Dear Sir Thank you for the opportunity to comment on the ICRP consultation document ¡¥Assessing dose of the representative individual for the purpose of radiation protection of the public¡¦. Nirex is the government-owned body in the UK responsible for the long-term management of radioactive waste. Our mission is: "In support of Government policy, develop and advise on safe, environmentally sound and publicly acceptable options for the long-term management of radioactive materials in the UK." Our current role is to: „h carry out scientific, engineering and social science research to help develop safe and environmentally sound options for dealing with radioactive waste in the long term; „h set specifications and standards based on a phased deep geological disposal concept and advise the industry on how to treat and package radioactive waste through the application of the Nirex Letter of Compliance Process; „h maintain an inventory of radioactive waste in the UK in conjunction with the Department for Environment, Food and Rural Affairs (Defra); „h communicate with all stakeholders, including the public, to build understanding and develop ways of addressing the wide range of concerns and views surrounding the management of radioactive waste, so that these influence our work. This work will be carried out adhering to Government policy at all times, ensuring no conflict or overlap with the work of Committee on Radioactive Waste Management (CoRWM) or the Nuclear Decommissioning Authority (NDA). We are broadly supportive of the technical guidance being proposed in this document. However, the current structure of the document makes it difficult to follow in places. There are also some areas where the wording of the guidance is confused or ambiguous and where it is not clear how it would be implemented in practice. The following provides some suggestions on how the document structure could be improved with respect to the clarity of the guidance. It would be helpful if the document made a clear and early statement that, to reflect the greater emphasis now being placed on source-related doses, rather than defining local habits, the ¡¥critical group¡¦ concept (defined to be homogeneous with respect to the habits and dose-determining characteristics of its members) is being replaced with the concept of an ¡¥exposed group¡¦ (defined to be homogeneous with respect to its exposure to radiation). Doses to the exposed group are assessed by consideration of a ¡¥representative individual¡¦ from that group. The ¡¥representative individual¡¦ is defined such that the doses it receives are based on the average characteristics of the highest exposed group. By then demonstrating that the dose constraint for this representative individual is met, the goal of protecting the general public can be considered to have been achieved, avoiding the need to measure or assess doses to the public directly. The document discusses the definition of the ¡¥representative individual¡¦ for both deterministic and probabilistic dose assessments. However, these discussions are often inter-mingled, the document would benefit from a structure that sets out the recommendations in the two cases more clearly. For example, deterministic assessments could be discussed first, before moving onto the more complex case of probabilistic dose assessment, to which the majority of the discussion relates. For prospective doses the ICRP now recommends using only three age categories (infant: 0 to <6 years; child: 6 to <16 years; and adult 16 to 70 years) with the use of habit data and dose coefficients for a 1-year-old infant, 10-year-old child and adult representing each of these categories. We believe this is a sensible recommendation, especially when the extended discussion on consumption of breast-milk by infants and foetal doses (see paragraph 79) is taken into consideration. The key recommendation of the document is that for prospective probabilistic dose assessments, ¡§the Commission recommends that the representative individual be identified such that the probability is less than about 5% that a person drawn at random from the hypothetical population will receive an annual dose exceeding the dose constraint. This hypothetical individual should be representative of, at most, a few tens of people who are the most highly exposed.¡¨ The document needs to explain more clearly, in one place, how such hypothetical populations and hence appropriate representative individuals are derived. The document would benefit from a discussion of how to identify the exposure pathways that may give rise to dose. This is an important aspect of any dose assessment, and an essential stage in identifying and assessing doses to a ¡¥representative individual¡¦ (as noted in paragraph 55). However, the topic receives relatively little mention and does not even appear in Figure 1, a schematic summary of the assessment of annual dose:¡V identification of exposure pathways or scenarios is a necessary precursor to the third stage of assessment shown in this Figure (estimate dose rates). The need for habit data to be reasonable, sustainable and homogeneous is stressed. To avoid the need for specific habit data for local populations (which may not be available), the document states that values may be derived from national or regional population data. It is suggested that 95th percentile data for behaviour can be considered to represent a cautious, but acceptable assumption, for defining a reasonable and sustainable intake rate in a deterministic approach, but there is a need to avoid selecting extreme percentiles for every variable to avoid excessive conservatism. This seems sensible advice as to what constitutes ¡§reasonable, sustainable and homogeneous¡¨ data. In a probabilistic assessment, it is stated that, ¡§care must be used to include all hypothetical individuals whose dose could possibly be representative of persons receiving the highest dose, including extremes.¡¨ In making this statement, it would be helpful if the document could clarify that these hypothetical individuals form part of the pool of which the representative individual represents the 95th percentile, i.e. the extreme hypothetical individuals are not themselves the representative individual. However, there does appear to be some confusion on this point, as at paragraph 53 it is stated that the ¡§representative individual is the hypothetical individual receiving a dose that is representative of the most highly exposed individuals in the population.¡¨ As noted above, a discussion on exposure pathways / scenarios may help to clarify the intent, which we understand to be that the ¡¥representative individual¡¦ should be exposed to the most extreme exposure pathways, but should not be assigned extreme habits. However, since exposure pathways and habits are intrinsically linked, clarity on what is intended, perhaps with some suitable examples, would be most helpful. It is stated (at paragraph 96) that ¡§it is necessary to characterise the probability that the individual is exposed¡¨, necessitating the need for a value for probability of exposure to be included in the calculations. Probabilities are notoriously difficult to justify and Nirex is concerned by the suggestion made in paragraph 96 that a probability distribution could be used for a probability value ¡V assigning probabilities to probabilities is not generally regarded as good practice and it is unclear what would be the statistical significance of such a distribution. The final area of discussion, is the recommendation to involve stakeholders in the characterisation of the ¡¥representative individual¡¦ for dose assessments. Nirex welcomes this, both in terms of the local knowledge that stakeholders can supply and as a means of engaging them in the assessment process and helping to ensure they understand and accept the process. Detailed points: „h In the Executive Summary (paragraph S6), the discussion on uncertainty in dose assessment also needs to acknowledge that there is also uncertainty associated with what situations, or scenarios, may give rise to a dose (particularly when considering future doses to hypothetical individuals). „h In Section 2.4 (paragraph 46) it states, ¡§Uncertainties associated with estimation of dose may be taken into account either deterministically by using single values for parameters chosen to take account of uncertainty, or probabilistically by incorporating distributions for parameter values. With either methodology, the goal should be to perform a realistic evaluation of dose.¡¨ If single values are chosen to ¡¥take account of uncertainty¡¦, it presumably means the values chosen are cautious, or conservative, in which case it is not clear how the resulting dose evaluation will be ¡¥realistic¡¦. On the other hand, if best estimate, or central, values are used to give a realistic assessment, it is not clear how uncertainty is taken into account, as there is no mention, for example, of the role of sensitivity calculations in deterministic assessments. Later text (paragraph 49) discusses the role of conservative screening to demonstrate compliance with a dose constraint, but this would clearly not provide a ¡¥realistic evaluation of dose¡¦. We would welcome the opportunity to discuss these issues with you in more detail. If you require clarification about any of the points raised please contact my office. We would be grateful if you could send us a copy of the outcome of the consultation and add us to any list of interested consultees that you hold. Yours faithfully Chris Murray Managing Director