1.4 Predicting environmental behaviour of explosives
It is possible to understand the behaviour of contaminants in the environment by studying contaminated environments; however, this can only be achieved after contamination has occurred. With increasing emphasis of pollution prevention in environmental law, the ability to predict and prevent environmental impact is becoming increasingly important [87]. The behaviour of contaminants in the environment, such as explosives, is affected by a multitude of interacting physical, chemical and biological processes. Therefore, controlled laboratory experiments have been developed to investigate their environmental behaviour, and are frequently used in conjunction with predictive modelling techniques to determine the potential impact of specific activities. The aim of modelling is typically to: (a) detect a site-specific problem; (b) conceptualise important features, processes and events so that proximity to receptors can best be determined (by defining the conceptual model); (c) implement a quantitative description; (d) collect and embrace field/experimental data that can be used to calibrate the model and assess its analytical assumptions; and (e) predict long term contaminant migration and transformation.
1.4.1 Laboratory experiments
Laboratory fate and transport experiments are often designed to simulate environmental conditions to ensure the generation of robust, representative and reproducible data under controlled conditions. Often, the dissolution, transport, degradation and toxicity of explosives are investigated independently to isolate the effect of environmental variables such as rainfall, UV exposure, soil types, temperature, physical degradation and biodegradation. Dissolution studies are used to determine the mechanism and rate of dissolution from the formulation matrix under environmental conditions by simulating rainfall [51]. Soil columns are designed to mimic the flow paths of contaminants in soil, and are often used to simulate the transport of contaminants in different soil types and climates [88]. Finally, photodegradation and biodegradation can be investigated in aqueous solutions and soil suspensions to isolate the effect of UV, biodegradation and adsorption, making it possible to identify associated breakdown products.
Rate of dissolution
The primary mechanism by which explosives migrate into the environment is by dissolution during precipitation [21]. However, only a limited number of methods are available to determine the dissolution rate of explosives in the environment often using different experimental approaches, making the direct comparison of results difficult. For example, methods used in the pharmaceutical industry, where dissolution rates are determined by introducing a solid to a large volume of agitated water and then measuring the time taken to achieve dissolution by sub-sampling at frequent intervals have been used to investigate the effect of variable temperatures (10 °C, 20 °C and 30 °C) on the dissolution rate of RDX and TNT [21]. However, explosives in the environment are unlikely to be submerged continually in agitated water, and are more likely to be dissolved through the action of sporadic rainfall. Further studies to determine changes to the three-dimensional structure of explosive formulations during dissolution by precipitation have been undertaken by submerging samples in water without agitation [89]. The samples were submerged in deionised water (DI), which is not representative of natural rainfall because the pH of the water is controlled (pH 7) and submerging the sample does not simulate the mechanical action of rainfall. This approach allowed the samples to be imaged by digital microscopy, which measured the dimensions of the samples as they dissolved.
To simulate the mechanical action of rainfall, indoor dripping tests have been carried out using glass frits holding pieces of explosive underneath the simulated water flow [76]. Water can be delivered using a gravity-fed system [79], or by rate-controlled systems such as peristaltic or syringe pumps [38, 90, 91]. The dripping is designed to mimic rainfall, although the rate of delivery is averaged against real weather conditions, e.g. to a consistent time and droplet size. However, artificial droplets are generally larger than natural rain falling at the same rate, highlighting the difficulties of simulating rainfall precisely [37]. Indoor rainfall simulations may be continuous or may include pauses for drying time, which increases the mechanical stress on the sample and is more representative of natural weather [90]. For example, drying the sample introduces cracks that allow the ingress of water, and may increase the rate of dissolution, but it is difficult to isolate and measure the effect of this variable in the laboratory. In addition, these indoor experiments only investigate the relationship between rainfall and dissolution, and do not consider other environmental variables such as temperature and sunlight. To account for all variables, similar experiments have been undertaken outside and have shown similar dissolution rates, but significant differences were observed in the mass of explosive recovered [90, 92]. This suggests that the dissolution rate may be largely dependent on water volume, but the explosives may be degraded through other mechanisms such as photodegradation when on the soil surface, which cannot be determined by laboratory-based dripping experiments.
Mass balance
Soil columns have been used as a standard method to evaluate the fate and transport of various chemicals and contaminants, ranging from pesticides to landfill waste [88, 93–99]. Soil columns are usually vertical tubes containing soil packed at a specific density, allowing the soil to be spiked with a contaminant. There is no standardised format, but the literature suggests best practice for building soil columns with different lengths and diameters, with the largest columns weighing up to 50 tonnes [88].
Columns can be extracted from the ground (monolithic), thus retaining the natural flow path characteristics such as micropores, root cavities and cracks. The method used to extract the monolithic column may influence performance, especially if there is any deformation or compression during extraction. Alternatively, homogeneous columns can be used, which are assembled manually from well-characterised soil samples. Preparation involves full soil categorisation (organic content, pH, particle size, etc), sieving the soil before filling, and tamping the soil in the column to ensure homogeneous packing without voids or inclusions. This method does not simulate the soil structure found naturally, but it standardises the soil structure making columns more reproducible and allowing comparisons to be drawn between different