ML20235J960

Ack Receipt of Informing NRC of Results of Safety Evaluation Re Possible Flooding of Oregon State Univ Triga Reactor Graphite Reflector

ML20235J960
Person / Time
Site:	Oregon State University
Issue date:	09/25/1987
From:	Wenslawski F NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V)
To:	Coate E Oregon State University, CORVALLIS, OR
References
NUDOCS 8710020177
Download: ML20235J960 (1)

Text

.

SEP 251987 Docket No. 50-243 Oregon State University Corvallis, Oregon 97331 1 Attention: Dr. E. Coate, Vice President Administration Gentlemen:

Thank you for your report, dated August 31, 1987, informing us on the results j of your safety evaluation associated with the possible flooding of the OSTR  ;

graphite reflector. This item will be reviewed during a future inspection, i Your cooperation with us is appreciated. l l

Sincerely, l Original signed by j

[.'A.NN5ikwski, Chief Emergency Preparedness and Radiological Protection Branch bec: I DN File I l

B. Faulkenberry J. Martin i M. Smith REGIONVdib d MCillis/ noack GYuh8 9M FWenslawski 9/pf/87 9/th/87 9/yf/87 R

f g (VL[S) /QUEST NO ](YE!COPY / / NO ] :JEl@ST YES / \N0/COPY .i REQUEST t0PY

^

v -

. END TO PDR J YESJ / N0 ]

/

pB700g00?N$l@;@y D A eo#g ,Lf7 )

r f g

g ,,

F 3 We" i corvams, oregon 97331 <s=> n n44 Radiation center Urovers.ty, 53 55 C~) g eO 5 w E30 o<

<>.9 -

August 31, 1987 -@

h 5 U.S. Nuclear Regulatory Commission Region V 1450 Maria Lane, Suite 210 Walnut Creek, CA' 94596 Ref: Oregon State University TRIGA Reactor (OSTR), License No R-106, Docket No. 50-243 Attn: Mr. Michael Cillis Gentlemen:

. The chronology in the letter and report, on the possible OSTR reflector flooding, submitted to you on August 28, 1987 was in error. Please find enclosed a revised report with our apologies, ours s* cerely, t

i rian Dodd for A. G. Johnson Director AGJ/of cc: T. V. Anderson, Reactor Supervisor, OSU S. E. Binney, Ch' airman, Reactor Operations Committee, 05U B. Dodd, Assistant Reactor Administrator, OSU Director, Oregon Department of Energy

-)'

C g MI gg 4

Oregon State University is an Affirmative Action / Equal Opportunity Employer

_________________________________j

i; REACTOR OPEPATIONS STAFF REPORT TO THE OSTR REACTOR OPERATIONS COMMITTEE EVIDENCE EVALUATED DURING OSU'S INVESTIGATION OF THE POSSIBILITY THAT THE OSTR REFLECTOR IS' FLOODED INTRODUCTION AND CHRONOLOGY l

In April of 1987, it was becoming apparent to members of the Radiation Center (RC) staff that neutron fluxes in certain reactor irradiation facilities were lower than normal. On April 15, the Reactor Operations Committee (ROC) was informed of this fact and it was suggested that a possible explanation might be thr.t the reflector was flooded. Since there were clearly no immediate safety implications, the ROC asked that various pieces of evidence i be assembled, and scheduled a special meeting for May 11, 1987.

As a result of the May 11 meeting, the Radiation Center staff was instructed: a) to make a preliminary notification to the NRC, b) to investigate the matter further by performing several suggested experiments, c) to contact GA Technologies about the potential flooding, and d) to prepare a safety analysis assumirg that the reflector was flooded. Region V of the NRC was initially notified of the potentially flooded reflector on May 12, 1987. I During June and the first part of July, various experiments were performed and evaluated by the reactor operations staff. At the quarterly July 15 meeting of the ROC, some of this data was presented to the committee, but it was incomplete. Therefore, a meeting dedicated to the reflector I flooding issue was scheduled for August 6,1987. At this August 6 meeting ,

old and new data and possible interpretations of the data were discussed at length. However, at that time the ROC was unable to come to a consensus as to whether or not the reflector was flooded.

(

1 Following the August 6 meeting, another meeting of the ROC was scheduled for August 20 to allow other members to return from vacation. At this- I meeting,-a draft of this report was accepted as sufficient evidence of the possibility that the Oregon State TRIGA Reactor (OSTR) reflector was flooded. The ROC also agreed that the safety implications of this situation  !

were as stated in this report, and that no further actions, apart from those already taken, were necessary. On the afternoon of August 20, 1987, Mr. Mike C1111s in the Region V Office of the USNRC was notified of these conclusions by the Assistant Reactor Administrator, Dr. Brian Dodd.

SUPPORTING DATA

1. Gamma exposure rates measured at a specific location near one of the beam port (BP #1) shutters during the daily reactor bay radiation survey have decreased from about 50 mR/hr to about 3 mR/hr at 1 MW.

The observed reduction in this exposure rate occurred during the time l 1

period September 1984 to July 1985. The exposure rate at the reference location currently remains at about I to 3 mR/hr at 1 MW. The shutter i

and shielding configurations have not changed since they were originally installed. An examination of the shutters and shielding in and around BP #1 showed no reason for this reduction in exposure rate. ,

With buildup considered, the attenuation of a collimated fission spectrum gamma ray beam would be approximately a factor of 2 through 10 inches )

of water (the thickness of the graphite reflector cutout facing BP #1). Although a significant reduction (about a factor of 17) in the exposure rate has been observed near BP #1, it appears that this reduction may not be entirely attributable to core gamma rays being attenuated by a flooded reflector. For example, Cf-252 neutrons experience

.,' ,4 "o

e an attenuation factor of about 17 in 10 inches of water. Thus, if the neasured gamma exposure rate is primarily neutron-induced (e.g.,

capture or activation gamma rays), the observed reduction would be about the right magnitude to suggest reflector' flooding.

2. The total neutron and gamma ray fluxes in the BP #1 radiography position have decreased significantly (see below). The first set of data consists of normalized (to Juli 1983 values) background densities from X-ray-films used in neutron radiography experiments. The film was exposed in the same location for the same period of time in each measurement.

Date Normalized Film Density July 1983 1.00 July 1984 0.91 February 1985 0.29 The above data represent a reduction in background density by a factor of over 3 in the period between July 1984 and July 1985. This reduction l

is about the magnitude that could be caused by reflector flooding.  ;

The second set of data consists of thermal and epi-cadmium neutron flux measurements. The units are neutrons /cm2-s at 1 MW.

The rmal

-Date Thermal Flux Epi-Cd Flux Epi-Cd Ratio 1976 2.4x107 3.7x107 0.65 July 1987 1.2x106 1.4x104 85 From the above data, it can be seen that the thermal neutron flux has decreased by a factor of about 20 and the epi-cadmium flux has

g R, -._ .

[ L also decreased considerably. The ratio of thermal to epi-cadmium l

neutrons has increased, which implies that significantly more ' neutron moderation is occurring than in the past, thus indicating reflector flooding. However, a simple calculation of fission neutron attenuation using the removal cross section concept indicates that 10 inches of water will decrease the fast neutron flux by a factor of about 14,

, which is smaller than the change measured. On the other hand, there is some reason to doubt the epi-cadmium flux measurement becabse a diffusion code calculation shows that even with large amounts of water the ratio of thermal to fast neutron flux does not exceed 4.

3. The recently measured BP#1 cadmium ratio (July,1987) based on gold was 6.4, compared to a value of 2.7 in BP #4. This comparison was made because BP #4 has a 10-inch water-filled aluminum plug in the inner end of the beam port in the reflector region (to enhance neutron scatter down the tangential BP #3). If the reflector were flooded, then one might expect the cadmium ratio in BP #1 and BP #4 to be about the same. However, the ratio for BP #1 is larger than BP #4 by a factor of over 2. This might be explained by the fact that the current core configuration is eccentric, and there is a ring of graphite elements q 1

and a vacant ring (i.e. , water) adjacent to BP #1. j 1

4. The neutron flux in the BP #3 radiography position has also decreased, '

as evidenced by a change in film density following comparable radio-graphic exposures prior to 1984 and then again after that time. The reduction has been approximately a factor of 10 in the thermal flux as measured by the need to use a film which is 10 times faster to achieve comparable film density. l i

l l

'(

(

L- .

5. The thermal neutron flux at 1 MW measured in thermal . column stringer position #4 (lower right stringer) has decreased by a factor of 5,

.as shown by the following data:

Date Thermal Flux 1980 2.5 x 1011 neutrons /cm2-s 1987 5.0 x 1010 neutrons /cm2-s This reduction may be of the order of magnitude that could be accounted for by reflector flooding, although the reflector only shields a small portion of the thermal column from the core. There are no significant voids in the reflector graphite on the side facing the thermal column.

If the reflector were flooded, there would be at most about 2 to 3 equivalent inches of water in the path, due to graphite porosity (see Safety Evaluation below) and small gaps inside the reflector cladding.

- This thickness of water attenuates Cf-252 neutrons by a factor of

. about 2.

OTHER MEASUREMENTS MADE

1. A series of neutron flux measurements were made in the (stationary) rotating rack to look for any flux decrease (due to increased absorption by water in the reflector) in the positions where reflector voids for bps #1 and #3 are located under the rotating rack. These measure-ments showed no abnormalities when compared to similar previous data.

However, water in the reflector could also give an increase in neutron flu'x at these locations due to neutron scattering, which might be an offsetting factor to the neutron absorption. Thus, this data gives no indication of reflector flooding.

l l

_ _ _ _ _ - . - i

.t. .

I SAFETY EVALUATION OF A FLOODED TRIGA REFLECTOR If the OSTR reflector is flooded, then it is probable that the leak occurred at a weld and that the ingress of water occurred over about a one-year period beginning around mid-1984 and continuing on into 1985.

There are small gaps in the reflector between the graphite and the aluminum housing, and there are also several cut-outs (voids) in the graphite where bps #1 and #3 intersect. In addition, the graphite has a porosity of about 20 to 25%, and all of that space is theoretically available for water to fill.

The OSTR Technical Specifications, section 5.2e, state that "The reflector. . .....shall be water, or a combination of graphite and water."

From a neutronics viewpoint, there are no safety implications for a reflector flooded with water. There is little difference neutronically between water and graphite homogeneously mixed together and a system of water / graphite / water.

If water were to flood the reflector housing, then the safety margin would actually be increased by raising neutron absorption and thereby reducing k..

There has been no indication of any abnormal reactivity changes in the reactor core. Core excess, control rod worths, rotating rack fluxes, and ?,neumatic transfer terminal fluxes all continue to be normal.

One of the reasons for keeping the reflector dry is to avoid the possibility of galvanic corrosion between the graphite and the aluminum; however, corrosion will only occur in the presence of an electrolyte.

The water in the OSTR has always been kept at a very low conductivity level (about 0.5 pmho/cm). Discussions with GA Technologies personnel indicate that impurities in the graphite may be Fe, Si, Ti, Zr, and Ca.

However, according to GA these are all well bound and therefore will not j change the conductivity of any water inside the reflector.

I

_ - - _ _ _ _ _ _ _ _ _ _ _ _ _ . - {

In view of the above, galvanic corrosion appears very unlikely; however, if there were any corrosion inside the aluminum housing around the reflector it would not become a problem unless sufficient corrosion occurred to significantly reduce the structural integrity of the reflector housing.

It is a well-known fact that galvanic corrosion produces pits and holes rather than an overall thinning of the material. Therefore, breakthrough corrosion would be easily detected by white powdery spots on the surface long before it has progressed far enough to weaken the reflector housing.

SU W RY CONCLUSION An examination of the data presented suggests that the OSTR reflector is probably flooded; however, the evidence is not conclusive. The measurements evaluated were taken by a number of different individuals using different equipment and techniques over a long period of time. In addition, the attempts to correlate measurements with calculations are variable; some of them show reasonable agreement, others do not.

If the reflector is indeed flooded with water, then we have concluded that there are no associated criticality or nuclear safety. issues, but there could be a slight possibility that corrosion of the reflector housing would occur. Any significant corrosion of the reflector housing would be easily detected by visual observation. The reactor tank, the core, and other components in the tank are inspected daily, and a recent change to OSTR operating procedures now requires a formal inspection of the tank, the core, and tank components on an annual basis.

It is therefore concluded that no further action is necessary at I

this time and that there are no unreviewed safety questions as defined

)

by 10 CFR 50.59(a)(2) with reference to the possibility of the OSTR reflector being flooded.

l