Register for Updates | Search | Contacts | Site Map | Member Login

ICRP: Free the Annals!

View Comment

Submitted by Harold Peterson, Consultant
   Commenting as an individual
Document Foundation docs Optimisation; Dose to Individual
 
Comments on ICRP Consultation Document, “Assessing Dose of the Representative Individual for the Purpose of Radiation Protection of the Public,” April 5, 2005 Version 18.1

Specific Comments

Page v, 2nd ¶,, last sentence — This is misleading. Retrospective doses are not doses to real individuals since they generally rely on Reference Man characteristics for the intake-to-dose and external concentration to dose characteristics rather than use parameters for a specific individual. Suggest “...while retrospective doses are estimated for reference individuals situated at actual locations.”

Page v, 3rd ¶, The use of the word “compliance” appears to be inappropriate when applied to ICRP recommendations rather than a statute or constraint issued by a regulatory body. Although it is recognized that the ICRP recommendations have a more formal status in the EU and IAEA, this is not universally true. Suggest replacing “compliance” with “conformance.”

Page v, 4th ¶, — Specific guidance on “compliance”including the confidence level appears to be more properly a matter for the regulatory agency than the ICRP.

Page vii, ¶ (S2), last line — As written this could be construed that the Commission advocates exceeding the dose constraint by a small amount. Suggest: “As a result, the Commission believes that the probability that some individual may exceed the dose constraint should be small.”

Page vii, ¶ (S4) — See above comment on Page 5, 2nd ¶,, last sentence.

Page x, ¶ (S17) — The following caveat should be added to the discussion of probabilistic risk assessment:

“Care should be taken when carrying out a probabilistic risk assessment of environmental doses. Often parameter value distributions are either taken from the literature or assumed de novo (e.g., a triangular distribution, a Gaussian distribution, or log-normal distribution). A problem can arise from this process because many environmental transfer parameters are not random variables – they are dictated by local environmental conditions. An example of the this is the bioaccumulation coefficient (CF) for some radionuclides in fish. For example, the CF for cesium in marine fish is around 60 whereas the CF in freshwater fish can exceed 2000. Arbitrarily choosing a value from a compilation of CF’s or based solely on a percentile of a distribution without regard to the nature of the environment can produce dose results that differ by orders of magnitude from the correct value.”

Page 15, ¶ (72) — One questions whether age-specific calculations are worth the added difficulty to perform when the dose constraints are a fraction of a millisievert. Age-specific doses used to be important when individual organ doses were the object of compliance because of the pronounced difference in dose with age for some organs (e.g., thyroid). However, the concept of effective dose generally entails assignment of small organ weighting factors to these organs so that the age-dependence of the dose factor becomes less important when effective doses are calculated. Despite the availability of age-dependent dose factors, recommending such calculations when the doses are far below the dose limits is adding unnecessary detail. For the purpose of environmental dose determinations, a single sexless reference individual should suffice. Also the use of age-dependent models implies that the facility should undertake an age-specific house-by-house census in order to properly apply the age-specific results.

The tables in Annex 2 show the variability of the dose within various age groupings. A similar table comparing the age-dependent numbers to a reference individual adult would probably show that, at most, there’s only a factor or 2.5 difference. This degree of uncertainty doesn’t warrant the added complexity of multiple age group calculations.

Page 17, Section 3.5 — As noted in previous comments above, “compliance” does not seem to be an appropriate term for use with ICRP recommendations.

Page 17, ¶ (83) — There is an apparent difference between U.S. practice and what is being advocated here in terms of when optimisation is applied. The ICRP recommendations appear to favor application of optimisation below some (presumably arbitrary) dose constraint. In U.S. practice, optimization has been used to set the dose constraints (e.g., EPA 40 CFR Part 190 [release limits for uranium fuel cycle facilities] and NRC 10 CFR Part 50, Appendix I [Release limits for power reactors].

Page 18, ¶ (91) — It might be worthwhile to note here that, when performing retrospective dose assessments, whether the doses were in compliance should be determined by comparison with the dose limits (constraints) that were in effect at the time the exposure occurred, and not with more recent limits.

Page 19, ¶ (92) — This paragraph should have an added caveat before the last sentence:
“This location should be one that is, or is likely to be, inhabitable.”

Page A-3 ¶ (A9), last sentence — It might be appropriate to add: “...both at the drinking water treatment plant and at the consumer’s tap.” Many U.S. households employ individual water purification devices.

Page ¶ (B7) Why does the Commission complicate the dose assessment process by defining a critical group of a ”few tens of people?” The purpose of this document appeared to be Assessing Dose of the Representative Individual. Doses to a representative individual can be readily evaluated. Doses to critical groups vary by exposure pathway and radionuclide so the doses to multiple individuals have to be evaluated. The Commission appears to be trying to avoid a single receptor that might have characteristics that result in higher-than-average doses by requiring an average over a few tens of people. If the “critical group” concept must be retained, at least go back to ICRP 7 and drop the number of individuals as a criterion.

Page B-4 ¶(B11) — the caveat suggested in the comment on Page x, ¶ (S17) might bear repeating here.

Page B-8 ¶(B26) — It might be noted that: “The distribution of radionuclide concentrations in environmental media are commonly fitted by a log-normal distribution.”