Nitrifikation i biobädd - Puls och stegsvarsexperiment utförda på Sjölundaverket i maj 1997

Report, 1997

In May 1997 a number of experiments were carried out on a pilot scale nitrifying trickling filter: one pulse response at a low flow, one pulse response at a high flow, a step increase in influent ammonium concentration at the higher flow, and a step increase in flow from the lower flow to the higher one. Samples were taken with short intervals from both the influent and the effluent. The pulse responses show that the mean residence time does not increase inversely proportional to the flow, which can be explained by an increase in the volume of bulk water with flow. Comparisons between data and simulations of model that has been derived to simulate the fast dynamics of nitrifying trickling filter indicate that the flow is not laminar. The step response experiments gave smooth responses in the measured effluent concentrations. The transients were similar for both step experiments. However, the effluent ammonium concentration increased faster in the experiment where the flow was increased. No overshoot in effluent nitrite concentration could be observed, something that could be expected if the increase in nitrite oxidation was slower than the increase in nitrification. Simulations, using the developed model, and data from corresponding experiments carried out on a very similar plant at Ryaverket in Göteborg closely agree. So does the pulse responses from Sjölunda, but the observed step response in effluent ammonium concentration was significantly slower than the model predicts. Further, even though the influent concentration of total dissolved nitrogen remained the same during the experiments the effluent concentration dropped after the step was introduced and reached the initial value again after approximately 30 minutes. This indicates an accumulation of nitrogen in the filter. A possible explanation is adsorption of ammonium on/in the biofilm. This phenomena was not observed in the experiments at Ryaverket. Reasons for this may be found in differences in water quality. The water at Sjölunda is much harder, has a higher alkalinity and a higher pH.