University
stringclasses 19
values | Text
stringlengths 458
20.7k
|
|---|---|
Virginia Tech | ACKNOWLEDGMENTS
The author would like to express his sincere appreciation to his advisor, Dr. Roe-Hoan
Yoon, for the guidance, inspiration, suggestions and criticisms. Special thanks are also given to
his committee members, Dr. Luttrell, Dr. Adel, and Dr. Ismail for their continued interests,
useful suggestions, constructive comments and discussion to understand the overall process.
The financial support from the National Energy Technology Laboratory and Department
of Mining and Minerals Engineering is deeply appreciated.
I owe special gratitude to Mr. David Brightbill, Steve Blubaugh and Steven Abbatello for
their helpful suggestions and continuous support. I am also grateful to Dr. Jinming Zhang, Mr.
Serhat Keles, for their friendship and support.
I would like to express my most sincere appreciation to my parents Ziynet and Ali
Eraydin for their inspiration, encouragements, moral support and continued suggestions. I would
also like to thank my sister Feyza for her continued support.
Finally, I would like to express my thanks to Emily A. Sarver for all her encouragements,
patience and dedication.
iv |
Virginia Tech | Figure 4.3 Engineering flowsheet for the POC-scale pond reclaim facility……..……… 160
Figure 4.4 Simplified process flow diagram for the pond reclaim facility……………… 163
Figure 4.5 Nearly completed plant incorporating the POC circuitry……………...…….. 164
Figure 4.6 Classifying cyclones used to provide a 100 mesh cutsize…………………… 164
Figure 4.7 Screen-bowl centrifuge used to dewater 100x235 mesh product……..…….. 164
Figure 4.8 Deslime cyclones used to perform a nominal 325 mesh cutsize…………….. 164
Figure 4.9 Three-stage conditions used for conditioning the dewatering aid…………… 165
Figure 4.10 Bank of disc filters used to dewater the fine coal froth product…………….. 165
Figure 4.11 Static thickener used to thicken solids and clarify process water…………… 166
Figure 4.12 Paste thickener used to further thicken wastes for disposal…………………. 166
Figure 4.13 Weight and ash distributions for samples collected from the shakedown
tests conducted on the coarse coal treatment circuits………….…………….. 168
Figure 4.14 Weight and ash distributions for samples collected from the shakedown
tests conducted on the fine coal treatment circuits……………..……………. 169
Figure 4.15 Moisture content versus dewatering aid dosage (RV)………..…….……….. 174
Figure 4.16 Moisture reduction versus dewatering aid dosage (RV)………..…………… 174
Figure 4.17 Effect of dewatering aid addition on vacuum filter pump power demand…... 175
Figure 5.1 Tergitol NP-7………………………………………………………………… 195
Figure 5.2 PPG-400……………………………………………………………………... 210
xi |
Virginia Tech | Table 2.14 Coal Clean coal (minus 325 mesh) dewatered using RW, RU, and RA
dispersed in Nalco 01DW110………………………………………………. 84
Table 2.15. Coal Clean coal (50% minus 325 mesh and 50% 100x325 mesh) dewatered
using RW and RU dispersed in Nalco O1DW110…………………….…….. 85
Table 2.16 Coal Clean coal (75% minus 325 mesh and 25% 100x325 mesh) dewatered
using RW, RU and RA dispersed in Nalco 01DW110……………...………. 86
Table 2.17 Effect of disc speed on product rate and moisture content for Coal Clean
coal (minus 325 mesh) from the filter module……………………………… 86
Table 2.18 Effect of Reagent 01DW133 on Coal Clean coal (50% minus 325 mesh and
50% 100x325 mesh) from the centrifuge module…………………….…….. 87
Table 2.19 Effect of reagent addition on Concord flotation feed (100 mesh x 0) using
RV and RW3…………………………………………………………..…….. 92
Table 2.20 Pilot-scale test results obtained on the Concord flotation feed sample (100
mesh x 0) using various vacuum pressures…………………………..……… 93
Table 2.21 Pilot-scale test results obtained on the Concord flotation feed sample (100
mesh x 0) using various reagent combinations………………………...……. 94
Table 2.22 Pilot-scale test results obtained on the Concord flotation feed sample (100
mesh x 0) using various reagent combinations...…...………..……….......... 95
Table 2.23 Pilot-scale test results obtained on the Concord flotation feed sample (100
mesh x 0) using various reagent combinations……………………………… 95
Table 2.24 Screen analysis of the Buchanan flotation product used for dewatering tests 98
Table 2.25 Screen analysis of the Buchanan filter feed used for dewatering tests…....... 98
Table 2.26. Effect of reagent addition on dewatering of Buchanan’s filter feed………… 99
xiii |
Virginia Tech | Table 2.27. Effect of reagent addition on dewatering Buchanan’s flotation product……. 99
Table 2.28. Effect of mixing intensity on dewatering of Buchanan’s flotation product… 101
Table 2.29 Effect of pilot-scale filter disc speed on filter cake production rates and
moisture content………………………………………....……….………….. 102
Table 2.30. Effect of reagent addition on the pilot-scale dewatering of Buchanan’s
flotation product …………………………………………………………...... 102
Table 2.31. Effect of pilot-scale filter vacuum level on filter cake moisture contents….... 104
Table 2.32. Effect of pilot-scale filter vacuum level on filter cake moisture contents (4
lb/t RW)…………………...………………………………………...……….. 104
Table 2.33. Effect of pilot-scale filter vacuum level on filter cake moisture contents (4
lb/t RU)…………….………………………………………………………… 104
Table 2.34 Effect of reagent addition on dewatering of Elkview’s filter feed sample
(standard metallurgical coal)………………………………………………… 107
Table 2.35. Effect of using RW, RU and RV on the dewatering of Elkview’s filter feed
(STD Met Coal)……………………………………………….........………… 108
Table 2.36 Effect of using RW, RU and RV for Elkview’s filter feed (Mid-Vol Met
Coal)…………………………………..………………….…………………. 109
Table 2.37. Effect of using RW, RU and RV on Elkview’s filter feed sample (Mid-Vol
Met Coal)…………………………….………………………………………. 109
Table 2.38 Effect of RW addition on the Pinnacle pond sample…………….………….. 119
Table 2.39 Effect of reagent addition on dewatering of Pinnacle thickener underflow
sample…………………………………………………………..…………… 120
Table 2.40 Effect of desliming on dewatering of Pinnacle thickener underflow sample. 121
xiv |
Virginia Tech | Table 2.41 Effect of reagent addition on dewatering of Pinnacle thickener feed sample. 121
Table 2.42 Effect of RW addition on the pilot scale dewatering of the Pinnacle-Smith
Branch Impoundment sample using the mobile units……………………….. 122
Table 2.43 Effect of RW addition on the pilot scale dewatering of the Pinnacle
thickener feed sample using the mobile test units……………..…………….. 123
Table 2.44 Effect of reagent addition on dewatering of Pinnacle thickener feed sample. 123
Table 2.45 Effect of reagent addition on the dewatering of the froth products obtained
using 400 g/t diesel as collector………………………………………..……. 128
Table 2.46 Effect of conditioning time on the dewatering of flotation product using
2 lb/t RV……………………………………………………………………... 129
Table 2.47 Effect of cake thickness on cake moisture in the presence of 3 lb/t RV…….. 129
Table 3.1 Effect of reagent type and dosage on dewatering of mixture sample……….. 139
Table 3.2 Effect of filtration time on dewatering of mixture sample (0.5 in. cake)…... 140
Table 3.3 Effect of vacuum pressure on dewatering of mixture sample (0.5 in cake)… 141
Table 3.4 Effect of cake weight and thickness on dewatering of mixture sample….….. 142
Table 3.5 Sample data collection sheet………………………………………………… 143
Table 3.6 Simulation results on Mingo Logan’s mixture sample………………...……. 150
Table 3.7 Simulation results on Mingo Logan’s flotation sample product……....……. 151
Table 4.1 Effect of RV and RW (dissolved in diesel at 1:2 ratio) at various
dosages………………………………………………………………………. 172
Table 4.2 Effect of RV and RW (dissolved in diesel at 1:2 ratio) at various dosages
on dewatering of floated sample……………………………………………. 173
xv |
Virginia Tech | Table 4.3. Effect of RV (dissolved in diesel at 1:2 ratio) at various dosages on
dewatering of deslimed samples…………………………..………………… 173
Table 5.1 Effect of amine (DAC) addition as the first step on dewatering of Georgia
clay……………………………………………………………..……………. 184
Table 5.2 Effect of RW addition as the second step (10 lb/t DAC & various dosages
of RW) on dewatering of Georgia clay………………………….………..… 185
Table 5.3 Effect of using 10 lbs/t amine and RW on the dewatering of PSD Filter Vat
kaolin clay sample after dilution to 15% solids at pH 7.0..…..................… 187
Table 5.4 Effect of using 10 lbs/t amine and RW on the dewatering of PSD Filter Vat
kaolin clay sample after dilution to 15% solids at pH 7.0…….................... 188
Table 5.5 Effect of using amine only on the dewatering of PSD Filter pre-leached
kaolin clay sample (pH 9.5& 30% solids)………..…………………….…… 189
Table 5.6 Effect of using amine only on the dewatering of the coarse pre-leach kaolin
clay sample from (pH 7 & 30% solids)………………...………………..….. 189
Table 5.7 Effect of using amine (10lb/t) and RW on the dewatering of PSD Filter pre-
leached kaolin clay sample at 30% solids at pH 9.5 and pH 7.0…………... 190
Table 5.8 Effect of foam on dewatering of Georgia clay at pH 7.0 when Tergitol NP-7
and NP-9 is used as foam generating agents……………………....………… 194
Table 5.9 Effect of Tergitol with thick cake (5-6mm) on dewatering of Georgia clay... 194
Table 5.10 Effect of foam on dewatering of Georgia clay at pH 3 when Tergitol NP-7
is used as foam generating agent……………………………………..……… 197
Table 5.11 Effect of alum addition on dewatering of Georgia clay when Tergitol was
used as foaming agent…………………………………………….…………. 197
xvi |
Virginia Tech | Table 5.12 Effect of reagent addition at (1:2) ratio on dewatering of Kentucky coal
sample at 17-19 inches Hg vacuum…………………………………..……… 202
Table 5.13 Effect of amine addition at (1:2) ratio on dewatering of Kentucky coal
sample at 17-19 inches Hg vacuum…………………………………..……… 203
Table 5.14 Effect of amine (10 lb/t) and RW addition at (1:2) ratio on dewatering of
Kentucky coal sample at 17-19 inHg vacuum………………….……..…….. 204
Table 5.15 Effect of amine addition as the first step on dewatering of fly ash sample at
20 in Hg Vacuum pressure……………………………….…....…………….. 205
Table 5.16 Effect RW, RU and RV addition at (1:2) ratio on dewatering of amine pre-
treated (3 lb/t of amine) Kentucky fly ash sample at 20 in Hg vacuum......... 205
Table 5.17 Effect of amine addition as the first step on dewatering of fly ash sample at
30psi pressure………………………………………………………………… 206
Table 5.18 Effect of amine (3 lb/t) and RW addition at (1:2) ratio on dewatering of
Kentucky coal sample at 30psi pressure……………………………..……… 206
Table 5.19 Effect of amine (1lb/t) and RV addition as the first step on dewatering of fly
ash sample at 20 in Hg vacuum pressure………………….……….………... 207
Table 5.20 Effect of using amine (1lb/t) and RV on the dewatering of the fly ash
sample at 30% solids………..………………………………………………. 208
Table 5.21 Effect of PPG addition on dewatering of Kentucky fly ash sample at 17-19
in Hg vacuum…………………………….………………………………….. 209
Table 5.22 Effect of Tergitol addition on dewatering of Kentucky coal sample at 17-19
in Hg vacuum………...…………………………...…………………………. 210
xvii |
Virginia Tech | CHAPTER 1 INTRODUCTION
1.1. INTRODUCTION
Fossil fuels, especially coal, oil and gas, are of great importance since they can be burned
to produce energy. The Energy Information Administration estimates that in 2005, fossil fuels
made up approximately 86% of United States’ energy production and 76% of world
consumption. The remaining sources are non-fossil sources such as hydroelectric, nuclear and
other (geothermal, solar, wind, wood and waste) at 6.3%, 6.0% and 0.9 % used for energy
production, respectively.[1, 2]
Fossil fuels are the most important energy sources; however, they are non-renewable.
According to Oil & Gas Journal estimates, years of production left in the ground for fossil
reserves are 45 years for oil, 72 years for gas and 252 years for coal. Of these fossil fuels coal
has the most widely distributed reserves and it is mined in over 100 countries and on all
continents except Antarctica. The total recoverable world reserve for coal was estimated by
International Energy Annual-2005 to be around 908 million tons. British Petroleum’s statistical
review of world energy data from 2007 shows that the United States has enormous coal
resources and recoverable reserves. Coal is classified into six types or ranks (peat, lignite, sub-
bituminous, bituminous, anthracite and graphite) which depend on the amount and the types of
carbon it contains and on the amount of heat energy it can produce. In the United States, the
most widely used coal types are lignite, sub-bituminous, bituminous and anthracite. Table 1.1
shows the proved recoverable coal reserves in millions of tons for the United States and the top
10 countries as of the end of 2006.[1-4]
1 |
Virginia Tech | In United States, coal represents approximately 95 percent of the nation’s fossil energy
reserves. Coal is mainly found in three large regions: the Appalachian Coal Region, the Interior
Coal Region, and the Western Coal Region (includes the Powder River Basin). It is currently
produced in 26 states, with most of it mined in Wyoming, followed by West Virginia, Kentucky,
Pennsylvania, and Texas. The Energy Information Administration reports that 1,162.8 million
short tons of coal were produced in 2006. Currently, approximately 50% of the electricity is
produced by using coal and there are approximately 600 power plants. Coal is also one of the
nation’s lowest-cost electric power sources (DOE). Thus, today the electric power sector drives
the coal demand for electricity production and is almost responsible for 90% of the coal
consumption (EIA). Figure 1.1 shows the comparison of electricity produced from the major
energy sources in United States.[1-3]
1.2
1
0.8
0.6
0.4
0.2
0
1940 1950 1960 1970 1980 1990 2000 2010
Figure 1.2 Coal consumption by sector (Energy Information Administration)
3
snoT
trohS
noilliB
Electric Power
Industrial
Residential, Commercial and
Transportation
Year |
Virginia Tech | The increase in demand resulted in tremendous expansions on coal mining and capacity.
After the 1970s this demand created more mines, especially above-ground. The National Mining
Association reports that, of the two main mining methods, surface and underground,
approximately two-thirds of today’s coal production is surface mining. The increase in the need
of coal also improved mining technologies to produce additional coal. This additional coal is
prepared to today’s standard specifications dictated by the users. Thus, coal processing
technologies is advanced to produce a cleaner and higher heating content end product.[5]
1.2. LITERATURE REVIEW
1.2.1. Coal Processing
Coal preparation includes four major methods, including blending and homogenization,
size reduction, grading and handling, and most importantly beneficiation or cleaning for high
quality product. Regardless of the intended utilization purposes, there are levels of cleaning to
which coal can be economically subjected.[6-10]
Coal preparation is a very important practice for processing industries, and market
demand determines the selection of preparation methods. Therefore, as shown in Figure 1.4,
there are different methods of economically cleaning coal.[6-11]
In coal processing, it is typical to employ an operation by starting with the crushing and
screening of run-of-mine (ROM) coal into coarse, fine, and occasionally intermediate sizes.
Being exclusive of fine sizes, the coarse and intermediate size fractions are then upgraded by
gravity concentration e.g. dense-medium baths, jigs, dense-medium cyclones, etc. Because the
differences in densities between pure coal particles and liberated mineral impurities are sufficient
to achieve an ideal cleaning process, these methods are dominantly used. The fine and ultra-fine
5 |
Virginia Tech | pass through a medium. Solids are retained on the surface of the medium, creating a cake
formation, which the liquid media flows through. Sedimentation is a separation method that
benefits from the differences in phase densities of solids and liquids by allowing solids to sink in
the fluid under controlled conditions.[20, 22-29]
In the conventional dewatering processes, thickeners, dewatering screens, vacuum filters
(drum, disc, and belt), centrifuges, pressure (hyperbaric) filters are utilized to remove the surface
water. Vacuum filtration is one of the dewatering processes widely used in the coal industry, its
advantage being a continuous operation that can be utilized under relatively simple mechanical
conditions. The removal of free water from the surface of fine particles is difficult and
unsatisfactory by mechanical methods. The problems associated with the dewatering of fine
particles are complicated. The finer particles have a larger total surface area than the coarse
particles, causing very high water retention and smaller capillaries in the filter cake. Eventually,
this results in high capillary pressures and slower dewatering rates.[19, 20, 23-26]
In general, the costs of cleaning fine particles are approximately 3 times higher than those
for coarse particle. This leaves coal producers only a few options as to what to do with the fines
in economical terms. The fine products also contain higher levels of impurities, ash and sulfur,
that lead to environmental concerns. Therefore, fine particles smaller than 500µm, and ultra-fine
particles smaller than 100µm, are abandoned with the discard streams if they constitute only a
small fraction (5% to 10%) of the product stream. This has been the case for many U.S. coal
producers. As a result, 30 to 40 million tons of fines have been discarded to waste ponds
annually, representing a loss of recognized, exploitable natural resources.[8, 17, 18, 21, 31, 36]
The key reason for not completely exploiting this energy resource is the cost of the
cleaning process as well as dewatering the high levels of moisture associated with the fine
7 |
Virginia Tech | fraction of coal. The dewatering of fines results in a significant expense reporting to overall
cleaning expenses because dewatering costs increase severely when the particle size is smaller
than 500µm. The cost also includes thermal drying, which is the only practical method of drying
fine coal to further decrease the moisture content. An acceptable level of moisture reduction is
usually below 10% (by weight). Even though preferred moisture levels can be achieved by
thermal drying, it is a capital-intensive and costly technology compared to mechanical
dewatering methods.[31, 33, 40, 44-46]
There have not been any significant technical innovations in fine particle dewatering in
decades because most of the fine fraction was sent to waste ponds. This lack of technological
knowledge generated approximately two billion tons of fine coal in waste ponds to date, and 500
to 800 million tons of the fines are still in active ponds. On the other hand, in recent times, the
industrial demand for coal has increased, and recovery of this size fraction has become more
important. Recent advances in the recovery of fine and ultra-fine particles by flotation have also
lead to more fine size coal production, creating an incompatibility between efficient, cleaner coal
production and insufficient, fine-particle dewatering techniques.[9, 23, 31, 46, 47]
The need for understanding and enhancing fine-coal dewatering will be a considerable
contribution to the performance of studies to meet the industry’s needs. Studies on fine-coal
dewatering will increase the availability of efficient dewatering processes that can provide lower
filter cake moisture, resulting in reduced thermal drying costs, reduced transportation cost,
improved product quality, increased calorific value, and minimized freezing during winter
storage.[31, 36, 48]
This study was carried out to achieve a better understanding to solve the problems
associated with fine particle dewatering. For this reason, two new dewatering technologies,
8 |
Virginia Tech | which have been developed at Virginia Polytechnic Institute and State University, were tested.
The first technology utilizes novel dewatering aids that have revealed promising options to
receive significantly lower moisture. The other technology is the utilization of foam-supported
dewatering, which is superior when compared to the current, commonly-used technologies.
1.2.2. Coal Dewatering
Because flotation is accomplished in an aqueous solution to produce clean coal, the
product contains approximately 80% water. Although, coal fine products may represent only
20% of the weight of a preparation plant feed, this fraction is accountable for almost two-thirds
of the final product moisture. From a utility viewpoint, a one-ton decrease in moisture can offset
four tons of steam coal. This steam coal can be added to clean coal product, strongly indicating
that the success of coal utilization is critically dependent on solid/liquid separation
technology.[23, 49, 50]
As the first step in dewatering, large settling tanks can be used to remove the free water,
where the slurry is thickened from 35% to 75% solid content. The second step involves
subjecting the pulp to filtration methods including vacuum, drum, disc, and belt filters,
centrifuges, and pressure (hyperbaric) filters to remove the remaining surface water. Despite all
of this, the fine coal, filtered by using these mechanical methods, may still include undesirable
amounts of water in their compositions. Thermal drying, the only fully-developed method to
lower the moisture to single digits, can further decrease cake moisture contents to attain desired
levels; however, the associated high energy intensives, operation costs, and special installation
permissions limit the employment of thermal driers.[20, 26, 51-53]
9 |
Virginia Tech | Thermal drying is the last operation conducted on dewatered solids when these two mechanical
solid-liquid separation methods are not sufficient.[22, 24, 54]
At processing plants, dewatering is practiced normally in a combination of the above
methods. First, the bulk of the water is removed by sedimentation methods, followed by
filtration methods. If needed, thermal driers are used to produce the desired final moisture
content. Several common factors influence all solid-liquid separation steps and equipment
selection, such as solid concentration, particle shape, specific gravity, surface characteristics,
liquid viscosity, and, most importantly, particle size and size distribution. Figure 1.6 describes
100
80
Static Thickeners
60
Cyclones, Sieve Bend, Vor-Sieve
40
.
Vac Drum Filter
30
Vacuum Filter Disc Type High Speed Vibrators
20
10 Solid Bowl/Screen Bowl Centrifuge
High Speed Vibrators
5 Positive Discharge Type Basket Centrifuge
Vibrating Basket
Centrifuge
Thermal Drying Ranges
2 Vibrating Screens
325M 200M 100M 48M 28M 20M 10M 6M 4M Mesh
0.001 0.005 0.01 0.05 0.1
0.50 1.0 2.0 5.0 Inches
Figure 1.6 Commonly used dewatering equipment in coal industry for various size ranges
and corresponding approximate moistures
11 |
Virginia Tech | the tools that are generally employed for different size fractions of coal to be dewatered.[8, 9, 20,
26, 51-53]
In general, sizes larger than 1.5 inches do not exhibit any particular/serious dewatering
problems since very basic types of shaker screens can produce products with low moistures. For
intermediate to 0.02inch particle size dewatering, high-speed vibrators may be used. Usually, for
dewatering of 3/8 to ¼inch size particles, centrifuges of various types and designs are used,
where the centrifugal force can be employed to promote/aid dewatering. For sizes smaller than
0.02 inch, solid bowl and screen bowl centrifuges can be used, even though higher moisture
values may result than what is desired. Vacuum filtration becomes increasingly important with
finer sizes (-30 mesh) for further dewatering.[8-10, 15, 41, 55, 56]
a) Sedimentation Techniques
Sedimentation is a collective term describing the gravity separation of the fine solids,
usually under quiescent conditions, resulting in the formation of a sedimentary layer of solids
and a relatively clear supernatant liquid. It is mainly used at the very early stages of the
dewatering process to increase the solid content of the slurry in large capacity thickeners.
Depending on the particle size and the solid percent, sedimentation processes involve the settling
of solids in slurry by either employing gravitational or centrifugal force. Because the settling
velocity of the very fine particles is extremely slow by gravity alone, the centrifugal force will
have greater affect on settling time and velocity. Alternatively, the particles may be
agglomerated into larger lumps to facilitate the dewatering. The settled solids are then collected,
removed, and introduced to filtration to further reduce the moisture content of the cake before
thermal drying.[20, 26, 41, 44, 52]
12 |
Virginia Tech | b) Centrifugation Techniques
Centrifugal Separation can be considered an extension of gravity separation because the
equipment creates high gravity forces to increase the settling rate of particles for the purpose of
solid-liquid separation. Compared to thickeners, centrifugal separation does not require a density
difference between the solid and the liquid. Even though high maintenance costs exist,
centrifugal dewatering is commonly the most effective mechanical method as a result of these
high forces’ ability to dewater the particles in a short time and a continuous manner.
Centrifugation is primarily utilized for mineral and coal processing industries because it can
collectively dewater a wide range of sizes (normally with a 37.5mm upper limit, where fine is
0.5mm x 0).[8, 9, 29, 41, 61]
Centrifugal separation can be executed by using cyclones or centrifuges. Cyclones are
very simple and cheap, but they suffer from limitations such as low efficiency when dealing with
fine particles and their inability to use flocculants due to high shear forces. Therefore, cyclones
are considered more like classifiers than thickening tools.
Centrifuges are generally classified into two groups.
a) perforate-basket type
- without transporting device
- with positive discharge system
- vibrating basket
b) bowl type
- co-current solid bowl
- countercurrent solid bowl
- screen bowl
13 |
Virginia Tech | In coarse coal dewatering, perforated basket centrifuges are most commonly used, and
bowl type centrifuges are most generally used for fine particles. Bowl type centrifuges are
commonly used to dewater coal at sizes from approximately 10mm to 1.0 mm. Two types of
bowl centrifuges may be used, solid bowl centrifuges and screen bowl centrifuges.[8, 9, 29]
c) Filtration Techniques
Filtration is a widely utilized dewatering application in mineral processing industries and
generally occurs after thickening. In filtration, there are four types of driving forces employed to
obtain flow through the filtering medium: gravity, vacuum, pressure, and centrifugal forces.
There are basically two types of filtration used in practice, surface filters, in which the solids are
deposited in the form of cake on the upstream side of the relatively thin filter medium, and depth
filters, in which particle deposition takes place inside the medium. In coal-preparation
applications, most filters are surface filters, employing vacuum and pressure forms of driving
force.[26, 29, 51, 52, 62, 63]
Vacuum filtration can be categorized into two groups: batch and continuous. In coal
dewatering, where continuous filters are widely employed, batch vacuum filters are not practical.
There are several types of vacuum filters that are used for fine particle dewatering. Three types
of vacuum filters are rotary drum, rotary disc, and horizontal belt (HBF), or disc, filter.[26, 44,
52]
Rotary vacuum drum filters are the most widely used continuous filters for fine coal and
mineral particles. They utilize a drum partially submerged into a tank of agitated slurry. Once
the vacuum is applied, cake is deposited on the drum surface and discarded by various types of
mechanisms, such as fixed knife and air blowing. Effluent is drained by different methods,
depending on the manufacturer’s design. Adjustable operating parameters, such as drum rotation
14 |
Virginia Tech | speed (rpm), applied vacuum pressure, and submergence, dictate the performance of the drum
filter. Changes in these conditions affect the cake formation, drying, throughput, and the degree
of dewatering achieved.[26, 44, 52, 64]
The key advantages of drum filters are i) effective washing and dewatering properties, ii)
low labor and operating costs, iii) wide operation variations, and iv) easy maintenance and clean
operation. The main disadvantages are i) high capital cost, ii) large space requirements, iii)
incompatible for fast-settling slurries, and low efficiency with ultra-fine particles that blind the
filter cloth.[26, 44, 52, 64]
Rotary vacuum disc filters have almost the same fundamental design as the drum filters.
Disc filters consist of a number of flat filter elements, mounted on a central shaft and connected
to a normal rotary vacuum filter valve. As the unit rotates, the discs are submerged in slurry
contained in a slurry tank and agitated. Gradually, cake is formed and dewatered as the unit
rotates out of submergence. The filter cake is usually removed by a combination of scraper
blades and a blowback mechanism. The disc filters have a low capital cost per unit area, and
they supply large filter areas in smaller floor areas; however, blowback systems may cause
higher moisture, and cake washing cannot be done efficiently.[26, 44, 52, 64]
Horizontal belt filters are continuous filters and consist of an endless reinforced
perforated rubber belt with drainage channels, where the vacuum is applied. The filter
medium/cloth sits on the rubber belt and moves along with it. The suspended slurry is fed from
one end of the filter to produce cake, and filtrate is collected in a tank to be pumped out as
effluent. The horizontal belt filters occupy large floor areas, and the installation costs per filter
area are high; however, being fully automatic, flexible, and having relatively high speeds of
operation offset these weak points.[22, 24, 29]
15 |
Virginia Tech | The pressure filters normally perform in a batch-wise manner under positive (air or
hydraulic) pressures to remove water and retain solids in the form of a cake. Pressure filters are
utilized very often in process industries that deal with fine, slow-settling particles exhibiting low
filterability and suspensions that contain higher solid contents. Pressure filters have advantages
over vacuum filters due to higher pressures and vertical incompressibility of solids. In these
units, high pressure creates an increased dewatering rate and lower filter cake moisture. On the
other hand, the discharge of the cake in a continuous manner from the inside of the unit is
difficult and therefore pressure filters are usually employed as batch units. The high capital costs
and inefficient returns associated with batch units create an additional economic disadvantage.
Batch, chamber filter presses and continuous, belt filter presses are two distinct types of pressure
filters most frequently utilized in coal dewatering.[25, 26, 52, 65]
d) Thermal Drying
Thermal drying of minerals and coal is the last and the most expensive unit operation
performed on the dewatered materials before transportation. For that reason, the surface area of
particles increases proportionally with the fineness of size and the final cake moisture. The
coal’s ultimate dewatering cost is strongly related to the amount of the fines. Thermal dryers are
utilized to generate low moisture, maintain high coal-pulverized capacity in power plant
applications, reduce heat loss, prevent freezing, and ease handling, storage and
transportation.[66-68]
There are different types of dryers available, but only a small number of them are used in
coal preparation. Coal thermal dryers can be categorized into two main groups, direct or indirect
heat exchange. The most common dryer in use today for coal preparation industries is the direct-
heated, fluid-bed type dryer. This type of dryer is generally used for fine particle drying, where
16 |
Virginia Tech | hot gas passes through the fine particles inside the dryer and removes water from the unit. In the
indirect heat exchange method, the fine particles in the chamber of the dryer are externally
heated by hot gas to obtain dry product. It is mostly employed when environmental concerns
arise. [66-68]
1.2.4. Dewatering Parameters
The performance of most of the dewatering tools that are used today depends strongly on
several parameters of particle-aqueous systems. Parameters which affect the dewatering process
include equipment properties, mineral type, particle size and distribution, physical and chemical
properties of the mineral surface, cake structure and thickness, impurity content, surface
oxidation, solid/liquid ratio, and the presence of chemical additives.[10, 14, 25]
a) Effects of Physical Properties
To improve dewatering to a large extent, understanding the characteristics and properties
of coal and their effects on dewatering behavior is of utmost importance. Coal is the most
abundant resource of fossil fuel available. It is 20 times more abundant than crude oil and over
1.5 times more than other fossil fuels and crude oil combined. Coal is an inherently
heterogeneous material, possessing organic matter, mineral matter, and has an extensive pore
structure. This is shown below in Figure 1.7.[1, 6, 10]
17 |
Virginia Tech | water that is not readily removable by mechanical methods, chemically bound to the particle, and
which is a part of the particle. It is also generally known as intra-particle moisture. This is
extensively seen in the structure of low rank coals, such as lignite, and it can only be removed by
thermal drying methods (over 100oC). The capillary water is trapped in small channels of the
filter cake, which, again, requires more complex and intensive methods for removal. Free water,
which is not associated with solids and behaves thermodynamically as pure water occupying the
bulk of the slurries, can be removed by any means of mechanical dewatering. Screens,
thickeners, cyclones, and centrifuges are widely-used tools to remove this type of water.[23, 24,
26, 38, 41, 56]
The existing relationship between water and particles is of interest in the dewatering
process, as well. Essentially, particle-water interaction has three main states. Figure 1.9 shows a
brief description of these 3 stages.
Liquid More Almost Middle Small None
Content Saturated
State Slurry Capillary Funicular Pendular Dry
-First- -Second- -Third-
Stage Stage Stage
Schematic
Diagram
Figure 1.9 Schematic of particle-water relationship in cake structure
In the saturated, or capillary, state, where all the pores and voids (or capillaries) of the
particles are filled with water, the liquid pressure is lower than the air pressure, and the surface
19 |
Virginia Tech | tension, capillary radii, and the contact angle of the system determine the magnitude of capillary
forces. Only if the applied forces of vacuum, pressure, centrifugal, and gravity are greater than
the capillary forces, can water be removed. This is called the funicular state, where the
remaining water starts to create bridges around the contact points of the particles. If the applied
force is further increased, it will lead to the formation of lenses of water between the particles.
Most of the time, this determines the final cake moisture. Figure 1.10 shows three main stages of
the water content of the particles under applied forces.[24, 30, 31, 66, 69, 70]
Figure 1.10 Relationship between applied pressure and moisture reduction
The deposition of solids on a filter medium is achieved by applying vacuum, pressure, or
centrifugal forces to a suspension. Throughout this process, several stages occur starting from
cake formation through the end of the drying cycle. The stages of the dewatering process are
illustrated in Figure 1.11. Figure 1.11(A) represents an enlarged slice of the filter medium and
slurry, where the initial bridging of particles begins. In this stage, the filtration commences, and
20 |
Virginia Tech | the first few layers of the cake start to emerge. Figure 1.11 (B) shows the formation of cake on
the filter media, and Figure 1.11 (C) shows the cake after it is formed. The time that passes
during these first three stages is called the cake formation time. Once the cake is formed, it gets
compacted and air starts to enter the cake structure, displacing the water from the largest pores
(shown in Figure 1.11 (D) and Figure 1.11 (E)). Finally, macropore and micropore channels
(capillaries) are formed where the air breakthrough is achieved, draining more water from the
cake. These three stages represent the dry cycle time.[23, 30, 70]
A B C D E F
Figure 1.11 Cake formations during dewatering stages
During the drying time, applied pressure, ∆P, is not capable of removing water
horizontally from the fine capillaries in the cake. Cheremisinoff, et.al, suggested a model with
zones between the particles in which water is located (shown in Figure 1.12).
21 |
Virginia Tech | Figure 1.12 Schematic of the moisture zones
Once the pressure is applied, the water in Zones 2 and 4 can be removed and displaced
with air. On the other hand, the water in Zones 1 and 3 will not be affected by the airflow, which
determines the final moisture, and water will still be located between the particles. To displace
water with air in these zones, chemicals can be used to increase the hydrophobicity of particles
and lower the surface tension of water, which, in turn, increases water removal.[24, 26]
b) Effects of Chemical Properties
The use of chemical additives as dewatering aids to increase the mechanical filtration
efficiency has become more crucial in mineral processing industries. No matter what dewatering
equipment is employed, it is almost a standard procedure to pre-treat the slurry with the addition
of chemicals. The applications show that using the appropriate chemicals may provide
significant improvements in dewatering efficiency as a means of reducing moisture content in
the filter cake and increasing the dewatering kinetics. Practically, chemical additives can be
fitted into two main categories, flocculants/coagulants and surfactants.[25, 44, 71-73]
Flocculants and coagulants are the chemicals that change the packing density and inter-
particle separation distances in the particle structure, modifying the compressibility and the
22 |
Virginia Tech | drainage features of the formed cake. Surfactants, on the other hand, are long-chain polymers
that are absorbed between two surfaces in order to change the surface properties. [25, 44, 71-73]
Coagulants or inorganic salts (electrolytes), such as aluminum, copper, chromic, and
ferric and calcium sulfates (or chlorides) affect the composition and the extent of the electrical
double layer surrounding the particles. They also change the zeta potential of the particles, as
well as the inter-particle electrostatic repulsion, which in turn lead to coagulation.[20, 21, 74, 75]
The balance between the repulsive, double-layer force and the attractive Van der Waals
force determines whether coagulation will occur. If the particle surfaces are not charged,
particles come closer to one another, which in turn help the attractive forces bring them together
to create small agglomerates. Conversely, the surfaces of the particles may be charged
electrically, creating repulsive forces between the particles and preventing the spontaneous
agglomeration brought about by Brownian motion. This phenomenon is shown in Figure 1.13.
A B
(Stabilized Particles) (Coagulated Particles)
Figure 1.13 Potential energy curves
23 |
Virginia Tech | The curve (A) in Figure 1.13 A represents Van der Waals energy. It is an attractive force
having an increasingly negative value, which is effective in small distances between particles.
Curve (B) represents the repulsive electrical force and curve (C) is the outcome of these two
forces, showing the maximum energy barrier for the colloidal system to become steady. At this
point, because the resultant force is repulsive, coagulation does not occur. To allow the
agglomeration to take place, chemical additives can be used to change the surface charge in favor
of Van der Waals forces. Figure 1.13 B shows the changes on the particle surface when
coagulant is introduced to the system. It reduces the electrical force and brings the curve (B) to
lower values. This causes the resultant curve (C) to fall below zero and allows the particles to
coagulate – if they come close enough – and the Van der Waals forces can be effective.[20, 31,
58, 71, 73, 74, 80]
The magnitude of the repulsive forces at the interface of the particles and the liquid
determines the colloidal stability. This stability can be explained by Stern’s double-layer theory.
Particle Surface Surface Potential
Shear Plane
Zeta
Potential
Diffuse Layer of Ions
Shear Plane in Liquid
Bound Layer of Ions
Distance into Liquid
Figure 1.14 Double layer model
24 |
Virginia Tech | The suspended particles, having a certain type of charge in an electrolyte solution, attract the
ions of opposing charges from the liquid, repel the ions that have similar charges, and develop an
electrical double layer of ions as shown in Figure 1.14. The ions that are located in the inner
stratum of the double layer are more strongly bound, and the ions that are located in the outer
stratum are weaker and more diffuse. The electrical potential between these two stratums is
called the zeta potential.[21, 50, 58, 71, 72, 80]
The ions of a coagulant compress the double layer by bringing the charges within the
plane of shear between the bound and the diffuse layers. This compression becomes more
effective when multivalent ions are in solution because they have greater charge concentration.
As a result, the zeta potential is decreased. The minimum ionic concentration required to
produce coagulation and its overall effect is described by the Shulze-Hardy rule. According to
the Shulze-Hardy rule, to start coagulation the minimum ionic concentration must be
proportional to the sixth power of reciprocal counter ion charge. The Al3+, Fe3+, and Ca2+ ions
are the commonly used ions as coagulants. The pH level of the slurry is important for the
hydrolysis of these salts. pH values of a slurry that are above or below the effective pH value
range for a given specific salt, or coagulant, may not allow hydrolysis, and chemical dosage
requirements may be higher than the ideal dosage.[21, 50, 58, 71, 72, 80]
Flocculants are long-chain polymers, or electrolytes, that cause the particles to aggregate
by forming bridges between particles. Typically, flocculants are categorized into two groups:
natural and synthetic. [71, 81]
Natural polymers, such as starches, gums, alginates, and polysaccharides, are mostly
short-chain, neutral, organic compounds. The effectiveness of these polymers is dependent on
the pH of the slurry as well. At alkaline and neutral conditions, polysaccharides are more
25 |
Virginia Tech | effective while, gums, alginates, and starches are more effective at acidic conditions. Because
natural polymers that are used as flocculants have short, rigid chain structure and low bonding
strength, their shear strength is very low. Thus, for flocculation applications, excessive amounts
of these polymers may be needed. In recent times, synthetic polymers, or polyelectrolytes, have
displaced these natural materials, as these polymers can be designed to give desired behaviors,
such as providing more durable flocs and more economical dosages when applied to a particular
problem.[25, 58, 77, 78, 80, 83]
The synthetic polymers are based on polyacrylamide or one of its derivatives that may
have very large polymer chains. These chains consist of anionic, cationic, or neutral groups,
causing the polymer to uncoil and bond to the surface of the coal or clay minerals. This may
result in a desired or undesired selectivity. Polymers with long chains have more contact with
particles and produce large and open floccules. However, they have low shear strength and
contain high residual moistures. In contrast, shorter and lesser-charged polymers generate more
compact granular floccules and improve the filtration characteristics. The high shear forces due
to dewatering methods might reduce the effectiveness of flocculation on fine particle dewatering.
When the flocculants/polymers in solution are introduced to the slurry, they work in two stages:
ion, or charge, neutralization and bridging. Although, the exact absorption mechanism is still not
fully understood, initial adsorption occurs by strong bonding between the polymer and the solid
particles. It involves a molecular bridge, or a series of bridges, between polymer and the solid
particles. The polymer chain from the solution adsorbs onto the solid particles and, when the
extended part of the chain or particles come close enough, creates bridges and continues to
adsorb onto the other particles. These basic floccules grow by bridging with other solid particles
until a most favorable floc size is formed. This is a quick and, unlike coagulation, irreversible
26 |
Virginia Tech | reaction, which again, needs low heat rates to avoid breaking the molecular bridges in order to
initiate a fast settling of solid. The use of flocculants can increase the filtration rate, as well as
the cake thickness, by several multiples. The combined use of anionic and cationic flocculants
exhibits further improvements. It is also reported that anionic flocculant is more effective in
promoting fine coal dewatering than cationic flocculant in vacuum filtration. On the other hand,
cationic flocculant was more effective in high shear centrifugal filtration. The positive increase
in kinetics is, however, accompanied by an increase in moisture content resulting from water
being trapped in the agglomerate structure. In addition, it was observed that the use of
flocculants increased final cake moisture due to increased kinetics, thicker filter cake, and water
trapped in the agglomerate structure.[25, 30, 50, 54, 58, 72, 73, 77, 78, 80, 82, 83]
Surfactants, also referred to as surface-active agents, are the chemicals that modify and
control interfacial interactions by adsorption. This can take place between any two phases or
immiscible components, including solid-vapor (S/V), solid-liquid (S/L), solid-solid (S/S), liquid-
vapor (L/V) and liquid-liquid (L/L) interfaces in a system. Surfactants consist of two
compounds, hydrophobic tail and hydrophilic head, and they are characterized by the chemical
structure of their hydrophilic groups as anionic, cationic, non-ionic, and amphoteric (shown in
Figure 1.15 below).[31, 80, 84]
Anionic
Cationic
Hydrophilic Hydrophobic
Amphoteric Head Tail
Non-ionic
Figure 1.15 Surface active agents
27 |
Virginia Tech | In the filtration process, these surfactants are used to control the characteristics of the
interface between a solid and a liquid by building molecular films, resulting in a reduction of the
viscosity and the surface tension of the liquid and an increase in the contact angle. This supplies
a better water drainage from the filter cake capillaries by simply lowering capillary retention
forces and increasing hydrophobicity, which provides lower final cake moisture. Changes in the
surface chemistry of the particles and filtrate by adding surfactants increases the contact angle
and makes the surface more hydrophobic. This relationship is shown in Figure 1.16.[31, 80, 84]
Figure 1.16 Contact angle and hydrophobicity
If the surface potential of a particle in a liquid has the same sign as the surfactants,
repulsive force is, in effect, between the particles and the surfactants which stabilizes the
additives in solution, preventing any absorption on the surface of the solid. When counter ions
are added to the system, these ions populate between particle surfaces and the head of the
28 |
Virginia Tech | surfactant, allowing adsorption. The adsorption occurs in two ways: ion exchange and ion
pairing. Ion exchange happens when the surfactant molecules are placed on the charged side of
particles which have not been occupied by the counter ions. In ion pairing, the surfactant
molecules displace the charge of the same counter ions and stay on the surface. Based on the
orientation of the surfactant molecules, the surface chemistry of the particles changes, and
becomes hydrophobic or hydrophilic. Depending on the surfactant concentration, the formation
of the additives forms either a neutralized monolayer or reversed, secondary bilayer (shown in
Figure 1.17).[24, 25, 31, 71, 77, 81, 86]
A B C
Figure 1.17 Schematic illustration of the layer formations
When the surfactant dosage is low, few or no surfaces may be coated, not causing a
change in hydrophobicity (Figure 1.17-a), while higher dosages form a close-packed monolayer,
representing neutralized surface charges (Figure 1.17-b). If the amount of chemical is further
increased or overdosed, there will be a bilayer or reverse orientation (Figure 1.17-c), which, in
dewatering, will decrease the hydrophobicity of the particles. [24, 31, 71, 77, 81, 84-86]
29 |
Virginia Tech | Polyelectrolytic polymers are highly charged, short-chain polymers that are used when
the solid content is very low. These types of polymers are used when the flocculants cannot be
effectively employed.[24, 31, 38]
1.2.5. Dewatering Theory
a) Dewatering Kinetics
Filtration kinetics is an important characteristic of fine particle dewatering that
determines the volumetric flow rate of liquid to be removed through a porous media created by
the particles and the filter throughput. Darcy first described the filtration kinetics in 1856 as the
rate equation for the dewatering process. Darcy’s basic filtration equation relating the flow rate
is [20, 26, 44, 88] :
dV A ∆P
= K [1]
dt η L
where V is the volume of fluid, t is the filtration time, ∆P is the pressure drop across the cake,
and L is the thickness. A is the cross-sectional area of the cake, η is the absolute viscosity of
liquid, and K is the rate constant referred to as the permeability of the cake. The equation reveals
a basic relationship; the rate of dewatering is proportional to the pressure gradient and the cross-
sectional area and is inversely proportional to the viscosity. Equation [1] is also written in the
form:
dV A ∆P
= [2]
dt η R
where R is the medium resistance (the medium thickness is divided by the permeability of the
cake). If the suspension does not include any solid particles, all the parameters in Equations [1]
and [2] will be independent of filtration time, t. This will result in a constant filtration rate for a
30 |
Virginia Tech | constant pressure drop. It also represents a linear increase of the cumulative filtrate volume;
however, if the suspension contains particles, resistance of the cake will increase gradually with
time and lead to a drop in the flow rate.[20, 22, 26, 39, 44, 88]
In batch filtration, resistance has two components, medium resistance, R, which can be
assumed constant, and cake resistance, R , which increases with time as a result of increase in
c
cake thickness (shown in Equation [3]).
dV A ∆P
= [3]
dt η(R+R )
c
If the resistance of the cake is assumed to be directly proportional to the amount of cake
deposited (w), then
R =αw [4]
c
where α is the specific cake resistance.
The mass of cake deposited is a function of time (t) and can be related to the accumulated
filtrate volume, V by
wA=cV [5]
where c is the concentration of solids in the system (mass per unit volume of the filtrate). By
integrating and rearranging Equations [3], [4] and [5], the general filtration equation can be
reached. The general filtration equation, Equation [6], is shown:
dt αηc ηR
= V + [6]
dV 2A2∆P A∆P
The fundamental filtration parameters, such as α and R, can be determined to evaluate the
effects of different conditions on filtration kinetics. The filtration kinetics can be altered using
chemical additives, which change filter cake properties, such as permeability (K), cake porosity
(ε), and the specific surface area of particles (S). [20, 22, 26, 39, 44, 88]
31 |
Virginia Tech | b) Dewatering Thermodynamics
To remove the liquid located in these capillaries, it is important that the applied pressure
be larger than the capillary pressure, (p), which is defined by Laplace:
2γ cosθ
p= 23 [10]
r
where, γ is the surface tension of liquid, and θ is the water contact angle. The Laplace equation
23
suggests that decreasing surface tension leads to a lower capillary force and improved cake
dewatering. Increasing the solid/liquid contact angle, θ, lowers the cosθ, and as a result, it
s
decreases the capillary force. This controls the water removal from the cake.[25, 31, 39]
By using an appropriate surface-active reagent, it is possible to hydrophobize the surface,
decrease γ , and increase contact angle by absorption in such a way that its hydrophobic part is
23
oriented away from the surface. The Young’s Equation describes the relationship between the
solid/liquid contact angle and the work of adhesion, a measure of how strongly water is bound by
the solid surface: [44, 71, 89]
γ =γ cosθ+γ [11]
13 23 12
where θ is the solid/liquid contact angle, γ is the surface free energy of the solid/air interface,
13
γ is the surface free energy of water/air interface, and γ is the surface free energy of the
23 12
solid/water interface. The free energy change per unit area, ∆G/dA, is determined by the
following equation in a solid/liquid/air system. This is also known as the Dupre equation.[31,
44, 71, 89]
33 |
Virginia Tech | dG
W = =γ +γ −γ [12]
adh dA 13 23 12
By substituting Equation [11] into [12], the following relationship can be obtained
(Young-Dupre Equation):
W =γ (1+cosθ) [13]
adh 23
The Young-Dupre equation suggests that the work of adhesion between solid and liquid
can be calculated from the liquid surface tension and the contact angle. The spontaneous
hydrophobization and effect of contact angle is illustrated in Figure 12, where A is the solid, B is
the liquid, and C is the air. [31, 44, 71, 90]
γ γ
lv lv
γ θ γ θ γ
sv γ sv sl
sl
A B C
Figure 1.18 Schematic representation of solid/liquid/air interfaces
For a completely hydrophilic surface, θ =0o, cosθ =1, and W will be at their
12 12 adh
maximum values. On the contrary, for a completely hydrophobic surface, θ =180o, cosθ =0,
12 12
and W will be at their minimum values. This, in summary, suggests that for a dewatering
adh
process, the work done on a hydrophilic solid surface can be minimized by simply decreasing
water/air surface tension and increasing the contact angle, as seen in Figure 12.[31, 44, 71, 89]
Novel dewatering aids introduced in this study are non-ionic, low hydrophile-lipophile
balance (HLB) number surfactants. The surfactant molecules adsorb on a hydrophobic solid
surface, such as coal, as a result of hydrophobic attraction and, thus, increase its contact angle.
34 |
Virginia Tech | CHAPTER 2 LABORATORY AND PILOT SCALE EVALUATION OF
DEWATERING AIDS
2.1. INTRODUCTION
The fine size coal production has increased as a result of several factors. These factors
can be broadly explained as follows[1, 2];
i) The mechanized ROM coal production methods have increased the fine coal
production
ii) The coal produced, for both coking and power industries, has to meet with certain
quality constraints, such as calorific value, ash, sulfur and moisture content. To
meet these requirements and produce a higher quality product, the raw material
should be crushed to a lower top size for improved liberation.
iii) The possibility of economical recovery of fine size fraction of coal that reports to
plants’ discard streams
iv) Recovery of the fine/ultrafine coal from the waste ponds which represents an
economically viable resource
Froth flotation is the widely used and the most effective separation method for fine coal
cleaning. It is a wet process and the separation results in two products in the form of slurry, i.e.
the concentrate (coal) and the tailing (ash-forming minerals). Of the products, the tailings are
discarded to the waste ponds and the coal is further processed before it reaches to the final
consumer.[2-4]
In processing plants, clean coal is dewatered by means of filters, such as vacuum disc,
horizontal belt and drum filters, and centrifuges. However, the larger surface area of fine/ultra
fine particles per mass results in more water adsorption. Additionally, the fine particles create
43 |
Virginia Tech | smaller size capillaries in the filter cake from which water cannot be removed easily during the
application of vacuum or air pressure. As a result, these dewatering methods cannot produce the
desired moisture levels if there is not a thermal drier. Therefore, fine coal dewatering is the most
difficult and costly operation in preparation plants.[5, 6]
Previous studies on fine coal dewatering addressed the problem and a number of
dewatering aids were developed to lower the final cake moisture to the extent that is not
achievable by mechanical means. In this study, laboratory and pilot scale tests were conducted
on a variety of fine coal samples to do engineering evaluation of these dewatering aids; and
possible industrial applications have been investigated.
2.2. GENERAL EXPERIMENTAL DETAILS
2.2.1. Samples
Laboratory- and pilot-scale dewatering tests were conducted on various samples
comprised of different mixtures of coarse and fine coal. Following are the samples that were
used in these tests:
i) Flotation feed (cyclone overflow or underflow)
ii) Flotation product
iii) Filter feed (blend of flotation and spiral products)
These samples were received from different preparation plants located in North America
region. Table 2.1 shows the selected coal samples for the dewatering tests.
44 |
Virginia Tech | Table 2.1 Names of the operations and their locations from where the samples received for
dewatering aid evaluation tests
Operation Location
Mingo Logan West Virginia
Coal Clean West Virginia
Concord Alabama
Buchanan Virginia
Moatize Mozambique
Elkview BC, Canada
Pinnacle Reclamation Virginia
These samples were tested
i) as received, if the sample was directly collected from filter feed stream,
ii) after upgraded with flotation, if it was cyclone overflow or cyclone underflow,
iii) after blended at different ratios.
For laboratory dewatering tests, the samples were collected or received as coal slurry in 5
gallon buckets. To minimize the adverse effects of superficial oxidation of the coal samples, the
dewatering tests were conducted within 24-48 hours of receipt. The samples were first subjected
to solid content determination (% solid) and particle size distribution determination. When
flotation step was carried out, Denver D-12 flotation equipment was used, where kerosene and
MIBC were added as collector and frother, respectively. After dewatering aid addition, each
sample was conditioned with a stand-alone mixer to ensure a proper dispersion and adsorption of
the dewatering aids.
45 |
Virginia Tech | 2.2.2. Methods and Procedures `
a) Laboratory Scale Dewatering Test Equipment and Procedure
The laboratory-scale dewatering tests were conducted using the following equipments:
i) 2.5 inch diameter Buchner funnel
ii) 2.5 inch diameter pressure filter
Buchner Funnel was used in the bulk of the tests fitted with various sizes of filter cloth
depending on the samples treated. It was mounted on a vacuum flask, which in turn was
connected to a larger vacuum flask to protect the pump itself and stabilize the vacuum pressure.
Before initiating the filtration, first, a known volume of coal slurry was transferred to container,
to which a known amount of a dewatering aid (or a mixtures thereof) was added by means of a
microliter syringe. The coal slurry was then subjected to mixing with a three-blade propeller
type conditioner for a given time to ensure that proper chemical dispersion and adsorption were
achieved. After conditioning, the slurry sample was poured into the funnel before opening the
vacuum valve. Filtration started when a vacuum was applied to the slurry. After the cake
formation, the vacuum pressure was kept on for a desired length of time to remove the remaining
water trapped in the capillaries. This period was called the dry cycle time. The amount of
volume added to the Buchner funnel determined the cake thickness. After the pump was stopped
a representative sample was removed from the cake and dried for a give time. The filter cake
was weighed before and after drying; and moisture content was determined from the dry-wet
weight differences. In each experiment, the cake thickness, set up and actual vacuum pressure
and cake formation time were recorded. Figure 2.1 shows the basic experimental set-up used for
the Buchner funnel filtration tests.
46 |
Virginia Tech | (Figure 2.2). The vacuum disc filter consisted of a disc mounted on a horizontal shaft. The disk
had interchangeable elements which could be changed for fitting and removing filter cloths. The
disc rotated in a sump into which the suspension was fed (the sump also has two agitators to
provide even cake formation), and a vacuum was applied through the core of the shaft. The
submerged sector of the disc collected the cake and then removed it by blow-back air cake
discharge system utilized in conjunction with a scraper just before re-entering to the sump. The
specifications of the equipment were as follows:
• 2 ft diameter with 10 removable sectors
• 0.2 - 2.0 ft2 of adjustable filter area by varying number of filter sectors used
• Peterson “Syncro-Blast” air cake discharge system
• 0.5 - 12 minutes per revolution
• 29 inches Hg vacuum pressure at 2.5 cfm.
• Dual filtrate sumps: 25 gal capacity each
• Connected HP: 2.25
• Dimensions: 5ft. High x 5ft wide x 4ft deep.
• Weight: 1,800 lb.
48 |
Virginia Tech | compressor. Coal slurry was fed to the column from an agitated tank using a variable-speed
centrifugal pump. Pulp level in the column was maintained by adjusting the tailings flow rate
using a pneumatic control valve. The valve actuated based on readings from a pressure
transducer mounted in the side of the column. Wash water was added to the froth to minimize
the entrainment of fine mineral matter. Chemical metering (reagent) pumps were used to add the
desired dosages of frother and/or collector to the feed slurry.
Head Pneumatic
Air
Froth Tank Sampler
Compressor Launder
Viewing Reagent
Feed Window Pump
Power
Mixer
Supply
Column
Speed
Tank
Feed Electronic
Controllers
Sump Timer
Sparger
Breaker
(In - Line
Mixers Box
Mixer)
Conditioner
Feed Sparger Circulation
Tanks
Pump Pump Pump
(a) (b)
Figure 2.5 Photographs of (a) the column unit, (b) the conditioner
A schematic diagram of the Conditioner Module is shown in Figure 2.5 (b). The module
incorporated two 20-liter conditioning tanks that were operated in series to provide up to 10
minutes of conditioning time. The conditioning tanks were equipped with single-impeller mixers
that could be varied in speed from 0 to 2500 rpm using electronic controllers. To ensure that
52 |
Virginia Tech | coarser particles did not settle when low feed rates were used, the slurry in the conditioning tanks
was continuously circulated through a head tank using a centrifugal pump. The head tank was
equipped with an automated sampling system that consisted of an electronic timer and a pneumatic
sample cutter. During operation, the sampling system was used to divert a defined portion of
circulated slurry to the Filter Module (or any other downstream operations). A chemical metering
(reagent) pump was used to add the proper dosage of dewatering aid to the feed slurry as it enters
the conditioning tanks. To obtain a consistent feed rate, a small peristaltic pump was installed at the
plant to pump feed slurry from the plant’s filter feed box to the conditioning module.
2.3. FIELD TESTING
2.3.1. Test Program Overview
A series of laboratory dewatering tests were conducted to identify the best chemicals and
optimize the dewatering parameters. After successfully completing the bench-scale, a diverse set
of field tests were undertaken using the mobile test modules described previously. The on-site
field tests were required in order to provide site specific operational information and scale-up
data for the development of the POC plant. The field tests were conducted using samples
comprised of different mixtures of coarse and fine coal obtained as flotation feeds (either cyclone
overflow or underflow), flotation froth products or filter feeds (blends of flotation and spiral
products). As a result, the various samples were tested as-received (if the sample was directly
collected from filter feed stream), after upgrading with flotation (if it was cyclone overflow or
cyclone underflow), or after blending at different mass ratios. In support of the field testing
effort, numerous laboratory tests were also performed. The specific procedures used for
conducting these laboratory dewatering tests have been described previously.
53 |
Virginia Tech | 2.3.2. Mingo Logan Site
a) Site Description
The Mingo Logan Coal Preparation Plant is located in the counties of Mingo and Logan,
WV. This is in Central Appalachia Region. As Arch’s principal source of metallurgical coal,
the plant produces approximately 3.9 million tons (2006) at a quality of 13,000 Btu/lbs. and, 1.1
lbs. SO /MM Btu. The preparation plant consists of two independent 800-ton circuits, each
2
provided with separate surge bins to ensure uniform splitting of the plant feed. Each of the
circuits consists of heavy media vessels for the 160 x 7.7 mm (6 x 0.3 inch) material, heavy
media cyclones for the 0.3 x 16 mesh material, spirals for the 16 x 65 mesh material, and froth
flotation for the 65 mesh x 0 material. Dewatering is accomplished through centrifuging for the
various size fractions. However, the dewatering efficiency was not satisfactory due to (i) the fine
particle size of flotation product, (ii) the screen bowl centrifuges were not capable of capturing
the fines, and (iii) the accumulation of excessively stable foam over the screen bowl centrifuges.
Also, the issues resulted in high final cake moistures and loss of considerable amount of valuable
material that might be recovered by other means. If recovered and dewatered more efficiently,
such losses could be turned into profits.
An extensive laboratory and pilot-scale dewatering test program was conducted to study
the feasibility of new approaches on more efficient dewatering methods and the recovery of the
fine particle size material which reports to the plant’s discard stream as a result of centrifuge’s
lack of fine particle capturing capability. To meet the objectives, exploratory laboratory
dewatering tests were conducted to evaluate the performance of various types and dosages of
newly developed fine particle dewatering aid technology. Data from the laboratory studies were
used to provide technical justification for the pilot-scale test program undertaken at a coal
54 |
Virginia Tech | preparation plant using a horizontal belt vacuum filter. Investigations incorporated the study for
the possible utilization of these technologies in industrial applications for fine coal beneficiation.
b) Reagents
The investigation, both laboratory and pilot-scale, was performed with varying amounts
of three types of dewatering aids, namely RW, RU, and RV. Since these dewatering aids are
insoluble in water, they were dissolved in a solvent. In both laboratory and pilot-scale
experiments diesel was used as the solvent. The ratio of reagents-to-solvent was optimized in
previous studies by varying the individual dosages (0.5 to 3lb/ton), while maintaining the total
blend dosage constant. For this test program, the optimum combination for a given dewatering
aid and solvent is one to two (1:2) dewatering aid-diesel ratio. The flocculants, Nalco-9822 and
Nalco-9806, were prepared at a 0.09 % solution and used both alone and in conjunction with
dewatering aids. After examining various dewatering aid and flocculant combinations and their
orders of additions, the best combination was found to be to add dewatering aids first, then to
add flocculants. As it is going to be discussed in the following sections, the agitation and mixing
intensity play an important role in determining the chemical performance. The dewatering aids
require certain amount of mixing intensity; however, it may be excessive shear for a given
flocculant. Excessive shear may lead to chain breakage, which reduces the effectiveness of a
flocculant for bridging. Therefore, flocculant was added after mixing the dewatering aids; and
the new combination was conditioned at a very low intensity for 15-20 seconds.
55 |
Virginia Tech | Table 2.2 Size distribution of Mingo Logan flotation product
Size (mesh) Weight (%)
65 10.95
65X100 9.25
100X150 11.91
150X325 22.25
325 45.65
Total 100
c) Coal Samples
The dewatering tests were conducted on feeds comprised of three different mixtures of
coarse and fine coal feeds. The first series (Series A) was conducted using a fine coal sample
from a classifying cyclone overflow that had been cleaned by flotation. The product solid
content was 23% by weight. Approximately 45%of the material was finer than 325 mesh size.
Table 2.2 shows the sieve analysis results on flotation product sample.
In some of the experiments, the cyclone overflow sample was also used to produce clean
flotation product for dewatering tests. The second series (Series B) was conducted using a
mixture of fine and coarse coal. The flotation product sample was mixed with a portion of the
spiral product at a ratio of 3:1 (i.e., 75% fine coal and 25% coarse coal) to prepare the
dewatering feed slurry for dewatering tests. The solid content of the combination of
Table 2.3 Size distribution of Mingo Logan mixture sample
Size (mesh) Weight (%)
35 17.80
35x100 22.25
100x325 23.25
325 36.71
Total 100
56 |
Virginia Tech | spiral/flotation product slurry was approximately 27% by weight. The third and final series
(Series C) was performed with a feed comprised of 50% coarse coal from the spiral circuit and
50% fine coal from the conventional flotation circuit. Table 2.3 shows the size distribution of
mixed clean spiral and flotation product (at 1:1 ratio) sample. The solid concentration of the
slurry was about 25% by weight, while 36% of the solids were minus 325 mesh.
d) Laboratory Procedures
All the samples including the flotation product, obtained in the lab using cyclone
overflow sample or collected directly from the preparation plant, and the mixture sample were
conditioned with the dewatering aid before the dewatering tests. As stated earlier, a conditioning
system is very important for the chemical dispersion and adsorption. The surfactants adsorb on
the surface of the solid leading to an increase in the solid/liquid contact angle and a decrease in
liquid surface tension. Each slurry sample subjected to mixing to ensure that proper chemical
dispersion and adsorption was achieved. The sample was first agitated in a 300 ml Plexiglas cell
equipped with a three-blade propeller-type mixer. The mixer was designed to control the mixing
strength by adjusting the input current and voltage of its motor. Once the sample was mixed, it
was transferred to a Buchner vacuum filter and subjected to dewatering tests. When flocculants
were used, proper mixing was supplied. This ensured that high shear conditions neither
degraded nor rendered the flocculant ineffective. Essential test parameters affecting the final
cake moisture and filtration performance, such as pressure level, specific cake weights, and
filtration time were recorded. Upon receipt, all the samples were subjected to dewatering tests
within 24 hours to minimize the effects of aging and artificial oxidation that can change surface
properties rapidly and affect the filtration behavior.
57 |
Virginia Tech | e) Pilot-Scale Procedures
Pilot-scale dewatering tests were conducted on feeds comprised of mixtures of coarse and
fine coal feeds. Feeds specifically used, for example, are fine flotation product or combinations
of coarse spiral and fine flotation products. The coal feeds, flotation product/spiral product coal
sample, were reasonably consistent for most test work. The blend sample contained
approximately 28% of solids by weight and 12% ash on dry basis. Table 2.4 provides the
particle size and ash analysis results. The particle size is almost uniformly distributed over the
range of 18 mesh x 0, but obviously, the fine particles contain more ash. As shown, 68% ash is
included in the fine fraction (325 mesh x 0). Table 2.5 shows the particle size and the ash
contents of each size class of the flotation product. Of the sizes, 43% of the particles was
passing 325 mesh (-44 µm), and more remarkably, 81% ash was distributed in this fine fraction.
Table 2.4 Particle size and ash content analysis of Mingo Logan plant spiral/flotation
product mixture (on-site samples)
Size Weight Ash Ash
(mesh) (%) (%) Distribution (%)
35 19.66 3.74 6.09
35x100 22.54 4.75 8.87
100x325 23.83 8.55 16.87
325 33.98 24.23 68.18
Total 100.00 12.08 100.00
Table 2.5 Particle size and ash content analysis of Mingo Logan plant flotation product (on-
site samples)
Size Weight Ash Ash
(mesh) (%) (%) Distribution (%)
35 1.77 2.53 0.48
35X100 15.06 2.08 3.45
100X325 40.22 3.38 15.00
325 42.96 17.1 81.07
Total 100.00 9.06 100.00
58 |
Virginia Tech | The on-site pilot-scale test work was conducted using a conditioning tank and the pilot-
scale horizontal belt filter built by WestTech/Delkor. The spiral/flotation product mixture
intercepted from the screen bowl feed pipe was fed to a 35 gallon conditioner. Flotation sample
was taken from the distribution box. After conditioning with the dewatering aids under test, a
desired amount of coal slurry was then fed to the filter. The total feed rate to the conditioning
tank was generally controlled at 13gal/min. This allowed approximately 3 minutes of
conditioning time for dewatering aids.
The dewatering aids, RU and RV were used. The flocculant, R9822 from Nalco, diluted
to 0.09% solution, was directly pumped to the horizontal belt filter’s feed pipe. The belt speed of
the filter was adjusted to allow the materials travel over the vacuum zone in the time interval
between 65 to 184 seconds, which enables to investigate the effect of different cake thicknesses
under certain filter feed rate, and of different dry cycle times. The cake sample was taken for
moisture analysis periodically from the ending point of the belt when a steady-state was achieved
after the test parameters were changed.
f) Laboratory Test Results (Series A – Fine Coal Only)
In this series of experiments, Mingo Logan’s clean coal product from the conventional
flotation circuit and cyclone overflow samples were used. The first dewatering tests were
conducted to determine whether the flotation product coal samples provided by Mingo Logan
Preparation Plant would respond well to the addition of the novel dewatering aids. The
preliminary results showed that with the addition of dewatering aids, it is possible to reduce the
final cake moisture content of fine coal cake by about 20%, while also increasing the rate of
dewatering. The surface moisture was reduced down to about 19% using 3 lb/ton RW, where it
was approximately 23% for control tests at about 8-11mm cake thickness. As many factors
59 |
Virginia Tech | influence the filtration performance, particle size distribution and corresponding ash contents
were the dictating parameters. Considering the amount of material under 325 mesh size in the
flotation product sample, a portion of ultra-fine particles was removed from the sample via
desliming. The tests results on the deslimed sample showed that moisture was further decreased
to 16.7 % at the same RW dosage. This is another significant moisture reduction from a baseline
of 21%. This moisture reduction in both baseline and with reagent tests is achieved by desliming
of the ultrafine fraction that results in more freely draining filter cake capillaries. This also
prevented the fines from forming an impermeable layer which might be positioned on the top of
the cake.
The screen analysis results also showed that 81% of the ash was in the minus 325 mesh
size fraction. This material, which consists of clays and slimes, is hydrophilic in nature and this
affects the dewatering performance negatively. The use of flocculants may be another way to
compensate the negative effects of the ultrafine particles. The principle of using of flocculants is
to bring the fine particles together in the coal slurry and create looser packing in the filter cake.
This loose packing results in larger capillaries between the aggregates and a more porous,
permeable cake. This allows a rapid drainage of water from these voids, which, in turn,
increases the filtration rate. To investigate the effects of flocculants on dewatering kinetics, a
series of tests was conducted. To optimize the dosage, various amounts of flocculant were
tested, from 5 g/ton to 75g/ton. It was determined that 25 g/ton was the most appropriate
flocculant amount when being used alone or in conjunction with other chemicals. The test
results showed that in most cases, the addition of flocculant did not improve the final cake
moisture, due to the water trapped inside the flocculants. Instead, the dewatering kinetics
increased significantly i.e., 30% to 75%.
60 |
Virginia Tech | Another way of lowering the moisture may be to clear out the excessive amount of ash-
forming minerals in the fines. To investigate this idea, prior to laboratory dewatering tests, the
plant’s clean coal product was subjected to another step of flotation using 1 lb/ton (454g/t) RV
(in Diesel (1:2)) as collector and 100g/t MIBC as frother. The particle size analysis results
showed that the 37.6% of the re-floated sample was minus 325 mesh. Dewatering test results
showed after its addition, RW, at about 3 lb/ton, can reduce the moisture to 15.5%, where the
baseline was approximately 20%. In addition, the dewatering kinetics of the coal sample was
also increased by 50% as a result of the increased hydrophobicity.
As stated earlier, the surfactant adsorption is very important so that it can lead to an
increase in solid/liquid contact angle and a decrease in liquid surface tension. For this reason, it
is very important to have an effective conditioning system. To investigate and optimize the
effectiveness of the dewatering aids and their performances, two series of filtration tests were
conducted at various mixing intensities and times. Table 2.6 gives the laboratory test results
obtained on the flotation product sample using RW at 3 lb/ton at about 20-24 mm cake
thicknesses. As shown, the moisture reduction was substantially improved when mixing
intensity was increased. The use of RW at 3 lb/ton reduced the filter cake moisture from
baseline value of 30.02% to 25.98% and 25.45% when using low and medium-energy agitation
at one minute, respectively. The cake moisture was further reduced to 20.26% moisture when
Table 2.6 Effect of mixing intensity and conditioning time on Mingo Logan flotation
product (20 inHg vacuum and 3 lb/ton RW)
Speed Moisture @ specified conditioning time
Level 1 min 2 min
0 30.02 30.02
Low 25.98 25.29
Medium 25.45 23.91
High 20.26 18.66
61 |
Virginia Tech | using high-energy agitation at the same reagent dosage of 3 lb/ton. The moisture reduction
showed a similar trend, but improved final cake moistures were obtained when the agitation time
was increased to two minutes. The final cake moisture was reduced to 23.91% and 18.66%
using medium and high-energy conditioning, respectively, at the same dewatering aid dosage. It
was found out that two minutes of conditioning at high intensity mixing was optimum. Beyond
two minutes of conditioning, there was no change in the residual moisture of the cake. These
results clearly demonstrate that the importance of proper conditioning when using the novel
dewatering aids.
In the light of the results obtained in the initial laboratory filter tests, using dewatering
aids were believed to be a promising method because it could not only remove water but also
increased the filtration kinetics. To evaluate the filtration performance in detail as a function of
cake formation time, dry cycle time, and specific cake weight in the absence and presence of
different types of dewatering aids, a series of vacuum filtration tests was carried out. It is very
informative to know the effects of these physical parameters for scale up of using chemicals,
optimization of total filtration time, production rate, and, consequently, the final cake moisture.
These parameters would also provide general suggestions to meet the desirable filtration
efficiencies.
The dewatering tests were carried out using a fixed amount of dewatering aids (3lb/ton)
and flocculants (25 g/ton) after the optimum dosage was determined. The tests were conducted
at a fixed setup vacuum pressure and pre-measured amount of slurry was added to the Buchner
funnel for dewatering tests. The dry cycle times were changed randomly, varying from 14 to 120
seconds. The cake weights were changed by increasing the slurry volume from 50 ml to 300 ml.
The cake formation time, dry cycle time, cake weights, amount of filtrate, and solid contents for
62 |
Virginia Tech | each test were recorded for production rate calculations. The filter cake production rates were
investigated in the absence and presence of dewatering aids and flocculants.
Figure 2.6 is a plot of the final cake moisture percent as a function of pounds of dry
solids per hour per square foot, or the production rate. The filtration kinetics without the
surfactants and flocculant addition were observed to be very slow, and the residual cake moisture
of the cake was found to be in the range of 31-36% at about 60-100 lb/hr/ft2 throughput, which
corresponds to approximately 5-20 mm cake thickness. The filtration time used for the
calculations was the sum of cake formation time and dry cycle time. In the absence of
dewatering aids, a production rate greater than 100lb/hr/ft2 was found to be impractical, as
achieving a dry cake was no longer possible. There was also segregation of particles in the filter
40
35
30
25
20
15
10
0 50 100 150 200
Figure 2.6 Normalized production rate (lb/hr/ft2) versus cake moisture
cake and a considerable amount of solid loss – up to 10% - in the filtrate. As expected, the cake
63
)%(
erutsioM
none
Reagent RU (3lb/t)
Reagent RV (3lb/t)
Filter Production Rate (lb/hr/ft2) |
Virginia Tech | moisture increased proportionally with increased production rate; however, the results showed
that using dewatering aids lead to very significant reductions in the final cake moistures and
increase in the filtration rate by several times. The addition of reagents outperformed baseline to
a great extent because it gave the same or lower cake moisture at a higher production rate with
almost 97% to 99% solid recovery. Also given in the Figure 2.6, RU and RV produced very
similar results.
The results obtained with the plants flotation product showed that moisture values as low
as 25-30% at improved throughputs – as high as 100-150 lb/hr/ft2 – could be achieved. When
the production rate was further increased at the expense of final cake moisture, it became
possible to increase the throughput almost 2.5 times (compared to baseline) in the presence of
dewatering aids. The test work demonstrated that the use of dewatering aids provided an
outstanding dewatering performance on Mingo Logan’s flotation clean coal product and
produced significantly lower final cake moisture values, as well as a higher rate of dewatering,
and improved the filtration efficiency.
If the other operating parameters are kept constant, the effectiveness in productivity of
filtration is related to the time required to complete a full solid-liquid separation cycle, consisting
of cake formation time and dry cycle time. The correlation between the cake moisture and total
filtration time, normalized with cake weight, is shown in Figure 2.7. Test results showed that
when the dewatering aids and flocculant were used in a combined manner, the time consumed
for filtration of a given amount of material was reduced, which, in turn, increased the production
rate by several factors. There was also significant moisture reduction in the presence of
dewatering aids; however, it appears that even when the filtration time was increased in the
absence of dewatering aids, the reduction in final cake moisture content was very small. The
64 |
Virginia Tech | 40
35
30
25
20
15
10
0 0.005 0.01 0.015 0.02
Figure 2.7 Cake moisture versus normalized drying cycle time
results suggest that there is an increase of 100% in solids throughput at a fixed moisture value or
5-10% moisture reduction at approximately the same solid throughput.
Figure 2.8 shows the effect of chemicals on cake formation time and, eventually, the
filtration kinetics. A strong correlation was observed between cake formation time and the
throughput of the filter. In the presence of RU and RV, even at higher specific cake weights, the
formation times were much shorter than what was seen in the baseline tests. Approximately 100-
140 seconds were required to produce 3.5-4.5 lb/ft2 of coal in baseline tests; however, for the
same coal production, using dewatering aids, approximately 30-60 seconds was needed. Above
5 lb/ft2, obtaining a baseline value was impossible because the filter time was impractically
prolonged, while 6-7lb/ft2 of coal could be produced when using dewatering aids. The test data
indicated that cake formation time was a significant parameter in throughput and residual cake
65
)%(
erutsioM
none
RV 3lb/t + 25 g/t F
RU 3lb/t + 25 g/t F
Time (hr/lbft2) |
Virginia Tech | moisture; however, formation time can be altered by using dewatering aids. In addition, in daily
practice, decreased cake formation time allows a longer dry cycle time to complete the full
filtration operation, increasing the throughput of the filter at lower final cake moisture.
19
17
15
13
11
9
7
5
3
1
0 50 100 150 200 250
Figure 2.8 Cake moisture versus cake formation time
After completing the tests using the fine clean coal sample from Mingo Logan, a second
set of tests were performed using a fine cyclone overflow sample that had been subjected to froth
flotation to minimize the adverse effects of high ash content. The tests were conducted by
floating the cyclone overflow sample using diesel (0.66 lb/t) as collector and MIBC (0.33 lb/ton)
as frother. Both conventional laboratory-scale flotation and bench-scale column flotation
equipment were used to produce ash-free clean coal. Both products were tested in the absence
and presence of dewatering aids.
66
2tf/bl
thgieW
eKaC
none
RV 3lb/t + 25 g/t F
RU 3lb/t + 25 g/t F
Cake Formation Time (sec) |
Virginia Tech | The first series of clean coal samples were produced using lab scale Denver cell. The
solid concentration of the product was about 15% by weight, while maintaining approximately
44% of minus 325 mesh particle size. The test results on this sample showed that the control
cake moisture was 25.31%, and with RW addition at about 3 lb/ton, the final cake moisture was
reduced approximately to 19.5% at 7-10 mm cake thickness. Another set of flotation tests was
run on the plant’s flotation product to discard ash-forming minerals, and the product of the
flotation tests was subjected to the same type of dewatering tests. The results indicated that the
floated cyclone overflow and re-floated samples, using the laboratory-scale Denver flotation
machine, gave considerably better results compared to the plant’s as-is flotation product sample.
This may be the result of discarding ash-forming minerals and removing ultra-fine particles more
efficiently from the coal slurry compared to plant operation.
Using a laboratory scale column flotation unit produced another series of clean coal
samples. In this set of tests, the flotation feed/cyclone overflow sample was floated using 0.66
lb/ton diesel as collector and 0.22 lb/ton MIBC as frother. The solid concentration of the column
flotation product was about 10% by weight, and approximately 39% of particles were minus 325
mesh size. The dewatering tests were conducted using a 2.5inch-diameter Buchner funnel at 20
inch Hg set-up vacuum pressure with 2 minutes drying cycle time and about 11 mm cake
thickness. RU and RV were used as dewatering aids with several of dosages ranging from 0.5-3
lb/ton. Vacuum filtration results on both column and Denver unit clean coal products showed
that RV alone was capable of both reducing the moisture and cake formation time significantly;
therefore, flocculant addition was unnecessary. The cake formation time for control tests was 94
seconds, whereas it was approximately 10 seconds in the presence of the dewatering aid, RV.
The presence of dewatering aids also made it possible to reduce the final cake moisture content
67 |
Virginia Tech | of the cake by about 35%. The dewatering test results from the column flotation’s clean coal
product are represented in Table 2.7. The final cake moisture for control tests was
approximately 22.80%. With RV addition, the moisture was reduced to 14.91%, and with RU
addition, the cake moisture was lowered to 14.55%, giving approximately 36% overall moisture
reduction. When the dry cycle time was lowered to one minute, the baseline was 26.9%, and
with about 1 lb/ton RU and RV addition, the moisture was lowered down to 20.44% and 17.95%,
respectively.
The influence of cake thickness on final cake moisture was also investigated. To study
the effect of dewatering aids on a thicker filter cake, the thickness was increased to
approximately 20 mm by increasing the slurry volume that was added to Buchner funnel. As
expected, thicker cake resulted in increased baseline moisture and cake formation time. The
addition of RU and RV at 3 lb/ton lowered the cake moisture to 17.81% and 15.08%,
respectively, from the baseline moisture 24.8% with 2 minutes of dry cycle time. These results
correspond to 22% and 33% total moisture reduction. The test results indicated that even with
thick cake, dewatering aids showed improvements with final cake moisture.
Another set of tests was performed to investigate the effect of flocculant on the
dewatering of the laboratory column flotation product (Table 2.8). The same procedure was
applied, and the same types of collector, frother, and dosages were used to produce clean coal.
Table 2.7 Effect of reagent addition on Mingo Logan flotation feed sample (cleaned using
bench-scale column).
Reagent Dosage Moisture (%)
(lb/ton) RV RU
0 22.80 22.80
0.5 16.51 19.20
1 16.99 17.02
3 14.91 14.55
68 |
Virginia Tech | For this series of dewatering tests, the moisture results for the baseline tests were considerably
higher than what was obtained previously. This might be attributed to the excessive amount of
fine material associated with this fine fraction. The baseline values for the tests were around
27.30% at approximately 10 mm cake thickness. The addition of flocculant alone at about 25
g/ton was capable of reducing the final cake moisture to 20% and decreasing the cake formation
time from 120 seconds to 25 seconds at about 7-9 mm cake thickness and 20 inch Hg vacuum
pressure. For lower vacuum pressure, about 15 inch Hg, flocculant additions were capable of
reducing the moisture to 23% from 28.1%, and significant improvement in dewatering kinetics
was present, as well. The cake formation time was lowered significantly to 26 seconds from 310
seconds. The final cake moisture was also lowered down to 15.52% and 17.02% at addition of 5
lb/ton RU and RV, respectively, in conjunction with flocculant addition. These results show
that, in terms of producing a cleaner product, column floatation is superior to the Denver test.
The results illustrate how dewatering aids can lower the final cake moisture significantly.
g) Laboratory Test Results (Series B – Mixture of 75% Fine and 25% Coarse)
A limited number of dewatering tests were conducted using a mixture of 75% fine coal
and 25% coarse coal as a function of various dosages of RW, RU and RV. The solid content of
the combination of spiral/flotation product slurry was around 27% by weight and it was
subjected to dewatering tests using novel dewatering aids at various dosages. The test results
showed that the final cake moisture was about 16% to 17% when dewatering aids were added at
about 1lb/ton dosage. The moisture content for the control test was approximately 20% at about
8-11 mm cake thickness. Tests were also performed in the presence of flocculant alone and in
conjunction with dewatering aids; however, increases in the dewatering kinetics were close to
69 |
Virginia Tech | Table 2.8 Effect of flocculant and dewatering aid addition on Mingo Logan floation feed
sample (cleaned using bench-scale column)
Reagent Flocculant Moisture (%)
(lb/ton) (lb/ton) RV RU
0 0 27.30 27.30
0 25 19.99 19.99
0.5 25 18.78 18.55
1 25 19.66 17.41
3 25 19.99 16.55
5 25 17.02 15.52
those obtained with reagents. Even though the preliminary dewatering test results were
promising, due to the Plant’s request dewatering tests were focused more on flotation or
flotation/spiral mixture samples. Thus, no further tests were done.
h) Laboratory Test Results (Series C – 50% Mixture of Fine and Coarse Coal)
The fine and ultra-fine size fraction of a stream to be dewatered is very influential on
dewatering, which affects the filtration performance. In the dewatering of such particles, lower
moisture percentages are always desirable; however, mechanically, there is a limit to the level of
moisture achievable, regardless of operating parameters, such as the length of the filtration time
or the applied pressure. In filtration, most of the water is held between and on the surfaces of the
particles. Finer particles will create a larger overall surface area and smaller inter-particle
openings, which, in turn, keep more water than coarser size distributions do. As the capillary
filtration model suggests, a filter cake consists of numerous capillaries with a range of diameters.
When the capillary radii are increased, the filtration rate should also increase.[5-8] It can be
achieved either by de-sliming or coarse particle addition into the stream or, in this case, into the
slurry. As mentioned earlier, when the coal slurry was partially or fully deslimed the drainage of
the filter cake was much more efficient, thus lowering the final cake moisture and increasing the
70 |
Virginia Tech | kinetics. Blending coarse particles with fine particles will also increase the capillary radii and
improve the filtration performance. In some coal preparation plants, this is already applied to
increase the dewatering efficiency as an alternative to other means. In fact, currently, Mingo
Logan Coal Preparation Plant has been blending spiral clean coal (18 X 100 mesh) with the finer
flotation clean product (smaller than 100 mesh) at about 1:1 ratio before filtration in dewatering
centrifuges.
The preliminary dewatering test results using RW, RU and RV on one-to-one
spiral/flotation blend product showed that significant cake moisture reduction can be obtained
using novel dewatering aids by about 20.5-28%, while also increasing the rate of dewatering so
much that the formation time could be reduced by as much as 50%. A set of three preliminary
tests was conducted to investigate the effectiveness of the dewatering aids (Table 2.9). The
dosages used were 1, 3 and 5 lb/ton. The baseline tests produced an average of 19% final cake
moisture at about 8mm cake thickness. Even at low dosage, 1lb/ton, when RW, RU and RV was
used, the cake moisture was decreased to 15.6%, 14.4%, and 14.8%, respectively, and the
filtration kinetics were increased by 30-50 %. Of the reagents and dosages being tested, RU was
the most effective for the cake moisture reduction. In this case, the addition of 5 lb/ton RU
reduced the cake moisture from 19.00% to 13.7% which corresponds to 20-30 % overall
moisture reduction.
71 |
Virginia Tech | Table 2.9 Effect of reagent addition on dewatering of Mingo Logan mixture sample
(50% flotation product and 50% spiral product)
Reagent Moisture Content (%)
Dosage
Reagent Reagent Reagent
(lb/ton)
RW RU RV
0 19.0 19.0 19.0
1 15.6 14.4 14.8
3 15.7 13.8 14.9
5 15.1 13.7 14.3
Similar dewatering tests were conducted to evaluate the filtration performance with
different types of dewatering aids as a function of filtration time and specific cake weight.
Filtration tests were done at the same vacuum level; however, the dry cycle times were varied
from 15 to 100 seconds. RU and RV were prepared at 2:1 active solvent ratio, and the dosage
amount was fixed at 3 lb/ton. The coal slurry was conditioned with the dewatering aids for 2
minutes. Then, flocculant was added at a 25g/ton dosage and conditioned at a very low intensity
for 15-20 seconds. Tests were conducted on a pre-measured amount of slurry to differentiate the
specific cake weight. This, in turn, varied the cake thickness from 5 to 20 mm. The cake
formation time, dry cycle time, and cake weights were recorded for each test. The filter
production rate was plotted as pounds of dry solids per hour per square foot, and the filtration
time used for filtration rate calculations was the sum of cake formation time and dry cycle time.
Figure 2.9 shows the relationship between production rate and cake moisture for the
spiral/flotation mixture sample.
72 |
Virginia Tech | 40
35
30
25
20
15
10
0 100 200 300 400 500 600 700 800
Figure 2.9 Cake moisture versus normalized filter cake production rate
The baseline tests were conducted without reagents and produced a cake of 29-34%
moisture at about 5-20 mm thickness, yielding a production rate in the range of 60-120lb/hr/ft2.
Throughput greater than 120 lb/hr/ft2 was found to be impossible because the cake formation
time was being prolonged to impractical limits. The additions of RV and RU, along with 25
g/ton flocculant, surpassed the baseline throughput to a great extent. The usage of RV reduced
the final cake moisture to 14-18%, and the usage of RU reduced the final cake moisture to 18-
21%. The production rates were also increased by multiples of 1.5 to 2.5, corresponding to 150-
190 lb/hr/ft2. The results also showed that the throughput could be increased at the expense of
cake moisture. When the production rate was increased to the 200-400 lb/hr/ft2 range, the cake
moisture increased to 20-25%; however, final cake moistures were still 5-15% lower compared
73
)%(
erutsioM
none
rv 3lb/t +f 25g/t
ru 3 lb/t +f 25gr/t
FilterProduction Rate (lb/hr/ft2) |
Virginia Tech | to baseline. Further increases in production rate were also achieved i.e., the 550-650 lb/hr/ft2
range was achieved at a 25-30% moisture range. It was obvious that the addition of reagent
outperformed baseline, and it was possible to increase the production rate with significantly
lower final cake moisture values. The results also showed that the use of RV represented better
performance than RU because it gave higher filtration rates at the same, or lower, final cake
moisture values.
Further data evaluation was carried out on the effect of filtration time on final cake
moisture. Figure 2.10 shows the correlation between the cake moisture and total filtration time
(normalized with cake weight) in the absence and presence of dewatering aids. In the baseline
tests the total filtration time varied between 130 and 360 seconds. On the other hand, the total
filtration times in the presence of RU and RV were 65 to 120 seconds and 50 to 118 seconds,
respectively. As seen, the dewatering aids that were tested decreased the filtration time, which,
in turn, produced higher production rates. It is also noteworthy that these outstanding times were
achieved while maintaining very low moisture levels.
The test data also indicated that cake formation time is an important factor in determining
dry cycle time and, thus, cake moisture and throughput, as well. Figure 2.11 shows the effects of
dewatering aids on cake formation time and, eventually, the filtration kinetics. The formation
times were found to be much shorter in the presence of RU and RV. When used with a
flocculant, even at higher specific cake weights, the formation times were shorter than those that
were recorded during the baseline tests.
74 |
Virginia Tech | As seen in Figure 2.11, only 10-15 seconds were required to produce 4.5 lb/ft2 of coal
when dewatering aids were used; however, in baseline tests, approximately 100-120 seconds
were needed for the same coal production. When the cake weight was increased to 7 lb/ft2, the
cake formation time was 20 seconds using dewatering aids, where it was 340 seconds for
baseline test. Cake weights above 5 lb/ft2 were found to be impractical for dewatering in the
absence of dewatering aids. Conversely, it was still possible to produce a cake weight of 11
lb/ft2 while maintaining a short cake formation time. The results showed that the use of
dewatering aids decreased the cake formation time by several multiples. The usage of
dewatering aids also substantially improved the throughput. This is a strong indication that in
the presence of dewatering aids, more material can be treated.
40
35
30
25
20
15
10
0 0.005 0.01 0.015 0.02
Figure 2.10 Cake moisture versus normalized filtration time
75
)%(
erutsioM
none
RV 3lb/t + 25 g/t F
RU 3lb/t + 25 g/t F
Time (hr/lbft2) |
Virginia Tech | 19
17
15
13
11
9
7
5
3
1
0 50 100 150 200 250 300 350 400
Figure 2.11 Cake moisture versus total cake formation time
i) Pilot-Scale Test Results (Series A – Fine Coal Only)
The pilot-scale dewatering tests were conducted using two feeds intercepted from the
plant’s main stream (i.e., flotation concentrate and blended products). The first set of tests
focused on dewatering of the flotation product, which is very difficult to dewater because of the
amount of fine size particles and high ash content. As mentioned earlier, nearly 43% particles of
this product was passing 325 mesh (-44 µm) and contained 81% ash. The tests were conducted
over a range of filter feed rate that typically varied at 2, 3, 4 and 7 gal/min of coal slurry. The
preliminary analysis showed that of the feed rates, 4-7 gal/min were at optimal ranges for the
filtration tests and also adequate to produce 0.33 to 1.0 inch cake thicknesses. The pilot-scale
76
2tf/bl
thgieW
eKaC
none
RV 3lb/t + 25 g/t F
RU 3lb/t + 25 g/t F
Cake Formation Time (sec) |
Virginia Tech | Table 2.10 Effect of RV on Mingo Logan flotation product sample
Cake Moisture (%)
Filter Cake
0.5 0.5 1.0 1.0
Time Thickness 25 g/t
Control lb/ton lb/ton lb/ton lb/ton
(sec) (mm) Floc
RV RV* RV RV*
184 25 31.49 34.61 27.11 26.13 - -
120 15 31.59 34.13 25.92 25.95 31.73 23.59
85 10 31.18 - 28.49 - 28.12 24.45
65 8 31.63 36.2 - - 28.9 26.67
*With Roller
test procedure was same as the mixture sample tests except that the flotation product was
intercepted from the distribution box pipeline.
The first set of dewatering tests was conducted with a pilot-scale horizontal belt filter
(HBF) at the feed rate of 4gal/min. Baseline tests provided a filter cake with 31~32% moisture.
With 25 g/ton flocculant addition to the feed, even though a slight decrease in cake formation
time was observed, the cake moisture increased to 34~36%. On the other hand, at 0.5 lb/ton RV
addition, the cake moisture was reduced down to approximately 26~28% (see Table 2.10). The
results also showed that, in this set of tests, the belt speed had a slight effect on the dewatering
when flocculant and reagent were added; however, it did not appear to influence the cake
moisture for the control tests (it only caused changes in cake thickness and formation time).
During the tests, it was observed that even though the filter feed rate was kept constant;
using RV increased the cake thicknesses. This can be attributed to increased porosity, which in
turn creates more permeable cake. This phenomenon also caused a cake-cracking problem. As a
result, the vacuum pumps also lost pressure. To overcome this problem, a roller was attached on
the belt to press down the cake, presumably to prevent the cracking. The test results clearly
showed that the cake moisture was further reduced to 23~26% when a roller was applied to the
77 |
Virginia Tech | Table 2.11 Effect of RV on Mingo Logan flotation product sample (with 5 g/t of flocculant)
Filter Cake
Final Cake Moisture (%)
Time Thickness
1 lb/ton 1 lb/ton 3 lb/ton 3 lb/ton
(sec) (mm) Control
RV RV* RV RV*
184 25 31.49 28.65 26.32 25.69 26.93
120 15 31.59 29.05 25.61 29.03 25.19
85 10 31.18 28.63 28.01 29.13 27.12
65 8 31.63 24.85 26.81 27.37 28.25
*With Roller
dry cake to help seal the pores inside the cake. Overall, the use of roller resulted in additional
moisture reductions from 3-4% to 15% at 0.5 lb/ton and 1 lb/ton of RV addition, respectively.
Table 2.11 shows another series of tests were conducted to investigate the effect of the
flocculant addition. The effect of roller was also tested in the presence of dewatering aid. The
cake moisture was in the range of 24 to 29% with combined use of 5 g/t of flocculant and 1
lb/ton of RV. Yet again, increased cake porosity and the cracking was a problem; however,
when the roller was applied in some cases moisture was lowered.
j) Pilot-Scale Test Results (Series B – 50% Mixture of Fine and Coarse Coal)
The next set of tests was conducting using an equal mixture of fine and coarse coal. The
fine coal was obtained from the froth concentrate, while the coarse coal was intercepted from the
screen-bowl feed containing approximately 28% solids by weight and 12% ash on dry basis. The
tests were conducted over a range of filter feed rate that typically varied from 2, 3, 4 to 7 gal/min
of coal slurry. Unlike the flotation product, it was found out that the 3 gal/min feed rate was
optimal for the blend filtration tests, which would be sufficient to produce 1/3 to 1 inch cake
thicknesses. Table 2.12 shows a summary of the pilot-scale test data obtained at the Mingo
Logan plant spiral/flotation product mixture sample using RU at 3lb/ton and flocculant at 25g/t
78 |
Virginia Tech | dosage at the feed rate of 3gal/min. Each series of tests were conducted as a function of total
filtration time.
This reagent reduced the cake moisture from about 30% down to 21% at 3 lb/ton dosage.
Meanwhile, the cake formation time decreased by 20~50% with the addition of RU as
dewatering aids. When used alone, it could produce a low-moisture cake (approximately
21~22% moisture), and the cake formation time was significantly reduced down to 14~75
seconds from approximately 85 seconds. The flocculant alone was not capable of reducing the
moisture content to the level that RU achieved; although, they could reduce the formation time
more significantly. The use of flocculants resulted in loss of vacuum pressure in the pump, an
indication of increased cake porosity, but did not help to remove the surface water that was
entrapped inside the flocs. However, the combined use of flocculants and RU could achieve a
very short cake formation time. When used together with 25 g/t flocculants, RU reduced the
cake formation time further to 7~24 seconds, while the final cake moisture remained at almost
the same level. In this case, the cake formation time was reduced by 50~75% over the belt speed
range under test, and the moisture was reduced from 30-34% down to 21~24%. Test work
performed using RU showed that a filter cake with good handling characteristics could also be
produced.
Table 2.12 Effect of RU and flocculant on Mingo Logan flotation product sample
Final Cake Moisture (%)
Filter Cake
25 g/t Floc
Time Thickness 3 lb/ton
Control 25 g/t Floc & 3 lb/ton
(sec) (mm) RU
RU
184 25 34.38 26.6 25.05 24.96
120 15 29.62 26.9 21.8 24.66
85 10 30.78 29.1 21.3 18.7
65 8 29.99 28.81 21.5 21.7
Vacuum (Inch Hg) 16-14 11-5 15-11 7-5
79 |
Virginia Tech | In comparison, RV was tested and the results given in Table 2.13 showed that RV was less
effective in reducing moisture and increasing the kinetics of dewatering. However, the difference
between RU and RV was narrowed when each of these two reagents were used together with low
dosage (25 g/t) of flocculants. It was also noticed that in the presence of dewatering aids, change in
belt speed made a slight difference in the moisture reduction, especially while operated at lower
speeds. When the filter feed rate was fixed, the short retention time of the materials over the
vacuum zone was compensated by the thin cake thickness at higher belt speed, and thick cake at
lower belt speed.
Table 2.13 Effect of RV and flocculant on Mingo Logan flotation product sample
Filter Cake Final Cake Moisture (%)
Time Thickness 25 g/t 3 lb/ton 25 g/t Floc &
Control
(sec) (mm) Floc RV 3 lb/ton RV
184 25 34.38 26.6 28.11 24.56
120 15 29.62 26.9 28.9 24.74
85 10 30.78 29.1 29.78 22.63
65 8 29.99 28.81 26.34 21.4
Vacuum (Inch Hg) 16-14 11-5 16-15 8-4
80 |
Virginia Tech | 2.3.3. Coal Clean Site
a) Site Descripton
The Coal Clean Panther Preparation Plant is located in Dry Branch, approximately 15
miles south of Charleston, West Virginia. The Panther Preparation Plant currently processes 63
mm x 0 raw coal for ultimate use in the metallurgical and steam markets. The minus 325 mesh
size fraction material reports to the discard stream without any type of processing method.
However, if this size fraction is recovered and dewatered, loss of valuable source could be turned
into profit. For this reason the plant management looked for alternatives for the recovery of the
minus 325 mesh coal that is presently discarded to refuse. To evaluate the feasibility of the
recovery and dewatering of this size fraction, extensive pilot scale flotation and dewatering tests
were conducted. A number of tests with pilot scale centrifuge unit were conducted for
comparison reasons.
b) Reagents
In this investigation, majority of the pilot scale dewatering tests were performed with
varying amounts of three types of dewatering aids, namely RW, RU and RA. Diesel was used as
solvent at one to two (1:2) ratio (dewatering aid: solvent). Each of the reagents were tested over
a range of dosages typically ranging from 0 to 20 pounds per ton for the ratios of the ultra-fine
and fine products in the feed to the Filter Module and the Centrifuge Module.
c) Coal Samples
For this particular test site, pilot-scale dewatering tests were conducted on feeds comprised
of different mixtures of coarse and fine coal feeds. These included (i) 100 mesh x 0 feed stream
from the overflow of the primary classifying cyclones, (ii) 325 mesh x 0 feed raw stream, and
(iii) blends of 100 mesh x 325 and minus 325 mesh product from the flotation column. The
81 |
Virginia Tech | Column Conditioner
Module Module Pilot-Scale
Test Circuit
Plant
Feed
Filter
Module
Centrifuge
Module
A
Dewatered
Product Dewatered
To Product
Conditioner
Figure 2.12 Configuration of the mobile unit test modules evaluated at the Panther
Preparation Plant site
blends were established at ratios of 1:1 and 3:1 of minus 325 mesh material to the 100 x 325
mesh product.
d) Pilot-Scale Procedures
The column flotation, conditioner, disc filter, and centrifuge modules were set up at the
Coal Clean Plant to accommodate the pilot-scale testing program. A total of 115 tests were run
over a one month period to establish optimum conditions for the proposed plant upgrades that
were under consideration to accommodate the proposed POC-scale test circuit. The arrangement
of the modules for the tests is presented in Figure 2.12. For the majority of the test work, the
plant supplied a relatively consistent feed stream of minus 325 mesh material to the Column
Module from the overflow of the secondary classifying cyclones. The secondary classifying
cyclones were fed from the overflow of the primary classifying cyclones which were separating a
82 |
Virginia Tech | 16 mesh x 0 slurry stream at a nominal size of 100 mesh. The underflow from the secondary
classifying cyclones reported to a bank of conventional flotation cells and the overflow was
discarded to the refuse thickener. A series of tests was also conducted on the 100 mesh x 0 feed
stream from the overflow of the primary classifying cyclones. The minus 325 mesh raw feed
was cleaned in the Column Module, with the clean coal routed to the Conditioner Module. The
conditioned slurry was then dewatered in either the Filter Module or the Centrifuge Module.
The Panther Preparation Plant offered a unique opportunity to conduct a series of tests with
various blends of 100 mesh x 325 mesh conventional flotation product added to the minus 325
mesh product from the Column Module. The projected blends were established at ratios of 1:1
and 3:1 of minus 325 mesh material to the 100 x 325 mesh product. Maintaining the blend at the
various ratios proved somewhat difficult due to the changing flows for the products, but most of
the tests were conducted within 5% of the projected blend. Although the Filter Module and the
Centrifuge Module were not operated at the same time, the tests included almost the same sweep
of reagent tests for both modules.
e) Pilot-Scale Test Results
Table 2.14 presents the results for dewatering tests with various reagents while
dewatering the minus 325 mesh product in the Filter Module. The sample was first floated using
a mixture of Nalco 01DU113 (0.25 unit) + 01DU145 (1.0 unit) as collector and Nalco-948 as
frother (9 units) in a flotation column. The solids content in the filter feed ranged from 4-5% by
weight. For dewatering tests, RW, RU and RA were used as dewatering aids at various dosages.
The vacuum pressure was approximately 20.5 to 23.0 inches Hg; and the cake thickness was
around 6-7 mm. The results indicate that the baseline moisture content of the filter cake (with no
reagent) was very consistent and it was approximately 29.0%. This value was reduced to 27.5%,
83 |
Virginia Tech | TTaabbllee 22..1154 . CCooaall CClleeaann ccooaall ((5m0i%nu ms i3n2u5s m32e5s hm) edsehw aantedr 5ed0 %us 1in0g0 xR3W25, mRUes,h a) ndde wRaAte dreisdp uesrsinegd
iRnW N aalncdo R01UD dWis1p1er0s ed in Nalco O1DW110
.
ReRaegaegnetn t MoistMuroei sCtounret eCnot n(%ten) t (%)
DoDsoasgaeg e RW RU RA
RW RU
(lb(l/bt)/ t) (1:4 Ratio) (1:2 Ratio) (1:2 Ratio)
(1:2 Ratio) (1:2 Ratio)
0.00 29.10
0.00 19.98
3.08 27.01
1.98 19.82
6.08 26.04
4.47 17.76
9.75 24.66
0.00 19.98
0.00 29.28
2.80 18.26
2.80 27.28
5.40 17.93
5.48 24.71
7.95 17.65
9.43 23.42
0.00 29.05
1.62 28.47
2.19 27.49
7.50 27.49
24.7% and 23.4% with RA, RW, and RU, respectively. As such, RU provided the best overall
performance for dewatering of the minus 325 mesh product. Considering the amount of the
minus 325 size fraction in this feed, RU was capable of reducing the final cake moisture by 20%.
Table 2.15 presents the results for dewatering tests using the Filter Module for a blend of
the minus 325 mesh product with the 100 x 325 mesh product from the conventional flotation
cells at a ratio of 1:1. For flotation tests, same procedure and collector and frother dosages were
employed. As shown, the addition of the coarser material had a dramatic effect on the moisture
content of the product. With no reagent added, the addition of coarser material reduced the
baseline moisture from 29% (Table 77) to 20% (Table 78). As the reagent dosage was increased,
the 20% moisture content was further reduced to 17.8% with RW and to 17.6% with RU.
Table 2.16 presents the results for the dewatering tests using the Filter Module for a blend
of the minus 325 mesh product with the 100 mesh x 325 mesh product from the conventional
flotation cells at a ratio of 3:1. The solids content in the filter feed ranged from 5% to 6%. As
84 |
Virginia Tech | Table 2.15. Coal Clean coal (50% minus 325 mesh and 50% 100x325 mesh) dewatered using
RW and RU dispersed in Nalco O1DW110
.
Reagent Moisture Content (%)
Dosage
RW RU
(lb/t)
(1:2 Ratio) (1:2 Ratio)
0.00 19.98
1.98 19.82
4.47 17.76
0.00 19.98
2.80 18.26
5.40 17.93
7.95 17.65
seen, when the amount of the minus 325 size material was increased, the baseline moisture value
was also increased. The results, as expected, lie between those obtained in the two test series
noted above. Of the dewatering aids, RU was the most effective and capable of reducing the
cake moisture from 24.78% to 20.77 % which correspond to 17 % overall moisture reduction.
Table 2.17 presents the results for dewatering tests for a minus 325 mesh product in the
Filter Module with the filter disc speeds ranging from 4 to 1.5 min/rev. The sample was floated
using a mixture of 01DU110 as collector and Nalco 01DU009 as frother in a flotation column.
The solids content in the filter feed ranged from 9-10% by weight. The product rate increased
from 43.78 to 59.18 lb/h with only a slight increase in the moisture content from 27.5 to 28.5%
when using RU at approximately 2.0 lb/ton.
85 |
Virginia Tech | Table 2.16 presents the results for pilot scale dewatering tests using RW for a blend of
the minus 325 mesh product with the 100 x 325 mesh product at a ratio of 1:1 using the
Centrifuge Module. The sample was floated using mixture of 01DU110 as collector and Nalco
01DU009 as frother in a flotation column. The solids content in the centrifuge feed ranged from
9-10% by weight. The 150 mm diameter centrifuge operated at approximately 480 rpm with a
differential of 95:1 for the conveyor. The moisture of the product increased from 33.2 to 34.5%
as the solids yield increased from 87.2 to 93.4% with the increase in reagent dosage. The higher
Table 2.16 Coal Clean coal (75% minus 325 mesh and 25% 100x325 mesh) dewatered
using RW, RU and RA dispersed in Nalco 01DW110
Reagent Moisture Content (%)
Dosage RW RU RA
(lb/t) (1:2 Ratio) (1:2 Ratio) (1:2 Ratio)
0.00 24.78
2 23.57
4.36 24.83
7.4 22.49
0.00 24.78
1.85 23.01
3.78 22.55
7.7 20.77
0.00 24.78
2.0 23.00
3.56 23.82
8.47 23.68
Table 2.17 Effect of disc speed on product rate and moisture content for Coal Clean coal
(minus 325 mesh) from the filter module
Disc Rotation Filter Cake Moisture
Speed Production Content
(min/rev) (lb/hr dry) (%)
4.0 43.78 27.49
3.0 44.86 27.76
2.0 52.94 28.63
1.5 59.16 28.47
86 |
Virginia Tech | Table 2.18 Effect of Reagent 01DW133 on Coal Clean coal (50% minus 325 mesh and 50%
100x325 mesh) from the centrifuge module
Reagent Nalco 01DW133 in 01DU110 (1:2 Ratio)
Dosage Moisture Solids
(lb/t) Content (%) Yield (%)
0.00 33.24 87.20
2.31 33.94 89.60
4.51 33.36 90.30
6.16 33.86 86.00
moisture and improved yield appear to be mostly due to an increase in the recovery of the ultra-
fine material.
Table 2.18 presents the results for dewatering tests using Reagent 01DW133 for a blend
of the minus 325 mesh product with the 100 x 325 mesh product at a ratio of 1:1 using the
Centrifuge Module. The sample was again floated using a mixture of 01DU110 as collector and
Nalco 01DU009 as frother in a flotation column. The solids content in the centrifuge feed
ranged from 9-10%. The moisture of the product increased slightly from 33.2 to 33.8% and the
solids yield increased from 87.2 to 90.3% with the increase in reagent dosage. Once again, the
increases in moisture and yield were attributed to increases in the recovery of ultra-fine material.
During the continuous test work conducted at the Panther Preparation Plant site, a series
of timed samples were collected periodically from various points around the pilot plant so that
complete mass balances could be established for the different unit operations. The samples
included representative splits for the column cell (feed, product and tails), filter (feed, product
and filtrate), and the screen bowl centrifuge (feed, product, effluent and drain). The series of
tests for the minus 325 mesh feed showed that the 300 mm (1 ft) diameter Column Module
produced an average of 68.2 lb/hr of concentrate. Depending on the particular operating
conditions, the clean coal capacity ranged from a low of 37.4 lb/hr to a high of 127.6 lb/hr. The
87 |
Virginia Tech | 2.3.4. Concord Site
a) Site Description
The Concord Coal plant is located near Birmingham, Alabama. This facility processed
50 mm x 0 coal for the metallurgical and steam markets, with a design plant feed rate of 1,000
tph and typical clean coal yields of 55%-60%. The intermediate/fine coal circuit consists of
primary classifying cyclones (PCC), spirals, secondary classifying cyclones (SCC), froth
flotation, and screen bowl centrifuges. The overflow from the PCCs is fed to the SCCs; the SCC
underflow is the feed stream to flotation (4 banks of five 180-ft3 cells), while the SCC overflow
is piped to a refuse thickener. The coal being processed is very soft and fine in size consist and
the feed to the flotation cells can be as much as twice that of the design flowsheet rate (design 54
t/hr vs. actual 80-100 t/hr). The feed to the flotation cells is approximately 80% minus 325 mesh
(0.045 mm). The flotation and spiral clean-coal products are combined and then dewatered via
four 44” x 132” screen bowl centrifuges with a total design feed rate of 2,200 gal/min and 242
t/hr.
The primary objectives of the test program were to determine whether (i) a thick and low-
moisture filter cake and (ii) a filter cake with good material handling characteristics could be
produced from the minus 100 mesh flotation feed stream (primary cyclone overflow) that is
currently processed in conventional flotation cells and centrifuges at this plant. To meet these
objectives, extensive laboratory and pilot scale flotation and dewatering test program was
conducted at Virginia Tech Mineral Processing Facility in Virginia, and at the Concord coal
preparation plant in Alabama. The laboratory tests included the performance evaluation of
various types and dosages of dewatering aids to collect dewatering data that could be used in
89 |
Virginia Tech | direct support of pilot-scale test work. The data obtained from the laboratory tests were used to
provide a technical guidance for the pilot scale test program.
b) Reagents
The laboratory and pilot scale tests included both the flotation and dewatering tests. For
laboratory flotation tests, diesel and MIBC were used as collector and frother, respectively. The
floated sample was then subjected to dewatering tests using RW3 and RV. For pilot-scale
flotation tests RW, RU, RV and diesel as collectors and Nalco DU009 as frother were used. The
same reagents were used as the dewatering aids in filter tests. Each of the reagents was tested
over a range of dosages that typically varied from 0 to10 lb/ton of dry coal. As mentioned
elsewhere, these dewatering aids are insoluble in water. For all the tests, flotation and
dewatering, diesel was used as solvent at one to two (1:2) dewatering aid-to-diesel ratio.
c) Coal Samples
The coal slurry samples for the flotation and dewatering tests were collected from the
minus 100 mesh flotation feed stream (primary cyclone overflow) that was processed. The same
coal slurry feed was used for the pilot-scale flotation and dewatering tests.
d) Laboratory Procedures
The flotation feed sample (minus 100 mesh) from the Concord plant was first floated in a
laboratory mechanical flotation cell using 0.66 lb/t of diesel and 0.33 lb/t of MIBC to remove
ash-forming minerals. The floated product was then subjected to dewatering tests at about
13.5% solids by weight. In these tests, RV and RW3 (both diluted to 33.3% in solvent) were
used as dewatering aids. Immediately prior to each filter test, a known volume of slurry was
conditioned for 5 minutes with the reagent in a mechanical shaker and then poured into a
90 |
Virginia Tech | Buchner funnel before applying vacuum of 20 inches Hg. A constant drying cycle time of 2
minutes was used in all tests. The thickness of the filter cake varied from 7 to 9 mm.
e) Pilot-Scale Procedures
The on-site test work was conducted using the Column, Conditioner, and Filter Modules.
In each test, the cyclone overflow from the plant was first upgraded using the Column Module (a
305 mm or 12-inch diameter Microcel column) using diesel and RW, RU and RV as collectors
and Nalco DU009 as frother. The plant feed was reasonably consistent for most of the test work
and contained 3.5-5.5% solids by weight and 23.3-26.5% ash. The column produced a clean coal
product with 7-11% ash and 84-87% combustible recovery, depending on the reagent type and
dosage used. The froth product was then dewatered using the Filter Module (i.e., 0.186 m2 (2 ft2)
filter area, 10 sector Peterson disc filter). The same reagents listed above, namely RW, RU and
RV, were also used as the dewatering aids in the filter tests. Each of the dewatering aids were
tested over a range of dosages that typically varied from 0 to 10 lb/ton of dry coal. Timed
samples were collected periodically from various points around the test modules to establish
typical material balances and reagent addition rates for this particular coal.
f) Laboratory Test Results
Several series of laboratory dewatering tests were conducted to collect data that could be
used as guidance for the pilot scale dewatering tests. Table 2.19 shows the laboratory test results
obtained on the Concord Plant flotation feed sample using RV and RW3 as the dewatering aids.
A 62.5 mm diameter Buchner funnel was used. The solid content of the flotation feed sample
was 6%. It was increased up to 13.5% after flotation. 0.66 lb/t diesel and 0.33 lb/t MIBC was
used during flotation. Vacuum setup point was 20 inches Hg. Cake thickness: 7-9 mm.
Conditioning time: 5 minutes; drying cycle time: 2 minutes. The volume of the slurry was 100
91 |
Virginia Tech | Table 2.19 Effect of reagent addition on Concord flotation feed (100 mesh x 0) using RV and
RW3
Reagent Moisture Content (%)
Dosage
RV RW3
(lb/t)
0 24.6 24.6
0.5 21.3 23.2
1.5 19.6 21.0
2.5 18.1 19.4
10.0 18.2 18.6
ml. As shown, the moisture content of the filter cake was reduced as the reagent dosage
increased. At dosages of 0.5 lb/t and 2.5 lb/t of RV, the moisture contents of the filter cake were
reduced from 24.6% down to 21.3% and 18.1%, respectively. Similar results were obtained
when RW3 was used as the dewatering aid. These values correspond to a 10-26% moisture
reduction in the filter product.
g) Pilot-Scale Test Results
The first series of dewatering tests were conducted at various levels of vacuum pressures
applied. The feed slurry was first floated using diesel collector and Nalco DU009 frother in
flotation column. As shown in Table 2.20, there was a strong correlation between the cake
moisture and vacuum pressure. The results showed that when the vacuum pressure was
increased the cake moisture was decreased from 31.12% down to 24.18%.
Table 2.21 provides an overall summary of the pilot-scale test data obtained at the
Concord plant. In the first series of tests, RW, RU and RV were evaluated over a wide range of
reagent dosages at a constant disc speed of 4 min/rev. The test data show that RW was the most
effective of the three dewatering aids (Table 2.21a). This reagent reduced the cake moisture
from 25.5% down to 20.2% at the highest reagent dosage of 2.7 lb/t. RU, which was rather less
92 |
Virginia Tech | effective than RW, was not capable of reducing the moisture content to less than 21.5% (Table
2.21b). However, this moisture level was achieved at a very low reagent dosage level of just
0.88 lb/t. In fact, higher dosages of RU did not appear to be as effective in reducing moisture in
this particular series of tests. RV was generally the least effective in reducing moisture (Table
2.21c).
Table 2.20 Pilot-scale test results obtained on the Concord flotation feed sample (100 mesh x
0) using various vacuum pressures
Vacuum Moisture
Pressure Content
(Inch Hg) (%)
10 31.12
15 26.83
20 24.18
Although less effective in reducing moisture, RV generally provided the best overall cake
thicknesses (up to 12 mm) when compared to RW and RU. The thick cakes produced using RV
also possessed the best material handling characteristics and were cited by plant personnel as the
most suitable for their particular needs. Therefore, several additional tests were conducted using
RV to determine whether lower moistures could be achieved by adding the reagent directly to the
flotation column feed in place of the diesel collector. As shown in Table 2.22, this strategy
significantly improved the moisture reduction and greatly reduced the total reagent requirement.
More importantly, the lower moistures were obtained at relatively large cake thicknesses (i.e., 9-
10 mm). In one test run, the cake moisture was reduced from 25.1% down to 20.9% (a moisture
reduction of 16.5%) while maintaining a cake thickness of 9 mm. The total reagent dosage
93 |
Virginia Tech | 2.3.5. Buchanan Site
a) Site Description
Consolidation Coal Company’s Buchanan Mine #1 is an underground coal mine located
two miles south of Route 460, adjacent to State Route 632, at Mavisdale, Buchanan County,
Virginia. Consol Energy Inc., located in Pittsburgh, Pennsylvania, is the parent company of
Consolidation Coal Company. The Buchanan preparation plant processes approximately 5
million tons of 37.5 mm x 0 raw coal (2006) from the Pocahontas 3 Seam for use in the
metallurgical and steam markets.
The objectives of this study were to (i) identify the best possible reagents and
combinations thereof for this specific coal and (ii) identify the conditions under which a given
dewatering aid can give the best performance. To meet the objectives, laboratory and pilot scale
tests were conducted to evaluate the performance of various types and dosages of dewatering
aids.
b) Reagents
The investigation, both laboratory and pilot-scale, was performed with varying amounts
of three types of dewatering aids, namely RW, RU, and RV. Since these dewatering aids are
insoluble in water, they were dissolved in a solvent. In both laboratory and pilot scale
experiments diesel was used as the solvent. The ratio of reagents-to-solvent was optimized in
previous studies by varying the individual dosages (0.5 to 3lb/ton), while maintaining the total
blend dosage constant. For this test program, the optimum combination for a given dewatering
aid and solvent is one to two (1:2) dewatering aid-to-diesel ratio.
96 |
Virginia Tech | c) Coal Samples
The test work was conducted on different samples taken from Consolidation Coal
Corporation’s Buchanan Preparation Plant in Mavisdale, Virginia. These samples included (i) a
flotation feed sample, (ii) a grab sample of current flotation product, and (iii) a slip-stream
sample of filter feed (all taken on various dates).
d) Laboratory Procedures
Prior to experiments sieve analysis was conducted for each sample by wet screening
using 600, 300, 150, 75 and 45 micron sieves. Dewatering tests were mostly conducted on filter
feed and flotation product samples. The solid concentration of Buchanan’s flotation product was
about 25% by weight while maintaining approximately 25-30% of minus 45 micron material.
Table 2.24 shows the sieve analysis results on flotation product sample. The flotation product
sample was occasionally mixed with a portion of the spiral product at the plant at a ratio of 1:1 to
have better dewatering kinetics. Table 2.25 shows the sieve analysis results of spiral/flotation
product slurry collected from the plant.
For the laboratory-scale batch dewatering tests, the samples were collected in 5-gallon
buckets and to be able to receive a representative sample, samples were homogenized by mixer.
When the plant flotation feed sample was used in the dewatering tests, the sample was first
floated in a laboratory mechanical flotation cell using 0.66 lb/t of kerosene and 0.33 lb/t of
MIBC to remove ash-forming minerals. The floated product was then subjected to the
dewatering tests at about 20% solids by weight. Immediately prior to each filter test, a known
volume of slurry (whether flotation product or mixture sample) was conditioned with the reagent
in a mechanical shaker and then poured into a Buchner funnel before applying vacuum. The
dewatering aids RW and RU, diluted to 33.3% in solvent, were used in these tests. The
97 |
Virginia Tech | Table 2.24 Screen analysis of the Buchanan flotation product used for dewatering tests
Particle Size Pipe 1 Pipe 2 Pipe 3
(Mesh) Weight (%) Weight (%) Weight (%)
Plus 28 3.6 3.4 7.7
28x45 19.0 15.8 27.3
45x100 23.0 18.0 24.1
100x200 14.9 13.2 13.2
200x325 8.1 25.3 6.9
Minus 325 31.4 24.3 20.7
following conditions were kept constant during the tests: 20 inches Hg of vacuum, 2 minutes of
drying cycle time, 10-15 mm of cake thickness, 100 ml volume of feed slurry and 5 minutes of
conditioning time.
e) Pilot-Scale Procedures
In the pilot-scale dewatering tests, the Conditioner Module and Filter Module were
required since the feed slurry for the tests was supplied directly from the Buchanan Preparation
Plant. The pilot-scale disc filter tests were conducted on flotation product samples.
Table 2.25 Screen analysis of the Buchanan filter feed used for dewatering tests
Particle Size Pipe 1 Pipe 2 Pipe 3
(Mesh) Weight (%) Weight (%) Weight (%)
Plus 28 6.0 6.3 5.5
28x45 23.4 23.3 21.8
45x100 21.7 22.5 19.9
100x200 14.7 14.8 14.6
200x325 8.2 6.9 8.9
Minus 325 26.0 26.2 29.3
98 |
Virginia Tech | Table 2.26. Effect of reagent addition on dewatering of Buchanan’s filter feed
Reagent Moisture Content (%)
Dosage
RW RU
(lb/t)
0 18.1 18.1
1 17.58 17.5
3 15.2 16.78
5 14.9 16.58
f) Laboratory Test Results
Table 2.27 gives the laboratory test results obtained on the Buchanan plant flotation
product using RU and RW as the dewatering aids at 20 inches Hg vacuum. As shown, the
moisture content in the filter cake was reduced with increasing reagent addition. At RU
additions of 1 and 5 lb/t, the moisture contents of the cake were reduced from 17.6% to 16.1%
and 14.4%, respectively. Similar results were obtained when RW was used as dewatering aid.
These values corresponded to a 15-20% moisture reduction in the filter product.
Similar dewatering tests were conducted on the filter feed sample (which contains
approximately 10 g/t of Nalco 9806 polymer flocculant as the dewatering aid) using RW and RU
as dewatering aids. Results for the filter feed sample are summarized in Table 2. The results
show that the moisture content of the filter product again decreases with increasing RW and RU
additions from 1 to 5 lb/ton. In this case, the addition of 5 lb/ton of RW reduced the cake
moisture from 18.1 to 14.9%, giving a percentage moisture reduction of about 20%. The
Table 2.27. Effect of reagent addition on dewatering Buchanan’s flotation product
Reagent Moisture Content (%)
Dosage
RW RU
(lb/t)
0 17.6 17.6
1 16.2 16.1
3 16.2 14.9
5 15.0 14.4
99 |
Virginia Tech | moisture reduction is quite similar to that obtained for the flotation product, except that the
moisture content of the filter cake product obtained using RU was almost 2 percentage units
lower, i.e., 14.4% vs. 16.6% moisture in the filter cake (see Table 2.27 and Table 2.26). The
reasons for the relatively poorer behavior of RU may be related to the presence of flocculant in
this particular sample. Apparently, the polymer flocculant has an adverse effect on the
performance of RU during dewatering. Those poor results are also due to the Ca2+ ions present
in Buchanan plant water.
Two series of laboratory filtration tests were conducted to determine the effects of
agitation intensity on the dewatering performance of the Buchanan filter feed. The first series of
tests were conducted using a laboratory shaker to condition the feed samples. The shaker was a
low-energy conditioner that uses reciprocating motion (similar to wrist-action shaking) to gently
mix slurry contained in a 100-ml glass conditioning flask. A second series of tests were
conducted using a 100-ml Plexiglas cell equipped with a three-blade propeller-type mixer at
1000 rpm. The rotary mixer provided an intense agitation that is necessary for high-energy
conditioning. The feed slurry was conditioned for 5 minutes in both series of tests.
As shown in Table 2.28, the moisture reduction was substantially improved when high-
energy conditioning was used. For example, the use of 5 lb/ton of dewatering aid reduced the filter
cake moisture from a baseline value of 18.2% (no reagent added) down to 14.6% when using low-
energy agitation. The cake moisture was further reduced to 11.7% moisture when high-energy
agitation was used at the same reagent dosage of 5 lb/ton. Similar results were obtained using RW
as the dewatering aid. With this reagent, the final cake moisture improved from 14.9% to 13.4% at
2 lb/t of dewatering aid and from 14.3% to 13.1% at 5 lb/t of dewatering aid. These results clearly
demonstrate the importance of proper conditioning when using the novel dewatering reagents. The
100 |
Virginia Tech | Table 2.28. Effect of mixing intensity on dewatering of Buchanan’s flotation product
Moisture Content (%)
Reagent
RU RW
Dosage
High Energy Low Energy High Energy Low Energy
(lb/t)
Mixing Mixing Mixing Mixing
0 18.2 18.2 18.2 18.2
1 15.4 17.1 13.5 15.5
2 13.2 14.9 13.4 14.9
3 11.9 14.6 13.1 14.7
5 11.7 14.6 13.1 14.3
results also indicate that the high-intensity conditioning increases the adsorption density of
dewatering reagents; as a result, lower moisture filter cake product can be obtained.
g) Pilot-Scale Test Results
Table 2.29 gives the results of pilot scale dewatering tests which were obtained using
various dewatering reagents at different addition rates. The data indicate the moisture content of
the filter cake decreased from a baseline (no reagent) value of 16.9% to 14.5% with 5 lb/ton of
RU and to 15.0% with RW. Likewise, Table 2.27 gives the laboratory test results obtained on
the Buchanan plant flotation product using RU and RW as the dewatering aids at 20 inches Hg
vacuum. As it can be seen from the table, the moisture content in the filter cake was decreased
with increasing reagent addition. At RU additions of 1 and 5 lb/t, the moisture contents of the
cake were reduced from 17.6% to 16.1% and 14.4%, respectively.
Similar results were obtained when RW was used as a dewatering aid. These values
correspond to a 15-20% moisture reduction in the filter product.
101 |
Virginia Tech | Table 2.30. Effect of pilot-scale filter disc speed on filter cake production rates and moisture
content
Filter Disc Product Moisture
Speed Rate Content
(min/rev) (lb/hr) (%)
3.0 182.6 16.9
2.0 238.9 16.7
1.0 256.5 17.3
The effect of filter disc speed was also investigated in the pilot-scale tests. Table 2.30
summarizes the results obtained by increasing the filter disc speed from 3 to 1 min/rev for
dewatering of the plant flotation product. The product rate increased from 182.6 to 256.5 lb/hr,
with a small change in the moisture content of the filter cake. The results show that it would be
possible to increase the filter capacity by 28% without adversely impacting the moisture content
of the filter cake. Tests were conducted on Buchanan plant flotation product without dewatering
reagents.
A series of pilot-scale dewatering tests were conducted to study the effects of different
Table 2.29 Effect of reagent addition on the pilot-scale dewatering of Buchanan’s flotation
product
Reagent Moisture Content (%)
Dosage
RU RW
(lb/t)
0.0 16.9
1.43 15.7
3.08 15.2
6.05 14.5
0.00 16.9
1.47 15.8
3.19 15.5
6.31 15.0
102 |
Virginia Tech | vacuum levels on moisture reduction. The initial results, which are presented in Table 2.31,
were obtained without dewatering aid addition. The data show that it is possible to reduce the
product moisture from 19.7% to 17.1% by simply increasing the vacuum level from 5 to 15
inches Hg. It seems that a further increase in vacuum is not advantageous in terms of further
lowering the moisture contents of the filter products. It should be mentioned here that the disc
filters in the Buchanan preparation plant are currently operated at vacuum levels of only 10.5-
11.0 inches Hg. Because of such low vacuum levels, the plant filter product typically contains
21-22% moisture. The present work shows that by increasing the vacuum levels from 5-11 in
Hg, the plant could probably obtain a filter product with 17-18% moisture. Besides dewatering
aid addition, vacuum level is one of the important operating conditions determining the final
product moisture in the filter cake.
Table 2.32 gives the pilot-scale test results obtained on the Buchanan plant flotation
product using RW as the dewatering aid at various vacuum pressures. As shown, the moisture
content in the filter cake was reduced with increasing vacuum pressure. At vacuum pressure
increasing from 10-20 in Hg, the moisture contents of the cake were reduced from 18.2 to 16.8%
and 16.3%, respectively.
The test results given in Table 2.33 indicate that a further improvement in filter cake
moisture (about two percentage points) was obtained when using RU as dewatering aid. As the
vacuum levels increased from 10 to 20 in Hg, the moisture content of the cake were reduced
from 16.8 % to 14.5% and 14.3%, respectively.
103 |
Virginia Tech | 2.3.6. Elkview Site
a) Site Description
The Elkview coal cleaning plant, B.C Canada, is processing 1,400 metric tons per hour
(t/hr) of run-of-the-mine (ROM) coals. The materials that are floated and dewatered are
classifying cyclone products. The cyclone overflows (O/F) are fed to five banks of mechanically
agitated flotation cells, while the underflows (U/F) are fed to sieve bends (60 mesh). The sieve
bend U/F joins cyclone O/F and are fed to the flotation cells, while the sieve bend O/F bypasses
the flotation cells. The froth product and the sieve bend O/F’s are combined and fed to vacuum
disc filters to reduce the moisture to approximately 21.5%. The filter cake is then fed to a
thermal dryer to further reduce the moisture to 8.4%. Typically, the thermal dryer is operating at
its maximum capacity, i.e., 65 t/hr of water evaporated, and cannot handle additional froth
product. Under this condition, operators cannot pull the flotation cells hard, causing a significant
loss of fine coal.
The primary objective of the project was to develop appropriate methods of reducing the
filter cake moisture to the level that can eliminate the situation where the thermal dryer is acting
as a bottleneck for increased production. These novel dewatering aids are designed to increase
hydrophobicity. As such, the dewatering aids can be added to a flotation cell displacing some, or
perhaps even all, of the conventional collector (kerosene) that is currently used. This can result
in a higher flotation recovery while at the same time improving dewatering. However, the novel
dewatering aids would work better if they were added to a separate conditioner with a strong
agitation since the energy dissipation in a flotation cell is generally less than that in a well-
designed conditioner. Therefore, the extent of moisture reduction may be less when the froth
cell is used for conditioning. Adding the dewatering reagent in place of collector for flotation
105 |
Virginia Tech | may be sufficient since relatively small moisture reduction may be sufficient in eliminating the
bottleneck at the thermal dryer and thereby allowing operators to pull the flotation cell hard and
increase the recovery. To meet this objective, a series of dewatering tests have been conducted
at Virginia Tech. The present work was limited to testing the novel dewatering aids to reduce
the moisture of the filter cakes produced from the vacuum disc filters at Elkview.
b) Reagents
The investigation, both laboratory and pilot-scale, was performed with varying amounts
of three types of dewatering aids, namely RW, RU and RV. Since these dewatering aids are
insoluble in water, they were dissolved in a solvent. In both laboratory and pilot scale
experiments diesel was used as the solvent. The ratio of reagents-to-solvent was optimized in
previous studies by varying the individual dosages (0.5 to 3 lb/ton), while maintaining the total
blend dosage constant. For this test program, the optimum combination for a given dewatering
aid and solvent is one to two (1:2) dewatering aid-to-diesel ratio. When flotation feed was used,
the sample was subjected to laboratory flotation tests using 0.66 lb/t of kerosene or RV as
collectors and 0.44 lb/t of MIBC as frother.
c) Coal Samples
Two types of samples were received from the Elkview site, i.e., a standard metallurgical
coal (Std-Met) and a medium-volatile metallurgical coal (Mid-Vol Met). In each case, both
flotation feed (minus 60 mesh, 2-3% solid by weight) and vacuum filter feed (67% froth product
and 33% sieve bend overflow, 25% solids by weight) samples were received.
d) Laboratory Procedures
Most of the laboratory filtration tests were conducted using a 2-inch diameter Buchner
vacuum filter at 20-inch vacuum pressure (68 kPa) and 2 minutes of drying cycle time. To
106 |
Virginia Tech | compare the effect of pressure drop on filtration, a few tests were also conducted using a 2-inch
diameter pressure filter at a 30 psi of compressed air. In each dewatering test, a coal sample was
conditioned in a mixing tank for 2 minutes. The cake thicknesses were varied in the range of 15
to 25 mm by varying the slurry volume.
To prepare the test samples, a series of flotation tests were conducted using a Denver
laboratory flotation cell. In each test, a known amount of a dewatering/flotation aid was added to
the flotation cell and the slurry was agitated (or conditioned) for 2 minutes before introducing air
to the slurry to initiate flotation. When the froth product was to be used for dewatering tests, the
flotation tests were conducted until exhaustion and the froth product was used for filtration tests
in the same manner as described above. In this procedure, the flotation cell was used effectively
as a conditioner. In this series of tests, the flotation products were not analyzed for ash to
determine the recovery.
e) Laboratory Test Results (Filter Feed – Standard Metallurgical Coal)
Table 2.34 gives the results obtained on the filter feed sample using RW and RV as
dewatering aids. The tests were conducted at 22-24 mm cake thickness by varying the reagent
dosages. The reagents were used as 1:2 blends with diesel, and the dosages given refer to the
Table 2.34 Effect of reagent addition on dewatering of Elkview’s filter feed sample (standard
metallurgical coal)
Reagent Moisture (%)
Dosage
RW RV
(lb/ton)
0 20.17 20.17
0.1 17.92 17.72
0.5 15.82 16.06
1 14.73 15.43
2 13.11 13.88
3 12.73 13.84
107 |
Virginia Tech | Table 2.35. Effect of using RW, RU and RV on the dewatering of Elkview’s filter feed (STD
Met Coal)
Reagent Moisture (%)
Dosage
RW RU RV
(lb/ton)
0 25.56 25.56 25.56
0.1 18.04 17.69 17.99
0.5 15.83 15.45 15.49
1 14.39 13.73 15.36
2 11.93 12.97 13.82
3 11.93 12.32 12.92
active ingredients only. As shown, the cake moistures decreased from 20.17% to 12-14% range.
Table 2.35 gives the results of the laboratory vacuum filter tests conducted on the filter
feed (STD met coal) sample using RW, RV and RU as dewatering aids. The tests were
conducted at approximately 15 mm cake thicknesses. The moisture was reduced from 25.56% to
the 12-13% range, which represents approximately 50% moisture reduction. The higher
moisture obtained at the control test was probably due to the fact that the tests were conducted a
few days after receiving the sample. The results showed that RW was most effective, followed
closely by RU and RV.
f) Laboratory Test Results (Filter Feed – Medium Volatile Metallurgical Coal)
Table 2.36 gives the laboratory test results obtained on the Mid-Vol filter feed using RW,
RV and RU as dewatering aids. The tests were conducted at approximately 15 mm cake
thickness by varying the reagent dosages. Cake moistures were reduced from 23.34% to 13-14%
range, representing approximately 43% moisture reductions.
108 |
Virginia Tech | g) Laboratory Test Results (Flotation Feed)
The two flotation feed samples (Std-Met and Mid-Vol-Met) were subjected to a series of
dewatering tests. As shown in the previous sections of this report, cake moistures can be reduced
to the 12-14% range by weight using 2 to 3 lb/ton (active ingredient) of the novel dewatering
aids. This was achieved by adding the dewatering aid to a conditioning tank so that it is readily
dispersed in the slurry. It was found in our previous work that moisture reduction improves with
increasing energy input during conditioning. Figure 2.13 shows a relationship between moisture
reduction and energy input. The results presented in this figure have been obtained on a clean
bituminous coal sample (from Moss 3 preparation plant, Virginia) that has been pulverized
before dewatering tests.
In the present work, it was decided that dewatering tests be conducted without using a
stand-alone conditioning tank. Instead, the coal samples were conditioned during flotation with
varying reagent dosages. The energy dissipation imparted by a flotation cell is substantially
16
14
12
10
8
6
4
2
0
0 10 20 30
Figure 2.13 Effect of conditioning on moisture (Middlefork coal)
110
)%(
erutsioM
Power x Time1/2 |
Virginia Tech | lower than that of a conditioner. Therefore, the results would not be as good as the case of using
a conditioner, but the moisture reduction may suffice the needs at Elkview. One concern we had
with this approach was the possibility that we would have to use a higher dose of frother when
using a higher dose of collector/dewatering aids.
Figure 2.14 shows the results obtained with the standard metallurgical coal and the Mid-
vol coal using RV. The reagent dosages given in the figure include both active ingredient and
solvent. The ratio of the active ingredient and solvent ratio was 1:2. Cake thicknesses were
approximately 20 mm and two minutes of drying cycle time was employed. In the control test
conducted with kerosene as collector, the cake formation time was one minute, which was
reduced to 25-35 seconds when using RV. The tests were conducted on the flotation products
without conditioning. The flotation tests were conducted using various amounts of collectors. In
all flotation tests, 0.33 lb/t of MIBC was used as frother. With this coal, it was not necessary to
increase frother dosage at higher collector dosages. As expected, moisture reductions improved
substantially with increasing collector dosage. The reagent dosages given are inclusive of the
solvent, which comprised 66% of the total reagent addition (the results are plotted in metric
units). At 1,500 g/t (3.3 lb/t), which was the highest dosage employed in these series of tests, the
cake moistures were 15.9% for the Mid-vol coal and 16.6% for the standard metallurgical coal.
These values are comparable to those obtained with the filter feeds. At the 1,500 g/t (3.3 lb/t)
dosage, which is equivalent to 500 g/t (or 1 lb/ton) active ingredient, the dewatering tests
conducted on the filter feeds after stand-alone conditioning gave moistures of 15.56% for the
standard metallurgical coal and 16.46% for the Mid-Vol coal. Note that the moistures of the
floated products are comparable to those of the filter feeds despite the facts that the separate
conditioning step was omitted and that particle size was finer. Recall that the filter feed was 0.6
111 |
Virginia Tech | mm x 0 while the flotation product was 0.3 mm x 0. It appears, therefore, that the conditioning
step can be omitted for the Elkview coal, which is a significant advantage.
23
22
21
20.6
20.3
20
19
18
17
16
15
100 400 700 1000 1300 1600
Reagent Dosage (g/t)
Figure 2.14 Results of the low-pressure pressure filter tests conducted on the Elkview
coal sample (filter feed).
It would be of interest to compare the results obtained with RV with those obtained with
kerosene. The froth products obtained using 700 g/t (1.4 lb/ton) of kerosene gave moistures of
20.6 and 20.3% for the standard metallurgical coal and medium volatile coals, respectively. At
present, Elkview is using 600 g/t (1.3lb/t) kerosene as collector and 60 g/t of MIBC as frother.
At 700 g/t RV and 120 g/t MIBC, we obtained 17.9 and 17.4% moistures for the standard
metallurgical (1.4lb/t) coal and Mid-vol coal, respectively. Thus, the use of RV can reduce
moistures by 2.7 to 2.9 percentage points absolute over the case of using kerosene as collector at
112
)%(
erutsioM
RV (MidVol)
RV (STD Met Coal)
Kerosene (Mid Vol)
Kerosene (STD Met Coal) |
Virginia Tech | 2.3.7. Smith Branch Site
a) Site Description
One of the most promising coal samples evaluated in this project was obtained from the
Smith Branch impoundment located near the Pinnacle Mine Complex. The complex, which is
owned by Cleveland Cliffs Mining, consists of an underground mining operation, a surface wash
plant, and the waste coal impoundment. The Pinnacle site contains approximately 100 million
tons of unmined coal reserves of which 3.3-4.0 million tons are processed annually by the wash
plant. The waste coal from the plant is diverted to the Smith Branch Impoundment, which is
believed to contain 2.85 million tons of potentially recoverable fine coal.
b) Coal Samples
A number of coal samples were used to conduct dewatering tests during the evaluation
and scale-up tests. Samples were taken from the PinnOak Company’s Pinnacle Plant site and
Smith Branch Impoundment near Pineville, WV. The samples consisted of a Vibracore
composite sample taken from the Smith Branch Impoundment, a grab sample of current
thickener underflow (taken in 2002), and a slip-stream sample of current thickener feed. All of
these samples were tested in both the laboratory and pilot scale using newly-developed
dewatering aids. Shown below is an overview of the samples used in dewatering tests.
i) Laboratory Tests
Smith Branch Vibracore Composite Sample (68% solid, 30% ash)
Plant’s Thickener Underflow Sample (12% solid, 33% ash)
Plant’s Thickener Feed Sample (1.6% solid, 46% ash)
ii) Pilot Scale Tests
Smith Branch Vibracore Composite Sample (68% solid, 30% ash)
Plant’s Thickener Feed Sample (1.6% solid, 46% ash)
114 |
Virginia Tech | Vibracoring is a technology used to extract core samples of underwater sediments and
wetland soils. The vibrating mechanism of a vibracorer, sometimes called the "vibrahead,"
operates on hydraulic, pneumatic, mechanical, or electrical power from an external source. The
attached core tube is drilled into sediment by gravitational force and boosted by vibrational
energy. When the drilling is completed, the vibracorer is turned off, and the tube is pulled out
with the aid of hoist equipment. Extracting core samples via the Vibracore sampling method
assessed the quality of fine coal contained in the impoundment. The 2.5 inch diameter cores
were extracted down to depths of 25 to 30 ft. and subjected to size and ash analyses.
Figure 2.15 shows a typical set of size-by-size analyses that were obtained from one set
of core samples. Error bars are provided to illustrate the high, average, and low values obtained
for each size class. This particular set of data shows that the minus 270 mesh fraction contains
about 60% of the coal tonnage, with an average ash content of about 33-35%. The raw quality
within the impoundment was found to vary greatly, dependent on the distance from the discharge
point into the impoundment. In general, coal extracted near the discharge point was found to be
80
70
60
50
40
30
20
10
0
Plus 28 28x100 100x270 Minus 270
Size Fraction (Mesh)
Figure 2.15 Weight and ash distributions of slurry from the Smith Branch impoundment
115
)%(
thgieW
40
35
30
25
20
15
10
5
0
Plus 28 28x100 100x270 Minus 270
Size Fraction (Mesh)
)%(
hsA |
Virginia Tech | coarser and higher in ash, while the coal extracted farther away from the discharge point was finer
and lower in ash. Also, coal fines extracted from a greater depth were found to be of better quality
than coal taken from more shallow locations. This was expected because the coal fines deposited
earlier in the life of the mine were discarded before several improvements were made to the fine
coal circuits in the existing preparation plant. The average ash content of the remaining 40% had
an ash content of less than 15%.
c) Laboratory Procedures
The laboratory flotation and dewatering equipment consists of a Denver laboratory
flotation cell, a vacuum pump, a mechanical shaker, a stand-alone mixer – in some cases – and a
Buchner Funnel with a fitted filter medium. All the vacuum filtration tests were conducted using
a 2.5 inch diameter Buchner Funnel at 20-25 inch Hg (68 kPa) with a 40X60 wire screen mesh
filter medium.
The dewatering tests were conducted to determine the best reagents and dosages for
different samples and to investigate under what conditions a given dewatering aid can give the
best performance. All the coal samples were tested within three days of receipt in order to
minimize any artificial negative effects of aging or surface property changes on dewatering.
Because the samples were taken from the waste pond, thickener feed discharge, and thickener
underflow discharge, they possessed high impurities. For this reason, a cleaning step was
employed before the filter tests. Depending on the solid content, the samples were prepared first
by either diluting or decanting to 16-17% solids and floated in a laboratory Denver flotation cell
approximately at 1000 rpm using 0.88 lb/t kerosene and 0.33 lb/t MIBC to remove ash. A
limited number of flotation feed and product samples were analyzed for ash. Then, the samples
were collected in a separate container to be used in filtration tests. The dewatering aids RW, RU,
116 |
Virginia Tech | and RV were tested and diluted to 33.3% in solvent. When using the dewatering aids,
conditioning is critical, and increased energy input in conditioning yields improved moisture
reductions. For this reason, prior to each filter test, a known volume of slurry was conditioned
with the dewatering aids in the mechanical shaker and, in some cases, the stand-alone mixer.
When the mechanical shaker was used, the conditioning time was kept at 5 minutes, and when
the stand-alone mixer was used, the conditioning time was kept at 2 minutes. Then, the
conditioned sample was poured into the Buchner funnel for filtration. During the tests, the cake
thickness was determined by the amount of the slurry sample used. Some of the conditions were
kept constant during the tests. For example, a 7-10 mm cake thickness was determined for 2
minutes of drying time.
d) Pilot-Scale Procedures
The continuous dewatering test rig used at the Smith Branch site incorporated three main
components, a column module, conditioner module, and a disc filter module, along with some
other ancillary components. The details about the modules were described previously. For the
pilot-scale tests, feed slurry was supplied from either the existing preparation pond reclaim
facilities or in barrels and directed into the circuit feed sump. During the tests, the slurry was fed
at a constant rate into a 12 inch diameter column by means of a peristaltic pump and flexible
piping. The flotation column was used to produce a filter feed and reject ultra-fine hydrophilic
clay that negatively impacts the effectiveness of some of the dewatering aids. Then, the clean
coal froth was routed to the multi-stage conditioner module by gravity while the column reject
slurry flowed, also by gravity, to a refuse sump. After adding appropriate dewatering reagents,
the conditioned slurry was pumped at a constant rate into the filter test module. Next, the
conditioned slurry was dewatered in the filter module. The filter cake was discharged into a
117 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.