THE ENVIRONMENTAL CHARACTERISATION OF PARTICULATE ASSOCIATED RADIOACTIVITY DEPOSITED CLOSE TO THE SELLAFIELD NUCLEAR REPROCESSING PLANT, CUMBRIA UK.

The Sellafield works, formerly known as Windscale has routinely discharged low level radioactive material to the atmosphere for over 45 years, (McCarthy and Nicholls 1990 ). Since operations began, the site has processed two types of nuclear fuel, uranium metal and ceramic uranium oxide. Spent uranium fuel from the civilian Magnox division has been processed from 1956 to the present day. The nuclear fuel reprocessing operation separates fission and actinide products from each other and also separates uranium from plutonium. In the early days of Windscale's operations, weapons grade plutonium was produced from the irradiation of uranium metal from the Windscale Pile reactors from 1950 up until 1957 when operations ceased due to a fire in Pile No. 1.

The second type of nuclear fuel to be processed at Sellafield was ceramic uranium oxide from the experimental Advanced Gas-Cooled Reactor (AGR) and this fuel was processed from 1969-73 and the reactor ceased operations in 1982, Jones S.R. ( pers. comm ). This type of fuel is currently processed at the new Thorp facility.

Throughout the years the site has expanded as new plants and processes have come on-stream. This has led to a complicated atmospheric /ground deposition pattern from both within the site and to the local environment. Over this period of time, operational and acute (short term accidental releases) of either ground level or atmospheric discharges of radioactive material have led to an inventory of nuclide activities that deposit to ground, adsorbs to soil or other ground material and subsequently, may become available for resuspension or re-entrainment into the near-surface air layer by processes such as mechanical intrusion, wind, pedestrian and vehicular resuspension. Thus, it is highly likely that historic material is associated with large, possibly soil-bearing particles because soils are the major 'sinks' for atmospherically deposited material.

The history and origin of large particle associated-radioactivity found within the Sellafield site is limited and indeed, difficult to apportion to specific processes or periods of site operations. It is possible that particle-associated activity may originate from on-site sources such as the old Windscale Ponds where fuel rods which are stored underwater have corroded giving rise to particulate-associated enriched radioactive material which concentrates at the air/ microsurface layers of the Ponds. These materials are subsequently ejected from the surface of the ponds and due to their large particle size may deposit locally, Felstead et al (1979), Woollam, (1981). Alternatively, much of the radioactive material found within and close to the site may originate from the Windscale Piles prior to the fire of 1957 due to the poor stack efficiencies of the early filters (Dibben and Howells 1955), (Chamberlain 1991). Other possible sources of locally deposited radioactive material may include a small proportion of marine-discharged insoluble radionuclides swept back to land via on-shore winds from the so-called 'sea-to-land' effect'. This material may also originate from fugitive emissions which periodically escape to atmosphere from contaminated abandoned buildings somewhere within the Sellafield complex.

LIMITATIONS OF MODEL PREDICTIONS:

The critical group dose assessment models that are currently used by BNFL sites assume that material released from stacks are emitted as small particles typically 1 mm (AMAD) with a dry deposition velocity (Vg) of 1E-03 ms^-1. Dry deposition velocity is defined as the downward activity flux divided by the atmospheric concentration, Chamberlain and Chadwick, ( 1953 ).

The assumed rate of Vg of 1E-03 ms^-1 is assumed from theoretical considerations of particle transport physics which is also consistent with weapons fallout values and with recent measurements of the long range transport of Chernobyl material, (Kownacka et al 1994 ).

However, recent deposition measurements at the site perimeter fence at Sellafield indicate much higher values of Vg of up to 5E-02 ms^-1 to 10E-02 ms^-1 for ^239+240Pu and ²³⁸Pu; total Vg (wet + dry ) were even higher, (Nicholson and Fulker 1994 ). Dry deposition velocity for ¹³⁷Cs, although lower than plutonium is still higher than the commonly assumed value of 1 E-03 ms^-1 , used in atmospheric dispersion modelling. High deposition velocities such as those found by (Nicholson & Fulker 1994) are more consistent with large particles > 10 m m (AMAD). Thus, in theory at least, particles this size should not be emitted from the Sellafield stacks.

If the assumed value of Vg of 1 E-03 ms^-1 is not representative of the material discharged from stacks, this may lead to ground-exposure dose assessment models under-predicting the radiological dose from ground-related exposure pathways such as external irradiation (to the skin and body) and that deposited onto grass or agricultural produce. On the other hand, inhalation pathways will be over-predicted because large particles rapidly deposit to a surface due to their high sedimentation properties.

Clearly, it is of some significance that (a) the association between radioactivity and particle size must be established (b) we need to know about the origin and nature of this material as defined by its radionuclide activities, composition and isotope ratios and (c) whether or not these materials originate from stacks or from the resuspension of previously deposited historic material. Data is discussed in relation to the first two points only.

The association between particle-reactive radioactivity and particle size was determined using two air sampler types, a Pm₁₀ to collect the less than 10 um size range and an impaction rod sampler housed in a wind tunnel to collect the greater than 11 um fraction of the aerosol. The deposition fluxes for radioactive material were collected in large area inverted Frisbees at 1 m above ground level. Two Frisbee types were deployed, one to collect dry deposited particles and the other collected total deposition (wet and dry deposited) material.

Large particles by virtue of their inertial and sedimentation properties, rarely enter a sampler inlet quantitatively because these particles have high deposition rates and impact onto areas of the sampler other than the inlet. Consequently, particles in excess of 10 um are not efficiently sampled by routine air samplers such as thePm_10' (Garland and Nicholson 1991). As a result of non-representative air sampling, Vg will be over estimated because the derived Vg value contains an underestimation of air concentration, due to that portion of radioactivity associated with un-collected large particles, (Nicholson & Fulker 1994). It is, however, unlikely that large radioactive particles are emitted directly from stacks, (Garland and Nicholson 1991), (Fry 1983) because the majority of the Sellafield stack discharges are filtered. If unfiltered large particles are present however, large particles will tend to dominate local deposition processes, ( Nicholson 1988 ). Chapter 5 details dry and total (wet and dry) deposition velocities for the Met Station during 1993 - 1994. (download Ph.D here).

It was also important to elucidate something of the nature of these particles, preferably on a particle by particle basis and this was attempted by (a) bulk radiometric measurement of soil and deposition samples and (b) individual hot particle analyses using alpha-autoradiography followed by scanning electron microscopy /energy dispersive x-ray analyses.

An initial task was to characterise the 0-2 cm surface soil layer spatial distribution and activity of the major nuclides, ¹³⁷Cs, ⁴⁰K, ^239+240Pu and ²³⁸Pu around the perimeter of Sellafield to determine the potential for resuspension. Soil parameters which may control the vertical and horizontal migration of these nuclides such as pH, organic matter and soil moisture were also determined.

Surface Activities and Spatial Distribution of Major alpha and gamma -emitters

Data (see Chapter 4 of my Ph.D) show that sites 4, 5 and 6 are characterised by the highest activities of major alpha and gamma-emitters with site 13 to the east of plant ( Calder Hall Farm ) having the highest activities of ¹³⁷Cs . Road dust sites to the S.E. of plant are characterised by significantly low levels of nuclide activity, hence resuspension from this area of the plant appears to be negligible. It is important to note, however, that the composition of road dusts are transient in that they may change on a daily basis and radionuclide inventories within road dusts should not be seen as either sinks or sources.

On the other hand, soil plutonium isotope ratios, ( ²³⁸Pu/^239+240Pu) for a number of sites have similar ratios to weapons test fallout of ~ 0.058, however, the plutonium activities are much higher than fallout values of ~ 5 Bq kg^-1 , (Livens et al 1988). The highest plutonium activities are associated with low ²³⁸ Pu/^239+240 Pu ratios of between 0.058 - 0.08 which is characteristic of late 1950's - early 1960's atmospheric and marine liquid effluent discharges, Jones et al (1995).

The relationship between plutonium isotope ratios and plutonium activities.

The plutonium isotope ratios suggest this material is at least 25-40 years old and is probably in an aged form, e.g. associated with and complexed by soil and other environmental matrices. However, plutonium isotope ratio dating is complicated by the fact that small quantities of low burn up fuel had been processed lately (Fulker 1995 pers. Comm. ). It is also probable that material which has been trapped within the interstices of stack ducts and channels for many years and may be emitted with present day stack emissions (Barlow 1994 ). It is also likely that the plutonium isotopes may be from different sources and consequent reprocessing histories and what is measured may be composite material. It is clear from Table 3 however, that the highest plutonium activities are associated with low ²³⁸ Pu/^239+240 Pu ratios, which are indicative of material from the early atomic weapons era of Windscale.

TOP OF PAGE

 REFERENCES :

Alika et al (1993) Euphytica 66 (1-2), pp. 65-71 Variation among maize (zea mays L.) accession of Bendel State Nigeria: multivariate analyses of agronomic data.

Anspaugh L.R. (1975) Health Physics Vol. 29, pp. 571-82 Resuspension and redistribution of plutonium in soils.

Barlow R. ( 1993) personal communication.

Bleichrodt J.F. (1978) J. Geophys. Res. Vol. 83, pp. 3058-62 Mean tropospheric residence time of cosmic-ray produced 7 Be at northern latitudes.

Chamberlain A.C. and Chadwick R.C. (1953) Nucleonics 11 pp. 22-25 Deposition of airborne radio iodine vapour.

Chamberalain A.C. ( 1991 ) Radioactive aerosols, Cambridge Environmental Chemistry Series ( 3 ), pp. 69-77 Pub. Cambridge.

Chepil W.S. ( 1945 ) Soil Sci. 60, pp. 397-411 Dynamics of wind erosion: initiation of soil movement

Coryell C.D. et al ( 1963 ) J. Geophys. Res. Vol. 68 No.2,pp. 559-566, A procedure for geochemical interpretation of terrestrial rare earth abundance patterns.

Dibb J.E. et al (1993) Atmos. Environ. 27 (17-18) pp. 2751-60 ⁷Be and ²¹⁰Pb in aerosol and snow in the Dye 3 Gas, Aerosol and Snow sampling program.

Dibben H.E. and Howells H. ( 1955) UKAEA, IGO-R/W9 The distribution, characterisation and origin of particulate activity found in the Windscale area, July-September 1955.

Dreicer M. et al ( 1984 ) Health Phys. 46 (1) pp.177-87, Rainsplash as a mechanism for soil contamination of plant surfaces.

Evans E.I. (1991) Msc. thesis, Imperial College of Science, Technology & Medicine. Sea-to-land transfer of radionuclides: Materials and methodologies of environmental air sampling

Evans E.I. (1998) Ph.D thesis, Imperial College of Science, Technology & Medicine. Environmental characterisation of particulate-associated radioactivity deposited close to the Sellafield works

Felstead et al (1979) CEGB Research Division RD/B/N4671 Job No. xj140, Evidence for the existence of a microlayer on the B30 Pond at Windscale.

Fry F.A. (1983) J. Aerosol Sci. 14, pp. 452-452 Airborne radionuclides in the vicinity of a coastal fuel reprocessing plant.

Fulker M.J. (1995) personal communication.

Garland and Nicholson (1991) J. Aerosol Sci. Vol. 22 pp 479-99 No. 4, A review of methods for sampling large airborne particles and associated radioactivity.

Geiss P. (1993) Ph.D thesis, Factors affecting the resuspension of monodisperse particles from grass swards. C.A.R.E. Imperial College, Silwood Park, Ascot Berkshire SLI5 7TE

Gillette D.A. et al ( 1974 ) J. Geophys. Res. 79 (27) pp.4068-75, The influence of wind velocity on size distribution of aerosols

Hall D.J. and Waters R.A.(1985) Atmos. Environ. Vol. 20 No. 1, pp. 219-222 An improved readily available dustfall gauge.

Hamilton E.I. and Clifton R.J. (1988) DOE Report No. DOE/RW/88001 The origin, composition and distribution of ‘hot’ particles derived from the nuclear industry and dispersed in the environment.

Hanson W.C. (1975) Health Phys. 28, pp. 529-37 Ecological considerations of the behaviour of plutonium in the environment.

Hermann ( 1972 ) from Rahn (1976) Graduate School of Oceanography, University of Rhode Island, USA pp.161.

Hoffman F.O. et al (1992) Atmos. Environ. Vol. 26A, No. 18 pp. 3313-3321, Quantification of the interception and initial retention of radioactive contaminants deposited on pasture grass by simulated rain.

Hotzl H. and Winkler R. (1987) J. Environ. Radioactivity Vol. 5 No.6 Activity concentrations of ²²⁶Ra, ²²⁸Ra, ²¹⁰Pb, ⁴⁰K and ⁷ Be and their temporal variations in air.

Jarvis K.E. ( 1988 ) Chemical Geology 68 pp. 31-39, ICP-MS: a new technique for the rapid or ultra trace level determination of the rare earth elements in geological materials.

Jolliffe I.T. et al (1989) Expl. Agric. Vol. 25, pp. 259-69 Comparison of variety means using cluster analysis and dendograms.

Jones S.R. et al ( 1995 ) in press, Deposition of actinides in the vicinity of Sellafield, Cumbria: accounting for historical discharges to atmosphere from the plant.

Jones S.R. ( 1992 ) personal communication.

Kownacka L. et al (1994) Sci. Tot. Environ. 144, pp. 201-15Nuclear weapons and Chernobyl debris in the troposphere and lower stratosphere.

Lambert B.E. ( 1990 ) in How safe is safe? Radiation controversies explained, pp. 220-26 Pub. Unwin Paperbacks.

Livens F.R. et al (1988) Sci. Tot. Environ. 70, pp. 1-17 Particle size and radionuclide levels in some W.Cumbrian soils.

McCarthy W. and Nicholls T.W. ( 1990 ) J. Environ. Radioactivity 12 pp. 1-12, Mass spectrometric analysis of plutonium in soils near Sellafield.

McLendon H.R. (1975) Health Phys. 28, pp. 347-54 Soil monitoring for plutonium at the Savannah River Plant.

Nicholson K.W. ( 1988) Atmos. Environ. Vol.22 No.12 pp.2653-666, Review article: the dry deposition of small particles, a review of experimental measurements.

Nicholson K.W. ( 1993 ) personal communication.

Nicholson K.W. and Fulker M.J. ( 1994 ) J. Aerosol Sci. No. 5 pp. 807-20 The atmospheric surface-exchange of radionuclides at the Sellafield reprocessing plant.

Noll K.E. et al (1988) Atmos. Environ. Vol. 22 pp. 1461-68 No. 7, Characterisation of the deposition of particles from the atmosphere to a flat plate.

Playford K. et al (1993) AEA-EE-0498 DOE/RAS/93.003 Radioactive fallout in air and rain: results to the end of 1991.

Rahn K.A. ( 1976 ) Graduate School of Oceanography, University of Rhode Island, Kingston U.S.A., The chemical composition of the atmospheric aerosol.

Sehmel G.A. (1980) Environ. International Vol. 4, pp. 107-27 Particle resuspension: a review.

Somerville M.C. et al (1994) Atmos. Environ. Vol. 28 No. 21 pp.3483-3493, Statistical approaches in wind sector analyses for assessing local source impacts

Strahler A.N. and Strahler A.H. (1973) Environmental Geoscience: Interaction between natural systems and Man. Publisher: Wiley International

Woolam P.B. (1981) CEGB Research Div. TPRD/B/0018/R82 Job No. XB140, Fuel cooling pond microlayers.