Journal of Environmental Radioactivity 128 (2014) 97e105

Contents lists available at SciVerse ScienceDirect

Journal of Environmental Radioactivity journal homepage: www.elsevier.com/locate/jenvrad

The corrosion of depleted uranium in terrestrial and marine environments C. Toque a, A.E. Milodowski b, A.C. Baker c a

Defence Science and Technology Laboratory, MBG23, Institute of Naval Medicine, Crescent Road, Alverstoke, Gosport, Hampshire PO12 2DL, UK British Geological Survey, British Geological Survey, Kingsley Dunham Centre, Keyworth, Nottingham NG12 5GG, United Kingdom c Defence Science and Technology Laboratory, i-SAT F, 102, Building 5, DSTL Porton Down, Salisbury, Wiltshire SP5 OJQ, UK b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 26 March 2009 Received in revised form 3 January 2013 Accepted 3 January 2013 Available online 5 December 2013

Depleted Uranium alloyed with titanium is used in armour penetrating munitions that have been fired in a number of conflict zones and testing ranges including the UK ranges at Kirkcudbright and Eskmeals. The study presented here evaluates the corrosion of DU alloy cylinders in soil on these two UK ranges and in the adjacent marine environment of the Solway Firth. The estimated mean initial corrosion rates and times for complete corrosion range from 0.13 to 1.9 g cm2 y1 and 2.5e48 years respectively depending on the particular physical and geochemical environment. The marine environment at the experimental site was very turbulent. This may have caused the scouring of corrosion products and given rise to a different geochemical environment from that which could be easily duplicated in laboratory experiments. The rate of mass loss was found to vary through time in one soil environment and this is hypothesised to be due to pitting increasing the surface area, followed by a build up of corrosion products inhibiting further corrosion. This indicates that early time measurements of mass loss or corrosion rate may be poor indicators of late time corrosion behaviour, potentially giving rise to incorrect estimates of time to complete corrosion. The DU alloy placed in apparently the same geochemical environment, for the same period of time, can experience very different amounts of corrosion and mass loss, indicating that even small variations in the corrosion environment can have a significant effect. These effects are more significant than other experimental errors and variations in initial surface area. Ó 2013 Published by Elsevier Ltd.

Keywords: Depleted uranium Corrosion Soil Seawater Eskmeals Kirkcudbright

1. Introduction Depleted Uranium (DU), a by-product of the enrichment process to produce fuel for nuclear reactors, is used in a number of civilian and military applications (such as counter-weights in aircraft, radiation shielding, armour and armour penetrating munitions). It is used to penetrate armour because of its metallurgical properties; in particular, the DU alloy most commonly used (DU-0.75%Ti) is very dense and hard and has self-sharpening properties on impact (Bleise et al., 2003). Armour penetrating DU munitions are often referred to as DU long rod penetrators (from here on referred to as DU penetrators), because it is solely the kinetic energy of a rod of DU flying at high velocity that causes armour penetration to occur. DU penetrators have been used in conflicts in the Balkans (Kosovo, Serbia and Montenegro, and Bosnia-Herzegovina) (UNEP,

E-mail address: [email protected] (A.C. Baker). 0265-931X/$ e see front matter Ó 2013 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.jenvrad.2013.01.001

2001, 2002, 2003) and the Gulf (Iraq and Kuwait) (Akerblom, 2008). In the first Gulf war approximately 300 tonnes of DU were fired, whilst in the second, between 170 and 1700 tonnes were used (UNEP, 2007). In addition, DU penetrators have been test fired on a number of ranges in various countries. In the UK, the two ranges on which they have been test fired are the Eskmeals Range in Cumbria and the Kirkcudbright Range, in Dumfriesshire. At the Kirkcudbright Range DU penetrators were fired at large cloth targets mounted on the cliff above the sea. The penetrators were intended to pass through the targets unhindered, to land several kilometres off shore. The firings at Eskmeals were conducted against hard targets over a short distance. There is evidence that in all the above conflict zones and on both the UK firing ranges, fragments of DU alloy and DU contamination are present in the environment (UNEP, 2001, 2002, 2003; Oliver et al., 2007). Following the deposition of DU alloy in the environment, corrosion is the first step leading to contamination of the wider environment. Movement in the environment is controlled by the

98

C. Toque et al. / Journal of Environmental Radioactivity 128 (2014) 97e105

rate of corrosion and rate of dissolution of the corrosion products (Royal Society, 2001). A number of authors have undertaken research into the corrosion of uranium, DU or DU alloys (see for example, Bullock, 1976; Trzaskoma, 1982; Ritchie et al., 1986; Greenholt et al., 1987; Haschke, 1998; Schimmack et al., 2005, 2007; Handley-Sidhu et al., 2009a,b, and c) and overviews of the corrosion of uranium under different conditions are given in Royal Society (2001), Laue et al. (2004) and Handley-Sidhu et al. (2010). Most corrosion experiments with uranium, DU or DU alloy have been conducted under idealised conditions in the laboratory and often at accelerated rates. It is clear, from the work undertaken that the corrosion of DU alloy is controlled by a large number of factors, including the presence of water, oxygen, chloride and other ions, and on physical parameters such as temperature. Only one study evaluates the rate of corrosion of DU penetrators in the natural soil environment (Oxenberg and Bouwer, 2003) in the field. Unfortunately this has only been reported to date in abstract form. No studies have been identified that assess the rate of corrosion of DU penetrators in the open marine environment. However, a number of studies have measured the rate of corrosion of DU penetrators in soils (Erikson et al., 1993; Schimmack et al., 2007; Handley-Sidhu et al., 2009a,b and c) or synthetic sea water or various NaCl solutions (Trzaskoma, 1982; McIntyre, 1988) in the laboratory. The work reported in this paper aims to assess the rate at which DU would corrode in the main UK environments in which it is present, these being soils at the Eskmeals and Kirkcudbright Ranges and the marine environment offshore from Kirkcudbright. The intention is to gain a greater understanding of the time period for a DU penetrator to corrode and release potentially mobile contaminants into the specific environments of these ranges. It is clear that whilst experiments can be undertaken in the laboratory, slight differences in the geochemical or physical environment may have a significant impact on the rate of corrosion. Hence, this study was undertaken in the field, directly in some of the environments of interest. The basis of the study was to place multiple uncorroded cylindrical pieces of DU-0.75%Ti alloy in terrestrial soil or the marine environment at specific locations and to retrieve them at timed intervals, to allow the extent of corrosion to be evaluated through time. The UK Ordnance Survey Grid References for the terrestrial experimental sites are: NX 70664 43621 (Kirkcudbright) and SD 08018 92924 (Eskmeals). The location of the marine experimental site was approximately 500 m offshore from Kirkcudbright at 54 460 01.1600 North and 004 01000.2900 East. The terrestrial soils corrosion study reported here was undertaken between 8th May 1999 and 9th October 2003. Soil and water samples for geochemical analysis were collected between 5th and 11th October 2003. The open water marine study was undertaken between 10th July 2000 and 24th April 2001. The marine sediments study was undertaken at the same location between 10th October 2000 and 1st October 2003.

Fig. 1. Cylinders of DU-0.75% Ti alloy showing the smooth, finely threaded and coarsely threaded finishes used in the experiments.

retrieved from) each environment and the mean masses of the cylinders are presented in Table 1. The threaded cylinders were machined from the rod of the same dimensions as that used for the smooth cylinders. Hence, the coarsely threaded cylinders, having had most material removed, were lighter than the finely threaded cylinders, which in turn were lighter than the unmilled smooth cylinders. The masses of the cylinders placed in the marine waters and sediments were lower than those for the terrestrial experiment because a 5-mm-diameter hole was drilled along the central axis of each cylinder to allow them to be supported in these more turbulent environments. The initial surface areas of the DU cylinders were calculated using the initial length and radius. For the threaded samples, an appropriate adjustment was made to allow for the additional surface area on the circumference of the cylinders. Prior to placement in an experimental area, each cylinder was thoroughly cleaned by mechanical brushing with a nylon brush in a beaker containing 1000 mL of alkali degreaser (sodium hydroxide) at 80  C for 3 min, rinsed twice in distilled water and once in ethanol before being air dried following the method described in International Standards (1987) and American Society for Testing and Material (1990). 2.2. Placement of DU cylinders in terrestrial and marine environments Each DU alloy cylinder to be buried in the terrestrial environment was placed in a plastic mesh bag and labelled. At each of the chosen terrestrial experiment sites at Eskmeals and Kirkcudbright, six pits of dimensions 0.8 m by 0.6 m by 0.3 m in depth were excavated by hand. During excavation care was taken to keep the top soil and subsoil separated. Twelve DU alloy cylinders (four of each surface finish) were placed directly on the undisturbed soil surface at the base of each pit, keeping a distance of approximately 20 cm between cylinders. The subsoil was then replaced and compacted by hand, followed by replacement of the top soil and root mat. A total of 72 DU alloy cylinders (24 of each surface finish) were buried in each of the two experimental sites. The DU alloy cylinders that were placed within the marine environment were either supported in the open water in an anchored steel frame or attached to lengths of marine cord and buried in the marine sediments. Cylinders to be supported in the

2. Methodology 2.1. Preparation of DU alloy cylinders The DU-0.75%Ti alloy cylinders were obtained from BAe Royal Ordnance Speciality Metals. The cylinders had three types of surface finish to investigate if the surface form of the metal has a significant impact on the rate of corrosion. The surface finishes were coarsely threaded, finely threaded and smooth (as shown in Fig. 1) to approximate the finishes present on the DU penetrators test fired at Eskmeals and Kirkcudbright ranges. Each cylinder was accurately weighed and given a unique identification number. The numbers of cylinders placed in (and also

Table 1 Summary placement data for the DU cylinders. Placement

Number of cylinders

Cylinder type

Mean mass (g)

Standard deviation (g)

Eskmeals soil Eskmeals soil Eskmeals soil Kirkcudbright soil Kirkcudbright soil Kirkcudbright soil Marine Marine Marine

24 24 24 24 24 24 16 16 16

Smooth Finely threaded Coarsely threaded Smooth Finely threaded Coarsely threaded Smooth Finely threaded Coarsely threaded

281 272 246 276 258 243 267 245 229

8 9 12 8 8 11 0.8 1.7 1.1

C. Toque et al. / Journal of Environmental Radioactivity 128 (2014) 97e105

open water were threaded in groups of six onto nylon rods, with each cylinder separated from the next by a 10-cm-long polytetrafluoroethene (PTFE) spacer of diameter equal to that of the smooth cylinders. The central hole of each cylinder was filled with silicon grease to minimise water entry and the cross sections sprayed with a layer of silicon grease. The rods were positioned on the galvanised steel frame onboard ship prior to placement of the frame on the sea bed. The nylon rod and PTFE spacer provided electrical insulation of the DU from the galvanised steel frame. The frame assembly was anchored to the sea bed and buoys were used to mark its location. The frame assembly is shown in Fig. 2. The DU alloy cylinders to be buried in the marine sediments were threaded onto a 6-mmdiameter nylon marine rope, and kept in position by knotting the rope at 20-cm intervals. Each rope carried three cylinders and was secured onto a locating rope extended between two 25 kg anchors and buried at depths of up to 15 cm within sediments. A total of 90 DU alloy cylinders (30 for each surface finish) were deployed in the marine environment (24 in the marine sediments and 66 suspended in open sea water). Deployment of the rig and DU alloy cylinders (and the later retrieval of the cylinders) was undertaken by the Royal Navy Northern Diving Group. Diving conditions at the site were very difficult; it was only possible for the divers to work at slack water at high or low tide (because of the strong currents) and visibility varied from zero to a few tens of centimetres. 2.3. Retrieval of samples The DU alloy cylinders were retrieved from the Eskmeals and Kirkcudbright terrestrial burial sites at intervals of 0.3, 0.5, 1.4, 1.7, 2.2, 2.5, 2.7, 3.0, 3.2, 3.8 and 4.4 years after placement. Retrieval comprised of careful excavation of half of a burial pit and removal of 6 DU alloy cylinders, along with surrounding soil and corrosion products. The 6 cylinders retrieved at each location comprised of two of each type of surface finish. The DU alloy cylinders were due to be retrieved from the marine environment at 0.25 year intervals. The first set of 6 cylinders was retrieved to plan. However, at the next scheduled retrieval date, it was apparent that the steel frame used to support the cylinders in the open water had suffered severe damage in the heavy winter seas and that many of the cylinders were no longer in their intended position or had detached entirely. This retrieval had to be aborted because of the poor weather. The divers returned approximately two months later during improved weather and

99

recovered the frame and most of the DU alloy cylinders. Unfortunately, of those DU cylinders recovered only 18 were still labelled and could be referenced back to their original uncorroded masses. The time of exposure of these 18 DU cylinders to the marine environment was 0.8 years. The DU cylinders buried within the marine sediments were recovered in batches of 3 (one of each surface finish) after periods of exposure of 0.56, 0.83, 1.04, 1.52, 1.75, 2.30 and 2.97 years. A total of 48 identifiable cylinders were retrieved from the marine environment and provided data that could be analysed. 2.4. Cleaning of samples and calculation of mass loss and corrosion rate The corroded cylinders retrieved from the terrestrial environments were separated from the surrounding soil, corrosion products and mesh bag by hand and allowed to air dry for 3 days. The DU alloy cylinders retrieved from the marine environment did not have significant adhered corrosion products or sediment. Any loose corrosion products would have been expected to have washed away, either due to the turbulent nature of the environment or by being unintentionally washed by the divers during retrieval. The DU alloy cylinders were separated from the plastic rods (for the open water experiment) or marine rope (for the sediments experiment) and, as for the terrestrial experiment, allowed to air dry for 3 days. Following drying, all corroded DU cylinders were brushed using a nylon brush to remove any non-adherent materials. This was followed by a chemical cleaning regime to remove corrosion products (following methods recommended by International Standards, 1987; American Society for Testing and Material, 1990). The loss of DU alloy mass was calculated for the cylinders by subtracting their final mass following cleaning from their initial uncorroded mass. Traditionally the corrosion rate is calculated as the mass loss per unit surface area over time. However, the surfaces of the DU cylinders rapidly became very irregular and procedures were not in place to measure the surface areas of such irregular objects. Hence, the corrosion rate is only reported for very early time and later results are presented in terms of mass loss. 2.5. Analysis of the corrosion environments Six samples of soil were obtained from each experimental pit during its excavation in May 1999 for particle size analysis. Samples of the backfill and surrounding undisturbed soils from within or adjacent to the Kirkcudbright and Eskmeals burial pits were collected between 5th and 11th October 2003, during the final retrieval of DU cylinders. These samples were used for mineralogical and geochemical analysis, including measuring the chemical composition of the pore water. Porewater was extracted from the Kirkcudbright clay soil by squeezing and from the Eskmeals sand by centrifugation. The geology of the soil within and surrounding the experimental pits was described in detail at this time and in-situ measurements of pH were made. The methodologies and detailed results of this work are presented in detail in an unpublished report by the British Geological Survey (Milodowski et al., 2005). A summary of the results is given here to allow the corrosion of the DU cylinders to be put into context. 2.6. Analysis of corrosion products

Fig. 2. Frame assembly used to support DU alloy cylinders in the open sea water.

Representative samples of the corrosion products removed from the DU cylinders or the surrounding soil on retrieval were analysed by X-ray diffraction analysis. This identified the phases present in the corrosion products. The data were collected using a Phillips

100

C. Toque et al. / Journal of Environmental Radioactivity 128 (2014) 97e105

PW1700 series diffractometer with APD1700 software, using a cobalt ka radiation source (40 kV, 40 mA), a peak angle range of 15e 70 , in step sizes of 0.02 2q and using count times of 5 s. The raw data was analysed using the Bruker/AXS Diffrac AT software package and the ICDD powder database. The corrosion products were also analysed by Milodowski et al. (2005) on the final set of DU cylinders retrieved in 2003. 3. Results and discussion 3.1. The corrosion environments The soil at the Kirkcudbright experimental site comprises a thin layer of organic litter and root mat (to approximately 0.07 m depth) overlying weathered boulder clay (a glacial deposit comprising clasts of weathered sandstone, low-grade metamorphosed siltstone, and granitic rock, with a matrix of silty clay). The boulder clay is very weathered to a depth of approximately 0.3 m. The soil matrix comprises of 22% sand-46% silt- and 32% clay-sized particles and is classified as silty clay loam. In-situ soil pH values varied between 5.3 and 6.0 in both the undisturbed boulder clay and disturbed silty clay pit backfill. The soil at the Eskmeals experimental site comprises a very thin organic layer to approximately 0.07 m depth over fine to medium dune-bedded sand with thin laminae of gravel. The soil profile is penetrated by roots to over 0.7 m deep and bioturbation by ants is observed to at least 0.25 m. The soil of the burial pit contains 99% sand-sized particles. In-situ soil pH values varied between 7.0 and 7.9 in both the undisturbed sand and the disturbed sand pit backfill. The cation exchange capacity, total inorganic carbon and in-situ pH results for the soils are presented in Table 2. Table 3 presents the results of pore water analysis of two samples (one from the backfill and one from approximately 100 mm beneath the corroding DU cylinders) from each experimental site (from Milodowski et al., 2005). It is clear that there are some significant differences in pore water chemistry between the sites. In particular, the concentrations of Naþ, Cl and total organic carbon are significantly greater in Kirkcudbright soil pore waters. In addition, in-situ and extracted pore water pH measurements indicate that pore waters are more acidic at the Kirkcudbright experimental site than at Eskmeals. No information on the chemistry of the sea water environment at the location of the marine experiments was collected as part of the experiments reported here. However, the chemistry of seawater is well known and relatively “uniform” in comparison to other environmental materials. Seawater in equilibrium with air saturated with water vapour contains approximately 14 g L1 Naþ and 21 g L1 Cl ions and an oxygen concentration of between 5.9 and 8.1 mg L1 for a temperature range of 0e15  C. This diminishes with depth. pH is well buffered and normally lies in the range 8.1e8.3 (Shrier et al., 2000). The mean sea bottom temperatures in

Table 2 Results of cation exchange capacity, total inorganic carbon and in-situ pH analysis for soils from the experimental pits. Eskmeals Mean

Range

Total organic 0.06 0.01e0.16 12 carbon (%) Cation exchange

The corrosion of depleted uranium in terrestrial and marine environments.

Depleted Uranium alloyed with titanium is used in armour penetrating munitions that have been fired in a number of conflict zones and testing ranges i...
946KB Sizes 0 Downloads 0 Views