Answer to Question #8933 Submitted to "Ask the Experts"

The radiation protection literature cites 1 Gy as the threshold for adverse health and deterministic effects from an acute radiation exposure. The literature also cites a 1 Sv acute threshold implying a 1:1 correspondence for Gy:Sv independent of biological considerations or quality factors for specific particle type. For mixed neutron/photon spectra, as noted in nuclear criticality accidents, the ratio of Gy:Sv is dependent on the neutron and photon fraction, energy spectra, and flux-dose conversion factors. Are aggregate acute deterministic effects at a specific energy imparted (e.g., 1 Gy) independent of the specific neutron and photon energy spectra and contribution of each particle type? That is, does 1 Gy from a pure photon dose result in the same threshold effect as 1 Gy from a neutron or mixed neutron/photon dose, since energy is constant, all things considered? Wouldn't 1 Gy absorbed dose from a neutron dominant spectra such as a metal accident result in close to 10 Sv, which is unacceptable—and twice the LD_50/60 of 5 Sv? Or is the dose equivalent unit restricted to long-term, stochastic effects? For acute neutron/photon mixed spectrum doses, is the gray unit preferred over the dose equivalent unit of sievert? When is it appropriate to use gray units and when would it be appropriate to use sievert units for personnel protection from an accident? The regulatory limits are expressed in rem (or Sv) units; however, this implies long-term, chronic exposures independent of particle type. The standards and regulatory literature are not consistent with dose units or application for acute dose thresholds.

The Gy and the Sv are not equivalent units; that's why we have both. The Gy refers to absorbed energy, or absorbed dose, but the absorbed dose does not correlate well with biological effects, either deterministic or stochastic, when comparing different types of radiation (for example, photons and neutrons) or even radiation of the same type (for example, neutrons) but different energies. Hence the use of weighting factors that allow absorbed doses to be expressive of biological effects on a common scale, regardless of radiation type or energy. For stochastic effects, the weighting factor is the radiation weighting factor, and gives us the equivalent dose in Sv. 

For deterministic effects, the weighting factor is the RBE (relative biological effectiveness) and gives us a quantity that has no generally agreed-upon name but is often referred to as the rad-equivalent, gray-equivalent, or RadEq, in rad or GyEq in Gy. These equivalent Gy quantities allow addition of Gy doses from different radiations. This was the quantity used in analysis of the recent criticality accident in Japan, and that allowed addition of the doses due to photons and neutrons from the fission reaction to obtain the total effect. The RBE is not a standard number for all effects, but varies with many parameters, including radiation type and energy as well as dose rate and biological end point, among others, and must therefore be chosen on a case-by-case basis depending on the details of the exposure.

The RBE values are generally quite a bit smaller than the radiation weighting factors for a given type of radiation. For example, for neutrons, the radiation weighting factor may be 10 but the RBE may be 3. This is one of the main reasons for not using the Sv for deterministic effects: it is likely to overestimate the probability and severity of a deterministic effect, with potentially unpleasant consequences.

The regulatory limits in the United States are expressed in Sv because these limits are meant mainly for protection against stochastic effects. The limits on organ doses are also expressed in Sv, although they are meant to protect against deterministic effects. Strictly speaking, they should be in Gy, but using Sv throughout is simpler for radiation protection purposes, as opposed to analysis of cases of accidental high exposures, and is also a conservative practice since the doses to organs will be multiplied by conservatively high weighting factors to obtain the equivalent doses in Sv.

Thanks for your question to the ATE Web site!

Sami Sherbini, PhD