Studies on Clarication of Coal Washery Euent Using Polymeric Flocculants and Settling Kinetics

Settling tests were conducted on the washery euent using three types of occulants namely cationic (Teloc-3674F), anionic (Magna-1011) and non-ionic (Nalco-83370 + ). For the study purpose coal washery euent (having pH of 8.72) was collected from a newly commissioned plant-thickener. Tests were conducted by varying the pH of the pulp at three levels of acidic (4.0), alkaline (11.0) and the natural pH (8.72) of the as collected euent, besides the occulent dosages varied at four levels (2, 4, 6 and 8gpt) for each of the occulants stated. The results of these tests, estimated in terms of initial settling rate and turbidity indicated that; among the three occulants tested Teloc-3674F and Magna-1011 gave best results as compared to Nalco-83370 + occulent at a washery water pH 8.72. The effect of molecular weight of the occulants on settling of solids in euent has been established in terms of the kinetics with a characteristic number for each of the type of occulants used.


Introduction
With advancements that have taken place in the eld of mining, proportion of nes generation in the ROM coal has gone up appreciably. This resulted in numerous environmental problems related to air-water pollutions. Further, Indian coals possess di cult-to-wash property which needs ner size reductions for improved liberation prior to washing. This has resulted in generation of excess nes in the coal being sent to washing circuits. Coal (hydrophobic) along with other mineral nes (hydrophilic) like shale, slate, clay and several other impurities are generated during stacking, handling, transportation, and dumpling at different points in CPP (Duong et al. 2000). Coal washing involves wet processes where coal gets mixed with water, and requires e cient dewatering and clari cation techniques. Coal being hydrophobic in nature, at ner sizes the particles remain suspended in water and do not settle even after several days, particularly in tailing ponds. This pollutes the ground water and also deteriorates the agricultural land/crops surround it. Besides this the unclari ed, when recirculated to coal washing plants, spoils the overall performance of the operating washery, because of accumulation of colloidal sized particles in excess. These problems are not limited only to the Indian coal washing circuits, but it is noticed globally in all the coal washeries (Moosavirad et al. 2010;Sabah and Erkan 2006;Gupta and Singh 1995;Sarioglu et al 2002).
To get-rid of all these problems, one alternative is to enhance the settling rate of particulate solids using occulants (Khandal 1985). Polymer occulants show higher e ciency and acceptability in wastewater/e uent treatments (Bolto and Gregory 2007;Lee et al. 2014;Teh et al. 2016;Dao et al. 2016) especially for removing suspended particulate matter that are accountable for high turbidity (Haiyang et al. 2017). Flocculants are high-molecular long chain water-soluble polymeric compounds that bind the surfaces of the dispersed particles physically and combine them to form agglomerates known as ocs that settle rapidly. The widely used occulants are; the polyelectrolytes of anionic or cationic or non-ionic types having high molecular weight with different structures of macromolecules which intensify the solid-liquid separations in coal washing plants (Yakovchenko et al. 2019). Long-back Slater and Kitchener reported that the application of certain polymers as occulants helped in promoting the settling rates of solids in industrial disperse systems (Slater and Kitchener 1966).
Water clari cation plays a major role in the overall economics of the coal preparation plants (CPP). As per the present environmental policies no e uent/slurry is allowed to be discharged outside the washery premises. Under these circumstances, water clari cation using occulants will help to abide by the guidelines set by MoEF, GoI, 2010, wherein the CPP are to operate the plants on a closed system watercirculation basis, and limit the fresh water consumption not exceed above 1.5 m 3 . Nowadays, the washeries are designed for closed-circuit system, where the water used in washing circuits to wash the raw coal is recycled and reused. Since this water detains appreciable percentage of colloidal size coal nes and clay particles, need to be clear-off dirt and contain only negligible amounts of suspended particles. Otherwise, huge volume of water would be required to process the coal. But, the removal of these colloidal nes (or water clari cation) from the tailing slurry has posed number of challenges.
As a general norm, nes having size less than 0.50 mm are generally responsible for forming slurry in the washery (Ash and Rao 1972). Thus, an appropriate solid-liquid separation has become an essential part of a coal washing plant. In practice, the prevalent method adopted is sedimentation with aid of occulants in thickeners, using polymeric occulants. This method has two signi cant strictures that evaluate occulation performance (Oruç et al. 2010).
i. The rst is the low water-turbidity, since water recovered at the thickener over ow is recycled to the plant. The suspended colloidal particles in the recycled water should be at the lowest level, eliminating any negative implications to the process.
ii. The second is, settling rate of the occulated particles, has a direct effect on thickener capacity as well as occulent performance.
Thus, the challenge is dual; precise determination of various properties of the washery water e uent, this includes mineralogy of the gangue particles, physic-chemical properties such as pH, conductivity, pulp density, TDS, alkalinity, hardness etc. besides electro-kinetic properties of the solid matter in washery e uents. Achieving the lowest water turbidity level while maintaining high settling rate typically in uenced and controlled by polymeric occulent properties, its dosages and the pH value of the washery e uent.
Moreover, the e ciency of the process is dependent on the use of accurate type of occulants (i.e. anionic, cationic and non-ionic polymer) as well as the addition of their dosage (Sabah and Cengiz 2004;Pal et al. 2008).
On the other hand, water is becoming a scarce commodity, and several governmental slogans are heard every day to save water under the government's national policy. Still, the problem of water puri cation persists and supply of good quality water remains as a major problem. Thus, coal-water slurry handling has become an integral part of a coal bene ciation plant. This is because, the amount of ne-coal tailings generated annually is very high. To overcome the di culties faced, the alternatives left are: i. Development of e cient solid-liquid separation systems, ii. Selection and addition of the correct occulent.
iii. Understanding of the interaction between the occulent and the slurry is vital in achieving optimum condition.
In view of the above, in this study the in uence of occulent types, their dosage effects and effect of pH of the pulp on the settling of ne coal nes has been attempted after the physico-chemical and mineralogical characteristics of an e uent collected from a thickener of an operating plant (Dahibari, India . Among the several containers, one of the container was selected, the slurry was ltered and dried. The solids obtained thereof were analysed for the characterisation studies in detail. About Dahibari coal washing plant: Dahibari washery was designed for washing of high-ash Indian coals, that are recently classi ed and categorised as washery grade IV (with 28 to 35% ash) and washery grade V (with 35 to 42% ash). This is a standalone-type coal washery commissioned in the year 2018 with a treatment capacity of 1.6 MTPA of coal that produces small quantity of washed coking coal (of 19% ash) and the power coal (middling) as its nal products. Both washed coal (of 19% ash) and the middling (with 33.5±0.50% ash to captive power plants) are sent to SAIL steel plants. Since the washery was designed to receive coals from different collieries, it falls under the category as NLW (Non-linked washery). Hence the feed ash of raw coal to the washery varies very widely ranging from 28% to 40%. Raw coal crushed to 30mm is washed with the help of heavy media cyclones and thickeners installed for the collections of comparatively large-sized fraction of nes. The supernatant water of the thickener impounded is recirculated to the plant while the sludge of the thickener is store in tailing ponds. The satellite location map (latitude=23 o 72'91.4"N and longitude=86 o 77'92.0"E) of Dahibari is given in Fig. 1.

2.2
Characterization of Sample: Characterization studies on the collected e uent sample was carried out to know the physico-chemical parameters of slurry viz. pH, conductivity and TDS were measured onsite while the other parameters in the laboratory. The collected e uent showed a natural pH of 8.72 and high turbidity value of 1370 NTU, which may be because of presence of colloidal nes necessitating addition of occulent for improved clari cation. All the physic-chemical properties of the e uent are presented in Table   1. For other characterization studies of coal in e uent, a small amount of slurry was dried at 40 o C and the results are presented in Table 2. Details of the major instruments used for characterization studies are mentioned in Table 3.

2.3
Water and Flocculants used: The washery-water slurry of natural pH 8.72 was used to study settling test. To study the effect of pH of pulp (4.0 and 11.0), the same washery-water was induced by addition of required amount of 1N Hydrochloric acid (HCL) supplied by M/s RFCL Ltd., New Delhi, India and 1N Sodium Hydroxide (NaOH) supplied by Merck Specialities Pvt. Ltd., Mumbai, India. For experimental work three types of occulants were selected namely viz., Cationic occulent -Tel oc-3674F (TF), Anionic occulent -Magna-1011(MF) and Non-ionic -Nalco-83370 + (NF). Settling tests were performed by varying each of the above stated occulants at 3 levels. TF is a medium molecular weight cationic polyacrylamide was supplied by M/s Ion Exchange India Ltd., Patancheru, India. MF is a high molecular weight anionic polyacrylamide supplied by M/s BASF, India. Whereas, NF is a non-ionic polymer supplied by M/s Nalco, India as presented in Table 4. For each test desired amount of freshly prepared occulent was used after diluting it to 0.1% w/w homogeneous stock solution to study the dosage variations. Fresh occulent solutions were prepared at an interval two days to improve the precession of results.

Flocculation Tests And Experimental Studies
A total of 36 Settling tests were performed using three different occulants. For each type of occulent, its dosages were varied at four levels and for each dosage of the occulent, three pH levels were studied. In all these tests a xed pulp solids concentration of 3.0% w/w was kept constant (as per prevailing practice at the plant). All these tests were performed at the room temperature varying between 25 -30°C. The details of the test conducted along with the levels of the variables studied are summarised in Table 5. Each of these settling test were performed in a 1000 cm 3 graduated measuring cylinder as per ASTM (ASTM 2008) standard jar-test method. For homogeneous mixing of slurry, a constant mixing time of 3min was applied to the collected grab sample prior to each tests. The desired amount of occulent was added to the slurry, followed by manual stirring for 30 sec. The change in the interface heights as a function of time was recorded for plotting of the settling curves. The interface heights of slurry for the chosen time intervals of 0, 15, 30, 45, 90, 120, 180, 300, 600 seconds were noted. Using this data, settling rates were established graphically from the slope of the straight line of settling curve obtained as per standard procedure (Fig.6c).
After 600 sec of settling time, an aliquot of the supernatant was taken for the turbidity measurements using turbidimeter.

Discussion On The Results Of Characterization Tests Conducted
The following characterization studies were carried out on the high-ash Dahibari e uent nes: a. Granulometric analyses of coal washery e uent: The particle size distribution patterns expressed in both differential and cumulative forms (by volume percentages) washery e uent are shown in Figs. 2a and 2b respectively indicating about 80% of the particles are below 60 µm in the e uent (Fig.2b).
b. Zeta potential studies: The zeta potential of the suspension plays an important role in occulation. It is a measure of surface charge acquired on the particles in washery e uent. The results of Zeta potential of Dahibari washery e uent as a function of pH is shown in Fig. 3. The zeta potential value of the particles at natural pH of 8.72 was 10.3 mV exhibiting positive surface charge at all pH values except at pH 12. The addition of OHions (from NaOH) is accountable to revert the positive surface charge into negative at alkaline pH 12 of e uent. The positive surface charge may be due the presence of aluminosilicates along with sodium and potassium. However, in the presence of occulants at a pH of 8 the value of the zeta potential gets increased to -70.1mV, -69.1mV and -52.8mV for TF, MF and NF occulants respectively. According to the ASTM D-4187 Standard Test Method, if the zeta potential of colloids (colloidal particles) in water or waste slurry is above -60mV, the stability of the system is classi ed as 'perfect' (Sabah and Erkan 2006). Since, the present case zeta potential for the coal slurry was observed to be 10.3mV at the pH 8, this was classi ed and considered as highly stable condition, suggesting the possibility of increase in settling characteristics through occulation.
c. XRD analysis of coal: The adsorption of a occulent depends on the mineralogical assemblages of the suspended particulate matter, in addition to the prevailing aqueous environment i.e., in the present case the quality of e uent collected from the washery. Therefore, it was necessary to identify the gangue mineral components present in the feed coal. The XRD pattern of the washery e uent coal is represented in Fig. 4, which indicates the presence of quartz, kaolinite and muscovite as the gangue minerals present in the order of increasing intensity of XRD.

Results And Discussion
To identifying a most suitable type of flocculent for improved clarification of coal washery effluent, the results of all the jar settling tests conducted have been analyzed in terms of (i) initial settling rates and (ii) the turbidity of the aliquot supernatant liquid. These two responses were determined as stated below: Estimation of Initial Settling rate: This was determined graphically (as per the standard practice) by plotting the height of the slurry-water interface measured as function of time. The value of the slope of the settling curve estimated graphically in the near straight-line region of initial settling periods is considered as the 'initial settling rate'. This is shown in Fig.6c as a typical example (for a test done at 4gpt of dosage at a pH of 8.72 of the flocculent TF). Together with the settling curve (Fig.6c), a comparison of settling data with the zones noticed in thickeners corresponding to the interfaced seen in a jar test is made in Figs. 6 (a and b).
Suspension Turbidity: Turbidity of the aliquot drawn from the supernatant liquids at the end of 600 seconds were measured by Turbidimeter. These results were used in conjunction with the initial turbidity of the washery effluent to measure the 'Turbidity Removal Efficiency' given as (Osborne 1978 improved bridging mechanism, despite the evidences where, increase in molecular weights beyond certain limit, results in deleterious effects. This in fact may be because; very high molecular weight flocculants generally are more viscous and do not mix properly or get distributed throughout the slurry. This affects the adsorption, which has to be very rapid. In view of this in the present study, three different molecular weight flocculants (of high; MF, medium; TF and low; NF) were selected as specified in Table- (Table 4). Therefore a natural pH 8.72 is found to be effective for all types of flocculants used in this study to enhance the settling rates.

Effect of turbidity analogous to settling rate
Since the Dahibari coal washery effluent consisted of kaolinite (clay) and muscovite minerals. Both being flaky/lath shaped particles, do not settle easily and remain suspended, leading to high turbidity of the aliquot. Turbidity of Dahibari washery gave a value of 1370 NTU at natural pH of 8.72 imparting black colour to the effluent (Table 1). Figures 8 (a to c) shows a maximum turbidity removal of 99.19% (11.1 NTU), 99.28 (9.8 NTU) and 93.20 (93.1 NTU) for TF, MF and NF respectively at 8gpt of floc dosage and natural pH of 8.72. This analogous to the initial settling rates mentioned in section 5.1. A highest turbidity reduction of 9.8 NTU was noticed with MF because of oppositely charged physical bonding between the particles in effluent and the anionic surface charge of flocculent.
Besides this, the variations noticed in turbidity results is due to high molecular weight of polymers leading to formation of large sized but less compact flocs, while the low molecular weight polymers show a reverse phenomenon resulting in formation of small and more compact flocs (Gregory 1989;Hogg 2000;Tao et al. 2000;Sabah et al. 2003).
In general, at slow initial settling rates, the colloidal particles in effluent get more time to interact with polymer and also more inter-particle collisions to form strong flocs resulting in more clarity of supernatant. But in the present case, for the cationic and anionic flocculants, even with higher initial settling rates the turbidity removal efficiencies noticed are high.
However figures 8 (d to f) and 8 (g to i) clearly represent that even with reduced initial settling rate values (noticed at acidic and alkaline pH levels) the turbidity removal efficiencies did not improve. This observations is noticed at all the dosage levels of the flocculants. In other words, even with increased flocculent dosage, the turbidity removal efficiency do not meet the high values observed with that of natural pH of 8.72. This is mainly because of the presence of excess of H + and OHions that hinder the polymer adsorption activity onto the surface of particles present in Dahibari washery effluent. It can be concluded from figure 8 that high supernatant clarity does correspond to high settling rate.

Settling kinetics
The effect of flocculants (TF, MF and NF) on clarification of Dahibari effluent has been assessed for the date of natural pH (8.72) using a standard second-order kinetic model given as (Raj 2001): Where 'a'-is the initial height of effluent in a jar test, and 'x'-is the final interface height of effluent after addition of flocculent measured at time 't'.
From Eq.2, the efficiency of flocculation can be estimated by plotting the second order rate constant K 2 (cm -1 sec -1 ) against time-'t'. Gregory (Gregory 2013)  that, all the curves show almost a linear variation with time upto 90sec, and then deviate with prolonged settling time, finally to remain constant after 120 sec of settling time. Whereas, the low molecular weight flocculent NF although shows the linearity in settling upto 90sec, but an exponential variation in settling is noticed till 240sec of settling time which eventually tends to an asymptotic value i.e. after 240sec of time.
In Figures 10 (a to c), the discrete settling rate constants (∆h/∆t) estimated at a different settling durations are plotted against time. This was felt necessary because, most of the settling curves overlapped each other indicating almost similar settling pattern in Figs.9 (a to c). However, the plots shown in Figs. 10 (a to c) clearly depict a distinguishable variations in their linearity following the second order kinetic settling rate for all the three types of flocculants tested. It is interesting to note that, from the slope of these plots,

Conclusion
Characterization studies conducted on Dahibari washery e uent by various methods indicated e uent is of highly heterogeneous nature witness from its physico-chemical properties.
The collected e uent showed presence of colloidal nes with D80 percent passing size=60 µm which results in high turbidity value of 1370 NTU. Thus hindered the natural settling of coal particles in thickener when observed at a natural pH of 8.72 of e uent without the addition occulent. This lead to conclude that addition of occulent as a necessary step.
The results of the occulation tests indicated that, all the three occulants tests gave satisfactory results for clari cation of the water of Dahibari washery e uent. However, among the three occulants tested, the cationic (TF) and anionic (MF) polymer occulants gave better results in terms of initial settling rates and the measured turbidity values.
Second order kinetic rate constants estimated elaborate the results of effective clari cation occurred between polymer occulants and the suspended particles of Dahibari e uent.
The order of effectiveness of polymer occulants is TF > MF > NF. The overall results of the study indicated that, the TF and MF are most suitable for clari cation of Dahibari coal washery e uent to reutilize the treated water in coal washing circuits and to reduce the fresh water consumption for coal washing process.

Declarations Ethical Approval
The authors of this paper hereby con rm that all the procedures described, tests performed, studies conducted, data presented and the interpretations made thereof on data are in accordance with ethical standards of the institution. However the site map of the Dahibari washery presented and the thickener diagram (Fig. 6a) from the website for which the reference has been cited.

Consent to Participate
Not Applicable.

Consent to Publish
All the authors of this paper hereby give their consent publish the contents of the paper after its peer review. Authors have no objection in publishing the data or analytical contents of the paper in the journal.     Figure 1 Please see the supplementary les section to view the gures.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.