SEPARATION CHARACTERISTICS OF FLY ASH ... Journal of Advances in Engineering & Technology, May 2013....

International Journal of Advances in Engineering & Technology, May 2013.

©IJAET ISSN: 2231-1963

1013 Vol. 6, Issue 2, pp. 1013-1025

SEPARATION CHARACTERISTICS OF FLY ASH PARTICLES IN

HYDROCYCLONE

Suresh1, Chandranath Banerjee2, A.K.Majumder2, S. N. Varma1 1Department of Mechanical Engineering, University Institute of Technology, Rajiv Gandhi

Proudyogiki Vishwavidyalaya, Bhopal, India 2Department of Mining Engineering, Indian Institute of Technology, Kharagpur, India

ABSTRACT

Graded fly ash particles have many useful industrial applications due to its fineness and spherical shape.

However, grading of fly ash in an industrial scale is a challenging task. The possibility of using hydrocyclones

for this purpose has been demonstrated here based on carefully conducted laboratory experiments. The primary

objective of this research is to study the effectiveness of a classifying hydrocyclone to produce graded fly ash

particles. To do this the principal challenge is on the optimization of the process and the design variables to

generate graded fly ash particles of various size ranges. The experimental data reveals that the ratio between

the underflow and the overflow discharge openings, popularly known as cone ratio, of a hydrocyclone actually

controls the particle size separation behavior in a hydrocyclone. Finally, the study reveals that the grading of

fly ash particles at desired size ranges is possible by properly controlling the cone ratio at an appropriate level

of feed inlet pressure.

KEYWORDS: fly ash, classification, hydrocyclone, cone ratio.

I. INTRODUCTION

Fly ash is a by-product of thermal power plants resulting from the combustion of pulverized coal in

the coal-fired furnaces. Inorganic matter present in coal solidifies while suspended in exhaust gases

and ultimately gets collected through electrostatic precipitator. Due to this rapid solidification process

fly ash particles are generally spherical in shape. The fly ash is mainly considered to be a

ferroaluminosilicate element (El-Mogazi et al., 1988; Mattigod et al., 1990) although presence of char

and some amorphous and crystalline phases is common. Carbonates, silicates, sulfates, hydroxide and

oxides of calcium, iron, aluminum, and other metals in trace amount (Adriano et al., 1980) also exist

in flyash. The pH of flyash can vary from 4.5 to 12.0 depending largely on the S content of the parent

coal (Plank and Martens, 1975).The physical and chemical properties of fly ash depend on the nature

of the geo-morphological nature of coal deposit, mining technology, conditions of combustion, type

of emission control devices and storage and handling methods.

Fly ash particles usually belong to size range of 0.01 to 100 μm (Davison et al., 1974). Particle size

distribution of coal fly ash shows different size range (Mahlabaet al., 2011)which arebe classified as

clay-sized particles (particles greater than 1 μm and smaller than 5 μm), silt-sized particles (particles

greater than 5 μm but smaller than 75 μm) and sand-sized particles (particles greater than 75 μm but

smaller than 425 μm). Material scientists from various fields have advocated that graded fly ash (very close size range

product) particles have excellent properties to find application in many potential areas. However, the

major challenge lies in the development of a commercially viable process capable of handling large

tonnages per unit time in grading of fly ash particles due to its fineness. Hydrocyclone is a very

popular unit operation in various engineering fields like chemical engineering, mineral engineering,

food processing etc. mainly for solid-solid and solid-liquid separations. It also finds applications in

solid-gas and liquid-liquid separation. Although hydrocyclone is a versatile unit operation due to its

apparent simplicity in operation, low floor space requirements and high throughput due to very low



1014 Vol. 6, Issue 2, pp. 1013-1025

residence time requirement, the major challenges lie in the selection of an appropriate design for a

particular requirement and in the optimization of the process as well as design variables. However,

application specific designs of hydrocyclones are already commercially available.

The objective of this work is to study whether hydrocyclone can be effectively used for the generation

of closely sized particles of fly ash or not. To do this an in-depth study on how the partitioning

behaviors of solid particles are influenced by the process and design variables of the hydrocyclones is

a pre-requisite. Detailed experiments were carried out using a laboratory model closed-circuit

hydrocyclone test rig varying the important process and design variables.

II. LITERATURE REVIEW

With increasing global population the demand for electricity is also increasing at an exponential rate.

The consumption of coal worldwide is projected to increase by 36% by the year 2020 (Jala and Goyal,

2006; Haynes, 2009). Therefore, the generation of fly ash is mounting at a high rate globally. Disposal

and proper utilization of fly ash are therefore, two very important issues. Fly ash disposal into the

environment causes air, water and soil pollutions. Primarily, fly ash is disposed off using either dry or

wet disposal scheme. In dry disposal, the fly ash is transported by truck, chute or conveyor at the site

and disposed off by constructing a dry embankment (dyke). In wet disposal, the fly ash is transported

as slurry through pipe line and disposed off in impoundment called "ash pond". Coal fly ash caused

groundwater contamination has been reported worldwide. Dumped fly ash contaminates surface and

groundwater (Praharajet al., 2002), soils and vegetation due to the presence of hazardous metals

(Sikka and Kansal, 1995). It also contains many radioactive elements, which causes some harmful

diseases. Therefore, to overcome all these problems research work is continuing on its bulk utilization

and value added product development from it.

Globally, more than 30% of the total annual fly ash produced is utilized (Haynes, 2009) in high-

volume-low-return areas like land filling, road making, cement making etc. at present and the rest of

the amount is being discharged in tailing ponds in the form of slurry. In the process large volume of

water is wasted as losses due to evaporation and the ground water is also getting contaminated due to

leaching of metals from the mineral substances present in fly ash. Large areas of valuable land in the

vicinity of power plants are also getting occupied to store enormous volume of fly ash particles in ash

ponds. This situation is worse in countries like India where the thermal coals generally have ash

contents in the range of 40-45% by weight. The percentage of flyash utilization is different in

different countries. In India, flyash production was 112 million tones in 2005–06 and it is expected to

generate about 150–170 million tons of flyash per year by end of the 2012 (MOEF, 2007).The

utilization of flyash in India has increased substantially in recent years from 13 million tons to51

million tons (MOEF, 2007). Research attention is, therefore, mounting to increase the utilization of

fly ash to partially overcome all those problems.

Many groups are working all over the world on fly ash utilization and many technologies have already

been developed and implemented. Fly ash in a bulk proportion is used mainly in the concrete and

cement manufacturing industries (Ecoba, 2002) due to its pozzolanic properties. Other bulk utilization

includes building roads, mine backfilling and land stabilization in mining areas (Jarvis and Brooks,

1996). There are a number of technical, economic, and legal barriers to the use of large quantities of

coal fly ash. In bulk utilization although high volume of fly ash can be utilized but net return on

investment is very low. On the other hand several value-added products have been developed where

less volume of flyash is utilized but return is very high.

Refractory bricks and ceramic products (Sevelius, 1997), heat and wear resistance ceramic products

(Vilches et al., 2001), cement substitute (Gao et al., 2008), porcelainised granite tiles, synthesis of

high cation exchange capacity (CEC) zeolites (Moreno et al., 2002), mullite as refractory aggregate

for refractory castables cordierite, additives for immobilization of industrial waste water treatment

(Dirk,1996),removal of heavy metal ions from wastewater (Somerset et al., 2008), ‘slash’ (fly

ash/sludge blend) production for soil amendment (Reynolds et al., 1999), sorbents for flue gas

desulfurization (Garea et al., 1997), filter material for the production of different products (Kruger,

1997) are some examples where less volume of fly ash is utilized but return is very high.

Fly ash particles can also be effectively utilized in many other sectors if the particles are graded

properly. For example, particle having the size greater than 45 micron is used as binders and the



1015 Vol. 6, Issue 2, pp. 1013-1025

particles of 0.1 to 5 microns are used for making mineral filler (Styron et al., 1999). Particles in the

size range of 40-150 micron is reported to be used as reactive filler in grout packs, roll Crete, shot

Crete asphalt and lightweight foamed concrete (Kruger 1997). The filler fly ash have three different

particle sizes (25-45µm, 90-105µm & 150-180µm).These particles can also be used in the cement,

concrete, bricks and polymer industries.

Therefore, one of the technical challenges lies in classifying the fly ash particles into different grades

conforming to various close size ranges.

In many engineering disciplines, hydrocyclones are effectively used as a dynamic particle separation

unit based on their size differences only. This is a continuously operating device, which utilizes

centrifugal force to enhance the relative settling velocity differentials between the particles. The

primary objective of this research was to study the effectiveness of a classifying hydrocyclone to

produce graded fly ash particles. To do this the principal challenge was on the optimization of the

process and the design variables to generate graded fly ash particles of various size ranges.

III. EXPERIMENTATION

3.1 Material

Flyash was collected from a local power plant for experimentation purpose. The specific gravity of

the collected fly ash particles was measured to be 2.56. The size analysis of the sample was carried

out using a LASER particle size analyzer (Model MSX-14) and the data is shown in Figure 1 below in

a log-normal scale

Figure1. Particle size distribution of fly ash

The measurement was repeated three times and the relative standard deviation was calculated to be

1.5 which ensures the reproducibility of the measurement technique. It may be observed from Figure

1 that the average size (d50) of the as-received fly ash sample is around 16 micron

A laboratory scale 76 mm diameter Mozley hydrocyclone made of polyurethane fitted in a closed

circuit test rig used for experimental purpose. Flexible arrangements, using flange, were made to fit

the units directly above the slurry tank. To transport the slurry (mixture of water and fly ash particles)

continuously from bottom of the tank to feed inlet of the cyclone a centrifugal pump (Type SP- 2H)

was used. The pump has rubber-lined impeller to minimize the abrasion of the solid particles. From

literature (Govindarajan, 1991 and Verghese, 1994) it was found that the slurry throughput in a 76

mm diameter cyclone varies from 0.5 to 4 liters/sec. Slurry feed rate to the separator and the pressure

at the inlet was adjusted by using the by-pass valve (gate valve). A heavy duty stirrer was fitted

horizontally at the bottom level of the slurry tank to keep the slurry in uniform suspension. The

schematic diagram of the closed circuit test rig is shown in Figure 2

Feed size distribution

0

10

20

30

40

50

60

70

80

90

100

0.1 1 10 100 1000

Particle size (in micron)

vo

lum

e %

pa

ss

ing



1016 Vol. 6, Issue 2, pp. 1013-1025

Figure2. Schematic diagram of the test rig

To collect the representative samples from the overflow and underflow streams for a desired duration,

a properly designed splitter system was fitted on the top of the slurry tank. The splitter consists of a

rectangular channel having a length of 1000 mm, width 500 mm and height 500 mm. A partition wall

of a height 400 mm was given to avoid mixing of overflow and underflow samples. In each

partitioned compartment an opening of 32 mm diameter was given to collect the samples separately.

Each opening was also connected with hose pipes having appropriate diameters to avoid free falling.

The entire arrangement of sample splitter was made up of stainless steel plates having thickness

around 5 mm. The dimensional detail of the sample splitter is shown in Figure 3.

Figure3. Top and front view of the Sample splitter

3.2 Experimental Procedure

The collected fly ash samples were dried in a drying oven at 105 ± 5 oC for 4 hours to remove surface

and entrapped moisture. The dried samples were then mixed thoroughly and representative samples of

10 kg each were sampled out using the conventional coning and quartering method. Slurry having a

TOP VIEW

FRONT VIEW

Slider

Slide

Handle



1017 Vol. 6, Issue 2, pp. 1013-1025

fixed pulp density of 10% by weight was prepared by adding calculated amount of water with

measured quantity of fly ash sample for all the experiments. Slurry temperature was maintained at 40

°C during the experiments throughout. Initially, slurry is pumped into the separator by keeping the by-

pass valve fully open, and subsequently, the opening of the valve was adjusted to obtain required feed

inlet pressure. The system was then allowed to run for a few minutes to attain a steady state. The

attainment of steady state was confirmed once the measured slurry flow rates through both the streams

reached at a constant level.To collect samples separately but simultaneously from the overflow and

underflow streams the sample splitter assembly was used. Before starting the next experiment,

equivalent amount of water and fly ash sample collected from previous experiments were added to

maintain the constant slurry head in the tank. The slurry thus collected for a fixed duration was dried

in an oven to determine the distribution of water and fly ash in the respective streams. The respective

weights of slurry collected through each outlet i.e. overflow and underflow were recorded for further

analyses. The same procedure was followed in all the experiments.

IV. RESULTS AND DISCUSSION

It was mentioned that following the experimental methodologies discussed in detail in previous

section, systematic experimental data was generated first for further analyses. The solid particles

collected through underflow and overflow of the hydrocyclone at different operating conditions were

subjected to particle size analyses. The experimental conditions and the data generated at those

conditions are presented in Table 1.

Table1. Experimental Conditions and the Data

Slurry wt. Solid wt. Water wt. Collection

Time

(sec) Exp.

No.

P

(psi)

VFD

(mm)

SPD

(mm)

O/F

(kg)

U/F

(kg)

O/F

(kg)

U/F

(kg)

O/F

(kg)

U/F

(kg)

1 10 13 10 6.61 4.65 0.27 0.90 6.34 3.75 12.99

1R 10 13 10 6.92 4.74 0.27 0.90 6.64 3.85 13.44

1RR 10 13 10 6.52 4.48 0.26 0.84 6.26 3.64 12.92

2 10 13 15 3.42 9.46 0.13 1.16 3.29 8.30 12.97

2R 10 13 15 3.17 8.67 0.12 1.05 3.05 7.62 11.75

2RR 10 13 15 3.49 9.52 0.14 1.13 3.35 8.39 13.28

3 15 13 15 3.72 9.50 0.14 1.18 3.58 8.32 10.79

3R 15 13 15 3.80 9.86 0.15 1.23 3.65 8.63 11.32

3RR 15 13 15 3.70 9.56 0.15 1.17 3.55 8.39 10.97

4 15 13 10 4.39 6.61 0.81 0.26 3.57 6.35 11.11

4R 15 13 10 4.38 6.70 0.81 0.25 3.57 6.44 11.1

4RR 15 13 10 4.32 6.38 0.79 0.25 3.53 6.13 11.14

5 20 13 10 7.192 4.88 0.27 1.08 6.92 3.80 10.76

5R 20 13 10 7.60 5.04 0.29 1.03 7.31 4.00 11.42



1018 Vol. 6, Issue 2, pp. 1013-1025

5RR 20 13 10 7.61 5.098 0.29 1.02 7.32 4.08 11.68

6 20 13 15 3.64 10.02 0.12 1.28 3.52 8.74 10.79

6R 20 13 15 4.044 10.15 0.13 1.29 3.92 8.86 10.99

6RR 20 13 15 4.06 10.35 0.13 1.30 3.93 9.05 10.48

7 10 18 15 8.00 7.76 0.27 1.24 7.74 6.52 12.36

7R 10 18 15 7.80 7.446 0.33 1.18 7.47 6.27 11.96

7RR 10 18 15 5.12 4.94 0.22 0.77 4.90 4.17 11.78

8 10 18 10 6.13 1.76 0.29 0.51 5.84 1.25 6.37

8R 10 18 10 7.23 2.03 0.37 0.58 6.86 1.45 7.57

8RR 10 18 10 6.88 1.96 0.33 0.55 6.55 1.41 7.43

9 15 18 10 7.74 2.10 0.34 0.68 7.40 1.42 6.61

9R 15 18 10 7.68 2.10 0.35 0.67 7.33 1.43 6.37

9RR 15 18 10 7.12 1.90 0.38 0.60 6.73 1.31 6

10 20 18 10 10.98 2.95 0.47 0.97 10.51 1.98 7.95

10R 20 18 10 9.63 2.56 0.40 0.76 9.23 1.80 7.82

10R 20 18 10 10.59 2.85 0.45 0.92 10.14 1.92 6.95

11 20 18 15 7.12 6.70 0.27 1.10 6.85 5.60 7.89

11R 20 18 15 9.41 8.72 0.35 1.41 9.06 7.31 9.75

11RR 20 18 15 9.77 9.09 0.36 1.45 9.40 7.64 10.27

12 15 18 15 7.68 7.24 0.31 1.21 7.37 6.04 9.52

12R 15 18 15 7.78 7.17 0.31 1.17 7.47 6.00 9.51

12RR 15 18 15 7.32 6.88 0.30 1.10 7.02 5.78 8.93

13 10 25 10 11.33 1.29 0.62 0.67 10.71 0.62 6.7

14 10 25 15 11.95 3.62 0.57 1.00 11.38 2.63 8.3

15 10 25 20 10.88 8.754 0.47 1.47 10.41 7.28 8.64



1019 Vol. 6, Issue 2, pp. 1013-1025

16 15 25 10 9.26 1.008 0.44 0.54 8.82 0.47 3.84

17 15 25 15 9.02 2.554 0.39 0.77 8.63 1.78 4.39

18 15 25 20 6.37 4.668 0.25 0.85 6.12 3.82 3.88

19 20 25 10 11.72 1.398 0.57 0.77 11.15 0.62 4.3

19R 20 25 10 16.77 1.87 0.81 1.02 15.96 0.85 5.93

20 20 25 15 12.45 3.476 0.52 1.09 11.94 2.38 5.31

20R 20 25 15 13.27 3.59 0.54 1.10 12.73 2.49 5.51

21 20 25 20 9.9 7.098 0.37 1.31 9.53 5.79 5.3

21R 20 25 20 9.77 6.786 0.36 1.25 9.41 5.54 5.23

Sometimes in a closed-circuit test rig that is used for the experimentation purposes the feed size

distribution changes because of abrasion and due to the improper design of the stirring system in the

slurry tank. Therefore, the comparative data at various experimental conditions generate misleading

information. To confirm the consistency in feed particle size distributions at various experimental

conditions, the reconstituted feed particle size distributions are compared with the actual feed particle

size distribution as presented in Figure 4. It is evident from this figure that the feed particle size

distribution remains unchanged throughout which reaffirms that the experimental data can reliably be

used for further comparison purposes.

Figure 4: Comparative Plots for Feed Particle Size Distribution

In the first column of Table 1, ‘R’ and ‘RR’ denotes repeat test 1 and 2 respectively. It is evident from

the repeat tests that the variation in the data at a particular experimental condition is marginal. For

example, if we calculate the standard deviations at a 95% confidence level of slurry weights, solid

weights and water weights collected through overflow and underflow at experiment number 1, the

0

10

20

30

40

50

60

70

80

90

100

0.1 1 10 100 1000

Vo

lum

e %

Particle size (in micron)

Feed Exp.5 Exp.6 Exp.7 Exp.8 Exp.9Exp.10 Exp.11 Exp.12 Exp.13 Exp.14 Exp.15Exp.16 Exp.17 Exp.18 Exp.19 Exp.20 Exp.21"Exp1" Exp 2 Exp3 Exp4



1020 Vol. 6, Issue 2, pp. 1013-1025

values are 0.41, 0.27, 0.01, 0.07, 0.4 and 0.21 respectively. Marginal values of the standard deviations

data establish the reliability and the reproducibility of the experimental data generated. For further

analyses, the average values of the repeat data are considered.

4.1 Partitioning behaviours of Slurry, Solid and Water

The spigot diameter (SPD) and the vortex finder diameter (VFD) are the two most important design

variables which affect the classification performance of hydrocyclones. Instead of studying the effects

of the SPD and the VFD independently, cone ratio (ratio between SPD and VFD) has been used to

plot the data presented in Table 1 to understand the particle classification behavior at different

operating conditions. The graphical representations of the data are given in Figures 5 to 7 hereunder

which reflects the trends of the data.

Figure 5: Effect of Cone Ratio on Slurry Split

Figure 6: Effect of Cone Ratio on Solid Split



1021 Vol. 6, Issue 2, pp. 1013-1025

Figure 7: Effect of Cone Ratio on Water Split

From Figures 5 to 7 it is observed that with increasing cone ratio the slurry split, solid split and water

split decrease monotonically. This has happened due to the increased cross-sectional area in the

underflow discharge opening with increasing cone ratio which allows more slurry to pass through.

From the Figures 5-7 it is also interesting to note that at a fixed cone ratio all the three splits remain

almost identical at all feed pressures (P). This is probably signifying that the relative proportions of

slurry, solid and water reporting through a particular discharge opening remain unchanged with

increasing inlet flow rates, which is directly dependent on P, although their individual values will vary

considerably. This is a very significant observation which may be used as a scale-up factor for

hydrocyclones for industrial applications.

A close look at the Figures 5-7 also reveals that the absolute values of slurry split and water split at a

given operating condition are almost identical whereas the solid split values differ considerably. As

the solid concentration for all the experiments were kept at a constant level of 10 % by weight, the

density and the viscosity values of slurry are expected not to differ much from water and as a result

there is hardly any difference in their respective partitioning behaviors. The differences between the

solid split values and slurry split or water split values strongly signifies the classification of solid

particles at all the operating conditions.

4.2 Particle Classification Behavior To understand the particle classification behavior at different operating conditions properly, the %

recovery of each particle size in underflow at each experimental condition has been calculated using

the following formula:

% Recovery in underflow = 100 x (weight of a particle size class in underflow) / (weight

of that particle size class in re-constituted feed)

% Recovery in underflow thus calculated has then been plotted as a function of individual particle

size, the curve thus generated is commonly known as partition curves in mineral processing literature

(Wills’ 1992). The partition curves generated for each experimental condition are then compared, as

shown in Figure 8.



1022 Vol. 6, Issue 2, pp. 1013-1025

Figure 8: Comparison of Partition Curves

Following interesting observations can be made from Figure 8:

(1) Partition curves in all experimental conditions do not follow the conventional ‘S’ shaped

pattern as mentioned in text books (refer wills’ book again), rather they have “Fish-hook”

patterns as described in recent literature (Nageswararao, 2000; Del Villar and Finch, 1992;

Finch, 1983 ; Flintoff, Plitt and Turak, 1987). This means that the recoveries of relatively

finer particle sizesare initially higher than the coarser particle sizes in underflow up to a

critical particle size and after that the recovery increases with increase in particle sizes.

(2) The changes in recovery patterns as mentioned above in point (1) are within a particle size

limit of 10 microns only in all the experimental conditions. The recoveries in all experimental

conditions are minimum at a very close particle size range of 4-6 microns and the recoveries

of 0.6 – 0.7 micron particles are almost equal to 9-10 micron particles. Although mechanistic

arguments behind this typical phenomenon in hydrocyclones have already been given by

Majumder et al. (2003, 2007) further in-depth studies are required in this direction which is

beyond the scope of this research work.

(3) From Table 1 it is obvious that the experimental conditions were different from each other

and hence, it was expected that the partition curves would be different for all the experimental

conditions. However, it is apparent from Figure 8 that all the partition curves can easily be

divided into only four types as shown in the same figure. This probably suggests that although

the experimental conditions were different from one another but the experiments carried out

which represent a particular type of partition curve are hydro-dynamically similar. To confirm

this observation further analysis of the partition curves are made as described hereunder.

The relevant experiment numbers belonging to each type of partition curve, as shown in Figure 8, are

first noted from Table 1. These are as follows:

Type 1: Cone ratio is 1.15 for experiments 2, 3 and 6

Type 2: Cone ratios are 0.77 for experiment numbers 1 and 5, 0.8 for experiment numbers 15,

18 and 21, and 0.83 for experiment numbers 7 and 12.

Type 3: Cone ratios are 0.56 for experiment numbers 8, 9 and 10, and 0.6 for experiment

numbers 17 and 20.

Type 4: Cone ratio is 0.4 for experiments 13,16 and 19

While looking at the experimental conditions representing each type of partition curves, as

mentioned above, interestingly it is observed that the ratios of SPD and VFD at those conditions

are either identical or within a very close range. These ratios are given hereunder.

Type 1: Cone ratio is 1.15 for all the three experiments as mentioned above.

0

10

20

30

40

50

60

70

80

90

100

0.1 1 10 100 1000

% R

eco

very

(Un

de

rflo

w)

Particle Size (micron)

Expt. 1 Expt. 2 Expt. 3 Expt. 6 Expt. 5 Expt. 7 Expt. 8

Expt. 9 Expt. 10 Expt. 11 Expt. 12 Expt. 13 Expt. 14 Expt. 15

Expt. 16 Expt. 17 Expt. 18 Expt. 19 Expt. 20 Expt. 21

Type 1

Type 3

Type 4

Type 2



1023 Vol. 6, Issue 2, pp. 1013-1025

Type 2: Cone ratios are 0.77 for experiment numbers 1 and 5, 0.8 for experiment numbers 15,

18 and 21, and 0.83 for experiment numbers 7 and 12.

Type 3: Cone ratios are 0.56 for experiment numbers 8, 9 and 10, and 0.6 for experiment

numbers 17 and 20.

Type 4: Cone ratio is 0.4 for all the three experiments as mentioned above.

It is interesting to note that at a fixed cone ratio the partition curves at different experiment numbers

are almost getting super imposed on each other. The overall standard deviations calculated between

the respective %recovery data amongst all size distributions at different experimental conditions are

1.46, 1.16, 1.7, 1.15, 2.08, 2.87 and 2.98 for cone ratios of 1.15, 0.83, 0.80, 0.77, 0.60, 0.56 and 0.40

respectively. As the standard deviation values are considerably low and may safely be assumed within

the range of experimental errors, it can be concluded that the cone ratio is the most significant variable

in a hydrocyclone which controls the particle classification behavior. It may further be concluded that

at a given cone ratio the pattern of the partition curves in a hydrocyclone does not change, at least

within the range of the variables studied, at all which means the hydrodynamic conditions inside are

mainly controlled by the cone ratio. This is a very significant observation from the designing and the

operational points of views of any hydrocyclone.

It can also easily be observed from Figure 8 that at experimental conditions representing type 1 and

type 4 curves there is hardly any particle classification. This has happened because the cone ratios are

either too large (1.15 for type 1) or too small (0.4 for type 4) in those experimental conditions. Too

large a cone ratio signifies a condition where the SPD is much bigger than VFD whereas too small a

cone ratio signifies a condition where VFD is much bigger than SPD. In both the conditions the

majority of the solids present in feed will escape through the relatively larger opening along with

water due to availability of more passage which will not yield any particle size separation inside the

hydrocyclone. However, type 2 and type 3 partition curves signify that the fly ash particle

classification at a d50 size of in between 5-12 micron is easily possible with a 76 mm diameter

hydrocyclone if the cone ratio is optimally selected in the range of 0.56 to 0.83.

V. CONCLUSIONS

Based on the above the following conclusions may be drawn:

(1) It was found that at a fixed cone ratio all the three splits remain almost identical at all feed

pressures (P). This signifies that the relative proportions of slurry, solid and water reporting

through a particular discharge opening remain unchanged with increasing inlet flow rates,

which is directly dependent on P, although their individual values will vary considerably.

(2) At a fixed cone ratio the pattern of the partition curves in a hydrocyclone remains identical, at

least within the range of the variables studied, which means that the hydrodynamic conditions

inside are mainly controlled by the cone ratio.

(3) Hydrocyclones can easily be used to classify fly ash particles at desired size ranges if the cone

ratio is selected within the appropriate range.

The most important conclusion of this research is to to establish the technical viability of a specific

design of a hydrocyclone to produce graded fly ash particles of various specifications by properly

controlling its design parameters which can be applicable to classify various other industrial wastes,

apart from fly ash, for their utilization or safe disposal of polluted wastes concentrated into a

particular size range (for example organic pollutants deposition on the surfaces of 20 micrometer or

below particle sizes in dredged sediments).

VI. FUTURE WORK

Bulk utilization of fly ash in various area are well established. A number of research articles have

been published till today based on utilization of fly ash. But utilization of graded fly ash is a new

research area. Most challenging is to grading the fly ash into closely size range product. Most of the

developments and usages are taking place in foreign countries. It became difficult to know the exact

status of technologies and its evaluation without physically carrying out a survey abroad. In India it is

still utilized in bulk application as mentioned above which is economically not profitable.



1024 Vol. 6, Issue 2, pp. 1013-1025

Considerable experimentation and testing is going on to classify the spent flyash available after

mineral extraction so that different fractions could be supplied to different markets. But very few

organizations have been involved in this area and limited research is going on in this area.

Therefore the future research should be the detailed study on the grading of fly ash particle into

different size range by physical separation techniques like hydrocyclone or flotation and application

potential of graded fly ash in making value added products and to study the effect of various grain

size distributions on the reactivity of fly ash or on the strength characteristics of fly ash. This will

increase the utilization of fly ash in many folds and in future fly ash can be considered a potential

resources rather than industrial waste.

REFERENCES

[1]. Adriano, D.C., Page, A.L., Elseewi, A.A., Chang, A.C., Straugham, I., 1980. Utilization and disposal

of fly-ash and coal residues in terrestrial ecosystem: a review. Journal of Environmental Quality 9,

333–344.

[2]. Bakharev T., Thermal behaviour of geopolymers prepared using class F fly ash and elevated

temperature curing. Cement Concrete Res 2006; 36(6):1134–47.

[3]. Bradley.D.1965.”The Hydrocyclone”, pargamon press. London (1965).

[4]. Castro, O., 1990. Pulp rheology effects for hydrocyclone models, M.Sc.Thesis, University of

Queensland (JKMRC), Brisbane, Australia.

[5]. Das SK, Kumar S, Singh KK, Das US. An improved process to produce ceramic tiles using industrial

waste. Patent No. DEL/1800/96; 1996.

[6]. Das SK, Kumar S, Singh KK, Das US. Process for the production of ceramic tiles. US Patent No.

6,743,383, 01.06.2004; 2004a.

[7]. Das SK, Kumar S, Singh KK, Das US. Process for the production of ceramic tiles. Patent No. AU

2.002 E 09, 28.10.2004; 2004b.

[8]. Davison, R.L., Natusch, D.F.S.,Wallace, J.R., Evans Jr., C.A., 1974. Trace elements in flyash:

dependence of concentration on particle size. Environmental Science and Technology 8, 1107–1113.

[9]. Dirk, G., 1996. Pulverized fuel ash products solve the sewage sludge problems of the wastewater

industry. Waste Management 16, 51–58.

[10]. Ecoba (European Coal Combustion Products Association e.V), 2002. http://www.ecoba.com.

[11]. El-Mogazi, D., Lisk, D.J., Weinstien, L.H., 1988. A review of physical, chemical and biological

properties of fly-ash and effects on agricultural ecosystems. Science of the Total Environment 74, 1–

37.

[12]. Finch J.A., Modelling a fish-hook in hydrocyclone selectivity curves. Powder Technol. 1983; 36:127–

129.

[13]. Garea, A., Viguri, J.R., Irabien, A., 1997. Kinetics of flue gas desulphurization at low temperatures:

fly-ash/calcium (3/1) sorbent behavior. Chemical Engineering Science 52, 715–732.

[14]. Govindarajan, B., 1991. Modeling Studies on Vorsyl Separator and Heavy Medium Cyclone, PhD

Tesis, Indian School Of Mines, Dhanbad.

[15]. Haynes, R.J., 2009. Reclamation and revegetation of fly ash disposal sites: challenges and research

needs. Journal of Environmental Management 90, 43–53.

[16]. Jala, S., Goyal, D., 2006. Fly ash as a soil ameliorant for improving crop production –a review.

Bioresourse Technology 97, 1136–1147.

[17]. Kelsall, D.F., 1952. A study of the motion of solid particles in a hydraulic cyclone. Trans Inst. Chem.

Eng. 30, 87-108.

[18]. Kruger, R.A., 1997. Fly-ash beneficiation in South Africa: creating new opportunities in the market

place. Fuel 76, 777–779.

[19]. Kumar R, Kumar S, Mehrotra SP. Fly ash: towards sustainable solutions. In: Proceedings of

international conference flyash India 2005; 2005a. p. X.11.1–12.

[20]. Lynch, A.J., Rao, T.C., 1975. Modeling and scale up of hydrocyclone classifiers. Paper Presented at

11th Int. Min. Process. Congress Cagliari, Italy.

[21]. Physical, chemical and mineralogical characterization of hydraulically disposed fine coal ash from

SASOL Synfuels

[22]. Majumder AK, Shah H, Shukla P, Barnwal JP. Effect of operating variables on shape of ‘‘fish-hook’’

curves in cyclones. Miner Eng.2007; 20:204–206.

[23]. Majumder AK, Yerriswamy P, Barnwal JP. The ‘‘fish-hook’’ phenomenon in centrifugal separation of

fine particles. Miner Eng.2003; 16:1005–1007.



1025 Vol. 6, Issue 2, pp. 1013-1025

[24]. Mattigod, S.V., Rai, D., Eary, L.E., Anisworth, C.C., 1990. Geochemical factors controlling the

mobilization of inorganic constituents from fossil fuel combustion residues: I. Review of the major

elements. Journal of Environmental Quality 19, 188–201.

[25]. Ministry of Environment and Forests (MOEF) Notification, 3rd April 2007. Fly AshNotification 2007.

Ministry of Environment and Forests, New Delhi.

[26]. Mishra D P, Das S K. A study of physico-chemical and mineralogical properties of Talcher coal fly ash

for stowing in underground coal mines. Materials Characterization 2010; 61:1252 –59.

[27]. Moreno, N., Querol, X., Plana, F., Andres, J.M., Janssen, M., Nugteren, H., 2002. Purezeolite synthesis

from silica extracted from coal fly-ashes. Journal of Chemical Technology and Biotechnology 77, 274–

279.

[28]. Nageswararao, K., 2000. A critical analysis of fish-hook effect in hydrocyclone classifiers. Chemical

Engineering Journal 80 (1–3),251–256.

[29]. Nageswararao, K., 1978. Further developments in the modelling andscale-up of industrial

hydrocyclones, Ph.D. Thesis, University of Queensland (JKMRC), Brisbane, Australia.

[30]. Plank, C.O., Martens, D.C., Hallock, D.L., 1975. Effect of soil application of fly-ash on chemical

composition and yield of corn (Zea mays L.) and on chemical composition of displaced soil solution.

Plant and Soil 42, 465–476.

[31]. Plitt, L.R., 1976. A mathematical model of the hydrocyclone classifier.CIM Bulletin 69 (776), 114–

123.

[32]. Sevelius, D., 1997. Key Engineering Materials, 132–136.

[33]. Somerset, V., Petrik, L., Iwuoha, E., 2008. Alkaline hydrothermal conversion of flyash precipitates into

zeolites 3: the removal of mercury and lead ions from wastewater. Journal of Environmental

Management 87, 125–131.

[34]. Styron, R.W., U.S. Patent No. 6242098, 1999.

[35]. Vilches, L. F., Leiva, C., Vale, J., Fernández-Pereira, C., 2005. Insulating Capacity of Fly Ash Pastes

Used for Passive Protection against Fire. Cement & Concrete Composites, 27, 776 -781.

[36]. Del Villar, R. and Finch, J.A., 1992. Modelling the Cyclone Performance with a Size Dependent

Entrainment Factor, Minerals Engineering, 5 (6), 661.

[37]. Wills, B.A., 1992, Mineral Processing Technology, 5th edition, Butterworth-Heinemann

BIOGRAPHIES

Suresh Boriah received his Bachelor of Technology in Mechanical Engineering and Master of

Technology in Mechanical Engineering. He is an Assistant Professor in Globus Engineering.

College, Bhopal. He is presently doing his Ph.D study in the Rajiv Gandhi Proudyogiki

Vishwavidyalaya, Bhopal, India. His area of research is hydrodynamic modeling of cyclone

separator.

Chandranath Banerjee received the Bachelor of Science degree in Chemistry and Bachelor of

Technology in Chemical Technology from Calcutta University and Master of Technology from

Indian Institute of Technology, Kharagpur where he is currently pursuing the Ph.D. He is

currently doing research on hydrodynamic modelling of Hydrocyclone, CFD analysis of

swirling flow and solid- fluid interaction in turbulence medium.

A. K. Majumder is presently working as an Associate Professor of Department of Mining

Engineering, IIT Kharagpur. He received his Ph.D degree from university of Queensland,

Australia in Mineral Processing. His area of research is Mineral Processing, Coal Washing,

Solid-Fluid Interactions and Fine Particle Processing. He has published more than 100 technical

paper in reputed international and National journal. Before joining IIT Kharagpur he worked as

a scientist in RRL Bhopal.

S. N. Verma is presently serving as a Professor in Mechanical Engineering Department in Rajiv

Gandhi Proudyogiki Vishwavidyalaya, Bhopal, India. He received his Bachelor of Engineering,

Master of Technology and Ph.D in Mechanical Engineering. His area of Specialization is Fluid

Mechanics and Operation Research.

SEPARATION CHARACTERISTICS OF FLY ASH ... Journal of Advances in Engineering & Technology, May 2013....

Documents

Transcript of SEPARATION CHARACTERISTICS OF FLY ASH ... Journal of Advances in Engineering & Technology, May 2013....