Evidence from the measurements of the Atomic Weapons Establishment, Aldermaston, Berkshire, UK

Chris Busby
Saoirse Morgan


Uranium weapons have been increasingly employed in battle action since their first use by the US and UK forces in the Persian Gulf War in 1991. Since then they have been used in the Balkans in the late 1990s, then Kosovo in 2000, probably in Afghanistan in 2002 and then also in the 2nd Gulf War (GW2) in March and April 2003. On impact, uranium penetrators burn fiercely to give an aerosol of sub micron diameter oxide particles which are largely insoluble and remain in the environment for many years. There is considerable public and scientific concern that these radioactive particles may remain suspended for long periods, or may become resuspended and are therefore available for inhalation by non combatants at some distance from the point of impact. Little research seems to have been carried out on the distance travelled by the uranium aerosols. The military maintains that the uranium remains near the point of impact, and the Royal Society report (2002) also states that the material does not travel more than some tens of metres. On the other hand, measurements of uranium in local populations in Kosovo some nine months after the use of uranium weapons all tested positive for depleted uranium in urine (Priest 2004) and The United Nations (UNEP) found uranium particles in air filters in Bosnia some years after its use. The question of the dispersion of uranium aerosols from the battlefield is of significant legal interest, since if a radioactive weapon resulted in the general contamination of the public in the country of deployment or elsewhere, the weapon would be classifiable as one of indiscriminate effect. There is now conceded to be no safe level of exposure to radiation. Further, there are major scientific questions over the risk models used to assess the health effects of uranium particle exposure from weapons use. In addition there is evidence of ill health in many of those exposed to uranium particles from Gulf veterans to the population of Iraq. In this paper we examine the trend in uranium shown by the measurements made on high volume air sampler filter systems deployed by the Atomic Weapons Establishment (AWE) Aldermaston Berkshire UK. AWE have been routinely monitoring uranium in air since the early 1990s but since 2000 have carried out filter determinations from high volume air samplers (HVAS) every two weeks. They were required to set up these monitors in the late 1980s following the discovery of a child leukaemia cluster near the plant. There are monitors onsite but they also deploy them at various other sites some 15km distance from the plant. We have obtained their results using the Freedom of Information Act. Examination of the trends in uranium reported here show that there was a statistically significant increase in uranium in all the filters beginning at the start of GW2 and ending when it ended. Levels in the town of Reading exceeded the Environment Agency Reporting threshold of 1000nBq/m3 twice during the period. We report the weather conditions at the time and show that over the period there was a consistent flow of air from Iraq northwards and that the UK was in the centre of a anticyclone which drew air in from the south and from the south east. On the basis of the mean increase in uranium in air of about 500nBq/m3 we use respiration data on standard man to calculate that each person in the area inhaled some 23 million uranium particles of diameter 0.25 microns. We suggest that health data, particularly birth data be examined for possible effects from this exposure. As far as we know, this is the first evidence that uranium aerosols from battle use have been shown to travel so far.


Depleted Uranium weapons have been employed in battlefields at least since the first Persian Gulf War in 1991. Since then, and since their further use in the Balkans in the late 1990s and possibly Afghanistan in 2002, there have been arguments about the health effects of exposure to the uranium oxide aerosols which are produced when the uranium burns in air upon impact. On the one hand, conventional assessments based on the radiological arguments of the International Commission on Radiological Protection (ICRP) have led to most official agencies and government departments to state that uranium exposure at the levels likely to occur after its use in battle is too low for any significant or measurable health effect. But further, it is argued that populations are not even exposed: contamination of the environment is localised to the positions where the strikes occurred.

“On the basis of these two arguments, the many reports of widespread ill health in areas where Depleted Uranium weapons have been used have been discounted by such authorities and thus the military has been absolved thereby of having used a weapon of indiscriminate effect. This is an important ethical, if not legal point since such use is similar to the use of chemical or biological weapons and is banned by the Geneva Convention. Regarding the radiological issue, the European Committee on Radiation Risk (ECRR), an independent radiation risk agency based in Brussels, has published a risk model which draws attention to the inadequacy of the ICRP radiation risk models for dealing with the health consequences on internal radionuclides (ECRR 2003). The concerns of ECRR have recently been echoed by the French IRSN agency who have agreed that the ECRR questions over the adequacy of the ICRP model for internal exposures to uranium are valid. (IRSN 2005). The errors in the ICRP model, which is based on external irradiation following an acute large dose, are particularly important when considering internal radioactive particles and DNA seeking isotopes. The uranium weapons aerosols are in both these categories since the particles have mean diameters below 1 micrometer and are respirable and when translocated to the tissue from the lungs via the lymphatic circulation can cause high uranium ion concentrations in cells. Uranium as uranyl ion UO2++ has enormous affinity for DNA phosphate. The affinity constant is about 1010, (Nielsen 1992) and uranium stains have been used for DNA imaging in electron microscopy since the 1960s (Zobel et al 1961, Huxley and Zubay 1961). Recently, one of us has pointed out that the uranium may focus external natural background radiation on the DNA and enhance its radiological effect (Busby 2005, Busby 2005b).

“There is considerable evidence that uranium is genotoxic and carcinogenic and is associated with a whole range of harmful health effects. However, this brings us to the second main point made about uranium weapons, that of the particle dispersion and possible exposure of those who are at some distance from the impact point, including non-combatants. The environmental dispersion of uranium particles after any battlefield use is a matter of considerable interest. However, little attempt has been made by any official agencies to determine this dispersion of uranium aerosols; rather it has merely been stated that the material remains near the site of impact and cannot contaminate those who are further than some tens of metres from this point.

“Since the 1990s, measurements of uranium in high volume air sampling filters have been routinely made by the Atomic Weapons Establishment at Aldermaston in Berkshire, UK. The requirement to measure uranium and plutonium followed a public enquiry in the early 1990s into releases of these substances to the local environment and the concerns of local people following the discovery of significant excess childhood leukaemia in the area around the plant (for a discussion and the main papers see Beral et al 1990). AWE made environmental measurements of radioactive contamination both on and offsite at various intervals. By 2000 they were routinely (generally at two week intervals) measuring alpha and beta activity in cloths (passive airshades) and also uranium and plutonium in high volume air samplers. These measurements were made onsite and offsite at various locations shown on the map in Fig. 1 and were intended to monitor the releases of uranium from the AWE site. The offsite control locations were some considerable distance from the plant. Thus comparison of levels of radiation at these various sites enables the detection of discharges from the AWE sites.

“The annual publication of the results of these measurements was discontinued in 1999 but the monitoring was continued, the results apparently being reported to the UK Environment Agency. It occurred to us to examine these data for any evidence of uranium from the Gulf War 2 which began in March 2003. The question we wish to address is whether uranium aerosols from the bombing of Iraq in March 2003 became sufficiently environmentally dispersed to reach Europe. In 2004 we applied to AWE for access to these data but the data were not released to us. In January 2005, the Freedom of Information Act (FOI) became UK law. A formal application under the FOI to AWE for results from 2000 to 2004 resulted in the release of the data on paper but curiously the period covering Gulf War 2, that is, early 2003 was the only section missing.

“Re-application resulted in a long wait, and then eventually we received these data from the Defense Procurement Agency in Bristol, and not from AWE. We report here the trend in uranium in high volume air samples on site and near the AWE Aldermaston as shown by these data.

“It is not the intention of this paper to spend much time addressing the health effects of uranium particles and other internal exposures. One of us has dealt with this in various places elsewhere (see Busby 2002, 2003, Committee Examining Radiation Risks from Internal Emitters CERRIE 2004) and there is a considerable literature drawing attention to anomalous mutagenicity associated with exposure to the uranium particles from weapons use (Craft et al 2004, Kuepker and Kraft 2004). The arguments about the health effects pivot upon the scientific validity of using radiation risk models obtained from studies of external acute high dose irradiation (mainly the Japanese A-Bomb studies) for chronic internal exposures to radioactive substances which produce anisotropic i.e. local doses. In addition, one of us has pointed out elsewhere that uranium will amplify natural background gamma radiation owing to its high atomic number and its ability to convert the gamma radiation into local photoelectrons (Busby 2005, 2005b). Uranium has a very high affinity for DNA (Nielsen et al 1992, Zobel et al 1961, Huxley and Zubay 1961, Constantinescu 1974) and in cells which have internalized a submicron uranium particle, the equilibrium ionic concentration of uranium will be high enough to have saturated the DNA in the cell by binding to phosphate. This focusing of the radiation on the DNA may be the cause of many anomalous mutagenic effects which show themselves in cell cultures (Miller et al 2002, 2004) in laboratory animals ( Paquet 2005, IRSN 2005) and in the many reports of ill health associated with exposure to uranium (Craft et al 2005, Zaire et al 1997).


The use of battlefield uranium weapons has been classed by some as weapons of indiscriminate effect; as such they would be implicitly illegal under various conventions of war. Those who defend or justify their use do so by arguing that the uranium is localized at the point of impact or nearby and that exposure of large populations does not occur. The history of the disclosures of the data in this case supports the idea that AWE were aware that their filters provided evidence of the long range movement of uranium. They were at first reluctant to release any data; it required a Freedom of Information Act request to force them to release the results of the monitoring. But significantly they did not send initially the block of data relating to the Gulf War period, and a second request was necessary. The long wait between this second request, and the appearance of the data, and the fact that the missing data came from a different organization, the Defense Procurement Agency in Bristol, suggests that there was significant attention being paid to the interpretation of the results, and decisions had to be made about what the data would show and its political implications for the military.

“Despite many pieces of evidence that the uranium aerosols are long lived in the environment and are able to travel considerable distances, this is the first evidence as far as we know, that they are able to travel thousands of miles. The distance traveled from Baghdad to Reading following the wind patterns implicit in the pressure systems at the time is about 2500 miles. Although this transport may be hard to believe at first, the regular desert sand events which occur in the UK should teach us that the planet is not such a large place, and that with regard to certain long lived atmospheric pollutants, no man is an island. This was a lesson first shown graphically and alarmingly by the atmospheric nuclear tests of the 1960s and the subsequent Strontium-90 in milk, and more recently by the Chernobyl accident. However, like the atmospheric tests, the use of battlefield uranium weapons, especially the new bunker busting bombs which are alleged to have more than 1 ton of uranium in the warhead, are events which are controlled by man: they are not accidents. The results from the AWE filters should teach us that the consequences are not restricted to the areas where they are used. Indeed, on the basis of the results reported here, there would have been a significant exposure to the public in many countries. Uranium is a powerful genotoxic stressor. Although the air concentrations are small in mass terms, the evidence suggests that the excess in the UK represents evidence of dispersion of a new type of uranium, the ceramic sub micron oxide particle. It seems likely that air concentrations in European countries closer to Iraq would have been exposed to higher levels than those found at near Aldermaston. In view of the many reports of heritable genetic effects in areas where uranium has been used and these particles generated, and in the illnesses reported in Gulf veterans, time series analysis of infant mortality and congenital malformation rates in European databases assuming exposures to the foetus or the pre-conception parents in mid March 2003 might be worth carrying out. We have applied to ONS in the UK for monthly data but apparently they are not ready yet.”

Occasional Paper 2006/1
January 2006
Aberystwyth: Green Audit
To read the full copy of this 18 page article, including two pages of references: