The antagonistic and synergistic effects of temperature during solar disinfection of synthetic secondary effluent

• We simulated solar disinfection of secondary effluent under controlled conditions.
• A Design of Experiments was implemented on an intensive poly-parametric study.
• The effects of time, temperature, intensity and initial population were evaluated.
• Temperature alters the disinfection kinetics in a dual, non-proportional manner.
• A temperature-dependent model was proposed to describe solar wastewater disinfection.

A 4-factor, multilevel, full factorial design of 240 experiments was performed in order to investigate the effect of temperature on the inactivation efficiency of spiked Escherichia coli in simulated solar disinfection of a synthetic secondary effluent. The initial population of the microorganisms was 103, 104, 105 and 106 CFU/mL, the exposure time 1, 2, 3 and 4 h, the treatment temperature 20, 30, 40, 50 and 60 °C and the sunlight intensity 0, 800 and 1200 W/m2. Radical changes in bacterial behavior, process efficiency and remaining populations were observed, while treating effluents in discreet temperatures. Elevating treatment temperature from 20 to 40 °C drastically impaired disinfection. Thermal inactivation with no regrowth predominated at 50 °C and total inactivation of microorganisms was observed at 60 °C in non-irradiated samples. Irradiation at 800 and 1200 W/m2 much increased inactivation efficiency, especially at 50 and 60 °C, proving sensitive light-temperature synergy at those temperatures. Total inactivation was achieved within 4 h under a range of treatment conditions, including all samples at 1200 W/m2, or 60 °C samples at 800 W/m2. Also, 99.9–100% efficiencies and final population below 1000 CFU/100 mL were obtained at 800 W/m2 and temperatures of 50 °C and above. Treatment time, temperature and intensity are the critical parameters for the disinfection process, while initial population is insignificant for removal efficiency. An explanation of the mechanism of the process as well as a general linear model predicting the outcome of the experiments is also suggested.

Figure optionsDownload as PowerPoint slide