Hot particle dosimetry and radiobiology—past and present

'Hot particles' are physically small radioactive particles that can pose a risk, primarily to the lung, skin and alimentary tract. This paper briefly reviews the origins and effects of 'hot particles'. A specific example of the evaluation of the potential health effects from 'hot particles' is given by consideration of fuel fragments in the environment around the Dounreay nuclear site in Scotland.

The term 'hot particle' has had a somewhat chequered history. The first recorded use was probably some 40 years ago in connection with the evaluation of the risk posed by the extremely non-uniform pattern of dose produced by the deposition of discrete beta-particle sources in the lung following exposure from radioactive fallout on a nuclear battlefield [1]. They are sometimes referred to as 'fleas', particularly in the USA, as a description of their mobility. 'Hot particles' originating from irradiated nuclear fuel are sometimes referred to simply as fuel fragments. Those 'hot particles' occurring most commonly in the nuclear industry are predominantly beta/gamma emitters. Due to rapid beta attenuation and the near-inverse square fall-off in fluence with distance for small sources, the spatial dose distribution around 'hot particles' is highly non-uniform. Conventional methods for evaluating dose and for predicting biological effects are inappropriate for 'hot particles'. 'Hot particles' with linear dimensions less than ~10 µm are capable of entering the deep lung, and long-term residence of insoluble material may lead to a potential long-term lung cancer risk. Large 'hot particles' with dimensions >10 µm can be inhaled but they will be deposited in extrathoracic airways or the first few generations of bronchi. They will not penetrate to the deep lung and are likely to be rapidly cleared. Large particles may also represent a hazard following ingestion or from external exposure such as to the skin. The potential hazard of large 'hot particles', rather than smaller respirable particles, has been the main concern in recent years regarding 'hot particle' risks in the nuclear industry. The exposure of skin has been the most dominant actual practical experience, particularly for ageing American water reactors.

In the nuclear power industry, high specific activity particles can arise by two main routes [2]:

'Hot particles' from both sources are of particular practical concern for ageing US water reactor plants [3, 4]. Both types of particle are seen more frequently on pressurised water reactors (PWRs) than boiling water reactors (BWRs) with the frequency of finding them during outages being approximately five times that during normal operation. The particles found on PWRs tend to be slightly more active, with isolated examples up to 40 GBq. Particles that have arisen by neutron activation in UK gas-cooled power reactors are more diverse, ranging in size from small microscopic fragments to identifiable components (such as activated washers) and containing such radionuclides as ⁵¹Cr, ⁵⁴Mn, ⁵⁶Mn, ⁵⁹Fe, ⁵⁸Co, ⁶⁵Zn, ^110mAg, ¹²⁴Sb, ¹⁹²Ir and ¹⁸²Ta, as well as ⁶⁰Co which dominates in water reactors. Figure 1 shows some examples of 'hot particles' that have been found on UK power reactor sites. In each case, the skin dose is largely due to beta radiation. The limited range of beta particles leads to the majority of the total energy being deposited within a skin area that is significantly less than 1 cm². Average doses over an area of 1 cm² may thus be up to several orders of magnitude (depending on particle size) less than the peak dose immediately beneath the particle.

Information relating to radiation-induced effects following uniform large-area/volume exposures (as in radiotherapy for example) is totally inappropriate for predicting the biological effects of non-uniform exposures from 'hot particles'. Similarly, radiation-monitoring instruments that have been calibrated using wide-beam gamma-radiation fields, or large-area beta sources will normally require the application of significant correction factors when used to measure doses from 'hot particles'. Recent advances in dosimetry and radiobiology have addressed these issues and they will be briefly discussed in the following sections.

Zoom In Zoom Out Reset image size

Controversy regarding biological effects of 'hot particles' first arose during the late 1960s in connection with lung irradiation by small particles—so-called 'hot particles', containing alpha-emitting radionuclides, particularly ²³⁹Pu. Claims were made [5–7] that lung irradiation by 'hot particles' was several orders of magnitude more carcinogenic than estimates made by assuming a uniform lung exposure—as advised by the International Commission on Radiological Protection (ICRP). A 'hot particle' thus signified a physically small radioactive particle that can deliver very high local doses to tissue such that adverse biological effects might be expected, and where average dose was not considered to necessarily be an appropriate parameter. The claim of increased carcinogenicity was referred to as the 'hot particle hypothesis'. After a decade of intensive experimental and theoretical investigation an ICRP [8] review of the biological effects of inhaled radionuclides in 1980 refuted this claim in the context of inhaled 'hot particles'. The 'hot particle hypothesis' in the context of skin exposure has been refuted by a series of skin cancer induction experiments in mice and in cell studies of induced malignant transformation [9–11]. The ICRP has maintained its advice that the use of mean organ or tissue dose is appropriate for the evaluation of carcinogenic risk for radiological protection purposes.

There has been considerable experimental work on pig skin in the past 20 years in the UK [12] and the USA [13] to provide radiobiological data regarding the effect of skin exposure from 'hot particles'. Data are available on pig skin exposures with various sizes of particles containing beta-emitting radionuclides. The incidence of early and late acute effects such as acute epidermal necrosis, ulceration and dermal thinning has been scored, as well as the time-course, severity and healing outcome. The combination of UK and US data shows remarkable consistency. When doses are averaged over an area of 1 cm² at a depth of³ 70 µm (as recommended by the ICRP [14]) the ED₁₀ and ED₅₀ values (the doses to produce an incidence of the effect of 10% and 50% respectively) for acute epidermal necrosis or acute ulceration vary by only a factor of ± 2 across a wide beta-energy range. The ED₅₀ (1 cm², 70 µm) values are approximately 15, 8, 12, 6, 6, 11, and 8 Gy for ¹⁴⁷Pm (E_max~0.25 MeV), ⁶⁰Co (E_max~0.32 MeV), ⁴⁶Sc (E_max~0.35 MeV), ¹⁷⁵Yb (E_max~ 0.47 MeV), ¹⁷⁰Tm (E_max~0.97 MeV), fissioned ²³⁵UC₂ micro-spheres (E_max~1.8 MeV), and ⁹⁰Sr/⁹⁰Y (E_max~2.3 MeV), respectively. These data are for small sources (dimensions <1 mm). For sources larger than ~1 mm the ED values are increased [12, 15]. In the case of higher-energy beta sources with diameters of ~3 mm (corresponding to the largest and most active of the Dounreay fuel fragments, 10⁸ Bq ¹³⁷Cs) the increase in ED₁₀ and ED₅₀ values is by a factor of about 2–3.

These studies have enabled the ICRP [14, 16] and NCRP [17, 18] to provide authoritative guidance on the limitation of skin exposure from 'hot particles'. The ICRP [14] has recommended an annual skin dose limit of 0.5 Sv, measured at a depth of 70 µm, for occupational exposures, with a factor of 10 less for members of the public. This limit is appropriate for general skin exposures as well as 'hot particle' exposures, in which case the dose should be evaluated over the most highly exposed area of 1 cm². The NCRP [18] has recently recommended a relaxation by a factor of 10 in the 'hot particle' skin dose limit, compared to the ICRP recommendation [19]. At this higher dose limit there is a significant risk of an observable skin lesion but the NCRP considers that these would be relatively innocuous.

Discrete fragments of irradiated nuclear fuel have been discovered on the foreshore at the Dounreay nuclear site in Scotland, offshore on the seabed and at nearby public-access beaches—primarily Sandside beach. The fuel fragments vary substantially in size but are most typically similar in size to grains of sand [20, 21]. The principal radionuclides contained within the particles are the fission products ¹³⁷Cs and ⁹⁰Sr/⁹⁰Y, and they also contain small amounts of ²³⁸Pu, ²³⁹Pu and ²⁴¹Am. The ¹³⁷Cs activity by which the particles are generally characterised is within the range of 10³–10⁸ Bq. The most highly radioactive particle found at Sandside Bay contains ~3 × 10⁵ Bq ¹³⁷Cs (it is the particle shown at the bottom of figure 1). Larger particles are found offshore.

There are two main types of particle: fragments of Materials Test Reactor (MTR) and Dounreay Fast Reactor (DFR) fuel. The more abundant MTR particles originated as swarf generated during milling to remove aluminium cases from fuel elements [20, 21]. In general DFR particles have smaller proportions of ⁹⁰Sr/⁹⁰Y and the actinides, relative to ¹³⁷Cs, than do MTR particles. They are also generally less soluble. The dose and risk assessments presented in this paper relate primarily to MTR particles and are likely to be conservative when applied to DFR particles.

Because of the high local dose rates delivered by the more active particles and because they are generally of low solubility, the principal concerns are possible skin contact and localised damage, and possible ingestion and damage to the alimentary tract, particularly the large intestine. A previous NRPB assessment of potential health effects by Wilkins et al [22] considered both these possibilities, but the dosimetric approaches used resulted in conservative estimates of doses and risks [23], taking no account of self-absorption of energy within larger particles and applying the standard ICRP [24] model of the alimentary tract. The Scottish Environment Protection Agency (SEPA) has commissioned a re-evaluation of doses and risks from Dounreay particles, carried out jointly by staff at the University of Birmingham and the Health Protection Agency, Radiation Protection Division (formerly the NRPB). In this more recent study [25, 26], self-absorption within particles was taken into account and alimentary tract doses were calculated using a new ICRP model of the alimentary tract [27]. In addition, the possibility of inhalation of particles was addressed and absorption to blood and doses to all organs and tissues were considered in the context of risks of cancer and hereditary effects. Some of the results of this most recent evaluation are presented here, giving most emphasis to possible exposures of the skin and extending consideration of risks from those particles entering the eye.

SEPA also commissioned a study of the probability of an individual encountering a fuel fragment on Sandside beach. Considering all potential uses of the beach, the probability of contact with a fuel fragment was estimated to be less than 10⁻⁶ y⁻¹ [28].

Measurements of dose distributions from Dounreay fuel fragments, over various skin areas and at several depths, have been carried out using the radiochromic dye film (RDF) technique [29]. The 37 fuel fragments selected for measurement had been subject to prior characterisation using scanning electron microscopy and energy dispersive x-ray microanalysis. Activity determination of¹³⁷Cs facilitated comparisons of measured doses and Monte Carlo calculated doses. Calculations of skin dose were made using the MCNP [30] Monte Carlo radiation transport code, assuming spherical particles of uniform elemental composition of U/Al (15% U), density of 3.1 g cm⁻³, activity 2 GBq g⁻¹ (¹³⁷Cs), and a ⁹⁰Sr/⁹⁰Y:¹³⁷Cs activity ratio for MTR particles of 0.9. Figure 2 shows the measured dose distribution around the highest-activity fuel fragment found to date at Sandside beach.

Figure 3 compares measured and calculated skin dose rates (1 cm², 70 µm) for all 37 particles, showing good agreement for MTR particles (lower solid line). Measured skin doses are intermediate between predictions with no self-absorption (upper dotted line, which is a linear extrapolation of the dose data from small particles) and self-absorption (lower solid line). This is compatible with calculated effects of self-absorption and deviations from spherical shape. Measurements for DFR particles give lower dose rates than predicted by calculations for MTR particles and most are close to calculations for particles with no ⁹⁰Sr/⁹⁰Y content (lower dashed line). Also shown in figure 3 are the higher dose rates assumed in earlier evaluations of health effects by Wilkins et al [22] (upper solid line), which took no account of self-absorption of energy within larger particles and assumed a higher ⁹⁰Sr/⁹⁰Y:¹³⁷Cs activity ratio of 2.1 than is now considered appropriate.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Considerations of the health effects of 'hot particles' have largely concentrated on effects of respirable particles (size<10 µm) in the lung and external exposure from larger particles on the skin. The NCRP [18] has also considered possible effects of particle entry into the ear or eye, although there are only two published examples of such exposures. The probability of ear and eye exposures is presumably much less than potential skin exposures, if only for the reason that the areas exposed to the environment are much less than for the skin.

Table 1 gives dose rates and estimated residence times of fuel fragments on the skin to reach threshold doses for acute ulceration and ED₅₀ [25]. For example, for fuel fragments of activity 10⁵ and 10⁸ Bq ¹³⁷Cs MTR particles (diameters of about 300 µm and 3 mm), dose rates were estimated as about 0.3 Gy h⁻¹ and about 70–140 Gy h⁻¹, respectively, averaged over 1 cm² at a depth of 70 µm. On the basis of a threshold for skin damage of about 2 Gy, 10⁵ Bq ¹³⁷Cs particles would have to remain in stationary contact with skin for at least about 7 h to cause any discernible effect. Movements of even a few millimetres during contact would increase the threshold dose by a factor of ~2–3.

Stationary 10⁵ Bq particles if trapped for longer periods of a day or two may cause small ulcerations that extend over areas of a less than 1 cm² and heal within a few weeks. Such ulceration could occur following short periods of contact of 15–30 min or more to 10⁸ Bq ¹³⁷Cs particles, and exposures of several hours would result in more serious ulceration that would take longer to heal.

Point opacities of the eye lens are present in all eyes and increase in number with age [31]. Detectable induced opacities of the eye lens can be observed months/years after acute exposure at absorbed doses of ~1 Gy to the whole lens [14, 18]. Opacities at these doses are likely to be small and cause little visual impairment. The incidence and severity of the opacification increases as the dose increases, and the ICRP [32] has in the past considered that for an acute exposure a threshold dose of ~5 Gy was appropriate for more extensive opacities (cataract) that could cause visual impairment. Fractionation of the exposure over long periods of weeks–months increases these threshold doses by a factor of up to about 3. Exposure of only part of the eye lens produces significantly reduced detriment [31]. More recently, in draft recommendations, the ICRP [33] refers to a dose value of ~1.5 Gy for a 1% incidence of cataract (visual impairment). The ICRP [14] recommends annual dose limits for workers and members of the public of 150 and 15 mSv, respectively, on the basis that continuous, year-on-year, exposure at these dose limits should not produce impairment of vision from cataract induction. No justification is given for the factor of 10 reduction for members of the public.

The mechanism of cataract formation involves damage to a target population of equatorial epithelial cells around the whole periphery of the lens (marked E on figure 4 [31]). These cells are more sensitive than other cells in the lens epithelium since they have a higher mitotic index and are also undergoing differentiation into lens fibres that will subsequently pass into the body of the lens. Radiation exposure leads ultimately to denaturation of proteins in these cells, which then undergo opacification as they progress to the rear central region of the lens, as part of the continual slow process of lens growth. Cataracts can be removed with a minor operation, and vision can be restored with a lens implant.

Zoom In Zoom Out Reset image size

Figure 4 gives the range of eye geometries seen in children and adults, extending from short-sighted children to far-sighted adults in old age [31]. The infant eye is somewhat smaller than in young children and the cornea and lens are subject to significant changes in shape during early infancy. The equatorial epithelial cells around the periphery of the lens lie at a depth of at least ~2 mm from the eye surface (figure 4) over a wide age range. For a 'hot particle' at any position on the corneal surface the majority of the equatorial region of the eye is much deeper than this, and the mean equatorial lens dose will be at least 2 orders of magnitude less than the skin dose rate (1 cm², 70 µm) from a fuel fragment. Damage to the cornea is therefore considered to be of greater concern [18].

The cornea is approximately 0.5 mm (500 µm) thick and consists of connective tissue sandwiched between an outer epithelial layer and an inner endothelial layer. The outer epithelial layer is about five cells thick. It is a mucous type in contrast to the outer keratinised layer of the skin. The cellular turnover time (time for a cell in the basal layer to divide, differentiate, and migrate to the surface and desquamate) is about one week. The inner endothelial layer consists of a single layer of cells which repairs injury by the elongation and spreading of surviving adjacent cells.

There is a considerable literature regarding beta irradiation of the cornea in radiotherapy. It is used for example in conjunction with local surgery for the treatment of pterygium—a raised, wedge-shaped growth of the conjunctiva. Beta irradiation prevents the regrowth of pterygium after eye surgery. Single acute surface doses of 30 Gy and fractionated doses up to 60 Gy have been used for this treatment using large area ⁹⁰Sr/⁹⁰Y beta sources in contact with the cornea. These treatments are considered to be effective, with few complications and little likelihood of cataract induction [34].

There is insufficient experience or experimental data to definitively predict the outcome of a high dose delivered to a small area of the cornea. However, it is reasonable to assume that small-area effects would not exceed the expectation of damage from the same dose delivered to a larger area. That is, if surface doses of at least 30–60 Gy to the cornea can be tolerated from large-area exposures then similar doses from 'hot particles', measured over small areas in the vicinity of the 'hot particle', should also be tolerated. NCRP (1999) [18] also considered that although 5 Gy (1 cm², 70 µm) represents the ED₅₀ for skin lesions, only a very small fraction of these would extend to a depth equivalent to the full corneal thickness. Non-trans-corneal wounds would not be expected to produce other than very small imperfections in corneal topography, which would have little likelihood of influencing visual function or acuity. On this basis NCRP (1999) [18] recommended for radiological purposes that eye dose from a 'hot particle' should be limited to 5 Gy (1 cm², 70 µm) averaged over the most highly exposed 1 cm² of ocular tissue. This was considered adequate to prevent effects such as loss of visual function, breaching of the eye (e.g. resulting from severe damage of the cornea), or breakdown of the barrier function of the eye-related skin (e.g. lid) with consequent possibility of infection. Actual doses obviously depend on residence times of particles in the eye. The largest Dounreay fuel fragments found at Sandside Bay have diameters of about 0.4 mm. This is similar to a medium-size grain of sand on the beach. It would seem reasonable to consider that toleration of this would not usually extend for more than a few hours, and early biological effects would be unlikely. Dose rates to the cornea are likely to be reduced due to particle movement around the eye, and movement of eye lids and eye ball. Extended corneal exposure to higher-activity particles could produce corneal ulceration, which may require medical intervention and treatment, as in the case of side effects of high-dose beta irradiation in ocular radiotherapy.

The external ear consists of the auditory canal, which is terminated at the ear drum (the tympanum). There is little information regarding the skin response at these sites, but even high doses of 60–70 Gy from low-dose-rate (0.3–0.9 Gy h⁻¹) ¹⁹²Ir interstitial radiotherapy sources produced little or no observable damage [35]. On the basis of similarities with skin responses, the NCRP recommended a dose limit for the ear from a 'hot particle' which is the same as they recommended for the skin (0.5 Gy averaged over 10 cm² at a depth of 70 µm, which is equivalent to a dose of 5 Gy over 1 cm²). The likely residence times of particles in the ear are difficult to ascertain. Personal ear-cleaning habits are likely to vary widely. Ulceration of the skin of the ear is likely to lead to medical help which may include treatment for infection and ear syringing. NCRP considers that the limited beta-radiation range in tissue will limit radiation effects to the skin of the ear or the tympanic membrane. Secondary infection of the middle ear could produce some hearing loss if infection in the tympanic membrane was untreated.

An assessment has been made of possible acute radiation damage to the alimentary tract following the ingestion of fuel fragments, concentrating on the colon as the region receiving the greatest doses. There is little evidence for long term retention in the crypts between villi in the small intestine [27] and this has been ignored. Doses have been calculated using a new ICRP model of the human alimentary tract [27]. The model has explicit consideration of doses to target cells in the various regions and includes age-related parameter values for dimensions and transit times [36].

Doses to the rectosigmoid region of the large intestine were calculated for MTR particles in transit through the region using the MCNP code, taking account of self-absorption within particles, with assumptions regarding particle composition and specific activities as for skin dose calculations. Absorbed dose estimates have been made for a particle moving randomly through the lumen of the rectosigmoid, with no fixed position relative to the wall, and maximum doses for a particle moving in contact with the wall. Doses from a 10⁸ Bq ¹³⁷Cs particle moving randomly were calculated as 0.3–0.4 Gy in adults and 1 Gy in one-year-old children. The corresponding maximum estimated doses were about 1–2 Gy in adults and 4 Gy in one-year-old children. Estimated doses for particles with a ¹³⁷Cs activity of 10⁵ Bq, typical of the highest activity particles found at Sandside Bay, were less than 10 mGy in adults and 20 mGy in one-year-old children. These doses compare with an estimated threshold dose for lethal damage to the colon, following protracted irradiation from ingested radionuclides, of 20 Gy and an LD₅₀ of 35 Gy [37, 38].

An additional consideration for ingested particles is that the flow of material through the colon is highly variable. Movement in the rectosigmoid in particular does not occur as a constant flow but rather as mass movements resulting from periodic contractions between longer periods of quiescence. Local doses within the rectosigmoid may therefore be substantially greater than the average dose within the region. As discussed by Harrison et al [25], such exposures can be considered to be similar in effect to exposures of skin.

The ICRP Human Respiratory Tract Model [39] was used to calculate doses from inhaled particles, using the model to determine the probability of particle deposition in different regions. The analysis showed that inhaled particles of sufficient activity to cause acute damage to the lungs would be too large to reach the airways of the lungs and would deposit in the extra-thoracic airways. Doses from particles deposited in the anterior nasal passages were estimated as around 300 mGy for a 300 µm particle (10⁵ Bq ¹³⁷Cs) and 300 Gy for a 3 mm particle (10⁸ Bq ¹³⁷Cs). These estimates are based on standard ICRP model assumptions including a clearance half-time of one day, and represent an average dose to the target tissue in the region. Potentially more important is the possibility of high local doses which may be substantially greater than the average dose within the region. For the purposes of assessment of possible effects of local irradiation, the epithelial lining of the extra-thoracic airways can be regarded as similar to skin. It should be noted that particles of 3 mm diameter are substantially larger than the range of particle sizes normally considered to be respirable; the probability of their inhalation will be very low and might only be expected to occur as a result of airborne sand caused by very high winds.

The calculation of committed equivalent doses to all body tissues and committed effective doses requires information on absorption of radionuclides to blood following particle ingestion or inhalation. The work commissioned by SEPA included in vitro and in vivo measurements of the solubility of ingested particles and the availability of their radionuclide content for intestinal absorption. Dissolution was measured in vitro in simulated stomach and intestinal fluids (at the Scottish Universities Research and Reactor Centre) and absorption was measured directly in rats given particles by gastric intubation. On the basis of these studies, typical values for absorption to blood from MTR particles were taken to be 1% for ¹³⁷Cs, 0.01% for ⁹⁰Sr and 0.001% for ²³⁸Pu, ²³⁹Pu, ²⁴¹Am and ⁹⁰Y.

Committed equivalent doses to all body tissues and committed effective doses were calculated for ingestion of MTR particles of typical solubility and for the very soluble particle, using standard ICRP models but taking account of self-absorption of energy within particles throughout the alimentary tract using a degraded spectra approach. For typical particles, committed effective doses following ingestion of 10⁵ and 10⁸ Bq ¹³⁷Cs particles were estimated as 0.1 mSv and 80 mSv, respectively, for an adult male, and 0.5 mSv and 300 mSv, respectively, for a one-year-old child. In each case, these estimates of dose are dominated (>70%) by contributions from committed equivalent doses to the alimentary tract, particularly the colon.

Thirty years ago there was such a lack of knowledge on the early and late effects of 'hot particles' that little guidance could be given by the ICRP and NCRP on their control. Recent advances in dosimetry, radiobiology and epidemiology now enable 'hot particles' to be characterised and their early and late biological effects to be assessed. Some uncertainties remain in specific cases. The current situation has been exemplified by the evaluations of doses and risks from fuel fragments in the environment around the Dounreay nuclear site.

Analyses of possible doses and risks from exposure to Dounreay fuel fragments indicate that the principal concern following skin contact, ingestion or inhalation is the possibility of localised ulceration of skin or of the mucosal lining of the colon or extra-thoracic airways. A 10⁵ Bq ¹³⁷Cs MTR particle, typical of the more active particles found at Sandside Bay, would deliver skin doses of about 0.3 Gy h⁻¹ (1 cm², 70 µm). A 3–6 h exposure to such a particle is not expected to result in any visible skin lesions. Stationary contact for longer periods of 15 h or more might result in small lesions. More active particles have the potential to cause more serious ulceration over small skin areas.

ICRP [14] recommends dose limits for localised skin exposure and committed effective dose. These dose limits apply to controlled sources and are not intended to apply to existing situations in which the only available protective action takes the form of intervention, such as is the case with these fuel fragments. Nevertheless, the limits provide values with which to compare possible doses from a fuel fragment. The dose limit for localised skin exposure of workers of 0.5 Sv (1 cm², 70 µm) can be regarded as conservative when applied to 'hot particle' exposures [40] since the threshold for effects is around 2 Sv. For the public, ICRP reduces the dose limit by a factor of ten to 50 mSv. However, this reduction has no scientific basis and it can be argued that no special limit for members of the public is necessary [40]. At a dose rate of 0.3 Sv h⁻¹, a 10⁵ Bq ¹³⁷Cs MTR particle held stationary on the skin would deliver doses corresponding to the ICRP worker and public dose limits in<2 h and 10 min, respectively. The limit on committed effective dose to members of the public is 1 mSv. Doses from ingestion of 10⁵ Bq ¹³⁷Cs MTR particle are likely to be<1 mSv in adults and children, but slowed colonic transit or uncharacteristic particle solubility could result in doses of a few mSv.

It can be concluded that a local skin dose from a 10⁵ Bq ¹³⁷Cs MTR particle is unlikely to cause ulceration, but the possibility of a small lesion cannot be ruled out for long residence times. More active particles are more likely to cause serious ulceration. Committed doses and cancer risks are of secondary importance.

Our understanding of the radiobiological effects of 'hot particles', and our ability to measure relevant doses from them, has increased dramatically over the past 25 years. This is true particularly for the exposure of skin. Effects in the skin have been used as surrogates for effects in other organs, but there remains a lack of direct information regarding the effects of 'hot particles' in the eye and ear. This is likely to remain the case since reported incidents of such events are very rare, and animal experiments to clarify the issue would be difficult to justify. The probability of a particle entering the eye or the ear is presumably much lower than of it locating on the skin. The likelihood of a casual encounter with a 'hot particle' in the environment, even in the environs of Dounreay, is very low [28, 41]. The evaluation of health effects commissioned by SEPA [25] has shown, on the basis of reasonable radiobiological and dosimetric assumptions, that even in the case of a human encounter with a 'hot particle' at Sandside Beach, there are unlikely to be any discernible biological effects. It is possible to predict significant health effects, but only by making assumptions which have a low probability, such as exceptionally long residence/transit times [25]. This low probability must be compounded with the already very low probability of encountering a particle (estimated as <10⁻⁶ y⁻¹ for Sandside beach [28]). The challenge for regulators remains, as in other situations of environmental contamination, to determine the extent to which remediation is required in the light of potential health effects with a low probability of occurrence.