Zeta Potential Measurement for Water Treatment Coagulation Control

Zeta potential was assessed as a tool to optimise pre chlorination and coagulation at Happy Valley WTP. Samples were obtained from jar tests using raw water spiked with green algal cells. Treatments including pre-chlorination, polyelectrolyte addition and alum dose were investigated. Results indicate that: pre chlorination of algal spiked raw water had a negative effect on zeta potential; alum dose was the most important factor in determining zeta potential; and, the alum dose that achieved the point of neutral charge was equivalent to the alum dose predicted using a predictive algorithm (WTCcoag) which used raw water UV254, colour and turbidity as input parameters. Results obtained from full scale trials did not reproduce jar test results. INTRODUCTION Raw water for Happy Valley water treatment plant (WTP) is indirectly supplied from the River Murray and natural catchment and includes open reservoir storage prior to treatment. Raw water quality typically contains elevated concentrations of natural organic matter (NOM) and is generally low in turbidity (<10 NTU). Happy Valley reservoir is subject to algal and cyanobacterial blooms and Cryptosporidium oocysts are detected at the inlet to the WTP. Physical treatment including alum coagulation, cationic (positively charged) polyelectrolyte addition as a flocculant aid, flocculation, sedimentation and granular media filtration are employed. These physical treatment processes rely upon modification of size, surface charge and density of particulate and colloidal contaminants in raw water by coagulation (Jefferson et al., 2004). Zeta potential provides a measure of the overall surface charge of all the particles and colloids in a water sample. Most particles and colloids, including NOM, algal cells, microorganisms, clay and silt, that are found in natural water at normal pH conditions (pH 6 to 9) are negatively charged possessing a zeta potential in the range -14 to -30 mV (Jefferson et al., 2004). These surface charges give rise to repulsive forces which prevents aggregation and results in a stable system. NOM has a great influence over the fate of colloids and particles in surface waters. NOM is a diverse group of chemicals including algogenic organic matter (AOM). The chemical properties of NOM is an important factor in determining whether colloids will be stabilised or destabilized (Wilkinson et al.,1997; Walker and Bob, 2001). In typical raw water conditions, fulvic acids will coat and impart a negative charge to colloids and prevent agglomeration while organic compounds with chain like structures can aggregate inorganic colloids through the formation of bridges (Wilkinson, 1997). Zeta potential is determined by measuring the velocity of particles and colloids while they are subject to an electric field (Jefferson et al., 2004). Charged particles and colloids migrate towards an electrode of the opposite polarity in proportion to the field strength and zeta potential (Jefferson et al., 2004). Recently developed, commercially available instrumentation enables automatic measurement of zeta potential within water samples; the ZetasizerTM is one such instrument. Edzwald and Tobiason (1999) described two mechanisms for alum coagulation of NOM. The first involves reactions with positively charged aluminium hydrolysis species and negatively charged NOM to form a complex. The second mechanism enables additional NOM removal as colloids and particles adsorb to the complex-metal hydroxide precipitate. The surface charge of particulate and colloidal contaminants in water are strongly influenced by pH and in most cases the zeta potential increases with increasing pH (Jefferson et al., 2004). Similarly the positive charge of aluminium hydrolysis species is also pH dependent, within the pH range that is required to minimize aluminium solubility. The miniumum pH of solubility is about 6.0 and 6.8 at 25C and 4C respectively (Eldzwald and Tobiason, 1999). Interestingly the surface charge of the aluminium species is greatest at low pH and deceases as pH increases. This explains why alum coagulation at low pH can be achieved with lower doses than at high coagulation pH (Eldzwald and Tobiason, 1999). Reports in the literature indicate that a range of optimum zeta potential conditions are required in order to achieve treatment objectives and site specific testing is recommended (Jefferson et al., 2004). Henderson (2008) claimed an operational zeta potential window between -10 to +2 mV is required in order to achieve optimum removal of both algal cells and extracellular organic matter. The author claimed that this can be achieved through a combination of coagulant dose and/or pH adjustment. Similarly Sharp et al., (2005) reported an operational window in the range -10 to +5 mv for the optimal removal of dissolved organic carbon (DOC). The WTC-coag software model (previously referred to as mEnCo) is employed at metropolitan Adelaide WTPs to predict the required alum dose for an operator set point removal of dissolved organic carbon (DOC). The model has been described previously (van Leeuwen et al., 2005 and 2009). Inputs to the model are raw water colour, turbidity and absorbance at 254 nm (UV254). In recent years operators at Happy Valley WTP have reported that nuisance algal blooms have upset filter performance resulting in increased filtered water turbidity and reduced backwash intervals as well as increasing the required alum demand. The preferred operational response to nuisance algal blooms is pre chlorination where a chlorine dose in the range 1 to 2 mg L is applied prior to coagulant (alum) addition. Jar tests have been undertaken by operators at Happy Valley WTP in an attempt to optimize the pre chlorine dose using conventional endpoints such as colour, turbidity and UV254. However, results obtained to date have been inconclusive. This, together with the instability of algae populations in the raw water, presents a challenge when attempting to select the optimum pre chlorine and alum dose required to achieve water quality requirements cost effectively. Zeta potential measurement was identified as a possible way to optimize treatment at Happy Valley WTP. This paper presents results from a trial undertaken to assess zeta potential as a tool to assist operators optimize pre chlorine and alum dose at Happy Valley WTP.