3. Appendix notes on intervention and derived intervention levels in relation to food and agriculture

Contents - Previous - Next

3.1 Intervention levels
3.2 Derived intervention levels for food moving internationally
3.3. Derived intervention levels as a function of time and food processing

3.1 Intervention levels

In the event of a nuclear accident the limitation of human exposure to radiation may only be achievable by "intervention" (44). Moreover, as demonstrated by Chernobyl this will probably involve constraints on, or changes in, agriculture, forestry and fisheries practices and upon livestock movement, product harvest, distribution, processing, and trade with implications for the dependent communities.

A post-accident intervention level of dose commitment for the public has to be translated into the corresponding and measurable levels of ground deposit (Bq m^-2) of the significant radionuclides present, and into the related levels in food and drink as consumed (Bq Kg^-1 or Bq 1^-1). This then provides for the application of derived intervention levels (DIL's) in the same units. These are the relatively easily monitored levels in food or environment, below which intervention would not be justified under the circumstances or, above which intervention would be indicated on the grounds of public health.

ICRP has tabulated radiation dose commitment conversion factors in terms of Sv Bq^-1, in consumed food and drink (204) in connexion with occupational exposure. The same approach is used and extended in the development of guidelines for intervention levels in post-accident situations by IAEA (81). A few dose per unit intake factors @#$ are illustrated in Table I below. The relatively higher H_eff for the 'critical' group of infants should be noted. While not within the scope of this review, ICRP (54; 205) has quantified the non-stochastic and stochastic relationships between health effects and radiation dose.

TABLE I
Conversion factors for committed effective dose equivalent per unit radionuclide intake (from ref. 81, p. 55) - nSv.Bq^-1 = 10-9Sv/Bq^-1

The same IAEA publication (81) includes guidelines on terminology, and on the relationships between intervention levels of radiation dose (ILs), and derived intervention levels (DIL's). In its general form a derived intervention level was defined (81, pp. 18 et seq):

DIL = IL/DCF

where IL is the intervention level of radiation dose commitment for the exposed person and DCF is the dose conversion factor, both for the particular situation, e.g., radiation exposure from the ingestion of contaminated food; or for external radiation exposure from a ground deposit (see ref. 81, pp. 2325).

While a consensus has emerged on the guidelines for establishing "lower-tier" and "upper-tier" ILs and DILs (e.g., see refs. 44; 205; 100; 81) those actually adopted will depend on the local or national situation after the accident or emergency as explained below (Section 3.2).

FAO, meanwhile, adopted a lower IL of total body dose commitment of 5 mSv for the first post-accident year for the purpose of lower-tier DILs for food moving in international trade (181; see Section 3.2), and a dose commitment of 1 mSv for subsequent post-accident years. Taking into account the "weighting factors" (54), first and subsequent post-accident-year dose commitments of 50 mSv and 10 mSv were respectively adopted for radionuclides affecting specific organs.

Chernobyl has also demonstrated the need for improved international agreement, not only on terminology and guidelines, but also on specified post-accident IL's (206) as a basis for DILs for food and feed moving internationally. FAO's initiative in this context was timely. If wide agreement could not be achieved, e.g., in relatively nuclear-power free areas such as Australia (207), then as wide agreement as possible might usefully be sought, e.g., in food trading zones which include significant nuclear power programmes and countries highly dependent on food imports such as Japan (208).

3.2 Derived intervention levels for food moving internationally

A major problem after "Chernobyl" were the disparate "action levels" adopted by different countries for the acceptability of food for import or domestic consumption For example, for cesium-137 in milk such levels ranged from 50 to 4,600 Bq 1^-1 (181).

Such disparities may have arisen, understandably, as a result of differing interpretations of two existing international guidelines: The "ALARA" principle (see ref. 121; Section IV), and "The establisment of............ intervention levels for particular circumstances is the responsibility of the competent national authorities" (34; 54; 44; 205).
Unfortunately, the detrimental consequences in terms of constraints on food harvest, movement (including that of livestock feed), international trade, and upon the welfare of the dependent communities had been underestimated or unforeseen.

FAO took a timely step towards a solution to this problem. This involved the elaboration and recommendation for future international recognition of "Interim International Radionuclide Action Levels for Food" (i.e., for food moving in international trade). The acronym "IRALFs" for these levels was proposed (181).

IRALF's were designed to provide for wide margins of post-accident radiological health safety for importing communities, for minimizing constraints in future, and for improved protection of the welfare of agriculture and fisheries communities against anomalous trade barriers. The levels can be applied to single radionuclides or mixtures (181, p. 11) and would lend themselves to simple monitoring, certification, and application. The levels represent those below which export or import restrictions would not be technically justified, taking into account "economic and social considerations" in the international context.

As implied by their definition, IRALF values provided for periodical international review after some accidental release, a provision already adopted for EC countries (85).

3.3. Derived intervention levels as a function of time and food processing

IAEA has addressed the problems of derived intervention levels (DIL's) for radionuclides in foods (81). The basis is comparable with that adopted for 'IRALF' levels (above) by FAO but refined to take into account changes with time in the radionuclide concentration in the foodstuff (and, therefore, the time of DIL application) during the year of intake. Thus, "on the assumption that only one nuclide and one foodstuff contributes to the diet", the DIL, in its simpler form, was defined as follows:

DIL = EDL/I.DCF.G Eq. 1

where: DIL (Bq kg^-1) is the intervention level to be applied at time t, EDL (Sv) represents the effective intervention level of dose and will represent that fraction of the total committed radiation dose equivalent allowed for the intake of food in the one year, e.g., allowing for possible additional sources of radiation exposure. I (kg yr^-1) is the rate of intake of foodstuff as consumed. DCF (Sv Bq^-1) is the appropriate radiation dose conversion factor for that radionuclide. G (yr) is the ratio of "the integral over the one year of the nuclide concentration in that foodstuff" (Bq-yr kg^-1) to the concentration at time t (Bq kg^-1).

Any realistic advantage in this refined approach will depend upon some foreknowledge or model of the radionuclide in the foodstuff as a function of "real" time. However, given that condition and the same basic assumptions above, the approach can be illustrated as follows:

Let the intervention level of dose for food restriction (EDL) selected by the appropriate authority for the post-accident year under consideration be L (Sv). Suppose the radionuclide concentration C (Bq kg-l) is some function of time t (yr) such that C = f(t) during the year under consideration (i.e., 0 < t < 1). Further, suppose that the C - t curve during the year is of the developing (i.e., as observed in "real" time) or predicted form illustrated in Fig. 1 below:

Fig. 1. Hypothetical concentration - time curve for radionuclide in foodstuff over one year period.

Such a curve might, for example, represent cesium-137 in the dried milk powder from dairy herds initially grazing on pasture following a "spike' deposition followed by a rise on transfer of the herd to winter shelter and feed which had been harvested at an earlier post-fallout state. The initial decline will be due to the short "effective" halflife of cesium in the cow following an initially high intake of recent deposit while the later rise and near-horizontal part of the curve will reflect the 'steady-state' situation as a result of the more constant feed-radionuclide intake when the decline in the feed will be the result of radioactive decay alone with a halflife of 30 yr.

It follows that the equivalent dose committed during the year will be W.DCF.C_m where W will be the total weight of foodstuff (kg) effectively taken in by the average exposed member of the population. C_m will be the time-weighted mean concentration of radionuclide (Bq kg^-1) in the foodstuff as consumed during the same year, i.e., A/t where A is the shaded area enclosed by the C - t curve of Fig. 1 or, formally,

C_m = 1/t x ò f(t).dt where t = 1 (yr).

Now suppose that DIL_t, is the DIL for application at time t',

i.e., when W.DCF.C_m = L. Let DIL_t' = C_m/G'

Then, by definition (above), DIL_t, = L/W.DCF.G' which represents a specific illustration of equation l (W = I/t where t = 1).

Likewise DIL_t' = L/W.DCF.G'' for the DIL applied at time t'' for the same limiting value of EDL. Note that in the illustration of Fig. 1, G will exceed unity for DIL_t' but will be less than unity for DIL_t''

It may also be noted that the ratio G will be unaffected by the actual value of Cm because, for a particular model, "the relationship between the derived quantity and the primary intervention level of dose is linear" (ref. 81, p. 4).

In developing "interim" DIL's for food moving in international trade (i.e., IRALFs) FAO (181) has ignored changes in C during a year of export or import since the IRALF would be applied at the particular time of export and import and the foodstuff would be subject to certified monitoring. Moreover, once imported, changes as a result of significantly short halflives are taken into account. This approach has the advantages of simplicity and conservative caution since under these conditions C can only decline with t after the importation. However, equation 1 might be useful at exporter level as an indicator as to when an agricultural product would become internationally acceptable for export.

Equation 1 (above) can be modified to take into account the effects of applying the DIL at different stages of food processing from the point of harvest or collection (e.g., milk) as follows (81, p. 31):

DIL = EDL.F / I.DCF.G Eq. 2

The new modifying factor F is the ratio of radionuclide concentration in the food or drink at the point of DIL application (and implied sampling) to that level in the food or drink as ready for consumption. Thus,

F will exceed or equal unity. For example, if the DIL were applied as the "IRALF" to imported milk powder which would only be consumed as reconstituted milk (with uncontaminated water) then F would have a value of approx. 6.5 because milk contains approx. 15 % dry weight of solids (Dr. J.G. Davis personal communication). Likewise, the values of F would be of the orders of the reciprocals of the fractions illustrated in Table IV of Section 1.2.4 which quantify the effects of processing on fruit and vegetables, e.g., if the DIL were to be applied to "internal" leaves of freshly-harvested green vegetables or flesh of fruits, before processing, and for the consumption of the processed food.

Contents - Previous - Next