This paper describes the hydraulic performance of a start-up, pilot-scale, horizontal sub-surface flow constructed wetland (CW), located outdoors at the Helmholtz UFZ, Leipzig. This paper aims to investigate the impact of the method of hydraulic calculation in a pilot-scale system. Impulse-response tracer tests were conducted at multiple depths and locations throughout the system and the uranine concentration was measured using a fluorometer. In addition, the volumetric flow rate was closely monitored and climatic data was gathered to support the hydraulic results. Werner and Kadlec's modified residence time distribution (RTD) theory (originally developed for systems with large flow rate and volume fluctuations) was applied and the results compared to those obtained using classic RTD theory.

Progressive uranine dispersion, broadening of the RTD base, a change in peak shape and extended tailing were observed with increasing distance. All of these factors indicated deviation from plug flow and mixing effects with low-to-moderate dead volume. As this was a non-steady flow system, application of modified RTD theory ensured that the first moments of the normalized breakthrough curves and RTD functions were always unity. The Student's t-test (95% confidence) showed that the outlet RTDs calculated assuming steady-flow were significantly different, but those determined using the modified theory were closely comparable.

In general, a decrease in flow rate from inlet to outlet was observed and fluctuations in the outflow were linked to climatic conditions. August was characterized by the highest temperatures, high global radiation and high rates of evapotranspiration. Low or no outflow was recorded in conjunction with high evapotranspiration. The lowest temperatures, low global radiation, low evapotranspiration and high humidity were recorded in October, as well as the second highest rainfall (82 mm) after June (115 mm). Surges in outflow were observed with rain events.