Dynamic sorption of ammonium by sandy soil in fixed bed columns: Evaluation of equilibrium and non-equilibrium transport processes

Abstract

The release of excess nitrogen-containing compounds into groundwater is a major concern in aquifer recharge by the Soil Aquifer Treatment (SAT) process. Ammonium ($N{H}_4^{+}$) is one of the most nocive and common nitrogen compounds in wastewaters. In order to assess the risk of wastewater use for aquifer recharge,$N{H}_4^{+}$adsorption onto Souhil wadi soil sampled from the SAT pilot plant (Nabeul, Tunisia) was studied using laboratory columns experiments. Several experiments were conducted using aqueous synthetic solutions under different aqueous ammonium concentrations and flow rates. Furthermore, a real wastewater solution was used to test the effect of competitive cations contents on $N{H}_4^{+}$ adsorption. Afterwards, the Hydrus-1D model was used in inverse mode to simulate the ammonium transport through the Souhil wadi soil.

For the synthetic solutions, the adsorbed ammonium amount varied from 1 to 30.7 mg kg⁻¹ for aqueous ammonium concentrations between 4.9 and 36.4 mg L⁻¹. The linear isotherm model was found to be the most suitable for describing this adsorption. The flow rate decrease from 45 to 15 mL min⁻¹ induced an increase in the ammonium adsorption capacity by 49%. Indeed, the lesser the flow rate is, the longer the residence time and the higher the exchange between the aqueous solution and soil matrix. The use of wastewater instead of aqueous synthetic solution decreased about 7 times the Souhil wadi adsorption capacity of ammonium because of its relatively high concentrations of competitive ions such as calcium and magnesium.

The use of the Hydrus-1D model showed that the chemical non-equilibrium model was the best to simulate the ammonium transport through the laboratory soil columns.

Introduction

Ammonium is a common subsurface contaminant that may originate from diverse sources. The related contamination could be a result of seepage from landfills, foul sewerage systems and contaminated industrial sites or generated by an induced percolation, such as aquifer recharge with pretreated wastewaters and plants fertilizing. Furthermore, ammonium is often considered as a key indicator of leachate contamination arising from landfill sites and Soil Aquifer Treatment (SAT) process. So, it is commonly selected for consideration during the assessment of pollution risks to groundwater. It is a List II substance under both the Groundwater Directive (80/68/EEC) and the Dangerous Substances Directive (76/464/EEC) (Buss et al., 2004).

In many regions of Tunisia, treated municipal wastewater reuse for various purposes is a strategic alternative to overcome water shortage. Wastewater is generally reused for agricultural and golf fields irrigation, urban watering and groundwater artificial recharge (Kallali et al., 2005). This process, adopted to increase aquifer-based water supplies, uses underground formations as natural filters to finalize the purification of conventionally treated municipal wastewater prior to reaching the aquifer. It consists in favoring water vertical infiltration in constructed basins through several meters of vadose zone. In typical SAT process, recharge basins are flooded for 3–7 days, and then rested for 4–21 days to allow soil re-aeration and avoid its clogging (Torrens et al., 2009). However, an optimized design and management of SAT basins are needed to insure an efficient complementary purification of the treated wastewater and avoid fresh underground water contamination by nitrogen and organic pollution (Jellali et al., 2009). During treated wastewater infiltration in unsaturated subsoil layers, a significant quantity of ammonium is transformed biologically to nitrites and nitrates. However, infiltrating a large quantity of ammonium through low-thickness unsaturated zone raises the groundwater pollution magnitude (Bouwer, 2000, Repert et al., 2006). According to the Tunisian wastewater reuse norms (NT106.02), the maximum ammonium concentration allowed for indirect aquifer recharge is relatively low (1 mg L⁻¹). This concentration is very stringent considering the relatively low efficiency of some wastewater treatment plants of the country (Kallali et al., 2005).

Nitrogen removal from wastewaters has been widely studied using biological (Satoh et al., 2000, Stief et al., 2003, Yamaguchi et al., 1996) and physico-chemical methods (Ben Ali et al., 2004, Bodaloa et al., 2005, Chimenos et al., 2003, Karadag et al., 2008, Wu et al., 2006). Among these methods, ammonium adsorption removal techniques from wastewaters based on cation exchange are well established and relatively high effective under certain experimental conditions (Demir et al., 2002, Karadag et al., 2008, Sarioglu, 2005, Wu et al., 2006). However, the need of further experimental data under Tunisian conditions, concerning the quantification of ammonium adsorption in groundwater and the determination of the role of the porous media in attenuating this pollution, has been identified by several investigators (Ghrabi et al., 2001, Kallali et al., 2005). On the other hand, very few works over the world have modeled the ammonium transport in the field soils because of the difficulty in considering nitrification and denitrification processes in the developed numerical models.

Field and laboratory experiments with conservative tracers and reactive solutes in soils have shown that preferential and heterogeneous flow and non-equilibrium transport conditions often exist (Ashraf et al., 1997, Moradi et al., 2005, Pernyeszi et al., 2006, Suarez et al., 2007). As a consequence, traditional transport models have been found to be inadequate in many cases to explain the experimental results (Gaber et al., 1995). This has led to the development of models that include chemical and physical non-equilibrium in order to have better simulation results of solutes transport. Chemical non-equilibrium models (CNEMs) assume that adsorption is not an instantaneous process, but is time dependent and can be represented by a first-order kinetic term (Grathwohl, 1998). In physical non-equilibrium models the water phase is divided into mobile and immobile regions and the mass transfer to adsorption sites occurs mainly by diffusion which is time dependent.

Different degrees of complexity in modeling non-equilibrium solute transport are incorporated in some popularly models. Non-equilibrium convective dispersive equations taking into account of the physical and chemical non-equilibrium are implemented in some common numerical models such as Hydrus-1D (Simunek et al., 1998), Hydrus-2D (Simunek et al., 1999), Leachm (Huston and Wagenet, 1992) and HP1 (Jacques and Simunek, 2005). The use of these models permits the evaluation of the groundwater potential pollution by ammonium and the determination of the principal processes involved in its transport in the subsurface.

The objectives of the present work were: (1) to experimentally quantify the impact of ammonium aqueous concentration, flow rate and the presence of competitive cations on its adsorption onto an aquifer soil matrix sampled from a SAT pilot plant site, (2) to evaluate the non-equilibrium processes importance in the ammonium subsurface transport under the experimental conditions cited above.