Non-pumping reactive wells filled with mixing nano and micro zero-valent iron for nitrate removal from groundwater: Vertical, horizontal, and slanted wells

Highlights

• Bench-scale NPRWs filled by micro/nano ZVI to ${{\mathrm{NO}}_3}^{-}$ reduction are studied for the first time.

• Linkage between the analytical results and laboratory experiments are also established.

• Slanted NPRWs have a higher ${{\mathrm{NO}}_3}^{-}$ reduction rate in comparison with vertical and horizontal.

• Hydraulic conductivity, mass of ZVI, and orientation of NPRWs affected nitrate removal rate.

• Horizontal NPRWs configuration is applicable for removing high risk pollutants.

Abstract

Non-pumping reactive wells (NPRWs) filled by zero-valent iron (ZVI) can be utilized for the remediation of groundwater contamination of deep aquifers. The efficiency of NPRWs mainly depends on the hydraulic contact time (HCT) of the pollutant with the reactive materials, the extent of the well capture zone (W_cz), and the relative hydraulic conductivity of aquifer and reactive material (K_r). We investigated nitrate removal from groundwater using NPRWs filled by ZVI (in nano and micro scales) and examined the effect of NPRWs orientations (i.e. vertical, slanted, and horizontal) on HCT and W_cz. The dependence of HCT on W_cz for different K_r values was derived theoretically for a homogeneous and isotropic aquifer, and verified using particle tracking simulations performed using the semi-analytical particle tracking and pathlines model (PMPATH). Nine batch experiments were then performed to investigate the impact of mixed nano-ZVI, NZVI (0 to 2 g l⁻¹) and micro-ZVI, MZVI (0 to 4 g l⁻¹) on the nitrate removal rate (with initial ${{\mathrm{NO}}_3}^{-}$=132 mg l⁻¹). The NPRWs system was tested in a bench-scale sand medium (60 cm length × 40 cm width × 25 cm height) for three orientations of NPRWs (vertical, horizontal, and slanted with inclination angle of 45°). A mixture of nano/micro ZVI, was used, applying constant conditions of pore water velocity (0.024 mm s⁻¹) and initial nitrate concentration (128 mg l⁻¹) for five pore volumes. The results of the batch tests showed that mixing nano and micro Fe⁰ outperforms these individual materials in nitrate removal rates. The final products of nitrate degradation in both batch and bench-scale experiments were ${{\mathrm{NO}}_2}^{-}$, ${{\mathrm{NH}}_4}^{+}$, and N₂(gas). The results of sand-box experiments indicated that the slanted NPRWs have a higher nitrate reduction rate (57%) in comparison with vertical (38%) and horizontal (41%) configurations. The results also demonstrated that three factors have pivotal roles in expected HCT and W_cz, namely the contrast between the hydraulic conductivity of aquifer and reactive materials within the wells, the mass of Fe⁰ in the NPRWs, and the orientation of NPRWs adopted. A trade-off between these factors should be considered to increase the efficiency of remediation using the NPRWs system.

Introduction

Nitrate is a universal, ubiquitous groundwater pollutant that constitutes a major health risk to humans and a burden on the environment (Gandhi et al., 2002). This widespread contaminant leaches into the groundwater from anthropogenic sources (e.g. nitrogen fertilizers, animal waste, and septic systems), irrigation and runoff from farmlands. Among common denitrification technologies (i.e. biological and chemical reduction) (Della Rocca et al., 2007), the abiotic chemical reduction of nitrate by zero-valent iron (ZVI) has been effectively implemented (Ludwig and Jekel, 2007). Chemical reduction of ${{\mathrm{NO}}_3}^{-}$ by nano-scale ZVI (NZVI) has gained its application since 1990s (Tratnyek et al., 2003; Grieger et al., 2010), although ${{\mathrm{NO}}_3}^{-}$ reduction by Fe⁰ was reported earlier by Young et al. (1964). One of the well-known approaches for in-situ nitrate treatment in groundwater is emplacing the ZVI particles in permeable reactive barriers (PRBs), with different configurations, namely continuous trenches, funnel-and-gate or reactive vessel (Day et al., 1999; Hosseini et al., 2011a). These configurations are typically limited to groundwater depths less than approximately 20 m (United States Environmental Protection Agency, USEPA, 1998), and consequently cannot be utilized for the remediation of deep groundwater contamination (Freethey et al., 2002). Arrays of non-pumping reactive wells (NPRWs) have been studied as an appropriate alternative for application to deeper aquifer systems (Wilson et al., 1997; Hosseini and Tosco, 2015).

The construction of NPRWs arrays is a promising and cost-effective alternative emplacement method for contamination plumes deeper than 50 m (Wilson et al., 1997; Puls et al., 1999; Wilkin et al., 2002). The efficiency of NPRWs filled by ZVI in treating a groundwater contaminated plume is controlled by several key aspects, including aquifer hydrology (e.g. flow velocity, prevalent flow direction, hydraulic conductivity, porosity, depth and fluctuations of water table, and heterogeneity), groundwater geochemistry (e.g. concentration of pollutants, geometry and depth of contaminated plume, pH, dissolved oxygen, type and concentration of dissolved salts), configuration of NPRWs (e.g. diameter and length of wells, space between wells, design of screen), characteristics of the reactive material (e.g. particle size, Fe⁰ content, eventual surface modifier), contact time (or residence time) needed for the reaction between contaminant and reactive agents, and difference in hydraulic conductivity between the reactive material and the aquifer medium (Wilson et al., 1997; Freethey et al., 2002; Painter, 2004; Hosseini et al., 2011b; Hosseini and Tosco, 2013).

The installation of horizontal and slanted wells has become popular since the 1930s, in the petroleum industry, and recently for the recovery of contaminated groundwater, collecting non-aqueous phase liquids and soil vapor from the subsurface, drainage, and mine dewatering (Hantush and Papadopulos, 1962; Tsou et al., 2010). Bearing in mind the higher cost of drilling horizontal wells, they have some advantages over vertical wells: their geometry improves the contact with contaminated groundwater, and they are suitable to install in thin aquifers (Zhan and Zlotnik, 2002); horizontal and slanted wells produce less pronounced drawdown cones, and are more suitable for groundwater flow with significant vertical velocity component and pronounced fluctuations of the water table (Morgan, 1992; Kompani-Zare et al., 2005; Huang et al., 2011); drilling operations of horizontal and slanted wells are a more doable option when surface structures (e.g. buildings) are present at the surface. Furthermore, due to the significant advances of the directional drilling technology over the last two decades and the cost reduction in their installation procedures (Liang et al., 2016), interest in the application of horizontal and slant pumping wells has been reignited. To date, the horizontal and slant pumping wells are commonly installed in shallow aquifers to withdraw a large amount of groundwater (Bear, 1979) or to remove a large amount of contaminant (Sawyer and Lieuallen– Dulam, 1998). Readers are encouraged to refer to Yeh and Chang (2013) for a recent and comprehensive review of groundwater flow hydraulics modeling of the horizontal and slant pumping wells. To the best of our knowledge, in spite of the aforementioned advantages of slanted or horizontal wells, no previous work could be found that investigates the efficiency of these wells for the emplacement of the reactive materials for groundwater remediation.

There are a very limited number of studies that investigate the efficiency of vertical NPRWs as reactive barriers. High removal rate of chromate from contaminated groundwater at North Carolina (Puls et al., 1999) and chlorinated hydrocarbon compounds at Denver Federal Center, Colorado (Wilkin et al., 2002) are reported for iron-filled NPRWs arrays. Mixtures of bone-char phosphate and iron oxide were deployed in arrays of NPRWs at two sites: Christensen Ranch In-Situ U Mine, Wyoming and Fry Canyon, Utah (Naftz et al., 2002). Initial U removal efficiencies exceeded 99.9% during a 7-month deployment period at the Christensen Ranch site. In a previous study, Hosseini and Tosco (2015) reported a successful application of biochemical remediation of a nitrate contaminated bench-scale aquifer by emplacing a combination of NZVI and carbon substrates in an array of vertical NPRWs.

The motivation of this study is to answer these questions: are the aforementioned advantages of horizontal and slanted pumping wells also extendable when these wells serve as non-pumping reactive wells to remediate nitrate contaminated groundwater? Precisely, how would the progress of the denitrification process through NPRWs be influenced by different well orientations in the presence of nano and micro Fe⁰ particles as reactive materials?

To answer these questions and to gain an understanding on the applicability of slanted NPRWs systems, connection between hydraulic parameters of aquifer, chemical reduction of ${{\mathrm{NO}}_3}^{-}$ by mixing NZVI and MZVI, and orientation of NPRWs should be analyzed. In this study, nitrate removal process was investigated in vertical, horizontal, and slant NPRWs filled by mixing nano/micro ZVI positioned in a homogeneous and isotropic bench-scale sand medium. Batch experiments were also performed to investigate the effect of mass of ZVI particles, type of reagents (i.e. micro or nano or a mixture of both) on the denitrification process under constant initial nitrate concentration. The main focus of bench-scale experiments was to modify the contact time of nitrate and reactive materials by changing NPRWs orientation, whereas other factors (e.g. initial nitrate concentration, mass of Fe⁰, pore water velocity) were considered constant.

This paper is organized as follows (Fig. 1): we first consider the hydrodynamic aspects related to the NPRWs, presenting analytical equations for computing the hydraulic contact time and capture area for a NPRW in three orientations (vertical, horizontal, and slanted with inclination angle of 45°) (2.1 Hydraulic contact time and area of well capture zone computations, 3.1 Results of hydraulic contact time computation). Homogeneous and isotropic aquifer and steady-state flow condition are assumed. The effect of relative hydraulic conductivity between aquifer material and reactive media is then investigated. Batch experiments are performed to investigate the effect of mixing two reagents NZVI (0–8 g l⁻¹) and MZVI (0–16 g l) on the nitrate reduction and nitrogen byproducts (2.2 Batch experiments, 3.2 Results of batch experiments). Kinetic analysis of nitrate reduction and ammonium stripping and production are also conducted based on batch experiments (Section 2.2). The best mixing ratio of micro and nano ZVI for nitrate reduction obtained in batch experiments is selected to implement NPRWs systems with vertical, horizontal, and slanted orientations of wells at a bench-scale (2.3 Bench scale experimental setup, 3.3 Nitrate removal through NPRWS).