The efficiency of various disinfection treatments against Legionella was tested on a hot water distribution system (HWDS) pilot unit. The results demonstrated clearly that most Legionella in the networks were fixed in the biofilm at the surface of the pipe (more than 98% for each loop). Chemical treatments (continuous chlorination, hyperchlorination, hydrogen peroxide and peracetic acid mixing) commonly used for the eradication of Legionella in hot water distribution networks appeared to be inadequate for eradicating the bacteria in the biofilm. Unfortunately, the biofilm contained most of the pathogens in an HWDS whereas legislation is only restricted to the Legionella concentration in the water phase. Thermal treatment appeared to be efficient to disinfect most of the biofilm but seemed to promote the biofilm re-growth as well. It was then concluded that the best solution to prevent Legionella contamination in hot water distribution systems would be to have perfect control of the temperature in the networks (temperature > 55 degrees C at all points). Nevertheless, in many cases it is difficult to have such control, so during the time necessary to modify networks, the best solution to control Legionella proliferation appears to be to apply a treatment shock (thermal or chlorination as a function of pipe characteristics). These treatments must be followed by a continuous chlorination that is totally controlled and equipped with alarm systems. This study demonstrates that biofilm sampling devices must be installed in hot water distribution systems to anticipate Legionella contamination and correctly determine the efficiency of the treatments.

The basic kinetic expressions developed in non-aerated biofilms (denitrification and RBCs) were verified on granular floating media in aerated filters. The limiting parameters of biofilms were verified on a small scale lab unit, using wastewater and media from full scale. It could be shown that the observed relationships corresponded to theoretical expectations and half-order kinetics as well as the significant constants were established. Maximum surface removal rates were measured as 1.7 g N-NH4/m2 d for nitrification, and the most limiting parameter was a half-order oxygen concentration close to saturation. Similar kinetics as on small scale were observed on a full-scale prototype of a new floating aerated biofilter, which can be used for both nitrification and denitrification, as well as complete nitrogen removal from settled wastewater in one reactor. The mixed biofilm nitrified 15 % less than pure tertiary treatment, and 20 % performance was lost between lab and full scale, due mainly to aeration limitations. Also, 35 % lower denitrification rates were observed between ethanol and sewage as carbon source, but addition of ethanol raised the full-scale performance to nitrate removal rates of 1,4 g N-NO3/m2 d. After pilot testing and full-scale demonstration, the process was implemented on several treatment plants, one of which was located in Denmark, to achieve nitrogen residuals below 8 mg TN/1. This plant uses 8 filters of 63 m2 to treat 12 000 m3/d, and the biological reactor occupies a surface similar to the existing conventional primary settler. The results of the first few months of operation are given: as long as the ratio between biodegradable COD and total nitrogen at the filter inlet is above 5, nitrogen removal of 80 % is achieved.

No abstract