Depleted Uranium in Kosovo:
Review of UNEP Report of 13th March 2001
Chris Busby PhD
Green Audit
Occasional Paper 3/2001
Aberystwyth
First given in the European Parliament at Strasbourg
Wednesday April 4th 2001

1. Introduction

Following concerns about the possible health effects of radioactive contamination from Depleted Uranium weapons used by NATO in the actions in Kosovo in 1999, a number of scientists and experts were assembled under the auspices of the United Nations Environment programme to visit Kosovo between 5-19th November 2000 to investigate levels of contamination and report on possible health hazards. Details of the expedition and its protocols and findings are to be found in the report [UNEP, 2001]. In this paper I briefly address the results which were reported by UNEP and examine the claims made, with regard to the results and also to measurements which I made in Kosovo and on samples collected by me in January 2001[Busby 2001a].

2. UNEPs Summary and Conclusions

UNEP made three main claims relating to the findings.

1. There was no widespread dispersion of DU in areas of Kosovo where the shells were fired. DU measurements showed only local contamination, i.e. there was no evidence of DU further than 10-50 metres from a direct hit site.
2. There was no contamination of water sources.
3. There was no health hazard to humans anywhere with the possible exception of some slight danger from handling shell fragments for a long period.

Examination of the tables of results shows that all three of these conclusions are incorrect and that the results showed the presence of widespread contamination by DU both by aerosol dispersion of particles greater than 0.2 micron diameter and decay products of U- 238.

I conclude that the analysis of the results given in the tables was either biased or badly interpreted. Significantly, the tables of results were not attached to the report when it was sent to the Press. Consequently it was only the conclusions which were addressed at the Press conference [Parsons, 2001, Fleming, 2001]

I will now address all three of UNEPs claims in turn.

3. Claims that there was no widespread dispersion of DU
a. Ratio tests and total Uranium

The tables of results show that 314 samples were analysed for total Uranium. 79 of these showed total Uranium > 10mg.Kg. Some of these samples were analysed for DU by examining the ratio of U-235 to U-238. Others were analysed for DU by measuring the ratio of U234/U238. Some of the samples had both measurements made. Table 1 below shows the numbers involved:

Table 1. Testing for DU in 314 samples: DU assumed present if percentage present is given as more than 5% in UNEP tables

Method	Total Samples	Percentage of all samples with DU
Total Uranium > 10mg/kg	79	33
U235/U238 ratio calculation	126	40
U235/U238 & U234/238 ratio	143	46

Thus my first conclusion is that the tables show that about 46% of samples showed DU contamination and 33% showed more contamination than 10mg/kg. How then is it that the report concluded that there was no widespread contamination? The answer is that ‘widespread contamination’ was defined in terms of risk to health. But ‘Risk to Health’ was in turn defined in terms of the exposure dose above 1mSv and following the definitions and advice of the models of the International Commission on Radiological Protection. These models were developed for external acute radiation doses and are inappropriate for the internal particle doses which are argued elsewhere to be the basis of the health hazard from DU [Busby 2001a].

b. Anomalous U-234 levels.

The use of U-234/U238 and U235/U238 ratios as an indicator of DU presence relies upon the fact that normal Uranium in the environment is a mixture of three isotopes whose activity and characteristics are reviewed at some length in the UNEP report.

Briefly, the ratio of both U235/U238 and U234/U238 is changed when the U235 and U-234 are removed in the process of manufacture of DU. Once the U-235 is removed, that fixes the ratio of U235/U238. However, the same is not true for U234, because this isotope is a disintegration product of the parent U238. Thus, in a closed system, it will grow back in to the DU over a long period of time. The decay of U238 to U-234 is shown in Table 2 below.

Table 2. Decay of U-238 to U-234

Isotope	Predominant Decay	Half life
Uranium-238	Alpha	4.4 x 109 yrs
Thorium-234	Beta	24 days
Protoactinium-234m	Beta	6.7hrs
Uranium-234	Alpha	2.45 x 105 yrs

Because the intermediate beta emitter daughters Th-234 and Pa-234 have relatively short half-lives, they are said to be in secular equilibrium with the parent. This is to say that their activity is rapidly established as the same as the parent. Thus all samples of U238 should have the same activity (Becquerels per kilogram)of Pa-234 and Th-234 after about twelve weeks. However, this is only true for U-234 after some hundreds of years, since U-234 has such a long half-life. But each decay of Th-234 produces one atom of Pa-234 and one atom of U-234. The reason for this small discussion will now be addressed.

Examination of the tables shows many examples where the activity of U-234 is greater than the activity of U-238. In the environment, for natural samples of Uranium, in rocks or the soil, this is impossible. The ratio of activity cannot be more than 1:1, since all the U-234 derives from the U-238 and has grown in over a very long period. But in the tables of results, the fact that the U-234 measurement is higher than the U-238 measurement, has resulted in the values for %DU which are negative. The frequency histograms of DU percentages in samples as measured by U234/U238 ratios given in Fig 1 show that most of the results are negative, which is impossible, unless there has been some addition of extraneous U234 to the sample. On the other hand the U235 ratios seem to give meaningful results in that the frequency distribution in Fig 2 shows samples which are either free of DU or contain DU.

Thus I conclude that the U-234 ratio is not a good method of analysing a sample for the presence of DU. But there is a more interesting deduction.

c. Causes of anomalous U-234 ratios leading to negative DU concentrations.

If the impact of a shell causes the production of micron-diameter particles of Uranium Oxide, then we need to follow such particles and their movement in the environment. The U-238 in these particles is decaying to Th-234 and Pa-234 but no measurements of environmental beta activity were reported by UNEP. Beta daughters inside the particle matrix may be trapped but surface atom decays will results in the sloughing off of atomic or molecular Thorium and Protoactinium, both insoluble materials. I measured beta levels of 4500Bq/kg in road dust in Gjakove and samples which I brought back and analysed by gamma spectrometry showed that the activity of the beta daughters was significant, at about 5000Bq/kg. However, in some samples, the activity of the daughters was ten times higher than the activity of the parent U-238. [Busby 2001b]. In addition, of course, these decays will lead to the formation of U-234 atoms, as thus there will, in general, be a different dispersion mechanism for U-234 derived from particulate DU than U-238. I believe that this is the explanation for the anomalous U-234 derived DU percentages. The negative DU percentages are usually present in samples where the total Uranium is low, below 10mg/kg. This is a region where small extra amounts of U-234 will show up as a negative DU percentage. For the high Uranium contaminated samples, in general, the two methods agree.

This interpretation supports the belief that the DU aerosols have become widely dispersed. There is further evidence for this from water sample analysis.

4. Claims there was no contamination of water sources

If inhalation of micron-sized particles is the major hazard, then the key question is what is the probability of this occurring. Therefore, the dispersion range and mechanism, the resuspension of precipitated aerosols, the air concentration: these are the areas of interest. This, in turn, means that analysis of air samples and water samples for particles of Uranium Oxide is critically important. The concentration of Uranium in soil is of less relevance, except in that it may indicate the dispersion of the particles from their source zones.

Unfortunately, UNEP did no sampling of the air, or look at air filters from cars, for example, although it would have been fairly straightforward. Nor were measurements of alpha activity in air reported, although they had instruments which would measure this (they covered the windows with plastic in order to keep out water).

Fortunately, there was one set of measurements which they made which is of some interest in that it indicates that high levels of Uranium particulate contamination of the air occurred. This was a set of measurements made on a sample from a pond in Vranovac, UNEP codes 128 and 327. The results are shown in UNEP report Table 4.1 p 23 and are reproduced below in Table 3.

Table 3 Uranium and Thorium in Vranovac pond and some other locations.

Sample	Bristol University Uranium (mg/kg)*	SSI Uranium (mg/kg)**	Thorium (ppb)
UNEP128,327 Vranovac pond	7.79e-5	2.38e-4	0.19
UNEP 080,125,330 Rznic channel water	4.7e-4	4.13e-4	0.01
UNEP 064,137,334 Bandera Farm 3 well	4.2e-5	7.7e-5	0.01
UNEP 077,139, 338 Planeja well	2.17e-4	2.67e-4	ND

*Filtered through 0.2 micron filter; ** Unfiltered

Uranium measurements by the Bristol University team followed a technique involving filtration of the sample through a 0.2 micron filter prior to analysis. Luckily, some samples were also analysed by other laboratories which did not pre-filter the water. In the UNEP table 4.1 reporting the results, most of the analyses show very similar results for samples taken from the same location within the range of experimental variation. However, the Vranovac pond sample shows a marked difference between the filtered sample result and the unfiltered one. The difference between the filtered result of 7.79 x 10-5 mg/kg and 2.38 x10-4 mg/kg is 1.6 x 10-4 mg/kg. Modelled as particles of Uranium Oxide, of 0.5 micron diameter (mass 5 x 10-9mg) this represents about 30,000 particles per litre, if this were the explanation for the discrepancy. I addition we find that this sample showed the presence of anomalously high levels of Thorium. The mean Thorium concentration found by chemical analysis and given in UNEP Table of Trace Elements found by Bristol University for 13 samples was 0.0146 parts per billion (standard deviation 0.0139). The Vranovac pond sample is the only sample which showed 0.19 ppb, which is highly significantly different from the mean. This raises the question of the origin of the Thorium. For Thorium-234 is one of the daughter isotopes of Uranium 238 and thus the particle dispersion model suggests the following explanation fro Vranovac pond.

Vranovac pond is a rainwater pond, and has become contaminated with particles of Uranium as a result of washout from the atmosphere of aerosols of DU of diameter > 0.2 microns. Since these particles are decaying to Thorium and Protoactinium (and U- 234), the surface layer of the particles is rich in Thorium, which becomes detached and separated owing to its molecular rather than particle nature. It is this Thorium which passes through the Bristol University filter and is measured in the sample they took. But because the parent particles are too large to pass through the filter, the Uranium content measured by Bristol is lower than that measured by the Swedish laboratory.

5. Claims that there was no health hazard to humans

The risks to human health were examined using the standard model of the ICRP. This model is not relevant for internal radiation doses from particles since the averaging process used to establish the exposure in Sieverts does not take into account inhomogeneities of energy deposition in tissue through space or in time. The matter has been addressed elsewhere [Busby 2001a and b] and will not be considered further here.

6. Conclusions

Data given in the tables of results appended to the UNEP report (Appendix X) show widespread Uranium contamination in Kosovo and is in disagreement with the conclusions of the report given in the text. Anomalous levels of the isotope U-234 in samples of low overall Uranium content indicate widespread contamination by DU particles. Measurements made by two separate laboratories on water from a pond in Vranovac suggest that the rainwater is contaminated with DU particles to at least 30,000 particles per litre.

7. Recommendations

1. UNEP should urgently take steps to measure and report on the concentration of DU particles and daughter isotopes in the air using both passive and active systems, such as are deployed at the Atomic Weapons Establishment (AWE) Aldermaston, Berkshire, UK.
2. UNEP should urgently take steps to measure and report on the concentration of DU particles and daughter isotopes in rainwater.
3. UNEP should examine the levels of DU in car air filters from the region.
4. UNEP should use Pa-234 gamma lines at 1001 and 765 keV to measure hot spots by gamma camera survey.
5. UNEP should publish results of the beta measurements they made in Kosovo is counts per second, specifying the machine used and its response to a standard source.
6. UNEP should take evidence and consider the validity of the external radiation models of the ICRP for predicting risk from internal DU particles.

References

Busby 2001 a: Science on Trial: On the Biological Effects and Health Risks following Exposure to Aerosols produced by the use of Depleted Uranium Weapons Chris Busby PhD. Invited Presentation to the Royal Society, London, July 19th 2000 Also given in part to the International Conference against Depleted Uranium Manchester, 4th/ 5th November 2000; Occasional Paper 2000/11 Aberystwyth: Green Audit October 2000

Busby 2001 b: Additional Evidence A1 to the paper: Science on Trial Chris Busby PhD. First Presentation to the Royal Society London, March 13th 2001 Occasional Paper 2001/2 Aberystwyth: Green Audit March 12th, 2001

Parsons and Fleming 2001: personal communication from two reporters.

UNEP 2001: Depleted Uranium in Kosovo Post-Conflict Environmental Assessment UNEP Scientific Mission to Kosovo 5 - 19 November 2000