ML20138N793

Responds to Violations Noted in Insp Repts 50-276/85-02 & 50-160/85-03 on 850826-29.Corrective Actions:New Procedure for Analyzing Liquid Waste Appended.W/Two Oversize Encls

ML20138N793
Person / Time
Site:	Neely Research Reactor, 05000276
Issue date:	12/09/1985
From:	Karam B Neely Research Reactor, ATLANTA, GA
To:	Walker R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II)
References
NUDOCS 8512240157
Download: ML20138N793 (19)

Text

-

E 041tr

-* doorgia institute to Technology j --

Neely Nuclear Research Center ~ - - -

Atlanta, Georgia 30332 (404) 894-3600 amonStATsCIE 1800 1906

_ _ . _ 55 CEC 17 P 3 : 51 December 9, 1985 Mr. Roger D. Walker, Director Division of Reactor Projects U.S. Nuclear Regulatory Commission Region II 101 Marietta Street Atlanta, Georgia 30303

SUBJECT:

INSPECTION REPORT NOS. 50-276/85-02 and 50-160/85-03 This letter is our response to the referenced inspection of the GTRR on August 26-29, 1985.

I tem 1 Technical Specification 3.5.a(5) required that during release of liquid radioactive effluento, a sample be taken from each tank prior to release and during release, and an independent sample shall be taken from the discharge line during release.

Contrary to the above, the requirement to perform sampling during releases of liquid radioactive effluents was not met, in that:

a. Between October 4, 1984, and August 15, 1985, tank samples had not been obtained during releases,

b. On October 4, 1984, November 9, 1984, April 24, 1985, June 10, 1985 and June 11, 1985, discharge line samples had not been obtained.

Technical Specifications 3.5.a(5) states, "During release of liquid radioactive effluents, two independent samples of each tank shall be taken, one prior to release and one during release. An independent sample shall be taken from the discharge line during release."'

The interpretation of this Technical Specification under which the GTRR was operated from day one and practiced until August 29, 1985 was to take one sample from the faucet at the bottom of the tank before release. If the results of the analysis on that sample were within Technica,1 Specifications 3.5.a.1 and 2, 8512240157 PDR ADOCK 0 B$h60 PDR I Q

An Equal Educaten and Employment Opporturuty Institution JL60) 1 \

A Unit of the University System of Georgia

Mr. Walker Page 2 December 9, 1985 pumping of the liquid waste out of that tank begins. At the midpoint of the pumping process, i.e., when half of the liquid content has been pumped out, another sample is taken from the discharge line and analyzed. Occasionally the pumping out of half the content was not achieved until the second day (pumping interrupted overnight) . This practice has contributed to and is perhaps the main reason why on the occasions cited in the reference inspection, the second analysis was never made Technical Specifications 3.5.a(5) is at best ambiguous. For example, the sentence starts with "During release . . . two independent samples of each tank shall be taken, one prior to release and one during release." Now, hos could one take a sample during release (the first during release in the sentence) only to be told that one sample must be taken prior to release at the end of the sentence. Another ambiguity comes from the words,

". . . of each tunk". We have three tanks and normally we pump liquid waste out from the large 5000 gallon tank only.

Consequently our practice has been that we analyze samples taken prior to and during release from that tank only.

Since August 29, 1985 we have modified our procedure in such a way that we analyze three samples: one taken prior to release from the bottom of the tank from which liquid waste is to be released; two taken during release, one from the bottom of the tank and one from the discharge line. The revised procedures which were approved by the Safeguards Committee are appended.

Please note that we do not take samples from each tank. If NRC disagrees with this interpretation, I would appreciate being informed about it.

Item II Technical Specification (TS) 3.5.b required that the maximum release rate of gross radioactivity in gaseous ef fluents not exceed 585 microcuries per second ( p Ci/sec) of Ar-41 equivalent, that both gaseous radioactivity monitors (Kanne

, and Geiger-Mueller) be set to alarm and automatically isolate the gaseous waste release prior to exceeding the 585 pCi/sec release rate, and that if the maximum release rate of Ar-41 equivalent gaseous effluent is exceeded, gaseous discharge from the facility shall not be resumed until the cause of the excessive discharge is identified and corrected.

Technical Specification 6.4.b(6) required that radiation control procedures be provided and utilized.

J

Mr. Walker Page 3 December 9, 1985 Contrary to the above, the requirements for the release of gaseous waste and maintenance of radiation control procedures and records of gaseous effluents were not met in that:

a. On January 6, March 9, September 21, December 11, 1984, and January 16, 1985, the release rate of gross radioactivity in gaseous effluents exceeded 585p Ci/ seconds.

b. On January 6, 1985, the Geiger-Mueller gross radioactivity monitor was set such that the monitor would not have alarmed and automatically isolated the release unless the release rate was a factor of 100 above the limit. c

c. The licensee failed to cease gaseous effluent releases until the cause of excessive releases were identified and corrected for the five releases in excess of 585 p Ci/second whicn occurred on January 6, March 9, September 21, December 11, 1984, and January 16, 1985.

d. The licensee failed to establish a radiation control procedure to determine gaseous effluent release rates immediately preceding and during exhaust duct isolation, and to document the actual release rates for the gaseous effluent releases which were in excess of the 585p Ci/second release limit.

Item b above refers to a January 6, 1984 occurrence.

The problem with Argon-41 has been appearing and reappearing for sometime. The primary reason for this is lack of quantitative and definitive data regarding what actually is being released through the stack after the containment building has isolated.

For example, NRC Inspection Neport No. 50-160/84-03 reporting results of an inspection performed May 11-17, 1984 did not find any violations with regard to gaseous releases even though our records for such releases were inspected (page 2 Report Details). The findings of the last inspection are different.

In order to quantitatively determine whether or not Technical Specification 3.5.b was violated we measured the activity of the air in the exhaust stack downstream from the 30,000 cfm dilution blower with a GM counter. Another GM counter was placed in the exhaust duct near the inlet of the sampling tube leading to the Kanne chamber. This position is located beyond the containment building (CB) isolation valve. When the CB is isolated, the Kanne chamber continues to sample the air in that section of the exhaust duct. However, the flow of air in this section practically ceases at the instant of isolation.

.1

Mr. Walker Page 4 December 9, 1985 The outputs of the two GM tubes were connected to a two-pen chart reco. der so that side by side comparison of the traces can be m s r. .

In order to have the CB isolate at a lower point than the set. point of the Kanne chamber, the isolation was actuated with the gas monitor. This instrument is part of the permanent GTRR equipment and is located some 50 feet upstream from the isolatian valve. The trippoint on the gas monitor is at the upper end of the range for each decade. The conditions leading to isolation were as follous:

1. GTRR was at a power level of one megawatt for one hour and 36 minutes. This is necessary to generate Ar-41.

2. After shutdown, the plug of experimental port V-43 was removed and re-inserted. This step simulates the action of a piston. displacing a tolume of air equivalent to that of the plug. The displaced air contains the Ar-41 formed during the time the reactor was at a power of 1 MW. The displaced air is pushed into the building exhaust stream toward the gas monitor and out through the stack. If the activity is high enough, the gas monitor will cause isolation. The air activity is measured by the two GM tubes placed in the exhaust duct and stack. The chart recorder traced the events. Figure 1 shows the results. The blue trace is due to the GM tube in the stack, the red trace is due to the GM tube in the exhaust duct. There is a physical displacement of about three divisions between the blue and the red lines; the blue is set ahead.

The chart speed was one small division /2.5 sec. The time corresponding to the removal and insertion of the plug from V-43 is marked on the trace in Figure 1. The time it took the puff of Ar-41 to reach the GM tube in the stack is approximately 15 seconds. It appears that both GM tubes began sensing Ar-41 at about the same time.

The instant the containment building isolated, the air activity in the stack started to drop toward zero and remained low. The activity in the duct beyond the isolation valve continued to increase after isolation (remember the red line trace is offset 3 divisions behind the blue). The increase is apparently due to a reverse in flow, i.e., back flow into the duct. This is brought about by the fact that when isolation takes place and the 4000 cfm flow from the CB stops, the increase in flow of outside air to maintain the normal 30,000 cfm flow through the stack causes the slight back flow. Additionally, the pump which draws air

]

Mr. Walker page 5 December 9, 1983 from the duct and into the Kanne chamber continues to siphon air from the duct at a rate of 2.5 cfm and this will most certainly cause back flow.

The air in the exhaust duct between the isolation valve (when closed) and the mixing chamber at the foot of the stack is drawn through the Kanne chamber at a rate of 2.5 cfm and is ultimately emptied into the mixing chamber and out through the stack. The dilution factor, however, is now 2.5 cfm to 30,000 cfm instead of the normal 4000 cfm to 30,000 cfm before isolation. This fact is apparent from the trace in Figure 1: Immediately after isolation the two traces diverge. The trace from the exhaust duct, comparable to that of the Kanne chamber continues to rise and slowly dies off whereas the trace from the stack falls off immediately and stays low.

The results shown in Figure 1 +ad to the conclusion that whenever the containment building is isolated by the gas monitor, the release rate of Argon-41 equivalent gases is highest at the isolation point and decreases rapidly thereafter.

Since the gas monitor is some 50 feet upstream and the Kanne chamber is some 20 feet downstream from the isolation valve, it was felt that it would be necessary to demonstrate that similar results would be obtained when CB is isolated by the Kanne chamber chart recorder (KCCR) actuation. Consequently the experiment was repeated and this time isolation was achieved by lowering the setpoint on the KCCR. The results are shown in Figure 2. Again, the blue line is the trace from the stack monitor and the red from the exhaust monitor. The blue trace is on a scale of ten lower than the red. Again we see that after about 20 seconds from the time the shielding plug is inserted back in V-43, the puf f of Ar-41 reaches both GM tubes causing an increase in response and an isolation on the KCCR. And again we see that upon isolation the stack monitor's response immediately drops to practically zero and stays there.

The conclusion that can be drawn from the results in Figure 2 is similar to the conclusion drawn from Figure 1. This leads to the ultimate conclusion that the release rate of Ar-41 equivalent gases -is highest at the isolation point and decreases rapidly thereafter.

In summary the data we obtained to verify whether or not Technical Specification 3.5.b was violated show conclusitely that no violations were incurred on any of the dates cited.

Mr. Walker Page 6 December 9, 1985 With regard to Item I, we agree that TS 3.5.a(5) was violated for the reasons stated abeve. The corrective steps taken are a new procedure for analyzing liquid waste (appended).

I hope that you will find our response satisfactory. If you have any quencions please let me know.

Sincerely yours,

.h-R.A. Kara.

(Ul6 ~--

Director RAK:jlr Enclosures I

FIGURE 1: Air activity as taeasured by two GM tubes: one in exhaust duct near the inlet of the sampling tube leading to Kanne chamber (red line) and the other in r the exhaust stack of the GTRR. Isolation achieved by gas monitor.

OVERSIZE DOCUMENT PAGE PULLED i

SEE APERTURE CARDS NUMBER OF PAGES:

ACCESSION NUMBER (5):

Rr~ taavois7-o/

i ,

i l

i l

l APERTURE CARD /MARD COPY AVAILABLE FROM RECORD SERVICES BRANCH,TfDC FTS 492 = 8959 i

i L - - - . -.. _ - _ _ _ __-

FIGURE 2: An activity as measured by two GM tubes: one located in the exhaust duct near the inlet of the sampling tube leading to the Kanne chamber (red line) and the other in the exhaust stack of the GTilR. Isolations achieved by Kanne chamber chart recorder.

~ _ _ - - _ _ - _ - _ - _ _ _ _ _ _

OVERSIZE O DOCUMENT PAGE PULLED SEE APERTURE CARDS MUMBER OF PAGES: /

ACCESSION NUMBER (S):

86/SSY O /S~7-D A i

l l

i APERTURE CARD /MARD COPY AVAILABLE FROM RECORD SERVICES BRANCH,TIDC FTS 492-8989 ,

)  !

t 1

'~

G ORGIA EG RESEARG REACTOR Chapter 11 Procedure 9025

• , Approved 12/10/85 Health Physics Latest Rev.

Operations Liquid Waste Tank Analysis Page 1 of 2 I. PURPOSE The purpose of this procedure is to determine if the liquid waste effluents are within the limits set by 10 CFR, Part 20 and GTRR Technical Specifica-tions for instantaneous liquid effluent release rates.

II. MATERIALS REQUIRED A. Parabella funnel B. Vacuum Erlenmeyer flask C. Celman membrane filter 0.45 pm X 47 mm diameter D. Liquid scintillation vial + 10 cc of standard liquid scintillation preparation E. Appropriate radiation counting equipment (Liquid Scintillation for H-3 and C-14, Gas Flow Proportional Counter for alpha and beta gamma, etc.)

f II

I. PROCEDURE

A. Obtain 500 cc sample from lab >perations.

B. Vacuum filter the sample twice.

C. Take 1 cc of filtrate and place in Liquid Scintillation vial with 10 cc of liquid scintillation preparation.

D. Count the membrane filter sample on the Gas Flow Proportional Counter (1 minute mode) and the 1 cc sample on the Liquid Scintillation Counter (1 minute mode 5 times).

E. Determine the concentration of gross radioactivity and tritium in the effluent (units of pCi/ml). See Procedure 9026.

F. Refer to Technical Specifications Section 3.Sa to see if sample is within pumping limits.

G. Notify lab operations to begin pumping operations if sample is within limits. Use Form RS-28.

a

+

e.

11 GEORGIA TECH RESEARCH REACTOR

r. Procedure 9025 P- Health Physics Apprcved 12/10/85 Operations Liquid Waste Tank Analysis Latest Rev.

Page 2 of 2 H. Obtain from Lab Operations a second 500cc sample from the tank during the release and analyze it utilizing steps III B through F above. If sample exceeds limits, notify lab operations to immediately suspend pumping operations.

I. Obtain from lab operations a 500cc sample during the release from the discharge line for analysis. If sample exceeds limits, notify lab operations to immediately suspend pumping operations.

J. After pumping is complete, fill out liquid waste logbook.

K. Place paperwork in liquid waste notebook for later inspection.

NOTE: Refer to Technical Specifications Section 3.Sa., Form RS-28 and Calculation Model, Procedure No. 9026.

9 4

e t

N

s

'~

Form RS-28 (10-64)

WASTE TANK PlfMPING Date Sampic No.

To:

From:

Subject:

Coorgia Tech Roccarch Reactor Waste Tank Pumping You are (are not) authorized to ptemp the centent of Tank No.

with a concentration of microctaries/cc to the Atlanta, Ccorgia, Sanitary Sewer based on sampic taken t ime date Radiological Safety Officer.

To: Radiological Safety Officer From: .

gallons from Tank No, were pumped,to the Atlant.a. Georgia, Sanitary Sewer. Pumping conpleted

~

time date Signatureofperrondoingptsnpin[~

I-- , .,

, ,ill, llll l

]O t

=

M

/

C d r e e

~

t c c i

e f r f r O o

C y t

. e f

e : : : : : : :

a dS y e

- r o

. ol Ba u

t N ,

c a

- e g Mg

.i gn -

l p ,B k o i tl S -

m a ro ei b S bd a oa L

:  : :  : :  : : RR -

e t

u  :

n n i o -

. M i

/ t -

s a t r n t -

u n o e r

C c R o n O s o F b C A

T r -

R o -

O f P

E  : :  : :  : :  : : n R o i

G t N

y a _

I t l S b n u S u c

. E d o l C e y C a _

O t t C R s i l P e r a

. u o t E q i o L e r T P R P M

A S

n o _

i t

a r

, u D

e _

m

) i 4 T _

7

- )

8 p : :  : :  : :  : :

( e e t r _

9 a r P

- D o S F e d R e l e l t p et _

m p n m t n r m u a au o a o S Do _

F S C ( C