ML20149H477
| ML20149H477 | |
| Person / Time | |
|---|---|
| Site: | Framatome ANP Richland |
| Issue date: | 11/02/1994 |
| From: | Wenslawski F NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| To: | |
| Shared Package | |
| ML20149H456 | List: |
| References | |
| 70-1257-94-06, 70-1257-94-6, NUDOCS 9411220224 | |
| Download: ML20149H477 (19) | |
Text
{{#Wiki_filter:4 APPENDIX _B U.S. NUCLEAR REGULATORY COMMISSION REGION IV Inspection Report: 70-1257/94-06 License: SflM-1227 Licensee: Siemens Power Corporation 2101 Horn Rapids Road P.O. Box 130 Richland, Washington facility Name: Siemens Power Corporation Inspection At: Richland, Washington Inspection Conducted: September 26-30 and October 12-14, 1994 Inspectors: F. A. Wenslawski, Chief, Materials Branch C. A. Hooker, Senior Fuel Facility Inspector Approved: N. h ' FraWA~ Wenslawski,' CfiEf7 Materials Branch
//
Date j-@ Inspection Summary Areas Inspected: Special, unannounced inspection to review the circumstances surrounding an event on September 18, 1994, involving the spill of approximately 41 kilograms of low enriched uranium dioxide (UO y) powder from a hammermill hood. The inspection also included a review of a violation of a transportation requirement identified by the State of Washington and open items. Although proprietary information was reviewed during this inspection, such information is not knowingly described in this report. Results:
. The licensee's initial response to the event and reporting to NRC were appropriate and timely. Conditions for restart of operations were adequately evaluated (Section 1.3).
. The inspection determined that the direct cause of the U0, powder spill on September 18, 1994, was improper installation of equipment. The root i cause of the event was the failure of the engineering staff to comply l with engineering change notice procedures for a plant modification that 9411220224 9411gg ;
l {DR ADOCK 07001257 PDR
in turn resulted in the failure to provide systematic training to operators (Section 1.6). Contributory causes were failure to follow operating procedures to ensure equipment was properly secured which resulted in the U0, powder spill from the hood (Section 1.8), and failure of operators to communicate observed system configuration changes to their supervisors (Section 1.7).
- The inspection also found disagreement with portions of the licensee's investigation report (Section 1.8).
Summary of Inspection Findings:
- Violation 70-1257/9406-01 was opened. This violation involved the failure to follow engineering change notice procedures (Section 1.6).
- Violation 70-1257/9406-02 was opened. This violation involved the failure of operators to ensure hood latches were fastened in accordance with established operating procedures (Section 1.8).
- licensee's response to a State identified violation of a transportation requirement was reviewed. Based on the review of the licensee's response to the State of Washington and documentation related to the shipment, the inspectors concluded that adequate corrective actions were taken by the licensee to preclude similar violations. Since this matter was administrative in nature and did not represent a potential safety problem to the public, there was no additional action taken by the NRC (Section 2).
- Inspection Followup Item 70-1257/9104-26 was closed. This item involved the licensee's development of procedural guidance for health physics audits (Section 3). No violation was identified related to this issue.
Attachment:
- Persons Contacted and Exit Meeting 3
e
DETAI_LS 1 REVIEW 0F LICENSEE EVENT (90712, 88015, AND 88020) 1 l 1.1 Event (Report No. 27800) On Monday September 19, 1994, at 5:00 p.m. (PDT), the licensee notified the NRC Operations Officer of an event in accordance with the 24-hour reporting requirements of NRC Bulletin 91-01, Supplement 1, " Reporting Loss of Criticality Safety Controls." The event involved a spill of approximately 41 kilograms (kg) of low enriched uranium out of a powder preparation hood that occurred at 6:00 p.m. (PDT), on September 18, 1994. I Just prior to the formal notification to the NRC operations officer, the licensee made a courtesy notification call to the Region IV Walnut Creek Field Office. The licensee stated that in addition to the 41 kg of uranium powder that spilled from the hood, approximately 320 kg of uranium powder had spilled inside of the hood. The licensee also stated that the spill occur red because a pipe plug had vibrated out of an inspection port on a transition chute to the feed system of the Line 2 hammermill. On September 20, 1994, a conference call was held between the licensee and Region IV, and members of the NRC headquarters Fuel Cycle Licensing and Operations Branches, Division of Fuel Cycle Safety and Safeguards, NMSS, to have a better understanding of the significance of the event. The licensee also discussed the criticality safety controls and criticality safety barriers related to the powder preparation system (PPS). Af ter arrival onsite on September 26, 1994, the inspector discussed details of the event with cognizant licensee management. The licensee informed the inspector that their root cause investigation team had not finalized its report, but expected a final draft of the report to be submitted for licensee management review within the next two days. The licensee also informed the inspector that one of the causes of the event was that a pipe plug (standard 2-inch steel pipe plug) had been inadvertently installed in the inspection j port of Line 2 PPS's bridge breaker transition assembly (BBTA), as opposed to ; an inspection valve that was to be in place. According to the licensee, the l inspection plug had vibrated out of the inspection port during operation of l the PPS, resulting in the uranium dioxide (UO2 ) powder spill. I 1.2 System and Equ_ipment Description The licensee processes low enriched uranium (equal to or less than 5.0 wt.% U-235) for the production of pressurized-water reactor and boiling-water reactor fuel. Following chemical conversion (two chemical process lines, Line 1 and Line 2), ammonium diuranate is heated in a calciner (rotary kiln) l and reduced to dry UO powder. The Line 1 calciner supplies UO, powder to its i respective PPS and is in a separate room from Line 2. The Line 2 calciner supplies powder to two PPSs (Line 2 PPS and Line 3 PPS) which are collocated within the same operating area of the facility. The chemical conversion and
I _4_ powder preparation operations are manned 24 hours per day, 7 days per week by I four rotating 12-hour operating crews (shifts, A-D). The VO, powder is vacuum transferred from the calciners through flexible ) plastic hoses and stainless steel piping to favorable geometry slab hoppers ! for quarantine and verification of moisture content before being transferred to unfavorable geometry blenders. Following blending, the U0, powder is vacuum transferred from the blenders to a PPS feed hopper, then dumped through a BBTA (transfer chute) to the hammermill's feed screw. The hammer mill is the f 4st stage of the powder preparation system. The BBTA is equipped with a , vibrating device to prevent the powder from clogging as it is dumped from the ; feed hopper. Following processing through the PPS, the UD, powder is I subsequently dumped into unfavorable geometry 45-gallon poisoned (neutron ! absorber boron inserts) storage drums for ultimate fuel pellet fabrication. l The Line 2 and Line 3 PPSs encompasses three floor levels of the Line 2 conversion area. Each of these PPS lines is encased with a ventilated lexan enclosure system (hood). The Line 1 PPS extends two levels. The Line 2 hammermill hood is located on a third level mezzanine approximately 15 feet i from the main floor. Within specified limits, when there are changes in the enrichment of uranium to be processed an enrichment clean out (ECO) is performed of the material previously processed. The extent of an ECO depends on the established enrichment control parameters. A major ECO involves the disassembly and vacuuming out the internals of the involved systems, including the removal of the BBTA on the PPSs. l 1.3 General Description of Event and Restart of Operations , At approximately 6:00 p.m. (PDT) on September 18, 1994, while the Line 2 rps was processing UD, powder enriched with 3.78 wt.% U-235-fram the mega blender, a health and safety technician (HST) changing air samples in the area discovered UO, powder spilling from the lower door of the Line 2 hammermill hood onto the mezzanine deck. The HST immediately informed the PPS operator on the main floor area who immediately shut down and locked out the PPS and informed the shift supervisor of the spill. The shift supervisor immediately ordered the shutdown and lockout of the Line 1 and Line 3 PPSs. By telephone, the shift supervisor notified the lead criticality safety specialist (LCSS) at 6:10 p.rr., the general supervisor, Chemical Operations (GSCO) at 6:20 p.m., and the manager, Plant Operations, at 6:25 p.m. at their homes. The LCSS determined that the accumulation of the U0, powder was within the confines of the moderation controlled area; however, the release of the UO, powder outside the confines of the PPS's hood represented a loss of the criticality safety barriers for one guarded contingency (Section 1.9 below) of the parameter being controlled (moderation). The LCSS also determined that less than a critical mass of U0, powder had spilled from the hood and due to other controls in place, a criticality accident was not possible.
Licensee actions taken prior to restart of operations were reviewed. These actions were: (1) the Line 1 and Line 3 PPSs were shut down and locked out, (2) the U0, powder spill outside and inside the hammermill hood was cleaned , up, (3) the ball valves were verified to be installed in the inspection ports of. the BBTAs on all three PPSs and their valve handles wi ed in place to , prevent them from vibrating out, (4) an authorized inspec1 ion plug on the Line 2 feed hopper above the BBTA was wired in place to prevent it from vibrating out, and (5) inspections of the PPSs for powder spills were directed to occur every 30 minutes. Following completion of the above corrective actions, the LCSS, operations manager and the plant manager concluded that system integrity had been reestablished and authorized restart of operations. The PPSs were restarted at approximately 1:00 a.m., on September 19, 1994. The plant manager . subsequently directed that a root cause investigation of the matter occur. l The inspectors concluded that the initial response by the HST and operator was prompt and in accordance with the licensee's procedure for response to off normal events. The affected system was immediately shutdown and notifications ' to the immediate supervisor, the criticality safety component, and upper management were prompt. Management's initial actions were prompt and appeared appropriate. The licensee's notification of the event to NRC was timely. . 1.4 Background for Installation of Ball Valves On May 30, 1993, a non-NRC reportable event occurred in the Line 2 PPS that ' involved approximately 4.5 kg of UO2 powder spilling into the hammer-mill hood when an inspection plug vibrated out of the BBTA. A Criticality Safety , Corrective Action Peport was initiated to review the incident and recommend j corrective actions to prevent reoccurrence. in June 1993, by telephone, the GSCO requested that engineering replace the pipe plugs in the BBTAs with ball valves to prevent UO powder spills from the I hammermill systems. Ball valves are commonly used for sampling and inspection ports on other powder systems. The GSCO also believed that the ball valves would allow easy access for vacuuming out the BBTAs during minor ECOs. 1.5 EnJ1 i neering Change Notice (ECN) Process The inspectors reviewed the requirements, ECN process and circumstances involved with the installation of the ball valves. License Condition No. S-1 of SNM License No.1227 requires that licensed material be used in accordance with the statements, representations, and conditions contained in Part I of the licensee's application dated July 1987 and supplements dated November 12, 1987, and supplements thereto, l I
I l
)
l 1 Section 2.5, " Operating Procedures, Standards and Guides," Part I of the ) license application states, in part, that the licensee is committed to l controlling activities in accordance with Standard Operating Procedures, l Company Standards and Policy Guides. Section 3.5, " Configuration Control," Chapter 3.0, " Nuclear Criticality Safety Standards," of Siemens Safety Manual EMF-30, Revision 12, dated May 1994 (previously Section 3.4 under Revision 11 dated December 1993) states, in part:
"All changes in equipment involving fissile material require review and approval prior to installation and require review and inspection and acceptance prior to the introduction of fissile material. " Changes" includes new installations, as well as modifications of existing designs and removals. Control is exercised through a Work Order and an Engineering Change Notice (ECN) system .. ."
The ECN process has evolved significantly over the past two years. Below is a summary 3f changes that have occurred to the ECN procedure (EMF-858, No.1.13,
" Engineering Change Notice [ECN]"):
- Revisions 10 and 11, issued December 10, 1992, and February 26, 1993, respectively did not require operator training or revision of operating procedures to startup a modification under " engineering control" and did not require operator training to close out an ECN.
- Revision 12 issued March 17, 1993, established the short form ECN for minor modifications. This revision also required operator training and revision of procedures prior to final acceptance of the ECN, but not prior startup.
- Revision 13 issued December 22, 1993, was prompted by an NRC identified weakness (Inspection Report 70-1257/93-11 dated October 17, 1993) involving the lack of a provision to assure that operating proced..res were revised and operator training was performed for modifications being
" operated" while under the concept of " engineering control." Although this concept was not discussed in the ECN procedure, the licensee considered it to be a step in the ECN process where the responsible engineer maintained custody of the ECN until all of the work had been completed before forwarding the ECN to operations. This included minor details of the ECN that had no direct effect on the equipment being operated by operations.
Page 2 of "e standard and short form ECN was revised under Revision 13 to include a specific section for "startup approval" for all modifications requiring ECNs, prior to the introduction of fissile material. The "startup approval" section requires operator training and
operating procedure revision to be completed prior to signing off that the work had been completed.
- Revision 14 issued May 11, 1994, added requirements for ECNs on changes to highly hazardous systems, allowed needed field modifications (documented and approved) identified during testing, and required pressure vessel code review.
= Revision 15 issued September 22, 1994, added computer software modifications to the procedure, clarified the use of "Out-of-Service" tags, and added as-built requirements to associate the criticality safety analysis with the piping and instrue nt diagram drawings on the form.
In conjunction with Revision 13 of the ECN procedure, training specific to the new quirements of the procedure as well as other changes was provided to the p ring staff. A memorandum announcing the training was issued by the n.a n ,
, Electrical and Instruments (person charged with maintaining the ECN proc.,ure), to the manufacturing engineering staff on December 30, 1993. The memorandum stated in part :
"In particular, the ECN procedure,1.13, was changed to incorporate recommendations made by several people to modify the practice previously referred to as " engineering control." Please note that startup approval on page 2 of the ECN form is now required prior to operation of a new or modified system even for the purpo.,e of testing. This is to ensure that operators have procedures, and that Operations and Safety organizations are aware of system operation. ..
The subject training was provided to all of the engineering staff during January 1994. The training related to the ECN procedure also included view graphs of the new changes. Regarding ECNs issued under previous revisions of ; the ECN procedure, the manager, Manufacturing Engineering, stated that although management made a decision to not recall the previously issued ECNs, he expected that the intent of the procedure be followed, specifically, that ; training and revision of operating procedures was expected to occur prior to i startup of the affected systems. The engineering staff was instructed on this , expectation. l l Section 6.0, " Filing" of Procedure EMF-858 No. 1.13, Revision 12, stated, in part:
"Once the work is completed, and the ECN has been signed as completed, as described in 7.0, a copy of the ECN text is marked
" Completed" and is sent to the User / Operator Manager. . . ."
Section 7.2 under " Acceptance of Work Completed" of Procedure EMF-858 No. 1.13, Revision 12, stated, in part that acceptance of work performed I l 1
under a short form ECN is shown by signatures shown on the bottom of Page 2 of the short form ECN. Section 7.4 stated, in part, that the project engineer , shall work with the affected user / operator manager to provide required ; operator training and operating procedure revision. l Section 8.1 under " Acceptance of Work Completed" of procedure EMF-858 No.1.13, Revision 13, required that acceptance of work performed be ; documented by signatures on Page 2 of the ECN form and " Prior to operation, l the Startup Approval signatures must be obtained." f 1.6 Installation of Ball Valves and Traininq r On August 24, 1993, an ECN (No. 2317S) was drafted by the responsible engineer - and approved to replace the pipe plugs in the hammermill inspection ports with , ball valves. At the time the ECN was generated, Revision 12 of the procedure was in effect. Page 2 of the short form ECN had a section of checkoff items ! for work completed with acknowledgement by the responsible person's initials : before being forwarded for final acceptance by the respective operating manager. One of the checkoff items under the work completed section was the operating group supervisor's acknowledgement, by initials, as to whether operator training and operating procedure revision was required. , Additionally, Section 7.4 of the procedure required that "The Project Engineer : shall work with the affected user / operator Manager to provide required operator training, S0P [ Standard Operating Procedures] revision, etc." ; Based on the review of licensee records and discussions with cognizant i licensee personnel, the inspectors noted that on February 14, 1994, ball valves were installed in the inspection ports of the Line 1 and Line 2 BBTAs ; and in the Line 3 BBTA on February 18, 1994. Based on an interview with the pipe-fitter who installed the valves on Line 1 and Line 2, the pipe plugs removed from these systems were subsequently screwed into the threaded outside port of the valves. The installation involved the removal of the inspection plugs and installing 2 inch full-port stainless steel ball valves. A pipe plug installed in the inspection port of the Line 2 feed hopper, directly , above the BBTA, was not planned to be replaced. t The inspectors noted that the " operator training and S0P revised" block on , Page 2 of ECN 2317S was not signed off by operations until September 20, 1994. The responsible engineer stated, after .the valves were installed he did not ; forward the ECN to operations because problems immediately surfaced with the i use of the ball valves. Even prior to installation, due to the size and weight of the new ball valves the pipe fitter discussed potential operator handling problems with the responsible engineer. According to the engineer, ; due to a vendor specification change, the valves purchased were much larger i and heavier than was originally expected. Immediately af ter the valves were , installed, operators begin to complain about their use. The valves had to be
- removed from BBTAs for disassembly during ECOs. The engineer stated that soon >
af ter the installation of the valves he knew that he would have to correct the operators' problem and considered the use of smaller valves. Upon further consideration, he contemplated alternative solutions and held onto the ECN. i l
, e -. -. - __ - ~ - - - _ _ . - _ _ _ _ _ _ _ _ . _
i As a result of the engineers deliberations he held up the ECN and did not l forward it to operations for the training and procedure sign off. The , engineer stated that the project was low priority and he forgot that he was holding the ECN sign off page. He also stated that in retrospect, since the work was completed he should have forwarded the ECN to operations as opposed to keeping it in his files, because any subsequent changes with replacing the , ball valves would require a new ECN. ; Based on discussions with cognizant licensee management personnel, interviews , with operators, and the review of licensee records, the inspectors determined - that there was no systematic training of the operators relative to the ball ' valve installation prior to them operating the PPS. When the work order was , completed, the resr,onsible engineer checked the " work completed section" of ; the ECN form and notified the GSCO by telephone that the valves had been j installed, but since the paperwork did not follow, no training occurred. Subsequent to the ball valve installation, communications relative to the i modification were haphazard. The powder prep operators on shift observed the installation of the ball valves. On February 14 the shift supervisor noted in his log that the Line 1 and Line 2 PPSs were shut down for installation of valves on the BBTA. One operator recalled that the valve installation had been discussed during a shift warmup meeting about the time of the modification. On or about February 19, 1994, operators on one of the shifts : questioned the shift supervisor of the modification. The shift supervisor had ! to call the engineer to find out about the modification. During the onsite inspection on September 26-30, 1994, the general supervisor, Chemical Operations, informed the inspector, that from his interviews of operators all of them were aware that the valves had been installed, but not all of them p were fully aware as to why the ball valves had been installed. . Based on the observations described above, the inspectors determined that ; operators were not provided training on the installation of the ball valves prior to the startup of the PPSs. This was identified as violation of License Condition No. S-1 for failure to comply with the applicable ECN procedure (70-1257/9406-01). On August 26, 1994, the responsible engineer initiated a short form ECN ! (No. 43885) to replace the ball valves with 2-inch pipe plugs equipped with a j locking device to prevent them from vibrating out of the BBTA. The { justification section of the ECN stated " Ball valves were installed to replace l plugs which vibrated loose. These ball valves have proven to be too large and ! cumbersome and pose a safety hazard during ECOs." The ECN was issued on i September 21, 1994, following the drafting department's completion of a drawing of the equipment to be installed. At the time of the inspection, the ECN was on hold pending the licensee's final review of the September 18, 1994, event. l w 4 ,-- , , - F
1.7 Installation of the Pipe Plun S'nce the event was caused by the original pipe plug vibrating loose, the
'nspectors . reviewed the circumstances of how the plug came to replace the ECN installed ball valve.
Based on the review of the licensee's investigation report, licensee records, l and independent interviews of operators and the responsible engineer, the inspectors determined that the operators had experienced problems with the removal of the valve from the BBTA during ECOs due to the weight and cramped quartm s within the hood. These problems surf aced directly af ter the valves were installed. According to the licensee, there was only one operator strong enough to remove the BBTA assembly with the valve attached in preparation for an ECO. The other operators had to unscrew the valve from the BBTA for ECOs. The ball valves weighed approximately 12 pounds. The weight of the BBTA with the ball valve was estimated to be approximately 35 pounds. Item 7 of a memorandum, " Process Improvement Meeting", dated April 14, 1994, from a shif t supervisor to the general supervisor, Chemical Operations, I l stated: "Do away with the large bulky valve installation on the bridge breakers. A ' pinned' cap should work for this application." During the" day shift on August 24, 1994, the ball valve was remaved from the inspection port of the Line 2 BBTA during equipment disassembly for an ECO. l The ball valve was placed on the hammermill inside of the hood. Following the ECO, the operator started to reassemble the equipment but he was interrupted to perform another task. During the night shift on August 25, 1994, an operator in training and a l' qualified operator completed the reassembly of the Line 2 PPS. Although the "Line 2 Powder Prep" log stated, " Prep Re-Assembled," neither operator could recall reassembly of the BBTA. Their only recall of the reassembly was that the operator in training (who was assigned to work on the system) recognized , that he had inadvertently reassembled a screen on the Line 2 PPS's granulator { ' incorrectly. He then asked the qualified operator to inspect the Line 2 granulator and assist in its reassembly. Its the licensee's belief that the operator in training inadvertently installed the plug while reassembling the system. The inspectors agree with the licensee's conclusion. During August 25 through September 10, 1994, the Line 2 PPS was down and no work of any kind was performed on the system. However on September 8, 1994, after changing working shifts (C to D) and being transferred from Line 1 to Line 2 PPS, an operator observed that a pipe plug was installed in the Line 2 BBTA. Since he was familiar with the ball valve being used in the Line 1 BBTA, he called the responsible engineer regarding this change. According to the operator, he was informed that the change was "ok" and that an ECH was coming. The engineer stated that he remembered the conversation differently, in that he informed the operator that there was an ECN in progress to remove the valves and install the pipe plug with a locking device, and he did not say that it was ok to use the plug.
On September 14, 1994, an operator observed the pipe plug in place and was going to question why the plug was there, but on his way to the control room he responded to an operating alarm and forgot about the plug. On September 15, 1994, one operator called the control room to question the use of the plug and was informed that the plug was being used. When interviewed by the inspectors, the control room operator stated that he said the plug could be used because he believed it was being used, but could not recall the reason. On September 16, 1994, during the morning of the day shift, an operator removed the a pipe plug from the Line 2 BBTA and performed a " quick" ECO. He observed the ball valve lying on the bottom of the hammermill hood, but believed there was something wrong with the valve and reinstalled the pipe plug handtight. His reasoning for installing the plug handtight was that he believed that during final " checkoff" by another operator, the BBTA would be reexamined. The final inspection of the system was performed by another . individual. According to statements made by the shift supervisors, they were not made aware that the pipe plug was being used. The inspectors concluded that the engineering department's failure to follow established ECN procedures prevented the training necessary for operators to have a clear understanding of the purpose for using the ball valves (to prevent powder spills). As evidenced by the above described confusion, operators were not aware that the system had undergone a formal design change l and the plugs were not supposed to be in place. The failure to conduct the i required train'ng is considered by the NRC to be the root cause of the powder i spill event. Additionally, it appeared that the operators did not communicate questions concerning system configuration changes to the shift supervisors who . may have prevented use of the pipe plug. l.8 Licensee's Investigation During the inspection on September 26-30, 1994, the inspector concluded that the licensee had not systematically or aggressively pursued interviews with operators for determining the details related to why the pipe plug was being i used as opposed the ball valve. Although the GSCO had interviewed some of the operators, several operators had not been interviewed. The inspector recognized that interviews of personnel while on their days off (two operating crews) was dif ficult, but not impossible as indicated by subsequent telephone interviews of operators by the GSCO with the inspector's participation. Based on discussions with cognizant licensee management, no person or group had been assigned the task of performing organized detailed interviews of operators. On late September 29, 1993, the licensee's root cause investigation team submitted its final draft investigation report to licensee management. The licensee informed the inspector that the report was not as detailed as expected. The report was returned to the investigation team for further
9 review. The inspector was not privileged with the details of the draft report. By facsimile on October 7,1994, the licensee provided the inspector with a copy of the finalized investigation icpert dated October 7,1994. By facsimile on October 11, 1994, the licensee sent a letter to the Region IV Regional Administrator that summarized their analysis of the event and corrective actions taken to resolve the problems that contributed to the event. Under the heading " Root Cause," the licensee's investigation report listed the following causes of the event: (1) the port inspection closure was not adequately designed, (2) inadequate oversight of a technician in training who may have initially installed the pipe plug, (3) the existence of the ball valve was not identified in the Start of Shift Powder Prep Checkoff List, (4) the Checkoff List procedures did not specify that all latches on hood doors should be locked in place when operating equipment, and (5) the failure of an audible alarm which would have alerted operators that loose powder was present in the hammermill hood. Under the heading " Incident Summary," the following are identified as causal factors: (1) the ball valve was so large and heavy that the BBTA could not be operated as intended, (2) the BBTA was reassembled incorrectly, (3) the 1 I wording on the Powder Prep Checkoff List did not specify a ball valve was to be in place, (4) the bottom latches on the hood door were not fastened which allowed the powder to spill onto the floor, and (5) the audible alarm that could indicate when loose powder was present in the hood failed to function properly. l l Under the heading "Other Contributory Causes," the licensee's investigation ! identified the following: (1) the concept of equipment ownership is not clearly understood, (2) the training trigger was not activated because the ECN i' f or installing the ball valves was not f orwarded to operations, (3) miscommunication between the powder prep technician and the responsible engineer regarding the use of the pipe plug, and (4) the lack of expanded operating practices training for operators to avoid mistakes. The investigation team recommended the following corrective actions to prevent reoccurrence:
- Have a powder prep technician, supervisor, and appropriate maintenance person review the final corrective design for the inspection port plug device before installing a new device.
= Reinstruct all experienced conversion employees who take part in on-the-job training that they are responsible for work performed by a trainee under their direction.
l
. Rewrite the operating procedures to specifically identify the ball valve. . Review operating procedures and checklists for accuracy, completenece, and usability. . Simplify existing procedures to cover the specifics of the operation covered. . Repair the annunciator panel such that alarms (both audible and lights) operate reliably.
Other actions taken by the licensee included that all ECNs dating back to March 1993 were reviewed to assure proper closure and that operations training was conducted in accordance with the current ECN procedure. Additionally, all active ECNs were reviewed and brought to the requirements of the current ECN procedure. Special training relative to configuration control was provided to operations and maintenance personnel and refresher training on the current (Revision No.15) ECN procedure was planned for the engineering staff. The investigation also concluded that because the ECN used to install the ball valves was initiated under Revision 12 of the ECN procedure, the ECN procedure was not violated. Operator training and procedure revisions were required to close ECNs, but not required to begin operating the equipment. The findings of the NRC inspectors were not altogether consistent with those of the licensee's investigation team. The following conclusions were discussed with licensee personnel either at the exit meeting on October 13, 1994, or during the inspection:
. The inspectors believed it was a mistake to identify an equipment reliability problem with an audible alarm as a root cause, especially when the primary design and purpose of the equipment was to provide a warning when the ventilation exhaust filter on the hammermill hood was starting to plug and not to detect powder spills. This observation was acknowledged by the licensee.
. Based on the observations described in Sections 1.5 and 1.6 above, the inspectors were not in agreement with the statement:
"The ECN procedure (EMF-858, 1.13, Rev. 12) in effect at the time ECN 2317s was written to install the ball valves, was not violated. Operator training and SOP revisions were required to close out ECN's, but not required to begin operating the equipment."
It was clearly the licensee's purpose, as presented in the training to the Engineering Department in January 1994 (prior to the ball valve installation), that the intent of Revision 13 of the procedure was in 1
~
l effect with issuance of the procedure in December 1993. In any event, Section 7.4 of Re"ision 12 of the procedure was not followed. This section requires the project engineer to work with the affected user / operator to provide required training.
- The inspectors believed that a statement about training appeared to be i inaccurate. The investigation report noted, in part:
"The formal training / communication, which would have been initiated by the December ECil modifications, was not conducted; however, investigation established that personnel on all shif ts knew that the ball valves had been installed and why they were installed."
The inspectors acknowledged that eventually all of the operators became aware that the ball valves had been installed and at least one operator on each shift knew the reason they had been installed. but NRC interviews of operators as well as by the GSCO determined that not all of the operators were aware of why the ball valves had been installed.
. The licensee's investigation did not recognize that the operators failed to surface the use of the pipe plug to the shift supervisors. From independent interviews, it appeared that none of the shift supervisors had been informed or questioned on the use of the pipe plug.
Regarding the latches on the hammermill hood door, Item 4.16 in Section 4.0,
" Radiological, Industrial and Fire Safety," of Operating Procedure EMF-22, flo. P66,813, " Preparation of UD, [ uranium dioxide] Powder as Press Feed,"
Revision 23, dated August 12, 1994, requires verification that equipment has been properly reassembled in accordance with Attachment C of the procedure prior to operating the powder preparation system. Attachment C, " Powder Prep Start of Shift Checkoff List," states for the latches, "All hood doors closed and latched." Although the inspectors noted that the checkoff sheet on September 18, 1994, indicated that the hood doors were closed and latched on the Line 2 PPS, the licensee concluded that the two bottom latches had not been fastened which allowed the powder to spill from the hood. The inspectors also noted during a walkdown of the PPS that there were three different kinds of latches. Some of the latches could be verified as f astened by visual observation. The ones on the Line 2 hammermill could appear to be in the closed position, but had to be physically checked to ensure they were fastened. There are several hood doors and about 40 latches on each PPS. From interviews of operators, the inspectors concluded that they primarily relied on visual observations that the latches are secured. At the exit meeting on September 13, 1994, some licensee personnel had conjectured that the latches on the hammermill hood door may have vibrated loose during operation of the PPS. On October 14, ; 1994, the GSCO informed the inspectors that from his reexamination of the ! latches. they could not have vibrated loose if properly secured. It was i 1 1
i S concluded by the NRC and the licensee that the two bottom latches on the hood door were not fastened prior to operating the PPS on September 18, 1994. The f ailure to secure the latches on the Line 2 hammermill hood was identified as a violation of the License Condition S-1 for failure to comply with the operating procedure (70-1257/9406-02). 1.9 Criticality and Radiological Safety Regarding criticality safety implications of the event, the Line 2 PPS was being fed from the mega blender that contained approximately 3,250 kg of U0, powder enriched with 3.78 wt.% U-235. The average moisture content of the UD, powder was approximately 847 parts per million (a total of 2.752 kg of water in the powder being processed). The inspectors concluded that the release of approximately 320 kg of U0, powder into the hammermill hood and approximately 41 additional kg onto the floor outside of the hood did not constitute an immediate nuclear criticality safety threat. The powder had previously been verified to contain less than 1.0 wt% water (10,000 ppm water). Industry acceptance is that 36 kg UO zpowder enriched to 5.0 wt.% U-235 homogeneously mi.xed with water in the most reactive geometry (a sphere) and fully reflected j (1 foot of water surrounding the sphere) can become critical (K.,, = 1.0). The industry standard-for nuclear criticality safety (ANSI /ANS 8.1, 1983) indicates that 32.8 kg of 5.0 Wt.% U-235 is the subcritical limit value. Tb ' spill of 41 k.g of UO, powder at 3.78 wt.% U-235 was more than a safe mass (45% l of a critical mass) of material, but less than a critical mass. Approximately I 60 kg of U0, at this enrichment would be reeded to approach critical at optimum conditions. Criticality safety in the blenders and PPS is assured by moderation controls. The unfavorable geometry blending and PPS operations are zoned as moderation controlled areas. As stated in NRC Bulletin 91-01, the basic tenet of this principle is that at least two, unlikely, independent, and concurrent changes in a process conditions must occur before a criticality accident is possible. The licensee considers the PPS system as one barrier and lexan hoods as a second barrier for one contingency (accident pathway), and that the pipe walls of systems containing moderating materials (water and chemical lines) in areas adjacent to the moderation zone as the barrier for the second contingency. The spill of the U0, spill outside of hood constituted a loss of the two barriers foi one accident pathway. The barrier for the other accident pathway (integrity of the systems containing moderating materials) remained intact. The inspectors had no disagreement with the licensee's designation of barriers relative to implementing the double contingency principle. The licensee's evaluation of personnel exposure to airborne radioactive material due to the spill indicated that no individual exceeded 80% of the derived air concentration limits specified in 10 CFR 20 Appendix B, Table 1, Column 3. No personnel contaminations occurred nor were there any detectable releases to the environment. Based on a review of the licensee's survey data, no disagreements or radiological safety concerns were identified.
2 TRANSPORTATION (86740) On August 4,1994, SPC shipped in a sea-land container, 88 reject, empty 55 gallon 17-H waste storage drums to Allied Technology Group (ATG) for their use or disposal. ATG noted that a small rusted section on the bottom of one drum had separated from the drum and detected low-level contamination where the drum had sat in the sea land container. ATG notified the State of Washington, Department of Health, Division of Radiation Protection, and a representative of the State visited ATG to review the matter. By letter dated August 15, 1994, the State issued SPC a violation of 49 CFR 173.427(b) for shipping an impaired empty radioactive waste container. A copy of the letter was provided to the NRC Region IV office. The licensee's response to the violation, letter dated September 6,1994, stated that an error had been made in the Bill of Lading and Radioactive Materials Shipping. The drums were shipped as individual packages as opposed to the sea land container being the package containing empty radioactive containers. Siemens Power Corporation informed the State that all future shipments will list the sea land container as the shipping package. By letter dated September 21, 1994, the State acknowledged that SPC's corrective action was adequate to prevent recurrence of this type of problem in the future. The inspectors had no concerns relative to the licensee's actions. 3 FOLLOWUP ON OPEN ITEMS (92701) (Closed) Inspection Followup Item 70-1257/9104-26: No Procedural Guidance for Health Physics Audits Based on the review of a new Administrative Procedures Document EMF-P81,004, Revision 0, dated September 21, 1994, " Safety, Security and Licensing Program for Environmental, Safety and Health Audits and Inspections" and lower tier inspection guides, the inspector concluded that the licensee had adequately addressed this Operational Safety Assessment team inspection identified weakness. The closecut of this item completes all of the inspection followup items from the Operational Safety Assessment team inspection conducted in October and November 1992 (Inspection Report 70-1257/91-04 dated February 6, 1992). No violation was identified from this issue. (-
ATTACHMENT l i 1 Pers_clLs Contacted 1.1 Licens;e Personnel f R. C. Arnn, Control Room Operator ;
*B. F. Bentley, Manager, Plant Operations ,
T. R. Blair, Manger, Plant Projects and Scheduling , R. K. Burklin, Health Physicist : D. C. Burris, Powder Prep Technician T. P. Collins, Powder Prep Technician
*J. B. Edgar, Staff Engineer, Licensing r D. J. Deschane, Powder Prep Technician *B. N. Femreite, Plant Manager, , *R. L. Feuerbacher, Manager, Materials and Scheduling +
E. L. Foster, Supervisor, Radiological Safety T. W. Good, Powder Prep Technician S. D. Haug, Shift Supervisor, Chemical Operations
*J. W. Helton, Manager, Plant Engineering W. G. Keith, Manger, Maintenance Engineering J. J. Payne, Shift Supervisor, Chemical Operations *J. H. Phillips, General Supervisor, Chemical Operations , *W. E. Niemuth, Acting Manager, Manufacturing Engineering l *L. J. Maas, Manager, Regulatory Compliance *C. D. Manning, Criticality Safety Specialist ,
M. A. Moberg, Shift Supervisor, Chemical Operations l J. Morales, Shift Supervisor l
*T. C. Probasco, Supervisor, Safety i S. W. Russell, Powder Prep Technician .
D. L. Smart, Pipe Fitter i D. A. Stallcop, Powder Prep Technician , M. K. Valentine, Manager, Manufacturing Engineering l
*R. E. Vaughan, Manager, Safety, Security and Licensing l L. O. Washington, Process Engineer !
P. M. Wees, Project Engineer j 1.2 NRC Personnel ; R. C. Pierson, Chief, Fuel Cycle Licensing Branch, NMSS f In addition to the personnel listed above, the inspectors contacted other personnel during this inspection, j
- Denotes personnel present at the exit briefing conducted on October 13, 1994. I 2 Exit Meeting On October 13, 1994, the inspectors met with the licensee representatives to discuss the scope and findings of the inspection. lhe licensee was informed of the observations described in the report.
7:40= = u s xaaa ar c,aca, coww ss" SPECIFY ALL THAT APPLY Page{ oil IFS Da a E try orm '
* * * ~ " " " " ~
tee C4'enee F/N) Eme' Log ute. R ea ey Jw w pg lg , pfg 16ss'erer kupew (it,s ops.an ti _,_. L E A res (irs opeen s, lat 0% .I.!/_ E me' LE R e:=, SeteName Sr & ht b t1>A. [Aut o L o'u' P N Repat Trarsm.nal Dale .d._Ib_ff Cear F% af p he Dccw % amo inae 4 5 opion e. gg bieterials Only Report NBR Docket NBR Lkenee NBR nespmmorg coao F/kbW A OMOO G_._ o700/pS7 f N M-/22 y Report Ena Dais /f.J /.IJp B A*9'on Update? (Y/N): .LL._ Opened tR'LER'P21 LOG'if 5 Numbef.
"* Sequence NBR: n/ Item Type: N / O " Severity: _ ul _ " Supplement: "TC~,_
Status 'UPD LH *Proj. Closeout
- Actual Closeout to CFR Uc n Co A 0 Ut)?]th _1.__J_
m Down (T ~ l yfC e _ _ __s__.i_ ___/_ _/_ _ c __j_ _ t., _ _ j_ ,j_._. Tstle._ f~ct e la - < 'T o f
- N i 'l 4 O N <oYor Ye s o H r y iss charecw w an Closeout Org '/'d.2O 'Closecut EMP:CllQ *Ccnut EMP;
- Procedure: 'Funcil Area: .
'Cause CD: _._ "EA Numtet "NOWNNC issue Date: / /_ _,
Text: b n Te <
- Y & / re n'*r 14 ("snc/rIk' S ~ /, 1/ .c lo c .e o., < c .<. -(m if,/ss $
~D 7'r su n op.r eTon o n a. m ed,' Sic =% *t~o M powdv p<wsuTay <; vc?remt on
- f. 6 ovv /%v4IY,/Sf4f,ac v-suon./ bv f/e la ysu r' fuy ue rnf f La
- o ns O f s [oc-*/H4.
Update? (YIN): .h/_ Opened IR'LER'P21 LOG' IFS Number.
"* Sequence NBR: 0;t _ ltem Type: w n " Severity: I u_ " Supplement: . i2Z;. ~
Status v ...r ,e', onor
*UPD L9 *Prol Closeout ' Actual Closeaut 10 CFR~
0 Ucense Coext Tle Dom'n A csp _It 9[ . _ i__.J _ $'l W'S E _._J_._I_ ___l-_J-C _.) J __ _ )_._)_ _ - Tst'e fez t/u r < 'l e _-{p//.?w .5 Tas, / cud (19snTY sy 6 et . 4 e.s _ gss et,e,scter Mtm Close>ut Org #/J2.0 *Ciosecut EMP. C //U
- Contact EMP. ' Procedure: 'Funct! Area: . __ __
_, 'Cause CD. . _ . - "E A Numbet _"NOWNNC lssue Date: / /_ Te11: Ovr75 # Y Co Lucos-e & /r Uw .S'~/> < s-> 3 H es75r der.fw / C o c usun H m37%o)A s/s24eCsn th -.c bp> - v $$$# d / 7% -e ), m s__2,. Nonnwm///' Uj n .s -fnAmf k l& "f1r 4 s' W5 ws < i3,31- ~ k -l es H s,u w-8 hv, - T1 - s Ths., devel/)m
, y k r& e aAss optxr i r.o.
semey. sumwrent. % Novwec cn, am.caue for v.otaws E A Nant.or onS eWoth for EscWed Enfmment furs
*" Svpencs NBA s not a;po'.4 la o d eire:ated P21. L E R. or rrrr&icket relate.a co rs ITEMS CONTINUED? (Y/N):
I'M Pcg21of 2:: Asport NBR O_.f. fCO4 Docket NBR 6700/2 e 7 IFS Data Entry Form (continued) sne< e -, u o,.s Update'(Y/N) _f_ Opened IRtER'P21 LOG"FS Number: 9/d 4/ -J/
'" Sequence NBR: Item Type: " Severity: " Supplement: ,
Status v. ..,..i. oow
'UPD t/R 'Prol Closeout ' Actual Cfoseout 10CFR UcEse Cond. 11e Down A I I hi I
_i_ 1_ - i._J-C / I
) -) _ ..
Titte: (55 charactw *ctn) Clos +out Org W 3 0
- Closeout EMP.C U ' Contact EMP: ' Procedure: 'Funct! Area: .
'Cause CD. _ < - "E A Numbon "NOV/NNC issue Date. / /
Tert: Update 9 (Y/N). Opened IRtER'P21 LOG'lFS Numbot
"' Sequence NBR: .__._ ttem Type: " Severity: " Supplement: . _ _ _ .
Status u . i .. .. : . oni,
'UPD 1/R ' Pro {. Closeout
- Actual Closeout 10CFR Ucense Cond. T1e Down A _f_ )_ _ )_ J_
8 C Tite: _)._.)__ ___I_, )_ _ )
) _J
)_
I (55 emade wth) Clos +out Org *Cicseout EMP. ' Contact EMP; ' Procedure: 'Funct! Area: ,
'Cause CD: , _ _ , _ . _ "EA Numbet "NOVINNC tasue Date: / /
Tert: Update? (Y/N): Opened IRtER'P21 LOG /lFS Numbet
"" Sequence NBR: Item Type: " Severity: " Supplement: ,
Status *UPD L'R 'Prol. Closeout ' Actual Closeout a u.i ....: . onir 10 CFR Ucense Cond. T1e Down A _/.,,,_ ,,J._ ,__/_, J__ 8 .___j _J__ _ )_.__/_ C _t_ ) _ __J _._J t Tate: (55 v. ara:tv wc:h) Closeout 0,g- toseout EMD. ' Contact EMP: ' Procedure. *Funct Area: ,
*Ceuse CD: _ . _ _ "E A Nr. bet "NOV'NNC issue Date: I I Text I
._.}}