Efficiency of a small artificial wetland with an industrial urban catchment

Abstract

The efficiency of a small artificial wetland in Walter Park, in an industrial urban area in the City of Lake Macquarie, Australia, was examined by monitoring lead levels and suspended solids contents in the water. The location was adjacent to a main road and near a lead–zinc smelter. Samples taken from the sediment were subjected to selective chemical extraction procedures to investigate the behaviour of lead in the sediment and mechanisms of lead removal in the wetland. The wetland was relatively effective in removing lead and suspended solids, except for heavy flows. The removal of lead in the wetland was attributed to sedimentary processes, adsorption of lead on exchange sites and lead bound to the carbonate and oxidizable phases.

Introduction

Artificial or constructed wetlands for wastewater treatment are practical alternatives to the conventional treatment of domestic and municipal sewage, industrial and agricultural waste, stormwater runoff and acid mine drainage. In comparison with conventional secondary and advanced wastewater treatment systems, constructed wetlands may have several advantages. They are able to effectively remove wastewater contaminants thus reducing BOD, suspended solids, nutrients, trace organics and heavy metals with low costs of construction and maintenance, they have low energy requirements and, as a `low-technology' system, they can be established and run by relatively untrained personnel. The systems are usually more flexible and less susceptible to variations in loading rate than conventional treatment systems (Bastian and Hammer, 1993).

Bastian et al. (1989)indicated that under appropriate conditions, constructed wetlands have achieved high removal efficiencies (40–90%) for BOD, suspended solids, nutrients, heavy metals, trace organic compounds and pathogens from municipal wastewater. Wetlands accomplish water improvement through a variety of physical, chemical and biological processes operating independently in some circumstances and interactively in others (Hammer, 1993). Of these, physical and chemical reactions are thought to be the primary mechanisms for trace metals removal, with vegetation uptake secondary (Sereico and Larneco, 1988). Similar chemistries suggest possible removal of Ni, Cu, Pb, Zn, Ag, Au and U by constructed wetlands based on limited results from mining and industrial drainage. Removal rates are high, but application rates are low and available data is from pilot projects or wetland systems that have operated for only short periods (James et al., 1990). Long-term constituent removal by adsorption, absorption and some complexation reactions may be limited by saturation.

There are several possible physical and chemical mechanisms for metal retention by sediment: adsorption on cation exchange sites, sedimentation, precipitation and complexation with soil organic matter (Faulkner and Richardson, 1990). For volatile elements (e.g. mercury), loss to the atmosphere may be the dominant removal pathway (Sherwood and Reed, 1990). Heavy metals readily adsorb on particulates in the water column and in sediments (Jenne, 1968, Jenne, 1976; Hilderbrand and Blum, 1974; Fred, 1975; Florence, 1986; Faulkner and Richardson, 1990; Campbell and Tessier, 1991; Schindler, 1991). Clay minerals and organic matter, in addition to hydrous oxide materials, are always present in wetlands and all contribute adsorption sites. Adsorption is affected by pH, Eh, ionic strength, the nature of particles and the presence of competing cations, and thus will vary considerably between wetlands systems.

Metal ions in soil may be strongly bound as insoluble combinations with organic matter, especially to components of the humic fractions, particularly humic acid (Stevenson and Ardakani, 1972). Since a major proportion of sediment organic fraction consists of humic compounds, these will play an important role in the formation of insoluble complexes in wetland systems. The wetland environment also provides many sources of inorganic ligands which form complexes. Precipitated metals include metal hydroxides, carbonates, phosphates and sulfides. Under the reducing conditions common in wetland sediments, heavy metals rapidly precipitate as insoluble sulfides, whereas oxides and hydroxides will form under aerobic conditions. Biochemical reactions are considered secondary mechanisms for removal of heavy metals in wetland systems. These processes are attributed to microorganisms living on and around the macrophytes. The macrophytes remove pollutants by: (1) directly assimilating them into their tissue; and (2) providing surfaces and a suitable environment for microorganisms to transport pollutants and reduce their concentration (Meiorin, 1990; Shutes et al., 1993). Uptake by vegetation accounts for less than 1% of total metal removal (James et al., 1990; Eger, 1992).

The distribution of specific metal forms varies widely according to the chemical properties of the individual metal, pH, oxidation–reduction potential, the presence of complexing agents and loading rate. In general, predominantly-flooded wetland soils and sediments tend to favour stabilisation of trace metals as sulfide precipitates or by complexing with humic matter. However, Gambrell (1994)states that upon long-term oxidation, many wetland soils with an initial near-neutral pH may become sufficiently acid to cause metal release. Since heavy metals readily adhere to particulate matter, runoff water from heavily developed areas commonly contain approx. 100 ppm of suspended sediment which can settle out with as much as 0.3% by weight of lead in runoff from shopping centres, although much less from residential areas (Whippe and Randall, 1983). Such sedimentary processes play an important role in removing heavy metals in wetland systems (Sereico and Larneco, 1988).

Constructed wetlands are effective reservoirs for metals (Sereico and Larneco, 1988; Gearheart et al., 1990; Dunbabin and Bowmer, 1992). The capacity for wastewater purification by constructed wetlands is well documented (Gearheart et al., 1990; Gersberg et al., 1990; Hammer and Bastian, 1990). However, most evaluations of the use and efficiency of constructed wetlands are little more than input–output analyses. Very little is understood about the capacity of physical, chemical and biological properties within the `black box' (Wetzel, 1993). As constructed wetlands are used to meet inland surface water standards for metals (Bastian et al., 1989), it is anticipated that the data collected from an investigation of metals in the sediment will be more useful in defining their fate.

The fact that metals are complexed in many different ways in sediments has led to the suggestion that movement and behaviour of metals in aquatic systems can be explained better by determining the amount of the metals in various pools or fractions of the sediment rather than a measurement of the total amount present. As the result of this line of thought, several researchers (Gupta and Chen, 1975; Tessier et al., 1979; Vuorinen et al., 1988) have presented partitioning schemes for the fraction of metals in sediments based on a selective chemical attack of sediments into each of the following phases: exchangeable, carbonate, Fe–Mn oxides, organic matter and sulfides and residual fractions.

This present study concerned the behaviour of the heavy metal lead in the artificial wetland which received urban runoff water from the Walter Park catchment area, an area which is well known to be contaminated with lead (Crawford et al., 1976; State Pollution Control Commission and Environmental Audit of Lake Macquarie, 1983; Sommon and Trengove, 1989; Batley, 1991). The wetland is one of the water quality control structures constructed by Lake Macquarie City Council in an attempt to reduce pollutant loading in the receiving waters of Lake Macquarie. Therefore investigation of the behaviour of lead as a pollutant in the sediment of the wetland is important. In addition to a study of the sediments, the efficiency of the wetland was determined by monitoring lead concentration in the water flow. Data obtained should have some practical value in developing process performance and control strategies in a wetland system.