Optimising pre-treatment and filtration

Surface water sources around the world are experiencing an increase in colour

caused by humic substances, which many experts say is a result of global climate

changes.

A granular filter bed for drinking water treatment with expanded clay aggregates

(Filtralite) as filter media is extensively tested in pilot scale experiments

at the Norwegian University of Science and Technology (NTNU ). The objective

of the work was to improve the direct filtration process for removal of humic

substances from water. The filter bed used consisted of a upper layer of 118cm

Filtralite NC (Normal density, Crushed) and a bottom layer of 60cm Filtralite

HC (High density, Crushed) with grain sizes of 1.6-2.5 mm and 0.8-1.6 mm respectively.

This filter bed is designed to give long filter runs (high storage capacity)

and good effluent water quality. For the results shown in the figure the raw

water had a colour level of 50 mg Pt./l and a turbidity of about 0.2 NTU. The

raw water was pH adjusted before coagulant addition and flocculation, with variable

detention times and a G value of 110 s^-1 in the tube flocculator. In the experiments

were polymer was used, it was added just after flocculation prior to filtration.

The production period for a filter is confined by a maximum allowed effluent

turbidity and/or a maximum head loss through the filter. In these three experiments

the predetermined maximum effluent turbidity was set to be 0,3 NTU, and all

the runs were limited by breakthrough (effluent turbidity exceeding maximum

value). The polymer used was a non-ionic high molecular weight polymer (Magnafloc

LT20).

Polymer efficiency

As can be seen in the figure the colour removal is high for all three runs.

There is a slight increase in colour removal for the two runs where polymer

is added, from 90 to 93% without and with polymer respectively.

When flocculating and adding polymer as filter aid, a major lengthening of

the run length and an improvement in effluent turbidity can be observed. This

is because the polymer makes the flocs bigger and stronger. The adding of long

chained polymers increase the efficiency of the adhesion in the filter, and

the storage capacity in the relatively coarse Filtralite filter is much better

utilized.

The run length increase significantly with an increasing detention time in

the flocculator. Longer detention times enhance the growth of flocs. Visual

observations during the experiments support this theory. For the experiment

with 5 minutes detention time in the flocculator and the addition of polymer,

large flocs could easily be observed above the filter media, for the shorter

detention time the flocs were smaller, and for the experiment with alum alone

(without flocculation), visible flocs were not observed.

The combination of a metal salt and polymer as filter aid with the Filtralite

filter have shown that it is possible to obtain long filter runs and good effluent

water quality when treating humic waters in a direct filtration plant.