Constructed Wetlands for Water Purification: Engineering Biology to Clean Water

What This Page Answers

The operator looking at constructed wetlands is usually solving one of three problems: complying with effluent discharge regulations for agricultural wastewater at a cost that does not consume margin; replacing a failing septic system on a rural property without connecting to a municipal sewer; or closing the water loop on a farm by treating and reusing greywater or process water rather than paying for licensed disposal. In all three cases, the question is the same: can an engineered biological system treat water to a legal or functional standard at a lower capital and operating cost than mechanical alternatives?

The answer, for flows below approximately 2,000 cubic metres per day, is reliably yes. Constructed wetlands have been used for municipal wastewater treatment since the 1970s and for agricultural applications since the 1980s. The International Water Association has published design manuals covering their performance. The US Environmental Protection Agency maintains case study databases with measured effluent quality data from hundreds of operating systems. The biology is not experimental; it is the engineering that determines whether a specific system meets its treatment targets.

This page covers the two primary constructed wetland types, the pollutant removal mechanisms, the sizing math, and the installation and operating costs a practitioner needs to evaluate whether a constructed wetland makes economic sense for their site. It connects to the broader water harvesting pillar because treated water becomes a reusable resource rather than a regulated waste stream, changing the economics of the entire farm water budget.

How Constructed Wetlands Clean Water

A constructed wetland is not a pond. It is an engineered bed of gravel or soil, planted with emergent macrophytes such as Phragmites australis (common reed), Typha latifolia (cattail), or Scirpus species, through which wastewater flows at a controlled rate. The treatment is accomplished by four simultaneous processes: physical filtration through the substrate, biological degradation by microbial communities attached to root surfaces and gravel, uptake of nutrients and some metals by plant tissue, and sedimentation of suspended solids. No energy input beyond gravity is required once the system is established and the inlet-outlet elevation differential is set correctly.

There are two primary system types with distinct hydraulic regimes. Free water surface (FWS) systems, also called surface flow wetlands, have open water above the substrate. Wastewater flows horizontally through the water column and over the substrate surface, exposed to air and sunlight. These systems perform well for BOD and suspended solids removal and for polishing secondary effluent, but they provide less nitrogen removal and carry a mosquito management burden in warm climates. Subsurface flow (SSF) systems route water through the gravel substrate, below the surface, where oxygen levels are controlled by substrate depth and flow rate. Horizontal flow SSF systems create anaerobic to anoxic zones that support denitrification; vertical flow SSF systems, where water is dosed intermittently from the top and drains downward, create aerobic conditions that support nitrification. Combining a vertical flow cell feeding a horizontal flow cell achieves simultaneous nitrification and denitrification, the configuration most relevant for agricultural effluent with high nitrogen loads.

The substrate material matters significantly for phosphorus removal, which neither plant uptake nor microbial activity addresses reliably at scale. Phosphorus binds to calcium, aluminium, and iron in the substrate through ligand exchange and precipitation. Standard gravel substrate has limited phosphorus sorption capacity. Systems designed for long-term phosphorus removal use blast furnace slag, zeolite, or other reactive substrates with high calcium or iron content. These materials cost more per cubic metre installed but extend the effective phosphorus-removal life of the wetland from 5 to 10 years on standard gravel to 20 or more years on reactive substrate (Kadlec and Wallace, Treatment Wetlands, 2nd ed., 2009).

Removal Rates, Sizing, and Cost Data

Measured performance data from operating constructed wetlands is the only reliable basis for system sizing. The US EPA Constructed Wetlands Treatment of Municipal Wastewaters Manual (2000) and the IWA Design Manual for Constructed Wetlands (2011) synthesise data from hundreds of systems. For subsurface horizontal flow wetlands treating domestic-strength wastewater (BOD5 200-400 mg/L), typical measured removal rates are: BOD5 80 to 95 percent, suspended solids 80 to 95 percent, total nitrogen 40 to 70 percent, total phosphorus 30 to 60 percent, fecal coliforms 1 to 3 log units. For free water surface systems treating secondary effluent (BOD5 20-40 mg/L), BOD removal is lower at 50 to 75 percent but pathogen removal can reach 3 to 4 log units in systems with adequate hydraulic retention time and sunlight exposure.

Capital costs vary by system type and local construction costs. European installed costs for subsurface flow constructed wetlands range from 50 to 200 EUR per cubic metre of daily treatment capacity (vault_atom_TBD: EU rural infrastructure cost registry). For a dairy farm producing 25 m3/day of parlour washwater, that is 1,250 to 5,000 EUR for the constructed wetland infrastructure, not counting liner, inlet distribution piping, or vegetation. Compare that to a packaged biofilm reactor for the same flow at 8,000 to 20,000 EUR installed, plus 2,000 to 5,000 EUR per year in electricity and maintenance. Over a 20-year operational life, the constructed wetland total cost is typically 30,000 to 60,000 EUR lower for this flow rate than a mechanical system of equivalent treatment performance.

Hydraulic loading rate is the primary design variable. Subsurface flow systems are typically sized at 0.04 to 0.08 metres per day hydraulic loading rate for the surface area of the bed. At 0.05 m/day loading, a 25 m3/day flow requires 500 m2 of bed area. At 30 cm substrate depth, that is 150 m3 of gravel. Gravel costs approximately 20 to 40 EUR per tonne delivered, at approximately 1,400 kg per cubic metre of bed volume: 150 m3 requires roughly 210 tonnes, or 4,200 to 8,400 EUR in substrate alone. Liner material for a 500 m2 bed at 1.0 mm HDPE adds approximately 2,500 to 4,000 EUR. Total installed system cost for this 25 m3/day farm scenario: roughly 15,000 to 30,000 EUR, including civil works, substrate, liner, inlet and outlet structures, and Phragmites planting stock.

Heavy metal removal in constructed wetlands is directly relevant to farms downstream of mining activity or those using copper-based fungicide programs. Measured zinc removal rates from constructed wetlands treating stormwater range from 50 to 90 percent depending on pH and substrate composition. Copper removal is similar: 50 to 85 percent in most studies (Vymazal, 2014, Ecological Engineering 68:52-60). These numbers make constructed wetlands viable as pre-treatment for water intended for aquifer recharge or pond storage, where metal accumulation in a closed-loop system would otherwise become a management problem over time.

Design and Operation at Farm Scale

The design sequence for a farm-scale constructed wetland follows five steps. First, characterise the influent: measure or estimate daily flow volume, BOD5, total suspended solids, total nitrogen, total phosphorus, and any target metals or pathogens. The treatment targets determine which system type and which substrate are appropriate. A parlour washwater stream with 800 mg/L BOD and 60 mg/L nitrogen requires different configuration than a domestic greywater stream at 200 mg/L BOD and 15 mg/L nitrogen.

Second, select the system type. For raw or primary-treated agricultural effluent with high BOD, a two-stage system is standard: a primary treatment pond or septic tank to remove gross solids, followed by a subsurface flow wetland for secondary treatment. Skipping primary treatment loads the wetland inlet zone with material that causes rapid clogging of the gravel substrate, reducing hydraulic conductivity and treatment performance within two to three years. Third, size the beds. The k-C* first-order area model from Kadlec and Wallace is the most widely validated sizing approach for treatment wetlands. Area (m2) equals Q times natural logarithm of (Cin minus C*) divided by (Cout minus C*), divided by k (the first-order rate constant for the target pollutant at the design temperature). Rate constants and background concentrations for most common pollutants are tabulated in the IWA manual.

Fourth, set out and excavate. Bed slope is typically 0.5 to 1.0 percent from inlet to outlet to ensure drainage without short-circuiting. Install a robust HDPE liner, 1.0 mm minimum thickness, with a 30 cm overlap at all seams. Place inlet distribution pipe (perforated, 100 mm diameter) across the full width of the bed inlet, connected to the primary treatment outlet. Place substrate in two layers: coarse gravel (20-40 mm) at base and finer gravel (10-20 mm) at the surface, each 15 cm deep. Install outlet collection pipe at bed base, connected to an adjustable weir in an external control chamber. The weir height sets the operating water level; typically 5 to 10 cm below the gravel surface for horizontal flow SSF.

Fifth, plant and establish. Phragmites australis is the most widely used species globally for constructed wetlands, tolerating wide influent strength ranges and surviving periodic dry periods. Plant at 4 to 6 rhizomes per square metre in spring or autumn. The system operates at reduced efficiency during the first growing season while roots penetrate the substrate. By the second growing season, a well-established Phragmites bed with dense root mass is operating at near-design efficiency. Vegetation harvesting removes a small fraction of nitrogen and phosphorus taken up by plant tissue, typically 5 to 15 percent of the nitrogen load and 10 to 20 percent of the phosphorus load per annual harvest cycle. Harvested biomass has a C:N ratio of approximately 40:1 to 60:1 and is suitable for mulch, thatch, or composting.

Operating the system is straightforward. Monthly checks: confirm flow is distributed evenly across the inlet width, check outlet weir level, observe surface for ponding that indicates substrate clogging. Annual: harvest vegetation to above-ground biomass 15 to 20 cm above the substrate surface, clear any inlet distribution pipe obstructions. Every five to ten years: rest the inlet zone by redirecting flow while the clogged substrate zone recovers biologically; in worst-case scenarios, replace the top 10 cm of inlet-zone gravel.

Where Constructed Wetlands Fit in Water Infrastructure

A constructed wetland is not a standalone system; it is a processing node that converts a liability (regulated wastewater) into an asset (treated water suitable for reuse). The output from a well-operated farm-scale constructed wetland meeting BOD below 30 mg/L and total nitrogen below 15 mg/L is typically suitable for irrigation reuse in most European and US jurisdictions, subject to local regulation. That reuse potential changes the economic calculation: instead of paying for disposal of process water, the farm recycles it to crops or pasture, reducing freshwater demand and irrigation infrastructure costs.

In the water harvesting stack, constructed wetlands connect most directly to managed aquifer recharge: treated wetland effluent that meets groundwater recharge standards can be directed to spreading basins or infiltration galleries rather than discharged to surface water. This closes the farm water loop entirely in jurisdictions where groundwater rights are attached to the recharging land. The treated effluent also pairs well with farm pond storage: a constructed wetland treating parlour washwater, with output directed to a storage pond for dry-season irrigation, is a complete greywater recycling system that eliminates both the disposal cost and the dry-season water purchase cost.

The connection to rainwater harvesting tanks and cisterns is through blending: in some climates, treated greywater blended with captured rainwater at appropriate ratios can meet irrigation quality targets that neither source alone would achieve. The nitrogen and phosphorus retained in constructed wetland output, typically 5 to 15 mg/L each after treatment, functions as a dilute fertigation solution when applied to land, reducing synthetic fertiliser requirements in a way that registers directly on the input cost line.

The role of constructed wetlands in regenerative aquaculture is well established: treatment wetlands are used upstream of fish pond systems to control organic loading and nitrogen, preventing the algal blooms that deplete dissolved oxygen in fish ponds. In integrated farm systems combining livestock, constructed wetlands, and aquaculture, the wetland functions as the nutrient management interface between the livestock waste stream and the pond ecosystem, allowing the farm to extract value from both without exhausting the pond system's biological capacity.

The practical constraint on constructed wetland adoption is not technical performance but land availability and regulatory approval time. Systems treating more than a few cubic metres per day typically require discharge consent or reuse permit in most jurisdictions, with application processes taking three to eighteen months. The planning horizon means constructed wetlands are most viable for operators building new infrastructure or replacing end-of-life systems, rather than as emergency compliance fixes. The operator who builds the system into a new dairy shed design from the start, sizing the bed as part of the total water management infrastructure, captures the full lifecycle cost advantage over a mechanical system installed reactively.