Powdered activated carbon (PAC) is a well proven technology for removal of organic
compounds from drinking water including natural organic matter (NOM), taste and odor
compounds and pesticides (Karanfil et al., 1999; Treguer et al., 2006). PAC is
normally added during the rapid mix stage along with the coagulant during coagulation so that
the PAC can be incorporated into the floc and be removed along with the sludge. However,
while this dosing regime enables the PAC to be removed with the floc aggregates, there is a
significant reduction in the surface area available for adsorption when the PAC is contained in
the floc and there is direct competition between the PAC and coagulant for the NOM being
removed.
Alteration of the dosing procedure has the potential of providing a significant improvement in the
overall NOM removed and the clarified turbidity. This paper focuses on a real operational problem
at a United Kingdom water treatment works (WTWs) and attempts to address the following key questions for
PAC dosing:
how and where should PAC be dosed for optimum NOM removal; what NOM does PAC remove over and above coagulation and is this dependent on
dosing position; and,
how is PAC removed during subsequent treatment?
Two waters were collected and tested during these experiments. Raw water from
Ewden water treatment works (WTWs) was taken (Sheffield, United Kingdom). This water is typical of a United Kingdom
brown colored moorland water being of high dissolved organic carbon content and low
alkalinity and low turbidity. Water from the same WTWs was also collected after coagulation
and dissolved air flotation (DAF).
Three PAC were tested: two types of Chemviron Pulsorb (Pulsorb C and Pulsorb 207 CP90, and
Norit SA Super). The adsorption capacity of the three PAC carbons was tested by assessing
NOM removal at a range of PAC doses and at six different pH. Raw and coagulated was added to a conical flask (250 mL). PAC was dosed at various concentrations and
80 mg/L at various pH levels. Adsorption was left for 6 hours, the time taken for
adsorption to reach equilibrium. Each condition was carried out in triplicate.
The two water types were tested on to determine the optimum removal of NOM using PAC and
coagulation using a jar tester (Phipps and Bird). The coagulant used was ferric sulphate (Ferripol
XL) and was dosed at a pre-determined optimum for this water at 10 mg/L at pH 4.0. PAC was
dosed at 20 mg/L. A range of coagulant PAC dosing protocols were tested. Includes 5 references, tables, figures.