Adaptation of two-stage vertical flow constructed wetlands for treating unscreened sewage in the UK.

The UK water industry has made a priority of embracing low energy sewage treatment, making of two-stage vertical flow constructed wetlands (VFCWs) a plausible alternative to conventional systems. However, for successful implementation of the technology in the UK a further understanding of any required adaptations in its design and operation is required, in particular linked to the UK use of spot sampling regulation of treated effluents and differences in rain profile affecting hydraulic loads in combined sewers. There is also a paucity of information regarding required time for full maturation of the technology and of performance during early operation. In response, the first UK full-scale WWTP based on two-stage VFCWs was built in Derbyshire, UK, and closely monitored during the initial three years of operation, allowing investigation of the system performance and robustness. The efficacy of the system for removal of total suspended solids (TSS), BOD₅ and NH₄ +- N has been evaluated by composite and spot samples analysis. The technology was found to efficiently remove pollutants to the level of 6 ± 3 mg·L⁻ ¹ , 5 ± 2 mg·L⁻ ¹ and 5.8 ± 3.8 mg·L⁻ ¹ for TSS, BOD₅ and NH₄+- N, for 24 hour composite samples. After one year of operation the system was retrofitted with forced aeration on the second stage, which enabled a higher ammonia removal with concentrations in final effluent of 95th percentile concentration of 5.3 mg∙L⁻ ¹comparing to 9.6 mg∙L⁻ ¹achieved in the system with passive aeration. The first stage filters were observed to be hydraulically limited during the first year of operation, with a hydraulic conductivity (HC) of <5.7× 10⁻⁶ m∙s⁻ ¹ which caused prolonged ponding occurrence. This in turn led to poor re-oxygenation of the first stage filter beds and low mineralization of the sludge layer deposited on its surface, which was characterized by TS and VS content of 11.5% ± 1.9% and 64.5% ± 7.3% respectively. A more mineralised sludge layer was present after three operational years, with HC <4.9 × 10⁻⁴ m∙s⁻ ¹ and sludge properties of 20.4% ± 3.5% TS and 62.3% ± 3.5% VS. These values are comparable to data reported for two-stage VFCWs operated in France and ponding was minimal after three years, suggesting this as the required length for system maturation. Close monitoring of the system during the first three years of operation evidenced the need to adapt rotating period lengths according to system age and sludge layer properties. Resting periods of up to 14 days were required for the young system to encourage sufficient hydraulic conductivity of the sludge layer, while the conventional 3.5 days feed/ 7 days rest cycle is appropriate after three years of operation. The influence of sludge properties on permeability was assessed in laboratory experiments, as well as artificial modifications of the sludge surface that could sustain increased permeability and accelerate sludge layer mineralisation. A positive impact of increased VS on permeability was found, with an increase in VS from 60% to 75% improving permeability from 6.14 × 10⁻ ¹⁷ m∙s⁻ ¹ to 2.00 × 10⁻ ¹⁶ m∙s⁻ ¹ in samples with TS of 9% and from 2.06E × 10⁻ ¹⁷ m∙s⁻ ¹ to 1.25 × 10⁻ ¹⁶ m∙s⁻ ¹ in samples with TS 15%. A greater relative contribution to permeability of preferential flow pathways (PFPs) present in the sludge layer than of sludge properties (TS, VS) was observed. PFPs were scarce in the young full scale primary VFCWs, which was believed to be the reason for their poor hydraulic acceptance. A positive effect on PFP formation and sludge drying rate was observed when artificially modifying the sludge layer by application of secondary activated sludge or by making initial cuts on its surface. Artificial modifications of the sludge layer could be implemented in full-scale systems to improve hydraulic acceptance and re-oxygenation, to shorten maturation of primary filters and to maintain prolonged operational periods without desludging.