ML20202A756

Forwards Required 30-day Event Rept ER-97-17 Re Actuation of Process Gas Leak Detector Head as Result of Leak of UF6.List of Commitments,Encl

ML20202A756
Person / Time
Site:	Paducah Gaseous Diffusion Plant
Issue date:	11/25/1997
From:	Polston S UNITED STATES ENRICHMENT CORP. (USEC)
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
GDP-97-1044, NUDOCS 9712020210
Download: ML20202A756 (7)

Text

.

o,uia we.

I nrwhment eivpwatum P.am.h si, otrue P O ikis 1410 Pada.h, K Y 41001 1et 302 441.sh03 Iat 502 44l 5801 November 25,1997 United States Nuclear Regulatory Commission SERIAL: ODP 971044 Attention: Document Control Desk Tashington, DC 20555-0001 Paducah Gaseous DHTusion Plant (PGDP)- Docket No. 70 7001 - Event Report ER 9717. Rev.1 Pursuant to the Safety Anansis Report (SAR), Section 6.9, Table 1 Criteria J2, Attachment 1 is the required 30-day written Event Report covering ER 9717. This event involves the actuation of a Process Gas Leak Detector (PGLD) head as a result of a leak of uranium hexafluoride (UF.).

The Nucicar Regulatory Commission (NRC) was notified of the event on October 7,1997 (NRC No. 33038). An initial investigative report was forwarded to NRC on November 3,1997.

l is a lis*. of commitments made in this report.

Should you require further information on this subject, please contact Bill Sykes at (502)441 6796.

Sincerely, BM Steve Polston General Manager Paducah Gaseous DifTusion Plant SP:WES:JLil:mel l

t Attachments (2)

] Q.

a uw ec:

NRC Region til NRC Senior Resident inspector, PGDP 9712O20210 971125 PDR ADOCK 07007001 ll ll ll.ll. ll ll C

PDR

Docket No. 70-7001

'Attachhient 1 Page1of5 EVENT REPORT ER 9717 (Rev.1) l BACKGROUND Cascade feed cylinders are transported to the cascade feed facilities in C 333 A and C 337 A.

The cascade feed vaporization involves heating the cylinders in steam heated containment type autoclaves to convert the solidified uranium hexafluoride (UF.) to liquid with a relatively high vapor pressure. The pressurized vapor is drawn ofrand controlled by valves and flow-measuring devices to maintain distribution of UF. gas through heated piping to appropriate points in the enrichment cascade. Although the autoclaves are designed to contain a UF. release, Process Gas Leak Detector (PGLD) heads are installed above the autoclave head locking ring, the heated housing at the autoclave head, above the. jet station piping, in the piping trench, and on the west wall in C-337 A. If a leak is detected, an alann is sourided in the autoclave area and on the UF.

d:tector alarm panel in the Operations Monitoring Room (OMR).

On October 6,1997, a small UF. leak occurred on the purge air pressure bleed line on autoclave 3 South at the C-333 A feed facility. As a result of the leak, PGLD head YE 61313, which is a "Q" safety system located in the autoclave housing, alanned. The 3 South autoclave and five other in service autoclaves in C 333 A were placed into containment, the piping was evacuated below atmosphere, and the piping leak was temporarily plugged.

The automatic actuation of a "Q" safety system resulting from a valid conditi'n is reportable in l

accordance with Safety Analysis Report (SAR), Section 6.9, Table 1, Criter;a J2. On October 7,1997, the Nuclear Regulatory Commission (NRC) headquarters was notified of the event (Event No. 33038).

DESCRIPTION OF EVENT On October 6,1997, at 1540 the C 333 A position 3 South PGLD head (YE 613-13) fired. The l

operators responded in accordance with the alarm response procedure (CP4-CO AR8333 A), and emergency procedure (CP4 CO-CE5017A), donned their protective suits and took hydrogen fluoride (llF) samples. None of the samples indicated the presence ofIIF. An attempt was made l

to reset the POLD head; however, the attempt was unsuccessful. At 1610 a smoke watch was l

l established and at 1625, YE 61313 was declared inoperable. At 1635 additional troubleshooting l

l reported the presence of smoke from a small pin hole leak from a three-eighths inch copper purge l

line. The area was immediately evacuated of all personnel and the operations supervisor and l

Plant Shift Superintendent (PSS) were notified. At 1654 the C-333-A autoclaves were placed l

l t

Docket No. 70 7001

' Attachment 1 Page 2 of 5 into the containment mode. At 1705 the Plant Emergency Squad was called out to report to l

C-333 A position 3 South for a small UF release. The Emergency Squad took additional air samples, which were negative, and plugged the source of the UP. leak. At 1945 the "all clear" I

l was given. At 2225 the C-333 A position 3 South was declared inoperable, due to the release l

from the purge lin. At 2300 position 3 South was placed in mode 2 and the smoke watch l

terminated.

On July 15,1996, an Engineering Service Order (ES0) Z o880 was developed to install pressure bleed points on select system piping to allow for full differential pressure testing of all autoclave containment block valves during autoclave containment testing. At the time, the existing autoclave configuration did not allow for full testing of outer autoclave containment valves.

A tap and valve was added at three places on piping at each autoclave to allow back pressure to be relieved for containment valve testing on 22 autoclaves. On December 18,1996, modification work was completed on autoclave position 3 South. Pressure was initially raised on the line to 110 psig. The pressure was then reduced to 100 psig, a soap test performed, and no leaks were evident.

Troubleshooting indicates that the leak occurred in a copper line which was installed in accordance v'ith ESO Z96880. The line is located between the purge containment valve (CV Sil) and the purge valve (CV 512). The purpose of the copper line is to bleed off pressure between CV 511 and CV 512 for pressure decay tests. Copper is considered compatible with UF.; therefore, the presence of UF. would not cause a deterioration of the copper. Additionally, there have been no other leaks, that were not discovered during installation testing, from copper piping installed on the other 21 autoclaves as a part of the improved containment testing modifications.

On November 5,1997, the Materials Technology Laboratory provided a repon relative to the I

copper piping and pipe fitting. An extract from the report follows:

l l

As part of the emergency response, the Emergency Squad applied UF. resistant putty to l

contain the leak. When the putty was separated from the piping in the Materials Lab, the I

entrapped uranium oxides were located near a braze joint which had been leaking. The l

leaking side of the joint appeared cracked and porous with no properly designed fillet.

l The joint was cross sectioned to reveal a leak path Fetween the fitting and the copper I

tubing. Although small amounts of brazing material were found at the base of the copper l

tubing, no brazing material was found wetted to the fitting in this area. Analysis of the l

fitting by energy dispersive spectroscopy on the scanning electron mir.oscope revealed l

the material of const uction to be a stainless steel alloy similar to 316 SS. The brazing l

material was analyzed to be a silver zine-copper alloy similar o bag-10. After j

metallographic etching, it was also discovered that the col,per tubing had been melted l

during the brazing process, indicating that poor brazing technique was used during i

fabrication. No signs ofcorrosion wer-found in the brazing material.

l

Docket No. 70-7001

'Attachhient i Page 3 of 5 Brazing of stainless steel alloys are complicated by the formation of chromium oxide l

films which form on the surface when stainless steels are heated in air. These oxides l

interfere with a brazing material wetting to the surfaces. Correct solder alloy, good l

surface pirparation, a flux which is suited for the material and rated for the necessary l

brazing temperatures, and proper brazing techniques are crucial for a sound joint. The l

pressure bleed points passed a leak test in December 1996. During the test, thejoint was l

physically tight, but not chemically bonded.

l l

As indicated by destructive analysis, the above examination was conducted by sectioning tubing l

and fitting. Ilowever, it was not possible to determine the poor quality of thisjoint by l

nondestructive examination.

l l

The worker who did the brazing on the joint was qualified under Lockheed Martin Utility l

Service (LMUS) brazing procedures, in LMUS shops by LMUS quality control personnel. The l

test to qualify under the brazing procedures includes a stainless steel to copperjoint.

l l

Currently, all autoclaves,in both C-333 A and C-337 A, are undergoing an instrumentation l

upgrade project associated with he Paducah Gaseous Diffusion Plant (PGDP) Plan for l

Achieving Compliance with NRC Regulation. Construction of these modifications requires l

bodily entry into close quarters of the instrument heated housing, such as where the UF leak l

occurred. While the copper tubing had been installed and in use since December 1996, the UF.

l leak did not occur until the second heat cycle following the completion of the instrument upgrade l

modifications to the autoclave. Therefore, there is a relatively high probability that the piping l

joint was inadvertently disturbed during construction activity.

l l

This was an isolated occunrnce, because a visual walkdown examination subsequent to the UF.

l leak revealed no joint damage inside the autoclave heated housing; the brazing work was l

completed by a plant trained, qualified brazer performing the job in accordance with plant l

specificdons; and a similar event has not occurred in the remaining 21 autoclaves which l

underwent identical modifications. The planned corrective actions will reduce the probability of l

further faulty or damaged joints within the autoclave heated housings.

l l

CAUSE OF EVENT A. Direct Cause The direct cause of the actuation of PGLD head YE-613 13 was the release of UF. from the l

copper piping.

l 1

Docket No. 70-7001

'Attachinent 1 Page 4 of 5

11. Root Cause i

The root cause of the UF. release and subsequent actuation of PGLD head YE 613 13 was the l

separation of the copper piping from the fitting. In this instance thejoint was physically tight, l

but not chemically bonded.

l C. Contributing Cause None CORRECTIVE ACTIONS A. Completed Corrective Actions

1. On October 7,1997, autoclave 3 South was declared inoperable and placed into mode 2.

2. On October 22,1997, the copper piping was sent to the plant laboratory for analysis.

I

3. On November 21,1997, Engineering completed a visual walkdown examination of l

piping inside the local autoclave heated housing to assure no damage had occurred as a l

result of the modifications to the feed autoclaves.

l l

B. Proposed Corrective Actions l

l

1. Ily January 5,1998, Maintenance will conduct pre-job briefings cautioning personnel to l

exercise extreme caution when performing work in the vicinity of piping which has been l

brazed / soldered to preclude inadvertent damage to the joints.

l l

2. Ily February 15,1998, Engineering will revise the instrument upgrade modification l

instructions to require a visual inspection of each individual remaining autoclave l

instrument upgrade to assure that no damage has occurred to each remaining autoclave l

(C-333 A position 2, North and South, and position 4, North and South; C-337-A l

position 3, East and West, and position 5, East and West).

l l

0. By August 15,1998, Operations will assure that any piping discovered damaged and in l

need of replacement as a result of corrective action No. 2 above, is removed and l

- Ibnvarded to the laboratory to detennine the existence of any negative trends.

l l

4. Ily February 15,1999, Engineering, using the analysis from corrective action No. 3 l

above, will assure problems regarding piping leaks in the feed facility local heated l

housing are trended to determine if additional failures have occurred due to instrument l

upgrade work. Trends will be evaluated fbr needed corrective actions.

l

~

Docket No. 70-7001

'Attac}dnent 1 Page 5 of 5 I ESSONS LEARNED Although the actuation of the PGLD head in this event was determined to have been en isolated l.

occurrence generated as a result of a separation of the copper piping from the litting, any further l

incidents involving the piping installed, as a part of the improved containment testing project l

- (ESO Z96880), will be evaluated to determine whether an adverse trend is developing.

l l

EXTENT OF EXPOSURE OF INDIVIDUALS TO RADIOACTIVE MATERIALS Two operators and one instrument mechanic had urine bioassay results of 4,14, and 15, respectively, micrograms of uranium per liter. When retested, the results were less than MDA (less than 3 micrograms of uranium per liter), and below the follow-up recall limit of 5 micrograms of uranium per liter. The intake from the event was not significant and well below regulatory limits, i

l l

l f

1 i

--w y

o i

Docket No. 70-7001

• Attach' ment 2 i

Page1of1 3.

List of Commitments l

j I

1. : By January 5,1998, Maintenance _ will conduct pre. Job briefings cautioning personnel to l

- exercise extreme caution when performing work in the vicinity of piping which has been l

brazed / soldered to preclude inadvertent damage to the joints.

l-I

2. 11; February 15,1998 Engineering will revise the instrument upgrade modification l

instructions to require a visual inspection of each individual remaining autoclave l

Instrument upgrade to assure that no demage has occurred to each remaining autoclave l

(C-333 A position 2, North and South, and position 4, North and South; C-337 A l

4

- position 3, East and West, and position 5, East and West).

l l

i

3. By August 15,1998, Operations will assure that any piping discovered damaged and in l

need of replacement as a result of currective action No. 2 above, is removed and l

forwarded to the laboratory to determine the existence of any negative trends.

l 4.'

By February 15,1999, Engineering, using the analysis from corrective action No. 3 l

above, will assure problems regarding piping leaks in the feed facility local heated l

housing are trended to determine if additional failures have occurred due to instrument

- l-upgrade work. Trends will be evaluated for needed corrective actions.

l l

l 1

(..

l' i

e - ~,

.,m-+,.-,_,i,__m....,-

,g._...

,_-,.,.3-,..m.,~,__,

z.,

_