Élimination des phosphates par filtration directe sur lit de sable Removal of phosphates by direct filtration on sand bed

Résumé

La présente étude porte sur la réduction des phosphates d'une eau synthétique à faible turbidité par filtration directe sur lit de sable. L'effet de plusieurs paramètres (vitesse de filtration, dose de sulfate d'aluminium et intensité d'agitation) sur la performance du procédé a été étudié. Des efficacités de rétention de phosphore supérieures à 80 % ont été obtenues pour des vitesses de filtration de 5 et 10 m/h. La filtration en ligne est applicable à 5 m/h mais devient beaucoup moins efficace à 10 m/h. Le rapport massique Al/P=2 est optimal et indépendant de la vitesse de filtration ainsi que de l'intensité d'agitation dans le floculateur.

Mots-clés:

• Pollution phosphorée,
• déphosphatation,
• vitesse de filtration,
• perte de charge,
• rapport massique Al/P,
• gradient de vitesse

Abstract

Phosphates and nitrates are the main nutrients responsible for disturbing aquatic ecosystem balance. In general, high levels of phosphate and nitrogen cause increases in algal proliferation, which has negative effects on water quality - the phenomenon is known as eutrophication. While nitrogen is in great quantities in the atmosphere, the sources of phosphorus are limited to two main sources: domestic waste waters and chemical fertilizers. Phosphorus is often known as the limiting factor for the growth of the algae in the aquatic areas.

The usual biological processes (activated sludge) are not very efficient in the removal of phosphorus from waste water (efficiency of less than 50%). Alternative biological processes require certain conditions that are difficult to create in practice. For that reason, we thought it useful to explore another alternative, based on a combination of the precipitation-flocculation of phosphorus by metallic cations brought by alum (aluminium sulfate) and direct filtration through a sand bed.

The aim of this work was to study the effect of filtration velocity, mass ratio (Al/P), and velocity gradient (G) on phosphorus removal efficiency in the flocculator. To this effect, a pilot-scale unit was set up. It consisted of one glass column (2 m long, with an inner diameter of 30 mm) that was filled with sand (0.85 m). The volumetric mass of sand and the porosity of the bed were 2.62 g/cm³ and 0.44 respectively.

The synthetic water used was obtained by dissolution of potassium dihydrogen-phosphate in tap water. The characteristics of this water were as follows: turbidity (varied from 0.5 to 1 NTU); pH 7-8 ; electric conductivity of 450 -550 µS/cm; temperature fixed at 25 °C; a phosphorus concentration 2 mg/L. The pH of coagulation was adjusted to 6.5 and the flocculation time was 10 min. The filtration cycle was arbitrarily fixed at 9 h. The first sample was taken 15 min after the start of the experiment and subsequent sampling was carried out every hour. Once the filtration phase was terminated, the filter was back-washed (combined air-water). The efficiency of phosphorus removal by direct filtration was determined by measurements of residual phosphorus, final turbidity and total head loss in the bed.

The results obtained show that :

- Phosphorus removal efficiency is strongly related to the filtration velocity.

- A mass ratio Al/P=2 seems optimal for the two velocities tested (5 and 10 m/h). For the ratio Al/P=3, we can see an increase in head loss and an increased turbidity in the filtered water, without any improvement in phosphorus retention. The increased turbidity is caused by the formation of fine particles of aluminium hydroxide and/or aluminium orthophosphate.

- When the filtration velocity is 5 m/h, the mixing intensity (G) in the flocculator does not increase the removal efficiency of phosphorus. The efficiencies at G=0 (in line filtration) and G=50 s^-1 are similar and development of head loss is essentially the same in both cases. Mixing in the flocculator becomes necessary when the filtration velocity is 10 m/h. In line filtration is much less efficient.

- Intense mixing (G>100 s^-1) had no beneficial effect on phosphorus removal, but rather led to a decrease in filtration time.

Keywords:

• Phosphorus pollution,
• eutrophication,
• phosphorus removal,
• filtration velocity,
• head loss,
• mass ratio Al/P,
• velocity gradient