Modification of jar testing protocol combined with mEnCo model predicted dose to predict dissolved organic matter removal for surface water

Abstract

In recent climatic events in Australia, the need for frequent jar testing to determine optimum coagulant and flocculant aid doses became apparent in order to determine optimum treatment efficiencies. The generally applied jar test procedure can be time consuming and this prompted investigations for capacity to test waters more rapidly, for a range of treatment conditions and raw water qualities. The aim of the study reported here was to develop a rapid procedure for determination of dissolved organic matter (DOM) removal, measured as dissolved organic carbon (DOC) and UV absorbance (254 nm/cm) for a range of treatment conditions (coagulant dose and pH). Using a previously established predictive model (mEnCo) for coagulant determination, several coagulant doses were then predicted for application in a modified jar test protocol. Best fitted and predicted data using the exponential decay function compared with data from a 6× jar test procedure were obtained using three doses, 0.5× EnCD, EnCD and a VHD (≥2× EnCD). It is proposed that this procedure may be used to more rapidly determine treatment efficiency based on the removal of organics compared with traditionally applied jar tests.

Comparing the log‐response curve and adsorption isotherm model for removing dissolved organic matter during La Nina event

“…Monthly samples were collected from the Morgan WTP between December 2010 and May 2011 during a strong La Nina event (Murshed et al ., 2013; Aslam et al ., 2013) and used to develop the predictive model detailed later in this study. Moreover, Murshed et al .…”

“…The modelling of coagulation dose prediction is still relatively new and most available models developed for conventional treatment plants are based on historical data‐driven analyses (Adgar et al ., 2000; Baxter et al ., 2002; Fisher et al ., 2004; Uyak and Toroz, 2007). With respect to models used for predicting coagulant dose for enhanced treatment (for removal) of NOM, there are several ways to estimate the coagulant dose required: (i) for maximum or near maximum removal (van Leeuwen et al ., 2003; 2005); (ii) it corresponds to a certain point of diminishing returns (PODR) (USEPA, 1999); (iii) it is based on percentage of NOM removal; and (iv) a decrease in NOM to an absolute target (Xie et al ., 2012; Murshed et al ., 2013). However, though these models can provide for rapid prediction of coagulant dose, validation of doses in response to rapid water quality changes still may require laboratory analyses of the raw water.…”

“…Models for predicting residual DOM or DOM removal after coagulation have enabled the water treatment industry to estimate and/or predict the performance of chemical treatment processes. These models are normally empirical and semi‐empirical based on a regression analysis that considers the amount and character of NOM in raw and treated water (Chiemchaisri et al ., 2008; Xie et al ., 2012; Murshed et al ., 2013; Sabia et al ., 2018). By establishing the empirical or semi‐empirical relationships of water quality parameters, the coagulation dose can be estimated to achieve targeted water quality.…”

State of the Art of Online Monitoring and Control of the Coagulation Process

“…In addition, pH changes due to hydrolysis reactions could be predicted. The software mEnCo has been applied to predict coagulant doses in a modified jar test protocol [27].…”

“…The data are normally used to adjust coagulant dosage manually by the operating staff based on empirical knowledge [5]. Although software tools have been developed that predict both dosages and finished water quality based on feed-forward control techniques [27,37], automatic processing of finished water parameters for direct feed-backward control has not yet been realized. For raw water sources with quality variations observed only after weeks or months, for example Scandinavian lakes, feed-backward control of finished water quality could be employed though many plants actually do not see the real need for online dosing control.…”