• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Comparison Between Radiological and Chemical Health Risks Assessments: The Nord-Cotentin Study

Posted on: Sunday, 31 July 2005, 03:00 CDT

ABSTRACT

In 1997, the French Ministries of the Environment and Health commissioned a detailed radioecological analysis of the Nord- Cotentin region in response to public concern about radiological risks associated with local nuclear facilities. This work was entrusted to the Groupe Radiocologie Nord-Cotentin (GRNC), a working group of experts from various origins (industrial facilities operators, public institutions, monitoring agencies, public interest and citizens groups, foreign experts). An epidemiology investigation in 1995 had reported an excess of two radiation-induced leukemia cases in an area near a nuclear reprocessing plant, a finding that attracted great interest in France, and which stimulated the need for further investigation. After the publication of its report in 1999, the GRNC was again commissioned to perform, inter alia, a corresponding assessment on the chemical releases of the local nuclear facilities. This second stage is now achieved and has revealed important similarities as well as some important differences between radiological and chemical risk assessments when applied to the specific case of the Nord-Cotentin nuclear facilities. Due to the considerable amount of work and results of the GRNC, the purpose for this article is to briefly describe the main developments of the risk assessment methodology followed by the GRNC in both cases, to detail some of the main results and to identify and explain, at each step, the similarities and the differences. The whole technical documents that support these works are available on the Internet at [left angle bracket]http:// www.irsn.fr/nord-cotentin/[right angle bracket].

Key Words: exposure assessment, dose reconstruction, risk assessment, nuclear facilities, environmental risk.

INTRODUCTION

Several epidemiological studies examining the risk of leukemia in the Nord-Colentin region of France have been published since 1989 (Guizard et al. 1997; Laurier and Bard 1999; Guizard el al. 2001). Especially, in 1995, Viel el al. (1995) published the results of a study on the incidence of leukemia among people aged from 0 to 24 years living in a 35-km radius around the La Hague nuclear reprocessing plant. The study suggested an excess of two leukemia cases in the district of Beaumont-Hague, an administrative unit corresponding approximately to a 10-km radius around the plant.1 Then, Pobel and Viel (1997) published the results of a case-control study that attempted to determine the factors associated with the risk of leukemia among the young people of the Nord-Cotentin. The authors of the study concluded that their results provided "some convincing evidence in childhood leukemia of a causal role for environmental radiation exposure from recreational activity on beach."

The publication of these two studies aroused a hot debate, locally and nationally. That is why the French ministries of the Environment and Health asked a commission of experts for a new epidemiological study in the Nord-Cotentin Peninsula. In January 1997, the commission was set up, headed until June 1997 by Professor Charles Souleau and then two complementary missions were devoted, respectively, to Professor Alfred Spira for epidemiological aspects and to Annie Sugier to study the effects of levels of radioactivity in the Nord-Cotentin Peninsula. The latter commission, set up by Annie Sugier, was named the Groupe Radiocologie Nord-Cotentin (GRNC). The GRNC brought together experts from various organizations: regulators, governmental experts, operators, experts from nongovernmental laboratories, and foreign experts.

First, a detailed radioecological analysis of the Nord-Cotentin region, in response to public concern about radiological risks associated with local nuclear facilities, was realized (Laurier et al. 2000; Rommensetal. 2000). More than 50 experts participated in this radiological health risk assessment, which was especially devoted to a study of the radio-induced risk of leukemia in response to epidemiological results. After the publication ofits report in 1999, the GRNC was commissioned again to perform a corresponding assessment on the chemical releases of the local nuclear facilities and the GRNC was broadened further for the chemical risk assessment.

This second stage is now achieved and has revealed important similarities as well as some important differences between radiological and chemical health risk assessments when applied to the specific case of the Nord-Cotentin nuclear facilities. The aims of this article are to briefly describe the risk assessment methodology followed by the GRNC in both cases, to detail some of the main results, and to identify and explain the similarities and the differences of both assessments. The main efforts of this article are devoted to the achievement of the third objective.

RISK ASSESSMENT METHODOLOGY

The risk assessment methodology is classically organized around what is generally called the risk paradigm: hazard identification, dose-response evaluation, exposure assessment, and risk characterization (NRC 1983).

Figure 1. Method of assessment of the risk of radiation-induced leukemia.

In the GRNC's experience, the four fundamental steps of the risk paradigm can be identified in both risk assessments (see Figure 1 for the radiological risk assessment and Figure 2 for the chemical risk assessment).

For the radiological assessment, four thematic groups were formed with respect to the risk assessment methodology (Figure 1):

* The first group (WG1) critically examined discharges declared by operators of the Nord-Cotentin facilities and reconstructed missing data when necessary.

* The second group (WG2) collected and interpreted measurements made in the environment by the various stakeholders since the facilities were put into operation.

* The third group (WG3) estimated environmental radionuclides activities obtained by comparing the predictions of transfer models with environmental measurements.

* The fourth group (WG4) identified the exposure pathways and estimated the doses received by the local population and the risk of radiation-induced leukemia.

For the chemical assessment, the working structure chosen by the GRNC was similar because the first experience confirmed that it allows each stakeholder to be involved in the assessment process. Thus, several specialized groups were formed. But the mission devoted to each working group differed from those of the radiological assessment due to the differences in the database available to perform the chemical risk assessment.

Figure 2. Method of assessment of the chemical risks.

* The group "Source term" was responsible for the inventory of chemical discharges from the Nord-Cotentin nuclear facilities.

* Because only very few results of environmental measurements of chemical concentrations were available, there was no equivalent of the previous WG2. However, during the work, it appeared necessary to think about a relevant campaign of measurements to confirm some of the hypotheses used in the model and a specific group, "Measurements," was formed that recommended further measurements to be carried out and designed the proposed measurements campaign.

* The working group "Risk for humans" was responsible for the local population health risk assessment whereas the working group "Risk for the environment" was responsible for the risk assessment for the terrestrial and marine ecosystems. Both working groups shared the responsibility of modeling the transfer of chemicals into the environment. The following only focuses on the health risk assessment for humans and not the terrestrial or marine ecosystem.

At this global stage, the comparison of Figures 1 and 2 illustrates that the health risk assessment went through the same steps when applied to radioactivity and to chemicals. We will now discuss, within each step, the similarities and differences in the practical methodology implementation for radioactivity and chemicals.

Hazard Identification

Four nuclear facilities are located in the Nord-Cotentin (Figure 3): the Navy Yard at Cherbourg (since 1958), the nuclear fuel reprocessing plant run by COGEMA at La Hague (since 1966), the shallow land disposal repository facility, run by ANDRA at La Hague (since 1969), and the EDF nuclear power plant at Flamanville (since 1985).

In both radiological and chemical assessments, a retrospective analysis of discharges from the Nord-Cotentin nuclear facilities was performed. The GRNC started from the figures for liquid and gaseous discharges supplied by the operators. These figures were verified when possible and the GRNC also reconstructed the discharges of some elements for periods during which they had not been measured. The analysis of the source term focused on 81 radionuclides and 63 chemicals (Table 1):

* For the 81 radionuclides studied in the radiological assessment, several situations were met depending on the facility concerned (GRNC 1999):

* In the particular case of the reprocessing plant of COGEMA, it was possible, first, to verify the figures supplied by the operators (33 radionuclides), and, second, to add radionuclides that had not been individually identified by COGEMA (48 r\adionuclides). This was possible because precise information on the amounts of radionuclides contained in the spent fuel that enter the reprocessing plant were available and because many studies examining their behavior during the process were available. Based on this information, the GRNC was able to make an evaluation of the movement of each radionuclide in the reprocessing plant and to verify the coherence between what enters in the spent fuel and what comes out in the discharges, the by-products and the reprocessed fuel. The 48 radionuclides added by this analysis did not modify the order of magnitude of the results supplied by the operators, but they helped to give more exhaustive information about the composition of the discharges.

Figure 3. Study area.

Table 1. Inventory of the pollutants considered in both risk calculations.

* In the case of the nuclear power plant of Flamanville, such an analysis was not possible because the discharges are strongly correlated with the nature of materials used for the construction of the plant and the treatment and the management of the effluents, which are site-specific. That is why the quantitative inventory of discharges used by the GRNC relied essentially on the operator's figures, except for ^sup 14^C for which amounts discharged were reconstructed by the use of a production rate from the literature, and ^sup 63^Ni for which amounts discharged were reconstructed from some measurements made by the regulators and the use of a constant activity ratio between ^sup 63^Ni and ^sup 60^Co.

* In the case of the shallow land disposal repository facility, there is no direct relation between the inventory of radioactive materials present on the site and discharges into the environment, because the function of the repository is to isolate the radionuclides present in the waste packages from the biosphere. That is why the risk assessment for this facility was carried out on the basis of the interpretation of observed activities in the local environment, especially the radionuclide content of the Sainte-Hlne river.

* In the case of the naval dockyards, the amounts of activity discharged were supplied by the Ministry of Defense for the period 1980-1997, but the nature of discharged radionuclides is only known since 1992, the date on which measurement results were computerized. In general, it was found that released activities are low compared to activities from others operators (especially COGEMA); these amounts were neglected in the risk calculations.

* For the 63 chemicals studied in the chemical assessment, the evaluation of the quantities discharged relied mainly on the results of measurements made by the operators between 1985-1989 and 2000. For periods during which they had not been measured (which was the case for the COGEMA's and ANDRA's sites), the quantities discharged were reconstructed for the main contributors based on the annual quantities of spent fuel reprocessed in the case of COGEMA or on the annual fluxes of run off water for ANDRA. Concerning, the naval dockyards, the information from the French Navy about their discharges is still under study. This step of the chemical assessment showed various limitations because some measurements are too general to be usable for the risk assessment (for example, the global measurement of the volatile organic compounds, VOCs), or because some elements were suspected to be in the releases but without any measurement (for example, polycyclic aromatic hydrocarbons, PAHs, in the release of COGEMA's boiler house). At the end of this step of the chemical assessment, a recommendation was made to acquire more precise information about the chemical releases of the plants. However, the GRNC decided to carry on the risk assessment with the existing data.

Available Data for Estimating Dose and Risk

The risk calculations performed by the GRNC were rather different according to radiological risk assessment or chemical risk assessment because of the difference in the data available for both cases.

For the radiological risk assessment, the individual excess leukemia risk can be quantified because data are available concerning the effects of radiation on the target organ for leukemia, which is the red bone marrow (RBM). In that case, the risk of radiation-induced leukemia is obtained from the exposure by multiplying it, first, by a dose coefficient that accounts for the dose delivered to the RBM and then, by the use of a risk model that assumes a no-threshold dose-risk relation and a 2-year latency period for leukemia. The dose coefficient depends on the individual's age, the exposure pathway, and the radionuclide considered, and the risk model used takes into account the variations of risk according to age at exposure and according to time since exposure. For the chemical risk assessment, the leukemia risk cannot be quantified because the existing toxicological values for individual chemicals for leukemia were not reported for the chemicals identified in the discharges of the local nuclear facilities. However, the GRNC decided to carry out the calculations and to quantify the whole cancer risk and the non-cancer risk for each chemical. For carcinogenic effects, it is assumed that the dose- response relationship is linear without threshold and the individual excess risk (ER) is calculated by the use of a reference toxicological value, named unit risk (UR), which is multiplied by the exposure dose.

The UR depends on the chemical and the exposure pathway. The influence of individual's age and time since exposure cannot be taken into account with the data currently available. For the non- carcinogenic effects, it is assumed that there is an effect only if the exposure dose exceeds the threshold defined by a toxicological reference value, which can be the reference dose (RfD) for ingestion or the reference concentration (RfC) for inhalation. The risk is then represented by an indicator, the hazard index (HI), which is the ratio between the exposure dose and the toxicological reference value. The toxicological reference value depends on the chemical and the exposure pathway. The influence of age is not taken into account.

Figure 4 illustrates the different concepts and relations between exposure, dose, and effect hidden behind the calculations performed for both risk assessments.

Figure 4. Schematic of exposure, dose, and effect.

The dose coefficients used to calculate the internal and external RBM doses from individual exposure to radiations (Eckerman and Ryman 1993; ICRP 1999) and the risk model used to estimated the number of radiation-induced leukemia cases (UNSCEAR 1994) came from the international literature. For each of the 81 radionuclides for which the releases were quantified (Table 1), the risk calculations were possible.

For the chemical assessment, there was no international consensus on the toxicological reference values to be used in risk assessment and a consensus had to be found within the GRNC. That is why this step needed much more effort than for the radiological assessment: first, to collect the toxicological data available in the literature and then, to make a choice. The main databases were examined by a panel of experts from the GRNC (OMS 2000; Health Canada 2001; TERA 2001). When several values were found, the panel chose the toxicological value for which the way of achievement is the most detailed and around which the uncertainty is the least. When several values matched these criteria, the most protective value was chosen. And finally, when no quantitative value was found for a known negative effect on health, the risk assessment calculations were performed up to the estimation of the exposure dose. The risk estimate for this effect will be calculated when a toxicological reference value becomes available.

The risk calculations were eventually possible for only 30 chemicals out of the 63 for which the releases were quantified (Table 1).

Exposure Assessment

To estimate individual exposure, it was necessary in both cases to reconstruct specific aspects of the population's lifestyle and to evaluate the transfer to humans of pollutants contained in the discharges of the local nuclear facilities.

Characterization of the population

Because the individual's risk of radiation-induced leukemia depends on all the doses received since birth, it was necessary for the radiological assessment to reconstruct the population who lived in the area where an excess of leukemia cases was observed and to reconstruct the way of life (dietary habits and time-budget) of this population. The population concerned, "the cohort," is constituted of young people aged from 0 to 24 years who have been living in the district of Beaumont-Hague between 1978 and 1996. The reconstruction of the way of life of the cohort was the subject of extensive and repeated discussions inside the GRNC due to the lack of data for the past. The objective was to determine, in the most realistic possible manner, the way of life of the average young person in the cohort.

For the chemical assessment, there was not the same time- constraint as for the radiological assessment. That is why the GRNC's calculations for the chemical risk relied on the most recent data to reflect actual and future levels of risk for scenarios representative of the local population. Three scenarios were studied: a reference scenario, S1, representing the average individual of the district of Beaumont-Hague (19 towns) and two scenarios, S2 and S3, which represent the individuals most exposed to atmospheric discharges and to marine discharges, respectively.

Finally, in both cases, the local data were used, but the chemical assessment relied on the most recent dietary surveys (Dufour 1998; CREDOC 1998), whereas the radiological assessment relied on the results of a previous food survey (Mathieu and Mathieu 1978) with assumptions for the non-available data, those \concerning the time-budget for the past. The food surveys results used were all based on the recording by the housewives of the consumption of the family during several days.

Estimate of environmental levels due to local nuclear facilities

Two complementary methodologies were considered to evaluate the transfer to humans of pollutants contained in the discharges of the local nuclear facilities:

* the first one is to measure levels in the environment and consequently to estimate the impact on the population, and

* the second consists of estimating the environmental levels, knowing the discharges, and making use of transfer models representing mechanisms for the dispersion and reconcentration of pollutants in the environment.

For the radiological assessment, the GRNC performed an exhaustive inventory of the radioactivity sampling and environmental measurements that had been carried out by all the stakeholders involved in the GRNC. This inventory includes over 500,000 pieces of data up to 1996, but despite the great numbers of data, measurements were not available for each radionuclide, each year, and each environmental compartment. That led the GRNC to use a transfer model (Figure 5) to fully reconstruct the contamination in the environment (air, sea water, sand, soil, and foodstuffs categories). Correction factors were introduced when necessary, to make the transfer model fit the existing environmental measurements.

Figure 5. The transfer model used for radiological and chemical risks assessments.

For the chemical assessment, because of the paucity of environmental measurements available, the GRNC relied totally on the transfer model already used for radioactivity transfer (Figure 5) (Rommens and Duffa 2003). For radioactive and chemical transfers, the same environmental compartments were taken into account, the same transfer coefficients were used for the elements that are common between radiological and chemical assessments (e.g., it was the case for the isotopes ^sup 57^Co, ^sup 58^Co, ^sup 60^Co and the element cobalt as a chemical pollutant), and the calculations were performed on an average annual basis for both assessments.

Risk Characterization

For the radiological assessment, the individual excess risk of radiation-induced leukemia associated with radiation from the local nuclear facilities was calculated for the average individual of the cohort and then the number of cases of radiationinduced leukemia in the cohort for the period 1978-96 was estimated by summing all the individual risks weighted by the number of young people in each birth cohort. This final number can be compared with the result of the cluster epidemiological study: four cases observed compared with two expected (Viel et al. 1995).

For the chemical assessment, no population risk was calculated. A set of results, related to the non-carcinogenic risk, is represented by hazard index (HI) for each chemical. If HI is less than 1, the chemical is not expected to present a health hazard. We chose not to sum all the HI or even part of them, because the illnesses involved by each chemical are different.

Another set of results, related to carcinogenic risk, is represented by individual excess risk of cancer (ER). It is the use to add the figures of ER for all the chemicals (USEPA 1986) and so we obtain only a figure of ER per scenario. These ER are compared with accepted risk values which lie between 10^sup -5^ and 10^sup - 6^, that is to say one case of cancer out of 100,000 or 1,000,000 persons.

The results for dioxins were presented separately because two different approaches to assess risk associated with dioxins are proposed in the literature. WHO (IPCS-WHO 2000) considers the dioxins as non-genotoxic and defines the reference dose between 1 and 4 pg TEQ.kg^sup -1^.day^sup -1 2^ whereas the USEPA (1998) considers a no-threshold dose-effect relationship and defines a unit risk of 5.10^sup -3^ per pg TEQ.kg^sup -1^.day^sup -1^. If we consider the acceptable risk value of 10^sup -5^, the equivalent reference dose for the USEPA unit risk is 2.10^sup -3^ pg TEQ.kg^sup -1^.day^sup -1^, that is to say a reference dose about 1,000 times lower than the one from the WHO.

RESULTS AND DISCUSSION

The radiological and the chemical assessments are based on many numerical data and produced a large number of results that are not presented here because this article focuses on health risk calculations. All the data and results of the GRNC can be consulted on the Internet at (http://www.irsn.fr/nord-cotentin/).

Figure 6. Contribution of each radionuclide (when above 1%) and each exposure pathway to the radiological risk.

Radiological Assessment

The number of cases of radiation-induced leukemia in the cohort attributable to discharges from the local nuclear facilities was estimated around 0.002. The contribution of each radionuclide and each exposure pathway is represented in Figure 6. The contribution of external exposure (irradiation by the plume, the seasprays, the soils, or the sand) is less than the internal exposure (ingestion of contaminated materials) with, respectively, 36 and 64%. The main radionuclides involved in the risk are ^sup 90^Sr, ^sup 106^Ru, ^sup 60^Co, ^sup 244^Cm, ^sup 85^Kr, and ^sup 137^Cs.

From this mean estimate, the probability of one or more cases occurring is low: probability of two per thousand for the onset of one case or more. It thus seems most improbable that the exposure due to radioactive discharges from local nuclear facilities might contribute significantly to explain the two cases in excess observed by the epidemiological study.

Chemical Assessment

In the following tables, the results are presented in two columns depending on the scenario taken into account for the incinerator. This comes from the decision of COGEMA, taken during the work of the GRNC, to stop the operation of its incinerator in 2002 whereas the initial scenario of exposure considered by the GRNC was in operation until 2030. The column "operation" considered an exposure of the population to discharges of the incinerator from 2000 to 2030 whereas the column "stop" considered an exposure of the population to discharges from the incinerator only for the period 2000 to mid- 2002. This led to noticeable changes only for the chemicals discharged mainly by the incinerator; that is, for arsenic, cadmium, chromium, copper, dioxins, hydrochloric acid, lead, manganese, and mercury.

For non-carcinogenic hazards (Table 2), all the HI values are below 1. The most important risks are obtained for scenario S2 by inhalation of nickel (0.7), SO^sub 2^ (0.3) and NO^sub x^ (0.06). By ingestion, the most important hazards are obtained for scenario S2 and for lead (0.23 if the incinerator operates until 2030, only 0.08 if it is stopped in 2002), nickel (0.04) and cadmium (0.03 if the incinerator operates until 2030, only 0.009 if it is stopped in 2002).

For carcinogenic risks (Table 3), the ER values corresponding to each scenario are less than 10^sup -5^, respectively; 8.10^sup -7^, 4.10^sup -6^, and 1.10^sup -6^ for S1, S2 and S3. The consideration of the operation of the incinerator, or not, after 2002 does not modify, significantly, the ER calculated because the main contributors to the cancer risk are chemicals discharged by the boiler house (nickel) and by the reprocessing plants (hydrazine).

For dioxins (Table 4), the results are contradictory according to the approach addressed. With WHO's approach, the risks seem acceptable because the His are below 1 for all scenarios, except for S2 when considering the operation of the incinerator until 2030. On the contrary, with USEPA's approach, all the ERs are above the acceptable risk of 10^sup -5^. The current knowledge does not allow the GRNC to choose which approach is more relevant but the GRNC recommended that dioxins in the environment required study by future measurement campaigns.

Concerning leukemia, the study performed by the GRNC did not allow the calculation of leukemia risk associated with chemicals. However, a qualitative analysis was made of the potential leukemogenicity of the chemicals discharged by the nuclear facilities and of the chemicals used by COGEMA. But, due to the paucity of data on environmental factors of causation of leukemia, the GRNC was not able to conclude whether chemicals discharged by the nuclear facilities can be involved in the observed leukemia cluster or not.

At the present time, summarizing risk assessments for both radioactive and chemical discharges from local nuclear facilities, we are still not able to explain the cluster of leukemia observed by the epidemiological studies.

Assessing Uncertainty

Assessing uncertainty involved different degrees of sophistication when addressing radiological assessment versus chemical assessment.

For the radiological assessment, the uncertainty analysis was realized during the two years after the deterministic calculations and was quantitative (Merle-Szeremeta et al. 2002). The parametric uncertainties were taken into account for more than 200 parameters of the transfer and exposure models, by the use of probabilistic distributions of values that were defined by expert panels of the GRNC (Rommens et al. 2002). Then, different methods of uncertainty propagation were used that led to several possible intervals for the leukemia risk results. With the classical method of uncertainties propagation, based on the theory of probabilities, the GRNC obtained an upper value for the leukemia risk that is a factor of three above the deterministic calculation. With an alternative method, based on fuzzy logic, this factor becomes five. These calculations do not refute the initial conclusion that it seems improbable that the exposure due to radioactive discharges from local nuclear facilities might contribute significantly to explain the two excess cases observed by the epidemiological study.

Table 2. Hazard index calculated fo\r non-carcinogenic chemical hazards for each of three scenarios.

Table 3. Excess risks calculated for carcinogenic chemical risks for each of three scenarios.

For the chemical assessment, the study of uncertainties was realized in parallel with the deterministic calculations, but was limited to an uncertainty characterization; that is, a qualitative discussion of the thought processes that lead to the selection or rejection of specific data, hypotheses, or scenarios.

Table 4. Risks calculated for the dioxins.

CONCLUSION

The main mission of the GRNC was to assess the overall risk for the populations related to the operation of the local nuclear facilities in order to provide information to compare with the results of local epidemiological studies.

This experience showed that the same scientific approach can be used for radiological and chemical risk assessments. The methodology in four steps is the same in both assessments and the approach of the environmental transfers of radionuclides and chemicals is very similar.

However, differences appear in the application of each step of the methodology. First, the content of the hazard identification step is quite different; for the radiological assessment, the objective is to identify and quantify, in the most exhaustive and precise manner, the radionuclides discharged by the nuclear facilities because each radionuclide according to linear no threshold model (LNT) is considered to contribute to the adverse health effect. For the chemical assessment, this step also implies an inventory of the chemicals but there is, in addition, a selection process over the chemicals because for some chemicals there is no demonstrated adverse health effect or because the available data for many chemicals are too scarce to reasonably engage any health risk calculations. Another difference between both assessments concerns the indicators of the risks calculated. In the case of the radiological assessment, the study protocol was set up to consider the same population and the same period as those of the epidemiological studies, and the results about radiation-induced leukemia risk can usefully be compared with the epidemiological findings. For the chemical assessment, the paucity of knowledge about causation of leukemia prevents the GRNC from the same protocol and the GRNC merely assesses the whole-cancer risk and the non- cancer hazards.

Experiences where radiological and chemical risks arc assessed for the same industrial site were already carried out and especially in United States (RAC 1999a,b,c) but for France, the GRNC is the first of that type. In addition to the technical challenge, the constitution of the GRNC, which comprises individuals from the various stakeholders and the audience in the nuclear French community, was also a social challenge. The GRNC experience has illustrated the advantages of an open exercise, even if the interpretation of the calculations figures can finally differ from one stakeholder to another.

ACKNOWLEDGMENTS

The authors thank all members of the GRNC for their contribution in these collective works and especially the animators of the thematic working groups, which where particularly involved in the assessment process and people that performed the computer calculations: Caroline Ringeard and Dominique Laurier from the French Institute for Radioprotection and Nuclear Safety, Jean- Pierre Degrange from the CEPN (Centre d'Etudes sur l'valuation de la Protection dans le domaine Nuclaire) and Laure Dlry and Eric Thybaud from the French Institute of Industrial Environment and Risks.

1 The observed and expected (O/E) ratios of childhood leukemia successively published for the canton of Beaumont-Hague were: 1978- 1992: O/E = 2.8 [0.8; 7.3], 1978-1996: O/E = 1.9 [0.5; 4.81, 1978- 1998: O/E = 2.2 [0.7; 5.1].

2 The toxicity of a mixture of dioxins is commonly expressed as a toxic equivalent (TEQ). This is a figure obtained by multiplying the concentration of individual dioxin species by a suitable toxic equivalent factor and summing the results.

REFERENCES

CREDOC. 1998. Les activits d'extrieur et d'intrieur dans le Nord- Cotentin. Rapport ralis la demande de COGEMA. Centre de Recherche pour l'Etude et l'Observation des Conditions de vie, Paris, France

Dufour A. 1998. Enqute sur la consommation alimentaire dans le Nord-Cotentin. Rapport ralis la demande de COGEMA. Centre de Recherche pour l'Etude et l'Observation des Conditions de vie, Paris, France

Eckerman KF and Ryman JC. 1993. External Exposure of Radionuclides in Air, Water and Soil. EPA Report 402-R-93-081. Federal Guidance Report 12. U.S. Environmental Protection Agency Washington DC, USA

Groupe Radiocologie Nord-Cotentin. 1999. Synthesis: Estimation of Exposure Levels to Ionizing Radiation and Associated Risks of Leukemia for Populations in the Nord-Cotentin. Institut de Radioprotection et de Sret Nuclaire, Fontenay-aux-Roses, France

Guizard AV, Spira A, Troussard X, et al. 1997. Incidence des leucmies de 0 24 ans dans le Nord Cotentin. Rev Epidemiol Sante Publique 45:530-5

Guizard AV, Boutou O, Pottier D, et al. 2001. The incidence of childhood leukaemia around the La Hague nuclear waste reprocessing plant (France: a survey for the years 1978-1998. J Epidemiology Community Health 55:469-74

Health Canada-site du ministre canadien de la sant. 2001. Available at [left angle bracket]http://www.hcsc.gc.ca/[right angle bracket]

International Commission for Radiation Protection (ICRP). 1999. The ICRP Database of Dose Coefficients: Workers and Members of the Public. CD Rom v 1.0. Elsevier Science, Stockholm, Sweden

IPCS-WHO. 2000. Environmental Health Criteria-Dioxines. World Health Organisation, International Programme on Chemical Safety. Available at [left angle bracket]htpp://www.inchem.org/ fullist.htm[right angle bracket]

Laurier D and Bard D. 1999. Epidemiologie studies of leukaemia among persons under 25 years old living near nuclear sites. Epidemiologic Reviews 21:188-206

Laurier D, Rommens C, Merle-Szeremeta A, et al. 2000. Evaluation du risque de leucmies radioinduit proximit d'une installation nuclaire: l'tude radiocologique Nord-Cotentin. Revue d'pidmiologie et de sant publique 48:34-6

Mathieu PY and Mathieu C. 1978. Etude du Rgime Alimentaire des Habitants de la Rgion de Flamanville. Etude Faite la Demande de la Section de Radiocologie du Centre d'Etudes Nuclaire de Cadarache. Bureau d'tude de gographie et d'cologie applique, Saint-Paul-lez- Durance, France

Merle-Szeremeta A, BrenotJ, Chojnacki E, et al. 2002. Nord- Cotentin radioecological study. In Forum Eurosafe 2002, November 4- 5, 2002, Germany

NRC (National Research Council). 1983. Committee on the Institutional Means for Assessment of Risks to Public Health. Risk Assessment in the Federal Government: Managing the Process. National Academy Press, Washington DC, USA

OMS. 2000. Valeurs guides pour la qualit de l'air. Available at [left angle bracket]http://www.who.dk/document/e71922.pdf[right angle bracket]

Pobel D and Viel JF. 1997. Case-control study of leukaemia among young people near La Hague nuclear reprocessing plant: the environmental hypothesis revisited. BMJ 1997 314:101-6

RAC (Radiological Assessment Corporation). 1999a. Technical Summary Report for the Historical Public Exposures Studies for Rocky Flats Phase II-Final Report, Neeses, USA

RAC (Radiological Assessment Corporation). 1999b. The Faisability of Performing a Chemical Dose Reconstruction Study at the Idaho National Laboratory (INEEL)-Final Report, Neeses, USA

RAC (Radiological Assessment Corporation). 1999c. Savannah River Site (SRS) Environmental Dose Reconstruction Project-Final Report, Neeses, USA

Rommens C and Duffa C. 2003. The Nord-Cotentin chemical risk assessment. Seventh International Conference on the Biogeochemistry of Trace Elements (ICOBTE), Upsala, Sweden. June

Rommens C, Laurier D, and Sugier A. 2000. Methodology and results of the Nord-Cotentin radioecological study.J Radiological Protection 20(4):361-80

Rommens C, Renaud P, Vray F, et al. 2002. Uncertainties of the transfer of radionuclides in terrestrial ecosystems. In Radioprotection. Actes du congrs ECORAD, September 3-7, 2001, Aix- en-Provence, France, 37: C1-1217/C1-1222

TERA (Toxicology Excellence for Risk Assessment). 2001. Available at [left angle bracket]http://www.tera.org/iter/[right angle bracket]

United Nations Scientific Committee on the Effects of Atomic Radiation. Sources and Effects of Ionising Radiation. 1994. United Nations (UNSCEAR 1994 Report to the General Assembly, with scientific annexes), New York City, NY, USA

USEPA (US Environmental Protection Agency). 1986. Guidelines for the health risk assessment of chemical mixtures. Fed Reg 51 (185):34014-25

USEPA (US Environmental Protection Agency). 1998. Exposure and Human Health Reassessment of 2,3,7,8-tetrechlorodibenzo-p-dioxin (TCDD) and Related Compounds. Part III: Integrated Summary and Risk Characterization for 2,3,7,8-Tetrechlorodibenzo-p-dioxin (TCDD) and Related Compounds. External Review Draft EPA/600/P-00/001Ag. NCEAORD, Washington, DC, USA

Viel JF, Pobel D, and Carr A. 1995. Incidence of leukaemia in young people around the La Hague nuclear waste reprocessing plant: a sensitivity analysis. Stat Med 14:2459-72

Catherine Mercat-Rommens, Didier Louvat, Cline Duffa, and Annie Sugier

Institute for Radioprotection and Nuclear Safety, IRSN, Bat 153, 13108 St Paul-Lez-Durance, Cedex, France

Received 22 April 2004; revised manuscript accepted 31 August 2004.

Address correspondence to Catherine Mercat-Rommens, Institute for Radioprotection and Nuclear Safety, IRSN, Bat 153, 13108 St Paul-Le/ -Durance, Cedex, France. E-mail: Catherine. mercat-rommens@irsn.fr

Copyright CRC Press Jun 2005

Source: Human and Ecological Risk Assessment

More News in this Category