ML20212A192
| ML20212A192 | |
| Person / Time | |
|---|---|
| Site: | Framatome ANP Richland |
| Issue date: | 09/10/1999 |
| From: | Maas L SIEMENS POWER CORP. (FORMERLY SIEMENS NUCLEAR POWER |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| IEB-91-001, LJM:99:102, NUDOCS 9909160107 | |
| Download: ML20212A192 (4) | |
Text
O y
SIEMENS.
September 10,.1999 LJM:99:102 U.S. Nuclear Regulatory Commission Document Control Desk Washington, D.C. ' 20555 Gentlemen:
Follow-up to NRC Operations Center Report No. 36030;Bulletin 91-01 Report on Over-Batching of Waste Drums Due to inaccurate Determinations of Uranium Contents (License SNM-1227; Docket 70 1257)
On August 13,1999, Siemens Power Corporation (SPC) reported a condition which met the 24-hour reporting criteria of NRC Bulletin 91-01. The condition involved the discovery that SPC's laboratory and non-destructive assay (NDA) uranium determinations were under-estimating the true uranium content in a particular waste stream generated at SPC, namely a filter cake material generated at a vacuum drum filter in SPC's Engineering Laboratory Operations (ELO)
Building. This in turn allowed two drums to be placed on SPC's waste storage pad that exceeded SPC's internal criticality safety mass control limits for individual waste drums (214.2 grams U-235 per container) in a three tier array. SPC internal procedures require a'30-day
~
follow-up report to all Bulletin 91-01 reportable conditions or events. This letter fulfills this requirement.
Backaround SPC's Mop Powder Dissolver Facility, located in the ELO Building basement, dissolves impure urania powders and then uses a vacuum drum filter to separate the resulting uranium-containing solution from the residual undissolved inert solids. These inert solids, mixed with the particulate filter media u6ized on the drum filter, are " shaved" off the rotating drum filter and collected in four-gallon.)lastic buckets lined with plastic bags. The buckets are sampled and then queued one foot epart. When acceptable laboratory sample results are received, the contents of the buckets, i.e. the bags of filter cake shavings, are transferred to a 55 gallon
~
drum. These drums are transferred to SPC's waste handling area in the Uranium Dioxide (UO )
2 Building, where trained waste handlers conduct final waste packaging operations (consolidation, compaction, etc.). In conjunction with these waste packaging activities, the drums are verified
.via NDA to contain less than 100 grams U-235 (a waste disposal limit). They can then be i
moved to SPC's waste storage pad, where they are allowed to be stacked three high. The mop fl
' powder dissclver facility has been in operation'at SPC for approximately two years (November 7
dFCM' b
' Siemens Power Corporation i
2101 Hom Rapids Road Tel:
(509) 375-8100 l
RicNand, WA 99352 Fw mnoi u r.8402 l
9909160107.990910 l
PDR ADOCK 07001257 L
C PDR
USNRC Docum:nt Control D:sk LJM:99:102 September 10,1999 Page 2 1997 startup) and SPC had accumulated thirty-seven (37) drums of filter cake from this facility on its waste storage pad.
Incident Descriotion and immediate Corrective Actions On August 10,1999, the supervisor from SPC's UO Laboratory contacted SPC's Lead 2
Criticality Safety Specialist with preliminary evidence that the laboratory method normally used to determine the uranium content of the filter cake shavings may be yielding erroneously low results. The discrepancy, estimated at that time at approximately thirty (30) percent, did not raise immediate criticality safety concerns, nonetheless the Criticality Safety Specialist requested that the laboratory continue its investigation on a priority basis and update him as appropriate. As a precautionary measure, the Lab Supervisor put a hold on reporting percent uranium results on these filter cake solids pending results of further investigation.
On August 12,1999, after follow-on laboratory investigations that involved varying the sample dissolution protocol by which the filter cake samples were prepped for analysis, the laboratory informed the Lead Criticality Safety Specialist that the laboratory results appeared to be low by a larger margin than previously estimated, possibly by a factor of up to three (3). Based on this information, the Lead Criticality Safety Specialist ordered an immediate shutdown of the entire Mop Powder Dissolver Process. In addition, SPC's Traffic and Warehousing organization was directed to reduce the exining three tier filter cake waste drum storage array to a single tier array. As previously noted, creation of additional 55 gallon waste drums had already,been stopped by the laboratory's August 10 decision to not report any more analytical results on the four-gallon filter cake accumulation buckets.
A number of reporting and corrective action tracking activities were also initiated. Condition Report 7866 was initiated by the Laboratory Supervisor to address remedial / corrective actions under SPC's Quality Assurance Program. The Lead Criticality Safety Specialist initiated Criticality Safety Corrective Action Request (CSCAR)99-045 to document the incident and track corrective actions relative to criticality safety. And as previously mentioned, SPC notified the NRC Operations Center on August 13,1999 under the requirements of NRC Bulletin 91-01.
SPC convened a formal incident investigation Board (llB) to investigate this occurrence on Monday, August 16,1999. The llB, chaired by SPC's Manager, Safety, initiated a number of near-term investigative actions to define the safety significance of this condition. Those actions included a rigorous sampling campaign of all thirty-seven (37) drums of the mop powder-derived filter cake material, analysis of the samples utilizing an aggressive sample preparation protocol that assured placement of all of the uranium into solution, and comparison of the data to previously generated analyticallaboratory and NDA data for those drums. In addition to confirming the low bias in the laboratory data, these comparisons revealed that the NDA data for this material were low by an average factor of 1.7. The net result of these systematic under-estimations of the uranium content in the filter cake material was the placement of two drums on SPC's waste storage pad that contained uranium in excess of SPC's internal criticality safety controllimit of 214.2 grams U-235 per drum (by margins of 32 and 15 percent, respectively). However, conditions did not exceed the criticality safety limit of 642.79 U-235
USNRC Document Control Dask LJM:99:102 September 10,1999 Page 3 per vertical stack and therefore did not approach those conditions required to cause an actual
- nuclear criticality event, incident Causes The llB conducted a comprehensive investigation of this occurrence, including an assessment of root cause utilizing the Taproot
- Incident Investigation System, coupled with the Root Cause Tree. The primary root cause was determined by the llB to be less than adequate qualification processes for the radiological characterization (lab analysis and/or NDA) of solid waste streams, particularly new cr changing waste streams. In the laboratory analysis area, this was manifested by the application of an inadequate sample preparation protocol to a new and unique waste stream. With respect to NDA, this lack of proper process qualification resulted in the utilization of NDA standards that may not have adequately matched the waste form with respect to geometry and/or material matrix.
The llB also identified the scope of the Mop Powder Dissolver Criticality Safety Analysis (CSA) as a potential root cause of this occurrence. Specifically, the CSA stopped with the filter cake shavings being placed into the four gallon containers and being stored. Placement of the bucket contents into 55 gallon drums was covered in a separate waste handling CSA. In retrospect, the llB felt that a cingle CSA covering the entire process may have resulted in a more robust criticality safety barrier relative to the bucket-to-drum transfer.
Follow-uo Corrective Actions Based on the findings of the 118, corrective actions have been identified to address the specific issues relative to the characterization of filter cake waste generated at the Mop Powder Dissolver Facility as well as to address generic issues related to the radiological characterization of other solid waste streams generated on-site. These corrective actions have been assigned to SPC's Analytical Services, Manufacturing Engineering, Plant Operations, and Safety, Security and Licensing organizations.
Specifically with regards to the Mop Powder Dissolver System, SPC is pursuing a formal process qualification of all methodology that will be utilized to provide radiological characterization of the mop powder-derived filter cake solids, as needed to support SPC's j
criticality safety and waste storage programs. These methods, which may include analytical laboratory, NDA, or other instrumental techniques, will be applied such that the collection, interim storage, and final waste packaging of this filter cake waste stream are covered by robust and fully qualified criticality barriers on at least a double contingency basis. SPC's criticality safety analysis and associated criticality safety specification will be revised to fully reflect the imposition of this fully qualified suite of measurement-based criticality safety barriers. Furthermore, all Plant Operations Standard Operating Procedures (SOPS) will be revised to reflect updated criticality safety limits and controls and required pre-startup operator I
training will be provided and documented.
The Mop Powder Dissolver Facility will not be operated in a mode that produces waste filter cake shavings until the corrective actions described above can be completed. Re-start of the l
l t
1 I
USNRC Documsnt Control Desk LJM:99:102
)
September 10,1999 Page 4 facility will be addressed via a formal SPC Startup Council per the protocol of SPC's Manufacturing Engineering organization. The Startup Council will determine whether initially a phased startup (initial accumulation in buckets only) or full process startup (accumulation in buckets with transfer to drums) will be authorized.
The llB also addressed the generic implications of this occurrence, i.e. the implications relative
)
to the radiological characterizations (laboratory, NDA) of other solid-phase waste streams generated on site. A number of initiatives / corrective actions are being pursued, including:
Implementation of a formal qualification process relative to the application / acceptance of sampling and analytical techniques for new solid phase waste streams; i
evaluation of chemical test method qualification status of all current sample types in e
which the current sample preparation protocol does not achieve total dissolution of sample material; craation of new laboratory operator aids and sample request forms that facilitate proper l
application of appropriate sample preparation / analytical methodologies to routine and non-routine (new, special, etc) samples submitted to the laboratory; implementation of a formal qualification process to assure correct NDA applications to e
current and new/ modified solid waste streams; modification of applicable SOPS as needed to assure that conditions conducive to e
accurate NDA measurements are achieved and maintained (proper waste packaging, appropriate waste geometry control, optimal assay operating parameters, etc) and that NDA results are compared to other analytical data, as available; review of the adequacy of CSA coverage for other solid waste streams which rely on e
laboratory determinations of uranium content to assure that problems similar to those encountered for the mop powder dissolver filter cake solids do not exist; i
placement of all off-site waste shipments on hold until the methodologies utilized to e
radiologically characterize the particular waste stream have been formally qualified.
If you have question regarding this occurrence or about the actions SPC has taken or is pursuing as a result of this occurrence, please contact me on 509-375-8537.
Ver/ truly yours, a
n OM %-
Loren J. Maas, Manager Regulatory Compliance
/las cc: E. W. Merschoff, NRC Region IV, Arlington, TX J