Optimization of the wastewater treatment capacity of a short rotation willow coppice vegetation filter

Highlights

• Evaluating the transpiration ratio (α factor) helps estimate evapotranspiration.

• New α factor can be used to fine-tune irrigation rate, once crops are established.

• High treatment efficiency achieved even with an increase in irrigation rate.

• Matching irrigation with estimated evapotranspiration reduced deep percolation.

• Nitrogen leaching observed in late summer when the evapotranspiration rate dropped.

Abstract

The objective of this study was to determine the conditions to optimize the wastewater treatment efficiency of a short rotation willow coppice (SRWC) plantation (Salix miyabeana ‘SX67’) used as a vegetation filter to treat small municipal primary effluents (with less than 800 population equivalent). With the aim of maximizing the annual amount of wastewater treated, the effect of adjusting the hydraulic loading rate (HLR) according to the estimated evapotranspiration was tested at demonstration scale under humid continental climate conditions. We proposed a new method to calculate the evapotranspiration rate from plant physiological data, introducing an α factor based on direct transpiration measurements. This method increased the accuracy of the water balance, with a prediction of the crop coefficient (k_c) based on either an seasonal approach (R² of 0.88) or a monthly approach (R² of 0.94). This led to a more precise estimation of the pollutant loading reaching the groundwater and could be used after plantation establishment as a fine-tuning tool. Adjusting the HLR to that of evapotranspiration between May and October led to an annual increase of 2 mm/d (around 0.35 m³/m² per growing season) in HLR, while maintaining a pollutant loading removal efficiency of at least 96% for organic matter, 99% for total phosphorus and 93% for total nitrogen. A high HLR at the end of the season caused nitrogen leaching into groundwater, indicating that the HLR should be decreased in October, when willow growth is greatly reduced.

Introduction

Short rotation willow coppice (SRWC) used as a vegetation filter represents a wastewater treatment process that could be used by small municipalities (e.g. 300 to 800 population equivalent) (Dimitriou and Aronsson, 2011; Guidi Nissim et al., 2015; Hasselgren, 1998; Lachapelle et al., 2019). The efficiency of such treatment depends on critical parameters such as soil conditions, plant physiological characteristics and wastewater hydraulic loading rate (HLR).

Soil texture and composition affect how water percolates and becomes available to plants. Clay-based soils retain more water than sandy soils but being denser, they reduce plant growth by restricting root penetration and elongation (Lafleur et al., 2016). With their typically high level of organic matter and nutrient content, clay-based soils improve soil fertility (Havlin et al., 2013). Sandy soils, however, have a higher hydraulic conductivity, which improves water infiltration and aeration, favoring organic matter decomposition, but do not retain water and nutrients as well as clay-based soils (Van Veen and Kuikman, 1990).

The selection of plants for use in wastewater treatment by short rotation coppice vegetation filter depends on the desired goal, which can be to maximize (e.g. zero-discharge wetland) or minimize the evapotranspiration (ET) rate (in an arid climate) (Headley et al., 2012). Fast-growing willow shrubs can be a good choice for humid climate where precipitations are well distributed throughout the growing season and for which maximizing ET is desirable to minimize water discharge (Guidi Nissim et al., 2015; Jørgensen and Schelde, 2001). Cultivars of Salix miyabeana have been studied extensively and have exhibited good performance for environmental purposes in various projects in Eastern Canada (Guidi Nissim et al., 2015; Guidi Nissim et al., 2013; Guidi Nissim et al., 2014; Jerbi et al., 2014; Lachapelle et al., 2019; Mirck and Volk, 2010).

The ET rate affects wastewater treatment efficiency. A high ET rate leads to a reduction in the mass of contaminants discharged but, in some cases, may result in an increase in the pollutant concentration in the water treated due to the reduction of dilution water (Zhao et al., 2012). Setting aside runoff and capillary layer rise, the deep percolation (DP) can be calculated from the water balance, including irrigation (Irr) and rain, according to Eq. (1) (Allen et al., 1998):

$$\mathrm{DP}={\left(\mathrm{Irr}.+\text{Rain}\right)}_{\text{IN}}-\mathrm{ET}$$

The ET rate is often estimated by the Penman-Monteith equation combined with crop coefficients (k_c) measured for a given species watered as needed (Allen et al., 1998). The k_c curve provides the seasonal tendency of the plant ET rate, but it is recommended that local data be collected to consider specific cultural practices and regional pedoclimatic conditions (Pereira et al., 2015). Without specific on-site data, the uncertainty in water balance becomes high, increasing the risks of discharging high loads of pollutants to the groundwater. Thus, there is a need to develop a method to adapt the crop coefficient derived from the literature to on-site conditions.

To optimize wastewater treatment efficiency of a vegetation filter, the HLR is more important to consider than plant selection or soil type (Jonsson et al., 2004). A high HLR favors a higher rate of pollutant loading removal but may result in increased deep percolation (or runoff). A low HLR leads to an improved groundwater infiltration water quality, but a lower treatment capacity (Jonsson et al., 2004).

When the HLR is kept constant over the growing season, it will often result in an imbalance between the needs of plants for water nutrients and their availability throughout the growing season (Lachapelle et al., 2019). The HLR can be adjusted according to ET and meteorological conditions according to daily, seasonal and annual variations. Increasing the HLR may change soil water saturation and aeration, and its effects on groundwater water quality is not well understood. Changes in soil conditions can affect biological treatment (Havlin et al., 2013). Soil moisture sensors can be used to monitor the amount of water needed to maintain adequate conditions and modulate the influent flow rate (Cardenas-Lailhacar and Dukes, 2010; Romano, 2014).

Willow biomass yield (which is related to ET) varies depending on the number of years of plant growth and coppicing (Volk et al., 2011). Variations make the design more difficult to implement reliably, when it is not possible to measure ET extensively. A better knowledge of the expected ET, however, would improve the design reliability and predictability. The originality of this work is to suggest a new approach to help fine-tune the irrigation HLR over the years with in-field ET measurements.

The objective of this study was to determine the conditions optimizing the wastewater treatment capacity of an SRWC vegetation filter to treat municipal primary effluent wastewater. It also aimed to develop a method to adjust the crop coefficient from plant physiological data and to determine the consequences on the quantity and quality of the deep percolation water. Four HLRs were tested during a two-year demonstration scale project. The water balance and the water treatment efficiency were characterised for every loading rate over the two growing seasons to optimize the SRWC vegetation filter treatment efficiency.