[Show abstract][Hide abstract] ABSTRACT: The accumulation of residual coagulant flocs and microbial substances on the surface of ultrafiltration (UF) membranes are major contributors to membrane fouling that reduces process performance. Previous approaches to reduce fouling (e.g. addition of an oxidant or disinfectant) have been only partly successful in reducing the formation of a cake layer and material deposits within membrane pores. In this study the performance of an integrated granular media – UF membrane process has been evaluated in which a hollow-fibre UF module was embedded within a sand layer in order to prevent fouling material reaching the UF surface (forming a cake layer). The evaluation involved comparing two laboratory-scale UF systems, operated in parallel for 74 days, with one incorporating the sand layer (CSUF, coagulation-sand layer filtration-ultrafiltration), and the other without (CUF, coagulation-ultrafiltration), serving as a reference conventional process. The results showed that the incorporation of the sand layer successfully prevented the formation of any significant cake layer on the membrane surface and substantially reduced inner membrane fouling, which lead to a much reduced trans-membrane pressure (TMP) increasing rate. The difference in performance was principally attributed to microbial growth and the release of extracellular polymeric substances (EPS) which was much greater in the conventional CUF system. Thus, in the CUF system, the deposition of coagulation flocs (consisting of precipitated nano-scale primary particles) and bacteria on, and within, the membrane produced substantial reversible and irreversible fouling. In contrast, the deposited material (flocs) in the sand layer of the CSUF system was easy to be washed away, resulting in fewer bacteria in the sand layer, and a much reduced production of biopolymer and other EPS, and their accumulation by the UF membrane.
[Show abstract][Hide abstract] ABSTRACT: This paper describes some results of mini-pilot-scale tests concerning the performance of potassium manganate (K2MnO4) as a pre-treatment chemical prior to ultrafiltration. Manganate is an intermediate in the commercial preparation of permanganate and in aqueous reactions MnO42− can act as both an oxidant and a coagulant/adsorbent arising from the formation of insoluble MnO2. In addition, the combination of ferrous sulfate and manganate (Fe/Mn), offers a potentially cheaper and effective combination of pre-oxidant and coagulant compared to the chemicals used currently in water treatment (e.g. ozone, chlorine, ferric sulfate). In comparative tests with conventional ferric sulfate and using simulated raw water, the results showed that Fe/Mn pre-treatment substantially reduced membrane fouling in terms of the rate of trans-membrane pressure development (arising from both external and internal fouling). Fe/Mn pre-treatment was effective in reducing bacterial activity, changing the characteristics of organic matter and decreasing the production of extracellular polymeric substances (EPS) by bacteria. The external fouling in this process was determined by the EPS concentration, and the internal fouling mainly determined by the adsorption of lower MW organic matter to the membrane pores. Fe/Mn pre-treatment reduced the amounts of both types of fouling material within the cake layer and membrane pores in comparison to conventional pre-treatment with ferrous sulfate, most likely through the formation of solid-phase Fe(III) and MnO2 and by MnO42− oxidation, thereby leading to a substantial increase in membrane run time.
[Show abstract][Hide abstract] ABSTRACT: Uncertainty regarding changes in dissolved organic carbon (DOC) quantity and quality has created interest in managing peatlands for their ecosystem services such as drinking water provision. The evidence base for such interventions is, however, sometimes contradictory. We performed a laboratory climate manipulation using a factorial design on two dominant peatland vegetation types (Calluna vulgaris and Sphagnum Spp.) and a peat soil collected from a drinking water catchment in Exmoor National Park, UK. Temperature and rainfall were set to represent baseline and future conditions under the UKCP09 2080s high emissions scenario for July and August. DOC leachate then underwent standard water treatment of coagulation/flocculation before chlorination. C. vulgaris leached more DOC than Sphagnum Spp. (7.17 versus 3.00 mg g−1) with higher specific ultraviolet (SUVA) values and a greater sensitivity to climate, leaching more DOC under simulated future conditions. The peat soil leached less DOC (0.37 mg g−1) than the vegetation and was less sensitive to climate. Differences in coagulation removal efficiency between the DOC sources appears to be driven by relative solubilisation of protein-like DOC, observed through the fluorescence peak C/T. Post-coagulation only differences between vegetation types were detected for the regulated disinfection by-products (DBPs), suggesting climate change influence at this scale can be removed via coagulation. Our results suggest current biodiversity restoration programmes to encourage Sphagnum Spp. will result in lower DOC concentrations and SUVA values, particularly with warmer and drier summers.
Water Research 12/2014; 67:66-76. DOI:10.1016/j.watres.2014.09.015 · 5.53 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Ozonation before chlorination is associated with enhanced formation of chloropicrin, a halonitromethane disinfection by-product (DBP), during drinking water treatment. In order to elucidate reasons for this, five natural organic matter (NOM) surrogates were treated using both chlorination and ozonation-chlorination under controlled laboratory conditions. Selected surrogates comprised two phenolic compounds, two free amino acids and one dipeptide; these were resorcinol, 3-aminophenol, L-aspartic acid, β-alanine and ala-ala, respectively. Quantified DBPs included chloropicrin, chloroform, dichloroacetonitrile and trichloroacetonitrile. Relative to chlorination alone, increases in the formation of chloropicrin from ozonation-chlorination varied from 138% for 3-aminophenol to 3740% for ala-ala for the four amine surrogates. This indicates that ozone is more effective than chlorine in mediating a rate-limiting oxidation step in chloropicrin formation, most plausibly involving conversion of an amine group to a nitro group. While both hydrophilic and hydrophobic surrogates acted as chloropicrin precursors, ala-ala was the most reactive precursor following ozonation-chlorination. Since peptides are far commoner in drinking water sources than free amino acids, further research into chemical oxidation of these species by ozone and chlorine is recommended. In contrast, oxidation with ozone prior to chlorination reduced chloroform formation moderately for the two phenolic compounds.
[Show abstract][Hide abstract] ABSTRACT: Climate change in the UK is expected to cause increases in temperatures, altered precipitation patterns and more frequent and extreme weather events. In this review we discuss climate effects on dissolved organic matter (DOM), how altered DOM and water physico-chemical properties will affect treatment processes and assess the utility of techniques used to remove DOM and monitor water quality. A critical analysis of the literature has been undertaken with a focus on catchment drivers of DOM character, removal of DOM via coagulation and the formation of disinfectant by-products (DBPs). We suggest that: (1) upland catchments recovering from acidification will continue to produce more DOM with a greater hydrophobic fraction as solubility controls decrease; (2) greater seasonality in DOM export is likely in future due to altered precipitation patterns; (3) changes in species diversity and water properties could encourage algal blooms; and (4) that land management and vegetative changes may have significant effects on DOM export and treatability but require further research. Increases in DBPs may occur where catchments have high influence from peatlands or where algal blooms become an issue. To increase resilience to variable DOM quantity and character we suggest that one or more of the following steps are undertaken at the treatment works: a) 'enhanced coagulation' optimised for DOM removal; b) switching from aluminium to ferric coagulants and/or incorporating coagulant aids; c) use of magnetic ion-exchange (MIEX) pre-coagulation; and d) activated carbon filtration post-coagulation. Fluorescence and UV absorbance techniques are highlighted as potential methods for low-cost, rapid on-line process optimisation to improve DOM removal and minimise DBPs.
Science of The Total Environment 01/2014; 473-474C:714-730. DOI:10.1016/j.scitotenv.2013.12.095 · 4.10 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Peatlands in the UK represent a large proportion of the soil carbon store, however there is concern that some systems may be switching from sinks to sources of carbon. The accumulation of organic material in peatlands results from the slow rates of decomposition typically occurring in these regions. Climate change may lead to faster decomposition which, if not matched by an equivalent increase in net primary productivity and litter fall, may tip the balance between source and sink. Recent trends have seen a greater flux of dissolved organic matter (DOM) from peatlands to surface waters and a change in DOM character, presenting challenges to water treatment, for example in terms of increased production of disinfectant by-products (DBPs). Peat systems border a large proportion of reservoirs in the UK so uncertainty regarding DOM quantity and quality is a concern for water utilities. This study considered five peatland vegetation types (Sphagnum spp., Calluna vulgaris, Molinea caerulea, peat soil and mixed litter) collected from the Exmoor National Park, UK where it is hypothesised that peat formation may be strongly affected by future changes to climate. A factorial experiment design to simulate climate was used, considering vegetation type, temperature and rainfall amount using a current baseline and predictions from the UKCP09 model. Gaseous fluxes of carbon were monitored over a two month period to quantify the effect on carbon mineralisation rates while 13C NMR analysis was employed to track which classes of compounds decayed preferentially. The DOM collected was characterised using UV and fluorescence techniques before being subject to standard drinking water treatment processes (coagulation/flocculation followed by chlorination). The effect of the experimental factors on DOM amenability to removal and propensity to form DBPs was then considered, with both trihalomethane (THM) and haloacetonitrile (HAN) DBP classes monitored. Initial results have shown a statistically significant (Mann-Whitney U) difference in THM formation (p<0.05) as well as the amount of DOM produced and specific UV absorption at 254nm (p<0.01) between vegetation classes.
Cite as: Author(s) (2013), Title, Abstract B33F-0542 presented at 2013 Fall Meeting, AGU, San Francisco, Calif., 9-13 Dec.
AGU Fall Meeting, San Francisco, California, USA; 12/2013
[Show abstract][Hide abstract] ABSTRACT: During drinking water treatment aqueous chlorine and bromine compete to react with natural organic matter (NOM). Among the products of these reactions are potentially harmful halogenated disinfection by-products, notably four trihalomethanes (THM4) and nine haloacetic acids (HAAs). Previous research has concentrated on the role of bromide in chlorination reactions under conditions of a given NOM type and/or concentration. In this study different concentrations of dissolved organic carbon (DOC) from U.K. lowland water were reacted with varying amounts of bromide and chlorine in order to examine the interrelationship between the three reactants in the formation of THM4, dihaloacetic acids (DHAAs) and trihaloacetic acids (THAAs). Results showed that, in general, molar yields of THM4 increased with DOC, bromide and chlorine concentrations, although yields did fluctuate versus chlorine dose. In contrast both DHAA and THAA yields were mainly independent of changes in bromide and chlorine dose at low DOC (1mg·L(-1)), but increased with chlorine dose at higher DOC concentrations (4mg·L(-1)). Bromine substitution factors reached maxima of 0.80, 0.67 and 0.65 for the THM4, DHAAs and THAAs, respectively, at the highest bromide/chlorine ratio studied. These results suggest that THM4 formation kinetics depend on both oxidation and halogenation steps, whereas for DHAAs and THAAs oxidation steps are more important. Furthermore, they indicate that high bromide waters may prove more problematic for water utilities with respect to THM4 formation than for THAAs or DHAAs. While mass concentrations of all three groups increased in response to increased bromide incorporation, only the THMs also showed an increase in molar yield. Overall, the formation behaviour of DHAA and THAA was more similar than that of THM4 and THAA.
Science of The Total Environment 10/2013; 470-471C:469-479. DOI:10.1016/j.scitotenv.2013.09.106 · 4.10 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Dissolution of elemental aluminium or iron in aqueous solutions is the essential reaction in electrocoagulation and other processes. These and several other such metals have been reported as exhibiting dissolution rates that are higher than expected from the current/charge passed, assuming electron stoichiometries of three (Al) or two (Fe). In batch dissolution experiments with aluminium electrodes, the addition of 30 g m−3 humic acid to a neutral solution of 0.5 mol m−3 Na2SO4 and 8.8 g m−3 NaCl led to an unexpected increase in the charge yield from 1.0 to ca. 1.5, i.e., the measured concentration of aluminium species in solution was 50% higher than expected from Faraday's law. This phenomenon may be explained when net currents are used for the determination of charge yields, whereas partial dissolution currents should be used. The latter may be significantly greater than the former when a concurrent reduction reaction, such as hydrogen evolution, occurs at the dissolving electrode.
Chemical Engineering Science 05/2013; 95:353–359. DOI:10.1016/j.ces.2013.03.016 · 2.34 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Passivation of aluminium electrodes is a common problem in electrocoagulation reactors used for the purification of natural waters, leading to high cell voltages and compromising the viability of such industrial processes. In order to identify possible mitigation strategies, cyclic voltammetric, potentiostatic and galvanostatic measurements were made to investigate the effects of electrode surface topography and solution composition on the electrochemical behaviour of pure aluminium and Al 1050 in neutral electrolyte solutions. In addition, electrochemical impedance spectroscopy was used to estimate thicknesses of passive layers before and after dissolution. Chloride ions, well-known pitting promoters, were found to facilitate dissolution of passive surfaces, though attenuated by the presence of the naturally-occurring contaminant humic acid. Under galvanostatic conditions, a smooth aluminium electrode showed a constant rise in electrode potential due to passivation. However, even in the absence of pitting promoters electrodes with a ‘rough’ surface finish (Ra-values>1 µm) exhibited an initial potential increase, followed by a dramatic decrease to relatively low steady state values due to ‘spontaneous de-passivation’. The precipitation of dissolved aluminium(III), releasing H+ ions, was identified as the probable cause of this effect, leading to acidic pH values, locally dissolving the passive layer and enhancing dissolution rates.
Chemical Engineering Science 05/2013; 95:301–312. DOI:10.1016/j.ces.2013.03.010 · 2.34 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Catchments draining peat soils provide the majority of drinking water in the UK. Over the past decades, concentrations of dissolved organic carbon (DOC) have increased in surface waters. Residual DOC can cause harmful carcinogenic disinfection by-products to form during water treatment processes. Increased frequency and severity of droughts combined with and increased temperatures expected as the climate changes, have potentials to change water quality. We used a novel approach to investigate links between climate change, DOC release and subsequent effects on drinking water treatment. We designed a climate manipulation experiment to simulate projected climate changes and monitored releases from peat soil and litter, then simulated coagulation used in water treatment. We showed that the 'drought' simulation was the dominant factor altering DOC release and affected the ability to remove DOC. Our results imply that future short-term drought events could have a greater impact than increased temperature on DOC treatability.
[Show abstract][Hide abstract] ABSTRACT: The aqueous ozonation reactions of surfactants have been reviewed, including the degree of reaction and ozonation byproduct identity. Compounds are classified into three groups: anionic, cationic and non-ionic surfactants. Experimental conditions for each study reviewed are summarized. Much work has been carried out under conditions unlikely to occur during drinking water treatment. Thus, most of the findings of the papers reviewed cannot be directly applied to potable water treatment. However this review can serve to indicate the likely reactivity of the individual surfactants towards ozone and the possible byproducts formed. This reactivity depends mainly upon the chemical structure of the surfactants. The removal of surfactants is pH dependent, with the best results being obtained in alkaline media. High ozone doses usually do not cause complete disruption of the surfactants, while smaller doses are sufficient to enhance their biodegradation.
[Show abstract][Hide abstract] ABSTRACT: This research note investigates the efficiency of ozone for the degradation of threeherbicides, imazapyr[2-(4-isopropyl-4-methyl-S-oxo-2-imidazolin-2-yl)-nicotinic acid], triclopyr (3,5,6-trichloro-2-pyridinyloxyacetic acid) and diuron [3-(3,4-dichlorophenyl)-1,1-dimethylurea] under controlled laboratory conditions. Experiments were conducted at pH 7.5, in a bubble contactor column, with a raw lowland surface water spiked with initial active ingredient concentrations of 2 μgL−1. The ozone doses applied and consumed were 4.79 and 2.9 mg O3L−1 respectively, and resulted in removals of 53%, 48% and 90% of imazapyr, triclopyr and diuron, respectively.
[Show abstract][Hide abstract] ABSTRACT: Granular activated carbon (GAC) production from waste materials is of considerable interest. However, waste-derived GACs may exhibit an inferior attrition resistance, inhibiting their use in pressure swing adsorption (PSA) processes or adsorption beds. This study investigated the use of pelletisation to improve the attrition resistance of a sewage sludge-derived GAC. Both extrusion and hydraulic compaction were investigated as pellet production techniques. A range of binders were evaluated, as follows: Ammonium LignoSulphonAte (ALSA); PolyVinyl Alcohol (PVA); phenolic resin; araldite resin; LignoSulphonic Acid, Calcium salt (LSAC); CarboxyMethyl Cellulose sodium salt (CMC). Hydraulic compaction (typically >500 kg/cm2) was found to be more effective than extrusion. Both low (40–110 °C) and high (500–800 °C) curing temperatures were assessed, as was the sequencing of the steam activation and pelletisation steps. The most effective pellet production method comprised coupling a lignosulphonate binder with post-pelletisation activation.
[Show abstract][Hide abstract] ABSTRACT: This study has investigated the comparative treatment and hydraulic performance of biosand filters (BSFs) of pumice and sand filter media, with Escherichia coli and turbidity as the principal indicators of water quality. The study has also assessed the effect of resting time on E. coli and turbidity removal. The performances of three filter columns consisting of sand, pumice, and sand/pumice dual media with a bed depth of 80 cm were evaluated over 4 months continuously. The columns were charged twice daily with local canal water. The pumice and the dual media filters achieved 24 and 14%, respectively, greater volume production per cycle compared to that of the sand filter. The pumice filter had consistently lower filtrate turbidity than the other filters with about 98.5% turbidity removal. Average E. coli removals were similar for all filters and corresponded to 0.9-1.8 log units for unripened media, and 1.4-3.3 log units for ripened media. It was observed that resting time of more than 4 h was necessary to achieve significant E. coli removal. Hydraulic and water quality profiles indicated that schmutzdecke development in the pumice layer was not effective as in the sand bed.
Aqua 06/2012; 61(4):201. DOI:10.2166/aqua.2012.100 · 0.84 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Monitoring the quality of drinking water from the treatment plant to the consumers tap is critical to ensure compliance with national standards and/or WHO guideline levels. There are a number of processes and factors affecting the water quality during transmission and distribution which are little understood. A significant obstacle for gaining a detailed knowledge of various physical and chemical processes and the effect of the hydraulic conditions on the water quality deterioration within water supply systems is the lack of reliable and low-cost (both capital and O & M) water quality sensors for continuous monitoring. This paper has two objectives. The first one is to present a detailed evaluation of the performance of a novel in-pipe multi-parameter sensor probe for reagent- and membrane-free continuous water quality monitoring in water supply systems. The second objective is to describe the results from experimental research which was conducted to acquire continuous water quality and high-frequency hydraulic data for the quantitative assessment of the water quality changes occurring under steady and unsteady-state flow conditions. The laboratory and field evaluation of the multi-parameter sensor probe showed that the sensors have a rapid dynamic response, average repeatability and unreliable accuracy. The uncertainties in the sensor data present significant challenges for the analysis and interpretation of the acquired data and their use for water quality modelling, decision support and control in operational systems. Notwithstanding these uncertainties, the unique data sets acquired from transmission and distribution systems demonstrated the deleterious effect of unsteady state flow conditions on various water quality parameters. These studies demonstrate: (i) the significant impact of the unsteady-state hydraulic conditions on the disinfectant residual, turbidity and colour caused by the re-suspension of sediments, scouring of biofilms and tubercles from the pipe and increased mixing, and the need for further experimental research to investigate these interactions; (ii) important advances in sensor technologies which provide unique opportunities to study both the dynamic hydraulic conditions and water quality changes in operational systems. The research in these two areas is critical to better understand and manage the water quality deterioration in ageing water transmission and distribution systems.
Water Research 11/2011; 46(1):235-46. DOI:10.1016/j.watres.2011.10.058 · 5.53 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The potential of using sludge based activated carbons (SBACs) for catalysing the wet air oxidation (WAO) of phenol, o-cresol, o-chlorophenol and p-nitrophenol was assessed in both a batch slurry reactor and a continuous trickle-bed reactor. In the batch reactor, the activity of two powdered carbons prepared from, respectively, dewatered raw (DRAW) sludge and dewatered, mesophilic anaerobically digested (DMAD) sludge was tested at 160 °C and 4.2 bar of oxygen partial pressure. Continuous, trickle-bed reactor experiments of up to 72 h were conducted at similar operating conditions to study the durability and catalytic wet air oxidation (CWAO) performance of three economically promising steam activated SBACs. Due to their low mechanical strength, the two DRAW derived SBACs tested were produced using two different attrition resistance enhancement techniques. A commercial activated carbon (Chemviron, AP4-X) was employed as the reference catalyst for all of the tests.
In the batch runs, the SBACs and AP4-X achieved high levels of pollutant conversion in the case of phenol, o-cresol and o-chlorophenol. However, irrespective of the carbon tested, p-nitrophenol was resistant to oxidation. When employed in the trickle-bed reactor, the DRAW derived SBAC pelletised using a lignosulphonate binder was found to be the most stable carbon. With this carbon the order of compound reactivity was as observed in the batch experiments.
[Show abstract][Hide abstract] ABSTRACT: In this study the use of novel CWAO catalysts, namely potentially low cost, sewage sludge derived activated carbons was explored. Two types of municipal sludge were used: dewatered raw filter cake and dewatered mesophilic anaerobically digested sludge. The carbons were produced by: carbonisation; physical activation (steam or CO2) and chemical activation (K2CO3). Hydrochloric acid washing of some of the carbons was also investigated. The carbons were characterised in terms of their surface area, contact pH, propensity towards metal leaching, surface chemistry (via FTIR), ash content and inorganic elemental composition. Their CWAO performance was assessed at 160°C and a partial oxygen pressure of 4.2 bar (25bar of air) within a stirred batch reactor containing a 5 g/L phenol solution. All the carbons exhibited catalytic activity, with the K2CO3 activated and HCl washed carbons attaining a phenol and TOC removal that matched the performance of an activated carbon specifically manufactured for oxidative wastewater treatment applications. A strong correlation was found between surface area and phenol or TOC conversion, suggesting that surface area is a primary requisite for their performance in the first batch cycle. Thus, the oxidation of phenol is thought to proceed via a free radical driven mechanism. The active sites necessary to facilitate this mechanism, whether present as surface functional groups or active metals (e.g., Fe), were detected on all of the sludge based activated carbons. However, no clear correlation between phenol conversion and these active sites could be established.
[Show abstract][Hide abstract] ABSTRACT: This paper describes a preliminary evaluation of the performance of carbonaceous materials prepared from sewage sludges (SBCMs) in a hybrid water treatment process based on adsorption and catalytic wet air oxidation; phenol was used as the model pollutant. Three different sewage sludges were treated by either carbonisation or steam activation, and the physico-chemical properties of the resultant carbonaceous materials (e.g. hardness, BET surface area, ash and elemental content, surface chemistry) were evaluated and compared with a commercial reference activated carbon (PICA F22). The adsorption capacity for phenol of the SBCMs was greater than suggested by their BET surface area, but less than F22; a steam activated, dewatered raw sludge (SA_DRAW) had the greatest adsorption capacity of the SBCMs in the investigated range of concentrations (<0.05 mol L(-1)). In batch oxidation tests, the SBCMs demonstrated catalytic behaviour arising from their substrate adsorptivity and metal content. Recycling of SA_DRAW in successive oxidations led to significant structural attrition and a hardened SA_DRAW was evaluated, but found to be unsatisfactory during the oxidation step. In a combined adsorption-oxidation sequence, both the PICA carbon and a selected SBCM showed deterioration in phenol adsorption after oxidative regeneration, but a steady state performance was reached after 2 or 3 cycles.
[Show abstract][Hide abstract] ABSTRACT: The aqueous degradation of iopromide, an iodinated X-ray contrast media (ICM) compound, by the combination of UV(254) irradiation and potassium peroxydisulfate (K(2)S(2)O(8)) has been studied in laboratory scale experiments. The influence of various parameters on the performance of the treatment process has been considered, namely the UV irradiation light intensity, the initial concentrations of iopromide and peroxydisulfate, and the initial solution pH. Iopromide degradation increased with UV light intensity and peroxydisulfate concentration, but decreased with initial pH. Under specific conditions complete removal of iopromide was achieved within 30 min, and near-complete mineralisation (loss of solution TOC) within 80 min. Degradation was believed to be caused by a combination of direct photolysis, sulphate radical attack, and, to a minor degree, direct oxidation by peroxydisulfate. Approximate values for the reaction rate constants have been determined and found to be equal to 1-2x10(4) M(-1) s(-1) for sulfate radicals, and 1-2 M(-2) s(-1) for S(2)O(8)(2-). Overall compound degradation was observed to follow first-order kinetics where the rate constant decreased with initial solution pH. During the reaction, the solution pH decreased as a consequence of sulfate radical scavenging.
[Show abstract][Hide abstract] ABSTRACT: Ferrate (FeO₄²⁻) is believed to have a dual role in water treatment, both as oxidant and coagulant. Few studies have considered the coagulation effect in detail, mainly because of the difficulty of separating the oxidation and coagulation effects. This paper summarises some preliminary results from laboratory-based experiments that are investigating the coagulation reaction dynamically via a PDA instrument, between ferrate and humic acid (HA) at different doses and pH values, and comparing the observations with the use of ferric chloride. The PDA output gives a comparative measure of the rate of floc growth and the magnitude of floc formation. The results of the tests show some significant differences in the pattern of behaviour between ferrate and ferric chloride. At pH 5 the chemical dose range (as Fe) corresponding to HA coagulation was much broader for ferrate than ferric chloride, and the optimal Fe dose was greater. Ferrate oxidation appears to increase the hydrophilic and electronegative nature of the HA leading to an extended region of charge neutralisation. A consequence of the ferrate oxidation is that the extent of HA removal was slightly lower ( approximately 5%) than with ferric chloride. At pH 7, in the sweep flocculation domain, ferrate achieved much greater floc formation than ferric chloride, but a substantially lower degree of HA removal.