Comments on the ICRU/ICRP draft report
C. Hranitzky, SSDL, Seibersdorf Labor GmbH, Austria
31st October 2017
The efforts of ICRU/ICRP concerning improvements of the system of operational quantities is very much appreciated. The communication of a new concept and the broad discussion with all radiation protection stakeholders seems to be missing up to now. I hope the possibility of commenting this draft report is a first step but not the last step in that process.
The title of the report may be reconsidered. I would like to propose the title ‘Operational dose quantities for personal and area monitoring of external radiation’.
The system of absorbed dose quantities can be simplified (simplification of the quantities is one of the stated goals of this report). The values of the conversion coefficients for directional and personal absorbed dose quantities are identical. This means that for all radiation fields the value of the directional absorbed dose is identical to the value of the personal absorbed dose. Therefore it seems obvious that the same quantity can be used for personal and area monitoring applications. The only difference is how an instrument is to be calibrated/used, i.e. with or without phantom/person. In conclusion I recommend to use only two absorbed dose quantities (additionally, handy names and symbols may be used), Dlens ‘eye lens dose’ and Dskin ‘local skin dose’ respectively.
Tables of conversion coefficients for photons with kerma-approximation will be used by dosimetry laboratories and shall include more detailed energy steps (similar to the detailed tables from Behrens for eye lens dosimetry) to reduce the risk of errors by the users (the proposed Lagrangian interpolation and logarithmic scaling may not be manageable by all users). It seems quite easy to solve this issue by performing additional simulations. Especially conversion coefficients with sufficient energy steps below 30 keV are important for laboratories using characterized reference radiation fields with measured x-ray spectra and tube-potentials starting from 10 kV.
The symbol h* shall be used for the conversion coefficients to ambient dose (instead of h*Emax). The symbol h*K (similar to the use of the current conversion coefficients) may be used for the conversion coefficients from air kerma to H*. The symbol dskin could be used for the conversion coefficients to the local skin dose. Additionally, the type of phantom related to the conversion coefficients may be given to get a clear description of the applied conversion coefficients, i.e. dskin,slab, dskin,pillar or dskin,rod.
Accredited laboratories have to provide measurement uncertainties for their results. The draft report does not address this important issue (except the statement of statistical uncertainties of the calculations of about 1 %). Additional uncertainties will be necessary for the personal dose quantities Hp and Dlens when using the new conversion coefficients with the ‘simplified’ calibration phantoms. A clear difference in the amount of backscatter between the computational phantoms and the ISO water phantoms can be expected. To solve this issue it will be necessary to simulate and compare backscatter factors depending on the energy and angle of radiation incidence.
The report shall clearly address the fact that there will be a significant change in the calibration factor of radiation protection instruments. At 662 keV (which is the photon energy of Cs-137 for calibrating radiation protection dosemeters) hp and h* conversion coefficients are about 20 % lower compared to the current operational quantities. In addition, new instruments for measuring Hp or H* will have a significantly decreased response at lower photon energies compared to the current instruments. Stakeholder should be aware that when performing long term monitoring statistics a decrease in measured dose or dose rate must not be assigned to radiation protection improvement actions but maybe to the use of a new instrument.
It seems that it is not clear up to now which biologically relevant volume to use for eye lens dosimetry. There are conversion coefficients based on the complete lens and additionally based on the sensitive cells of the lens (or the maximum absorbed dose of both of them). In the final report this medical issue shall be solved and only one set of conversion coefficients shall be included.
Reference dosemeters (e.g. secondary standard ionization chambers) for operational quantities are used by dosimetry laboratories to directly measure dose and dose rates. Dosemeter manufacturers may think about developing new instruments if changes in the quantities can be expected in the future (due to changes in effective dose based on developments in computational reference phantoms or biological knowledge regarding weighting factors). A stable definition of a measurement quantity is considered as absolutely important. As a consequence if a ‘new’ effective dose will be introduced (even if the conversion coefficients are not automatically changed) the dosemeters for Hp and H* are then designed to measure a sort of ‘old’ effective dose. It can be expected that ‘old’ quantities will remain in use for quite a long period of time due to the use of ‘old’ instruments.
Due to the complex geometry of the computational phantoms the simulation of the new conversion coefficients cannot be verified easily. It is important that mistakes and wrong values are excluded from the report. I checked CPE conversion coefficients from table A.5.3b dlens(0°) and found significant differences to the published results (Behrens 2017a table A11): 3 % difference at low photon energies and up to 16 % difference at high photon energies. Other data may be carefully checked as well (especially the unpublished data Daures 2017 and Otto 2017 personal communications). Generally, a difference in most non-CPE data compared to CPE data can be found even in the lower-energy region where no differences are expected. Such statistical differences (up to 4 %, e.g. in table A.4.1.1b and A.5.4.1b at 70 keV and 75°) shall be removed to provide consistent data tables. Finally, all the data tables could be made available as simple data files.