ML20203H659
| ML20203H659 | |
| Person / Time | |
|---|---|
| Site: | Portsmouth Gaseous Diffusion Plant |
| Issue date: | 02/26/1998 |
| From: | Morgan J UNITED STATES ENRICHMENT CORP. (USEC) |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| GDP-98-2009, NUDOCS 9803030283 | |
| Download: ML20203H659 (8) | |
Text
{{#Wiki_filter:_ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - t, CSEC A Global Energy Company Fcbruary 26,1998 GDP 98 2009 United States Nuclear Regulatory Commission Attention: Document Control Desk Washington, D.C. 20555 001 Portsmouth Gneous Diffusion Plant (PORTS) Docket No. 76 7002 Event Report 98-02 The enclosed written Event Report for an occurrence involving the actuation of the Calciner Discharge Collector Probe Detection System in the X-705 Duilding at the Portsmouth Gaseous Diffusion Plam is being submitted as a voluntary report. Encle=ure 2 is a list of commitments made in the report. Should you require additional information regarding this event, please contact Scott Scholl at (740)S97 2373. Sincerely, 2011 k James D. Morgan d Acting General Manager Portsmouth Gaseous Diffusion Plant Enclosures cc: NRC Region 111 Office NRC Resident inspector - PORTS I 9003030203 900226 {DR ADOCK_0700 0 3 _pf f f! ! !!!, ! ! ! P.O. Ikix 804, Portsmouth,011 45661 Telephone 614 897-2255 I:as 614 897-2644 http://www.uwe.com Omces in 1.ivermore, CA Padteah, KY Portsmouth. OH Washington, DC
GDP 98 2009 page 1 of 6 Event Repor* 98 02 llackground Uranyl nitrate solution from the Uranium Recovery process is converted to uranium oxide by electric calciners. The Uranium Recovery post evaporator concentrates the uranyl nitrete to a specific gravity of 1.4 to 1.5. The concentrated solution is stored in one of three storage eclumns (reference Figure 1,"A","11", and "C" Loop). Each column has an overflow pipe and a high level alami which is used to indicate when the column is full. A valving nianifold is used to direct the solution from a particular column to either #1 calciner t #2 calciner eed system. To feed #2 calciner, the solution r then goes through the feed control valve (FSV.1606) to the feed metering pump (P C2F). In the event of a 1.ntem alarm, the feed valve automatically closes and the feed pump shuts down. The feed metering pump is a positive displacement pump in which a piston drives a diaphragm pumping a consistent volume for each stroke for a particular setting. Double check valves on the pump inlet and outlet prevent solution from flowing backwards, if the inlet pressure is higher than the outlet, there are no mechanical obstructions to prevent continuous flow through the pump. To maintain positive pressure on the pump outlet, the piping is looped above the level of the storage column high level alarm. There is an overflow vent pipe at the top of the loop (at a height of 24 feet above the suction piping), which empt:,:s into an overflow glass, to serve as a vent and as a siphon break. The feed solution passes through several manual valves, through a sight glass flow indicator, and imo a packed feed sembber column. De feed solution is preheated by trickling over metal rings in the scrubber column prior to entering the top of the calciner. The scrubber also washes particulate matter from the hot gas exiting the top of the calciner. The feed decomposes by heating to approximately 1400*F in the presence of air within the tube. Nitrous oxides and water vapor are exhausted from the top of the calciner through the feed scrubber column. Three metal milling reds inside the calciner tube grind the uranium oxide into a fine powder, which rolls down the inclined tube into the discharge chute. Two level probes monitor the discharge chute to assure there is no excess accumulation. This is addressed in TSR 2.6.3.2 which states, "If the uranium oxide leaving the calciner would back up into the calciner discharge throat which is an unfavorable geometry, a criticality could occur if the U 235 assay and the amount of moderation were high enough." The operator can mon.or the probes and the interior of the calciner through a sight glass on the chute.
l GDp 98 2009 Page 2 of 6 Event Report 98-02 ne powder oxide product falls through the discharge pipe valve (C2DV) into a product can. The can is monitored by weight and by a level probe to assure it does not overfill and allow oxide to overflow onto the floor of the glovebox. Description of Event On January 27,1998, at 1118 hours, X 705 #2 calciner was in Mode 111, processing uranyl nitrate solution to uranium oxide, when an audible and visual alann for high level in the calciner discharge chute was received. Operators responding to the alamt found that the safety system had ac.uated due to moisture, rather than to an accumulation of uranium oxMe. Operators inspected the discharge chute through the sQht glass and no uranium was apparent on the probes. There was moisture visible on the pros:s and on the eight glass. Operators changed the product can and observed no evidence of moisture or weight gain attributable to moisture. Since the actuation was caused by steam contacting the probes and not by uranium oxide building up in the discharge chute, this safety system actuation was determined to be invalid and not reportable in accordance with the Safety Analysis Report (SAR), Section 6.9, Table 6.91, J (2), llowever, a voluntary event report is being submitted to describe the system conditions that led to this invalid safety system actuation, ne system had been operating normally until about i115 hours, when the "C" loop column became empty. The operator shut off the feed pump and performed the necessary valving to start feeding from "B" loop storage, in accordance with procedure XP4-CU Clll200," Operation of the Electric Calciner", the operator started the feed pump and checked the flow indicator sight gage for proper flow. 'lhe operator observed an unusual amount of flow through the sight gage, which showed a steady stream rather than the usual spurt or spray from the metering pump. The operator turned off the feed pump and closed the manual feed block valve. The operator observed continued flow through the sight ghtss and closed the sight glass inlet valve, isolating further feed to the packed feed column. At 1118 hours, the operator heard the audible ce'.ner alarm sound and went to the control room to check the control panel, ne alami indicators Wowed a loss of calciner vacuum due to floodico of the packed feed column and a calciner discharge chute high level probe actuation. The operators followed the alarm response procedure XP4 CU AR8950.C8 "Calciner Annunciator Panel. Window C8," verified that there was no material contacting the high level prows, verified that the safety
GDP 98 2009 Page 3 cf 6 Event Report 98-02 system actuation had closed the feed valve FSV 1606 and tumed off pump P C2F. and replaced the oxide can. The operators observed moisture on the probes and droplets and streaks of material on the discharge chute sight glass. The operators did not observe any moisture in the oxide can. The First Line hianager (FLhi) responded at 1125 hours. FLhi verified that the system was in a stable condition and that all required immediate actions had been accomplished. FLhi reported the alarm to the Plant Shin Superintendent (PSS) and Nuclear Criticality Safety (NCS). The calciner continued to process the residual feed >mution on heat until it was all converted to oxide in accorduce with calciner procedure XP4 CU Clll200, section 8.10 (Standby Operation). At 1230 hours the last of the oxide exited the calciner. The #2 Calciner was allowed to cool down and was then tagged out of service and declared inoperable. The safety systems functioned as designed and there was no release of hazardous material or radiological exposures associated with this event. Cause of Event The direct cause of the safety system actuation was blockage of the calciner exhaust vemilation caused by concentrate solution flooding the feed scrubber column. With the exhaust flow blocked the water evaporating from the feed solution filled the interior of the calciner with steam, ne steam began to condense in the discharge chute and on the level probes, actuating the safety system. The root cause for overfeeding the feed scrubber column was a lack of procedtral detail for controlling head pressure across the metering pump. Engineering analysis of the incident indicated that the inlet head pressure excertd the outlet pressure allowing solution to push through the pump. The inlet head pressure depends on the level in the concentrate storage columns and ths density of the solution. If the level of the solution in the concentrate storage column is above the height of the feed line loop or if the density of the solution in the column is greater than the density in the feed line, it is possible to start a continuous flow of solution to the feed scrubber column which will continue until the feed pump inlet head pressure is lower than the outlet head. Aner the solution was valved oft following the incident, the solution in "B" storage column was measured and found to be 7 inches above the high level alann point (approximately 10 inches below the level of the feed line loop). The solution density was 1.430, while there are indicatiem from the results of the previous run that the "C" colunm solution still in the feed line was a lower density (no "C" column solution remains for analysis), it is believed that a combination of a high solution level in the storage column
GDP 98 2009 Page 4 of 6 Event Report 98 02 ami the difference in density of the feed solutions caused the flooding of the packed feed scrubber column. 1 Procedure Xp4 CU-Clll200 does not restrict the solution height in the storage columns. Procedure XP4 CU Cll 2803,"Operaticn of A, B, and C Loops in Recovery", states "To prevent overfill of concentrate storage, monitor concentrate storage while filling", but does not mention the high level alarm or the storage column sight glass level readings. The procedures do not address differences in solution concentration with feed from different columns. For a 24 foot piping height, there is n difTerence in effective head height of over 20 inches between a 1.4 density solution on the pump outlet end a 1.5 density on the pump inlet. Since the top of the feed loop is 17.5 inches above the e lumn high level alarm, density difTerences alone could i cause increased fee.1 solution flow. NJ t ilciners will remain out of service until applicable procedures are revised. A contributing cause was the des!gn of the #2 calciner piping. Measurements indicate that the . highest point of the everhead loop was not at the vent, but was at the bend of the pipe dropping down to the feed scrubber. When the feed line is full, this configuration allows the vent to be partially filled by solution. When this occurs, the exhaust on the feed scrubber causes a vacuum on the scrubber, siphoning a stream of solution into the scrubber instead of the solution being metered by the pump. The siphoning will continue until the solution clears the vent and breaks the siphon. The
- 2 calciner will remain out of service until piping modifications are complete, Corrective Actions 1.
By April 30,1998, Operations will modify the calciner procedure (XP4 CU-Cill200) to include verification of the Concentrate Storage Column level end appropriate feed pump setting. 2. By April 30,1998 Operations will modify the Recovery 1.oop procedure (Xp4 CU-Cl12803) to enhance control of the solution specific gravity prior to transfer to the Concentrate Storage Columns. 3. By April 30,1998, Engineering will complete a design to ensure the #2 calciner feed line vent is above the highest point on the line and that the line has the proper slope. z'm m- -Y--r.-e.ww-- _y .,.yy w r py,, 9-.m-er- -,---+y, .-m,-qr,y-.7_- o-- r--- 9 e -+--
4 Enclosure ODP 98 2009 Page5of6 Event Report 98-02 4. By May 30,1998 Operations will assure piping modifications have been completed in accordance with the pproved design. Estent cf Exposure ofIndividuals to Radiation or Radioactive Materials There were no exposures to individuals from this incident to radistion or radioactive materials. Lessons Learned A combination of high storage column level and dif ferential solution densities caused the flooding of the feed system resulting in the safety system actuation. There was no unsafe mass accumulation or configuration. The actuation was due to moisture accumulating on the probes during a system upset. v 9
o ODP 98 2009 Page 6 of 6 Event Report 98 02 Figure 1 Calciner Flow Diagram s L% i e 5 ig.;................. =
- .:: p r ' -l
+ li-. i 7 7
- s l-
! !,b,, - aq gt - g; e \\gl 1.J. o= W g j 5 ll i i a y o= 3.!a
- _J 3--. 3 y
2: Y j y; ~--h 3 Il .d M 2 ~~~j I Nnmoo i T smuSCH IN3A j ]k. ]s= g <!N r.... k l' I 2 y3 5.....s C= u' ,) g g! .. 2( l l I 5 7 nuouw f j 5-- NWrnOOO33d 03MOYd l jk 5 S A 6 y n a gI p 'g ssvioMou W3Ao dl 3 C 3 j U g1yoggi 5 mu 7.. + D e t w g r f ]5 )...... el u bhgit-e g o e tm 5 n 1g e 3 m 9 B u n n u: nn u= nn d ee o o o o m m <b b b bb b u b bb bb -w-u u- .-w-u
- : - u --
-u--w I.:: m c-{S o <0-k, h k
- m
- =
~. ~~
,o ODP 98 2009 Page1of 1 Event Report 9f. 02 List of Commitmats 1. By April 30,1998, Operations will modify the calciner procedure (XP4-CU-Cill200) to include verification of the Concentrate Storage Column level and appropriate feed pump setting. 2. Ily April 30,1998, Operations will modify the Recovery Loop procedure (XP4 CU Cll2803) to enhance control of the solution specific gravity prior to transfer to the Con:entrate Storage Columns. 3. Ily April 30,1998, Engineering will complete a design to ensure the #2 calciner feed line vent is above the highest point on the line and that the line has the proper slope. 4. By May 30,1998. Operations will assure piping modifications have been completed in accordance with the approved design. = = ________________.-----__----w}}