures specific to the cask.

d. Once the cask is- loaded, perform an air' sample and survey;l check' temperature and pressure inside cask, if necessary. .

.s. ,

. c. If elements' are placed in . temporary storage away from core monitoring,x J

ensure that criticality monitoring in accordance with 10 CFR 73 is in place, i

f 1 ,

I.

t

.] '

l

'13  ;

-'i._ .

' REFERENCES

1. General Atomics, letter to Mr. Mark Mcore on fuel follower control rods,' 28 October,1988.

2. El-Wakil, M.M., Nuclear Heat Transnor ,d The Amench Nuclear Society, Lagrange Park, IL,1978. '

3. Defense Atomic-Support Agency, AFRRI/USAEC Facility 14nse R-84.

Conmlete with Anolientions and Amendmesta, Bethesda, MD, IM.

4. Wallace, W.P. and Simnad, M.T., Metallurry of TRIG A Ftel Elements. GA- l 1949, General Atomic, San Diego, CA,1961, i l

l L

i i

'I 4

4

-l I

1 k

i

~

.