Your ICRP "gold standard" risk models are wrong

Thanks for bringing that report to my attention -- they bring up some good points about dose calculations and suggest some possible solutions. Without getting into it too deeply, here are just a couple of notes I have on the topic of uncertainty in dose conversions: (1) The CERRIE report says that the ICRP dose conversion factors for I-131 and Cs-137 (in particular) are not too uncertain. From pages 27-28 of that report:

Overall, uncertainties in dose coefficients can be regarded as small for some radionuclides (e.g., 137Cs and 131I) but substantially larger for other radionuclides, particularly when considering intakes by children or doses to the fetus (eg 210Pb, 210Po, 239Pu and 241Am).

And a related footnote on page 28:

A recent detailed analysis of uncertainties in dose coefficients from ingestion of 131I, 137Cs and 90Sr resulted in 95% confidence intervals for organ dose coefficients for adults (males and females) of about 3–6 for 137Cs (all tissues), 7–8 for thyroid dose from 131I, and 20–40 for bone surface and red bone marrow doses from 90Sr (Apostoaei and Miller, 2004). In a large exercise on uncertainties (Goossens et al, 1997), involving aggregation of uncertainty ranges obtained by a panel of experts, combined 90% confidence intervals quoted by Harrison et al (1998) for adult ingestion and inhalation dose coefficients varied from factors of less than ten for 137Cs and 131I thyroid dose, to factors of several thousand for 90Sr lung dose after inhalation and 239Pu bone marrow dose after ingestion. These larger ranges were partly attributable to uncertainties over the chemical forms that might be ingested or inhaled (Harrison et al, 1998). The estimated uncertainty ranges differed substantially between the study experts

(2) Even if the dose conversion factors are uncertain by a factor of 10, and if we happen to be using a number that is 10 times too low -- or even 1000 times too low -- the doses would still be very, very tiny. For example, taking as a worst case our single highest measurement of I-131 (20 Bq/L in rainwater on 3/23), we calculated that one would need to drink about 140 liters in order to have a dose equivalent to a round-trip cross-country plane flight. If our dose conversion factor is 10 times too low, one would need to drink 14 liters of the water for the same dose. If our dose conversion factor is 100 times too low, it is 1.4 liters. If 1000 times too low, then 0.14 liters. But keep in mind that the airplane flight that we are comparing to is already a very small dose with almost zero health effects, so the overall magnitude of the risk would still be very small. Let me know if this makes sense to you. Mark [BRAWM Team Member]