Answer to Question #12133 Submitted to "Ask the Experts"

To estimate dose through shielding penetrations, would a deuterium-tritium (D-T) generator resulting in 14 megaelectronvolt (MeV) neutrons be useful if the actual source has D-D interactions resulting in neutrons with average energy of 2.5 MeV?

For many instances of dose estimation from neutrons leaking through wall penetrations (I am assuming that the wall material is concrete), the results from the 14 MeV study could be used, with some adjustments, to provide an estimate of the dose from the lower average-energy 2.5 MeV neutrons. The instances in which I would expect this to be true are those in which (1) the inner aperture of the penetration is not impinged upon by the primary neutrons, i.e., only neutrons that have been scattered within the irradiation room impinge on the aperture; and (2) the penetration through the wall has at least one bend, and preferably two or more bends, within the wall thickness. Under these conditions, the neutrons that emerge from the penetration will be essentially thermalized as a consequence of scatter from the concrete as the neutrons make their way through the penetration.

A modifying consideration relates to the fact that the concrete reflection coefficients of the lower energy 2.5 MeV neutrons are significantly higher than those of the 14 MeV neutrons, roughly by a factor of two. As a result, assuming equal source-emission rates for the 2.5 MeV and 14 MeV neutrons, the fluence rates of scattered neutrons produced in the irradiation room from the 2.5 MeV neutrons will exceed those produced by the 14 MeV neutrons, and this could result in a commensurate increase in the thermal neutron fluence emerging from a given penetration. Considering this, and assuming equal fluences of neutrons from the low- and high-energy sources and equivalent irradiation conditions (same geometries, same scatterers, etc.), I would multiply the 14 MeV results by a factor of two to account for the initially increased scattered fluence in the room (and the associated increased fluence at a given penetration).

As should be evident from the above, we are dealing with some approximations and implicit assumptions in making the comparison between the expected results for the two situations you describe. This means that there is a notable level of uncertainty in the exactness of the projected results. This may be acceptable for your purposes, but if you require more certainty, the problem becomes more complicated, since the types of calculations required to predict neutron dose rates through shield penetrations may involve relatively complex Monte Carlo simulations that require experience in modeling and use of the appropriate computer codes.

George Chabot, PhD