back to Appendix 4
UNEP Balkans Report of March 2001Appendix V(page 119 - 122)Possible effects of DU on groundwaterV.1 Possible effectsUranium is ubiquitous in all rocks, soils, rivers and groundwaters on the Earth's surface. Average values for uranium in rocks are about 2-3 mg/kg (20-40 Bq/kg). However, it is not unusual that uranium concentrations are much higher, e.g. in uranium-rich granites the uranium concentration is 10-30 mg/kg. Some black shales that constitute large bedrock areas, such as the Chattanooga shale, have uranium concentrations of 10-80 mg/kg and the Swedish alum shale has uranium concentrations of 50-300 mg/kg.A rock that contains 3 mg/kg uranium contains 8.1 g of U per cubic metre of rock (1 m 3 ). If one penetrator is added into this 1 m 3 of rock the concentration of uranium increases to 308 g or 115 mg/kg of uranium. This value is thus several times higher than the average concentration in soils and rocks, but still not higher than the uranium levels of soil and bedrock in areas where many people live. The uranium concentration in natural waters is much lower than in natural soils and rocks. In natural waters, typically the concentrations vary from less than 1 µg/l (12.4 mBq/l) to 100 µg/l (124 mBq/l) or more. In many countries groundwater in uranium rich areas have concentrations of up to 1000 µg/l, and, in areas with uranium mineralization, of up to more than 1 mg/l (UNSCEAR, 2000). The results of the UNEP missions field study in Kosovo suggests that the majority of the 10 tons of depleted uranium penetrators introduced into the Kosovo environment are probably buried deep in the soil, although the presence of DU in lichen in many of the locations indicates that some dust has also spread over the environment at the time of the attack. One possible effect that the depleted uranium could have on the local population is exposure to enhanced levels of uranium in water due to the dissolution of the penetrators by water in the soil, and transport of the dissolved uranium down through the profile to the groundwater table. This uranium could thus enter drinking water wells. During one interdiction (a steep diving attack) against a target by a single A-10 aircraft, 100-150 DU rounds are likely to be fired. Observations by the UNEP mission showed that the penetrators hit the ground in long lines at an interval of 1-3 m. Test range data from the US indicate that in soft soil the penetrators can go as deep as 6-7 m and that they will sometimes shatter. Therefore as available data is reviewed it is assumed that each penetrator contaminates a minimum of 1 m 3 of soil. The depth to the groundwater table will then be taken into account. The UNEP mission observed that the shallowest groundwater wells had depths to the groundwater table of 2 m. The deepest one was 35 m. If we imagine a column that goes down to the groundwater table that is 1m 2 in area, each penetrator has the capacity to contaminate 2 m 3 to 35 m 3 of soil before the groundwater is affected. Under this assumption, the average concentration of uranium in 1 m 3 of soil (the natural uranium in the soil plus 1 penetrator of 300g uranium) is 115 mg/kg, in 2 m 3 of soil it is 57 mg/kg, and in 35 m 3 of soil 3.3 mg/kg (1% of uranium in a rock containing U-average crustal abundance) . Below, the available scientific data for the behaviour of uranium in the natural environment is summarised in order to explore whether the drinking water is likely to become contaminated by DU in future years. Uranium metal is unstable when in contact with oxygen and water, and therefore uranium oxides form on the surface of the penetrators or fragments. In the presence of water these oxides are hydrated (i.e they contain water). The maximum solubility of oxidised uranium phases (surface layers on penetrators - e.g. schoepite, UO3 . nH2O at near-neutral pH (as found in the water in Kosovo) is about 10 mg/l (ppm). However, there are many processes in nature that can retard the transport of uranium (sorption to minerals and organics in the soil; co-precipitation with calcite) and reduce (if the soil contains iron-II-bearing minerals, bacteria, or organic matter) the uranium from its hexavalent soluble form to its tetravalent insoluble form (UO2 - solubility at neutral pH 0.1 µg/l). Before one evaluates the dissolution and transport of penetrators it is important to consider the composition of soils and rocks in Kosovo. The rocks that comprise the areas that the UNEP mission visited were largely limestone. The UNEP mission also visited Kuke/Kukovce area that was largely comprised of metamorphic rocks. Altered basalts form the bedrock at the dam at Radoniq/Radonjic Lake. At the sites, the depth of soil cover varied, from 4 m in some areas, to only a few cm at Ceja Mountain and Planeje/Planeja. As stated above, the depth to the groundwater table was measured to be 2 m to 35 m in areas where water was collected from private wells. Of note is that most of the wells are less than 10 m in depth and can thus be considered to be in unconfined (or surface) aquifers. The only deeper wells were on hills that rose above the flat valleys. Therefore there exists no confining layer that could protect the aquifers from depleted uranium. The climate in Kosovo can be considered humid-continental with precipitation close to 75 cm/yr. This precipitation value represents the infiltration rate into the surface aquifers. The penetrators that were retrieved in Kosovo showed clear signs of two alteration phases, one black and the other yellow. Analysis of the penetrators at Bristol University using spectroscopy (Raman spectroscopy and X-ray Photoelectron spectroscopy) and scanning electron microscopy with an energy dispersive spectrometer (SEM/EDS), indicate that the alteration phases only contain uranium and oxygen, in addition to low levels of other metals (iron, titanium, chromium, silicon and aluminium) that are known to be present in DU ammunition. XPS analysis of a penetrator fragment shows that the uranium in these oxides is both in the form of U(6+) and U(4+) as determined from reference spectra by Allen et al. (1982; 1984). These yellow and black alteration phases are thus uranium oxides. A French study of depleted uranium metal found on a test site in Southern France (Crançon, 2001) used X-ray diffraction studies of the alteration phases they observed (also black and yellow) to show that these two phases are UO2(OH)2(s) and UO3 . 2H2O (schoepite). The former is likely to be the black alteration phase and the latter the yellow alteration phase (Allen, personal communication). Since the yellow phase is more abundant on the surface, we conclude that the black phase (U 4+ ) is an intermediary step in the alteration from uranium metal (U 0 ) to the fully oxidised yellow phase (U 6+ ). At Gjakove/Djakovica Garrison a penetrator was found under 5 cm of soil. From studies of the uranium concentrations in soils sampled below this penetrator and the amount of uranium that could be easily smeared off the penetrator in the field, it can be concluded that the penetrator had lost 2-8% (5-20g) of the original weight (about 295g) by the effects of being fired and hitting the ground and then by subsequent oxidisation and weathering over the 18 months since it had been fired. About two thirds of this uranium was found in the soil profile, to the depth of 12.5 cm below the penetrator. This indicates that in soil solutions the rate of schoepite dissolution is about the same as the rate of oxidation of the penetrators. This can be concluded from comparison with laboratory dissolution studies of schoepite in the presence of CO2 from the atmosphere (Duro, 1996), which is 32g/300 g of schoepite. This assumes that the surface area of the penetrator is 27 cm 2 . If the penetrator forms an aerosol on impact, its surface area is increased and the dissolution rate increases accordingly. Possibly all of the penetrator will dissolve in 15 to 30 years. Instead of the solid uranium metal penetrators, the soils and rocks will initially contain elevated concentrations of schoepite. With time the schoepite will dissolve and uranium will move downward through the soil. The distance that the uranium will be transported, however, is limited if the penetrator is embedded in organic-rich soil. Once the uranium comes into contact with either organic matter or minerals (<100 µm) further down the soil profile, the uranium will initially be sorbed into these minerals (Waite et al., 1994) and organic matter (Nash et al., 1981). Due to the presence of divalent iron in the soil minerals (Liger et al., 1999) and bacteria (e.g. Loveley et al. 1991), uranium will be reduced to its insoluble tetravalent form (see for example Ragnarsdottir and Charlet, 2000, for a summary of uranium behaviour in the natural environment). In the presence of oxygen the surfaces of UO2 will oxidise to UO2+x and this effect is enhanced by the radiolysis of water. However, soil water normally contains very low levels of oxygen (0.6 mmol O2 per litre), and therefore uranium remains reduced, unless oxygen rich waters are introduced or a high volume of water flows through the rocks. This conclusion is supported by the above quoted French study of a depleted uranium test site in Southern France. There it was found that uranium had only been transported 30 cm down the soil profile (in approximately 30 years) and was entirely held within the A1 profile of the soil (Crançon, 2001). The complementary experimental studies show that the distribution coefficient (Kd=concentration of U in solid divided by the concentration of U in water) is 2000 in the presence of humic colloids, or that uranium moves 3000 times slower (retardation factor, Rf=3000) than the water that percolates through the soil. Crançon does conclude, however, that about 10% of the uranium is able to move further down the profile and reach the underlying groundwater by the formation of uranium-humic colloids which aid in the transport of uranium. This is the cause of the somewhat elevated uranium levels in the groundwater and canal waters of the test site that reach 25 µg/l (ppb) during extreme droughts. Usually the maximum concentration does not exceed 8 µg/l (ppb). The groundwater uranium values in the Southern French site are likely to be higher than in some areas of Kosovo due to the different type of rocks in that area. In France, the soil pH is somewhere between 3.5 and 4.5 for the A1 profile, whereas the soil pore water pH for most of the sites in Kosovo is likely to be higher due to the presence of carbonate minerals from the limestone, which increases the pH. In future years, in areas in Kosovo where the penetrators are embedded in thick soil, the maximum concentration of uranium in groundwater is therefore likely to be below 25 µg/l. When the uranium values found at the French DU test site are compared with acceptable drinking water values, it emerges that the latter vary according to regulating agencies around the world. The drinking water standards for public waters set by WHO is 2 µg/l, whereas the US Environmental Protection Agency has a standard of 20 µg/l, Canadian Health has a value of 10 µg/l and SSK in Germany 300 µg/l (but in Hessen it is 2 µg/l). WHO is currently revising their value of 2 µg/l because it is thought to be unreasonably low. EU has an indicative dose limit for radioactivity of public water of 0.10 mSv/year, which roughly corresponds to a concentration of U-238 of 110 µg/l (EU, 1998). Of note is that many bottled mineral waters have high uranium contents, up to about 100 µg/l. No drinking water standards exist for private drinking-water wells. All of the drinking water samples that were collected by the UNEP mission had uranium values of 2 µg/l and below. Of note is that the water samples not filtered contain measurable uranium in some cases, whereas the filtered water samples are below detection limits. This indicates that some of the uranium is transported as colloids. It is possible that uranium concentrations will increase as more of the penetrators and their alteration phases dissolve. From the observations in France, it may be concluded that uranium will be retarded if the penetrator is trapped in soil. According to the observations made by the UNEP mission, most of the rounds penetrated deep into the ground. If the soil cover is very thin or the penetrators go through the soil down to the bedrock, the capacity for uranium retardation is lower than if the penetrator is trapped in a thick soil layer. There is one exception, namely where the groundwater is over-saturated with respect to calcite, uranium can sorb/coprecipitate with calcite (Kitano and Oomori, 1971; Carroll and Bruno, 1991; Meece and Benninger, 1993). It is also important to consider the depth to the groundwater in drinking water wells. V.2 Conclusions for sites visited[...] Complete UNEP report and appendixes (PDF) - More UNEP DU reportsback to Appendix 4 |