Haining, China

Haining No.2 drinking water treatment plant (DWTP) is one of two DWTPs in Haining, which are practicing the DOSCON coagulant control system since 2012. Cost reductions of approx. 15% are achieved.

The Haining drinking water treatment plant has early 2013 ordered Doscon also on one more plant.

The treatment scheme is as follows: Sequentially, inlet water goes to oxidation pretreatment, coagulation process, filter tank, ozone contact tank and carbon filter. For coagulation process, it consists of reaction tank and sedimentation tank, the sludge of tank bottom is removed into sludge restore pool.

Inlet of the plant is from Changshan River. The quality of inlet can vary much during summer season, because of the varied inlet water it is difficult for plant operator to adjust dosage. Especially in heavy raining time if the dosage cannot be increased in time, it always caused high turbidity in outlet of coagulation process, which resulted in high turbidity in effluent of the plant. While the quality of inlet of plant turn to normal, because of high dosage cannot be back to normal, it can cause lower turbidity than normal level, which resulted in coagulant waste.

The plant consists of two parallel treatment processes with same inlet water. DOSCON system control coagulant dosage in one treatment process.

The plant sensors are connected to controllers and the output signals from the controller are shared by the SCADA system of the WWTP and the DOSCON control box. The DOSCON control box is functioned as a ‘Modbus Slave’ to the main SCADA system. The dosing pump is controlled by the SCADA system using the information received from the DOSCON.

Conclusions

  • Except heavy rain time, DOSCON control system is successfully operated with well controlled optimal dose prediction in Haining No.2 DWTP.
  • DOSCON controller was installed in April 2012, and has achieved the expected treatment qualities under condition of coagulant saving.
  • According to current saving rate 8.6% in line 2, DOSCON can be expected to save 37 000 RMB/year on coagulant costs comparing to line 1. And the saving rate can be increased with increasing water quality data set and accuracy of the model.
  • The initial model is uncalibrated to periods of heavy rain. After collecting data for several periods of heavy rain, the performance is expected to improve.
  • A system based on multiple models was used to eliminate malfunctioning parameters from dose prediction.