Laboratory scale experiments were performed in a non-continuous batch reactor system with 0.8-41.2 l domestic sewage exposed to constant light intensity, temperature and humidity. The treatment performance of duckweed (L. gibba)-covered sewage lagoons (DSL) was studied within a CODtotal range of 200-500 mg/l (113-294 mg CODfilt/l), in 10, 30, 70 and 95 cm deep reactors, and liquid mixing intensity (power dissipation) of 0, 0.3, 1.0, 2.3 and 34.1 W/m(3). The duration of each experiment was 20 days with biomass harvesting every 5 days. COD removal at extreme depths and extreme mixing intensities was compared with that in covered control reactors without duckweed. Removal of CODfilt did not differ in duckweed-covered and control reactors. The role of duckweed cover was marginal in changing the redox potential or the DO. COD removal lambda(r) (kg CODfilt/ha 20 days) correlated strongly with initial surface load lambda(s) (kg CODfilt/ha). Concentration removal (as mg CODfilt/l) was also proportional to initial CODfilt concentration. For a given CODfilt mass input. increasing depth up to 1 m affected DSL performance only by increasing surface load, and not by hampering oxygen transfer. Mixing (up to 2.3 W/m(3)) raised CODfilt removal. Therefore, at depths beyond 70 cm, moderate mixing is recommended. The first-order kinetic removal rate coefficient for CODfilt was 0.04-0.06 d(-1). (C) 2000 Elsevier Science Ltd. All rights reserved.
