Membrane fouling is a chronic problem for many nanofiltration (NF) membrane plants. Foulant material can range from colloidal, particulate, inorganic minerals and natural organic matter (NOM) (Schäfer et al., 2006). This research project worked with a small community integrated membrane facility (low-pressure membrane followed by high-pressure) in Nova Scotia with membrane fouling concerns associated with dissolved NOM as the primary foulant. Membrane autopsies conducted in our laboratory have demonstrated that NOM deposits on the NF membrane decreased pore space on the membrane (Lamsal et al., 2012). The membrane fouling resulted in a requirement for increased pressure to produce a constant permeate flow. By adding in-line coagulation prior to low-pressure filtration in an integrated membrane system, the goal was to remove more organic material by MF thereby improving the quality of the feed-water entering the NF membranes. Previous work has shown that for some IMS installations there is a need to reduce the amount of dissolved organic matter prior to NF (Cho et al., 2000; Lamsal et al., 2012; Nilson and DiGiano, 1996; Schäfer et al., 2001). An improved membrane feed-water quality reduces fouling on the membrane and membrane operating cost, and increases productivity and lifespan of the membrane (Choi, 2008). A negative aspect to adding in-line coagulation is it adds another step to the treatment process and sludge removal is required. This study examined the use of in-line coagulation using coagulants aluminum sulphate, ferric chloride and polyaluminum chloride to improve membrane feed-water quality. The addition of in-line coagulation prior to microfiltration will remove NOM with the MF producing improved feed water quality for NF. After determining the optimal dose of each coagulant, 20 L of post-coagulation MF permeate was batched and run through the bench-scale NF membrane for 200 hours. The water quality of the feed tank, concentrate and permeate were monitored constantly as well as the operational properties of pressure and flow. To simulate a full-scale plant the operating conditions of Collins Park water treatment plant on Fletchers Lake were used in the bench-scale set-up. After the 200h NF run time the membranes were analyzed to assess the fouling on the membrane and the performance of each coagulant. Coagulation was found to reduce NF pressure fouling by reduction of NOM in the NF feed-water. Ferric chloride was found to perform best of the three coagulants at a low dose of 0.5mg/L of Fe at a pH of 5.0. / n/a

Abstract The removal of Natural Organic Matter (NOM) from water supplies is crucial for the provision of clean, safe drinking water. Lab-scale experiments have been extensively utilized in water treatment facilities to optimize this process. The most used lab-scale method is known as jar test. However, with emerging membrane filtration, lab-scale inline coagulation has been recently utilized to mimic the inline coagulation over membrane filtration in water treatment plants (WTPs).     This study aims to compare the jar test with inline coagulation micro pilot methods from different aspects, including the NOM removal, time, and cost needed for each process. Three case studies were conducted using different water sources in Sweden: Katrineholm, Mälaren, and Mjörn.     In the experiments conducted to find the optimal NOM removal condition, the inline coagulation micro pilot method outperforms the jar test in terms of removal efficiency. In the Katrineholm case, the micro pilot method with a coagulant dosage of 4.5mg/l Al at pH 6.7 achieved removal efficiencies of 63% for fluorescent dissolved organic matter (fDOM), whereas the jar test showed 60% for fDOM but with a significantly higher coagulant dosage (7mg/l) . In the Mälaren case, the micro pilot method using a dosage of 2.5mg/l Al at pH 6.4 yielded removal efficiencies of 62% for UV254, absorbing organic matter and the jar test showed almost the same removal (64%) for UV254 but with a significantly higher coagulant dosage. Finally, in the Mjörn case, the micro pilot method with a coagulant dosage of 3.5mg/l Al at a pH of 6.6 resulted in removal efficiencies of 76% for UV254, the jar test also resulted in 76% for UV254 at the same pH level.     While the micro pilot method showed better removal efficiencies, it consumed more chemicals compared to the jar test. The micro pilot experiments required higher volumes of coagulant and additional chemical backwashing after each set of experiments. In contrast, the jar test method used lesser quantities of coagulant and chemicals due to the smaller scale of the experiments.     The findings of this study highlight the superiority of the inline coagulation micro pilot method over the jar test for optimizing NOM removal in water treatment processes. Despite the jar test being quicker and more cost-effective, it often required higher coagulant dosages to achieve comparable results. This was mainly because jar test demands heavy and large enough flocs to be removed from the water. Conversely, the micro pilot method, although more time-consuming and costly, provided more precise coagulant dosage control, resulted in higher removal efficiencies, and offered a more comprehensive understanding of the coagulation process. This is achieved by using a sensor called EXO sensor, which allows for immediate monitoring of the treatment results.