WATER POLLUTION CONTROL RESEARCH SERIES • 14010 EJT 09/71
Studies on Densification of
Coal Mine Drainage Sludge
ENVIRONMENTAL PROTECTION AGENCY • WATEK QUALITY OFFICE
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe
the results and progress in the control and abatement
of pollution in our Nation's waters. They provide a
central source of information on the research, develop-
ment, and demonstration activities in the Environmental
Protection Agency, Water Quality Office, through inhouse
research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.
A tripilicate abstract card sheet is included in the
report to facilitate information retrieval. Space is
provided on the card for the user's accession number
and for additional uniterms.
Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Head, Project Reports
System, Office of Research and Development, Environmental
Protection Agency, Water Quality Office, Washington, D. C.
20242.
-------
Studies on Densification of
Coal Mine Drainage Sludge
Bituminous Coal Research, Inc.
350 Hochberg Road
Monroeville, Pennsylvania 15146
for the
ENVIRONMENTAL PROTECTION AGENCY
Project #14010 EJT
September 1971
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402- Price $1.25
Stock Number 5601-0101
-------
EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does not signify that the contents necessarily
reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recommen-
dation for use.
-------
ABSTRACT
The purpose of this research was to alleviate present problems in han-
dling and disposal of sludges obtained by lime neutralization of coal
mine drainage, through the investigation of various sludge densifica-
tion techniques. The scope of the work was restricted to bench-scale
batch experiments. Tests were largely of an exploratory nature and, as
such, did not afford sufficient data to permit detailed cost comparisons
among the various techniques.
In the first approach, conditions commonly employed in the lime treat-
ment procedure (lime neutralization and aeration) were altered to pro-
duce a dense, fast-settling, ferromagnetic sludge. Although the proper-
ties of this sludge are sufficiently unique, the magnetic sludge
conversion process includes a requirement for sludge heating and is
quite sensitive to the presence of small amounts of aluminum in the
original mine water.
In the second approach, recognized conditioning methods were applied to
sludges obtained by the usual treatment procedure involving lime
neutralization and aeration. These methods included the use of coagu-
lant aids, sludge bulk additives (filter aids), seeding materials, and
sludge heating and freezing.
In addition, exploratory tests were conducted on the introduction of
carbon dioxide into the mine water to promote coprecipitation of
calcium carbonate during lime addition.
Among the sludge densification methods tested, only three--magnetic
sludge preparation, sludge freezing, and C03 pretreatment--appeared to
be promising in terms of results obtained. Each method led to sludge
volume reductions on the order of 90 percent and increases in sludge
solids contents of from 0.5 to about 5 percent after SO-ininute settling
periods. Further developmental work would be desirable to determine
the general usefulness and comparative costs of each method as applied
to coal mine drainage treatment.
This report was submitted in fulfillment of Program No. I^OIO EJT
under the joint sponsorship of the Water Quality Office of the
Environmental Protection Agency, the Commonwealth of Pennsylvania, and
the coal industry through its research agency, Bituminous Coal Research,
Inc.
Key Words: Coal mine drainage, sludge, sludge conditioning, magnetic
sludge, iron removal, sludge densification, sludge disposal,
stream pollution, pollution abatement, acid mine water
treatment.
111
-------
CONTENTS
Section Page
I Conclusions 1
II Recommendations 5
III Introduction 7
Nature of the Problem 7
Objectives and Scope of the Research 7
IV Experimental Procedure 11
Materials, Equipment, and Analytical
Techniques 11
Magnetic Sludge Studies 1^-
Sludge Conditioning Studies 17
V Results and Discussion 23
Magnetic Sludge Studies 23
Sludge Conditioning Studies 63
VI Acknowledgments 109
VII References Ill
-------
FIGURES
No.
1 Schedule of Work - Recovery of Iron in a More Dense Form
from Coal Mine Drainage 10
2 Magnetic Response of Various Ferromagnetic Materials with
Increasing Sample Weight 16
3 Magnetic Response and Settled Sludge Volume Percent as
Functions of the Mg/Fe Molar Ratio 36
k Magnetic Response and Settled Sludge Volume Percent as
Functions of the Al/Fe Molar Ratio UO
5 Magnetic Response and Settled Sludge Volume Percent as
Functions of the Al/Fe and Mg/Fe Molar Ratios 1*1
6 Variation in Magnetic Response with Increasing Al/Fe and
Mg/Fe Molar Ratios ^
7 Results of Seeding Tests with Keystone Magnetic Sludge .. U8
8 Conceptual Design of Mine Drainage Treatment Process for
the Formation of Magnetic Sludge 5^
9 Relative Sludge Volumes After a Thirty Minute Settling
Period 55
10 Schematic Representation Showing Amenability of Coal Mine
Discharges from Two Watersheds to Magnetic Sludge
Treatment 6l
11 Effect of Various Anionic Coagulant Aids on Zeta
Potential of Synthetic Coal Mine Drainage Sludge 65
12 Effect of Various Anionic Coagulant Aids on Zeta
Potential of Synthetic Coal Mine Drainage Sludge 66
13 Effect of Zeta Floe WA on Zeta Potential of Synthetic
Coal Mine Drainage Sludge 68
Ik Effect of Two Nonionic Coagulant Aids on Zeta Potential
of Synthetic Coal Mine Drainage Sludge 69
15 Effects of Filter Aids Alone on Solids Content of
Synthetic Coal Mine Drainage Sludge 79
VI
-------
FIGURES (Continued)
No. Page
16 Effects of Filter Aid Concentration on Filtration
Properties of Synthetic Coal Mine Water Sludge 85
17 Effect of Freeze-dewatering on Volume of Synthetic Coal
Mine Drainage Sludge 98
VI1
-------
TABLES
No. Page
12 Results of Seeding Tests with Magnetic Sludge from
Keystone Mine Drainage
1 Representative Compositions of Coal Mine Waters Used
in Sludge Densif ication Studies ........ . ................ 13
2 Results of Factorial Experiment on Variables Affecting
Magnetic Sludge Formation .............................. 2k
3 Codified Design and Calculation Matrix ................. 25
k Main Effects and Interactions Based on Data from
Magnetic Sludge Factorial Experiment ................... 26
5 Comparison of Results Obtained by Partial Replication of
Magnetic Sludge Factorial Experiment ................... 29
6 Revised Calculation of Main Effects and Interactions
Based on Data from Magnetic Sludge Factorial Experiment 30
7 Main Effects and Interactions for the 2s Factorial
Experiment Involving High Temperature Tests Only ....... 32
8 Effect of the Mg/Fe Molar Ratio on Magnetic Sludge
Properties ............................................. 35
9 Effect of Increasing Mg/Fe Molar Ratio on Chemical
Composition of the Magnetic Sludge ..................... 38
10 Effect of the Al/Fe Molar Ratio on Magnetic Sludge
Properties ............................................. 39
11 Effect of Increasing Al/Fe Molar Ratio on Chemical
Composition of the Magnetic Sludge ..................... k-2
13 Results from Initial Tests on Preliminary Sludge
Concentration .......................................... 50
lU Results of Additional Preliminary Sludge Concentration
Tests During Magnetic Sludge Preparation Studies ....... 51
15 Effect of Coagulant Aids on Sludge from Lime Treatment
of WQO-EPA Synthetic Coal Mine Water ................... 6k
16 Effects of Coagulant Aids on Settling Rates, Sludge
Volumes, and Solid Contents of Synthetic Coal Mine Water
Sludge ................................................. 70
viii
-------
TABLES (Continued)
No. Page
1? Materials Used in Filter Aid Studies ................... 75
18 Effects of Filter Aids Alone on Properties of
Synthetic Coal Mine Water Sludge ....................... 76
19 Effects of Filter Aids Plus 0.5 ppm Calgon 2^0 on
Properties of Synthetic Coal Mine Water Sludge ......... 78
20 Effects of Various Filter Aids on the Filtration
Properties of Synthetic Coal Mine Water Sludge ......... 82
21 Effects of Filter Aid Concentration on the Filtration
Properties of Synthetic Coal Mine Water Sludge ......... 8k
22 Synthetic Coal Mine Water Seeding Experiments - Effect
of Seed Crystal Size ................................... 86
23 Synthetic Coal Mine Water Seeding Experiments - Effect
of Seed Crystal Amount ................................. 88
2k Synthetic Coal Mine Water Seeding Experiments - Effect
of Stirring Rate ....................................... 89
25 Results of Synthetic Coal Mine Water Seeding Experiments
with Dried Sludge Samples .............................. 91
26 Effect of Heating Mine Water Before Lime Addition on
Behavior and Properties of Synthetic Coal Mine Water
Sludge ................................................. 92
27 Effect of Heating on Behavior and Properties of
Synthetic Coal Mine Water Sludge ....................... 9^
28 Effect of Heating and Subsequent Cooling on Behavior
and Properties of Synthetic Coal Mine Water Sludge ..... 95
29 Results of Freeze- dewatering Experiments with Actual
Coal Mine Waters ....................................... 99
30 Effect of COS Addition on Lime Requirement and Sludge
Properties During Lime Treatment of Synthetic Coal Mine
Water .................................................. 102
31 Effect of C02 Sparging Period on Lime Requirement and
Sludge Properties During Lime Treatment of Synthetic
Coal Mine Water ........................................ 10k
IX
-------
TABLES (Continued)
No.
32 Effect of Dissolved C02 Content of Sludge Properties
During Lime Treatment of a Carbonated Mine Water 106
33 Effect of COS Sparging in Conjunction with Coagulant Aid
Addition on Sludge Properties During Lime Treatment of
Synthetic Coal Mine Water 107
x
-------
SECTION I
CONCLUSIONS
Studies on coal mine drainage sludge densification have led to the
following conclusions:
1. Coal mine waters containing iron principally in the ferrous (Fe3"1")
state can be treated by a modification of the commonly practiced lime
neutralization process to yield a ferromagnetic sludge. The sludge,
presumably composed in large part of a hydrated ferrosoferric oxide
(Fe304-xBgO), is much denser and faster settling than the ordinary
"yellowboy" sludge obtained by lime neutralization and aeration at
ambient temperatures.
2. The magnetic sludge preparation technique involves (quantitative)
precipitation of iron (largely as Fe(OIi)2) by the addition of hydrated
lime to a pH of 8.5, separation of the resulting sludge from the
aqueous phase, and heating of the sludge concentrate to at least 80 C
(176 F), preferably higher, followed by mild aeration.
3- Magnesium, present as Mg3* in most coal mine drainage waters,
inhibits the formation of the magnetic sludge. This inhibition can be
adequately minimized by keeping the pH below 9-6 during lime addition
(thus avoiding the coprecipitation of Mg(OH)2) and by separating the
sludge from the bulk of the aqueous phase before the heating step, as
indicated in (2) above.
k. Aluminum, commonly present in coal mine drainage waters, also
inhibits the formation of the magnetic sludge. This inhibition is of
practical significance when the molar ratio of aluminum/total iron is
greater than 0.18. Below this value, sludges are still appreciably
magnetic, but a,re bulkier and slower settling than those obtained when
aluminum is absent. Unlike magnesium inhibition, aluminum inhibition
cannot be minimized by the scheme indicated above, since aluminum
precipitates as Al(OH)3 near pH k and remains dispersed within the
sludge.
5. The limitation on aluminum content is probably the most serious
restriction affecting the feasibility of the magnetic sludge process as
presently conceived. Nevertheless, a limited review of available water
quality data has indicated that, in terms of relative acid and iron
loadings, a significant number of major coal mine discharges in the
eastern Ohio River basin may be amenable to treatment by a magnetic
sludge process.
6. Data from tests on the use of magnetic seeding materials indicate
that seeding did not enhance the formation of a magnetic sludge under
the experimental conditions employed.
1.
-------
7. Based on the available experimental evidence, the magnetic sludge
process appears to be technologically feasible. Preliminary cost con-
siderations indicate that costs for sludge heating in the process
would not be prohibitive.
8. The use of coagulant aids alone, of the anionic or nonionic
polyelectrotyte types, will not result in an increase in the density of
sludges obtained by lime neutralization of (synthetic) coal mine drain-
age. However, considerable increases in sludge settling rates and
decreases in turbidity of the supernatant liquid are possible through
the use of coagulant aids.
9. The use of filter aids or sludge "builders" such as fly ash,
magnetite, "red dog," sand, blast furnace slag, sawdust, or gypsum can
result in increased sludge densities and faster settling rates. The
use of filter aids in combination with a coagulant aid appears to offer
several advantages over the use of filter aids alone, such as faster
settling rates, lower supernatant liquid turbidity, and a reduced
coagulant aid requirement.
10. Among all the filter aids tested, only blast furnace slag and
gypsum differed from the others in their proportionate ability to
increase sludge solids contents as their concentrations increased.
The blast furnace slag appears to offer certain added benefits as a
sludge densifier. It is a relatively cheap waste product which also
contains a reactive alkaline component, thereby reducing the lime re-
quirement during neutralization of the mine water-
11. Filtration tests failed to show any dramatic improvements in
sludge dewaterability when the filter aids were added as a "body feed."
However, differences in behavior among the various materials evaluated,
as well as effects due to filter aid concentration, indicate that some
of the cheaper, more readily available materials might be successfully
employed in a sludge densifying/dewatering application.
12. Tests involving the use of anhydrous ferric oxide, as well as
freeze-dewatered and ordinary dried sludges (obtained by lime neutral-
ization of synthetic coal mine water) as seeding materials failed to
show any strong dependency of sludge properties on either the size of
the seed crystals or the stirring rate employed, within the ranges
50 to 325 mesh ajad 300 to U80 rpm, respectively. The mere presence of
the seed material was sufficient to promote increases in sludge densi-
ties and settling rates, and these effects became increasingly apparent
as larger amounts of seed material were added to the suspension. How-
ever, these benefits in sludge properties were counteracted by a marked
increase in the turbidity of the more heavily seeded samples after
sludge settling. In all respects, the results of tests with seeding
materials were quite similar to those obtained during the filter aid
studies, and it is believed that whatever mechanism is involved in
effecting sludge densification through the use of seeding materials,
2.
-------
it Is probably not the usual mechanism involving growth of an ordered
crystalline lattice on the seed crystal nuclei.
13. Attempts to effect densification of a synthetic coal mine water
sludge by either heating the mine water before lime addition, or heat-
ing the concentrated sludge itself, were unsuccessful under the condi-
tions employed (temperature range of ^0 to 80 C, at atmospheric pres-
sure.) Slight increases in sludge density were apparent when the
sludge was allowed to settle immediately after heating. However, these
increases were negated if the sludge was resuspended after cooling to
room temperature, indicating that sludge dehydration by heating is a
reversible process within the temperature limits employed.
lU. In contrast to the results of sludge heating experiments, the
effects of sludge freezing were rather striking, and significant in-
creases in sludge density were observed after the frozen sludge had
been thawed.
15- Dissolved C02 in coal mine waters, whether occurring naturally or
added Intentionally by C03 sparging, results in the coprecipitation of
calcium carbonate during lime neutralization. This in turn results in
a significant densifying effect on the sludge in terms of reduced
sludge volumes and increased sludge solids contents. The effluents
from lime treatment of mine waters containing appreciable amounts of
dissolved C0a tend to be much more turbid than those in which COg Is
essentially absent. This problem can be easily overcome through the
use of coagulant aids, without any apparent attenuation of the sludge
densifying effect.
16. Of all the approaches to sludge densification explored, three
appear potentially promising in terms of the results obtained. These
are magnetic sludge preparation, sludge freezing, and pretreatment of
the mine water with carbon dioxide to effect coprecipitation of calcium
carbonate during lime addition. Each of these methods resulted in
settled sludge volume reductions of approximately 90 percent and in-
creases in sludge solids contents of from 0.5 to about 5 percent after
30-minute settling periods. The magnetic sludge process has the
advantage of possible recovery of a useful product. Sludge freezing Is
conceptually the simplest approach, and is apparently effective for any
type of sludge resulting from lime treatment. The C03-addition method
offers the possibility of partial recovery and recycling of the reagents
(both CaO and C02) through sludge calcination.
-------
SECTION II
KECOMMEKDATIONS
The following additional studies are recommended:
1. Future development of the magnetic sludge process should involve
optimization of the process variables; these include stirring rate,
sludge heating time, aeration rate, method of preliminary sludge de-
watering, and the effects of magnetic separation for final sludge proc-
essing. In addition, the matter of amenability of coal mine waters to
treatment should be considered more thoroughly by means of a more com-
prehensive literature review as well as actual tests on mine drainage
samples.
2. The technique of sludge freezing should be explored further.
Emphasis should be given to conservation of energy in the process, and
the filtration characteristics of the resulting sludges should be
evaluated.
3- Further studies should be conducted on the feasibility of the C03-
addition approach to mine drainage sludge densification. The filtration
characteristics of product sludges should be evaluated, and the possi-
bility of recovering and recycling the residual alkalinity of the
sludge, either directly as CaC03 or as CaO plus C03 after calcination,
should also be explored further.
k. Efforts should be made to develop cost data for the conceptual
processes involving magnetic sludge formation, sludge freezing, and COS
pretreatment, with the purpose of selecting at least one of these
processes for further testing at a field site.
5. The use of inexpensive waste materials (e.g. fly ash, "red dog,"
slag) as sludge additives, possibly in conjunction with coagulant aids,
should be considered further. Results of the present studies have
indicated that such materials, employed in relatively large amounts,
increase the density and enhance the filterability of sludges from lime
neutralization of coal mine drainage. The use of pulverized blast
furnace slag in this rega.rd appears particularly promising, and studies
should be made to ascertain its availability, relative cost, residual
alkalinity, ajid behavior as an additive in conventional mine drainage
treatment processes involving lime neutralization and aeration.
-------
SECTION III
INTRODUCTION
This is the final report on Project CR-102 under the auspices of the
Department of Environmental Resources, Commonwealth of Pennsylvania.1
The project was activated on January 5, 1970, with financial support
from Bituminous Coal Research, Inc., the Pennsylvania Coal Research
Board, and the Water Quality Office of the Environmental Protection
Agency (WQO-EPA)2 through Grant 1^010 EJT to the Commonwealth of
Pennsylvania. Work on the project was conducted according to the pro-
cedures outlined in BCR Research Program Proposal RPP-l6^-R, dated
April 2^, 1969, and a subsequent Revision of Scope, dated February 25,
1970. This report covers the entire project period from January 5>
1970, through March 25, 1971.
Nature of the Problem
Most lime neutralization processes presently employed for the treatment
of coal mine drainage yield a valueless, flocculent, gelatinous sludge
which is slow-settling and difficult to dewater. The problem of han-
dling and disposing of this sludge has been considered by some workers
to be the greatest single obstacle associated with coal mine drainage
treatment operations. Moreover, as the number of treatment plants con-
tinues to increase, increasingly larger volumes of sludge are produced,
further aggravating the problem.
Objectives and Scope of the Research
The major objective of this research was to alleviate present sludge
handling and disposal problems through the development of densification
techniques for coal mine drainage sludge.
The work was divided into two broad areas. The first involved an
approach whereby the usual conditions prevailing in the lime neutraliza-
tion technique were altered so as to produce a dense, magnetic sludge.
These studies were based on exploratory work (l)3 conducted under an
earlier BCR research program.
1 Formerly under the auspices of the Pennsylvania Coal Research Board.
2 Formerly the Federal Water Quality Administration, Department of the
Interior.
3 Numbers in parentheses indicate references given in Section 7.
7-
-------
During these earlier studies it was found that dense, black or dark
brown magnetic sludges could be obtained from both synthetic and
natural mine waters by partial oxidation of ferrous hydroxide (precipi-
tated by neutralization with hydrated lime) under controlled conditions
of pH and temperature. Unfortunately, however, this conversion of
ferrous hydroxide to a magnetic form is hindered by relatively small
amounts of several other species common to most coal mine waters.
Other workers have shown that magnesium (2, 3, U) and aluminum (2, 5) _
inhibit the formation of magnetite in aqueous suspension. This inhibi-
tion is reportedly confined to pH ranges in which these ions exist as
insoluble hydroxides, namely pH k.L to 10.8 for aluminum and pH > 9-6
for magne s ium.(2)
Available experimental evidence indicates that small amounts of these
impurities can be tolerated during the magnetic sludge transformation.
Thus, Stauffer and Lovell (5) report that the ferromagnetic properties
of sludges obtained by mixing heated suspensions of Fe(OH)2 and
Fe(OH)3 were strong up to an Al/Fe molar ratio of 0.085, intermediate
when this ratio was 0.123, and essentially negligible above a molar
ratio of 0.184. On a parts per million (ppra) basis, for each 100 ppm
of iron present the aluminum concentrations corresponding to these
molar ratios are U.I, 5-9, and 8.9 ppm, respectively. In similar
experiments, Kakabadse et a.l. (k) found that dense, black magnetic
precipitates prepared in the absence of magnesium became dark brown and
less magnetic as magnesium was introduced up to a Mg^/Fe2"1" molar ratio
of 1.0. Beyond this value, the precipita,te was lighter in color,
gelatinous, and nonmagnetic. This molar ratio corresponds to ^3-5 ppm
of Mg3* per 100 ppm of Fe2* .
The data of these other investigators provided useful guidelines re-
garding the feasibility of a coal mine drainage treatment process based
on the concept of magnetic sludge formation. Nevertheless, it was felt
that the development of any such conceptual process necessitated a more
complete study of reaction variables, the effects of interfering
species, and possible means by which these interferences could be alle-
viated. Studies of this nature were included in the scope of the pro-
gram as proposed, as well as studies to define the types of natural
coal mine waters (or mixtures thereof) which would be amenable to
treatment by a process resulting in magnetic sludge recovery.
The second general area of investigation involved the application of
sludge conditioning techniques to densify sludges obtained by ordinary
lime neutralization of coal mine drainage. Several of these techniques
have been tested in sewage treatment studies as a means of enhancing
the dewaterability of the sludge, and a few have been adapted to actual
in-plant operations. In the area of coal mine drainage sludge treat-
ment, however, these techniques have received only cursory attention as
indicated by the few scattered references in the pertinent literature.
8.
-------
Those techniques chosen for study in the present investigation included
the use of coagulant aids, filter aids, seeding materials, sludge heat-
ing and freezing, and the use of dissolved carbon dioxide gas to pro-
mote the coprecipitation of calcium carbonate during lime neutraliza-
tion. The relative effectiveness of each of these sludge conditioning
techniques was judged on the basis of changes in sludge settling rate,
settled sludge volume and solids content, and turbidity of the superna-
tant liquid after sludge settling. In addition, some limited tests on
sludge filterability were conducted in conjunction with the filter aid
studies.
The studies on magnetic sludge preparation and sludge conditioning were
carried out according to the work schedule shown in Figure 1.
It should be noted that the scope of the program did not include an
evaluation of the economic feasibility of these various sludge densifi-
cation techniques as applied to the treatment of coal mine drainage
sludge. However, this admittedly important aspect of the problem was
considered in a qualitative manner in judging the potential usefulness
of the various approaches, and pertinent remarks regarding cost factors
are included throughout this report.
9.
-------
Activity
A. Magnetic .Sludge Studies
1. Determination of Inter-
ference Tolerance Limits
2. Adaptation to Mine
Water Treatment
3. Evaluation of Magnetic
Sludge Properties
B. Sludge Conditioning Studies
1. Coagulation
2. Filter Aids
3. Seeding
4. Heating and Freezing
5. Miscellaneous Approaches
C. Comparison of Results
and Reporting
Month, 1970-1971
Sep Oct Nov Dec
May Jun Jul
I I I I II
I 1 I I
I 1 I I I I
I I I I J
I I I I I I
I I I I I
I I I I I I I I I I
Bituminous Coal Research, Inc. 2038G18
Figure 1. Schedule of Work —Recovery of Iron in a
More Dense Form from Coal /Wine Drainage
-------
SECTION IV
EXPERIMENTAL PROCEDURE
Materials, Equipment, and Analytical Techniques
Four actual coal mine waters were used in various phases of the research
program. All were from the Sewickley Creek area in Westmoreland County,
Pennsylvania. The BCR designations, site locations, and nature of the
mine water samples are described briefly in the following paragraphs.
Keystone
An inactive drift mine in Sewickley Township located near the junction
of Sewickley Creek with the Youghiogheny Pxiver, approximately 200 feet
west of Pennsylvania 6^10^ and 0.5 mile south of the junction of this
road with Pennsylvania T-UOO. Samples are characterized by a relatively
high ferrous iron and sodium content and by a relatively high pH. The
ferric iron present tends to hydrolyze and precipitate fairly readily
during storage of the mine water.
Brinkerton
An inactive shaft mine in Mount Pleasant Township located on the south
side of railroad tracks, 0.2 mile west of their junction with
Pennsylvania T-778 and at a point 0.2 mile south of the junction of this
road with Pennsylvania 6^130. Except for lower iron (and sodium) con-
centrations, this water is quite similar to the Keystone discharge.
Both Keystone and Brinkerton are relatively large-volume discharges,
each of about 3,000 gpm.
South Greensburg
An inactive drift mine in Hempfield Township located on the north side
of Pennsylvania T-681 northwest of its junction with Pennsylvania 6
Samples are characterized by the usual absence of ferric iron and a
fairly constant composition in spite of variations in flow rates.
Tarrs
An abandoned mine site in East Huntingdon Township about 0.5 mile south-
west of the town of Tarrs, Pennsylvania. Samples contain relatively
high concentrations of silicon and aluminum.
Mine water batch samples were collected in 5-gallon carboys and used in
the experiments the same day.
Studies on sludge conditioning reported herein were conducted using a
synthetic coal mine water. This synthetic mine water was prepared from
11.
-------
directions furnished by Fir. R. D. Hill of the WQO-EPA, and is essen-
tially a 0.01 N sulfuric acid solution containing specified amounts of
the sulfate salts (chemical reagent grade) of iron, aluminum, magnesium,
calcium, and manganese. The synthetic mine water was prepared as needed
in 15-liter batches and stored in a 5-gallon polyethylene carboy.
Representative analytical data for the four actual coal mine waters as
well as the WQO-EPA synthetic mine water are shown in Table 1. Values
for the Fe8*/total iron ratios indicate that all of these mine waters
with the exception of the Tarrs discharge can be considered as "ferrous"
waters.
Measurements of pH were accomplished using an Orion Model 1+01 Specific
Ion Meter. A Sargent combination pH electrode was used for measurements
at or near room temperature; pH measurements at elevated temperatures
were accomplished with a Sargent high-temperature glass electrode and a
reference electrode containing silver-silver chloride internals. When
appropriate, changes in pH with time were recorded using a Houston
Instrument Omnigraphic T-Y recorder, Model HR-80.
Experiments involving sludge heating were conducted with the aid of a
Cole-Farmer Model 3^31 hot plate regulated by an immersion probe for
automatic temperature control.
A Hach Model 2100 Laboratory Turbidimeter was employed for turbidity
measurements during the sludge conditioning studies. All turbidity
values reported herein are in Jackson Turbidity Units (JTU).
Electrophoretic mobility measurements were accomplished using a Zeta-
Meter (Zeta-Meter Inc., New York, New York).
X-ray diffraction analyses were conducted using a Picker x-ray generator
with the powder camera technique.
Ferrous iron concentration was determined by the colorimetric method
using ortho-phenanthroline. Sample absorbance was measured at 510 mp.
using a Gary Model lU spectrophotometer. Total iron as well as aluminum,
magnesium, calcium, manganese, silicon, and sodium were determined by
emission spectrographic analysis (solution technique) using a Jarrell-
Ash Model 78-000 1.5-meter Wadsworth grating spectrograph.
"Hot" acidity was determined by titration of the sample with 0.1 N
WaOH to pH 8.2. The sample was first treated with 5 drops of 30 per-
cent H203 to oxidize all iron, heated to near boiling for four to five
minutes, and cooled to room temperature. "Cold" acidity was determined
in^a similar manner except that the heating step in the procedure was
omitted. All acidity values reported herein are expressed as ppm CaC03
equivalents, and unless indicated otherwise, all values are from "hot"
acidity determinations.
12.
-------
TABLE 1. REPRESENTATIVE COMPOSITIONS OF COAL MINE WATERS
USED IN SLUDGE DENSIFICATION STUDIES
Source
Keystone
Spectrographic Analysis_, ppm
pH Pe Al Mg Ca Mn SI Na
!.2 255 < 2* 100 2^2 k 7
Fe2+/Total Fe
Ratio
0.91
Brinkerton
5.7 131 < 2* 81 198 U 8 59
South Greensburg U.9 110
85 210 k 12 80
0.88
0.91
Tarrs
3.1 95 k8 8k 20k k ko < 50*
WQO-EPA Synthetic 2.2 200 15 2k 80
1.00
Below the limits of accuracy of the analytical technique
-------
Magnetic Sludge Studies
Magnetic Sludge Factorial Experiment
The first phase of the research program involved a study of important
variables affecting the formation of a magnetic sludge, particularly
those relating to interferences caused by the presence of foreign
cations such as A13+ and Mg2 + in the system. Toward this end, a facto-
rial experiment was designed to determine the effects of temperature,
pH, Al/Fe ratio, Mg/Fe ratio, and interactions of these variables on
the magnetic properties, settling rate, and volume of sludge produced
during neutralization of a synthetic mine water with hydrated lime.
The methodology involved in this factorial experiment was based on a
recent publication on statistically designed experiments.(6) The exper-
iment consisted of 16 tests; initial iron concentration (300 ppm),
stirring rate (l88 rprn), and aeration rate (50 cc/min) were held con-
stant throughout the series of tests.
The experimental procedure was as follows: Aliquots of stock solutions
of ferrous sulfate, aluminum sulfate, and magnesium sulfate were added
to an 800-ml beaker and diluted to a total volume of 500 nil to obtain
the desired Al/Fe and Mg/Fe molar ratios. Stirring was commenced and
the solution was brought to the desired temperature (25 or 90 C). The
pH was then adjusted to the desired value (8 or 10) by dropwise addi-
tion of a 10 percent w/w Ca.(OH)s suspension and aeration wa.s started.
The pH of the suspension was maintained at the desired level during the
test, and changes in color were noted. In addition, a small portion of
the suspension was withdrawn early in the test a.nd filtered; aliquots
of the filtrate were analyzed for residual dissolved Fe2+.
At the end of a 60-rninute reaction period, the beaker was removed and
solids were allowed to settle by gravity for one hour. The supernatant
liquid was then decanted, and the sludge was transferred to a. 100-ml
graduated cylinder for measurement of settling rate. The volume of the
suspension was made up to 110 ml, and timing was started when the
sludge-liquid boundary passed the 100 ml mark of the cylinder. After
60 minutes, final settled sludge volume was recorded. The settling
rate was determined by plotting the sludge volume versus time and cal-
culating the slope of the linear portion of the settling rate curve.
All values were converted to centimeters per minute based on the actual
height of the graduated cylinder.
Magnetic response of the settled solids was checked qualitatively with
a hand magnet, ajid if the sludge was magnetic, the solids were filtered
in a 30-ml medium porosity fritted glass funnel with suction, washed
several times with acetone, and dried at room temperature. The solids
were then pulverized to a granular powder, and a small sample (0.20 g)
was transferred to a tared 15-ml beaker. The beaker was placed on the
pan of a modified Torbal torsion balance directly a.bove a cylindrical
pot magnet, which was 1-3/16 inches high and 1-3/8 inches in diameter.
Ik.
-------
The weight required to overcome the pull of the magnet, minus the tare
weights of the beaker and sludge sample, was determined, as a quantita-
tive measure of magnetic response. Although this technique was
admittedly crude compared to the usual methods for determining magnetic
susceptibility, it was rapid and sufficiently reliable for our purposes.
This is indicated further by the data in Figure 2, which show the mag-
netic response of various ferromagnetic materials measured by this
technique as a function of sample weight. The materials tested included
reagent grade iron powder, a synthetic magnetite (about 92 percent
Fe304), a synthetic maghemite (essentially pure Y-Fea03), and two
magnetic sludge samples prepared from Keystone mine drainage. Although
the curves in Figure 2 tend to tail off with increasing sample weight,
the data indicate that the relationship of magnetic response to sample
weight is, as shown by the dashed lines, reasonably linear below a
sample weight of about O.kO g. In view of these results, it was felt
that the procedure was suitable as a means of comparing magnetic proper-
ties of various materials on a relative basis, as long as sample weight
did not exceed 0.^0 g.
Effects of Magnesium and Aluminum Interference
The interference of magnesium and aluminum with magnetic sludge forma-
tion was studied using essentially the same experimental procedure as
that described above for the factorial experiment. Initial experiments
were conducted using synthetic coal mine waters containing appropriate
aliquots of stock solutions of ferrous sulfate, magnesium sulfate,
and/or aluminum sulfate diluted to a total volume of 500 ml. The ini-
tial iron concentration (300 ppm), stirring rate (l88 rpm), and aeration
rate (50 cc/min) were held constant during all experiments. Based on
results from the earlier factorial experiment, it was decided that
preferred reaction conditions were pH 8.0 and 90 C. Accordingly, mag-
netic sludge preparations were carried out, under these conditions.
Originally, these tests involved heating the mine water sample to 90 C,
adding hydrated lime slurry to attain a pH of 800, and aerating the
suspension during a 60-minute reaction period. Later in the investiga-
tion, it was found that interference from magnesium could be minimized
through a preliminary sludge concentration step in the procedure, and
the experimental procedure was altered slightly as follows:
To a 6,000-ml batch sample of the synthetic or actual mine wa.ter,
sufficient hydrated lime in the form of a 10 percent slurry was added
to the sample with stirring to raise the pH to 8.5 and precipitate iron,
primarily as Fe(OH)3 . The sludge wa,s allowed to settle for a predeter-
mined time period, after which the supernatant liquid was carefully
removed by siphoning. During the sludge settling period a portion of
the supernatant liquid wa.s withdrawn, filtered, and analyzed for
residual dissolved iron. In cases where centrifugation was employed,
the sludge was first allowed to settle by gravity for one hour, then
the settled sludge was dewatered as noted above and centrifuged in an
15-
-------
60-.
55-
50-
45-
40-
AAAGNETIC 35~
RESPONSE,
grains
30 -\
25-
20-
15-
10-
_. Iron Powder
(B&A Reagent Grade, > 95.0% Fe)
II Magnetic Sludge from
Keystone Drainage
(Prepared 2/18/69)
£ 8 (Synthetic Iron Oxide Pigment;
^ Mapico Brown 422)
0.0 0.2 0.4 0.6
FeaO4 (Mineral Mills
"Floatkleen" Magnetite,
Grade No. 1)
ic Sludge from
Keystone Drainage
(Prepared 3/18/70)
0.8
SAMPLE WEIGHT, grams
Bituminous Coal Research, Inc. 2038G19
Figure 2. Magnetic Response of Various Ferromagnetic
Materials with Increasing Sample Weight
16.
-------
International Model V trunnion-head centrifuge (international Equipment
Company, Boston, Massachusetts) for ten minutes at 1,500 rpm.
The volume of the gravity-concentrated (or centrifuged) sludge was
measured, and a sample was taken for solids content determination. The
bulk of the sludge was then converted to a magnetic form by heating at
90 C for 30 minutes with stirring and aeration. At the end of the
heating period the magnetic sludge was transferred to a graduated
cylinder, and settled sludge volume was recorded after one hour, after
which the supernatant liquid was removed and a sample of the sludge was
taken for solids content analysis once again. A portion of the magnetic
sludge was filtered with suction, dried by acetone washing at room
temperature, and pulverized by grinding with a glass rod. A 0.200 g
sample of the dried sludge was used for magnetic response measurement.
A second portion (about 70 mg) was dissolved in 5 nil of concentrated
hydrochloric acid ajid the resulting solution was diluted to 100 ml in a
volumetric flask. A 25-ml aliquot of this solution was then taken for
the determination of Mg/Fe and Al/Fe molar ratios by emission spectro-
graphic analysis.
The experiments on magnetic sludge seeding involved the earlier approach
of preheating the mine water sample to 90 C. Weighed amounts of the
seed material were a,dded to the heated sample, just before lime addition
(to pH 8.0) and aeration (for 60 minutes) were commenced. The seed
material used in these tests was a magnetic sludge prepared from the
Keystone mine water on March 18, 1970, as follows: Hydrated lime was
added with stirring to 90 gal of the raw mine water until a pH of 8.5
was attained. Wo residual dissolved iron was detected at this pH» The
sludge settled overnight to a volume of about 2.5 ga.1. The supernatant
liquid was removed and the sludge concentrate was heated to 90 C and
maintained at that temperature for one hour with stirring but no aera-
tion. A dark brown, strongly magnetic sludge was obtained which settled
overnight to a volume of 0.53 gal. A portion of the settled sludge was
filtered on a Buchner funnel, dried at 107 C for two hours, ground to a
fine powder, and sieved to a particle size fraction of 325 x kOO mesh
for use in the magnetic sludge seeding tests. (This material was also
used for magnetic response measurements, shown earlier by Figure 2. It
is interesting that this magnetic sludge prepared from the Keystone
discharge exhibits essentially the same response in a magnetic field as
do commercially available ferromagnetic iron oxides.)
Sludge Conditioning Studies
Coagulant Aid Studies
For the coagulant aid addition studies, the following procedure was
developed for sludge preparation. To a 1,000-ml sample of the WQO-EPA
synthetic mine water, hydrated lime (Ca(OH)2) was added as a dry powder
with stirring. The amount of lime added was equivalent to the exact
stoichiometric requirement based on the mine water total acidity.
17.
-------
Aeration was commenced at a. rate of 2,500 cc/min, and aeration and
stirring were continued during a 30-minute reaction period. Analyses
for residual dissolved iron showed that complete iron oxidation and
precipitation occurred within the 30-minute reaction period. The final
pH of"the sludge suspension was about 7-8-
Aliquots (100 ml) of the sludge suspension were transferred to 250-ml
beakers and the desired amount of coagulant aid was added. The suspen-
sion was stirred for one minute at low speed, after which samples of
the suspension were withdrawn for electrophoretic mobility measurement
with the Zeta-Meter. Coagulant aids were prepared as 1,000 ppra stock
solutions, and aliquots of these stock solutions were diluted as
necessary to obtain the desired coagulant aid concentrations in the
sludge suspensions.
For the determinations of settling rate and solids content, 100-ml
aliquots of the synthetic mine water sludge suspension were taken and
treated with selected amounts of coagulant aids as described above.
The samples were then transferred to 100-ml graduated cylinders and the
cylinders were stoppered and inverted three times. Sludge settling
rates were measured and the settled sludge volume was recorded after a
30-minute period. Samples of the supernatant liquid were then with-
drawn for turbidity measurements. Finally, the remainder of the
supernatant liquid was drawn off by careful siphoning, and a sample of
the wet sludge was taken for solids content determination. Sludge
samples were dried in an oven at 105 0 to constant weight.
Filter Aid Studies
In the studies on filter aid addition, the requisite amount of the
selected filter aid was added to a 500-ml batch sample of the synthetic
mine water with stirring. Changes in pH produced by the filter aids,
if any, were monitored for a one-minute period, then hydrated lime was
added to the suspension and aeration was commenced. In cases where the
filter aid material effected an initial increase in pH, the amount of
lime added was that fraction of the stoichiometric amount necessary to
achieve a final pH of 7.8. After a 30-minute aeration period, 100-ml
aliquots of the suspension were taken for electrophoretic mobility
measurement and for determinations of settling rate, 30-minute settled
sludge volume, solids content, and turbidity, by the procedures
described earlier. In addition, the behavior of sludges containing the
filter aids only was compared with that of the same sludges after
treatment with a coagulant aid.
Filtration Tests
In conjunction with the filter aid studies, a series of batch filtra-
tion tests was conducted using a modified leaf test procedure.(7) The
sludge suspension was prepared by adding the desired amount of filter
aid to 1,000 ml of the synthetic mine water, followed by hydrated lime;
18.
-------
the mixture was then stirred and aerated for a 30-minute period, result-
ing in complete oxidation and precipitation of the iron. The amount of
lime added was that sufficient to achieve a final pH of 7-8 in the
sludge suspension.
The sludge sample was then transferred to a magnetic stirrer (except
for those samples containing magnetite, in which case a small mechani-
cal stirrer was employed) and the leaf filter was immersed about 2
inches below the surface. Stirring was commenced, and a vacuum of 18
inches of mercury was applied to the filter.
Filtration was carried out for a 10-minute period, then the filter was
withdrawn from the suspension, mounted in a vertical position, and
suction was continued for an additional 5-minute period to simulate a
cake dewatering cycle. The filtrate volume was measured and the weight
of the wet filter cake was recorded. The filter cake was then dried in
a dessicator at room temperature to constant weight.
The filter used for the filtration tests consisted of a standard Milli-
pore glass filter base connected by rubber vacuum tubing to a 1,000-ml
filtration flask. Millipore membrane filters (MF type) of 8p. pore size
were used throughout the tests, and were held to the filter base by two
grooved Plexiglas rings connected by small springs. The filtration
flask was also connected to a U-tube mercury manometer and was equipped
with an air bleed valve to regulate vacuum, which was obtained using a.
Model k¥L Neptune Dyna-Pump.
It was found difficult to effect a quantitative transfer of the small
amounts of damp filter cake obtained (on the order of i+00 mg) to a
separate container for weighing. Consequently, the weight of wet sludge
was determined by subtracting an average tare weight of the damp filter
paper. Measurements made on six filter disks showed an average dry
weight of 6^.8 t 1.3 mg and an average wet weight (after filtration
cycles with deionized water alone) of 2^-7-7 - 9-6 mg. Only those filter
disks whose initial dry weights were within the limits indicated were
chosen for use in the filtration tests.
Seeding Experiments
The nature of the precipitate formed in the presence of seed crystals is
known to be influenced to a. la.rge extent by seed crystal size, seed
crystal amount, and agitation rate.(8) Accordingly, tests on the use
of seeding materials were designed to consider the effects of these
three variables, as well as the presence or absence of the seed material
itself, on sludge properties. Most of the tests involved the use of
anhydrous ferric oxide (Fisher Certified Reagent grade) which was ob-
tained in the form of a red powder. The material was sieved to obtain
the following particle size fractions (U.S. Sieve Series): 50 x 80 mesh
(297 x 177[i, average 237|a), 100 x ikO mesh (1^9 x 105|_i, average 127(j),
and 200 x 325 mesh (7^ x 44[i, average 59f-0 - Thus the average particle
sizes for each fraction were in the ratio of U:2:l, respectively.
19-
-------
In the tests on the effect of seed crystal size, the various particle
size fractions of the seed material were added to 1,000 ml of the syn-
thetic mine water to a concentration of 100 ppm. This amount corre-
sponded to approximately six percent of the total sludge weight on a
dry basis. The air flow rate was reduced from 2,500 to about 280 cc/rnin,
and the stirring rate was held constant at 300 rpm. Under these condi-
tions, the degree of agitation necessary to keep the seed material in
suspension was minimized. Sufficient hydrated lime was added to the
stirred, aerated suspension to achieve a final pH of 7-8. It was found
necessary to increase the reaction time from 30 to 60 minutes to achieve
complete oxidation and precipitation of iron; this may have been due to
the lower aeration rate and the fact that, at the lower speed, the
stirrer was less effective in breaking up small agglomerated particles
of lime in the suspension. After the 60-minute reaction period, the
entire sample was transferred to a 1,000-ml graduated cylinder and the
sludge was allowed to settle for 30 minutes. Measurements were made of
settling rate, settled sludge volume, sludge solids content, and super-
natant liquid turbidity after the 30-minute settling period.
In tests on the effect of seed crystal quantity, the 200 x 325 mesh
sieve fraction was used throughout and the amount of seeding material
employed was increased through the range of 100 to 2,000 ppm. All other
steps in the experimental procedure remained as described above.
In tests on the effect of stirring rate, the stirring rate was increased
through the range 300 to ^80 rpm. The aeration rate was kept at about
280 cc/min and the 200 x 325 mesh sieve fraction of the seed material
was used in all tests at a concentration of 2,000 ppm. All other con-
ditions remained as described earlier. For purposes of comparison, a
few tests were conducted to observe the effect of stirring rate alone
on unseeded (control) samples.
Finally, a few experiments were conducted using as seeding materials
two actual sludge samples which had been obtained from lime neutraliza-
tion of the synthetic mine water. One of the samples had been recovered
after a freeze-dewatering experiment. Both sludge samples were filtered,
air-dried, and ground to a powder. After they had been sieved, the
samples were found to consist predominantly of minus 325 mesh material,
thus the 325 mesh x 0 particle size fraction was used for these tests.
The aeration and stirring rates employed were 280 cc/min and 300 rpm,
respectively.
Sludge Heating and Freezing
In the sludge heating studies, tests were grouped into three series
according to experimental procedure. The first series involved heating
the synthetic mine water samples before lime addition and iron precipi-
tation. Hydrated lime, in an amount sufficient to produce a final PH
of 7-8, was added with stirring and aeration to 1,000-ml batch samples
of the mine water which had been previously heated to the desired
20.
-------
temperature. Temperature was maintained constant during a 30-minute
reaction period, after which the sample was gradually cooled to room
temperature over a one-hour period with the aid of a cold tap water
bath. The sludge suspension was then transferred to a 1,000-ml gradu-
ated cylinder, and measurements were made of sludge settling rate,
30-minute settled sludge volume, sludge solids content, and supernatant
liquid turbidity after the 30-minute settling period. A portion of the
sludge was filtered using a Millipore filter assembly, and the filter
cake was air-dried at room temperature, pulverized to a powder, and
subjected to x-ray diffraction analysis.
The second series of experiments involved preparation of the sludges at
room temperature and heating the sludge concentrates. Sludges were
prepared by adding hydrated lime to 1,000-ml samples of the snythetic
coal mine water, followed by a 30-minute aeration period with stirring.
The sludge was allowed to settle for 30 minutes, then the supernatant
liquid was siphoned off and the remaining sludge concentrate (about
200 ml) was transferred to a 250-ral beaker and heated to the desired
temperature with stirring. The sludge was kept at the elevated tempera-
ture for 10 minutes, and was then cooled to room temperature again.
Finally, the sample was stirred, a 100-ml portion was transferred to a
100-ml graduated cylinder, a.nd settling rate, 30-minute settled sludge
volume, and sludge solids content were measured. In addition, samples
of the settled sludge were prepared for x-ray diffraction analysis as
described above. It was not possible to conduct turbidity determina-
tions in this series of tests, since the volume of supernatant liquid
remaining in the 100-ml graduated cylinder after 30 minutes was not
sufficient to provide the required amount of sample (about 3° ml) for
the turbidimeter.
The third series of experiments involved essentially a variation on
those conducted during the second series. Sludge samples were prepared,
concentrated, and heated in the same manner as described in the preced-
ing paragraph. However, instead of being first cooled to room tempera-
ture, the sludges (100-ml portions) were transferred immediately to the
100-ml graduated cylinder while still warm and settling behavior was
observed for the usual 30-minute period. After the sample for solids
content measurement had been taken, the sludge was left undisturbed for
an additional 30-minute period to allow further cooling to room tempera-
ture. The sample was then made up to volume (100 ml) using portions of
the supernatant liquid which had been saved from the original sludge
concentration step; the sludge was resuspended, and settling rate,
sludge volume, and solids content were measured again.
In the sludge freezing experiments, sludges were prepared in the usual
manner by hydrated lime treatment of synthetic and actual mine waters.
Sludges were allowed to settle for some specified time period, then the
settled sludge was recovered by decantation and frozen at -k.k C (2)4 F)
in a refrigerator freezer compartment. Once frozen, the samples were re-
moved from the freezer and allowed to thaw gradually without agitation
21.
-------
at room temperature. No attempt was made to control the rates of
sludge freezing or thawing.
Miscellaneous Approaches - Coprecipitation of Calcium Carbonate
In the experiments on the effect of calcium carbonate coprecipitation
during lime neutralization, 1,000-ml batch samples of both synthetic and
actual (Keystone) mine waters, were used. Into one sample, carbon
dioxide gas (Matheson "Coleman Instrument Grade," 99-99 percent minimum
purity) was bubbled at room temperature for a predetermined time period
(C02 flow rate = 2.0 scfh or <$kk cc/min). After the C02 purge period,
hydrated lime was added and the sample was aerated with stirring for
30 minutes. The lime was added as a dry powder in the earlier experi-
ments, but it tended to agglomerate and pH control was difficult. Con-
sequently, in later experiments the lime was sieved to minus 325 mesh
and added to the mine water sample as a slurry in deionized water.
After the 30-minute reaction period, the sample was transferred to a
1,000-ml graduated cylinder for measurement of settling rate, settled
sludge volume, sludge solids content, and supernatant liquid turbidity
after the 30-minute settling period. In most cases, control samples
without prior C02 treatment were tested simultaneously for comparison.
22.
-------
SECTION V
RESULTS AND DISCUSSION
Magnetic Sludge Studies
Magnetic Sludge Factorial Experiment
The initial results from the magnetic sludge factorial experiment are
summarized in Table 2. For ease in interpretation, the four separate
variables are identified with the high and low level of each variable
being represented by + and - signs, respectively. Several important
relationships are evident from an inspection of the data in Table 2.
Foi1 example, magnetic sludges were obtained only during the high temper-
ature tests. Secondly, with the exception of the last three tests,
there is an apparent correlation between settled sludge volume (or
sludge settling rate) and pH.
Data on residual dissolved Fe2 + indicate that precipitation of iron is
not complete at pH 8 and room temperature (Tests 1, 3> 5j and 7)- Al-
though this should not have any confounding effect on the interpretation
of the results, it is an important observation and an aspect of the
overall conceptual process that should be considered carefully during
any future developmental work.
The data in Table 2 were employed more fully in the calculation of main
effects and variable interactions for each of the three responses. For
this purpose, a codified design and calculation matrix, shown in
Table 3? was used. The derivation of this matrix is described by
Davies.(9) The positive or negative signs for variable interactions are
obtained simply by algebraic multiplication of the signs for each single
variable involved in the interaction. Main effects and interactions are
calcxilated by multiplying the response values (Table 2) by the corre-
sponding effect elements (Table 3)? summing the results, and dividing
by 8. The resulting value for single variables is also sometimes desig-
nated as an average effect, since (for two levels of the variable only)
it is simply the difference between the average response of all trials
ca.rried out at the first level of the variable (factor) and that of all
trials at the second level. If the effect of one variable changes at
different levels of another variable, the two variables are said to
interact. The relative importance of the variable or variables in ques-
tion is reflected by the absolute magnitude of the calculated main
effect or interaction. First-order (two variable) interactions are
usually easy to interpret in terms of the physical behavior of the sys-
tem. Second- and higher-order interactions are often less readily
understood, and are often of minor significance in any case.
The main effects and interactions derived from the present study in the
manner described above are listed in Table k-. It should be noted that
23.
-------
TABLE 2. RESULTS OF FACTORIAL EXPERIMENT ON VARIABLES AFFECTING
MAGNETIC SLUDGE FORMATION
Test
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Responses
Variables*
123 4
-t-
60-Min Settled
Sludge Volume
Percent**
15.0
19.0
18.2
18.7
13-0
18.8
18.3
19.0
5-1
13.6
11.6
18.8
17.2
11.8
18.4
12.1
Sludge
Settling
Rate , cm/min
0.08
0.02
0.03
0.02
0.11
0.02
0.03
0.02
2.43
0.16
0.15
0.02
0.05
0.19
0.02
0.34
Magnetic
Response
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
4.06
0.29
2.46
< 0.01 4-
2.10
0.09
0.61
0.00
* Variable Identification
Variable
1 = pH
2 = Al/Fe Ratio
3 = Mg/Fe Ratio
4 = Temperature
Levels
+
lo.o 8
0.15 o
1.5 0
90 c 25
-
.0
.05
.5
C
Residual
Dissolved
Fes+ ppm
16
< 1
6
< 1
< 1
3
< 1
< 1
< 1
< 1
< 1
< 1
< 1
< 1
< 1
** Based on a total suspension volume of 500 ml
4- Very weak magnetic response, below the sensitivity
of the balance, was detected by visual observation.
24.
-------
ro
VJI
TABtE 3. CODIFIED DESIGN AND CALCULATION MATRIX
Variables Interactions
Test
1
2
3
4
5
6
7
8
9
10
11
12
13
l4
15
16
1234 (1,2) (1,3) (
+ +
4- - - -
4- - 4-
+ 4- - 4-
- - 4- - 4-
4- - 4- - - +
— 4- 4- — — ..
4- + + - 4- +
- 4- 4- 4-
4- - - 4-
- 4- - 4- - 4-
4- 4- - 4- 4-
- - 4- 4- 4-
4- - 4- 4- - 4-
-4-4-4-
4-4-4-4- 4- 4-
;i,4) (2,3) (2,4) (3,4) (1,2,3)
4-4-4-4-
-4-4-4- +
+ - - + 4-
+
4- - 4- 4-
+
4- 4-
4- 4-
4-
4- 4- 4-
4- 4-
4- - 4-
+ 4-
4- - - 4-
-4-4-4-
4- 4- 4- 4- 4-
(2,3,4)
-
4-
4-
4-
4-
-
-
4-
4-
-
-
_
-
4-
4-
(1,2,4)
4-
+
-
—
4-
4-
-
+
-
-
4-
4-
-
-
4-
(1,3,4) (1,2,3,4)
4-
4-
-
4- 4-
4-
4-
4- 4-
-
4-
4-
4- 4-
-
4-
4-
-
4- 4-
Designation of Variables: 1 = pH
2 = Al/Fe Ratio
3 = Mg/Fe Ratio
4 = Temperature
-------
TABLE 4. MAIN EFFECTS AND INTERACTIONS BASED ON DATA FROM MAGNETIC SLUDGE FACTORIAL EXPERIMENT
Responses
Settled Sludge
Volume Percent
Variable(s)
Single Variable:
Two Variable
Interactions:
Three Variable
Interactions;
Four Variable
Interaction:
Main Effects
and
Interactions
k
2
1
3
1,3
1,2
3,4
2,4
2,3
1,4
1,3,4
2,3,4
1,2,4
1,2,3
-4.22
3.00
2.17
0.77
-2.87
-1.65
1.22
0.90
-0.65
-0.57
-3.37
-1.30
0.50
-0.45
Sludge
Settling Rate
Variable(s)
1,2,3,4
-0.05
1,2,3,4
Main Effects
and
Interactions
4
2
1
3
1,3
1,2
2,3
2,4
3,4
1,4
1,3,4
2,3,4
1,2,4
1,2,3
0.38
-0.31
-0.27
-0.26
0-35
0.31
0.31
-0.27
-0.27
-0.22
0.36
0.32
0.28
-0.24
Magnetic Response
Main Effects
and
Variable(s) Interactions
k
1
3
2
1,4
3,4
1,3
2,4
1,2
2,3
1,3,4
1,2,4
2,3,4
1,2,3
1.08
-0.99
-0.62
-0.32
-0.99
-0.62
0.57
-0.32
0.22
0.16
0.57
0.22
0.16
-0.11
-0.24
1,2,3,4
-0.11
Designation of Variables:
1 = pH
2 = Al/Fe Ratio
3 = Mg/Fe Ratio
4 = Temperature
-------
these initial results were refined by data from later replications (to
be discussed below). As indicated by Table U, the main effects and
interactions for each response are grouped according to the number of
variables involved and are listed in order of decreasing absolute magni-
tude in each group. The results will be discussed below in terms of
each response.
Settled Sludge Volume
The data in Tables 2 and h indicate that the most significant variable
affecting settled sludge volume is temperature. The negative value for
the temperature main effect simply means that settled sludge volume was
significantly greater at the lower level of temperature (25 C in this
experiment). This effect is particularly apparent if one compares
Test 6 with Test lU (settled sludge volumes of 18.8 and 11.8 percent,
respectively) and Test 8 with Test 16 (settled sludge volumes of 19.0
and 12.1 percent, respectively). The sludges from each of these four
tests were all essentially nonmagnetic. It is believed that these dif-
ferences are due to coprecipitation of gypsum during the higher temper-
ature tests, and this will be discussed further below.
Other single variables significantly affecting settled sludge volume
are Al/Fe ratio and pH, in order of decreasing importance. Both main
effects are positive, indicating an increasing sludge volume with both
increasing pH and aluminum concentration. Ostensibly, there is no
reason why an increase in pH should result in greater sludge volumes,
if all other conditions are held constant. The explanation lies in the
appearance of a significantly large first-order interaction between pH
and the Mg/Fe ratio. This interaction was expected, and is consistent
with the fact that Mg(OH)3 precipitates at about pH 9-6. This same
situation is reflected by the relatively large second-order interaction
involving pH, Mg/Fe ratio, and temperature.
Sludge Settling Rate
A relationship between sludge settling rate and settled sludge volume
would be expected, and this is evidenced by the close correspondence in
the ranking of main effects and interactions within a given group.
Thus, statements made previously regarding sludge volume would hold
generally in this case, with the understanding that effects or inter-
actions contributing to high sludge volumes also lead to low settling
rates (indicated by the reversal of signs).
Magnetic Response
Again, the data indicate that temperature is the most important vari-
able affecting magnetic response. This would seem obvious from the
fact that no magnetic sludges were obtained in the lower temperature
tests (Table 2). In this case, however, pH appears to be the second
most important variable, and the main effect of -0.99 indicates that
27.
-------
increasing pH to the higher level produced a marked decrease in the mag-
netic properties of the product. The effect of increasing magnesium
concentration is third in importance, and is also reflected in the
relatively large 1,3 and 1,3,4 variable interactions. It was^antici-
pated that these interactions would be significant, in view of the
reported ability of magnesium to inhibit the formation of magnetite
from aqueous suspension when the pH is high enough (> 9.6) to allow
precipitation of Mg(OH)3.(2,4)
Other major interactions affecting magnetic response are the pH, tem-
perature [1,4] and the Mg/Fe ratio, temperature [3,4] interactions.
The 1,4 interaction (negative) may be interpreted as follows: the
effect of higher temperature in increasing the magnetic response of the
product is greater at the lower pH level, or conversely, the effect of
higher pH in decreasing the magnetic response is greater at the lower
temperature level. The 3,4 interaction may be interpreted in a similar
manner if one substitutes "Mg/Fe ratio" for "pH" in the preceding sen-
tence. The positive value for the 1,3 interaction indicates that both
variables are operating in the same direction, that is, magnetic
response is affected adversely when both variables are at their higher
levels. This is presumably related to the precipitation of Mg(OH)2 at
the higher pH level, as was indicated previously. All other variable
interactions, and those involving the Al/Fe ratio in particular, appear
to be less important in terms of their effect on magnetic response „
All tests at the higher temperature level (Tests 9 through 16, Table 2)
were replicated. The results of these replications are shown in
Table 5 under the headings "Trial 2." Also shown in Table 5 are the
mean values obtained for each response. These mean values were used in
a complete recalculation of the main effects and interactions of the
four independent variables considered in this study (pH, Al/Fe and
Mg/Fe molar ratios, and temperature). The results of these calcula-
tions are shown in Table 6.
A comparison of the results in Table 6 with those derived earlier
(Table 4) reveals that, in nearly every case, the rankings and signs of
main effects and interactions within a given group remain virtually
unchanged from those determined before the data were refined. The only
exceptions to this statement involve a few first-order (two-variable)
and higher-order interactions which were judged to be not significant
in the original interpretation of the results.
The important trends noted as a result of these revised calculations
are as follows:
In general, the absolute magnitudes of all main effects and interactions
decreased slightly. In other words, there was a tendency for small
effects and interactions to more closely approach a value of zero
(nonsignificance).
28.
-------
TABLE 5. COMPARISON OF RESULTS OBTAINED BY PARTIAL REPLICATION
OF MAGNETIC SLUDGE FACTORIAL EXPERIMENT
Responses
ro
vo
Settled Sludge
Variables* Volume Percent**
Test
9
10
11
12
13
14
15
16
123 Trial 1 Trial 2
- - - 5-1
+ - - 13-6
- + - 11.6
+ + - 18.8
+ 17.2
+ - + 11.8
- + + 18.4
+ + + 12.1
6.5
17-3
11.2
18.7
17,4
12.6
18.0
12.8
Mean
5.80
15-45
11. 40
18.75
17.30
12.20
18.20
12.45
* Variable Identification
Levels
Variable +
pH 10 . 0 8 .
Al/Fe Ratio 0.15 0.
Mg/Fe Ratio 1-5 0.
0
05
5
Sludge Settling
Rate , cm/min
Trial 1
2.43
0.16
0.15
0.02
0.05
0.19
0.02
0.34
Trial 2 Mean
1
0
0
0
0
0
0
0
.02 1.
.04 0.
.16 0.
02 0.
.04 0.
.19 o.
.03 o.
.28 0.
725
100
155
020
045
190
025
310
Magnetic Response
Trial 1 Trial 2 Mean
4.
0.
2.
< 0.
2.
0.
0.
0.
06
29
46
01 <
10
09
61
00
3-87
0.25
2.61
0.01
1.77
0.00
0.79
0.00
3.965
0.270
2.535
< 0.0104-
1-935
0.045
0.700
0.000
** Based on a total suspension volume of 500 ml.
4- Very weak magnetic response, below the sensitivity
of the balance, was detected by visual observation
(Temperature = 90 C for all tests)
-------
TABLE 6. REVISED CALCULATION OF MAIN EFFECTS AND INTERACTIONS BASED ON
DATA FROM MAGNETIC SLUDGE FACTORIAL EXPERIMENT
Responses
Single Variable:
Two Variable
Interactions:
Three Variable
Interactions:
Four Variable
Interaction:
Settled Sludge
Volume
Variable(s)
4
2
1
3
1,3
1,2
3,4
2,3
1,^
2,4
1,3,4
2,3,^
1,2,4
1,2,3
Percent
Main Effects
and
Interactions
-3-55
2.30
2.15
0.8?
-3-22
-1.45
1.32
-0.62
-0.60
0.12
-3-72
-1.27
0.70
0.02
Sludge
Settling Rate
Variable(s)
4
2
1
3
1,3
1,2
2,3
3,4
2,4
1,^
1,3,4
2,3,4
1,2,4
1,2,3
Main Effects
and
Interactions
0.28
-0.21
-0.19
-0.17
0.27
0.22
0.22
-0.18
-0.18
-0.15
0.28
0.22
0.19
-0.17
Magnetic
Variable! s)
4
1
3
2
1,^
3,4
1,3
2,4
1,2
2,3
1,3,4
1,2,4
2,3,4
1,2,3
Response
Main Effects
and
Interactions
1.18
-1.10
-0.51
-0.37
-1.10
-0.51
0.45
-0.37
0.29
0.05
0.45
0.29
0.05
0.00
1,2,3,4
0.42
1,2,3,4
-0.17
1,2,3,^
0.00
Designation of Variables:
1 = pH
2 = Al/Fe Ratio
3 = Mg/Fe Ratio
4 = Temperature
-------
With regard to the settled sludge volume, the main effect of the Mg/Fe
ratio (variable 3) increased slightly. This was also reflected by in-
creases in the absolute values of certain interactions involving this
variable, notably the 1,3 (pH, Mg/Fe) interaction. Thus, these inter-
actions which were found to be important in the original analysis
involving single trials only, now assume an even greater importance
based on the revised calculations.
With regard to magnetic response, revised calculations show slight in-
creases in the absolute values of the main effects for variables 1, 2,
and h. Correspondingly, all interactions involving combinations of
these variables were also found to be slightly larger. These findings
do not alter the conclusions drawn from the original analysis of the
data from single trials. However, the revised calculations do indicate
that the effect of aluminum on magnetic response is slightly more impor-
tant (relative to the effect of magnesium) than was suggested by the
original analysis.
The data obtained from the partial replication of the factorial experi-
ment may be employed to compute error terms for each of the three re-
sponses. To accomplish this, and since only the high temperature tests
were replicated, it is appropriate to consider the last eight tests
(Tests 9 through 16) as a separate factorial experiment involving three
variables (temperature held constant) at two levels each, i.e., as a 2s
factorial experiment. Main effects and interactions of the three vari-
ables are then computed in the usual manner for each response.
The results of such computations are shown in Table 7> together with the
error terms for the 95 percent confidence interval (95 percent C.I.).
These error terms are derived by calculating the variance for each
test, pooling these variances, and determining the variance of main
effects or interactions, according to the procedure outlined by Pavelic
and Saxena.(6) Main effects and interactions which are larger than the
error term, indicated by the asterisks in Table 7> are assumed, to be
significant at the 95 percent confidence level.
Several interesting comparisons can be made between the results of the
full (24) factorial in Table 6 and those of the partial (2s) factorial
in Table 7- For example, the importance of the 1,3 interaction is
evident in all cases. For magnetic response, the ranking of single
variables in both cases is identical. However, for the other two re-
sponses, variable 3 apparently assumes greater significance when temper-
ature is held constant at the higher level. This fact, in conjunction
with the significant 2,3 interaction for sludge volume, suggests that
the physical behavior of the sludge may be more sensitive to changes in
both Mg and Al concentrations at the higher temperature than at the
lower temperature. One possible explanation for this involves the tend-
ency toward coprecipitation of gypsum at the higher temperature level
and its effect on sludge properties (discussed below).
31.
-------
TABLE 7. MAIN EFFECTS AND INTERACTIONS FOR THE 2* FACTORIAL
EXPERIMENT INVOLVING HIGH TEMPERATURE TESTS ONLY
Responses
OJ
ro
Single Variable:
Two Variable
Interactions:
Three Variable
Interaction:
Error Term
C.I.)
Settled Sludge
Volume Percent
Variable(s)
2
3
1
1,3
2,3
1,2
Main Effects
and
Interactions
2.50*
2.20*
1.55*
-6.95*
-1.90*
-0.75
Sludge
Settling Rate
Variable (s_)
2
3
1
1,3
2,3
1,2
Main Effects
and
Interactions
-0.39
-0.35
-0.33
0.5*1*
O.UU
0.1+1
Magnetic
Variable(s)
1
3
2
1,3
1,2
2,3
Response
Main Effects
and
Interactions
-2.20*
-1.02*
-0.75*
0.91*
0.59*
0.11
1,2,3
1,2,3
+1.19
-0.3^
+0.53
1,2,3
0.00
+0.13
* Probably significant effects or interactions based on error term.
Designation of Variables: 1 = pH
2 = Al/Fe Ratio
3 = Mg/Fe Ratio
-------
Overall Significance of the Results
The magnetic sludge factorial experiment and replications yielded valu-
able data concerning certain variables which affect the formation and
settling behavior of the magnetic product. In addition, certain design
parameters for a conceptual treatment process for coal mine drainage
may be inferred from the results.
For example, it is significant that temperature was found to be the
most important variable affecting all three responses, indicating that
a conceptual magnetic sludge process should include a sludge heating
step. The effect of pH is also relatively important, particularly as
related to magnetic response of the product. The pH effect is in turn
reflected in the 1,3 interaction, which was found to be significant for
all three responses. These findings indicate that the pH should be high
enough to afford complete precipitation of iron (as Fe ) but low enough
to avoid coprecipitation of Mg(OH)a. Experience has shown that the
optimum pH is about 8.5.
It should be noted that another phenomenon was observed during these
experiments which complicated interpretation of the results. This in-
volved the occasional presence of coprecipitated gypsum (CaS04'2HgO) in
the sludge and its effect on the measured settling rates. In each of
the high temperature, high pH tests (including replications) a whitish,
granular material was observed during settling rate measurements. This
material was usually dispersed throughout the bulk of the sludge volume,
but in at least one case (Test 16) it settled rapidly to deposit as a
discrete layer at the bottom of the graduated cylinder. A portion of
this material was separated by centrifugation and identified as gypsum
by x-ray diffraction analysis. The tendency of this coprecipitate to
settle at a faster rate than the bulk of the sludge probably accounts
for the deviations from the established trends for sludge volumes and
settling rates encountered in the later tests, especially Tests 1^- and
16 (Table 2). This belief is reinforced by another observation: in at
least two tests (Tests 10 and lU) the sludge settled rather slowly for
about 40 minutes, but as settling rate measurements were commenced
(after the sludge-liquid boundary passed the 100 ml mark of the gradu-
ated cylinder) the rate increased considerably. This behavior may have
been the result of gradual crystallization of gypsum from a supersatu-
rated solution, a tendency which is characteristic for this compound.
Coprecipitated gypsum was not observed in any of the lower-temperature
tests, consistent with the fact that its solubility increases with de-
creasing temperature.
Ideally, aluminum and magnesium should be absent from the system, al-
though the data show tha,t it is possible to obtain a sludge exhibiting
some magnetic response when small amounts of aluminum and magnesium are
present. Consequently, the next logical step in the program was to
ascertain the limits to which these impurities could be tolerated based
on their effects on magnetic response and other properties of the sludge
obtained under the conditions cited above.
33-
-------
Effects of Magnesium and Aluminum Interference
Tolerance Level for Magnesium
A series of tests was conducted at pH 8 and 90 C to study the effect of
increasing the Mg/Fe molar ratio on magnetic sludge properties. ^ The
experiments were run using synthetic mine water solutions^containing
only dissolved ferrous and magnesium sulfates (i.e., aluminum was
absent from the system). The results of these tests are shown in
Table 8 and Figure 3. The data reveal that the effect of magnesium^on
magnetic response and other sludge properties is certainly not negligi-
ble at pH 8. This result was somewhat unexpected in view of reports
that the interference is restricted to pH values at which magnesium is
precipitated as the hydroxide (pH > 9.6).(2,U).
As reflected in Table 8, there was a gradual lightening in color and a
rapid increase in settled volume of the precipitate as the Mg/Fe molar
ratio increased. A marked discontinuity in the settled sludge volume
curve, indicated by the dashed line in Figure 3, is believed to be due
to the effect of coprecipitated gypsum at the higher magnesium concen-
trations . Small, granular gypsum particles were observed dispersed
throughout the sludge in the tests at Mg/Fe molar ratios of 5 and 10.
The effect of this coprecipitate in increasing settling rates and de-
creasing sludge volumes has been mentioned previously; Figure 3 shows
striking graphic evidence of the nature of this effect.
Based on the results of these experiments on the effect of magnesium on
magnetic sludge formation, it would appear that there is no well-defined
tolerance limit for this impurity. Figure 3 shows that the decrease in
magnetic response is gradual as the Mg/Fe molar ratio increases. In'
fact, when magnetic response was observed qualitatively with a small
hand magnet, the behavior of all samples prepared' at a Mg/Fe molar
ratio of 2.5 and below appeared to be essentially identical.
The results shown by Figure 3 suggest that a practical (although perhaps
conservative) tolerance level for magnesium would be that corresponding
to a Mg/Fe molar ratio of 1.0. This value is in accord with that re-
ported by Kakabadse et al.,(if) although it should be recognized that
these investigators employed substantially different experimental con-
ditions than those of the present study. Our data (Figure 3) show that
magnetic response at the molar ratio Mg/Fe = 1.0 is roughly only ho per-
cent of that measured when magnesium was completely absent from the
system.
On the basis of the present studies, the interference due to magnesium
would appear to involve the cation Mgs+ via a solution mechanism. This
actually would seem to be more consistent with the proposed mode of
magnesium interference, which involves substitution by Mgs+ for Fes+ in
the magnetite lattice,(k) than would a mechanism involving solid magne-
sium hydroxide. At any rate, magnesium interference is presumably of
-------
TABLE 8. EFFECT OF THE Mg/Fe MOLAR RATIO ON MAGNETIC SLUDGE PROPERTIES
60-Minute
UJ
vn
Test No.
Control
397-60
397-59
397-58
397-56
397-53
Mg/Fe
Molar Ratio
0.00
0.50
1.50
2.50
5.00
10.00
mg/1 Mg per
100 mg/1 Fe
0.0
21.8
65-3
108.8
217.6
i+35.2
Settled Sludge
Volume Percent^
0.8
5.0
12.6
16.7
8.2**
12.5**
Sludge Settling Magnetic
Rate, cm/min Response*
4-
k.I
0.12
0.05
0.6k
0.16
8.58
U.90
2.76
1.65
0.3^
0.12
Color of
Dried Sludge
black
dark brown
medium brown
medium brown
light brown
beige
£ Based on a total suspension volume of 500 ml.
* Measured on a 0 200 gram sample.
4- Indistinct sludge-liquid boundary prevented settling rate measurement.
** Granular gypsum was observed dispersed throughout the sludge.
-------
r-18
MAGNETIC
RESPONSE,
grams
Settled Sludge
Volume Percent
SETTLED
SLUDGE
10 VOLUME
PERCENT
Mg/Fe MOLAR RATIO
Bituminous Coal Research, Inc. 2038G2C
Figure 3. Magnetic Response and Settled Sludge Volume Percent
as Functions of the Mg/Fe Molar Ratio
36.
-------
much greater consequence at higher pH values. This is indicated by the
observations of Kakabadse e_t al. (2,U) as well as by the results of our
own factorial experiment.
Results of chemical analyses of sludges obtained in this series of
tests are presented in Table 9- Interestingly, although all of the
magnesium was presumably in solution under the experimental conditions
employed, the data indicate that approximately one-half of the magne-
sium available during magnetic sludge formation is present in the
sludge. The magnesium content of the sludges increased regularly with
increasing Mg/Fe ratios, although the presence of coprecipitated gypsum
in the last two sludges in the series led to anomalously low values for
both iron and magnesium contents.
Tolerance Level for Aluminum
Another series of tests was conducted at pH 8 and 90 C to study the
effect of increasing the Al/Fe molar ratio on sludge properties. Sny-
thetic mine water solutions containing only ferrous and aluminum sul-
fates (no magnesium) were used in these experiments. The results are
shown in Table 10 and Figure h.
As indicated by Figure ht magnetic response decreases and settled sludge
volume increases as the Al/Fe molar ratio increases. These results are
very similar to those from the tests on increasing Mg/Fe molar ratios,
presented in the preceding section. However, there is an important
difference in the magnitudes of the two effects. This is illustrated
in Figure 5j which shows the Al/Fe and Mg/Fe molar ratios plotted on the
same scale. It is obvious from Figure 5 that, on a mole-to-mole basis
and under the experimental conditions employed, the effect of aluminum
on magnetic properties of the sludge is much more pronounced than that
due to magnesium. Interestingly, 60-minute settled sludge volumes were
roughly the same at a given value of the Al/Fe or Mg/Fe molar ratio; in
no case was coprecipitated gypsum observed during this series of tests
on increasing Al/Fe ratios.
Again, there is no clear-cut tolerance limit for aluminum. However, in
line with the example given earlier for magnesium interference, the
point at which magnetic response is ho percent of that of the control
sample (from Figure U) corresponds to an Al/Fe molar ratio of about 0.18
It is noteworthy that Stauffer and Lovell (5) have reported, based on
qualitative observations, that the ferromagnetic properties of products
prepared in a similar manner diminished rapidly as the Al/Fe molar
ratio increased, and were "so slight as to be inconsequential" at an
Al/Fe molar ratio of 0.18U.
Results of chemical analyses of sludges obtained in this series of
tests are presented in Table 11. The data show that, as expected,
essentially all of the aluminum present in the original solution reports
to the sludge, presumably as coprecipitated Al(OH)3.
37-
-------
co
00
TABLE 9 EFFECT OF INCREASING Mg/Fe MOLAR RATIO ON CHEMICAL
COMPOSITION OF THE MAGNETIC SLUDGE
Mg/Fe Molar Ratio Mg/Fe Molar Weight Percent of Indicated
During Sludge Ratio in Constituent in Sludge
Test No.
397-60
397-59
397-58
397-56
397-53
Preparation
0.50
1.50
2.50
5.00
10.00
Sludge
0.25
0.67
1.13
1-95
k.96
Fe
53-2
39-7
32. k
13-1
5.0
Mg
5.8
11.6
16.0
11.1
10.8
Ca
3.5
2.U
2.7
11.1*
16.3*
* Coprecipitated gypsum was detected visually in
these two samples
-------
U)
TABLE 10. EFFECT OF THE Al/Fe MOLAR RATIO ON
MAGNETIC SLUDGE PROPERTIES
60-Minute
Al/Fe
Test No. Molar Ratio
Control
397-61
397-64
397-62
397-63
397-65
0.00
0.05
0.10
0.15
0.30
0.50
mg/l Al per
100 mg/l Fe
0.0
2.4
4.8
7.2
14.4
2k. 0
Settled Sludge,
Volume Percent*
0.8
1.0
1.2
1.5
2.9
6.6
Sludge Settling Magnetic Color of
Rate, cm/min Response** Dried Sludge
4-
4-
4-
4-
1.95
0.88
8.58
8.24
6.36
4.84
0.65
0.10
black
black
black
brownish-black
reddish-brown
yellowish-brown
* Based on a total suspension volume of 500 ml.
** Measured on a 0.200 gram sample.
4- Indistinct sludge-liquid boundary prevented settling rate measurement
-------
7-
6-
MAGNETIC
RESPONSE,
grams
5-
4-
3-
1-
0.0
I Magnetic
Response
Settled Sludge
Volume Percent,
0.1
T
\
0.3
0.5
1-18
-16
-14
-12
-10
-8
-6
-4
-2
SETTLED
SLUDGE
VOLUME
PERCENT
0.6
0.2 0.3 0.4
AI/Fe MOLAR RATIO
Bituminous Coal Research, Inc. 2038G2
Figure 4. Magnetic Response and Settled Sludge Volume Percent as
Functions of the Al/Fe Molar Ratio
ij-0.
-------
AAAGNETIC
RESPONSE,
grams
Legend:
0,0 Data from Mg/Fe
ratio runs
•,• Data from AI/Fe
ratio runs
'Settled Sludge
'Volume Percent
Magnetic Response
vs. Mg/Fe Ratio
i Magnetic Response
vs. AI/Fe Ratio
0.0
i.o
1.5
1
2.0
I
2.5
AI/Fe or Mg/Fe MOLAR RATIO
SETTLED
SLUDGE
VOLUME
PERCENT
-16
-14
-12
-10
-8
-6
-4
-2
3.0
Bituminous Coal Research, Inc. 2038G22
Figure 5. Magnetic Response and Settled Sludge Volume Percent as
Functions of the AI/Fe and Mg/Fe Molar Ratios
-------
TABLE 11. EFFECT OF INCREASING Al/Fe MOLAR RATIO ON CHEMICAL
COMPOSITION OF THE MAGNETIC SLUDGE
Al/Fe Molar Ratio Al/Fe Molar Weight Percent of Indicated
During Sludge Ratio in Constituent in Sludge
Test No. Preparation Sludge Fe Al Ca
397-61 0.05 0.05 57. ^ I.1* 2.9
ro 397-6U 0.10 0.09 61.1 2.6 h.O
397-62 0.15 0.13 55.2 3-5 U.2
397-63 0.30 0.30 U5.7 6.6 3.14.
397-65 0.50 0.1+6 35.1 7.8 5.2
-------
Correlation Between Magnesium and Aluminum Interferences
For the purpose of predicting whether a given mine water might be ame-
nable to treatment by a magnetic sludge process, some correlation be-
tween magnesium interference and aluminum interference would be desir-
able. The data suggest that the effects are additive, but that the
effect of aluminum must be weighted more heavily than the effect of
magnesium.
Through a consideration of the results shown by Figure 5, it was found
that the following relationship existed for a given value of the mag-
netic response:
Mg/Fe = k(Al/Fe)a
where k was approximately constant with a value of 30 + 5- Consequently,
data from the tests on the effect of aluminum were revised according to
the above relationship and plotted as a function of the magnetic re-
sponse of the samples, as shown by Figure 6. In a similar fashion, data
from the tests at pH 8 and 90 C obtained during the earlier factorial
experiment (Tests 9> H> 13» and 15, Table 5) were included in the cal-
culations and are also shown in Figure 6. These data are significant
in that both aluminum and magnesium were present at varying concentra-
tions during the magnetic sludge preparations, thereby affording a test
of the hypothesis that the interference effects are additive. In spite
of some scattering of the points around the smooth curve, it is apparent
that a reasonably good correlation is possible. Of particular signifi-
cance is the fact that the data points from the factorial experiment
also tend to follow the curved line.
The results shown by Figure 6 suggest that, as a first approximation,
the tolerance levels for aluminum and/or magnesium can be defined empir-
ically in terms of their molar ratios with iron by means of the follow-
ing expression:
Mg/Fe + 30(Al/Fe)3 < 1.0
In the absence of aluminum, the tolerance level for magnesium would be
that concentration corresponding to a Mg/Fe molar ratio of 1.0. This
value was suggested earlier as a practical limitation for magnesium
interference. In the absence of magnesium, the corresponding value for
the Al/Fe molar ratio would be the square root of 0.0333 or 0.182,
which was also derived earlier from a consideration of the data on
aluminum interference alone.
It should be emphasized that the relationships shown above are only
valid for the experimental conditions employed, i.e., where the entire
mine water sample was heated before lime alkalization. Under these
conditions, magnesium was always present in the system, presumably as
at pH 8. In later studies, to be discussed below, it was found
-------
MAGNETIC
RESPONSE,
grams
6-
4-
3-
f\
1-
Legend:
a Data from Mg/Fe Ratio Runs
• Data from AI/Fe Ratio Runs
Data from Factorial Experiment
(both Mg and Al present in sample)
I I 1 1 1 1 1 1
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
MOLAR RATIO 30(AI/Fe)! + Mg/Fe
Bituminous Coal Research, Inc. 2038G23
Figure 6. Variation in Magnetic Response with Increasing
AI/Fe and Mg/Fe Molar Ratios
-------
that magnesium interference could be largely eliminated by a prelimi-
nary precipitation/sludge separation step. In view of this later find-
ing, therefore, no further attempts were made to verify the validity of
the relationships shown above through additional tests with either syn-
thetic or actual coal mine waters.
Magnetic Sludge Seeding Experiments
It was reasoned that the interfering effects of magnesium and aluminum
on magnetic sludge formation might be reduced or eliminated, if small
amounts of a magnetic sludge were added to the reaction mixture to pro-
vide a substrate upon which the freshly formed magnetic sludge could
nucleate and grow. This possibility was tested in a series of experi-
ments where small portions of a magnetic sludge, prepared from an actual
coal mine discharge, were added to synthetic mine water samples prior
to iron precipitation and magnetic sludge formation.
The system chosen for study was that represented by Test 15 of the fac-
torial experiment (Table 5), in which aluminum and magnesium were at
their higher levels (corresponding to Al/Fe and Mg/Fe molar ratios of
0.15 and 1.5» respectively). In other words, the effect of the seed
material was tested under conditions found by previous experimentation
to be definitely inimical to the formation of a strongly magnetic sludge,
The seed material, as noted earlier, was prepared from the Keystone mine
water, and was found by emission spectrographic analysis to be essen-
tially free of magnesium and aluminum.
The results of these seeding tests are shown in Table 12. All product
sludges were filtered, washed with acetone, dried at room temperature,
and weighed. By maintaining weight balances on the sludge and seed
material, it was possible to calculate a minimum expected magnetic re-
sponse based on the assumption that the product material was simply a
homogeneous mixture of the seed material and the weakly magnetic sludge
produced in the absence of seed material. This calculation also in-
volved the assumption that magnetic response is directly proportional
to sample weight, an assumption which was justified by the results of
earlier experimental work (see Figure 2). An example of this calcula-
tion for Test 39?-?6 is given below:
-------
TABLE 12. RESULTS OF SEEDING TESTS WITH MAGNETIC
SLUDGE FROM KEYSTONE MINE DRAINAGE
Weight of Sludge
D-Min Settled Sludge Obtained, grams
Magnetic Response
of 0.200 gram
Seed Material
Test No. Added, grams
397-71
397-74
397-76
397-8044
397-88#
397-82
0 . 0000
0.0500
0.1000
0.2000
0.4000
0.8000
Seed Material, grams
Actual*
--
1-13
2.29
4.31
8.48
14.76
CorrectedJ-
--
1-13
2.26
4.52
9.04
18.08
Sludge Volume
Percent**
18.4
18.4
18.8
17.3
18.6
14.2
Settling
Rate,, cm/min Actual
0.02
0.02
0.06
0.04
0.02
0.10
0.8411
o . 9110
0.9179
1.0033
1.2756
1.6539
Minus Wt.
of Seed
0.8411
0.8610
0.8179
0.8033
0.8756
0.8539
Sludge Sample, grams
Actual
0 52
0.35
0.72
1.45
1.51
2.45
Calculated*
--
0.74
0.96
1.32
1.78
2.46
o\
* Measured on total amount of seed material used.
4- Calculated based on 0.0500 g seed material and assuming direct
proportionality between seed sample weight and magnetic response.
** Based on a total suspension volume of 500 ml.
£ Calculation explained in the text.
-1-4- Averaged data from two tests.
ir Averaged data from three tests.
-------
Wt. seed material = 0.1000 g
Magnetic Response (M.R.) of 0.0500 g seed material
by actual measurement = 1.13 g
M.R. of 0.1000 g seed material = °'J:°°° x 1.13 g = 2.26 g
W • \J ^ \J\J
Wt. sludge formed during reaction = 0.8179 g
M.R. of sludge formed
during reaction4 = tj^SjL x 0.8179 = 2.13 g
U . C.UU
Combined M.R. of sludge mixture containing 0.1000 g seed
plus 0.8179 g fresh sludge = 2.26 + 2.13 = ^-39 g
Calculated M.R. of 0.2000 g sample
O POOO
of sludge mixture = ''oa x ^-39 = 0.96 g
Calculated magnetic responses for 0.200 g portions of the sludge samples
from all tests were derived in a similar fashion and are listed in the
right-hand column of Table 12.
From a comparison of the calculated and actual (measured) magnetic re-
sponses of the product sludges, it seems evident that the presence of
the seed material did not result in any appreciable enhancement of mag-
netic properties of the final sludge. In other words, the data indicate
that the product sludge was simply a homogeneous mixture of the original
seed material and the weakly magnetic sludge formed in the absence of
seed material. This conclusion is supported by the fact that settled
sludge volumes remained virtually unchanged until the weight of seed
material became so great as to approximate 50 percent by weight of the
dried sludge .
The results of these tests, shown graphically- in Figure 7, illustrate
that the measured magnetic response of the seeded sludge is not much
different than that calculated, assuming separate contributions from the
seed material and the weakly magnetic, unseeded sludge. A line showing
the magnetic response of the seed material alone as a function of sample
weight is included in Figure 7 for comparison.
4 The measured response of a 0.200 g sludge sample from the unseeded
control test (Test 397-71) was 0.52 g.
-------
16-1
14-
12-
10-
MAGNETIC RESPONSE,
grams
8-
6-
4-
rt __
Legend:
Magnetic Response of:
1. Seed Material Alone
Theoretical
• Measured
2. Seeded Sludge
T^ I I I I I I I I
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
WEIGHT OF SEED MATERIAL ADDED, grams
Bituminous Coal Research, Inc. 2038G24
Figure 7. Results of Seeding Tests with Keystone Magnetic Sludge
-------
Reduction of Magnesium Interference by Preliminary Sludge Concentration
Results of earlier tests (Table 8) showed that even at pH 8, magnesium
interference with magnetic sludge formation is still significant; i.e.,
under conditions where magnesium remains in solution as Kg3*. It was
reasoned that this interference might be reduced if the aqueous phase
containing Mgs+ was separated from the sludge before the magnetic con-
version (heating) step.
Initially, a few experiments were conducted using actual coal mine
waters from the Keystone and South Greensburg sites (see Table l). The
Mg/Fe and Al/Fe molar ratios in both the raw mine waters and the mag-
netic sludges prepared therefrom were determined by emission spectro-
graphic analyses. The results of these initial tests are shown in
Table 13, and the data show a decided increase in magnetic response
when the sludge was concentrated by overnight settling and separated
from the supernatant liquid by decantation (Test 397-33). As shown,
magnesium was virtually absent from the magnetic sludge product in this
case. It is also interesting that in the tests where no prior sludge
concentration was employed, about half of the magnesium (and all of the
aluminum in the case of Test 397-79) available in the raw mine water
reported to the sludge, as indicated by the molar ratios of these
impurities with total iron. These results are in agreement with those
from the tests with synthetic mine water samples, presented earlier
(Tables 9 and 11).
These initial tests were followed by a more extensive series of experi-
ments on the effects of preliminary sludge concentration and separation
in reducing magnesium interference during the magnetic sludge conversion
step. Tests were conducted using a synthetic mine water solution con-
taining both magnesium and aluminum, and with actual mine drainage
samples from the Keystone and Brinkerton sites. As indicated earlier
in Table 1, these two actual mine discharges are essentially free of
aluminum. Three methods of sludge concentration were employed. The
first two involved gravity settling periods for 1 and 16 hours, respec-
tively, followed by decantation of the supernatant liquid. The third
method involved a one-hour gravity settling period, followed by decan-
tation of the supernatant liquid, centrifugation of the settled solids,
and decantation of the centrate liquid.
The results of these later sludge concentration tests are summarized in
Table ih. The data show that protracted settling periods (l6 hours)
resulted in increases in initial sludge solids contents and decreases
in settled sludge volumes, although these changes were generally small
in comparison to those obtained by centrifugation of the settled sludge.
Conversion of the concentrated sludges to a magnetic form resulted in
further increases in solids contents, but with the exception of one
possibly anomalous value of 15.8 percent solids for the magnetic sludge
obtained from the Brinkerton discharge after preliminary centrifugation
(Test ^09-30), these increases were of a relatively small order of
-------
TABLE 13. RESULTS FROM INITIAL TESTS ON PRELIMINARY SLUDGE CONCENTRATION
Test No.
397-79
Mine
Water
Sample
1:1 Mixture of
Keystone and
South Greensburg
Sludge
Concentration
Method
None
Molar Ratios-l-
Raw Mine Water
Mg/FeAl/Fe
1.17 0.05
Magnetic Sludge
Mg/FeAl/Fe
0.51
0.05
Sludge
Magnetic
Response,
grams*
vn
o
397-77
397-83
Keystone
Keystone
None
None
0.95 < 0.02
0.71 < 0.02
0.57 < 0.02 1.25
0.32 < 0.02 I.kk
397-33
Keystone
Overnight Settling 1.07 < 0.02 < 0.02 < 0.02
. 85
4- Fe = total iron
* Measured on a 0.200 gram sample of the dried sludge.
-------
TABLE ll*. RESULTS OF ADDITIONAL PRELIMINARY SLUDGE CONCENTRATION TESTS
DURING MAGNETIC SLUDGE PREPARATION STUDIES
Test No.
409-27
409-29
409-25
409-40
409-37
409-18
409-32
409-34
409-30
Mine
Water
Sample*
Synthetic
Synthetic
Synthetic
Keystone
Keystone
Keystone
Brinkerton
Brinkerton
Brinkerton
Sludge
Concentration
Method-)-
A
B
C
A
B
C
A
B
C
* All
-I- A -
B -
C -
Concentrated Sludge
Weight
Percent Sludge
Solids Volume, ml
0.4 876
1.1 721
3-8 85
i.o 692
1.5 622
6.2 60
0.3 651
0.7 471
5.2 35
Magnetic Sludge
Weight
Mg/Fe Al/Fe
Percent Sludge Molar Molar
Solids* Volume,
1.1 392
1.6 252
5.0 40
3.8 98
5.1 102
6.0 4o
1.1 134
2.3 101
15.8 10
ml* Ration Ration
0.21 0.04
0 . 21 0 . 04
< 0.03 O.o4
< 0.04 < 0.02
< 0.04 < 0.02
< 0.04 < 0.02
0.15 < o.o4
0.11 < o.o4
0.08 < o.o4
Magnetic
Response,
Grams
2.31
2.49
2.78
1.92
2.52
2.4l
3-05
3-30
4.06
Iron Content,
Weight Percent
41.5
47.8
42.7
31.6
29.4
30.7
27-3
25.1
32.0
tests conducted on 6,000 ml batch samples.
1 hour gravity settling
period.
16 hour gravity settling period.
1 hour gravity settling
period followed by
centrifugation for 10 min at 1,500 rpm.
* Measured after 1 hour gravity settling period.
f Molar ratios in raw mine waters :
Synthetic :
Keystone :
Brinkerton :
Mg/Fe Al/Fe
1.53 0.04
0.88 < 0.02
1.63 < o.o4
-------
magnitude. It is perhaps more significant that the one-hour settled
sludge volumes of the magnetic sludges were usually less than half
those of their nonmagnetic precursors, and in most cases considerably
greater reductions in sludge volumes were obtained during the magnetic
conversion process.
The most significant results, in terms of the purpose of this study,
are reflected by the data for Mg/Fe molar ratios in the magnetic
sludges. In all cases the Mg/Fe molar ratios for the sludges^are con-
siderably smaller than those for the raw mine waters, indicating^that
under the experimental conditions most of the magnesium remains in
solution and is removed from the sludge during the preliminary sludge
concentration-separation step. Consequently, these data support the
conclusion that magnesium interference with magnetic sludge formation
can be minimized if the initally precipitated (at pH 8.5) ferrous
hydroxide sludge is concentrated and separated from the bulk of the
mine water before the magnetic sludge conversion. In terms of a prac-
tical coal mine drainage treatment process, such a preliminary sludge
concentration step has the additional advantage of reducing the volume
of that portion of the total flow requiring thermal treatment. These
results are in agreement with those obtained during the much earlier
exploratory work, where it was observed that preliminary sludge concen-
tration by centrifugation tended to enhance the formation of a magnetic
sludge.(l) However, the reasons for this enhancement were not apparent
at that time.
The effect of preliminary sludge dewatering on the sludge Mg/Fe molar
ratio is probably shown best by the data for the magnetic sludges ob-
tained from the Brinkerton discharge. There is a regular decrease in
the Mg/Fe ratio as the method of sludge dewatering becomes more effi-
cient. The results for the Keystone magnetic sludges are more difficult
to explain, since the amounts of magnesium in the samples analyzed were
consistently below the detectable limit of the method (3 ppm), regard-
less of the procedure employed for sludge concentration and dewatering.
Composition of Magnetic Sludges Prepared from Actual Coal Mine Waters
It was observed that all of the dried, pulverized magnetic sludges pre-
pared from actual mine waters effervesced strongly when treated with
dilute hydrochloric acid. The sludge sample from Test 397-33 (Table 13)
was selected for further analytical studies.
X-ray diffraction analysis of this sludge revealed both magnetite
(Feg04) and calcite (CaCQg). Analysis of the sludge by a combustion
train procedure for ultimate carbon yielded a value of k.Q percent
carbon, corresponding to approximately i+0 percent CaC03 by weight. This
unexpectedly high percentage of calcium carbonate is indicative of the
extent of carbonation of the Keystone discharge. In addition, the
amount of calcium corresponding to this percentage of CaC03 is 16.0 per-
cent. Since emission spectrographic analyses revealed that the calcium
content of this sludge was 15.k percent, there is an indication that,
52.
-------
for this particular sludge at least, all of the calcium in the sludge is
present in the form of CaC03.
Calcite in those magnetic sludges derived from carbonated mine waters
may be considered as an impurity which decreases the bulk magnetic
properties of the sludge. This may account for the fact that, in those
experiments where the entire mine water samples were heated (Tests 397-
79, -77, and -83, Table 13), the data did not fit the relationship
developed from similar experiments with synthetic mine waters (Figure 6)
Conceptual Magnetic Sludge Process for Coal Mine Drainage Treatment
Based on the findings discussed above, it is felt that a treatment
process for coal mine drainage leading to a magnetic sludge product is
technologically feasible under certain conditions. A conceptual process
suggested from the results of these studies is presented schematically
in Figure 8. The potential usefulness of such a process, as well as the
important cost factors involved, will be considered in the following
paragraphs.
Process Evaluation
The proposed conceptual magnetic sludge process would be expected to
result in the following advantages over existing methods of coal mine
drainage treatment:
Reduced sludge volumes: The data from tests with actual mine
waters (Table lU) indicate that approximately five- to seven-
fold reductions in settled sludge volumes can be achieved in
a magnetic sludge process, compared with sludge volumes ob-
tained by conventional lime treatment. This is further illus-
trated in Figure 9? which shows the relative volumes of
sludges obtained from 1,000-ml batch samples of the Keystone
mine water by three different methods of treatment. The
magnetic sludge was prepared by heating the solids, previously
concentrated by gravity-settling for one hour, at 90 C for
30 minutes. (The material which appears to be in suspension
above the magnetic sludge actually consisted of small parti-
cles adhering to the inner walls of the Imhoff cone.) Consid-
erably greater reductions in sludge volume may be possible if
mechanical dewatering (e.g. centrifugation) is employed during
the preliminary sludge concentration step.
Increased solids contents: On the order of four-fold increases
in solids contents of the settled sludges were observed to
result from the magnetic sludge conversion process. Again,
larger increases may be possible with preliminary mechanical
dewatering as well as with the use of magnetic separation
equipment for final sludge processing.
53-
-------
Hydrated Lime
(Ca(OH)2)
Heated Water
Lime Slurry
Feeder
VJl
Lime Slurry
Mixer
Lime Slurry (10-30%)
Sand Filter
(optional)
Coagulant Aid
Addition (Optional)
Steam Lance
or
Immersion
Heater
CT~pH
Controller
Product
Water
(Sludge
Settling)
Magnetic
Sludge
Thickener
(option-
al) / Wet Magnetic
Separator
Settled
Sludge
(Fe(OH)o)
CD
Mixing Tank
Reaction
Tank
\
Centrifugate
Treated AMD
(PH 8.5)
Solid Bowl
Centrifuge
(continuous)
Magnetic Sludge
Air Sparger
Concentrated
Magnetic Sludge
Bituminous Coal Research, Inc. 2038G25
Figure 8. Conceptual Design of Mine Drainage Treatment
Process for the Formation of Magnetic Sludge
-------
Sludge Obtained by
Lime Neutralization
and Aeration
Sludge Obtained by
Lime Neutralization Alone
(No Aeration)
2038P1
Magnetic Sludge
Figure 9. Relative Sludge Volumes After a Thirty Minute Settling Period
(Sludges Prepared from Keystone Mine Discharge)
55-
-------
More rapid sludge-liquid separations: The denser magnetic
sludge resulting from a process similar to that shown in
Figure 8 presumably would be faster settling and more readily
dewatered than sludges from lime treatment, leading to higher
treatment capacities for the magnetic process. Moreover,
magnetic separation devices could be employed to obviate the
need for sludge settling lagoons.
The possible shortcomings of the conceptual magnetic sludge process are
as follows:
Necessity for sludge heating: The need for elevated temper-
ature in the process cannot be fully explained based on the
available experimental data, although it is known from results
of the earlier factorial experiment that temperature is a
critical variable. Possible explanations for the beneficial
effect of elevated temperature might include enhanced nucle-
ation and growth of the magnetic phase and/or dehydration of
the hydrous iron oxides involved in the reaction leading to
the magnetic product. Another possible explanation is that
high temperature limits the solubility of air in the slurry
and minimizes the possibility of over-oxidation of the mag-
netic ferrosoferric oxide to the nonmagnetic hydrous ferric
oxide. In fact, this latter explanation was cited by the de-
velopers of the DuPont process (for obtaining magnetic sludge
from waste steel pickle liquors) as the reason for maintain-
ing reaction temperatures in the range 82 to 100 C in that
process.(10)
Close pH control: The necessity to quantitatively precipitate
iron, yet avoid coprecipitation of magnesium, imposes rela-
tively tight limitations on pH in the initial neutralization-
precipitation step of the treatment process. With modern pH
control instrumentation, however, adequate pH control should
not prove difficult.
Process restricted to certain types of coal mine waters: Al-
though it has not been stated explicitly before, the magnetic
sludge treatment process is presumably limited to those mine
waters containing iron principally in the ferrous (Fea+)
state, although theoretically a magnetic product would be
possible from a mine water with a Fe3+/total Fe ratio as low
as 0.33. Undoubtedly, a more serious limitation is imposed
by the sensitivity of the process to aluminum interference,
and the fact that many coal mine discharges contain dissolved
or suspended aluminum admittedly tends to militate against
the practicality of a magnetic sludge process. This matter
will be considered further in a subsequent section.
56.
-------
Unfortunately, there does not seem to be any practical way in this
application to separate aluminum from iron in solution. The usual pro-
cedure in analytical chemistry is to add an excess of a soluble alkali
(e.g. NaOH); this causes the aluminum (Al(OH)3) to redissolve as the
aluminate ion, AlOg, and the insoluble iron hydroxide remaining is sep-
arated by filtration or other means. Such an approach is not feasible
in the magnetic sludge process for several reasons, not the least of
which is the necessity to avoid coprecipitation of Mg(OH)2 above pH 9.6,
It might be possible to keep aluminum in solution near pH 8 through the
use of a chelating agent. However, it is doubtful whether any of the
known chelating agents are sufficiently specific for aluminum (i.e.,
iron would also be complexed to some extent), and in any event, use of
such materials would probably involve extraordinary conditions and pro-
hibitive costs. Our original concept of controlling the Al/Fe ratio,
by blending various mine waters or adding soluble ferrous iron salts,
probably offers the most expedient solution to this problem.
Cost Considerations
The conceptual magnetic sludge treatment process might provide the
following cost benefits over conventional lime neutralization.
With the utilization of magnetic separation equipment, costly
real estate for sludge settling lagoons would be unnecessary.
Since the process would yield a denser sludge, the cost per
unit weight for sludge storage and/or disposal should be
reduced.
Heavy-duty aeration equipment of the type commonly used in
lime treatment might be replaced with lighter equipment or
eliminated altogether.
The magnetic sludge might have some saleable value to offset
treatment costs.
On the other hand, certain economic disadvantages are inherent in the
magnetic sludge process:
Reagent (lime) costs would be essentially the same as for
conventional lime treatment.
The more sophisticated flowsheet for the process, indicated
by Figure 8, suggests that a higher capital investment might
be required. In particular, the purchase and installation of
specialized equipment such as a continuous centrifuge or mag-
netic separators would tend to make the process initially
more expensive than direct lime neutralization.
57-
-------
The cost of sludge heating would tend to increase operating
expenditures. While at first glance the cost of thermal
energy for the process might seem to be prohibitive, it may
actually be quite reasonable when one realizes that only the
sludge itself (which has presumably undergone an initial con-
centration/dewatering step) and not the bulk of the mine
drainage will be heated. For example, in the experiment
described earlier involving the preparation of a magnetic
sludge for use in seeding experiments, a volume of 2.5 gal of
sludge was obtained from an initial 90 gal batch of the Key-
stone mine drainage after lime treatment and overnight gravity
settling. This 2.5 gal of sludge was separated and further
processed to yield 0.53 gal of magnetic sludge. By direct
proportion," the amount of sludge from the Keystone discharge
requiring thermal treatment would be 27-8 gal or 231.8 Ib per
1,000 gal of the raw mine water. If it is assumed that the
sludge must be heated from $k F (12 C) to 19^ F (90 C), the
energy requirement to heat the sludge (assuming the specific
heat of the sludge as that of water, i.e. 1 Btu/lb/°F) will
be 32,^52 Btu. If the average calorific value of bituminous
coal is taken as 12,000 Btu/lb, then approximately 3 Ib of
coal will be consumed per 1,000 gal of raw mine water. If
the cost of coal is taken as 0.5^ per Ib, the cost for sludge
heating will be about 1.5^ per 1,000 gal in this particular
case. Note that in the example given, no mechanical dewater-
ing (e.g. centrifugation) of the sludge was attempted prior
to heating; such practice would have the effect of reducing
sludge heating costs even further.
In summary, based on available experimental evidence it would appear
that a magnetic sludge treatment process for coal mine drainage is
feasible. The limitations of such a process have been defined, and
certain restrictions (e.g. magnesium interference) can be minimized
through proper process design and control of reaction variables. From
a cost standpoint, the magnetic sludge process appears to be competitive
with conventional lime neutralization, although a meaningful cost evalu-
ation can only be made after further developmental work on the process.
Amenability of Coal Mine Waters to Magnetic Sludge Treatment
Because of the restrictions involving aluminum contents, many coal mine
waters may not be amenable to treatment by a magnetic sludge process.
This thought led to a review of two WQO-EPA surveys of coal mine drain-
age sources within particular watershed areas. Each report contains a
considerable amount of tabulated analytical data on the various mine
discharges sampled.
The following criteria were used in analyzing the data from these re-
ports: Only those discharges containing 7 ppm or more iron were con-
sidered, based on current Pennsylvania stream standards. A mine water
58.
-------
was considered treatable by the magnetic sludge process if the Al/Pe
molar ratio was 0.18 or less, a value indicated as a practical limit by
previous experimental work. A mine water with an Al/Pe molar ratio
between 0.19 and 0.26 was considered a borderline case (the value of
0.26 corresponds approximately to the point above which magnetic re-
sponse decreased sharply with increasing Al/Fe ratios; see Figure k).
Since no data were given on the oxidation state of iron in the mine
waters, it was assumed that iron was principally in the ferrous state.
The first report deals with coal mine drainage pollution in the McMahon
Creek watershed in Ohio. (11) The McMahon Creek watershed drains 91.2
square miles of Belmont County in southeastern Ohio; McMahon Creek it-
self, which is "seriously degraded" by mine drainage pollution, flows
into the Ohio River at Bellaire, Ohio.
Data on 51 coal mine discharges were tabulated. Of this number, ^2 con-
tained > 7 ppm iron (one source was not included in the present analysis
since no flow data were given). These h-2. discharges contained an aver-
age iron content of 589.0 ppm Fe (range: 10.3 to 3,228 ppm), an average
aluminum content of 90.8 ppm Al (range: 1.0 to 798 ppm), and the aver-
age Al/Fe molar ratio was 0.76 (range: 0.01 to 17.06). Eleven of the
lj-2 discharges (26.2 percent) contained less than 100 ppm Fe, and some-
what more remarkable is the fact that 8 of the ^2 discharges (19.0 per-
cent) contained more than 1,000 ppm Fe.
The data revealed that 15 of the k2 discharges (35.7 percent) would be
treatable by a magnetic sludge process according to the criteria men-
tioned above, and an additional 3 discharges were borderline cases.
Twelve of the 15 treatable discharges are from inactive drift mines,
two are from inactive strip mines, and one is from a railroad cut.
The 15 treatable discharges are responsible for 16.6 percent (3,168
Ib/day) of the total net acidity load and ^4-0.6 percent (1,671.2 Ib/day)
of the total iron load contributed by the ij-2 selected discharges to the
McMahon Creek watershed. If the 3 borderline cases are included, these
percentages are increased slightly to 17.0 percent and *tl.7 percent,
respectively.
Seven of the 15 treatable discharges flow directly into McMahon Creek,
and represent 68.5 percent (2,68U Ib/day) of the total net acidity load
and 76.8 percent (1,^32.8 Ib/day) of the total iron load contributed by
all discharges flowing directly into McMahon Creek. Of further interest
is the fact that 3 of the 15 treatable discharges are among the 12
"principal mine drainage sources" listed in the report as contributing
the largest acidity loadings to the watershed (over 92 percent). These
3 discharges, potentially treatable by a magnetic sludge process,
account for only 1^.0 percent (2,505 Ib/day) of the total net acidity
load but 37-1 percent (1,32^.0 Ib/day) of the iron load attributed to
these 12 "principal" discharges.
59-
-------
It was noted from the data that, generally, those discharges with the
highest Al/Fe molar ratios were from refuse piles.
The second report reviewed deals with coal mine drainage pollution in
the Sewickley Creek watershed in Pennsylvania, (12) and is perhaps more
pertinent to the present discussion since some of the sources listed
were sampled and utilized routinely during these studies (Table l).
The Sewickley Creek watershed drains 169.0 square miles of Westmoreland
County in southwestern Pennsylvania; Sewickley Creek is a major tribu-
tary of the Youghiogheny River, which is a tributary to the Monongahela
River. Thus, as was the case with McMahon Creek in Ohio, coal mine
drainage in the Sewickley Creek watershed utimately flows into the Ohio
River.
Data on 6l coal mine discharges were tabulated. Of this number, k8
contained > 7 ppm iron (the data from one source with an apparent Al/Fe
ratio of 31.3 were rejected as being possibly anomalous). These k&
discharges contained an average iron content of 128.9 PPm Fe (range:
7.9 to 9*4-8.0), an average aluminum content of kl.2 ppm Al (range: 1.3
to iV/.O ppm), and the average Al/Fe molar ratio was 1.13 (range: 0.06
to 4.85). Twenty-nine of the kQ discharges (60.k percent) contained
less than 100 ppm Fe
The data revealed that 6 of the 48 discharges (12.5 percent) would be
amenable to treatment by a magnetic sludge process, while an additional
3 discharges were borderline cases. One of the borderline cases is a
rather large discharge from an active underground mine. The other 2
borderline cases and the 6 potentially treatable cases are all dis-
charges from inactive underground mines of the shaft or drift type.
The 6 treatable discharges are responsible for 37=5 percent (27,278
Ib/day) of the total net acidity load and 58.6 percent (13,456 Ib/day)
of the total iron load contributed by the kQ selected discharges to the
Sewickley Creek watershed. If the 3 borderline cases are included,
these percentages are increased substantially to 77.2 percent and
89.7 percent, respectively.
It should be noted that of the 10 principal discharges contributing
97-0 percent (70,454 Ib/day) of the total net acidity load to the water-
shed, 3 are treatable and 2 are borderline cases. As a matter of fact,
the second and third major discharges in terms of daily acidity loads
are, respectively, the Keystone and Brinkerton discharges. Together
they contribute 38 percent of the total acid load and 52 percent of the
total iron load to the Sewickley Creek watershed. The preparation of
dense, strongly magnetic sludges from these two discharges has been
successfully demonstrated during the present studies (see, for example,
Table l4). '
The major findings from this brief literature review are represented
schematically in Figure 10. An examination of this pictorial representa-
tion will reveal the fallacy of judging the practicality of the magnetic
60.
-------
ON
H
Legend:
PERCENT
Amenable to Treatment
Borderline Cases
Not Amenable to Treatment
100-,
90-
80-
70-
60-
50-
40-
30-
20-
10-
0-
SEWICKLEY SEWICKLEY SEWICKLEY
McMAHON McMAHON AAcMAHON
35.7
7.1
12.5
6.2
81.31
40.6
1.1
58.3
57.2
NUMBER OF DISCHARGES
58.6
31.1
10.3
IRON LOAD
16.6
0.4
37.5
83.0
39.7
22.8
NET ACIDITY LOAD
Bituminous Cool Research, Inc. 2038G26
Figure 10. Schematic Representation Showing Amenability of Coal Mine
Discharges from Two Watersheds to Magnetic Sludge Treatment
-------
sludge process based solely upon the number of potentially treatable
discharges, without any consideration of the quality or volume of those
discharges. Based on the data for the McMahon Creek and Sewickley
Creek watershed areas, it appears that a magnetic sludge process would
indeed be feasible for treatment of several major discharges in these
watersheds, and it is felt that further work on the magnetic sludge
process could be readily justified on that basis.
Future Developmental Work
Work in the area of magnetic sludge preparation was limited to rela-
tively small, bench-scale batch experiments, and further developmental
work is clearly desirable. Although such work was originally planned,
revisions in scope during the course of the program necessitated that
these studies be abandoned. Some suggestions regarding optimization of
the process conditions are listed below, in the event that additional
developmental work is undertaken at a later date.
Stirring rate is a variable which conceivably could have an
effect on the properties of the magnetic sludge. Further
work would be desirable to ascertain the effect of stirring
rate in the process or, in fact, whether stirring is actually
necessary in the process during sludge heating.
The effect of sludge heating time should be investigated. In
the DuPont process, mentioned earlier, the optimum heating
period was found to be h hours. In our own experiments,
heating periods of 30 and 60 minutes were employed without any
apparent difference in the properties of the sludge. As a
first approach, sludge samples could be withdrawn periodically
during a protracted heating period and changes in solids con-
tent, magnetic response, etc., could be observed with time.
Although magnetic separation has been mentioned as an attrac-
tive method for magnetic sludge dewatering, no actual tests
were conducted. Since the magnetic sludge from coal mine
drainage may have quite different properties (e.g. particle
size, specific gravity, etc.) than materials ordinarily han-
dled by wet magnetic separators, as for example in a heavy
media circuit, magnetic separation tests should be conducted
at an early date if future developmental work is undertaken.
The probable increase in sludge solids contents as a result
of magnetic separation is also a factor worthy of further
investigation.
The procedure for preliminary sludge concentration ahead of
the heating step deserves further attention. Although re-
sults of some work in this area are contained in this report,
it is felt that additional work is necessary. For example,
perhaps the use of coagulant aids in conjunction with gravity
62.
-------
settling would prove as good or better than any of the
methods tried so far-
The aeration rate, like stirring rate, might influence sludge
properties during conversion of the sludge to a magnetic form.
Therefore, tests should be conducted to optimize the aeration
rate in the process. Aeration during heating may actually
not be necessary; in fact, it has been observed in past tests
that conversion of the sludge to a magnetic form sometimes
occurs readily with heating alone, especially if the sludge
has been first concentrated by mechanical dewatering.
Sludge Conditioning Studies
Coagulant Aid Studies
Electrophoretic mobility measurements showed that the sludge particles
formed by lime neutralization of the WQO-EPA synthetic coal mine water
had an electrostatic charge of +17 to +25 millivolts. Therefore, the
use of anionic coagulant aids was indicated. The objective in coagu-
lant aid addition is to add just enough of the coagulant to reduce the
electrostatic forces of mutual repulsion to a minimum, i.e., to or near
a value of zero (the isoelectric point). When this is accomplished,
the particles are more susceptible to agglomeration and tend to settle
more rapidly. It was thought that this more rapid settling, together
with the possibility of greater sludge compaction, might result in the
formation of a denser sludge after settling.
In these studies, 12 anionic polyelectrolytes and 2 nonionic polyelec-
trolytes were tested. These coagulant aids are representative of the
type commercially available today. The coagulant aids tested are listed
in Table 15, which also shows the concentration of each coagulant aid
necessary to achieve the isoelectric point. Six of the anionic poly-
electrolytes were effective at concentrations below 1 ppm. The zeta
potential curves for these six most effective coagulant aids are shown
in Figure 11. Another group of five coagulant aids were effective with-
in the concentration range 1 to 5 ppm, and results with these samples
are shown in Figure 12. Although the curves in Figures 11 and 12 would
indicate that charge reversal should be effected by further increased
concentrations of the coagulant aids, it was actually found for nearly-
all coagulant aids tested that further additions up to the 20 ppm level
did not affect the apparent sludge zeta potential, i.e., it remained at
zero.
Zeta Floe WA is rather unique in that, in addition to containing a
"strongly anionic" polyelectrolyte, it also contains an appreciable
percentage of insoluble aluminum silicate as a bulk additive. On a
weight-for-weight basis, therefore, one would expect that a substan-
tially larger amount of this coagulant aid would be required to achieve
the isoelectric point compared with the other anionic coagulant aids
63.
-------
TABLE 15. EFFECT OP COAGULANT AIDS ON SLUDGE FROM LIME TREATMENT
OF WQO-EPA SYNTHETIC COAL MINE WATER
Additive
Magnifloc 836A
Magnifloc &37A
Polyfloc 1130
Magnifloc 835A
Genfloc 155
Superfloc 16
Nalcolyte 673
Calgon 2^0
Polyhall M-295
Reten AM
Primafloc A-10
Zeta Floe WAt
Superfloc 1274-
Polyox WSR-3014-
Source
American Cyanamid Co.,
Industrial Chemicals Division
American Cyanamid Co.,
Industrial Chemicals Division
Betz Laboratories, Inc.
American Cyanamid Co.,
Industrial Chemicals Division
General Mills, Inc.
American Cyanamid Co.,
Industrial Chemicals Division
Nalco Chemical Co.
Calgon Corporation
Stein, Hall & Co., Inc.
Hercules, Inc.
Rohm & Haas Cb.
Narvon Mining & Chemical Co.
American Cyanamid Co.,
Industrial Chemicals Division
Union Carbide Corp.,
Chemicals Division
Concentration at
Isoelectric Point,
ppm
0.1
0.5
0.6
0.7
0.7
0.8
i.o
i.o
1.5
2.0
5-0
•K
* Contains insoluble aluminum silicate.
-I- Nonionic potyelectrotyte - all other coagulant aids are
anionic polyelectrolytes.
* Isoelectric point never achieved; sludge remained
electropositive.
6k.
-------
ON
0.0
Polyfloc 1130
Superfloc 16
0.2 0.4 0.6
COAGULANT AID CONCENTRATION, ppm
0.8 1.0
Bituminous Coal Research, Inc. 2038G27
Figure 11. Effect of Various Anionic Coagulant Aids on Zeta
Potential of Synthetic Coal Mine Drainage Sludge
-------
+ 25-1
+ 20-
> +15
E
Z
LU
o +10
a.
UJ
N
+ 5-
Primafloc A-10
0.0
1.0 2.0 3.0
COAGULANT AID CONCENTRATION, ppm
4.0 5.0
Bituminous Coal Research, Inc. 2038G2(
Figure 12. Effect of Varjous Anionic Coagulant Aids on Zeta Potential
of Synthetic Coal Mine Drainage Sludge
-------
tested, and this is borne out by the results, shown in Figure 13- On
the other hand, according to the manufacturer's literature, competitive
polyelectrolytes "generally sell for ten to twenty times as much as
Zeta Ploc on a cost per pound basis." On this basis, therefore, Zeta
Floe WA could probably be considered comparable in efficiency with those
coagulant aids whose zeta potential curves are shown in Figure 12.
As expected, the isoelectric point was not achieved using the two non-
ionic polyelectrolytes, Superfloc 12? and Polyox WSR-301, as indicated
by Figure lU. Relatively large dosages (> 50 ppm) appeared to result
in a stable zeta potential in the range +7 to +9 millivolts.
The effects of coagulant aids alone on settling rates, sludge volumes,
and solids contents of the sludge obtained by lime neutralization of the
WQO-EPA synthetic water are shown by the data in Table 16. These re-
sults indicate that in practically every instance, sludge settling rates
increased as the coagulant aid dosage was increased. This effect was
especially pronounced when a relatively large excess of the coagulant
aid (e.g. 20 ppm) was employed. These settling rate increases were also
accompanied by marked improvements in the clarity of the supernatant
liquid, as reflected by the data on turbidity. The one exception in
these tests was the nonionic Polyox WSR-301, which failed to effect any
appreciable changes in sludge settling rate or supernatant liquid tur-
bidity up to a dosage of 100 ppm. Use of the other nonionic coagulant
aid, Superfloc 127, led to results quite similar to those obtained with
the anionic polyelectrolytes.
When the anionic polyelectrolytes were employed at the minimum concen-
trations necessary to achieve the isoelectric point (see Table 15), the
fastest settling was obtained with Genfloc 155, while three coagulant
aids (Polyfloc 1130, Genfloc 155, and Nalcolyte 673) resulted in minimum
turbidity values of 3 JTU. However, the data are characterized more by
similarities than by differences, and where differences do appear they
are often so small as to be within the limits of experimental error.
Thus, it should be emphasized that these results do not imply an endorse-
ment for any particular coagulant aid, and any such judgment should be
predicated on a much more exhaustive series of flocculation tests under
conditions more closely approaching those used in actual plant practice.
At best, our data indicate that there are not significantly large dif-
ferences among the results obtained with the various anionic polyelec-
trolytes tested; in addition, the data suggest that the use of excess
amounts of the coagulant aids (i.e., dosages above the minimum necessary
to achieve the isoelectric point) is desirable, and the additional rea-
gent cost thereby incurred might be justified on the basis of substan-
tially improved results.
Slight increases in 30-minute settled sludge volumes were observed with
increasing coagulant aid dosages. This result is quite likely due to
the fact that a greater proportion of the suspended solids were settled
in the presence of the coagulant aids.
67.
-------
+ 20-,
+15-
>
E +10-
z
LU
I—
O
°- +5-1
LU
N
0-
-5
~T
o
i l
10 20
ZETA FLOC WA CONCENTRATION, ppm
1 1
30 40
Bituminous Coal Research, Inc. 2038G2
Figure 13. Effect of Zeta Floe WA on Zeta Potential of
Synthetic Coal Mine Drainage Sludge
-------
ON
MD
+20-1
+15-
+10
o
a- +5
LLJ
N
0-
-5
Polyox WSR-301
Superfloc 127
10 20 30
COAGULANT AID CONCENTRATION, ppm
40 50
Bituminous Coal Research, Inc. 2038G30
Figure 14. Effect of Two Nonionic Coagulant Aids on Zeta Potential
of Synthetic Coal Mine Drainage Sludge
-------
TABLE 16. EFFECTS OF COAGULANT AIDS ON SETTLING RATES, SLUDGE VOLUMES,
AND SOLID CONTENTS OF SYNTHETIC COAL MINE WATER SLUDGE
Coagulant Aid
Magnifloc 836A
Magnifloc 837A
Polyfloc 1130
Magnifloc 83 5A
Coagulant Aid.
Dosage, ppm
Sludge Settling
Rate, ml/min
30-Minute
Settled Sludge
Volume, ml
Sludge Solids
Content, Weight
Percent
Turbidity of
Supernatant
Liquid, JTU
0
0.05
0.10*
1.0
5-0
20.0
0
0.30
0.50*
1.0
5.0
20.0
0
0.50
0.60*
0.80
1.0
20.0
0
0.20
0.70*
1.0
5-0
20.0
19
28
4o
±9
71
78
19
28
36
ill
78
84
19
33
33
33
3^
98
19
28
^9
57
78
84
14
14
14
17
19
20
13
14
14
14
17
17
15
16
18
18
18
17
14
13
13
14
17
16
o.4o
0.45
0.51
0.47
0.53
0.45
0.37
0.39
0.43
0.4l
0.44
0.33
0.53
0.44
0.47
0.49
0.37
0.54
0.49
0.50
0.39
0.41
0.49
0.41
15
8
8
3
3
l
17
5
6
2
2
1
18
4
3
2
2
2
21
6
5
3
5
3
* Minimum concentration required to achieve the isoelectric point.
-------
TABLE 16. EFFECTS OF COAGULANT AIDS ON SETTLING RATES, SLUDGE VOLUMES,
AMD SOLID CONTENTS OF SYNTHETIC COAL MINE WATER SLUDGE (Cont.)
Coagulant Aid
Genfloc 155
Superfloc 16
-3
H
Nalcolyte 673
Calgon
Coagulant Aid
Dosage, ppm
Sludge Settling
Rate, ml/min
30-Minute
Settled Sludge
Volume, ml
Sludge Solids
Content, Weight
Percent
Turbidity of
Supernatant
Liquid. JTU
0
0.30
0.70*
1.0
5.0
20.0
0
0.50
0.80*
1.0
5-0
20.0
0
0.50
1.0*
3-0
5-0
20.0
0
0.50
1.0*
3-0
5-0
20.0
18
52
63
63
75
78
20
33
^3
ho
79
79
20
33
^5
69
71
79
20
30
32
38
39
77
14
16
17
17
18
17
Ik
1k
11
Ik
15
15
13
15
15
17
19
19
Ik
13
16
16
16
16
0.6l
0.^7
0.29
oAl
0.^3
0.31
O.kk
0.58
0.76
0.55
0.70
0.82
0.20
0.27
0-35
0.29
0.25
0.32
0.38
0.55
0.52
0.52
0.59
0.76
18
3
3
2
2
6
18
6
11
k
k
6
16
5
3
2
3
2
17
7
5
2
2
3
* Minimum concentration required to achieve the isoelectric point.
-------
TABLE 16. EFFECTS OF COAGULANT AIDS ON SETTLING RATES, SLUDGE VOLUMES,
AND SOLID CONTENTS OF SYNTHETIC COAL MINE WATER SLUDGE (Cont.)
Coagulant Aid
Polyhall M-295
Reten AM
ro
Primafloc A-10
Zeta Floe WA
Coagulant Aid
Dosage, ppm
0
0.50
1.5*
3-0
5-0
20.0
0
1.
2.
3-
5-
o
0*
0
0
20.0
0
1.0
3-0
5.0*
10.0
20.0
0
10.0
20.0
30.0
1+0.0*
50.0
Sludge Settling
Rate, ml/min
21
27
33
56
59
65
22
35
60
69
61
80
20
22
30
30
1*3
23
28
37
37
30-Minute
Settled Sludge
Volume, ml
Sludge Solids
Content, Weight
Percent
Turbidity of
Supernatant
Liquid, JTU
11+
Ik
15
16
17
17
15
15
15
16
16
17
Ik
15
15
15
16
17
Ik
Ik
Ik
15
15
16
0.36
o.ko
0.33
0.36
0.57
0.39
0.30
0.32
0.25
0.32
0.85
0.14.5
0.55
0.51
0.68
0.69
0.62
0.68
0.35
0.21
0.31+
0.1+6
oM
0.36
21
1+
k
2
3
2
16
2
5
1
1+
1
23
16
12
9
8
7
16
5
1+
3
k
k
Minimum concentration required to achieve the isoelectric point.
-------
TABLE 16. EFFECTS OF COAGULANT AIDS ON SETTLING RATES, SLUDGE VOLUMES,
AND SOLID CONTENTS OF SYNTHETIC COAL MINE WATER SLUDGE (Cont.)
Coagulant Aid
Superfloc 127
Polyox WSR-301
U)
Coagulant Aid
Dosage, ppm
0
0.10
0.50
1.0
5-0
20.0
0
3-0
5-0
20.0
50.0
100.0
Sludge Settling
Rate, ml/min
20
3^
39
69
81
85
21
2k
2k
29
25
26
30-Minute
Settled Sludge
Volume, ml
13
13
13
13
15
12
13
13
13
13
13
13
Sludge Solids
Content, Weight
Percent
0.51
0-36
O.kO
0.25
O.kl
0.6k
0.6k
0.60
O.k7
o.k3
o.ko
0.55
Turbidity of
Supernatant
Liquid, JTU
22
10
6
6
k
3
31
28
26
16
2k
28
-------
With regard to sludge solids contents, there is some indication of a
trend toward slightly higher solids contents with increasing coagulant
aid dosage. However, it is felt that such an effect is negligible, if
present at all. This is apparent when one considers the solids content
data for the control samples to which no coagulant aid was added. These
values range from a low of 0.20 percent solids to a high of 0.6*4- percent
solids, with a mean deviation of 0.10. Since practically all of the
values determined (with and without coagulant aids) lie within this
range, it is evident that no significant increases in sludge solids
contents were obtained through the use of coagulant aids. In summary,
then, the combined data on sludge volumes and solids contents indicate
that the use of coagulant aids alone will not result in increases in
sludge density.
Filter Aid Studies
The purpose of this series of tests was to determine whether or not
sludge density could be increased significantly by the addition of
various bulk additives or sludge "builders." These materials might
also be of use as filtration aids, affording more rapid and efficient
sludge dewatering.
The materials tested are listed in Table 17- Results of preliminary
tests with the fly ash indicated slightly increased settling rates and
decreased turbidity when the filter aid was added to the synthetic mine
water before rather than after the lime, during sludge preparation.
Results of a previous BCR study (l) on the use of coagulant aids with a
ferrous hydroxide sludge indicated a slight advantage was gained in
terms of a lower coagulant aid requirement when the fly ash was added
before the lime. This earlier study also involved tests with fly ash-
(and other "inert solids") in combination with Calgon 2^0, and for the
sake of comparison with these previous results, Calgon 2kO was likewise
chosen for use in the present studies.
Preliminary electrophoretic mobility measurements indicated that the
presence of fly ash alone had a somewhat erratic effect on the zeta
potential of the sludge prepared by lime treatment of the WQO-EPA
synthetic mine water, although there was a general tendency for the
sludge zeta potential to decrease as the amount of fly ash added was
increased. Relatively high concentrations of fly ash, on the order of
2 to 5 grains per liter, were required to lower the zeta potential of
the sludge to the isoelectric point.
The results of tests using the filter aids alone are shown in Table 18.
The data reveal that as the amount of filter aid added to the suspen-
sion was increased, there was an increase in sludge solids contents and
a decrease in settled sludge volume in practically all cases, indicating
that an overall increase in sludge density had been effected. In addi-
tion, sludge settling rates were increased by a factor of about 1.5 to
2 over those measured in the absence of the filter aids (Table 16).
-------
TABLE 17. MATERIALS USED IN FILTER AID STUDIES
Filter Aid
Source
Density,*
Remarks
ash
Blast furnace slag
(agricultural grade)
Magnetite
(heavy media grade)
Sawdust
Gypsum
Sand
"Red Dog"
Magnetite
(pigment grade)
Colfax Power Station
Duquesne Light Co.
U.S. Steel Corp.
Duquesne Works
Mineral Mills, Inc.,
"Floatkleen," Grade 6
Rosensteele and Kunkle Co.,
Spring Church, Pa.
Fisher Scientific Co.
Rosensteele and Kunkle Co.,
Spring Church, Pa.
Sewickley Township
Westmoreland County, Pa.
Columbian Carbon Division,
Cities Service Co.
Mapico Black, Lot 1296
2.29
2.81
5.10
2.83
2.75
5.62
Sieved as received to minus 200
mesh.
Dried at 105 C, ground and sieved
to minus 200 mesh.
Used as received, 98+ percent minus
200 mesh.
Sieved as received from saw mill to
minus 20 mesh.
Reagent grade CaS04-2H^O powder,
used as received.
Dried at 105 C, ground and sieved
to minus 200 mesh.
Product of coal refuse combustion]
ground and sieved to minus 200
mesh.
Used as received, 99-9 percent minus
325 mesh (predominant particle
size range 0.2 to 0.8 microns).
Determined with a Beckman Model 930 Air Comparison Pycnometer.
-------
TABLE 18. EFFECTS OF FILTER AIDS ALONE ON PROPERTIES OF SYNTHETIC COAL MINE WATER SLUDGE
Filter Aid
Fly ash
Slag
Magnetite
(heavy media
grade )
Sawdust
Gypsum
Sand
"Red Dog"
Magnetite
(pigment
grade)
Filter Aid
Dosage, ppm
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
Sludge Settling
Rate, ml/min
U5
k5
ko
29
35
^3
35
30
35
k5
ko
kl
30
32
35
26
28
29
27
28
29
32
38
33
30-Minute
Settled Sludge
Volume, ml
11
9
16
12
11
8
7
12
12
Ik
12
12
12
10
9
10
10
9
12
10
9
Sludge Solids
Content,
Weight Percent
1.22
3-23
6.82
0.91
1.88
7.21
0.9^
k.k2
8.50
0.70
0.93
2.3*1
0.68
I.kk
5.78
1.07
1.72
3-82
1.80
3-30
6.32
1.13
2.90
5.66
Sludge Zeta
Potential,
mv
+ 11
+ 16
0
+ 12
+ 13
+ 13
+ 111
+ 17
+ 13
+ 13
+ 16
+ 17
+ 18
+ 17
+ 1
+ 12
+ 12
-i- 12
-H 11
+ 15
+ 17
+ 29*
+ 23*
+ 2k*
Turbidity of
Supernatant
Liquid, JTU
2k
32
37
27
35
36
27
k2
Ul
26
2k
2k
22
18
12
19
32
33
27
29
35
3^
31
19
Presence of the filter aid apparently interfered with electrophoretic
mobility measurements, and values may be anomalous.
-------
On the other hand, turbidity data indicate that an appreciable amount
of material remained in suspension after the 30-minute settling period.
The data on sludge zeta potential indicate that the electrophoretic
mobility of the sludge was affected in a rather erratic and inconsistent
way in the presence of the filter aids alone. Only the fly ash and
gypsum seemed to have a definite lowering effect on the sludge zeta
potential at the highest concentrations employed (5,000 ppm).
The results of tests using the filter aids in conjunction with 0.5 ppm
of the anionic polyelectrolyte Calgon 2^0 are shown in Table 19. As
expected, use of the coagulant aid resulted in further increases in
sludge settling rates and decreases in supernatant liquid turbidity,
compared with the results of the same tests where the coagulant aid was
absent (Table 18). In addition, it is significant that the isoelectric
point was achieved in nearly all cases with 0.5 ppm of Calgon 2kO.
This is only one-half the amount that was required in the absence of
the filter aids (Table 15). Thus, there is evidence that for a given
concentration of the coagulant aid, the combination of the filter aid
and coagulant aid has a more pronounced effect on the sludge zeta poten-
tial than does the coagulant aid alone. This finding is in agreement with
that reported previously,(l) however, similar tests with other coagulant
aids would be necessary to establish whether it applies generally in
all cases.
In view of the settling rate data from tests with Calgon 2*4-0 alone
(Table 16), it is felt that the filter aid-Calgon 2*4-0 combination re-
sulted in significant increases in sludge settling rate. With regard
to settled sludge volumes and solids contents, the differences between
the results with filter aids alone and with the filter aid-coagulant
aid combination are not sufficiently large to be meaningful.
The effects of the filter aids alone on sludge solids contents are
illustrated graphically in Figure 15. The curves show a generally reg-
ular increase in solids contents with increasing filter aid concentra-
tion for most of the materials evaluated. The two notable exceptions
are the curves for gypsum and blast furnace slag; it is apparent that
the increases in sludge solids contents become comparatively larger as
the concentrations of these two materials increase. The role of gypsum
in promoting coal mine drainage sludge densification has been implied
by a number of workers, and the presence of gypsum seed crystals has
been suggested as the factor responsible for the effectiveness of
sludge densification by recycling techniques.(13) In addition, rather
striking evidence of the effects of coprecipitated gypsum on sludge
density was observed during the magnetic sludge studies (for example,
see Figure 3). Thus, it is perhaps not surprising that results of the
present studies indicate that gypsum is an effective sludge densifier.
The other material which appears to be effective, blast furnace slag,
has to our knowledge not been employed previously in an application of
this type. Our results suggest that its use should be further explored,
particularly since it is a relatively cheap waste product which also
contains a reactive alkaline component (see below).
77.
-------
co
Filter Aid
Fly ash
Slag
Magnetite
(heavy media
grade)
Sawdust
Gypsum
Sand
"Red Dog"
Magnetite
(pigment
grade)
TABLE 19.
Filter Aid
Dosage, ppm
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
500
2,000
5,000
EFFECTS OP FILTER AIDS PLUS 0.5 PPM CALGON 2^0 ON PROPERTIES
OF SYOTHETIC COAL MINE WATER SLUDGE
Sludge Settling
Rate, ml/min
52
48
60
26
56
59
48
65
90
65
63
70
38
55
35
43
50
35
42
65
37
43
44
30-Minute
Settled Sludge
Volume, ml
11
9
16
11
10
7
7
12
12
13
12
11
12
9
9
10
10
9
ll
8
7
Sludge Solids
Content,
Weight Percent
1.11
3.20
6.64
0.87
2.43
6.53
1.15
4.31
8.60
0.68
0.90
2.47
0.64
1.4o
4.86
,26
,46
6.29
18
6.70
1-13
2.91
5-99
Sludge Zeta
Potential,
mv
+ 7
0
0
0
+ 1
+ 10
+ 10
+ 1
+ 1
1
1
0
+ 1
+ 1
0
0
0
0
0
0
0
12*
24*
16*
Turbidity of
Supernatant
Liquid, JTU
16
12
14
17
25
12
25
11
13
8
11
10
10
11
12
19
16
14
15
20
17
18
Presence of the filter _aj.d apparently interfered with electrophoretic
mobility measurements, and values may be anomalous.
-------
z
LLJ
U
I
O
z
O
u
t/>
Q
_j
O
O
Q
4-
3-
2-
1-
AAagnetite
(Heavy Media
Grade)
V '
O \
Magnetite
(Pigment
Grade)
Sawdust
T
5.5
0 1.0 2.0 3.0 4.0 5.0
FILTER AID CONCENTRATION, GRAMS PER LITER
Bituminous Coal Research, Inc. 2038G31
Figure 15. Effects of Filter Aids Alone on Solids Content
of Synthetic Coal Mine Drainage Sludge
79-
-------
Several of the filter aids tested exhibited various peculiarities which
are worthy of note. There was a tendency for all of the filter aids to
separate from the sludge during settling, although this tendency was
reduced to some extent when the coagulant aid was employed. Sludges
containing fly ash or slag were judged to be more homogeneous than the
others.
The f3y ash had no measurable effect on the initial pH of the synthetic
mine water, despite the fact that it produced an alkaline solution when
added to deionized water (pH 9.U at 500 ppm fly ash). The addition of
either the sand or the "red dog" to deionized water resulted in slight
increases in pH (on the order of 1.0 and 0.5 pH units, respectively, at
the 5,000 ppm level). However, as was found with the fly ash, this
effect was apparently masked in the presence of the synthetic mine
water, and only negligible changes in pH (< 0.1 unit) were observed
under these conditions. The blast furnace slag was the only filter aid
tested which had a significant effect on the initial pH of the synthetic
mine water. For example, an increase of 2.6 pH units was observed when
5,000 ppm of slag was added to the synthetic mine water.
Although the fly ash sample contained a small percentage of ferro-
magnetic material, sludges containing the fly ash exhibited only very
weak magnetic response when tested with a small hand magnet. By contrast,
sludges containing magnetite were strongly magnetic, as expected. It
was possible to separate much of the magnetite from the sludge by manip-
ulation of the hand magnet. Interestingly, however, the sludge parti-
cles themselves were moderately magnetic, and this effect seemed greater
when the coagulant aid was present. These observations indicate that
the smaller particles of magnetite were entrapped within the sludge
floes, giving rise to bulk magnetic properties in the sludge.
The blast furnace slag had' an alkaline component, as evidenced by its
definite effect on pH when added to the synthetic mine water. The lime
requirement necessary to attain a terminal pH of 7.8 during sludge
preparation was reduced to 0.78, 0.7^-, and 0.60 times the stoichiometric
amount with slag additions of 0.5, 2.0, and 5-0 grams per liter, respec-
tively. A minor problem was encountered primarily at the lowest slag
concentration employed (0.5 g/l); when the slag was added to the syn-
thetic mine water, the suspension was still sufficiently acidic to de-
compose sulfides in the slag, and some H2S was evolved. At the higher
slag concentrations, this problem was much less evident, presumably
because the pH of the suspension increased rapidly enough to inhibit
sulfide decomposition.
The sawdust was probably the least satisfactory of the filter aids
tested. It behaved as a pH depressant, and some irregularities were
noted in the changes of the suspension pH with time after sawdust addi-
tion. This may have been due to the presence of resinous materials
leached from the wood, which was indicated by a considerable amount of
foaming at the liquid surface during aeration. A portion of the sawdust
also tended to float in the container, rather than settling with the
sludge.
80.
-------
The usefulness of pigment grade magnetite in this application must be
discounted since, in addition to its relatively high cost ($0.1? per
pound), problems were encountered in handling the material. With a
consistency much like that of carbon black, it tended to coat all sur-
faces •with which it came in contact. It also tended to foul the pores
of the air sparger, and considerable difficulty in subsequent cleaning
of the apparatus was experienced.
Filtration Tests
In the first series of filtration tests, each of the filter aids was
employed at a concentration of 2,000 ppm (2.0 g/l) in the synthetic
coal mine water sludge suspension. The results of these tests are sum-
marized in Table 20. The data represent averaged values from duplicate
experiments. The filter aids are listed in order corresponding to in-
creasing cake moisture of the filtered solids. It should be noted that
this ranking also corresponds to the filtration rate in terms of pounds
of dry solids/sq ft/hr. Thus, on the basis of cake yield and moisture
content, the more effective filter aids are at the top of the list.
The data for filtrate yields in Table 20 do not appear to conform to the
trends shown by the other results. It should be pointed out, however,
that of all the experimental measurements, the values for filtrate
yields were the least reproducible. Differences on the order of 10 ml
in filtrate volumes from duplicate experiments were not uncommon, and
this difference corresponds to 1.3 gal/sq ft/hr. Consequently, with the
possible exception of the result from the test with "red dog," it is
felt that there are essentially no significant differences among the
values for filtrate yields. In all of these tests, the filtrates
appeared to be quite clear and free of suspended solids.
Recent results reported by investigators of the Johns-Manville Corpora-
tion afford an interesting comparison with the data from the present
study. These workers observed filtration rates on the order of
30 gal/sq ft/hr and cake solids contents of between 25 and ^5 percent
during precoat filtration tests with diatomaceous earth (Celite 501) at
one field site, where various combinations of lime, limestone, and mag-
nesite were used to treat the coal mine discharge.(l^) Filtered sludges
obtained by lime-limestone or lime-magnesite treatment had noticeably
lower solids contents (29.6 and 2^.8 percent, respectively) than did
those obtained by treatment with the limestone-magnesite combinations.
These values compare favorably with those obtained using some of the
filter aids in the present study, where the sludges were produced by
lime treatment alone. On the other hand, our measured filtration rates
are considerably lower than those reported by the Johns-Manville workers,
It should be recognized, however, that there were significant differ-
ences in both the methods and materials employed in these two studies,
and comparisons among the results may not be entirely appropriate.
81.
-------
Co
ro
TABLE 20. EFFECTS OF VARIOUS FILTER AIDS ON THE FILTRATION
PROPERTIES OF SYNTHETIC COAL MINE WATER SLUDGE
Filter Cake Properties
Filtration Rate
Filter Aid*
Magnetite
(pigment grade)
Fly ash
"Red Dog"
Sand
Slag
Magnetite
(heavy media grade)
Sawdust
Gypsum
Control}-
Moisture,
Percent
61.7
70.7
73-1
76.8
78 A
81.3
85-9
86.5
87-3
Solids ,
Percent
38.3
29-3
26.9
23.2
21.6
18.7
14.1
13-5
12.7
Ib dry solids /sq ft/hr
0.265
0.201
O.llj-6
0.112
0.105
0.092
0.060
0.053
0.053
gal filtrate/sq. ft/hr
19.8
21.7
26.2
20.0
22.5
20.7
23.1
23.7
21.2
* Filter aid concentration was 2,000 ppm (2.0 g/l) in all tests.
4- Raw sludge only; no filter aid added.
-------
In the second series of filtration tests, two of the filter aids were
employed in successively increasing amounts to determine the effect of
filter aid concentration on dewatered sludge properties. The filter
aids chosen were blast furnace slag, based on its performance as indi-
cated by Figure 15, and "red dog," which performed well in the prelimi-
nary filtration tests (Table 20). The results of these tests are shown
in Table 21 and Figure 16. All data represent averaged values from
duplicate determinations. The data show an apparent increase in filter
cake solids contents with increasing filter aid concentration, and the
results with the "red dog" are slightly better than those with the slag.
With regard to filtration rates, a similar trend is indicated by the
data for filter cake yields, although the differences between the two
filter aids are not as pronounced. In fact, in view of the variations
in cake yields shown by Figure 16, it would appear that there are no
substantial differences between the effects of the two filter aids at
each concentration level up to and including 5.0 g/1. The sudden diver-
gence of the curves at 6.0 g/1 is unexpected, and indicates that either
one or both of the values at this concentration may be anomalous.
The data for filtrate yields show slight (although probably negligible)
increases in the volume of filtrate with increasing filter aid concen-
tration. As indicated by Figure 16, this•increase was rather regular
when slag was employed, but the results for "red dog" were erratic. In
fact, there is actually a tendency toward decreased filtrate volumes at
the higher "red dog" concentrations, and this situation might be ex-
pected in view of the greater cake accumulations (reflected by cake
yield data) at the higher concentrations of filter aid.
In summary, it is felt that the results of these filtration tests have
failed to show any dramatic improvements in sludge dewaterability
through the use of filter aids in the manner described (i.e., in a
"body feed" rather than as precoat materials). However, the slight im-
provements which were observed may be sufficient to justify the use of
some of the cheaper, more readily available materials for this applica-
tion, especially in cases where such materials may be considered margin-
ally useful waste products themselves.
Seeding Experiments
The results of experiments on seed crystal size are presented in
Table 22. Compared with the results for the control test where no seed
material was present, the data show that measurable increases in sludge
settling rate and decreases in settled sludge volume were effected by
the presence of the seed material. However, the similarities among the
results for the three particle size fractions indicate that there were
no significantly different effects on sludge properties due to seed
crystal size. In this regard, it should be noted that microscopic exam-
ination of the sieved material revealed that all size fractions actually
consisted of spherical agglomerates of even smaller sized particles.
Consequently, it is quite possible that many of these agglomerates were
83.
-------
OO
-p-
TABLE 21. EFFECTS OF FILTER AID CONCENTRATION ON THE FILTRATION PROPERTIES
OF SYNTHETIC COAL MINE WATER SLUDGE
Filter Aid Filter Cake Properties
Filtration Rate
Filter Aid
Control*
Slag
"Red Dog"
Concentration ,
g/1
—
0.5
2.0
3-0
4.0
5-0
6.0
0.5
2.0
3-0
4.o
5.0
6.0
Moisture,
Percent
87-3
84.8
78.4
74.8
71-5
66.7
63-5
81.8
73-1
67-4
62.6
60.3
56.4
Solids,
Percent
12.7
15-2
21.6
25.2
28.5
33-3
36.5
18.2
26.9
32.6
37.^
39-7
43.6
Ib dry solids/sq ft/hr
0.053
0.078
0.105
0.155
0.172
0.217
0.277
0.064
0.146
0.162
0.169
0.225
0.211
gal filtrate/sq, ft/hr
21.2
20.2
22.5
22.4
23-3
22.9
25.4
25.8
26.2
27.2
22.3
21.8
23-5
* Raw sludge only; no filter aid added.
-------
O u
U of.
LU
to o.
O
-; o
Q ""
Ho
CO
LL. CO
9" 51
s: o
< CO
a: o
45-i
40-
30-
20-
10-
40-,
30-
20-
10-
0-
30n
20-
I
1
•"Red Dog"
o Slag
—r~
5.0
6.0 6.5
1.0 2.0 3.0 4.0
FILTER AID CONCENTRATION, GRAMS/LITER
Bituminous Coal Research, lnc.2038G32
Figure 16. Effects of Filter Aid Concentration on Filtration Properties
of Synthetic Coal /Wine Water Sludge
85.
-------
TABLE 22. SYNTHETIC COAL MIKE WATER SEEDING EXPERIMENTS -
EFFECT OF SEED CRYSTAL SIZE*
o>
Seed Material Sludge Settling
30-Minute
Settled Sludge
Sludge Solids
Content,
Turbidity of
Supernatant
Test No.
if 18- 38
418-40
\
418-42
U18-44
Mesh Size4-
50 x 80
100 x 140
200 x 325
Control
Rate, ml/min
86
87
87
80
Volume, ml
125
130
125
145
Weight Percent
0.89
0.57
0.63
0.57
Liquid, JTU
14
14
12
12
* Anhydrous ferric oxide used as a seed material at a concentration of 100 ppm
in all tests except control. Stirring rate maintained at 300 rpm.
4- U.S. Sieve Series
-------
broken down by attrition during sludge preparation, even though care was
taken to minimize agitation of the suspension. The apparent absence of
a particle size effect must therefore be considered as a conditional
result subject to further experimentation.
The results of tests on seed crystal amount are shown in Table 23. The
data reveal a definite increase in sludge settling rate and solids con-
tent and a decrease in sludge volume with increasing amounts of seed
material added. Offsetting these improvements, however, is the fact
that residual turbidity increased significantly as the amount of seed
material was increased. This observation suggests that a considerable
portion of the relatively finely divided seed material remained in sus-
pension, i.e., the seed particles had failed to function as precipita-
tion nuclei for the growth of larger, heavier sludge floes. This like-
lihood is further supported by the fact that very similar overall
results were obtained in the series of tests with filter aids; compari-
son of the results in Table 18 at filter aid concentrations of 500 and
2,000 ppm, for example, shows that in most cases sludge volumes and
solids contents were of the same order of magnitude as those measured
using ferric oxide as a seeding material. (The differences in sludge
volumes by a factor of about 10 is explained by the fact that in the
filter aid studies, settling tests were conducted on 100-ml portions of
the sludge suspensions, whereas in the seeding tests the entire sample
was transferred to a 1,000-ml graduated cylinder. This same considera-
tion probably also explains the lack of agreement between measured
settling rates in the two series of tests.) Thus, it is concluded that
the relatively dense (about 5-1 g/ml) ferric oxide behaved more as a
sludge "builder," similar to the filter aids previously employed, than
as a seed material providing precipitation nuclei. The situation is
admittedly ambiguous, and interpretation of the results is complicated
by the fact that the filter aids were also introduced prior to lime
addition, and could therefore have provided nucleation sites during
sludge precipitation However, it is generally recognized that the
addition of a strong alkali to a solution containing iron results in the
extremely rapid precipitation of a disordered (ferrous hydroxide) or
amorphous (hydrous ferric oxide) precipitate, and it seems unlikely that
seeding materials would be beneficial under such conditions. Seed
crystals are ordinarily employed to provide sites for the growth of
ordered crystalline phases, often from pure solutions, in which condi-
tions of temperature, pressure, degree of supersaturation, and other
critical factors are carefully controlled.
The results of experiments on the effect of stirring rate are shown in
Table 2k, and the data indicate that within the rather narrow range of
stirring speeds employed, no significant changes in sludge properties
were apparent. The tendency towards slight increases in solids contents
(seeded tests) and decreases in sludge volumes (seeded and control
tests) with increasing stirrer speed may have been due to the production
of smaller particles at the higher stirring rates and, subsequently,
their more efficient packing in the settled sludge. The relative dif-
ferences in the results are not large, however, and may be within the
87-
-------
TABLE 23. SYNTHETIC COAL MINE WATER SEEDING EXPERIMENTS -
EFFECT OF SEED CRYSTAL AMOUNT*
00
CO
Concentration 30-Minute Sludge Solids Turbidity of
of Seed Sludge Settling Settled Sludge Content, Supernatant
Test No. Material, ppm Rate, ml/min Volume, ml Weight Percent Liquid, JTU
108-1*4
108-42
1+18-55
418-53
108-51
108-1*9
0 (Control)
100
200
500
1,000
2,000
80
87
101
103
118
iio
ll*5
125
110
100
85
75
0.57
0.63
0.79
1.03
1.71
3-10
12
12
17
22
36
59
* 200 x 325 mesh anhydrous ferric oxide used as a seed material in
all tests except control. Stirring rate maintained at 300 rpm.
-------
TABLE 24. SYNTHETIC COAL MINE WATER SEEDING EXPERIMENTS -
EFFECT OF STIRRING RATE*
00
VD
Stirring Rate, Sludge Settling
30-Minute
Settled Sludge
Sludge Solids
Content,
Turbidity of
Supernatant
Test No.
418-49
418-57
418-59
418-61
Control
Control
Control
rpm
300
360
420
480
300
360
480
Rate, ml/min
141
123
150
121
80
91
96
Volume, ml
75
70
70
65
145
125
125
Weight Percent
3-10
3-4o
3.46
3-86
0.57
0.46
0.54
Liquid, JTU
59
78
84
90
12
12
13
* 200 x 325 mesh anhydrous ferric oxide (2,000 ppm) used
as a seed material in all tests except control tests.
-------
limits of experimental error. Nevertheless, one result which is be-
lieved to be significant is the marked increase in turbidity of the
seeded samples with increasing stirrer speeds. This is believed to be
related to increasing breakdown of the initially agglomerated seed par-
ticles, as discussed earlier, especially since no similar effect was
noted during the control (unseeded) tests. These findings suggest that
at least in this instance, the most efficient use of the sludge additive
(seeding material) will involve some optimum stirring rate which is
great enough to maintain the material in suspension during sludge pre-
cipitation, yet not so great as to degrade the material by attrition
during mixing. This same statement might also apply in the case of
other sludge additives (e.g. certain of the filter aids tested earlier),
and should be considered in any future tests of this nature.
The results of tests using dried sludge samples as seeding materials are
shown in Table 25. In general, the data fail to show any appreciable
differences between the effects of the ordinary sludge and the freeze-
dewatered sludge at a given concentration level. Furthermore, a com-
parison of these data with the data from tests using anhydrous ferric
oxide (shown in Table 23) reveals that, with few exceptions, the overall
results are quite similar for each seed material at each concentration
level. The main discrepancies appear at the 1,000 ppm level, where the
data indicate that a greater sludge density was obtained using the
ferric oxide. On the other hand, the use of ferric oxide as the seed
also resulted in a higher supernatant liquid turbidity after the
30-minute settling period.
Prom a comparison of the results from the seeded and unseeded (control)
tests shown in Table 25, it is apparent that at least small increases
in sludge density can be effected by adding a portion of previously
obtained sludge to the system before lime neutralization. This observa-
tion may in fact be the most significant finding based on the results
summarized in Table 25, and is quite likely a major factor in the
reasoning which has led to various coal mine drainage treatment schemes
involving sludge recirculation.
Sludge Heating and Freezing
Heating Experiments
The results of the first series of experiments on heating the mine
water before lime addition are summarized in Table 26. The data show
that increasing temperature resulted in reduced sludge volumes and in-
creased sludge solids contents, i.e., an overall increase in sludge
density was effected. On the other hand, the sludge settling rate was
apparently unaffected by temperature, and the trend to higher turbidity
values indicates a tendency toward dispersion or peptization of the
sludge with increasing temperature. These observations, and the fact
that the overall increases in sludge density were relatively small,
suggest that preheating the mine water before lime addition is not a
90.
-------
TABLE 25. RESULTS OF SYNTHETIC COAL MINE WATER SEEDING
EXPERIMENTS WITH DRIED SLUDGE SAMPLES*
VD
H
Concentration
of Seed SIu
Test Wo. Seed Material Material, ppm Rate, ml/min
418-44 None (Control)
4l8-69 Air-dried sludge 100
418-71 Air-dried
sludge after
freeze-dewatering 100
4l8-68 Air-dried sludge 1,000
: Settling
ml/mi n
80
96
106
157
30-Minute
Settled Sludge
Volume, ml
145
120
125
115
Sludge Solids
Content,
Weight Percent
0.57
0.59
0.71
1.36
Turbidity of
Supernatant
Liquid, JTU
12
12
»
16
418-70 Air-dried
sludge after
freeze-dewatering
1,000
110
115
1.34
26
* Minus 325 mesh sieve fraction of seed material used in all seeding tests;
stirring rate was maintained at 300 rpm in all tests including control.
-------
TABLE 26. EFFECT OF HEATING MINE WATER BEFORE LIME ADDITION ON BEHAVIOR AND PROPERTIES
OF SYOTHETIC COAL MINE WATER SLUDGE
K>
Test No. Temperature, C
Sludge Settling
Rate, ml/min*
30-Minute
Settled Sludge
Volume,
Sludge Solids
Content,
Weight Percent
Turbidity of
Supernatant
Liquid, JTU
418-3
108-5
418-7
108-9
22 (room)
4o
60
80
101
126
100
107
130
100
105
75
0.56
0.64
0.60
0.90
9
14
26
24
* Measured in a 1,000-ml graduated cylinder after
sludge had cooled to room temperature.
-------
feasible approach to sludge densification. Furthermore, such an ap-
proach would almost certainly be impractical from the standpoint of
cost.
The results of the second series of tests, involving direct heating of
the concentrated (gravity-settled) sludge, are shown in Table 27.
Again, the results are not encouraging, and the data indicate that no
really significant changes in sludge volumes, solids contents, or set-
tling rates were effected by heating and subsequent cooling of the con-
centrated sludge. (The extremely low settling rates indicated are due
to the fact that, because of preliminary sludge concentration and sepa-
ration, the sludge was already in the compression zone of settling
after being transferred to the 100-ml graduated cylinder.)
Portions of all sludges obtained during the first two series of heating
tests were dried and subjected to x-ray diffraction analysis. It was
anticipated that the sludges might become dehydrated during heating, re-
sulting in the formation of more crystalline iron oxide species. How-
ever, the only iron compound identified was evidently a-FeOOH (synthetic
goethite), and this compound was present in all samples including those
prepared at room temperature. The diffraction pattern lines for the
samples from Tests ^t-18-7 and Ul8-9 (at 60 and 80 C, respectively; see
Table 26) seemed to be slightly stronger and more intense than those for
other samples, indicating that at least some improvement in crystal
ordering was induced at the higher temperatures. In addition, it was
observed that the color of the sludge prepared at 80 C (Test 4l8-9) was
darker than those prepared at the other temperatures.
In the third series of tests, sludge concentrates were heated as in the
second series, but settling rates and sludge properties were measured
before the sludge suspensions were allowed to cool, and measurements
were repeated after the sludge samples had cooled to room temperature
and were resuspended. The results of these tests are presented in
Table 28. The data show slight although measurable increases in solids
contents and more appreciable decreases in settled sludge volumes for
sludges allowed to settle immediately after heating. Under these condi-
tions, therefore, it is evident that sludge density can be increased by
heating alone. On the other hand, when the sludges were resuspended
after first being allowed to cool to room temperature, settled sludge
volumes and solids contents reverted to approximately the same values
as were found for the control test at room temperature (Table 27). Con-
sequently, it is apparent that sludge dehydration (or densification) by
heating is a reversible process, at least within the temperature range
employed in these studies (22 to 80 C) and this behavior would seem to
militate against the feasibility of direct heating as a means of sludge
densification.
It should be noted that the technique of sludge densification by heat
treatment has been applied successfully to the dewatering of sewage
sludges. A process, known as the Porteous Process, was developed in
93-
-------
TABLE 27. EFFECT OF HEATING ON BEHAVIOR AND PROPERTIES OF SYMHETIC
COAL MINE WATER SLUDGE
30-Minute Sludge Solids
Sludge Settling Settled Sludge Content,
Test Mo. Temperature, C Rate, ml/min* Volume) ml* Weight Percent
415-90 22 (room) 0.06 98.0 0.50
415-94 40 0.13 96.0 0.52
415-86 60 0.12 96.5 0.54
415-88 80 0.08 97.5 0.67
* Measured in a 100-ml graduated cylinder after sludge
had cooled to room temperature.
-------
\o
TABLE 28. EFFECT OF HEATING AND SUBSEQUENT COOLING ON BEHAVIOR AND PROPERTIES OF
SYNTHETIC COAL MINE WATER SLUDGE
Resuspended Sludge After Cooling
Heated Sludge to Room Temperature
30-Minute Sludge Solids Settling 30-Minute
Tempera- Settling Rate, Settled Sludge Content, Rate, Settled Sludge
Test No. ture, C ml/min* Volume, ml Weight Percent ml/min* Volume, ml
418-32 40 0.25 91-5 0.51 0.18 94.5
418-30 60 1.60 78.0 0.66 0.27 92.0
1+18-34 80 3.60 66.0 0.73 0.12 96.5
Sludge Solids
Content ,
Weight Percent
0.46
0.45
0.50
* Measured in a 100-ml graduated cylinder.
-------
England and has been marketed in the United States since 1968 by BSP
Corporation, San Francisco, California.(15 through 1?) The process has
been adopted by several major European cities but has found only limited
acceptance so far in this country. The major step in the process in-
volves steam-heating of the sewage sludge at temperatures of 350 to
390 F (177 to 199 C) and pressures of 180 to 210 psi (12.2 to lU.3 atm),
and it will be noted that these conditions are considerably different
from those employed in the present studies. The costs of the Porteous
Process in this country have been reported at between $1.50 and $3-50
per ton of dry solids, and part of the overall cost may be offset by
recovering the calorific value of the filtered sludge. Although filter
cakes from the process reportedly contain as high as 60 percent solids,
it is doubtful whether this rather sophisticated approach could be
adapted economically for dewatering of coal mine drainage sludge. This
conclusion is supported by the fact that work done elsewhere on the
pressure-heat treatment of mine drainage sludges led to discouraging
results.(18)
Freezing Experiments
The technique of freeze-dewatering has been investigated by several
workers for the treatment of sewage sludges.(19 through 26) Moreover,
this approach has been incorporated at several full-scale sewage treat-
ment plants in England and sludge volumes processed are on the order of
^,800 gallons per day at one location.(22,23) The overall costs (capi-
tal and operating) of the sludge freezing process reportedly ranged
between $1.^4-7 and $0.35 per thousand gallons of water treated, depending
to a large extent on the solids content of the thickened sludge sub-
jected to freezing. (23) Sludge freezing costs were minimized by recy-
cling the refrigerant (ammonia) through the sludge thawing tanks; in
this way, heat acquired by the refrigerant during the freezing step was
liberated during the thawing step. Once a particular batch of sludge
had been frozen, the refrigerant flow was simply reversed. The result-
ing sludge could be easily dewatered to 60 to 70 percent solids by
gravity draining.
One limitation of the sludge freezing technique, mentioned by several
investigators, is that the sludge must be frozen at a relatively slow
rate.(23,2^) This is so because water is in effect "squeezed" out of
the gelatinous-sludge floes as formation of the solid state is ap-
proached. Eapid or "flash" freezing of the sludge is undesirable since
physically entrapped water is retained within the sludge floes, result-
ing in little or no dewatering after thawing. It should be noted, how-
ever, that the results of a study reported very recently indicate that
substantially higher sludge freezing rates may be feasible through the
use of a film-freezing technique.(25) Sludge freezing rates of about
U.2 ml/min were successfully achieved, whereas earlier studies had
indicated that practical freezing rates were on the order of O.Uo
ml/min.
-------
The present studies involved two preliminary experiments on the effects
of freezing a synthetic coal mine drainage sludge. In the first experi-
ment, the sludge prepared by lime treatment of a 1,000-ml batch sample
of the WQO-EPA synthetic mine water was allowed to settle for 30 minutes
to a volume of 120 ml. The supernatant liquid was then siphoned off and
the settled sludge was frozen. After it had thawed, the sludge was
transferred to a 100-ml graduated cylinder, and after a second 30-minute
settling period the sludge volume was 3 ml. Thus, a reduction in sludge
volume of 97.5 percent was evidently effected by sludge freezing. The
30-minute settled sludge from the freezing treatment had a solids con-
tent of 9.2 percent. In addition, the sludge was darker in color and
had a decidedly granular texture. X-ray diffraction analysis of the
dried sludge failed to reveal any well-crystallized iron compounds, how-
ever, and gypsum was the only sludge component positively identified by
this technique.
These results were substantiated in a second experiment in which dupli-
cate k-liter samples of the synthetic mine water were treated with lime
and the sludges obtained thereby were allowed to settle for 16 hours.
The bulk of the supernatant liquid was removed from the settled sludges
and one sample was subjected to the freeze-dewatering procedure. The
sludges were then transferred to one-liter Imhoff cones and the volume
of each sample was brought to 1,000 ml by adding portions of the super-
natant liquid obtained during the previous solid-liquid separation step.
Both samples were left undisturbed for a 2^-hour period, and the final
results are illustrated by Figure 17. The sample subjected to freezing
is on the left, and the differences in settled sludge volume are obvi-
ous. Again, the freeze-dewatered sludge was darker in color, granular
in texture, and had a solids content of 11.6 percent. By contrast, the
solids content of the control sludge sample (on the right in Figure 17)
was only 0.90 percent after the 2^-hour settling period.
The preliminary experiments were followed by additional tests using the
actual mine waters listed in Table 1.. The purposes of these tests were
to determine whether comparable results could be achieved with actual
mine water sludges, and to investigate the effect of dissolved aluminum
during the process. With regard to the latter, it has been reported
that the presence of at least 20 ppm of aluminum (presumably present in
the sludge as coprecipitated aluminum hydroxide) is essential for effi-
cient dewatering during (sewage) sludge freezing.(25) Since at least a
few actual mine waters (e.g., Keystone and Brinkerton) contain essen-
tially no measurable aluminum, this apparent constraint regarding alumi-
num content could be significant factor in the conditioning of mine
drainage sludges by freeze-dewatering.
The results of these sludge freezing experiments are summarized in
Table 29. Sludges were prepared in the usual manner (addition of hydra-
ted lime to a final pH of 7-8) and aluminum (as A1-, (S04)3 -iSEfeO),
when added, was dissolved in the 1,000-ml mine water sample before lime
neutralization.
97-
-------
2038P2
Sludge Volume after
Freeze-dew atering
Sludge Volume before
Freeze-dew atering
Figure 17. Effect of Freeze-dewatering on Volume of Synthetic
Coal Mine Drainage Sludge
98.
-------
TABLE 29. RESULTS OF FREEZE-DEWATERING EXPERIMENTS WITH ACTUAL COAL MINE WATERS
Test No.
421- 5C
it SI- 6
421-6A
1+21-7
421-45
421-46
421- 46A
\0
\O
14-21-25C
421-26
421-26A
1+21-27
421-35C
1+21-36
1+21- 36A
Mine Water
Keystone
Keystone
Keystone
Keystone
S . Greensburg
S . Greensburg
S . Greensburg
Brlnkerton
Brinkerton
Brlnkerton
Brinkerton
Tarrs
Tarrs
Tarrs
Ala+ 30-Minute Settled
Added, Sludge Volume
ppm Before Freezing, ml
None
None
20
100
None
None
20
None
None
20
100
None
None
20
70
65
95
225
1+0
1+0
70
45
50
70
210
100
100
130
30-Minute Settled Reduction
Sludge Volume in Sludge
After Freezing, ml Volume, Percent
—
11
11+
20
--
1+
5
--
5
7
12
--
6
7
--
83.1
85.3
91-1
--
90.0
92.9
--
90.0
90.0
91+. 3
--
91+.0
91+. 6
Sludge Solids
Content,
Weight Percent
l.llt
5.17
1+.50
5.10
0.51
3-97
1+.53
0.53
1+.29
1+.57
5-13
0.52
1+.81+
5.11+
Turbidity of
Supernatant
Liquid, JTU
62
1+0*
32*
42*
1+1+
12*
18*
1+5
25*
21*
18*
33
14*
12*
* After settling of freeze-dewatered sludge.
-------
The data in Table 29 show that, as with the tests involving the synthetic
mine water, significant reductions in settled sludge volume (on the order
of 90 percent) were effected by sludge freezing. In addition, decreases
in turbidity were noted as a result of the freeze-dewatering process.
Solids contents of the 30-minute settled sludges after freezing were on
the order of k to 5 percent. Although these solids content values are
somewhat lower than those observed for the synthetic mine water sludges
after freeze-dewatering, it is felt that they still represent substantial
increases compared with the corresponding values for the unfrozen
(control) sludges.
Concerning the effect of added aluminum on sludge properties, there is a
trend toward slightly greater sludge volume reductions and increased
solids contents in the sludge after freezing as the amount of aluminum
increases in the system. However, it should be noted that the addition
of aluminum also has a pronounced effect on the original volume of the
settled sludge before freezing For example, the addition of 20 ppm of
A13+ to each of the four mine waters led to increases in initial sludge
volumes of from 30 to 75 percent. In view of these findings, therefore,
there seems to be no justification for the intentional addition of alu-
minum salts as a means of controlling mine drainage sludge properties
during the freeze-dewatering process.
In summary, the data in Table 29 show surprisingly good agreement among
the results for each of the four mine waters tested. It would appear
that the technique of freeze-dewatering holds considerable promise as a
potential method of coal mine drainage sludge densification. Results of
these tests, as well as those referred to earlier involving sewage
sludge treatment, suggest that freeze-dewatering is a relatively simple
operation which could be applied to any type of sludge resulting from
various coal mine drainage treatment processes. From this standpoint,
therefore, freeze-dewatering might find wider acceptance than other
potential sludge densification processes (e.g., magnetic sludge prepara-
tion) which are conceptually more complex or are limited to certain types
of coal mine discharges. The technology of freeze-dewatering seems to
be rather well developed, and the only major obstacle to this approach
apparently involves the question of economic feasibility In view of
the fact that this is one of the most successful methods of sludge con-
ditioning (and possibly the simplest) explored during this program, it
is felt that further work on this technique would be appropriate.
Miscellaneous Approaches - Coprecipitation of Calcium Carbonate
As described earlier, it was observed that magnetic sludges obtained from
certain of the carbonated mine waters tested, e.g., Keystone and Brinker-
ton, contained calcium carbonate. These sludges generally occupied
smaller volumes after settling and had higher solids contents than did
similar sludges prepared from mine waters essentially free of dissolved
COS, e.g., synthetic mine waters containing comparable iron concentra-
tions. In addition, it is known from work at BCR and elsewhere that
100.
-------
limestone treatment of a given mine water usually results in a denser
sludge than that obtained by hydrated lime treatment. It was reasoned,
therefore, that the intentional addition of C02 to mine water prior to
lime neutralization might have a pronounced effect on the properties of
the resulting sludge.
This was verified during one preliminary experiment where a synthetic
mine water was sparged with COg for 15 minutes before lime was added
(as a slurry) to a final pH of 7.8. The resulting sludge settled to a
volume of 20 ml in 30 minutes and had a solids content of 8.9 percent
by weight. In contrast, a control sludge obtained by the usual lime
neutralization procedure without prior C02 sparging settled to only
180 ml and had a solids content of 0.4 percent.
Because lime is consumed in reacting with COs to form coprecipitated
calcium carbonate,
Ca(OH)2 + C02 = CaC03
it is obvious that the presence of dissolved COg will increase the lime
requirement during treatment. This aspect was investigated in a series
of experiments in which lime was employed (as a slurry) in multiples of
the stoichiometric requirement based on the hot acidity5 of the sny-
thetic mine water (931 ppm) . In all the experiments involving CO., addi-
tion, a 15-minute C02 sparging period was employed previous to lime
addition and the 30"minute reaction period. Control experiments (no
prior C02 sparging period) were also conducted for comparison.
The results of these experiments are summarized in Table 30- The data
show that significant reductions in sludge volume, on the order of 90
percent in all cases, were achieved as a result of calcium carbonate co-
precipitation. (The presence of CaC03 in the sludges from COa -addition
tests was confirmed through x-ray diffraction analysis.) Interestingly,
settled sludge volumes were essentially the same for all the C02 -addition
tests despite the fact that solids contents of the sludges increased reg-
ularly with increasing amounts of lime utilized. Counteracting the
apparent beneficial effects of C03 addition is the fact that turbidities
in these samples were unusually high compared to the control samples.
This tendency toward increased turbidity has also been noted in past BCR
studies on limestone treatment of coal mine drainage. In fact, the
sludges obtained from the C02 -addition tests bore a striking resemblance
to those obtained from limestone neutralization experiments in terms of
color and texture, in addition to settling behavior.
B
The term "hot acidity" refers to the mine water acidity, as ppm cal-
cium carbonate equivalents, determined by the usual titration procedure
involving heating of the sample to remove C03 content. Conversely, the
"cold acidity" is determined with no prior heating of the sample, and
consequently dissolved COS> when present, is also titrated, leading to
higher values.
101.
-------
TABLE 30. EFFECT OF C02 ADDITION ON LIME REQUIREMENT AND SLUDGE PROPERTIES DURING LIME
TREATMENT OF SYMHETIC COAL MINE WATER
Test No.
Control 418-86C
Tests
421- 8C
418-96C
4l8-94c
418-90C
418-92C
COS 418-86
6 Tests
ro 421-8
418-96
418-94
418-90
418-92
Amount Sludge
of Settling Rate,
Ca(OH)s* ml/min
1.00 X 107
1.10 X 118
1.25 x 83
1.50 X 48
2.00 X 64
3.00 X 60
1.00 X *
1.10 X *
1.25 X *
1.50 x *
2.00 X *
3.00 x 203
30- Minute
Settled Sludge
Volume, ml
120
145
160
190
180
330**
15
20
20
15
20
35
Sludge Solids
Content ,
Weight Percent
0.57
0.49
0.64
0.59
0.79
0.75
1.90
1.95
2.42
2.86
4.96
5.62
Turbidity of
Supernatant
Liquid, JTU
11
11
10
9
9
7
74
81
58
71
36
28
Final pH Reduction
After Lime in Sludge
Treatment!- Volume, Percent
5-78
7.82
9.02
11.08
12.00
12.31
6.02
7.70
8.15
8.15
8.09
7-93
--
--
--
--
--
87.5
86.2
87.5
92.1
88.9
89.4
* Multiple of stoichiometric lime requirement based on hot acidity determination.
4- pH of suspension after 30-minute lime neutralization/aeration period.
** The sludge was cream-colored; coprecipitated Mg(OH)s was probably present, accounting for larger sludge volume.
* Settling rate could not be measured due to extreme turbidity of the sample.
-------
As indicated by the data in Table 30, noticeable improvements in sludge
solids contents were effected by the pretreatment of the samples with
COs.
With regard to the lime requirement, the data indicate that possibly as
much as twice the stoichiometric amount (based on the hot acidity)
should be employed in conjunction with COS pretreatment to effect the
desired changes in sludge properties. It should be noted that the use
of the exact stoichiometric amount (l.OO X) was not sufficient in either
case since the terminal pH value was only about 6 after the 30-minute
reaction period; furthermore, significant amounts (15 to 25 ppm) of Fe2+
remained in solution under these conditions at the end of the reaction
period. At all higher lime concentrations employed (1.10 X and above),
residual dissolved iron was reduced to the 1 to 2 ppm level.
The presence of relatively large amounts of dissolved C03 apparently
tends to buffer the system against an overdose of lime, as indicated by
the fact that terminal pH values were all close to 8.0 (except for the
test at 1.00 X lime) during the 003 pretreatment series. By contrast,
it may be noted that for the control tests the terminal pH increased
appreciably as the amount of lime was increased.
In another series of tests, the lime requirement was based on the cold
acidity of the synthetic mine water determined after the COg sparging
period, which was varied in duration from 15 to 60 minutes. The results
of these tests are summarized in Table 31. These results followed the
same general trends as were observed in the earlier tests (Table 30)5
discussed above. The results of cold acidity determinations shown in
Table 31 indicate that only relatively small increases in acidity were
effected by increasing the C03 sparging period. Since changes in the
cold acidity values are presumably an indirect measurement of changes
in dissolved C02, the data suggest that there is little advantage to be
gained by prolonging the C03 sparging period beyond 15 minutes. Thus,
although increased C08 sparging periods may lead to higher solids con-
tents in the sludge as indicated, this gain is offset by increasing
lime requirements and overall treatment times.
The effects of calcium carbonate coprecipitation were further substanti-
ated in tests using an actual mine water from the Keystone site. As
indicated earlier, this mine discharge appears to have an unusually high
dissolved C02 content. Three 1,000-ml samples of the mine water were
treated, each by a difference approach. With the first sample, nitrogen
gas was sparged into the mine water for 15 minutes before lime addition
in an attempt to reduce the original dissolved C08 concentration. Dur-
ing this period, a portion of the ferrous iron was precipitated as evi-
denced by the appearance of a blue-green precipitate, and the pH rose
from an initial value of 6.3 to 7.5- The cold acidity of the sample
was then determined, and 1.00 X lime was added based on this value. The
second sample (control) was treated as received using 1.00 X Ca(OH)2
based on the cold acidity value for the raw mine water. The third sample
103.
-------
TABLE 31. EFFECT OF COg SPARGING PERIOD ON LIME REQUIREMENT AND SLUDGE PROPERTIES
DURING LIME TREATMENT OF SYNTHETIC COAL MINE WATER
Test No.
421-10C-I-
421-10
421-18
421-12
C03 Sparging
Period,
minutes
None
15
30
60
Cold Acidity After
C03 Sparging Period,
ppm CaC03
--
2,026
2,061
2,143
Amount of
Ca(OH)s Used,
grains /liter
0.7270*
1.4999
1.5258
1.5865
Final pH
After Lime
Treatment*
7-75
7.96
8.09
7-80
H
Test No.
421-10C-I-
421-10
421-18
421-12
Sludge
Settling
Rate, ml/min
100
i4o
157
i4o
30-Minute
Settled Sludge
Volume, ml
125
35
30
35
Sludge Solids
Content,
Weight Percent
0.70
3-61
4-37
5.11
Turbidity of
Supernatant
Liquid, JTU
6
33
30
42
Reduction
in Sludge
Volume , Percent**
--
72.0
76.0
72.0
-I- Control test; no pretreatment with C0g .
£ Amount of lime for control test was 1.05 X stoichiometric requirement based
on hot acidity determination; all other amounts were 1.00 X stoichiometric
requirement based on cold acidity determination after indicated COg sparging
period.
* pH of suspension after 30-minute lime neutralization/aeration period.
** Based on settled sludge volume measured during control test.
-------
was sparged with COS for 15 minutes before lime addition in an effort to
determine whether additional dissolved CQg could be introduced into the
mine water. The lime requirement (l.OO X) was based on a cold acidity
determination immediately after the C03 sparging period.
Preliminary tests showed that it was not possible to completely oxidize
the iron in the nitrogen-purged sample during the usual 30-minute reac-
tion period (presumably due to depletion of dissolved oxygen during the
nitrogen purge period). Consequently, the reaction time was increased
to two hours for all tests in this series.
The results of these tests are shown in Table 32. The influence of
content on both the acidity values and sludge properties is apparent
from the data. Although the volume of settled sludge from the nitrogen-
purged sample was identical to that for the control test, one will note
that there is a difference (although small) in the solids contents of
the two samples. In addition, the relatively high turbidity in the
nitrogen-purged sample reveals that an appreciable amount of sludge was
still in suspension, thus in effect decreasing the apparent volume of
the settled sludge.
It is clear from these exploratory studies that the effect of dissolved
COg is significant, and perhaps has not been given sufficient considera-
tion by other workers involved in mine drainage treatment research. Its
importance is particularly noteworthy as pertains to the lime treatment
of carbonated or the so-called "alkaline" mine waters (such as Keystone
and Brinkerton), although these may be comparatively few in number. Al-
though the presence of dissolved C02 as it affects the raw mine water
acidity (and therefore, the lime requirement) is probably generally
understood, its effect on sludge properties may not have received ade-
quate recognition in earlier studies.
Finally, because of the apparent increases in turbidity caused by the
use of COfe pretreatment, a brief series of experiments was conducted to
ascertain whether the use of coagulant aids could overcome this problem.
The anionic polyelectrolyte Genfloc 155 was added at a concentration of
1 ppm to both synthetic mine water control samples and those which had
been subjected to a 15-minute COg sparging period. The polyelectrolyte
was added during a 1-minute stirring period immediately before settling
tests were commenced.
The results of these experiments are shown in Table 33. The data show
that significant improvements in clarity of the supernatant liquid were
obtained through the use of the coagulant aid. Thus, although C02-
treated samples were still slightly more turbid than the control samples
after the 30-minute settling period, they were considerably less turbid
than those prepared in the absence of the coagulant aid (cf. Table 31).
Furthermore, although the addition of the coagulant aid resulted in
substantial increases in the settling rates (and slight decreases in
settled sludge volumes) of the control samples, the relative reductions
105-
-------
TABLE 32. EFFECT OF DISSOLVED C02 CONTENT OF SLUDGE PROPERTIES DURING
LIME TREATMENT OF A CARBONATED MINE WATER (KEYSTONE)
Test No.
421-37
421-38
421-39
Mine Water
Pre treatment
Ng - 15 min
None^-
C02 - 15 min
Cold Acidity After
Pretreatment ,
ppm CaC03
23k
549
2,174
Amount of
Ca(OH)s Used,
grams/liter
0.1732
o.4o64
1.6095
Final pH
After Lime
Treatment*
8.35
8.45
8.36
H
O
OA
Test No.
421-37
421-38
^21- 3 9
Sludge
Settling
Rate , ml/min
110
123
164
30-Minute
Settled Sludge
Volxime, ml
70
70
35
Sludge Solids
Content ,
Weight Percent
0.88
1.37
6.34
Turbidity of
Supernatant
Liquid, JTU
95
43
48
4- Control test.
* pH of suspension after 2-hour lime neutralization/aeration period.
-------
TABLE 33. EFFECT OF COg SPARGING IN CONJUNCTION WITH COAGULANT AID ADDITION ON SLUDGE PROPERTIES
DURING LIME TREATMENT OF SYNTHETIC COAL MINE WATER
o
Amount of Multiple of Final pH
Mine Water Ca(OH)2 Used, Stoichiometric After Lime
Test No. -1-4- Pretreatment grams/liter Amount Treatment**
Test No.
421- 22C
1421-21
421- 24C
1421-24
421- 22C None 0.7270 1.05 X*
421-21 COj, - 15 min 1-5739 1.00 X*
421- 24C None 0.7655 1.10 X*
421-24 CO-, - 15 min 1.7802 1.10 X*
Sludge 30-Minute Sludge Solids Turbidity of
Settling Settled Sludge Content, Supernatant
Rate, ml/min Volume, ml Weight Percent Liquid, JTU
380 120 0.57 1
-I- 40 3.34 7
350 120 0.55 1
4- 30 4.92 6
7-09
8.05
7.81
7-95
Reduction
in Sludge
Volume, Percent
--
66.7
--
75.0
4-4- 1.0 ppm of Genfloc 155 added in all tests prior to settling period.
*•* pH of suspension after 30-minute lime neutralization/aeration period.
t Based on hot acidity determination.
* Based on cold acidity determination after COS sparging period.
4- Sludge floes settled rapidly without a well-defined solid-liquid boundary.
-------
in sludge volumes and increases in solids contents for the C02-treated
samples were of the same order of magnitude as those observed earlier
in the absence of the coagulant aid.
It should be noted that although this approach involving C02 pretreat-
ment may be a novel one as applied to coal mine drainage treatment, it
has been tested and patented in conjunction with sewage treatment.(27,28)
The procedure involved sparging C03 (from incinerator flue gas) into the
activated sludge for 10 to 30 minutes, followed by the addition of
hydrated lime (27) or hydrated lime plus calcium chloride.(28) With
lime alone as the precipitant, the resulting sludge could be vacuum
filtered to about 35 percent solids content. With both processes, the
filtered sludge was calcined to recover some CaO which could be recy-
cled. In this regard, any extension of this approach to coal mine
drainage treatment should involve studies on the use of the residual
sludge alkalinity to offset increased lime requirements as a result of
preliminary C02 addition. Further investigations would be desirable to
determine the filtration characteristics (dewaterability) of the sludge,
its amenability to CaO and C03 recovery by calcination, and the possi-
bility of using the CaC03 constituent of the sludge with or without
additional lime for direct mine water neutralization. Since the CaC03
is formed via a precipitation reaction, it is presumably in a very
finely divided state; because this is a desirable property in the selec-
tion of limestones as neutralizing agents for coal mine water,(29) the
sludge containing coprecipitated calcium carbonate might compete favor-
ably with pulverized limestones in such an application. On the other
hand, it is possible that sludge recycling might negate the original
densification effect obtained through the mechanism of CaC03 coprecipi-
tation. It seems clear that there are several facets of this approach
which warrant further investigation.
108.
-------
SECTION VI
ACKNOWLEDGMENTS
Work on this program was supervised by R. C. Streeter, Project Scientist,
R. K. Young, Principal Investigator, and R. A. Glenn, Project Director.
Technicians involved in the conduct of the experimental work were
J. R. Allender, D. R. Wolber, G. L. Maclay, and A. C. Rupert.
The assistance of Project Scientists C. T. Ford in the electrokinetic
studies, and R. R. Care in the spectrographic analyses, is also acknowl-
edged.
The valued advice and guidance during the course of this program from
R. D. Hill, EPA Project Officer, and Dr. D. R. Maneval, Director of
Planning and Coal Research for the Department of Environmental Resources,
Commonwealth of Pennsylvania, is sincerely appreciated.
109.
-------
SECTION VII
REFERENCES
1. Bituminous Coal Research, Inc., "Studies on limestone treatment of
acid mine drainage," Water Pollution Control Res. Series DAST-33,
1^010 EIZ, 01/70, Water Quality Office, Environmental Protection
Agency, Washington, D. C. (1970).
2. Kakabadse, G. J. and Whinfrey, P., "The effect of certain ions on
the formation of magnetite from aqueous solution," Ind. Chim.
Beige 2g (2), 109-12 (196*1).
3. Riddoch, J., "The hydrolytic and redox properties of ferric iron,
with special reference to the effect of magnesium ions above pH 10,"
Ph.D. Thesis, Victoria University, Manchester, England,
k. Kakabadse, G. J., Riddoch, J., and Bunbury, D. St. P., "The effect
of magnesia on the formation of magnetite from aqueous solution,"
J. Chem. Soc. (A), 1967, 576-9 (1967).
5. Stauffer, T. E. and Love 11, H. L., "The oxygenation of iron(ll) -
relationship to coal mine drainage treatment," Special Research
Report SR-69 to the Coal Research Board, Commonwealth of
Pennsylvania (Nov. 1, 1968).
6. Pavelic, V. and Saxena, U., "Basics of statistical experiment-
design," Chem. Eng. 76 (21), 175-80 (1969)-
7. Dorr-Oliver Inc., "Filtration leaf test procedures," Bulletin
No. 251 LT, 1955-
8. Matusevich, L. N., "Crystallization in presence of seed crystals,"
J. Appl. Chem. U.S.S.R. ^±, 952-7 (I96l).
9. Davies, 0. L. , ed., "The Design and Analysis of Industrial Experi-
ments," New York: Hafner Publishing Company, 1956. Chapter 7-
10. Aarons, R. and Taylor, R. A., "The DuPont waste pickle liquor
process," Proc. 22nd Ind. Waste Conf . , Purdue Univ., 1967, pp. 120-5.
11. Anon., "Extent of coal mine drainage pollution, McMahon Creek
watershed, Ohio," U.S. Dept. Interior, FWQA Ohio Basin Region,
Work Document No. 3^, June 1970.
12. Anon., "Sewickley Creek area, Pennsylvania; Monongahela River mine
drainage remedial project," U.S. Dept. Interior, FWQA, June 1966.
111.
-------
13- Parsons, W. A., "Chemical Treatment of Sewage and Industrial Wastes,"
Washington, D. C.: Bulletin No. 215, National Lime Association,
1965. pp. 64-5-
lh. Brown, T. S. and Long, B. W., "Neutralization and precoat filtration
of concentrated sludge from acid mine water at the Rushton Mining
Company in Osceola Mills, Pa. Project CR-82," Johns -Manville Kept.
No. E412-8087-S1 to Pennsylvania Coal Res. Board (Nov. 3, 1969).
16 pp.
15. Porteous, I. K., "Mechanical treatment of sewage sludge by the
steam- injection method," Municipal Engineering, Dec. 16, 1966. 2 pp.
16. Bulletin No. SllU, BSP Corporation, San Francisco, California.
17. Cardinal, P., Jr., "Porteous process," BSP Corporation, San Francisco,
California, undated. Unpublished. 12 pp.
18. Glover, H. G., Natl. Coal Board, South Yorkshire Area, England, 1970.
Private communication. 2 pp.
19- Clements, G. S., Stephenson, R. J., and Regan, C. J., "Sludge de-
watering by freezing with added chemicals," J. Inst. Sewage Purif. k,
318-37 (1950).
20. Doe, P. W., "The treatment and disposal of wash water sludge," J.
Inst. Water Engrs. 12, ^09-^5 (1958).
21. "The sludge freezing plant at the Stocks reservoir," The Fylde
Water Board, J. Brit. Waterworks Ass. ^3, 1*3^ and ^36 (1961).
22. Doe, P. W., Benn, D., and Bays, L. R., "The disposal of wash water
sludge by freezing," J. Inst. Water Engrs. 19, 251-91 (1965).
23. Benn, D. and Doe, P. W., "The disposal of sludge by the freezing and
thawing process," Filtr. Separ. 6 (4), 383-9 (1969).
2U. Katz, W. and Mason, D. G., "Freezing methods used to condition
activated sludge," Water Sewage Works 117, 110-1^ (1970).
25. Cheng, C., Updegraff, D. M. , and Ross, L. ¥., "Sludge dewatering by
high- rate freezing at small temperature differences," Environ. Sci.
Technol. k (12), 11^5-7 (1970) .
26. Farrell, J. B., Smith, J. E., Jr., Dean, R. B., Grossman, E., Ill,
and Grant, 0. L., "Natural freezing for dewatering of aluminum
hydroxide sludges," J. Am. Waterworks Assoc. 62 ( 12) , 787-91 (1970).
27. Davis, W. S. and Foust, 0. J. (to North American Rockwell Corp.),
"Waste treatment with recycling of flocculating agents," U.S. Pat.
3,^0,166 (Apr. 22, 1969). 6pp.
112.
-------
28. Davis, N. S., Foust, 0. J., and Withers, T. W. (to North American
Rockwell Corp.), "Waste treatment with recycling of lime," U.S. Pat
3,¥K),l65 (Apr- 22, 1969). 8 pp.
29. Ford, C. T., "Selection of limestones as neutralizing agents for
coal mine water," 3rd Symp. Coal Mine Drainage Res. Preprints,
Pittsburgh, Pa., by Coal Ind. Advisory Coram. to Ohio River Valley
Water Sanit. Comm., 1970. pp 27-51.
113-
-------
BIBLIOGRAPHIC:
Bituminous Coal Research, Inc. Studies on Densification
of Coal Mine Drainage Sludqe, Final Report WQO-EPA
Grant No. 14010 EJT,
ACCESSION NO.
ABSTRACT
The purpose of this research was to alleviate present
problems in handling and disposal of sludges obtained by
lime neutralization of coal mine drainage, through the
investigation of various sludge densification techniques.
The scope of the work was restricted to bench-scale
batch experiments. Tests were largely of an exploratory
nature and, as such, did not afford sufficient data to per-
mit detailed cost comparisons among the various tech-
niques.
In the first approach, conditions commonly employed in
the lime treatment procedure (lime neutralization and
aeration) were altered to produce a dense, fast-settling,
ferromagnetic sludge. Although the properties of this
sludge are sufficiently unique, the magnetic sludge con-
version process includes a requirement for sludge heating
and is quite sensitive to the presence of small amounts of
aluminum in the original mine water.
In the second approach, recognized conditioning meth-
ods were applied to sludges obtained by the usual treat-
KEY WORDS: Coal mine drainage, sludge, sludge con-
ditioning, magnetic sludge, iron removal,
sludge densification, sludge disposal,
stream pollution, pollution abatement,
acid mine water treatment.
BIBLIOGRAPHIC:
Bituminous Coal Research, Inc. Studies on Densification
of Coal Mine Drainage Sludge, Final Report WQO-EPA
Grant No. 14010 EJT,
ACCESSION NO.
ABSTRACT
The purpose of this research was to alleviate present
problems in handling and disposal of sludges obtained by
lime neutralization of coal mine drainage, through the
investigation of various sludge densification techniques.
The scope of the work was restricted to bench-scale
batch experiments. Tests were largely of an exploratory
nature and, as such, did not afford sufficient data to per-
mit detailed cost comparisons among the various tech-
niques.
In the first approach, conditions commonly employed in
the lime treatment procedure (lime neutralization and
aeration) were altered to produce a dense, fast-settling,
ferromagnetic sludge. Although the properties of this
sludge are sufficiently unique, the magnetic sludge con-
version process includes a requirement for sludge heating
and is quite sensitive to the presence of small amounts of
aluminum in the original mine water.
In the second approach, recognized conditioning meth-
ods were applied to sludges obtained by the usual treat-
KEY WORDS: Coal mine drainage, sludge, sludge con-
ditioning, magnetic sludge, iron removal,
sludge densification, sludge disposal,
stream pollution, pollution abatement,
acid mine water treatment.
BIBLIOGRAPHIC
Bituminous Coal Research, Inc. Studies on Densification
of Coal Mine Drainage Sludge, Final Report WQO-EPA
Grant No. 14010 EJT,
ABSTRACT
The purpose of this research was to alleviate present
problems in handling and disposal of sludges obtained by
lime neutralization of coal mine drainage, through the
investigation of various sludge densification techniques.
The scope of the work was restricted to bench-scale
batch experiments. Tests were largely of an exploratory
nature and, as such, did not afford sufficient data to per-
mit detailed cost comparisons among the various tech-
niques.
In the first approach, conditions commonly employed in
the lime treatment procedure (lime neutralization and
aeration) were altered to produce a dense, fast-settling,
ferromagnetic sludge. Although the properties of this
sludge are sufficiently unique, the magnetic sludge con-
version process includes a requirement for sludge heating
and is quite sensitive to the presence of small amounts of
aluminum in the original mine water.
In the second approach, recognized conditioning meth-
ods were applied to sludges obtained by the usual treat-
ACCESSION NO.
KEY WORDS: Coal mine drainage, sludge, sludge con-
ditioning, magnetic sludge, iron removal,
sludge densification, sludge disposal,
stream pollution, pollution abatement,
acid mine water treatment.
-------
ment procedure involving lime neutralization and aera-
tion. These methods included the use of coagulant aids,
sludge bulk additives (filter aids), seeding materials, and
sludge heating and freezing.
In addition, exploratory tests were conducted on the
introduction of carbon dioxide into the mine water to
promote coprecipitation of calcium carbonate during
lime addition.
Among the sludge densification methods tested, only
three—magnetic sludge preparation, sludge freezing,
and CO2 pretreatment—appeared to be promising in
terms of results obtained. Each method led to sludge
volume reductions on the order of 90 percent and in-
creases in sludge solids contents of from 0.5 to about 5
percent after 30-minute settling periods. Further de-
velopmental work would be desirable to determine the
general usefulness and comparative costs of each method
as applied to coal mine drainage treatment.
This report was submitted in fulfillment of Program No.
14010 EJT under the joint sponsorship of the Water
Quality Office of the Environmental Protection Agency,
the Commonwealth of Pennsylvania, and the coal in-
dustry through its research agency, Bituminous Coal
Research, Inc.
ment procedure involving lime neutralization and aera-
tion. These methods included the use of coagulant aids,
sludge bulk additives (filter aids), seeding materials, and
sludge heating and freezing.
In addition, exploratory tests were conducted on the
introduction of carbon dioxide into the mine water to
promote coprecipitation of calcium carbonate during
lime addition.
Among the sludge densification methods tested, only
three—magnetic sludge preparation, sludge freezing,
and CO2 pretreatment—appeared to be promising in
terms of results obtained. Each method led to sludge
volume reductions on the order of 90 percent and in-
creases in sludge solids contents of from 0.5 to about 5
percent after 30-minute settling periods. Further de-
velopmental work would be desirable to determine the
general usefulness and comparative costs of each method
as applied to coal mine drainage treatment.
This report was submitted in fulfillment of Program No.
14010 EJT under the joint sponsorship of the Water
Quality Office of the Environmental Protection Agency,
the Commonwealth of Pennsylvania, and the coal in-
dustry through its research agency, Bituminous Coal
Research, Inc.
ment procedure involving lime neutralization and aera-
tion. These methods included the use of coagulant aids,
sludge bulk additives (filter aids), seeding materials, and
sludge heating and freezing.
In addition, exploratory tests were conducted on the
introduction of carbon dioxide into the mine water to
promote coprecipitation of calcium carbonate during
lime addition.
Among the sludge densification methods tested, only
three—magnetic sludge preparation, sludge freezing,
and CO2 pretreatment—appeared to be promising in
terms of results obtained. Each method led to sludge
volume reductions on the order of 90 percent and in-
creases in sludge solids contents of from 0.5 to about 5
percent after 30-minute settling periods. Further de-
velopmental work would be desirable to determine the
general usefulness and comparative costs of each method
as applied to coal mine drainage treatment.
This report was submitted in fulfillment of Program No.
14010 EJT under the joint sponsorship of the Water
Quality Office of the Environmental Protection Agency,
the Commonwealth of Pennsylvania, and the coal in-
dustry through its research agency, Bituminous Coal
Research, Inc.
-------
-4 <.'ces-s; on Number
Subject Field &, Group
05E
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
Environmental Protection Agency
Office of Research and Monitoring, Washington, D. C,
6 Ti
Studies on Densification of Coal Mine Drainage Sludge
10
Authors)
R. C. Streeter
R. K. Young
R. A. Glenn
Bituminous Coal Research, Inc.
16
Project Designation
EPA Grant No. 1*1010 EJT
21
Note
22
Citation
Water Pollution Control Research Series
Publication No. 1*1010 EJT 06/71
Environmental Protection Agency, Office of Research and Monitoring, Washington,DC
Descriptors (Starred First)
Sludge, Sludge Disposal, Acid Mine Water, Water Pollution Treatment,
Pollution Abatement, Mine Drainage*
25
Identifiers (Starred First)
Sludge Conditioning, Magnetic Sludge, Sludge Densification, Iron
Removal
27
Abstract The purpose of this research was to alleviate problems in handling and disposal
Jof sludges from lime neutralization of coal mine drainage, through studies of various
sludge densification techniques. Tests were exploratory in nature and did not afford
sufficient data to permit detailed cost comparisons among the various techniques.
As one approach, conditions commonly employed in the lime treatment procedure (lime
neutralization and aeration) were altered to produce a dense, fast-settling, ferro-
magnetic sludge. The magnetic sludge conversion process includes a requirement for
sludge heating and is quite sensitive to the presence of small amounts of aluminum in the
original mine water.
As a second approach, recognized conditioning methods were applied to coal mine
drainage sludges. These methods included the use of coagulant aids, sludge bulk additives
(filter aids), seeding materials, and sludge heating and freezing. In addition, tests were
conducted on the introduction of carbon dioxide into the mine water to promote coprecipi-
tation of calcium carbonate during lime addition.
Among the sludge densification methods tested, only magnetic sludge preparation,
sludge freezing, and C08 pretreatment appeared to be promising in terms of results
obtained. Each method led to sludge volume reductions on the order of 90 percent and
increases in sludge solids contents of from 0.5 to about 5 percent after 30-minute
settling periods. Further developmental work would be desirable to determine the general
usefulness and comparative costs of each method as applied to coal mine drainage treatment.
Abstractor
R. C. Streeter
Institution
Bituminous Coal Research, Inc.
WR:'02 (RE\. JULY 1969)
WRSI C
SEND TO: WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 20240
* GPO: 1969-359-339
------- |