[ RadSafe ] TLD questions

Sun Oct 19 14:27:04 CDT 2008

Kai,

I will attempt to answer some of your questions, but some are so specific
that more information is needed before an answer can be given. Hopefully
this will give you some background you can use when discussing these things
with your processor.

First, the TLD cards you purchased have space for four TLD chips, and in the
Thermo Electron product those four positions each have a standard name:
Positions 1, 2, 3 and 4. A "cut" on the otherwise rectangular card at
position 2 ensures the card always goes through the reader, and into a
holder, in the proper orientation. The "xxxx" notation can be one of two
specifications. It may be the set of numbers that indicates the TLD material
in each of the four positions: a "1" indicates TLD-100 (LiF:Mg,Ti with
natural enrichment), a "6" would be TLD-600 (Li enriched in 6Li, so neutron
sensitive), a "7" indicates TLD-700 (Li depleted in 6Li, or mostly 7Li so
neutron insensitive), a "4" I believe is one of the CaF phosphors, etc. The
"0" indicates no chip in that position, so "0110" would indicate no TLDs in
positions 1 and 4, and natural LiF:Mg,Ti in positions 2 and 3. A second set
of four numbers indicates the thickness of the TLD chip in each of the four
positions. A "0" means, again, no chip, and a "1" means there is a
0.015-inch thick (100 mg/cm2) chip, a "4" designates a 0.0036-inch chip, and
a "6" means a 0.006-inch chip. So you are correct that you have 2 TLDs per
card, and in fact those TLDs are in positions 2 and 3 of the card. Which of
the two is intended I cannot tell, but it is probable that both are true,
and the fuller specification would be 0110-0110. In my own program, we use
7776-1141, 7776-1161 and 6776-1111 cards. I am not able to comment on the
meaning of #53196#; perhaps others can shed some light on this.

[Question 1] Thermo Electron cards can be annealed either in the reader
where one card at a time is subjected to a controlled heating profile,
and/or as a batch in an annealing oven where a 100-degree C, 16 hour cycle
is applied (or some variation thereof). The latter eliminates the presence
of lower-energy traps in the phosphor, reducing the corrections required due
to fading. A lab that uses this method must carefully control all aspects of
the heating and cooling of the cards, and it is not a technique for the
fainthearted. While batch annealing is very common for loose chips, when the
chips are permanently encapsulated in Teflon (which melts at about 300
degrees C and is the reason a standard 400-deg C followed by the 16-hour low
temperature anneal is never done), and in the relatively massive aluminum
card, great care must be taken to ensure uniform heating and cooling of the
cards in order not to introduce variations across the batch. Regardless of
the technique(s) used, it sounds like your processor is performing necessary
quality control to ensure that the TLDs do anneal properly, and for some
reason the cards are failing those checks. 

[Question 2] The only reason I can think of for monthly versus quarterly
exchange cycles for LiF:Mg,Ti would be administrative in nature. The service
may automatically bill you or generate missing dosimeter reports for
dosimeters that are not returned within an expected time period, or perhaps
they use a fixed fading correction or background correction that is based on
a monthly wear period.

[Question 3] Roii and Roiii I believe are indeed the data deriving from
fixed regions of interest in the glow curves. Thermo Electron software
generally allows up to four regions of interest within a 200-channel
acquisition of the light collected from each TLD chip during the readout
process. (I'm familiar with the NET-REMS software, but others are available
so I can't say this authoritatively for all of them.) I would guess this
acronym translates Region of Interest i and ii, but exactly which portions
of the readout are being summed is anyone's guess. The processor should have
no problem providing the information to you. I suspect one of those two
regions is the set of channels in which the main glow peaks are expected.
The second could be a region near the end of the heating cycle where only a
small response is expected if the chip heats acceptably, or a region near
the beginning of the heating cycle where non-TL-induced response might be
seen, say, from dirt or chemical contaminants. But, the first and second
regions might both be within the main glow peak regions and some summing
technique used to report dose. Finding out from your processor the meaning
of these two regions is probably key to understanding what is happening with
your dosimeters.

[Question 4] For reporting occupational personnel dose one would always want
to subtract the contribution from the natural background. Background can be
determined by one of two general methods: control dosimeters or a fixed
"mrem per day." With the second, control dosimeters should be used to verify
that the "mrem per day" value is correct, and also to identify when stored
personnel dosimeters might have been irradiated by a source when personnel
were not wearing them. Note that some customers use dosimeters for
environmental monitoring, so no background would be subtracted. The terms of
your contract with the processor should say which one of these options is
used.

[Question 5] No matter how the regions of interest are set up, for a "good"
glow curve I would expect a reasonably constant ratio between the regions.
The fact that yours are not, and the reported dose levels are above what
could be attributed to noise, indicates a problem with the reading.

[Question 6] I would not expect to find any significant temperature-related
dose-response correlations for your TLDs. In a very cold environment the
fading might be a little less than expected, but not noticeably so.

[Question 7] Yes, it is possible to permanently (or essentially permanently)
damage a TLD card. Some of the more common modes of failure are: dirt that
cannot be (or is not) cleaned off; chemical contamination; a relaxing of the
Teflon encapsulation so that the chip flexes during the readout process
(contacting the PMT window which then acts as a heat sink, resulting in
incomplete heating - I wouldn't expect this in newer cards); and cracking of
the chip to the extent that the heat transfer during processing is affected
(not common in the 100 mg/cm2 chips). 

Since you work in the mining industry my first inclination would be to look
at how well-sealed the cards are inside the holders, and whether dirty
holders are cleaned before the cards are removed for processing. 

Best regards,
Kim

Kim McMahan
External Dosimetry Program
Oak Ridge National Laboratory

-----Original Message-----
From: radsafe-bounces at radlab.nl [mailto:radsafe-bounces at radlab.nl] On Behalf
Of Kai Kaletsch
Sent: Saturday, October 18, 2008 1:17 PM
To: radsafe at radlab.nl
Subject: [ RadSafe ] TLD questions

Friends,

So far, I have always just taken the dose value that the TLD readout service
generates and assumed it is correct. Recently, it has become apparent that I
need  some education regarding TLDs.

I am the radiation safety consultant for a uranium mine in a developing
country. The local readout service did not have enough TLDs, so we purchased
300 cards. (I was not involved at that time.) They were TLD-100 LiF:Ti,Mg
Type "0110" #53196# (don't know what that means) from Thermo Electron for a
model 4500 reader. I think there are 2 TLDs per card. The cards are on site
(on a different continent). So, I can't confirm the configuration right now.

The readout service took some time making enough cards available to us. The
reason was that some were not annealing properly.

Question 1: What is the annealing process and is it reasonable that there
should be difficulty with this process? Keep in mind that we don't know in
what shape the equipment used by the service is.

The TLDs were on a monthly changeover schedule. I suggested that we go to a
quarterly changeover, instead. That way, we could keep the TLDs that were
issued for another 2 months while the annealing issue was being sorted out. 
(Expected gamma exposures were very low.) The readout service indicated that
these were monthly TLDs and should not be used on a quarterly basis. We
ended up getting enough TLDs by the end of the month. However, I would like
to eventually go to a quarterly change over for other reasons.

Question 2: What makes a TLD monthly, rather than quarterly?

The results we get have headings of: Name, TLD No. Roii, Roiii, D mSv. I
assume that Roii and Roiii are the raw light outputs of the 2 TLDs in the
card, I.e. "Region of Interest i and ii".

Question 3: Does my interpretation of Roii and Roiii make sense?

The ore grade is low and there has been very little activity near the ore
lately. The control dosimeters are reading around 0.05 mSv and most worker's
dosimeters are below 0.1 mSv. (Control readings are not subtracted from
worker's readings.)

Question 4: Is there consensus on whether control dosimeter readings should
or should not be subtracted from worn readings.

Some readings are considerably higher (up to 1 mSv). It is pretty much
mathematically impossible for anyone to get these exposures in a month. 
Furthermore, the individuals that seem to get the highest readings are the
ones that spend the least time near the ore. For the "well behaved" 
dosimeters, Roii is approximately equal to Roiii and is between 5 and 10. 
For the outliers, there is normally a large difference between Roii and
Roiii and either Roii or Roiii can be larger. For example, one worker had
Roii = 84.51 and Roiii = 14.31 and was assigned 0.98 mSv while another had
Roii = 14.84 and Roiii = 24.15 and was assigned 0.33 mSv.

Question 5: Would one expect these Roii and Roiii readings from properly
functioning dosimeters, or is this an indication that something is wrong.

Some of the people with higher readings spend more time outside than others
and it can get pretty cold on site.

Question 6: Are TLDs cold sensitive?

I only have a subset of the data here and am not sure if the same people
always end up with elevated results or if it is the same TLDs that always
end up with elevated results. My next steps would be to get all the data and
look more closely at these correlations. However, I would appreciate input
from other people before I go on a wild goose chase. (...and, yes, I am also
planning to talk to the TLD manufacturer...)

Question 7: Is it possible to permanently damage a TLD so that it
consistently reads high.

Thank You in advance and Best Regards,
Kai

Kai Kaletsch
Environmental Instruments Canada Inc.
#202 135 Robin Cr.
Saskatoon, Saskatchewan
Canada S7L 6M3
Tel: 406 686 0081

_______________________________________________
You are currently subscribed to the RadSafe mailing list

Before posting a message to RadSafe be sure to have read and understood the
RadSafe rules. These can be found at:
http://radlab.nl/radsafe/radsaferules.html

For information on how to subscribe or unsubscribe and other settings visit:
http://radlab.nl/radsafe/