Land characterisation for soil-based constructed wetlands: Adapting investigation methods to design objectives

SBCWs are systems located between wastewater treatment plants (WWTP) and receiving water bodies. The literature on these systems is quite abundant on their ability to enhance the water quality for major pollutants, pathogens (Ayaz 2008; Tzanakakis et al. 2009) and micro-pollutants (Li et al. 2014; Verlicchi & Zambello 2014). They have also recently gained an increased interest from wastewater treatment companies and stakeholders (Kampf et al. 2007; Blin et al. 2014; Pagotto et al. 2014).

SBCWs aim at protecting receiving water bodies, especially at minimum river flow periods. Indeed, according to the literature, SBCWs: (1) can be considered as buffer zones; (2) allow infiltration (decrease in discharged treated wastewater); (3) potentially improve the treated wastewater quality. Besides, the potential enhancement of biodiversity aspects seduces designers and contracting authorities, and partly explains the large development of SBCWs (in France, more than 500 SBCWs have been built, most of them in the last 5 years, Prost-Boucle & Boutin 2013).

A previous study on French SBCWs (Prost-Boucle & Boutin 2013) showed no clear link between their design and aims. Moreover, the sizing of these systems is not established and almost always corresponds to the remaining space available after the WWTP has been built. Hence, more research is needed to assess the efficiency of SBCWs as regards to their assigned objectives and to establish design and sizing methods.

SBCWs receive effluents of variable quality depending on the treatment performed upstream (effluent from primary, secondary or tertiary treatment) as influent. Water parameters regarding pollutants will evolve throughout the SBCWs depending on the different processes occurring. Pollutant removal takes place in one or several of the three components of SBCWs: (1) free water; (2) soil; (3) plants. Regarding SBCWs, the infiltration of treated wastewater is one of the main goals. Thus, soil is supposed to be the compartment where most of the treatment occurs. However, infiltration of treated wastewater may also change soil mechanical and chemical characteristics.

Water flow in soils is controlled by (1) porous media initial conditions (e.g., water content and temperature); (2) boundary conditions (inflow i.e. loading rate, and outflow – water table level), and (3) soils intrinsic hydrodynamic characteristics (e.g., pore size distribution, pore connectivity, matrix hydrophobicity). Unlike the hydraulic loading rates that can be controlled, water table level and rainfalls vary according to seasons, thus influencing the effective infiltration capacity.

Large influent particles are filtered at the filter surface, whereas small particles (e.g., colloids) may migrate within soils and sometimes be adsorbed to the solid matrix (Keller & Auset 2007).

Nutrients transported in the soil water phase due to treated wastewater loading promote growth of attached microorganisms (Thullner 2010). The biomass activity can be aerobic, anoxic or anaerobic. The biomass is responsible for degradation of macro-pollutants and of some micro-pollutants. However, the resulting biofilm build-up can reduce soil permeability.

Adsorption and desorption processes are ruled by equilibrium laws between solid and liquid concentrations of the pollutants. These laws also largely depend on physico-chemical parameters (e.g., matrix adsorption sites, temperature, redox, pH, chemical species characteristics).

The strong interplay between the aforementioned processes is obvious, and combinations of them lead to an increase or decrease in subsurface water quality. Two main characteristics define an efficient and sustainable SBCW: its ability to maintain the infiltration capacity and to enhance the influent quality. That said, two main phenomena could affect SBCWs operation: clogging and pollutants leaching.

Clogging matter (or biomat, Gette-Bouvarot et al. 2014) is the result of particle filtration and biological activity. The biomat formation leads to an increase in pore-water quality and to a decrease in soil infiltration capacity. It needs to be controlled to avoid clogging issues. Indeed, attached growth treatment systems like soil treatment units cannot be seen as systems undergoing stationary phenomena. These systems constantly evolve, and a proper kind of design and/or operation can lead to a resilient balance between pro and con effects of clogging, while ill-considered kinds may accelerate critical clogging.

Leaching is due to pollutants desorption from the soil matrix and can create a decrease in underground water quality. It can apply to adsorbed pollutants coming from the influent, as well as to pollutants coming from the soil matrix (most of them metal ions in natural soils). It is due to a change in the adsorption-desorption equilibrium, caused either by a change in redox or pH condition (e.g. due to bacterial activity), hydrolysis of organic matter, or an excess in water flow (Ollivier et al. 2013).

Soil is often considered as an efficient medium for wastewater remediation when using low hydraulic loading rates (fractions of the saturated hydraulic conductivity, Siegrist 2014). However, the sustainability of systems operated at higher loading rates remains uncertain. Systems using undisturbed soil for wastewater treatment are rather simple and cheap to build, but involve a large variety of interconnected processes. This makes soil-based systems a challenge for researchers.

In this paper, the authors propose an overview of available methods for the study of SBCWs: (1) studies required for design and building; (2) the choice of a comprehensive set of methods to be applied for their long term (3–4 years) study. These methods are presented thereafter, then classified and discussed. An original strategy is then proposed.

The first soil layers depth (and its spatial variation inside the chosen plot) needs to be determined, until reaching either a highly impermeable soil layer or the permanent (independent on seasons) water table. This can be proceeded through piezometers installation (and subsequent transient water table evaluation) and infiltration capacity tests. If an impermeable soil layer is found, the receiving water body has to be identified (e.g. a watercourse nearby the plot).

Methods detailed in this chapter were applied in Clos de Hilde WWTP, located in Bègles (33, France) near Bordeaux before the building of a set of semi-industrial scale SBCW pilots.

Identifying vadose zone soil structure has become significantly easier with the development of geophysical methods. Using these ‘non-invasive’ methods, the screening of the studied zone is less time consuming than traditional auger investigations. However geophysical signals do not only depends on the structure and the texture of the soil layers. It is therefore necessary to perform auger tests to identify precisely the nature of the soil layers and also decipher interactions (like biochemical reactions). Three geophysical methods were tested in association with auger tests: (1) Electrical Resistivity Tomography (ERT), (2) controlled-source electromagnetic induction method (EM) and (3) Ground-Penetrating Radar (GPR).

Auger test (Figure (2(a)) is a spot measurement that can be mechanical or manual. It gives access to the depth of the first soil layers and their textures.

ERT involves the measurement of a set of apparent resistivities, and the post-processing of these data allows the determination of a probable spatial-distribution of the resistive characteristics of soil. Soil electrical resistivity depends on lithology, pore fluid chemistry and water content (Binley & Kemna 2005) and therefore provides interesting information for both site description and the monitoring of transient phenomena (e.g., variation of pore fluid phenomena or water content).

EM measures the electrical conductivity averaged over a volume of soil. Both methods are complementary. EM allows performing measurements over larger areas but vertical variations in soil electrical conductivity are averaged. ERT can only be carried out on smaller areas but provides vertical cross-sections of the soil electrical resistivity.

GPR uses radar pulses to image the subsurface in a nondestructive way. The method uses electromagnetic radiation in the microwave band of the radio spectrum, and detects the reflected signals from subsurface structures/layers due to a spatial change in the soil dielectric constant. It permits high resolution data acquisition (about 5 cm in depth) between 0 and 1.5 m depth.

Constant head infiltrometry method (NF EN ISO 22282-5, AFNOR 2014) is a spot measurement that was used to determine the effective clear water infiltration capacity of soil layers.

5 m depth piezometers (following the French standard NF P94-157-2, AFNOR 1996) were installed at the borders of the plot. They were equipped with in line water level monitoring systems (STS DL/N) in order to follow the water table seasonal variation.

Soil samples were analyzed for Li, Mn, As, Cd, U, B, Fe, Se, Sn, Al, Co, Rb, Sb, Ti, Ni, Sr, Ba, V, Cu, Tl, Hg, Cr, Zn, Ag and Pb using Inductively Coupled Plasma Mass Spectrometry (ICP-MS). Soil samples were also analyzed for hydrophobic organic substances (ex. polycyclic aromatic carbons, alkylphenols, polybrominated diphenyl ethers, polybrominated biphenyls …) by GC-MS-MS or LC-MS-MS.

Figure 3 shows that two main soil layers could be found in the plot section: the first one is a backfill layer with a maximal 50 mm/h infiltration capacity, and the second one is a homogeneous clay layer with 2 mm/h infiltration capacity. In the case of high hydraulic loading rates (like in SBCW systems operation), this proved that most of the water flow will happen in the backfill layer. The water will be guided by gravity to lower heights parts of the backfill-clay interface.

Clogging and pollutant removal/leaching are impacted by water flow processes. Thus, two kinds of hydrodynamic parameters in variably saturated conditions will be monitored: (1) water content and (2) water pressure.

Methods aiming at evaluating clogging will be run at the SBCW scale.

Tracer tests with potassium bromide injection and ERT monitoring have been chosen as a nondestructive method for the infiltration dynamics assessment in ditch systems. The method allows acquiring a 3D image of the infiltration bulb at different time steps. The tracer tests will be performed at the very beginning of the SBCW operation and every 6 months throughout the project monitoring duration (4 years). As wastewater infiltration creates biomat, the infiltration bulb should grow more slowly as a consequence of a decrease in soil infiltration capacity (indirect clogging assessment).

At the same time and period, direct measurement of infiltration capacity (NF EN ISO 22282-5, AFNOR 2014) on SBCWs surface will be done. The main advantage of using ERT is assumed to be the capacity of localizing the clogged zone and estimating its extent.

Three methods will be used to assess soil pollutants leaching: (1) soil chemical analysis, (2) subsurface water sampling and chemical analysis and (3) piezometers water chemical analysis.

Soil chemical analysis (trace pollutants) has been used for the SBCWs pre-implementation study, and will be run from time to time inside SBCWs during their long-term assessment (SBCW scale).

Silicon carbide (SiC) porous cups (Figure 2(c)) will be used to sample subsurface water inside SBCWs (Figure 5). Both major and trace pollutants will be analyzed (SBCW scale).

Water from piezometers installed at the borders of the SBCWs site will also be analyzed in order to assess a potential broader pollutants contamination (site scale).

Table 1 summarizes the methods, results and aims of comprehensive SBCW pre-implementation study and long term assessment. A classification is proposed for methods only used for the pre-implementation study (in blue), for the long term assessment (in green), or for both (in orange).

Some of the methods (specific to the pre-implementation study) presented here give results that are redundant (EM, ERT, GPR), and were run in our case study in order to compare them (that is not the purpose of this communication).

The authors emphasize the importance of the combination of each type of methods (water flow and spatialized soil types measurements – data fusion) in order to gain information about three-dimensional water flow inside the system, and this before any SBCW building.

The knowledge of initial soil trace contaminants (organic and inorganic) is also a criterion regarding the potential operational risks (pollutants mobilization due to water infiltration) induced by soil-based SBCWs. Long term assessment of these parameters will allow us to evaluate the impact of SBCWs operation.

Design choice (trenches or meadows), footprint (in term of acceptable hydraulic loading rate), operation and placement of SBCWs inside the plot could be guided by results of these studies. However, as no validated approach exists so far, this remains a field of future research.

SBCWs such as meadows and ditches use soil for treated wastewater infiltration, thus questioning the fate of the infiltrated wastewater in term of quality and flow.

A detailed methodology for pre-implementation studies and long-term monitoring of these systems was presented. Special attention was given on processes occurring in SBCWs like pollutants removal/leaching and biomat formation. Perspectives of this work are to draw sustainable rules for SBCWs design and operation.

We emphasize that, in the near future, SBCWs should be conceived as a kind of soil treatment units.

The authors thank the Onema (the French National Agency for Water and Aquatic Ecosystems) for providing financial support to this work (BIOTRYTIS project, Bègles, France). We are also grateful to Bordeaux Métropole (33, France) for funding and support.