University
stringclasses 19
values | Text
stringlengths 458
20.7k
|
|---|---|
Virginia Tech | container and immediately transferred to an oven for moisture analysis while the filter effluent
was pumped to the reject sump. In addition, if any excess feed problems occurred, the column
feed sump and conditioner tanks were allowed to overflow in a controlled manner.
e) Laboratory Results (Vibracore Samples)
The objective of the test series was to conduct preliminary investigations that evaluate the
performance of dewatering aids. The laboratory tests were conducted on the Pinnacle Smith
Branch Vibracore Composite Sample. Prior to the dewatering tests, the sample was cleaned
using a Laboratory Denver Flotation Cell using 0.88 lb/t kerosene and 0.33 lb/t MIBC at 16.7 %
solid content. Of the samples, the vibracore sample was very important because the data from
the laboratory studies was also used to provide practical justification for the pilot scale test
program which would soon to be carried out using Peterson Disc Filter.
A series of initial batch laboratory filter dewatering tests was conducted to determine
whether the floated vibracore coal would respond well to the addition of the novel dewatering
aids. A 62.5 mm diameter Buchner filter was used for these tests. The solids content of the
sample was 16.0%. and a vacuum setup point of 20 inches Hg was utilized. The tests typically
provided a cake thickness of 7-8 mm without reagent and 8-10 mm with reagent. A conditioning
time of 5 minutes and a drying cycle time of 2 minutes was used. The volume of the slurry was
100 ml. The impoundment sample was floated at 16.7% solids using 0.88 lb/t kerosene and 0.33
lb/t MIBC before the filter tests. Table 2.38 gives the laboratory test results when RW was used
as the dewatering aid. As shown, the moisture content in the filter cake was reduced when
reagent addition was increased. At 1 and 5 lb/t RW additions, the moisture contents of the cake
were reduced from 28.7% to 24.6% and 14.3%, respectively. The lower value corresponds to a
50% moisture reduction in the filter product.
118 |
Virginia Tech | Table 2.38 Effect of RW addition on the Pinnacle pond sample
RW Moisture Moisture
Dosage Content Reduction
(lb/t) (%) (%)
0 28.7 --
1 24.6 14.28
2 22.9 20.20
3 22.5 21.60
5 14.4 49.82
*62.5 mm diameter Buchner filter was used. The solid content of the sample was 16.0%. Vacuum
setup point was 68 kPa (20 inches Hg). It was not changed during the experiment. Cake thickness;
base: 7-8 mm, with reagent: 8-10 mm. Conditioning time: 5 minutes; drying cycle time: 2 minutes.
The volume of the slurry was 100 ml. The impoundment sample was floated at 16.7% solids using
400 g/t kerosene and 150 g/t MIBC before the filter tests.
f) Laboratory Results (Thickener Underflow Samples)
At the preparation plant after cleaning the coal, the residue was sent to a thickener and
then pumped to the slurry impoundment. Because it was an active impoundment, the feed
sample, thickener underflow, was also investigated, and similar tests were conducted on this
sample using RW and RU. Results for the thickener underflow sample are given in Table 2.39,
and the moisture content of the filter product again decreases with increasing RW and RU
additions from 1 to 5 lb/t. Here, the RW addition rate of 5 lb/t reduced cake moisture contents
from 31.4% to 22.1%, giving a percentage moisture reduction of about 30%. This is much less
than what was observed in the case of the impoundment sample and can attributed to the
different mean particle sizes of these two samples. Mean particle size, as determined by
Microtrac particle size analyzer in Beard Technologies’ laboratory, was approximately 32
micron for the impoundment sample and 15 micron for the thickener underflow sample. The
reasons for the relatively poor behavior of RU are unclear but may be related to the presence of
flocculant in the sample.
119 |
Virginia Tech | Table 2.39 Effect of reagent addition on dewatering of Pinnacle thickener underflow sample
Reagent Moisture Content (%) Moisture Reduction (%)
Dosage
RW RU RW RU
(lb/t)
0 31.4 31.4 -- --
1 30.3 28.8 3.5 8.28
2 28.0 28.2 10.82 10.19
3 22.7 -- 27.71 --
5 22.1 27.8 29.61 11.5
*A 62.5 mm diameter Buchner funnel was used. The solid content of the sample was 16.0%.
Vacuum setup point was 68 kPa (20 inches Hg). It changed to 61-68 kPa during the experiment.
Cake thickness: base: 7.0 mm, with reagent: 8-8.5 mm. Conditioning time: 5 minutes; drying cycle
time: 2 minutes. The volume of the slurry was 100 ml. The underflow sample was first floated
using 400 g/t kerosene and 150 g/t MIBC before the filter tests.
Another set of tests was conducted on a deslimed thickener underflow sample to
investigate the effect of particle size distribution. Before the dewatering tests 50% of minus 45
micron material was removed by wet screening. Table 2.40 shows the effect of the desliming on
dewatering of this sample. The final cake moisture was reduced down to 22.0% moisture 5 lb/t
dosage of RW. When RU was used as dewatering aid, the moisture could be reduced down to
21.4% moisture at the same dosage of 5 lb/t.
120 |
Virginia Tech | Table 2.40 Effect of desliming on dewatering of Pinnacle thickener underflow sample
Reagent Moisture Content (%) Moisture Reduction (%)
Dosage
RW RU RW RU
(lb/t)
0 29.38 29.38 -- --
1 23.98 25.28 18.38 13.96
2 23.39 23.17 20.38 21.13
3 22.77 22.16 22.49 24.58
5 22.00 21.37 25.11 27.26
*A 62.5 mm diameter Buchner funnel was used. The solid content of the sample was 16.0%.
Vacuum setup point was 68 kPa (20 inches Hg). It changed to 61-68 kPa during the experiment.
Cake thickness: base: 7.0 mm, with reagent: 8-8.5 mm. Conditioning time: 5 minutes; drying cycle
time: 2 minutes. The volume of the slurry was 100 ml. The underflow sample was first floated
using 400 g/t kerosene and 150 g/t MIBC before the filter tests.
Table 2.41 Effect of reagent addition on dewatering of Pinnacle thickener feed sample
Reagent Moisture Content (%) Moisture Reduction (%)
Dosage
RW RU RW
(lb/t)
0 29.5 29.5 -- --
1 26.5 27.3 10.17 7.46
2 24.3 25.2 17.63 14.58
3 22.2 23.1 24.75 21.70
5 21.1 21.8 28.48 26.10
*62.5 cm diameter Buchner filter was used. The solid content of the sample was 11.0%. Vacuum
setup point was 68 kPa (20 inches Hg). It changed to 61-68 kPa during the experiment. Cake
thickness; base: 7.0 mm, with reagent: 8-8.5 mm. Conditioning time: 5 minutes; drying cycle time:
2 minutes. The volume of the slurry was 100 ml. The thickener feed sample was floated using 400
g/t kerosene and 150 g/t MIBC before the filter tests.
g) Laboratory Results (Thickener Feed Samples)
Table 2.41 gives the results for the thickener feed sample. In this case, RW and RU
reduced the cake moistures from 29.5% to 21.1% and 21.8%, respectively, and percentage
moisture reductions of about 28% and 26% at addition rates of 5 lb/t. Again, these reductions
are much less than those that occurred with the impoundment sample. Due to the preferential
121 |
Virginia Tech | removal of fine clay during flotation, the mean size of the test sample was probably also around
15 microns.
h) Pilot-Scale Results (Vibracore Samples)
RW was used as the dewatering aid in these sets of pilot scale dewatering tests. Table
2.42 gives the pilot scale dewatering results obtained on the Smith Branch Impoundment sample.
These results show that the moisture contents of the filter cakes are substantially decreased in the
presence of dewatering aid. For example, at reagent dosage rates of 2-5 lb/t of RW, filter cake
moisture content was reduced from 28.4% to 17.7%-16.3%. Cake thicknesses were as high as 16
mm when using RW as the dewatering aid versus 3-6 mm without a reagent. Even at this cake
thickness, moisture reductions were significant. Additionally, filter effluents were much cleaner
when dewatering aids were used, indicating that filter recoveries increased substantially in the
presence of the dewatering aid.
Table 2.42 Effect of RW addition on the pilot scale dewatering of the Pinnacle-Smith
Branch Impoundment sample using the mobile units
Reagent RW Moisture Moisture
Dosage Content Reduction
(lb/t) (%) (%)
0 28.4 --
1 19.6 30.99
2 17.7 37.68
3 17.2 39.43
5 16.3 42.61
* 10 sector single disc Peterson filter was used in the experiments. The solid content of the filter
feed was 17.0%. The vacuum setup point was 81.27-84.66 kPa (24-25 inches Hg). It dropped to
67.73-77.89 kPa (20-23 inches Hg) during the experiment. Cake thickness; base: 6-8 mm, with
reagent: 12-16 mm..
i) Pilot-Scale Results (Thickener Feed Samples)
122 |
Virginia Tech | Table 2.43 gives the pilot scale dewatering results obtained on the Pinnacle coal thickener
feed sample tests. In this series of tests, RW was used as the dewatering aid. As shown below,
the moisture content of the filter cake is reduced from 29.4% to 21.5% by the addition of 3 lb/t of
RW. Likewise, the test results given in Table 2.44 show that the moisture content of the filter
cakes are significantly decreased when RW and RU are used as dewatering aids. For example,
Table 2.43 Effect of RW addition on the pilot scale dewatering of the Pinnacle thickener
feed sample using the mobile test units
Reagent RW Moisture Moisture
Dosage Content Reduction
(lb/t) (%) (%)
0 29.4 --
1 -- --
2 -- --
3 21.5 27.2
5 -- --
* 10 sector single disc Peterson filter was used in the experiments. The solid content of the filter
feed was 12.3%. Vacuum setup point was 81.27-84.66 kPa (24-25 inches Hg). It dropped to
71.11-74.50 kPa (21-22 inches Hg) during the experiment. Cake thickness; base: 6-8 mm, with
reagent: 12-16 mm.
Table 2.44 Effect of reagent addition on dewatering of Pinnacle thickener feed sample
Reagent Moisture Moisture
Reagent
Dosage Content Reduction
Type
(lb/t) (%) (%)
0 28.0 --
0.95 23.4 16.43
RW
3.7 20.3 27.5
4.45 21.8 22.14
0 28.0 --
RU 3.1 20.6 26.43
5.5 21.0 25
* 10 sector single disc Peterson filter was used in the experiments. The solid content of the filter
feed was 8-9%. Vacuum setup point was 71.11-88.05 kPa (21-26 inches Hg). It dropped to 47.41-
77.89 kPa (14-23 inches Hg) during the experiment. Cake thickness; base: 5-6 mm, with reagent:
10-20 mm.
123 |
Virginia Tech | 2.3.8. Moatize Site
a) Site Specific Information
A new reserve is being developed by Companhia Vale do Rio Doce (CVRD). The coal
sample was from the Section 2A of the reserve. CVRD is in the process of designing a coal
preparation plant with a throughput capacity of 4,000 metric tonnes per hour (t/h). The coal
sample received was a fine coal (0.25 mm x 0), which will be fed to a single-stage flotation
circuit. Columns are chosen over mechanically agitated cells, because the former can produce
cleaner froth products. The throughput capacity of the flotation circuit will be around 440 to 560
t/h.
The CVRD is looking for the possibility of reducing the moisture of the column flotation
product to less than 15% by weight using horizontal belt filters (HBF). It has been a challenge to
achieve such a low level of moisture with a by-zero froth product using a vacuum filter. It is
possible, however, to achieve the objective, if the HBF is used in conjunction with the novel
dewatering aids developed at Virginia Tech. For this reason a series of laboratory dewatering
tests have been conducted on a coal sample from the Moatize region in Mozambique.
b) Experimental Design
Flotation and Dewatering Aids
During this investigation, dewatering tests were performed with varying amounts of two
types of dewatering aids, RU, and RV. Since these dewatering aids are insoluble in water, they
were dissolved in diesel which was used as a solvent. The ratio of reagents-to-solvent was
optimized in previous studies by varying the individual dosages (0.5 to 3lb/t), 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. When flotation tests
125 |
Virginia Tech | were conducted, RV, Nalco-9855 and kerosene were used as collectors while Nalco-9840 was
used as frother.
Samples and Procedures
Coal Samples: The coal sample was received as wet cake in a bucket, which had been
obtained by decantation. Prior to each series of dewatering tests, a portion of the wet cake was
removed from the bucket, mixed with a volume of tap water, and agitated for a minimum of 3
hours. After the agitation, part of the slurry was transferred to a Denver D-12 laboratory flotation
cell and subjected to flotation. The froth product was then used as a feed to a series of laboratory
vacuum filtration tests.
Procedure: The froth product from a single-stage laboratory flotation tests was placed in
container, and agitated to obtain a homogenous suspension of clean coal. A volume of the
suspension was removed by means of a cup of known volume. The suspension was then
subjected to vacuum filtration tests using a 2-inch diameter Buchner filter using a fabric filter
medium. The thickness of the filter cake formed on the filter cloth was varied by controlling the
volume of the suspension filtered. When the slurry volume was varied from 100, 150 to 200 ml,
the cake thicknesses varied from approximately 10, 15, and 20-22 mm. When the slurry volume
was varied from 100, 150 to 200 ml, the cake thicknesses varied from approximately 10, 15, and
20-22 mm. At a given filtration experiment, the initial vacuum pressure was set at 20-inch Hg,
which was reduced to 15- to 17-inch Hg at the end of the 2 minutes of filtration time employed
in the present work.
Initially, the flotation tests were conducted at 5 to 6%. In later experiments, the solid
concentration was increased to 10% to obtain froth products with a higher solid content. In some
126 |
Virginia Tech | experiments, the feeds to filtration experiments were conditioned in a standalone conditioner. In
others, the flotation cell was used as a conditioner.
c) Results and Discussions
Using Flotation Cell as Conditioner
The first set of dewatering tests was conducted by adding the dewatering aids directly
into the flotation cell without using a stand-alone conditioning tank. In this procedure, the
flotation cell was used effectively as a conditioner. This was possible because the dewatering
aids being tested in the project can increase the hydrophobicity of coal, as noted previously. The
froth products obtained from the flotation tests were then subjected to dewatering tests using the
Buchner filter. Each test was conducted at 15 mm cake thickness and 2 minutes of filtration
time, including cake formation and drying cycle times.
For comparison, a series of dewatering tests was also conducted using kerosene as
collector. The use of Nalco 9855 and RV gave substantially lower cake moistures than the diesel
at low reagent dosages. At 0.88 lb/t, diesel produced a cake with 22.8% moisture, while using
RV resulted in 18.1% moisture reduction.
Using a Standalone Conditioner
In the second set of tests, flotation tests were conducted using 0.88 lb/t of kerosene as
collector. The froth product (approximately 10% solids) was conditioned in the presence of a
dewatering aid for 2 minutes in a small, rectangular-shaped conditioner prior to dewatering test.
The results obtained using varying amounts of two different dewatering aids (RV and RU) are
given in Table 2.45.
127 |
Virginia Tech | Table 2.45 Effect of reagent addition on the dewatering of the froth products obtained
using 400 g/t diesel as collector
Reagent Dosage Cake Moisture (%)
(lb/t) RV RU
0 22.30 22.30
0.25 17.35 17.50
0.5 16.28 16.03
2 15.93 15.91
3 14.78 15.00
The control tests conducted without any dewatering aid gave 22.30 % moisture. At
increased reagent dosages, the cake moisture was reduced to the 14-15% range, which
represented approximately 33% reduction in moisture. The two different dewatering aids tested
gave no significant differences. All tests were conducted at 15 mm cake thicknesses.
Effect of Conditioning Time
The 2 minutes of conditioning time employed in the test work as described in the
foregoing sections of this test work were based on the previous experiences with other coal
samples. In order to determine more specific conditioning times needed for the Moatize coal, a
series of dewatering tests were conducted by varying the conditioning time. At each dosage
level, 15, 30 and 60 seconds of conditioning times were employed. It was found that at the lower
reagent dosage, a short conditioning time was sufficient. At 3lb/t, however, cake moistures
varied in the range of 16.19, 15.30, and 14.84% at 15, 30 and 60 seconds of conditioning times,
respectively.
Another set of dewatering tests were conducted at 2 lb/t RV In this series, the
conditioning times were varied from 15 to 120 seconds. The results are given in Table 2. As
shown, the conditioning time makes no significant difference beyond 15 seconds. This finding
128 |
Virginia Tech | suggests that the conditioner can be much smaller than originally anticipated. The data
represented in Table 2.46 suggest that at reagent dosages of up to 2 lb/t, only a short conditioning
time was necessary. At higher reagent dosages, longer conditioning times were helpful.
Table 2.46 Effect of conditioning time on the dewatering of flotation product using 2 lb/t
RV
Conditioning Cake
Time Moisture
(seconds) (%)
0 22.83
15 16.55
30 16.21
45 16.10
60 16.26
90 16.05
120 16.08
Effect of Cake Thickness
Another important parameter affecting filtration is the cake thickness. Therefore, a series
of dewatering tests were conducted by varying cake thicknesses in the range of 10 to 22 mm in
the presence of 3 lb/t RV. In each test, a 1-minute conditioning time and 2 minute filtration time
were employed. The results showed that that cake thickness is critical in controlling cake
moistures. As shown in Table 2.47, the targeted 15% moisture can be readily obtained by
decreasing cake thickness.
Table 2.47 Effect of cake thickness on cake moisture in the presence of 3 lb/t RV
Cake Thickness (mm) Cake Moisture (% wt)
20-22 19.60
15 14.83
10 12.33
a)
129 |
Virginia Tech | 2.4. SUMMARY AND CONCLUSIONS
Laboratory- and pilot-scale dewatering tests were conducted on various fine coal samples
received from several coal preparation plants. The objectives of the dewatering tests were to
maximize the effects of novel dewatering aids in terms of final cake moisture and the dewatering
kinetics in order to evaluate possible industrial applications. The evaluation tests were
conducted using various types and dosages of novel dewatering aids while varying operating
parameters (e.g., cake formation time, dry cycle time, vacuum and air pressure and cake
thickness.) Laboratory tests were conducted using a Buchner Funnel and air pressure filters
while the pilot-scale tests were conducted using vacuum disc and horizontal belt filters. During
pilot-scale tests, a two-step conditioner was also used to ensure effective mixing of the aids. In
some of the dewatering tests, a column flotation unit was utilized to produce fresh coal.
Currently, fine particle coal reports to the Mingo Logan Preparation Plant’s discard fine
stream due to the inability of centrifuges to capture this material. A series of laboratory- and
pilot-scale dewatering tests were also conducted to study the feasibility of using the novel
dewatering aids in conjunction with horizontal belt filters to recover this fine coal. The tests
varied types and dosages of dewatering aids added to feeds comprised of different mixtures of
relatively coarse and fine coal particles. Dewatering test results showed that a 20% moisture
reduction along with almost 50% increase in kinetics could be achieved on fine coal feeds
(nearly 43% particles of this product was passing 325 mesh (-44 µm) and contained 81% ash).
When fine coal samples were subjected to dewatering tests after re-cleaning to reduce the
amount of hydrophilic ash forming minerals, the moisture reduction was further increased to
35%. Similar results were obtained when tests were conducted on relatively coarser samples
130 |
Virginia Tech | comprised of 50% fine and 50% spiral product: the moisture could be reduced by 20-30% while
increasing the kinetics by 30-50%.
On-site pilot-scale tests using HBF showed that, when the operating parameters are
optimized, it is possible to lower the moisture of fine coal samples by 18% and coarser samples
by 30%. Results also indicated that adequate mixing time and intensity is very critical and
should be optimized. Using flocculants along with the novel dewatering aids additionally
increased the effectiveness by increasing kinetics. If the dewatering aids are used along with the
required equipment (i.e., column and HBF), the additional annual net revenue was calculated to
be $1.57 million with the payback time of 3.1 years.
Several pilot-scale dewatering tests were also conducted on Coal Clean Preparation Plant
feeds (100% -325 mesh) to investigate the possible application of novel dewatering aids. The
fine coal was treated by using a column flotation unit and fed to a pilot-scale disc filter. The
results showed that under optimized conditions, it is possible to lower the moisture by almost
20% while increasing the dewatering kinetics. Based on the results that were obtained on coal
from this site, it appears possible to increase the fine circuit’s recovery while lowering the
moisture.
Coal samples obtained from Concord Preparation plant were subjected to flotation and
dewatering tests to investigate the recovery of the fine coal at acceptable moisture levels.
Laboratory test results showed that the moisture could be reduced by 26%, and to verify these
results, a number of on-site pilot-scale tests were also conducted. Pilot-scale results confirmed
that dewatering aids are capable of lowering the final cake moisture by 21%. Additionally, a
number of tests were conducted using the dewatering aids as collectors, wherein the dosage was
only 50% of a normal diesel dosage. Under these conditions, it was observed that recovery was
131 |
Virginia Tech | increased and the moisture was lowered by 7%. When the same dosage as diesel was utilized,
the moisture was further reduced, giving a 14% overall reduction. The dewatering kinetics were
also increased.
Several dewatering tests were conducted on Buchanan filter feed samples. The final cake
moisture was found to be 18.1% for the control tests and 14.9% in the presence of a novel
dewatering aid. Similar tests were conducted with the pilot-scale unit and comparable moisture
reductions were obtained. However, considering the relatively coarser size of the coal, the
results were not good as expected. This may be attributed to water chemistry (i.e., excessive
amounts of Ca2+ ions present in Buchanan plant water.) Additionally, a series of tests were
conducted to investigate the effect of mixing intensity and time. It was found that supplying
proper mixing intensity and time may lower the final cake moisture substantially. One study
showed that total moisture reduction was around 36% under proper mixing conditions as
compared to 19% when the mixing was not sufficient.
A number of coal samples were received from Elkview Coal Preparation Plant, where
insufficient moisture reduction causes a bottle neck in production. Laboratory dewatering test
results indicated that the final cake moisture may be lowered by 35-50% depending on the coal
feed type. Using dewatering aids as collectors at a 700 g/t dosage, the recovery was the same or
better as compared to diesel. Without any additional dewatering aids, a moisture reduction of
about 13% was observed.
An investigation was also conducted to determine if the novel dewatering aids could be
utilized in industrial applications for fine coal beneficiation at the Smith Branch Reclamation
Site. To this end, a series of laboratory- and pilot-scale dewatering tests were conducted on
samples from this site. The results showed that the moisture content of the filter cake was
132 |
Virginia Tech | reduced as reagent addition was increased. At 1 and 5 lb/t RW additions, the moisture contents
of the cake were reduced from an initial 28.7% to 24.6%, and to 14.3%, respectively. The latter
value corresponds to a 50% moisture reduction overall in the filter product. A number of
laboratory tests were also conducted on thickener feed and thickener underflow samples, wherein
the moisture was reduced by 25-27%. Pilot-scale tests produced similar results.
In a coal preparation plant that will be built in Moatize, Mozambique, Companhia Vale
Rio Doce seeks to lower the moisture of it’s column flotation product to less than 15% (by
weight) using horizontal belt filters (HBF). Achieving this low moisture level on a by-zero froth
product is very challenging using a vacuum filter. It is possible, however, to achieve this level if
the HBF is used in conjunction with the novel dewatering aids. This was proven during a series
of laboratory dewatering tests on a coal sample from the Moatize region. Results indicated that
considerably low moistures are possible when using the novel aids as only dewatering agents.
Furthermore, when using the aids as collector, it is also possible to obtain low moisture contents.
Additionally, the Moatize coal can be treated using relatively shorter chemical mixing times.
Overall, test results clearly showed that when operating parameters are optimized, use of
novel dewatering aids can generate substantial moisture reductions that cannot be achieved by
mechanical means. Additionally, the dewatering kinetics can be increased, which in turn may
greatly increase the throughput of filter operations. The dewatering aids can also be used as
collectors where mixing is not possible. They not only produce similar or better results than
conventional collectors, but also lowering moisture contents.
133 |
Virginia Tech | CHAPTER 3 SIMULATION AND SCALE-UP
3.1. INTRODUCTION
There are certain difficulties in coal processing, especially with fine coal cleaning
and dewatering. When coal is cleaned by flotation – the most dominant fine coal
cleaning method – the product is in slurry form and has to be dewatered to acceptable
moisture levels before sale.[1-3] As stated in the previous chapter, the final cake
moisture and the production rate of a filter unit are important factors which must be
optimized for an effective operation.[2] The Mingo Logan Preparation Plant is a
particular site which has suffered from inefficiencies in flotation and especially
dewatering. The plant has been losing valuable coal because its current operation design
does not allow recovering the entire fine coal fraction from centrifuges. Typically,
centrifuges are very effective at coarse sizes, but the fine coal recovery is relatively low;
and when the fine size fraction is increased, the moisture is increased to levels that are
unacceptable.
Taking these facts into account, a detailed dewatering study was undertaken to
investigate the feasibility of a vacuum filtration operation, a vacuum disc filter or a
horizontal belt filter (HBF). Based on the promising results obtained using novel
dewatering aids, it was proposed that vacuum filtration may be used to decrease the loss
of fines fraction and effectively lower the moisture.
In addition to the feasibility study, a number of parallel dewatering tests were
conducted to develop empirical models and simulations for a vacuum disc filter or HBF.
The results from these dewatering tests, including the plots, tables and the general
135 |
Virginia Tech | information were used to create the empirical model by correlating and combining the
filtration parameters. Development of an effective computer simulation will be valuable
in handling the complexities involved in data analysis and general performance
evaluation.
Data from the dewatering tests require moderately straightforward analysis in
order to make them useful for the simulation development. As a starting point, some
qualitative screening tests were carried out to determine what effects certain reagents and
other fundamental dewatering parameters have on final cake moisture. After establishing
the key parameters and their influences on cake moisture, more specific tests were
undertaken to evaluate overall performance.
It should be noted that there is not a standard procedure for dewatering modeling
and simulation. However, it is essential to perform the analysis to assess the variations in
cake thickness, cake weight, vacuum pressure, filtration time, concentration in the slurry,
particle size and distribution, filtration rate, etc.; experience has shown that these
parameters have a marked effect on the dewatering performance and profound influence
in designing the simulation.
3.2. EXPERIMENTAL
The filtration tests were carried out with a simple apparatus to determine the
effects of the parameters on final cake moisture. For these tests, a Buchner Funnel
equipped with a filter cloth is used. The general procedure to obtain the dewatering data
is as follows:
136 |
Virginia Tech | 1. the solid concentration in the slurry is determined,
2. a well mixed-known amount of slurry sample is poured into the
funnel containing a filter cloth,
3. the vacuum pressure is adjusted to a pre-decided level and the
vacuum valve was opened,
4. the filtration characteristic were observed and recorded (i.e., pressure
drop, cake formation time, dry cycle time, cake weight and
thickness),
5. chemical pre-treatment was done using dewatering aids,
6. the final cake moisture is determined in each test.
Dewatering aids are used to enhance the solid-liquid separation efficiency by
aggregation and hydrophobization mechanisms. The role of novel dewatering aids is
mainly to lower the surface tension and increase the coal hydrophobicity and thus
enhance the ease of the water removal.
3.3. DATA COLLECTION AND RESULTS
The dewatering process is controlled by several factors, including reagent type
and dosage, particle size, filtration time, cake thickness and weight, and vacuum pressure
during filtration. These parameters are required for preliminary filtration calculations and
empirical modeling. The effects of these parameters were investigated and a limited
number of test results are represented as examples below.
The tests were conducted on two samples obtained from the Mingo Logan
Preparation Plant. One sample was a ‘mixture’ and the other was a flotation product.
The mixture sample was a blend of flotation product and spiral product at 50:50 blend
137 |
Virginia Tech | ratios. The second sample was 100% flotation product. The -325 size fraction was
determined to be about 30-35% for the mixture sample and 40-45% for the flotation
product sample.
The first series of evaluation tests were conducted to determine the effects of the
filter cloth and particle size on dewatering performance. As expected, the kinetics and
moisture values were more acceptable with the coarser material (i.e., the mixture sample)
compared to flotation product. Results also showed that among the filter cloths that were
tested, 60X40 mesh screen produced the best results. After establishing the optimum
filter cloth type and how the two samples responded to filtration, another set of tests was
conducted to investigate the effects of reagent type and dosage, filtration time, vacuum
pressure cake weight and thickness. The following tables represent the general trend of
the effects of these parameters on the mixture sample. Identical tests were conducted for
the flotation product sample as well, but only the mixture sample results are given as
examples.
Example 1: Effect of Reagent Type and Dosage
A series of tests were conducted to investigate the effect of reagent type and
dosage. The results showed that RU and RV performances were encouraging at 1 to
3lb/t. To increase the filtration rate Nalco 9806 was also used. Table 3.1 shows the
effect of reagent type and dosage.
138 |
Virginia Tech | Table 3.1 Effect of reagent type and dosage on dewatering of mixture sample
Reagent Dosage Moisture Content (%wt)
(lb/ton) RW RU RV
0 19.00 19.00 19.00
1 15.66 14.4 14.86
3 15.7 13.81 14.91
5 15.12 13.70 14.35
Reagent conditioning time: 5 minutes, Set up Vacuum: 68 kPa (20 inches
Hg), Actual Vacuum: 16-18.5 inches Hg, Drying cycle time: 2 minutes,
Cake thickness: 9-12mm, Solid Content: 25.0%, Dewatering Reagents: RW,
RU, RV in Diesel (1:2)
Example 2: Effect of Filtration Time (CFT + DCT)
After establishing the optimum reagent type and dosage, a series of dewatering
tests were performed to investigate the effects of the dry cycle time. An example of
simplified and summarized series of tests is shown in Table 3.2. Tests were conducted at
various intervals in the presence and absence of the dewatering reagents. As seen from
Table 3.2, the dry cycle time has an important effect on final cake moisture. It is also an
important parameter when designing the HBF’s. The dry cycle time is directly related to
the belt speed (i.e., the longer the dry cycle time is, the slower the belt speed or vice
versa.) Generally, in real applications, if the feed to the filter is constant, slower belt
speed allows a thicker cake formation but higher cake moistures. Higher belt speed
results in greater solids production rates by forming thinner cake and lower cake
moisture.
During the tests vacuum set-up pressure was adjusted to 20 in Hg and 100 ml of
slurry was used to obtain approximately 1/2in filter cake. It was observed that there is a
good correlation between drying time and the cake moisture. It was also found that the in
139 |
Virginia Tech | Table 3.2 Effect of filtration time on dewatering of mixture sample (0.5 in. cake)
Reagent Filtration Time (sec) Moisture
Moisture
Dosage Reduction
CFT DCT (%)
lb/t (%)
59 2 30.32 -
56 30 22.62 25.40
Baseline
51 60 20.55 32.22
(No Reagent)
66 120 19.67 35.13
77 300 18.88 37.73
40 2 24.31 -
37 30 19.77 18.68
RU
45 60 18.85 22.46
(3 lbs/t)
48 120 16.90 30.48
47 300 14.84 38.96
29 2 31.91 -
31 30 21.72 31.93
RV
25 60 17.76 44.34
(3 lbs/t)
25 120 18.29 42.68
27 300 15.53 51.33
* slurry tested: 100 ml, 34% solid; conditioning time: 5 min.; filter media: wire mesh screen
the presence of RU and RV the formation times were much shorter compared to baseline
tests. In the presence of the dewatering aids, the moisture content is also lowered by 3 -
5%.
Example 3: Effect of Vacuum Pressure
During the filtration, the vacuum pressure is unquestionably related to the
moisture content of the final cake. The higher the pressure, the lower the final cake
moisture content and the better the cake consolidation. Table 3.3 shows the effects of the
vacuum pressure on final cake moisture at various dry cycle times. The results show that
the cake moisture is very sensitive to pressure changes. For instance, the cake moisture
drops from 23.29% to 20.55% when the vacuum pressure is increased from 15 inHg to 20
140 |
Virginia Tech | inHg at 60 sec dry cycle time. The same trend can be observed at other dry cycle times
as well.
Table 3.3 Effect of vacuum pressure on dewatering of mixture sample (0.5 in cake)
Reagent Filtration Time (sec)
Cake Thickness Vacuum Moisture
Dosage
(mm) inHG (%)
CFT DCT
(lb/t)
95 2 30.37
75 15 27.43
Control (no
87 30 11-14 15 inHG 25.67
Chemical)
83 60 23.20
72 120 21.11
59 2 30.32
Baseline 54 15 25.49
(No 56 30 13~14 20 22.62
Reagent) 51 60 20.55
66 120 19.67
Example 4: Effect of cake weight and thickness
The effect of cake weight and thickness on final cake moisture at various drying
times is represented in Table 3.4. In this set of tests, the cake thickness was controlled by
changing the slurry volume. RU was used as dewatering aid at 3lb/t. As seen under the
same operating conditions, increasing the cake thickness increased the final cake
moisture. A cake thickness of 5-7 mm resulted in approximately 14% moisture, and a
thickness of 12-15 mm resulted in 17.3% moisture. When the thickness was further
increased, the moisture was increased to 21.27%.
141 |
Virginia Tech | Table 3.4 Effect of cake weight and thickness on dewatering of mixture sample
Slurry Filtration Time (sec) Cake
Moisture
Volume Thickness/Weight
(%)
CFT DCT
(ml) (mm/gr)
2 15 16.78
50 2 30 (5~7)/(15) 15.02
3 60 14.04
18 15 18.90
100 18 30 (12~15)/(30) 19.39
19 60 17.32
61 15 23.63
150 69 30 (21~24)/(50) 22.29
68 60 20.01
91 15 26.59
200 83 30 (21~24)/(65) 24.16
88 60 21.27
*slurry tested: 30% solid; conditioning time: 5 min.; filter media: white filter cloth
After establishing the general effects of operating parameters, a series of tests
were also conducted to be used for the modeling and simulation. Dewatering tests were
conducted on the same two samples that were previously used for parameter evaluation
tests. Approximately 60 dewatering tests were conducted varying the slurry volume and
Dry Cycle Time (DCT). All parameters such as cake formation time (CFT), pressure
during cake formation (P ) and pressure during drying cycle (P ) were recorded. The
F D
specific cake weight (W ), normalization factor (NF) and the percent moisture were
S
calculated from the data. In this modeling study, instead of cake thickness, the specific
cake weight is used; this is because, in the presence of the dewatering aids which increase
porosity, the cake thickness was artificially increased and created unreliable results. It
was found that the cake weight gave more consistent and reliable results. Table 3.5
represents an example of the data collection sheet.
142 |
Virginia Tech | greater submersion time allows for greater accumulation on the filter surface (i.e., greater
cake weight.) Likewise, for a given submersion time, increased pressured will also
increase the accumulation of coal on the filter medium. To determine the relationship
between time, pressure and cake weight, various amounts of coal slurry (30-35 % solids)
were filtered at a pre-adjusted pressure. The cake formation times and pressure changes
were recorded. As shown in Figure 3.1 and Figure 3.2, the specific cake weight (W ) is
S
plotted as a function of CFT x P for mixture and flotation product samples. When these
F
parameters were plotted for the baseline condition, for RV dosage of 3lb/t, and for RU
dosage of 3lb/t, power equations are obtained. Using these power equations, it is then
possible to predict the amount of cake pickup or amount of cake generation during
submersion as a function of CFT and P . From Figure 3.1, for example, W for the
F S
mixture sample in the presence of RV can be calculated according to the model if P and
F
CFT are known:
(3)
(cid:24).(cid:25)(cid:26)(cid:27)(cid:25)
(cid:2)(cid:3) = 1.8961×(cid:5)(cid:6)(cid:7)(cid:8)×(cid:10)(cid:11)(cid:12)
Using the above and the other obtained power equations, W can be solved for
S
both mixture and flotation samples in the absence and the presence of the dewatering
aids.
144 |
Virginia Tech | 40
Baseline
y = -3.731ln(x) + 38.833
R² = 0.9378
RU (3 lb/t)
y = -4.128ln(x) + 35.42
R² = 0.9776
30
RV (3 lb/t)
y = -5.311ln(x) + 36.657
R² = 0.9585
20
10
0.01 0.1 1 10 100
Figure 3.4 Moisture(%) vs. (1/W)(DCTxP /µ) for flotation product sample
D
Unlike the disc filters, for HBF application, the cake formation rate is not dictated
by the pressure and CFT. Instead, the feed rate (t/hr) can be mechanically controlled by
simply changing the feed pump speed to the filter, which in turn determines the
throughput of the filter, and thus W . For this reason, if the pressure is kept constant, the
S
HBF’s moisture is mainly controlled by the cake formation time and dry cycle time. The
filtration time, sum of the cake formation time and dry cycle time, can be controlled by
changing the belt speed. At a set pressure and feed rate, slower belt speed may generate
thicker cakes with relatively high cake moistures; whereas higher belt speed may result in
forming thinner cake with low cake moistures. To determine the final cake moisture for
RV-treated mixture sample at a given feed rate and pressure, the dry cycle time should be
determined. This time the difference between the cake formation time and the total
148
)%(
erutsioM
Baseline
RU (3lb/t)
RV (3lb/t)
(1/W)(DCT x P / µ)
D |
Virginia Tech | 3.5. SUMMARY AND CONCLUSIONS
The complexities associated with data and general performance analysis of coal
dewatering processes have typically required an abundance of tests to predict final
moistures under a given set of conditions. As such, an empirical model was developed
for vacuum filtration applications based on data obtained from two comprehensive tests.
The tests were conducted on two samples with different size distributions by varying
parameters such as dry cycle time, vacuum pressure during cake formation, and cake
drying time, formation time, cake weight and thickness. Additionally, different types and
dosages of dewatering aids were utilized. Based on the test data, models were established
and a scale-up simulation was created for vacuum disc filter and horizontal belt filter.
Simulation results indicated that dewatering aids are most effective when used in
horizontal belt filtration due to increased dewatering kinetics, which allow the filter to
employ a longer cake drying time. This also allows more flexibility in controlling the
cake moisture and throughput capacity when the horizontal belt filter is utilized. On the
contrary, when using disc filters, the total filtration time is almost equally shared (1:1
ratio) between cake formation and drying times. Thus, the disc filter is not as effective at
dewatering. It was also found that cake thickness may sometimes be deceiving because of
increased cake porosity caused by dewatering aid addition. A more reliable evaluation
can be made, however, by using specific cake weight instead of cake thickness in the
simulation.
152 |
Virginia Tech | CHAPTER 4 INDUSTRIAL DEMONSTRATION OF DEWATERING
AIDS
4.1. INTRODUCTION
4.1.1. General Background
The United States possesses approximately 25.1 % of the world’s coal reserves with 400
coalfields and small deposits. Coal alone is the largest source of fuel for domestic energy
production, and 90% of that energy source is consumed in power plants. Almost 52% of the
domestic electric power generation is supplied by coal, which accounts for about 33% of the total
energy production. Coal is also used directly by the manufacturing industries that produce
chemicals, cement, paper, ceramics and various metal products. [1, 2]
Annually, more than 1.1 billion tons of coal is mined in the United States. About 600-650
million tons are processed yearly, and, typically, 350-400 million tons are handled in wet
processing. Most of the coal preparation plants only process fine coal from the size fraction
greater than 100 mesh (150 micrometer). For large size material, 16 X 100 mesh (1.0 X 150
micrometer), water-only cyclones and spirals are utilized. Depending on the quantity of the
material present, material smaller than 100 mesh in the total plant feed may either be treated or
discarded. If it is not treated, the fine material – material smaller than 100 mesh – is sent to a
thickener. Then, it is pumped to a slurry impoundment. If the fine material is to be treated,
classifying cyclones are utilized to remove the coal particles finer than 325 mesh (45
micrometer), and the 100 X 325 mesh size material is treated by utilizing flotation; however,
processing and associated costs are significantly increased when dealing with the finer material.
This fact has been a necessary limiting factor for coal processors, sometimes leaving no
154 |
Virginia Tech | economical choice but to discard the fines to a slurry impoundment. Increased mechanization in
the mining industry has also decreased the selectivity and increased the volume of refuse. As a
result, 70-90 million tons of fine coal is discarded yearly to refuse impoundments. As of 2001,
approximately 500-800 million tons of coal had been discarded in 713 active waste
impoundments. In the United States, the majority of the coal waste impoundments are found in
the eastern states – mainly in West Virginia, Pennsylvania, Kentucky, and Virginia.[1-6]
The coal waste impoundments have usually been considered permanent disposal sites;
however, it can also be viewed as an unexploited energy resource, representing a loss of profit.
In the past, recovery of fine coal was not as efficient, resulting in many older slurry
impoundments containing significant amounts of coal refuse larger than 28 mesh (600
micrometer). Today, processing technologies have improved, and efforts have been made to
reduce the coal loss by increasing the quality control during mining, optimizing the processing
systems by improving fine coal recovery, minimizing the mass of solids for disposal, and
utilizing dewatering effectively. As a result, the fraction of material smaller than 100 mesh in
refuse slurries has increased, and the amount of coal being deposited in refuse slurries has
decreased; however, the coal present in impoundments still represents a massive recoverable
energy value. [2, 4-7]
Typically, if an impoundment contains at least 1 million tons of in-situ slurry, a recovery
rate of at least 30% of a marketable, fine coal product from the slurry can prove to be a profitable
venture. For this reason, re-mining impoundments can be considered as a promising method to
recover coal and reduce the slurry volume. Many successful and unsuccessful efforts have been
made over the last twenty to fifty years to re-mine the numerous waste sites. There are a number
155 |
Virginia Tech | tons of potentially recoverable fine coal (see Figure 4.1). The impoundment is an active site and
continues to receive coal slurry refuse from the existing Pinnacle preparation plant as thickener
underflow. Approximately 200,000 tons of additional fine coal is discharged into the
impoundment annually by the existing Pinnacle preparation plant.
This site was ideally suited for a demonstration project since the dewatering aids made it
possible to convert the waste coal impoundment at this site from environmental liability into a
profitable resource.
4.1.3. Preliminary Analysis
Considering the amount of the fines that are going to be fed from the waste pond to the
preparation plant, the success of this project was strongly depended on an effective cleaning and
dewatering operation. Thus, Virginia Tech personnel conducted preliminary assessments on
equipment specification, preliminary circuit design and preliminary cost analysis. A conceptual
circuit design was also developed by personnel at Virginia Tech using standard process design
and cost estimation procedures (Figure 4.2).
The circuit was developed to recover the waste coal fines therefore it included an
advanced flotation processes together with the dewatering technologies developed. The
flowsheet was designed based on average feed rate (dry basis) of 103 tph for the pond reclaim
and was assumed to operate for two 8-hour shifts per day and 250 days per year with 90%
availability. The projections indicated that the POC circuit installed for pond reclamation would
produce clean coal around 68.7 tph and required two 8-disc filters to achieve the target capacity
of 34.0 tph. The raw plant feed will be screened at 6.3 mm and the oversize discarded. The
157 |
Virginia Tech | Beard Technologies and Virginia Tech. The detailed flowsheet included mass and flow balances,
equipment specifications, and size distributions for the various process streams. A proof copy of the
POC flowsheet is provided in Figure 4.3 as an AutoCAD drawing.
Once the flowsheet was completed, detailed listings of required unit operations, such as
equipment type, unit size, throughput capacity, reagent/chemical requirements, power
requirements, air/water requirements, operating limitations, vendor cut-sheets, were produced by
technical personnel at Beard Technologies. Detailed plant layout diagrams were prepared by
Boyce, Graybeal and Sayre (BGS), Inc. of Slab Fork, West Virginia. The layout diagrams
specified the physical arrangement of all primary operations, ancillary processing units,
connecting streams, location of electrical wiring, arrangement of piping and plumbing, and other
pertinent electrical/mechanical requirements. The engineering drawings and specifications were
of sufficient detail to permit mechanical/electrical subcontractors to fabricate, construct, and
assemble the proposed POC circuitry.
4.2.2. POC Fabrication and Installation
Bid packages were prepared for soliciting bids for major purchases of equipment,
materials, fabricated components, and services necessary to complete the installation of the POC
circuitry. Upon receipt, the bid packages were reviewed, and appropriate vendors were selected
based on cost, availability, and suitability. This work included (i) fabrication of all required
components associated with the various POC circuits, (ii) shipping of POC modules, ancillary
equipment and construction materials to the POC site, (iii) inspection of all purchased POC
modules, ancillary equipment and materials to ensure that they are of suitable workmanship and are
161 |
Virginia Tech | structurally, mechanically and/or electrically operational, and (iv) preparation of operation,
maintenance, and safety manuals for each unit operation.
After developing the flowsheet, the installation of all unit operations, piping, electrical
wiring, and instrumentation were undertaken. The on-site construction work was contracted and
managed by Boyce, Graybeal and Sayre (BGS), Inc. of Slab Fork, West Virginia. The fabrication
and on-site construction activities required approximately one year to complete. The plant
incorporates some of the most advanced technologies available to the coal preparation industry
as POC circuits. The most significant of these included a two-stage advanced column flotation
circuit for fine coal separation, a three-stage bank of agitated mixing tanks to condition the
dewatering aids and a paste thickener to convert the fine high-ash wastes into a high-solids
product for disposal with minimal environmental impact.
A simplified process flow diagram of the as-built plant incorporating the POC dewatering
circuitry is provided in Figure 4.4. A photograph of the nearly completed plant is shown in
Figure 4.5. The plant was designed with a raw feed nameplate capacity of approximately 200
tph of dry solids. The plant sizes/classifies the feed from the dredge into four nominal size
fractions, plus28 mesh, 28 x 100 mesh, 100 x 325 mesh, and minus 325 mesh. The initial
separation occurs on a sieve bend and single deck vibrating screen. The plus28 mesh material
(screen oversize) is discharged to a stockpile outside the plant, and the minus 28 mesh reports to a
30,000 gallon surge tank. Material from the surge tank is pumped into 20 inch diameter classifying
cyclones, which nominally sizes the feed at about 100 mesh.
162 |
Virginia Tech | column flotation bank. The deslime overflow is currently discharged as waste, but provisions are
included in the flowsheet design to incorporate a secondary advanced flotation bank, dewatering
aid conditioners, and disc filter to recover the ultrafine coal that is now lost in this stream. The
cleaned froth product from the column units passes into a de-aeration tank to promote breakdown
of any residual froth.
The froth product is then passed through three agitated mixing tanks (Figure 4.9) where
dewatering aids are added. After conditioning for several minutes, the treated slurry is fed to a
bank of disc filters (Figure 4.10) for final dewatering. Previous test work conducted in this
project has shown that adequate mixing, both in terms of time and intensity, is critical to the
performance of the dewatering aids. The dewatered froth product is discharged to a reversible
product collection belt. The coal product can be directed to either a clean coal collection belt or
a noncompliant conveyor system, if the moisture of the filter product is higher than the
specifications allow. The total product moisture is continuously monitored using an on-line
Figure 4.9 Three-stage conditions used Figure 4.10 Bank of disc filters used to
for conditioning the dewater the fine coal froth
dewatering aid product
165 |
Virginia Tech | Figure 4.11 Static thickener used to
thicken solids and clarify Figure 4.12 Paste thickener used to
process water further thicken wastes for
disposal
moisture analyzer located on the clean coal collection belt.
The refuse from the various plant circuits is treated using a two-stage thickener system.
The first stage consists of a 90 ft. diameter high-rate static thickener (Figure 4.11) into which
coagulant and flocculant is added to promote aggregation and rapid settling of the fine solid
waste. The overflow from this unit is taken back into the plant as clarified process water. The
underflow is pumped to a 60 ft. diameter paste thickener (Figure 4.12) for secondary
densification. The bed depth is maintained between 12 and 25 ft and has a retention time
between 11 to 15 hrs.
4.2.3. POC Circuit Testing
a) Shakedown Testing
At the completion of the POC design and construction, preliminary shakedown tests were
conducted to resolve operational problems that arose during start-up of the POC plant. Initial
test runs were conducted to ensure that pumping capacities, pipe sizes, electrical supplies, control
166 |
Virginia Tech | systems, and instrumentation were adequate. After completing start-up activities, exploratory
tests were conducted to validate the design capacities of the various unit operations used in the
POC circuits. Data obtained from these tests were used to identity key operating parameters that
should be investigated in detailed testing. This work was followed by detailed testing of the
plant circuitry and, in particular, evaluation of the fine coal vacuum filters where the dewatering
aides were utilized.
In general, the in-plant testing of the POC circuitry showed that the sizing, cleaning,
dewatering and disposal circuits performed as expected. During these tests, the plant was fed the
design capacity of 200 ton/hr of raw feed and produced an average of 58 ton/hr of fine clean
coal. Of this tonnage, 39 ton/hr was plus 100 mesh and 17 ton/hr was nominal 100 x 325 mesh.
The final product (nominal plus 325 mesh fraction) generally met the product quality
specifications of 5.5% ash, 0.8% sulfur, and 17% moisture; however, some fluctuations in
product qualities were observed from time to time due to the widely varying characteristics of
the feed material extracted from the impoundment. Size-by-size summaries of the performance
data for the shakedown tests are provided in Figure 4.13 and Figure 4.14 for the coarse and fine
plant circuits, respectively.
The test data indicate that the 20-inch diameter classifying cyclones provided a nominal
cut size of 100 mesh. The underflow contained about 75.05% of the plus 100 mesh solids at an
ash content of 7.70%. 24.95% of the plus 100 mesh solids containing 36.36% ash was misplaced
material. The overflow contained 2.87% of the plus 100 mesh solids at 2.50% ash, which
reported to the de-slime cyclone circuit. The compound spirals generated a coal product
containing 80.15% plus 100 mesh at 2.98% ash, with 19.85% minus 100 mesh at 21.49% ash
reporting with the product. The spiral middlings product was found to be of a sufficient quality
167 |
Virginia Tech | 20" Cyclone 20" Cyclone
U/F U/F
20" Cyclone 20" Cyclone Plus 28
O/F O/F 28x100
100x325
Spiral Clean Plus 28 Spiral Clean Minus 325
28x100
100x325
Spiral 1 Refuse Minus 325 Spiral 1 Refuse
Spiral 2 Refuse Spiral 2 Refuse
15" Cyclone 15" Cyclone
U/F U/F
15" Cyclone 15" Cyclone
O/F O/F
Sieve 1 O/S Sieve 1 O/S
Sieve 2 O/S Sieve 2 O/S
0 20 40 60 80 0 20 40 60 80 100
Weight (%) Ash (%)
Figure 4.13 Weight and ash distributions for samples collected from the shakedown tests
conducted on the coarse coal treatment circuits
to be recovered without additional treatment. So, the splitter was closed to direct this stream into
the product. The primary and secondary spiral reject streams contained 50.73% and 64.11% plus
100 mesh solids at 59.08% and 13.42% ash, respectively, and 49.27% and 39.89% minus 100
mesh solids at 72.09% and 42.72% ash, respectively.
The spiral product was pumped into the clean coal cyclone circuit to provide desliming of
the plus l00 mesh product. The circuit incorporates 15-inch diameter classifying cyclones and a
two-stage rapped sieve bend system. The cyclone underflow contained 83.25% of plus 100 mesh
solids at 4.10% ash. A total of 16.75% of minus 100 mesh solids at 49.73% ash were contained
168 |
Virginia Tech | in this stream. The overflow, which reports to the de-slime cyclone circuit, contained 13.51% of
plus 100 mesh at a 2.84% ash. The two-stage rapped sieve was used to remove minus 100 mesh
contamination and associated ash from the coarse circuit product. The higher quality of the
oversize product produced by the primary and secondary rapped sieve bends reduced the amount
of minus 100 mesh in the product from 16.75% to 7.46% with a corresponding ash reduction
from 11.74% to 5.47%. The undersize solids from the rapped sieve bend circuit were circulated
back to the clean coal sump.
The product from the cyclone/spiral circuit was dewatered through two 36 x 72 inch
screen bowl centrifuges. The coal product from the screen-bowl was 5.18% ash, which
4" Cyclone U/F 4" Cyclone U/F
4" Cyclone O/F 4" Cyclone O/F
Rougher Froth Rougher Froth
Rougher Tails Rougher Tails
Cleaner Froth Cleaner Froth
Screen-Bowl Screen-Bowl
Cake Cake
Plus 28
Disc Filter Disc Filter 28x100
Cake Cake 100x325
Plus 28 Minus 325
Combined Combined
28x100
Clean Clean
100x325
Minus 325
Thickener U/F Thickener U/F
0 20 40 60 80 100 0 20 40 60 80
Weight (%) Ash (%)
Figure 4.14 Weight and ash distributions for samples collected from the shakedown tests
conducted on the fine coal treatment circuits
169 |
Virginia Tech | contained 78.29% plus 100 mesh at 4.17% ash and 21.71% minus 100 mesh at 8.82% ash.
The shakedown test data showed that the 4-inch diameter classifying cyclones provided a
nominal cut size of 325 mesh. The deslimed underflow contained 73.06% plus 325 mesh (6.52%
ash) and 26.94% minus 325 mesh (47.12% ash). The deslime overflow, which was discarded to
the static thickener as waste, contained 12.05% plus 325 mesh (4.77% ash). The loss of low-ash
material in this stream was expected because the ultrafine cleaning circuit had not yet been added
to the plant circuitry. The clean froth product from the entire column flotation circuit was found
to contain 6.66% ash. This product contained 78.34% plus 325 mesh at 2.84% ash and 21.66%
minus 325 mesh at 20.45% ash. The rougher column reject (i.e., feed to the cleaner column) was
found to contain 39.12% ash. Of the rougher column reject, 51.72% was plus 325 mesh (11.17%
ash) and 48.28% was minus 325 mesh (69.05% ash). The ash content of the cleaner column
froth was 26.69%, and of that percentage, 60.81% at a 5.72% ash is plus 325 mesh and 39.19%
at 59.22% ash is minus 325 mesh. The high ash content of the minus 325 mesh fraction can be
attributed to problems associated with the entrainment of fine clay in the column froth. As such,
the froth quality is expected to improve after modifications are made to the wash water system.
So, more minus 325 mesh clay is eliminated from the froth products.
As indicated previously, the product from the deslime cyclone/column flotation circuit is
dewatered by using a twelve-disc 12.5 ft diameter vacuum disc filter. In the shakedown tests,
the filter product was found to contain 8.76% ash (74.88% plus 325 mesh at 3.21% ash and
25.12% minus 325 mesh at 25.30% ash). The higher ash in the minus 325 mesh size fraction
was found to have a major impact on both the final ash quality of the filter product and on the
overall product moisture. On average, the moisture of the filter cake was found to be
170 |
Virginia Tech | approximately 26% in the absence of dewatering aids. The addition of dewatering aids reduced
the filter moisture to below 18%.
The 90 ft diameter, high-rate thickener required modifications to allow optimum
operation during the shakedown tests. The required modifications were to raise the flume above
the liquid level in the thickener unit to avoid surface froth entering into the clarified water tank,
flatten the slope and increase the width of the flume for capacity to slow the velocity and
minimize the agitation of the slurry material, and divert the slurry flow down into the center well
at the end of the flume to reduce short-circuiting of slurry out of the center well into the main
body of the thickener. The underflow from the static thickener provided the feed for the paste
thickener. The paste thickener, which is still being subjected to shakedown testing, appears to be
capable of handling the varying feed rates, size distributions, and qualities of waste slurry from the
processing plant.
b) Detailed Testing
After completing the shakedown tests, several series of detailed tests were performed for the
POC dewatering circuit. For comparison, a set of laboratory dewatering tests were conducted
during the detailed testing. Both the laboratory and POC-scale tests were completed using RV.
The laboratory filtration tests were conducted using a 2.5- inch diameter Buchner funnel. After
establishing the optimum operating conditions and reagent blends using the laboratory filter, the
dewatering aid was added to the feed of the three-stage conditioners that were located just ahead
of the vacuum disc filter. As indicated previously, adequate mixing is critical because previous
studies showed that the moisture reduction improves with increasing energy input during
conditioning. For this reason, the full three-stages of conditioners were used in all of the POC-
scale dewatering tests.
171 |
Virginia Tech | It was also found that the quality and size distribution of the coal within the
impoundment varied greatly at different locations. These variations affect the cleaning and
dewatering processes. For that reason, some parallel dewatering tests were also conducted
during the plant testing. In this case, the sample was collected directly from the plant’s cyclone
overflow stream and floated at the lab using the Denver Flotation cell to produce the filter feed
sample to be used in laboratory filtration tests. Table 4.1 shows the results of dewatering test
using RW and RV as dewatering aids. The moisture reduction corresponds to an approximate
40% decrease; however, as shown below, the results are seven to eight points higher compared to
previous results. This change in the moisture can be attributed to the particle size distribution.
The flotation feed sample collected from the cyclone overflow contained approximately 70 %
minus 325 mesh particles.
Another set of tests was conducted with another sample collected a different day. The
results are shown in Table 4.2. The moisture reduction was again around 40%; however, the
minus 325 percent material in the feed was too high. The reagents were greatly effective, but the
sample was not representative of their effectiveness. To simulate the average plant filtration
Table 4.1 Effect of RV and RW (dissolved in diesel at 1:2 ratio) at various dosages
Reagent Moisture Content (%) Moisture Reduction (%)
Dosage
RW RV RW RV
(lb/ton)
0 34.03 34.03 -- --
0.5 33.05 24.58 2.9 27.77
1.0 26.03 22.91 23.50 32.68
3 23.97 20.76 29.56 38.99
5 23.33 20.15 31.44 40.79
172 |
Virginia Tech | Table 4.2 Effect of RV and RW (dissolved in diesel at 1:2 ratio) at various dosages on
dewatering of floated sample
Reagent Moisture Content (%) Moisture Reduction (%)
Dosage
RW RV RW RV
(lb/ton)
0 32.92 32.92 -- --
3 20.59 19.21 37.45 41.67
5 19.27 19.14 41.46 41.86
operation the sample was deslimed by removing either all (totally deslimed) or just two-thirds
(partly deslimed) of the minus 325 mesh solids from the test sample. Table 4.3 shows that the
deslimed samples gave moisture contents below the target value of 17% moisture.
Table 4.3. Effect of RV (dissolved in diesel at 1:2 ratio) at various dosages on dewatering of
deslimed samples
Reagent Moisture Content (%) Moisture Reduction (%)
Dosage Total Partly Totally Partly
(lb/ton) Deslimed Deslimed Deslimed Deslimed
0 20.90 26.41 -- --
3 13.22 17.01 36.75 35.56
5 13.08 16.99 37.42 35.67
4.2.4. Scale-Up Assessment
A comparison of the laboratory and POC-scale filtration test results is given in Figure
4.15. Because the baseline moisture was different in each case (i.e., 26% vs. 24%), the data have
been plotted again in Figure 4.16 as percentage moisture reduction for each reagent dosage. The
experimental data clearly demonstrate that the addition of dewatering aid substantially reduced
the moisture contents of the filter products. In the POC-scale tests, the total moisture content
173 |
Virginia Tech | 28
26
24
22
20
18
16
0 0.5 1 1.5 2
Reagent Dosage (lb/t)
was reduced from 26% to 20% at a dosage of approximately 0.5 lb/ton and farther down to
17.5% at about 1 lb/ton. The 17.5% moisture represents a moisture reduction of nearly one-third
compared to the baseline moisture of 26%. For example, the cake thicknesses observed in some
of the POC-scale tests were as large as 3 inches. These POC-scale results compare favorably
with the laboratory data. The apparent gap between the laboratory and POC-scale results
obtained at higher dosages of dewatering aid may be explained by differences in particle size,
cake thickness, drying time, etc., used in the two test programs.
Another important observation made during the detailed test program was that the power
draw for the disc filter vacuum pumps dropped dramatically upon addition of the dewatering aid.
An example of this behavior is shown in Figure 4.17 for one of the test runs performed at the
plant. The power draw dropped from a normal baseline value of about 160 to 115 amps after the
174
)%(
erutsioM
35
Laboratory
30
In-Plant
25
20
15
10
5
0
0 0.5 1 1.5 2
Reagent Dosage (lb/t)
Figure 4.15 Moisture content versus
dewatering aid dosage (RV)
)%(
noitcudeR
erutsioM
Laboratory
In-Plant
Figure 4.16 Moisture reduction versus
dewatering aid dosage (RV) |
Virginia Tech | 4.3. SUMMARY AND CONCLUSIONS
In the present work, capabilities of novel dewatering aids were successfully tested on the
fine coal that is processed at the Smith Branch impoundment site. In United Stated, annually,
approximately 1.1 billion tons of coal is mined and 600-650 million tons of the production is wet
processed while remainder is utilized as is. Typically, wet processing methods are responsible
for processing 350-400 million tons of coal. However, especially for fine materials smaller than
100 mesh, cleaning and dewatering costs have significantly been high. This fact has been a
limiting factor for the coal processing industry. As a result, depending on the quality and the
quantity of the fine size fraction, coal has been either treated and sold or discarded to waste
slurry impoundments.
As of 2001, there are 713 active waste slurry impoundments in which approximately 500-
800 million tons of coal lies. Considering today’s economical conditions however, the coal in
the waste slurry impoundments may now be an unexploited energy source. Consequently, re-
mining the impoundments may be considered a promising method to recover this coal which
previously represented a loss of profit.
The Smith Branch impoundment, located in West Virginia, is estimated to contain 2.85
million tons of potentially recoverable fine coal. At the time of this work, Beard Technologies
Inc. owned the permit to re-mine and recover coal from this site. Considering the amount of fine
size fraction, the success in recovering the coal from this impoundment was strongly dependent
on successful cleaning and a very effective dewatering operation. Thus, this site was ideally
suited for a demonstration project since the novel dewatering aids made it possible to convert the
waste coal impoundment at this site from environmental liability into a profitable resource.
176 |
Virginia Tech | Based on the feasibility studies, Beard Technologies constructed a POC circuit. The
plant was designed with a raw feed capacity of approximately 200 tph of dry solids. The plant
sizes/classifies the feed from the dredge into four nominal size fractions, plus28 mesh, 28 x 100
mesh, 100 x 325 mesh, and minus 325 mesh. The circuit incorporates the novel dewatering
technology together with advanced flotation process to ensure maximum moisture reduction and
coal recovery for the 100 x 325 mesh material.
Cleaning is done by a two-stage advanced column flotation bank. The cleaned froth
product from the column units passes into a de-aeration tank to promote breakdown of any
residual froth. The froth product is then passed through three agitated mixing tanks where
dewatering aids are added. Previous test work has shown that adequate mixing, both in terms of
time and intensity, is critical to the performance of the dewatering aids. After conditioning for
several minutes, the treated slurry is fed to a bank of a twelve-disc, 12.5ft diameter vacuum disc
filter for final dewatering.
In the POC-scale tests, the total moisture content was reduced from 26% to 20% at a
dosage of approximately 0.5 lb/ton and further to 17.5% at approximately 1 lb/ton of novel
dewatering aid; this represents a total moisture reduction of nearly 33%. As well, dewatering
kinetics were also increased i.e., the cake thicknesses observed in some of the POC-scale tests
were as much as 3 inches. These POC-scale results also compare favorably with the laboratory
data.
Another important observation made during the detailed test program was that the current
required for the disc filter vacuum pumps dropped dramatically upon addition of the dewatering
aid. The normal baseline current value was about 160 amps, but fell to 115 amps after the
177 |
Virginia Tech | CHAPTER 5 DEWATERING OTHER MINERALS
5.1. INTRODUCTION
In the previous chapters, the feasibility of using novel dewatering aids was investigated in
detail. These dewatering aids adsorb on the surface of the hydrophobic particles such as coal,
and increase the hydrophobicity. The results obtained from laboratory, pilot, and full scale tests
that were conducted on several coal samples confirmed that the dewatering aids are capable of
decreasing the final cake moisture by 20-50 % and increasing the filtration kinetics to a point that
cannot be achieved with existing mechanical capabilities. In order to investigate and confirm
that the novel dewatering aids can also be effective on other minerals and hydrophilic particles, a
number of dewatering tests were conducted on minerals such as copper and hydrophilic materials
such as kaolin clay and fly ash.
5.2. KAOLIN CLAY
5.2.1. Background Information
Kaolin clay, a hydrated aluminum silicate (Al Si O (OH) ) is a white, soft mineral and is
2 2 5 4 ,
mainly composed of fine-grained, plate-like particles. It is commonly referred to as "China
Clay," which refers to its discovery in Kao-Lin, China. Kaolin clay deposits are classified as
either primary or secondary. Primary kaolin results from residual weathering or hydrothermal
alteration of feldspar and muscovite. Secondary kaolin is sedimentary in origin.[4, 1, 2]
Kaolin is a distinctive industrial mineral due to its unique physical, physiochemical, and
chemical properties. For example, it remains chemically inert over a relatively wide pH range, it
is non-abrasive, and it has low thermal and electrical conductivities. As a result, it has broad
applications in many industries including (by percent): paper and filling (45%), refractory (16%),
180 |
Virginia Tech | ceramics (15%), fiberglass (6%) cement (6%), rubber and plastic (5%) and paint catalyst and
others (7%). In the paper industry, high-brightness kaolin is used for coating, and low-brightness
kaolin is used as filler.[4, 5, 6]
The total global kaolin production is estimated (2007) to be 37.8 million tons per year,
and the US has approximately 20% of the clay production share in the world markets. The other
major kaolin producers are Brazil, Czech Republic, Germany, Republic of Korea and United
Kingdom. In the US, a stream of kaolin-containing rocks reaches from Eufaula, Alabama
approximately 250 miles to Aiken, South Carolina. This stream contains 7 to 10 billion metric
tons of sedimentary kaolin. 90% of US kaolin production is concentrated in a 150-mile-long
segment of this stream that extends from Macon, Georgia to Aiken, South Carolina. Georgia is
responsible for producing 92% of the tonnage and 96% of the dollar value, while South Carolina
produces 4% of the tonnage and 2% of the dollar value. Kaolin from this region of the US is of
high quality due to its high brightness and relatively low viscosity at high solids concentration
(70%).[2-4, 7]
By and large, all kaolin is mined using open pit methods utilizing shovels, draglines, and
backhoes. Two processing methods – one dry, the other wet – are used in the production of
kaolin. The dry process is rather simple and inexpensive, while the wet process employs
sophisticated separation techniques to improve the color brightness and size consistency.[1, 5]
In the wet process, kaolin clay is treated with magnetic separation, flotation, selective
flocculation, leaching, or combinations of multiple methods, and usually contains more than 80%
moisture content. The most difficult phase of processing is dewatering of this moisture content,
due to the naturally ultra-fine particle size of the clay. Fine particles require more complex and
intense chemical and mechanical treatment due to the larger surface area to be dewatered and
181 |
Virginia Tech | smaller capillaries. This, along with the kaolin clay’s natural hydrophilic characteristics, gives
higher capillary pressures. In industry, kaolin clay is dewatered using vacuum drum filters and
various types of pressure filters, which can lower the moisture to 40% to 60%. In kaolin clay
dewatering, a unique function of the filtration process is also to remove chemicals from the clay.
Depending on the market requirements the filtration is followed by spray drying, which,
although costly, is the only method capable of producing a very low moisture product
(approximately 10% to 30%).[4, 8-10]
In this section, the effects of novel dewatering aids were investigated using two new
methods, namely Two Step Hydrophobization and Foam Assisted Dewatering Method.
5.2.2. Two Step Hydrophobization Method
a) Experimental
A number of kaolin clay samples with various properties were obtained from different
companies for use in dewatering tests. The tests utilized a two-step hydrophobization technique,
described below. The samples are:
i) Thiele Kaolin Company’s Kaolin Clay Sample (25% solid, pH 7.2)
ii) J.M.H Corporation’s PSD Filter Vat Clay Sample (30% solid, pH 4)
iii) J.M.H Corporation’s PSD Filter pre-leached Clay Sample (30% solid, pH 9.5)
b) Results and Discussion
1) Thieles’ Sample (30% solid at pH 7)
In the first set of laboratory dewatering tests, East Georgia kaolin clay sample received from
Thiele Clay Company was subjected to dewatering tests using the two-step hydrophobization
method. The as-received sample was first subjected to some simple analyses, such as pH value
and solid content determination. The measured pH value for the clay slurry sample was 7.14 and
182 |
Virginia Tech | had 25% solids with a size that was finer than 2µm. In this set of tests, a laboratory pressure
filter unit was used for all tests at 60psi pressure. The as-received sample was first agitated for a
given period of time to keep the clay particles suspended and well-dispersed in the slurry,
followed by dilution using Blacksburg tap water from 25% to 8%. In the first dewatering test, a
known volume of the diluted sample was taken and tested at 60psi pressure without any surface
treatment. During this control test, the dewatering kinetics were noticeably low, giving
approximately 25 minutes cake formation time, where the adjusted drying cycle time was 5
minutes. The cake thickness was 1.5 to 2mm, having an average of 39.14% final moisture. The
following tests were done in the presence of various dosages of dodecylamine chloride (DAC) to
create a slightly hydrophobic surface as the first step of two-step hydrophobization. After DAC
was added, the slurry was agitated for 5 minutes with the stand-alone 3cm, three-bladed,
propeller-type mixer, transferred to the pressure filter, and subjected to the same tests. During
these tests, it was observed that the slurry was becoming more viscous by further addition of
DAC after about 10 lb/ton. This may be attributed to the spontaneous flocculation facilitated by
amine during mixing. As shown in Table 5.1, in the presence of DAC, the moisture contents
were reduced to 32.14% and 31.90% at 5lb/ton and 10lb/ton reagent addition, respectively. The
cake formation time, which is an indicator of dewatering kinetics, was also cut down to 4-8
minutes, even though the cake thickness was increased to 2mm to 4mm. The increased cake
thickness may be an indication to an increased diameter of capillaries, which in turn, lower the
capillary pressure and ease the water removal from the cake.
183 |
Virginia Tech | Table 5.1 Effect of amine (DAC) addition as the first step on dewatering of Georgia clay
Reagent
Dosage Moisture Content (%)
(lb/t)
0 39.14
5 32.14
10 31.90
15 33.23
20 34.23
*The clay sample was diluted from 25% to 8%.The measured pH was 7.14. Conditioning time was 5 minutes. Applied pressure of
the filter was 60 Psi. The cake thickness for baseline was 1.5- 2 mm, with reagent 2-4 mm. Cake formation time for baseline was 25
minutes, with reagent 4-8 minutes. Drying time was 5 minutes.
As mentioned above, this is a two-step hydrophobization process, and the addition of
various dosages of RW was investigated as the second step on the amine-treated sample.
Table 5.2 gives the laboratory test results obtained on the clay sample using amine
(DAC) and RW under the given conditions. For this series of tests amine dosage was kept
constant at 10lb/ton. The mixing time for RW was 2 minutes. The addition of RW as the second
hydrophobization step increased the kinetics. The cake formation time was further decreased to
3-5 minutes from 4-8 minutes depending on the RW dosage. The final cake moisture for 3 lb/ton
RW addition was reduced to 28.3%, giving 27.6% overall moisture reduction. These results
clearly indicate that the two-step hydrophobization process is a very effective way of lowering
the final cake moisture and increasing the dewatering kinetics and thus throughput of a filter unit.
Fundamentally, it may be also concluded that the reagents that are used, either alone or in
combination with each other, enhance the clay dewatering by controlling contact angle, surface
tension, and capillary radius. These quantities are related by the Laplace equation, and are
discussed in detail in the preceding literature review.
184 |
Virginia Tech | 2) J.M.H Corporation’s PSD Filter Vat Kaolin Clay Sample (30% solid, pH 4)
In this set of dewatering tests, East Georgia kaolin clay samples received from J.M.
Huber Corporation were subjected to dewatering tests to investigate the effect of using the two-
step hydrophobization method. The same test procedure was followed, except that the 2.5inch-
diameter, 6inch-height, Buchner funnel vacuum filter at 25in. Hg vacuum pressure was used
instead of a pressure filter. The filter cloth was supplied from the same company. The first
sample tested was PSD filter Vat Kaolin Clay sample that had 30% solid content, and a pH of
about 4.
Tests results obtained on the as-received clay sample show that varying amine addition
did not make any significant difference in final cake moisture where the baseline moisture was
approximately 40-41%. However, dewatering kinetics was observed to increase with increasing
amine addition. When RW was added to the amine-treated sample as the second step of
hydrophobization, the moisture reduction was not significant, i.e., 1-2 %. This might be
explained by the fact that the magnitude of zeta potential has an important influence on the
aggregation of particles in water. When the magnitude of zeta potential is low, the repulsive
force between the particles is reduced so that the particles come closer and form agglomerates,
which occur at acidic pH. Agglomeration increases the settling rate of particles, resulting in fast
dewatering kinetics. On the contrary, it may induce the trapping of water in the agglomerate
structures, which directly increases the moisture content of the final cake. The solid
concentration of the slurry is of interest, too, because dewatering kinetics and final cake moisture
are affected by the settling rate. As the suspended solid increases, a lower settling rate will occur
up to a certain concentration level, which will, in turn, affect moisture content. High solid
content also makes reagent conditioning difficult because of the high viscosity. This may be
186 |
Virginia Tech | another reason that the as-received sample of 30% solid content did not yield desirable moisture
reduction, even after two-step hydrophobization.
Taking these facts into account, another series of laboratory dewatering tests was
conducted at a reduced solid content using the same sample. For these tests, the as-received
sample was first diluted to 15% solid content using Blacksburg tap water prior to dewatering to
exclusively inspect the effect of solid concentration on dewatering of fine clay samples.
Accordingly, the pH was also increased to 4.5-5. Table 5.3 shows the results obtained to
investigate the effect of two-step hydrophobization on the dewatering of the diluted clay slurry
sample. For filtration tests, again, a 2.5-inch diameter, 6-inch tall Buchner funnel vacuum filter
was used at 25in. Hg vacuum pressure. During the tests, amine dosage was kept constant at a
pre-determined dosage i.e.,10lb/ton, while RW dosages varied from 1lb/ton to 20lb/ton. For RW
addition of 10lb/ton, the results indicate that lowering the solid content decreases the final cake
moisture significantly, from 40-41% to 34.58%, even when the slurry is at acidic pH.
The same tests were repeated on the diluted sample with pH further adjusted to about 7
by adding sodium carbonate (Na CO ). Table 5.6 shows the effect of dilution and pH adjustment
2 3
on dewatering of the clay sample. Test results show that the cake moisture was significantly
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 4.5-5.0
Reagent Moisture Content (%) pH 4.5-5
Dosage
Amine+RW
(lb/t)
0 40-41
5 40.61
10 42.73
15 34.58
20 35.22
187 |
Virginia Tech | 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
Reagent Moisture Content (%) pH 7.0
Dosage
Amine+RW
(lb/t)
0 40-41
5 38.85
10 32.00
15 31.85
20 28.57
reduced from 41% to 28.57% when RW was added at pH 7 and 15% solid content on amine
treated sample. This is a 30% overall moisture reduction. The results also indicate that
filterability is better at lower solid concentrations. When the solid content of the slurry is
increased, it is more difficult to filter kaolin clay. It is also evident that pH has a considerable
influence on the filterability and final cake moisture of clay.
3) J.M.H Corporation’s PSD Filter pre-leached Kaolin Clay Sample (30% solid, pH 9.5)
To investigate the effect of two step-hydrophobization dewatering technique the sample
was first tested as it is received i.e. 30% solid at pH 9.5. It was observed that at this higher value
of pH, the particles were so well dispersed in the slurry that there was not any significant cake
formation. Also, the filter cloth had an increased tendency to blind during the control tests. As a
result of filter cloth blinding, no cake formation was observed. Test results obtained on this
sample indicated that amine addition tended to help cake formation in the expense of increasing
the final cake moisture (Table 5.5). At low amine dosage (1-5lbs/ton), the cake formation was
very poor, and the cake was extremely thin, rendering the results not reproducible. In contrast,
an amine addition of 10lb/ton or above provided decent cake thicknesses and produced more
188 |
Virginia Tech | Table 5.5 Effect of using amine only on the dewatering of PSD Filter pre-leached kaolin
clay sample (pH 9.5& 30% solids)
Reagent
Dosage Moisture Content (%) pH 9.5
(lb/t)
0 -
1 29.30
3 28.86
5 32.80
10 37.66
15 40.00
20 42.24
reliable moisture data. The best slurry characteristics for dewatering tests were obtained at 10
lb/t of amine dosage.
In the previous tests, promising results were obtained at pH 7 and for that reason some
parallel tests were conducted on the same sample by lowering the pH to 7 using sodium
carbonate (Na CO ). Table 5.6 shows the test results obtained on the same sample at 30% solids
2 3
and pH 7. At this pH, the particles were still so well dispersed in the slurry that there was not
any significant cake formation for the control tests. Observations during dewatering tests on this
sample indicate that at low amine dosage (1-5 lbs/ton), the cake formation was again poor and
Table 5.6 Effect of Using Amine Only on the Dewatering of the Coarse Pre-leach kaolin
Sample from Huber (pH 7 & 30% Solids)
Reagent
Dosage Moisture Content (%) pH 7
(lb/t)
0 -
1 24.79
3 30.14
5 30.67
10 37.01
15 41.92
20 42.98
189 |
Virginia Tech | the cake was thin. As in the previous test, the amine addition of 10 lb/ton, or above, improved
cake thicknesses and produced more reliable moisture data. Further addition of amine was
observed to make the slurry more viscous and, therefore, slurry conditioning with reagents
became difficult. As a result, the final cake moisture actually increased. For this reason, a
10lb/ton amine dosage addition was chosen for the first hydrophobization for the tests that were
done at pH 7 and 9.5.
Table 5.7 gives the overall laboratory test results using amine and RW under the given
conditions, 30% solid content and pH 7 and 9.5. As mentioned above, this is a two-step
hydrophobization process, where the addition of RW constitutes the second hydrophobization
step.
In this series of tests, amine dosages were kept constant at 10lb/ton, while RW dosage
was varied from 1 to 10 lb/t. At pH 9.5, with 10 lb/ton amine and 10 lbs/t RW the final cake
moisture was significantly reduced from around 37.66% to 28.3%; at pH 7, the moisture was
reduced from 37.01% to 29.11%. These results correspond to 24% and 21% overall moisture
reduction, respectively. The addition of RW in conjunction with amine also increased the
kinetics. A shorter cake formation time and lower final cake moisture indicated that the two-step
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
Reagent Moisture Content (%)
Dosage
pH 9.5 pH 7.0
(lb/t)
0 37.66 37.01
1 36.87 36.02
3 33.60 32.02
5 33.92 31.03
10 28.31 29.11
190 |
Virginia Tech | 5.2.3. Foam Aided Dewatering Method
There are no previously published studies on foam dewatering and the mechanisms which
render this method effective are not completely understood at this time. However, tests
conducted at Virginia tech suggest the following: first, the foam increases the void spaces
(porosity) within the cake structure and, second, foam lowers the surface tension of particles
such that in a clay slurry system, for example, the foam may actually increase the dewatering
efficiency. This idea is in agreement with Laplace’s theory, i.e., lowering surface tension,
increasing capillary radius and contact angle. To investigate this new concept, some tests were
conducted using foam.
a) Experimental
For this study, the Thiele Kaolin Company sample was used. It is a flotation product and
contains approximately 25% solids. In some of the tests, the clay slurry is diluted with tap water
down to 5~8% solid for dewatering testing. A set of tests was conducted at natural pH of about
7.2.
Two procedures were tested to apply foam to clay dewatering. These methods are:
i) Method 1: A foam-generating agent (Tergitol) was added to the slurry, and then
agitated strongly with air blown into the slurry until it was filled with foam. The
foamy slurry is then subjected to the pressure filtration tests.
ii) Method 2: The foam was generated separately by agitating the foam-generating
agent containing water in presence of blowing air. Then the foam that was
generated was added onto the top of clay cake inside the pressure filter chamber
after the cake formation time.
192 |
Virginia Tech | In foam dewatering tests Tergitol (with different NP numbers) were used as foaming
agent. One of the advantages for using Tergitol as foaming agent is that it does not contaminate
the clay surface because it does not adsorb onto the clay surface.
b) Results and Discussion
The preliminary test results showed that the vacuum filtration was not effective with the
foam dewatering of clay. With consideration of the natural difficulty in clay dewatering, and
unsatisfactory moisture reduction obtained from vacuum filtration, all the tests were conducted
using a laboratory-scale pressure filter. In each dewatering test, a desired volume of clay slurry
was prepared using the methods that was described, and then transferred to the pressure filter.
During the filtration, cake formation time and drying cycle time were recorded. The air pressure
for filtration was set at 60 psi for all the tests. #44 filter papers manufactured by Whatman
Company were chosen as filtration media. The filtrate obtained with #44 filter paper was always
very clear.
Foam Aided Clay Dewatering Tests at pH 7.2
Method 1: To investigate the effect of foam on clay dewatering, clay sample was
conditioned with various dosages of Tergitol NP-7 and Tergitol NP-9 (surfactants that are
capable of generating foam in a solution) and aerated to create a foamy environment. In this set
of test, after adjusting the slurry ph to 7.14, Tergitol, foam generating agent, was added to the
slurry and agitated strongly while injecting air. 5 minutes was given for this process. The
pressure for filtration was adjusted to 60 Psi. Table 5.8 shows that Tergitol NP-7 or Tergitol NP-
9 alone could reduce the cake moisture down to as low as 25%. This moisture reduction
correspond to approximately 35 % overall reduction. The results also showed that, Tergitol was
very effective at low dosage i.e., 3 lb/t and higher dosages did not make a lot difference.
193 |
Virginia Tech | Table 5.9 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.
Reagent Moisture Content (%) pH 7.14
Dosage
Tergitol (NP-7) Tergitol (NP-9)
(lb/t)
0 39.14 38.04
3 26.69 27.01
5 - 26.52
10 26.35 26.11
30 25.52 25.33
Slurry was diluted from 25% to 8%.The measured pH was 7.14. Conditioning time was 5minutes. Applied pressure was 60
Psi. The cake thickness for baseline was 1.5- 2 mm, with reagent 2-4 mm. Cake formation time for baseline was 25 minutes,
with reagent 20-25 minutes. Drying time was 5 minutes
The cake formation time at pH 7.14 was 25 minutes for baseline test and 20 minutes in
the presence of different types of Tergitol. The drying time was 5 minutes. During the tests, it
was observed that, when Tergitol added into the slurry along with air and agitated, the cake
thickness increased from 1- 1.5 mm to 2-4 mm.
Another set of tests were conducted to investigate the effect of foam on a thicker cake.
The tests discussed in the above section were conducted at 2~4 mm thick cake by using 50 ml
slurry. Table 5.9 shows the test results obtained with approximately 5-6 mm thick cake when
100 ml slurry was used for filtration tests in the presence of Tergitol NP-7. As expected, the
increased cake thickness gave higher cake moisture, but the moisture reduction from 43% to
29% is still impressive, giving 33% overall moisture reduction.
Table 5.8 Effect of Tergitol with thick cake (5-6mm) on dewatering of Georgia clay
Reagent
Dosage Moisture (%)
(lb/t)
0 43.45
1 29.19
5 29.94
10 30.23
194 |
Virginia Tech | Method 2: Another way of using foam is adding foam that is generated separately on the
top of the cake. Once the cake is formed, it is possible to add foam onto the top of the cake. The
test results showed that creating foam inside the slurry or adding foam on the top of the cake for
the dewatering tests gave approximately same moisture reductions. When Tergitol added on top
of the cake, the moisture again drops from baseline moisture of 38-39 % down to 25-26 %. This
moisture reduction can be attributed to the decrease in the in the surface tension of the liquid that
has to be dewatered. As seen in Figure 5.1 Tergitol is very effective in reducing the surface
tension even at very low dosages.
Figure 5.1 Tergitol NP-7
As explained in the Laplace equation, the cake moisture can be lowered by lowering the
surface tension, increasing capillary radius and contact angle. Tergitol is probably controlling
the surface tension. Addition to this, applying Tergitol in foam method enlarges the capillary
radius, thus lowering the moisture. This phenomenon may also explain the moisture values level
195 |
Virginia Tech | out at very low dosages. As presented in Table 5.8 and Table 5.9 the moisture reduction is
almost same after the first dosages. Increasing the dosage does not change the final cake
moisture.
It should also be noted that, when Tergitol is used as a foaming agent in water, dosage
added on top of the pre-formed cake cannot determined. Both of the techniques of introducing
foam gave significantly low moistures. However, one undesirable point about the method 2 is,
when the foam is added on top, an extra time is given for the foam to travel within the cake.
This extra time increased the total filtration time.
Clay Dewatering Tests at Acidic pH (pH 3-3.5)
As introduced above, clay is usually further treated by chemical bleaching after
separation in order to achieve the required brightness. In industry, the leaching process takes
place in the presence of sodium hydrosulfite at an acidic pH (pH 3) to keep the dissolved iron in
a soluble ferrous state and prevent the formation of ferric hydroxide.
In this series of test, the effect of Tergitol as foaming agent was investigated after the
clay sample was bleached in the similar manner as in industry. The pH of slurry was adjusted to
3 using sulfuric acid, and then sodium hydrosulfite was used as a reducing agent. Alum is also
added to the clay slurry to help coagulation and increase the dewatering kinetics. Table 5.10
shows the dewatering results at pH 3.
196 |
Virginia Tech | Table 5.10 Effect of foam on dewatering of Georgia clay at pH 3 when Tergitol NP-7 is
used as foam generating agent
Reagent Moisture Content (%) pH 3
Dosage
Tergitol (NP-7)
(lb/t)
0 38.04
3 32.87
5 32.63
10 32.14
30 30.84
*Solid Content was 5-6 %. Tergitol was added to the slurry (mixed and aerated to produce foam). Applied pressure 60
Psi., Cake thickness was 1.5-2 mm. with 50 ml slurry.
The cake moisture was increased after the clay sample was bleached at acidic pH.
However, the filtration kinetics was improved at acidic pH because of the coagulation effect.
The cake formation time was lowered down to 5 minutes when the medium is acidic. The cake
moisture can still be lowered to 30% when foam was generated in the slurry using Tergitol as
foaming agent.
A series of tests were conducted to improve dewatering kinetics further by using Alum as
coagulating agent. Alum addition at 5 lb/ton increased the kinetics dramatically (Table 5.11); the
Table 5.11 Effect of alum addition on dewatering of Georgia clay when Tergitol was used
as foaming agent
Moisture Content (%) pH 3
Reagent
Tergitol (NP-7)
Dosage
+
(lb/t)
5lb/t Alum
0 41.10
1 37.01
3 35.48
5 35.26
197 |
Virginia Tech | 5.3. FLY ASH
5.3.1. Background Information
When coal is burned in a power generation facility or in an industrial boiler, it produces
noncombustible mineral matter called ash. This residue is partitioned into bottom ash (or slag),
and fly ash. Bottom ash is coarse in nature and falls to the bottom of the combustion chamber,
where it is easy to collect during routine cleaning. The properties of bottom ash make it a good
road base and construction material, so it can be readily given away or sold. On the other hand,
fly ash is lighter and very fine in nature, and remains suspended in the flue gas which is not so
easily disposed of. Most fly ash is captured by particulate emission control devices, such as
electrostatic precipitators or filter fabric collectors – commonly referred to as baghouses – before
release to the atmosphere. There are also two other byproducts of coal combustion air control
technologies, namely flue-gas desulfurization (FGD) and fluidized-bed combustion (FBC)
wastes. Collectively, all the residue materials are named as coal combustion products (CCP).[11,
12, 13]
In 2004, approximately 64 million metric tons of fly ash was produced, 25.5 million
metric tons of which were used in a variety of engineering applications. Approximately 61% of
the useable fly ash finds its application as an addition to cement in concrete production.
Depending on the intended final use, quality requirements for fly ash vary. For concrete
applications, there are four important characteristics: loss of ignition (LOI), fineness, chemical
composition and uniformity. LOI is especially critical for concrete application. It is a
measurement of unburned coal (carbon) that remains in the fly ash. Basically, carbon content in
fly ash may result in significant air-entrainment problems in fresh concrete, and it may also
unfavorably affect the durability of concrete.[12, 13, 14] At high LOI, carbon can be either
199 |
Virginia Tech | reclaimed or discarded utilizing flotation which is widely demonstrated. Clean fly ash that is
hydrophilic in nature is dewatered to generate a dry product for end use.[12, 14-16]
5.3.2. Two Step Hydrophobization Technique
a) Experimental
Several fly ash samples with various carbon contents were provided for flotation and
dewatering tests. The samples were received from stockpiles at two power plants, one located in
Kentucky and the other in Mexico.
i) Kentucky Fly Ash Samples
• 65 % Ash-35 % Carbon (LOI); (91 % -325 mesh material)
• 100% Fly Ash; (100% -325 material)
Only laboratory dewatering tests were conducted on the above samples, utilizing the two-
step hydrophobization technique previously described.
ii) Mexico Fly Ash Sample
The carbon content (LOI) for this sample ranged from 2.59% to 5.85% with an average
of 4.03% at 10-12 % moisture. The amount of the -100 mesh material was approximately 76%,
the +1 mm was 8%, and the -1mm x 100 mesh was 16%. On this sample a number of flotation
and dewatering tests were conducted to produce clean and dry fly ash product. The tests
included both laboratory and pilot scale tests, described below.
A number of laboratory flotation tests were conducted to lower the LOI using a procedure
that was suggested by the company which supplied the fly ash samples. The samples were
prepared at 25 % solids, and conditioned with the collector (2.67 lb/t) for 8 minutes at 2000 rpm
in Denver cell. This step is designed as a conditioning step. After conditioning, a known
amount of water was added to lower the slurry solid content down to 20%. After frother addition
200 |
Virginia Tech | at about 0.18 lb/t, the flotation was initiated at 950 rpm. After 12 minutes of flotation and
sample collection, another 0.67 lb/t of collector was added as the second step and conditioned for
4 more minutes. Right after the mixing the flotation was continued for another 4 minutes and
bulk samples were collected. Subsequent to sample generation, dewatering tests were conducted
using various types and dosages of chemicals.
Pilot scale tests were conducted on the same sample and tests were run to produce clean
fly ash samples for dewatering tests. The general layout of the pilot scale circuitry includes
• A 12” wide fine sieve to classify the raw feed at nominally 100 mesh. The screen feed
was delivered by a pump and buckets from the barrels
• A conical bottom slurry pump, approximately 250 gallons capacity, received the
underflow from the sieve screen. A mixer was used in the sump to aid in keeping the
solids in suspension
• A centrifugal pump to circulate the slurry from the sump to the upper floor level and back
into the sump using 3” diameter PVC piping.
• A bank of 4, 3 cu ft conventional flow-through type flotation cells utilized as primary
cells. Feed to these cells was supplied by a slip stream taken from the 3” diameter pump
discharge pipe.
• A bank of 4, 3 cu ft conventional cell to cell type flotation cells utilizing the first cell as
conditioner (air valve closed) and the remaining cells as secondary cells.
• Barrels were used to collect the floated product and tails product. The tails product was
flocculated in barrels and the water decanted to thicken the product.
• A 2 sq ft, single disc vacuum filter to dewater the thickened product.
201 |
Virginia Tech | Following the screening process, the slurry was circulated/blended in the sump using the
pump and the mixer for several minutes. The collector was added to the slurry and was
conditioned for 20 minutes (using the pump and the mixer). After conditioning period, water
was added to the sump to dilute the slurry to 20% solids (by weight). The slurry was again
circulated/blended in the sump for several minutes.
The supplier company of the fly ash sample established 16 minutes of retention time in
the laboratory. It was suggested that the scale up to the pilot-scale cell would require 2-2.5 times
the lab cell retention time. Considering the total capacity of the seven pilot scale cells is 20.1 cu
ft (active) or about 150 gallons the slurry feed rate was set to 4 gpm to accommodate a 32-40
minutes retention time.
5.3.3. Results and Discussion
a) Kentucky Fly Ash Samples
65 % Ash-35 % Carbon Sample
Typically fly ash contains low amounts of carbon. However, the sample received from
the Kentucky location contained 35% carbon matter. Considering this, a limited number of
dewatering tests were conducted directly using dewatering aids such as RW, RU and RV,
without first using dosing amine. However, the cake moisture did not change significantly with
Table 5.12 Effect of reagent addition at (1:2) ratio on dewatering of Kentucky coal sample
at 17-19 inches Hg vacuum
Reagent Moisture (%)
Dosage
RW RU RV
(lb/ton)
0 41.81 41.81 41.81
1 41.45 41.97 41.41
2 41.48 41.77 41.44
3 39.99 41.30 40.88
*dissolved in diesel at 1:2 ratio
202 |
Virginia Tech | addition of reagents. Table 5.12 shows the dewatering test results with the addition of various
dewatering aids. RW addition lowered the moisture by only 2%. For filtration tests, a 2.5-inch
diameter, 6-inch tall Buchner funnel vacuum filter was used at 25 in Hg vacuum pressure
This result was not surprising because, despite the presence of carbon, this sample was
primarily comprised of fly ash. Since ash is hydrophilic, it has no affinity for the dewatering
reagents tested. (RW, RU and RV were developed for coal dewatering and are low-HLB
surfactants, designed to adsorb on hydrophobic surfaces, like coal particles.) From these results,
it was clear that the first step of the two-step hydrophobization process would be necessary to
dewater this sample. Amine is a cationic surfactant, and is known to interact with hydrophilic
particles, such as clay.
The second set of tests was conducted using amine. Table 5.13 shows the effect of amine
addition as the first step of the two-step hydrophobization technique. The cake moisture was
reduced from 41.81% to 29% at the addition of 20 lbs/t amine, representing a 30% moisture
reduction. Even though moisture reduction was considerably high, the viscosity was increased
when the amine dosage was more than 10 lb/t and caused handleability problems.
Table 5.13 Effect of amine addition at (1:2) ratio on dewatering of Kentucky coal sample
at 17-19 inches Hg vacuum
Reagent
Dosage Moisture Content (%)
(lb/t)
0 41.81
0.5 41.7
1 38.3
2 41.0
3 38.1
5 35.4
10 33.4
20 29.0
30 28.9
203 |
Virginia Tech | As mentioned above, this is a two-step hydrophobization process and addition of various
dosages of RW was investigated as the second step on the amine treated sample. Table 5.14
gives the laboratory test results obtained on this sample using amine and RW. For this series of
tests, amine dosage was kept constant at 10lb/t. The mixing time for RW was 2 minutes. The
addition of RW as second hydrophobization step increased the kinetics, i.e., the cake formation
time was decreased almost by half depending on the RW dosage. The final cake moisture for
3lb/t RW addition was further reduced to 28.5%, indicating that the two-step hydrophobization
process is superior to the one-step process for both moisture reduction and filtration kinetics.
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
Reagent Moisture Content (%)
Dosage
Amine+ RW
(lb/t)
0 33.4
1 33.0
2 30.5
3 28.5
5 28.0
100% Fly Ash Sample
Another set of dewatering tests was conducted on the 100 % fly ash sample. During the
tests, it was observed that the viscosity of the fly ash slurry was gradually increased with respect
to increased amine dosage. When amine dosages of 5lb/t or greater of were used, the viscosity
was increased to the point that the conditioning of the fly ash slurry became almost impossible.
Also, very high viscosity yielded poor dewatering performance. With respect to handleability,
kinetics and viscosity, the best performance was obtained with 3lb/t of amine addition; however,
204 |
Virginia Tech | Table 5.15 Effect of amine addition as the first step on dewatering of fly ash sample at 20
in Hg Vacuum pressure
Reagent
Dosage Moisture Content (%)
(lb/t)
0 22.6
1 22.13
3 22.44
5 19.41
varying amine dosage did not significantly change the moisture reduction. Table 5.15 shows the
results obtained using different amounts of amine as the first step of hydrophobization.
Subsequent to treating the sample with amine, addition of various dosages of RW, RU
and RV was investigated as the second step. In this series of tests, amine dosage was kept
constant at 3lb/ton. As seen in Table 5.16 the addition of the dewatering aids lowered the
moisture content from 22% down to approximately 17%, giving 22% overall moisture reduction.
RV gave the least moisture reduction, but the kinetics was considerably fast.
Another set of tests were conducted using a 2.5 inch pressure filter. The pressure was
adjusted to 30 psi and the rest of the test parameters were kept similar to the vacuum filtration
tests. As expected, there was not any significant moisture reduction with the amine addition
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
Reagent Moisture (%)
Dosage Reagent Reagent Reagent
(lb/ton) RW RU RV
0 22.44 22.44 22.44
1 22.63 - 23.38
3 18.41 19.17 21.51
5 17.07 17.26 20.60
*dissolved in diesel at 1:2 ratio
205 |
Virginia Tech | Table 5.17 Effect of amine addition as the first step on dewatering of fly ash sample at
30psi pressure
Reagent
Dosage Moisture Content (%)
(lb/t)
0 13.11
1 12.19
3 12.50
5 12.16
(Table 5.17), but increased kinetics was observed. The amine dosage was varied from 0 to 5lb/t
and best kinetics and viscosity were observed at 3lb/t of dosage. Baseline moisture was around
13.11% which was considerably lower compared to vacuum filtration moisture values. This is
directly related to the difference in the pressures applied.
As the second step, dewatering aids, RW, RU and RV were added at various dosages
from 1 to 5lb/t (Table 5.18). The best moisture reduction was achieved using RW which resulted
in a total moisture reduction of 22% moisture reduction followed by RU with a 16.6% reduction;
RV was the least effective in moisture reduction.
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
Reagent Moisture (%)
Dosage Reagent Reagent Reagent
(lb/ton) RW RU RV
0 12.50 12.50 12.50
1 12.52 11.92 11.72
3 11.07 11.13 11.88
5 9.86 10.48 11.74
*dissolved in diesel at 1:2 ratio
206 |
Virginia Tech | b) Mexico Fly Ash Samples
The sample from Mexico contained average carbon content (LOI) of 4.03%. To lower
the carbon content of the fly ash and produce bulk samples for dewatering tests, a number of
laboratory scale flotation tests were conducted. RV was used as collector and Nalco DVS4U013
was used as frother and added as previously described. In this way, the carbon content was
lowered from to an average of 0.74 %. The RV dosage was 2.67 lb/t for the first and 0.67 lb/t for
the second step of the flotation. After enough samples were produced, dewatering tests were
conducted. Amine was used as the first hydrophobizing step at a dosage of 1lb/t. The baseline
moisture values did not change; however, the kinetics was increased in the presence of the
amine. As seen in Table 5.19, the moisture was decreased from 18.2% to 13.2 %, corresponding
to a 27 % overall moisture reduction.
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
Reagent
Dosage Moisture Content (%)
(lb/t)
0 18.20
1 16.10
3 14.70
5 13.20
Similar flotation and dewatering tests were conducted using the pilot scale units. The
flotation tests were conducted using collector dosages of 2lb/t and 0.5lb/t, and frother
(DVS4U013) dosages of 0.75lb/t and 0.25lb/t in the first and second steps of flotation,
respectively. Flotation yielded tailings of 92.7% and the LOI value was 0.68%. After producing
a sufficient amount fly ash sample, a number of dewatering tests were conducted. The first set of
207 |
Virginia Tech | Table 5.20 Effect of using amine (1lb/t) and RV on the dewatering of the fly ash sample at
30% solids
Reagent Moisture Content (%)
Dosage
Test 1 Test 2
(lb/t)
0 17.90 18.50
3 15.20 15.60
3 15.60 15.40
tests was conducted using 1lb/t of DAC and 3lb/t of RV. The amine was mixed for
approximately 4 to 5 minutes, and RV was mixed for 10 minutes at 25% solids. The slurry was
then diluted to 20% solids and subjected to dewatering tests. I n this set of tests, unlike the
laboratory dewatering tests, an adequate particle pick up was not possible for a decent cake
formation.
To be able to increase the cake formation and thickness, after treating the sample with
amine and RV, two sets of tests were conducted in the presence of flocculant (Test 1) and a
combination of flocculant and coagulant (Test 2). The results are presented in Table 5.20. The
filter cake without the dewatering aid had a moisture content of 17.90% when only flocculant
was used. When treated with the dewatering aid at 3lb/t, the moisture was decreased to 15.20%.
In the second test, wherein a flocculant and a coagulant were used along with the dewatering aid,
a moisture content of 15.40% was obtained. As shown, the addition of coagulant increased the
baseline moisture to 18.50%; however, the filtration kinetics and cake thickness were also
increased. The pilot scale results clearly showed that it is possible to lower the final cake
moisture by approximately 15%.
208 |
Virginia Tech | 5.3.4. Foam aided dewatering method
a) Experimental
As previously mentioned foam can be applied in two ways, by adding the foam-
generating agent into the slurry or generating externally and adding on top of the cake. In this set
of test the method 1 employed. The tests were conducted on samples that were obtained from a
stockpile in Kentucky.
b) Results and Discussion
The first set of tests was conducted on the sample which contains 65 % fly ash-35%
carbon matter. These tests were carried out using PPG at various dosages. The foam is
generated in situ and dewatering tests were conducted 20 in Hg vacuum pressure. The results are
presented in Table 5.21. The results showed that in the presence of foam the cake moisture was
dropped down to 32.04 % from 42.03% of baseline moisture %. This decrease corresponds to
24% overall moisture reduction.
It was also observed that at 50 g/t of PPG dosage, the moisture was leveled out and
increasing PPG dosage did not have any impact in final cake moisture. This can be explained by
the reduction in surface tension (Figure 5.2) and drastically increased porosity in the cake
structure.
Table 5.21 Effect of PPG addition on dewatering of Kentucky fly ash sample at 17-19 in
Hg vacuum
Reagent Moisture Content (%)
Dosage
PPG-400
(lb/t)
0 42.03
50 33.06
200 32.66
150 32.04
200 32.22
209 |
Virginia Tech | 5.4. COPPER
5.4.1. Background Information
Copper use has evolved over millenniums, often with the aspects of civilization that
define the times. Used in ancient weapons, as currency, in ornamental fixtures of the
Renaissance, and now as an essential material in building infrastructure and electronics, the
metal has been an important commodity since its discovery. In addition to being malleable and
ductile, copper forms favorable alloys, because it is corrosion resistant, biostatic and easily cast,
and is an exceptional electrical and thermal conductor. Today, copper is nearly exclusively
exploited for its conductivity and resistance to corrosion, and the rapid progression of
construction and technology maintains a high demand for the metal. The copper worldwide
usages in various sectors are: building construction (37 %), electrical components (27%),
industrial machinery (15%), transportation (11%), and consumer products (11 %). [17-19]
Although there are variations on processing methods for copper (e.g. bioleaching, flash
smelting), the most widely used, by far, are the conventional concentrate-smelting-refining
(pyrometallurgy) and leaching-solvent extraction-electrowinning (hydrometallurgy) methods.
Conventional concentrating includes crushing, grinding, screening, flotation, thickening and
dewatering. [17, 20]
5.4.2. Experimental
Copper mineral samples from Sweden were used for dewatering tests in the presence and
absence of dewatering aids. The samples were flotation products and received from preparation
plants in dry form. All tests were conducted at 2 minutes of total dewatering time. This time
included the cake formation and dry cycle times.
211 |
Virginia Tech | 5.4.3. Results and Discussion
It was determined in previous studies that RW is an effective dewatering aid for particles
with hydrophobic surfaces. To investigate its suitability for use on flotation products from
copper processing, a series of dewatering tests were performed on the clean copper samples.
These tests included both pressure and vacuum filtration methods. Table 5.23 shows the copper
dewatering results obtained with RW at varying dosages for tests using vacuum filtration. At 1
lb/ton RW, cake moisture was reduced from 14.98% (baseline) to 12.61% for medium cake
thickness (8mm). At 3lb/ton RW, the moisture was further reduced to 8.08%. When a thicker
cake (14 mm) was used, the baseline moisture was 17.01% and an addition of 3 lb/t of RW
reduced the moisture to 12.39 %.
Table 5.23 Effect of using RW on the dewatering of the copper sample at 20 in Hg
vacuum pressure
Reagent Moisture Content (%)
Dosage
8 mm 14 mm
(lb/t)
0 14.98 17.01
1 12.61 15.91
3 8.08 12.39
A second set of tests was conducted using a pressure filter and the results obtained are
presented in Table 5.24. The baseline moisture levels were much lower in these tests. For
example, with an RW dosage of 3lb/t and a cake thickness of 8 mm, the moisture was reduced
from 8.48 % to 5.07%. At the same RW dosage and a cake thickness of 14mm cake, the
moisture was reduced from 9.66 % to 6.38 %.
212 |
Virginia Tech | 5.5. SUMMARY AND CONCLUSIONS
Dewatering tests were conducted to effectively reduce moisture content in kaolin clay,
copper and fly ash samples. The experimental work included a variety of parameters (e.g.,
vacuum and air pressure, cake thickness, pH, and LOI) and two new dewatering techniques were
utilized: a two-step hydrophobization method using novel dewatering aids, and a foam-aided
dewatering method.
Kaolin clay is one of the most difficult minerals to dewater due to hydrophilic nature and
ultrafine particle size. It has been shown that kaolin clay samples can be dewatered using both of
the new methods. Two-step hydrophobization was tested on samples with various pH values.
During pressure filtration (60psi) tests on Thiele’s clay samples (pH 7), the moisture was
lowered from 39.14% to 31.90% and 28.33% when using amine and Reagent W, respectively.
This corresponds to approximately 28% overall moisture reduction. Using amine and RW
together, the kinetics were also increased such that the cake formation time was lowered from 25
minutes to 3-5 minutes, for a cake thickness of 2-4 mm at 60psi. The results showed that the
slurry pH and the solid content are two very important parameters when using two step
hydrophobization methods.
It was confirmed that the two-step hydrophobization method can also be very effective in
vacuum filtration when the slurry is diluted. For PSD Filter Vat kaolin clay sample diluted to
15% solids, it is possible to lower the final cake moisture by 16% at pH 4.5 and 31% at pH 7.
Foam-aided dewatering may be applied using two different methods. The foam can either
be generated in the slurry or added on top of the filter cake separately. Tergitol was used as the
foam generating agent. The results confirmed that, when the final cake moistures are considered,
both methods of foam application gave the same moisture reductions. However, considering the
214 |
Virginia Tech | filtration time constraints, it was found to be more favorable to generate foam in the slurry. The
foam-aided dewatering test results also showed that very low dosages of Tergitol addition
lowered the final cake moisture by 19% at pH 3 and 35% at pH 7. Also, Alum can be used to
increase the filtration kinetics, even though it may cause a slight increase in final cake moisture.
For example, in one test, cake formation time was lowered from 5 minutes to 50 seconds.
In addition to utilizing the above methods in clay dewatering, similar tests were
conducted on various fly ash samples. These test results have shown that depending on the fly
ash properties, the two-step hydrophobization method is significantly effective in reducing the
final cake moisture and increasing the dewatering kinetics. For samples obtained from
Kentucky, it was possible to lower the overall cake moisture by 24-32 % in vacuum filtration and
11-24 % in pressure filtration.
Dewatering tests utilizing two-step hydrophobization conducted on another fly ash
sample from Mexico confirmed similar results. This sample was treated with 1 lb/t of amine and
3lb/t of RV. The tests were conducted using a 2.5-inch diameter, 6-inch tall Buchner funnel
vacuum filter, operated at vacuum pressure of 25 inches of Hg. The moisture was decreased
from 18.2% to 13.2%, corresponding to a 27% overall moisture reduction. A brief study was
also conducted using a pilot-scale vacuum disc filter. The results showed that a 15-17%
moisture reduction was obtained when the two-step hydrophobization method was used in the
presence of dewatering aids.
The foam-aided dewatering method was tested on various fly ash samples. PPG-400 and
Tergitol were used to produce foam. At low dosages of these foaming agents, the final cake
moisture was lowered by 23-27% when the foam was generated in-situ. This confirms that high
moisture reductions are obtainable with relatively low dosages of foaming agents.
215 |
Virginia Tech | CHAPTER 6 CONCLUSIONS
In the preceding chapters, the results of the dewatering test work have been discussed to
determine the effectiveness of using novel dewatering aids for reducing the moisture of the fine
coal and other mineral products. Overall, the aids can provide an economically feasible means
of dewatering fine particles, often producing more desirable results than can be achieved by
mechanical means alone. Additionally, a completely new dewatering method – foam-aided
dewatering – was developed and tested, also with favorable results. This chapter summarizes the
general conclusions from these investigations.
Laboratory- and pilot-scale dewatering tests were conducted on various fine coal samples
to do engineering evaluations of the novel dewatering aids. The aids were capable of increasing
the contact angle and lowering the surface tension of the water, while increasing the capillary
radius, all of which are responsible lowering the final cake moistures. The evaluation tests were
conducted using different types and dosages of the novel dewatering aids while varying
operating parameters. A Buchner Funnel vacuum filter and an air pressure filters were used for
laboratory tests while a vacuum disc filter (VDF) and a horizontal belt filter (HBF) were utilized
for pilot-scale tests. Laboratory test results showed that the use of dewatering aids can
substantially decrease the final cake moisture by as much as 50%, while increasing the
dewatering kinetics by 30-50%. Due to the improvements in moisture reduction and increased
kinetics, it was estimated that the Mingo Logan Preparation Plant can realize additional annual
net revenue of $1.57 million with a payback time of 3.1 years.
Other results also indicated that adequate conditioning time and intensity are critical to
obtain desired moisture reductions. Thus, mixing should be optimized. In one test, moisture
219 |
Virginia Tech | reduction was around 36% after proper conditioning as compared to 19% when the conditioning
was not sufficient. For this reason, for both laboratory- and pilot-scale test work, conditioning
tanks were used.
A number of tests were conducted using the dewatering aids as collectors where
conditioning was not possible. When the reagent dosage was only 50% of a normal diesel
dosage, it was observed that coal recovery was increased and the moisture was lowered by 7%.
In the pilot-scale tests, when the reagent dosage was the same as diesel, the moisture reduction
was further reduced, giving a 14% overall reduction in addition to increased kinetics.
At the Buchanan coal preparation plant, VA, the use of novel dewatering aids improved
both the moisture reduction and throughput, although the moisture reduction was less than
anticipated. This was primarily due to water chemistry and possibly due to insufficient
conditioning, which may have negatively affected the performance. The plant water had large
amounts of Ca2+ ions, and conditioner was not equipped with a properly-sized impeller.
An empirical scale-up model was developed to predict final moistures under a given set
of operating conditions i.e. the reagent type and dosage, filtration time, vacuum pressure, and
cake weight. The model was for the scale-up of vacuum disc filters and horizontal belt filters.
The model shows that dewatering aids can be most effective when used in conjunction with a
HBF which allows control of cake thickness and drying cycle time independently.
The benefits of using the novel dewatering aids have been demonstrated in full-scale at
the Smith Branch impoundment site. In the full-scale operation, the total moisture content was
reduced from 26% to 20% at a reagent dosage of approximately 0.5 lb/ton, and further to 17.5%
at approximately 1 lb/ton; this represents a total moisture reduction of nearly 33%. The use of
220 |
Virginia Tech | the dewatering aids also reduced the power consumption by 30% for the vacuum disc filter.
Furthermore, the use of the novel dewatering aids improved handlability of the coal.
Two new dewatering techniques were tested for minerals that are hydrophilic in nature: a
two-step hydrophobization method using novel dewatering aids, and a foam-aided dewatering
method. The results showed that the use of these new dewatering methods can effectively reduce
the moisture contents in kaolin clay and fly ash samples
The two-step hydrophobization process was tested on kaolin clay samples at different pH
and solids contents. A pressure filtration test conducted at 60 psi showed that substantial
moisture reductions could be obtained. Using the addition of amine at the first hydrophobization
step and a novel dewatering aid at the second hydrophobization step, the moisture was reduced
and the kinetics were increased. It was confirmed that the two-step hydrophobization method
can also be very effective in vacuum filtration when the clay slurry is diluted.
The foam-aided dewatering method can be implemented in two different ways. The foam
can either be generated in the slurry or added on top of the filter cake separately. Even though
both of the methods are very effective in reducing the cake moisture, when considering filtration
time constraints, it was found that generating the foam in the slurry is more favorable. Test
results showed that significant moisture reductions could be obtained in a wide range of slurry
pH values.
Similar dewatering tests were also conducted on different fly ash samples. These test
results have shown that depending on the fly ash properties, the two-step hydrophobization
method is effective in reducing the final cake moisture and increasing the dewatering kinetics.
221 |
Virginia Tech | Additionally, the foam-aided dewatering method was tested on various fly ash samples. It was
confirmed that high moisture reductions are obtainable with relatively low dosages of foaming
agents.
Novel dewatering aids were also tested on copper samples in both vacuum and pressure
filtration tests. The vacuum filtration test results showed that 30-45% overall moisture reduction
can be obtained depending on the cake thicknesses. The pressure filtration tests showed that, in
the presence of dewatering aids, the moisture can be lowered to 5% and 6.3% at 8 mm and 14
mm cake thicknesses, respectively.
Overall, test results clearly showed that when operating parameters are optimized, use of
novel dewatering aids can generate substantial moisture reductions in mechanical dewatering.
Additionally, the dewatering kinetics can be increased, which in turn should increase the
throughput of filter operations. The dewatering aids can also be used as collectors where
installation of a conditioner is not possible. They not only produce similar or better results than
conventional collectors, but also lower moisture contents. Both the two-step hydrophobization
and foam-aided dewatering methods are capable of reducing the moistures of ultrafine mineral
particles that are difficult to be dewatered using the currently available mechanical dewatering
devices.
222 |
Virginia Tech | FUTURE WORK
In view of the results and conclusions of the experimental work included in this dissertation,
some recommendations regarding potential future work are provided below:
• Novel dewatering aids were successfully tested on fine coal and significant moisture
reductions were obtained. The dewatering aids effectively increased the hydrophobicity
of the coal surface. The same approach could also be applied to coarse coal, i.e., 2 inches
x 2 mm size. Typically, after density separation, the coarse coal is spray-washed and fed
to dewatering screens to remove excess water. At this stage the novel dewatering aids
might be sprayed on the coal travelling on the screen. To increase the effectiveness of
the dewatering aids for coarse coal application, an appropriate surface tension reducing
agent might also be blended with dewatering aid.
• The novel dewatering aids are water insoluble and needs a certain amount of
conditioning. Test results confirmed that conditioning intensity and time are two
important parameters for dewatering aid dispersion and adsorption. In laboratory- and
pilot-scale tests, the proper conditioning was supplied by in-house build mixing tanks.
However, conditioning in this way may not be possible for full-scale applications.
Alternatively, inline mixers might be investigated to properly condition the dewatering
aids for real operations. It would be very beneficial to conduct dewatering tests to
investigate the effect of different types of inline mixers at various solid contents and flow
rates.
223 |
Virginia Tech | • A two-step hydrophobization dewatering technique was successfully tested in the
laboratory on clay samples. The results confirmed that significant moisture reductions
and increased kinetics can be obtained. To be able to confirm the moisture reductions
that were obtained in laboratory tests, pilot-scale tests should be conducted under various
operating parameters.
• In the two-step hydrophobization technique, the first step involved treatment with a
cationic reagent and the second step involved use of novel dewatering aids. Dewatering
aids are effective at relatively low dosages; however, in some cases, the required dosages
of cationic reagent were high. This may not be acceptable for full-scale applications.
Taking this fact into account, a series of laboratory tests should be conducted to
investigate new types of cationic reagents which may be used at considerably lower
dosages.
• Foam aided dewatering tests were conducted on clay samples. The results showed that
this technique can be successfully utilized over a wide range of pH values to dewater the
clay particles. Additional laboratory tests should be conducted using various solid
contents and particle size distributions. Afterward, pilot-scale work would be needed to
better define the effectiveness of the technique.
• While lowering the moisture content of fine particles, dewatering aids have been shown
to also increase the filtration kinetics. This is especially important for coal processing
plants that utilize in horizontal belt filters (HBF), because they are able to take full
224 |
Virginia Tech | RESIDUAL DIESEL RANGE ORGANICS AND SELECTED FROTHERS IN PROCESS
WATERS FROM FINE COAL FLOTATION
Joshua Powell Morris
ABSTRACT
The purpose of this thesis is to examine some of the potential fates of processing reagents
in a coal preparation plant. The focus is specifically on petro-diesel (termed “diesel” in this
paper), which is used as a collector in the flotation of fine coal. Diesel range organics (DRO)
and polyaromatic hydrocarbons (PAHs) are measured in aqueous samples using gas
chromatography equipped with either a flame ionization detector (GC-FID) or a mass
spectrometer (GC-MS). Samples are subjected to a variety of test conditions in order to
understand the behavior of diesel compounds in coal processing streams.
Results show that frother and collector reagents are not likely to partition completely to a
single fraction of the process slurry. Further test work has shown that sub-ppm levels of DRO
dominated by the water soluble fraction of diesel are expected to be present in process waters;
however, PAHs and insoluble DRO may be removed via volatilization and/or degradation. DRO
and PAHs are also expected to be desorbed from coal particles when contacted with fresh water.
Flotation tests have revealed that low levels of DRO are found in both the concentrate and
tailings processing streams with slightly higher concentrations being found in the concentrate
stream. From the tests performed in this thesis, it appears as though there is no apparent
environmental concern when coal preparation plants are operating under normal conditions. |
Virginia Tech | 1 LITERATURE REVIEW
1. Introduction
Coal preparation is often required in order to produce clean, efficient “thermal” coal for
steam generation, or to produce high quality “metallurgical” coal suitable for coking (Pitt and
Millward 1979; Lu et al. 2012). In order to complete this requirement, ash minerals (e.g., sulfur
and ash) must be removed. Unwanted minerals are separated from coal particles through a range
of processes depending on the particle size. Coarse and intermediate sized particles are
separated based on gravity and do not require reagents whereas fine and ultra-fine sized particles
are separated via froth flotation which does require the addition of chemical reagents, as
described below. Following flotation, or other wet processing, excess moisture can be removed
from coal product via dewatering mechanisms such as screens, centrifuges, filters, and
thickeners, and thermal dryers (Bratton and Luttrell).
Since water is used as a continuous phase for particulate processes within the preparation
plant (Adel 2012), it may concentrate residual reagents from various preparation circuits,
including flotation. Flotation feed slurries usually consists of around 5% solids, meaning that for
every ton of feed entering the flotation circuit, roughly 17,000 liters of water is required. Water
is recycled within a preparation facility as much as possible. It can be either recycled within the
plant itself, or re-circulated back from a tailings impoundment after being pumped there with
fine refuse material; ideally, the fine particles settle in the impoundment and the water being
pumped back to the plant is “clear.” While in the impoundment, water, and any chemicals
contained within it, could interact appreciably with the environment (e.g., via release into surface
waters, seepage into underlying soil or vaporization to the atmosphere).
1.1 Environmental Context
Federal water contaminant levels for petro-diesel (i.e., total diesel range organics or
DRO) do not currently exist in the US (EPA 2003); however, some states have specified
contamination levels of DRO for drinking water standards (DEP 2002). As stated above,
environmental concerns with low-level petro-diesel (termed “diesel” in this paper)
concentrations in water are primarily associated with potential PAH release. The US EPA has
1 |
Virginia Tech | developed a list of 16 priority PAHs, one of which (i.e., benzo-[a]-pyrene) does have a maximum
contaminant level (EPA 2011). Benzo-[a]-pyrene was not found in the diesel used in this paper
and only trace levels of six PAHs were detected. No PAHs were detected in the biodiesel used
as well as an obtained sample of pine oil. Unlike diesel, ideally biodiesel does not contain any
aromatics as well (Demirbas 2007). PAHs have been found in the liquid phase of coal waste
slurry; however, it is difficult to determine the origin of these compounds (i.e., naturally or from
pollutants).
Diesel is currently restricted in coal processing under certain conditions, such as when
coal waste slurry is planned to be disposed by underground injection. This ban exists on any
substance that is or contains what is classified as a hazardous waste by toxicity under the
Resource Conservation and Recovery Act (RCRA). In these instances, collectors such as
biodiesel and pine oil represent potential alternatives. Studies have shown that DRO can be
measured in waste slurry, and the source of the DRO is from coal preparation (WV-DEP 2009).
Although diesel is banned from being used where underground injection takes place in West
Virginia, there are no criteria for diesel under the West Virginia water quality criteria for warm
water fishery nor is it subject to a drinking water standard (Ducatman et al. 2010).
1.2 Performance Characteristics
Diesel is commonly used as a collector in fine coal flotation due to its wide-spread
availability, low cost, and well-established performance. Recent laboratory and full-scale results
have shown that alternative collectors, made of renewable and biodegradable materials, can
outperform diesel regarding yield and product ash even at lower dosage levels. A higher yield
corresponds to a lower cost for water clarification as well as a reduction in material heading to
the impoundment. The flotation kinetics of these alternative collectors was also better than
diesel, which would allow for higher throughput (Eraydin et al. 2012). Other studies have shown
that waste vegetable oils can be used to agglomerate coal cleaning fines wastes; however, these
studies used a large amount of oil – 5% by weight (Valdes and Garcia 2005). Alternative
collectors have also been developed from the pulping of trees. The resulting crude tall oil, a
byproduct of pine trees in making paper and tissue, has shown improvements in plant
performance regarding an increase in combustible recovery (Hines et al. 2011). Patents have
2 |
Virginia Tech | recently been placed on the use of fatty acids, the main component of biodiesel, and rosin acids,
regarded as pine oil, as well as a combination of the two (Hines et al. 2009).
2. Research Questions Addressed in this Thesis
Chapter 1: Literature Review
What components of diesel are currently regulated?
Are there any restrictions on when diesel can be used in coal processing?
How does the performance of alternative collectors compare to diesel?
What makes up diesel?
How are trace levels of DRO measured in water samples?
Chapter 2: Reagents In Coal Preparation: Where Do They Go?
What are the potential fates of reagents used in coal processing?
Do frother and collector reagents partition completely to either the solid or
liquid fraction of coal slurry?
Does any frother sorb to the coal surface or does it completely remain in the
water?
Does any diesel remain in the water after it is contacted with coal?
Do residual DRO results differ depending on the dewatering technique (i.e.,
filtration vs. centrifuge)?
Are results reproducible for measuring low level DRO?
What is the effect of percent solids on residual DRO?
What is the effect of ash on residual DRO?
Chapter 3: Diesel Range Organics In Coal Preparation
Are PAHs measured in process water after coal has been contacted with
diesel?
Are PAHs concentrated relative to total DRO in process water samples?
What is the persistence of residual DRO and PAHs once it is in process water?
What is the desorption of DRO and PAHs from coal surfaces after being
exposed to fresh water?
Chapter 4: Flotation Tests
3 |
Virginia Tech | Is DRO measured in process water after flotation?
Does the amount of DRO measured in process water vary between the
concentrate and tailings?
Are naphthalenes measured in process water after flotation?
Does flotation concentrate the naphthalenes found in the process water after
flotation?
How does flotation performance compare with residual DRO?
Chapter 5: Conclusions: Challenges And Lessons Learned
What sample size should be used for measuring residual DRO?
What is the preferred technique for concentrating samples?
What internal standard should be used when measuring residual DRO?
3. Literature Review
3.1 Use of Reagents in Coal Preparation
Chemical reagents are used in coal preparation for particle separation and water
clarification. The primary use of reagents in coal preparation is for froth flotation, a technique
that enhances the hydrophobicity of coal such that the coal will attach to air bubbles and rise to
form a stable froth and be removed while the unwanted minerals sink. Two of the main types of
reagents used in flotation are collectors and frothers. The insoluble collectors coat the surface of
the coal particles, which helps to enhance their hydrophobic behavior, allowing them to float
more easily. The frothers are water-soluble surfactants used to create a stable froth on top of the
flotation cell or column, which collects the floated particles and acts as a mode of transport to
remove them to the concentrate stream. Three of the main groups of frothers include aliphatic
alcohols, polyglycols, and hydroxylated polyethers. The two frothers used in this thesis are
methyl isobutyl carbinol (MIBC), an aliphatic alcohol, and polypropylene glycol (PPG), a
polyglycol (Laskowski 2001). Although frothers are briefly mentioned, diesel collectors are the
focus of this thesis.
4 |
Virginia Tech | 3.2 Petro-Diesel as a Collector
Over 40 years ago, collectors were coke-oven by-products; but these collectors have been
largely discontinued since they contain an abundance of potentially toxic aromatic hydrocarbons.
Most collectors are now derived from crude oil, a more environmentally friendly alternative to
the coke-oven by-products (Laskowski 2001); however, they still contains a small amount of
aromatics (Morris et al. 2013). Common collectors for coal preparation include petroleum
products like petro-diesel (termed “diesel” here), kerosene and fuel oil. Diesel, developed from
the fractional distillation of crude oil, is currently the most commonly used collector for coal
flotation. Diesel is produced all over the world from highly variable crude feed stocks and, as a
result, is subject to significant variability (AMMA).
Diesel is a complex mixture of compounds spanning a range of roughly C to C
10 19
hydrocarbons, and varying not only with the crude oil source, but also with the refining
process(es) (ATSDR 1999). Because of the complexity and variability of diesel, characterization
is difficult; however, broad categories are typically assigned to the diesel range organics (DRO)
(ATSDR 1995). Saturated hydrocarbons (e.g., alkanes, cycloalkenes) usually make up 90% of
diesel by weight, whereas aromatics (e.g., naphthalenes, acenaphthenes, acenaphthylenes) make
up about 10% (Wang et al. 2003). Waxes and resins are typically not quantified since they
account for a very small amount of the diesel. In small volumes or concentrations, saturated
hydrocarbons are not typically considered significant environmental concerns, because they tend
to break down easily and are relatively non-toxic. However the aromatics, specifically polycyclic
aromatic hydrocarbons (PAHs), may be harmful even in small volumes or concentrations and
have received a great deal of attention by the research and regulatory communities over the past
several decades.
The US EPA has developed a list of 16 priority PAHs, which are considered hazardous
due to their potential toxicity, including carcinogenicity, mutagenicity, and teratogenicity. Low
molecular weight PAHs (i.e., 2 or 3 aromatic rings) have a relatively lower toxicity than high
molecular weight PAHs (i.e., five or six aromatic rings) due to their structure (ATSDR 2009).
Of the 16 priority PAHs, 7 are classified as being probably carcinogenic while the remaining 9
are not classifiable, meaning that the carcinogenic levels are significantly lower. Studies on mice
show that the 7 probable carcinogenic PAHs have toxicity equivalency factors, or relative
toxicities, that are 100 to 1000 times higher than the other 9 non-classifiable PAHs (Lee and Vu
5 |
Virginia Tech | 2010). In addition to being potentially toxic, PAHs are also known to be relatively difficult
and/or slow to degrade. Degradation rates vary from a few hours to days in air to several months
to years in soil (SCF 2002). Diesel is known to contain mostly lightweight PAHs (e.g.,
naphthalenes, fluorenes, phenanthrenes) that are typically more volatile and soluble, as well as
easier to degrade, than the higher molecular weight PAHs (Wick et al. 2011).
While ideal (or “parent”) PAHs do exist in diesel and other petroleum products, it is
important to note that modified (e.g., alkylated) PAHs are also present, and generally at much
higher concentrations. These modified PAHs are essentially variations on parent PAHs, whereby
various functional groups have been added (Irwin et al. 1997). The range of possible
modifications is immense, particularly as the number of potential functional group sites increases
with increasing aromatic ring number, which leads to a high degree of variability in terms of
quantity and diversity of individual compounds in any petroleum product. For example, there are
22 individual compounds in just the class of methylated naphthalenes, which is the simplest class
of alkylated PAHs (i.e., only two aromatic rings, and only methyl group additions). In diesel, di-
and tri-methylated naphthalene compounds may be present at about three times the concentration
of pure naphthalene; whereas mono- and tetra-methylated compounds may be present at similar
concentrations to pure naphthalene (Abraham et al. 2005).
Since naphthalene is the most prevalent PAH in diesel, it provides an illustrative example
of the challenges that arise when attempting to characterize specific components of the variable
product. Naphthalene is typically classified based on the number of methyl groups, ranging from
C0-C4 with C-0 being the parent compound (i.e., pure naphthalene) and C1-C4 being alkylated
naphthalenes. It is important to note that since these alkylated naphthalenes are so complex,
making them difficult to quantify, they are often not measured. When they are, naphthalene is
sometimes reported as “total naphthalenes”, which is the sum of the C0-C4 compounds (Irwin
1997).
Studies have shown that the alkylated compounds in diesel account for 93% of total
PAHs. The focus for the work described in this thesis regarding PAHs is on naphthalene because
total naphthalenes make up 80% of total PAHs, with alkylated naphthalenes making up 77% of
total PAHs (Wang et al. 2003). By measuring the aniline point, an inverse relation to the amount
of aromatics, studies have shown that mono-alkylated naphthalene compounds tend to have a
higher aromatic content than the poly-alkylated naphthalene compounds (Hessell 2003; Hourani
6 |
Virginia Tech | 2007). It is expected that the lighter compounds would be more aromatic, and so it is important
to understand the behavior and characteristics of the alkylated compounds since they may behave
slightly different. Since aromatic hydrocarbons are more water soluble than aliphatics (Schein et
al. 2008), it is likely that the mono-alkylated naphthalenes are more water soluble than the poly-
alkylated naphthalenes.
3.3 Biodiesel as a Collector
Biodiesel may represent a potentially “greener” alternative to petro-diesel as a coal
collector. Biodiesel is generally marketed as a renewable energy source made from recycled
vegetable oil or from agricultural co-products and by-products such as soybean oil. It is a fuel
consisting of long-chain mono alkyl esters, also known as fatty acids, designated by the
requirements of ASTM D 6751 (NBB 2013). Biodiesel is made by a process called trans-
esterification, where oil or fat is reacted with a short-chain alcohol such as methanol with a
catalyst, typically sodium hydroxide or potassium hydroxide. The result of reacting 100 pounds
of oil or fat with 10 pounds of alcohol is approximately 100 pounds of biodiesel and 10 pounds
of glycerin (DOE 2013).Typically, when biodiesel is produced for fuel, quality is standardized
by limiting free glycerol and total glycerol where the values must be less than 0.02 and 0.24% by
weight, respectively (Gerpen et al. 2004). However, it is not clear if such standards are
necessarily in force in the context of coal preparation, since quality control measures could be
different depending on the intended use of the biodiesel product.
Environmentally, the main benefits of biodiesel are generally regarded to be its
biodegradability, non-toxicity, and renewability as energy source, making it “carbon neutral”
(RyeBiofuels 2013). Other advantages include lower sulfur content and – particularly interesting
for its use as an alternative coal collector – theoretically no aromatic content (Demirbas 2007).
One of the drawbacks to using biodiesel is its incompatibility with rubber – it will eventually
dissolve the rubber – which presents fouling issues in coal preparation facilities where rubber
components (e.g., tubing or reactor liners) are common. Another mechanical issue with biodiesel
is that it has the tendency to gel at low temperatures, especially biodiesel derived from animal fat
(TriangleBiofuelsIndustries 2007). These mechanical issues mean that physical changes to the
processing circuit may need to be made to switch from diesel to biodiesel, such as replacing
rubber tubing or installing heaters near biodiesel storage tanks.
7 |
Virginia Tech | Since operating conditions will most likely be different when using biodiesel than it is
using diesel, coal operators that have used diesel for many years may be resistant or unwilling to
accept this new change. Another drawback of biodiesel, similar to that of diesel, is variability.
The composition of biodiesel not only varies based on the refining process, but also on the type
of oil or fat that is used (Gerpen et al. 2004). Biodiesel production also has very large water and
land use requirements. A study in China showed that the land footprint of biofuel ranges from 2-
28 m2 of land per liter of fuel produced and the water footprint ranges from 1.5-15 m3 of water
per liter of fuel produced (Yang et al. 2009).
Other alternative collectors are available in addition to biodiesel such as pine-based
products. Regarding production, the benefit of pine oil is that it is a byproduct of paper and
tissue making (Hines et al. 2011). Information regarding production and performance of pine oil
in the context of coal flotation is very limited; however, like diesel and biodiesel, variability in
composition is presumably high, and performance is likely linked quite closely to operating
parameters.
3.4 Analytical Methods
Measuring residual reagents in process water from coal preparation is a difficult
undertaking given the range of potential compounds and relatively low concentrations that may
be present. Thus, it is important to understand the processes by which these reagents are
measured so that the results can be viewed with a better understanding of the challenges leading
up to those results. Gas chromatography is a technique used to separate volatile substances from
one another. In gas-liquid chromatography (referred to as “GC” here), the sample is injected into
the chromatography column with a carrier gas, and the non-volatile solvent-coated column
selectively partitions the sample components; the result is that elution of each component from
the column occurs at a different time. By reference to known, characteristic elution times of
compounds of interest, inference can be made regarding the compounds present in the sample.
To quantify compound proportions in the sample, gas “chromatographs” are utilized. As
the sample elutes from the column, a detector then records the time at which components leave
the column and the result is a chromatograph showing millivolts as a function of time. Ideally,
each unique compound is represented by an individual peak, where the millivolt reading begins
at zero, spikes, and then returns back to zero, granted there is adequate separation between peaks.
8 |
Virginia Tech | for quantification of mixtures. With this being said, studies performed showed that for certain
mixtures (i.e., TPH in soil samples), results appear similar between the MS and FID detectors
(Haddad and MacMurphey 2006).
Two different modes can be used on a MS detector: Full Scan and selective ion
monitoring (SIM). Full Scan mode is used to detect unknown compounds by comparing their
mass to charge ratio over a given time duration (i.e., mass spectra) to that contained in a
computer library. SIM mode can then be used to enhance the sensitivity by 10 to 100 times by
only searching for specific compounds of interest. This is accomplished by searching for two to
four ions per compound and the ratios of these ions represent a unique compound. Used in
conjunction with one another, GC-MS Full Scan and SIM modes are highly effective tools for
identification and quantification of trace compounds (ALS 2008).
3.7 Sample Preparation
Before a sample can be run on a GC-FID or GC-MS it must be concentrated and prepared
in an acceptable solvent (e.g., hexane, but not water). This can be accomplished using a variety
of techniques such as solid-phase micro extraction (SPME), solid-phase extraction (SPE), and
liquid-liquid extraction (LLE) (Koning et al. 2008). The work described in this thesis deals
primarily with SPE; however, LLE is also used. LLE selectively partitions the component of
interest from the rest of the sample using two phases, or solvents. Sometimes multiple steps are
required to increase recovery as well as using an additional solvent(s), which can be costly as
well as time-consuming. The solvent is then evaporated to concentrate the analyte, but the large
sample volume will increase the time required for this step (McDonald 2001). Along with these
problems, LLE also requires the use of breakable glassware. For these reasons, particularly
focusing on the amount of solvent and time required, SPE is the preferred method of separation
(SigmaAldrich 1998).
Solid-phase extraction deals with the separation of a component of interest via two
phases: a solid phase and a liquid, emulsion, gas, or supercritical fluid phase. The work
described in this thesis deals with extraction between a solid and liquid phase. The partitioning
between the two phases occurs when the analyte adsorbs to the solid phase or remains in the
liquid phase. If the analyte adsorbs to the solid phase, a separate solvent is then used to elute the
sample which is then concentrated by evaporation of the solvent. The principle of SPE is similar
10 |
Virginia Tech | to LLE, with the solid phase replacing one of the liquid phases. Problems can arise with SPE
when the sample capacity of the cartridge is exceeded or liquid remains trapped in the solid
phase (McDonald 2001). Flow rate can also impact the separation efficiency by not allowing the
analyte sufficient interaction with the cartridge which can result in a low recovery (Waters
2009).
Different modes of extraction exist for SPE such as reversed phase and normal phase.
Each of these modes utilizes the chemistry of the sample matrix and the analyte of interest.
Reversed phase involves a polar sample matrix, a nonpolar stationary phase, and a nonpolar
analyte whereas normal phase involves a nonpolar matrix, a polar stationary phase, and a polar
analyte. The procedure used in this paper is reversed phase since the samples are in water, a
polar phase, and the analytes, collectors comprised of hydrocarbons, are nonpolar. The
recommended solvent used with this method for extraction is hexane, due to its non-polarity
(SigmaAldrich 1998). The C18 cartridge was used in this paper for this separation mode because
it contains a strongly hydrophobic phase and is recommended for measuring trace organics in
environmental water samples (McDonald 2001).
The steps associated with SPE vary in terminology; however, the process is similar
regardless of the cartridge manufacturer. For the work presented in this thesis, the first step is to
wash the cartridge (i.e., run methanol through the cartridge, followed by hexane.) This process
rids the cartridge of any potential contaminants. The next step is to condition the cartridge with
methanol and then water. This process is to essentially wet the entire surface area of the
cartridge, and without it, results would not be reproducible. The next step is to extract the
sample by running the sample through the cartridge. The sample must be acidified prior to
extraction; this is said to be done in order to prevent microbiological activity (Hach 2009) as well
as to increase the capacity of the cartridge, which will help to avoid sample breakthrough. Most
reversed-phase cartridges have a sample capacity of up to 100 mg, much larger than the sub-ppm
levels observed in this testing (McDonald 2001). Finally, the last step in SPE is to elute the
sample with hexane and concentrate by evaporation.
Sample size is one of the more difficult decisions for sample preparation by SPE. Based
on experience gained through the research reported here, it is recommended that a sample size of
200 mL be used for future work, particularly for flotation test samples. A larger sample size
results in better detection and quantification of analytes; however, the tradeoff is the time it takes
11 |
Virginia Tech | to prepare these samples. Fine particles may be able to pass through filter paper during
dewatering, but the particles can become trapped in the cartridge used for SPE which makes it
more difficult for liquid to pass. The pump pressure can be increased, but the rate at which
liquid passes through the cartridge becomes increasingly difficult to maintain between cartridges.
For these reasons, 200 mL samples have been determined to be a proper balance between
tradeoffs.
At least one blank should always be extracted during SPE, but two blanks are
recommended. Along with extracting a blank, it is advised that a standard solution (i.e., known
concentration of diesel) be extracted as well in order to determine the extraction efficiency. In
some tests reported here, a surrogate was also used with the intent of determining the extraction
efficiency. However, it was determined that this method was not desirable for two reasons: 1)
the surrogate chromatogram results interfered with the DRO results, and 2) without further
testing, it is unclear whether or not the surrogate extraction efficiency corresponds to the DRO
extraction efficiency.
When analyzing samples by GC, it is important to run hexane between each set of
samples to ensure that the GC is clean and there is no carryover between samples; the hexane
essentially purges the column of contaminants and resulting chromatograms can be utilized to
confirm that compounds which may interfere with sample analysis are not present. If there are
multiple GC users, and particularly if a variety of sample types are being analyzed on the same
column, it is advised that the hexane purge be performed twice at the beginning of a sample run.
It is also important to add an internal standard to each sample prior to analysis on the GC.
While sample area can be used to compare results to one another from the same GC run,
it is also advisable to consider the internal standard since it can also be used to compare results
from one sample run to another. If any variation of the equipment exists, the sample area
relative to the internal standard area should remain consistent. For the work in this thesis, 20 µl
of phenol-d6 was used as the internal standard. If an internal standard is not available the sample
area can still be analyzed; however, this can reduce confidence in results, particularly when
comparing one set of samples to another in cases where any conditions might change within the
GC between runs.
12 |
Virginia Tech | 3.8 Data Analysis and Challenges
Data analysis is probably the most difficult aspect of gas chromatography. It is important
to achieve adequate separation efficiency between compounds if individual compounds are to be
measured. Separation efficiency is dependent upon the column type, length, and temperature,
and carrier gas and airflow rate (UMass ; UniversityofWashington). The sample size and
injection time can also affect the results of the chromatograph, creating taller and wider peaks for
larger samples and longer injection times, respectively (McNair and Bonelli 1968). A 30 m x
0.25 mm ID x 0.25 um film thickness Rxi-5sil MS with 5% diphenyl and 95% dimethyl
polysiloxane column was used for the work described in this thesis. The FID uses hydrogen at
45 mL/min, air at 450 ml/min, and nitrogen as the make flow at 30 ml/min, while the carrier gas
has a flow rate of 1 ml/min. The injector temperature is 250 oC and the GC begins timing at 40
oC and after three minutes increases by 10 oC per minute until reaching a final temperature of
300 oC after 29 min. The GC maintains this temperature for another six minutes and then the
program ends.
The area under each peak is calculated by the computer based on the user defined
baseline. Either a baseline can be drawn for each individual peak as shown in Figure 1.1 or one
baseline can be drawn as shown in Figure 1.2. The challenge with drawing one baseline as
shown in Figure 1.2 is that sometimes the peaks do not return to one baseline, so this can create
user variability. To quantify the example shown, the total area calculated in Figure 1.1 is 5.75
mV min whereas the total area calculated in Figure 1.2 is 11.15 mV min. The advantage of
● ●
drawing a baseline for each peak is that the variability is removed; however, this increases time.
A method can be set up on the GC to draw the baseline for each peak, which removes the
variability and ultimately decreases time required if many samples are going to be analyzed. The
method parameters for this paper inhibit integration from occurring until after 10 minutes, and
then the minimum area to be considered is 0.01 mV min. The purpose for not integrating until
●
after 10 minutes is so that neither hexane nor methylene chloride are calculated in the total area
and results show that under the GC operating conditions, DRO do not elute from the column
until after 10 minutes. The purpose for not considering areas less than 0.01 mV min is to
●
eliminate noise. The data in Chapters 2 and 3 were calculated using the user drawn baseline, but
then after learning about the program method technique; this was used for calculating the data in
Chapter 4.
13 |
Virginia Tech | Internal Standard
Figure 1.3: Chromatogram showing the internal standard and diesel
Chromatographs are qualitatively analyzed based on retention times, since each
compound has a unique retention time. Typically a library is used to match compounds with the
MS spectra. The best way to quantitatively measure a mixture of compounds is to first develop a
calibration curve based on the ratio of the peak areas of the known concentration mixture to that
of the internal standard. Then when a mixture’s concentration is unknown its ratio of peak area
to internal standard is compared to the calibration curve to determine a measured concentration
(Jeffery et al. 1989). If an internal standard is not present, the peak area of the unknown mixture
can simply be compared to the calibration curve based on the peak areas for mixtures of known
concentrations. This is not the preferred method since the internal standard is used to help
alleviate any changes that may exist within the GC between runs.
4. Conclusions
Water quality is an issue of increasing concern for the coal mining industry. Although
processing reagents have not previously been scrutinized for environmental transport and fate, it
is an important topic requiring further research for reasons previously stated, particularly
regarding PAHs. While external research is underway regarding the performance of renewable
and biodegradable collectors, the tests described in this thesis for diesel as a collector can be
used as a guideline for future test work using these new, alternative collectors. The analytical
techniques described in this chapter can also be adapted to test for other organic processing
reagents in water samples.
15 |
Virginia Tech | 2 REAGENTS IN COAL PREPARATION: WHERE DO THEY GO?
Josh Morris, Emily Sarver, Gerald Luttrell, John Novak
Paper peer-reviewed and originally published in proceedings of the 51st Annual Conference of
Metallurgists (Canadian Institute of Mining, Metallurgy and Petroleum), October 1-3, 2012.
Niagara Falls, Ontario, paper 7391. Reproduced with permission of the Canadian Institute of
Mining, Metallurgy, and Petroleum. www.cim.org
1. Abstract
A variety of reagents are utilized in coal preparation, but aside from performing their
desired function relatively little is known about the behavior of these reagents within the
processing circuits. Where exactly do reagents go once dosed? In this paper, we present
preliminary results of partitioning studies on frother (i.e., MIBC) and collector (i.e., petro-diesel)
chemicals commonly used in coal flotation, and examine implications for water management
(e.g., in closed-loop systems). Additionally, we discuss the usefulness of such data in predicting
environmental transport and fate of chemicals – which is currently a top priority for industry.
2. Introduction
The purpose of coal preparation is to upgrade mined coal into more valuable products.
Since coal is primarily used as a fuel source for electricity generation, product specifications are
typically contracted to minimize unwanted constituents that detract from the overall heat value
(e.g., ash and moisture) or that add to environmental pollution or other problems like corrosion at
a power plant (e.g., sulfur) (Pitt and Millward 1979). Failure to meet specifications results in a
financial penalty for the coal producer (Szwilski 1986), and thus preparation processes have
evolved to simultaneously optimize recovery of valuable “clean” coal with rejection of mineral
matter and moisture. In addition to advancements in equipment and circuitry, development and
application of various chemical reagents has dramatically improved the performance of coal
preparation processes.
Contemporary preparation plants typically include multiple circuits that can be
categorized by the size of particles they process: coarse, intermediate, and fine/ultra-fine (Figure
21 |
Virginia Tech | 2.1). Coarse and intermediate circuits generally rely on size classification and gravity separations
(e.g., dense-media cyclones), and do not require significant chemical reagents. However, fine
and ultra-fine circuits often use froth flotation to separate coal from impurities, which requires
chemical additives (Table 2.1). The primary additives include collectors, which coat the surface
of the coal particles to render them (more) hydrophobic and thus more likely to attach to air
bubbles and float; and frothers, which aid in the formation and stability of the froth that will
accumulate the floated coal particles. Modifiers are also commonly added to flotation circuits to
regulate pH in instances where coal or impurity characteristics may change water chemistries
(Laskowski 2001). Following flotation, coagulants and flocculants are often utilized in solid-
liquid separations (i.e., dewatering or clarification) for coal products, and for tailings slurries
prior to their disposal in impoundments. Coagulants function via double-layer compression1 to
bring colloidal particles together, while flocculants promote bridging between the grouped
colloids – and the combined result is enhanced sedimentation (Wills 2006). Defoaming or anti-
foaming agents may also be required to avoid fouling of dewatering operations.
1 Double-layer compression refers to the action of added ionic species on the electrical double layer surrounding a
colloid or fine particle. In the case of negatively charged coal, the addition of a cationic coagulant effectively
reduces the (repulsive) electrostatic forces between particles such that Van Der Waals’ forces may attract the
particles together Scott, J. H. (1976). Coagulation Study of a Bound Water Bulked Sludge. Master of Science,
Virginia Polytechnic Institute and State University..
22 |
Virginia Tech | The goal of this paper is to begin answering these questions. The following sections
review the potential fates and impacts of coal preparation reagents, and present preliminary data
regarding the partitioning of frothers and collectors between coal and process water.
Table 2.1: Common reagents in coal preparation (McIntyre 1974; Knapp 1990; Pugh 1996;
Laskowski 2001)
Type Group Reagent
Fuel Oil No. 1 - Kerosene
Collectors Hydrocarbons Fuel Oil No. 2 - Diesel
Fuel Oil No. 6
Aliphatic Alcohols Methyl Isobutyl Carbinol (MIBC)
Polyglycols DF 250
Dowfroth M150
Frothers
Hydroxylated Nalco 8836
Polyethers Polyoxyl Sorbitan Monolaurate
(PSM)
NaCl
Promoters CaCl
2
Modifiers Na SO
2 4
H SO
pH Regulators 2 4
CaO
Organic Starches
Coagulants (cationic) Inorganic Salts
Polyamines
Dewatering/Clarification Flocculants (non- Organic Starches
Reagents ionic) Polyacrylamide
Organic Starches
Flocculants (anionic) Acrylamide/Acrylate Copolymers
Polyacrylates
Tributyl Phosphate (TBP)
Defoaming Reagents Defoamers
Polydimethylsiloxane (PDMS)
3. Reagent Fates and Implications
Determining the fate of coal processing reagents necessitates tracking those reagents
from their addition points in a preparation plant (e.g., Table 2.1) to some ultimate destination.
Based on a simple materials balance approach, only a fixed number of possibilities exist for
reagents leaving the plant: they may end up with the clean coal products, with the tailings by-
products, or with recycled water, or they may be lost (e.g., via volatilization or spills).
24 |
Virginia Tech | 3.1 Environmental Fate and Transport
The environmental fate and transport of processing reagents has been scarcely examined.
It is generally expected that collectors (e.g., petro-diesel) substantially partition to coal products
because their chemistry promotes sorption to the coal particles (Watts 1998). Any collector that
does not sorb may remain with water, either floating on the water surface, as an emulsion, or as a
dissolved species – although water solubility is likely low. Frothers, on the other hand, are not
expected to significantly sorb to coal (or other solids), and thus should follow water streams.
Alcohol-based frothers like MIBC tend to have relatively low water solubility and low to
moderate volatility (Howard 1993), which indicate that they may remain at the water-air
interface; whereas glycol-based frothers like Dowfroth M150 are much more soluble in water
and are relatively non-volatile. Coagulant and flocculant reagents are of course expected to
partition to fine coal or tailings particulates, at least in the short-term. These chemicals may well
remain with dewatered coal products; but in the case of reagents associated with tailings solids, it
is difficult to predict how they might react or mobilize under disposal facility conditions.
Reagents that partition to coal products are likely to be combusted with the coal – unless
they volatilize during handling and transport. The combustion by-products of the reagents may
enter the atmosphere as either gaseous or particulate emissions, which may then be returned to
the earth via either wet or dry deposition. In the case of petro-diesel collector (termed “diesel” in
this paper), for example, it is expected that much of the alkane fraction2 will be completely
combusted and converted to carbon dioxide and water; however, PAHs that occur naturally in
the diesel or that form as a result of incomplete combustion might also be released.3 In addition
to atmospheric emissions, reagents or combustion by-products of reagents might become part of
the solid fly ash (i.e., waste from coal combustion) and eventually disposed (e.g., in landfills),
either because the reagents were associated with the mineral fraction (i.e., noncombustible) of
the coal or because their aerosols were scrubbed from flue gases. In the example of diesel
2 Diesel is not a specific compound, but rather a range of compounds collected from fractional distillation of
petroleum (i.e., between 200-400 °C). Its general composition includes primarily moderate weight alkanes (i.e., C -
15
C ), and also cycloalkenes and polyaromatic hydrocarbons (PAHs) Watts, R. J. (1998). Hazardous Wastes: Sources,
25
Pathways, Receptors. New York, NY, John Wiley and Sons..
3 PAHs are an environmental concern because they pose human and ecological health risks ATSDR. (2009). 2012..
However, the bioavailability of PAHs derived from diesel combustion is not well understood Scheepers, P. and Bos,
R. (1992). "Combustion of diesel fuel from a toxicological perspective. II. Toxicity." Int Arch Occup Environ
Health 64(3): 163-177..
25 |
Virginia Tech | collector that partitions to coal products, this is another likely scenario for some PAHs (Liu et al.
2008). Following atmospheric deposition or disposal of fly ash, coal processing reagents or their
by-products could move through terrestrial and aquatic ecosystems via hydraulic or biologic
transport processes.
For reagents that partition to either the water or solid fractions of coal tailings,
environmental fate and transport is heavily dependent on the tailings disposal conditions. If
tailings are disposed via underground injection, reagent fate will be governed by chemical
conditions of the storage cavity (i.e., atmosphere, water chemistry, and wall rock mineralogy);
and reagent transport will depend on the degree to which groundwater interacts with the cavity.
More often, tailings are disposed above ground in impoundments or ponds, where the water
fraction is expected to clarify as the solid particles slowly settle. Some of the water is generally
recycled back to the preparation plant and used as make-up water, but a portion of it is released
to the environment via evaporation, engineered discharges (i.e., through decant structures or
spillways) (MSHA 2009), or percolation to the subsurface since impoundments for coal refuse
are rarely lined (USEPA 1999). If reagents or reagent by-products are present in impoundments,
water releases could possibly mobilize them. Other possibilities include photo- or bio-
degradation within the impoundment (e.g., MIBC), or sorption to soils beneath the impoundment
(e.g., diesel).
In the context of environmental fate and transport, it is also important to note that coal
processing reagents are seldom pure products with constant composition. For instance, diesel can
vary with the properties of the petroleum feedstock used to produce it, and some frother reagents
are actually acquired as by-products from the manufacture of other products (e.g., brake fluids).
While variability in reagent quality will not be discussed in detail here, it is a topic that deserves
further attention.
3.2 Residuals in Operations
In addition to tracking processing reagents to better understand environmental
implications, it is becoming increasingly important to understand implications for preparation
plants that utilize large volumes of recycled water. Use of closed water systems (i.e., zero
discharge from site) is growing in response to calls for both water efficiency and water resource
protection. For coal preparation facilities, such systems generally combine the plant and tailings
26 |
Virginia Tech | water circuits, such that “clear” water from an impoundment is recycled back to the plant as
makeup water. Water may also be recycled within the plant (e.g., from the coal product thickener
back to cyclone or flotation circuits).
To the extent that processing reagents (or their by-products) remain in the recycled water,
chemical concentration may have significant impacts on plant operation. While residual
chemicals could potentially reduce the rate of new chemical addition in some cases, it is also
possible that reduced efficiency or fouling of some unit processes may occur. For example,
residual frothers may impact processes that cause significant agitation (e.g., dense media cyclone
separations) (Lahey and Clarkson 1999), or where water chemistry promotes foaming (e.g.,
where recycling has caused increased salt concentrations). Even at sites where only a portion of
water is recycled throughout the plant, it is already well established that such problems lead to
preventative under-dosing of frother in flotation circuits, which in turn sacrifices recovery of fine
coal (Coffey and Lambert 2008). For closed water systems, the implications may be far more
significant, and additional water treatment efforts might be required to maintain efficient
operations.
In light of the environmental and operational implications of processing reagent fates, it
is important to understand how they partition between solid and liquid fractions in preparation
plants.
4. Experimental Methods
Partitioning studies were carried out to obtain preliminary data on the potential fates of
common frother and collector reagents for fine coal flotation4. The frothers were MIBC,
polyoxyl sorbitan monolaurate (PSM), Dowfroth M150, and Nalco 8836, and the collector was
diesel. Raw coal samples were ground using a laboratory hammer mill, and sized by wet
screening for the desired test conditions (Tables 2.2 to 2.4). Full proximate analysis was not
conducted on any of the raw coal samples, however approximate ash contents were determined
(see below). For each test, a slurry sample was prepared by adding the required weight of sized
raw coal to distilled water, followed by the required volume of reagent. Slurries were mixed for a
4 The frother partitioning tests were partially reported in an MS thesis (Knapp, 1991), but have not been published
elsewhere.
27 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.