In Japan various biological phosphorus removal processes have recently been researched by laboratory or pilot plant scale studies and most of them have shown good results. Based on these results, the Japan Sewage Works Agency has conducted a full scale study of the biological phosphorus removal process from June 1982 until February 1983, which was the first full scale operation of this process in Japan. The main purpose of the study was to evaluate phosphorus removal efficiency and also nitrogen removal efficiency of the process and in addition, to ascertain the important operating factors of the process.
For the study a treatment train of a large scale sewage treatment plant was remodelled. The aeration tank of 3.825 m3 volume was divided into four equal cells. The whole train including return sludge line was operated entirely independently of the other trains. During the experiment the train was operated under two different modes, Mode 1 and Mode 2. In Mode 1, the train was operated as an A/O process, the first cell of the aeration tank being anaerobic and the other cells oxic. In Mode 2, the train was operated as a Modified Phoredox process. In this case, the first cell was anaerobic, but the second cell was anoxic and nitrified liquor was returned to it from the end of the oxic cells. Mode 1 and Mode 2 were further divided into many ‘runs' and the flow rate varied between 12,550 m3 d−1 and 25,270 m3 d−1 , corresponding to retention times of 7.3 hours and 3.6 hours, respectively.
Throughout the experimental period the mean value of influent (primary effluent) total-P concentration was 3.38 mg 1−1 , and that of the final effluent was 0.47 mg 1−1 .
A cumulated frequency curve of the data showed that about 93% of measured effluent total-P was below 1.0 mg l−1 . Therefore, it can be concluded that with these influent total-P levels, biological phosphorus removal processes can sufficiently satisfy the effluent standard of 1 mg 1−1 total-P. Even when the process was operated as a Modified Phoredox Process, no obstruction to phosphorus removal because of nitrification was observed and phosphorus removal remained good.
However, since the sewage treatment plant treated influent from a combined sewerage system, phosphorus removal was sometimes affected by heavy rainfalls. In such cases phosphorus release in the anaerobic cell was insufficient because of increased influent NOx concentration and accordingly increased denitrification level in the anaerobic cell. Therefore, as a result, enhanced phosphorus uptake in the following cells could not be observed. Higher process stability can be expected if an effective countermeasure to high influent NOx concentration can be made.
Influence of flow rate fluctuation on the process was also studied. The treatment train was operated for a week under a daily flow rate fluctuation pattern which ranged between 460 m3 hr−1 and 820 m3 hr−1 . Nevertheless, the effluent total-P concentration showed no increase and stayed constantly lower than 0.5 mg 1−1.
The oxidation reduction potential (ORP) was an effective control index to evaluate the degree of phosphorus release in the anaerobic cell. Water temperature did not affect phosphorus release and uptake rates.
Recent advances in bio-P modelling – a new approach verified by full-scale observations