ML20205M553
| ML20205M553 | |
| Person / Time | |
|---|---|
| Site: | 07001113 |
| Issue date: | 03/27/1986 |
| From: | Dan Collins, Stoddart P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| To: | |
| Shared Package | |
| ML20205M527 | List: |
| References | |
| 70-1113-86-01, 70-1113-86-1, NUDOCS 8604150308 | |
| Download: ML20205M553 (9) | |
Text
_
?
- p3 Kiroq.
UNITE 3 STATES f
- o NUCLEAR REGULATORY COMMISSION REGION il k
101 MARIETTA STREET.N.W.
- I 2
ATL ANTA, GEORGI A 30323
\\
/
APR 0 81986 Report No.: 70-1113/86-01 Licensee: General Electric Company Wilmington, NC 28401 Docket No.: 70-1113 License No.:
SNM-1097 Facility Name: General Electric Company Inspection Conducted:
Februar'
-5, 1986 Inspector:
Mdk>
'd,+('
44 PhMlip G. Stoddart ~
Date Signed Approved by:
hh*
N*
3/R7h Dougfas M. Collins, Branch Chief Date Signed Division of Radiation Safety and Safeguards
SUMMARY
Scope: This special, announced inspection entailed 14 inspector-hours onsite in the area of inquiry into the circumstances of a process upset and release of radioactive material to the environment.
Results:
Two violations were identified - Eight HEPA (high efficiency particulate air filters) had been improperly installed, resulting in degradation of the rated efficiency of the gaseous effluent treatment system, and failure to approve changes to transportation procedures prior to use.
8604150308 860408 PDR ADOCK 07001113 C
PDR.
1
REPORT DETAILS 1.
Persons Contacted Licensee Employees
- C. Vaughan, Manager, Regulatory Compliance
- B. Beane, Senior Fabrication Engineer
- A. Dada, Manager, Chemical and Ceramic Engineering
- S. Murray, Senior Nuclear Safety Engineer R. Schaeffer, Filter Change Supervisor
- C, Schiltz, Manager, Components and Fuel Fabrication Engineering
- P. Winslow, Manager, Licensing and Nuclear Material Management
- Attended exit interview 2.
Exit Interview The inspection scope and findings were summarized on February 5, 1986, with those persons indicated in paragraph 1 above.
A violation described in paragraph 4, eight HEPA filters were improperly installed resulting in degradation of the rate efficiency of the gaseous effluent treatment system, was discussed in detail. The licensee acknowledged the findings and took no exceptions.
On March 25, 1986, the licensee was informed in a telephone conversation between Mr. K. P. Barr of the Region II Office and Mr. P. Winslow, that the failure to have changes to the procedures for the transportation of radioactive material formally approved by licensee management prior to use would be considered a violation of 10 CFR 71.0(d).
This issue was previously discussed in Inspection Report No. 70-1113/84-04, 3.
Problem Description (93710)
On January 22, 1986, at approximately 3:00 p.m., the General Electric Company's Nuclear Fuel Manufacturing Department at Wilmington, North Carolina, experienced a uranium hexafiuoride conversion process problem and a stack effluent discharge of radioactive material. The problem conditions and discharges were of a nature and magnitude which did not require formal reporting under federal regulations or license conditions.
However, a courtesy notification was made to RII on January 23, 1986. Preliminary dose calculations based on effluent stack sample analyses indicated both onsite and offsite doses to be less than 10 CFR Part 20 limits.
The licensee's preliminary investigation indicated that improper installation of high efficiency particulate air (HEPA) filters was a contributing factor in the effluent discharge.
On February 3-5, 1986, a Region II inspector went to the GE - Wilmington facility to inspect the circumstances involving the discharge of radioactive
=
2 4
material and the plant conditions which led to the discharge. The inspector discussed the details of the process problem with licensee system engineers,
'i discussed the. associated ventilation and effluent treatment systems with plant engineering personnel, examined the affected filtration system i
accompanied by licensee personnel and discussed effluent radioactivity
- sampling, analysis, and dose determination with licensee regulatory compliance and health physics personnel. The inspector also reviewed the licensee's preliminary report prepared by a review committee consisting of plant staff.
The initiating event, which ultimately led to the release of radioactive materials, was what plant personnel termed a " flame-out" in the reaction 4
vessel of a proprietary process system for the direct dry conversion of uranium hexafluoride (UF ) t uranium oxide.
6 The event occurred in one of three process lines in operation at the facility. The licensee stated that all.three of the process lines had been shutdown and that none of the three lines would be restarted until any problems identified by the internal investigation are resolved.
No violations or deviations were identified.
4.
Radiation Protection (83822)
The inspector reviewed the radiation protection aspects of the event, including steps taken by the licensee to locate, identify, and quantify material which escaped from the process stream. Based on collected material and sample analyses, the following quantities were located and identified.
0.56 kg was found in the secondary hood around the reaction vessel. 0.28 kg was found in the water and sludge of the scrubber unit.
Based on weight measurements of representative prefilters, 2.5 kg was estimated to be on the i
prefilters upstream of the HEPA filter.
Based on the effluent air sample, operated at 56 liters per hour for 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> total sample time, approximately -
l 0.81 kg of uranium was calculated to have been exhausted through the stack.
I.
The HEPA filters could not be evaluated for uranium content until the radioactive decay products reach equilibrium; this required a 60-day delay l
before the HEPA's could be accurately scanned for uranium content.
The inspector discussed the health physics aspects of the event with j
licensee personnel. The initiating circumstances of the event resulted in the sounding of several audible alarms; the lead operator immediately instructed all personnel to-put on full face mask respirators. The event i
was brought under control by automatic control functions of the system and by the lead operator who quickly identified the cause of-a discharge from the reaction chamber and cut off nitrogen gas flow to the process system.
Several high volume air samples were collected in the immediate area of the f
event. The highest of these was 52 x 10 " pCi/cc (gross alpha activity) as compared to the 10 CFR 20, Appendix B, concentrations of 1 x 10 " uCi/ml for insoluble uranium and 5 x 10 " uCi/mi for soluble uranium.
Stationa ry n
4 n
3 work area samplers showed a high of 28 x 10 22 pCi/cc, averaged over an eight-hour run; the high sample was not the closest to the event location.
Urine samples were collected from all persons who had been in the work area and who were considered to be potentially exposed. Twenty-six (26) samples were submitted; the highest three samples were 17 to 18 ug/ liter.
The
{
internal action guide for urine samples was 35 ug/ liter.
4 The area was decontaminated and declared "off full-face masks" by 4:10 p.m.,
i based on high volume samples of less than 7.1 x 10 22 pCi/cc.
i Four individuals who were working in the process area at the time of the event were counted on the whole body counter. All results were at or below 1
detection limits.
Surface contamination surveys were made on the top of the building roof near the stack, and on the pavement downwind from the stack and smears were taken 1
on the pavement near the site warehouse. All surface meter readings were below detection limits.
Highest smear count was 22 dis / min (gross alpha) with a detection limit of 5 dis / min. Several samples of soil and vegetation were all found within normal ranges of 0.67 to 2.6 ppm (uranium).
Four continuous environmental air monitor samples were analyzed.
The highest value was upwind of the release and was 9.7 x 10 1' pCi/cc, for the South monitor station, based on an undecayed week-long sample.
No violations or deviations were identified.
1 I
5.
Environmental Protection (88045)
Exhaust gases from the process hood were ducted into the building ventilation exhaust system and were processed through prefilters and high efficiency particulate air (HEPA) filters. The filter system in use at the time was identified as No. 546-X, a Mine Safety Appliance (MSA) " caisson" system, with 32 prefilter-HEPA filter units, arranged in four rows of eight HEPA filters each. A redundant system was available for use as an alternate l
in the event of blower failure or excessive dust loading of the filters.
The process event condition was quickly brought under control, with no 3
j significant uptake of uranium by exposed process personnel, as demonstrated j
by results of urine samples from all those involved.
In the immediate post-event time period, from about 3:10 p.m. to 6:00 p.m.,
the release of measurable quantities of uranium oxide dust from the plant was not considered likely, based on the presence of the HEPA (high efficiency particulate air) filter system in the ventilation: exhaust train.
However, when the effluent stack air sample was removed at 6:00 p.m.
to-verify the asstration, an abnormally high level' of alpha activity was detected when an immediate-count was made.
The presence of naturally-occurring short-lived alpha-emitters in air samples made prompt analysis of the effluent sample difficult. Rather than 4
A 4
e.
wait the usual seven days for decay of the short-lived materials, the air sample was dissolved in nitric acid and the uranium content analyzed by X-ray fluorescence.
Based on a sample run (collection) time of ten hours, the analysis indicated d
an average 10-hours concentration of 3700 10 22 pC1/cc, representing approximately 810 grams of uranium discharged. Actual site meteorological data was used to calculate concentrations at points downwind and a 24-hour release was assumed for additional calculations.
Concentration at the restricted area fence (259 meters downwind) was calculated to be 1.1 x 10 22 1
l pCi/cc, (24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />) assuming a ground level release.
The closest point inside the fenced area at which the plume was expected to touch the ground was 13.5 x 10 22 pC1/cc at 150 meters downwind. The projected concentration at the site boundary (945 meters) was 0.27 x 10 22 pCi/cc.
The closest
=
occupied dwelling in the downwind direction was approximately k mile from the site boundary. The unrestricted annual average MPC is 4 x 10 22 pCi/cc I
for insoluble uranium.
The inspector reviewed sample analytical data and printouts of computer calculations of onsite and offsite concentrations of airborne particulates and discussed the results with staff engineers and health physicists.
The inspector determined that the results were satisfactory.
No violations or deviations were identified.
6.
Improper Installation of HEPA Filters (88045)
]
When an abnormal level of airborne effluent activity was identified at 6:00 p.m. (January 22, 1986) building exhaust airflow was diverted from the 546-X filter system to the 541-X filter system as a precautionary measure. A work order was initiated on January 23, 1986, to replace-all of the filters in the 546-X system. On the afternoon of January 23, 1986, during preparations for the filter change operation, it was discovered and reported that one bank of eight.HEPA filters was found in a condition which indicated that the filter clamping devices had apparently not been employed when the filters were installed This resulted in the existence of a gap of approximately 7/16 to 1/2 inch thickness on all four sides of all eight filter units (24 inch x 24 inch).
This provided a large leakage or bypass path whereby unfiltered air was able to bypass the filtration media.
In the GE installation, the HEPA filters are mounted horizontally, in MSA
" caisson" units, with the intake side of the HEPA filter down, and with the gasketed edges at the top (downstream side). When inserted, the bottom of the filter rests on two metal bars and there is about 1/2 inch clearance at i
the top (in this case, the bypass path).
To-complete the installation, a special tool is used to turn a " scissors-jack" arrangement, which applies spring-loaded tension to the bars on which the filter is supported, raising the filter to contact the sealing face of the filter housing. The jack mechanism is turned until the filter edge gaskets are compressed to about one-half their original thickness, completing.the sealing of the filter to the filter housing.
m m..
m
o 5
On inspection of the other 24 filters in the filter system (546-X), two filters in one bank were observed to be installed such that the gaskets were apparently in contact with the housing seal surface but with the gaskets not fully compressed. A licensee representative stated that there was visible evidence that some bypass had occurred; however, the extent of bypass could not be determined but was considered small compared to the eight HEPA filters where the gaskets were 1/2 inch from the seal face.
The inspector reviewed the procedure for installation of HEPA filters, No. 340-FME-85-70, "HEPA Filter Change and Certification," Revision 0, dated March 18, 1985. The inspector noted that this procedure was titled "HEPA Filter Change and Certification" but that the title was misleading in that no certification was provided as assurance that filters had been properly installed.
At one point in the procedure, instructions were: " Install new filter and tighten filter housing locking clamps.
Check to see that the filter is seated correctly." In discussions with licensee personnel, the inspector noted that this step of the procedure lacked several specifics, such as which side of the filter should be up, and guidance as to how tight the filter should be in the housing and to what degree the gasket material should be compressed.
Licensee representatives agreed to revise the procedure, and stated that the procedure would provide "one-over-one" certification for each filter installed by th installer and by the job foreman or supervisor.
The inspector discussed the conditions of the filter change task during the installation in question. Licensee representatives stated that two persons normally did all of the plant's filter change work.
On this occasion, however, due to the number of filters to be changed in 546-X and other units in the same time period, a crew of nine persons was assembled, including the two regular filter changers. The seven additional persons had previously been trained in this task and had satisfactorily changed filters before; however none of the seven had been involved in filter changing over the most j
recent period of several months.
Licensee representatives stated that it was considered likely that one or more of the seven persons on the crew who had not recently participated in filter change operations had been working on the affected filter bank and had neglected to complete a step in the procedure. As of the time of this inspection, no specific identification of any responsible person (s) had been made.
At the time of the event, filter unit 546-X was operating at a total flow of approximately 28,000 cubic feet per minute with a Ap (pressure drop) of approximately 1.4 inches water gauge (1.4" w.g.) across the 32 combined prefilters and HEPA filters.
The prefilters and HEPA filters had been changed approximately 10 weeks earlier and initial pressure drop had been approximately 1.0" (w.g.).
The inspector calculated that the difference in pressure drop resulting from the failure to properly install th eight HEPA filters would have been on the order of 0.05 to 0.10" (w.g.).
Since
o 6
s, variations in pressure drop across each filter are also of this magnitude, the error in installation would not expected to be observed by this measurement alone.
The current licensing conditions for the G.E. facility do not require leak testing with dioctyl phthalate (DOP) smoke and GE does not perform this test. When questioned by the inspector as to why DOP leak testing was not used, a licensee representative stated that it was the opinion of the plant staff that leak testing was not sufficiently sensitive to assure the_ degree of efficiency which HEPA filters were capable of - and that careful installation of filters could assure a greater degree of filter effectiveness than leak testing. When the inspector pointed out that the site procedure had fai. led to assure filter effectiveness in this. case, it was stated that a revision to the procedure was being proposed which would add several quality control steps to assure that improper installations of filters would not happen again.
Licensee representatives stated that a program had been initiated to re-check the current installations of all HEPA filters in all filter units at GE-WMD.
As of January 28, 1986, the program was approximately 60%
complete; no additional defective installations had been detected.
License Condition 9, of License No. SNM-1097, Docket No. 70-1113, requires the licensee to operate in accordance with the statements, representations, and conditions of Part 1 of the License Application dated May 14 and June 20, 1984. Part 1, Section 2.7 of the License Application, " Operating Procedures - Administrative Controls," states, in part:
" Area managers shall assure preparation of written nuclear safety control procedures incorporating limitations established by the criticality and safety functions." Section 2.7.1 further states, "The radiation protection program i
is designed to establish and maintain a comprehensive set of written instructions for radiation health and safety practices so as to maintain occupational radiation exposures at levels as low as reasonably achievable."
Environmental Protection Instructions are included in a tabulation of items covered by Section 2.7.1.
Section 3.2.2.4,
" Exhaust Systems" of Part I states, " Exhaust systems in potentially contaminated airborne effluents are equipped with filter media which are selected to maintain its integrity when subjected to chemicals, solvents, and abnormal operations or processes.
These filters are at least 99.97% efficient for removal of 0.3 micron particles."
In November 1985, plant personnel working under the instructions of Procedure No.
340-FME-85-70, "HEPA Filter Change and Certification,"
Revision 0, dated March 18, 1985, incorrectly installed eight HEPA filters, of a total of 32 HEPA filters installed in. the 546-X filter system, in a manner such as to allow substantial bypass of the filter media. This action resulted in the compromise of the rated filtration efficiency of the entire bank of 32 HEPA filters.
The inspector determined' that Procedure No. 340-FME-85-70 lacked several specific details essential to the correct
~
installation of HEPA filters and despite the use of the word " certification" in the pr;cedure title, contained no provision or requirement for
T 1
?
l 7
i f
i supervisory or worker certification, either by initial, signature, or
" check-off" box, to provide certification, confirmation, or assurance that the HEPA filters had been satisfactorily installed. The inspector concluded that the procedure was inadequate for the purpose for which it was written I
in that it (1) did not provide sufficient guidance to filter change personnel to assure that the task could be performed and (2) failed to provide for certification or other means of verification that the task had been performed in a satisfactory manner.
The circumstances detailed above violated License Condition 9 in two respects.
First, the filter change procedure provided under Section 2.7,
" Operating Procedures - Administrative Controls" of Part 1 of the License Application was determined by the inspector to be inadequate for its intended purpose, and second, that the integrity of the Ventilation Exhaust System HEPA filter media specified in Section 3.2.2.4, " Exhaust Systems," of Part 1 of the License Application was compromised by the incorrect installation of eight HEPA filters, which directly contributed to the release to the environs of approximately 800 grams of uranium oxide on January 22, 1986.
The two sets of circumstances detailed above were combined into a single violation of License Condition 9 (70-113/86-01-01).
I Licensee representatives stated that procedural revisions had already been drafted which would minimize the potential for a recurrence of the release of radioactive materials and that two internal investigations were being conducted from which other relevant corrective action recommendations could be expected.
The inspector informed licensee representatives that the l
procedure revisions, reports, and corrective actions would be reviewed at a later date.
One violation was identified. No deviations were identified.
i i
i 7.
Approval of Transportation Procedures (86740)
.i 1
I During a previous inspection (Inspection Report No. 70-1113/84-04), the
]
inspector noted that the licensee's transportation-procedures FS-7, i
Radioactive Material Shipments. and FS-36, Low Level Radioactive Waste-Shipments, had not been formally revised to incorporate the-1983 changes to the Department of Transportation (DOT) regulations contained in 49 CFR and NRC regulations contained in 10 CFR 71. The individual responsible for the shipment of radioactive materials had a draft copy of the-procedures which 4
did include the changes to the DOT and NRC regulations. The shipping forms-for radioactive material had been revised. In addition, review of shipping records by the inspector indicated that shipments made by the licensee had
~
been in accordance with the revised DOT and NRC regulations.
During this inspection, the inspector concluded that no violations or deviations of NRC l
regulations had occurred, since the inspector did not identify an NRC requirement for licensee management to formally approve the procedure uad-by the licensee for the transportation of radioactive material.
a A recent review of this issue by the Regional Office has identified a specific NRC requirement that changes to transportation procedures are to be formally reviewed by the licensee and approved.
10 CFR 71.0(d) states that the transport of licensed material or delivering of licensed material to a carrier for transport is subject to the operating controls and procedures requirements of Subpart G of this part, to the quality assurance requirements of Subpart H and the general provisions of Subpart A,
including DOT regulations referenced 'in 10 CFR 71.5.
10 CFR 71.113, under Subpart H, requires that the licensee establish measures to control.the issuance of documents such as instructions, procedures, and drawings, including changes, which presents all activities effecting quality. 10 CFR 71.113 further requires that measures must assure that documents, including changes, are reviewed for adequacy, approved for release by a*.:thorized personnel, and distributed and used at the location where the prescribed activity is performed.
Failure of the licensee to establish measures to assure that changes to the transportation procedures had been approved for release was identified as an apparent violation of 10 CFR 71.0(d) (70-1113/86-01-02).
i