ML20248K001

Insp Rept 70-1257/98-02 on 980413-17 & 980504-08.No Violations Noted.Major Areas Inspected:Operational Safety, Radioactive Effluents & Environ Protection

ML20248K001
Person / Time
Site:	Framatome ANP Richland
Issue date:	06/02/1998
From:	NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
To:
Shared Package
ML20248J985	List: ML20248J983 ML20248K001
References
70-1257-98-02, 70-1257-98-2, NUDOCS 9806090390
Download: ML20248K001 (23)

Text

_ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ - _ _ - _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ ___-__ ____ _ _.___ _ ____-_-__ _ _ - _ - _ __ _ _

e ENCLOSURE U.S. NUCLEAR REGULATORY COMMISSION REGION IV Docket No.: 70-1257 License No.: SNM-1227 Report No.: 70-1257/98-02 Licensee: Siemens Power Corporation Facility: Siemens Power Corporation Location: Richland, Washington Dates: April 13-17 and May 4-8,1998 Inspectors: C. A. Hooker, Senior Fuel Facility inspector, Region IV D. A. Seymour, Senior Fuel Facility inspector, Region 11 W. L Britz, Radiation Specialist, Region IV Approved By: Frank A. Wenslawski, Chief J Materials Branch

Attachment:

Supplemental Inspection Information l

WO6090390 990602 PDR ADOCK 07001257 C pg

a. >

l EXECUTIVE

SUMMARY

l i Siemens Power Corporation NRC inspection Report 70-1257/98-02 This onsite inspection included a review of radioactive effluents and environmental protection.

During the course of the inspection, a fire occurred in the waste processing area of the UO, Building, which resulted in the licenses activating its Emergency Plan (EP) and declaring an Alert. Much of the insoectors' activities were diverted to the response to the fire and the

inspection was extended to a subsequent onsite visit which focused on the review of the licensee's radic!ogical assessments, the status of the root cause investigation, and the corrective actions re!ated to the fire. The inspection also included followup on open items from previous inspections.

Operational Safetv

. Although scmo momentary confusion developed as a result of fire alarm testing occurring coincident at the time of the fire, the matter of a real fire was promptly identified and the licensee promptly initiated actions outlined in its EP. The Plant Emergency Director (PED) appropriately classified the event as an Alert. The accountability of onsite personnel and notification of offsite agencies appeared timely.

Initial actions taken by the Plant Emergency Response Team (PERT) and the Richland Fire Department (RFD) appeared appropriate. Although changing the stack effluent samples did not appear timely, all other actions taken to determine the potential radiological consequences were appropriate and timely. There were no personnel injuries, no apparent appreciable release of radioactivity within the facility or to the environment, and no adverse impact on criticality safety (Section 1.1).

. The licensee's preliminary data indicated that the radiological consequences to plant personnel, RFD personnel, and the environment were minor and well below NRC regulatory limits (Section 1.2).

. The licensee's identification of acid soaked rags in the fire debris coupled with having previously identified acid soaked and alcohol soaked rags / mop heads in the waste bin is considered an unresolved item (Section 1.3).

. The licensee's finalization of the Incident investigation Board (llB) report was considered premature without having all the facts in hand relative to the cause of the fire.

Recommended corrective actions for the potential cause appeared appropriate (Section 1.3).

Radioactive Effluents

. Liquid effluent releases to the sanitary sewer system were well below the limits specified in 10 CFR Part 20, and the licensee was adequately controlling radioactive gaseous effluent releases. Data indicated that releases were being maintained as low as is reasonably achievable (ALARA) (Sections 2.1 and 2.2).

3-Environmental Protection

. The licensee's overall environmental monitoring program appeared adequate, and offsite releases did not appear to have any adverse effect on the environment (Section 3).

L f.

I

j Reoort Details Summary of Plant Status The plant was shut down for its routine annual maintenance and special nuclear material inventory during the April 13-17.1998, visit. There were no ongoing transfers of special nuclear material and all process systems were secured to prevent inadvertent transfer of material.

During the May,4-8,1998, visit the new dry conversion facility was operating and the Line 2 chemical conversion system was reprocessing clean scrap uranium. The Line 1 chemical conversion system had been permanently shut down in preparation for dismantlement,

. consistent with the licensee's plans for fuel manufacturing with its dry conversion process.

1 Operational Safety Review 1.1 Operational Event - Fire in UO, Buildino Solid Radioactive Waste Processino Area Overview During the past year, the licensee moved its solid waste processing operation from the Line 2 chemical conversion to the south end of the " hot shop"(uranium-contaminated equipment repair area)in the UO2 Building. The move was to facilitate the use of a new airlock hallway between the new dry conversion facility and the Line 2 chemical conversion area, a location where solid waste was processed. The new solid waste operation was separated from the hot shop work area by ceiling-to-floor and room-width welding curtains. The combined dimensions of the area is 48 ft. long by 18 ft. wide with 12 ft. ceilings. The hot shop portion also has an offset area approximately 18 ft. long and 8 ft. wide. The waste processing area occupies an area approximately 24 ft. long and 18 ft, wide. There are three exits / entries for the hot shop / waste area (HSWA): (1) a small transfer airlock on the south end that leads to the outside yard area, (2) a door at the west side of the waste area that leads to a hallway for access to an adjacent uranium powder storage room and an upper-level U3 0, facility (directly over the HSWA), and (3) a door at the north end of the hot shop area leading to a uranium powder drum  :

transfer hallway / ramp for the Line 2 conversion area.

The air exhaust system (K-37) for the HSWA also exhausts air from the powder storage room, the U30, facility, and the tank canyon between the Line 1 and Line 2 chemical conversion areas at the north side of the UO2 Building. Outside air is supplied to the tank l canyon from a separate system (K-38) and the HSWA, the U0, 3 facility and powder storage room are supplied by another system (K-39). Room air from each of these areas is exhausted through a roughing pre-filter and a primary high-efficiency particulate air (HEPA) filter and subsequently filtered through a secondary HEPA filtering bank before being exhausted through a stack to the outside atmosphere. The K-37 system duct-work is made of galvanized steel and is equipped with a water fog deluge system near the final HEPA filtering bank. The deluge system is automatically activated by a rate-of-rise / heat detector located in the exhaust duct. There are seven other separate f heating ventilation and air conditioning systems (HVAC)/ exhaust systems that service the UO2 Building. The HVAC systems are equipped with temperature, pressure and

flow sensors that automatically maintain damper positions to control the temperature and applicable pressure differentials for the various building areas. The air supply systems for each respective exhaust system are interlocked so that upon failure of an exhaust fan the supply fan automatically shuts down. Also, the HVAC systems are interlocked with the respective area fire detection system. If the fire detection system is ac4vated the respective area supply fans automatically shut off.

To aid in moderation control for criticality safety, there are no fire suppression systems in the UO, Building process areas. Areas where moderation controlis a factor are posted with water cxclusion emblems to prevent the use of water for firefighting. These exclusion areas are also identified in the licensee's Pre-Emergency Plan (Pre-Fire Plan) which has been provided to the fire department. Criticality safety for the waste processing area was based on surface density and the area is not a water exclusion area.

The licensee's fire detection system consisted of rate-of-rise / fixed temperature and/or smoke detectors. Manual fire alarm pull stations are also located throughout the licensee's facility. The HSWA area utilized a rate-of-rise / fixed temperature fire detection system. The actuation of any of these devices provides an automatic fire alarm in the affected area and at the annunciator panel in the central guard station. An alarm signal is also automatically transmitted by a radio signal to the Benton County Emergency Dispatch Center (EDC,911 calls) and, by a slave unit at the EDC, to the RFD. The nominal response time from the RFD to the licensee's site is less than 10 minutes.

Relative to solid waste operations, onsite generated uranium-contaminated waste is segregated into combustible and noncombustible waste and processed through the UO2 Building. " Hazardous" waste (pH < 2.5 and > 8.0) and chemically and non-chemically contaminated radioactive waste are collected in small, designated satellite collection containers labeled for the respective waste constituent where waste is generated. Hazardous waste is processed through a neutralization station prior to transfer to its respective satellite accumulation area in the waste processing area.  ;

Satellite accumulation drums in the waste processing area are segregated according to the specific chemical compound. Examples of segregation include: (1) washed and neutralized nitric acid contaminated mop heads and rags, (2) wet combustibles, (3) ammonium fluoride pre-filters, (4) ammonium nitrate pre-filters, (5) ammonium fluoride contaminated combustibles, and (6) solvent contaminated rags. Note: Due to the ongoing special nuclear material inventory during the week of April 13-17,1998, these satellite accumulation drums had been secured from use. Uranium-contaminated materials suitable for onsite incineration or offsite burial are placed in a portable wire waste coilection bin approximately 49 inches long,39 inches wide, and 32 inches deep.

The criticality safety requirements and posting used for the waste collection bin were

" bagged, pre-sorted waste fle_e from any uniquely identifiable accumulations of uranium."

After assay in the waste processing area for uranium content, combustible waste is subsequently incinerated in the licensee's Solid Waste Uranium Recovery facility for volume reduction and ultimate uranium recovery. Noncombustible materials are compacted onsite or super-compacted by a nearby vendor and shipped to a local burial site.

l

-d-l l a. Insoection Scoce (93702 and Tl2600/00.3) l At approximately 8:17 a.m. on April 15,1998, while onsite, the inspectors heard an announcement over the public address (PA) system for the PERT to assemble at the back of the UO2 Building. At this time, one of the inspectors went to where the PERT was assembling and noted an ongoing fire event. Within the same time frame there was an announcement over the PA system for the Plant Emergency Response Management Team (PERMT) to assemble in the Emergency Operations Center (EOC), and the second inspector went to the EOC. At this time, the inspectors diverted their attention from a routine onsite inspection to review the circumstances surrounding the event.

The review of this event included observing initial actions taken by the licensee, RFD, EOC operations, and actions taken by the licensee following the event. The inspectors interviewed personnel involved with the fire and personnel near the area when the fire started. Tne inspectors also reviewed sequences and consequences relative to the event, conditions of the area where the fire occurred, personnel and facility contamination survey data, facility work area air sample and effluent sampling data, selected licensee procedures, and the licensee's notification to NRC of the event (Event No. 34074).

b. Observations and Findinas Summarv of Event On April 14,1998, the licensee had replaced "firmware" in selected fire alarm processors and the software in its fire detection system under engineering change notice (ECN)

No. 6533L," Fire Alarm Firmware Upgrade." Following the upgrade of the system, the ECN required a 100 percent test of all fire alarms to verify and validate the software. ,

New upgraded software had been provided by the vendor to correct false alarm l problems encountered by the licensee.  !

At 6 a.m. on Apnl 15,1998, the licensee initiated testing of the fire alarm system in the Specialty Fuels Building which is adjacent to the UO2 Building. The testing process involved personnel actuating fire detectors in the area being tested, and the responsible engineer observing (panel and system activity printout) and acknowledging alarms as they were received. Both parties were maintaining radio communication. Because two false fire alarms from the testing had been sent to the EDC before the portion of the system undergoing testing was bypassed, and due to the uncertainty of more false ,

alarms, at 7 a.m. the licensee notified the EDC of the testing and provided instructions to ignore allincoming fire alarms from the site. The alarms from systems not being tested were not bypassed and were capable of sending a fire alarm signal to the EDC.

l The sequence of events on April 15,1998, described below are a best fit, approximate time sequence from licensee records, personnel interviews, personal written statements, and personal observations.

l l

J

i l At 7:45 a.m., while performing inventory activities in the l40, facility, an individual smelled smoke and thought it may have been from welding in the hot shop below. At

! 7:50 a.m., this individual left the area and did not observe anything unusual as he passed the door (windowed) to the waste area and exited through the powder storage room. At 8 a.m., two maintenance personnel delivered some doors to the hot shop but did not smell any smoke.

I At 8:06 a.m., a fire alarm signal from Zone 15 (UQ Building - hot shop, U3 0, facility, and powder storage room) was received at the fire alarm control panet. The fire alarms in the UO, Building were activated, and the alarm signal was also transmitted to the EDC. The responsible engineer at the fire system control panel noted that the signal was from an area not being tested but believed it may have been related to the testing and new software changes. Also, due to an earlier announcement of testing site fire alarms, there was some initial confusion by personnel in the UO2 Building as to the validity of the alarm.

Within the same time frame, an engineer and maintenance person in the powder transfer hallway opened the door to the hot shop, smelled smoke and evacuated the area.

Shortly after the fire alarm, an operator in the Line 2 control room, initially confused as to the validity of the alarm, began checking his areas to verify the realism of the alarm.

Upon opening the door from the ramp area into the hot shop he observed dense smoke emanating from the welding curtains to the waste area and rapidly filling the hot shop.

He immediately left the area, pulled a nearby manual fire alarm and notified security using the onsite emergency phone extension. He made a sweep of the conversion area to see that all personnel had evacuated and subsequently evacuated the building. In the meantime, several other employees had called security confirming a fire in the HSWA, which tied up the phone system and resulted in a delay of security notifying the EDC. At 8:08 a.m., Siemens security notified the EDC that the fire alarm signal was from a real fire at the facility.

With the fire alarm sounding and word of smoke in the UQ Building, the PERT began to initiate actions. A PERT leader and PERT member began checking air locks at the back of the UO, Building. A view through the airlock doors leading into the Line 2 conversion area did not reveal any evidence of fire. These individuals proceeded to the air lock leading into the HSWA and observed, through the inner door window, with the lights out, black smoke and flames about 3 ft. high in the area where the waste bin was normally located.

At 8:15 a.m., the RFD arrived onsite, and announcements were made over the PA system for the PERT to assemble at the south end of the UQ Building and the PERMT to assemble in the EOC. With the incident Commander at a nearby Department of Energy training facility, the General Supervisor, Chemical Operations, assumed the position of incident Commander and began directing the PERT.

At 8:18 a.m., the PED (alternate) declared the event as an Alert in accordance with Siemens EP (Emergency Plan), "any fire in any area onsite that has not been c'eclared extinguished within 15 minutes."

l

• l

)

l-8-

I At 8:19 a.m., two PERT members wearing normal plant protective clothing and each equipped with a self-contained breathing apparatus (SCBA) entered the waste area, just inside the inner door, to assess the situation. With the lights out, the room filled with smoke, and the fire apparently beyond the incipient stage, the two PERT personnel immediately left the area back through the air lock.

At 8:23 a.m., a RFD fire fighter and PERT member (to provide guidance) entered through the interlock door into the waste room with a CQ fire extinguisher and water fire hose to extinguish the fire. The fire fighter was equipped with fire fighting turnout gear and a l SCBA. The PERT member was equipped with normal protective clothing and a SCBA.

! With the fire hose stretching through the inner and outer door of the airlock, a small l amount of visible smoke started escaping from the room at the upper level of the airlock outer door into the outside atmosphere and the radiological PERT began setting up two goose neck air samplers outside the airlock door. PERT members zoned off the area to keep personnel away from the action area and smoke coming from the room. The fireman initially expelled the CQ extinguisher on the fire and noted the flames were not completely out. The fireman then used a small amount of water on the remaining fire, left the area and reported that the fire was out. The fireman and PERT team member went back inside and removed the waste bin into the yard area near the air lock to " overhaul" the smoldering debris. The normal PED who was at a nearby Department of Energy training facility arrived at the scene to relieve the acting PED.

At 8:30 a.m., two portable goose neck air samplers were online with one stationed down wind and one upwind of the bin. Within the same time frame, the smoke started to increase from the bin. At this time arrangements were being made to install a plastic )

tarp under the waste bin and use water if further extinguishing action was needed.

Shortly after there were two minor flare ups of the fire which were extinguished with small amounts of water. With the fire completely out, the remaining debris was subsequently inspected by the RFD and simultaneously placed into a 55 gallon drum.

At 8.33 a.m., with the report from the fireman that the fire was out and all personnel had been accounted for, the licensee declassified the event from an Alert and began notifying i I

offsite agencies and nearby facilities. At the time of the fire alarm, most of the operating personnel were in training classes and the UQ Building was being manned with only a few personnel.

Two PERT members were sent back into the area equipped with SCBAs to check for smoke. With the fire source removed from the room, most of the smoke had dissipated. l Some smoke was also noted in the powder storage room and the Line 2 conversion area. Damage was noted to various polyvinyl chloride (PVC) piping in the area. The PERT members posted no entry signs at all doorways of the affected areas and left the building. j At 8:45 a.m., air balance personnel and health and safety technician (HST) personnel entered the area to assess conditions in the plant and commence steps to evacuate  !

remaining smoke from the effected areas. By 9:10 a.m., (1) all smoke had been cleared j from the UO2 Building, (2) all notifications to off site agencies had been completed l I

J

(NRC notified at 8:50 a.m.), (3) the licensee was responding to the local media, (4) personnel contamination surveys had been completed, (5) the licensee had appointed a chairman to head an llB to determine the cause of the event, (6) smear surveys indicated that no radioactive materials had spread from the process areas into the normally clean areas of the UO2 Building, and (7) the RFD was preparing to leave the site.

The barricaded zone outside the HSWA airlock was surveyed for loose contamination.

2 Loose contamination just outside the airlock door averaged 4000 dpm/100 cm

(-12 ft 2area),1200 dpm/100cm 2(-16 ft area) 2 where the bin had been placed, and 2 2 400 dpm/100cm (-120 ft area) within the zone. A small area of fixed contamination measuring 7000 dpm/100cm2 (~400 cm2 area) was found on the ground. The loose contamination was removed and the area was conditionally released until a more detailed survey for any remaining fixed contamination on the ground could be performed.

At 9:10 a.m., preliminary alpha counting sample results (no decay time allowed for natural occurring alpha emitting radionuclides) were obtained from the air samplers setup near the waste hin. The air concentration (30 minute collection time) from the downwind sampler indicatt d 7.6E-12 uCi/ml and the upwind sampler indicated a concentration of 5.7E-12 UCi/ml. These results do not include a dilution factor (approximately 1,000) for concentrations expected at the boundary of the restricted area fence line. The 10 CFR Part 20, Appendix B, Table 2, Column 1 unrestricted release annual average limit using Class "Y" material (U-234 is the most predominant activity for low enriched uranium) is 5.0E-14 uCi/ml.

By 9:30 a.m., the licensee had determined that with the exception of the K-37, -38, and

-39 HVAC systems which had automatically shut down, all ventilation systems were operating as designed and differential pressures across HEPA filtering systems were normal.

At 9:32 a.m., the RFD Battalion Chief arrived at the EOC and provided a summary of his perspective. He noted that no preliminary causes were identified and that the waste bin being mobile helped in extinguishing the fire. He also reported that the communication interfacing between the licensee and the RDF was good.

Radiological surveys of personnel who had evacuated from the UO2 Building and those who responded to the fire indicated no intakes of radioactive material from nasal smea:s and no skin contaminations. A small sport of uranium, measuring 4000 dpm, was found i on the right boot of the fireman that had entered the HSWA. The boot was 4 decontaminated and returned to the fireman. No contamination was found on the fireman's turnout gear or >ther protective equipment. Equipment used to extinguish the fire and to inspect the debris was wiped down as it was removed from the area, and no contamination was found on the equipment during subsequent surveys. At 9:45 a.m.,

the HST staff began changing air samples from the HSWA and other potentially affected areas of the UO, Building.

l At 9:47 a.m., the fireman who extinguished the fire addressed the PERMT in the EOC i and stated, in part, that he observed that there were approximately 3 ft. high flames coming from inside the bin and no evidence of any fire exterior of the bin. Heat damage was evident in the room (melted plastic piping), and 4-5 waste drums had to be moved to access the fire and remove the bin. He also stated that the use of the fire hose was primarily for his personal protection. From his observations in the room and inspection of the debris remaining in the bin (paper, rubber gloves, and cloth rags), he did not see any fire code violations.

At 9:55 a.m., the PERMT was released but reassembled at 10:10 a.m. as a suggested precaution by the PED because there had been no report of the results of an inspection of the primary HEPA filters for the HSWA. At 10:14 a.m., the licensee verified that there was no damage to the HEPA filters and the PERMT was deactivated.

Relative to observations of EOC activities, the inspector noted that the PERMT assembly in the EOC was done in an orderly and organized fashion. There was some initial confusion as to the proper channel to use on the emergency portable radios, but the matter was quickly resolved. Notifications to offsite agencies were completed within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> after declaration of an emergency as required by implementing Procedure 1.1 of Siemens' EP. The inspector also noted that individuals in the EOC referred to their procedures throughout the event, and that an excellent event chronology was maintained.

Preliminary results from the air samplers (upwind and downwind) outside the air lock from 9 a.m. to 10:15 a.m. averaged 7 4E-13 uCi/ml. All of the air samples in the UO2 Building process areas had been changed at approximately 6:00 a.m on April 15, 1998. Four fixed air samples from the HSWA (two in each area) averaged approximately 1.4E-11 uCi/ml. The 10 CFR Part 20, Appendix B, Table 1, Column 3, derived air concentration (DAC) using Class "Y" material (U-234) is 2.0E-11 uCi/ml. Since the samples were heavily sooted, the sample results observed would not be representative i I

of what was collected due to attenuation of the alpha activity by the soot. The samples were sent to the analytical laboratory to be analyzed for uranium content which involved dissolving the sample, extraction of the uranium, and analysis with an inductively coupled plasma-mass spectrometer.

Approximately 89 air samples located throughout the UO, Building process areas were I

changed within the same time frame (9:45 a.m.) as those from the HSWA. Preliminary alpha counting results without any decay time indicated concentrations that ranged from 4.4E-11 uCi/ml to 9.7E-13 uCi/ml. Most of these samples were sooted and were being held for subsequent analysis in the chemistry laboratory. j The licensee's inspection of conditions in the room and HVAC system revealed that the K-37, -38, and -39 systems had automatically shut down due to loss of control power.

The control power was supplied from a 208/120 volt breaker panel located on the wall  ;

within 3.0 ft. of where the waste bin was stored. The licensee noted that all of the  ;

breakers utilized in the panel had tripped; heat from the fire was suspected as the cause.

Also, alongside the 208/120 volt panel was a 480/277 volt panel with breakers tripped.  !

l

_ ____--_-_-_____ _ - ______ _ __ O

I In addition to the HVAC control system, the 208/120 volt panel supplied power for the local fluorescent lighting, an actuation solenoid for the K-37 deluge system, and roll back doors for the transfer ramp. The 480/277 volt panel supplied power to welding outlets and furnaces in the U3 0, facility. Visible damage observed by the licensee included a melted PVC vacuum line and PVC back-flow preventer for the K-37 deluge system, a melted exit sign above a door near the fire zone, damage to an overhead fluorescent l

lighting fixture and its respective wiring, exposed wiring from an unsealed wire pull box from the 480/277 volt system and an unsealed wire pull box from the 208/120 volt system located in the overhead above the fire zone. A fire detector in the overhead above the area where the waste bin had been stored was also damaged. Due to the apparent heat damage to the electrical systems, the HVAC system was not operable.

Using established procedures to jumper, the K-37 exhaust fan was started (approximately 12:30 p.m.) in the manual mode to maintain negative pressure in the affected areas. The supply fans were not started.

At 11 a.m., stack effluent samples were changed from all eight exhaust stacks and counted. The K-37 stack sample results exhibited a preliminary result of 1.1E-12 uCi/ml with a corresponding fence line concentration of 1.3E-15 uCi/mi using a predetermined dilution factor (818). The results from the other stacks were less than 5.0E-13 uCi/ml, the licensee's action level for calculating the effluent concentration at the fence line. The data indicated no appreciable release of radioactive material to the environment.

Although the K-37 exhaust automatically shut down during the initial stages of the fire and the system was equipped with a double HEPA filtration system, the inspectors viewed the changing of the effluent samples was not as timely as other actions taken to evaluate the potential radiological consequences.

c. Conclusions f

Although some momentary confusion developed as a result of fire alarm testing I occurring coincident at the time of the fire, the matter of a real fire was proniptly identified and the licensee promptly initiated actions outlined in its EP. The PED (alternate at the time) appropriately classified the event as an Alert. The accounting for onsite personnel and notification of offsite agencies appeared timely. Initial actions taken by the PERT and the RFD appeared appropriate. Although changing the stack effluent samples did not appear timely, all other actions taken to determine the potential radiological consequences were appropriate and timely. There were no personnel injuries, no apparent appreciable release of radioactivity within the facility or to the environment, and no adverse impact on criticality safety.

1.2 initial Followuo to Event

a. Insoection Scoce (93702 and Tl2600/003)

The inspectors observed actions taken by the licensee to further evaluate the radiological j consequences and its continuing investigation as to the cause of the fire.  !

l l

i

I

}

b. Observations and Findinas At 1:30 p.m. on April 15,1998, the HST staff again exchanged all of the air samples that had been changed earlier at 9:45 a.m. Preliminary alpha counting results for the four air samples from the HSWA averaged approximately 9.0E-12 uCi/ml with no significant sooting problem. The air samples from the other areas ranged from 1.0E-11 uCi/ml to 1.3E-13 uCi/ml. Respiratory protection requirements were lifted at this time.

Between 1:30 to 2:30 p.m., the llB began to collect and evaluate data, and interview personnel who had been in or near the area of the fire. As part of the investigation, the licensee also requested the assistance of a fire inspector from the City of Richland Fire and Emergency Services. With the fire detection system out of service in the HSWA, the licensee established a roving fire watch for the area.

An NRC inspector, while in the HSWA, noted significant sooting of the waste processing area, especially the ceiling, and the heat damage discussed above. The inspector also noted that the exhaust duct from the waste area to the K-37 system and nearby surfaces were only slightly sooted. This indicated that the venti ation systems had shut off early in the event. Also, the heaviest sooting was confined to he waste area of the room (HSWA). It appeared that the welding curtains had ar influence on retaining most of the soot to the waste area. Since the ventilation systems had shut down, the general air flow, as evidenced by the soot, appeared to have been toward the door to the Line 2 conversion area. Although there was some sooting r'oted from wipes of surface areas, there was no significant sooting in the powder storage room or 40, facility. At the time, there was no observable physical damage to equipment outside of the waste area. The sealed waste containers in the room did not show any signs of loss of integrity from the fire.

At 9 p.m., the licensee completed its survey for fixed contamination on the outside ground area near the HSWA airlock. The licensee's survey data using a thin window Geiger-Muller pancake probe indicated four small areas of fixed contamination that measured (1) 1400 dmp/100cm for a 1000 cm' area, (2) 7000 dpm/100 cru' for a 2

400 cm' area, and (3) 1500 to 1600 dpm for two respective 60 crd areas. Utilizing the licensee's survey data from the loose contamination discussed above and the data from the fixed ground contamination surveys, the licensee estimated that approximately 0.5 uCi of uranium had been spread to the outside ground area near the airlock doors.

The areas were subsequently cleaned to less than the licensee's action limits.

On April 16,1998, the licensee informed the inspectors that based on its records, a {

melte:1 exit sign above the door to the hallway leading to the 40, facility and powder storage contained tritium. Since most of the exit signs onsite do not contain tritium, this matter was not recognized during or immediately following the fire. Because the state of Washington is an Agreement State, the inspectors informed the licensee that the matter would be under the jurisdiction of the state which should be notified of the matter. The licensee initiated actions to evaluate the potential radiological consequences from the damaged sign and reported the matter to the State. Subsequently, the NRC Region IV office also informed the State of the matter.

I

Relative to the licensee's ongoing investigation, samples of ash and water caught in the plastic tarp placed under the waste bin during the overhauling of the fire debris were collected and sent to the chemistry laboratory for analysis of the chemicalingredients of the material sampled. The licensee's initialinspection of the debris from the waste bin identified:

. cloth rags, including three rags with a pH of <1.0 (determined by pH paper)

. Masslinn cleaning cloths

. a plastic bag containing pieces of electrical wiring

. pieces of 2 inch by 4 inch lumber

. latex and vinyl gloves

. cotton liner gloves

. melted PVC piping

. record cards

. a small piece of "unistrut" After the event, the inspectors reviewed selected portions of the licensee's EP to determine if the EP was correctly implemented. This review included: (1) declaration of the Alert, (2) use of EP implementing procedures, (3) termination of the Alert, (4) use of alternates in the EOC, (5) use of the PERT for fire control, (6) the accountability process, (7) radio usage, and (8) building evacuation. The inspectors noted that although personnelin the EOC referred to the EP implementing procedures, the accompanying checklists were not filled out by all of the PERMT members. This was a matter also identified by the licensee and was to be addressed in a self-assessment of EOC activities during the event. The inspectors determined that the licensee had adequately implemented the requirements of its EP.

Shortly before an exit briefing on April 17,1998, the licensee provided the inspectors with the analytical results of the sooted air samples from the HSWA. The data indicated significantly higher results than the preliminary alpha counts. An in-office review of the data indicated that the two samples from the waste area averaged approximately 8.7E-10 uCi/ml, and the two samples from the hot shop area averaged 7.27E-10 uCi/ml.

The highest concentration for both areas was 1.1E-9 uCi/mi from a sample located at a welding hood in the northwest corner of the hot shop. Although these samples indicated a concentration higher than the 10 CFR Part 20 DAC, the SCBA used by the fireman and PERT member provided a protection factor of 10,000, and the stay time for these individuals was approximately 4 minutes. Thus, any intake of low enriched uranium would be well below the 10 CFR Part 20 annual limit of intake.

c. Conclusions The licensee's preliminary data indicated that the radiological consequences to plant personnel, the RFD, and the environment were minor and well below NRC regulatory limits.

E 1

' 1.3 Subseauent Followuo of Licensee Actions i

A subsequent onsite followup of !icensee actions related to the fire event was performed on May 4-8,1998.

a. Insoection Scoce (93702 and Tl2600/003) l The inspection included a review of the status of the licensee's radiological assessments, results of laboratory analysis of debris from the fire and the licensee's llB Report, l "UO, Building Waste Handling Area Fire," dated May 1,1998, and approved May 6, 1998. The inspectors also toured the HSWA to observe current condit'ons and reviewed the startup of the facility with cognizant licensee personnel.

b. Observations and Findinas Licensee's Radiological Assessment With the exception of the analyces for the air samples from the HSWA, the licensee had not completed all of the analyses of the sooted air samples changed during the fire event. The laboratory procedures for treating and analyzing the samples involved about 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> of preparation time. Also, the air samples were being analyzed by the same equipment used for its uranium bioassay urinalysis program, which limited the number of air samples that could be processed daily.

The licensee informed the inspectors that when all of the sample analyses were completed a comprehensive report of the licensee's final assessment of the radiological consequences would be included in the llB investigation file.

Licensee's Investigation Reoort The licensee's llB investigation report included discussions on: (1) the damage incurred; (2) a summary of the incident that includad a brief description of the radiological consequences, initial conditions, the event, initiating event, immediate corrective actions, and root cause; and (3) recommended corrective actions.

The damage incurred was similar to that described in the above sections. As for the radiological consequences, the licensee determined from information collected to date, that the incident did not result in any personnel radiological exposures or effluent releases that exceeded any regulatory limits. Regarding the initiating event, no definite cause for the fire had been determined as of the time of the report (May 6,1998).1he licensee considered several pathways for the fire to have started which included:

(1) cigarette smoking in the area, (2) UQ " burn-back," (3) welding / cutting activities, (4) zirconium end caps and turnings in waste, (5) electrical short, (6) deliberate act of arson, and (7) incompatible chemicals resulting in spontaneous combustion. Based on the licensee's inspection of the debris from the fire and interviews with personnel, the llB ru!ed out all items except a deliberate act of arson, electrical short, and incompatible chemicals.

{

i Regarding arson, the licensee's interviews with personnel in the area prior to the fire disclosed no suspicious activities on the part of employees or contractors. No materials l

were identified in the fire residue that would indicate a deliberately set fire. However, the l licensee did not rule out this possibility.

Regarding an electrical short, the city senior fire inspector had checked all likely sources l of electrical shorts and based on his initial assessment did not identify any source of the I ignition. However, since his official report would not be available for a few weeks, the licensee did not rule out this source. In addition to the fire inspector's investigation, the licensee's examination of the electrical systems did not disclose any evidence that the fire had started from an electrical short.

Regarding incompatible chemicals, several combinations of chemicals in sufficient quantities within a particular environment can generate sufficient heat to cause a fire.

This would include nitric acid mixed with alcohol or with other organic compounds. Nitric acid is used in the chemical conversion process along with lubricating oils and grease for equipment. Rags containing these materials are generated in the chemical conversion and pellet pressing (grease and oil) areas. Alcohol, in small quantities, is used in the pellet pressing area; therefore, rags containing alcohol are also generated. The licensee's examination of the fire debris identified rags with a pH of less than 1.0 (acidic),

but no alcohol or other organic volatiles were detected. Based on personnel interviews, the llB believed that it would be highly unlikely that enough acid, alcohol and organic or flammable material could have been in the waste bin to cause an exothermic reaction of a magnitude capable of causing a fire. However, because the satellite collection waste containers for alcohol, solvent contaminated waste, and acid soaked waste had been removed from the process areas due to the uranium accountability inventory, the only available location for this material was the waste bin. Therefore, the llB could not rule out the possibility that the fire could have started from incompatible chemicals.

For root cause, the licensee used the "TapRootM Incident Investigation System" and listed arson, electrical short, and incompatible chemicals for possible root causes. Since there was no evidence pointing to arson or an electrical short, the llB focused on incompatible chemicals. Under incompatible chemicals the llB delineated the following possible causes.

• Procedures not used or not followed, and lack of adequate training, because acid soaked and alcohol soaked rags / mop heads had been found in the waste bin from time to time in violation of Procedures (1) EMF-22, P66-1055, " Wet Waste Uranium Recovery Station," (2) EMF-22, P66-083, "Clast A Waste Generator Requirements," and (3) EMF-1787, "SPC Satellite Accumulation Area Control Document."

• No procedure was written to cover the disposal of waste generated after the removal of the proper waste containers during the annualinventory.

f l The llB listed for recommended corrective actions:

- Conduct a self-assessment of Plant Operations pre-disposal accumulation and management of acid and solvent soaked waste including procedures, training and collection points. To be completed by Plant Operatiuns by June 15,1998.

. Evaluate the purchase or fabrication of an enclosed combustible waste storage bin container with a self-closing lid activated by a fusible link. To be completed by Facilities Engineering by May 19,1998.

Relative to the licensee's investigation of ignition mechanisms for the fire, the inspectors noted that not all of the laboratory analyses had been completed for the liquid collected from the tarp used when overhauling the fire and the fire debris. This included analyses for zirconium, nitrate ions or ion analysis (acid), and volatile organic analysis (to be sent offsite for analysis). Since the licensee had not received the fire inspector's report and the chemical analyses had not been completed and reviewed, the inspectors questioned licensee llB members as to their reasoning for completing the llB report prior to having finalized information. The licensee representatives stated, in part, that with the information at hand, they believed there would be no changes in the fire inspector's report from that verbally communicated during the investigation and the chemical analyses would not result in any significant differences from the visual inspections and the acid rags found in the fire debris. The licensee representatives also informed the inspectors that if there were any changes to information relat 1 to the identification of cause of the ignition source, a supplemental report would be added to the current IIB report.

The licensee's identification of acid-soaked rags in the fire debris coupled with having previously identified acid-soaked and alcohol-soaked rags / mop heads in the waste bin is a possible violation of the licensee's operating procedures for controlling chemical waste.

Pending the licensee's completion of the chemical analysis of the fire debris, this matter is considered an unresolved item (URI 70-1257/9802-01).

Regarding the licensee's evaluation of its response to the fire event, on April 17,1998,a critique of the PERT's actions was performed, but had not been formally documented.

Also, a lecsons learned report from EOC activities had not been completed.

Since the licensee had not (1) finalized its radiological consequences report, (2) received the fire inspector's report, (3) completed its laboratory analyses of the fire debris, or (4) completed the evaluations of its emergency response actions, these matters will be reviewed in a subsequent inspection and are identified as an inspection followup item (IFl 1257/9802-02).

During facility tours, the inspectors noted that the licensee had cleaned most of the soot from the HSWA, had repainted the walls of the waste area, and made the necessary repairs to restart the facility. The startup of the facility was being performed in accordance with the ECN process and Startup Council reviews and approvals. No concerns were identified by the inspectors.

c. Conclusions The licensee's finalization of the 11B report was considered premature considering that all the facts relative to the cause of the fire were not yet available. Recommended corrective actions for the potential cause appeared appropriate. An unresolved item involving a possible violation of the licensee's operating procedures for controlling chemical waste was identified.

1.4 Miscellaneous Operational issues (Closed) Violation 70-1257/9707-03: This issue involved the licensee's identification of unauthorized storage of uranium oxide powder in temporary storage locations. Based on discussions with cognizant licensee representatives, a review of licensee inventory records, and a physical examination of the selected temporary storage locations (sea / land containers), the inspectors verified that the licensee had completed the corrective actions described in the licenseo's letter dated February 13,1998. No i additional concerns were identified by the inspectors.

(Closed) Insoection Followuo item 70-1257/9303-06: This item involved a review of the licensee's criticality safety analyses update program that was initiated in 1993.

Consistent with the NRC Fuel Cycle Operations Branch assuming the primary responsibility for this matter as documented in Headquarters inspection Report 70-1257/98-201 (IFl 70-1257/98-201-02), this matter is considered closed.

2 Radioactive Effluents 2.1 Radioactive Liauid Effluents

a. Insoection Scoce (88035)

The inspection included a review of liquid effluent sampling data, selected licensee procedures, and a tour of the liquid effluent sampling station to observe the status of the sampling and effluent measurement equipment.

b. Observations and Findinas The inspectors noted that composite samples of discharges from Lagoon SA to the onsite sewer system line for the past year continued to indicate that uranium concentrations are typically less than 0.1 ppm. Composite samples of the retention tank from the laundry facility were typically less than 0.2 ppm uranium. Daily composite samples of the combined site effluents indicated uranium concentrations were less than the licensee's detection level of 0.05 ppm.

During a tour of the licensee's liquid effluent sampling station, the inspectors r.oted that the liquid effluent flow measurement system (pressure bubbler type), the in-line pH measurement system and automatic composite sampling system were fully operational

l and currently calibrated. A view inside the manhole for the sampling equipment and effluent flow fiume indicated that the effluent stream was free of any debris that could have a negative impact on the sampling system.

The inspectors noted that there had been no changes in the licensee's analytical program during the past year. The licensee continued to use a nitrogen ion laser kinetic phosphorescence analyzer (KPA) to determine the total uranium concentration released in liquid effluents. As currently used, the KPA's detection limit is about 0.05 ppm uranium which corresponds to approximately 8.8E-8 microcuries (uCi) per milliliter. The performance of the KPA unit is monitored by plotting daily standards against a weekly generated calibration curve.

c. Conclusions Liquid effluent releases to the sanitary sewer system were well below the limits specified in 10 CFR Part 20 and appeared ALARA.

2.2 Gaseous Effluent Releases

a. Insoection Scoce (88035)

The inspectors toured licensee facilities to observe operations associated with gaseous effluent sampling equipment and reviewed selected gaseous effluent sampling data for the past year.

b. Observations During facility tours, the inspectors noted that stack sampling equipment appeared fully operational. Current calibration stickers for the flow meters and a tag that showed the calibrated sample flow rates and corresponding sample flowmeter readings were appropriately located at each sample collection station. Heat tracing systems appeared fully functional as there was no observed moisture in stack sample flow meters.

The licensee's sample data and evaluation of gaseous effluent releases at the site boundary indicated that releases of radioactive materials were well below the limits specified in 10 CFR Part 20 and the reporting level in Part I, Section 5.1.1 of the license.

The inspectors noted that the licensee's calculated annual dose for 1997 gaseous effluent discharges was less than 0.03 millirem to the hypothetical maximally exposed person living at the nearest site boundary. This dose was calculated using Environmental Protection Agency " COMPLY" computer code. The licensee's 1996 annual dose was 0.011 millirem. The increase noted in 1997 was attributed to operating the incinerator facility which had been shut down most of 1996, new operating facilities such as the new dry conversion facility and mop water processing facility, and some minor increases in releases from the chemical conversion area that were corrected by changing the HEPA filters. The licensee's 1997 offsite dose from gaseous effluents was well below the 10 CFR 20.1101 dose constraint of 10 millirem per year.

-______a

19

c. Conclusions ,

The licensee was adequately controlling radioactive gaseous effluent releases and the data indicated that releases were being maintained ALARA.

2.3 Effluent Reoorts

a. Insoection Scoce (88035)

The licensee's semiannual effluent reports for 1997, dated August 6,1997, and February 10 and 13,1998, were reviewed.

b. Observations and Findinos The reports provided a summary of the radioactive gaseous and liquid effluents released from the licensee's facility, and a summary of the radioactive solid waste sent for burial.

The report dated February 13,1998, provided a correction of a licensee identified typographical error in the February 10 report for total radioactivity released from two stacks, which did not affect the reported total radioactivity released. The licensee's data indicated that a total of 16.7 uCi of low enriched uranium had been released from gaseous effluent discharges in 1997, which was an increase from 8.9 uCi released in 1996. Regarding liquid effluents, the licensee's data for 1997 indicated that 10.0 mci of low enriched uranium had been released to the sanitary sewer which was an increase from the 8.0 mci released in 1996. This increase was attributed to increased site activities as discussed above.

c. Conclusions The licensee's reports were submitted in accordance with the requirements of 10 CFR 70.59. Liquid and gaseous effluent releases appeared minimal.

3 Environmental Protection 3.1 Insgetion Scoce (88045)

The inspectors reviewed selected licensee procedures and records of environmental sampling results and licensee internal audits for the past year. The inspection also included discussions with cognizant licensee representatives, and a tour of the lagoon system to observe the status of the inter-liner sampling system.

3.2 Observations and Findinas Responsibihties, controls, and the environmental surveillance program were adequately described in Chapter 4.0, " Environmental Standards," of the licensee's safety manual (EMF-30). Controls for liquid and gaseous effluent releases were also included in this stanciard. Collection and treatment of each type of sample were provided in lower-tier radiological and operating procedures.

20-l l The inspectors noted that monthly inter-liner samples taken from the lagoons indicated i no lagoon leakage. During the past two years, no measurable liquid had been observed l from the lagoon leak detection system.

The licensee's sample results of water samples from nearby test wells (GM-1 up-gradient and down-gradient GM 5-8, and TW 6-7 and 21) have shown no significant impact from lagoon operations in the groundwater (non-drinking water). The licensee's trending of vendor-supplied analytical results of 1997 quarterly groundwater well samples continue to indicate a slight increare in gross alpha and beta concentrations in down gradient well water sample results. As an example, for well GM-5 immediately down-gradient of Lagoon 1, the average 1996 gross alpha activity was approximately 78.8 pCi/l. The 1997 average was 88.3 pCi/l.

There was an overall decrease and/or leveling trend for chemical constituents.

Trichloromethane (TCE) ranged from 1.0 ug/l to 6.0 ug/l. Test well GM-8 had the highest concentration of TCE with a 1997 average of 3.6 ug/l. TCE in down gradient wells has shown a decrease from previous years and appeared to be level for the past 4 years.

Nitrate was detected in all of the wells sampled during 1997 and ranged in concentration from 9.8 to 42.1 mg/l which showed a slight increase from 1996, but lower than in previous years. Ammonia was only detected in Wells GM-5 and 8 and ranged in concentrations from 5.8 to 18.6 mg/l ammonia (as nitrate). Ammonia levels continued to show a decreasing trend. Fluoride was detected in all wells sampled and ranged in concentrations from 0.21 mg/l to 6.15 mg/l, with the highest concentrations in Well GM-5.

The fluoride concentrations in 1996 ranged from 4.9 mg/l to 12.0 mg/l.

The inspectors noted that analytical results of quarterly soil samples indicated uranium concentrations of less than 1 picocurie per gram (pCi/g). Sample data for fluoride in air and forage continue to indicate values slightly above or less than the respective detection limit.

Related to liquid effluent releases to the sanitary sewer system, the licensee's analytical results of monthly sewer sludge samples taken at the Richland sanitary treatment facility continue to indicate an average of less than 10.0 pCi/g, which represents a steady downward trend during the past few years.

3.3 Conclusions The licensee's sampling program was consistent with Section 5.2, Part I, of the license application and licensee procedures. The licensee's overall environmental monitoring program appeared adequate, and offsite releases did not appear to have any adverse effect on the environment.

4 Exit Meeting Summary The inspectors presented the inspection results to members of licensee management at the conclusion of the onsite inspection on April 17 and May 8,1998. The licensee

21-acknowledged the findings presented. Although proprietary information was reviewed during this inspection, such information is not knowingly described in this report.

l l

l l

b'; ~

m . L ATTACHMENT SUPPLEMENTAL INSPECTION INFORMATION PARTIAL LIST OF PERSONS CONTACTED Licensee B. F. Bentley, Manager, Plant Operations l J. M. Deist, Criticality Safety Analyst J. B. Edgar, St.iior Engineer, Licensing B. N. Femreite, Vice President, Manufacturing E. L. Foster, Supervisor, Radiological Safety S. S. Knegler, Manager, Waste Management Engineering D. C. Kilian, Manager, Manufacturing Engineering L. J. Maas, Manager, Regulatory Compliance l C D. Manning, Lead Criticality Safety Specialist l G. A. McGehee, Criticality Safety Specialist l J. H. Phillips, General Supervisor, Chemical Ope:ations T. C. Probasco, Manager, Safety K. H. Tanaka, Environmental Engineer

1. J. Urza, Manager, Manufacturing Technology R. E. Vaughan, Manager, Safety, Security and Licensing Fire and Emeroency Services. City of Richland. Washinaton J. C. Stewart, Senior Fire inspec'or INSPECTION PROCEDURES USED Tl 2600/003: Operational Safety Review IP 93702: Prompt Onsite Response to Events at Operating Power Reactors IP 88035: Radioactive Waste Manageme '

IP 83045: Environmental Protection IP 92701: Followup IP 92702: Followup on Corrective 6 is for Violations and Deviations ITEMS OPENED, CLOSED, AND DISCUSSED Goened 70-1257/9802-01 URI Failure to follow waste handling procedures 70-1257/9802-02 IFl Review of licensee's Onal assessments of the April 15,1998, fire event Closed 70-1257/9303-06 IFl Review of criticality safety analyses update program 70-1257/9707-02 VIO Unauthorized storage of uranium oxide powder in temporary

storage locations

.x _ . _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ ______ _ ___ __

1 LIST OF ACRONYMS USED ALARA As low as is reasonably achievable DAC derived air concentration ECN Engineering Change Notice EDC Emergency Dispatch Center EOC Emergency Operations Center EP Emergency Plan HEPA high-efficiency particulate air HST health and safety technician HSWA hot shop / waste area HVAC heating ventilation and air-conditioning IIB Incident Investigation Board KPA kinetic phosphorescence analyzer PA public address PED Plant Emergency Director PERMT Plant Emergency Response Management Team PERT Plant Emergency Response Team PVC polyvinyl chloride RFD Richland Fire Department SCBA self-contained breathing apparatus TCE Tnchloroethane

_ _ _ _ _ _ _ _ _ _ _ _ _ - - _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _