Recommendations on Dose Coefficients for Assessing Doses From Accidental Radionuclide Releases to the Environment

Exposure Pathways

Internal Exposure from Inhalation of Radionuclides in the Plume

Individuals immersed in a radioactive plume can receive internal exposures following inhalation of airborne radioactive material. Inhaled radionuclides irradiate the tissues of the respiratory system, as well as those of other organs. The committed effective dose to an individual is determined by numerous physical, chemical, and biological factors, including the amount and type of material inhaled, its deposition and retention in the respiratory system, and the breathing rate of the individual.

In 1994, the ICRP published its new respiratory tract model, which incorporates many of the factors affecting the dose assessment of inhaled radioactivity (ICRP 1994). One of the major changes in this model is the replacement of the previously used average lung dose with the explicit calculation of dose to each of five separate respiratory regions. The scope of the model has also been extended to apply explicitly to all members of the population. Reference values for age-dependent parameters have been given for six age groups covering infants to adults. The ICRP has used this model to derive inhalation dose coefficients for a number of radionuclides.

Human Physiological Parameters

One of the features of the new respiratory tract model is the incorporation of respiratory physiology, which affects the rates and volumes of air inhaled and exhaled, and determines the amount of radioactive material deposited in, and cleared from, the respiratory tract (ICRP 1994). To account for age-dependency, the ICRP has defined a number of age-dependent parameters, including reference respiratory values such as ventilation, or breathing, rates for 3-month-old infants, 1-, 5-, 10- and 15-year-old children, and adults.

The Working Group compared the ICRP default breathing rates (ICRP 1994, 1995b) with those recommended by a Health Canada working group on reference values (Health Canada 1993). ICRP values are based on published respiratory data and surveys of average times spent at several activity levels (sleep, sitting, light exercise, heavy exercise) for various ages. Health Canada values were derived from the breathing rates specified for ICRP Reference Man (ICRP 1975) at two activity levels, with some adjustment to account for size differences in the Canadian population. Recommended age-adjusted values were then given for five age groups. The Working Group initially considered the Health Canada values as these attempted to represent Canadian populations. However, since these values are based on ICRP Reference Man rather than actual population surveys, and are given for age groups different from the current ICRP classifications, it was concluded that ICRP breathing rates were the most relevant and consistent. Recommended age groups and default breathing rates are reproduced in Table 1 (ICRP 1995b).

Inhalation Dose Coefficients

Age-dependent committed effective dose coefficients for inhalation by members of the public have been calculated using the new respiratory tract model and are compiled in ICRP Publication 72 (1996) for radioisotopes of 91 elements. Calculations for the public for inhaled particles are based on an activity median aerodynamic diameter of 1 µm, using the default breathing rates listed in Table 1 . Age-specific biokinetic models describing tissue distribution, retention, and excretion of systemic activity are used for radioisotopes of 31 of these elements. For radioisotopes of the remaining 60 elements, biokinetic models are based on those given in ICRP Publication 30, Parts 1-4 (1979, 1980, 1981, 1988). Allowances were made for age-specific changes in gut uptake, body mass and geometry, and urinary excretion rates, but not in the biokinetics of systemic activity. Tissue and radiation weighting factors used to calculate effective dose coefficients are those given by ICRP in its 1990 recommendations (ICRP 1991).

The Working Group recommends that the dose coefficients for inhalation compiled in ICRP Publication 72 be used as the standard for assessing inhalation doses, as they represent the latest internationally recognised values. These dose coefficients have also been adopted in the International Basic Safety Series (IAEA 1996) and in the Euratom Directive (EC 1996). It is also recommended that age-dependent doses be explicitly calculated for the six ICRP age groups, using the age-specific breathing rates of Table 1 as default values.

Inhalation dose coefficients for radionuclides of potential importance following a radiological emergency are given in Table 2 , based on the default lung absorption types (the rate of absorption from the respiratory tract to body fluids) recommended by the ICRP for those situations when no specific information is available on the chemical form of the radionuclide. The ICRP has developed a quality-assured database on CD-ROM of inhalation dose coefficients for the public and for workers for a range of particle sizes and integration times, as well as ingestion coefficients for various gut absorption values (ICRP 1998, in press). This database, or ICRP Publication 72, should be referred to when assessing the doses from radionuclides not listed in Table 2 , or when a different lung absorption factor is assumed.

External Exposure from Radionuclides in the Plum e and on the Ground

Reference conversion coefficients incorporating the latest ICRP recommendations have recently been compiled in ICRP Publication 74 (1996a) for irradiation by monoenergetic radiation. These coefficients are expressed as the effective dose per unit air kerma ¹ for several parallel and non-parallel monoenergetic radiation field geometries applicable to occupational radiation exposure. The ICRP has not compiled radionuclide-specific effective dose coefficients for external irradiation from radionuclides distributed in the environment, with the exception of dose rate coefficients for the exposure of adults to inert gases (ICRP 1996) ² .

Since, in practice, radiation fields resulting from environmental radionuclide contamination are not monoenergetic, mean or effective conversion coefficients must be determined by integration over the entire radiation energy spectrum, and summed for all types of radiation present. Differences in the irradiation geometry can also be expected between uniform radiation fields, and radiation sources distributed in the environment. Finally, the units for conversion coefficients used in environmental dose assessments (effective dose per unit time-integrated exposure to a radionuclide) are not easily derived from effective dose per unit air kerma. Based on these considerations, the Working Group reviewed existing compilations of external dose coefficients that incorporated the recommendations and changes of ICRP Publication 60.

Two compilations of dose coefficients for cloudshine and groundshine were considered by the Working Group on account of their completeness, applicability, and availability, namely those of Macdonald and Laverlock (1996), and Eckerman and Leggett (1996), the latter which is an extension of the work of Eckerman and Ryman (1993). Both compilations are based on Monte Carlo modelling of the doses in the organs of a mathematical phantom, resulting from photons and electrons emitted by radionuclides distributed in air, water, soil, and on the ground surface.

Dose Coefficients from Macdonald and Laverlock (1996)

The data set of Macdonald and Laverlock (1996) is a revision to the work of Holford (1988, 1989). Holford calculated dose coefficients for the 24 organ systems of Reference Man (ICRP 1975) using the EDEFIS code of Barnard and D'Arcy (1986), and employing ICRP Publication 26 tissue weighting factors (ICRP 1977), but also including the later recommendation of a tissue weighting factor of 0.01 for the skin. Seven exposure scenarios were considered, including the three pathways of interest to the Working Group. Details of the computational method are given in Barnard and D'Arcy (1986). In brief,

• the absorbed dose rate to an element of the medium at the point of interest was calculated, and converted to a dose rate in an element of tissue-equivalent material at the body surface;
• this dose rate was multiplied by organ dose ratios computed by Monte Carlo simulations of the interaction of photons with a mathematical model of standard man, to determine the dose rates to individual organs. For electrons, only the skin dose was calculated;
• the effective dose equivalent was estimated from the individual organ dose coefficients.

The external dose from immersion in air was calculated for Reference Man assuming a centroid at 1 m from a surface in a semi-infinite volume, and an air density of 1.189 kg m-3. Exposure from a contaminated ground surface was calculated assuming a two-dimensional distribution of the radionuclide on the surface, with a disc-shaped receptor suspended in parallel, 1.6 m from the surface.

In Macdonald and Laverlock's revision, effective dose coefficients were recalculated for external exposures by applying the new ICRP tissue weighting factors to the organ dose coefficients calculated by Holford (1989). A minor deviation was made from the ICRP (1991) guidance on the application of the tissue weighting factor for the remainder tissues under certain cases. Specifically, the ICRP weighting of the remainder fraction for exposure scenarios where one of the remainder organs receives an equivalent dose greater than in any of the principal organs was not used, based on the judgement that this type of dose distribution does not occur for external exposures.

Dose Coefficients from Eckerman and Legett (1996)

Eckerman and Legett (1996), and Eckerman and Ryman (1993) calculated coefficients for effective dose equivalent and effective dose for external exposures based respectively on ICRP Publications 26 and 60 tissue weighting specifications (ICRP 1977, 1991). Values conforming to ICRP Publication 26 appeared in Federal Guidance Report 12 of the U.S. Environmental Protection Agency (Eckerman and Ryman 1993), for use in radiation protection programs in the United States. Although not appearing in the EPA guidance, effective dose coefficients based on ICRP 60 methodologies were calculated using the same modelling techniques, and have been made available in the computer software package DCFPAK: Dose Coefficient Data File Package (Eckerman and Leggett 1996). The Working Group obtained this compilation and accompanying documentation directly from the authors.

Details of the calculational methods can be found in Eckerman and Ryman (1993), of which the three major steps were:

• computation of the distributions of the radiations incident on the body for a number of initial energies of monoenergetic sources distributed in environmental media of interest;
• evaluation of the transport and energy deposition in organs and tissues of the body of the incident radiations by Monte Carlo methods, for each of the initial energies considered; and
• calculation of the organ or tissue dose for specific radionuclides, considering the energies and intensities of the radiations emitted during nuclear transformations of those nuclides.

For photons, organ doses were computed at each of 12 monoenergetic photon energies for 25 organs in an adult hermaphrodite phantom (Christy and Eckerman 1987), modified to include the oesophagus, and to improve the modelling of the neck and thyroid. For electrons, values were tabulated for skin only. The computational methods were chosen to give an accurate characterisation of the energy and angular dependence of the radiation field incident on the body. The contribution from bremsstrahlung was also included for all exposure modes.

For each mono-energetic photon energy, coefficients for immersion in air were calculated assuming a semi-infinite cloud source containing a uniformly-distributed mono-energetic photon emitter surrounding a human phantom standing on the soil at the air-ground interface, under conditions of 40% relative humidity, a pressure of 760 mm Hg, air temperature of 20ºC, and a density of 1.2 kg m-3. Groundshine coefficients for a contaminated surface are based on an infinite isotropic source of monoenergetic photons, located at the air-ground interface, with a standing human phantom at the interface.

Comparison of the Macdonald - Laverlock and Eckerman - Leggett Data Sets

The two sets of external dose coefficients were compared in terms of both the calculated values for a number of radionuclides, and the quality of the models from which the values were derived. The models were compared based on the types of mathematical phantoms used, the types of processes considered (eg. bremsstrahlung), the material referenced by the authors, and the type and completeness of the models used.

Ratios of cloudshine and groundshine dose coefficients from Macdonald and Laverlock (DCML ), and Eckerman and Legett (DC EL ) were calculated for 10 8 radionuclides, including those listed in CSA guideline documents N288.1 and N288.2 (CSA 1987, 1991) as being of potential radiological importance under normal and accidental situations, as well as their decay products. Radionuclides for which either source gave a value of zero were excluded from this comparison, although they were considered in the general analysis of the two data sets.

Figure 1 shows the frequency of occurrence of the various values of DCEL :DCML , which are predominantly in the range of 1.00-1.25 for both exposure pathways, indicating a generally good agreement between the two data sets, although the values of DCEL tend to be greater than those of DCML . From Figure 2, the best agreement occurs for those radionuclides which contribute the greatest effective dose per unit activity concentration, and are therefore of most importance in radiological assessments. Differences between the two data sources generally increase as the dose coefficient decreases. In some cases, Macdonald and Laverlock's values were significantly greater than those of Eckerman and Legett, particularly for low-energy beta emitters for which the dose per unit activity concentration is small. Such differences are expected due to the previously discussed differences in the models.

Figure 1: Frequency distributions of DC_EL :DC_ML for cloudshine and groundshine dose coefficients

Figure 2: Ratio of DC_EL :DC_ML versus DC_EL

Eckerman and Ryman (1993) compared their data with other published results, and in general, agreement was good. Discrepancies were traced to differences in both the types of phantoms used, and certain features of the computational models. As a final check of the Eckerman and Legett data set, their dose coefficients based on ICRP Publication 26 (1977) were compared with those of Kocher (1983). Kocher's values were previously recommended for use in the Canadian Standards Association's N288 series of guidelines (CSA 1987, 1991). The two sets were in reasonable agreement, with the values of Eckerman and Legett typically 10-20% greater than those of Kocher, again due to the greater sophistication of their model.

Based on these considerations, the Working Group concluded that the dose coefficients of Eckerman and Legett (1996) represent the best values available to date for cloudshine and groundshine assessments. External dose coefficients for those radionuclides that might be of radiological importance following a nuclear accident are given in Table 2 . Appendix C contains an extensive list of external dose coefficients reproduced from Eckerman and Legett's DCFPAK software, and should be referred to for dose coefficients of radionuclides not listed in Table 2.

Effect of Gender and Age on External Dose Coefficients

Organ equivalent dose coefficients and effective dose coefficients given by Eckerman and Legett were calculated for an hermaphrodite anthropomorphic adult phantom derived by Cristy (Cristy and Eckerman 1987) from ICRP Reference Man (ICRP 1975). Doses to individuals of different size and gender can be expected to be somewhat different due to differences in radiation transport through the body. In light of this, the Working Group reviewed the recommendation contained in the Canadian Standards Association's 1987 guidelines for derived release limit calculations (CSA 1987) to increase the values of the adult external dose coefficients by a factor of 1.5 when applied to infants.

Gender-specific and age-specific aspects of external dose have been investigated by several researchers, including Drexler et al (1989), Petoussi et al (1991), Yamaguchi (1994), and Schultz and Zoetelief (1997). Although dose coefficients in these studies are frequently given as air kerma to effective dose equivalent (or effective dose) conversion coefficients for monoenergetic photon radiation fields, the conclusions on gender and age dependency are relevant. These studies show that, typically, the dose to individual body organs increases with decreasing body size. This effect is more pronounced at low photon energies, and for organs located near the middle of the body, which are shielded by overlying tissues. This also applies to the differences arising from the use of hermaphroditic versus sex-specific phantoms, as discussed by Eckerman and Ryman (1993).

Petoussi et al (1991) have indicated that infant organ doses for cloudshine and groundshine may be as much as about 40% greater than those in the adult male at photon energies greater than 100 keV. Below 100 keV, the difference may approach a factor of 3 for deeper-seated organs. Yamaguchi (1994) calculated coefficients for anthropomorphic phantoms using the six ICRP age groups under 5 irradiation geometries. For isotropic radiation fields, effective dose coefficients for 0- and 1-year-old infants normalised to kerma in free air were about 20-30% higher than those for adults at energies above 115 keV. For lower energies, differences between infant and adult values approached factors of 3 to 4. From Schultz and Zoetelief (1997), ratios of child to adult effective dose coefficients for electrons ranged from about 2- to 20-fold for energies above 600 keV. The relative differences were less than 0.2% only for energies below 600 keV.

In light of these results, the modifying factor of 1.5 as recommended by the CSA (1987) is very conservative for some radionuclides and age groups. However, due to the complexity of deriving radionuclide- and age-specific corrections, the Working Group recommends that this factor be retained as a default to the external dose coefficients for the two youngest ICRP age groups (i.e. 3-month- and 1-year-old infants). Adult values should be used for the remaining four age groups. It is also recommended that if the identity of the nuclides and radiation spectra exposing the critical group are known, then a more appropriate factor may be calculated. The Working Group has chosen not to recommended a correction factor based on gender.

Other Modifying Factors

The Working Group considered the issue of modifying factors, such as shielding and exposure factors, and finite cloud corrections. It was decided that these were beyond the scope of the present report. More information on this topic can be found elsewhere, such as Kocher (1983), and Eckerman and Ryman (1993).

---