ML20236T707
| ML20236T707 | |
| Person / Time | |
|---|---|
| Site: | Portsmouth Gaseous Diffusion Plant |
| Issue date: | 07/23/1998 |
| From: | Jonathan Brown UNITED STATES ENRICHMENT CORP. (USEC) |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| GDP-98-2035, NUDOCS 9807280345 | |
| Download: ML20236T707 (6) | |
Text
_ - _ _ _ _ _ _ _ _ _
-?
b l
USEC A Global Energy Company July 23,1998
)
GDP 98-2035
]
l United States Nuclear Regulatory Commission Attention: Document Control Desk Washington, D.C. 20555-0001 l
Portsmouth Gaseous Diffusion Plant (PORTS)
Docket No. 70-7002 Event Report 98-07, Revision 1 l
- Pursuant to Safety Analysis Report (SAR), Section 6.9, Table 6.9-1, J (2), Enclosure 1 provides a revised 30 day Event Report (ER) for an event involving a high condensate level shutoiTactuation at the Portsmouth Gaseous Diffusion Piant. The revised event report includes the root cause and corrective actions. Enclosure 2 is a list of commitments made in the report. Changes from the previous report are marked with a vertical line in the right margin.
' Should you require additional information regarding this event, please contact Scott Scholl at (740) 897-2373.
Sincerely, J. Morris Brown General Manager
/
Portsmouth Gaseous Diffusion Plant j/ 'h l
Enclosures:
(1)
,,,1s3 cc: NRC Region Ill Office NRC Resident inspector-PORTS 9007200345 980723 PDR ADOCK 07007002 C-PDR 110. Box 800, Portsmouth,011 45661 Telephone 740-897-2255 Fax 740-897-2644 http://www.usec.com OMces in Livermore, CA Paducah, KY Portsmouth, OH Washington, DC
j l
GDP 98-2035 Page1of4 Event Report 98-07, Revision 1 l
Description of Event On May 2,1998, at 0344 hours0.00398 days <br />0.0956 hours <br />5.687831e-4 weeks <br />1.30892e-4 months <br />, X-343 Autoclave (AC) #6 was in Mode 11 heating a 48-inch
{
Uranium Hexafluoride (UFJ cylinder when the audible alarm for steam shutdown was received.
6 Operators at the autoclave local alarm panel saw the "B" condensate level probe light on, indicating the high condensate level shutoff (IICLS) safety system had actuated. The autoclave local alarm panel indicated that steam supply block valve FV-613 was closed, stopping steam flow to the autoclave as designed. A HCLS actuation is reportable in accordance with the Safety Analysis Report (SAR), Table 6.9-1, J(2).
Autoclave #6 had just been returned to service fbilowing an outage of approximately six mon:hs duration whose primary focus was to perform a series of modifications under the Autoclave Nuclear Safety Upgrade Project. Among these modifications were improvements to both the steam and condensate systems intended to minimize the atential for liCLS actuations.
The steam system modifications provided increased steam flow to the autoclave. The modifications included the addition of higher capacity 2" parallel pressure reducing valves coming off of the main 6" steam supply header. The pipe size from the main header to the pressure reducing valves was also increased, increased steam flow was needed to prevent the autoclave internal pressure from decreasing during the highest heat demand conditions. In the past, HCLS actuations have occurred because the internal autoclave pressure decreased to the point where there was inadequate motive force to expel condensate from the autoclave.
i The condensate drain system modifications included increasing the existing common condensate l
header from 3" to 6", all other condensate piping was increased from 2" to 3", including new 3" l
containment valves. The existing bucket type steam trap was replaced with a new float /thennostatic
)
trap to increase the condensate drain flow capacity. A new 3" strainer was also installed to prevent i
debris from fouling the steam trap.
l Several tests were performed on AC #6 as part of the post installation testing following the completion of the project modifications on the autoclave. One test involved heating the autoclave for approximately three hours without a product cylinder, to determine if the steam and condensate systems were operating satisfactorily. These tests included a confirmation that the condensate temperature control valve (TCV) was operating correctly. Condensate flow through the steam trap was confirmed by using an ultrasonic listening device and the absence of alarms. All testing was successfully completed. Because of the extensive work performed on the autoclave during the six month outage, the drain lines were boroscoped. The boroscopic in&pection found the drain lines
l' Enclosum 1 GDP 98-2035 u
Page 2 of 4 i
Event Report 98-07, Revision 1 acceptable for operation. The strainer was also inspected after the live steam test to ensure no residual material collected in the strainer before the autoclave was returned to service. No material was found in the strainer. Following the successful completion of post installation tests and Technical Safety Requirement surveillance, AC #6 was placed back in service.
t Approximately 24 minutes following the initiation of Mode 11 (heating), Operators received a IICLS actuation by the "B" condensate level probe. Approximately 2 minutes later, the high level alarm cleared. Operations personnel at the autoclave were closely monitoring the shell pressure, which was at approximately 3 psig and the cylinder temperature readings on the two thermocouple, which were l
approximately 209 and 211 degrees Fahrenheit.
The condensate level shutofTsystem is provided to prevent over pressurization or a nuclear criticality in an autoclave following a postulated UF release. Excess water is undesirable in the event of a UF 6
6 release from the cylinder that could cause either high 11ydrogen Fluoride pressure as the result of the reaction between UF and water or the excessive moderation of an unsafe mass of uranium thereby 6
causing a criticality within the autoclave. The system ftmetion is to detect either a drain line plug or restriction and to shutoff the steam flow to the autoclam Cause of Event Engineering has determined that the direct cause for the HCLS safety system actuation was the l
installation of the wrong orifice in the condensate drain line steam trap. The model number of the l
trap that was installed was not the model number specified in the engineering drawings or equipment /nuterial listing. The 14/2" TLV steam traps specified for the NSU project were to be furnished with size #10:ificembich have a hole diameter of approximately IQ/64". The installed l
l trap, however, contabed a s!w #10 orifice which has a hole diameter of appmximately 37/6f'. The l
smaller size #10 c:Hice only has approximaMy 1/5th the flow area of the larger size #1 orifice.
l Engineering determined that a size #10 orifice is not large enough to meet the calculated maximum I
condensate flow rate of 16 gpm, which corresponds to the estimated maximum 8000 lbs/hr steam l
supply during normal operation. Testing of the steum trap with a #10 orifice and a #1 orifice was l
performed with the TCV isolated and the bypass valve closed. The test involved flowing metered I
cold water down the autoclave drain and through the trap to shnulate condensate flow. The testing l
showed that the smaller #10 orifice pssed approximately 8-3/4 gpm of condensate. The initial as
{
found flow readings through the trap with the #1 orifice varied from approximately 12-1/8 to 14 l
gpm. Additional testing revealed that flow increased to 21 gpm when the trap bypass valve was l
opened approximately 1/4 tum. After the bypass valve was closed the trap maintained a flow of 20 l
i I
\\
GDP 98-2035 Page 3 of 4 Event Report 98-07, Revision 1 gpm without any alarms. Air binding in the pipe is the most probable reason for this occurrence.
l Engineering was unable to determine if the air binding occurred because of the cold water testing l
method or whether it may also be present under steam conditions. The testing demonstrated that the l
steam trap passed more than the calculated maximum condensate flow with only gravity head l
pressure.
l During the condensate flow testing the drain line was borescoped several times to investigate the l
lower than expected flow readings. When parts of the drain line were disassembled for inspection, I
a plastic shoe cover was discovered lodged in the TCV. The TCVs main function is to remove initial I
condensate and non-condensibles from the autoclave during startup until a temperature of I
approximately 210 degrees Fahrenheit is reached. At the time of the liCLS actuation it is not known l
if the TCV was closed or open. The pre-operational inspection of the condensate drain line did not l
verify that all of the line, including the TCV, were free of debris or obstructions prior to placing the l
autoclave in service. An ultrasonic listening device was used during pre-operational testing to verify l
flow through the TCV, but the amount of flow was not quantified. Engineering was unable to l
determine the extent the shoe-cover may have blocked flow through the TCV or contributed to the l
t IICLS actuation. Since subsequent testing was done with the TCV isolated, the effect that the l
choe-cover had on the combin, i flow through both the trap and TCV is unknown. However, the l
discovery of the shoecover indicates that improved foreign material exclusion controls are needed I
for the autoclaves.
l The root cause of the event was the failure to maintain adequate configuration control of non-safety l
related components whose failure could result in safety system actuations. Engineering's l
investigation into why the wrong model steam trap was installed has revealed several weaknesses in the procurement process for non-safety related equipment. It was determined that a typographical error existed on the original engineering equipment specification sheet. This error was resolved through informal conversations but the original drawing and equipment material list were not corrected prior to ordering the. steam trap. The model number listed on the uncorrected engineering documents and on the purchase order description ec,rrectly indicated a size #1 orifice. Ilowever, j
additional information was added to the purchase order description that requested a steam trap with l
an incorrect size #10 orifice. When the part arrived this -error was not detected by receiving personnel. Once the steam traps were delivered to the field, the trap discrepancy was not identified i
l by field personnel examining the parts prior to installation.
A contributing cause for the event was that the testing that was done to confirm correct operation I
of the steam supply and condensate removal systems on AC #6 was inadequate to identify the wrong orifice. The testing demonstrated that condensate passed through the trap, but did not closely I
i
e i
GDP 98-2035 Page 4 of 4 Event Report 98-07, Revision i simulate actual / calculated condensate production during heating and UF phase change conditions.
l l
6 As result, the problem with the new installation was not detected until aller the autoclave was placed back in service.
Corrective Actions 1.
On June 11,1998, the correct orifice was installed in the AC #6 steam trap.
l i
2.
By September 1,1998, Engineering will develop and document design controls for l
{
non-safety components in the autoclave condensate system.
I d
3.
By December 15,1998, Engineering will develop and implement a system for the control l
l of specific non-safety components. This system will define by what criteria components are l
determined to be in need of additional control and what additional rigor is to be taken during l
l their procurement and installation.
l l
4.
By September 15, 1998, Maintenance will evaluate and revise as necessary procedure l
XP2-GP-GP1070 " Foreign Material Exclusion" to prevent the accumulation of foreign l
l material inside autoclaves.
l 5.
By September 15,1998, Engineering will revise the Autoclave Nuclear Safety Upgrade l
l project pre-operational testing requirements to include a water flow test and boroscopic l
examination of the condensate drain lines on upgraded autoclaves before returning them to l
service.
l Extent of Exposure ofIndividuals to Radiation or Radioactive Materials l,
l There were no exposures to individuals from this incident to radiation or radioactive materials.
Lessons Learned
' Inadequate configuration control of certain non-safety components can result in safety system l
' actuations. Appropriate control must be established for these components to ensure reliable l
3 operation.
l L___________________-_
1 O
GDP 98-2035 Page1of1 Event Report 98-07, Revision 1 List of Commitments 1.
By September 1,1998, Engineering will develop and document design controls for l
non-safety components in the autoclave condensate system.
l 2.
By December 15,1998, Engineering will develop and implement a system for the control l
of specific non-safety components. This system will define by what criteria components are l
I determined to be in need of additional control and what additional rigor is to be taken during l
their procurement and installation.
l 3.
By September 15, 1998, Maintenance will evaluate and revise as necessary procedure l
XP2-GP-GP1070 " Foreign Material Exclusion" to prevent the accumulation of foreign l
material inside autoclaves.
l 4.
By September 15,1998, Engineering will revise the Autoclave Nuclear Safety Upgrade l
project pre-operational testing requirements to include a water flow test and boroscopic l
examination of the condensate drain lines on upgraded autoclaves before returning them to l
service.
l l
l i
I
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - -