UC Berkeley

Constructed wetlands are nature-based treatment systems that offer an attractive means of removing nutrients and trace organic contaminants from effluent-dominated surface water and treated wastewater due to their low cost and ability to provide additional benefits. One recently developed nature-based treatment system, the horizontal levee, consists of a gradually-sloped subsurface flow treatment wetland built on the seaward side of conventional storm control levees. Native vegetation grows on its surface and an underground treatment layer consisting of sand, gravel, and woodchips, removes nutrients and trace organic contaminants from wastewater effluent. Although the mechanisms controlling nitrate removal in the horizontal levee have been previously reported, ambiguity about the mechanisms through which trace organic contaminants are removed in horizontal levees may hinder their implementation due to uncertainty about how design and operational variables impact their performance.

Efforts to improve upon the design of horizontal levees require insight into contaminant removal mechanisms. Often, studies of constructed wetland performance have been limited to monitoring the concentrations of contaminants flowing into and out of them (i.e., a black box approach). In contrast, in this research, a combination of high-resolution porewater monitoring, sampling of wetland vegetation and sediments, and microcosm experiments were used to assess the relative contribution of different removal mechanisms. Results of this research are germane to other subsurface flow treatment systems, to natural riparian zones and wetlands of ecological importance, and to new applications of the horizontal levee technology (e.g., the treatment of reverse osmosis (RO) concentrate from potable water reuse).

To develop a mechanistic understanding of trace organic contaminant removal in horizontal levees (Chapter 2), a demonstration-scale experimental horizontal levee that received nitrified, treated wastewater from an activated sludge wastewater treatment plant was monitored over the course of 2.5 years. High-resolution sampling of porewater, soil and plants was paired with isotherms, microcosms, and one-dimensional transport modelling to quantify the contribution of sorption, transformation, and plant uptake to the removal of trace organic contaminants. Results indicated that persistent trace organic contaminants were removed primarily through biotransformation and that their removal was linked to redox conditions. In particular, sulfate- and Fe(III)-reducing conditions facilitated rapid transformation of sulfamethoxazole and carbamazepine, two contaminants that have proven difficult to remove under aerobic and sub-oxic conditions. Transformation of relatively labile contaminants occurred in the presence and absence of nitrate in the wetland, implying that subsurface treatment systems that include a variety of redox conditions may effectively remove a broad range of trace organic contaminants via transformation reactions.

Although plant uptake and translocation did not contribute significantly to the mass balance of trace organic contaminants in the horizontal levee (Chapter 2), the presence of trace organic contaminants in plants may raise concerns about potential exposures to herbivorous and detritivorous wildlife and invertebrates. Results presented in Chapter 3 indicated that concentrations of the compound with the highest affinity to be taken up into the shoots of plants, carbamazepine, rapidly dropped in plants growing along the flow path and that concentrations varied widely among plant species. These results indicate that subsurface flow constructed wetlands could be designed to minimize plant uptake through careful selection of plants along the flow path or by minimizing the concentrations of trace organic contaminants in porewater. These design options may be particularly helpful if horizontal levees are used to treat wastewater that has higher concentrations of contaminants (e.g., RO concentrate generated by advanced wastewater treatment plants used as part of potable water reuse projects).

The plants in the horizontal levee grew under conditions similar to those found in natural riparian zones and wetlands. Therefore, monitoring results from the horizontal levee may provide insight into the fate of trace organic contaminants in ecosystems at the aquatic/terrestrial interface. Results suggested that uptake and translocation of carbamazepine may be a widespread and overlooked terrestrial exposure pathway because rivers in waster-stressed regions are often comprised primarily of wastewater effluent. Although riparian zones are small in area, they are disproportionately important for the survival of most terrestrial vertebrates, especially in arid regions. In addition, the concentrations found in plants in the horizontal levee may exceed levels that cause ecotoxicological effects. Considering these factors, research conducted in Chapter 3 indicated that there is a need to understand the extent of plant uptake as a terrestrial exposure pathway of wildlife to wastewater-derived trace organic contaminants.

Under the design and operating conditions employed at the experimental field site, the horizontal levee removes most contaminants from treated wastewater within approximately 20% of its length, implying that it may be able to process higher contaminant loads, such as those produced when RO is used to treat municipal wastewater effluent. Management of RO concentrate is a major barrier to the adoption of reuse in most water stressed regions due to a lack of access to suitable discharge infrastructure or receiving waters and the absence of low-cost and/or multi-benefit treatment systems. In Chapter 4, the ability of horizontal levees to remove nitrate and trace organic contaminants from RO concentrate was assessed. Results indicated that the horizontal levee removed nitrate and a diverse suite of trace organic contaminants via biotransformation at rates that exceeded those reported in other nature-based RO concentrate treatment systems on an areal basis. Because sulfate- and Fe(III)- reducing conditions were not reached until the last 5-7 meters of the system, persistent trace organic contaminants that are resistant to biotransformation under aerobic or anoxic conditions were not removed. However, if horizontal levees were designed to reach sulfate- and/or Fe(III)-reducing conditions when they are used to treat RO concentrate, they could provide an extremely effective means of removing nitrate and trace organic contaminants.