The physico-chemical behaviour of purified water turbidity in distribution networks

The groundwater treatment processes at Brabant Water are designed for the removal of iron, ammonium and manganese. These processes work perfectly. The water quality generally fluctuates around the stringent internal company standards. Nevertheless, the purified water still contains particles (turbidity).

The behaviour of these particles in the distribution network is important because it has an influence on the network’s fouling and on discoloured water complaints.

Behaviour and composition of particles from the treatment

The behaviour and composition of particles in the distribution network has an influence on the network’s fouling and complaints about discoloured water. The particles’ composition depends on the treatment process. For instance, after a deterioration in manganese removal, the fouling in the distribution network and the number of discolouration complaints increase significantly. But the special role played by manganese (both Mn2+ and MnO2) in the turbidity in the purified water and, subsequently, in the physical behaviour of the particles, is not fully understood.

With the new understanding there will be an extra incentive to examine whether or not to further optimise the existing treatment regarding water quality, because there will be more clarity about what needs to be changed in the water quality to reduce the number of discoloured water complaints and flushing intensity. This will possibly also shed new light on the company standards for turbidity in the purified water – these should perhaps be relaxed or tightened, or perhaps it depends on the Mn:Fe ratio in the purified water.

Experiments with different iron-manganese ratios

To begin with, batch experiments will be carried out on a bench scale, so as to gain an understanding of the relationship between the Mn:Fe ratio and the settling characteristics of the precipitation. Turbidity measurements will be taken at different time intervals. For every determination of the turbidity, a sample will also be taken to determine the total suspended solids (TSS), which provide a good measure of the total quantity of precipitated iron and manganese.

Depending on the results of the first bench tests, a decision will be made on the subsequent experiments to be conducted. These could for example include other water matrices, or the influence of parameters like calcium, magnesium, silicate, phosphate, bicarbonate and dissolved organic carbon (DOC).

In parallel to the bench tests, pilot-scale experiments will be conducted on-site by the water utility. These experiments are intended to acquire insight into the behaviour of particles in the distribution network, with a particular focus on the effects following dosing with manganese.


  • Understanding of the physico-chemical behaviour of turbidity (colloids, Fe and Mn) in the distribution network.
  • Understanding of turbidity, Fe and Mn threshold values for optimal composition of purified water in relation to settling behaviour, pipe fouling and discoloured water reports.
  • Understanding of the velocity required to prevent particles with different properties from settling under normal operational conditions (optimisation criteria self-cleaning networks).
  • New input for distribution modelling.