EPA -660/2-74-018
March 1974
Environmental Protection Technology Series
Storage and
Disposal of Iron Ore
Processing Wastewater
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology*. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and .non-point sources of
pollution. This work provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards.
-------
EPA 660/2-74-018
March 1974
STORAGE AND DISPOSAL
OF
IRON ORE PROCESSING WASTEWATER
by
C. Robert Baillod
George R. Alger
Michigan Technological University
Civil Engineering Department
Houghton, Michigan 49931
Project 14040FVD
Program Element IBB040
Project Officer
Clifford R is ley, Jr.
Office of Research and Development
Region V
Chicago, Illinois 60606
for the
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D. C. 20460
Vat lale by the Superintendent of Documents, U.S. Government Printing office, Washington, D.O. SM03 Price |1.80
-------
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
recommendation for use.
ii
-------
ABSTRACT
The overall objective of this investigation was to advance and improve the
applied technology related to the storage and disposal of wastewater resulting
from the concentration of low grade iron ore. The investigation was based both
upon laboratory studies and upon field studies conducted at the tailings
impoundment and clarification systems associated with two iron ore concentrating
plants located at the Republic and Empire Mines in the Upper Peninsula of Michigan.
Annual water balances were formulated for the tailings impoundment systems
associated with each plant. Particular attention was directed toward estimating
design flows and associated seasonal storages required during the critical spring
melt period. Dye dispersion tests conducted on three clarification basins
indicated considerable stagnant volume and short circuiting. For flocculent
suspensions, settling column analyses were used in conjunction with the dye
dispersion curve to predict effluent solids concentrations with a correlation
coefficient of 0.882. Application of basin volume correction factors ranging
from 2.5 to 10. was suggested when the settling column analysis was applied by
assuming a plug flow hydraulic regime.
A series of jar tests was conducted using the Republic Mine tailings basin
effluent to investigate the cost effectiveness of alum and several polymeric
coagulants. Under most conditions, the cationic polymer coagulant, Calgon M-510
was found to be the most economical, with chemical costs ranging from $2.50/MG
to $8.00/MG depending on initial turbidity and temperature. However, other
coagulants tended to produce less sludge volume. An investigation of several
sludge thickening aids showed the anionic polymer, Calgon M-560, to be the most
effective in reducing the unit area required for thickening. Cationic polymer,
Calgon M-510, was an effective vacuum filtration aid when applied to a sludge
prethickened with Calgon M-560.
A possible system to collect, thicken, filter and dry the 16.9 tons/day of
coagulated solids produced at the Republic Mine was synthesized. The initial
investment was estimated as $870,800, and the cost per ton* of solids was estimated
as $37.65. The synthesis of other alternative systems for dealing with these
solids was hampered by the lack of information on water quality requirements for
reuse within concentrating processes.
Settling column experiments in which the fine tailings were diluted
with various natural waters were conducted to determine the influence of
temperature, dilution, dissolved solids and alum residual. These factors were
found to have greater effects in the case of fine particles (average size of
1.3 microns) resulting from a hematite ore than in the case of larger particles
(average size of 6.5 to 8. microns) resulting from a magnetite ore.
This report was submitted in fulfillment of Project Number 14040 FVD under the
partial sponsorship of the Environmental Protection Agency.
iii
-------
CONTENTS
Section Page
I Conclusions 1
II Recommendations 7
III Introduction 9
IV Basin Design and Management Criteria 13
Hydro logic Aspect 13
Description of Study Areas 13
Water Balances 22
Water Balance Results 24
Performance of Tailings Clarification Basins 28
Quiescent Settling Tests: Performance 29
of an Ideal Plug Flow Basin
Residence Time Distributions: Performance 32
of a Non-Ideal Basin
Comparison of Predicted and Actual Effluent 38
Concentrations
Application to the Design and Management 41
of Tailings Storage and Clarification Basins
Water Reuse 46
V Treatment of Tailings Basin Overflows 49
Investigation of Alternative Coagulants 49
Extended Evaluation of Most Effective 56
Coagulants
Characterization and Analysis of Republic 63
Mine Coagulation - Flocculation System
Handling of Coagulated Solids 66
Feasibility of a Possible Solids Handling System 80
at the Republic Mine
Alternative Strategies for Handling Tailings 85
Basin Overflows
VI Settling Characteristics of Tailings Particles in 87
Natural Water Systems
Quiescent Column Tests 87
Analysis of Factors Affecting Tailings Removal 92
-------
CONTENTS CONTINUED
Section Pa9e
Field Study: Tailings Removal in Republic Mine 96
Tertiary Effluent Stream
Particle Size Distribution "
VII Acknowledgements 103
VIII References 1°5
IX Publications 107
X Appendices 109
v!
-------
FIGURES
Page
1. Location Map 14
2. Republic Mine, Simplified Concentration Process Diagram 16
3. Republic Mine, Tailings Impoundment and Clarification 17
System
4. Empire Mine, Simplified Concentration Process Diagram 19
5. Empire Mine, Tailings Impoundment and Clarification 21
System
6. Results of Typical Quiescent Settling Test, Republic 30
Tailings
7. Typical Performance Curve for an Ideal, Plug Flow Basin 30
8. Dye Dispersion Curve for Republic Pond 3, Northern Portion 33
9. Dye Dispersion Curve, Republic Tertiary Pond 35
10. Dye Dispersion Curve, Republic Tertiary Pond 35
11. Dye Dispersion Curve, Empire Clarification Basin 36
12. Correlation between Actual and Predicted Effluent Suspended 40
Solids Concentration
13. Conceptual Sketch of Hydrograph during Spring Melt Conditions 44
14. Comparison of Turbidity Removal for Several Anionic Polymer 52
Coagulants
15. Turbidity Removal of Various, Equal Cost Combinations of Alum 54
and Anionic Polymer Coagulant Aid
16. Comparison of Turbidity Removals of Several Cationic Polymer 55
Coagulants
17. Turbidity Removal of Various, Equal Cost Combinations of 57
Alum and Cationic Polymer Coagulant Aid
vii
-------
Page
18. Turbidity Removal of Various/ Equal Cost Combinations of 57
Alum and Won Ionic Polymer Coagulant Aid
19. Comparison of Turbidity Removals for Most Effective Ah ionic, 58
Cationic and Won Ionic Polymer Coagulants
20. Turbidity Remaining as a Function of Coagulant Cost 60
21. Calculated Values of G and GT as a Function of Flow Rate 64
for the Rapid Mixing Portion of the Republic Effluent Channel
22. Calculated Values of G and GT as a Function of Flow Rate 65
for the Slow Mixing Portion of the Republic Mine Effluent
Channel
23. Relationship between Jar Test Supernatant Turbidity and Slow 67
Mixing G Value for Republic Pond 3 Effluent
24. Relationship between Jar Test Supernatant Turbidity and Slow 68
Mixing GT Value for Republic Pond 3 Effluent
25. Interface Subsidence Curves for Coagulated Solids Generated 70
at Republic
26. Interface Subsidence Curves for Coagulated Solids Generated 70
at Republic
27. Influence of Anionic Polymer/ Calgon M-560, on the Interface 73
Subsidence Pattern of Coagulated Solids Generated at Republic
28. Influence of Filter Aid Dosage on Specific Resistance of 77
Filter Cake
29. Gravity Dewatering and Drying of Coagulated Solids 79
30. Possible System for Handling Coagulated Solids at the 81
Republic Mine
31. Typical Iso Percent Removal Curves for Republic Fine Tailings 90
32. Typical Iso Percent Removal Curves for Republic Fine Tailings 90
33. Effect of Temperature and Dilution on Removal of Republic Fine 93
Tailings in Lake Michigan and Lake Superior Water
34. Effect of Dilution and Natural Water Source on Removal of 94
Republic Fine Tailings at 4Q°F
viii
-------
Page
35. Effect of Various Diluent Waters on the Removal of 95
Republic Fine Tailings at 40°F
i'
36. Effect of Temperature and Dilution on Removal of Empire 97
Fine Tailings in Lake Michigan and Lake Superior Water
37. Effect of Various Diluent Waters on the Removal of Empire 98
Fine Tailings at 40°F
38. Particle Size Distributions for Fine Tailings 101
-------
TABLES
No. Page,
1. Comparison of Annual Water Budgets for Republic and 25
Empire Tailings Systems Based on Unit Areas of the
Tailings System
2. Relative Importance of Precipitation, Evaporation, 26
Storage and Seepage
3. Comparison of Maximal Tailings Ice and Snow Storage 26
Based on Unit Areas of the Tailings Impoundments
4. Summary of Peak, 10 Day Mean Direct Surface Outflows 31
5. Comparison of Hydraulic Characteristics of Tailings 37
Clarification Basins
6. Summary of Actual and Predicted Effluent Suspended 39
Solids Concentrations for Northern Portion of Republic
Pone 3 at a Return Flow Rate of Approximately 35 cfs
7. Approximate Ranges of Dissolved Solids Concentrations (ppm) 46
Observed at Various Points in Tailings Systems
8. Coagulants Subjected to Initial Screening Tests 50
9. Polymer Coding System 50
10. Comparison of Sludge Volumes Formed after 30 Minutes of 62
Quiescent Settling in Imhoff Cones
11. Solids Concentrations Attained in Batch Thickening Tests 71
12. Estimate of Unit Areas, ft2/lb/sec, No Thickening Aid 71
13. Estimate of Unit Areas, ft2/lb/sec, Calgon M-560 Polymer 74
Employed as a Thickening Aid in the Presence of Slow Stirring
14. Effect of Filter Aids: Summary of Specific Resistance Values 75
of Various Filter Aid Dosages
15. Summary of Leaf Test Results 78
-------
TABLES, CONTINUED
No. Page
16. Summary of Solids Handling Cost Estimate 86
17. Quiescent Settling Column Tests Performed 88
18. Dissolved Solids and Conductivity of the Natural Waters 89
Used in the Settling Analyses
19. Summary of Quiescent Settling Column Data 91
20. Results Obtained from Field Study at Republic Mine 99
xi
-------
SECTION I
CONCLUSIONS
1. For tailings impoundment and clarification systems located in the Lake
Superior iron mining district/ the critical flows for hydraulic and
clarification design occur during the spring snow melt period.
2. Annual water balances on two operating impoundment systems indicated that:
a) The volume of excess flow released during the spring melt period
resulted from snow accumulated both on the impoundment and on
the tributary natural watershed and from tailings ice and possible
ground water storage.
b) Based on unit areas of the impoundment system ,the accumulation of
tailings ice and possible ground water storage ranged from 10.1
inches to 13.7 inches of water equivalent. The higher value
was associated with a system in which overland sheet flow occurred
over a greater area.
c) Based on unit areas of the impoundment system the average seepage
rate varied from 60 inches/year for a system with a total hydraulic
loading of 552 inches/year to 93 inches/year for a system with a
total hydraulic loading of 159 inches/year. At the lower hydraulic
loading/ seepage assumed greater significance in the water balance.
3. In the design of a tailings clarification system/ the design outflow rate is
related to the seasonal storage provided. Both the design flow for
clarification and the associated seasonal storage can be estimated by a
straightforward hydro logic analysis of the system. This analysis involves
the synthesis of an approximate spring hydrograph associated with some
recurrence interval. Data required for this analysis include records of
precipitation/ snowpack/ and runoff applicable to the natural watershed
tributary to the tailings system. In addition allowances must be made for
tailings ice and seepage. At present/ there are few data which can be used
to formulate estimates of tailings ice and seepage.
4. The results of dye dispersion tests on three clarification basins indicated
considerable stagnant volume and short circuiting as the ratio of the time
to the peak dye concentration divided by the theoretical detention time ranged
from 0.071 to 0.29. The degree of short circuiting was qualitatively
related to the plan configuration of the basins.
5. The Republic Mine tailings waste was generated from a hematite ore by a
flotation process. This waste with an initial solids content of about 8%
and no added coagulant.settled as a flocculent suspension and exhibited a
color similar to tomato juice. For this waste,settling column analyses could
-------
be used in conjunction with the basin residence time distribution to predict
the effluent suspended solids concentration with reasonable accuracy. The
correlation coefficient between observed and predicted values was 0.882.
For design purposes, in the absence of information on the expected residence
time distribution, the quiescent settling column analysis could be employed
directly. However, the theoretical basin detention time determined in this
manner (for a given basin depth) must be multiplied by a correction factor.
A correction factor of approximately 2.5 was determined for critical conditions
at the Republic Mine clarification basin. The residence time distributions
of the other basins studied suggested correction factors ranging from 6 to
10.
6. The Empire Mine tailings waste was generated from a magnetite ore by a
magnetic separation process. This waste, with an initial solids content
of 46% and a residual concentration of thickening aid, quickly developed
an interface during settling tests and subsequently settled as a hindered
suspension. The supernate above the interface contained less than 5mg/l
of suspended solids whereas the effluent from the acutal pond, operated
at an overflow rate of approximately 1% of the interface subsidence rate,
ranged from 10 mg/l to 20 mg/l. In this case, the suspended solids
concentration remaining within the supernate could be used to approximate
the concentration in the pond effluent within 15 to 20 mg/l.
7. A series of j'ar test was conducted using the Republic Mine tailings
basin effluent to investigate the cost-effectiveness of alum and several
polymeric coagulants at an initial turbidity of 1400 JTU and temperature
of 13° Centigrade. Alum, certain cationic polymer coagulants (Calgon
M-500, M-510 and M-520) and certain cationic, anionic and non-ionic
coagulant aids (Nalco Na-603, Dow A-22, and Dow N-ll) used in
conjunction with alum were found to be effective in reducing the initial
turbidity by 90% or more. Anionic and non-ionic polymer coagulants were
less effective. Of the coagulants or coagulant combinations which reduced
the initial turbidity by 90% or more, the most effective cationic polymer
coagulant was Calgon M-510 and the most effective coagulant aid was
Dow non-ionic polymer N-ll.
8. A series of jar tests was conducted using the Republic Mine tailings
basin effluent to investigate the cost effectiveness of alum, cationic
polymer coagulant Calgon M-510 and non-ionic polymer coagulant aid
Dow N-ll plus alum at various levels of initial turbidity and termperature
(400 to 1400 JTU and 1°C to 250C). No single coagulant or combina-
tion of coagulants was found to be most economical in achieving a given
supernatant turbidity under all conditions of initial turbidity and temperature.
Chemcial costs to achieve a specified turbidity varied greatly with initial
turbidity and temperature. However, for supernatant turbidities greater
-------
than approximately 60 JTU, Calgon M-510 was the most economical
under all conditions of initial turbidity and temperature. Approximate
chemical costs for Calgon M-510 required to attain a supernatant
turbidity of 60 JTU after 30 minutes settling time ranged from $2.50
per million gallons ($3.00/ton dry solids/ dosage of 0.68 ppm)atan
initial turbidity of 400 JTU and temperature of 25°C to $8.00 per
million gallons ($3.10/ton dry solids/ dosage of 2.18 ppm) at an
initial turbidity of 1400 JTU and temperature of 1°C. At low initial
turbidities (400 JTUHhis cost, to achieve a supernatant turbidity of
60 JTU was virtually equal to the cost of alum plus coagulant aid
Dow IM-11 and was about 80% of the cost of alum alone. However, at
high initial turbidities (1400 JTU)/ the cost of Calgon M-510 to
achieve a supernatant turbidity of 60 JTU was roughly 50% of the cost
of alum plus coagulant aid Dow N-ll and roughly 35% of the cost of alum
alone.
9. Differences were observed between the volumes of sludge generated
during j'ar tests employing various coagulants. At an initial turbidity
of 1400 JTU (700 mg/l suspended solids) the cationic coagulant
aid/ Nalco l\la-603 employed in conjunction with alum (3.1 ppm l\la-603/
50 ppm alum, chemical cost of $18.70/MG/ supernatant turbidity of
47 JTU) generated about 10.5 ml sludge/liter. This amounted to about
45% of the sludge volume generated by the most economical coagulant/
Calgon M-510 (3 ppm/ chemical cost of $11.00/MG/ supernatant
turbidity of 28 JTU) under identical conditions. Volumes of sludge
generated by other coagulants generally fell between the volumes
produced by Calgon M-510 and Nalco Na 603.
10. An analysis of the coagulation - flocculation system at the Republic Mine
suggested that improvements could be made to bring the actual values of
the mean velocity gradient and detention time closer to their optimal values.
Since the hydraulics of the present system are largely set by the natural
terrain/ a significant improvement would probably require the construction
of a mechanical rapid mix - flocculation system.
11. At the Republic Mine/ a large volume (1.44 x 10° cubic feet) of sludge with
an average solids content of 22.6% had accumulated in an impoundment as
a result of chemical coagulation of the highly colored red water leaving the
tailings basin. During the 1 year study period these solids accumulated at
an average rate of 16.9 tons of dry solids per day. An investigation of
the thickening and filtration characteristics of this sludge was conducted
to assess the feasibility of thickening and dewatering this and any future
solids accumulation.
-------
a) An investigation of several thickening aids showed the anionic
polymer Calgon M-560 at a concentration of 0.05% (by weight
of solids) to be the most effective in reducing the unit area re-
quirement. A unit area of 2080 ft2 sec/lb was estimated for a
feed solids concentration of 10% and an underflow concentration
of 20% in the presence of 0.05% Calgon M-560. This amounted
to 23% of the unit area estimated for identical conditions in the
absence of thickening aid. The chemical cost for the thickening
aid was estimated as $1.38 per ton of dry solids. The other
thickening aids investigated were ineffective.
b) An investigation of several filtration aids applied singly showed
that the most effective in reducing specific resistance (24.8 x 10'
sec vg without filter aid) was cationic polymer Calgon M-510
(76% reduction at concentration of 1% by weight of solids). Other
filtration aids were also effective in reducing specific resistance.
To investigate the effect which might result by adding Calgon M-510
to a sludge which had been thickened using Calgon M-560, a
series of filtration experiments in which the two polymers were add-
ed in equal amounts was conducted. A synergistic effect was
found as the reduction in specific resistance by the combination was
greater than the sum of the individual reductions. A reduction of
84% was observed at a total dosage of 0.5%.
c) Both Blichner test and leaf test data were used to estimate a vacuum
filter loading rate of 10.45 Ib/hr. ft2 for dewatering a sludge
thickened to 20% solids with 0.05% Calgon M-560 and further
conditioned with'0.05% Calgon M-510. The filter cake solids
content was estimated to be 49%. The cost of Calgon M-510
employed as a filtration aid amounted to $0.44 per ton of dry solids.
12. A possible system to collect, thicken, filter and dry the coagulated solids
produced from the Republic Mine overflow was synthesized. In doing this,
it was assumed that water quality requirements for reuse prohibited both
the addition of coagulants within the reuse system and increasing the
average percentage of process water recycled. It was estimated that this
system would require an initial investment of $870,800. The average
cost per ton of dry solids (including interest, depreciation, labor, thicken-
ing and filtration chemicals but excluding coagulation chemicals) was
estimated as $37.65. It is questionable whether this system or any
other system of comparable cost could be justified unless some economic
value could be attached to the dried solids. In the absence of any
economic value, the dry or semi-dry solids would probably have to be
disposed of by burial. In this case drying would not be necessary
and th,e solids processing cost would decrease to about $26.00 per
ton exclusive of final disposal.
-------
13. The synthesis of other alternative systems for dealing with the fine
particulates contained in tailings basin overflows was hampered by:
a) The lack of information on water quality requirements for
reuse within ore concentrating processes.
b) The lack of information which could be employed to predict
any gradual build-up of important water quality parameters
which may occur within a reuse system.
14. Settling column experiments in which the fine tailings particles
contained in the Republic Mine basin overflow were diluted by factors
of 5 and 10 with various natural waters showed that:
a) An increase in temperature caused the effective settling velocities
to increase. In certain instances/ the concomitant decrease in
viscosity accounted for less than half of the measured increase in
settling velocity.
b) Dilution generally tended to increase the effective settling
velocities. As the dilution factor increased from 5 to 10/
effective settling velocities increased by 0% to 50%.
c) The effect of dilution was generally greater for natural waters
higher in dissolved solids. The dilution effect was generally
the greatest for a water containing a residual alum concentra-
tion. The relation between increase in effective settling
velocity and diluent water dissolved solids content did not
appear to be linear.
d) The addition of 2 mg/l of alum generally increased the effective
settling velocity by about 0 to 10%.
Similar experiments conducted using Empire fine tailings particles showed
less effect for changes in temperature/ dilution and dissolved solids content
of the diluent water. In fact/ for Lake Superior water increasing the
dilution from a factor of 5 to a factor of 10 caused solids removals to
decrease by 2% to 10%. No consistent effect of diluent water dissolved
solids content on effective settling velocity was observed. Likewise/
no consistent effect of the addition of 2 ppm alum was observed.
15. One of the factors contributing to the difference in behavior between the
Republic and Empire fine particles is believed to be the difference in
particle size. Particle size distributions were determined for both the
Republic and Empire fine tailings fractions using two methods; settling
data and Stokes Law and direct measurement from photomicrographs. For
the Republic tailings/ the mean particle sizes (50% smaller by weight)
determined by each method were nearly identical and equal to 1.3 microns.
For the Empire particles/ the optical method yielded a mean particle size
-------
of 6.5 microns whereas the settling data yielded 8.0 microns.
Generally, the optical method tended to estimate fewer numbers of
very small and very large particles.
16. A field study was conducted to compare suspended solids removal
predicted from column data with that actually occurring over 3,800
feet of the Republic Mine Tertiary Effluent Stream. The predicted
removal was 20% whereas the actual removal was 24%.
6
-------
SECTION II
RECOMMENDATIONS
1. Design of tailings storage and clarification systems should include
a thorough hydrologic analysis of the proposed system. Particular
attention should be given to the system water balance, storage
capacity and clarification performance during the critical melt period.
2. More data should be gathered concerning the hydrology of other tailings
impoundment systems so that expected quantities of water released
during melt periods could be estimated with more confidence. Likewise,
more data should be gathered concerning the hydraulic and sedimentation
efficiencies of other tailings clarification basins.
3. More field data should be gathered concerning the transport of fine
particles in natural water systems, particularly in thermally stratified
impoundments.
4. A comprehensive investigation should be conducted to develop water
quality criteria for reuse in ore concentrating processes. These criteria
should be determined experimentally by pilot studies of various types of
flotation and magnetic separation processes. It is generally recognized
that the concentration of dissolved solids, particularly Caf^ , Mg"*"1",
and other multi-valent cations may be significant.
5. An applicable methodology should be developed to predict any gradual
build-up of important water quality parameters which may occur within
a reuse system.
6. Potential uses of the dried solids resulting from the chemical coagulation
operation at the Republic Mine should be investigated. Any economic
value attached to this material would help to justify the possible solids
handling system discussed in this report. In assessing the feasibility of
any system for handling the coagulated solids, due consideration should
be given to the results of studies similar to those recommended under
Items 4 and 5 above. It may prove to be more economical to eliminate
the overflow than to treat it and handle the resulting solids.
7. If a solids handling system similar to the one discussed in Section VI
of this report is judged to be the optimal solution, a continuous flow
pilot scale demonstration study should be conducted to develop design
data for a full scale operation.
-------
SECTION III
INTRODUCTION
Over the past 25 years/ the iron mining industry has witnessed a trend toward
the exploitation of low grade ore formations. Formerly, nearly all of the ore
shipped consisted of richer, natural ores which required little or no beneficia-
tion or concentration at the mine site. As the high grade ore deposits became
scarcer, technological improvements in beneficiation processes made the
exploitation of low grade ore deposits economically attractive. The term
taconite, although descriptive of a hard, dense Minnesota rock with 25% to 35%
iron content, has come to be frequently used to refer to low grade iron formations
in general. In 1945, only 23% of the ore shipped resulted from beneficiation
processes. However, in 1965, 77% of the 87 million tons of useable ore had
been beneficiated. The Lake Superior district accounted for three-fourths of
this production. It is anticipated that this trend will continue in the future.
In the concentrating process the low grade ore (257<> to 35% Fe) is concentrated
to 55% to 70% Fe. This results in a waste material containing 10% to 20% Fe.
Thus, for every four tons of ore at 35% iron, 1.8 tons of concentrate at 65%
iron would be produced. The 2.2 remaining tons at 10% iron would be discharged
to the tailings deposit. The concentrated ore is palletized in a heat hardening
operation with the aid of a suitable binding material such as bentonite. The
fact that this pelletized iron ore is an extremely desirable blast furnace feed
is one of the factors responsible for the increased exploitation of low grade ore
deposits.
A variety of unit operations such as gravity separation, screening, cycloning,
magnetic separation and flotation are employed in the concentrating process.
Nearly all of these operations are wet processes in which the ore materials are
suspended in water. Naturally, considerable quantities of water are used in
these processes. To produce one ton of concentrate from four tons of crude
ore can require from 600 to 6,000 gallons of water depending on the process.
Upon leaving the process, this water serves to transport the remaining waste
materials to the tailings basins.
A system for handling tailings wastewater serves chiefly to separate the suspended
load from the liquid so that the water can be either discharged to a natural water-
shed or reused directly. Since the tailings wastewater contains on the order of
70,000 to 500,000 ppm suspended solids (98% of which settles very rapidly),
large volumes of solids are deposited. As pointed out above, the volume of this
deposit is roughly equivalent to 50% to 75% of the volume of ore mined. In any
tailings disposal system, therefore, a certain amount of land would be required
for the permanent disposal of these solids. This has led to the development of
-------
huge tailings deltas in the proximity of concentrating plants. In some cases
(the Reserve Mining Co. in Minnesota and some inactive Michigan copper mining
operations), these deltas extend into Lake Superior. In beneficiation operations
located further inland, these deposits may be better contained. Nevertheless,
a substantial amount of supra-colloidal solids (500 ppm to 1,500 ppmjmay
remain in suspension and settle very slowly. Since most concentrating operations
are located in areas of relatively low land values, the most economical solution
has been to create large impoundments and to rely upon gravity sedimentation to
accomplish separation of these solids,
One advantage of this impoundment system for secondary treatment is that the
partially clarified wastewater may be reused in the plant operation. This reuse
minimizes the quantity of pond effluent discharged into the natural water course
and tends to make the plant water balance approximate a closed system. Naturally,
this minimizes the cost of any treatment required for the pond effluent. However,
in a particular situation, water reuse may be limited by the build-up of dissolved
substances which may interfere with the concentrating process.
Some inherent problems associated with impoundment systems are caused by the
areal extent of the ponds and their associated watersheds. Taken as a whole,
these areas can be viewed as artificial watersheds which respond to meteorologic
changes. The Lake Superior iron mining district receives a heavy snowfall; and,
during the spring thaw, large quantities of water are released from the impoundments
and their associated watersheds. Thus, even if water reuse is practiced, unless
the impoundment has an adequate storage capacity, pond water may be discharged
to the receiving stream at a high rate. Moreover, at high discharge rates accompanied
by low water temperatures, the clarification capacity of the impoundment is
impeded and the suspended solids content of the pond effluent tends to increase.
Thus, in certain cases, further treatment may be required for the pond discharge.
The overall objective of this investigation was to advance and improve the
applied technology related to the storage and disposal of wastewater resulting from
the concentration of low grade iron ore, and, specifically:
A. To develop basin design and management criteria which recognize the
influence of tailings settleability as well as the effects of meteorologic
and hydro logic phenomena on the pond effluent quality.
B. To investigate alternative tertiary treatment methods for eliminating
the fine particulate materials from the basin discharges with a view
toward developing a tertiary treatment system for subsequent demon-
stration .
C. To investigate the physical and transport characteristics of the supra-
colloidal particles contained in the tailings basin discharges with a view
toward predicting the rate at which these particles might be removed from
a natural water system.
10
-------
These objectives were attained through a comprehensive study of the tailings
impoundment systems associated with the iron ore concentrating plants asso-
ciated with the Republic and Empire Mines/ located in the Upper Peninsula of
Michigan and operated by the Cleveland Cliffs iron Company. In addition,
complementary laboratory studies were carried out at the Sanitary Engineering
Laboratory of Michigan Technological University. Subsequent sections of
this report treat each aspect of this comprehensive study in detail.
At the outset of the study, it was also intended to study the impoundment
system associated with the Humboldt Mine. However, midway through the
study period, mining operations at the Humboldt site ceased and the concentrating
plant was shut down for modifications. Therefore, the intended study of the
Humboldt impoundment system was abandoned and is not included in this report.
11
-------
SECTION IV
BASIN DESIGN AND MANAGEMENT CRITERIA
Hydrologic Aspect
As pointed out in the introduction, tailings basins and their surrounding areas
can be viewed as vast artificial watersheds. An understanding of the hydrology
of these watersheds is fundamental to the development of sound basin design
and management criteria. To develop this understanding, water balances were
formulated for the tailings basins associated with the Empire and Republic mines.
In this way, the relative importance of precipitation (both rain and snow), snow-
melt, evaporation, seepage and surface outflow could be judged.
Description of Study Areas
Climate and Meteorology
The two main study sites, the Republic Mine and the Empire Mine,are separated
by approximately 15 miles, as shown in Figure 1. The two main river basins,
the Michigamme near Republic and the Escanaba near Empire, drain into Lake
Michigan. For this area, the mean monthly temperature ranges from 15°F in
January to 66°F in July with an average annual temperature of about 41°F.
The average date*of the first fall temperature of 32° or colder is September 21
and the last day in the Spring is May 28. Average annual precipitation is about
31 inches with snowfall averaging about 110 inches. Snowfall during the
1970-71 study period totaled 135.5 inches with a maximum snow depth on the
ground of 36 inches. Monthly evaporation potentials and a class A pan coefficient
of 80% have been determined for this area by Wiitala et j*|. (1).
Two important factors affecting the magnitude and duration of spring runoff to be
expected from tailings basins and their associated watersheds are the thaw duration
and water equivalent of the snowpack at the beginning of thawing. A frequency
analysis of the duration of the spring thaw (time elapsed between the onset of melt-
ing conditions and the disappearance of the snowpack) for this region based on 20
years of record (1951-1970) indicated a mean thaw duration of 22 days. The thaw
duration was equal to or less than 39 days in 90% of the years and equal to or less
than 12 days in 10% of the years. A similar analysis of the water equivalent of
the snowpack at the onset of thawing conditions indicated a mean value of 8.5
inches. The snowpack water equivalent was equal to or greater than 6 inches for
90% of the years and equal to or greater than 11 inches for 10% of the years. Snow-
pack water equivalents did not appear to be correlated with thaw duration. The
length of melt period and the water equivalent of the snow just prior to the melt for
the 1970-71 study period were 15 days and 9.5 inches respectively. A water
equivalent greater than this would be expected about once in three years. Likewise,
a thaw duration shorter than this would be expected about once in 5 years,
13
-------
CANADA
LAKE SUPERIOR
HOUGHTON
UPPER
MICHIGAN
0 ""EMPIRE MINE
REPUBLIC MINE
LAKE
MICHIGAN
WISCONSIN
LOWER
£ MICHIGAN
Figure I. Location Map
-------
Tailings Handling Operations at the Republic Mine
The Republic Mine is operated by the Cleveland Cliffs Iron Co (CCI) for
CCI Jones & Laugh I in Steel, Wheeling Steel, and International Harvester
Companies and is located on the Marquette Iron Range approximately 30
miles southwest of Marquette, Michigan as shown in Figure 1. The Republic
ore body consists of hematite (major economic material), magnetite, martite,
quartz, jasper, iron silicates, and minor secondary carbonates. All of the
constituents appear in the tailings deposits. The concentration plant processes
roughly 20,700 long tqns (2240 Ib./long ton) per day of low grade hematite
at 35;.5% iron to produce roughly 9,700 long tons per day of concentrated ore
at 65.5% iron. The remaining 11,000 long tons per day at roughly 10%
total iron are discharged to the tailings basin. In the concentration process, the
ore is first ground to a fine state (80% - 325 mesh). Argillaceous slime
materials are then removed by wet eye Ion ing. Subsequently, the concentrated
ore is floated while the flotation underflows are discharged to the tailings stream.
Thickening and vacuum filtration are employed to dewater the concentrated ore.
This primary concentrate is then further concentrated in a regrind flotation process.
The concentrate from the regrind operation is then dewatered and palletized.
A simplified diagram of the concentration process and waste sources is shown in
Figure 2. It can be seen from this figure that the waste arises from six principal
sources. The total flow to the tailings basin amounts to roughly 59 cfs of which
58 cfs is water. Because of the supra-colloidal hematite particles in suspension,
the tailings stream and ponds exhibit a red color similar to that of tomato juice.
The tailings slurry is conveyed in an open channel to the impoundment system.
The Republic impoundment system covers an area of about 938 acres which is
divided into 3 ponds and a return flow system. The average basin condition is
about 30% water area and 70% tailings (land) area. The entire basin is enclosed
by dikes and, in general, has an interior elevation higher than the surrounding
land area. There is, therefore, no surrounding watershed. Figure 3 shows a
plan view of the impoundment and clarification system. Pond 1 consists of a
consolidated tailings deposit, and, including the Northeast Area, covers an area
of 593 acres. The tailings stream spreads out and meanders over this deposit in
sheet flows. During this process, the bulk of the solids settle out and become part
of the consolidated deposit. The total annual tailings deposit, spread uniformly
-over the 593 acres of Pond 1, would cause an increase in the surface elevation
of the deposit of approximately 3.5 to 4.0 feet. Spot checks of the tailings
elevation at the beginning and end of the 378 day study period indicated that
this was the case. Thus, it can be concluded that, over the study period, nearly
all of the tailings deposit was stored in the Pond 1 area.
Pond 2 consists of an intermediate settling area of 78 acres. During the study
period, the flow entered Pond 2 through a rock-fill dike. The water surface
elevation is controlled by a weir structure ahead of the culverts leading to Pond 3.
15
-------
INPUT
CRUDE ORE
20,700 LJ/DAY
WATER
FATTY ACID
CONDITIONER
WATER
WATER
BINDING
MATERIAL
CRUSHING
GRINDING
WASTE
(MISC. S.Scfs)
1
HYDROSCILLATORS
1
CYCLONES
1
>IOcfs
/v|4,000ppm-
CONDITIONING
i
FLOTATION
PROCESS
-*24.5cfs
/~l 50,000 ppm-
THICKENER
\.
VACUUM
FILTRATION
y^/*/l,OOOppin
PRIMARY
CONCENTRATE
REGRIND PROCESS
4-
SECONDARY
CONCENTRATE
PELLETIZING
OPERATION
PELLETS
0.5cfs
^5,0 00 ppm
TO TAILINGS
59 cfs,8.l«& SOLIDS
AVERAGE OF36.4cfs
RECYCLED TO PLANT
Figure 2. Republic Mine, Simplified Concentration Process Diagram
16
-------
V**MAIN FLOW FROM MILL
"""'">
A
PUMPHOUSES
ETURN FLOW
POND
RETURN FLOW
COUNTY
ROAD
FFW
POND 2
3B INLET
TERTIARY POND
1000 2000 FEET
OPEN WATER
Figures. Republic Mine,Tailings Impoundment and Clarification System
17
-------
The dikes surrounding Pond 3 enclose an area of 190 acres. Sixty-nine acres of
this surface is occupied by consolidated tailings and 90 acres by water. The
remaining area of 31 acres consists of peninsular projections of higher, natural
forested terrain. The tailings stream enters Pond 3 at the point labelled 3B Inlet
and tends to flow in a channel along the southern boundary before flowing north-
east toward the open water area where secondary clarification takes place. The
performance of this clarification area is analyzed in this report.
The flow leaves Pond 3 at two points labelled Return Flow and 5B Outlet. The
flows at each of these points varied considerably and were measured during the
study period. The return flow is pumped to a reuse reservoir before being reused
in the concentrating process. The flow rate at the 5B outlet is controlled by
adding or removing weir planks at the outlet structure. During times of high flow/
alum and polyelectrolyte are added to the flow leaving the 5B outlet. Coagulation
and flocculation takes place in the channel between the 5B outlet and the tertiary
pond. Upon entering the tertiary pond, nearly all of the remaining solids settle
out. The tertiary pond effluent is discharged to the receiving stream, Gambles*
Creek,which has an average upstream flow of 4 cfs. The coagulation and floccula-
tion of this overflow is analyzed in this report.
A considerable amount of unconsolidated muck accumulated in the tertiary pond
as a result of the settling of the coagulated solids. The thickening, dewatering
and drying characteristics of these solids are investigated in this report.
s
Tailings Handling Operations at the Empire Mine
The Empire Mine is operated by the Cleveland Cliffs Iron Co. (CCI) for CCI, Inland
Steel, McLouth Steel, and Internation Harvester Companies and is located on the
Marquette Iron Range approximately 17 miles southwest of Marquette, Michigan.
The Empire orebody consists of magnetite (major economic material), iron carbon-
ates, iron silicates, martite, earthy hematite, chert and quartz. All of these
constituents appear in the tailings. The concentration plant processes roughly
28,500 long tons per day of crude ore at 33% total iron to produce roughly 9,700
long tons per day of concentrate at 66.5% iron. The remaining 18,800 long tons
per day at 16.5% total iron is discharged to the tailings basin.
Figure 4 shows a simplified diagram of the concentration process. In this process,
the ore is ground to a fine state (90% - 500 mesh) by means of an autogenous
grinding process. A wet magnetic cobbing process serves as the first stage of
concentration. The cobber concentrate is classified in cyclones. The cyclone
underflows are fed into the pebble mill while the overflows are subjected to further
stages of concentration.
The siphonizer serves to dewater and deslime the cyclone overflows thereby
upgrading the iron content. The final concentration steps consist of further
magnetic concentration and an amine flotation. The process stream is then
18
-------
INPUT
CRUDE ORE
28,500 LT/DAY
WATER
WATER
CONDITIONER
WATER
BINDING .
MATERIAL
WASTE
ROCK MILL
PEBBLE MILL
MAGNETIC
COBBER
CLASSIFYING
CYCLONE
cfs ,
1
SIPHONIZER
£
-20 4 cfs, 1.0
^
MAGNETIC
FINISHER
1
f
27.5 cfs,2.i
5.0 cfs, 8.0*fo -»
THICKENER
-*8.5cfs,0.03fr-»
VACUUM
FILTRATION
TOT.= 308cfs,2.
CONCENTRATE
1
PELLETIZING
OPERATION
1
PELLETS
TAILINGS
THICKENER
REUSE
TAILINGS
12 cfs, 464fe
9-3 cfs
Figure 4. Empire Mine, Simplified Concentration Process Diagram
19
-------
thickened and dewatered by vacuum filtration before being pelletized. A poly-
electrolyte (American Cyanamid Co. Superfloc 16, 0.087 mg/l) is added ahead
of the thickener to improve the settling characteristics of the suspension.
Figure 4 shows that the tailings originate from five principal sources within^the
concentration process. The total tailings stream of 308 cfs at about 2.68%
solids is passed into three tailings thickeners where about 96% of the water
is reclaimed for reuse. The thickener underflow, at about 46% solids/ is then
pumped a distance of about 5 miles to the tailings disposal area.
The Empire tailings basin consisted of an average water surface area of 406 acres,
a tailings area of 326 acres and a surrounding tributary watershed of 1,315 acres.
Figure 5 shows a plan view of this area. During the study period, the flow, for
the most part, was introduced along the northern boundary of the tailings deposit.
Occassionally, the tailings stream was discharged at points along the southern
boundary. The total flow to the tailings impoundment area averaged about 12.8
cfs of which 9.9 cfs was water. This flow was carried to the tailings area by a
system of parallel pipelines in which a velocity of 7 feet per second was maintained.
The tailings stream meandered over the consolidated tailings deposit in sheet flows.
However, this overland sheet flow occurred over a distance of roughly only 2,000
ft./ whereas, at Republic, the overland sheet flow occurred over a distance of
roughly 6,000 ft. Thus, at Empire, more of the tailings particles remained in
suspension when the tailings reached the open water area. As a result, most of the
Empire tailings particles were deposited into the open water area and actually caused
water to be displaced. Field studies conducted near the beginning and end of the
study period showed that the tailings front advanced about 2,000 feet over the
course of a year. Figure 5 shows the position of the tailings front at a point mid-
way during the study period. The survey also showed that about 80% of the total
tailings deposit was deposited into the open water area. The water which was dis-
placed by these particles was considered as a negative storage quantity in computing
the water balance.
The visual appearance of the water area at the Empire site was not markedly different
from a natural lake. In contrast to the Republic basins, the Empire basin contained
a lower suspended solids concentration (usually less than 25 mg/l) and did not
exhibit the red color of the Republic basins.
Direct outflow from the pond area occurred at two points as indicated in Figure 5.
During the early portion of the study period, the East outlet carried the major portion
of the flow. However, during the fall of 1970, the weir level at the East outlet was
raised and the West outlet began to carry nearly all of the flow. Flow at each out-
let was measured during the course of the study.
20
-------
WATERSHED BOUNDARY
TAILINGS FRONT
SECONDARY
PONDS
TAILINGS FROM
PLANT^
TAILINGS PONDS
UNDER CONSTRUCTION
: OPEN WATER
SCALE*, i" = 2,000'
Figures. Empire Mine, Tailings Impoundment and Clarification System
-------
Upon leaving the main tailings basin, the flow passed through two secondary
settling ponds where additional clarification took place. During the period of
this study, construction was underway on a larger system of tailings ponds and
dikes to be located adjacent to the south and east boundaries of the existing
tailings area.
Water Balances
Water balances were formulated for the Empire and Republic sites over the period
of July 1970 to June 1971 using the equation
| + P-E-L=Q+AS 1.
Where:
I = Plant Inflow
P = Precipitation
E = Evaporation
L = Seepage
Q = Surface Outflow
AS = Total Change in Storage
Plant Inflow and Precipitation
Monthly average plant inflows were taken from plant operation records., Precipita-
tion was measured at both sites utilizing standard 8 inch non-recording rain gages
during the entire study period as well as recording type rain gages during the
summer months. During periods when the recording type gages were inoperative,
long period catches in the non-recording gages were broken down into their
corresponding daily values by utilizing the cIimatological records of the U.S.
Environmental Data Service at Marquette, Michigan.
Evaporation
Average monthly evaporation was estimated by utilizing mean monthly temperature
as recorded at the Marquette station in conjunction with monthly evaporation
potentials reported by Wiitala et al. (1). The full evaporation potential was
applied to open water areas. For the land areas, however, the evaporation potential
was compared with the monthly precipitation and the smaller of the two values was
applied.
Surface Outflow
Surface outflow was measured at both sites. At Republic, the surface outflow
consisted of the return flow, which was determined from the operating records of
calibrated pumps, and the 5-B outflow which was continuously measured by
22
-------
recording the water elevation abovea weir. At Empire the surface outflow
was continuously measured by recording the water elevation ahead of hydraulic
control sections.
Storage
The total change in storage can be expressed as
AS= ASp +ASS + ASj + AS + ASt 2.
Where:
AS = Total Change in Storage
ASp = Change in Pond Water Storage
ASS = Change in Snowpack Storage
ASj = Change in Tailings Ice Storage
ASg = Change in Groundwater Storage
= Change in Tailings Storage
The change in pond water storage, AS p/ was determined by multiplying the recorded
changes in pond water surface elevations by the pond areas. This quantity also
reflected the volume of ice and snow floating on the pond surface. In cases where
solid tailings were deposited into open water a negative storage correction was
applied to account for the pond water displaced by the tailings .
Tailings storage, St, 'refers to the interstitial water stored within a saturated
tailings deposit as it builds up. This was determined by multiplying the change
in volume of saturated tailings by the average porosity.
Snowpack storage, Ss/ was determined from periodic snow surveys conducted at
the sites during the winter months. Snow depths and water equivalents were
measured on both the tailings deposits and surrounding watersheds. The samples
taken on the tailings deposits did not include any tailings ice which existed under
the snow. Tailings ice refers to ice formed on the tailings deposit as a result
of the meandering overland sheet flows . Winter accumulations of tailings ice
storage, Sj, were estimated by closing the water balance.
i
Groundwater storage refers to subsurface water stored in the natural watershed
tributary to the tailings system. No field measurements of changes in ground water
storage were made. Since the study commenced during the summer and lasted 1
year it was assumed that the ground water storage at the beginning of the study
period was equal to that at the end of the study period. Likewise, no snow or
ice were present at the beginning and ending dates . Thus over the entire study
period:
ASS = 0 3.
23
-------
AS: = 0 4,
ASg = 0 5.
and AS = ASn + ASt 6-
Seepage
Since there was no practical way to directly measure seepage, total seepage
during the study period was estimated by closing the water balance; Combining
Equations 1 and 6 gives:
) - ( ASp+ ASt+Q) 7.
which was used to estimate total seepage during the study period.
At Empire, no drastic changes in basin configuration took place during the course
of the study. Thus, it was assumed that seepage occurred at a constant rate
throughout the year. At Republic, however, flow was diverted into the northeast
portion of Pond 1 at the beginning of April 1971. This visibly increased the
opportunity for seepage along the northeast boundary. Therefore, it was assumed
that seepage occurred at a constant rate over the period of June 15, 1970 to
March 31, 1971 and at another, constant, but greater rate, over the period of
April 1, 1971 to June 28, 1971. The first rate was estimated by applying
Equation 7 over the period from June 15 to October 31, 1970. In developing
Equation 7, it was assumed that ASg = 0. Fluctuations of ground water stage
in this area indicate that, over this period, ASg would be negative causing this
seepage rate estimate to be somewhat low (2). The second rate was estimated '
by applying Equation 7 on the period of April 1 to June 28, 1971. In this case,
the assumption that ASg = 0 would cause the seepage rate estimated to be some-
what high.
Water Balance Results
Tables Al and A2 (Appendix A) show the water balance data for Empire and
Republic respectively. The basic volume measure employed here is the second-
foot-day (sfd) which is equal to the volume accumulated if 1 cfs were to flow for
1 day, or 8.64 x 104 cubic feet. These data are summarized by the hydrographs
and mass curves shown in Figure A.I to A.3 (Appendix A). Combining Equations
1 and 2 gives:
ASs+ASj +ASg =(I+P-E-L)-(Q+ASp + ASt) ? 8.
Thus the quantity of water stored as snow, tailings ice, and groundwater is shown
by the difference in ordinates between the Inflow + Precipitation - Evaporation
- Seepage line and the Outflow + Storage line on the mass curve. This quantity
is important because it represents an additional volume of water released during
the spring thaw.
-------
Change in Storage,ASp+ASt, inches +18
Table 1. Comparison of Annual Water Budgets for Republic and Empire
Tailings Systems Based on Unit Areas of the Tailings System
Republic Empire
Area of Tailings System sq. mi.
Area of Surrounding Watershed sq. mi.
Plant Inflow/ I, inches
Precipitation, P, inches
Evaporation, E, inches
Net Input, I+P-E, inches
Direct Surface Outflow, Q, inches1
Seepage, L, inches
Net Outflow + Change in
Storage, Q+L+ASp+ASt,
inches 552 159
Table 1 shows a comparison of the Republic and Empire water budgets based on
unit areas of the impoundment system. The precipitation and evaporation figures
include the volumes contributed by the tributary watershed. However, all the
figures are based on unit arias of the impoundment system. The quantities, given
in inches, represent the height of a volume of water spread over the area of the
tailings basin. Thus, the annual plant inflow to the Republic basin amounted
to a volume of water 542 inches high spread over 1.46 sq. miles; whereas the
annual Empire plant inflow amounted to a column of water only 119 inches high
spread over 1.14 sq. miles. These large differences in annual loading between
the two systems are partially a result of the fact that water is reused from the
Republic basins.
Because the direct hydraulic loading of the Empire basin is so much less,
precipitation, evaporation and seepage play a much larger role in its overall
water balance. Moreover, the importance of precipitation is accentuated by
the 2.06 sq. mi. of tributary watershed at Empire. Table 2 shows the relative
25
-------
Table 2. Relative Importance of Precipitation, Evaporation
Storage and Seepage
Percent of
Annual Inflow
Annual Change in
of Inflow
Percent of
Annual Outflow
Plant Inflow
Precip. less Evap.
Storage as Percent
Direct Surface Outflow
Seepage
Republic
98.1%
1.9%
3.3%
87.7%
11.3%
Empire
75%
-4.7%
44.4%
55.6%
Table 3. Comparison of Maximal Tailings Ice and Snow Storage
Based on Unit Areas of the Tailings Impoundments
Republic Empire
Unaccounted Storage, ASj + ASS +
ASg inches, water equiv. 15.5 25.8
Snow Accumulation, ASS inches,
water equiv. 1.8 15.7
Accumulation of Tailings Ice and
Ground Water Storage, ASj + AS
inches, water equiv. 13.7 10.1
importance of precipitation and seepage at each site. Precipitation less evaporation
accounted for only 1.9% of the total input to the Republic basin, but amounted to
257° of the total input to the Empire basin. Likewise, seepage accounted for only
11.37o of the outflow at Republic, but made up 55.6% of the outflow at Empire.
On a unit area basis, the seepage rates compared much better as the 93 in./yr. of
seepage at Empire was only about 1.5 times the 60 in./yr. at Republic. It would
be expected that the seepage be greater at Empire since the natural terrain under-
lying the tailings deposit dips to the southeast at roughly 20 to 30 feet per 1,000
feet. Thus, the dike along the southeast boundary of the impoundment reaches a
26
-------
height of about 90 feet. The natural terrain underlying the Republic basin is
generally flatter, and except for one small section along the west boundary/
the dikes are generally less than 30 feet in height.
Table 3 shows the maximal accumulations of tailings ice and snow. These
quantities are given in inches and are based on the areas of the tailings impound-
ments (938 acres at Republic and 726 acres at Empire) exclusive of any
tributary natural watershed. The unaccounted storage represents the additional
amount of water (above and beyond the average flow) which was released during
the spring melt. The snow accumulation represents the portion which existed as
snow on the impoundments and/or on the surrounding natural watershed. Sub-
tracting the snow accumulation from the total unaccounted storage gives the
accumulation of tailings ice and ground water storage.
4'.-.
At Empire, the 25.8 inches of unaccounted storage/ or additional water released
during the spring melt, was significanly greater than the 15.5 inches at Republic.
This was caused by the heavy snow accumulations on the 1,315 acres of surround-
ing watershed. Republic had no tributary watershed surrounding the diked area
and accumulated only 1.8 inches of water equivalent as snow on the impoundment.
In the absence of a surrounding natural watershed/ ground water storage may be
neglected. Thus/ the remaining 13.7 inches represents the accumulation of
tailings ice.
A study of Tables A.I and A.2 and the hydrographs/ mass curves and temperature
data given in Figures A.I and A.4 (Appendix A) shows that, at Republic, the spring
melt begain in early March and was essentially completed uy late April/ only a
few days after the snow was gone. This is what would be expected in the absence
of ground water movement into the impoundment system.
During March/ the average daily air temperatures seldom exceeded 32°F and the
10 day mean temperatures were consistently below 30°F. Therefore/ it appears
that the melting conditions may have been caused by a combination of condensation
and the adsorption of solar radiation by the dark colored deposits of tailings.
A study of the same figures reveals that the Empire site behaved quite
differently during the melt period. The surface outflow rate increased markedly
at the beginning of April when the 10 day mean air temperature exceeded 32°F.
Moreover/ the high outflow rate continued through June even though the snow was
essentially gone by April 20. This is indicative of a substantial amount of sub-
surface water movement into the impoundment system.
The interpretation of the hydrographs shown in Figures A.I and A.2 (Appendix
A) is complicated by the fact that changes in storage were also occurring. It is
estimated that/ in the absence of storage/ the 10 day mean surface outflow at
Republic would have reached 84 cfs during early April 1971. However/ a
considerable amount of water (252 sfd) was stored at Republic between April 4
27
-------
and April 10 as a result of the diversion of the tailings stream into the
Northeast Pond 1 area on April 4. This is a good example of how prudent
management can mitigate the effect of peak flows. Likewise, at Empire,
it is estimated that in the absence of storage, the 10 day mean flow would
have reached 50 cfs.
Table 4 summarizes the peak 10 day mean outflows at the two sites. A
comparison of the ratios of peak 10 day mean flows to average flows shows
that the effect of the high spring flows is relatively much greater at Empire
than at Republic. The incremental flow rates (computed peak 10 day mean
flow in the absence of storage minus the average annual flow) were more
comparable. Expressed on a unit area of impoundment basis, the incremental
flow rates amounted to 22.6 and 38.5 cfs/mi2 for Republic and Empire
respectively. The higher value for Empire reflects the contribution of the
tributary watershed.
Performance of Tailings Clarification Basins
The development of sound basin design and management criteria requires an
understanding of both basin hydrology and clarification performance. It is
essential to estimate both quantities of water to be handled and effluent quality
attained by these basins. Information presented in the preceeding paragraphs
can be used with engineering judgement as a guide to estimate water quantities.
In general, the quality of a water is determined by temperature and by the
concentration of various dissolved and suspended materials present. The
major parameter affecting the quality of tailings basin effluents is suspended
solids concentration. In certain cases, temperature and the concentrations of
dissolved organic and inorganic substances may appreciably affect water quality.
For the purposes of this report, tailings pond effluent quality is judged by
suspended solids concentration. Thus, the objective of this research aspect was
to develop a method, based on laboratory settling data, by which the suspended
solids concentration of a tailings basin effluent could be predicted and to evaluate
this method by comparing predicted and actual concentrations obtained from operating
tailings clarification basins.
A brief description of the research plan employed to achieve this objective is as
follows:
Republic Mine
The Republic tailings suspension settled as a classical flocculent suspension.
The settling characteristics of the waste were quantitatively described through the
use of quiescent settling column tests (3). From the data developed in these tests,
it was possible to predict the effluent concentration for an ideal plug flow basin.
28
-------
To better predict the performance of the actual basins/ residence time distributions
were determined by using dye tracer techniques. By combining the residence time
distribution data with the settling column data, it was possible to predict effluent
concentrations for the actual basins. In order to compare predicted and actual
concentrations/ suspended solids concentrations were routinely measured at
various points within the Republic tailings system during the course of the study.
Periods of approximately steady state operation were selected from these field data.
Knowing the flow rates/ temperature/ and mean residence times during these periods
of pseudosteady-.stateoperation/ it was possible to predict the basin effluent
suspended solids concentration from quiescent settling column data gathered under
similar conditions.
Empire Mine
Unlike the Republic tailings/ the Empire tailings with an initial solids content of
46% did not exhibit flocculent settling. Almost immediately after the start of each
column test/ an interface was formed and hindered settling took place. Because
flocculent settling did not occur/ the settling curve analysis employed for the
Republic tailings could not be applied here. The supernate above the interface
contained less than 5 mg/l of suspended solids whereas the effluent from the actual
pond operating at an overflow rate of approximately 1% of the interface subsidence
rate ranged from 10 to 20 mg/l. In this case, the suspended solids concentration
remaining within the supernate could be used to formulate a rough estimate of that
in the pond effluent. It is likely that the concentration increment in excess of that
obtained in the column test was caused by wind action and scour at high flows.
Quiesfieotc'fottling Tests; Performance of an Ideal Plug Flow Basin
Methodology and experimental conditions employed in performing the quiescent
settling tests are described in Appendix B. A total of 17 individual column
tests employing the Republic tailings were conducted at various temperatures
(40°F to 75°F)/ initial suspended solids concentrations (265 mg/l to 12/600
mg/l) dissolved solids content (114 mg/l to 150 mg/l) and coagulant dosages.
The dissolved solids content of the tailings stream varied from day to day and was
carefully noted since settleability 'had been observed to b& a function of dissolved
solids (4). A wide range of conditions was investigated so that comparisons could
be made between the column data and the field data.
The results of the settling tests can be conveniently represented by noting the
measured concentrations on a plot depth versus time. Figure 6 shows the results
of a typical settling test. The curves represent iso-concentration lines located by
interpolation between the measured concentrations. For discrete settling/ the iso-
concentration lines would be linear. Curvilinear iso-concentration lines indicate
the flocculent nature of a suspension (3).
29
-------
10
20
30
ui
U
Z 40
Q.
UJ
Q
50
60
70
iniO
II48Q
50
150
100
TIME IN HOURS
Figures. Results of Typical Quiescent Settling Test,Republic
Tailings.
200
x 1000
o»
2
o
i
o
. 800
600
a- 400
S
u.
UJ
200
0 20 40 60 80 100 120 140 160 180
DETENTION TIME IN HOURS
Figure?. Typical Performance Curve for an Ideal,Plug Flow Basin
-------
Table 4. Summary of Peak, 10 Day Mean Direct Surface Outflows
Republic Empire
Average Annual Flow/cfs 51 6.15
Actual Peak - 10 Day Mean 60 20.2
Computed Peak - 10 Day Mean
Assuming No Storage, cfs 84 50
Ratio of Actual Peak - 10 DM. 1>18 3<28
Average Annual
Ratio of Computed Peak-10 P.M. 1>65 8|15
Average Annual
Incremental Flow/ cfs
(Computed Peak 10 D.M. - Ave.) 33 43.9
Incremental Flow Rate cfs 22 6 38 5
Unit Area of Impoundment System sq.mi.
The iso-concentration lines constructed from a quiescent settling column test can
be used to predict the overall removal to be expected in a continuous flow basin
under ideal plug flow conditions. Ideal plug flow conditions exist when every
part of the fluid entering the basin is evenly distributed over the entire vertical
cross section of the entrance zone/ and the flow advances at a uniform and constant
horizontal velocity to the outlet zone. Such an hypothetical continuous flow settling
basin will be termed an "ideal basin". The procedure employed in predicting effluent
concentrations as a function of detention time for an ideal basin of a particular depth
is developed and explained in standard reference works (3). Figure 7 was developed
from the data given in Figure 6 and shows the effluent concentration for an ideal
basin 57 inches deep (average depth of Republic Pond 3) as a function of detention
time.
31
-------
Residence Time Distributions: Performance of a Non-Ideal Basin
Application of the foregoing analysis of an ideal basin to an actual situation
requires consideration of the factors which affect settling in an actual basin.
The efficiency of the sedimentation operation is largely a function of the
hydraulic characteristics of the basin in which the operation takes place. The
term hydraulic characteristics as used in this report refers to the flow pattern
produced in the basin. This flow pattern or regime is related to the physical
features of the basin such as length, width, depth, inlet and outlet conditions.
The prevailing flow pattern in a basin determines the hydraulic efficiency of that
particular basin. The term hydraulic efficiency as used here refers to the
residence time distribution of the flu id in the basin. A basin's flow regime.
seldom consists of only one type of flow, but is usually a combination of several
types ranging from idealized plug flow at one extreme to a completely mixed
pattern on the other depending on the hydraulic characteristics of the basin.
The flow regime and residence time distribution of an actual basin can be deter-
mined both qualitatively and quantitatively by use of tracing methods. A tracer
is introduced into the influent to the basin and its concentration at the outlet is
determined as a function of time. From the results of a tracer test employing a
one shot injection of tracer, a curve of effluent tracer concentration versus time
may be plotted. The shape of the dispersion curve is a qualitative measure of
the flow pattern through the basin. The curve itself can be interpreted as a
residence time distribution of the fluid in the basin. If the dispersion curves are
plotted in dimensionless terms, they can be used to compare the hydraulic
characteristics of basins with different shapes or of the same basin with different
flow rates.
In order to determine the residence time distributions, tracer studies were per-
formed on both Pond 3 and the Tertiary Pond at Republic and on the main tailings
pond at Empire. In each case, a slug of fluorescent tracer was injected into the
influent and the concentration in the effluent stream was measured as a function
of time. The methodology employed in these studies is described in Appendix C.
The dispersion curve for the northern portion of Republic Pond 3 is shown in
Figure 8. During the time of the dispersion test, the average return flow rate
was determined to be 36 cfs and the flow rate at 5B to be 2.2 cfs. In view of
the pond configuration shown in Figure 3, it is reasonable to consider the
northern 75 percent of the pond area (above the dashed line in Figure 3) as
carrying the return flow, and the southern portion as carrying the 5B overflow. With
this assumption the theoretical detention times for the northern portion and the south-
ern portion were determined to be 117 and 545 hours, respectively. Since the
majority of the flow was in the direction of the return flow outlet the dispersion
curve for the northern portion, measured at the return flow outlet was of primary
interest. The irregular shape of the declining portion of the curve shown in
Figure 8 was very likely caused by a portion of the dye being trapped in a small
pond at the tailings front and later being released.
32
-------
Dispersion Curve, Republic Pond 3, Northern Portion
Flow = 36cfs (at Return Flow Outlet)
Surface Area « 70 acres
Volume = 337 acre ft.
Theoretical Det. Time - 117 hours
Dye* Rhodamine WT, 25lb. of 20% Solution
80 120 160 200
Time in Hours
Figure 8. Dye Dispersion Curve for Republic Pond 3, Northern Portion.
240
-------
Dispersion curves for the Republic Tertiary Pond are shown in Figures 9 and
10. In view of the theoretical detention times indicated on the figures it is
evident that considerable short circuiting took place. Moreover, a comparison
shows a lack of geometric similarity between the two curves measured at
different flow rates. This indicates an unstable flow pattern which hampered
the prediction of effluent suspended solids content.
Figure 11 shows the dispersion curve observed for the Empire clarification
basin. The irregularities in the curve here were considerably greater than in
the proceeding curves. This was probably caused by wind action over the large
water surface and by the relative large "dead" area in the northwest portion (see
Figure 5). Considerable short circuiting is indicated as the peak, tracer con-
centration occurred at a time equal to 7% of the theoretical detention time.
Various investigations(5)(6)(7) have proposed dimensionless parameters for
judging the hydraulic characteristics of sedimentation basins. Three of these
parameters have been employed here to compare the hydraulic characteristics of
tailings clarification basins.
Initial Time Ratio = time to the initial appearance of tracer
theoretical detention time
(Measures severe short circuiting, equals 1.0 for
plug flow and zero for ideal mixing.)
Modal Time Ratio = time to peak tracer concentration
theoretical detention time
(Measures dead or stagnant regions, equals 1.0 for
plug flow and zero for ideal mixing.)
Dispersion Index = time for 90% of tracer to pass
timeror 10% of tracer to pass
(Measures longitudinal mixing, equals about one for plug
flow and 23.6 for ideal mixing.)
Optimal hydraulic characteristics for sedimentation are indicated when these
parameters approach their plug flow values.
Table 5 compares the values of these parameters calculated from the dispersion
curves shown in Figures 8 to 11. It can be seen that both the initial and modal
time ratios are more indicative of ideal mixing than plug flow. Moreover, the low
values of these parameters for the Republic Tertiary Pond and Empire Pond in- >
dicate a considerable amount of short circuiting and stagnant regions. The values
for Republic Pond 3 are indicative of significantly better hydraulic characteristics
than the others. In view of the plan configuration shown in Figures 3 and 5 ,
34
-------
Dispersion Curve, Republic Tertiary Pond
Flow = 28 cf s
Surface Area = 2.0.1 acres
Volume = 91 acre feet
Theoretical Det. Time = 34.5 hr.
Dye: Rhodamine B, 700 g
16
20
34 8 12
Time in Hours
Figures. Dye Dispersion Curve,Republic Tertiary Pond.
24
.Q
a.
a.
c
0)
o
O
a>
>»
O
Dispersion Curve, Republic Tertiary Pond
Row = 22 cfs
Surface Area= 20.1 acres
Volume - 91 acre feet
Theoretical Det. Time = 50 hr.
Dye: Rhodamine B, 300g
0 10 20 30 40 50
Time in Hours
Figure 10. Dye Dispersion Curve,Republic Tertiary Pond.
35
-------
Dispersion Curve, Empire Tailings Pond
Flow = 5.3 cfs
Surface Area = 320 acres
Volume = 500 acre feet
Theoretical Det.Time =1130 hr.
Dye: Rhodamine B, 700g.
o*
Figure II.
100 200
Time in Hours
Dispersion Curve, Empire Clarification Basin.
300
-------
Table 5 . Comparison of Hydraulic Characteristics of Tailings
Clarification Basins
Initial Modal
Flow Time Time Dispersion
Republic Rate/cfs Ratio Ratio Index
Pond 3 36 O.U 0.29 5.9
(Northern Portion)
Tertiary Pond 28 0.056 0.11 4.1
Tertiary Pond 22 0.072 0.14 6.7
Empire Pond 5.3 0.033 0.071 5.2
Theoretical
Plug Flow 1 1 1
Ideal Mixing 0 0 23.6
Representative Values
for Rectangular Basins (5) 0.2-0,7 3-8
these observations seem reasonable. The calculated values for the dispersion
index compare well with representative values for rectangular basins (5). Thus/
it seems that the relatively poor hydraulic characteristics observed are caused
primarily by short circuiting and stagnant regions.
It is obvious that improvements in plan configuration and inlet positions could
produce hydraulic characteristics more closely approximating plug flow. How-
ever, in general/ plan configuration and Inlet position are largely dictated by
topographical features. It follows, then/ that under conditions producing poor
hydraulic characteristics/ only a fraction of the basin volume should be considered
as effective.
37
-------
Comparison of Predicted and Actual Effluent Concentrations
The method by which the settling column and dye dispersion data were combined
to predict the effluent concentration for a non-ideal basin is explained by
Baillod and Christenson (26) and outlined in Appendix D.
As stated earlier, data on the performance of the Republic Mine clarification
basins were gathered over a one year period. These data included suspended
solids concentrations,dissolved solids concentrations and flows at various points
within the system and are summerized in Figures A.2, A.4 and A.5 of Appendix
A and Table 7 on page 46. These data were used to compare the actual per-
formance of the clarification basins with that predicted from the settling column
analyses. The field performance was compared both with that predicted for an
ideal basin and that predicted from the residence time distribution for a real flow
basin. This was done to estimate the correction or scale-up factors involved in
using settling column data to design large clarification basins.
Ideally, the comparisons between predicted and actual performance should be
made for periods of steady-state operation during which the flow was equal to
that existing at the time of the dispersion test. In addition, the other field
conditions (e.g. influent suspended solids, dissolved solids, temperature)
should correspond to those employed for the column settling test. However,
with the exception of the periods during which the dispersion tests were con-
ducted, true steady state conditions seldom existed for periods longer than a
few hours. Thus, periods of approximately steady conditions were used for
comparison. The column test conditions were chosen to reflect the conditions
existing during these periods. Linear interpolation was employed to predict
performance for field temperatures falling between two column tests.
Comparisons were made for the northern portion of Republic Pond 3 for 7 periods
during which the flow rate was approximately equal to that existing during the
dispersion test. Table 6 summarizes the actual and predicted effluent concen-
trations. The concentrations predicted by considering the residence time
distribution agreed reasonably well with the actual concentrations. The con-
centrations predicted far an ideal plug flow basin with a volume equal to that
of the actual basin were considerable less than the actual concentrations. The
correction factors listed in column 6 represent factors by which the volume of
the actual basin would have to be multiplied to produce an effluent concentration
equivalent to that of the ideal basin. The reciprocal of this factor could be
loosely interpreted as an effective fraction of the actual basin volume. In
designing a basin based upon column settling data, the detention time required
for an ideal basin to attain the specified concentration should be multiplied by
this correction factor.
Figure 12 shows a correlation between the observed and predicted effluent
concentrations. It is evident that the residence time distribution analysis was
fairly successful in predicting the actual concentration. However, since few
data related to residence time distributions in large impoundments are presently
available, it would be difficult to apply this analysis in design. Although the
effluent concentrations predicted for the ideal basin were considerably less than
38
-------
w
Table 6. Summary of Actual and Predicted Effluent Suspended Solids Concentrations for Northern Portion of
Republic Pond 3 At a Return Flow Rate of Approximately 35 cfs
Period
(1)
Aug. 3-16
Sept. 21-28
Nov. 3-8
Feb. 23-27
Mar. 19-23
May 20-26
Sept. 1-14*
Water
Temp.
Op
(2)
72
55
43
32
32
52
65
Actual Effl.
Susp. Solids
mg/L
(3)
280
320
500
720
760
520
245
Predicted Effl. Susp. Solids mg/L
Considering
Residence Time Ideal Basin
Distribution
(4)
329
430
550
741
565
455
252
(5)
146
191
244
328
315
202
157
Corr. Factor
For Ideal
Basin
(6)
1.92
1.68
2.05
2.20
2.41
2.57
1.56
*Period of Dispersion Test
-------
6
.£
§
O
O
o
C0
o
m^t
"5
CO
o.
CO
3
CO
*-
c
-------
those attained in the actual basin, a reasonable correlation existed. The
correction factors for the ideal basin analysis ranged from 1.56 to 2.57 with
the higher values occurring during the initial melt conditions. Thus, for
basin performance under critical melt conditions, the larger factors would be
appropriate. These factors would apply only to basins similar to Pond 3 at
Republic. For different basin geometries, these factors would be expected to
change.
The modal time ratios shown in Table 5 are indicative of the dead or stagnant
regions within a basin. The ratio of 0.29 for Republic Pond 3 is significantly
greater than the ratios for the other ponds, indicating that the effective fraction
of the Pond 3 volume was greater than the effective volume fractions of the
other ponds. If, as a first approximation, the ideal basin correction factors are
assumed to be inversely proportional to the modal time ratios, rough estimates
of ideal basin correction factors applicable to flocculent settling in the Republic
Tertiary Pond and Empire Pond can be formulated. Taking a correction factor
of 2.5 as corresponding to a modal time ratio of, 0.29 yields correction factors
of 6 and 10 for the Tertiary Pond and Empire Pond respectively.
In the case of the Republic Tertiary Pond a comparison of Figures 9 and 10 had
shown the flow pattern to be unstable. Thus, for the Tertiary Pond, performance
comparisons were made only for the periods during which the dispersion tests were
conducted. At a flow rate of 22 cfs the actual effluent concentration was 18
mg/l. Under similar conditions, the settling column data predicted an effluent
concentration of 3 mg/l for an ideal basin and 20 nig/1 considering the residence
time distribution. However, at a flow rate of 28 cfs, the actual concentration
was 50 mg/l while the predicted concentration was 4 mg/l for both the ideal and
non-ideal flow basins. The unstable flow pattern hampered the predictions in
this case.
Application to the Design and Management of Tailings Storage and Clarification
Basins
Topographical Considerations
*
In the design of a tailings storage and clarification system, the general configura-
tion of the basins is influenced by the natural topography of the area. From an
economic viewpoint, the most desirable site would be one which requires a minimal
amount of dike construction, is situated in an area of low land value, and minimizes
the operating expense.
Taken together, these three site attributes will minimize the cost per ton of
tailings stored. According to these criteria, a natural lake is an extremely
desirable site since dike construction can be avoided and, in certain instances, the
lake bottom, upon which the tailings are stored, may be "free" government land.
Thus, in the Lake Superior region, natural lakes have often been employed for
tailings storage.
-------
A second type of desirable site from a topographical viewpoint would consist
of natural valley with steep walls. With this topography, the tailings deposit
could be retained by a single dike across the lower end of the valley. In
practice/ however, because of land costs and transportation expense, the
most economical site may have topographical features which necessitate more
dike construction than the sites mentioned above. The topography of the Republic
tailings site, for example, is characterized by several small hills or knolls
surrounded by relatively flat land. In this situation, the amount of dike
cpnstruction can be minimized by taking advantage of the higher areas of
natural terrain. Thus, to a certain extent, the plan configuration of the impound-
ment and clarification system has been influenced by the site topography. The
Empire tailings site is located in a valley forming the headwaters of Green Creek.
The principal dike, therefore, has been constructed across the lower end of the
valley and forms the southeast boundary of the tailings impoundment.
It is evident that, in any given location, the plan and profile configuration of a
tailings impoundment and clarification system will, to a great extent, be governed
by the site topography.
Hydrologic Considerations
From the proceeding discussion it can be seen that the impoundment and clarifica-
tion sites most desirable as far as minimizing dike construction is concerned generally
result in a certain area of natural watershed being tributary to the impoundment and
clarification system. This tributary watershed, in itself, does not affect the
volume available for tailings storage. However, the runoff from this tributary water-
shed will affect the volume required for clarification. Table 3 indicated that,
at the Empire site, the excess volume (above and beyond the average) released
during the spring amounted to 25.8 in. spread over 726 acres, or 1,560 acre
feet. Nearly 50% of this resulted from snow accumulated on the 1,315 acres of
tributary watershed. The excess flow at the Republic system, with no tributary
watershed amounted to only 15.5 inches spread over 938 acres, or 1,210 acre feet.
Thus, as far as the design and operation of the clarification system is concerned,
it is preferable to avoid large areas of tributary watershed.
t
The manner in which the tailings deposit is formed can also influence the excess
flows released during the spring melt and in turn affect the volume required for
clarification. The overland sheet flow at Republic caused a build-up of tailings
ice and ground storage amounting to 13.7 inches over 938 acres, or 1,070 acre
feet. At Empire the overland sheet flow covered a smaller area and most of the
solids were deposited into open water. This resulted in a maximal tailings ice and
ground storage accumulation of 10.1 inches over 726 acres/or 610 acre
feet. Thus, with all other considerations being equal, it is preferable to avoid
the occurrence of overland sheet flow during the cold months. However, it should
be pointed out that the overland sheet flow is effective in increasing the elevation
42
-------
of the tailings deposit. Therefore/ in many cases this mode of operation
may be necessary.
In the rational design of a tailings clarificatbn system, it is necessary to
determine a design flow for clarification. (i .e. the maximum effluent flow
rate which the clarification system can pass while still maintaining a satis-
factory effluent quality.) This will normally be less than the flow rate employed
for hydraulic design. If provision for storage is to be made in the clarification
basin, the maximal flow rates into the basin will be greater than the maximal'
effluent flow rates. In this case, it is logical to base the clarification design
on the basin effluent flow rate. The design storage will then be determined by
both the influent and effluent flow rates. In addition, an averaging time must
be associated with the clarification design flow rate. In waste treatment
practice, the averaging times associated with process design flows are generally
approximated by the detention time of the unit to be designed. For tailings
clarification basins, this averaging period could vary from 1 to 15 days
depending on the size of the basin and the flow rate.
It is evident from the hydrographs given in Figures A.I and A.2, (Appendix A)
that, in the Lake Superior Region, the critical flow conditions will occur during
the spring melt. The designer is faced with constructing an approximate hydrograph
for the expected inflow to the tailings clarification basin during the critical melt
period. If storage is to be provided, the instantaneous peak inflow rates are not
overly important. Thus, this approximate hydrograph might be based upon 10 day
mean inflow rates. The hydrograph would be constructed by superimposing the
excess flow released during the melt on the base flow (plant inflow less uniform
storage less seepage) as illustrated in Figure 13. The total area above the
base flow represents the volume of excess water released during the melt period
and would consist of snow, tailings, ice, precipitation during the melt period
and possible ground water storage. The volumes of snow (as water equivalent)
and precipitation to be expected for any given recurrence interval can be easily
estimated by a frequency analysis. The volumes of tailings ice and ground water
storage are more difficult to estimate. The values listed in Table 3 might be
judiciously employed as guides in estimating tailings ice and ground water for
systems closely related to the Republic and Empire systems.
The exact shape of the inflow hydrograph during the melt period is not critical.
Of more importance is the base width, which is difficult to accurately estimate.
In the case of a basin dominated by a tributary watershed it may be possible to
gather flow data before construction of the tailings basin begins. Alternatively,
values similar to those observed in this study might be cautiously applied to
closely similar systems.
43
-------
LU
o
_l
Q.
DESIGN STORAGE
DESIGN
OUTFLOW
' -ESTIMATED
INFLOW'
4
|«-BASE WIDTH-
PLANT INFLOW RATE DESIGN FLOW
LESS SEEPAGE RATE LESS FOR
UNIFORM STORAGE RATE
CLARIFICATION
4*
WINTER
SUMMER
SPRING
TIME
Figure 13. Conceptual Sketch of Hydrograph
during Spring Melt Conditions.
44
-------
Once the base width is established and the total volume of excess water to
be released during the spring is estimated, it is possible to sketch the estimated
inflow hydrograph. The design storage required in order that any specified
design outflow not be exceeded is given by the doubled shade area illustrated
in Figure 13. It can be seen from the figure, that, for a given situation,
increasing the design flow for clarification decreases the design storage. At the
same time, however, increasing the clarification design flow increases the
required surface area of the clarification basin. The design storage would
normally be provided by raising the surface elevation of the clarification basin.
Thus, greater storages would necessitate higher dikes whereas a greater
clarification area would require more land area and possibly longer dikes. The
optimal design outflow is that flow which minimizes the overall cost while still
yielding an effluent of the desired quality.
Based on data collected during the study period, the base flows for the Republic
and Empire tailings systems can be estimated as follows:
Plant Inflow
Base Flow = Rate - Seepage Rate - Uniform Storage Rate
At Republic,
Base Flow = 58 cfs - 8.9 cfs - ( + 1.30 cfs - 1.41 cfs)
*
Base Flow 49.2 cfs
The base flow at Empire can be expressed as:
Base Flow = 9.9 cfs- 7.75 cfs -(- 1.31 cfs) =3.46 cfs
Balance calculations indicate that, in the absence of storage, outflow rates
higher than these base flows would have occurred over a period of 45 days at
Republic and 40 days at Empire. As pointed out earlier, however, the high
flow period began earlier at Republic.
In certain instances, the designer may be constrained by maximum (and perhaps
minimum) basin outflow rates specified by regulatory agencies. The minimum
design flow for clarification would be the average annual outflow. At this
outflow rate, storage would be maximized and the accumulated storage would
be gradually released during dry periods.
'i *''
In the operational management of a tailings impoundment and clarification
system, it is essential that attention be given to the hydrology of the basin,
particularly during snowmelt conditions. Prior to, and during melt conditions,
the operator should be aware of the water equivalent of the snow pack and
tailings ice resting on the tributary basin. However, not even the most refined
operational management can result in satisfactory operation of a system in which
adequate storage volume and clarification area are not available.
45
-------
Water Reuse:
From consideration of the overall water balances, it is obvious that reuse of
the clarified process water can reduce the quantity of water discharged from
the tailings basins and thereby lessen the impact on receiving waters. Figure 2
indicates that at the Republic Mine an average of about 63% of the process
water is reclaimed from the tailings basin. Likewise, Figure 4 shows that,
at the Empire Mine, 96%,of the process water is reclaimed from the tailings
thickener.
In general, reuse of spent process water may be limited by the following
factors:
1. A certain quantity of water must remain in the tailings stream to
effect fluid transport of the waste solids.
2. The presence of dissolved and suspended material may interfere
with concentration processes. It is known for example, that di and
tri valent cations can precipitate certain flotation reagents thereby
increasing process chemical requirements.
At both sites, the dissolved solids in the mill effluents were considerably
higher than that in the fresh make-up water. Approximate ranges of dissolved
solids concentrations observed at various points in the tailings systems are
indicated in Table 7. The concentrations measured in the pond effluents tended
to vary seasonally being high in the winter and summer and low in the spring
because of the dilution from melting ice.
Table 7. Approximate Ranges of Dissolved Solids Concentrations (ppm) Observed
At Various Points in Tailings Systems
Empire
Make-Up Mill Pond
Water Effluent Effluent
75 300-380 150-240
Republic
Make-Up
Water
Mill
Effluent
Pond
Effluent
60
110-150 90-130
It is evident that certain materials are solubilized during the concentration process
and precipitated in the impoundment system.
46
-------
The problem of water reuse does not lend itself to a simple analysis/as one is
immediately faced with cost trade-offs between treatment of pond discharge,
concentration process cost, and treatment of the reuse water. Assuming that
a concentrating plant is faced with a set of effluent standards, techniques of
systems analysis (5) could be applied to determine the optima) reuse strategy.
However, before this can be done, some basic information related to reuse
would have to be established. This necessary information would include:
1. Water quality criteria for reuse. It would be necessary to develop
information on allowable concentrations of various substances in
water to be reused in various concentration schemes applied to
various ore types. Information on incremental concentrate produc-
tion costs associated with higher concentrations of these sub-
stances would also be required.
2. An applicable methodology to predict any build-up of important
substances which might occur within a reuse system.
Further research on this topic is obviously needed since this information is
not presently available.
47
-------
SECTION V
TREATMENT OF TAILINGS BASIN OVERFLOWS
The primary objective of this aspect of the research was the investigation of
alternative treatment methods which might be applied both to remove particulate
material from the basin discharge and to facilitate final disposal of the collected
solids. A related, secondary objective, was to assess the feasibility of in-
corporating certain treatment methods into a treatment system for subsequent
demonstration.
The research'plan employed to achieve these objectives relied heavHy on
experimentation with the highly colored 5B outlet stream and tertiary pond at
the Republic Mine. The Empire Mine tailings pond overflow contained very low
suspended solids concentrations (on the order of 10 ppm) and thus did not lend
itself to this aspect of the research.
Investigation of Alternative Coagulants
It is difficult to conceive a process which could effectively replace chemical
coagulation and sedimentation in the treatment of tailings basin overflows. Thus/
some effort was devoted to the development of refinements in the coagulation
flocculation process. The basic experiemtnal procedure employed here consisted
of the well-known jar test. A detailed description of the methodology is given
in Appendix E.
Initial Screening of Various Coagulants
Alum and thirteen polymer coagulants or coagulant aids were initially selected
for jar testing. The polymers consisted of: 5 anionic coagulants, 1 anionic
coagulant aid, 5 cationic coagulants (one of which was also tested as a coagulant
aid) 1 non-ionic coagulant, and 1 non-ionic coagulant aid as shown in Table 8.
The term "coagulant aidljdenotes an agent employed in conjunction with another
coagulant. In this investigation, coagulant aids were evaluated in conjunction with
alum. Several other polymers were excluded from testing for one of the following
reasons: possible toxicity; extreme difficulties involved in preparing standard
polymer solutions; or, in the case of one company, no information was received
regarding type, recommended usage, or price of the polymer samples.
The initial screening jar tests were conducted under the following conditions;
Temperature = 13°C
Initial pH= 7.5
Initial Turbidity = 1,400 Jackson Units
49
-------
Table 8 . Coagulants Subjected to Initial Screening Tests
Type
Alum
Anionic Polymer Coagulant
n n n
n n n
Anionic Polymer Coagulant Aid
Cotionic Polymer Coagulant
ii n M
n n n
n M n
Non-Ionic Polymer Coagulant
Non- Ionic Polymer Coagulant Aid
*Also tested as a coagulant aid.
Manufacturer
Calgon
n
n
Nalco
Dow
Calgon
n
Nalco
Calgon
Dow
Manufacturer's
Designation
M-560
M-570
M-580
M-590
D-2332
A-22
M-500
M-510
M-520 .
Na-603*
Na-607
M-550
N-ll
Given Cost
$/lb.
$0.03
$1.38
$1.38
$1.38
$1.38
$0.95
$1.15
$0.43
$0.44
$2.00
$0.241
$0.30
$1.38
$0.95
To facilitate subsequent comparisons between coagulants/ a system of coding the
various polymers was developed. This method is explained in Table 9 .
Table 9 .
1st Letter
A = an ionic polymer
C = cationic polymer
N = non-ionic polymer
Polymer Coding System.
2nd Letter
C = Calgon
D = Dow
N = Nalco
Manufacturer's
Designation
Refers to a company's
own polymer number.
For example, polymer coagulating agent ACM-560 is an anionic polymer manu-
factured by the Calgon Corporation, with a model number of M-560.
50
-------
Results of Initial Screening Tests
Anionic Polymer Coagulants. Polymers ACM-560 to ACM-590 (model numbers
pertain to relative values of polymer molecular weights) exhibited certain identi-
fiable characteristics, summarized below.
(1) All polymers exhibited rapid, massive floe formation following
their addition to the samples being jar tested. The speed of
floe formation, and the size of the irregularly-shaped floe
clumps increased with polymer dosage and polymer molecular
weight.
(2) Floe particles formed by higher dosages of all polymers ex-
hibited significant deposition during slow mixing operations.
Higher dosages of polymer ACM-590 formed extremely massive,
irregularly-shaped, floe particles which settled oat during
rapid mixing operations.
(3) Increasing dosages and molecular weight of all polymers re-
sulted in increasing supernatant color remaining after
coagulation-flocculation-quiescent settling. This was not
related to levels of residual supernatant turbidity, however,
as turbidity removals increased (or decreased slightly) with
increasing polymer dosages.
(4) Coagulation-flocculation with the polymers tested did not
significantly reduce sample pH or total alkalinity.
Anionic polymer coagulant AND-2332 exhibited the poorest floe formation
and settling characteristics of aH the anibnic polymer coagulants tested.
Figure 14 depicts turbidity removal efficiency as a function of increasing polymer
dosage. It is observed that at dosages above 6 ppm, the lowest molecular-weight
pblymer, ACM-560, gave the highest removals. Increasing dosages (beyond 10
ppm) of the higher modecular-weight an ionic polymers generally resulted in decreased
removal efficiencies.
Anionic Polymer Coagulant Aid. An additional variable, alum concentration, is
introduced when comparing efficiencies of coagulant aids. It proved convenient
to make these comparisons at equal levels of chemical cost per million gallons
treated. Thus, the turbidity removal efficiencies of various, equal cost, com-
binations of alum and coagulant aid were compared. These comparisons were made
over the cost range of 0 to $25 per million gallons. The cost of $25 per million
gallons was equal to the chemical cost at the optimal alum dosage of 100 ppm.
Cost data for the various coagulants are given in Table 8 .
51
-------
Ul
ro
o
E
-------
The turbidity removal efficiency of anionic polymer coagulant aid ADA-22
was dependent on the amount of alum added to the system, as can be seen in
Figure 15. At a constant alum dosage, of 10 ppm, no amount of the
coagulant aid added to the system resulted in turbidity removals comparable
to those obtained with increasing dosages of alum alone. Alum dosages of
30 to 50 ppm/ together with increasing dosages of coagulant aid, resulted in
turbidity removal values higher than identical-cost alum dosages. No further
improvement in supernate was obtained at the highest coagulant aid dosages.
Cationic Polymer Coagulants. Cationic polymer coagulants CCM-500 to
CCM-520 (model numbers pertain to relative values of polymer molecular
weights) exhibited certain common characteristics, relative to polymer dosage
and molecular weight; these are summarized below.
(1) As the polymer molecular weight increased, the turbidity
removal efficiency obtained with low polymer dosages
increased.
(2) The speed of floe formation, and the size of floe particles
formed during a jar test, increased with increasing polymer
molecular weight.
(3) The density of floe particles decreased with increasing
polymer molecular weight.
(4) As molecular weight of the cationic polymer coagulants increased,
amounts of standing floe (surface scum) formed during jar testing
decreased.
(5) All 'COM1 series cationic polymer coagulants were effective in
removing fine turbidity-causing particles, with CCM-510 being
most effective.
(6) All polymers exhibited extremely rapid floe particle settling,
leaving a clear supernatant within 2-3 minutes of quiescent
settling.
Polymer coagulants CNNa-603 and 607 were significantly less effective
than coagulants CCM-500 to CCM-520. The higher molecular-weight polymer,
CNNa-607, exhibited the poorest floe formation and floe particle settling
characteristics of all cationic polymer coagulants evaluated.
Figure 16 depicts per cent turbidity removal as a function of increasing cationic
polymer coagulant dosages. Decreased removal efficiency (and subsequent
increases in sample supernatant turbidity) occurred only with higher dosages of
53
-------
100
o
o
E
a>
JQ
u.
3
H
80
60
!i 40
20
Alunw
p
X
&
f
^
X
. *""""
:
^
Test Conditions:
Initial Turbidity = 1355 JTU
Temperature = 13° C
Initial pH = 7.5
O Alum Alone
Polymer Coagulant Aid ADA-22 plus,
A lOppm Alum
0 30 ppm Alum
n 50 ppm Alum
0 5 10 15 20 25
Coagulant Cost $/mg
Figure 15. Turbidity Removal of Various, Equal
Cost Combinations of Alum and Anionic
Polymer Coagulant Aid.
54
-------
Ol
Ul
100
90
_- 80
o
o
E
o>
o: 70
| 60
£
50
40
Test Conditions:
Initial Turbidity = 1400 JTU
Temperature = 13° C
Initial pH = 7.5
Polymer
OCCM-500
ACCM-510
e CCM-520
D CNNa-603
OCNNa-607
8
10
I 234567
Coagulant Dose,ppm
Figure 16. Comparison of Turbidity Removals for Several Cationic Polymer Coagulants.
-------
CCM-520, the highest molecular-weight polymer1. Polymer CNNa-603 was
further evaluated as a coagulant aid, as it was being used in this manner at
the Republic Mine.
Cationic Polymer Coagulant Aid. Polymer coagulant aid CNNa-603 exhibited
a stronger dependence on the amount of alum coagulant added than the anionic
polymer coagulant aid, ADA-22. Figure 17 illustrates this dependency. Only
at a constant alum dosage of 50 ppm were increasing amounts of the coagulant
aid successful in removing more per cent turbidity than identical-cost dosages
of alum alone.
Non-Ionic Polymer Coagulant Aid. Turbidity removal efficiency of
coagulant aid NDN-11 was dependent on the amount of alum coagulant added
to the system, as illustrated in Figure 18. At alum dosages of 30 and 50 ppm,
increasing amounts of coagulant aid greatly improved coagulation-flocculation
process efficiencies; 10 ppm of alum plus increasing amounts of coagulant aid
were not as effective as identical-cost dosages of alum alone.
Non-Ionic Polymer Coagulant. Figure 19 depicts turbidity removal efficiency
as a function of increasing dosages of polymer coagulant NCM-550, and
illustrates three distinct zones of coagulation-flocculation efficiency: (1) an
initial zone, in which removal efficiency greatly increased as more coagulant
was added to the samples; (2) a narrow zone of optimum turbidity removal
efficiency; and (3) a final zone, in which turbidity removal efficiency greatly
decreased as more than optimum amounts of polymer coagulant were added to
the samples.
Increasing dosages of polymer coagulant NCM-550 formed the largest floe
of any polymer coagulants or coagulant aids . Floe formed immediately after
coagulant addition in rapid mixing operations. Subsequent slow mixing operations
caused the breakup of these massive, clumped floe particles. Substantial
amounts of standing floe (surface scum) were formed at all coagulant dosages.
No effect on sample pH or total alkalinity was observed.
For purposes of comparison, Figure 19 also shows, coagulant dosage - turbidity
removal relationships for the anionic and cationic polymer coagulants judged to be
most effective. In general, the cationic polymer coagulants were most effective in
removing turbidity while the anionic polymer coagulants were least effective.
Extended Evaluation of Most Effective Coagulants
Figure 19 shows that, of the polymer coagulants, CCM-5 10 was the most
effective. Likewise, an examination of Figures 15, 17, and 18 indicates that,
of the coagulant aids, NDN-11, was somewhat more effective than the others,
in augmenting the removal by alum. Consequently, these two polymers, along
56
-------
100
80
i 60
ui
IT
,UM
TEST CONDITIONS:
INITIAL TURBIDITY =I365PPW
TEMPERATURE =I3"C
INITIAL pH = 7.5
OALUM ALONE
POLYMER COAGULANT AID,
CNN«-603 PLUS,
10 PPM ALUM
U30 ppM ALUM
50 ȣu ALUM
0 5 10 15 20 25
COAGULANT COST |/MG
Figure 17. Turbidity Removal of Various, Equal Cost Combinations of
Alum* and Cationic Polymer Coagulant Aid.
<
o
bl
ee
Q
£
ec.
o
oc
20
ALUM-^
/
)
>
fi'
^&2
"~~
_ ~1
TEST CONDITIONS:
INITIAL TURBIDITY = 1355 JTU
TEMPERATURE = I3*C
INITIAL pH = 7.5
OALUM ALONE
POLYMER COAGULANT AID,
NDN-II PLUS,
10 PPMALUM
50 p^ ALUM
0 5 10 15 20 25
COAGULANT COST |/MG
Figure 18. Turbidity Removal of Various, Equal Cost Combinations of
Alum and Cationic Polymer Coagulant Aid.
57
-------
00
IOC
90
* 80
9*
"o
O
E
£ 70
>»
T»
S 60
50
40
I
Test Conditions:
Initial Turbidity ^
Temperature = I3°C
Initial pH=7.5
Coagulant
1400
OACM - 560
OCCM-5IO
DNCM-550
Cost/lb.
i 1.38
J0.44
ft 1.38
8
234567
Coagulant Dose.ppm
Figure 19. Comparison of Turbidity Removals for Most Effective Anionic, Cat ionic
and Non Ionic Polymer Coagulants.
10
-------
with alum, were selected for further testing. The purpose of these detailed
evaluations was to determine the optimal coagulants for various temperature
and turbidity levels. Thus/ the extended evaluations consisted of turbidity
removal comparisons at various equal cost combinations of coagulant and
coagulant aid/ and at various temperatures and initial turbidity levels. Initial
turbidity levels of approximately 400, 700, 1000, and 1400 JTU were
selected; 1400 JTU closely approximated the maximum yearly initial turbidity
level recorded at the Republic Mine tailings basin overflow while 700 JTU
was approximately equal to the average yearly initial JTU level. 1, 13, and
25° Centigrade represented the minimum, average and maximum overflow
temperatures respectively. All jar tests were conducted at optimum slow mixing
and initial pH conditions. Rapid mixing conditions were standard (see Appendix
E).
Figure 20 shows turbidity remaining as a function of coagulant cost for each
of the three coagulants. The effects of sample initial turbidity and temperature
levels upon comparative turbidity reductions of the three coagulating agents
are evident. No single coagulant or combination of coagulants was found to
to be most economical in achieving a given supernatant turbidity under alf
conditions of initial turbidity and temperature. Chemical costs to achieve a
specified turbidity varied with initial turbidity and temperature. However,
for supernatant turbidities greater than about 60 JTU, Calgon M-510 was
the most economical coagulant under all conditions of initial turbidity and
temperature. Approximate chemical costs for Calgon M-510 required to attain
a supernatant turbidity of 60 JTU ranged from $2.50/million gallons at an
initial turbidity of 400 JTU and temperature of 25°C to $8.00/million
gallons at an initial turbidity of 1400 JTU and temperature of 1°C. At low
initial turbidities (400 JTU), this cost to achieve a supernatant turbidity of
60 JTU was virtually equal to the cost of alum plus coagulant aid Dow N-ll
and was about 80% of the cost of alum alone. However, at high initial turbidities
(1400 JTU), the cost of Calgon M-510 to achieve a supernatant turbidity of
60 JTU was roughly 50% of the cost of alum plus coagulant aid Dow N-ll and
roughly 35% of the cost of alum alone.
Comparison of Sludge Volumes Formed by Various Coagulants
In an overall coagulation-fiocculation-sedimentation treatment process, con-
sideration of the thickening capacity of the sludge produced is, perhaps, just
as important as the economic considerations involved in selecting a coagulant
for clarification purposes. Therefore, observations were made to qualitatively
compare the volumes of sludge generated by the various coagulants.
Upon completion of jar tests in which reasonably high turbidity removals were
obtained, the samples were gently hand-mixed and allowed to settle in graduated
Imhoff cones for 30 minutes. The volumes of sludge were routinely noted. Know-
ing the initial and final turbidities, it was possible, with the aid of the suspended
59
-------
300
200
3 100
i-
-D
2 0
o
* 200
o
o
100
- 0
0>
Q.
3
CO
200
100
JTUj
JTUj ac.700
25°C
CCM- 510
Alum
NDN-IUAIum
I3°C
25 °C
CCM- 510
Alum
NDN-ll+Alum
\
I3°C
_L
2 4 6 8 10 12
Coagulant Cost ft/MG
4 8 12 16 20 24
Coagulant Cost $/MG
Figure 20. Turbidity Remaining as a Function of Coagulant Cost.
-------
300
200
100
0
200
100
IOOO
JTU, *I400
\
25° C
\
CCM-510
\
\
Alum
NDN-IU Alum
\
25° C
CCM-510
Alum
NDN-H + Alum
Q
5
o
o
-------
solids-turbidity correlation given in Figure E.I (Appendix E), to estimate the
solids content of the sludge.
Table 10 lists the sludge solids concentrations obtained for various coagulants.
Data used in the constructing of this table were selected so that the supernatant
turbidities were roughly comparable. Thus, comparisons can be made between
coagulant costs and sludge volumes for equal initial turbidities. The CGM-510
coagulant which was, under most conditions, the most economical from a clarifi-
cation viewpoint produced the largest sludge volume. The CNI\la-603, on the
other hand, employed as a coagulant aid, was more expensive but produced the
least sludge volume. During the period of this study, the polymer CNNa-603 was
employed as a coagulant aid at the Republic Mine.
Table 10= Comparison of Sludge Volumes Formed After 30 Minutes of
Quiescent Settling in Imhoff Cones. Initial JTU = 1400.
Temperature = 13°C.
Supernatant Coagulant Approx.
Turbidity, JTU Cost ml Sludge/ Sludge Solids
Coagulant after 30 min. $./MG liter Conc.%
NDN-11, 1 ppm +
30 ppm Alum
CCM-510, 3 ppm
Alum, 100 ppm
ADA-22, 0.89 ppm
30 ppm Alum
CCM-510, 3 ppm
Alum, 100 ppm
ADA-22,0.89 ppm
30 ppm Alum
CN,Na-603, 3.1 ppm + 47 18.70 10.5 6.4%
50 ppm Alum
27
28
66
38
28
66
38
15.40
11.00
25.00
16.00
11.00
25.00
16.00
20
24
15
17
24
15
17
3.6%
3.0%
4.5%
4.2%
3 .0%
4.5%
4.2%
62
-------
Characterization and Analysis of Republic Mine Coagulation-F[peculation System.
The coagulation-flocculation system at the Republic Mine consisted of:
(1) Coagulant addition.
(2) A weir, hydraulic jump and a relatively steep section of stream channel
for rapid mixing (slope = 0.12 ft/ft for 50 ft).
(3) A relatively flat section of stream channel for slow mixing (slope =
0.0024 ft/ft for 800 ft).
In order to judge the effectiveness of this system, comparisons were made between
the f locculation parameters attained in this system and those judged to be optimal
based on laboratory jar tests. The mean velocity gradient, G, and the dimension-
less product of G times residence time -, GT, were the.flocculatton parameters
employed (9).
Equation 9 developed for baffled channels (9) was used to estimate the mean
velocity gradient, G, and GT for the rapid and slow mixing portions of the Republic
Mine effluent channel.
- / Q Ah \ 9.
G = ( = I
where:
G = mean velocity gradient fps/ft
g = acceleration of gravity, 32.2 ft/sec2
Ah = drop in water surface, ft
v = kinematic viscosity, ft2/sec
T = mixing time, sec
The mixing time, T, was taken to be the mean residence time determined from a
dye dispersion test at a flow rate of 28.3 cfs. The mixing time at other flow
rates was estimated by assuming that channel roughness controlled the water
depth and applying the Manning Equation (9) with channel depth equal to hydraulic
radius.
Figures 21and 22 show the calculated values of the mean velocity gradient, G,
and GT for the rapid and slow mixing portions of the effluent channel. Recommend-
ed values of these parameters for rapid mixing are (10):
63
-------
2.0
1.6
o
X
H
O
1.2
0.8
- 1800
0)
(O
eT
I 1400
o
L.
O
° 1000
0>
o
-------
12
10
8
g
x
i-
To 100
o>
m
-------
G = 700 to 1000 sec"1
G T = 2 x 104 to 3 x 104
Figure 21 indicates that, except for low flow rates and low temperatures, the
rapid mix values of G were considerably higher than recommended. On the other
hand, because of the short residence time in the steep channel section, the
rapid mix values of GT were considerably less than recommended.
Figures 23 and 24 show relationships between jar test supernatant turbidity and
slow mixing G and GT values for the Republic Pond 3 effluent. Values of the jar
test velocity gradient were determined by using Camp's (11) correlation. These
figures indicate that, for slow mixing, the optimal ranges are:
G = 35 to 48 sec"1
G T = 3 x 104 to 6 x ID4
Comparing these values with those shown on Figure 22 indicates that, except for
low flow rates and low temperatures, the slow mix values of G were higher than
optimal. Likewise, over most flow and temperature ranges, the slow mix GT values
were higher than optimal. However, over the most critical range, (high flow rates
and low temperatures) the GT values fell within the optimal range.
Based upon the foregoing analysis, it appears that there is room for improvement in
the existing coagulation-flocculation system at the Republic mine. These improve-
ments would consist of modification to bring the G and GT values for both the slow
and rapid mixing operations closer to the optimal values. Since the hydraulics of
the present system are largely set by the natural terrain, a significant improvement
in G and GT would probably require the construction of a mechanical rapid mix-
flocculation system.
Handling of Coagulated Solids
As indicated in Section IV, the coagulated solids settled out upon entering the
Republic Mine tertiary pond. This caused the accumulation of a large delta of urrconsolid-
ated slurry near the mouth of the tertiary pond. The solids content of this
slurry varied from about 15% at the top to 30% at the bottom. This material pre-
sented a problem as it was situated in the stream path and was subject to scour at
high flows. Moreover, the volume of this deposit was rapidly increasing and en- (
croaching upon the pond volume available for clarification. An investigation of
thickening and filtration characteristics was conducted to assess the feasibility of
collecting and dewatering this and any future solids accumulation. Field surveys
showed that, on May 13,1971 , the unconsolidated muck in the tertiary pond
66
-------
240
V)
+-
"E
c
o
CO
je
o
o
o
c
200
160
120
80
40
0
Initial Turbidity = 670 JTU
Alum Dosage = 50ppm
6T= 3.25 X I04
o GT=5.5 XIO4
20
13° C
25°C
70
30 40 50 60
Mean Velocity Gradient, G, Sec"1
Figure 23. Relationship between Jar Test Supernatant Turbidity
and Slow Mixing G Value for Republic Pond 3 Effluent.
67
-------
240
200
£160
o
m
120
Initial Turbidity = 670 JTU
Alum Dosage = 50 ppm
Slow Mixing at 40rpm
Rapid Mix,7min. at 90rpm
I°C, 6 = 40sec"1
25°C, G=55sec-'
GT X 10
Figure 24. Relationship between Jar Test Supernatant Turbidity
and Slow Mixing GT Value for Republic Pond 3 Effluent.
68
-------
occupied a total of 1,438,100 cubic feet. The average density of the muck was
71.6 Ib/ft* and the average solids content was 22.6%. The total weight of dry
solids contained in the unconsolidated deposit, therefore, amounted to approxi-
mately 23.28 x 100 (b. or 11,650 tons. Over the period of June 1970 to June
1971, 6,170 tons accumulated.
The unconsolidated slurry employed in both the thickening and filtration experiments
was obtained from the deposit existing in the Tertiary Pond. Thus, the material
contained some residual alum and polymer. (The quantity of alum added ahead of
the tertiary pond amounted to about 5% of the weight of solids treated; the quantity
of polymer, Nalco, Na-603, amounted to about 0.5% of the weight of solids
treated.) Because of this, the results of both the interface subsidence and filtra-
tion experiments may not be applicable to a sludge generated by other coagulants.
Thickening Characteristics
To determine the thickening capacity of the unconsolidated muck, several interface
subsidence experiments were performed using 1000 ml graduated cylinders. The
control variables in these experiments were:
1. Initial solids concentration, 5%, 10%, 20%.
2. Amount of slow stirring, none,4 revolutions per hour.
3. Addition of thickening aids.
FeCU 3%, 6%, 9% (based on wt. of solids)
Polygalacturonic Acid 0.1%, 0.5%, 1%
CCM-510 polymer 0.1%, 0.5%, 1.0%
CCM-560 polymer 0.01%, 0.05%, 0.1%, 0.5%, 1.0%
In all cases, the direct response measured was the interface subsidence curve.
Unit areas were calculated for the conditions which gave the most promising inter-
face subsidence curves. All interface subsidence tests were conducted at room
temperature.
Slow stirring or slow raking action was achieved by using an apparatus similar
to that described by Eckenfelder (12). The stirrers were powered by synchron
motors which revolved at 4 rev/hr.
Figures 25 and 26 show interface subsidence curves for initial concentrations of
10 and 20% with and without slow stirring. The final solids concentrations attained
69
-------
CO
UJ
I
o
H
X
CD
UJ
X
UJ
o
<
u.
(T
UJ
16
14
12
10
8
6
4
2
INITIAL SOLIDS CONC. =10%
WITHOUT STIRRING
WITH STIRRING
100
200 300 400
TIME IN HOURS
500
700
800
Figure25. Interface Subsidence Curves for Coagulated Solids
Generated at Republic.
16
Z 14
x
o ,
* 10
i-
i e
u
UJ
o
cc
UJ
I I r
WITHOUT STIRRING
WITH STIRRING
NITIAL SOLIDS CONCE NTRATION = 20%
0
100
zoo
300
400
500
600
700
800
TIME IN HOURS
Figure 26. Interface Subsidence Curves for Coagulated Solids
Generated at Republic.
70
-------
at 800 hr. are summarized in Table 11.
Table 11. Solids Concentrations Attained in Batch Thickening Tests
Final Solids %
Initial No
Solids % Stirring Stirring
5% 17% 19.5%
10% 22.5% 35%
20% 24% 41%
Following the minimum solids flux method outlined by Dick (13) it was possible
to estimate thickener unit areas required for various underflow concentrations from
the interface subsidence curves. Estimates of the unit areas required to attain
underflow concentrations of 20% and 30% are shown in Table 12.
Table 12. Estimate of Unit Areas, ft?/lb/sec , No Thickening Aid
Feed Cone. %
i 10% 20%
Underflow
Concentration Stirred Not Stirred Stirred Not Stirred
20%
30%
9,100 117,000
90,000 - 32,300
Influence of Thickening Aids
A series of interface subsidence experiments was conducted to investigate the
effect of various thickening aids. Here, the concentrations of thickening aids em-
ployed are expressed as a percentage of the weight of solids. In each experiment,
an appropriate amount of a concentrated solution of the thickening aid was added to
a 10% suspension of slurry. After gentle mixing the slurry was placed into 1000
71
-------
ml graduated cylinders and allowed to settle usually under the influence of slow
stirring. In several cases, the effect of slow stirring was observed by inserting
the stirring apparatus when the interface appeared to reach an equilibrium level in
the absence of stirring.
Ferric Chloride; As a thickening aid, ferric chloride was ineffective at concentra-
tions of 3%, 6%, and 9%. It may have been that any additional benefit to be gained
by adding a tri-valent metal ion was precluded by the alum residual present in the
slurry. *
Polygalacturonic Acid: The investigation of this compound was prompted by a
suggestion received from the E.P.A. project representative. This compound was
ineffective at concentrations of 0.1%, 0.5% and 1.0%.
Cationic Polymer, Calgon M-510; This substance was the most economical
coagulant as far as turbidity removal was concerned. In addition, it was the most
effective of the vacuum filtration aids tested. However, it was only slightly
effective as a thickening aid at concentrations of 1%, 0.5% and 0.1%. Under
quiescent conditions, the initial rate of interface subsidence, for concentrations
of 0.5% and 1.0%/was increased markedly over that of the control. However,
the interface reached an equilibrium height of 10 inches at 50 hours. Further sub-
sidence took place only when stirring was applied at 140 hours. The raking action
resulted in an interface subsidence pattern very similar to that of the stirred control.
Overall, this substance did not produce a significant improvement over the interface
subsidence pattern of the stirred control.
Anionic Polymer, Calgon M-560: This polymer was selected for investigation as a
thickening aid on the basis of visual observations made during coagulant jar tests.
Overall, the Calgon M-560 polymer was the most effective thickening aid tested.
Figure 27 shows the effect of this substance on the interface subsidence curves at
concentrations of 0.01%,and 0.5%.
At an initial solids content of 10%, the maximum sludge concentration of approxi-
mately 35% was reached. This was comparable to that attained in the correspond-
ing experiment with no thickening aid. At concentrations of 0.1% and above, this
polymer caused serious "clumping" in the stirred thickening cylinders. This clump-
ing interfered, with the raking action as the clumps tended to move ahead of the
raking mechanism. A higher stirring speed tended to eliminate this problem.
Fortunately, this clumping action did not occur at the 0.05 % level. Thus,
0.05% of Calgon M-560 was taken to be the optimal thickening aid.
Table 13 lists the unit areas determined from the interface subsidence curve.
72
-------
TEST CONDITIONS*
TEMPERATURE AT 22° C
STIRRING AT 4 RPH
(0% INITIAL SOLIDS CONCENTRATION
POLYMER USED'- CALGON M-560
QUIESCENT CONTROL
O STIRRED CONTROL
0.01% POLYMER
A 0.05% POLYMER
30 40
TIME IN HOURS
Figure 27. Influence of An ionic Polymer, Calgon M-560, on the Interface
Subsidence Pattern of Coagulated Solids Generated at Republic.
73
-------
Table 13. Estimate at Unit Areas, ft2/lb/sec, Calgon M-560 Polymer
Employed as a Thickening Aid in the Presence of Slow Stir-
ring. Feed Solids Concentration = 10%.
M-560 Polymer Concentration
Underflow
Concentration 0.01% 0.05%
20% 8,400 2,080
30% - 9,530
The beneficial effect of this polymer is easily seen by comparing the unit areas
listed in Tables 12 and 13. To thicken from 10% to 20% under stirred conditions
required a unit area of 9,100 ft2/lb/sec without thickening aid. The same volume
reduction in the presence of 0.05% Calgon M-560 required a unit area at only
2,080ft2/lb/sec.
Vacuum Filtration Characteristics
Bu'chner Test Results. The vacuum filtration characteristics of the slurry were in-
vestigated using both the Bu'chner Funnel Test and the Filter Leaf Test. The
Bu'chner test was conducted according to the procedures outlined by Eckenfelder
(12) and Rich (3). The chief objectives of these tests were:
1. To evaluate certain constants used in filter design.
2. To evaluate the effect of various filter aids.
A total of 97 individual Bu'chner tests were conducted. The control variables in
these experiments were feed solids content, filtration vacuum, and type and con-
centration of filter aid. During each test, the volume of filtrate, V, was recorded
as a function of time, t. A plot of (t/V) versus V for each test normally yielded a
straight line. The slope of the line was then measured graphically and employed
to calculate the specific resistance of the filter mat 02).
The observed effects of filter aids on the specific resistance of the filter cake are
summarized in Table 14. The feed solids concentration used in these experiments
was 10%. In the case of ferric chloride, specific resistance was determined for
filter aid concentrations of 3%, 6% and 9%, whereas, for the polymers, specific
resistance was normally determined at dosages of 0.01%, 0.05%, 0.10%, 0.25%,
74
-------
and
?u FT the,aidS applied singly' and for the eclual dosa9es of M-510
the values listed in Table 14 represent the optimal dosage.
Table 14. Effects of Filter Aids: Summary of
at Various Filter Aid Dosages
Filter Aid
NONE
Ferric Chloride
Cationic Polymer
Calgon M510
Non-Ionic Polymer
Calgon M550
An ionic Polymer
Calgon M560
Combination
Calgon M510
Calgon M560
Combination
Calgon M510
Calgon M560
Combination
Calgon M510
Calgoh M560
Dosage of
Filter Aid
(% based onwt.
of solids)
3%
1%
0.5%
0.05%
0.25%
0.25%
0.2%
0.05%
0.4%
0.05%
Specific Resistance Values
Specific Resistance
at Given Filter
Aid Dosage, sec^/g
24.8 xlO7
10.64 xlO7
6.01 x107
8.58 xlO7
14.9 xlO7
4.02 xlO7
18.8 x 107
11.75 xlO7
In the thickening experiments, the non-ionic polymer Calgon M-560 was found to
be the most effective aid at a concentration of 0.05%. When used alone, as a
filter aid, the same concentration produced some improvement in specific resistance,
Higher concentrations, however, were ineffective. On the other hand, the cationic
polymer, Calgon M-510, was ineffective in thickening but was extremely effective
75
-------
in reducing specific resistance. To investigate the effect which might result by
adding M-510 to a sludge which had been thickened using M-560, a series of
experiments was conducted in which the two polymers were added in equal amounts.
The corresponding curve shown in Figure 28 indicates a synergistic effect. This
suggested that it would be practical to use the M-560 in a thickening process
ahead of a filtration process employing M-510.
Leaf Test Results. To determine the practicality of dewatering the sludge by
vacuum filtration, a total of 53 filter leaf tests were conducted using an Eimco
Filter Leaf Test Kit. The control variables in these experiments were: the filter
fabric, filter aid and concentration, form time and dry time. In nearly all experi-
ments, the feed solids content was maintained at 20%. The general test procedure
followed that outlined by Eckenfelder (12).
The results of these experiments are summarized in Table 15. Of the various
fabrics supplied with the test kit, the nylon fabric (Eimco No. NY 3I7F) was
judged to be the best on the basis of percent solids remaining in the filtrate and
ease of cake removal. It was found that a slight positive air pressure greatly
aided in cake removal. On the basis of these experiments, the optimal form time
was judged to be 30 seconds. Likewise, the minimum dry time was established to
be 30 seconds. Allowing 30 seconds for cake removal would give a cycle time of
1.5 minutes at 33% submergence. Based on a 1.5 minute cycle time and 20%
feed concentration, the filter loading rates calculated directly from the leaf test
results ranged from 13.9 to 28.1 Ib/hr/ft2 for the M-510 used alone, from 12.3
to 20.1 Ib/hr/ft2 for the various combinations of M-560 and M-510 and from
10.3 to 12.6 Ib/hr/ft2 with no filter aid.
Drying Characteristics in Air
Several drying experiments were conducted in still air under controlled temperature
and humidity conditions employing a sludge slurry with an initial solids content of
20%. The drying pans used in these tests contained a layer of coarse sand and
were equipped with drains so that gravity dewatering could be distinguished from
drying. Slurry depths of 0.5 in. and 1.0 in. were used. The temperature and
humidity conditons were: 37 F at 95% R.H. and 64°F at 40% R.H. Figure 29
shows the results of these tests. Gravity dewatering was effective only over the
first 25 hours. From 25 hours to 50 hours, the drying rate generally decreased
until shrinkage caused large surface cracks to develop. This aid.ed the drying
process by increasing the effective contact area.
Field drying experiments were conducted adjacent to a fully instrumented weather
station. In these tests, both covered and uncovered pans were employed. The
covers allowed circulation of air but eliminated precipitation. The results of these
tests showed that precipitation caused the solids in the uncovered pans to liquify
and seal the sand surface. Subsequent drying was hindered because large shrinkage
76
-------
E
o»
rf*
u
0)
V)
r-
o
X
UJ
o
z
<
h-
00
CO
UJ
a:
o
o
UJ
a.
CO
INITIAL SOLIDS CONC.= IO%
CALGON M560
CALGON M5IO
CALGON
M5IO 8 M560,
(EQUAL AMOUNTiS)
0 .2 -4 .6 .8 1.0
TOTAL COAGULANT DOSE AS A PERCENT
OF SLUDGE SOLIDS
Figure 28. Influence of Filter Aid Dosage on Specific
Resistance of Filter Cake.
77
-------
Table 15. Summary of Leaf Test Results, Concentration of Feed
Solids = 20%
Fabric
(Eimco No,)
Nylon
(NY317F)
Polypropylene
Dynel
(CY 453)
Nylon
(NY317F)
Nylon
/MV ^1 7F^
\\\ Y J J. /r /
Polypropylene
(popr 913F)
Polypropylene
(popr 913F)
Dynel
(DY453)
Dynel
(DY453)
Filter
Calgon
No.
M510
M510
M510
M510
M560
M510
MKAD
IVUOU
M510
M560
M510
M560
M510
M560
M510
M560
Aid
Conc.%
0.5
0.5
0.5
0.15
0.15
0.05
On*?
.U->
0.15
0.15
0.05
0.05
0.15
0.15
0.05
0.05
% Solids in Filtrate
with aid w/o aid
0.19
0.25 0.40
0.35 0.80
0.05
01 n
0.05 0.40
0.17 0.40
0.07 0.8
0.17 0.8
% So lids
cake with
(30 sec.
time)
40.7
42.6
39.9
44.8
AQ -2
*T7 *J
42.0
46.9
42.5
45.0
in
aid
dry
cracks did not develop. These results suggested that effective air drying would be
attainable only on covered drying beds. A mechanical heat-drying operation appears
to be more practical if some use for which a relatively pure material is required to be
made of the dried solids.
78
-------
NO
o
\_
o
Q.
Depth: I
64° F
40% RH
Depth' 1/2
64° F
40% RH
Depth: 1/2
37° F
95%RH
Depth: I
37°F
95% RH
20
100
250
ISO 200
Time, Hours
Figure 29. Gravity Dewatering and Drying of Coagulated Solids.
300
350
-------
Feasibility of a Possible Solids Handling System at the Republic Mine
The results obtained above suggest that it may be feasible to handle the coagulated
solids by a process system similar to that outlined in Figure 30. The data show
that it definitely is possible to coagulate, thicken and filter the solids. Although
no experimental data were obtained relative to heat-drying, the results of the air
drying experiments suggest that mechanical heat-drying would be possible.
Based on the results of the settling, thickening and filtration characteristics of
the solids, reasonable estimates of the areas required for settling, thickening and
vacuum filtration can be made. For design of a full scale system, however, more
accurate estimates might be made based on continuous flow pilot studies. Never-
theless, these reasonable estimates can be used in conjunction with available cost
information to roughly estimate the average annual cost of this solids handling sys-
tem .
In synthesizing this system, it was assumed that water quality requirements for
reuse prohibited both the addition of coagulents within the reuse system and in-
creasing the average percentage of process water recycled. It was also assumed
that storage within the impoundment system would be sufficient to produce a nearly
constant effluent flow rate close to the average annual rate of 14.7 cfs observed
over the study period and that coagulated solids would be generated at the average
annual rate of 6170 tons/yr. or 16.9 tons/day observed over the study period.
As previously stated, the coagulated solids employed in the thickening and filtra-
tion experiments were obtained from the Tertiary pond and contained some residual
alum and Nalco, Na-603 polymer. Therefore, it was assumed that the same
coagulants would be employed ahead of the system shown in Figure 30. Also,
since coagulation is presently practiced, the cost of coagulant chemicals was not
included in the cost estimate.
The quantities of alum and polymer added amounted to approximately 5% and 0.5%,
respectively, of the weight of solids treated. At the average initial solids content
(427 ppm, 850 JTU) this dosage resulted in a chemical cost of $9.60/MG or
$5.41/ton of solids. If coagulated solids produced by Calgon M-510 thickened
adequately, a 40% to 60% saving in chemical cost might be realized (see
Figures 17 and 20).
The cost estimates developed in the following section have been adjusted to reflect
September 1971 levels. In most instances, this adjustment was based on the
Engineering News Record Building Cost Index. (For Sept. 1971 the ENR Building
Cost Index = 990) Capital Costs have been converted to average annual costs
using a capital recovery factor of 0.11683 based upon an interest rate of 8% and
an amortization period of 15 years. Ten percent of the first cost was assumed to
cover engineering costs.
80
-------
SB Overflow
Circular
Clarifier
>
20%
Solid
wwll U
r
Vacuum
Filter
^
49%
f Soli
Dryer
I
Dry Solids
To
Existing
Tertiary
Pond
Figure 30. Possible System for Handling Coagulated Solids at the
Republic Mine.
81
-------
Clarification: The chief purpose of the circular clarifier shown in Figure 30 is to
remove the bulk of the solids and to allow the deposited sludge to be collected and
transferred to the thickener. This operation would be located near the inlet to
the existing tertiary pond. Since the effluent would subsequently pass through the
tertiary pond/ a conservative design would not be required.
In a properly designed basin it might be possible to collect the deposited solids
without a mechanical sludge collector. This could result in a considerable savings
in first cost. However, the following analysis is based upon a circular clarifier
equipped with mechanical sludge collection.
Settling column analyses of the 5-B overflow coagulated with alum (25 ppm) and
Nalco Na 603 (2 ppm) indicated that the effluent concentration could be reduced
to about 50 ppm with detention time of about 1.67 hr. in a basin 6 ft. deep.
(Overflow rate = 650 gal/day/ft2). The sludge concentration attained in the bot-
tom of the column after 1.67 hr. was about 5%. For this type of flocculent settling
a deeper basin would give a higher removal efficiency and facilitate sludge thicken-
ing. Hence iUvas assumed that a basin 10 ft. deep operated at an overflow rate
of 650 gdp/ft would give the desired removal and produce an underflow solids
concentration of 10%. The clarification area required is then estimated as 1 .46 x
square feet, equivalent to two circular clarifiers, each 96 ft. in diameter.
The unit cost, based on September 1971, is estimated at $25.70/ft2 includ-
ing appurtenances and engineering (14). Thus:
Capital Cost = 1.46 x 104ft2 ($25.70/ft2)= $373,000
Annual Capital Cost = $373,000 (0.11683)= $43,800
Capital Cost per Ton of Dry Solids = $43,800
6,170 Tons
11 = $7.10/ton
Thickening; Since the solids content of the thickener underflow affects the required
filter area, the thickening and filtration operations must be considered together. A
comparative cost analysis of both operations showed that it is more economical to
thicken to 20% solids than to thicken to 30% solids. Thus, the estimates are
based on thickening to 20% solids.
Table 13 shows the required unit area to be 2080 ft2/lb/sec with 0.05% Calgon
M-560. Using an excess capacity or scale-up factor of 2, gives a unit area of
4160 ft2/lb/sec. The mass rate of solids flow is 6170 tons/yr. or 0.392 Ib/
sec on a dry basis. Thus,
82
-------
A = 4160 ft2 (0.392 lb/sec)= 1635 ft2
Ib/sec
(1 circular thickener, 46 ft in diameter)
The unit cost, (September 1971) including appurtenances and engineering was
estimated (14) as $52.80 /ft2 Thus:
Capital Cost =1635 ft2 ($52.80/ft2) = $86,300
Annual Capital Cost= $86,300(0.11683)= $10,060
Capital Cost/Ton Solids = $10,060 = $1.63/ton
6170 Ton
The Calgon M-560 polymer costs $1.38 per Ib. Thus, the annual chemical cost
for thickening would be:
Ib
6170 (2000) lb/yr(5x 10-4"JF)($1.38/lb)= $8,550
or $ 1.38 per ton of dry so I ids.
Filtration: Equation 10 was employed to estimate filter loading rates (12).
L = 35.7
P(|-s)y
R
o
0.5 in
m iu
c
m
tn
where: Ib/ft2
L = loading rate, hr
... .P = vacuum, psi
p - filtrate viscosity, centipoise
R0 = specific resistance, sec2/g x 10~7 (determined from
Buchner Test) divided by p (vacuum during Buchner Test)
raised to the s power.
s = coefficient of cake compressibility (determined from Buchner
data by plotting log specific resistance vs log P)
C = cake solids deposited per unit volume of filtrate, g/ml
m = exponent (determine from leaf test data by constructing a
log-log plot of loading vs. initial feed solids at constant
cycle time)
t = cycle time, minutes
n = exponent (determined from leaf test data by constructing
plot of loading vs. form time at constant initial feed solids)
83
-------
y= fraction of cycle time devoted to cake formation
The constants m, n and s were evaluated from the laboratory data as:
m = CL8Q
n =0,88
s = 0,43
The filter loading rate was estimated for the following conditions:
Filter Aid: 0.05% Calgon M-510 Polymer
0.05% Calgon M-560 residual from thickening
Feed Solids Content = 20%
Temperature = 55 deg. F.
Vacuum = 9.8 psi
Cycle Time = 1.5 minutes, 0.5 minutes form time
Figure 28 indicates that at dosages of 0.05% Calgon M-560 and 0.057? Calgon
M-510, the specific resistance would be no greater than 16 x 10? sec2/g. Thus,
the loading rate can be estimated from Equation 10 as 10.45 Ib/hr. This value
ft*
is somewhat more conservative than the values at 12.3 to 20.1 Ib/ft^ determined
directly from the leaf test data for various combinations of Calgon M-510 and M-
560 polymers. In view of this, an excess capacity factor was not applied here.
If it is assumed that the filter operates 6 hr/shift or 18 hr/day, the solids loading
rate during operation is 1880 Ib/hr and the required filter area is 181 ft2.
The capital cost of the filter including building, appurtenances and engineering,
was estimated (14) as $159,500. The annual capital cost, then is, $159,500
(0.11683) = $18,380. The Calgon M-510 polymer costs $0.44 per IB result-
ing in a chemical cost of $2,720 per year. The power is estimated as 05)
$1,790 per year so that the total capital plus power and chemical cost amounts
to $22,890 per year or $3.71 per ton of solids.
Mechanical Heat Drying; A rough estimate of the probable cost of heat drying the
filter cake at 49% solids can be made by using literature sources.
It will be assumed here, for purposes of estimating, that a direct heat rotary dryer
will suffice. The cost of a direct heat rotary dryer, including buildings, appurten-
ances and engineering cost was estimated using information from Perry (16) as
$252,000. This amounts to $29,040 per year or $4.70 per ton.
84
-------
The dryer would handle 1410 Ib solids/hr and 1475 Ib water/hour. Assuming
the specific heat of the solids to be 0.2, the initial temperature to be 55°FL and
neglecting heat recovery, the heat requirement would be roughly 1.73 x 10
BTU/hr. If the overall thermal efficiency is assumed to be 60% (17), 2.89 x
10° BTU/hr would have to be supplied. If fuel oil with a heat value of 1.4 x
BTU/gal can be obtained for $0.10/gallon, the annual fuel cost would be
BTU $0.10/gaj. (365)(24)=$18/100
clO'°TRr" 1.4 y lCP"""BTD7gal
or $2.93 per ton of dry solids.
For comparison, the fuel cost for drying phosphate slimes at an initial solids con-
tent of 30% in a fluid bed operation has been estimated as $2.50 per ton of dry
solids (17).
Summary: Solids Handling Cost Estimate; The solids handling cost estimate is
summarized in Table 16. The operation and maintenance costs were estimated on
the basis of 8,760 hr of operation per year requiring an average of 1.5 men at an
average wage cost of $7 per hour per man. An additional $7,500 per year was
allowed for maintenance. The total cost of $37.65 per ton compares well with
costs encountered in processing sewage sludges.
Ultimate Disposal of Dewatered Solids
In view of the relatively plentiful amount of land available it is questionable whether
this system can be justified unless the 16.9 tons/day of dried solids can be shown
to possess some economic value. The dried solids, if pulverized, would consist
of a red powder. It is conceivable that this material might be useful as a component
of concrete building materials, as a soil sealant, as a binding material, or perhaps,
as a pigment. Any economic value attached to the dried solids would help to offset
the cost of the processing system considered here.
In the absence of any economic value, the dry or semi-dry solids would probably have
to be ultimately disposed of by burial. Direct spreading on the land surface would
probably not be suitable unless special precautions were taken to prevent resuspen-
sion and erosion. If land disposal or burial were to be practiced/drying of the solids
would not be necessary and the cost per ton of dry solids would decrease to about
$26 per ton (assuming that 25% of the 0 & M cost is attributable to drying) exclu-
sive of final disposal.
Alternative Strategies for Handling Tailings Basin.Overflows
In synthesizing the system discussed above, it was assumed that water quality
requirements for reuse prohibited both the addition of coagulants within the reuse
85
-------
Table 16. Summary of Solids Handling Cost Estimate
Initial Average $/Ton Dry
Capital Cost Annual Solids
Clarification and Primary
Solids Collection
Capital Cost $373,000 $ 43,800 $ 7.10
Thickening ,
Capital Cost 86,300 10,060 1.63
Chemical Cost 8,550 1,38-
Filtration
Capital Cost 159,500 18,380 2.98
Power 1,790 0.29
Chemical 3,720 0.44
Heat Drying
Capital Cost 252,000 29,040 4.70
Fuel 18,100 2.93
0peration and Maintenance 100,000 16 > 2 0
TOTAL $870,800 $232,440 $37.65
system and increasing the average percentage of process water recycled. However,
it was pointed out in Section IV that little information is presently available on water
quality criteria for reuse. Thus in a particular case, the optimal strategy wpu Id de-
pend on the reuse water quality requirements.
The polymeric thickening aid employed with the reuse system at the Empire Mine
apparently does not interfere with the magnetic concentration process. If a coagulant
were to be added to the Republic tailings upon entering Pond 1, the fine particles
could be induced to settle with the coarse fraction and the red water problem might
be eliminated. However, before the practicality of this and other schemes involving
changes in the recycle system can be assessed, more information on recycle water
quality requirements must be developed.
86
-------
SECTION VI
SETTLING CHARACTERISTICS OF TAILINGS
PARTICLES IN NATURAL WATER SYSTEMS
The objective of this aspect of the study was to investigate the physical and trans-
port characteristics of the supra-colloidal particles contained in tailings basin
discharges with a view toward predicting the rate at which these particles might
be removed from a natural water system.
The research plan employed to achieve this objective relied on experimentation with
both the Republic and Empire. Mine tailings .The general approach involved column
studies conducted to investigate the effects of diluent water, temperature and alum
on settling properties. These studies were complemented by a field study conducted
on the Republic Tertiary Pond effluent stream. The purpose of the field study was
to compare removals predicted from column data with that occurring in the stream.
Finally particle size distributions determined by microscopic techniques were com-
pared with these calculated from gravity settling data and Stokes Law.
Quiescent Column Tests
In these analyses, the tailings water containing the fine particles was diluted by
factors of 5 and 10 with various natural waters. Upon dilution, the tailings-
natural water mixture was allowed to settle under quiescent conditions. The pro-
cedure followed was similar to that described in Appendix B, except that the
settling columns were only 30 inches high and had only 4 sampling ports spaced
at 6 inch intervals.
The Republic tailings water used in these experiments was obtained from the 5B
outlet before alum addition. Since the Empire tailings pond effluent generally con-
tained less than 20 mg/l of suspended solids, the Empire tailings used in these
experiments were obtained by presettling a sample of the influent to the Empire
Basin until the supernate contained a suspended solids concentration comparable to
that attained at Republic. Table 17 summaries the conditions employed during
each series of quiescent settling experiments.
Various natural waters were used in the settling analyses to determine the effects,
if any, of the natural water source on fine tailings particle setteability. Natural
waters were obtained from the following sources:
1. Gambles Creek located near the Republic tailings basins.
2. Tertiary effluent stream located at the Republic tailings basin (contained some
residual alum and polymer).
87
-------
Table 17. Quiescent Settling Column Tests Performed
Series
1
2
3
4
5
6
7
8
Fine
Tailings
Sample
Republic
n
it
n
Republic
ii
11
H
Republic
n
H
»
Empire
Empire
Empire
11
Republic
n
Empire
n
Republic
11
Empire
n
Column
No.
1
2
3
4
1
2
3
4
1
2
3
4
1
2
1
2
3
4
1
2
3
4
Natural
Water
Lake
Michigan
n
1!
II
Lake
Superior
n
n
n
Gambles
Creek
n
Tertiary
Effluent
n
(Same as
(Same as
Schweitzer
Creek
n
Lake
Michigan
n
11
n
Lake
Superior
n
n
H
Dilution
Factor
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
Series 1)
Series 2)
V5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
Water
Temp.°F
40
40
70
70
40
40
70
70
40
40
40
40
40
40
40
40
40
40
40
40
40
40
Alum
Added
none
11
n
ti
none
n
n
11
none
n
n
n
none
n
2 ppm
{
H
II
II
II
II
II
If
88
-------
3. Schweitzer Creek located near the Empire tailings basin.
4. Lake Michigan (from the water treatment plant intake at Gladstone, Michigan).
5. Lake Superior (from the water treatment plant intake at Marquette, Michigan).
Table 18 gives the dissolved solids concentration and conductivity for each of the
natural waters.
Table 18. Dissolved Solids and Conductivity of the Natural Waters Used
in the Settling Analyses
Dissolved Solids Conductivity
Natural Water ppm micromhos/cm@20°C
Gambles Creek 71.2 41.0
Schweitzer Creek 88.0 66.3
Tertiary Effluent 123.0 95.0
Lake Michigan 146.6 122.0
Lake Superior 75.8 56.5
Figures 31 and 32 show typical iso-percent removal curves. Each of these figures
summarizes the data obtained from a single settling column experiment. These
curves are analogous to the iso-concentration curves presented in Section IV.
However, instead of plotting concentrations remaining at a particular time and depth,
percent removals based on the initial suspended solids concentration were plotted.
The iso-percent removal curves were then drawn by interpolation. Each of the iso-
percent removal curves traces the path of a particular size fraction of particles as
the fraction is being removed from suspension over a period of time. Downward
curvature of the removal lines indicates the flocculating nature of the suspension.
A suspension of non-flocculating particles would produce linear removal lines.
Table 19 summaries the settling column data. Total removals were calculated from
the plots at 100 hr., 200 hr., and 300 hr. settling times for Republic tailings
suspensions and 8 hr., 16 hr., and 24 hrs. for Empire tailings suspension. These
total removals represent the percent removals which would occur in an ideal plug
flow stream, 27 inches deep after the stated passage times. It has been assumed
that particle scour can be neglected. These removals were computed from the corres-
ponding iso-percent removal curves according to the procedure given by Rich (3).
89
-------
-------
Table 19. Summary of Quiescent Settling Column Data
Init. S.S. Total Removal. %
Natural Water Tailings
L. Michigan Republic
L. Michigan Republic
L. Michigan Republic
L. Michigan Republic
L. Superior Republic
L. Superior Republic
L. Superior Republic
L. Superior Republic
Gambles Cr. Republic
Gambles Cr. Republic
Tertiary .Effl . Republic
Tertiary Effl. Republic
L. Michigan Empire
L. Michigan Empire
L. Michigan Empire
L. Michigan Empire
L. Superior Empire
L. Superior Empire
L. Superior Empire
L. Superior Empire
Schweitzer Empire
Schweitzer Empire
Dilution
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
Tempop
40°
40°
70°
700
40°
40°
70°
70°
400
400
40°
40°
40°
40°
700
70°
40°
40°
70°
70°
400
400
2 ppm Alum Addition
L. Michigan Republic
L. Michigan Republic
L. Superior Republic
L. Superior Republic
L. Michigan Empire
L. Michigan Empire
L. Superior Empire
L. Superior Empire
1/5
1/10
1/5
1/10
1/5
1/10
1/5
1/10
40°
40°
ft
40°
400
40°
40°
40°
40°
cone . , ppm
88
45
85
40
90
44
86
45
90
42
80
40
85
43
85
45
90
44
85
43
90
46
80
40
85
42
90
48
90
43
300 hr
80.6
84.6
87.6
91.3
74.2
81.6
85.0
89.8
69.3
77.6
79.0
90.4
24 hr
S973-
87.7
93.2
87.8
88.9
87.9
89.4
93.4
90.2
88.3
300 hr
77.6
90.7
82.4
88.1
24 hr
87.7
88.4
88.5
89.0
200 hr
64.1
76.1
79.6
85.3
63.8
74.3
74.8
85.6
52.4
55.9
65.7
83.0
16 hr
£570"
84.7
92.1
85.3
85.8
'84.0
86.1
88.0
86.0
84.7
200 hr
71.6
81.7
69.1
76.4
16 hr
84.4
84.1
83.6
83.2
100 hr
50.2
61.1
66.0
70.0
45.1
52.0
56.6
65.0
34.9
33.6
37.6
56.3
8hr
7676"
78.3
77.5
74.1
80.2
75.8
80.9
82.1
76.7
76.7
100 hr
ai^_l_pBlllll^^_HI
58,0
64.5
43.9
57.1
8hr
70.2
76.4
78.4
69.4
91
-------
Analysis of Factors Affecting Tailings Removal
To examine the effects of temperature, dilution, and natural water constituents on
the removal efficiency of Republic and Empire fine tailings particles/ plots of the
total removal versus settling time were made using the data given in Table 18.
Republic Tailings Removal
Figure 33 shows the effect of temperature and dilution on the removal of Republic
fine tailings in Lake Michigan and Lake Superior water. The overall effect was an
increase in total removal with an increase in the water temperature. However/ the
increase in total suspended solids removal cannot be entirely accounted for by the
decrease in fluid viscosity alone. Increased settling rates apparently resulted
from a greater tendency for particles to flocculate at high temperatures. For example,
for the settling of Republic fine tailings at a 1/5 dilution in Lake Michigan water/
it took approximately 250 hours for 50% of the particles to settle through a depth
of 27 inches at 40 degrees. The same removal through the same depth required
only 100 hours at 70 F. This represents a 150% increase in settling velocity.
Stokes Law predicts that the decreased viscosity would cause the settling velocity
to increase by only 52.5%. Therefore/ the increase in settling velocity caused by
the decrease in viscosity only accounts for approximately 35% of the total increase.
The effect of temperature at a dilution factor of 1/10 was similar to the 1/5 dilution
runs. However/ the differences between the temperature curves were slightly
less pronounced.
Figure 34 shows the effect of dilution on the removal of Republic fine tailings
from several natural waters. In nearly all cases/ increased percent removals occur-
red at increased dilutions. The effect of dilution was the greatest for the tertiary
effluent stream and the smallest for Gambles Creek water. Interpreting this effect
in light of Table 18 suggests that it was related to the dissolved solids content of
the diluent water. Figure 35 compares the effects of various diluent waters and
offers further evidence that the dissolved solids content of the diluent water is
directly related to the percent removal obtained. It is reasonable to expect this
since the dissolved solids would act to lower the zeta potential of the negatively
charged particles. The Tertiary Effluent Stream (D.S.=123 ppm) tended to
produce higher removals than Lake Michigan (D.S.=146.6 ppm). This was prob-
ably a result of the residual concentration of alum and polymer remaining in the
Tertiary Effluent Stream.
Empire Tailings Removal
The Empire fine tailings were removed from suspension at a much faster rate than
Republic fine tailings particles. For example/ after 32 hours settling time/ over
92
-------
sO
1/10 DILUTION
LAKE MICHIGAN
O INITIAL S.S. CONC. = 40 ppm
INITIAL S S CONC = 45 Dom
1/5 DILUTION
LAKE MICHIGAN
O INITIAL S.S. CONC. = 85 ppm
INITIAL S.S. CONC = 88 DDITI
1/5 DILUTION
LAKE SUPERIOR
1/10 DILUTION
LAKE SUPERIOR
O INITIAL S.S. CONC.= 45ppm
INITIAL S.S. CONC. =44 ppm
O INITIAL S.S. CONC.= 86ppm
INITIAL S.S. CONC. = 86 ppm
100 200 300 0 100 200 300
SETTLING TIME, HOURS SETTLING TIME, HOURS
Figure 33. Effect of Temperature and Dilution on Removal of Republic Fine Tailings
in Lake Michigan and Lake Superior Water.
400
-------
100
GAMBLES CREEK
O INIT. S.S. CONC. = 42 ppm
INIT. S.S. CONC. = 90 ppm
LAKE MICHIGAN
o INIT. s.s. CONC.= 45 ppm
INIT. S.S. CONC.=85ppm
LAKE SUPERIOR
O INIT. S.S. CONC. = 40 ppm
INIT. S.S. CONC, = 85 ppm
TERTIARY EFFLUENT
OINIT. S.S. CONC.-40 ppm
INIT. S.S. tONC.= 80ppm
100 200 300 40010 100 200 300
SETTLING TIME,HOURS SETTLING TIME, HOURS
Figure 34. Effect of Dilution and Natural Water Source on Removal of Republic Fine
Tailings at 40° F.
400
-------
S 90
>
O
y so
__||__|1|__||||||__(__nn__1|||| imr^M^mM
LAKE MICHIGAN(D.S.= I46.6ppm)
O TERTIARY EFFLUENT(0,S.= 123.0 ppm)
D LAKE SUPERlOR(D.S.= 75.8ppm)
AGAMBLES
100 200 300
SETTLING TIME, HOURS
1/10 DILUTION
INITIAL S.S.CONC.»44ppm
(SAME NOTATIONS AS ABOVEf
P'
100 200 300
SETTLING TIME, HOURS
400
Figure 35. Effect of Various Diluent Waters on the
Removal of Republic Fine Tailings at 40°F.
95
-------
80% removal took place in the Empire tests. The same removal took 300 to 400
hours in the Republic tests.
The iso-percent removal plots for Empire fine tailings particles indicate predomin-
ately discrete settling especially at the 40°F temperature levels. Thus, it appears
that the relatively fast removals of Empire tailings over Republic tailings would not
be accounted for by increased coagulation and flocculation of particles of compar-
able size. However, the particle size distribution of Empire fine tailings indicated
a wider range of particle sizes than that for Republic fine tailings. Also/ the
specific gravity of the Empire tailings particles used in the analyses was greater
than that for the Republic tailings particles (i.e., 2.63 as compared to 2.00).
Figure 36 shows the effect of temperature and dilution on the removal of Empire
fine tailings in Lake Michigan and Lake Superior water. Comparison of this figure
with Figure 33 (note difference in time scales) shows that temperature exerted less
effect on removal of the Empire particles. Likewise, Figure 37 indicates that the
dissolved solids c'ontent of the diluent water exerted less effect on removal of the
Empire particles.
Influence of Alum Addition
Several additional experiments in which 2 ppm of alum was added were conducted
on both Republic and Empire fine tailings particles using both Lake Michigan and
Lake Superior water. This small alum concentration was used to simulate what
might happen when an effluent containing 10 to 20 ppm of alum was diluted by
factors of 5 or 10. The results showed that alum slightly increased the removals
for Republic fine tailings in both Lake Michigan and Lake Superior waters by about
0 to 10%. However, no consistent effect could be detected for the Empire fine
particles.
Field Study: Tailings Removal in the Republic Mine Tertiary Pond Effluent Stream
The purpose of this field study was to compare removal predicted from column data
with that occurring in the stream. The section of the Republic Mine tertiary effluent
stream below the tertiary pond outlet and above County Road FFW (see Figure 3)
was chosen for this study. The stream distance between these points is about
3,800 feet. The upper 2,000 feet consists of a channel roughly 10 feet wide with
a velocity of about 1 fps. The lower 1,800 feet consists of an area of impounded
water with a much lower velocity. Station 1 was located at County Road FFW,
Station 3 at the Tertiary Pond outlet and Station 2 was located 2,000 feet below
Station 3.
A dye dispersion test was performed to determine the mean residence time between
Stations 3, 2 and 1. At the same time replicate suspended solids analyses were
performed at these stations and samples were taken at Station 3 and subjected to
96
-------
\O
lOOi
70 °F'
-40°F
60
40
20
70°F^
7o°FJ:* 1)
-" "" ^ - ' teni^__
40°F
1/5 DILUTION
LAKE MICHIGAN
O INITIAL S.S. CONC. =85ppm
INITIAL S.S. CONC. =85ppm
1/10 DILUTION
LAKE MICHIGAN
O INITIAL S.S, CONC.= 45 ppm
INITIAL S.S. CONC.= 43 ppm
70'
40°F
1/5 DILUTION
LAKE SUPERIOR
O INITIAL S.S. CONC,= 85ppm
INITIAL S.S. CONC. = 90ppm
1/10 DILUTION
LAKE SUPERIOR
O INITIAL S.S. CONC.= 43ppm
INITIAL S.S. CONC.= 44ppm
0 8 16 24 0 8 16 24
SETTLING TIME, HOURS SETTLING TIME, HOURS
Figure 36. Effect of Temperature and Dilution on Removal of Empire Rne
Tailings in Lake Michigan and Lake Superior Water.
-------
100
o
kJ
O
90
o
H
o
o
ut __
o 80
z
ui
a
1/5 DILUTION
INITIAL S.S. CONC.»90 ppm
O SCHWEITZER CREEK (D.S. = 88.0 ppm)
LAKE MICHIGAN(D.S. = 170.2 ppm)
D LAKE SUPERIOR (D.S. = 66.8 ppm) _
8 16
SETTLING TIME, HOURS
100
o
u
O
vt
o
o
V)
o
u
o
z
Ul
a
cn
90
80
1/10 DILUTION
INITIAL S.S. CONC. « 45ppm
O SCHWEITZER CREEKtD.S. = SS.Oppm)
LAKE MICHI6AN(D.S.= 170.2 ppm)
D LAKE SUPERIOR(D.S. =66.8ppm)
70
8 16
SETTLING TIME, HOURS
Figure 37. Effect of Various Diluent Waters on the
Removal of Empire Fine Tailings at 40° F.
98
-------
quiescent settling column tests. The column depth of 27 inches approximated the
mean depth of the section under study. Suspended solids reductions were predicted
from the column data for a time equal to the mean hydraulic detention time of the
section. The results of this study, conducted at a flow rate of 12.8 cfs, are
shown in Table 20.
Table 20. Results Obtained
Station
Average Suspended Solids
Cone., ppm
Percent Suspended Solids
Removed in Stream
Dentention
Time, Mrs.
Predicted Susp. Solids %
Reduction
Average Velocity fps
Between Stations
from Field Study at Republic Mine
3 2
55.7 55.4
0 0
0 0.5
0
1.1 0.077
1
42.0
24.2
7.0
20.0
No removal was either obtained or predicted over the section between Stations 3 and
2 because the stream velocity was considerably in excess of the scour velocity for
the particles under consideration. The predicted removal over the section between
Station 2 and 1 agreed reasonably well with the actual removal.
Particle Size Distributions
Particle size distributions were determined for the Republic and Empire fine tailings
by two methods:
L. By applying Stokes1 Law to gravity settling data, and
2. By direct measurement of gross dimensions shown on photomicrographs.
According to Stokes1 Law:
Vc = d2 t Y - y ) 11-
(ys /w
99
-------
where: Vs = settling velocity, d = particle diameter, ju = viscosity, X s = weight
density of solids, 7 w = weight density of the fluid. The mass fraction of parti-
cles settling at velocities equal to or less than a given depth divided by a given
time was read directly from time settling or iso-percent removal curves e.g.
Figures 31 and 32. Then, for known values of u, r s and ^ w/ Stokes1 Law was
applied to calculate the diameter of particles settling at the given depth/time.
The specific gravity of the Republic and Empire fine tailings particles was deter-
mined by standard pycnometer techniques (18) as:
Republic fine tailings: SG = 2.00
Empire fine tailings: SG = 2.63
This method has a few inherent drawbacks when applied to flocculent suspensions
exhibiting curvilenear time -settling curves. The particle size distribution deter-
mined for these suspensions is affected somewhat by the particular point chosen for
analysis on a given time settling curve. In general, choosing points at greater
depths and times results in a larger particle size distribution. Likewise, the result-
ing particle size distribution may be influenced by the characteristics of the diluent
water.
The photomicrographs were taken using an American Optical Series 10 Phasestar
microscope equipped with a 35 mm camera. The particles shown on the photomicro-
graph were grouped into 8 size classes and between 15 and 70 particles were
counted in each size class.
Figure 38 summarizes the particle size distribution obtained. It can be seen that
the diluent water influenced the distribution obtained from gravitational settling
data. Generally, the optical method tended to estimate fewer numbers of very small
and very large particles. However, the mean particle sizes (50% smaller by weight),
determined by each method, agreed very closely for the Republic particles yielding
a mean diameter of 1.3 microns. For the Empire particles, the optical method
yielded a mean particle size of 6.5 microns whereas the settling data yielded 8.0
microns.
100
-------
O
E
CO
u
1_
0>
0.
A Photomicrographs
Diluent Water
Lake Michigan
O Lake Superior
D Schweitzer Creek
Average,Settling Data
Republic
Photomicrographs
Republic
Settling
Data
Empire
Photomicrographs
20
0.5 1.0 2.0 5.0 10
Particle Diameter, Microns
Figure 38. Partical Size Distributions for Fine Tailings.
-------
SECTION VII
ACKNOWLEDGEMENTS
The bulk of the field and laboratory work necessary for the completion of this pro-
ject was performed as part of advanced degree studies at Michigan Technological
University by graduate research assistants, Finn B. Christensen, Donald J. Greiner,
Timothy J. Me Clellan, James A. Visintainer and Donald P. Weaver, Jr. Dr.
Robert C. Polta, Assistant Professor of Civil Engineering, Michigan Technological
University, contributed to Section V.
Sincere appreciation is extended to Mr. Ralph Magnusen, Assistant to the Vice
President, Cleveland Cliffs Iron Co. for his support in instigating this project. The
overall coordination handled by Mr. Mel Viant, Chief Mining Engineer, C.C.I., is
gratefully acknowledged. Messrs. James Fegan, James Hanninen and Albert Nelson
of the Republic Mine and Mr. John Belling of the Empire Mine provided assistance
in gathering field data. Mr. Jack LaBelle, Senior Research Metallurgist, C.C.I.,
and Mr. John Meier, Hydrologist, C.C.I, supplied useful supplementing informa-
tion.
Sincere thanks are extended to Messrs. Thomas P. Evans, Leo Lucchesi and
Robert Olson of the Research Office at Michigan Tech for their coordinating efforts.
Appreciation is also extended to Dr. James D. Spain for the loan of several pieces
of research equipment. The advice and cooperation of E.P.A. Program Representa-
tive, Stephen Poloncsik, and Project Officer, Clifford R is ley Jr. are gratefully
acknowledged.
103
-------
SECTION VIII
REFERENCES
1. Wntala, S.W., Newport, T. G. and Skinner, E. L., "Water Resources of
the Marquette Iron Range Area, Michigan, "U.S. Geological Survey Water
Supply Paper, 1842, U.S. Department of the Interior, 1967.
2. "August 1971 Water Resources Summary Michigan11 U.S. Geological Survey,
Water Resources Division, Lansing District, U.S. Dept. of the Interior, 1971
3. Rich, L.G., Unit Operations of Sanitary Engineering, John Wiley and Sons,
Inc. New York, N.Y.,
4. Baillod, C.R., Alger, G. R., and Santeford, H.S., "Wastewater Resulting
from the Concentration of Low Grade Iron Ore," Water and Sewage Works,
^17:359, 1970.
5. Murphy, K.L., "Tracer Studies in Circular Sedimentation Basins," Proceed-
ings, Eighteenth Industrial Waste Conference, Purdue University Engineering
Extension Series No. 115, p. 374, 1963.
6. Villemonte, J.R. and Rohlich, G.A., "Hydraulic Characteristics of Circular
Sedimentation Basins," Proceedings, Seventeenth Industrial Waste Conference,
Purdue University Engineering Extension Series No. 112, p. 682, 1963.
7. Thirumurthi, D., "A Break-Through in the Tracer Studies of Sedimentation
Tanks," JWPCF, 41^ R405, 1969.
8. Rudd, D.F. and Watson, C.C., Strategy of Process Engineering , John Wiley
&Sons Inc., New York, 1968.
9. Fair, G.M., and Geyer, J.C., Water Supply and Waste- Water Disposal,
John Wiley & Sons Inc., New York, 1954, p. 637-639.
10. ASCE, AWWA, CSSE, Water Treatment Plant Design, AWWA, New York,
1969, p. 65-76.
11. Camp, T.R., "Floe Volume Concentration," JAWWA, 60j656, 1968.
12. Eckenfelder, W.W., Jr., Industrial Water Pollution Control, McGraw Hill
Book Co., New York, 1966, p. 22S-24V.
13 . Dick, R . I . , "Fundamental Aspects of Sedimentation/2 ," Water and Wastes
Engineering, March, 1969, p. 44.
105
-------
14. Smith, R., "Cost of Conventional and Advanced Treatment of Wastewater,"
JWPCF,40_:1546, 1968.
15. DiGregorio, D., "Cost of Wastewater Treatment Processes/' FWPCA Report
No.TWRC-6, 1968.
16. Perry, J.H., ed.7 Chemical Engineers' Handbook, 4th Edition, McGraw Hill
Book Co., New York, 1963, p. 20-20.
17. International Minerals & Chemical Corp., "Utilization of Phosphate Slimes,"
EPA Report 14050 EPU 08/71, 1971.
18. Boutilier, O.D., Bituminous Laboratory Manual, Michigan Technological
University, Houghton, Michigan, 1966.
19. Winneberger, J.H., Austin, J.H., and Klett, C.A., "Membrane Filter
Weight Determinations," JWPCF,_35; 807, 1963.
20. Standard Methods for the Examination of Water and Wastewater, 12th Edition
APHA, WPCF, AWWA, New York, 1965.
21. Wilson, J.F., Jr., "Fluorometric Procedures for Dye Tracing, "Techniques
of Water-Resources Investigations of the United States Geological Survey,
Chapter A12, Book 3, United States Government Printing Office, Washing-
ton, 1968.
22. Feuerstein, D.L., and Selleck, R.E., "Fluorescent Tracers for Dispersion
Measurements," Jour. San. Engr. Div., Proc. ASCE, 89 (No. SA4): 1,
1963. ~~
23. Levenspiel, 0., Chemical Reaction Engineering, John Wiley and Sons, Inc.,
New York, 1962";
24. Kramers, H., and Westerterp, K.R., Elements of Chemical Reactor Design
and Operation, New York, Academic Press, 1963.
25. Cohen, J.M., "Improved Jar Test Procedure," JAWWA, 49: 1425, 1957.
26. Baillod, C.R., and Christensen, F.B., "Hydraulic and Sedimentation
Efficiencies of Tailings Clarification Basins," Proceedings, 27th Industrial
Waste Conference, Purdue University Engineering Extention Series (in press)
1972.
27. Kawamura, S., and Hanna, G.P., "Coagulant Dosage Control by Colloid
Titration Technique," Proceedings, 21st Industrial Waste Conference, Purdue
University Engineering Extension Series, p. 381, 1966.
106
-------
SECTION IX
PUBLICATIONS
The following publications have resulted from this research project.
1. Christensen, F.B., Hydraulic and Sedimentation Efficiences of Tailings
Clarification Basins, Masters Thesis, Michigan Technological University.
Houghton, Michigan, 184 pages, 1971.
2. Gfefner, D. J., Vacuum Filtration Characteristics of a Slurry Resulting from
the Chemical Coagulation of Tailings Basin Overflows, Masters Thesis,
Michigan Technological University, Houghton, Michigan, 184 pages, 1972.
3. McClellan, T.J., Coagulation and Flocculation of Supra-Colloidol Particles
Contained in a Tailings Basin Overflow, Masters Thesis, Michigan Techno-
logical University, Houghton, Michigan, 237 pages, 1971.
4. ;yisintainer, J.A., Settling Characteristics of Fine Tailings Particles in
Natural Water Systems, Masters Thesis, Michigan Technological University,
Houghton, Michigan, 169 pages, 1971.
5. Weaver, D.P. Jr., Thickening Characteristics of Slurry Resulting from the
Coagulation of Tailings Basin Overflows, Masters Thesis, Michigan Techno-
logical University, Houghton, Michigan, 125 pages, 1972.
6. Baillod, C.R., and Christensen, F.B., "Hydraulic and Sedimentation Effi-
ciencies of Tailings Clarification Basins," Proceedings, 27th Industrial
Waste Conference, Purdue University Engineering Extension Series (in press)
1972.
7. Alger, G.R. and Baillod, C.R., "Mine Tailings Basins and their Associated
Watersheds," Proceedings, American Water Resources Assn. Symposium on
Watersheds in Transition, (in press), Fort Collins, Colorado, June, 1972.
107
-------
SECTION X
APPENDICES
Page
A, Water Balance and Pond Performance Data ~TTO
Table A.I. Empire Tailings Area, Water Balance Data 113
Table A.2. Republic Tailings Area, Water Balance Data 115
Figure A.I. Empire Tailings System, Outflow Hydrograph and 117
Mass Curves
Figure A.2. Republic Mine Tailings System, Hydrograph for 118
Direct Surface Outflow
Figure A.3. Republic Tailings System, Mass Curves 119
Figure A.4. Variation in Water and Air Temperature for the 120
Study Sites
Figure A.5. Republic Mine, Summary of Pond Performance Data 121
Figure A.6. Empire Mine, Pond Performance Data 122
B. Quiescent Settling Tests: Methodology 123
Table B.I. Summary of Quiescent Settling Test Conditions, 125
Republic Tailings
Table B.2. Mixing Conditions Used for Addition of Alum and 126
Polymer
C. Dye Tracer Methodology 127
D. Prediction of Effluent Concentration by Use of Settling and Dispersion 129
Curves
Figure D.I. Prediction of Effluent Concentration by Combining 130
Residence Time Distribution and Performance
Curves
E. Methodology Employed in Coagulent Evaluations 131
Figure E.I. Correlation Between Turbidity and Suspended Solids 132
Concentration for Republic Pond 3 Overflow
Table E.I Optimal Slow Mixing Conditions for Given 134
Jar Test Temperatures
F. Glossary of Terms
109
-------
APPENDIX A
Water Balance and Pond Performance Data
Storage Calculations for A Sp and A St
Republic
The cumulative change in water storage at Republic was determined by considering
the changes in water surface levels occurring in the various ponds throughout the
year. In 413 acres of the Pond 1 area, the elevations of both the tailings and
water surface increased by approximately 4.0 feet during the June 15, 1970 to
June 15, 1971 study period. It was determined that the average porosity of the
deposit was 46%. This storage was assumed to occur uniformly throughout the
period at a rate of:
4ft.(413ac)(0.46)
365 days (1.985 ac. ft.) '
sfd
An area of 49.7 acres located at the southern end of Pond 1 (originally part of
Pond 2) experienced a water level and tailings rise of 7.2 feet. The average
porosity of the tailings in this area was 507<>. This storage was assumed to occur
uniformly throughout the period at a rate of:
7.2ft. (49.7ac)(0.50)
365 days (1.985 ac. ft.) °'247 cfs
sfd
On April 3, 1971, the main tailings stream was diverted into the Northeast Pond
1 area. This area of 106 acres impounded the entire 58 cfs for a period of 4.35
days until an average water depth of 4.72 feet was attained. At this point water
overflowed into the Pond 1 area and near normal conditions were attained. However,
the main tailings stream continued to flow into this area of impounded water until
June 15, 1971. Thus, the solids were being deposited into water and causing a
displacement or negative storage of 1.41 cfs. The net storage in this northeast
Pond 1 area over the period of April 3 to June 15, 1971 was, therefore:
106 ac (4.72 ft.) . .. , ,-.0 . . ,_, ,,
a ft - 1-41 cfs (72 days)= 151 sfd
ac.ft.
sfd
110
-------
On July 26, 1970, a new set of outlet controls and culverts were put into service
at the 3B location. This caused the water surface of Pond 2 to rise. It was esti-
mated that this accounted for 60 sfd of storage over the period of July 26-27.
The water surface elevation of the Return Flow Pond was measured and recorded
daily. For Return Flow Pond surface elevations below 1507.15, the water surface
elevation of Pond 3 was held constant and by the overflow structure. For higher
elevations/ the changes in surface elevations of Pond 3 were equal to those of the
Return Flow Ffond.
The total cumulative change in water storage shown in Column 8 of Table A.2 was
determined by adding the cumulative changes calculated for Pond 1, 2, 3 and the
Return Flow Pond.
Empire
The cumulative change in water storage at Empire was determined by considering
both the changes in water surface elevation of the pond area and the amount of water
displaced by the advancing tailings front. The water surface elevation was recorded
continuously at the West outlet and the changes in storage caused by changes in the
water surface elevation were calculated by multiplying the change in elevation by the
average 406 acres of water surface.
The annual tailings accumulation amounted to approximately 2270 acre ft. Field
surveys showed that about 80% of this volume was deposited into water as the tail-
ings front advanced. The average porosity of the tailings deposit was determined
to be 38%. Thus, the rate at which water was being displaced was calculated as;
0.8(2270)ac.ft. (0.62) _ _ ,
1.985ac.ft. (365 days)
sfd
This was considered as a negative storage.
Because the pond water surface rose by nearly 2 feet during the study period, most
of the remaining 20% (which was not deposited directly into the water) resided at a
position below that of the final water surface. Thus, the rate at which water was
retained in this fraction was estimated as:
0.2 (2270)ac. ft. (0.38) = Q 24 rfs
1.985ac.ft. 365 days
-Ira-
-------
Thus, the net uniform storage was
0.24 cfs - 1.55 cfs = - 1.31 cfs
or, 1.31 cfs was continuously released by displacement.
The total cumulative change in water storage shown in column 6 of Table A.I was
determined by subtracting the cumulative amount of water released by displacement
from the cumulative volume added by the rise in the water surface.
Seepage Estimate
Republic
Applying Equation 7 over the period of June 15 to October 31, 1970 gives
L = (l+P-E)-(ASp + ASf + Q)
t t
J^= 8257 sfd- 7454 sfd = 5.77 cfs
t 139 days
The seepage over the pond of June 15, 1970 to March 31, 1971 is
L = 5.77 cfs (290 days) = 1672 sfd
Since the total seepage was 2463 sfd, the seepage rate over the period of April 1
to June 28, 1971 is given by
2463 sfd- 1672 sfd 0 rt ,
=8.9 cfs
Empire
Applying Equation 7 over the period of July 8, 1970 to June 28, 1971 gives
JL = 4715 sfdI - 1962 sfd
t 355 days = 7-75
112
-------
Table A.I. Empire Tailings Area - Water Balance Data
NET INPUT sfd
(0) (1) <2T (3) (4)
Date Plant Precip. Evap. Cumul.
Inflow Total
IP E
7/8/70
7/10
7/20
7/31
8/10
8/20
8/31
9/10
9/20
9/30
10/10
10/20
10/31
11/10
11/20
11/30
12/10
12/20
12/3L
1/10/71
1/20
1/31
2/10
2/20
2/28
3/10
3/20
3/31
4/10
4/20
4/30
5/10
5/20
5/31
6/10
6/20
6/28
218 356
324 81
.t ' '
306 470
303 495
289 230
299 150
305 306
294 218
303 177
270 121
327 217
292 353
338 236
137 504
223 1057
137 1718
95 2142
43 2548
0 3159
86 3585
178 3877
148 4130
240 4434
364 4715
H3
OUTFLOW sfd
(5)
Cumul.
Total
IQ
8
24
89
137
174
220
290
374
447
517
562
598
660
702
733
777
828
868
908
950
977
1022
1062
1103
1135
1169
1205
1245
1308
1510
1645
1755
1848
1959
2046
2125
2182
(6)
Cumulative Change In
Water Storage
I(AS + AS )
0
-11
-50
-78
-91
-103
-68
-94
-107
-95
-30
-31
+2
-11
-3
-16
-38
-60
-63
-103
-54
-107
-121
-134
-145
-159
-172
-220
-46
-36
-60
-122
-116
-147
-177
-159
-220
-------
Table A.I. Empire Tailings Area - Water Balance Data (Continued)
(0)
Date
7/8/70
7/10
7/20
7/31
8/10
8/20
8/31
9/10
9/20
9/30
10/10
10/20
10/31
11/10
11/20
11/30
12/10
12/20
12/31
1/10/71
1/20
1/31
2/10
2/20
2/28
3/10
3/20
3/31
4/10
4/20
4/30
5/10
5/20
5/31
6/10
6/20
6/28
(7)
Cumul. Tot.
Outflow +
Storage sfd
(5) + (6)
2(Q+ASp+ASt)
13
39
59
83
117
222
280
340
'422
532
567
662
691
730
761
790
808
845
847
923
915
941
969
990
1010
1033
1025
1262
1474
1585
1633
1732
1812
1869
1966
1962
(8)
Cumul.
Total
Seepage
sfd
2L
,
186
426
658
899
1130
1372
1612
1828
2070
2300
2540
2753
(9)
Cumul. Tot.
Input Less
Seepage
sfd
(4) - (8)
S(lfP-E-L)
50
78
399
819
1012
1176
1547
1757
1817
1830
1894
1962
(10)
Unaccounted
Storage
sfd
I(ASs+ASj
*
-9
-144
-23
+157
+251
+331
+632
+767
+7^2
+245
+82
0
(11)
Snow Storage
on Watershed
& Tailings
sfd
2(ASS)
0
0
0
0
68
212
466
!
480
480
0
0
0
114
-------
(0)
Table A.2. Republic Tailings Area - Water Balance Data
-NET INPUT, sfd
^""^"^^^^*"
Dates
6/15/70
6/20
6/30
7/10
7/20
7/31
8/10
8/20
8/31
9/10
9/20
9/30
10/10
10/20
10/31
11/10
11/20
11/30
12/10
12/20
12/31
1/10/71
1/20
1/31
2/10
2/20
2/28
3/10
3/20
3/31
4/10
4/20
4/30
5/10
5/20
5/31
6/10
6/20
6/28
(1)
Plant
m
Inflow
1
^^ ^^ *
926
1800
1800
1740
1800
1740
1800
1800
1625
1800
1740
1800
1625
(2)
Precip.
i
P
31
212
37
216
228
106
69
141
100
80
56
100
151
(3)
Evap.
E
65
201
102
102
63
43
20
0
39
80
87
127
163
(4)
Cumul.
Total
Z(H-P-E)
892
2703
4438
6292
8257
10060
11909
13850
15536
17336
19045
20818
22431
V/UIV
(5)
5-B
Outlet
1.6
8.4
16.4
78.9
186.0
307.4
493.5
633.1
759.6
938.6
1102.9
1316.2
1445.3
1589.7
1833.6
2022.5
2105.6
2201.5
2340.6
2457.7
2496.3
2569.9
2646.1
2708.6
2711.6
2737.2
2859.8
3067.3
3264.5
3542.9
3759.3
4086.8
4354.2
4600.9
4726.0
4909.0
5105.9
5296.0
5441.2
I(J1_/-\I IV C. UU
(6)
% » /
Return
Flow
50.1
291.3
777.3
1141.5
1521.6
1958.1
2330.4
2739.9
3165.9
3530.7
3916.2
4271.7
4636.2
4987.5
5344.2
5710.8
6195.0
6570.9
6956.1
7372.8
7856.4
8244.9
8648.7
9039.3
9497.4
9856.8
10140.0
10417.2
10748.4
11148.3
11495.5
11794.5
12003.0
12217.2
12574.2
12888.0
13177.8
13491.9
13779.0
1 1 ru i , sra
(7)
\ * /
Total
Output
2Q
51.7
299.7
793.7
1220.4
1707.6
2265.5
2823.9
3373.0
3925.5
4469.3
5019.1
5587.9
6081.5
6577.2
7177.8
7733.3
8300.6
8772.4
9296.7
9830.5
10352.7
10814.8
11294.8
11747.9
12209,0
12594.0
12999..8
13484.5
14012.9
14691.2
15254.8
15881.3
16357.2
16818.1
17300.2
17797.0
18383.7
18787.9
19220.2
115
-------
Table
A. 2. Republic Tailings Area - Water Balance Data (Continued)
CUMULATIVE, sfd
(0)
Dates
6/15/70
6/20
6/30
7/10
7/20
7/31
8/10
8/20
8/31
9/10
9/20
9/30
10/10
10/20
10/31
11/10
11/20
11/30
12/10
12/20
12/31
1/10/71
1/20
1/31
2/10
2/20
2/28
3/10
3/20
3/31
4/10
4/20
4/30
5/10
5/20
5/31
6/10
6/20
6/28
(8)
Change
In Water
Storage
2(ASD+
AStP)
0
12.1
1.7
71.8
95.3
164.8
140.9
132.1
155.1
182.4
186.1
204.7
227 .7
220.1
276.1
270.4
255 A
304.6
302.0
315.0
327.5
347.3
334.7
334.7
329.0
447.3
434.0
392.6
427.0
422.2
742.4
712.8
705.5
695.2
730.4
729.2
738.8
742.4
747.7
(9)
Total
Output
+Storage
S(Q+ASD+
ASt) P
51.7
311.8
795.4
1292.2
1802.9
2430.3
2964.8
3505.1
4080.6
4651.7
5202.2
5792.6
6309.2
6797.3
7453.9
8003.7
8556.0
9077.0
9598.7
10145.5
10680.2
11162.1
11629.5
12082.6
12538.0
13041.3
13433.8
13877.1
14439.9
15113.4
15997.2
16594.1
17062.7
17513.3
18030.6
18526.2
19022.5
19530.3
19967.9
(10)
Total
Seepage
ZL
(11)
Total
Input
-Seepage
(4H10)
(12)
Unaccounted
Storage
(11H9)
sfd
SU+P-E-L) 2(ASS+
92
271
450
623
803
975
1155
1335
1495
1675
1942
2217
2463
800
2432
3988
5669
7454
9085
10754
12515
14041
15661
17103
18601
19968
AS.+ASJ
1 y
+ 4.6
+ 1.7
- 93
-124
0
+ 8
+ 74
+432
+607
+548
+ 41
+ 75
0
(13)
Snow
Storage
sfd
Z(ASS)
0
0
0
0
0
38
81
86
73
60
0
0
0
116
-------
Figure
JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN
A.I. Empire Tailings System,Outflow Hydrograph and Mass Curves,
117
-------
DIRECT SURFACE C
JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN
JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN
- 40
JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN
1970 1971
Figure A.2. Republic Mine Tailings System, Hydrographs for Direct Surface
Outflow.
118
-------
220
JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN
1970 1971
Figure A/3. Republic Tailings System, Mass Curves.
119
-------
ou
70
60
50
30
Q/">
ou
70
~cr\
TEMPERATURE, DEGREES F.
_L _roco-^cnc7>-joo GJ £» in c
oooooooooo oooc
x
-.';
,*
/v
;
;
* /
"-F*^
\
^.
Ss
.
i * /\
v\
'
TEN DAY
s ,
TEN DAY
*
?\:
^s.
.
TFN DAY
MEAN--'
MfV^
-
"
rv. ' «.
.
^FAN-^-
FV
S^'
RFF
\,'
V
.. ..... ,
PIRE V
"
UB1 1C
^^^
AIR TFI
.
' V\-
-.
/ATER '
.HI SJT^
^A/ATFR
?
/IPFRA-
*
'Ns
FEMPE
TFMF
-,; ;
. .... .
!'
JRF
; '
-
' f
/
RATUR
"*I~^
FRATUI
;-!-.
- -*.., , -...«,
r i.:
*
-------
I
0
o
u
in
9
I
u
O
U
ui
a
z
UJ
a
120
I
u
o
u
in
o
o
in
in
u
i
O
z
z
o
a
0
u
QC
a
TERTIARY
MEASURED AT
l
12200 -
8200 r
4200 -<- -
200
2.0
1.0
I . ^^^^^^^^^^£J^m|^^w^^^U^^^^BMiHM«H^Uri»«B^^»«*M«^l^H>^i^i^^B«i«^B^>"l*"l^^ln^"«MWH«IMI^l^l«M^^M^MMMl
JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN
Figure A.5. Republic Mine,Summary of Pond Performance Data.
121
-------
60
E
o
u
CO
o
o
CO
o
UJ
o
UJ
O.
CO
CD
CO
(0
«*»
o_
uT
£
OL
$
O
_l
u.
I-
o
10
JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN
1970 1971
Rgure A.6, Empire Mine Pond Performance Data.
122
-------
APPENDIX B
Quiescent Settling Tests: Methodology
Samples of the tailings waste used for quiescent settling tests were collected in
5 gallon polyethylene containers from the main waste flow immediately after leaving
the mills. The following procedure was used to investigate the settling character-
istics of the tailings material.
(1) A representative sample of the main tailings waste was
stored in the collecting container until the suspension
reached its test temperature.
(2) After being thoroughly mixed, the sample was presettled
for various time periods depending upon the desired initial
suspended solids concentration.
(3) The supernatant from the presettled suspension was poured
into a 4 gallon bucket and mixed before being introduced
into the column.
(4) Compressed air was then supplied through the valved outlet
in the bottom of the column to create a uniform suspended
solids concentration over the entire length of the column.
(5) Immediately after the application of compressed air a hori-
zontal screen was dragged slowly upward from the bottom to
maintain uniform particle distribution and to damp out any
turbulence and recirculation eddies.
(6) Settling time was started when the screen reached the top.
(7) Immediately after starting the settling time a sample was
withdrawn from the top, middle and bottom sample port to
bring the surface level to a predetermined distance from the
bottom. The collected samples were used to determine
initial suspended solids concentration, dissolved solids
concentration, pH-value and to establish the uniformity of
the suspended solids concentration.
(8) The suspension was then allowed to settle under quiescent
conditions.
123
-------
(9) At each time interval, column samples (60 ml) were extracted
by use of a syringe and a specially designed sampling port
permitting the extraction of a representative sample from an
undisturbed cross-sectional area at different depths with
negligible leakage.
(10) The suspended solids concentration of each sample was
determined using the membrane filtration technique of
Winnebergeret. al. (19). A Mettler, H-20, analytical
balance was employed in the determination of the suspended
solids concentrations. The dissolved solids concentrations
were determined in accordance with the procedure outlined
in Standard Methods (20). A Corning, Model 12, research
pH meter was used to measure the pH-values of the sample
suspension.
A list of the quiescent settling tests performed is given in Table B.I. The test
temperatures were maintained at * 1°F by placing the settling column with its
suspension in^a closed room where near constant temperature conditions were main-
tained. The given temperatures were chosen because the clarification ponds at
Republic Mine was found to have a maximum temperature of approximately 75°F
during the summer and the fact that water reaches its maximum density at 39.2°F.
To determine the effect of addition of a coagulant material and in order to predict
the performance efficiency of the Tertiary Pond, certain settling tests were performed
with the addition of 10 to 25 ppm Alum and 2 ppm Polymer (Nalco 603-Cationic).
The above dosages are typical of those used at Republic Mine to treat the discharge
from Pond 3. When coagulants were added, the following column test procedure was
followed.
(1) After filling the columns with a representative sample of fine
tailings from Pond 3 overflow the suspension was mixed by
compressed air and the surface was brought to a predetermined
level by withdrawing samples from top, middle, and bottom
sampling port. The samples were analyzed for initial suspended
solids concentration, dissolved concentration, and pH-value.
(2) A predetermined flow of air was then supplied through the valved
outlet at bottom of column and the exact dosage of alum was
injected by a syringe through the bottom sampling port.
(3) The suspension was mixed rapidly and then slowly by com-
pressed air for a certain period of time after which a hori-
zontal screen was used to damp out any turbulence and the
settling time started.
124
-------
Ol
Table B.I. Summary of
Run
No.
1
2
3
4
5
6
7
8
Column
No.
1
2
3
4
1
3
3
1
3
1
3
1
2
3
4
1
3
4
Presettl ing
Time
2hrs.
2 hrs.
15 min.
15 min.
15 min.
15 min.
20 min.
20 min.
20 min.
20 min.
0
0
0
0
0
0
0
Quiescent Settling Test Conditions -
Alum
mg/l
0
10
0
0
0
0
0
0
0
0
0
0
25
0
25
25
25
25
Polymer
mg/l
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
2
Initial
Fluid Depth
Inches
72
72
69
69
71
71
71
71
71
71
71
71
71
70.5
70
71
71
70
Republic Tailings
initial
s.s.
mg/l
3,450
3,450
7,733
7,957
8,195
8,014
7,298
11,480
9,729
12,674
12,447
425
303
400
402
293
265
451
Dissolved
Solids
mg/l
150
150
_-
364
134
134
130
130
114
114
114
114
120
117
120
Average
Fluid Temp.
'°F
75
75
41
41
39
72.5
72
39
74
39
72
41
41.5
73
73
41
75
75
PH
7.8
7.8
..
7.1
7.1
7.1
7.1
7.2
7.2
7.5
-------
(4) If alum and polymer were added/ the polymer was added
with a syringe through the bottom sampling port after a
short period of rapid mixing with alum.
(5) The suspension was continuously rapid mixed during injection
of the polymer and for a certain time period thereafter.
(6) After a period of slow mixing the turbulence was damped and
the settling time started.
A list of the mixing times and the amount of air used for the various tests are given
in Table B.2. The data in Table B.2. are based on equations for mean velocity
gradient as given by Fair and Geyer (9).
Table B.2. Mixing Conditions Used for Addition of Alum and Polymer
Rapid Mixing Slow Mixing
Time
Coagulant
Alum
Alum and
Polymer
Test
Temp.
Time
Velocity
Gradient
Air
Flow
75
40
75
40
min. ft/sec/ft chrymin min.
7
7
4+3
4+3
153
113
153
113
943
1035
943
1035
17
17
17
17
Velocity
Gradient
ft/sec/ft
47
34.9
47
34.9
Air
Flow
cm^/min
106
98.5
106
98.5
Settling Columns. The quiescent settling tests were performed in 411 I.D. columns
made of 1/4" thick transparent acrylic resin tubing approximately 74" long and
mounted on 1/2" thick acrylic resin base plates. Sample ports were located at 12"
intervals.
The average depth, defined as Volume/Area, of the Republic Mine clarification
basins was determined to be approximately 57". Thus a total column length of 74"
with an effective settling depth of maximum 66" was used.
Effective Settling Distance. The water surface drop caused by extracting samples
from the suspension at different time intervals changed the settling distance for
each sampling run. This decrease in the settling distance was accounted for when
making iso-concentration plots in terms of depth and time. A sampling volume of 60
ml was used for all tests. This resulted in a surface level drop of approximately 5/16
inch for each sample extracted.
126
-------
APPENDIX C
Dye Tracer Methodology
u 'r'S P°nd 3 and at the Empire Pond the tpacer was injected into a channel
which had formed near the tailings front. The injection was accomplished by using
a 5 gallon polyethylene container with a 3 inch valved pipe connected to the con-
tainer mouth. The tracer was injected in less than 10 seconds at a depth equal to
60% of the total depth.
Continuous sampling and fluorescence measurement were employed for the relatively
short term (less than 70 hours) tests conducted on the Tertiary Pond. However,
for the larger ponds, the tests covered periods ranging from 250 hours to 350 hours.
Since it was impractical to maintain the portable alternator and other equipment
associated with the continuous sampling and flourescence measurement over this
time period, individual samples were collected and analyzed in the laboratory.
Of the many different fluorescent dyes available, rhodamine B and rhodamine WT
were chosen for this study. Rhodamine WT was chosen because of its relatively
low adsorption on the tailings fine particles and its high fluorescent strength.
Rhodamine B was chosen because of its low cost and high fluorescent strength. It
has been established that rhodamine B exhibits a moderate sorptive tendency (21).
Rhodamine B was used to determine the hydraulic characteristics of both the Tertiary
Pond and Empire Pond which had relatively low concentrations of suspended solids.
Rhodamine WT was used for Pond 3 which had a high suspended solids concentra-
tion throughout its entire volume.
Rhodamine WT is a relatively new fluorescent tracer and little information is avail-
able in the literature on its use. The manufacturer ( du Pont) describes rhodamine
WT as a dark red 20 percent solution with a specific gravity of 1.19 + 0.02 at
20°C and a low tendency to stain silt, dirt and other suspended mattenn shallow
and inland waters. The effect of temperature on the fluorescent intensity of rhodamine
WT has been established by Wilson (21). To eliminate the effect of the temperature
all rhodamine WT analyses were performed at the same temperature. No information
on the photochemical decay rate of rhodamine WT could be found, but it is believed
to be similar to that of rhodamine B (21). The effect of photochemical decay was
negligible since little light penetrated the Pond 3 suspension because of the high
solids concentration. Because the specific gravity of the 20% solution or rhodamine
WT (1.19) differed from that of water, methyl alcohol was added to the dye solution
to adjust the specific gravity to approximately 1.00. This promoted effective mix-
ing between the water and tracer so that the tracer dispersed uniformly throughout
the cross section of the influent stream before entering the pond.
127
-------
Rhodamine B is available in powder and solution form. The powder form was used
for this study. The behavior of rhodamine B as a tracer material has been investi-
gated by Feuerstein et. al. (22). They found the fluorescent intensity of rhodamine
B to vary markedly with sample temperature, but the photochemical decay rate was
found to be low. Temperature correction curves constructed by Wilson (21) were
employed in determining the actual concentrations of rhodamine B.
The determination of fluorescent intensities was made with a G.K,Turner Asso-
ciates Model 111 Fluorometer. A far UV lamp was used as light source. Filter
546 was used as a primary filter and filter 590 was used as a secondary filter for
the analyses of both rhodamine B and WT. For high fluorescent intensities a 10
percent neutral-density (ND) filter was placed over the secondary filter. The fluoro-
meter was equipped with a high volume continuous-flow-through door. A sample
inlet device consisting of a glass funnel and 1/4" tygon tubing was employed to
measure fluorescent intensities of individual samples under conditons of constant
temperature. Before the fluorescent intensities in the samples from Pond 3 could be
analyzed/ the samples were filtered using Millipore Filters (0.45 micron pore size).
After adjusting the filtrate to the standard test temperature, 72°F, the fluorescent
intensity was determined by continuously adding the filtrate to the sample inlet
device. In this way, the sample in the flow cell was continuously replaced to
prevent any increase in the sample temperature caused by heat from the light
source. All fluorescence measurements were corrected for natural background
fluorescence.
128
-------
APPENDIX D
Prediction of Effluent Concentration by Use of Settling and Dispersion Curves
The method employed to predict basin performance by combining residence time dis-
tribution and settling curves is based on theory described by Levenspiel (23)/
Kramers et. al. (24), and Baillod and Christensen (26). In this analysis, vertical
mixing is neglected and fluid elements are assumed to travel through the basin as
small, intact columns.
The dye dispersion curve consists of a plot of c versus t. The residence time dis-
tribution curve consists of a plot of relative tracer concentration, E = C/C0/ versus
relative time, T - t/0 . The time 0 is determined from the dis-
persion curve by
and C is also determined from the dispersion curve by
o 0
The performance curve for an ideal plug flow basin consists of a plot of Cco| versus
T . CCo| is. determined from quiescent settling data for various values of
t. Finally, the predicted effluent concentration is calculated by graphical integra-
tion as shown in Figure D.I.
Predicted /°°
Effluent = I C . EdT
« * A- o COI
Concentration
129
-------
ccol.
Residence Time
Distribution
Relative Time
Performance Curve
Relative Time
Graphical Integration
Area= Effl. Cone.
Relative Time
Figure D.I. Prediction of Effluent Concentration by Combining
Residence Time Distribution and Performance Curves.
130
-------
APPENDIX E
Methodology Employed in Coagulent Evaluations
Equipment, Materials, and Analyses. Equipment used, and analyses performed
throughout the research included: a Corning Model 12 Research pH Meter, used in
sample pH and total alkalinity measurement, a Hach Model 1860-A Turbidmeter,
used to measure sample turbidity, Phipps-Bird multiple stirring machine, used in
'jar testing' with the various coagulants and coagulant aids; and standard Imhoff
cones, used to measure volumes of sludge settled in 30 minutes (following coagu-
lation-flocculation). The alum used in the research was a reagent grade of hydrated
aluminum sulfate. Anionic, cationic, and non-ionic polymer coagulants or coagulant
aids were donated by the Nalco Chemical Company, the Calgon Corporation, and
the Dow Chemical Company.
Sample Collection and Preparation. Samples that were used in the laboratory coa-
gulation-flocculation tests were collected at the 5-B outlet, located immediately
upstream of the alum addition point. Figure E.I. illustrates the relationship be-
tween sample turbidity and suspended solids concentration.
Coagulant Preparation. Standard solutions of polymer coagulants were prepared
according to manufacturers1 instructions. Alum and polymers were added as solutions
of 1 and 0.1 percent, respectively, to the several water samples tested in a series
of j'ar tests.
Standard Testing Procedures. The jar test was selected as the primary method for
the evaluation of the coagulation-flocculation-quiescent settling process. The pro-
cedure was a combination of several practices, largely being patterned after Cohen
(25). Raw sample pH levels were adjusted to a predetermined initial pH value by
the addition of hydrochloric acid (HCI) or sodium hydroxide (NaOH), and the samples
were placed into 1 liter beakers. The beakers were then placed in a plexiglass
water bath, and ice or tap water was added to the bath. Circulation of the tank con-
tents by air-induced mixing, and further additions of ice or tap water aided in main-
taining water bath and sample temperatures at a specified temperature. Temperatures
of 1, 13, and 25° Centigrade were used in laboratory jar testing procedures. At no
time did measured temperatures deviate from specified levels by more than + 0.5°C.
Coagulants were added to the series of beakers simultaneously, by the use of test
tubes mounted in a plastic bar. These were turned upside down, with the tip protrud-
ing just underneath the water surface as the coagulant was added to the several
beakers. Alum was added at the commencement of rapid mixing, whereas polymers
were added 4 minutes after the start of rapid mixing.
131
-------
g
CD
tr
15
H
UJ
_J
CL
1800
1600
1400
1200
1000
800
600
§00
200
7
0 200 400 600 800 1000
SAMPLE SUSPENDED SOLIDS, ppm
Figure E.I. Correlation between Turbidity and Suspended
Solids Concencentration for Republic Pond 3 Overflow.
132
-------
Supernatant turbidity "samples were simultaneously withdrawn by the use of vacuum
lines. The withdrawal point was located at mid-depth of a beaker. Unless other-
wise specified, a standard quiescent settling time of 30 minutes elapsed between
the termination of slow mixing and the measurement of residual turbidity.
Residual pH and residual alkalinity values of the supernatant samples were usually
measured, after quiescent settling. Following quiescent settling, beaker contents
were stirred to homogenity with a minimum amount of hand stirring. Beaker contents
were then gently poured into Imhoff cones, where they were allowed to settle for a
period of 30 minutes. Comparisons of sludge volumes formed by various coagulants
were possible with this procedure.
Rapid Mixing Operations . In order to facilitate comparisons between coagulation
process efficiencies predicted in the laboratory and those attained in"the field, rapid
mixing conditions*were selected to approximate field conditions at the Republic Mine.
A rapid mixing speed of 90 rpm (machine maximum = 100 rpm) was selected. It
was originally estimated that the effective rapid mix time obtained under field con-
ditions was about 7 minutes. Residence time studies conducted later indicated
that the actual rapid mix time was on the order of 4 minutes. Nevertheless, 7 min-
utes was employed for rapid mixing in all experiments. Thus, the standard rapid
mix conditions were 7 minutes at 90 rpm. At 13°F, this resulted in a temporal
mean velocity gradient, G = 135 fps/ft, and GT = 5.67 x 104. The value of G
was somewhat less than that normally employed in rapid mix design whereas the
value of GT was somewhat greater than that normally employed (10).
Slow Mixing Conditions. In order to determine the optimal slow mixing conditions
for a given temperature, a series of experiments was conducted to evaluate the
effects of various levels of slow mixing and temperature upon turbidity removals at
an initial pH of 7.0 and 50 ppm of alum. Slow mixing speeds of 30, 40 and 50
rpm and temperatures of 1DC,13°C and.25°C were employed.. The values of the
temporal mean velocity gradient, G, were determined using data published by Camp
(11). Mixing time and temperature were varied to obtain values of GT ranging from
104to 105.
The results of these experiments indicated that turbidity removal efficiency, at any
slow mixing GT level, was strongly dependent on the coagulation-f Peculation
temperature and somewhat less dependent on the slow mixing speed. The effect
of mixing speed was more significant at the lower temperature. Optimal values of
mixing speed and mixing time determined for each temperature are shown in Table E.I.
Unless specified otherwise, the optimal slow mixing conditions for a given tempera-
ture were employed in all other jar tests.
133
-------
Table E.I. Optimal Slow Mixing Conditions for Given Jar Test Temperatures
Temperature Optimal Mixing Optimal Mixing Optimal GT
C° Speed/rpm Time/ min. v
1 40 17.4 4.17xl04
13 40 18.6 5.31xl04
25 30 21.2 4.83x104
Initial pH. A series of experiments was conducted to determine the influence of
initial pH on turbidity removal. Initial pH levels of samples were established within
approximately the same range of raw pH values found in the Republic Mine tailings
basin overflow over a period of one year- this range being from 7.3 to 8.1. Tests
were conducted using alum dosages of 50 and 100 ppm at temperatures of 1 and
13°C. The results clearly indicated an optimal initial pH close to 7.5. Unless
specified otherwise/ this pH value was employed for all subsequent experiments.
Particle Electrokinetic Properties. At the outset of the study, it was intended to
investigate particle electrokinetic properties using a Waters Associates Streaming
Current Detector. However, Waters Associates had temporarily discontinued the
manufacture of this instrument. Consequently it was not available and not employed
in this study. Subsequently an attempt was made to investigate particle charge
phenomena using the colloidal titration technique (27). However, this effort was
unsuccessful as the necessary reagents could not be obtained from the Japanese
manufacturer.
134
-------
APPENDIX F
Glossary of Terms
acre ft.:
argillaceous:
autogenous grinding:
ore beneficial!on:
second-foot-day (sfd):
snowpack:
seepage:
snowpack water equivalent:
supra-colloidal solids:
43,560ft
clayey
Process in which grinding is achieved in a
tumble mill through Interparticle contact
between the particles being ground.
Process by which the economic value of
an ore is enhanced through concentration
86,400ft3 = 1.983 acre. ft.
layer of snow covering the land surface
Subsurface movement of water.
Height of a column of water (covering a
given area) with a weight equal to the
weight of the snowpack covering the same
area.
Suspended solids consisting of particles
with diameters ranging from 0.5 microns
to 10 microns.
«U.S. GOVERNMENT PRINTING OFFICE: 1974 546-319/413 1-3
135
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
}, Report No.
3. Accession No.
w
4. Title
Storage and Disposal of Iron Ore
Processing Wastewater
7. Aathor(s)
Baillod, C. R. and Alger, G. R.
5. Report Date
8. P..ftotnii^g Organization
RepottNo,
9. Organization
Michigan Technological University, Houghton, Michigan
Department of Civil Engineering
10. Project No.
1404QFVD
11. Contract/Grant No.
1404OFVD
t2.', Sponsoring Organisation
U, Type .dRepoftand
Period Coveted
is. supplementary Notes U. S. Environmental Protection Agency, Research and Development
Report Number EPA 660/2-74-018, March, 1974.
is. Abstract Thjs study was concernecj wjth the handling of wastewater resulting from the con-
centration of low grade iron ore. Specific objectives were to develop tailings basin design
and management criteria, to investigate methods of treating basin effluents, and to investigate
transport characteristics of participate materials contained in basin discharges. These objectives
were accomplished through laboratory and field studies conducted at the tailings impoundment
systems associated with two iron ore concentrating plants located in Upper Michigan. .Annual
water balances were formulated for the tailings system at each plant to show the relative
importance of precipitation, surface outflow and seepage. Settling column and dye dispersion
tests were employed to predict concentrations of suspended material remaining in basin effluents.
The cost effectiveness of various coagulation systems for treating basin effluent was investigated.
Particular attention was devoted to a study of the thickening and dewatering characteristics of
the slurry produced by chemical coagulation of the basin effluent. Synthesis of alternative
systems for handling tailings wastewater was hampered by a lack of information on water
quality requirements for reuse within ore concentrating processes. Finally, settling column
experiments in which fine tailings particles were diluted with various natural waters were
conducted.
na. Descriptors |\/|jne Wastes*, Sedimentation*, Hydrology*, Coagulation, Industrial Wastes,
Sediment Transport, Impoundments, Michigan, Lake Superior
i7b. identifiers 7aj|jngs Basins*, Iron Ore Beneficiation*, Taconite Waste*, Thickening,
Vacuum Filtration.
17c. COWRR Field & Group 02 B, 02J, 04A, 05 D
18. Availability
J 9. ',.. Security Class.
- (Repor?)
M. Security Cl,,ss.
(Page)
Abstractor C. Robert Baillod
21, No.oi
Pages
72. Pne*
Send To:'
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 2O24O
institution Michigan Technological University
WRSIC IO2 (REV. JUNE 1371}
------- |