TL: The discharge of radioactive waste into coastal waters and its subsequent dispersal: a case of transfrontier pollution (GP) SO: Greenpeace UK DT: June, 1983 Keywords: oceans ocean dumping radioactive waste nuclear power uk europe germany conferences gp reports / A paper to the Paris Commission meeting in Berlin, June, 1983. Submitted by Greenpeace (UK) and Greenpeace (France), and prepared by Peter Taylor, consultant. Summary In normal operation the majority of land-based nuclear installations (e.g. reactors, fuel fabrication plants, enrichment plant), discharge very small amounts of radioactivity into the aquatic environment. In the case of coastal stations radioactive contamination is normally localised and not measurable above background levels more than a few kilometres away. A major exception occurs with reprocessing operations (recovery of plutonium and depleted uranium by chopping, acid leach, and chemical separation processes). Reprocessing accounts for over 90% of all radioactivity released to the environment from nuclear power programmes, weapons development and research activities. Thus, the degree of decontamination of waste streams before discharge is of major importance. This paper draws attention to very great disparities operating with regard to control of discharges from reprocessing plant. In Europe, two installations carry out large scale commercial operations: Cap de la Hague in France, and Windscale, in England. These two installations differ by a factor of 100-1000 in site discharge standards, and the tighter control operated by the French plant coincides with an environment that favours greater dilution and less risk than Windscale. The disparities are not explicable in terms of throughput, age or reprocessing technology, but reflect site management philosophy and different interpretation of international guidelines. Moreover, experience in other countries where reprocessing is or has been carried out, or is in the planning stage, shows that higher standards can be achieved at negligible cost if instigated at the design stage. This is an important factor as both Cap de la Hague and Windscale have expansion schemes with planning consent which do not make use of the best available technology. The matter is not simply precautionary. Effluent from the Windscale plant is now responsible for marked rises in the radioactivity of the Irish and North Sea, and hence of various commercially important seafoods. In some cases levels are now higher than those resulting from the testing of nuclear weapons in the atmosphere and which led to the test ban treaty. Ireland, Norway and Denmark are particularly affected. The currently accepted scientific opinion is that very small doses of radioactivity, although presenting risks that any individual might regard as negligible, should be assessed for their collective and cumulative effects on populations; that these adverse health effects be costed and cost benefit analysis applied to the control of discharges. In the case of Windscale, the UK Ministry of Agriculture, Fisheries and Food has carried out such assessments. Windscale's pollution is so widespread that health effects are assessed in millions of pounds, and œ100 million pounds is to be spent on some aspects of control. However, these calculations and decisions are being made within a national framework and with limited participation of those populations at risk, (local councils have some representation). Where effects on neighbouring states are concerned assumptions are made of a cost-benefit nature without consultation. Moreover, current plans for control at Windscale will not significantly restrict certain long-lived pollutants such as plutonium and americium. These radioactive elements will remain toxic for tens of thousands of years and their future behaviour in the environment is difficult to predict with precision. International guidelines (of the International Commission for Radiological Protection) are open to wide interpretation - for example, several reprocessing nations (Japan, USA, and USSR) will not permit plutonium (and americium) discharges at any significant level. Regulations in force in Japan, and planned control technology in Germany and the USA, demonstrate plutonium effluent control up to 1 million times that of Windscale and several thousand times that of Cap de la Hague, and yet unit costs per tonne of reprocessed material are not significantly different. Similar differences in environmental protection philosophy and technology employed also apply across-the-board with other radioactive contaminants. It is argued that the current international regulatory bodies: the Commission of t;he European Communities, the International Commission for Radiological Protection (ICRP), the International Atomic Energy Agency (IAEA), the Nuclear Energy Agency (NEA) of the OECD, have not proven effective in establishing a common standard with regard to discharges to the marine environment, and that at present these bodies are not constituted for the task. The IAEA has no statutory function, and is in any case a promotory body; the NEA represents the OECD nations, but is also a promotory body; the CEC does not represent certain Nordic interests; the ICRP guidelines already embody a social cost- benefit assumption applicable only to states with nuclear power programmes and in any case provides a framework which allows very great disparity. It is therefore concluded that the Paris Commission provides a forum for realising a standard of control for Windscale and Cap de la Hague, which is not currently available: the Commission includes all states affected by the discharges. Furthermore, the matter is not a problem of specialist scientific interpretation, for although there are areas of scientific controversy, the main question of standardisation can be approached adequately within the bounds of currently accepted scientific knowledge. There are sound reasons why regulation and standard setting, although requiring scientific data, should not be left to scientific working groups (e.g. ICRP), nor promotory bodies (NEA,IAEA). The nature and significance of the discharges from Cap de la Hague and Windscale Monitoring and Assessment A reasonably accurate record exists of the discharges from the two reprocessing plants. Data for Windscale is in greater detail than for the French plant. Monitoring data for the Windscale environment is far more extensive than for Cap de la Hague, and extends to the North Sea. A large number of radioactive elements are discharged and each has different characteristics which affect its ecotoxicity. The toxicity of an element will depend upon its half-life, the nature of the radiation emitted (alpha, beta or gamma), its distribution in the environment, and in particular, concentration in food chains or by biogeochemical cycles. The behaviour of some elements is relatively well known and predictable, (e.g. caesium-137, as a result of fall-out studies, and its ready uptake in biological materials), whereas others are less readily predictable (e.g. plutonium which binds to sediments and is subject to geochemical and hydrological processes over a very long time span). Nevertheless, a substantial monitoring and assessment program me exists for assessing the effects of these pollutants in the marine environment. The UK spends several million pounds annually and maintains a specialist laboratory within the Ministry of Agriculture, Fisheries and Food (MAFF) at Lowestoft in Suffolk. The MAFF laboratory publishes its surveys: the 1979 data was available in 1981. In addition, the West German Hydrographic Institute samples levels in the North Sea and English Channel. The situation at Cap de la Hague is less satisfactory, but data for local contamination of the Cherbourg peninsula is available. The MAFF sampling now includes the Channel Islands. The type of monitoring varies according to two principles: sampling intended to determine levels of exposure to man (e.g. food species), and indicator species. The latter are sampled at intervals. Scientific analysis is often made difficult by the incompleteness of yearly sampling, but generally speaking, sampling is adequate to gain a picture of the dose levels. It is generally agreed that MAFF use maximising assumptions and thus are liable to overestimate doses, particularly for the more highly exposed local population in Cumbria. However, when calculating the collective dose (the sum of all the very small individual doses received by consumers of fish contaminated by dilute radioactivity far from the plant), there is room for quite large errors and the error band is not given by MAFF nor are the detailed assumptions of the model published. Such collective doses involve detailed knowledge of fish landings and consumption patterns for the Irish Sea, North Sea, Norwegian coastal waters and arctic ocean fisheries, together with pathways arising from industrial processing for feed stock. Nevertheless, because of the advances made in the epidemiological study of the effects of low level of radiation, MAFF have been able to provide rough estimates of the dose to populations within the UK and other European countries, and more accurate estimates of the dose to local consumers. We have seen no comparable estimates for Cap de la Hague. However, in the latter case, the effects are likely to be overshadowed by Windscale. There are two views of the significance of the levels of exposure that arise from the discharges to the marine environment: Local Consumers It is generally agreed that the discharges from Windscale in the 1970's were unacceptably high, having led to significant increases in the radiation exposure of local consumers and to the population of NW England, particularly the Cumbrian coastal region. At its worst, the small number of people in the 'critical group' were exposed to an additional 200% of their normal background, and the coastal population in general to approximately a 10% increase. This arose primarily from radioactive caesium contamination of fish and shellfish. In the late 1970's BNFL started to build a new effluent treatment plant, variously estimated at œ50-100 million to control the caesium. Management of spent fuel stocks led to reductions in caesium discharges also. Distant consumers Outside of heavy consumers of fish and shellfish from landings in the northern Irish Sea, who may receive 1% or more additional radiation to background, consumers of fish stocks further afield, (see map of the extent of pollution of North Sea and arctic waters), especially in northern Europe, receive far less than 1% additional radiation burden. Thus, the individual risk (of cancer, genetic damage etc) is very small in relation to the normal risk in daily life. The UK regulatory bodies have argued, therefore, that such levels can be regarded as 'insignificant' and not summed in the same way as higher doses, or at least the sum of collective doses not be evaluated in an equivalent way. However, current international guidelines (ICRP) do not differentiate, and recommend that all low doses are summed to produce a collective dose figure which gives a measure of the total health damage to a population. MAFF have followed these guidelines and produced a collective dose figure for the UK and for western Europe, (the damage arising from one year's discharge but integrated over the life of the pollutant in the environment). In radiological jargon the damage is estimated to be: U.K. 130 man-Sv (13,000 man-rem) Europe 170 man-Sv (17,000 man-rem) The evaluation of these figures in human terms is subject to some controversy (of the hazards of low doses) and a range of dose effect estimates exists. However, the range is only a factor of 10: the ICRP is generally regarded as conservative, with the US Academy of Sciences providing a range to a factor of 3 above ICRP. However, recently the chairman of its committee on the biological effects of radiation (BEIR committee), Prof.E.P. Radford, has argued for a factor of 10 reassessment. Using this range, 10,000 manrems produces estimates of 1-10 fatal cancers, 1-10 major genetic defects, 1-10 non-fatal cancers and an unknown number of hidden genetic effects. ICRP Recommendation No 22 recommends that regulatory bodies apply quantitative cost-benefit analysis to discharge control. MAFF in UK have attempted to ascribe monetary values to a man- rem (at $100), and the UK NRPB have suggested a sliding scale according to dose level, with $1000 for a man-rem at doses at a significant percentage of background and much less for 'insignificant' doses, with very small doses not counted at all. This approach has certain advantages in that attention and resources are concentrated on those people most at risk. However, such an approach has not found universal support because of the arbitrary nature of deciding what is an 'insignificant' or 'negligible' risk of cancer, and also because other less easily quantifiable factors are involved in the assessment of pollution: people are concerned with intangible aesthetic aspects of clean food rather than mathematical calculations of risk. Indeed, the ICRP have stated the problem quite clearly (ICRP Report no 22) and recommend that such factors be taken into consideration, however, they provided no guidelines as to how this could be achieved. The US Environmental Protection Agency has proposed a value of $1000 per man rem be used in cost-benefit calculations with no sliding scale. If this figure is used, then the Windscale effluent leads to the following 'costs': Annual discharge: $13,000,000 (UK) $17,000,000 (Europe) Cumulative effects to date after 25 years of discharge: approx. $300,000,000 (UK and Europe) Future effects if levels maintained for 10 years: $130,000,000 U.K. $170,000,000 Europe On this basis BNFL would be required to install control technology if its installation costs were less than ca L190 million for the usual ten year life of the process plant. At present they are installing a Site Ion Exchange Plant (SIXEP) to control caesium-137 discharges at a reported cost of ca ~100 million, but according to the UK Department of the Environment (who set discharge limits), it is not yet clear what the SIXEP plant will achieve and the DOE intends to wait until BNFL have an indication of its performance before they set the authorization. This approach has led to some criticism. Thus it can be seen that current policy has allowed widespread pollution at a not-insignificant level when assessed by accepted international guidelines. It should be noted that eminent authorities are arguing for 3- to 10-fold revisions of the dose- effect assessment, and hence these cost figures could be revised by those factors. Adequacy of control and regulatory procedures The situation above speaks for itself, in terms of the philosophy and interpretation of international guidelines. The crucial question of control hinges upon the following precepts: i) is the discharge necessary (i.e. are there alternatives or is the activity in itself of benefit), ii) does it lead to doses to individuals above the internationally recommended limits, iii) dose should be as low as reasonably achievable (ALARA), social and economic factors taken into account. These precepts are recommended by the ICRP, who also recommend the 'dose limit' for members of the public (500 mrem, 5 mSv, per year). There are several problems with this approach: a) how is to be decided whether a discharge is resulting from a beneficial activity ? b) if it is so decided that an activity, in this case reprocessing, is of benefit to the community, then are there alternatives to discharging wast;e: e.g. decontaminating effluent and storing the waste, c) on what basis are 'international' limits recommended when it is known that any increase in doses represents a risk (i.e. any recommendation of a limit must therefore embody a social judgement of what is an acceptable risk), d) how is the ALARA process, which patently involves value judgements to be carried out: in particular, who is to participate and how are differences in evaluation to be mediated (this becomes important when the populations who receive the 'detriment' are different from those receiving the 'benefit', and especially so if those populations are far afield and from states with no nuclear power program me who may have so opted, precisely because of the risks involved. This matter is made doubly difficult when the activity is reprocessing, because there are substantive differences of opinion as to the merits of this on waste management grounds, and its only defendable argument is the production of plutonium for weapons and/or Fast Breeder Reactors. Neither of these activities can be said to command universal support either between states or within nuclear equipped nations, especially when plutonium stockpiles at Windscale amount to 10-20 tonnes and nuclear warheads may require as little as 5 killogrammes. Even were agreement reached on the net benefit either for defense, or on environmental grounds, and even if the 500 mrem dose-limit is accepted, then the ALARA principle requires a procedure to decide what is reasonable in economic and social terms. At present the U.K. and French governments assume that they have agreement on items (i) and (ii) above, and proceed to 'optimise' their discharges. The results have been indicated. They differ enormously between the two plants. The procedures are not clear in either case, but the matter is clearly left to specialists and with Cogema and BNFL having overriding influence in terms of stating what is 'reasonable' to achieve (see further critique below). Whatever procedures are used, it can be seen from Tab]e 1 Appendix l and Table l Appendix 2, how this has translated in practice: Caesium-137/l34 The most important nuclide in terms of the widespread effects (see Figs 1), particularly where Windscale effluent is concerned. There was a very large increase (a factor of 5) at Windscale in 1974, due to management problems of spent fuel held in pond storage: corrosion led to contaminated pond water having to be discharged in order to protect workers. In the period prior to this particular event, which has affected Windscale's figures for the whole of the period since then, Cap de la Hague discharged about 1/10th to 1/20th the Windscale levels for about 1/5th of the throughput. Other nuclides (e.g. Sr-90) In general, other nuclides have much the same order of magnitude difference. Sr-90 from Cap de la Hague varied in the 1970's from 100-2000 Ci/annum, whilst Windscale ranged from 6000-15,000. Similar factors operate for ruthenium-106, another major constituent of the effluent. Plutonium and other alpha emitters In the period 1971-1977 Cap de la Hague discharged from 4-30 Ci of alpha emitters (plutonium primarily), whereas Windscale discharged from 1600 to 4~300 Ci, a factor of 150-250 difference. Factors affecting discharges A number of factors in relation to the activities of the plant affect the discharges: throughput, the nature of the fuel (especially the length of time in the reactor in Megawatt- days/tonne), the nature of the process technology, hold-up times of waste water, and any other activities at the site (e.g. weapons laboratories and isotopic enrichment programmes for plutonium). These may account for a factor of 3-5 and it would be expected that Windscale with its larger throughput would have the higher levels. The factor of 100+ on Cs-137 is due to lack of foresight in pond-water treatment installations in the event of hold-ups due to difficulties in the process plant, as well as a long delay in deciding to tackle the discharge. The 100+ factor for plutonium and other alpha emitters may be due to military activities at the Windscale site, but two conflicting statements have been made by BNFL, and because of national security it is not possible to ascertain precisely where the plutonium and americium comes from. However, it is known that Cap de la Hague has a flocculation treatment plant for alpha emitters, whereas Windscale does not. The authorised limit at Cap is 90 Ci/annum, whereas at Windscale it has been between 4000-6000 Ci/annum. Clearly, a number of managerial decisions have been made in relation to control technology. BNFL have stated that the different environments must be taken into account (in reply to the document Appendix 1), but it is clear from environmental sampling that the Cap environment is more dispersive and that Windscale should be adopting a stricter standard on environmental grounds. European Reprocessing P]ants in comparison with others Details of discharges of other plant of a similar scale are sparse. Only Japan now operates reprocessing for commercial stations on an industrial scale. However, one thing is clear. The technology to control low level liquid discharges does exist and does not markedly affect the cost of plant if built in at the design stage. Japanese discharge standards for beta/gamma emitters are hundreds of curies per year, rather than thousands or tens of thousands. For plutonium and other long lived alpha emitters, standards are much stricter - of the order of millicuries. The early KEWA plans for reprocessing in Germany provided for a mCi discharge of plutonium for a plant 50% larger than Wind scale. In the USA and Russia, discharges of plutonium to the aquatic environment are effectively outlawed, with the US Rocky Flats process plant having achieved a total discharge of 0.75 mCi over a 22 yr period, with a plutonium production well in excess of that of Windscale. Thus, control technology exists to reduce beta/gamma emitters of 100 and alpha by a factor of 1 million, at marginal cost if built in at the design stage. This last factor underlies the urgency of setting a common standard at this present time, as both Windscale and Cap de la Hague are undertaking massive investment programmes (in excess of $1000 million) in refurbishing old process plant and building new capacity for LWR fuels. Managerial attitudes to environmental levels As already noted, BNFL have already agreed to reduce caesium-137 discharges. However, in evidence to the 1977 Public Inquiry on Windscale, they stated that overall beta/gamma levels would remain of the same order. Moreover, they saw no reason to adopt the stringent regulations with regard to plutonium that other states had adopted. At BNFL, managerial attitudes are determined entirely by adherence to ICRP dose-limits, and to a 'reasonable' factor below that representing their own cost-benefit assessment, usually at 5-10% of the dose limit. In fact, historically, local consumers have been regularly exposed to 25-50% of the ICRP dose limit. It should be noted that standards in the USA, Sweden and Germany are stricter, with site limits of 25-30 mrem/annum compared to ICRP's 500 mrem/annum. In Japan, site limits are 5 mrem for nuclear reactors, and most probably reprocessing will follow the US EPA 25 mrem limit. Dose data for Cap de la Hague is not readily available, but local environmental levels are 1000x less than at Windscale, representing both lower discharges and greater dispersion. In terms of long distance pollution, Cap de la Hague is not measurable north of the English Channel because of the overriding effect of Windscale pollution rounding the Scottish coast and entering the North Sea (see Figs). One major criticism of relying on annual dose-limits, is that nuclides which at first disappear from the immediate environment of man and thus do not give a dose (e.g. those incorporated on sediments), are not subject to control. In this case, some prediction of their future behaviour has to be made. This is especially difficult for plutonium with a half-life of 24,000 years. (It is for this reason that other states have adopted stricter standards). There is therefore no adequate international guideline on this matter other than subjective judgement on the adequacy of marine ecology predictions and the future values of generations to come. Conclusions Discharges from the reprocessing plants at Cap de la Hague and Windscale represent the major proportion of radioactive discharges to the seas around Northern Europe. There are very large differences in discharges from the two plants, particularly with regard to plutonium. In the case of Windscale, pollution is now widespread and affecting fisheries in international waters (especially, Norway,Denmark,and Ireland). This pollution is not insignificant and has been costed according to internationally recommended procedures by the UK authorities, and figures range to several hundred million dollars per annum. Over 50% of this detriment occurs outside the U.K. The technology to control discharges exists, is readily available, and has been employed in Japan, USSR and USA. It does not affect overall processing costs if built in at the design stage. In the special case of plutonium, other states have adopted standards of control a million times stricter than Windscale and 1000x Cap de la Hague. Managerial and government controls in U.K. and France are based upon the assumption that ICRP standards apply, whereas ICRP guidelines already subsume social and economic judgements that other states may not share: e.g. the necessity to reprocess, or the overall benefit of nuclear power itself. ENDS