Achievements in workplace neutron dosimetry in the last decade: lessons learned from the EVIDOS project.

The availability of active neutron personal dosemeters has made real time monitoring of neutron doses possible. This has obvious benefits, but is only of any real assistance if the dose assessments made are of sufficient accuracy and reliability. Preliminary assessments of the performance of active neutron dosemeters can be made in calibration facilities, but these can never replicate the conditions under which the dosemeter is used in the workplace. Consequently, it is necessary to assess their performance in the workplace, which requires the field in the workplace to be fully characterised in terms of the energy and direction dependence of the fluence. This paper presents an overview of developments in workplace neutron dosimetry but concentrates on the outcomes of the EVIDOS project, which has made significant advances in the characterisation of workplace fields and the analysis of dosemeter responses in those fields.

Application of workplace correction factors to dosemeter results for the assessment of personal doses at nuclear facilities.

Ratios of H(p)(10) and H*(10) were determined with reference instruments in a number of workplace fields within the nuclear industry and used to derive workplace-specific correction factors. When commercial survey meter results together with these factors were applied to the results of the locally used personal dosemeters their results improved and became within 0.7 and 1.7 of the reference values or better depending on the response of the survey meter. A similar result was obtained when a correction was determined with a prototype reference instrument for H(p)(10) after adjustment of its response. Commercially available survey instruments both for photon and neutron H*(10) measurements agreed with the reference instruments in most cases to within 0.5-1.5. Those conclusions are derived from results reported within the EC supported EVIDOS contract.

Nanodosimetric measurements and calculations in a neutron therapy beam.

A comparison of calculated and measured values of the dose mean lineal energy (y(D)) for the former neutron therapy beam at Louvain-la-Neuve is reported. The measurements were made with wall-less tissue-equivalent proportional counters using the variance-covariance method and simulating spheres with diameters between 10 nm and 15 microm. The calculated y(D)-values were obtained from simulated energy distributions of neutrons and charged particles inside an A-150 phantom and from published y(D)-values for mono-energetic ions. The energy distributions of charged particles up to oxygen were determined with the SHIELD-HIT code using an MCNPX simulated neutron spectrum as an input. The mono-energetic ion y(D)-values in the range 3-100 nm were taken from track-structure simulations in water vapour done with PITS/KURBUC. The large influence on the dose mean lineal energy from the light ion (A > 4) absorbed dose fraction, may explain an observed difference between experiment and calculation. The latter being larger than earlier reported result. Below 50 nm, the experimental values increase while the calculated decrease.

Nanodosimetry in a clinical neutron therapy beam using the variance-covariance method and Monte Carlo simulations.

Nanodosimetric single-event distributions or their mean values may contribute to a better understanding of how radiation induced biological damages are produced. They may also provide means for radiation quality characterization in therapy beams. Experimental nanodosimetry is however technically challenging and Monte Carlo simulations are valuable as a complementary tool for such investigations. The dose-mean lineal energy was determined in a therapeutic p(65)+Be neutron beam and in a (60)Co gamma beam using low-pressure gas detectors and the variance-covariance method. The neutron beam was simulated using the condensed history Monte Carlo codes MCNPX and SHIELD-HIT. The dose-mean lineal energy was calculated using the simulated dose and fluence spectra together with published data from track-structure simulations. A comparison between simulated and measured results revealed some systematic differences and different dependencies on the simulated object size. The results show that both experimental and theoretical approaches are needed for an accurate dosimetry in the nanometer region. In line with previously reported results, the dose-mean lineal energy determined at 10 nm was shown to be related to clinical RBE values in the neutron beam and in a simulated 175 MeV proton beam as well.