ML19327B292
| ML19327B292 | |
| Person / Time | |
|---|---|
| Site: | 07000036 |
| Issue date: | 10/18/1989 |
| From: | France G, Mallett B, Sreniawski D, Sturz F NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| To: | |
| Shared Package | |
| ML19327B291 | List: |
| References | |
| 70-0036-89-03, 70-36-89-3, CAL-RIII-89-20, NUDOCS 8910300109 | |
| Download: ML19327B292 (70) | |
Text
p I! iLied U.S. NUCLEAR' REGULATORY COMMISSION o
REGION III
/[
Report No.
70-36/89003(DRSS).
Docket No.
70-36 License No. 'SNM-33 e
Licensee:
Combustion Engineering, Incorporated Nuclear Power Systems i
Windsor, CT.06095 i
Facility Name:
Hematite Facility E
F
-Inspection At:
Hematite, Missouri L
Inspection Conducted:
August 30, 31, and September 1, 7 2, and 23, 1989 i
Inspection Team Leader:
D. J. Sreniaws 1 hief
/t4////
Nuclear Materials Safety' Date 5
Inspection Team Members:
. France, III
/0[/k/?U 1
gi 1 III
.Date l WIN h
. Sturz, NMSS
/o/f3//9d?
kbidd6 Approved By:
Bruce S. Mallett, Ph.D., Chief
/0[//[P/
Nuclear Materials Safety Branch Date Inspection Sumnary Inspection Conducted August 30, 31, and September 1, 7, 8, 22, and 23, 1989 (Report No. 70-36/89003(DRSS))
Areas Inspected:
An Augmented Inspection Team (AIT) composed of Region III and NMSS personnel investigated the circumstances regarding the unplanned release of enriched uranium containing approximately 4% uranium-235 and the licensee's actions to protect onsite personnel and the public.
The AIT also reviewed the licensee's operating procedures, process equipment and monitoring
. systems.
Results:
The AIT concluded, after review of information provided by CE and by direct observation, that the licensee's initial estimate of 274 grams of uranium released appears to be a valid estimate based on a stack sample and survey readings outside the building.
The release was due to failure of a nitrogen valve which pressurized the conversion reactor.
Prompt action was taken to assess worker exposure.
Weaknesses were identified in the detection of conversion process failure, communication of inoperable equipment status and in the scope of the inf tial environmental surveys to characterize the
- release, e910300109 891019 ApoCK07oog6 PDR C
s
Y lc c
i-c DETAILS i
1.
Persons Contacted R. Betlock, Production Supervisor B. Pigg, Quality Control Laboratory Supervisor L. Deul, Manufacturing Engineer H. E. Eskridge, Manager, Nuclear Licensing, Safety and Accountability R. W. Griscom, Plant Engineering Supervisor C. Heisker, Engineering Specialist R. Miller, Manager,-Administration and Production Control A. Noack, Production Superintendent J. A. Rode, Plant Manager G. Uding, Quality Assurance Engineer During the onsite investigation members of the NRC Augmented Inspection Team (AIT) discussed the events regarding the release of uranium with the above listed members of the licensee's staff.
In addition, several operators were interviewed and several members of the licensee's Radiation Protection Department assisted the AIT in the collection of environmental samples.
2.
Attendees of Meeting at Region III Office which was held on September 19, 1989 Combustion Engineering l
J. A. Rode, Plant Manager, Hematite l
H. E. Eskridge, Manager, Nuclear Licensing Safety and Accountability l
C. R. Waterman, Vice President and General Manager, l
Nuclear Fuel Manufacturing A. E. Scherer, Director, Nuclear Licensing NRC NMSS/ Region III L. C. Rouse, NMSS, Chief, Fuel Cycle Safety Branch, Division of Industrial and Medical Nuclear Safety Branch F. C. Sturz, NMSS Senior Project Manager / Health Physicist, Irradiated Fuel Section, Fuel Cycle Safety Branch D. McCaughey, NMSS, Nuclear Process Engineer, Uranium Fuel Section, Fuel Cycle Safety Branch A. B. Davis, RIII, Administrater C. E. Norelius, RIII, DRSS Division Director R. Lickus, RIII, State Liason Office L
D. J. Sreniewski, RIII, Section Chief G. M. France, III, RIII, Fuel Facility Inspector 3.
Normal Licensee Program The Combustion Engineering (CE) facility of Hematite, Missouri, produces uranium dioxide (U0 ) fuel for the commercial nuclear power industry.
2 I
Low enriched uranium (maximum 5% U-235) is received from uranium l
enrichment facilities as uranium hexafluoride (UFe) in 2 ton, 30 inch l
diameter cylinders.
ufo is converted to UO2 powder and/or pellets.
1 1
2 C
g L
o, h
a.
Uranium Hexafluoride Conversion to Uranium Dioxide (UFs to 002)
Prior to-startup the conversion reactors are purged with nitrogen to remove condensation and to purge the off gases.
During startup, the e
L nitrogen purge is replaced with steam when the operating temperature is reached.
Concurrently, a cylinder of UFe is heated within a steam chamber until the vapor pressure allows the vaporized ufo to l
flow to reaction vessel R-1, the fluidized bed reactor (Attachment D, Incident Report, Page Five, Figure 2,' Hematite Oxide Conversion i
Process).
In R-1, the UFs reacts with process steam to form uranyl fluoride (U0 F ), hydrogen fluoride (HF) and water vapur.
L 22 Gaseous HF'and H 0-(as excess steam)-exit the reactor through two 2
sets of porous metal filters.
The gaseous materials are routed to the HF removal system.
The HF removal-system, consists of five t
cylindrically shaped towers each packed with about 1500.lbs of pebble sized limestone (calcium carbonate, Caco ).
HF is removed by the a
reaction of F with Ca to form CaF.
Excess H 2
2 in the off gas stream is burned in an in-stack burner located after the limestone fluoride removal system.
Within the process UO F2 2 particles which are formed in reactor R-1 pass to a second and, third reactor R-2 and R-3 in series, where UO F2 2 reacts with hydrogen (obtained from cracked ammonia) to form UG.
2 The off gases from R-2 and R-3 are also filtered through porous metal filters, after which the gases pass to the limestone scrubber /HF removal system and out the conversion process plant stack.
This stack effluent is sampled by passing through a haat-traced stainless steel sample line at about 400 to 900 cubic centimeters i
per minute through two liters of KOH 10% solution which is analyzed for fluoride weekly.
Product UO is removed from the R-3 reactor via pneumatic transfer 2
g to the storage silos, for subsequent use in the U02 pelletizing operation.
Progress of the chemical reaction (HF + Caco 3) through the limestone scrubber bed is monitored by noting the scrubber bed temperature.
In addition, a portion of the spent limestone is removed from one or two of the towers during each shift so that about 1 1/3 operating r
days complete the replacement.
Each tower bed is replaced normally in sequence of one to five, oy closing the selective off gas manifold valve and dumping the contents into two 55 gallon drums per tower.
The scrubber loading system is used to replace the limestone.
3
A drum of spent limestone from every tower is rotated to mix the contents and surveyed for radiological content.
The reported activity of this spent material is usually less than 150 dpm/100 cm,
2 In normal operations, the porous metal filters (located inside R-1) along with the secondary filter (located in the off gas stream) removes UO F2 2 which prevents the buildup of radioactivity.
- However, the licensee indicated that unreacted UFs gas could pass through both sets of porous-metal filters associated with R-1 and collect on the limestone.
b.
Routine Health Physics Practices to Protect Onsite Staff Health and safety procedures are provided to protect workers engaged in plant operations.
They describe in detail how operating personnel are protected from excessive internal expoLure from uranium by controlling ventilation air, sampling the air in work areas, using protective clothing and respiratory protection.
equipment, and surveying for and decontaminating radioactive surface contamination.
In addition, workers are scheduled for bioassay and whole body counting.
Process workers submit urine bioassay specimens.each ironth (for analysis at CE's Windsor, Connecticut facility).
Whole body counting is performed at least annually.
The ventilation system provides a negative pressure so that air l
flow is from work areas into process equipment, hoods, or glove boxes.
The system is designed to move air from areas of low I
L potential contamination to potentially higher contaminated areas.
I' The workers breathing air is continuously sampled using fixed sampler heads mounted at various locations.
These samples are l
changed and analyzed for each shift.
Continuous air monitors (CAM)
I with alarms are also used to warn of high uranium air l
concentrations.
1 c.
Licensee's Offsite Survey Program Effluents are monitored to determine the amount of material released from the facility and compliance with NRC release limits.
Monitoring for uranium and fluorides in air occurs at the points of i
discharge.
In the Pellet Plant eight building stacks which exhaust process ventilation air are continuously sampled for uranium.
Exhausts j
from laboratory fume hoods which handle wet chemicals, and two of the three room air exhausts in the Pellet Plant dewaxing and sintering furnace area are not continuously sampled for uranium.
The Oxide Building (scrubber system) off gas is continuously sampled for fluoride.
The licensee's environmental radiological monitoring program
)
consists of collecting air, water, soil, and vegetation samples at L,
various onsite and offsite locations.
Continuous particulate air l
samples a m taken at two locations.
The air samples are collected l
4 l
(
[7 4
generally north and south of the site, and are analyzed weekly and composited quarterly.
Surface water is sampled monthly at Joachim Creek above and below the site creek outfall and quarterly from E
Joachim Creek both up and down stream of the property.
Well water samples are drawn quarterly from onsite wells and one well in Hematite.
Soil and vegetation samples are collected quarterly at five locations onsite surrounding the plant.
Attachment E dated January 24, 1983, (location of monitoring sites around Hematite Facility) shows the sampling locations.
4.
Description of Event On Friday, August 25, 1989, during the midnight to morning shift after i
shutting the conversion process down for the weekend, the Production Supervisor observed an air leak on the solenoid that supplies air to operate the R-2 nitrogen valve.
He closed the air valve to reduce plant air consumption and noted his actions on page 81 in the foreman's log.
He also noted that the valve needed repair prior to system startup on Monday, August 28, 1989.
Shutting the air supply caused the nitrogen p
valve t) fail open.
On Monday before 7 a.m., both the day shift Production l
Supervisor and Production Superintendent read pages 82 through 87 of the foreman's log, but failed to see the entry on page 81.
Consequently, the air valve was not opened nor was the solenoid repaired.
At 11: 50 a.m.
the control panel showed " normal" steam flow and normal ufo flow at 110 lbs/ hour.
Both nitrogen and steam have control valves on a common line.
However, the flow indicator cannot differentiate between steam or L
nitrogen flow.
With the nitrogen valve in the failed /open mode, the l-nitrogen pressure exceeded the steam pressure (30 vs. 25 lbs.) and permitted process steam in R-1 to be diluted with nitrogen.
As a result, l
ufo was not converted to UO F.
The unreacted UFe gas passed through 22 the internal R-1 and the secondary porous metal filters and reached the limestone bed scrubber system, where most of the material was trapped.
I At 1:05 p.m., on August 28, 1989, the continuous air monitor (CAM) on the fourth floor of the Oxide Plant alarmed and the ufo flow was shut down and the upper floors were posted as respirator areas.
The alarm was assumed to have resulted from loading of the bed into R-1.
The bed refers to the starting material which is comprised of UO F.
22 After finding no unusual conditions, at 1:40 p.m., the shift supervisor f
restarted the conversion process (ufo flow was restarted).
At 2:20 p.m.,
the CAM alarmed again and the system was shut down.
Dense fumes from the dry scrubber exhaust blower were noted to be drifting in a northeast direction disappearing into the trees.
At 3:00 p.m., after hearing of the startup difficulties, the Production Supervisor who first made the log entry on Friday, August 25, had the solenoid air valve repaired.
5
j
- k At about the same time a heavy thunderstorm drenched the area, washing away any soluble portion of the release thus making detection more difficult.
The system was inspected for leaks, purged with nitrogen, brought up to operating temperature with steam and at 8:40 p.m., the conversion process was resumed and no other problems were encountered.
At.7:30 a.m., August 29, 1989, the Production Superintendent noted a low UO2 production rate and requested Health Physics to survey the spent limestone that had been unloaded from the scrubber column.
The Plant Manager was informed that the rock from three scrubbers read 30,000 to
-i l
70,000 dpm/100 cm2 2
(normal is 150 dpm/100 cm ) and at 2:10 p.m., he ordered the Production Superintendent to shutdown the conversion process.
On August 29, 1989, the fluoride bubbler stack sample was analyzed for l'
uranium and the results indicated a release of 274 grams of uranium.
At
.3:10 p.m., the Nuclear Licensing Safety, and Accountability (NLSA) Manager notified the NRC (Region III Fuel Facility Inspector) of the event.
Initially, the' licensee thought leakage from a break in the porous metal filters caused the unplanned release.
On A'ugust 30, 1989, after a technical review of the incident, the licensee concluded that the amount of material was too high to have been caused by leakage from small breaks in the metal-filter.
l Meanwhile, the licensee collected initial urine bioassay samples from l
nine workers for emergency evaluation.
Followup fecat samples (three) and urine samples (nine) were collected on August 31 and September 1, 1989.
Following the discovery (8/29) of the accidental release of uranium, the licensee collected three soil, three vegetation, and three standing water samples from seven locations outside the fence line, but within the plant site boundary.
The samples were collected in the general direction (N-NE) that the exhaust plume was observed to have touched ground the preceding day.
The particulate air sample from the offsite east station, as well as a sample of onsite creek water, were also collected.
The area to the north east of the Oxide Building was surveyed with portable survey instruments to locate potential contamination.
Additional soil, water, and vegetation samples were collected onsite and outside the fenced area by both the ' licensee and the AIT on September 1 and September 7.
On August 31, 1989, CE determined that the cause of the release was due to incomplete conversion of ufo to UO2 because of nitrogen dilution of the process steam in reactor R-1.
Nitrogen was present because the failed solenoid valve allowed nitrogen to enter R-1 through R-2.
This evaluation was based on a material balance of uranium product and an analysis of the chemical and physical form of the uranium found in components of the conversion process and the HF scrubber system.
Continuing efforts were made to account for the 64.8 KgU which was fed into the reactor for the one hour and 55 minutes run time.
From about August 29 to September 18, 1989, the reactor was disassembled and the components (piping, etc.) were flushed and the contents were collected 6
v.
a 3
' 4 I
and weighed for uranium accountability. As of September 19,.1989, all but 3.1 Kg were recovered.
The licensee stated that such a level of entrained 1
materia 1'would not be unusual for normal operations.
NMSS (HQ NRC) concurred that it is reasonable to assume that 3.1 Kg of uranium could be retained in the hardware of the conversion process. '
5.
Evaluation of Event a.
Plant Operation The AIT review of the foreman's logbook and interviews with the Production Supervisor and Production Superintendent and Plant r
Manager confirmed that the logbook entry as described in Section f
Four was not adequate to assure that repairs on vital equipment were completed.
In addition, the Production Supervisor failed to recognize the safety significance of the failed solenoid in the s
nitrogen control valve.
This resulted in management not being informed of the problem in a timely manner.
It wasn't until the production superintendent noted a reduction in product output that a significant effort was made to evaluate the situation.
This reduction in U02 product was noted between 7:30 e.m. and 9:50 a.m.,
(
on August 29, 1989.
L Interviews with the Plant Manager, control room operators, and observations of the control room instrumentation showed that the flow indicator could not differentiate between steam and nitrogen.
As a result, the control room operators were unaware of the degraded uteam supply which in turn affected the hydrolization process which converts UFs into UO F.
22 b.
Onsite Exposure The AIT also evaluated workers' potential internal exposure by reviewing the results from the licensee's urine and fecal bioassay samples and data with the highest results from fixed air samples s
located at the work stations.
The highest uranium levels from the nine workers were reported as 3.1 pg/ liter (urine) and 0.47 pg U/g (fecal).
(See Pages 34, 35, Attachment D, Incident Report) The licensee has an action level of 25 pg U/ liter (urine) before any formal investigation is performed.
Inhalation exposure based on the fixed air sampler data showed a maximum of 22 MPC-brs and compared favorably with exposure levels that the workers experienced during routine operating conditions.
Whole body counting is scheduled for October 10, 1989, as an additional backup.
Based on both the licensee's findings and the inspectors' investigation, the safety significance and/or the radiation exposure levels to workers were minor and the licensee's action to evaluate them were acceptable.
7
p
- z e
o c.
Environmental Activities I
When the AIT arrived onsite on August 31, 1989, they were initially informed that the unplanned release was caused by a break in the l
R-1 filter.
The licensee was also uncertain as to the duration of the release and the amount of uranium that had been released to the l
environment.
The conversion process was shut down and had been since August 29, 1989.
A plume characterized by the licensee predominantly t
as steam a;.d HF had been sited drifting towards the north east on
+
August 28, 1989.
A review of the licensee's results for environmental direct surveys showed no levels greater than background.
The licensee's initial soil, water, and vegetation samples also showed no evidence of any depositions from this release.
After review of the meteorological data (wind speed and direction vs. time, Attachment D, Page 21, Figure 5, Page 22, Page 23, Figure 6), and discu5 sir.a with the licensee additional soil, vegetation, water temples, and direct surveys were taken in other areas to account for wind shifts.
After the discovery of the unplanned release, the licensee was not aggressive in its environmental sampling efforts.
At the time of the initial sampling, the time, duration, and arount of the release was not fully known.
The release could have occurred over about a 26 hour3.009259e-4 days <br />0.00722 hours <br />4.298942e-5 weeks <br />9.893e-6 months <br /> period during which weather data indicated the wind to be from almost all directions.
Yet the northeastern quadrant was the only one surveyed because of one observation of a plun In retrospect, this assumption proved valid.
However, more t w iles should have been obtained in different directions around the facility to firmly establish the existence of or lack of deposited material from the release.
The licensee did not pursue more sampling until directed by the AIT.
The licensee's emergency procedures / radiological contingency plan should include a more comprehensive environmental sempling program.
The number and location of air sampler stations is not adequate.
For example, the site boundary to the north has changed since the sampling system was implemented.
Consideration should be given to adding an air sampler or relocating the existing station more to the northeast to provide better coverage of the newer site boundary and the closest resident in that predominate wind direction.
Consideration should also be given to adding more air samplers to provide more directional coverage surrounding the plant.
The environmental air samplers are out in the open and appeared weathered.
The cffsite east air sampler was not operating.
Eventually environmental sampling and surveys were conducted at all points of the compass and at one of the nearby residents.
The soil, water, and vegetation samples were
'it with the licensee.
The NRC samples have been submitted to a DC contractor for an independent analysis.
Results reported by the licensee for both initial and split samples (Attachment D, Pages 26-28), do not differ substantially from the results of routine environmental samples.
According to the licensee's report, the highest value (20 1 7 pCi/g alpha) for soil samples was not indistinguishable fro:n background.
8
.a
L
}
es i
The scrubber stack sampler is used to monitor HF, (fluoride), but g.
not' uranium releases.
As a result, there is no routine evaluation of uranium releases from the conversion process stack.
The direct surveys of the limestone in the scrubber served as a general o
indicator as to the lack of uranium in the effluent.
The principal uranium compounds that constituted the plume could not be determined.
Because of the potential for a mixture of soluble and L,
insoluble uranium compounds released, and because solubility determines the amount of inhaled uranium which is transferred and retained in a particular body organ, critical organ doses for both soluble and insoluble uranium exposure were estimated.
For low enriched soluble uranium compounds inhaled, the critical organ for an acute uranium intake is the kidney, based on chemical toxicity.
The critical organ for the effects from radioactivity is the osteogenic cells, l
particularly those on the endosteal surfaces of bone.
For insoluble uranium compounds inhaled, the critical organ is the lung.
The maximum impact of the accidental release of uranium was assessed by calculating uranium concentrations at various distances towards the nearest resident and in the direction the plume traveled.
The amount of uranium that could have been inhaled by an individual located at those points offsite was then estimated to determine the dose.
Fifty year committed organ dose equivalents and committed effective dose equivalent that would occur from acute (two hour exposure in the plume) intake of uranium isotopes during the lifetime (50 years) of the individual were calculated.
For soluble uranium, a bone dose was calculated; for insoluble uranium, a lung dose was calculated.
The committed effectiva dose to the total body was also calculated for both soluble and insoluble uranium.
Approximately 300g of uranium was assumed (for calculation purposes) to be released to the atmosphere over a two hour period.
The isotopic composition (activity) of the uranium was assumed to be that of uranium enriched to four weight percent of the isotope uranium-235 (U2ss).
A lung solubility class "D" was assumed for soluble uranium compounds and Class "Y" for insoluble compounds.
Atmospheric dispersion factors were estimated based on the equations for relative concentrations at an area boundary for two hours immediately following the accident found in NRC Regulatory Guide 1.145.
(Atmospheric Dispersion Models for Potential Accident Consequence Assessments at Nuclear Power Plants.) These values were calculated using meteorological conditions prevailing during the accident.
Numerical results for selected points in the plume path downwind at distances up to 1000 meters were estimated.
The plume resulting from the accident generally traveled in a NE direction but shifted up to 60 degrees to either side.
9 l
y o
t l
At the location of the nearest res' dent, approximately 400 meters r
NE, an individual (assumed present) would have been exposed to an estimated average uranium air concentration of about 9E-04 mg-U/m8 i
(or 2E-12 pCi/ml).
This concentration is about 13 percent of the i
NRC's annual average permissible concentration for uranium (0.007 8
mg/m ) in air for unrestricted areas, listed in 10 CFR Part 20, Appendix B.
Exposure to this concentration for two hours would result in an inhalation intake of about 0.002 mg of uranium.
The NRC previously reported in NVREG-1189 (Assessment of the Public Health Impact From the Accidental Release of UFe at the Sequoyah j
Fuels Corporation Facility at Gore, Oklahoma) that an absorbed dose of 25 pg U/kg body weight (i.e., total intake of 1.8 mg) can be considered a minimal dose for inducing nephrotoxicity in the i
kidney.
The estimated intake of 0.002 mg uranium is far below this i
threshold, and therefore no effects due to chemical toxicity would i
be detectable, j
For an assumed 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> inhalation in the plume, the estimated maximum uranium intake for an individual is about 3 pCi.
If the uranium were all soluble, then this intake would result in a maximum bone dose of about 0.1 mrem and an effective total body dose less than 0.01 mrem.
If it were all insoluble uranium, then this intake would result in a maximum lung dose of about 3 mrem and an effective total body dove about 0.4 mrem.
Doses due to chronic exposure to j
inhalation of deposited uranium resuspended in the air and to ingestion of vegetables, milk, and meat contaminated from uranium deposited in the soil would contribute only a small fraction of that calculated for the acute plume inhalation pathway.
External exposures from submersion in the radioactive plume and from surface contamination of the soil via uranium deposition are insignificant (less than 0.1 percent) compared to other exposure pathways.
Factors such as plume meander, wake effect of wooded areas, and particle disposition would be expected to further reduce these i
calculated maximum doses.
Regardless, no measurable radiological impacts are expected.
Even if all of the uranium unaccounted for (3 kg V) had been released, any resultant offsite exposure would still not cause significant risk to an individual.
Details of the dose calculations used for this section are located in Attachments A, B, and C to this report.
6.
Licensee Meetina On September 19, 1989, Combustion Engineering and NRC staff met in Region III to discuss operational and engineering changes that were to be completed in antic 6ation of restart of the conversion process.
CE presented three problems that were identified as a result of the event:
10
y c
Failure to recognize a potential problem in the nitrogen and steam header feed line.
I Failure in the system for communicating and documenting the need for maintenance.
t Failure in training.
Operating staff was not experienced or trained to recognize the lack of ufo conversion.
CE also indicated that the following corrective actions would be implemented prior to start up:
j Interlocks for each nitrogen valve position have been installed to shut off the UFc flow automatically if the nitrogen valve is not in the closed position.
The lockout mechanism for ufo was installed on the control panel.
A maintenance log is maintained in the control room and requires a release of all entries by signature.
It is the responsibility of the shift supervisor and/or production supervisor to confirm that maintenance service is completed prior to startup.
Training was given to the operating staff to assure that every oxide operator is aware of complete / incomplete UFc conversion.
A dual purpose off gas sampling system was installed to perform isokinetic sampling for particulates and low volume sampling for fluroide.
The system will be tested in place.
CE provided a report dated September 1989 and received in Region III on September 28, 1989, describing the problems and their proposed corrective actions (Attachment D, Incident Report, Pages 36-38).
7.
Pre-Startup Inspection On September 22 and 23, 1989, a Region III inspector was dispatched to CE Hematite to observe the startup of the conversion process.
The licensee had not operated the process since the day shift on August 29, 1989.
During the interim period, the licensee installed / implemented a number of engineering and/or administrative l
controls as part of the corrective action effort to prevent a recun ence l
of the unplanned release.
(Discussed under Section 6. Licensee Meeting).
The inspector reviewed the licensee's procedures for preparing the conversion process for startup and observed prestart operat. ions in the control room.
The inspector confirmed that the ufo lock out key was in possession of the Production Supervisor and the lockout mechanism was installed on the control room panel. It was observed that the supervisor i
reviewed and initialed the maintenance sheet tc assure that all entries had been noted and that process components were operable prior to 1
u o.
unlocking the valve to allow UFe to flow into the R-1 reactor.
The maintenance sheet is used as an administrative control to assure that j
each operating shift supervisor can identify critical items for service.
The sheets are stored in the control room for viewing by management.
The licensee noted that entries in the foreman's log will norina11y address j
the less critical items for service.
Green and red indicator lights were installed on the control panel to identify individual interlocking for each of three valves that control nitrogen to reactors R*1, R-2, and R-3.
By engaging the interlocking i
devices the appropriate indicator light will show the operator whether the nitrogen valve is opened or closed.
The Engineering Supervisor discussed the modifications that were installed to improve stack sampling of the HF gas scrubber system, i
The inspector observed that the licensee had added an isokinetic stack i
sampler for particulate uranium analysis.
At least once each shift, the filter paper is exchanged and after a brief storage to eliminate decay products, is analyzed for uranium.
The stack sampler has been modified to enable the licensee to determine the levels of uranium and fluoride discharged from the HF serubber.
Fluoride determinations will continue to be made on a weekly basis.
The licensee agreed to monitor the stack i
scrubber solution for soluble uranium prior to first use and again after i
replacement to detect uranium vapor that may have bypassed the particulate filter.
The inspector also noted that the ventilation system blower was upgraded to provide more air flow.
At about 2:00 p.m., the licensee restarted the conversion process.
At 10:00 p.m., via telecommunications, the inspector was informed that
)
the process was still ongoing.
The licensee noted that the steam /HF l
plume that normally is visible during the conversion process seemed significantly smaller.
This may have been caused by greater air flow due to the upgraded blower. On September 23, 1989, the second cylinder of UFc was brought on stream.
The conversion process was continued at 2:38 a.m.
The inspector also confirmed by discussion and record review that 83 plant workers received training about the safety significance and communication problems identified in the licensee's investigation of the unplanned release.
The records also indicated that the oxide plant operators received training on the instrument panel in the control room to include the use of the UFc and N2 locking devices, the adjusting and recording of stack sampler flow i
rates, and requirements of the maintenance sheet.
The inspector concluded by discussion and observation that the licensee had implemented those engineering and administrative controls which were discussed at the Region III meeting on September 19, 1989.
12
7
~
i' '
- 24
.o; i
Q Attachments l.
A.
Calculated Source Term i
o t'
B.
Atmospheric Dispersion and Meteorology t
C.
Assessment of Off-Site Uranium Concentrations and Doses-V 1
[-
D.
Incident Report, Combustion Engineering, Inc., September 1989
]
E.
Location of Monitoring Sites Around Hematite Facility j
g Revision 1, January 24, 1983 j
F.
Confirmatory Action Letter, Dated September 01, 1989 l
1 L',
t I
t e.
I, t
\\
i
~
i
.I k
i f
o i
P P
l f
u I
t k
1
~
13
ATTACHMENT A e
g APPENDIX A
['
CALCVLATED SOURCE TERM The' estimated amount of uranium released from the stack at the top of the Oxide Building is based on the Licensee's alpha counting analysis of the
[
r fluoride sampling solution. While questions remain about uranium particulate collection efficiency of the stack fluoride sampling system, any uranium in stack gasses passing through the bubbler sampler is considered to be retained I
inthe10percentpotassiumhydroxide(KOH) solution. A 10 ml aliquot of the fluoride sampling solution was evaporated on a planchet and counted for 10 min j
r in the Licensee's Tennelee low background alpha counting sytsem to determine the amount of uranium in the solution, and thus, released out the stack. This i
method provides a lower limit and also the best reasonable estimate on the amount of uranium released.
5 l
The following are the calculations for determining the amount of uranium f
released based on the licensee's 51 dpm net count rate from the sample, i
f Amount of uranium in the sample.
10 ml 0 2 5 eff = 18.6 dpm/ml r
18.6 dpm/ml
= 8.4 x 10'0 mci /ml 2.22 x 10-6 dpm/mC1 l
l 8.4 x 10'0 mci /ml x 2 x 10 ml 8.4 x 10'3 g-U 3
=
2 mci /g-U l
The total uranium released is proportional to the ratio of the stack flow to l
the sample flow:
~
760 ft / min x 28.32 1/ft3 (stack flow) x 8.4 x 10-3 9-U 3
= 274 g-U 0.66 1/ min (sample flow)
The amount of uranium released was approximately 0.3 kg.
P A-1
--n
Y
~
i W
i
'_l'1 -
The' release occurred over two periods on August 28,1989; from 11:50 a.m. to 1:05 p.m. and.1:40 p.m. to 2:20 p.m.
For assessment' purposes the' release was assumed to be uniform over the total 115 minutes.
The uranium v:as enriched to j
4 weight percent of the isotope uranium-235 (U235) and had a Specific Activity of 2 pCi/g. The following table shows the release rates of the l
uranium isotopes used.-to assess offsite impacts.
j 1
i 1
i TABLE A-1 URANIUM ISOTCPIC RELEASE RATES
(.
Percent Release Release Amount Amount Activity Rato Rate Released Released Nuclide
(%)
(Bq/sec)
(pCi/sec)
(pci)
(ma)
U-234 78.0%
2.51E+03 6.78E-02 4.68E+02 U-235 4.2%
1.35E+02 3.65E-03 2.52E+01 i
U-236 0.8%
2.70E+01 7.30E-04 5.04E+00 U-238 17.0%
5.47E+02 1.48E-02 1.02E+02~
8 Total 3.22E+03 8.70E-02 6.00E+02 3.00E+05 j
i
'l i
1
?
I f
A-2
p-ATTACHMENT B l
l APPENDIX B j
ATMOSPHERIC DISPERSION AND METEOROLOGY c
L i
Meteorological conditions in the vicinity of the site on the day of the f
I accident were typical of a hot sumer afternoon with scattered heavy thunderstorms moving through the area later in the afternoon. At the time of I
the release, onsite meteorological me6surements indicated unsteady winds. The j
wind was flowing generally from the southwest (resulting in plume transport to j
the northeast), but shifting up to 60 degrees to either side. Wind speeds were about 5-10 miles per hour (mph) (average about 3 rr/s), with gusts up to 18 mph.
Because of these conditions, atmospheric stability was considered to be unstable (Pasquill type "B" or "C").
i Based on these meteorological conditions, atmospheric dispersion factors
{
3 (X/Q, sec/ra ) were estimawa for Pasquill stability type "B" and "C", then averaged. X/Q values were estimated for several distances downwind in the direction of the nearest resident out to 1000 meters. The equations for relative concentrations at an area boundary for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> immediately following an accident found in U.S. NRC Regulatory Guide 1.145, " Atmospheric Dispersion
{
Models for Potential Accident Consequence Assessment At Nuclear Power Plants,"
were used to calculate X/Q's.
Building wake efftict is considered for l
distances out to 800 meters.
The X/Q value used in dose calculations was the higher value calculated from Equations 1 or 2 shown below.
j Equation 1.
X/Q=
U(fs s + A/2) y Equation 2.
X/Q=
U(3 s s )
yy B-1 l
[
ATTACHMENT B
.o i.
L f
Whr.re 3
X/Q is relative concentration, in sec/m,
f 3.14159, U
is windspeed, 3 m/sec, s#
is lateral plume spread, in m, as a function of atmospheric stability and distance, s#
is vertical plume spread, in m, as a function of atmospheric stability and distance, A
isthysmallestvertical-planecross-sectionalareaofthebuilding, 100 m The downwind atmospheric dispersion factors in the direction of the nearest j
resident (NE) were adjusted for the wind blowing in that direction approximately 36 per cent of the time. The calculated X/0 values and concentrations of uranium in air for that direction are shown in Tables B-1 through B-3.
Because of the wide fluctuations in wind speed and direction over the duration of the accident, a computer code was used to model the release in order to determine potential uranium air concentrations at locations not necessarily directly downwind.
TRI AD: A Puff-irajectory Model For Reactive Gas Dispersion With Application to UF Releases Into the Atmosphere," National 6
Oceanic and Atmospheric Administration, february 1989, was used to correlate a uranium air concentration, based on an air sample, with the release.
The TRIAD code is a numerical model designed to simulate the dispersion of gases that react exothermically with moisture in the atmosphere.
It combines a Gaussian puff model with an objective wind field scheme. The wind speed and direction measurements during the time of the release were divided into five minute increments as input to TRIAD.
The 0.3 kg of uranium were assumed to be l
released uniformly over a two hour time period.
Uranium air concentrations i
B-2
Mf~
ATTACHMENT B y
i
'were calculated for both Pasquill stability classes "B" and "C" at several
]
locations rurrounding the facility. including the location of the Off-site East j
[
air sampling station (approximately 150 meters north of the release point).
e i
-The calculated 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> average uranium air concentrations from TRIAD are shown i
in Tables C-4' and C-5 of Appendix C.
l f
i I
i L
l
[1 i
j l
l t
- t i
i i
I t
I l
l i
. i
)
' i I
i B-3 J
m.
,,.,r.
,m....
m x'
ATTACHMENT B va c_
L.
l i
TABLE B-1 ESTIMATED DISPERSION DURING CE HEMATITE ACCIDENT RELEASE STABILITY CLASS B X/Q X/Q Distance Sigma y Sigma z Equatiog)1 Equatiog)2 (m)
(m)
(m)
(sec/m (sec/m i
1.00E+02 1.60E+01 1.10E+01 1.54E-04 7.23E-05 t
2.00E+02
' 3. 30E401 2.10E+01 5.27E-05 1.84E-05 i
L 3.00E+02 5.00E+01
- 3. 20E+01 2.34E-05 7.96E-06 L
4.00E+02 6.70E+01 4.20E+01 1.34E-05 4.52E-06 5.00E+02 8.20E+01 5.30E+01 8.73E-06 2.93E-06 i
L 8.00E+02 1.25E+02 9.10E+01 3.35E-06 1.12E-06 1.00E+03
- 1. 50E+ 02 _
1.20E+02 2.12E-06 7.07E-07 TABLE B-2 ESTIMATED DISPERSION DURING CE HEMATITE ACCIDENT RELEASE l
STABILITY CLASS C X/Q X/Q i
Distance Sigma y Sigma z Equatiog)1 Equatiog)2 (m)
(m)
(m)
(sec/m (sec/m i
1.00E+02 1.30E+01 7.60E+00 2.92E-04 1.29E-04 2.00E+02 2.50E401 1.50E+01 9.39E-05 3.40E-05 3.00E+02 3.60E+01
- 2. 20E+01 4.64E-05 1.61E-05
{
4.00E+02 4.90E401 2.90E+01 2.63E-05 8.96E-06 5.00E+02 6.00E+01 3.50E+01 1.79E-05 6.06E-06 l
8.00E+02 9.10E401 5.30E+01 7.87E-06 2.64E-06 1.00E+03 1.10E+02 6.50E+01 5.32E-06 1.78E-06 i
r l
TABLE B-3 I
ESTIMATED DISPERSION DURING CE HEMATITE ACCIDENT RELEASE i
STABILITY CLASS B - C i
Average of B and C Distance X/Q 3
(m)
(sec/m )
1.00E+02 2.38E-04 2.00E+02 7.33E-05 3.00E402 3.49E-05 4.00E+02 1.99E-05 i t 5.00E+02 1.33E-05 f
8.00E+02 5.61E-06 1.00E+03 3.72E-06 4
l B-4
ATTACHMENT C 1
APPENDIX C ASSESSMENT OF 0FFSITE URANIUM CONCENTRATIONS AND DOSES o
Offsite concentrations' of uranium in air, in the direction of the nearest resident downwind, were calculated at various distances out to 1000 meters based on the 0.3 kg uranium release calculated in Appendix A and dispersion k
values in Appendix B. Tables B-1 through B-3.
Tables C-1 through C-3 show the L
calculated uranium isotope concentrations in air for Pasquill stability classes t
"B", "C", and an average of the two, "B-C".
The closest site boundary to the north-northeast is approximately 200 meters.
As can be seen from these tables, the concentration of uranium in air beyond the site boundary was not expected to have exceeded the annual average permissible concentration for uranium for unrestricted area, listed in 10 CFR Part 20, Appendix B.
l Based on calculations by the " Triad" computer code, the concentrations of urarium in air at the Off-site East air sampling station (about 150 m N) would be expected to be about 8x10~I3 mci /mi for a 0.3 kg uranium release (Tables C-4 and C-5).
Results f rom analysis of the licensee's particulate air sample, f rom this location during the period of release, indicated a uranium air concentration of about 1.3x10~I4 mci /mi over the sample period 10:12 a.m. on 8/25 through 3:51 p.m. on 8/29.
This value is about 5 to 10 times higher than past routine air samples.
If the radioactivity on the air particulate sample is assumed to have been deposited during the two hour period of release on 8/28, then a corrected uranium air concentration from the accident would te about 5x10-13 mci /ml.
The relatively good correlation between a measured air concentration and a calculated concentration gives additonal support that the amount of uranium released was on the order of 0.3 kg.
C-1
[
ATTACHMENT C i
h Based on the uranium air concentrations for the average of Pasquill stability Class "B" and "C", an intake of uranium was estimated for a hypothetical l
l individual located approximately 400 meter northeast and exposed to the plume-l l
for about 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />. The breatning rate was taken from U.S. NRC Regulatory i
Guide 1.109, " Calculation of Annual Doses to Man from Routine Releases Of Reactor Effluents for The Purpose of Evaluating Compliance With 10 CFR Part
{
50, Appendix I" (1.52x10~2 m/ min). The estimated amount of uranium inhaled j
3 would be about 1.5x10-3 mg(1.5mg).
The minimal amount >:onsidered for I
inducing nephrotoxicity previously reported in NUREG-1189, " Assessment of l
6 Public Health Impact From the Accidental Release of UF at the Sequoyah Fuels Corporation Facility at Gore, Oklahoma," is an absorbed dose of 25 mg U/kg body weight, or a total intake of 1.8 mg.
l f
I Based on the previously calculated air concentration for stability class "B-C" and the breathing rate above, fifty-year committed doses were estimated for this acute exposure (2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> inhallation) for the various distance to the northeast. Dose conversion factors were taken from the Environmental Protection Agency report: " Limiting Values of Radionuclide Intake And Air Concentrations And Dose Conversion Factors For Inhalation, Submersion And Ingestion," Federal Guidance Report No. 11 September 1988.
Because of the potential for a mixture of soluble and insoluble uranium compounds to have been released, and because solubility determines the amount of inhaled uranium which is transferred and retained in a particular body organ, critical organ doses for both soluble and insoluble uranium exposure were calCJlated. Table C-6 list the comnitted dose equivalents to the bone and effective whole body for soluble (Class D) uranium isotopes and to the lung and and effective whole body for insoluble (Class Y) uranium isotopes. The dose to a hypothetical C-2
n.
ATTACHMENT C g
i it 5
individual located 400 meter northeast would be about 0.1 mrem to the bone and i
I i
en effective whole body dose less than about 0.01 mrem if the release were f
[
entirely soluble uranium compounds.
If the release were' entirely insoluble L-uranium compounds then the dose would be about 3 mrem to the lung and an
['.
effective whole body dose of about 0.4 mrem.
a L
I
' \\
l l'
l l'
C-3
~
ATTACHMENT C i..
I TABLE C-1 l
CONCENTRATI')NS FOR STABILITY CLASS B t
Total i',
Distance ~
U-234 U-235 U-236 U-238 Uranium i
(m)
(uC1/ml?
(uC1/ml) fuci/ml)
(uti/ml)
(mg/m3) 1.00E+02 1.25E-1; 6.71E-13 3.34E-13 2.72E-12 7.99E-03 l
2.00E+02 3.57E-12 1.92E-13 3.85E-14 7.79E-13 2.29E-03 3.00E+02 1.59E-12 8.55E-14 1.71E-14 3.46E-13 1.02E-03 4.00E+02 9.10E-13 4.90E-14 9.80E-15 1.G8E-13 5.84E-04 i
5.00E+02 5.92E-13 3.19E-14 6.37E-15 1.29E-13 3.79E-04 8.00E 02 2.27E-13 1.22E-14 2.45E-15 4.95E-14 1.46E-04 1.00E+03 1.44E-13 7.74L-15 1.55E-15 3.13E-14 9.21E-05 MPC 2.00E-11 2.00E-11 2.00E-11 3.00E-12 7.00E-03 l
i TABLE C-2 l
}
CONCENTRATIONS FOR STABILITY CLASS C 4
Total l
Distance U-234 U-235 U-236 U-238 Uranium f
(m)
(uci/ml)
(uCi/ml)
(uC1/ml)
(uC1/ml)
(mg/m3) 1.00E+ 02 1.98E-11 1.07E-12 2.14E-13 4.32E-12 1.27E-02 l
2.00E+02 6.37E-12 3.43E-13 6.86E-14 1.39E-12 4.08E-03 3.00E+02 3.14E-12 1.69E-13 3.39E-14 6.85E-13 2.02E-03 i
4.00E+02 1.78E-12 9.60E-14 1.92E-14 3.89E-13' 1.14E-03 5.00E+02 1.22E-12 6.54E-14 1.31E-14 2.65E-13 7.79E-04 8.00E+02 5.34E-13 2.87E-14 5.75E-15 1.16E-13 3.42E-04 1.00E+03 3.61E-13 1.94E-14 3.88E-15 7.86E-14 2.31E-04 i
MPC 2.00E-Il 2.00E-11 2.00E-11 3.00E-12 7.00E-03
.i i
TABLE C-3 i
L CONCENTRATIONS FOR STABILITY CLASS B-C i
Total Distance U-234 U-235 U-236 U-238 Uranium (m)'
(uCi/ml)
(uC1/ml)
(uC1/ml)
(uCi/ml)
(mg/m3) 1.00E+ 02 1.61E-11 8.70E-13 1.74E-13 3.52E-12 1.04E-02 1
2.0')E+ 02 4.97E-12 2.68E-13 5.35E-14 1.08E-12 3.19E-03 3
3.00E+02 2.37E-12 1.27E-13 2.55E-14 5.16E-13 1.52E-03 4.00E+02 1.35E-12 7.25E-14 1.45E-14 2.94E-13 8.63E-04 5.00E+02 9.04E-13 4.87E-14 9.73E-15 1.97E-13 5.79E-04 8.00E+02 3.80E-13 2.05E-14 4.10E-15 8.29E-14 2.44E-04 1.00E+03 2.52E-13 1.36E-14 2.72E-15 5.50E-14 1.62E-04 L
MPC 2.00E-11 2.00E-11 2.00E-11 3.00E-12 7.00E-03
~
i p
L C-4 l
p ATTACHMENT C i
p~
TABLE C-4 TRIAD OUTPUT FOR STABILITY CLASS B l
i:
'I 2.00 HR AVG. CONCENTRATION AT RECEPT 0RS FOR ALL SIMULATION PERIODS RECEPTORS CONCENTRgTION NO.
X(KM)
Y(KM)
Z (M)
(mg/m) l l-Source.
.500
.500 20.000 1
.550
.500
.000 1.205E-03 2
.500
.550
.000 7.058E-04 3
.600
.500
.000 1.376E-03 L
4
.500
.600
.000 7.078E-04 L
5-
.650
.500
.000 9.663E-04 6
.500
.650
.000 4.334E-04 7
.700
.500
.000 6.572E-04 8.
.500
.700
.000 2.545E-04 9
.800
.500
.000 3.222E-04 10
.500
.800
.000 9.215E-05 11
.535
.535
.000 1.609E-03 l
12
.571
.571
.000 1.871E-03 13
.606
.606
.000 1.306E-03 14
.641
.641
.000 8.864E-04 15
.677
.677
.000 6.121E-04 i
16
.712
.712
.000 4.382E-04 17
.747
.747
.000 3.179E-04 18
.783
.783
.000 2.276E-04 19
.818
.818
.000 1.614E-04 20
.854
.854
.000 1.101E-04 21
.571
.429
.000 3.944E-04 22
.924
.076
.000 7.352E-06 i
t f
l i
i
+
C-5
f.: <.
j.f....i ATTACHMENT C
?
1 I
~l 1
TABLE C-5 TRIAD OUTPUT FOR STABILITY CLASS C i
4g 2.00 HR AVG. CONCENTRATION Kt RECEPTORS FOR ALL SIMULATION PERI 00S j
1 i
RECEPTORS CONCENTRgTION L.
'NO.-
X(KM).
Y.(KM)
Z tM)
(mg/m )
I Source
' 500
.500 20.000
(
1
.550
.500
.000 6.093E-04
{
2~
.500
.550
.000 3.091E-04 l-3
.600
.500
.000 3.159E-03 i
r 4
.500
.600
.000 5.223E-04 5
.650
.500-
.000 9.459E-04
[-
6
.000
.650
.000 3.609E-04 i
I 7.
.700
.500
.000 6.930E-04 l
8
.500-
.700
.000 2.201E-04 9.
.800
.500
.000 3.614E-04 f
10
.500
.800
.000 7.896E-05 11
.535'
.535
.000 8.008E-04 0
12
.571:
.571
.000 1.579E-03 13
.606
.606
.000 1.292E-03 i
14
.641
.641
.000 9.439E-04 15
.677
.677
.000 6.831E-04 f
16
.712
.712
.000 5.011E-04 l
'17
.747
.747
.000 3.724E-04 18-
.783
.783
.000 2.749E-04
}
19
.818
.818
.000 1.997E-04 i
20
.854
.854
.000 1.368E-04 21,
.571
.429
.000 3.695E-04 22
.924
.076
.000 1.044E-05 Il l
'I l
1 I
i C-6
[
s LE...
e ATTACHMENT C TABLE C-6 COMMITTED DOSE EQUIVALENTS i
BONE DOSE (Solubility Class D)
Distance U-234 U-235 U-236 U-238 Total i
-(m)
JSv)
(SV) fSv)
(SV)
(SV)
}
l 1.00E402 1.a4E-05 5.68E-07 1.27E-07 2.23E-06 1.43E-05 L
2.00E+02 3.50E-06
- 1. 75 E-07 3.60E-08 6.85E-07 4.40E-06 i
3.00E402 1.67E-06 8.32E-08 1.71E-08 3.26E-07 2.09E-06 l
4.00E+02 9.49E-07 4.74E-08 9.76E-09 1.86E-07 1.19E-06 5.00E+02 6.37E-07 3.18E-08 6.54E 09 1.25E-07 8.00E-07 i
8.00E+02 2.68E-07 1.34E-08 2.76E-09 5.24E-08 3.37E-07 l
L 1.00E+03 1.76E-07 8.87E-09 1.83E-09 3.48E-08 2.23E-07 f
EFFECTIVEDOSE(SolubilityClassD)
Distr.nce U-234 U-235 U-236 U-238 Total
'(m)
(Sv)
(Sv)
(Sv)
(Sv)
(Sv) 1.00E+02 7.70E-07 3.85E-08 7.88E-09 1.51E-07 9.67E-07 2.00E+02 2.37E-07 1.19E-08 2.43E-09 4.64E-08 2.98E-07 l
l 3.00E+02 1.13E-07 5.64E-09 1.16E-09 2.21E-08 1.42E-07 4.00E+02 6.42E-08 3.21E-09 6.58E-10 1.26E-08 8.06E-08 i
5.00E402 4.31E-08 2.16E-09 4.41E-10 8.43E-09 5.41E-08
- 8. 00E+ 02 1.81E-08 9.07E-10 1.86E-10 3.55E-09 2.28E-08 l
1.00E-03 1.20E-08 6.02E-10 1.23E-10 2.35E-09 1.51E-08 i
l LUNG DOSE (Solubility Class Y)
Distance U-234 U-235 U-236 U-238 Total (m)
(Sv)
(Sv)
(Sv)
(Sv)
(Sv) l 1.00E402
- 3. ale-04 1.55E-05 3.17E-06 6.06E-05 3.90E-04 2.00E+02 9.58E-05 4.78E-06 9.76E-07 1.86E-05 1.20E-04 3.00E+02 4.56E-05 2.27E-06 4.65E-07 8.87E-06 5.72E-05 4.00E+02 2.60E-05 1.29E-06 2.65E-07 5.05E-06 3.26E-05 i
5.00E+02 1.74E-05 8.68E-07 1.77E-07 3.39E-06 2.18E-05 1
8.00E+02 7.33E-06 3.66E-07 7.47E-08 1.43E-06 9.20E-06 i
1.00E+03 4.86E-06 2.42E-07 4.95E-08 9.46E-07 6.10E-06 j
EFFECTIVE DOSE (Solubility Class Y) i Distance U-234 U-235 U-236 U-238 Total (m)
(Sv)
(Sv)
(Sv)
(Sv)
(SV) 1.00E+02 3.74E-05 1.87E-06 3.81E-07 7.28E-06 4.69E-05 2.00E+02 1.15E-05 5.75E-07 1.17E-07 2.24E-06 1.44E-0C j
3.00E+02 5.48E-06 2.74E-07 5.59E-08 1.07E-06 6.88E-06 l
4.00E+02 3.12E-06 1.56E-07 3.18E-08 6.08E-07 3.91E-06 r
5.00E402 2.09E-06 1.04E-07 2.13E-08 4.08E-07 2.63E-06 l
8.00E+02 8.81E-07 4.40E-08 8.98E-09 1.72E-07 1.11E-06 1.00E+03 5.84E-07 2.92E-08 5.96E-09 1.14E-07 7.33E-07 1
C-7
ATTACHMENT D ConsusTION ENGINEERING, INC.
HEMATITE NuCLean Furt MANuFACTunING FACILIrv l
$PECIAL NUCLEAR MArenIAL RELEAst
)
INCIDENT REPORT l
4 September, 1989 l
11 3
[
I 3
l 4
f l
i 7
f i
l t
Ie Anadmen.T 9 SEP 2 81939
, -......,.... -,... _... _. ~.
, _, _. ~.
V U
^
TABLE OF CONTENTS PAGE
.II Chronology 2
III Process Description 3
l Figure 2.
Hematite Oxide Conversion Process 5
p IV Analysis of Events 7
Figure 3.
Conversion Reactor Gas Supply System 9
[
Figure 4.
Incidence of Uranium in the Reactor 11 Off-Gas System l'
N V
Material Balance 13 i
l VI Environmental Impact 20 Figure 5.
Wind Direction and Velocity 21 During Release Figure 6.
Map of Hematite Plant Area 23
[-
Distance to Resident Nearby Plant Figure 7.
Area Survey and Samples 25 Taken August 29, 1989 Figure 8.
Samples Taken by the NRC and 28 Combustion Engineering Figure 9.
General Area Survey Map 29 Figure 10.
Smear and Alpha Reading Locations 32 n
VII Personnel Exposure 34 VIII Root Cause Analysis 35 IX Consequential Items 37 X
Corrective Actions 38 Figure 11.
Revised Health Physics Stack 39 Sampler for Oxide Off-Gas
L I
l
- 11. CHRON0 LOGY August 25. 1989 Friday Midnight Shift The conversion plant is down for the weekend.
Production
]
3 Supervisor il observes an air leak on the solenoid supplying air to the actuator on the R 2 nitrogen valve, closes the j
l air valve ahead of the solenoid to reduce plant air consumption, and notes in the foreman's log, page 81, at the
'1 bottom of the page that the valve needs to be repaired.
i This action causes the loss of air to the nitrogen valve l
actuator and the nitrogen valve opens.
l-Auaust 28. 1989 Monday A.M.
Production supervisors and production superintendent read pages 82 through 87 of the foreman s log, but fail to see d
the note on page 81 and neither open the air valve nor D
repair the solenoid.
i 7:00 Heatup of the conversion line started.
M 11:36 Bed loaded into R 1.
ld 11:40 Steam and ammonia are already on R 2 and R 3.
(Nitrogen is u
actually being fed to R 2 and R3 since the Nitrogen i
pressure exceeds the steam pressure.) Nitrogen supply valve f*
to R 1 is switched to closec position and the steam valve to i
R 1 switched to open position. The flow indicator on FIC 6 i
(steam or Nitrogen to R 1) drops and then slowly rises. The operator assumes this indicates steam condensing in the i
steam lines as it enters R-1.
11:50 Indicated ' steam' flow is normal and UF flow to R-1 is 6
started at 110 lbs/ hour.
12:30 Operators #1 and #2 observe overflow from R 1 to the weigh hopper and collect a sample at 12:30.
Size distribution is t
normal.
l
.{
13:05 The continuous air monitor (CAM) alarms and the UF IS 6
n' shutdown.
13:15 Health Physics posts the upper floors (2, 3, and 4) at 13:15 as respirator areas and collects samples.
The highest i
F samples are in the north end of the top floor of the oxide building near R 1.
The samples are grey to black. The CAM chart shows an increase starting at about 12:00 when startup t
began.
No leaks are found and it is assumed the release occurred during loading of the bed into R-1.
13:40 UF flow is restarted.
Steam flow was not changed during th shutdown interval.
1
. Z*
't Auoust 28. 1989 Monday (continued) 14:20 The CAM alarms and the system is shutdown with HF fumes l
leaking from the dry scrubber exhaust blower noted to be drifting in northeast direction disappearing into the trees.
Hf is also leaking into the third floor of the Oxide Building. Maintenance is called to repair the leaks in the blower housing.
15:00 Production Supervisor #1 returns, has the solenoid air valve
.j to the R 2 nitrogen valve repaired and cycles the steam and nitrogen valves to establish they are operational.
L A heavy thunderstorm drenches the plant.
No leaks are found.
I Samples collected look white but next morning they are b
yellow, indicating the possible presence of UO F '
22 20:40 Conversion was resumed. No problems encountered.
Auoust 29. 1989 Tuesday this week to silo, 105kg, after 6:45 1:30 First unloading of U02
,q hours running time.
3 to the silo: 100kg.
3:30 Second unloading of 002
[,
to the silo: 105kg.
5:30 Third loading of U02 7:30 Fourth unloading of 00 to the silo: 85kg.
2 4
13 Production superintendent requests Health Physics check the spent limestone unloaded on nidnight shift from scrubber fl.
Meanwhile, the second and third scrubbers are unloaded.
9:50 UF, flow is suspended to allow work on the #2 R-1 off-gas lL vaTve which is causing high pressures.
E i
After the limestone unloaded from scrubbers 2 and 3 cools, ll A the rock from the three scrubbers is counted and is hot -
l {
30,000 to 7hg00 dpm compared to normal levels of 150 or less dprr/100cm.
I 14:10 Plant Manager is apprised of the conditions and conversion is shutdown.
15:10 George france, Region 111 inspector, is advised of a suspected re b.se to the environment. Environmental samples l
are taken dov. wind before night f alls.
The solution from the fluoride sampler is analyzed for uranium.
Avaust 30. 1989 Wednesday 9:30 Release confirmed to NRC Region til based on analysis.
I l
.. a
1 m e
!!! PROCESS DESCRIPTION l
i Process Chemistry The Hematite process for the conversion of Uranium Hexafluoride (UF ) to 6
Uranium Dioxide is based on two simple chemical reacti6ns -- the reaction of steam with UF6 gas to produce Uranyl Fluoride (UO F ) and Hydrofluoric Acid
[
22 (HF).
R-1 Reaction UF (g) + 2H 0(g) e UO F (s) + 4HF(g) 6 2
22 L
Anhydrous UO F is white, but when exposed to the moisture in the air, quickly j
22 absorbs water and develops a yellow color. UO I is very soluble in water, 22 The second reaction fundamental to the Hematite conversion process is the reduction of UO F to Uranium Dioxide 00, by hydrogen (H )*
22 2
2 i
R-2 & R-3 00 F (s) + H (g)
- 00 (s) + 2HF(g) l 22 2
2 to Uranium Tetrafluoride A competing reaction can convert a portion of the UO2 Igl (UF orgreensalt).
__ l 4
F Off-gas System 00 (s) + 4HF(g) @ UF (s) + 2H 0(g)
~
2 4
2 This reaction, however, is an equilibrium reaction and can be prevented by feeding steam into the reactor and by operating at elevated temperatures.
to contact the Xj Under normal conditions, there is no opportunity for UF6 b
hydrogen used to reduce the UO F.
However, hydrogen will react with UF6 to 22 produce UF and hydrofluoric acid 4
y UF (9) + "2(g)
UF (s) + 2HF(g) si Potential Reaction 6
4 and this reaction is important in the description of this incident.
The hydrofluoric acid is removed from the reaction process off-gas stream by I
the reaction of the hydrofluoric acid with crushed limestone (CACO ) in 20 3
i foot tall scrubber towers by the reaction:
HF Scrubbing 2HF(g) + CACO (s)
- CaF (s) + H 0(g) 4 C0 (9) 3 2
2 2
under accident The limestone scrubbers also function well as a trap for UF6 conditions producing a bright yellow calcium diuranate (CaV 0 ),
carbon 27 dioxide and calcium fluoride, an insoluble and inert white rock, by the reaction:
j UF Trapping 2VF (g) + 7 CACO (s)
- CaV 0 (s) + 7 C0 (g) + 6CaF (s) 6 6
3 27 2
2 JAR / ear /16107
i a
a O
,t C i
l i
(
Eautoment 1
The Hematite conversion process shown in Figure 2 utilizes fluidized bed reactors for conversion.
UF cylinders heated with steam feed UF6 gas to the conversion reactor, 6
t
'I t
The reactor carries a bed of approximately 100 mesh UO F2 2 particles fluidized l
by steam fed in through a bubble cap plate at the bottom of the bed.
The steam reacts' with the UF in a surface catalyzed reaction which coats the 6
particles with additional UO F '
l 22
)
' yp As the particles coat, the bed grows and overflows to a load cell mounted b
weigh hopper which unloads to R 2 reactor for 1 minute at three minute i
intervals metering the UO F into the R 2 reactor.
~~
22 Seed, small particles of UO F, are pulsed into the. reactor as required to 22 prevent the particle growth from causing loss of fluidization.
,3 i
The fine UO f is filtered out of the reaction gases on porous inconel filters 22 f_#
(R-1 internal filter) and blown back into the bed by timed pulses of nitrogen.
These gases are passed through a second porous metal filter (the R 1 secondary l
p filter) which is not pulsed operating at lower temperature, (g,
The lower temperatures and the absence of a blowback system increases the life l
I h; of these filters and assures, via pressure drop measurements across these i
filters, that a primary filter failure will not go unnoticed.
Iq The R-2 and R-3 reactors employ a mixture of cracked ammonia (hydrogen) and to U0. Approximately steam, fed in as fluidizing gases, to reduce the UO F22 2
l4 90% of the UO F is converted in R-2 and the last of the fluoride is removed 22 in R-3 at higher temperatures and higher hydrogen concentrations.
[
l The automatic dump cycle from the weigh hopper to R 2 is interrupted every two j
j hours and the excess bed in R-3 is unloaded to a U0 cooler and pneumatically 2
transferred to storage silos.
I 1
_. _ _ _ _ _.. _ _ _ _ - _ _ _ _. _. _ _ _. _ _ _ _ _ _ _. _ _ ~ -... -.... _ _.. _ _.... _.. _ -.
-_..____.m_-____...-_.._.
5-
\\
\\*
k E\\0 4
ggg(er DUIU O!!-C B# "'
y
\\
([
\\
~
(lI
,,,,e-..,
~'
t pict
~
\\
\\
h N\\"
\\
Seed ICC0 gg Reactof i
hte
\\
gP6**U e a'***
, c,cw.'
\\
\\
yo2 suo, i
Yd It (Y
C\\
t Y;!
R2 BCSC\\*#
G
(
s g3 ge ctor
\\ I
\\
(
8 V02 CoolCT yspothers i
D R~
- yigure 2 Berna (de Oxide Conversion Process t
-b" 0
i i
When R 3 unloading is complete, R 2 is unloaded to R 3 and the weigh hopper j
unloading cycle is resumed, i
The offgas products from R-2 and R 3 phss through cyclone separators. The R 2 cyclone is unloaded to the weigh hopper and the R 3 cyclone dust is unloaded into transport hoppers.
.)
l The gases then pass from the cyclones to the primary filters above each reactor. These filters are cleaned by pulsing with nitrogen and the dust is l
unloaded froin the housings as required.
The gases from the two reactors are combined and passed through the R-2/R 3 secondary filter which, like the R 1
)
secondary filter, is operated at reduced temperatures and is not equipped with (e.
a blowback system.
After double filtration through porous metal filters, the reaction products i
y pass through steam heated lines to five 20' tall,12" diameter monel scrubbers L
filled with 5/8" limestone chips. These scrubbers serve a dual purpose. They effectively scrub out 90 to 95% of the HF from the off-gas system and trap UF6 released to the off-gas system.
I One of these scrubbers is unloaded and refilled each shift.
Every other day on day shift, two scrubbers are unloaded and refilled.
After cooling, the limestone is scanned for activity, the reading recorded and the limestone is dispositioned.
+
e, I
. t The scrubbed gases are burned and exhausted through a heated 30' stack to the 1
atmosphere.
4
' r t
l JAt/ ear /16106 i
t t
- 1.
IV. ANALYSIS OF EVENTS On the morning of August 29, 1989, the production superintendent asked Health Physics to check the limestone.
The surface contamination of six drums of linestone unloaded from scrubbers
- 1, 2 and 3 ranged from 30,000 to 75,000 dpm (roughly 200 times the normal level and 100 times the level occasionally observed in 3% of the limestone).
1.
I The more contaminated limestone was dark green to black with only an occasional tinge of yellow.
Since the reaction of UF with limestone 6
p produces a strong yellow CaV 0 (the color characteristic of the Uranyl 27 ion), it seemed apparent that no substantial quantity of UF6 had entered
)l the scrubber system.
This resulted in the obvious, though erroneous, conclusion that, in spite of the indication that some UF might have 6
it passed through R-1 unreacted, the contamination more likely resulted from 2
broken filters.
P" The scrubber off-gas sample was removed and checked by alpha count techniques for uranium. This showed 8mg of Uranium which was calculated to be a 274 gram
- release, n3 p"
Environmental samples of soil, vegetation and standing water, including U"
water from a puddle just beyond the fence, were collected downwind and north and east (the primary wind direction) of the plant.
No significant
,y]
radiation was detected in any of the samples.
The R-1 secondary filter was opened overnight to inspect for damage.
Approximately Ikg of green powder was found on the filters, apparently UF.
4 No leaks were detected initially, but one filter showed significantly less resistance to flow and was found by water immersion to have a 3/4" long crack in one of the six tubes. The downstream side of the filter was heavily coated with green powder.
}
p 4
8-T;,.
..e
\\
Scrubber #5 was later unloaded and found to have surface contamination 2
Q 1evels of '300,000 dpm/100cii.
The rock also contained substantial quan-I tities of light brown dust.
q, Cleanup continued with the cleanup water being collected and dispositioned for recovery, i
y A grab sample of the limestone from scrubber #5 was analyzed and found to
~'
. l.
contain substantial quantities of Uranium (2.5%).
This indicated that ai; much as 10kg of Uranium might have been captured in this scrubber alone, tu, Meanwhile, both the R 2/R 3 secondary filters and the R 1 primary filtars were removed and found to be intact.
e.y Ory scrubber #1 - loaded on midnight shift 08/28/89 - was unloaded and found
~
2 iE to be clean, i.e. < 150 dpm/100Tm of contamination.
ld i
f',
There was no Uranium in the R-2/R-3 off-gas lines.
The following conclu-l sions were reached based on this information.
- 1. The Uranium passed through the R-1 filters as a gas, i.e.
as UF '
6 Small amourts of Uranium solids might have penetrated through the R-1 L
secondary filter, but not through the R-1 primary filter, and no Ld Uranium had passed through the R-2/R-3 off-gas line or into the R-2/R-3
)
i-secondary filter.
- 2. The incident occurred prior to reloading scrubber #1 on the midnight f
a shift.
p
- 3. The probable cause was the previously identified problem with the R-2 Nitrogen valve actuator solenoid which flushed Nitrogen througt. the steam header to R-1.
(See Figure 3.)
g I
Subsequent interviews with the operators responsible for startup August 28, 1989, indicated no irregularities in the startup other than setting off the CAM on the fourth floor and a slower than normal overflow from R-1 i
to the weigh hopper of UO f '
22
' j.
o.
9 l
6 To Off-Cas To Off-Cas System System i
l i
Detail of I-Gas Supply I
Control Valves v'.N Rt Reactor m
R2 Reactor D. R. Rohd,
)
9/18/89 1.
/
)
'y n
m
' Q't n
e rev I
Uf5 c,
~.
e h!
[
.Eristing Interlocks tN on UF6 Valve aj,,
l 3
m
'Ur
'31 Base of Reactor $2 y
)
e Q' 30 pg, a
l
/
1r-d.
jg c'
l, h
O Steam now Control Valve l
) *li',
Steam Flow Control Element i
50 N2 Purge Valve. Air Operated (Fall Open)
Steam i
i.-1 DA Shutoff Valve Electrically Operated Solenoid l
ifall Valve to Send Air to Close Closed)
N2 Purge Valve l f' qw.grr 1-Supply N2.furse.
Instru b inst. Air Shutoff Valve m
Ai! Supp;.
Detail of Nitrogen Purge System for Steam Supply Figure 3 Conversion Reactor Gas Supply System
r f'.G L.
,, p p.
c However, on examination of the weigh hopper recorder, it appeared that was transferred to the weigh hopper.
Part, or all of, only 2-3kg of UO F22 this transfer may have resulted from splashing of the fluidized bed rather than from the conversion of UF to UO F '
6 22 In any case, it is clear in retrospect that little, if any, conversion occurred in the R-1 reactor and that this occurred because there was little or no steam fed to R-1.
The question then is, what happened outside the R-1 reactor in the off-gas b
lines and scrubbers?
h The condition which prevented steam from entering R-1 would also prevent steam from entering R-2 and R-3.
However, the flow of cracked ammonia
- ljf, (hydrogen) through a separate flow control valve would not be affected.
Consequently, R-2-and R-3 supplied 12 lbs/ hour of cracked ammonia to the s
mi system - 1.059 lb. moles of hydrogen / hour. Theoretically, this would provide sufficient hydrogen to the off-gas system to reduce 372 lbs/ hour of UF6 4
to UF
--- a 238% excess.
,a It seems probable that the greensalt (UF ) observed in the off-gas lines l
4 f
resulted from the reaction of this hydrogen with UF.
The UF found on the S
4 R-1 secondary filter probably resulted from back diffusion of Hydrogen into
(
the filter as well.
3
'1 The dry scrubbers appear to have had substantially unequal distribution of gas
]
based on the contamination observed in the limestone.
As much as 40% of the
[.g reaction products may have passed through the #5 dry scrubber and 15% through each of the other scrubbers.
The four scrubbers receiving less gas contained discolored limestone and limited amounts of dust.
The rock was generally stained green though spotty with some areas of black. Both stains probably resulted from greensalt.
l l
The rock in scrubber #5 clearly contained more light brown dust with 8 - 9%
!l uranium - probably a mixture of U0, U0, CaV 0, Ca0 and Caco.
This may 2
3 27 3
have resulted from the heat of reaction which caused decomposition of the i
later in the process (after 20:40) limestone and pyrohydrolysis of the UF4 when the steam feed was reinstituted.
. _._.. _. ~... _... _ _
_ _ _... _. _ _ _ ~ _ _ _ _ _.. _. _ _. _ _ _ _. _. _.. _.
l 0 I-
)
i I
?
o Pitot Tube
)
(~10 au Insulated Stack'
'e
" MISSING BLOWER HOUSING CLEAN 007, Bur LESS Stack Sampler THAN - 100 Gu, (113 ou 37Gh i
r Uninsulated SCRUBBER B10Wer Stack Burner jM Orr-GAS LINES Transition f
- 3,016 ou off-sas Burner th..
6 10.
gu
' TOTAL URANIUM h-
- ]-
4 eu REMovEn sY
_r IN Orr-GAS LINES 3.
gu WASHING LIMESTONE
- 26,191'ou lPu
- 11,509 su 4
u RESIouAt ilRANIuM l-AFTER WASHING
- 19,831 su 1
m Dry Scrubbers b, ',.
L "'
(748 Gu)
R1 Secondary l
wj.
N Filter tt R2 & R3 Secondary Filter g
nt w
I l Fil-
'R2 Primary Filter l
R1 Primary p
Filter R2 C c one Filter'
,p (Internal)
R3 Cyclone 1 Reacto R2 Reactor R3 Reactor FIGURE 4. INCIDELLE OF URANIUM IN THE REACTOR Orr-GAS SYSTEM f
S/fe/es*
l,
of o'
Subsequent' leaching of the rock with dilute nitric acid and water to remove the surface contamination removed only 57% of the uranium is the limestone.
This strongly suggests that significant quantities of uranium penetrated the rock, presumably as UF, in addition to that which was filtered out of the gas 6
stream as UF.
4 Additional uranium removed from the scrubber of gas lines 3.016 kgU may have resulted from dust carryover.
The limited quantity removed from the blower intake and unheated sta'ck I
transition (0.69kgU) was yellow and low in assay-58-65% U, suggesting some h!
mixture of UO and/or UO F with limestone dust and/or calcium oxide, 3
22 limestone dust or calcium diuranate. No UF was found.
4 Any residual UF, CaV 0, or UF entering the off-gas burner probably 6
27 4
decomposed in passing through the moisture laden flame of the off-gas burner.
The 30' heated stack wall was clean ekeept for a mall (appro$1mately log) v deposit of yc1?cw on the pitot tube wnich assayed 67-77%U.
This probably contained some U0 F or 00, knd mixed with a dilutent other than calcium.
22 3
- n I
l!r
'b
- !i
,1.
LII Y
L i\\
JAR / ear /16105
-l $ '
s-V. MATERIAL BALANCE Two types of material balances were conducted. The first, to determine the apparent loss during the critical 1:55 hours without steam and the second, around the complete 4.05% campaign.
The mass error around the critical time interval would be expected to be much smaller since it covers only about 1% of the material involved in the larger campaign balance. The campaign balance, however, would be expected to have a b
smaller error as a percent of the materials measured, for the most part, with considerable accuracy, i
b Both material balances are heavily dependent on the assay of the spent lime-stone which is singularly difficult to sample since it is a mixture of dust
- ggi,
'{di and rock of varying sizes and substantially different assays, ranging from 8 -
9% in the dust to 0.5% in the rock. The sampling technique probably results in biased low assays since the samples were of necessity removed from the top of la f
l' the drums-and some of the dust slid to the bottom when -the -drums were righted for sampling. The error involved due to the presence of dust is probably less f
than 3kg however, since some of the dust adheres to the limestone and the total dust content seems to be about 1.5%.
There are also noticeable variations in the uranium concentration within the limestone drums.
m l
'The UF flowmeter was assumed to accurately reflect the flow rate.
6 l fii In any case, the results of the two material balances indicate during the
' b critical 1:55 hour period that the release to the environment was 3kg or less and the " sanity check" balance over the entire 4.05% (21.8kg or <0.3%)
l'q campaign does not invalidate this conclusion.
l' JAR / ear /16111 f
l l
5.
s
t 14 J
MATERIAL BALANCE I 1
FoR THE CRITICAL 1:55 MINUTES i
Inout 64.80 kgU 64.80 kgU Outout l.
R-1 Secondary Filter solids 0.748 ADU from Line Cleanout 14.001 Cleanup Liquor Solids 0.112
)
i,n Miscellaneous Liquors 0.412 Limestone Leach Solution 26.191 Limestone Af ter Leach 19.831 Scrubber Blower Intake Solids
.325 t
Scrubber Stack Cleanou'.
.113 lb
~
61.733kgU CarTzcAL Loss 3.067xcU L
MATERIAL BALANCE II m!!h FOR THE 4.05% CAMPAIGN 0
c.m laat 1
UF 7214.885kgU 6
p.
lead 62.141kaU 7,277.026kgU Outout Conversion System 7193.476kgU Offaas System 61.733 7,255.209kgU
==
j CAMPAIGN Loss 21.817xcu JAR /eer/16111
.m,
-I S -
w MATERIAL BALANCE DATA FOR 4.05% CAMPAIGN i
llNPUT l
7 LOT / RESIDUE NO.-
UF6 lbs UF6 HEEL NET UF6 KG %U KGS U A-728-S 0204 4982 1.0 2259.332 67.611 1527.557 A-729-S 0132 4985 6.0 2258.425 67.611 1526.943 A-730-S 0194 4980 3.0 2257.517 67.614 1526.398 A-731-S 0168 4984 2.0 2259.785 67.627 1528.225 A-732-S 0109 4966 1360.0 1635.646 67.604 1105.762 d.
1-STARTING BED
. NET WT KGS - %U 52294-CF-687-M 18.665 78.32 14.618
{
52295-CF-687-M 22.645 78.44 17.763 52296-CF-687-M 21.145 80.29 16.977 g
SEED m
l 52402-CG-687-M 15.690 81.47 12.783 l TOTAL 7277.026 l m
Q l OUTPUT l
.sg LOT / RESIDUE NO.
%U KGS U B-156-S 2078.215 126.085 1952.130 87.80 1713.970 I
B-157-S 2078.005 122.460 1955.545 87.80 1716.969 B-158-G 2060.630 130.025 1930.605 87.80 1695.071 B-159-S 1833.545 304.135 1529.410 87.80 1342.822 1
l R-2 BUFF l
53248-CA-728-S 20.930 83.08 17.389 j
,h 53249-CA-728-S 15.275 83.08 12.691 d-53255-CA-728-S 11.435 83.08 9.501 l
53256-CA-728-S 19.750 83.08 16.409
' 1 53258-CA-729-S 14.225 83.08 11.819 53269-CA-729-S 13.060 83.08 10.851 53270-CA-729-S 18.355 83.08 15.250 y
53271-CA-729-S 9.575 83.08 7.955 l l 53273-CA-730-S 10.920 83.08 9.073 1
53278-CA-730-S 8.805 81.23 7.153 53280-CA-'730-S 19.540 81.23 15.873 l
53284-C A-731-S 16.920 81.23 13.745 53285-CA-731-S 13.360 81.23 10.853 l
53307-CA-731-S 26.410 81.23 21.454 53317-C A-732-S 13.140 81.23 10.674 53325-CA-732-S 19.505 81.23 15.844 f
, 53o28-CA-732-S 9.945 81.23 8.079 53332-CA-732-S 13.135 77.11 10.128 I
l t
1
,e f
!).
3
.j ho a:
o u
i R-3 CYCLONE FINES NET WT KGS
%U KGS U 53274-CC-729-S '
'13.870 87.23 12.099 53276-CC-730-S 12.690 87.23 11.069 53290-CC-731-S 19,290 87.45 16.869 53365-CC-732-S 1.900 87.45 1.662 i
R-3 BUFF 53286-CA-730-S 14.770 87.230 12.884 ~
- R-3 BUFF C/O
- [
53318-BD-731-S 23.700 87.230 20.674
- L 53364-CA-732-S 5.035-87.230 4.392 R-3 CYC C/O fg 53363-CC-732-S 20.060 87.230 17.493 l
l 'n -
R-2 BED Lh 53366-BD-732-S-1 28.600 85.600 24.482 53366-BD-732-S-2 23.630 85.600 20.227 53366-BD-732-S-3 21.115-
'85.600 18.074 i
t 53366-BD-732-S-4 19.230 85.600 16.461 I
53366-BD-732-S-5 14.580 85.600
.12.480 ClO FROM R-2 BOTTOM 4.010 85.600 3.433
?
5 R-3 BED 53383-CB-732-S-1 22.430 87.703 19.672 53383-CB-732-S-2 17.375 87.703 15.238 d
53383-CB-732-S-3 18.400 87.703 16.137.
53383-CB-732-S-4 20,410-87.703 17.900
'!Q 53383-CB-732-S-5 22.420 87.703 19.663
. lik.
53383-CB-732-S-6 17.735 87.703 15.554 53383-CB-732-S-7 17.150 87.703 15.041 Wi 53383-CB-732-S-8 9.345 87.703 8.196 di C/O FROM R-3 BOTTOM 3.915.
67.703 3.434 3
R-1 BED 53382-CF-732-S-1 22.750 79.310 18.043 53382-CF-732-S-2 24.000 79.310 19.034 53382-CF-732-S-3 24.000 79.310 19.034 y
ClO FROM R-1 BOTTOM 14.400 79.310 11.421 RETURN SAMPLES A-732-S 11.100 77.860 8.642 53385-CF-730-S 19.500 77.860 15.183 R-1 SEED A-732-S 19.700 81.950 16.144 1
t i
p l
.p 7-a
~
1
.j' NET WT KGS
%U KGS U 1AiSCEliNEOUS
. UTILITY HOOD & DS BLOWER C/O 14.120 70.690 9.981 53372-DE-000-S I
53373-BD-000-S SILO C/O BLENDER 1,2,3,4 C/O BLENDER K/O 18.665 87.394 16.312 SILO K/O I
DIVERTER VALVE C/O 53374-DJ-000-S SPLASH TANK CIO 5.160 74.410 3.840 53375-DD-000-S VACUUM SWEEPS PREFILTER K/O '
10.665 72.000 7.679 J.
UTILITY HOOD PREFILTER K/O l
53377-DD-000-S E.B. C/O VAC SWEEPS W.B PREFILTER K/O F
MICRON!ZER FILTER K/O 23.780 67.650 16.087 yll.
OXIDE PLANT VAC SWEEPS 53378-DB-000-S E.B. FILTER K/O 11.850 51.510 6.104 tm 53379-BD-000-S BLENDER HOOD CIO 2.075 85.010 1.764 I.)
UTILITY HOOD CIO OVS (LIQUID) 19.8 GM/KG
'11.900 0.236
. UTILITY HOOD CIO OVS (SOLIDS) 3.490 GAMMA 0.030 EAST BANK FILTER GAMMA 1.188 r-b WEST BANK FILTER GAMMA 5.284
~~ MISCELANEOUS GAMMA COUNT GAMMA 0.758
'4 l
TOTAL PRODUCT CONVERSION SYSTEM 7193.476 l
- r
)l
'b IP tl 1
4
C.
x
.Isw l
1
(
l
}l.
URANIUM RECOVERY FROM DRY SCRUBBER LIMESTONE BY 1
ACID LEACH j
DRUM NO.
SCRUBBER NO.
KG UMESTONE
%U KG U 7
1T 89 1.178 1.048 8
1T 99 0.766 0.758 L
9 1T 63 0.776 0.489 13 1B 96 0.473 0.4 54 14 1B 100 0.605 0.605 15 18 105 0.728 0.764
' lj '.
97%l"** [E @~ % "~"* N N T*3^*S T * ""'7F?? " Y4.118" 16 2T 108 1.076 1.?62 L
17 2T 56 0.920 0.515 18 2T 94 0.479 0.450 19 2B 111 0.351 0.390 20 2B 53 0.726 0.385 y
', 'l 21 2B 73 1.474 1.076 FtypM*?g'"? 'HTT g ~"*P? W Egp?Q*;7?
q.p.: 3 7'; '7T73.978 22 3T 79 0.681 0.538 y
O 23 3T 87 0.986 0.858 10 ST 96.
- -0.552
' O 530 11 3B 82 1.375 1.128 r
q -
12 3B 55 1.368 0.752 T%TMM5? %@Wty:#t&&M B:RyMt%WT QW5&T R W 3.806' l
27 4T 113 0.697 0.788 I'
28 4T 105 1.021 1.072 m
29 4T 29 1.440 0.418 24 4B 90 0.522 0.470 25 48 99 0.712 0.705 f'3 26 4B 66 1.157 0.764 QWMU WWW 777t'4;217.I l
7 W " 7 ?T F G 7 [?'$*Tg'CW p 7
j j 1
58' 35 1.632 0.571 J
2 SB 116 1.445 1.676 3
SB 93 2.232 2.076 l
j1 4
ST 95 2.139 2.032 5
ST 105 1.791 1.881 l
6 ST 99 1.855
,1 B2.h a
zwm;pgry;p qqqyyymp pyyp q mgg gyy cry,~59nyg
.f 2491 KGS 5493 LBS
[fT5535I~
WElGHT BEFORE LEACHING 4.776KG WEIGHT AFTER LEACHING 4.093KG WEIGHT LOSS IN LEACHING 0.693KG PRE-LEACHING ROCK 0.927%U f
IOb. SiOI51 PR5fE CNISG FIOCk ".ibi8'31kG7
IO.
i DRY SCRUBBER /OFF GAS STACK RECOVERY DATA NET WT KGS
%U KGS U r
R-1 BUF CIO 1.010 74.070 0.748 DRY SCRUBBER STACK ClO 0.195 58.160 0.113 DRY SCRUBBER BLOWER INT AKE CIO 0.495 65.580 0.325 CLEAN UP LIQUOR SOLIDS 4.395 2.540 0.112 INPURE ADU FROM R-1 OG LINES 8.165 43.950 3.589 1.740 45.080 0.784 6.950 49.340 3.429 1,
6.520-53.080 3.461 4.795 57.100 2.738 DRY SCRUBBER LIMESTONE LEACH SOLUTION 2491.000 26.191
g DRY SCRUBBER LIMESTONE AFTER LEACH (COMPOSITE) 2139.250 0.927 19.831
)
ADU FILTRATE 900 LTS 0.023 925 LTS 0.009
_ ml
'(jj ADU PRESS / FILTER CLEANUP 15.020 0.156 5.225 0.040 17.050 0.115 y
).
ADU PRESS CLOTHS 1.310 0.027
~~
... 4.420 0.042 og-l DRY SCRUBBER & OFF-GAS CLEANUP PRODUCT 61.733 l L
l l
v I1
.l 1
1
.-=
j.
'10-j VI. ENVIRONMENTAL IMPACT The immediate action taken to assess environmental impact caused by the event was to survey the area most likely to have received deposition of released uranium. Samples for urinalysis from all employees who potentially could have been affected were requested.
Expanded surveys were conducted on a more systematic basis in the days following discovery of the event.
The following f
sections discuss specific areas of investigation.
Meteoroloaical Data
'l4 The Hematite plant has wind direction and velocity indicators on the roof top y
l l il and continuous recording of this information on a strip chart recorder.
An
' evaluation of this infaination shows that the wind direction during the
(
release was predominantly to the east an average velocity of approximately 9 mph.
Figure 5 depicts the wind character during the release pei16ds.
Addi-
~
~'
tionally, a strong rainstorm started near the end of the second release period. An estimated 3/4" of rain fell during the storm, which lasteri about.
two hours.
Prior to the storm, a white discharge from the scrubber emission i
stack was observed by several employees to curl downward and impinge on the trees and ground to the northeast of the plant and about 50' outside the l
p Lh perimeter fence.
'f Neiahborina Residences The closest residence is approximately 300 meters to the west, and an addi-tional residence is about 400 meters to the northeast.
One residence to the f.
southwest is about 600 meters from the plant. An aerial photograph (Figure 6) depicts the plant and surrounding environment and residences.
4
J eZfe:
1 tI
- i 11 a
I gy, @ 9.5 MPH L'
[N
.. 7
^
V.
My
,/
~
"h.a 7,/,,
gy, @ 9,6 MPH
'/
un
,/
( qi!.
/ / '
/ '-
18% @ 7.5 MPH h})
!,/'f[/
/
r-
"E
-- - fa
,;7,\\ ~'
(i r/
x
+
20% j j
\\
1
'N N
-30%
/
"Di N'
18% @ 12.5 MPH
!I-
/
Q?
SW N -40%
s L>. J d
607.
- 97. @ 6.6 MPH T
S r
i FIGURE 5 WIND DIRECTION AND VELOCITY DURING RELEASE
' l.
OZ L' Wird Direction Ard Velocity From the August 28, 1989, Uranium Release i
'~
gigt HEM Direction l
10:50 5.0 NE 11:00 11.5 DE 11:10 9.5 ESE 11:20 5.0 NE 11:30 10.0 ESE 11:40 7.5 NE
.L ~
11:50 13.0 E
First cart of release becins,11:50 l,
12:00 6.5 DIE 12:10 10.0 E
12:20 12.0 E
'N 12:30 9.5 NE 12:40 6.0 E
' f; 12:50 8.5 D4E f;
13:00 9.4 SE First Dart of release ends 13:05 Average 9.3 I
S-13:10 8.0_
DE 13:20 8.0 ESE
~
"'I 13:30 5.5 NE l
13:40 8.0 E
hmnd cart of release becins 13:40 I
1 13:50 6.5 E
14:00 16.0 SE l'3 14:10 9.5 NNE
(
14:20 E
SSW Sgoond nart of release ends.14:20 lf,i Average 9.3 7,,
1' L
i#
14:30 3.5 SSE
[J" l
14:40 7.0 SW 14:50 13.0 NE 15:09 3.5 S
3 15:10 6.5 W
l, 15:20 3.5 NW 15:30 5.0 W3W l',
15:40 7.0 W
l l
Notes: There release was probably in two parts.
The first occurred frun 11:50 to 13:05.
4tw secord occurred from 13:40 to 14:20.
The L
weather data recorded here brackets both releases.
The velocity measurements are in units of miles per hour (MP!Q.
The direction is the direction toward wt11ch the wird is blowirg.
i I,
l,.j
,3
- 23*
m.
s i
I North Air Sample Station
/
Plant Site
/
Boundary j
South West Air
/
l Sample Station
'g 428 Yards
)
I l.
j
-I
/
T..-
4-s318 Yards
/
- ul 9-
/
/
Ran Road
,/
/
/
I v
j
[ h.
Wind Direction j
N I
g NE,,,,
E i P3 N
I 636 Yards
=" a E..
l 4
- i.,
l 3e" -
c#
e, 3,
f'd goeC a*u
,j
/
4 i
L1
/N
\\
s lr 1
1 Figure 6 Yao 0:'
.ema;ite h
?an; Area ll
\\N
' Distance to Residents ll.
s Nearby Hematite Plant-l 1
1.
ll:....... -
y 4 ',
- M*
Nearest Residence Calculation i
The nearest residence in the general wind pattern during the release periods is 428 yards from the plant.
The calculation performed assumed that the maximum quantity of 3kg of insoluble uranium compound was released and that an individual was at the nearest residence was initially and remained outdoors and inhaled it for the total duration of the release.
Assuming that 12% of.
L the intake - was retained in the lungs, and taking credit for known wind direction during the event, this dose is below 0.2 mR.
(Considering virtual lc source - concentration at the stack, plume meandering within sector, wake
""l effect. of ~ wooded area between stack and residence, and particle deposition from the plume, the dose would be further reduced by more that a factor of w
M 10.)
Ty NEAREST RESIDENT DOSE:
i
% Wind i
Building in NE Luna Dose (mR)
Distance (m)
- /0 Wake Effect Grid 3_k.gM 3000U k
400 2.6E-5 1.5 0.36 0.12 0.01 I
1 a.,
REMOTE AIR SAMPLING:
g Concentration (uci/ml) k, NNE Station 1.3 E-14 SSW Station 2.0 E-IS
~
3 Survey Data F
1.
Initial Sampling and Surveys - Tuesday, August 29, 1989 Shortly after discovery of the contaminated limestone scrubber beds, surveys were made and samples were taken in the area where a release would most likely have deposited uranium ccapounds. All sampling and surveys at this time were outside the outer perineter fence, and locations are shown in Figure 7.
Surveys for gamma radiation were made with a Ludlum Model 19 micro R mehr in order to detect areas of radioactivity greater than i
background.
Surveys for alpha contamination were taken with a portable r
PAC-4G survey meter.
These surveys, as can be seen in Figure 7, covered an area in the predominant wind direction which subtended over 220' of arc around the release point.
Standing water, soil and vegetation samples l
were taken from an area which ranged about 130' of arc around the release point and included the predominant wind direction.
ll:,.
o j,(,
l d
h.tl:
AREA SURVE AND SAMPLES TAKEN 8-29-89 l
i J
4
': : s
- )
Plant Site Loundary (HeavyLine)
/
Hissouri-Pacific l
Shaded Area Surveyed
!6 Railroad by Micro-R & PAC.-4G
$, 9 =
h n,'~
o
<*< < a p
g.
- ' @J ?z j-
, a'
~ t, a
v
(
d
/
's f
c,
,, f
\\, #
W l
>yk
- i 1
l
/ /
/
/
f;[
Highway "P"
/. '
j I
q,1
./
t(4 '
j
\\
(
i t.
t Joachim l'
P~
Creek A
s
/.
Hematite l
[i
^
I bi O
300 600 900 1200 r
A FIGURE 7
Scale of Feet I
w r
+
w wg--+e w-m.
m
O 26 j
a i
The results of these early samples and surveys showed no indication of measurable release of uranium from the scrubber stack; results were l
generally indistinguishable from baccground. The specific results are:
]
Sam)le Num)er Samole location Values l
1.
Water pool east of trailer
Background
2.
1911 bank near SW corner of limestone pile 13 pico C1/g 3.
Loll.along fence line due E of 0xide scrubbers 16 pico Ci/g 3a Veoetation along fence line due E of 0xide scrubbers 4 pico Ci/g*
p 4.
Water pool. in roadway to limestone pile
Background
5.
Malgt puddle 50' E of fence, in NE corner
Background
6.
Egil middle of field, N of 0xide Building
Background
6a. Veaetation middle of field, N of Oxide Building 1.2 pico Ci/g*
- 7.. Veaetation leaves from tree 20' from NW corner 10 pico Ci/g*
- wet basis
],4 M
2.
Subsequent Sampling and Surveys j
I Additional samples and surveys were taken on Friday (September 1), Satur-I l
day (September 2), Wednesday (Septece-6) and Thursday (September 7, 1989).
jy U
- a. The samples taken on September 1 and 7,1981, were shared by CE and the l
NRC AIT for later comparison of results. Figure 8 depicts the location of samples taken and are seen to encircle the plant site restricted are as well as mere distant locations (#7 and #8) along Joachim Creek.
A These samples were sent out to Teledyne for evaluation and results are I
shown in Table A.
l b.
On Saturday, September 2,1989, a radiation survey was performed that centered on the scrubber stack from which the release emanated, and Ii went in three concentric circles of radii 100, 200, and 300 yards, respectively.
There were twelve survey points per circle, beginning with the north and going clo:kwise.
Figure 9 shows the General Area Survey Map at the plant location just described. Two instruments were used in the survey, the Ludlum micro R meter (gamma) and a portable PAC-46 alpha survey meter. Background for the PAC-4G meter is between 50 and 150 cpm.
I
... - - - -, _. _ _. ~. -. - - -..., -
27 i
3 TABLE A
1 I
F TELEDYNE PCI/G Dnv BASIS SOIL ALPHA-BETA ST-1 14 1 6
28 1 3
-2 14 i 6 43 1
3
-3 16 1 6 47 1
3
-4 17 i 6 45 i
3 t-
-5 20 1 7 48 1 3
(
-6 20 1 7 40 1
3
-7 9.1 1 5.2 17 1 2
[
-8 5
8.7 1
1.7 VEGETATION PCI/G WET BASIS p
ALPHA BETA
~
SP-1 1.0 1 0.4 9.9 1 0.4
-2 1.5 1 0.6 19.0 1 1.0 l jj
-3 1.4 1 0.4 10.0 1 1.0
-4 4.1 1 1.0 64.0 1
1.0 ld
-5 1.0 1 0.3 15.0 1
1.0
?
-6 1.7 1 0.7 15.0 1
1.0
(
r R2 E PCI/L ALPHA BETA SP-7
<4 6.4 1
1.9 i
<4 6.8 1 2.0 JAR / ear /16115
- o13, X
- )
r 4!
SAMPLES TAKEN BY THE NRC A.ND COMBUSTION ENGINEERING
.(
Samples 1 thru 8 taken 9-1-89 Sampic31b thru 9b taken 9-7-89 All samples taken were soit and vegetation except numbers 7 and 8 which were water and sediment.
Plant Site Boundary (llcavy Line)
'l
,/
/
S 3
Hissouri-Pacific i
g Railroad
?
N
.s n,. ;,,
vg d'^y,,,,
<o
,... -s, g,
v,
- jil, A
s e
u s
~
/',
g>
' ~
{,
j 99'/-
e-1 g
o liighway "P" f8 j
f.-
i "T0 l
i*
9 i
l [
Joachim Crcck i
s l.
,I l
itematite e
j.
l L!
600 900 jeoo FIGURE 8
r-2 m
Scale of Feet
'i i
1
- s 2.9-p GENERAL AREA SURVEY MAP a
I
/
Plant Site Boundary (Heavy Line) 3 11 2
I h
/
, 2.
to '
Hissouri-Pacific Raiiroad hn s
'JD*
/'s r N <t5 a
d""
,1]l
-- ' $ %j
' /^,g M
!jj!
i-
}
c.'
{f,
- .. {,, ' 7 ' <
(
b 8
i---
g.
o,.
s, %.
... * *s.,
v l \\
j h.I..
's
[
s
,/
I h,{,
i Highway "P"
/,
j p
5 t
^)
..g Qj ll '
((
Joachim s
Creek A
r 4
f lL l
/-
Hematite n
/ :l 4
I
// !
O 300 600 900 1200 I
FIGURE 9
sceie or reet
SO~
i i
y Table B presents the readings taken at each survey point. Examination
'(
of the readings shows n6 discernible trends for either gamma'or alpha,
,1 -
regardless'of distance or direction from the stack. All readings are within anticipated background ranges.
.i-c.
In addition to soil and vegetation samples, six smear samples were taken on the plant rooftop. Two were taken on the Oxide Building roof and four were taken on the new Pellet Plant roof (Building 254). The 1.
results are as follows:
Sample Alpha 2
Number Location CPM DPM/100cm l-1.
Stack in center of Oxide roof 25 89
!T 2.
Near door in Oxide roof 109 388 q
3.
NE corner, new Pellet Plant 4
14 4.
Center E edge, new Pellet Plant 5
18 5.
SE corner, new Pellet Plant 4
14
'6.
NE3rner, Warehouse 3
11 m
d.
On September 6, the roadway area just inside the perimeter fence around the UF conversion area was surveyed and smeared for contami-6 38 nation.
Figure 10 shows the locations sar,eled and Table C indicates the results obtained.
All readings are within the normal range m.a lj expected for the area.
3 d
3*
Y
(-
5-JAR / ear /16109 e
--e-w-
,---nn-
.----,-.---<--o---
-e<
- ~, - -
a-
- = - - -, -
- + 1 y
g-- m m p g-m
~-
9
_
l GENERAL AREA SURVEY 1
d MICRO-R PAC-4-G MICRO-R1 PAC-4-G 2 MICRO-R
- PAC-4-G 1
2 LOCATION READINGS ALPHA READINGS LOCATION READINGS ALPHA READINGS LOCATION READINGS ALPHA READINGS 100-1 5
75 200-1 4
50 300-1 6
75 100-2 4
50 200-2 7
100 300-2 7
75 i
100-3 6
50 200-3 6
75 300-3 7
75 100-4 7
100 200-4 9
75 300-4 7
75 100-5 7
100 206-5 7 I 100 300-5 7
100 8
100-6 7
75 200-6 6
75 300-6 6
' 50 he 100-7 N/A N/A 200-7 9
100 300-7 7
100 100-8 4
100 200-8 6
75 300-8 6
75 100-9 7
75 200-9 4
125 300-9 5
50 100-10 5
75 200-10 6
100 300-10 6
75 100-11 7
100 200-11 7
50 300-11 6
75 100-12 6
125 200-12 5
f 75 300-12 5
50 ALL METER READINGS ARE PEAKREADINGS.
(1) Microroentgens per hour.
TABLE E
(2) Counts per minute.
~
-w a
2-
. e-.-.
n., -
m g=...-
=.
=
=
I I
e--
.34.n> 2 4-3.
orgiee m
.jb @
$ik [
I Bldt ill
.ParUhg lot
.,hjui? -
1;in;L :
segrasse Q
g.'
i j
t x -
,1
~
4.. -.=.ms.f=Z: pt=
Urs g
=,..,.. -
tb eer55:fbS$
)h Coa'enfon
" ?
m....
hea 2Mii:;;;;
remet m
sur
_ - - g, p gg g PHAS51f'
{
l l
7
~r c
4 ge=
zi,c..c g e
=ihPelle z ng;-
tii l
~
? !~/.MF g E.g;r==:._._,
-u h'!
- PHASE Ill BuildingJ~
A I
ng; WErfurF#. StoraSe s :::.... ": pi!!!h h
- "=="i N
=
=
.,.. ~.
i r;a;c.r4 i:
. Utilities
' nee =:Mb
(!
'" O '3,
' Ki
.Buildi
= anis 0FN50iils u02 4
=m:m em g7 i
],
n]i ejmMe, c!
s; me:
=
u,.
=
7MIbq ms yggpiy jyjj j
)l.i[
T l
ik-
=
ivarehouse' warehous j'j
-1 l
{ h!
1 " *ii!lli._---- y_,hmu.s ylj l
i se u.= =;;==
u=:n== :=
l a
- f.,,J,
li E Existing Buildings u
c
)
.s?
j O Ex.Eng Roadway i
31
- - Exisag Fence g
l l
New Buildings l
i New Roadway New Fence j ' Temporary Fence N
l x
ema;1.e 4 ac11;y e
4,z.,e, SMEARANDALP,bREADINGSLOCATIONS l
~
33
)
t, 1-
)
CLEAR AREA ROADWAY (FENCED)
CONTAMINATION CHECK SAMPLE -
FIXED READING SMEARS I
NUMBER CPM (NO BKGD SUBTRACTED)
DPM 1
100 11 2
75 0
3 100 4
b 4
100 0
5 175 4
i 6-100 0
7 50 GRASS
- y 8
75 GRASS 9
100 GRASS T
10 50 GRASS i
- lj
~
~~
i' THE FIXED READINGS ARE PEAK READINGS.
' f*R '
SEE MAP FOR SAMPLE LOCATIONS.
{)
l l lj[
TABLE C
l T
l
\\;:
l lb l.
i lf 1
4 i
l
eff.
l.'
(I j
l VII. PERSONNEL EXPOSURE Durina the Incident l
Six employees had potential radiological health involvement in the August 28, 1989, release.
Their radiological exposure was minimal, with the majority of internal depositions being below detectable levels. The radiological exposure for these six people was determined by two different methods.
Inhalation exposure in MPC hours was calculated from fixed air samplers located in the area of the scrubber.
Urine samples were taken several times after the release discovery for radiological bioassay. These results are shown below.
L* \\
Employee MPC-hrs Microarams U/ liter in Urine g.
~
Employee 6
<l.0 Il
'08/29/89 l $-
08/30/89
<1.0
<1.0
~~
08/31/89 Employee B
<l.0 08/29/89
<l.0 j
08/30/89
<1.0 a
08/31/89
<l.0 08/31/89 ll1;j Employee C 08/28/89 3.39 n
4[
08/29/89 1.67
<l.0 l
08/30/89 2.20
<l.0 08/31/89 2.42
<l.0 l
Employee D
<1.0 y
08/31/89
<1.0 i
09/01/89 Employee E 08/29/89 1.31
<1.0 08/30/89 2.91
<1.0 08/31/89 2.35
<1.0 Employee F
<1.0 08/30/89 08/31/89
<l.0
55-o-
j i *:
Emoloyee NPC-brs Microcrams U/ liter in Urine Emoloyee G i
08/28/89 3.19 (1) 08/29/89 21.89 (1) 1.4 (2) 08/30/89
<1.0 08/31/89
<l.0 09/01/89 (1) Administrative action level is 32 MPC-hrs in any week.
(2) Administrative action level is 25 micrograms U/ liter.
Durina the Limestone Unloadina
^
Three employees were involved in the scrubber limestone unloading on August 29, 1989.
Subsequent urine and fecal samples were taken. for radiological
'i bioassay.
Their radiological exposure was minimal, with the majority of S
internal depositions being below detectable levels. The results are set forth below:
L Micrograms U/ -
Micrograms U/
Emolovee MPC-hrs liter in Urine aram in Fecal Emolovee G l
08/28/89 3.19 (1) 1 08/29/89 21.89 (1)
- i 1.4 (2) j 08/30/89
<1.0 08/31/89
<l.0 4
09/01/89 i
bl.
09/06/09 0.47 l
L Employee H l
08/30/89 2.91
<1.0 08/31/89 2.35
<l.0 l
i
l 09/01/89 2.25
<1.0 i
09/06/89 0.10 Emolovee I
-p 1
08/29/89 13.09 (1) 08/30/89 4.85
<1.0
- [
08/31/89 0.58 3.1 (2) 09/01/89 1.74 cl.0 j
0.02 09/07/89 I
(1) Administrative action ievel is 32 MPC-brs in any week.
(2) Administrative action level is 25 micrograms U/ liter.
9 9
~
-..n-
+
m 36
- r 0
Vill. ROOT CAUSE ANALYS15
[
1hree fundamental problems are responsible for the August 18, 1989, release
)
incident.
l
' Lack of recognition of a potential system failure mode, i
' Inadequacies in the system for communicating and documenting the needs for maintenance.
p
' failure in training the opertting staff of the need to assure that
[
conversion of the Uf was actually occurring.
6 a
Q tack of recoonition of a notential <.,vstem f ailure mode.
3 f
Until the August 28 incident, the potential for feeding nitrogen into the 3
process steam header was not identified or analyzed, a
m lhts problem is clearly the most fundamental cause of the release.
l Jnadeouncies in the system for communicatina and documentina the needs for l
Lg maintenance.
l i
1~
None of the supervisors saw the note that maintenance work was required and
!M the note was not available to the operators in the control room.
It is clear that the conversion system would never have been operated with a disabled jj nitrogen valve.
f G
Failure in trainina the operatina staff of the need to assure that conversion I
of the UF was actually occurrina, 6
,f The startup crew recognized that the overflow to the weigh hopper was unusually slow.
They were primarily concerned that the overflow line was plugged and assumed incorrectly that when they collected a sample and saw l
material collecting in the weigh hopper that the system was operational.
While there vere some indicators available that conversion was nel oce,urring, the operators failed to properly interpret that information since they had no experience or tr0ining that would suggest that a total lack of conversion was pos sible.
JAt/est/16112
E I
o87o i
f.
r i
i e
IX. GQMS[QUlNII AL !TEMS Two additional potential improvements were also identified:
j 1
First, the sampling system was not designed to handle particulates.
a Second, the environmental sampling rate should be expanded - both in terms I.
of ability to sample to provide remote air sampling in the normal downwind j
direction (which would have allowed more definitive statements on the l
environmental impact) and in terms of collecting prompt data after the O
fact to more accurately assess potential environmental impact (and help I
reassure the public).
. ':I f
9 I
l i
I.!
f
{il
.j 1
k f
.p t
I l
r-o3Co o
4 X. CORRECTIVE ACTIMS An extensive review of the conversion line system was conducted. As a result of these reviews, interlocks connected to the R 1, R 2 and R 3 nitrogen valve flow automatically positions have been installed which will shut off the UF6 if the nitrogen valve is not in the closed position.
A maintenance requirements log has been prepared for posting in the conversion
[
plant control room.
All maintenance requirements will be posted on this log and the entries will be signed. A decision and the authorized individual who L
made the decision will be included on the log if the maintensnce is not considered critical to operation.
When restarting, the Uf control valve 6
switch will not be unlocked from the closed position untti all critical
'9[]
maintenance requirements have been released.
All conversion cperators are now aware of the necessity of watching closely for any indication that conversion is not occurring and this information will
^
m be included in all oxide training programs in the future to assure that every 0
oxide operator is aware of the potential _ problems.
a,,
A new dual purpose scrubber off gas sampling system is being installed prior to startup and will be tested in place (see Figure 11).
This system will operate on a more dilute stream to minimize condensation and will include both an isokinetic sampler for particulates and a low volume sampler for fluoride.
3d A third remote sampling site will be installed east of the plant to provide emergency sampling capability by March,1990.
4 Additional technicians for emergency environmental sampling will be trained U
from the Quality Control staff to free Health Physics technicians for other emergency work and another Micro R survey meter will be purenased to expedite f
future environmental survey efforts if future emergencies trise.
This tt hing will begin this year and be completed by March,1990.
I
(
I J A R / e e r /1 f,114 I
i
1 I
1 4
1
\\
k
\\
Stac k
\\
i 00 ' 1350
~
s
\\
fW
\\
Tee 9ef"tur*
- '*5Tri controled h
oven Soda h"
Ash
\\
\\
Scrub W i
i W
f yacuum pgTtet
\\
goldef Causttc
\\
g N
sk flow
()
geter (0.9 t.Muid Tra9 g ye',Te 1
g i
FIGURE 11 REVISED HEALTH Pnyszcs STACK SAMPLER FOR OXIDE OFF-gas i
. -, - +. - - -. -.. _.. - - -. _
.. -. -., -, -.,. - - - - -,. - - - -,,,, -,.. - -....., - -. -, - - -, -. ~. -., - - -.
ATTACHMEM E o
}
I e WAfsn a 80eLAhovse PLANT seTa mouNpAny (M4AVY LsNE) l.
l
\\
ts e
MYW8LLI PsNote asiatounwActre l
Ansa AMLROAD t
g' a le 7,a
""8*8 a
v 3
i j
NITORING WELLS 13 i
N
$1TEP0N0;'"n}*
ts Ts o Anst tecNFLusNcs) 1 STATS nouTs 31.A I
/
g
/,
I m eMeu f
entax i
HsMATITE "I#I 7
I s
?
o o
aos eco soo taos J
l I
I i
1 0
SCALE 0? FEET (MEMATITE WELL) p i
Location of lionitoring Sites Around Hematite Facility License No. SNM-33, Docket 70-36 Revision:
1 Date: 1/24/83 Page: 1.5-6
,---...-.-,,,.~..__.--_,,_,....___.,.~,.-,-.,,...,,-_..-,-..--.,__.-an_.-,-
ATTACHMENT F c.; ;
- CONFIRMATORY ACT!] LETTER CAL-RIII-89-020
- siseg,g#
gwgTED STAtts 7
\\
NUCLEAR CECULA7 DRY CONIMisslON
(
)
useow in m.
m v....
.o.= su.va. itu is m
$EP 011999
]
Docket No. 70-36 i
i CombustionEngineering,Inc.(CE)
ATTN: Mr. James A. Rode Plant Manager i
Hematite fuel Manufacturing Post Office Box 107
. Hematite, M0 63047 Gentlemen:
This refers to the conversation between Mr. James A. Rode and Dr. Bruce Mallett cf m) staff on August 30, 1989, concerning the recent inadvertent release of uranium from the conversion process system at the Hematite facility. As i
discussed, the NRC is concerned about causes of the release, damage to equipment, potential exposure to workers onsite and potential for offsite i
deposition. Consequently, we have dispatched an Augmented Inspectio) Team (AIT) on August 30, 1989 to begin review of the event and your actions.
Mr. Donald Sreniawski of my staff is the team leader.
Based upon the conversation with Dr. Mallett, it is our understanding that you have ceased operations invniving the affected conversion process line as i
o of~ August 30, 1989, other than investigating causec and damage to components.
l and will take the following actions, i
1.
Prohibit restart of the portions of the UFs to U0,, process line in the l
oxide plant that were involved with the reTease etent until (a) the root i
cause of the release has baen determined (b) any damaged portions of the i
process line have been repaired, and (c) your conclusions regarding the cause, adequacy of repairs, and basis for restarting the process line have been provided to the AIT Team Leader and the NRC Region 111 Administrator.
i 2.
Maintain records of all activities associated with the release and followup irvestigation by CE for ieview by the NRC.
(
i r
i CONFIRMATORY ACTION LETTER i
~-
[., 4 ATTACHMENT F CONF!RMA10RY ACTION LETTER CAL-Rill-89-020 f, --
Combustion Engineering. Inc.
2 SEP 011999 Issuance of this Cnnfirmatory Action Letter does not preclude the NRC from taking other actions regarding this matter including issuance of an Order requiring implementation of the above connitments.
If your understanding differs from that set forth above, please call me innediately, Sincerely, (b 7 % d OJ a A. Bert Davis Regional Administrator cc: Dr. P. L. McGill, Combustion f.ngineering. Inc.
DCD/DCB(RIDS) l i
h i
1 CONF 1RMATORY ACTION LETTER i
-