RESPONSE OF THE HEALTH PROTECTION AGENCY RADIATION PROTECTION DIVISION TO THE ICRP DRAFT DOCUMENT FOR CONSULTATION ON THE OPTIMISATION OF RADIOLOGICAL PROTECTION (42/105/05) The International Commission on Radiological Protection (ICRP) published a series of documents on its web site for consultation in April 2005. The document considered here is entitled ‘The optimisation of radiological protection’. Staff of the UK Health Protection Agency Radiation Protection Division (HPA-RPD; formerly the National Radiological Protection Board) have considered this document and their comments are given below. HPA-RPD welcomes the opportunity to comment on the draft document and feels that it covers an important part of the system of radiation protection. HPA-RPD staff are broadly supportive of the discussion in this document, although it would benefit from containing more in the way of practical guidance. A number of minor areas for clarification or revision are given below. In developing this response views were sought from all HPA-RPD staff, particularly those who were not directly involved in the work of ICRP. Both general and detailed technical comments are given but not mainly editorial comments as it assumed that they will be picked up in the process of producing the final ICRP publication. General comments • Overall, it is considered that the document is a useful addition to the ICRP set of publications. Although quite a lot of the content has been addressed previously in a range of ICRP publications, it is useful to draw this together. However, the discussion lacks focus in places and is so general that it is hard to draw conclusions from it. It would benefit from more practical guidance, in particular real examples and case studies. • The Executive Summary states that this report addresses optimisation for all situations except medical exposure of patients, which is dealt with separately by the Commission. It seems inappropriate not to include advice on the optimisation of such a major source of population exposure in this Foundation Document (after all, medical exposures outweigh occupational exposures by at least a factor of 100 in most developed countries). It is to be hoped that the forthcoming Foundation Document from Committee 3 on ‘Protection of the Patient in Medical Procedures’ will give this subject the attention it deserves and is available in time to influence the final version of ICRP’s new recommendations. • The document suggests ways in which disaggregation of collective doses could be done but gives no real guidance as to how the results should be used. It is important to recognise that although the definition of collective dose is as given collective doses for the public are not usually calculated this way in practice. For estimating doses from consumption of radioactivity in food account has to be taken of the fact that people do not obtain their food from where they live. For both aquatic and terrestrial foods collective doses are often estimated based on distributions of the production of the food and do not give information on who eats the food. It is not therefore possible to obtain accurate information on the breakdown of the collective dose into the contributing individual doses. The timescales of collective dose estimation are also generally longer than individual lifetimes which further complicates the issue. • RPD welcomes the recognition that a broad range of factors need to be taken into account in optimisation and the important part that ‘stakeholders’ can play. • The emphasis on distinguishing between exposed groups on the basis of age and gender seems to contradict the discussion on averaged effective dose in Section 5.5 of the Committee 2 Foundation report on dosimetric quantities. As reference values are used to estimate effective doses, which are averaged over both sexes, it does not seem consistent to then talk about distinguishing between different groups by gender. • One thing missing from the report is any mention of networking. The European ALARA Network (whose work is referenced in the report) has, for many years, focussed on the practical implementation of the optimisation principle. The Network involves international co-operation and sharing of information, and has helped advance optimisation in a number of practical ways. The role of networking, as an efficient means of promoting optimisation and good practice, has been recognised by IAEA who are now actively establishing networks in other global regions. It is suggested that some reference to this is included in the report. Specific comments Paragraph 9. The use of the phrase “precautionary principle” may be misinterpreted. This phrase is commonly used in relation to agents for which no adverse health effects have been properly demonstrated, but for which protection measures are nevertheless introduced. In contrast, for ionising radiation, adverse effects have been demonstrated, albeit at higher doses, and there are reasonable grounds for assuming a risk at low doses. Also, it is suggested that the word “immediately” (from “raised immediately”) be deleted. Paragraph 11. In the second sentence, it would be better to add the word “help”, ie “introduced a cost-benefit model to help implement the principle in practice.” Paragraph 14. This paragraph assumes an over-reliance on the ICRP 22 cost-benefit model. In practice, many users were already realising that this would only be part of the optimisation process. It is also suggested that the word “concrete” is deleted. Paragraph 15. A reference is required. Paragraph 20. Many of the issues listed here have also been identified by Workshops organised by the European ALARA Network, and additional references could be included here. Paragraphs 21 and 22. These two paragraphs have the potential to cause confusion, in particular due to the terminology used. The phrase “quantitative individual dose restrictions” is unclear: is this a reference to dose limits? (Paragraph 22 implies this). If so, this should be made clear. Paragraph 23. The phrase “basic levels of protection” requires more explanation, especially as to how it can be used to describe (equally?) both dose limits and dose constraints. In the wider context, the concept of dose constraints remains very unclear, especially in light of the 2005 ICRP Draft Recommendations. Dose constrains were previously well-defined in relation to the optimisation principle (for example, to avoid inequity in individual doses, to represent doses from well-managed operations, etc). It would be helpful to remind the reader of this previously philosophy, and to clearly indicate whether this had changed as a result of the 2005 Recommendations. Paragraph 24. The first sentence is not always true. In operational protection, the “single source” to be optimised is usually much more specific - a particular source, work area, task, etc. The introduction of the phrase “exposure situation” is potentially confusing, and it might be simpler just to state that the examples given are equivalent to “a source” for the purposes of setting dose constraints. The final part of this paragraph refers to the dose constraints recommended by the Commission, and how they help in choosing values for specific situations. However, as previously noted (in comments on the draft 2005 recommendations), the Commission has provided little indication of how the system will operate in practice. It would have helped to provide some additional practical guidance here. Paragraph 25. See comments on paragraph 23. In this paragraph there is mention of dose constraints being upper bounds to the optimisation process. However, further explanation is needed, especially with regard to the role of dose limits. Figure 1 is potentially of use, but needs to be better presented, and perhaps developed further. For example, the term “dose” is very vague – does it mean individual (annual?) effective dose? Do the graphs apply to occupational and public exposure? What is an authorised level? Also, the time-line needs further consideration. For example, for a planned operation would it not be possible to change the dose constraint in the future (normally downwards, but could increase under certain circumstances)? The graphs should also include the dose limit, where applicable. For emergencies situations it is the intervention that should be optimised not the dose reduction procedure to get to an intervention level. A figure such as that given for controllable existing situations would be better, replacing “existing exposure” with “dose resulting from an accident”. Paragraph 26. The phrase “tend to” is too weak, and “are intended to” is a suggested replacement. Paragraph 27. It is not clear why “conceptually the process is broader”. The definition, as given in paragraph 26, does include accidents/potential exposures. Safety culture and other approaches are indeed the key in terms of operational protection, but these are a logical outcome of the original definition and concept. It is suggested that the word “conceptually” is changed to “operationally” or “in practice” (or similar). Paragraph 28. Safety culture is essential, but it is not “the basic role of the concept of optimisation”. The basic role is dose restriction, and this is an important distinction. The report is right to stress the importance of safety culture, but is in danger of focussing on this too much. Practical protection often relies as much on “hardware” (shielding, engineering controls, etc) especially where “safe thinking” is an insufficient precaution. It would help if the reader were reminded of this. It would also help to emphasise the need for a balanced approach carrying on reducing doses at any cost is not optimising. Paragraph 29. The meaning of the second sentence “this process presents several well-defined characteristics” is unclear, and does not link well with the following text. What is meant by “characteristics” and how are they “presented” by the process? The subsequent text suggests two “characteristics”, not several, and neither appears well-defined. Paragraph 30. The first step in the process should be to define the problem and the scope of the optimisation process to follow. The phrase “scientific considerations” could be regarded as biased, i.e. suggesting that other considerations are (by default) “unscientific”, a phrase associated with unreliability or inaccuracy. The phrase “technical considerations” may be better. Paragraph 31. “When, where, how and by whom” is a good starting point, but it seems optimistic to assume that this will yield the full set of attributes listed. In terms of the characteristics of the exposure situation, “how much?” and “from what?” are key questions to help establish what is causing the particular exposures of concern. This is by no means obvious in many working environments, and this information is needed to frame the process and to choose effective protection options. By the same token, a full set of technical, economical, social, environmental and ethical considerations are very unlikely to be produced from a simple “when, where, how and whom”. It is suggested that the text be modified to indicate that the simple questions are a good starting point, but that there may be a wide variety of things to consider, and that further investigations will often be required. The last sentence (participation of stakeholders) is correct, but more detail would be welcome. For example, which stakeholders might typically be involved? What method of participation is envisaged? What is the advice for dealing with (the inevitable) disputes between different parties? Paragraph 32. Although the list is not exhaustive, it would also be helpful to indicate that it may be too extensive in many situations, and that the process should be proportionate to the scale of the problem. For example, a much shorter list of attributes will be appropriate when tackling most occupational exposure problems. In the list, the characteristics of the exposure should, where possible, identify and characterise the source(s). In some cases this is obvious, but often it is not. The list of “social considerations and values” contains many undefined (and unclear) elements. For example, the difference between “equity” and “fairness” is hard to understand. Phrases such as “sustainability” and “intergenerational consideration” need explaining, etc, etc. The list of “Environmental Considerations” could be wider. For example it could include impacts on climate, or (on a local scale) amenity/nuisance issues such as visual impact, noise, etc. Paragraph 37. This paragraph would benefit from more detail/explanation. In general, it is correct to say that the aim is to select the “best” protection option, but this naturally raises further questions. For example, best for whom? (See also comments on paragraph 55). Similarly, transparency and clarity are clearly desirable, but more detail on how this might be applied in a range of practical situations would be useful. Paragraph 39. This begins by talking about environmental actions but it is not clear where this leads in the remainder of the paragraph - this seems to be confined to source-related controls. Paragraph 40. This paragraph is too vague. For example, the meaning of the phrase “professional disciplines” is very unclear: what are co-ordinates? What is the difference between working groups and committees? In practice, organisational structures refer to stakeholders such as managers, line-managers, safety advisers and officers, employees and their representatives. The last sentence refers to “production” and “radiological activities” neither of which are clearly defined. It is suggested that this sentence is deleted. Paragraph 42. This paragraph seems rather self-evident and could be deleted. Paragraph 43. When discussing the relative merits of quantitative and qualitative approaches, there is an important distinction between methods/tools and issues/factors. The text refers to tools such as real-time software and operational dosimetry systems. Such tools provide insight into the pattern of occupational exposures and (most importantly) help identify protection options. In practice, they have been shown to be capable of making significant advances in operational optimisation. The text in this paragraph, however, combines such operational tools with wider decision-aiding tools (CBA, etc) and it is suggested that these be separated. Operational tools (ie that provide insights into exposure patterns and protection options) are usually quantitative, and may often be the sole input at certain stages of the process. In contrast, other quantitative (decision-aiding) tools should rightly be regarded as only one input, with quantitative issues being a greater consideration. Paragraph 44. The approach to low levels of (public) exposure is a key issue, for example when considering discharges to the environment, or remediation of contaminated sites. Some guidance from the Commission would be extremely useful – if not here, then in a future report. Paragraph 46. This paragraph states that effective implementation requires that “all stakeholders involved know and agree with the basic assumptions of radiological protection”. This would seem to be a key issue worthy of further discussion, especially in view of the increasing involvement of stakeholders (see paragraphs 48 to 54). Is the Commission suggesting that such agreements are pre-requisites, ie should stakeholders that do not agree with the basic risk estimates be prevented from being involved? There are examples involving public exposures where some stakeholders have insisted on a “zero dose” endpoint as the only acceptable solution. In some cases, this is due to a complete rejection of the “basic assumptions”, in others it is due to overriding societal issues. Resolution of these issues is far from easy, and further guidance would be welcome. Paragraph 47. This paragraph is rather all-embracing but in terms of responsibilities for those exposed seems to be more concerned with intervention and long-term exposure to existing contamination. Involuntary exposure as a result for authorised discharges does not seems to place a great deal of responsibility on the people exposed. Is it realistic to expect them to inform the site operator or regulator when circumstances that affect doses change? Surely it is still primarily for the operator to keep local conditions under review. Paragraph 52. “Plurality of views… will generally enhance the quality and effectiveness of the final decision” – it is hoped that this is the case, but it may not be so easy in practice. Stakeholder involvement is strongly supported, but conflicts are inevitable and the approach to resolving these needs to be considered now. It is accepted that the issue may be mainly outside the Commissions remit. However, in promoting stakeholder involvement, the potential problems should at least be acknowledged. It should also be stated that involvement of local stakeholders is also likely to make any final decision selected more acceptable to the wider community rather than relying on just technical experts to generate the optimum process Paragraph 53. We are pleased to see a description of what constitutes stakeholder involvement. However, it would be better to see the recognition that it can cover consultation processes, blank sheet approaches etc earlier on when the concept is first introduced. Say after paragraph 20. Paragraph 55. Reference is made here to the “most appropriate” protection options, whereas paragraph 37 refers to the “best” protection option. Although these could mean the same thing, most people would consider “most appropriate” to carry less force. The phrase “the best option under the prevailing circumstances” is preferred, notwithstanding the previous comments on paragraph 37. Paragraph 56 (and summary). It is not clear what is meant by stating that the exclusion levels should not be considered as a relevant endpoint for optimisation. Is it that the optimum level can be above the exclusion level or that it is acceptable to optimise below the exclusion level? Paragraph 58. This paragraph contains a number of debatable statements. There is no absolute reason why the results of optimisation will demonstrate that the process complements the use of constraints – indeed, it is not clear why this requires demonstrating. Similarly, “the residual dose will be the primary tool to demonstrate this compliance” – compliance with what? In any case, the preceding paragraphs makes it clear that the achievement of ALARA is not a question of the residual dose per se, but of “what else can be done?” It is suggested that this paragraph be deleted. The term ‘residual dose’ is introduced here for the first time and it would have been helpful to have definitions such as this in an earlier single section. Paragraph 59. The argument in this paragraph appears to be left hanging, and further explanation would be useful. Paragraph 62. Source may not be the correct term to be used here. There is potential confusion between a source of radioactivity and the source of discharges, i.e. a practice. This is illustrated by the definition of exposed individuals, where it refers to distance from the source. A source could be a contaminated area of land, in which case, the distance is a difficult value to determine. The final sentence also refers to the length of time since exposure. It would be clearer to say length of time following release of radioactivity to the environment, as time since exposure would imply there is no ongoing exposure. It would also be helpful to cover existing situations, where measurements can be used to estimate dose distributions. In this situation accuracy does not necessarily decrease with distance. Paragraph 70. The introduction of a group dose referred to as a "sub-collective" dose leads to confusion and its usefulness is not clear. Surely what is needed is to identify population groups that are exposed in terms of their characteristics (e.g. are we concerned principally with a particular location). It is certainly true that the size of the group may be an input to optimisation (ie doing the most good for the most people) but this is something different. Paragraph 71. Care needs to be taken to ensure the level of detail in choosing sub-groups is not excessive and that therefore the results cannot be relied upon. For example, the majority of consumers do not obtain all of their food from where they live. Therefore, the ingestion pathways should not be calculated for sub-groups based solely on their location. Surely all pathways should be considered, at least initially. Source-related optimisation is likely to affect more than one pathway. Paragraph 73. The use of weighting factors should be clearly identified in any dose matrix. For instance, the weighting factors may differ between assessments, e.g. different stakeholders. It would be unfortunate if two separate assessments were then compared without accounting for differing factors having been applied. Any decision-making should be clear in understanding the effects of such factors. This paragraph suggests weightings according to (mean) dose and time of exposure, and refers to examples in Annex 2. However, only time-weighted dose is considered in Annex 2. Weighting according to individual dose (mean, or bands) is considered a bigger issue, and an example of this (in Annex 2) should be given. Paragraph 79. The first sentence seems deliberately obscure – “an obligation of means” is a meaningless and empty phrase. And it is probably wrong (and certainly naďve) to downplay the importance of results. Optimisation may be a “state of mind” and an ever-ongoing process, but its effectiveness and value are determined by results. ‘What does ‘codified’ mean? Annex A1 (NB: paragraph numbering would help) Introduction. This states that occupational exposure is incurred at work and principally as a result of work. Does this is include exposure to discharges from work practices? I.e. public exposure to waste released from work with sources. Surely occupational exposure is just related to exposure when at work. Introduction Paragraph 2. It is not true to say that the individual dose received in the workplace is always well known. It is true for Category A workers, but the Commission may be over estimating the number of Category B workers that are subject to individual monitoring. Introduction Paragraph 3. Optimisation does not have a long lasting history in all workplaces, as is made clear in subsequent sections. Page 35, Paragraph 1. The list of things that can result from “Work Management” should include “ a reduction in the number of accidents”. Section 2.2 Medical Uses. It is true that some new technologies involve higher doses and new challenges. However, it is also the case that higher doses still arise from the use of old equipment in well-established techniques. Optimisation depends on the equipment as well as working procedures, and it would be useful to make this clear. Section 2.3 Natural Radiation. This paragraph suggests that workers be given the same level of protection (ie as for artificial sources), then immediately contradicts this argument. In practice, the level of protection is rarely the same, and it is suggested that the sentence “Workers exposed to ….” be removed. Section 3. Distribution of exposures. Paragraph below Figure 4: Data on occupational exposures from natural sources must be interpreted with great care. Available data (eg from ESOREX) is extremely limited and represents a very small fraction of exposed workers. The percentages quoted (for 10-20 mSv y-1, and > 20 mSv y-1) are almost certainly gross overestimates. See the “SMOPIE” report (www.nrg-nl.org) produced for the EC for more details. Figure 5 – is a useful summary, but in terms of presentation a vertical down arrow is preferred to a diagonally down-sloping arrow. Second paragraph below Figure 5. Suggest delete “be aware of” and replace with “provide information on”. It is not clear what the purpose of each type of dose matrix is. For example, how are age and gender of any relevance in practice? The impression is that the concept of dose matrices has been imported from the public dose arena. For occupational exposures, the use of such matrices should be gauged in terms of whether they can actually help the optimisation of doses in practice. Section 5. Conclusions. The conclusions seem overly brief, and miss out many of the points raised in the text. For example, more mention of stakeholder involvement would be appropriate. Annex A2. RPD welcomes the recommendations of a realistic approach for modelling exposures. However, this annex lacks focus and requires practical guidance and examples. Much of this relates to individual dose, which is contrary to the approach based on sub-collective doses given in the main text. A dose constraint of 1 mSv/y is proposed when a single source contributes the principal route of exposure. However, this does not include contributions to doses from discharges in earlier years. One might therefore have expected a value below 1 mSv, or a suitable comment about historical discharges. For accidents and long-term contamination, a dose constraint of 20 mSv/y is suggested. This does not seem to be consistent with the 10 mSv value used for prolonged exposures. Section 3 on the distribution of exposures refers to the radioactive half-life of the radionuclide affecting the age group that should be considered. Of importance here is also the time that the exposure will continue, e.g. continuous discharges over the lifetime of a plant will lead to the exposed population changing ‘age-group’ even for short lived radionuclides. It is not clear that it would be acceptable to weight differently doses delivered in the future and those delivered now – this does not seem to be consistent with the concept of sustainable development. Annex A3. Page 51, first full paragraph Epidemiology has now demonstrated radon risks below 200 Bq m-3. This paragraph should make it clear that national authorities may reasonably set action levels below the lower bounds of the ranges specified in Publication 65. Indeed, it would generally be regarded as unacceptable to set a protection standard above a level at which risks were demonstrable rather than based on extrapolation. (Note that Publication 65 was less rigid than present ICRP drafts in allowing national authorities flexibility) It may not fit well with ICRP’s general approach, but cost effectiveness analysis can be used to guide the selection of national radon policy. It might be better to acknowledge that this will sometimes be done. Page 52, line 10 of text “... process may include two steps...” (not “should include”). It may well be perfectly reasonable to proceed directly to remediation without further investigations. Page 52, lines around 28 In fact the radon risk has been demonstrated only in adults; a larger risk in children is hypothetical. Page 53, line 7 of text “Aeration” is an odd word; “increased ventilation” would be more usual and surely means the same thing? Page 53, paragraph 2 Preventative measures in new build are much simpler and more cost-effective than remedial measures in existing dwellings. Preventative measures should have the effect of significant reduction in all new buildings in which they are incorporated; not just preventing very high levels. These points should be made. Editorial comments Paragraph 22. First sentence change “it” to “there”. Paragraph 34. It is suggested that the start of the third sentence be changed to “This process should……” Paragraph 47. It is not clear why there is a reference to IAEA 2002 at the end of this paragraph. Paragraph 54. The word “occulted” is grammatically correct, but is rarely used in this context. It is suggested that it be replaced with “obscured”. Paragraph 63. There is an extra ‘dose’ in the sentence. Annex A1. Paragraph below Figure 5. First sentence, typographical error – “build”. Annex A1. Section 4. The role of operators and authorities. Second paragraph; The third sentence (“In many countries….”) is not grammatically correct. Last paragraph: In the last sentence (“The degree of stakeholder involvement…”), it is suggested that “could” is replaced with “should”. Annex A2. Section 2. The paragraph at the end of page 44 contains the phrase ‘must not be occulted’ this needs to be redrafted to be more understandable. Annex A3 Page 50, line 3 of text“... - Bq m-3 or less)” Page 50, line 11 of text “... its harmful effects in dwellings have been recognised ...” Page 52, line 5 of text “... (not in unoccupied cellars or ...”. Note that the English needs a polish around here.