Part 21 Report - 1996-020
ACCESSION #: 9511080110
NOTE: This text document was processed from a scanned
version or an electronic submittal and was processed
as received. Some tables, figures, strikeouts,
redlines, and enclosures are not included with this
submittal, and are omitted due to ASCII text
conversion limitations. In order to view this
document in its entirety, you may wish to use the
NUDOCS microfiche in addition to the electronic text.
September 18, 1995
U.S. Nuclear Regulatory Commission, Region III
Inspection and Enforcement Section
ATTN: Ms. B. J. Holt
801 Warrenville Rd.
Lisle, Illinois 60532
RE: Incident Involving J.L. Shepherd Model 109 Irradiator, BML 48-09843-
Dear Ms. Holt:
The UW Madison has 5 closed-beam irradiators under license number 48-
09843-28. Three of these are J.L. Shepherd irradiators, two Mark Is and
one Model 109. The Model 109 located in Rm. K4/358B Clinical Sciences
Center, is relatively unique in that the radiation source is actually
about 12 Cs-137, each nominally about 240 Ci for a total current activity
of 2880 Ci.
Because this incident involved authorized repairs to the door/cover to
the sample chamber, a description of that system may be in order. In the
Model 109, the sample chamber is a cylinder approximately 6 inches in
diameter and about 12 inches long. It is stainless steel with shielding
at both the bottom (or base) and the top. Because there is scatter
radiation, two thick scatter shields in the shape of French doors. There
is a solenoid interlock at the top of the left-hand door which is
activated by an arm from the right-hand door depressing the interlock
when the two doors are together. The interlock prevents irradiation of
the sample chamber while the scatter shield is open.
There is a door/cover (Figure 1) which provides access to the sample
chamber. It is a machined piece of stainless steel with a thumb hole
centered approximately 1 inch from the top which is used to remove the
door. When in place, the door is secured at the top by a spring loaded
screw fastened to the column through a tang on the door and at the bottom
by a small (approximately 1/4 x 1/4 tab welded to the door which is
inserted into a groove in the bottom of the sample chamber. The entire
cylindrical sample chamber is lowered into the source chamber for the
irradiation. Tolerances between the sample chamber and chamber housing
and the walls of the irradiator are less than 1 mm.
Approximately 1 month ago the tab on the bottom of the door broke off
(perhaps the second time in 15 years). The door/cover was removed (as it
must be to insert
University of Wisconsin-Madison 103 North Lake Street
Madison, Wisconsin 53715-1212 608/262-8769 FAX: 608/262-6767
samples into the chamber) and a new tab was welded on to the door.
Unfortunately, this tab was not quite aligned with the slot causing the
door to be slightly out of plumb. When the chamber was lowered, it would
drag and stick, resulting in the sloshing and spilling of some sample
liquids in the sample chamber. To reduce the drag on the system, the it
was decided to file / grind down portions of the door which were sticking
out and rubbing. This maintenance had the concurrence of J.L. Shepherd
and could be performed by taking the door to a machine shop.
Description of the Event (attached letter of 12 Sept)
Tim Sailor, a maintenance worker was contacted to perform the grinding
operation. Because he did not know what was actually wrong and where the
misalignment points were, he desired to be shown the problem in action.
On 11 September, Jim Refsguard, the maintenance manager in Human Oncology
brought the worker into the irradiator room to demonstrate the problem.
Because the scatter shield doors hide the movement of the sample chamber
column, Mr. Refsguard decided to by-pass the interlock using a screw
driver and watch the column descend, noting the areas of contact. The
two persons stood about 1 meter from the front of the unit (Figure 2) to
observe the transit, Mr. Refsguard selected a time of 0.1 minutes on the
timer and pressed the "irradiate" button. The sample chamber column
began descending. Within a second or so the Prime Alert area monitor
(set to alarm at 1 mR/hr) alarmed. Knowing that the alarm indicated
radiation exposure, Mr. Refsguard immediately pressed the "load" button
and the sample chamber returned. The door was then removed, the high
points ground down, and the door replaced.
Although the Jim Refsguard had been responsible for proper functioning
and certain routine maintenance of this system since about 1981 and was
familiar with its operation, there is no documentation of his receiving
formal training on the system. As Mr. Refsguard noted in his report,
this lack of formal radiation safety training resulted in his assuming
that the radiation would shoot up the column and not scatter.
Additionally, neither he nor Tim Sailor wore any dosimeters when in the
Thus, none of the workers involved had dosimetry nor documented training.
Since September, 1993, it is Safety's policy that all new closed-beam
irradiator workers first receive the 4-hour Radiation Safety for
Radiation Workers training (offered weekly) followed by an irradiator
training block which has both a 2-hour academic portion (usually offered
monthly) and a hands-on, irradiator specific (e.g., Mark I or Model 109)
portion. These workers then apply for dosimetry to monitor radiation
exposure. As noted in Item 8 of the 48-09843-28 license renewal packet
(control No. 398620) Safety is in the process of extending this program
to include the Model 109. Additionally, whenever maintenance workers
enter an irradiator room, our policy is
Model 109 Incident - 2
to provide that worker with training commensurate with the potential
hazard, but as a minimum one that covers the radiation hazard and any
alarms present, and to provide oversight while the worker is in the room.
On 12 September, we attempted to obtain some relatively firm exposure
numbers. We taped 3 TLD chips on the inside of the left scatter shield
approximately 6.5, 15, and 18 cm from the bottom of the scatter shield.
We then ran the system through the scenario outlined by Mr. Refsguard
(i.e., timer set for 0.1 min, irradiate button pressed, sample column
went to irradiate position, load button was pressed -- a total TLD
exposure time of approximately 9.8 seconds, 0.024 minutes exposure and
about 0.07 minutes chamber travel). These chips were then sent to our
dosimeter vendor for reading. The exposures to the chips next to the
sample column were 350, 45, and 27 milliroentgen, respectively.
Because the sources are stationary, we determined that the persons could
not have been exposed to the primary beam, however, it is possible that
the bottom chip (at 6.5 cm) may have been exposed partly to the primary
beam. Assuming a worst case scenario, that the exposure to this chip was
due entirely to scatter (assuming part exposure to the primary beam would
reduce the exposure to the persons involved), we calculated (attached
letter from Bruce Thomadsen dated 18 Sept) a maximum exposure at the 1
meter distance where both workers were standing to be approximately 0.4
Although calculated exposures are trivial, we believe that the bypassing
of a safety interlock by persons not completely aware of the potential
radiation hazards involved is a significant safety issue. Therefore, our
corrections are addressed to insuring such safety incidents do not occur
with this system, nor with the other closed-beam irradiator systems we
have on campus. Specifically:
1. Safety is currently integrating the training for all irradiator
workers into a single program which will enable us to implement a
uniform program with appropriate personnel controls for all three
research irradiators. The first of these classes is scheduled for
20 September. Mr. Refsguard indicated that he had always intended
taking these classes and would make attendance a priority. He did
attend the 4-hour training block on 13 Sept.
2. Mr. Refsguard was informed that all maintenance requests would need
to be staffed through the Safety Department (much as is done for the
other irradiators on campus). Safety will coordinate any necessary
training and on-site supervision. We have been communicating with
J.L. Shepherd and have tentatively arranged to have J.L. Shepherd
train and certify several of our workers to perform certain (non-
source handling) maintenance functions on these irradiators.
Model 109 Incident - 3
3. All irradiator owners on campus were appraised of the incident and
the need to channel maintenance functions through the Safety
Department to insure proper training and dosimetry are provided.
If you have any questions pertaining to this information, please call me
at (608) 262-9178 or FAX me at (608) 262-6767.
Ronald R. Bresell
Radiation Safety Officer
NRC Operations Center
Model 109 Incident - 4
UNIVERSITY OF WISCONSIN
TO: Ron Bresell
From: Jim Refsguard
UW Cancer Center
RE: K4/358 Alarm
September 12, 1995
Per our conversation, this is my report on the incident in CSC
K4/358 on September 11th.
I had received word that the drawer of the Cesium Irradiator was
again running roughly and needed to be checked. We have had our best
results working with the machine shop at the Physical Sciences Lab, and
Tim Sailor arrived to work on the unit.
We discovered that both the outer surface of the door to the sample
chamber and the inner surface of the lead outer doors were scratched and
grooved, apparently because the sample chamber door was out-of-round and
binding as the drawer was lowered.
In an attempt to determine what portion of the sample chamber door
was catching we manually closed the interlock switch for the lead doors,
leaving them open while the drawer descended. I assumed since the source
was buried in the stack the only escape of radiation would be directly up
and did not anticipate the scattering effect.
When the drawer was about half lowered the alarm went off for about
1 second then silenced as the drawer finished its descent. As soon as I
heard the alarm I pressed the load button to bring the drawer back up.
As the drawer returned the alarm again was heard for about 1 second.
When the drawer reach the load position we immediately closed the outer
doors. I understand now that the alarm occured when the hollow sample
section of the drawer was situated above the source and did not offer any
shielding. I now understand that some of this radiation scattered off
the insides of the lead stack until it exited the stack and scattered
about the room, thus activating the alarm.
Both of us were about 3 feet away from the unit positioned to look
horizontally at the level of the sample chamber cover as it descended.
I immediately reported the alarm to Dr. Thomadsen for his
evaluation. If you have any questions or need clarification, please
CSC K4/566 263-5365
K4/566, 600 Highland Avenue o Madison, WI 53792-0001 o
(608)263-5363 o FAX (608)263-6062
UNIVERSITY OF WISCONSIN MEDICAL SCHOOL
Date: 18 September 1995
To: Ron Bresell
Radiation Safety Officer
From: Bruce Thomadsen
re: Exposure from the cesium Hot Box
Based on the information gathered from Jim Refsguard and from
measurements on the cesium unit, we have calculated the exposure that he
and the worker from PSL might have received. As can be seen on the
scaled figure, neither person could have been in the primary beam.
Considering scattered radiation, the hand-written sheet contains some
detailed calculations based on the TLD results. Note that the TLD
measurements should be divided by 1.87 as the backscatter factor for Lif
in contact with lead. From that analysis, the dose to the workers is
likely to be approximately 0.00124 mrem.
Ignoring the likely assumptions, and instead assuming that all of the
dose to the TLD chip was due to scatter from the top lead plug of the
drawer, the dose to the closest worker would be
D= 187 mR (7.65 cm/165.5 cm)**2 = 0.4 mR.
These two values probably bracket the true dose, with lower value being
Department of Human Oncology
K4/B100 Clinical Science Center 600 Highland Avenue
Madison, WI 53792 608/263-8500 FAX 608/263-9167
Figure "Landauer Glow Curve Report Process #U8142A Card #0644955"
Handwritten page omitted.
*** END OF DOCUMENT ***
Page Last Reviewed/Updated Friday, January 31, 2020