> Thanks June for keeping the old dosimetry write-up!! And Sandy of course for > commenting.
> It was great that June sent this back to Powernet because it was 1999 and, > yes, some of these ED issues have been resolved. However, the magnitude and > scope of the issues related to bringing EDs and TLDs together is still as > large. Bringing the individual and man-rem totals together for record and > incremental dosimetry is a difficult task. It can't or shouldn't be answered > with a single factor incorporated onto the EDs. > > In the past few years at the Dosimetry Symposiums and in many conversations > with fellow HPs, I have presented the following issues to study the impacts > to dosimetry at your facility: > > SYSTEMATIC ISSUES > Dose incurred when worker was not in RCA or not on RWP (in PA at BWRs) > TLD processing problems (in conjunction with algorithm issues) > Generic algorithm problems (use of LiBO at doses that are too low) > TLD LLD > ED rounding (“log out if you get to 0.4 mR”); i.e., Facility not > recording the ED tenths > Numerous short RCA entries with 0.1, 0.2, 0.3, etc (an ED at 0.1 or 0.2 > could actually be 0.19 or 0.28) > Improper or no fade correction > Personnel TLDs not stored properly with controls (controls not > representative) > Controls stored in cars (heat = fade) > Incorrect control values used (TLDs annealed at different times, > anomalous control elements) > In-transit exposures with minimal control badges (the controls may not > detect the full magnitude of the transit exposure) > The original ED calibrations (the initial calibration should be checked > at Battelle along with TLD tests) > Overall TLD calibration > Manufacturing issues (dopant irregularities, variable fade, lead > contamination, filter thickness, incorrect phosphor types, phosphor > cross-contamination, TLD filter mix-ups) > Angular response of larger EDs > Energy response characteristic differences (N-16 exposures (e.g., RCP > oil addition), low En photons like Xe-133 exposure) > Internal ED software problems, including computer tracking problems of > ED results > > INDIVIDUAL ISSUES > TLD element readings do not fit the algorithm norms (flyers, gradients) > ED spikes from radio frequency interference, welding, moisture, sweat, > micro-phonics, static charge > ED moved on body in a dose gradient, but the TLD not moved (or opposite) > > ECF shifts in particular TLDs > Physical damage (moisture, dirt, loose phosphor, contamination) > TLD history (previous high doses) > RCA entry tracking errors or record badge worn when multiple TLDs worn > Medical uptakes and the worker wore TLD > Worker not wearing ED or TLD during an entry > Worker not wearing ED and TLD close together (e.g., TLD on lanyard and ED in > pocket) > Worker storing TLD in different location (not with controls – his locker or > desk) > TLD mix-ups at issue > Worker wearing another worker’s TLD on occasion > Intermittent TLD Reader malfunctions > Erratic response of EDs > Dose rate, energy, and angular under-response of EDs, or in-field failures of > EDs > Internal ED software problems, including computer tracking problems of ED > results > EDs not turned on, or EDs storing dose for several entries > Many problems when TLDs are taken home - - fading from storage in hot places, > wearing TLD to doctor, radioactive material at home, family members having > medicals > > This may be the longest Powernet note in a while. But it is a deserving > topic. > > Mike > Michael Lantz > Manager, Technical Services > Mirion Technologies (GDS) 602 677 3020 On Nov 30, 2011, at 2:11 PM, Perle, Sandy wrote: > Thanks June for re-posting, > > One must also recognize that the EADs of the 1999 vintage (and earlier) have > some different characteristics than compared to some of the EADs today. Their > design and technical advancements have "minimized" some of Mike's points. As > you may be ware, Mike has been working with me at Mirion Technologies for > about 2 years now and routinely consults to many of the reactor sites (be > they our client, in-house or other provider). > > Regards, > > Sandy > > ----------------------------------- > Sander C. Perle > President > Mirion Technologies > Dosimetry Services Division > 2652 McGaw Avenue > Irvine, CA 92614 > > +1 (949) 296-2306 (Office) > +1 (949) 296-1130 (Fax) > > Mirion Technologies: http://www.mirion.com/ > > From: June Scott <[email protected]> > Reply-To: "[email protected]" <[email protected]> > Date: Wed, 30 Nov 2011 15:00:55 -0600 > To: powernet <[email protected]> > Subject: Re: Powernet: RE: period TLD-ED dose discrepancy criteria > > <image001.gif> > This list of EAD response is an excerpt from a radsafe discussion from Mike > Lantz (1/29/1999) in response to Sandy Perle. It defines some of the EAD > response characteristics and the effect on CRE and EAD individual response. > It provides testing data for EAD response and the resulting bias that is > being used. > > Also, if you look at the TLD-EAD bias by binning the data in 30 mrem > increments you distinctly see the changes in EAD to TLD response from low > doses to high doses. In a plant setting it is quite often the perceived need > to “match” the doses less than 30 mrem that require the EAD over response to > be set high. This is confirmed also by the data on John’s graph in the lower > left of the page. Tracking of dose >100 mrem in a month and in photon fields > close to the calibration energy correlate closely. > > 1) Exposure to high energy photons from underneath, such as standing on = > a high integrity container of rad waste, TLD =3D 600 mrem, SRD =3D 575 = > mrem and ED =3D 250 mrem because of significant angular dependence. = > This factor of more than 2 increases to a factor of 4 underresponse if = > the transmitter is added to the ED.=20 > > 2) ED's have underresponded by a factor of 3 to noble gas submersion = > (dominated by Xe-133) even though testing shows the ED responds = > accurately to 80 kev photons at perpendicular incidence. Again, a = > serious angular dependence. > > 3) Low energy photons from the side, top or underneath have been = > measured to respond a factor of 14 low (!) in laboratory testing by the = > vendor. In my article a few years ago, I worried if these ED's were = > being used by flouroscopists because of the likelihood of significant = > underresponse.=20 > > 4) The vast majority of ED's in use in the US will significantly = > underrespond to any photons below approximately 60 keV at perpendicular = > incidence, and worsens dramatically as the angle moves from 0 degrees. > > 5) ED's will saturate and underrespond in VHRA's; and the underresponse = > will be relatively unknown. We recently tested a brand new dosimeter = > type that responded well at 90 R/hr, but at 120 R/hr indicated NO = > response, nothing! Its secondary dead time correction in its = > complicated dose rate to dose conversion algorithm, which is known to NO = > users, failed and the dosimeter stopped working until the dose rate was = > reduced to below 100 R/hr! We also tested ED's to NVLAP Proficiency = > Testing and 2 dosimeters read 54% and 65% lower than expected because = > they were irradiated to a high dose rate. > > 6) Drills, motors, and magnetic fields have placed ED's into a latent = > state where they are completely UNRESPONSIVE to radiation, regardless of = > the dose rate. > > 7) The energy response characteristics of individual pin diode detectors = > have been documented to be much more variable than I have ever seen in = > TLD or film testing. Variable responses, certainly more notable as the = > energy of photons is decreased, have been found, possibly related to = > dopants in the pin diode materials; errors up to 60%. > > 8) ED's have been found to change calibration factors by a factor of 2 = > in the field; and then correct themselves. > > 9) The current crop of ED's are fairly new. I find this the hardest = > part to swallow that people think they just work; and will continue to = > work or be supported by the vendor for years, even as the vendors move = > to newer designs. As they have aged, loss and repair rates have been = > significant. Speakers for the alarms fail. And so little has been = > published.=20 > > 10) Computer software and hardware that is not in a mature state yet = > controls the tracking of all doses to be recorded by ED's. Examples of = > data loss (ie, dose) include a recent situation where a terminal to = > verify that the ED was turned on actually turned off random ED's. Other = > software problems continue to lose dose by ED's. > > This is part of the list of problems with EDs that I wrote about in = > 1996. One of our many duties must be to continue to set standards for = > personnel dosimetry that guarantee our facility personnel high quality = > dosimetry that works in all spectra, angles and environments that they = > will encounter. The vendors have continued to improve their products = > and we should be proud that we were part of the push for that = > improvement. > > Mike Lantz, CHP > > [email protected] > > > > >
