Enhancing soluble phosphorus removal within buffer

Environ Sci Pollut Res DOI 10.1007/s11356-014-3164-5

RESEARCH ARTICLE

Enhancing soluble phosphorus removal within buffer strips using industrial by-products Reza Habibiandehkordi & John N. Quinton & Ben W. J. Surridge

Received: 3 March 2014 / Accepted: 5 June 2014 # Springer-Verlag Berlin Heidelberg 2014

Abstract Using industrial by-products (IBPs) in conjunction with buffer strips provides a potentially new strategy for enhancing soluble phosphorus (P) removal from agricultural runoff. Here, we investigate the feasibility of this approach by assessing the P sorption properties of IBPs at different solution-IBPs contact time (1–120 min) and solution pH (3, 5.5, 7.5), as well as possible adverse environmental effects including P desorption or heavy metal mobilisation from IBPs. Batch experiments were carried out on two widely available IBPs in the UK that demonstrated high P sorption capacity but different physicochemical characteristics, specifically ochre and Aluminium (Al) based water treatment residuals (Al-WTR). A series of kinetic sorption–desorption experiments alongside kinetic modelling were used to understand the rate and the mechanisms of P removal across a range of reaction times. The results of the kinetic experiments indicated that P was initially sorbed rapidly to both ochre and AlWTR, followed by a second phase characterised by a slower sorption rate. The excellent fits of kinetic sorption data to a pseudo-second order model for both materials suggested surface chemisorption as the rate-controlling mechanism. Neither ochre nor Al-WTR released substantial quantities of either P or heavy metals into solution, suggesting that they could be applied to buffer strip soils at recommended rates (≤30 g kg−1 soil) without adverse environmental impact. Although the rate of P sorption by freshly-generated Al-WTR applied to buffer strips reduced following air-drying, this would not limit its practical application to buffer strips in the field if adequate contact time with runoff was provided. Responsible editor: Bingcai Pan R. Habibiandehkordi (*) : J. N. Quinton : B. W. J. Surridge Lancaster Environment Centre, Lancaster University, Lancaster, UK LA1 4YQ e-mail: [email protected]

Keywords Buffer strips . Desorption . Industrial by-products . Phosphorus . Sorption . Water quality Abbreviations IBPs Industrial by-products P Phosphorus WTR Water treatment residuals SRP Soluble reactive phosphorus

Introduction Diffuse phosphorus (P) pollution from agricultural land has been well documented as a major threat to aquatic ecosystems in many parts of the world (Carpenter et al. 1998; Ulén et al. 2007). Increasing P loads and concentrations in rivers, lakes and estuaries can lead to eutrophication and associated adverse impacts on drinking water supplies, recreational uses of water and aquatic ecosystem status (Carpenter et al. 1998; Sharpley et al. 2006). However, the risk of eutrophication can be reduced through a range of best management practices applied to agricultural land (Sharpley et al. 2006). A critical issue for best management practices is their effectiveness in terms of reducing dissolved, rather than solely total, P concentrations and loads (Kleinman et al. 2011a, b). The dissolved P fractions within runoff are usually assumed to be readily bioavailable and to play a key role in the development of eutrophic conditions in freshwaters (Haygarth and Sharpley 2000). Best management practices have demonstrated variable effects on the concentration and load of dissolved P exported from agricultural land. For example, Sharpley and Smith (1994) found that no-tillage practices reduced annual total P loss by 70 %, but resulted in a three-fold increase of dissolved P loss. Increased loss of dissolved P resulted from two factors: firstly, accumulation of P in the surface of no-till

Environ Sci Pollut Res

soil thereby increasing P release to surface runoff, and secondly incidental loss of P due to reduced interaction between soil particles and applied manures (Kleinman et al. 2009). Constructed wetlands are a further example of a best management practice that successfully reduces sediment transport and particulate P export; although their efficacy in terms of dissolved P is highly variable (Ballantine and Tanner 2010). Vegetated buffer strips are widely used for tackling diffuse P pollution from agricultural land. However, these features are less effective for removal of dissolved P from runoff compared to particulate P, with retention ranging from −71 to + 95 %, with a negative retention indicating release of dissolved P to runoff (Hoffmann et al. 2009). The variability in buffer strip performance with respect to dissolved P is associated with a number of factors, including insufficient contact time between runoff P and buffer strip soil, and remobilisation of dissolved P from buffer strip soil due to either geochemical (e.g. reductive dissolution of Fe-bound P) or biochemical processes (e.g. dissolved P release from microbial pools) (Dorioz et al. 2006; Roberts et al. 2012). Therefore, there is a need to enhance the effectiveness of mitigation measures for controlling dissolved forms of P, and in particular, soluble reactive P (SRP). One possible approach to enhance SRP removal is to incorporate the use of materials such as gypsum and Polkemmet ochre, that enhance P sorption, within best management practices (Watts and Torbert 2009; Heal et al. 2005). Among possible sorbent materials, industrial by-products (IBPs) have received increased attention (Barca et al. 2012; Leader et al. 2008). Their use may also offer a route to recycle waste products, such as ochre and water treatment residuals (WTR), which are often otherwise disposed of in landfill sites (Babatunde et al. 2009). This is increasingly important due to recent stringent environmental regulations for waste disposal and water pollution, increasing disposal costs and decreasing landfill capacities (Babatunde and Zhao 2007). Industrial byproducts contain substantial quantities of reactive Al, Fe and Ca which are able to create insoluble P forms via sorption or precipitation reactions (Penn et al. 2011). Batch experiments have been used to characterise the potential of a wide range of IBPs to be used in combination with mitigation measures for removal of dissolved P (Cucarella and Renman 2009; Xu et al. 2006). Most of these studies have focused on selecting suitable IBPs for mitigating dissolved P export from best management practices, based on the maximum P sorption capacities of IBPs. Industrial by-products with low P sorption capacities are not practically useful as their sorption capacity will become exhausted over a short period of time, depending on P loads delivered to the mitigation measure, thereby requiring replacement or disposal of the reactive material. However, the broad applicability of IBPs in conjunction with a mitigation measure also depends on key geochemical parameters, beyond maximum sorption capacity, that have received

less attention to date. For instance, while buffer strips are established to improve water quality worldwide, their design and placement varies substantially, depending on catchment hydrology, topography, soil conditions, and vegetation type (Dorioz et al. 2006). Such variability drives different residence times and pH conditions for runoff entering a buffer strip. As a result, ideal IBPs should not only possess a high maximum P sorption capacity, but also sorb P during short contact times with runoff, as is likely with short hydraulic residence time in the relatively narrow, mandatory buffer strips, and under a wide range of pH conditions. The IBPs should also not pose a risk of adverse environmental effects, such as P desorption or heavy metal mobilisation. Understanding the rate and mechanism of P removal by IBPs is necessary in order to evaluate their performance in conjunction with mitigation measures. The rate of P removal from runoff can determine the contact time required for reducing P concentration to an acceptable level, therefore suitable mitigation strategies can be selected or designed to achieve the required outcome. For example, buffer strips should be designed to be wide enough in order to provide sufficient hydraulic residence time for effective interaction between runoff and IBPs. The feasibility of including IBPs within mitigation measures also depends on their availability at local level, free of charge, as well as possibility of using raw IBPs without further processing. Unlike some IBPs, such as iron ochre which needs to be dried at minewater treatment sites for handling and transport (Heal et al. 2004), others such as many types of WTR can be easily handled in their original wet form. The water content of IBPs applied to a mitigation measure in the field may also vary through time due to variations in rainfall and runoff delivered to a mitigation measure at any particular site. For example, without rainfall, freshly-generated WTR applied to buffer strip soils may dry through time, although it is not clear whether the P sorption rate of WTR can be altered as a result of air-drying. The pore-water content of WTR could potentially affect its P sorption characteristics by influencing the degree of access for phosphate to micropores within the IBP (Makris and Harris 2006). Therefore, the extent to which changes in the water content of IBPs may affect P sorption rate must be understood (Redding et al. 2008). In this study we selected two different IBPs for analysis, including Aluminium-based WTR (Al-WTR) and ochre, both of which are freely and locally available across the UK. Our previous research has shown that both Caphouse ochre and Al-WTR are promising substrates for P removal, due to their high maximum P sorption capacity. We therefore conducted further batch experiments on these materials to: (i) understand the P sorption–desorption characteristics of the IBPs given different solid-solution contact times; (ii) evaluate the performance of the IBPs under a range of solution pH conditions and, for the Al-WTR, using different IBP moisture contents;

Environ Sci Pollut Res

and (iii) identify whether heavy metal release from the IBPs could occur.

Materials and methods Material collection and preparation Two locally available IBPs, iron ochre and Al-WTR, with high P sorption capacities but contrasting chemical composition and production processes, were selected for this study. Ochre was collected from a UK-based, Coal Authority site (Woolley, Yorkshire), which stores large quantities of raw iron ochre that originated from the Caphouse Colliery minewater treatment plant, and was several years old at the time of collection. The Al-WTR was collected from Franklaw water treatment works in Garstang (Lancashire) in which aluminium sulphate is used as coagulant. The IBPs were air-dried, crushed and passed through a 2 mm sieve prior to analysis. Some freshly-generated Al-WTR was also kept separately to investigate changes in P sorption–desorption characteristics during air drying. In addition, to explore realistic application scenarios for the IBPs, ochre and Al-WTR were also mixed with soil at different soil-IBP ratios. The soil used in this research was silt loam (25.5 % clay, 64 % silt, and 10.5 % sand), Banbury association, Banbury series (Ragg et al. 1984) from Kettering, a common arable soil type in the UK. We selected an arable soil because buffer strips are primarily used in arable systems. All the following analyses on either individual IBPs or mixture of soil-IBPs were performed in triplicate. Characterization of IBPs The elemental composition of the materials used in this study was analysed by X-ray fluorescence spectrometry (PANalytical Axios-Advanced PW4400 XRF spectrometer). The maximum P sorption capacities of the IBPs were estimated using the linearized Langmuir model. The pH was determined in a 1:2 solid:0.01 M CaCl2 solution, using a Mettler Toledo FE20 pH meter. The loss on ignition (LOI) was determined after ashing approximately 5 g of oven-dried material at 550 °C for 2 h in a muffle furnace. Loss on ignition was calculated as the loss of weight at 550 °C as the percentage of the oven-dry (105 °C) weight of the sample (Allen 1989). Kinetic P sorption experiments Kinetic P sorption experiments were carried out through batch experiments following the procedure of Nair et al. (1984). One gram of air-dried material (