The Pennsylvania State University The Graduate School Civil and Environmental Engineering PROMOTING LOW-PH BIOLOGICAL IRON(II) OXIDATION OF ACID MINE DRAINAGE AT HUGHES BOREHOLE: FIELD AND LABORATORY EXPERIMENTS A Thesis in Environmental Engineering by Trinh Cong DeSa © 2009 Trinh Cong DeSa Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Science August 2009
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Pennsylvania State University
The Graduate School
Civil and Environmental Engineering
PROMOTING LOW-PH BIOLOGICAL IRON(II) OXIDATION OF ACID MINE
DRAINAGE AT HUGHES BOREHOLE: FIELD AND LABORATORY
EXPERIMENTS
A Thesis in
Environmental Engineering
by
Trinh Cong DeSa
© 2009 Trinh Cong DeSa
Submitted in Partial Fulfillment of the Requirements
for the Degree of
Master of Science
August 2009
ii
The thesis of Trinh Cong DeSa was reviewed and approved* by the following:
William D. Burgos Professor of Environmental Engineering Thesis Advisor Graduate Program Chair of Environmental Engineering
Rachel A. Brennan Assistant Professor of Environmental Engineering
Michael N. Gooseff Assistant Professor of Civil Engineering
*Signatures are on file in the Graduate School
iii
ABSTRACT
Fe(II) oxidation at a low-pH acid mine drainage (AMD) site in Pennsylvania was
enhanced by physical modifications to the existing iron mound. Hughes Borehole
discharges approximately 1000 gallons per minute of pH 4 AMD that contains high
concentrations of Fe (100 mg/L) and numerous trace metals. Long-term monitoring of
the site showed that biological Fe(II) oxidation occurred across the mound without
human intervention, but during the majority of the year very little Fe(II) was oxidized
before the AMD reached the effluent end of the mound. Maximizing the removal of Fe
and other metals on the preexisting iron mound could substantially increase the efficiency
of conventional passive treatment systems such as limestone drains or wetlands.
On-mound channel reactors along with laboratory-scale “gutter” reactors were
constructed to determine optimal conditions for passive biological Fe(II) oxidation.
Dissolved Fe(II) was much more efficiently oxidized from gutter reactors that contained
iron mound sediment than ones without any sediment. Residence times of 5-10 hours
were shown to oxidize close to 100% of dissolved influent Fe(II) and remove 75% of
dissolved total Fe. Additionally the reactors performed better as the age of the sediments
increased and consequently shorter residence times of 1-2 hours were also capable of
oxidizing substantial amounts of the influent Fe(II). The addition of surface area to the
on-mound reactors improved Fe(II) oxidation at residence times of 30 minutes or less.
The results of this study can be used to help design and implement large-scale treatment
systems for low-pH AMD discharges.
iv
TABLE OF CONTENTS
List of Figures ..............................................................................................................vi
List of Tables ...............................................................................................................xi
Acknowledgments........................................................................................................xii
1 INTRODUCTION /LITERATURE REVIEW.....................................................1
1.1 Generation of Acid Mine Drainage ................................................................1 1.2 History and Setting of Hughes Borehole ........................................................5 1.3 Active and Passive Treatment Methods .........................................................8 1.4 Biological Fe(II) Oxidation ............................................................................9 1.5 Fe(II) Oxidizing Bacteria................................................................................11
2 OBJECTIVES.......................................................................................................13
3 MATERIALS AND METHODS .........................................................................15
3.1 Field Site Characterization .............................................................................15 3.1.1 Field Site Sampling locations...............................................................15
3.1.2 On-mound Channel Reactor Construction ...........................................18 3.1.3 Channel Reactor Sampling Locations ..................................................20 3.1.4 Field Meter Measurements ...................................................................20 3.1.5 Water Sample Collection and Preservation Techniques ......................21 4.1.6 Salt Tracer Slug Tests...........................................................................22
3.2 Analytical Procedures.....................................................................................22 3.2.1 Dissolved Iron Measurements ..............................................................22 3.2.2 Dissolved Trace Metal Analysis...........................................................23 3.2.3 Elemental Analysis of the Iron Mound Sediment ................................23 3.2.4 Non-purgeable Organic Carbon and Total Nitrogen ............................24 3.2.5 Microbial Population Direct Counts.....................................................24 3.2.6 Sulfate and Reactive Phosphate............................................................25 3.2.7 Acidity ..................................................................................................25 3.2.8 X-ray Diffraction ..................................................................................25
3.3 Experimental Methods....................................................................................26 3.3.1 Organic Amendment Leachates ...........................................................26 3.3.2 Batch Reactor Design ...........................................................................27 3.3.3 Laboratory-Scale “Gutter” Reactor Design..........................................28 3.3.4 Laboratory Gutter Reactor Sampling Procedure ..................................29 3.3.5 Coconut Fiber Insertion........................................................................30
3.4 Iron(III) Speciation Modeling.........................................................................30 3.5 Iron(II) Oxidation Modeling...........................................................................31
v
4. RESULTS.............................................................................................................33
4.1 Hughes Borehole Chemistry Data ..................................................................33 4.1.1 Transect Chemistry...............................................................................33
4.1.2 Fence and Toe Observations ................................................................38 4.2 On-Mound Channel Reactor Data ..................................................................48 4.3 Laboratory-Scale Gutter Reactor Results .......................................................55
4.3.1 Variable-Residence Time Experiment ................................................56 4.3.2 Repeat of Variable-Residence Time Experiment .................................64 4.3.3 Coconut Fiber Experiment ...................................................................66
4.3.4 Carbon Dioxide Purge Experiment ......................................................72 4.4 Gutter Reactor Fe(II) Percent Remaining.......................................................73 4.5 Modeling Results for Fe(OH)3 Solubility .......................................................78 4.6 Modeling Results for Fe(II) Oxidation Kinetics Gutter .................................79
5 DISCUSSION.......................................................................................................81
6 CONCLUSION.....................................................................................................91
Bibliography ................................................................................................................94
Appendix A Microbial Population Data for the Variable-Residence Time Experiment............................................................................................................98
Appendix B Tabulated Data for Hughes Borehole Chemistry ...................................99
Appendix C Tabulated Data for the On-mound Channel Reactors ............................101
Appendix D Tabulated Data for the Laboratory-Scale Gutter Reactors.....................103
Appendix E Tabulated Data for the Discussion..........................................................108
vi
LIST OF FIGURES
Figure 1-1. Rates of pyrite oxidation by Fe(III) and dissolved oxygen (DO). At pH values below 4, pyrite oxidation occurs predominantly by reaction with Fe(III). At pH values above 4, pyrite oxidation occurs primarily by reaction with oxygen (adapted from Williamson et al. 2006) ............................................3
Figure 1-2. Picture of the mixing zone of the Little Conemaugh River and the AMD discharge from Hughes Borehole. The orange coating on the right is commonly referred to as yellowboy. ...................................................................4
Figure 1-3. The Hughes Mine Complex and Hughes Borehole location near Portage, PA (GAI Consultants, 2007). The underground mines cover an area of 7,302 acres and the Lilly and Piper mines are hydrologically up-gradient of Hughes Mine.........................................................................................................6
Figure 1-4. Bituminous and anthracite coal fields of Pennsylvania (DCNR, 1992). Shades of yellow and orange represent types of bituminous coal while pink regions designate anthracite coal fields. The approximate location of Hughes Borehole is marked by the blue circle ..................................................................7
Figure 1-5. Biological and abiotic rates of iron(II) oxidation of acid mine drainage (adapted from Williamson et al., 2006). The circles in the diagram are data correlated from field studies and the O2 green lines are from a theoretical model by Pesic et al., (1989)...............................................................10
Figure 3-1. Pictures of Hughes Borehole showing the fence (top photo) and toe (bottom photo) sampling locations .......................................................................16
Figure 3-2. Topographic survey map of Hughes showing the locations of the fence, toe, and on-mound channel reactor ............................................................17
Figure 3-3. Pictures of the field site channel reactors (looking upstream) at two different stages of the experiment; October 2008 (upper photo) and June 2009 (lower photo). The channels were labeled A – H (from left to right)..................19
Figure 3-4. Picture of laboratory scale gutter reactors showing the feed tank in the background and the four gutter reactors in the foreground. The gutter reactors are labeled 1, 2, 3, and 4 from left to right in the picture with reactors 1 and 2 as the controls and 3 and 4 as the experimental sediment reactors..........29
Figure 4-1. Map of Hughes Borehole showing the sampling campaign locations with corresponding distances from the borehole or source. The B transect designates channel flow and the C and D transects represent sheet flow. The C section was sampled twice in 2007, in August and December, and the D section was sampled in May 2009. The B transects were sampled in both years and the points are labeled with either a 7 or 9 to designate the years 2007 and 2009, respectively .................................................................................35
vii
Figure 4-2. Transect data showing Fe(II) concentrations, pH, and dissolved oxygen(DO) for separate sampling dates as function of distance from the source. The B transect represents channel flow and the C and D transects represent sheet flow ..............................................................................................36
Figure 4-3. XRD patterns from three locations on the iron mound at Hughes Borehole. The top black line is from a terrace, the middle red line is from the main channel, and the bottom blue line is from a pool.........................................37
Figure 4-4. Daily flow rate data recorded from the pressure transducer at Hughes Borehole. Upper panel contains daily flow values for 2 ½ years and average monthly rainfall data from Johnstown, PA. The red lines indicate the available flow data from the study period. The Lower panel displays the flow rate data for the study period with the Fe load calculated from the specific sampling dates ......................................................................................................41
Figure 4-5. Upper panel: Fence and toe dissolved Fe(II) concentrations versus calendar date representing the emergent and effluent ends of the Hughes Borehole iron mound. Lower panel: Available daily flow measurements from the pressure transducer weir at Hughes Borehole versus the same calendar dates as the sampling events. .................................................................45
Figure 4-6. Dissolved Fe(II) and Fe(III) concentrations at the fence and toe locations of Hughes Borehole. Fe(II) is in green and Fe(III) is in red ................46
Figure 4-7. Top figure: Dissolved metal concentrations at the fence (orange) and toe (blue) locations of Hughes Borehole. Bottom figure: Normalized metal concentrations from the fence and toe. As, Cr, Cu, Pb, and Ti were also measured but all concentrations were <0.01 mg/L, or non-detect........................47
Figure 4-8. Stacked plot with Fe(II) concentrations for the channel reactors at Hughes Borehole. Top figure presents the normalized Fe(II) concentrations for the average of the treatment and control channels. Bottom figure plots the actual Fe(II) concentration for the influent, treatment controls, and control channels. Period I refers to the period with no steps. Period II refers to the period when all the steps were installed. Period III refers to the physical treatment period with the plastic media. Period IV refers to the chemical treatment period with coir mats and netting. The red dashed line indicates no change in Fe(II)out / Fe(II)in...................................................................................51
Figure 4-9. Normalized dissolved Fe(II) concentrations for the separate on-mound channel reactors sets at Hughes Borehole during the coir period (IV). Plastic media was left in channels A-B, whereas channels C-D, and E-F received coir netting and coir mat, respectively. The red dashed line indicates no change in Fe(II)out / Fe(II)in. .............................................................52
viii
Figure 4-10. Dissolved non-purgeable organic carbon and total nitrogen concentrations from the channel reactor at Hughes Borehole during the coconut fiber (coir) treatment phase, period IV....................................................54
Figure 4-11. Residence time of the channel reactors at all four periods of the experiment, no modifications (I), step period (II), plastic media period (III), and coconut fiber period (IV) ...............................................................................55
Figure 4-12. Gutter reactor experiment testing variable residence times, 10, 5, 2, and 1 hour. The sediment reactors contained sediment from Hughes Borehole and the control reactors did not contain any sediment..........................57
Figure 4-13. Actual concentrations of dissolved Fe(II) from the variable-residence time experiment. The upper graph shows the influent and effluent values for the control reactors, and the lower graph shows values for the sediment reactors ..................................................................................................58
Figure 4-14. pH measurements from the variable-residence time experiment for both the control and sediment reactors. ................................................................59
Figure 4-15. Dissolved oxygen measurements from the variable-residence time experiment for both the control and sediment reactors ........................................60
Figure 4-16. Dissolved Fe(II) and Fe(III) measurements from the variable-residence time experiment. The experiment reactors are graphed above the control reactors. Fe(III) is in red and Fe(II) is in green .......................................61
Figure 4-17. Dissolved Fe(II) oxidation efficiencies for the repeat of the variable-residence time experiment which was conducted at 5 and 2 hour times. The red dashed line indicates no change in Fe(II)out / Fe(II)in.. ................65
Figure 4-18. pH values for the second residence time experiment which was conducted at 5 and 2 hour times ...........................................................................66
Figure 4-19. Dissolved Fe(II) oxidation efficiency kinetics for organic amendment batch reactors, including live (no-amendment) and sterile control reactors..................................................................................................................67
Figure 4-20. Dissolved Fe(II) oxidation for the gutter reactor experiment with the addition and removal of the coir mat at a residence time of 1 hour. Period I corresponds to the gutter reactors at a 1 hour residence time with no coir. Period II corresponds to the addition of coir to the sediment reactors (referred to as coir reactors in this section). Period III corresponds to the removal of the coir. Period IV corresponds to the second addition of the same coir mats from Period II. Period V corresponds to the removal of the coir a second time. The red dashed line indicates no change in Fe(II)out / Fe(II)in.. ..................69
ix
Figure 4-21. pH values for the coconut fiber experiment at the influent and effluent of the control and coir reactors. Period I corresponds to the gutter reactors at a 1 hour residence time with no coir. Period II corresponds to the addition of coir to the sediment reactors (referred to as coir reactors in this section). Period III corresponds to the removal of the coir. Period IV corresponds to the second addition of the same coir mats from Period II. Period V corresponds to the removal of the coir a second time. ..........................70
Figure 4-22. Dissolved oxygen for the coir fiber experiment at the influent and effluent of the control and coir reactors. Period I corresponds to the gutter reactors at a 1 hour residence time with no coir. Period II corresponds to the addition of coir to the sediment reactors (referred to as coir reactors in this section). Period III corresponds to the removal of the coir. Period IV corresponds to the second addition of the same coir mats from Period II. Period V corresponds to the removal of the coir a second time. ..........................71
Figure 4-23. NPOC and TN concentrations for the gutter reactors and during section II of the coconut fiber experiment ............................................................72
Figure 4-24. Dissolved Fe(II) oxidation for the experiment with 15% CO2:N2 balance purge of feed tank graphed against number of pore volumes. Period I refers to a 2 hour residence time with the N2 purge. Period II refers to a 1 hour residence time, also with the N2 purge. Period III refers to a 1 hour residence time with the15% CO2 gas mixture. The red dashed line indicates no change in Fe(II)out / Fe(II)in. .............................................................................73
Figure 4-25. Dissolved Fe(II) percent remaining at pseudo-steady state for varying hydraulic residence times during times of no modifications to the sediment gutter reactors. .......................................................................................74
Figure 4-26. Dissolved Fe(II) percent remaining for the age of sediments during times of no modifications to the sediment gutter reactors. ...................................76
Figure 4-27. Dissolved Fe(II) and total dissolved Fe percent remaining for the gutter reactor experiments. All reported values are under “original” conditions, except for the 1 hour w/coir mat. Both variable residence time experiments contained similarly aged sediments whereas the Coir and CO2 experiments contained older sediments ................................................................77
Figure 4-28. Dissolved Fe(II) oxidation efficiency for initial residence time of both residence time experiments, RT1 and RT2. RT1 began with a 10 hour time and RT2 began with a 5 hour. The red dashed line indicates no change in Fe(II)out / Fe(II)in. ..............................................................................................78
x
Figure 4-29. Fe(OH)3 solubility versus pH for varying levels of SO42-. The figure
was created with equilibrium equations and pKa values in Microsoft Excel. Fe(III) concentrations with corresponding pH values are also plotted on the figure. Hughes Borehole refers to fence and toe data, and RT1 refers to the first variable residence time experiment...............................................................79
Figure 5-1. Batch reactor data for sterile and live reactors with no iron mound sediment. The filter sterilized and 1% v/v formaldehyde reactors were also under aerobic conditions.......................................................................................82
Figure 5-2. Dissolved Fe(II) and total dissolved Fe percent remaining for select experiments. The fence/toe data was from August 14, 2008 to September 18, 2008. All gutter reactor measurements are from the sediment reactors at pseudo-steady states and are under original conditions of no modifications and N2 purging of the feed tank, with the exception of the 1 hour w/coir mat.....85
Figure 5-3. Dissolved Fe(II) percent remaining for the control and treatment channels from on-mound channel reactors. The average Fe(II) percent remaining for each channel set (A-B, C-D, E-F, G-H) were averaged for each period and graphed against the mean residence from the corresponding period. ...................................................................................................................88
Figure 5-4. Biological and abiotic rates of iron(II) oxidation of acid mine drainage (adapted from Williamson et al., 2006). Squares represent the sediment reactors from RT1, and the triangle represents the sediment reactors from the Coir experiment. Red is 10 hour, blue is 5 hour, green is 2 hour, and black is 1 hour residence times. The two lines with O are taken from Pesic et al., 1989 and the circles are from various published oxidation rates from field studies...........................................................................................................89
2
xi
LIST OF TABLES
Table 3-1. Equilibrium equations and log K values for various Fe and SO42-
species used to generate a plot of Fe(OH)3 solubility versus pH .........................31
Table 4-1. Water quality parameters from the fence and toe locations at Hughes Borehole. The standard deviation (± #), range (# - #), and number of samples (N) are given for each parameter ..........................................................................42
Table 4-2. Elemental analysis from the top 2 cm of sediment from the fence and toe locations at Hughes Borehole. Metal oxide values are in weight percent (%) of the original sample ....................................................................................43
Table 4-3. Dissolved oxygen (DO), pH, temperature, and conductivity for the channel reactors and influent splitter box at Hughes Borehole ............................53
Table 4-4. Elemental analysis of the sediment from the effluent end of the sediment reactors following the Coir experiment. The average metal oxide values are in weight percent(%) and the standard deviation (± #) is given for each .......................................................................................................................62
Table 4-5. Dissolved average metal concentrations for the feed tank and effluent of the gutter reactors at pseudo-steady state of each residence time. As, Cr, Pb, and Ti were also analyzed, but all concentrations were non-detect (<0.01 mg/L). The standard deviation (± #) is displayed for the gutter reactors; only one sample from the feed tank was analyzed at each residence time...................63
Table 4-6. NPOC and TN concentrations for the organic amendment leachates that were used for batch experiments ...................................................................67
Table 4-7. Initial and final pH values and Fe(II) oxidation rates (mol/L-s) for the organic amendment batch reactors. ......................................................................68
Table 4-8. Approximate experimental age of sediments, in days, at time of pseudo-steady state for varying hydraulic residence times during times of no modifications for the sediment gutter reactors .....................................................75
Table 4-9. Zero-order Fe(II) oxidation rates during pseudo-steady state of the gutter reactors from the first-variable residence time experiment and the insertion of the coir mat in the coir experiment....................................................80
xii
ACKNOWLEDGMENTS
I give me sincere thanks to my Master’s committee, especially my advisor Dr. Bill
Burgos, for their patience and guidance through my education. Special thanks to Juliana
Brown for her assistance and contributions to this study. I would also like to thank Dave
Faulds and Matt Hassinger who helped me with all my construction needs and Dave
Jones who answered numerous questions about the analytical equipment. Thanks to
Adam Dryburgh, Michael Adelman, the PADEP, and Brent Means who aided in my
collection and analyses of data. Lastly, I appreciate and all the other students, staff, and
family who supported me along the way.
1
1. INTRODUCTION / LITERATURE REVIEW
1.1 Generation of Acid Mine Drainage
Acid mine drainage (AMD) from coal mines, both active and abandoned, causes
significant detriment to the environment worldwide. The northern Appalachian Plateau
of the eastern United States contains more than 5,000 miles (8,000 km) of streams that
are affected by drainage from abandoned coal mines (Boyer and Sarnoski, 1995). Many
of these streams are located in Pennsylvania where approximately half of the discharges
are acidic, having a pH<5 (Brady et al., 1998). Anthracite coal fields in northeast
Pennsylvania and bituminous coal fields in the west, impact 45 of the 67 counties,
including at least 2,400 miles of streams and 250,000 acres of land (Rossman et. al, 1997;
PADEP, 2003).
AMD is created when metal sulfides, mainly iron sulfide (pyrite) are exposed to
water and oxygen through mining operations. The dissolution of pyrite (Eq. 1.1) creates
sulfuric acid and dissolved ferrous iron as the sulfide oxidizes to sulfate (Baker &
Banfield, 2003). The acidic water further dissolves metals in the surrounding rocks. The
main constituent of concern in Appalachian AMD is ferrous iron, Fe(II), but dissolved
sulfate (SO42-), aluminum (Al), manganese (Mn), and numerous trace metals are also
found in AMD discharges, particularly in strongly acidic low-pH waters (Cravotta,
2008a). Following pyrite oxidation, the presence of oxygen causes aqueous Fe(II) to
become oxidized and form ferric iron, Fe(III) (Eq. 1.2 ). Pyrite can also be oxidized by
Fe(III) (Eq. 1.3 and Figure 1-1). At pH < 4, pyrite dissolution is controlled by reaction
with Fe(III) and becomes autocatalytic as more and more pyrite is oxidized. At pH > 4,
2
pyrite oxidation occurs primarily by reaction with oxygen. Once oxidized, Fe(III)
hydrolyzes to produce Fe(III) hydroxides and additional acidity (Eq. 1.4). Some
precipitation of the Fe(III) hydroxides will occur at low-pH, such as in the form of
goethite, α-FeOOH (Cornell and Schwertmann, 1996). However, at low-pH the high
concentration of SO42- typically found in AMD discharges have been shown to increase
the solubility of Fe(III) by formation of FeSO and FeHSO species (Cravotta, 2008b). 4+
42+
FeS (s) + 3.5 O (aq) + H 0 Fe + 2SO + 2H2 2 22+
42- + 1.1
Fe + 0.25 O (aq) + H Fe + 0.5 H 02+2
+ 3+2 1.2
FeS (s) + 14 Fe + 8H 0 15Fe + 2SO + 16H23+
22+
42- + 1.3
Fe + 3H 0 Fe(OH) (s) + 3H3+2 3
+ 1.4
3
Figure 1-1. Rates of pyrite oxidation by Fe(III) and dissolved oxygen (DO). At pH values below 4, pyrite oxidation occurs predominantly by reaction with Fe(III). At pH values above 4, pyrite oxidation occurs primarily by reaction with oxygen (adapted from Williamson et al. 2006).
A common feature of AMD-impacted surface waters is the appearance of
‘yellowboy’, which is an Fe(III) hydr(oxide) precipitate that is generally orange or red
(Figure 1-2). Yellowboy occurs when the Fe(II) and Fe(III)-laden AMD enters
circumneutral pH surface waters which allows swift abiotic oxidation of dissolved Fe(II)
and hydrolysis of Fe(III) (Stumm and Morgan, 1996). The Fe(III) hydroxides are very
insoluble at neutral pH values and precipitate almost immediately upon mixing with
more-neutral surface waters, such as streams. The yellowboy coats stream beds which
inhibits plant and algae growth, and creates a slippery surface on which
4
macroinvertebrates cannot attach. In addition, the metals can clog the gills of fish and
lower the pH of receiving waters and thus creates stretches of “killed” streams.
Healthy natural stream
AMD impacted stream
Figure 1-2. Picture of the mixing zone of the Little Conemaugh River and the AMD discharge from Hughes Borehole. The orange coating on the right is commonly referred to as yellowboy.
An area of Pennsylvania highly affected by AMD is the Stonycreek-Conemaugh
watershed located in south-central PA which drains to the Ohio River. This watershed is
only 467 square miles but the United States Geological Survey in 1994 identified 270
abandoned coal mine discharges located here with the majority exceeding effluent
standards for total iron and manganese concentrations (Zink et al., 2005). A subdivision
of this watershed includes the Little Conemaugh watershed in which a state funded study
found 197 coal mine discharge points, with seven major discharges that attribute over
73% of the metal load to the watershed (Zink et al., 2005). One of these seven large
discharge points is Hughes Borehole.
5
1.2 History and Setting of Hughes Borehole
Hughes Borehole is an artesian AMD discharge located in Cambria County, PA,
near Portage, PA. It is approximately 100 feet from the Little Conemaugh River;
however, the majority of the flow from Hughes Borehole enters the Little Conemaugh
River a few thousand feet downstream of the site. The flowrate ranges from 300 to 2000
gallons per minute (gpm) throughout the year with a yearly average of approximately
1000 gpm. The emergent discharge has an average pH of 4 and dissolved concentrations
of iron, aluminum, and manganese of 90, 8, and 2 mg/L, respectively. A 1.5 acre “iron
mound” (area of metal hydroxide deposition) surrounds the borehole and has depths up to
5 - 6 feet.
Hughes Borehole drains a mine complex consisting of several interconnected
underground coal mines (Figure 1-3): Hughes Mine, W.H Piper-Sonman #2 deep mine,
and Lilly #3 and #3A deep mines (GAI consultants, 2007). The mines are located along
the Lower Kittanning bituminous coal seam (Figure 1-4). Together, these mines comprise
an area of 7,302 acres with Hughes mine as the largest at 3,657 acres. Operations at
Hughes started before 1923 and continued until the mine was closed in 1954. The other
three mines, Sonman, Lilly #3, and Lilly #3A, remained in production until 1958, 1954,
and 1968, respectively. Production records for all four mines were only recorded for 33
of the 43 years that the mines were in production. The yearly average of coal production
for these 33 years was 324,406 tons, with a total production of 10,705,400 tons.
6
Emergent discharge
Figure 1-3. The Hughes Mine Complex and Hughes Borehole location near Portage, PA (GAI Consultants, 2007). The underground mines cover an area of 7,302 acres and the Lilly and Piper mines are hydrologically up-gradient of Hughes Mine.
7
Figure 1-4. Bituminous and anthracite coal fields of Pennsylvania (DCNR, 1992). Shades of yellow and orange represent types of bituminous coal while pink regions designate anthracite coal fields. The approximate location of Hughes Borehole is marked by the blue circle.
The borehole at the Hughes site was drilled sometime during the operation of
Hughes Mine in the 1920s for drainage purposes. The other three mines, which are all
hydrologically up-gradient of the Hughes site, were deliberately connected to allow for
gravity drainage into the Hughes mine and all the flow emerges at Hughes Borehole. The
borehole was capped in the 1950s but blew out in the 1970s and has been discharging
ever since (Zink et al., 2005).
Monitoring of Hughes Borehole started in 1992 by the EPA and continues to the
present day. The Susquehanna River Basin Commission (SRBC) installed a rectangular
weir with a pressure transducer at Hughes Borehole in 2006 to obtain daily flow
measurements.
8
1.3 Active and Passive Treatment Methods
AMD can be treated by either active or passive methods. Active treatment of
AMD generally involves the collection of water followed by the addition of an alkaline
material to precipitate metals. This process can effectively remove dissolved metal
concentrations and increase the alkalinity and pH of the effluent, but requires the disposal
of sludge. This results in high capital and operation & maintenance costs. Due to the
numerous large AMD discharges across the Appalachia region, active treatment can not
economically be implemented at all of these sites.
At Hughes Borehole, an active treatment system, consisting of an equalization
pond, chemical treatment, clarification, and sludge removal would cost $5,263,500 for
capital costs alone, and $658, 900 each year for operation and maintenance (GAI
Consultants). This is a costly treatment option for just one of many acidic discharges;
therefore, a low cost solution that utilizes passive methods to remediate the AMD is
desired.
Passive treatment technologies generally include running the AMD over an
alkaline material to raise the alkalinity and pH to precipitate the metals. This results in
“armoring” of the limestone by the Fe(III) hydroxide precipitates which hinders the
neutralizing capacity of the limestone. Other passive treatment technologies such as
vertical flow wetlands or successive alkalinity producing systems (SAPS), and anoxic
limestone drains have been developed, but these technologies show varying treatment
efficiencies (Demchak et al., 2001). These options may also be hindered by armoring or
clogging by metal precipitates. In order to prevent or limit armoring, dissolved Fe can be
removed before the AMD waters are neutralized with limestone (Nengovhela, 2006).
9
1.4 Biological Fe(II) Oxidation
Singer and Stumm (1970) demonstrated that of all the potential catalysts they
studied, microorganisms appeared to have the greatest effect in increasing the rate of
Fe(II) oxidation, in some cases by a factor of more than 10 .6 At low pH (2.5 – 4.5) the
abiotic oxidation of Fe(II) is limited, but biological oxidation of the Fe(II) by iron
oxidizing bacteria (IOB) occurs in this acidic environment. At pH values less than 3.5,
biological Fe(II) oxidation has been shown to dominate over abiotic oxidation, with
biological Fe(II) oxidation rates a few orders of magnitude higher than abiotic oxidation
rates (Figure 1-5) (Kirby et al., 1999; Williamson et al., 2006). Although there is no
consensus on which variables are most important to determine Fe(II) oxidation rates, a
number of studies have shown that Fe(II) oxidation is much faster in the field than sterile
laboratory settings (Noike et al., 1983; Kirby and Brady, 1998). A first order rate law for
homogeneous Fe(II), which involves the oxidation of dissolved iron(II) species such as
Fe2+, FeOH+, or Fe(OH)2, was developed by Stumm and Morgan (1981) (Eq. 1.1).
d[Fe(II)] = -k [Fe(II)] [O ] [OH ]1 2- 2
1.5 dt
Published first order rate constants, which depend on Fe(II) concentration, for
oxidation in low-pH conditions range from 5 x 10-7 to 10-5 mol L-1s-1 (Noike et al., 1983;
Nordstrom, 1985). Kirby and Brady (1998) published rates of 10-9 mol L-1s-1 to 3.27 x
10-6 mol L-1 s-1, but found no statistical correlation between pH, Fe(II), or DO. Fe(II)
oxidation rates may depend on more variables than pH, Fe(II) and DO, and could be
10
affected by temperature, concentration of bacteria, particle surface area, incident sunlight,
and complexation (Pesic et al., 1989; Kirby and Brady, 1998; Kirby et al., 1999).
Figure 1-5. Biological and abiotic rates of iron(II) oxidation of acid mine drainage (adapted from Williamson et al., 2006). The circles in the diagram are data correlated from field studies and the O green lines are from a theoretical model by Pesic et al., (1989).
2
The theoretical model determined by Pesic et al., (1989), which was used to
derive the two diagonal green lines for the different concentrations of DO in Figure 1-5
was based on the following pseudo-first order rate expression(Eq. 1.6):
-d[Fe2+] / dt = 1.62 x 1011Cbact[H+][Fe2+] pO2e-(58.77/RT) 1.6
11
where Cbact is the concentration of bacteria, H+ is the concentration of hydrogen ion, pO2
is the partial pressure of oxygen, R is the universal gas constant, and T is temperature.
1.5 Fe(II) Oxidizing Microorganisms
Diverse biotic communities including autotrophic and heterotrophic bacteria,
protists, fungi, algae, and yeasts have been found at acidic and circumneutral-pH mine
drainage sites (Wichlacz and Unz, 1981; Emerson and Moyer, 1997; Johnson 1998;
Baker and Banfield, 2003;). Two of the most well-known Fe(II) oxidizing bacteria are
Acidothiobacillus ferrooxidans and Leptospirillum ferrooxidans but mine drainage sites
support a wide range of iron-oxidizing chemolithotrophs as well as heterotrophs
(Johnson, 1998). Chemolithotrophic bacteria such as additional Acidithiobacillus spp.
and obligately heterotrophic bacteria such as Acidiphilium spp. and Ferrimicrobium have
been isolated from AMD environments (Johnson et al., 2001; Rowe and Johnson, 2008).
Favorable conditions for the growth of IOB varies between species, but the
microorganisms must be capable of living in harsh environments. Sufficient amounts of
oxygen and carbon must be present along with lower concentration of specific nutrients,
such as nitrogen and phosphorus. Both optimal and maximum temperatures for
biological Fe(II) oxidation have been shown to be pH-dependent, with temperatures
decreasing as pH decreases (Nemati et al., 1998). Efficient bio-oxidation of Fe(II) was
found in the temperature range of 20-44 °C and the pH range of 1.8-2.3 (Malhotra et al.,
2002).
Obligate aerobic autotrophs, like A. ferrooxidans and L. ferrooxidans, require
both carbon dioxide and oxygen for growth. In a study by MacDonald and Clark (1970),
12
CO2 had no effect on the growth rate of A. ferrooxidans at sparge volume concentrations
of 0.035 – 10%. DO has been shown to limit growth at concentrations of 0.29 mg/L or
less (Nemati et al., 1998). Smith et al. (1988) proposed an overall stoichiometric
relationship for bacterial Fe(II) oxidation and biomass synthesis (represented by
C5H7O2N) by A. ferrooxidans at pH 2 with a typical electron transfer efficiency of 30%
(Eq. 1.7 ).
1.7Fe2+ + 0.224 O2 + 0.0045 CO2(g) + 0.0011 HCO3- + H+ +
0.0011NH4+ Fe3+ + 0.0011 C5H7O2N + 0.4989 H2O
Most AMD environments have low concentrations of dissolved organic carbon at
less than 20 mg/L (Johnson and Hallberg, 2003). Therefore, many of the acidophilic
heterotrophs are considered oligotrophic and can live on organic carbon originating from
leakage or lysis products from chemolithotrophic acidophiles (Johnson 1998). The most
critical limiting factor for the biotreatment efficiency of AMD is the availability of
carbon (Gibert et al., 2004). Increasing the availability of carbon at a site like Hughes,
which has little canopy coverage, could possibly stimulate the heterotrophs to more
efficiently oxidize iron.
Senko et al. (2008) showed that bacterial communities of both autotrophic and
heterotrophic bacteria capable of efficient removal of Fe(II) from AMD were present at
two Pennsylvania bituminous AMD sites, Gum Boot Run and Fridays-2. They suggest
that maximizing O2 concentrations and residence time would maximize the oxidation
efficiency of Fe(II) at these sites.
13
2. OBJECTIVES
Acid mine drainage (AMD) is a huge problem for the quality of surface waters in
the Appalachia region of the United States. Typical passive treatment systems focus on
the neutralization of AMD by alkalinity additions to raise the pH and promote
precipitation of dissolved metals. However, the precipitates can quickly coat the
treatment systems and compromise the neutralizing capacity of the system. Since iron is
the most dominant metal in many AMD discharges, biological oxidation could be used to
remove iron before the addition of alkalinity, thus improving the efficiency of the
treatment systems.
Existing iron mounds surrounding some AMD discharges show that natural
precipitation of certain metals can occur before the discharges reach surface waters.
Therefore, modifications of these mounds could be developed that would allow even
greater metal precipitation in a localized area and help maintain healthy aquatic
ecosystems.
This study attempted to characterize and modify the iron mound at Hughes
Borehole to address the following list of objectives:
1. The first objective was to better understand the geochemistry of the existing iron
mound at Hughes Borehole by conducting long term monitoring of various
locations on the mound.
2. The second objective was to determine the effect of biological oxidation of iron
and investigate various organic amendments with batch reactors.
14
3. The third objective was to construct on-mound treatment systems to analyze the
effect of different modifications to the mound, including physical and chemical
treatment options.
4. The fourth objective was to construct laboratory-scale “gutter” reactors to mimic
field conditions and further analyze possible modifications to the existing iron
mound, such as extended residence times.
5. The fifth objective was to develop design parameters for biological low-pH
passive treatment systems at AMD impacted areas.
15
3. MATERIALS AND METHODS
3.1 Field Site Characterization
At Hughes Borehole, locations were chosen to analyze the geochemistry of the
iron mound from the emergent to the effluent ends of the mound. Transects were also
conducted to compare different physical characteristics of the iron mound.
3.1.1 Field Site Sampling Locations
The majority of long-term monitoring data were collected from two locations
along the flow path across the iron mound and are referred to as the “fence and toe”
locations (Figures 3-1). The “fence” location was 8 feet downstream of the emergent
artesian discharge and the “toe” location was at the effluent end of the mound, 200 feet
downstream. Three transects were preformed on the mound that included areas of
channel flow and sheet flow. Channel flow was defined as deep and wide areas that
carried a significant amount of the surrounding flow. Sheet flow was defined as areas
where the AMD spread out across the surface of the mound and was fairly shallow, or no
more than a couple inches deep. The sheet flow area contained terraces, sections of step-
like structures where the AMD would cascade over the sediment, and pools, sections
where the AMD would become trapped and was relatively stagnant. There was one main
channel at the Hughes Borehole iron mound which developed into variable sheet flow
sections further downstream.
Toe
16
Figure 3-1. Pictures of Hughes Borehole showing the fence (top photo) and toe (bottom photo) sampling locations.
Fence
Toe
17
Toe
Fence
Channel Reactor
Toe
Fence
Channel Reactor
ToeToe
FenceFence
Channel ReactorChannel Reactor
Figure 3-2. Topographic survey map of Hughes showing the locations of the fence, toe, and on-mound channel reactor.
18
3.1.2 On-mound Channel Reactor Construction
On-mound channel reactors were constructed on the Hughes Borehole iron mound
in June 2008 (Figure 3-3). The reactors were comprised of a splitter box, measuring 4
feet x 8 feet x 4 feet high, and eight 40 foot channels. The walls of the splitter box
consisted of 5/8” marine plywood interconnected with pressure treated 2 x 4 boards. A 2
foot high weir was placed along the center of the box to evenly distribute the AMD
flowing into it. The box and weir were leveled upon construction. Hughes AMD flows
into the box via a 3” PVC schedule 40 pipe through the back of the splitter box 1.5 feet
above the bottom of the box and 0.5 feet below the mound surface. The influent pipe was
connected to the main channel flowing from the emergent discharge at Hughes Borehole
approximately 10 feet downstream of the SRBC weir. The box was placed down-
gradient of the intake structure of the influent pipe to promote gravity flow.
The eight channels were made with pressure treated 2 x 8 boards measuring 8 feet
in length. The boards were connected with metal screw strips after placement on the
mound. The boards were initially placed on the existing mound surface and then
sediment from the top 1 to 2 feet of the surrounding area of the mound was shoveled into
the channels. The final sediment depth within the channels was 3 to 4 inches, thus
leaving 4 to 5 inches of freeboard space. An average slope of 1.5 feet / 40 feet, or 0.375
was maintained to allow a sufficient drop in elevation for gravity flow. The channels
were labeled A, B, C, D, E, F, G, and H (from left to right looking upstream). Channels
G and H were control reactors and received no modifications besides the initial
construction.
19
Figure 3-3. Pictures of the field site on-mound channel reactors (looking upstream) at two different stages of the experiment; October 2008 (upper photo) and June 2009 (lower photo). The channels were labeled A – H (from left to right).
In August 2008, wooden blocks were added as small steps to the 6 experimental
channels to increase the residence time in each. Each channel modification was
conducted in duplicate in adjacent channels (e.g. A and B, C and D, E and F). The blocks
were pressure treated 2x6 boards and cut to fit within the channel. They were screwed
into the channel walls leaving 1.5 inches of freeboard above each step.
In October 2008, plastic media was added to the experimental channels to further
increase the residence time and, more specifically, to increase the surface area within the
channels. Brentwood cross flow media, CF 1900, with an effective surface area of 48 ft2
20
/ ft3, was selected. The media was cut with a chainsaw to 4-5 inches high and 10-10.25
inches wide. The pieces were pushed 0.5-1.0 inches into the top layer of sediment.
In June 2009, exactly half of the plastic media was removed by removing every
other piece in channels A and B. All the plastic media was removed from channels C, D,
E, and F. Coconut fiber (coir) mats were then placed on top of the mound sediments that
were newly exposed. The coir mats were added to the experimental channels to evaluate
whether the addition of a carbon (and possible nutrient) source could promote low-pH
Fe(II) oxidation. The addition of coir mats are referred to as a chemical modification of
the channels as compared to all the previous physical modifications of the channels.
3.1.3 Channel Reactor Sampling Locations
Samples for the channel reactor influent analyses were taken from the top few
inches of the AMD in the splitter box. Samples for the effluent analyses were taken from
immediately downstream of the last wood blocks in channels A-F, and a few inches
upstream from the end of the 2 x 8 wood boards in channels G and H.
3.1.4 Field Meter Measurements
Dissolved oxygen (DO), pH, electrical conductivity (EC), and temperature were
measured on-site with field meters. DO was measured with an Oakton 300 series meter
and was calibrated in the lab to 0% saturation then in the field to 100% air saturation.
The pH and temperature were measured with a Beckmann Ф200 series pH meter and was
calibrated in the lab and as needed in the field. Temperature was consistently measured
from this meter. EC was measured with an Oakton 400 series conductivity meter and
21
was calibrated in the lab at temperature similar to field. A Baski cutthroat flume was
used to measure the effluent discharge of the individual channels. AMD discharge from
the borehole was measured by a rectangular weir with a pressure transducer installed by
the SRBC.
3.1.5 Water Sample Collection and Preservation Techniques
The samples for ferrous and ferric iron measurements were filtered with 0.2 µm
syringe filters, and acidified to pH < 2 with hydrochloric acid (HCl). Samples for ICP-
AES were collected in the same manner except acidified with nitric acid (HNO3) to
pH<2. Samples for organic carbon and nitrogen analysis were acidified with sulfuric acid
to pH < 2 and then filtered with a 0.45 µm filter before being analyzed. Samples for
acridine orange direct counts (AODC) and sulfate were neither acidified nor filtered. All
samples were collected in either 50 mL or 15 mL sterilized, acid washed centrifuge tubes
and transported back to the lab in coolers. All samples were refrigerated at 4 oC upon
return to the laboratory.
Feed water for the laboratory gutter reactors was collected from the fence
location. Five gallon carboys were used to transport the raw AMD. The samples were
nitrogen flushed in the lab and then refrigerated at 4 oC, capped and stored for no more
than 3 weeks. Occasionally the feed tank AMD was spiked with ferrous sulfate ([Fe(II)]
in stock = 500 mg/L) and 0.5 M NaOH to have a more consistent Fe concentration and
pH for the lab reactors.
22
3.1.6 Salt Tracer Slug Tests
Salt tracer slug tests were conducted during the four periods of the on-mound
channel reactors to determine the mean residence time. 50 g of NaCl was mixed with 1
liter of Hughes AMD and poured into the influent end of the channels approximately 1
foot from the edge of the splitter box. An Oakton 400 series conductivity meter was
placed at the effluent end of each channel and the conductivity was recorded at intervals
of 10-30 seconds. The values were corrected to account for background concentrations.
3.2 Analytical Procedures
Most analytes from Hughes Borehole samples were measured in the laboratory
with exception of the field meter measurements.
3.2.1 Dissolved Iron Measurements
Dissolved ferrous iron (Fe(II)) was measured by the ferrozine assay. Samples
were centrifuged at 13,400 rpm to remove solids, and typically diluted 1:2 with 0.5 M
HCl to bring the values within the range of the Fe(II) standards. 20µL of diluted sample
was added to 1 mL of ferrozine reagent (50 mM HEPES buffer with pH adjustment to 6.8
- 7.0 with NaOH and 1 g/L of ferrozine iron reagent). Dissolved total iron was also
measured by the ferrozine assay after the ferric iron was reduced. Samples were added
to 0.5 M Hydroxylamine HCl by a ratio of 1:2 and allowed to react for 1.5 hours (Luu et
al., 2003). 20µL of diluted sample was then added to 1 mL of ferrozine reagent.
A Fe(II) concentration standard curve was generated with 0.025, 0.1, 0.25, 0.5,
and 1 mM FeCl2. The absorbance was measured at a wavelength of 562 nm on the
23
Shimadzu – UV-visible Spectrophotometer UV-1601. Blanks were created with 20µL of
0.5M HCl in order to zero the spectrophotometer. Dissolved Fe(II) concentrations were
calculated directly from the standard curve and dissolved Fe(III) concentrations were
calculated as the subtraction of Fe(II) from the dissolved total Fe.
3.2.2 Dissolved Trace Metals Analysis
Samples for metal analysis were filtered (0.2 µm) and acidified to pH < 2 with
concentrated HNO3 prior to storage and were analyzed within 6 months. The metals
analyzed included Al, As, Ca, Co, Cr, Cu, Fe, K, Mg, Mn, Na, Ni, Pb, Si, Sr, Ti, and Zn.
Samples were run on a Perkin-Elmer Optima 5300 ICP-AES (inductively coupled plasma
atomic emission spectrometer) by the Penn State Materials Characterization Laboratory.
Standards for each metal were included in each run to create a standard curve.
3.2.3 Elemental Analysis of the Iron Mound Sediment
Lithium metaborate fusion was conducted by the Penn State Materials
Characterization Laboratory on solid sediment samples to dissolve the sample and then
ICP-AES was performed on the mixture. The sediment samples were air dried then 0.25
g was mixed with 1.25 g lithium metaborate flux and heated to 1000oC in ultra pure
graphite crucible for 30 minutes. The mixture was then pored into 2% v/v nitric acid
prior to ICP-AES analysis.
24
3.2.4 Non-purgeable Organic Carbon and Total Nitrogen
Non-purgeable organic carbon (NPOC) and total nitrogen (TN) samples were
analyzed using a Shimadzu TOC-VCSN Total Organic Carbon Analyzer and Shimadzu
TNM-1 Total Nitrogen Measuring Unit, respectively. Zero-moisture compressed air was
used as the carrier gas and a Shimadzu ASI-V autosampler measured the samples.
Standards for NPOC and TN were created with a 1000 mg/L stock solution of potassium
hydrogen phthalate (2.125 g dried at 103 °C then mixed with 1 L MilliQ water) and 1000
mg/L stock solution of potassium nitrate (7.219 g dried for 3 hours at 105 °C, then mixed
with 1 L MilliQ water), respectively. The stock solutions were diluted to 10 and 50 mg/L
for NPOC and 10 and 30 mg/L for TN to create calibration curves. Five point calibration
curves were generated with these standards and automated dilution. The standard stock
solution and calibration curves were regenerated periodically.
3.2.5 Microbial Population Direct Counts
Microbial population counts were conducted by the acridine orange direct count
(AODC) method. 1 mL of sample was diluted with 1 mL of 0.2 µm filter sterilized
distilled water. 0.2 mL of acridine orange was used to stain the microbes and then was
pumped through a 0.2 µm black carbonate filter. The microbes were counted by hand on
an Olympus BH2 fluorescent microscope with a mercury vapor lamp.
Microbial counts from sediment samples were extracted with a sterile solution of
0.1 % sodium pyrophosphate adjusted to pH 3.5 with 0.5 M HCl. 1 g of sediment was
mixed with 9.5 mL of sodium pyrophosphate and placed on a shaker for 30 mintues
(Hurst, 2002). The samples were then centrifuged at 1000 x g for 10 minutes and 2 mL
25
of supernatant was removed for the AODC procedure. Dilutions were conducted
accordingly to bring the microbial counts within the acceptable grid range on the
microscope.
3.2.6 Sulfate and Reactive Phosphate
Sulfate and reactive phosphate were measured using HACH Test N Tube
procedures. HACH procedure 8051, SulfaVer 4 method, was used for sulfate, and
HACH procedure 8048, PhosVer3 method, was used for reactive phosphate
(orthophosphate). Both were measured on a HACH DR/2800 spectrophotometer along
with appropriate standards.
3.2.7 Acidity
Acidity was measured indirectly using a calculation based on pH and analytical
concentrations of dissolved Fe(II), Fe(III), Mn, and Al (Eqn 3.1)(Kirby and Cravotta,
2005). The equation does not count negative contributions of acidity, but this equation
was shown to be appropriate for waters having pH < 4.5 and low to no alkalinity.
Aciditycalculated = 50{1000(10-pH) + [2(FeII) + 3(FeIII)]/56 + 2(Mn)/55 + 3(Al)/27 3.1
3.2.8 X-ray Diffraction
Powder X-ray diffraction patterns were collected using a Rigaku DMAX/Rapid
Micro-Diffraction System with a Mo X-ray source and a 0.3 mm collimator. Intensities
were measured with phi axis oscillation from -20o to 20o and a speed of 1o per second for
26
a ten minute exposure. Prior to X-ray diffraction analysis, sediments were ground, sieved
and packed into 0.7 mm glass capillaries.
3.3 Experimental Methods
Batch reactors and gutter reactors were conducted in the laboratory. The batch
reactors were created to test different organic amendments and the effects of sterile
versus live reactors on Fe(II) oxidation. The gutter reactors were created to mimic the
field reactors and conditions and to test variables that could not be easily controlled in the
field.
3.3.1 Organic Amendment Leachates
Organic amendment leachates were prepared to test the effects of various nutrient
waters on Fe(II) oxidation rates. These leachates were created with autoclaved 1 liter
bottles. Solid amendments, such as shredded hardwood mulch, straw, and coconut fiber
(coir), were added to filtered (0.2 µm) Hughes Borehole AMD and placed on a shaker for
1-2 days. The three mixtures for mulch, straw, and coconut fiber contained 100 g mulch,
2.8 grams straw, and 2.8 grams of coir, respectively, in 600mL of the filtered AMD. The
mixture was centrifuged at 5,000 rpm for ten minutes and the supernatant was removed.
25 mL of his supernatant was added to 25 mL of live Hughes Borehole AMD and 10 g of
sediment in the batch reactors.
27
3.3.2 Batch Reactor Design
All batch reactors were conducted in 120 mL sterile (autoclaved at 123 oC and 16
psi for 15 minutes) serum bottles. The reactors were sealed with butyl rubber stoppers
and aluminum crimp tops, and sampled with sterile needles and syringes. Additional
oxygen was added to the aerobic reactors via air that was injected at each sampling time.
Sterile controls were created with the addition of 1 % (v/v) formaldehyde or were filtered
sterilized with 0.2 µm cellulose acetate membranes (Senko et al., 2008).
The organic amendment batch reactors contained 10 g of sediment and a mixture
of Hughes AMD and organic leachate. Control reactors without sediment contained only
50 mL of Hughes AMD or a mixture of AMD and leachate.
Other batch reactors used to observe the effects of sterile versus live reactors were
created in a similar manner, but did not contain sediment. They contained only 50 mL of
Hughes AMD and corresponding sterile control. An anaerobic batch reactor was purged
for 10 minutes with 100% nitrogen.
The batch reactors from both experiments were placed on a mixer (100 rpm)
throughout the entire experiment. At each sampling time the reactors were inverted, 1
mL of sample was removed, and then was centrifuged at 13,400 rpm for 2 minutes to
remove solids. 300 µL of supernatant was diluted by a ratio of 1:2 with 0.5 M HCl or 0.5
M Hydroxylamine HCL to measure dissolved ferrous and dissolved total iron,
respectively, by the ferrozine assay. NPOC, TN, and phosphate were measured by
aforementioned methods.
28
3.3.3 Laboratory-Scale “Gutter” Reactor Design
The laboratory-scale gutter reactors were created to mimic the on-mound channel
reactors on a smaller, more controllable scale (Figure 3-4). The reactors consisted of four
square PVC square tubes measuring 1” x 1”x 36” long. Control reactors with no
sediment were designated as reactors 1 and 2. Experimental reactors contained 100 g of
air dried sediment from Hughes Borehole and were designated as reactors 3 and 4. Weirs
consisting of rubber molding were glued at the influent and effluent ends of the gutters.
The weirs were cut to have a ¼” water column height of AMD in each gutter and a
volume of 125 mL.
The gutter reactors had a slope of 1/2” over 36”, or 0.0138 to allow for gravity
flow. The influent structure contained a small collection pool prior to the AMD flowing
into the gutter section which allowed for the influent sampling location. A peristaltic
pump was used to pump the AMD from the feed tank to the four gutters to ensure
precisely controlled flow to each. The effluent from the channels was pumped through
the same pump and collected in waste containers.
The feed water was collected directly from Hughes Borehole and was poured into
a 5 gallon Pyrex glass tank. The tank was sealed with a rubber stopper and was
constantly purged with a N2 gas mix to minimize oxidation of the ferrous iron. Six
separate tubes were placed through the rubber stopper to allow for the four tubes to carry
AMD, one for the N2 purge, and one for sampling. A seventh hole was drilled and filled
with copper tubing stuffed with glass fiber to allow for gas to escape the tank. The feed
water was replenished at the start of experiments or when the water level was less than
29
four inches in depth. Between experiments the feed tank was acid washed to remove any
precipitates.
Figure 3-4. Picture of laboratory scale gutter reactors showing the feed tank in the background and the four gutter reactors in the foreground. The gutter reactors are labeled 1, 2, 3, and 4 from left to right in the picture with reactors 1 and 2 as the controls and 3 and 4 as the experimental sediment reactors.
3.3.4 Laboratory Gutter Reactor Sampling Procedure
The frequency of sampling depended on the specific experiment. At each
sampling event, 1 mL samples were removed from the influent and effluent ends of each
channel and from the feed tank, and then were placed in semi-micro centrifuge tubes.
Each sample was centrifuged for 2 minutes at 13,400 rpm, then 0.3 mL of supernatant
was removed and added to 0.3 mL of either 0.5 M HCl or 0.5 M hydroxylamine HCL to
measure ferrous and total iron, respectively, by the ferrozine assay. pH was measured
with the remaining sample using a Thermo Orion 550A benchtop pH meter and semi-
micro Thermo Orion pH probe. Dissolved oxygen was directly measured in the gutters
with a Cole-Parmer benchtop dissolved oxygen meter and a Cole-Parmer glass
30
polarographic probe. On a limited basis, samples were removed for AODC, ICP-AES,
sulfate, TOC, and TN and were preserved as per previous stated methods.
3.3.5 Coconut Fiber Insertion
Coconut fiber (coir) from Rolanka landscaping mats was added to the gutter
reactors during the course of the experiment. 10 g of coir mat was added to the each
sediment gutter reactor. The coir mat was cut to 4/5 inches wide and 1/4 inches in depth,
to fit within the constraints of the reactors and was placed on top of the sediment.
3.4 Iron(III) Speciation Modeling
Microsoft Excel was used to generate a plot of Fe(OH)3 solubility versus pH for
and varying concentrations of sulfate. Equilibrium equations and log K values were
obtained from the PHREEQC modeling software database (Table 3-1).
31
Table 3-1. Equilibrium equations and log K values for various Fe and SO42- species used
to generate a plot of Fe(OH)3 solubility versus pH.
Equilibrium Equations Species log K
pFe3+ = pK(amph) + 3pH Fe3+ 0
p[FeOH2+] = pk + p[Fe3+] - pH FeOH2+ -2.19p[Fe(OH)2
+] = pk + p[Fe3+] - 2[pH] Fe(OH)2+ -5.67
p[Fe(OH)30] = pk + p[Fe3+] - 3[pH] Fe(OH)30
-12.53
p[Fe(OH)4-] = pk + p[Fe3+] - 4[pH] Fe(OH)4
- -21.6
p[HSO4-] = pK + pH2SO4
- - pH HSO4- 3
p[FeSO4+] = pK + p[SO4
2-] + p[Fe3+] SO42- -1.9
p[FeHSO42+] = pK + p[HSO4
-] + p[Fe3+] H2SO4 0
p[SO42-] = pK + pHSO4
- - pH FeSO4+ 4.04
FeHSO42+ 2.48
3.5 Iron(II) Oxidation Rate Modeling
Zero-order oxidation rates were calculated from the pseudo-steady states of each
residence time during the first variable-residence time experiment and the batch reactors.
Lucas (2008) determined that zero order Fe(II) oxidation rates had a better correlation
than first order rates in batch reactors with sediment and AMD from Gum Boot and
Fridays-2.
The gutter reactors were modeled as follows:
32
Accumulation = in – out ± rxn
Gutter reactors QinCin QoutCout
0 = (QC)in – (QC)out – koV
Since: Qin = Qout, V/Q = t, and accumulation = 0:
k = ∆C / ∆t, where C is in moles, t is in seconds, and k is in mol/L-s.
Williamson et al., (2006) presented Fe(II) oxidation rates in molal/s, but since 1 L is
assumed to equal 1 kg, molar and molal oxidation rates were assumed to be the same.
33
4. RESULTS
4.1 Hughes Borehole Chemistry Data
Characterization of the geochemistry of the iron mound at Hughes Borehole was
conducted via transects of samples taken from the emergent and effluent end locations of
the mound. Modifications of the existing iron mound were tested with the channel
reactors.
4.1.1 Transect Chemistry
Three flow path sampling campaigns were conducted at Hughes Borehole on
August 21, 2007, December 7, 2007, and May 22, 2009. The B transects correspond to
the main channel and the C and D transects correspond to sections where the AMD
spread out across the mound as thin sheet flow (Figure 4-1). Compared to the consistent
location and depth of the main channel, the sheet flow sections were variable and
changed from month to month. At the time of the D transect, in May 2009, the C transect
sheet flow section and the furthest point from the 2007 B transect, B7-5, were not
flowing. The areas without flow during the May 2009 transect are designated as orange
in Figure 4.1. Location D5, 201 feet from the source, was the effluent end location of the
mound.
In both B transects in 2007, dissolved Fe(II) and pH decreased with distance
(Figure 4-2). In the C transect in December 2007, there was a sharp drop in dissolved
Fe(II) at point C5, but this did not occur in August 2007. However, the pH decreased
both times from 4.0 to around pH 3.4 at this location. In the other sheet-like section, the
34
D transect from May 2009, dissolved Fe(II) did not significantly decrease, and the pH
decreased only slightly with distance. Likewise, in the B transect from May 2009,
dissolved Fe(II) and pH did not decrease.
The overall dissolved Fe(II) concentration from the source decreased from 102
mg/L in 2007 to 60 mg/L in 2009. The pH also decreased from 4.10 in 2007 to 3.83 in
2009. In all three sampling campaigns, the DO increased with distance from the source
from about 1 mg/L to 11 mg/L at the farthest point, D5.
35
B7-1 (27 ft)
B7-2 (36 ft)
B9-2, B7-3 (46 ft)
B7-5 (65 ft)
B7-4 (60 ft)
B9-1 (16 ft)
B9-3 (76 ft)
A (O ft) Borehole
C4 (93 ft)
C2 (62 ft)
C3 (75 ft)
C1 (54ft)
C5 (101 ft)
D1 (97 ft)
D3 (153 ft)
D2 (106 ft)
D4 (161 ft)
D5 (201 ft)
N
Figure 4-1. Map of Hughes Borehole showing the sampling campaign locations with corresponding distances from the borehole or source. The B transect designates channel flow and the C and D transects represent sheet flow. The C section was sampled twice in 2007, in August and December, and the D section was sampled in May 2009. The B transects were sampled in both years and the points are labeled with either a 7 or 9 to designate the years 2007 and 2009, respectively.
36
0
20
40
60
80
100
120
0 30 60 90 120 150 180 210
Dis
solv
ed F
e(II)
(mg/
L)
B transect 8-07 B transect 12-07 B transect 5-09
C transect 8-07 C transect 12-07 D transect 5-09
3.0
3.4
3.8
4.2
0 30 60 90 120 150 180 210
pH
0
2
4
6
8
10
12
0 30 60 90 120 150 180 210
Distance from source (ft)
DO
(mg/
L)
Figure 4-2. Transect data showing Fe(II) concentrations, pH, and dissolved oxygen(DO) for separate sampling dates as function of distance from the source. The B transect represents channel flow and the C and D transects represent sheet flow.
37
X-ray diffraction (XRD) was conducted at three locations on the iron mound
including a terrace, the main channel, and a pool (Figure 4-3). The terrace and pool
locations were sampled from the D transect and the main channel was sampled from the
B transect. Based on the XRD patterns, schwertmannite was the predominant iron
mineral at the terrace and pool locations, approximately 100-105 feet downstream of the
emergent discharge. Goethite was more dominant in the main channel, approximately 50
feet downstream of the emergent discharge.
Figure 4-3. XRD patterns from three locations on the iron mound at Hughes Borehole. The top black line is from a terrace, the middle red line is from the main channel, and the bottom blue line is from a pool.
38
4.1.2 Fence and Toe Observations
Two locations on the Hughes Borehole iron mound were chosen to provide a
representation of the processes occurring across the entire mound. The fence location is
located directly adjacent to the fence around the borehole and is approximately 8 feet
downstream of the emergent discharge. The toe location is the effluent end of the
mound, or where “iron waterfalls” convey the AMD into a large pool that eventually
flows into the Little Conemaugh River. The exact location of the toe varied throughout
the sampling period due to fluctuations of the natural flow regime of the mound, but
despite these changes it was always the effluent end of the mound. The hydraulic
residence time of the mound was not measured due to the multiple and variable flow
paths from the fence to the toe locations. It was estimated that the residence time of the
main channel section was on the order of minutes and the residence times of the sheet
flow sections were on the order of tens of minutes. Therefore, an educated guess for the
average residence time from the fence to the toe locations ranged from 10 – 30 minutes.
A multitude of analyses were conducted at the fence and toe locations. Dissolved
Fe(II) concentrations, pH, DO, conductivity, and temperature were measured at various
times from July 2008 to June 2009, at an average of approximately two sample points per
month, with higher frequency in the warmer months (Table 4-1). In addition, dissolved
Fe(III) was analyzed for half of these samples. Dissolved sulfate (SO42-) was also
analyzed and acidity was calculated from Eq. 3.1, but with less frequency. In addition,
NPOC, TN and Reactive PO43-, were measured on three occasions. Furthermore, average
deposition of the fresh, or new, metal hydroxide sediments onto the iron mound surface
was measured by placement of glass slides on the existing surface and determined to be
approximately 0.6 inches over a 1.5 month period, or 0.01 inches per day.
39
Trace metal concentrations, including Al, As, Ca, Co, Cr, Cu, Fe, K, Mg, Mn, Na,
Ni, Pb, Si, Sr, Ti, and Zn, were measured for 10 selected samples from July 2008 to
March 2009 (Table 4-1). Elemental analysis of the iron mound sediment at both
locations was also measured to determine its metal oxide composition (Table 4-2).
Daily flow rate measurements, along with rainfall data, and total Fe load
calculations from Hughes Borehole are presented in Figure 4-4. The data was only
available for about half of the study period, from July 2008 to February 2009. Data
dating back to October 2006 was available and was employed to observe overall flow rate
trends. The flow at Hughes ranged from approximately 300 to 2000 gallons per minute
(gpm) since 2006. During the study period the flow ranged from 300 to 1000 gpm. A
trend of increasing flow was evident from December 2008 until the last available date of
data.
On average, the pH decreased across the mound from 3.96 at the fence to 3.50 at
the toe. The dissolved oxygen and conductivity both increased from 1.13 to 9.94 mg/L
and 1,080 to 1,150 µS/cm, respectively. The average water temperature increased
throughout the year, but this was dependent on the air temperature and amount of
sunlight during the sampling events. The average temperature at the fence was 12.7 oC
and was relatively constant year round.
The overall stoichiometric relationship for autotrophic biological Fe(II) oxidation
(Smith et al., 1988), predicts that 0.0011 mM NH4+ is utilized for every mole of Fe2
+
oxidized (see Eq. 1.7). For Hughes Fe(II) concentrations of 100 mg/L, or 1.8 mM, this
would require 0.04 mg NH4+ /L, which is much less than the 1.0 mgN/L available at
Hughes. In addition, they used a Fe2+:PO43- ratio of 558:1 to account for the amount of
40
PO43- needed for cell synthesis. At this ratio, 0.18 mg PO4
3-/ L is needed for complete
Fe(II) oxidation at Hughes Borehole which is less than the measured average of 0.59
mgPO43-/ L. Assuming that both N and P are in forms that bacteria can utilize, neither
should be limiting. Total inorganic carbon measurements were not measured, so the
amount of available TIC was not known. However, Lucas (2008), found at both
Gumboot and Fridays-2 AMD sites, TIC was not limiting for autotrophic iron oxidizing
bacteria. Finally, these calculations are for autotrophic microorganisms and do not
account for the possible consumption of organic carbon for heterotrophic
microorganisms.
41
0
500
1000
1500
2000
2500
10/1/06 12/31/06 4/1/07 7/1/07 10/1/07 12/31/07 3/31/08 7/1/08 9/30/08 12/30/08 4/1/09 7/1/09
Dai
ly F
low
(gpm
)
0
2
4
6
8
10
12
14
Mon
thly
Rai
nfal
l (in
) .
Flow (gpm)
Rainfall (in)
0
200
400
600
800
1000
1200
7/3/08 7/24/08 8/14/08 9/4/08 9/25/08 10/16/08 11/6/08 11/27/08 12/18/08 1/8/09 1/29/09 2/19/09
Dai
ly F
low
(gpm
)
0
100
200
300
400
500
600
Fe L
oad
(kg/
d)
Flow (gpm)
Fe Load (kg/d)
0
500
1000
1500
2000
2500
10/1/06 12/31/06 4/1/07 7/1/07 10/1/07 12/31/07 3/31/08 7/1/08 9/30/08 12/30/08 4/1/09 7/1/09
Dai
ly F
low
(gpm
)
0
2
4
6
8
10
12
14
Mon
thly
Rai
nfal
l (in
) .
Flow (gpm)
Rainfall (in)
0
200
400
600
800
1000
1200
7/3/08 7/24/08 8/14/08 9/4/08 9/25/08 10/16/08 11/6/08 11/27/08 12/18/08 1/8/09 1/29/09 2/19/09
Dai
ly F
low
(gpm
)
0
100
200
300
400
500
600
Fe L
oad
(kg/
d)
Flow (gpm)
Fe Load (kg/d)
Figure 4-4. Daily flow rate data recorded from the pressure transducer at Hughes Borehole. Upper panel contains daily flow values for 2 ½ years and average monthly rainfall data from Johnstown, PA. The red lines indicate the available flow data from the study period. The Lower panel displays the flow rate data for the study period with the Fe load calculated from the specific sampling dates.
42
Table 4-1. Water quality parameters from the fence and toe locations at Hughes Borehole. The standard deviation (± #), range (# - #), and number of samples (N) are given for each parameter.
pH Temp DO Conduct. Fe(II) Fe(III) Fe* Acidity(oC) (mg/L) (µS/cm) (mg/L) (mg/L) (mg/L) (mg/LCaCO3)
3.96 ± 0.25 12.7 ± 0.6 1.10 ± 0.80 1080 ± 99 100.2 ± 17.4 0.5 ± 2.31 88.5 ± 3.3 228 ± 4.9 Fence (3.49 - 4.38) (11.4 - 13.7) (0.28 - 3.46) (915 - 1213) (63.3 - 124.8) (0.0 - 4.5) (85.3 - 94.8) (222 - 236.1)
N = 29 N = 30 N = 21 N = 23 N = 25 N = 15 N = 7 N = 73.52 ± 0.33 16.0 ± 3.8 9.99 ± 1.22 1140 ± 126 66.7 ± 23.0 14.6 ± 11.7 66.6 ± 16.2 200.2 ± 21.2
Toe (2.98 - 4.24) (9.8 - 22.1) (7.86 - 11.38) (947 - 1324) (31.6 - 102.3) (0.0 - 34.3) (48.0 - 87.5) (167.3 - 220.9)N = 22 N = 21 N = 16 N = 17 N = 21 N = 14 N = 7 N = 7SO4
2- NPOC TN PO43- Al Mn Ca Co
(mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L)573 ± 63 0.868 ± 0.22 1.00 ± 0.14 0.59 ± 0.30 8.19 ± 0.39 2.42 ± 0.02 113 ± 4 0.19 ± 0.01
Fence (500 - 640) (0.66 - 1.10) (0.86 - 1.14) (0.25 - 0.78) (7.47 - 8.57) (2.37 - 2.46) (110- 120) (0.18 - 0.19)N = 5 N = 3 N = 3 N = 3 N = 7 N = 7 N = 7 N = 7
572 ± 53 8.25 ± 0.44 2.48 ± 0.03 113 ± 3 0.19 ± 0.01Toe (510 - 620) (7.55 - 8.73) (2.45 - 2.52) (110 - 115) (0.18 - 0.19)
N = 5 N = 7 N = 7 N = 7 N = 7K Mg Na Ni Si Sr Zn
(mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L)6.23 ± 0.17 41.5 ± 0.95 8.53 ± 1.33 0.40 ± 0.01 17.7 ± 0.24 0.50 ± 0.01 0.27 ± 0.01
Fence (6.00 - 6.58) (40.1 - 43.2) (7.63 - 11.50) (0.38 - 0.42) (17.2 - 18.0) (0.47 - 0.51) (0.25 - 0.29)N = 7 N = 7 N = 7 N = 7 N = 7 N = 7 N = 7
6.23 ± 0.23 41.9 ± 0.87 8.21 ± 0.28 0.41 ± 0.01 18.1 ± 0.23 0.50 ± 0.01 0.27 ± 0.01Toe (6.06 - 6.75) (40.4 - 42.8) (7.88 - 8.71) (0.39 - 0.43) (17.8 - 18.4) (0.49 - 0.50) (0.26 - 0.29)
N = 7 N = 7 N = 7 N = 7 N = 7 N = 7 N = 7 *Concentration determined by ICP-AES.
43
Table 4-2. Elemental analysis from the top 2 cm of sediment from the fence and toe locations at Hughes Borehole. Metal oxide values are in weight percent (%) of the original sample.
Metal Oxide (%) Fence Toe
Al2O3 0.39 1.28 CaO 0.04 0.03
Fe2O3 66.7 67.0 MgO 0.18 0.16 P2O5 0.09 0.23 TiO2 0.08 0.08
< 0.01 or
non-detect
< 0.01 or
non-detect BaO, CoO, Cr2O3, K2O, MnO,
Na2O, NiO, SiO2, TiO, ZnO Loss on Ignition (1000 oC) 32.50 31.20
Every sampling event showed a decrease in dissolved Fe(II) at the toe location as
compared to the fence (Figure 4-5). In August and September 2008, there was 40-50 %
oxidation of Fe(II) across the mound. Little Fe(II) oxidation occurred during all other
times of the year. During the times of significant Fe(II) oxidation, the Fe(III)
concentrations at the toe location increased and therefore the total Fe removal was not as
efficient (Figure 4-6). Additionally, the dissolved Fe(III) concentrations at the fence
location were very low throughout the entire year. The available daily flow data from
Hughes Borehole shows a large decrease in flow from September 2008 to January 2009.
However, this decrease does not match the trend of increased Fe(II) oxidation from the
fence to the toe and therefore was probably not the primary reason for the increased
Fe(II) oxidation during August and September 2008.
A trend of decreasing dissolved Fe(II) concentration was evident from December
2008 to the June 2009, the last month of sampling. An increase in flow started around
44
the same time as the Fe(II) decrease, but since flow data after February 2009 was not
available, the influence of this increase in flow on the decrease in Fe(II) can not be
determined (Figure 4-4). The concentration of Fe did not change substantially from July
3, 2008, to January 19, 2009, despite the large changes in flow rate. Thus, the daily total
dissolved Fe load ranged from about 150 to 500 kg/d and followed the same trend as the
emergent discharge flow rate.
45
0
20
40
60
80
100
120
7/3/08
7/31/0
8
8/28/0
8
9/25/0
8
10/23
/08
11/20
/08
12/18
/08
1/15/0
9
2/12/0
9
3/12/0
94/9
/095/7
/096/4
/097/2
/09
Dis
solv
ed F
e(II)
(mg/
L)
Fence Toe
0
200
400
600
800
1000
7/3/08
7/31/0
8
8/28/0
8
9/25/0
8
10/23
/08
11/20
/08
12/18
/08
1/15/0
9
2/12/0
9
3/12/0
94/9
/095/7
/096/4
/097/2
/09
Dai
ly F
low
(gpm
)
Flow (gpm)
Figure 4-5. Upper panel: Fence and toe dissolved Fe(II) concentrations versus calendar date representing the emergent and effluent ends of the Hughes Borehole iron mound. Lower panel: Available daily flow measurements from the pressure transducer weir at Hughes Borehole versus the same calendar dates as the sampling events.
46
Fence
0
20
40
60
80
100
120D
isso
lved
Fe
(mg/
L)Fe(III)
Fe(II)
Toe
0
20
40
60
80
100
120
7/22
/08
8/7/
088/
14/0
89/
4/08
9/18
/08
9/25
/08
10/3
/08
10/9
/08
11/7
/08
12/5
/08
1/19
/09
2/20
/09
4/3/
096/
23/0
97/
3/09
Dis
solv
ed F
e (m
g/L)
Figure 4-6. Dissolved Fe(II) and Fe(III) concentrations at the fence and toe locations of Hughes Borehole. Fe(II) is in green and Fe(III) is in red.
Despite the removal of total Fe across the mound, all other aqueous trace metals
showed no reduction in concentration across the mound (Table 4-1 and Figure 4-7). The
concentrations of dissolved Al and Mn, two of the main contaminants of concern in
AMD, actually increased slightly at the toe location. Iron oxide, Fe2O3, was the most
47
abundant metal oxide by weight percent in the iron mound sediment for both the fence
and toe locations at 66.7 and 67.0 %, respectively (Table 4-2). The second most
abundant metal oxide was Al2O3, at 0.39 and 1.28 %, for the fence and toe, respectively.
Even though the aqueous grab sample data showed that dissolved Al increased slightly at
the toe location, the elemental analysis of the sediment showed that a little aluminum did
precipitate. All other trace metals were either non-detect or present in very low
concentrations.
0.01
0.1
1
10
100
1000
Al Ca Co Fe K Mg Mn Na Ni Si Sr Zn
Dis
solv
ed C
onc.
(mg/
L)
Fence Toe
0.7
0.8
0.9
1
1.1
Al Ca Co Fe K Mg Mn Na Ni Si Sr Zn
[Me]
_toe
/ [M
e]_f
ence
0.01
0.1
1
10
100
1000
Al Ca Co Fe K Mg Mn Na Ni Si Sr Zn
Dis
solv
ed C
onc.
(mg/
L)
Fence Toe
0.7
0.8
0.9
1
1.1
Al Ca Co Fe K Mg Mn Na Ni Si Sr Zn
[Me]
_toe
/ [M
e]_f
ence
Figure 4-7. Top figure: Dissolved metal concentrations at the fence (orange) and toe (blue) locations of Hughes Borehole for ten sampling events from July 2008 to March 2009. Bottom figure: Normalized metal concentrations from the fence and toe. As, Cr, Cu, Pb, and Ti were also measured but all concentrations were <0.01 mg/L, or non-detect.
48
4.2 On-Mound Channel Reactor Data
The on-mound channel reactors were constructed in June 2008 and sampling
began on July 3, 2008. On average, the channels were sampled every other week for the
fiscal year from July 2008 to July 2009. Initially, no modifications were made to the
channels besides the addition of iron mound sediment. On July 23, 2008, wood “steps”
were added to increase the residence time within the channels. Four blocks (2x6 pressure
treated wood beam cut to 10.25” wide) each were added to channels A, B, C, D, E and F,
and the treatment channels, G and H. Blocks were placed roughly 10 feet apart within
the channels. On August 7, 2008 additional steps were added in an attempt to create
greater differences in the residence times within the channels; 2 more steps were added to
channels D and E, and 4 more steps were added to channels E and F. Channels G and H
received no further modifications throughout the entire experiment.
Pressure treated wood has been shown to potentially contain toxic chemicals,
including chromated copper arsenate, and leaching of these chemicals from the wood
may impair biological communities and aquatic ecosystems (Stook et al., 2005).
However, due to the long-term life of the channel reactors, pressure treated wood was
chosen to prevent decay of the wood structure.
On October 9, 2008 Brentwood cross-flow plastic trickling filter media was added
to the six treatment channels. Each of these six received 30 pieces of media cut to 4” x
10.25” x 12” and the water depth in the channels covered approximately 3” of the media
for an effective surface area of 3,690 ft2 per channel from the plastic media alone.
On May 6, 2009, all the plastic media in channels C, D, E, and F were removed,
and Rolanka landscaping coconut fiber (coir) mats were added to channels E and F. The
mats covered 35 feet of the 40 foot channel length and were ¼” in depth. In addition,
49
half of the plastic media was removed from channels A and B. On May 22, 2009,
Rolanka coconut fiber erosion control “netting” was added to channels C and D in place
of the coir mats. The netting had a 48% open area and covered 35 feet of the 40 foot
channels.
The measurements for the entire on-mound channel reactor experiments are
presented in Figure 4-8. Period I refers to the period with no steps. Period II refers to the
period when all the steps were installed. Period III refers to the physical treatment period
with the plastic media. Period IV refers to the chemical treatment period with coir mats
and netting.
There was no significant decrease in dissolved Fe(II) effluent concentrations for
any of the channels until November 13, 2008, over a month after the plastic media was
added. None of the individual treatment channel sets, A-B, C-D, and E-F, were more
effective at Fe(II) oxidation than the other treatment channels until April 3, 2009. After
this date, channels A-B showed slightly better Fe(II) oxidation, especially on May 22,
2009. The May 6, 2009 sampling event occurred directly after the removal of the plastic
media and a clear increase in effluent Fe(II) was evident.
The results for only period IV when the coir was added to the reactors are
presented in Figure 4-9. The addition of the coir mats in channels E and F did not
significantly enhance the Fe(II) oxidation. However, the coir netting in channels C and D
helped oxidize 40-50% of the Fe(II) and performed the most effectively out of all the
channel sets during June and July 2009.
The actual Fe(II) concentration from the emergent discharge showed a continuous
decline after April 3, 2009. The Fe(II) concentration stayed relatively constant at around
50
100 mg/L from July 2008 to March 2009, then decreased to approximately 55 mg/L in
June 2009. The reason for this decrease is currently unknown, but may be inversely
related the increase in groundwater flow into the mines that feed Hughes Borehole.
The coconut fiber initially increased the dissolved non-purgeable organic carbon
(NPOC) of the treatment effluent by a factor of two compared to the control reactors, but
only slightly increased the dissolved total nitrogen (TN) concentration (Figure 4-10).
After 8 weeks of treatment, there was no significant increase in NPOC or TN in the
treatment reactors.
The pH, DO, conductivity, and temperature were measured at each sampling
event for the influent box and the effluent ends of the eight channels (Table 4-3). The pH
decreased across each channel during all four periods. Overall, pH decreased from 3.80
to 3.47 in the treatment channels (A-F), and from 3.80 to 3.55 in the control channels (G,
H). Dissolved oxygen increased throughout every channel with the largest increase, from
2.28 to 7.99 mg/L, in the control channels (G, H). Conductivity increased slightly from
1,110 to 1,120 µS/cm, with no clear trend in any individual set of channels. Temperature
generally increased during the warmer months.
51
0.2
0.4
0.6
0.8
1.0
1.2
Fe(II
)_ou
t / F
e(II)
_in
Figure 4-8. Stacked plot with dissolved Fe(II) concentrations for the channel reactors at Hughes Borehole. Top figure presents the normalized Fe(II) concentrations for the average of the treatment and control channels. Bottom figure plots the actual Fe(II) concentration for the influent, treatment controls, and control channels. Period I refers to the period with no steps. Period II refers to the period when all the steps were installed. Period III refers to the physical treatment period with the plastic media. Period IV refers to the chemical treatment period with coir mats and netting. The red dashed line indicates no change in Fe(II)out / Fe(II)in.
Treatment
Control
0
20
40
60
80
100
120
140
7/3/
087/
8/08
7/17
/08
7/23
/08
8/7/
088/
14/0
88/
21/0
88/
28/0
89/
4/08
9/12
/08
9/18
/08
9/25
/08
10/3/
0810
/9/08
11/7/
0811
/13/0
812
/5/08
3/20
/09
4/3/
095/
6/09
5/22
/09
6/23
/09
7/3/
09
Fe(II
) (m
g/L)
Influent
Treatment
Control
I IVIIIII
I IVIIIII
0.2
0.4
0.6
0.8
1.0
1.2
Fe(II
)_ou
t / F
e(II)
_in
Treatment
Control
0
20
40
60
80
100
120
140
7/3/
087/
8/08
7/17
/08
7/23
/08
8/7/
088/
14/0
88/
21/0
88/
28/0
89/
4/08
9/12
/08
9/18
/08
9/25
/08
10/3/
0810
/9/08
11/7/
0811
/13/0
812
/5/08
3/20
/09
4/3/
095/
6/09
5/22
/09
6/23
/09
7/3/
09
Fe(II
) (m
g/L)
I IVIIIIII IVIIIII
I IVIIIIII IVIIIII
Influent
Treatment
Control
52
0.0
0.2
0.4
0.6
0.8
1.0
1.2
5/6/09 5/22/09 6/23/09 7/3/09
Fe(II
)_ou
t / F
e(II)
_in
Plastic Media (A, B) Coir netting (C, D)Coir mat (E, F) Control (G, H)
Figure 4-9. Normalized dissolved Fe(II) concentrations for the separate on-mound channel reactors sets at Hughes Borehole during the coir period (IV). Plastic media was left in channels A-B, whereas channels C-D, and E-F received coir netting and coir mat, respectively. The red dashed line indicates no change in Fe(II)out / Fe(II)in.
53
Table 4-3. Dissolved oxygen (DO), pH, temperature, and conductivity for the channel reactors and influent splitter box at Hughes Borehole. Period I refers to the period with no steps. Period II refers to the period when all the steps were installed. Period III refers to the physical treatment period with the plastic media. Period IV refers to the chemical treatment period with coir mats and netting.
7/3/08 - 8/7/08 8/14/08 - 10/3/08 10/9/08 - 3/19/09 4/3/09 - 7/3/09DO (mg/L) I II III IV
1.38 ± 0.32 2.35 ± 1.45 1.78 ± 0.28 2.34 ± 0.76Influent (1.00 - 1.75) (1.32 - 3.37) (1.51 - 2.15) (1.62 - 3.41)
N = 5 N = 2 N = 4 N = 46.54 ± 0.92 6.60 ± 1.94 3.69 ± 1.13 3.81 ± 1.68
Treatment (4.74 - 8.35) (4.22 - 8.94) (1.39 - 5.51) (0.56 - 6.01)N = 30 N = 12 N = 21 N = 24
7.48 ± 0.56 8.46 ± 1.48 8.04 ± 1.74 9.58 ± 0.78Control (6.84 - 8.60) (6.82 - 9.8) (4.00 - 9.23) (8.49 - 10.88)
N = 10 N = 4 N = 8 N = 8
pH I II III IV3.98 ± 0.12 3.80 ± 0.19 3.85 ± 0.09 3.80 ± 0.10
Influent (3.86 -4.13) (3.53 - 4.08) (3.78 -3.97) (3.71 - 3.94)N = 5 N = 7 N = 6 N = 4
3.84 ± 0.10 3.38 ± 0.19 3.39 ± 0.18 3.36 ± 0.15Treatment (3.67 - 4.05) (3.00 - 3.76) (3.15 - 3.78) (3.07 - 3.66)
N = 30 N = 42 N = 32 N = 303.82 ± 0.13 3.33 ± 0.23 3.62 ± 0.21 3.61 ± 0.05
Control (3.59 - 4.04) (2.97 - 3.73) 3.22 - 3.85) (3.54 - 3.69)N = 10 N = 14 N = 12 N = 10
Temp. (oC) I II III IV13.3 ± 0.4 13.4 ± 1.1 12.4 ± 1.3 13.0 ± 0.1
Influent (12.8 - 13.8) (11.6 - 15) (10.4 - 14.3) (12.8 - 13.1)N = 4 N = 7 N = 6 N = 5
14.8 ± 1.0 13.9 ± 2.7 11.9 ± 2.8 15.0 ± 1.5Treatment (13.4 - 16.2) (9.2 - 17.5) (4.6 - 16.6) (13.1 - 18.0)
N = 30 N = 36 N = 33 N = 3315.7 ± 1.9 14.2 ± 2.5 12.1 ± 3.3 15.5 ± 1.7
Control (13.6 - 18.1) (9.9 - 17.0) (6.3 - 16.4) (13.5 - 18.7)N = 10 N = 12 N = 12 N = 12
Conduct. (µS/cm) I II III IV1194 ± 93 1136 ± 69 1110 ± 102 1071 ± 37
Influent (1073 - 1300) (999 - 1181) (989 - 1215) (1050 - 1137)N = 4 N = 6 N = 5 N = 5
1175 ±67 1163 ± 95 1159 ± 138 1134 ± 60Treatment (1067 - 1248) (998 - 1309) (914 - 1360) (1063 - 1254)
N = 24 N = 36 N = 33 N = 361180 ± 68 1170 ± 97 1094 ± 128 1068 ± 53
Control (1076 - 1263) (1012 - 1307) (854 - 1283) (936 - 1140)N = 8 N = 12 N = 12 N = 12
Experiment Period and Dates
54
0
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Influent Treatment Control
Dis
solv
ed C
onc.
(mg/
L)
NPOC: Start Coir Mat
NPOC: After 8 weeks
TN: Start Coir Mat
TN: After 8 weeks
Figure 4-10. Dissolved non-purgeable organic carbon and total nitrogen concentrations from the channel reactor at Hughes Borehole during the coconut fiber (coir) treatment phase, period IV.
The residence time in the channels ranged from approximately 2 minutes in the
control channels to 33 minutes in the treatment channels (Figure 4-11). The longest
residence times were recorded on November 7, 2008 with the addition of the plastic
media. The removal of the plastic media and the addition of coir decreased the residence
times of each set of channels. The variability of the mean residence time in channels G
and H, which received no modifications, followed the same trend as the other channels.
There were no modifications conducted in any channel during period I and thus the
residence time of each channel set was very similar. The relative increase in residence
time (θtreatment / θcontrol) for the step period (period II) for channels A-B (4 steps), C-D (8
55
steps), and E-F (6 steps), was 3.6, 9.6, and 5.6, respectively. The relative increase in
residence time for the plastic media period (period III) in A-B, C-D, and E-F, was 6.3,
7.4, and 6.1, respectively. The relative increase in residence time for the coir period
(period IV) for A-B, C-D, and E-F was 6.3, 10.5, and 5.3 respectively.
10.7
18.2
6.8
1.9
26.9
32.5
27.6
4.4
12.1
24.3
14.6
2.3
2.5
2.5
2.4
2.6
0 5 10 15 20 25 30 35
Treat. (A, B)
Treat. (C, D)
Treat. (E, F)
Ctrl (G, H)
Hydraulic Residence Time (min)
7/8/2008 I8/21/2008 II11/7/2008 III6/23/2009 IV
Figure 4-11. Residence time of the channel reactors at all four periods of the experiment, no modifications (I), step period (II), plastic media period (III), and coconut fiber period (IV). 4.3 Laboratory-Scale Gutter Reactors Results
Laboratory gutter reactors were used to test the modifications to the Hughes
Borehole iron mound in a more controlled setting. These experiments included two
variable-residence time experiments, a coconut fiber experiment, and a carbon dioxide
(CO2) experiment.
56
4.3.1 Variable-Residence Time Experiment
Experiments were conducted in the laboratory gutters to study longer residence
times than were possible with the on-mound reactors. The two treatment reactors
contained sediment and the control reactors had no additions besides the non-sterile
AMD. The experiment was conducted sequentially at residence times of 10, 5, 2, and 1
hour, starting at the longest residence time and finishing at the shortest residence time.
The flow rates in each reactor for these residence times were 2.08, 1.04, 0.415, and 0.207
mL/min, respectively and were verified by gravimetric analysis. The flow rates
controlled the residence times because the volumes and slope of the reactors remained
constant throughout all the gutter reactor experiments.
The first experiment started with a 10 hour residence time and continued until a
pseudo-steady state rate of Fe(II) oxidation had been reached. After this time, the
residence time was decreased to 5 hours and the system was subsequently challenged
with shorter times (Figure 4-12). In the treatment gutter reactors, an acclimation period,
where the Fe(II)out was greater than the Fe(II)in occurred at the onset of the experiment
and lasted for approximately thirty 10 hour residence times (300 hours) before the system
reached a steady state. At this point, 97% of the influent Fe(II) was oxidized during the
10 hour time. During the 5 hour residence time there was a slight increase in effluent
Fe(II) concentration followed by a shorter acclimation period. The effluent Fe(II)
concentration once again decreased which corresponded to a dissolved Fe(II) oxidation
efficiency, calculated as {1 - [Fe(II)out / Fe(II)in]}*100, of approximately 93%. The
system was challenged again and run at a 2 hour residence time. This time the increase in
the effluent concentration was higher than the switch to the 5 hour residence time and the
57
effluent concentration reached a pseudo-steady state condition that amounted to a Fe(II)
oxidation efficiency of 70%. The system was challenged once more at a 1 hour residence
time. There was a large increase in the effluent Fe(II) concentration and then the effluent
concentration reached a pseudo-steady state condition equivalent to a oxidation efficiency
of only 25%.
The control reactors stayed relatively constant throughout the experiment and
there were no distinguishable differences in Fe(II) oxidation efficiency for the different
residence times.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0 20 40 60 80 100 120 140 160 180
Pore Volumes
Fe(II
)_ou
t / F
e(II)
_in
Ctrl 1 Ctrl 2 Sed 1 Sed 2
10 hr 5 hr 1 hr2 hr
Ctrl 1 Ctrl 2 Sed 1 Sed 2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0 20 40 60 80 100 120 140 160 180
Pore Volumes
Fe(II
)_ou
t / F
e(II)
_in
10 hr 5 hr 1 hr2 hr10 hr 5 hr 1 hr2 hr
Figure 4-12. Gutter reactor experiment testing residence times of 10, 5, 2, and 1 hours. The sediment reactors contained sediment from Hughes Borehole and the control reactors did not contain any sediment.
The influent dissolved Fe(II) concentration ranged from about 70 to 100 mg/L
(Figure 4-13). The changes were due to varying Fe(II) concentrations in the stored
Hughes AMD that was used to refill the feed tank. Normalized values calculated by the
58
effluent dissolved Fe(II) concentration divided by the influent dissolved Fe(II)
concentration, are presented on most graphs to account for this fluctuation.
0
20
40
60
80
100
120
140
0 20 40 60 80 100 120 140 160 180
Pore Volumes
Fe(II
) (m
g/L)
Sed 1 in Sed 1 out Sed 2 in Sed 2 out
0
20
40
60
80
100
120
140
0 20 40 60 80 100 120 140 160 180
Fe(II
) (m
g/L)
Ctrl 1 in Ctrl 1 out" Ctrl 2 in Ctrl 2 out
10 hr 5 hr 1 hr2 hr
0
20
40
60
80
100
120
140
0 20 40 60 80 100 120 140 160 180
Pore Volumes
Fe(II
) (m
g/L)
Sed 1 in Sed 1 out Sed 2 in Sed 2 out
0
20
40
60
80
100
120
140
0 20 40 60 80 100 120 140 160 180
Fe(II
) (m
g/L)
10 hr 5 hr 1 hr2 hr10 hr 5 hr 1 hr2 hr
Ctrl 1 in Ctrl 1 out" Ctrl 2 in Ctrl 2 out
Figure 4-13. Actual concentrations of dissolved Fe(II) from the variable-residence time experiment. The upper graph shows the influent and effluent values for the control reactors, and the lower graph shows values for the sediment reactors.
The pH and dissolved oxygen were recorded at most sampling events (Figure 4-
14 and 4-15). The influent pH values changed periodically due to the addition of more
Hughes AMD to the feed tank. The effluent pH of the sediment reactors decreased by 1
pH unit immediately, but only decreased slightly, from 2.85 to 2.60, as the removal rate
of the reactors increased during the 10 hour period. Throughout each experiment, the pH
59
values at the effluent end of the gutter reactors did not drop below 2.6, even at times
when almost 100% of the Fe(II) was oxidized.
The dissolved oxygen in the feed tank stayed fairly constant but the reactor
effluent concentrations decreased slightly as the residence times increased. The control
reactors consistently had 3-4 mg/L higher concentrations of DO than the sediment
reactors.
2.4
2.8
3.2
3.6
4.0
4.4
0 20 40 60 80 100 120 140 160 180
Pore Volumes
pH
Ctrl_in Ctrl_out Sed_in Sed_out
10 hr 5 hr 1 hr2 hr
2.4
2.8
3.2
3.6
4.0
4.4
0 20 40 60 80 100 120 140 160 180
Pore Volumes
pH
Ctrl_in Ctrl_out Sed_in Sed_out
10 hr 5 hr 1 hr2 hr10 hr 5 hr 1 hr2 hr
Figure 4-14. pH measurements from the variable-residence time experiment for both the control and sediment reactors.
60
0
2
4
6
8
10
12
0 20 40 60 80 100 120 140 160 180
Pore Volumes
DO
(mg/
L)Ctrl_in Ctrl_out Sed_in Sed_out Feed
10 hr 5 hr 1 hr2 hr
0
2
4
6
8
10
12
0 20 40 60 80 100 120 140 160 180
Pore Volumes
DO
(mg/
L)Ctrl_in Ctrl_out Sed_in Sed_out Feed
10 hr 5 hr 1 hr2 hr10 hr 5 hr 1 hr2 hr
Figure 4-15. Dissolved oxygen measurements from the variable-residence time experiment for both the control and sediment reactors.
Total Fe was measured at most sampling events to determine the concentration of
dissolved Fe(III) and was analyzed by the ferrozine assay (Figure 4-16). The control
reactors showed little to no change in concentration of dissolved Fe(III) throughout the
experiment; however, the sediment reactors contained higher concentrations of dissolved
Fe(III) with corresponding lower concentrations of dissolved Fe(II).
61
Sediment Reactors
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0 3.1 7.3 12.9
17.6
21.4
31.4
33.6
35.6
39.8
54.5
92.3
100.6
102.4
106.1
116.6
134.4
144.1
146.1
163.1
Tota
l Fe_
out /
Tot
al F
e_in
Control Reactors
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0 3.1 7.3 12.9
17.6
21.4
31.4
33.6
35.6
39.8
54.5
92.3
100.6
102.4
106.1
116.6
134.4
144.1
146.1
163.1
Pore Volumes
Tota
l Fe_
out /
Tot
al F
e_in
Fe(III) Fe(II)
10 hr 5 hr 1 hr2 hr
Sediment Reactors
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0 3.1 7.3 12.9
17.6
21.4
31.4
33.6
35.6
39.8
54.5
92.3
100.6
102.4
106.1
116.6
134.4
144.1
146.1
163.1
Tota
l Fe_
out /
Tot
al F
e_in 10 hr 5 hr 1 hr2 hr10 hr 5 hr 1 hr2 hr
Control Reactors
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0 3.1 7.3 12.9
17.6
21.4
31.4
33.6
35.6
39.8
54.5
92.3
100.6
102.4
106.1
116.6
134.4
144.1
146.1
163.1
Pore Volumes
Tota
l Fe_
out /
Tot
al F
e_in
Fe(III) Fe(II)
Figure 4-16. Dissolved Fe(II) and Fe(III) measurements from the variable-residence time experiment. The experiment reactors are graphed above the control reactors. Fe(III) is in red and Fe(II) is in green.
A suite of dissolved metals including Al, As, Ca, Co, Cu, Cr, K, Mg, Mn, Na, Ni,
Si, Sr, Ti, and Zn were analyzed from samples taken from the reactor effluents and feed
tank during the pseudo-steady state period attained at each residence time (Table 4-5). In
addition, elemental analysis of the iron mound sediment at the effluent end of the
62
sediment gutter reactors was conducted (Table 4-4). Average concentrations are given
for each set of reactors. Fe was the only metal to be removed during any of the four
residence times. Even with the high Fe(II) oxidation efficiencies for the sediment
reactors at 10 and 5 hours, none of the other metals were significantly removed. The
sediment composition shows that iron oxide composed 75% of the weight percent of the
sediment. Some aluminum oxide was also present, but only composed <1% of the total
weight.
Table 4-4. Elemental analysis of the sediment from the effluent end of the sediment reactors following the Coir experiment. The average metal oxide values are in weight percent(%) and the standard deviation (± #) is given for each.
Metal Oxide (%) Average ± S.D.
Al2O3 0.52 ± 0.08
CaO 0.06 ± 0.00
Fe2O3 75.0 ± 0.21
MgO 0.15 ± 0.01
P2O5 0.18 ± 0.02
TiO2 0.08 ± 0.01
BaO, CaO, CoO, Cr2O3, K2O, MnO, Na2O, NiO,
SiO2, ZnO
< 0.01 or
non-detect
Loss on Ignition (1000 oC) 24.1
63
Table 4-5. Dissolved average metal concentrations for the feed tank and effluent of the gutter reactors at pseudo-steady state of each residence time. As, Cr, Pb, and Ti were also analyzed, but all concentrations were non-detect (<0.01 mg/L). The standard deviation (± #) is displayed for the gutter reactors; only one sample from the feed tank was analyzed at each residence time.
Fe Al Mn Ca Co Cu K Mg Na Ni Si Sr Zn10 hour
Influent 94 9.81 2.63 120 0.20 0.07 6.57 44 9.09 0.43 19.0 0.53 0.32Control 77.95 ± 3.75 9.66 ± 0.06 2.67 ± 0.02 122.5 ± 3.5 0.20 ± 0.01 0.07 ± 0.00 6.56 ± 0.04 44.55 ± 0.07 9.26 ± 0.18 0.44 ± 0.01 19.20 ± 0.14 0.54 ± 0.01 0.32 ± 0.01
Sediment 20.30 ± 2.40 15.90 ± 0.14 2.66 ± 0.01 120.0 ± 0.0 0.20 ± 0.01 0.03 ± 0.00 6.68 ± 0.03 45.15 ± 0.35 9.08 ± 0.03 0.44 ± 0.01 23.25 ± 0.21 0.52 ± 0.00 0.30 ± 0.01
5 hourInfluent 92.00 7.99 2.50 110.0 0.18 0.19 6.50 42.10 9.16 0.39 18.00 0.51 0.38Control 92.1 ± 2.69 8.18 ± 0.04 2.60 ± 0.09 112.5 ± 3.5 0.19 ± 0.01 0.22 ± 0.01 6.54 ± 0.06 42.90 ± 0.85 8.96 ± 0.33 0.41 ± 0.01 18.65 ± 0.49 0.52 ± 0.01 0.45 ± 0.01
Sediment 33.5 ± 0.57 9.65 ± 0.05 2.55 ± 0.00 110.0 ± 0.0 0.18 ± 0.00 0.14 ± 0.01 6.43 ± 0.14 42.62 ± 0.21 9.28 ± 0.62 0.40 ± 0.01 18.05 ± 0.35 0.51 ± 0.01 0.40 ± 0.01
2 hourInfluent 84 7.71 2.45 110 0.18 0.03 6.32 41 8.36 0.38 17.6 0.49 0.29Control 82.75 ± 0.35 7.78 ± 0.06 2.48 ± 0.01 110.0 ± 0.0 0.18 ± 0.00 0.03 ± 0.00 6.50 ± 0.06 42.00 ± 0.00 8.85 ± 0.08 0.39 ± 0.00 17.75 ± 0.07 0.51 ± 0.01 0.30 ± 0.01
Sediment 43.6 ± 1.41 7.84 ± 0.01 2.48 ± 0.01 107.5 ± 3.54 0.18 ± 0.01 0.02 ± 0.00 6.42 ± 0.02 41.05 ± 0.49 9.44 ± 0.94 0.38 ± 0.01 17.25 ± 0.07 0.50 ± 0.07 0.29 ± 0.07
1 hourInfluent 82 7.79 2.45 110 0.18 0.17 6.30 42 9.05 0.39 17.5 0.49 0.38Control 77.95 ± 1.48 7.62 ± 0.01 2.42 ± 0.07 110.0 ± 0.0 0.18 ± 0.01 0.17 ± 0.00 6.17 ± 0.02 40.70 ± 0.28 8.51 ± 0.12 0.38 ± 0.01 17.45 ± 0.49 0.49 ± 0.01 0.37 ± 0.03
Sediment 64.3± 9.05 7.76 ± 0.07 2.49 ± 0.03 110.0 ± 0.0 0.18 ± 0.00 0.15 ± 0.01 6.24 ± 0.08 41.45 ± 0.21 8.57 ± 0.19 0.39 ± 0.00 17.65 ± 0.35 0.50 ± 0.01 0.37 ± 0.00
64
In addition to the metals analysis, microbial counts were conducted in the feed tank
and effluent ends during the pseudo-steady state periods for the 10, 5, and 2 hour residence
times. Counts from both aqueous and sediment samples were conducted (Appendix A). The
aqueous microbial numbers for each channel increased during the 5 hour time but decreased
slightly during the 2 hour time. However, the microbial numbers in the feed tank displayed a
similar increase as the numbers from the effluent end. The microbial counts from the
sediment samples increased slightly during the 10 hour time as compared to the initial counts.
4.3.2 Repeat of Variable-Residence Time Experiment
A second residence time experiment (RT2) was conducted that utilized similar
conditions of the first residence time experiment (RT1). Fresh sediment was collected
from the same location on the iron mound as in RT1. However, this time the gutter
reactors were started with a 5 hour residence time and not 10 hour, as in RT1. At this 5
hour time, the sediment reactors reached pseudo-steady state at around eighty-five pore
volumes (400 hours), which corresponded to a Fe(II) oxidation efficiency of 90% (Figure
4-17). The reactors where challenged and the residence time was decreased to 2 hours.
This caused an immediate increase in Fe(II) concentration at the effluent end of the
sediment reactors. However, the effluent Fe(II) oxidation efficiency did not improve
after the initial increase, as was the case in the switch to the 2 hour residence time in
RT1.
65
0
0.2
0.4
0.6
0.8
1
1.2
0 20 40 60 80 100 120 140 160 180
Pore Volumes
Fe(II
)_ou
t / F
e(II)
_in
Ctrl 1 Ctrl 2 Sed 1 Sed 2
5 hr 2 hr
0
0.2
0.4
0.6
0.8
1
1.2
0 20 40 60 80 100 120 140 160 180
Pore Volumes
Fe(II
)_ou
t / F
e(II)
_in
Ctrl 1 Ctrl 2 Sed 1 Sed 2
5 hr 2 hr5 hr 2 hr
Figure 4-17. Dissolved Fe(II) oxidation efficiencies for the repeat of the variable-residence time experiment which was conducted at 5 and 2 hour times. The red dashed line indicates no change in Fe(II)out / Fe(II)in.
The pH was measured at most sampling events for RT2 (Figure 4-18). The pH of
the effluent end of the sediment reactors stayed relatively stable at 2.7 to 2.8 throughout
the entire RT2 experiment even with changes to the Fe(II) effluent concentrations, which
was similar to the pH trend during the 10 and 5 hour times of RT1. The dissolved
oxygen was not measured during this experiment but was assumed to be similar to the
values from RT1.
66
2.50
2.75
3.00
3.25
3.50
0 20 40 60 80 100 120 140 160 180
Pore Volumes
pH
Ctrl in Ctrl out Sed in Sed out
5 hr 2 hr
2.50
2.75
3.00
3.25
3.50
0 20 40 60 80 100 120 140 160 180
Pore Volumes
pH
Ctrl in Ctrl out Sed in Sed out
5 hr 2 hr5 hr 2 hr
Figure 4-18. pH values for the second residence time experiment which was conducted at 5 and 2 hour times.
4.3.3 Coconut Fiber Experiment
Coconut fiber was chosen as suitable organic amendment to test the effect of organic
carbon and nitrogen on Fe(II) oxidation efficiencies in the laboratory scale gutter reactors.
Batch reactors were conducted to determine the effect of various organic amendments on
Fe(II) oxidation efficiency. Leachates were created from three organic amendments;
hardwood mulch, straw, and coconut fiber (see section 3.3.1). The straw leachate had the
highest concentrations of NPOC, TN and PO43- and the mulch had the lowest (Table 4-6).
Zero-order Fe(II) oxidation rates (k) in mol/L-s were calculated for each reactor (see
section 3.5). The mulch reactor had the lowest k value of 6.83E-09 mol/L-s, out of the
three leachate conditions (Figure 4-19 and Table 4-7). The “Live” reactor, the one with
only Hughes Borehole water, had a similar Fe(II) oxidation rate as compared to the coir
67
reactor. The sterile reactor had the lowest Fe(II) oxidation rate out of any of the
conditions with a k value of 9.19E-09 mol/L-s.
Table 4-6. NPOC and TN concentrations for the organic amendment leachates that were used for batch experiments.
Leachate NPOC (mg/L) TN (mg/L) PO4-3 (mg/L)
Mulch 15.0 ± 0.3 2.79 ± 0.17 0.18Straw 403 ± 7.2 19.1 ± 0.19 4.66
Coconut Fiber 43.9 ± 0.3 4.45 ± 0.18 0.31
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
0 5 10 15 20 25
Time (hrs)
Fe(II
)_t /
Fe(
II)_i
nitia
l
Sterile
Mulch
Straw
Live
Coconut Fiber
Figure 4-19. Dissolved Fe(II) oxidation efficiency kinetics for organic amendment batch reactors, including live (no-amendment) and sterile control reactors.
68
Table 4-7. Initial and final pH values and Fe(II) oxidation rates (mol/L-s) for the organic amendment batch reactors.
Condition Initial pH End pH k (mol/L-s) R2
Sterile 3.38 3.03 9.19E-09 0.465Mulch 2.69 2.43 6.83E-09 0.934Straw 4.00 2.50 1.01E-08 0.971Live 3.38 2.52 3.82E-08 0.960
Coconut Fiber 3.63 2.53 2.17E-08 0.991
Fe(II) Removal Rate
The gutter reactors were run to test the effect of the Rolanka coconut fiber mat on
Fe(II) oxidation efficiency in a plug flow reactor. The first experiment consisted of
running the gutters at a residence time of 1 hr with treatment phases involving the
addition and removal of coir to the sediment reactors. The control reactors did not
receive any modifications. In Figure 4-20, Period I corresponds to the gutter reactors at a
1 hour residence time with no coir. Period II corresponds to the addition of coir to the
sediment reactors (referred to as coir reactors in this section). Period III corresponds to
the removal of the coir. Period IV corresponds to the second addition of the same coir
mats from Period II. Period V corresponds to the removal of the coir a second time.
The Fe(II) oxidation efficiency greatly improved with the addition of the coir mat.
The Fe(II) removal efficiency decreased immediately when the coir was removed in
period III, from 97% to 60% oxidation. The second addition of the coir mat, period IV,
required a similar number of pore volumes (~80) to stabilize and reach pseudo-steady
state, but it also demonstrated 97% oxidation of Fe(II). The two periods when the coir
was removed, periods II and V, showed similar oxidation efficiencies (50-60%) as
compared to the original efficiency in period I.
69
0
0.2
0.4
0.6
0.8
1
1.2
0 40 80 120 160 200 240 280
Pore Volumes
Fe(II
)_ou
t / F
e(II)
_in
Controls CoirI II III IV V
0
0.2
0.4
0.6
0.8
1
1.2
0 40 80 120 160 200 240 280
Pore Volumes
Fe(II
)_ou
t / F
e(II)
_iControls
nCoir
I II III IV V
0
0.2
0.4
0.6
0.8
1
1.2
0 40 80 120 160 200 240 280
Pore Volumes
Fe(II
)_ou
t / F
e(II)
_iControls Coir
I II III IV VI II III IV Vn
Figure 4-20. Dissolved Fe(II) oxidation for the gutter reactor experiment with the addition and removal of the coir mat at a residence time of 1 hour. Period I corresponds to the gutter reactors at a 1 hour residence time with no coir. Period II corresponds to the addition of coir to the sediment reactors (referred to as coir reactors in this section). Period III corresponds to the removal of the coir. Period IV corresponds to the second addition of the same coir mats from Period II. Period V corresponds to the removal of the coir a second time. The red dashed line indicates no change in Fe(II)out / Fe(II)in.
The pH and dissolved oxygen were measured throughout this experiment. The
influent pH value was quite variable due to addition of larger volumes of Hughes AMD
to the feed tank due to the higher flow rate (Figure 4-21) needed for the 1 hour residence
time. Nevertheless, the pH in effluent of the coir reactors decreased as the Fe(II)
concentrations decreased. The pH in the effluent of the control reactors was consistently
slightly lower than the influent value, but did not decrease as much as the effluent pH of
the coir reactors.
70
2.7
2.9
3.1
3.3
3.5
3.7
0 40 80 120 160 200 240 280
Pore Volumes
pHCtrl_in Ctrl_out Coir_in Coir_out
I II III IV VI II III IV V
Figure 4-21. pH values for the coconut fiber experiment at the influent and effluent of the control and coir reactors. Period I corresponds to the gutter reactors at a 1 hour residence time with no coir. Period II corresponds to the addition of coir to the sediment reactors (referred to as coir reactors in this section). Period III corresponds to the removal of the coir. Period IV corresponds to the second addition of the same coir mats from Period II. Period V corresponds to the removal of the coir a second time.
Similar to the first variable-residence time experiment, the DO in the effluent of
the control reactors was near saturation and 2-3 mg/L higher than the effluent of the
sediment-containing coir reactors (Figure 4-22).
71
2
4
6
8
10
12
0 40 80 120 160 200 240 280
Pore Volumes
DO
(mg/
L)Ctrl_in Ctrl_out Coir_in Coir_out
I II III IV V
2
4
6
8
10
12
0 40 80 120 160 200 240 280
Pore Volumes
DO
(mg/
LCtrl_in Ctrl_out Coir_in Coir_out
I II III IV VI II III IV V)
Figure 4-22. Dissolved oxygen for the coir fiber experiment at the influent and effluent of the control and coir reactors. Period I corresponds to the gutter reactors at a 1 hour residence time with no coir. Period II corresponds to the addition of coir to the sediment reactors (referred to as coir reactors in this section). Period III corresponds to the removal of the coir. Period IV corresponds to the second addition of the same coir mats from Period II. Period V corresponds to the removal of the coir a second time.
Effluent concentrations of NPOC and TN were measured during period II (Figure
4-23) of the coir experiment. There was a slight increase in NPOC, from 1.3 to 1.4
mgC/L at the effluent of the sediment reactors, as compared to the feed tank, but TN
concentrations stayed relatively constant at around 1 mgN/L. In the control reactors,
NPOC decreased to around 0.4 mgC/L and TN also remained constant.
72
0.0
0.20.4
0.60.8
1.0
1.21.4
1.6
Ctrl 1 Ctrl 2 Coir 1 Coir 2 Feed
Effluent Location
Con
cent
ratio
n (m
g/L)
NPOCTN
Figure 4-23. NPOC and TN concentrations for the gutter reactors and during period II of the coconut fiber experiment.
4.3.4 Carbon Dioxide Purge Experiment
A mixture of 15% CO2 with N2 balance was used to determine if the introduction
of CO2 would affect Fe(II) oxidation rates. The feed tank was purged with the CO2:N2
gas mix and run at a 1 hour residence time. Before the tank was purged with 15% CO2,
the gutters were re-run at a 2 hr residence time under the original experimental
conditions. In Figure 4-24, period I refers to a 2 hour residence time with the N2 purge.
Period II refers to a 1 hour residence time, also with the N2 purge. Period III refers to a 1
hour residence time with the15% CO2 gas mixture. This gas mixture had no significant
affect on the Fe(II) oxidation efficiency at a residence time of 1 hour.
73
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 20 40 60 80 100 120 140
Pore Volumes
Fe(II
)_ou
t / F
e(II)
_in
Control Sediment
I II III
Control Sediment
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 20 40 60 80 100 120 140
Pore Volumes
Fe(II
)_ou
t / F
e(II)
_in I II IIII II III
Figure 4-24. Dissolved Fe(II) oxidation for the experiment with 15% CO2:N2 balance purge of feed tank graphed against number of pore volumes. Period I refers to a 2 hour residence time with the N2 purge. Period II refers to a 1 hour residence time, also with the N2 purge. Period III refers to a 1 hour residence time with the15% CO2 gas mixture. The red dashed line indicates no change in Fe(II)out / Fe(II)in.
4.4 Gutter Reactor Fe(II) Percent Remaining
The dissolved Fe(II) percent remaining, defined as the relative concentration of
Fe(II) retained in the effluent AMD of the reactors, and calculated as {[Fe(II)out /
Fe(II)in]*100, was computed from the pseudo-steady state of each hydraulic residence
time for all of the gutter reactor experiments. The Fe(II) percent remaining was the
converse of the Fe(II) oxidation efficiency. The Fe(II) percent remaining under
“original” conditions for the sediment reactors is presented in Figure 4-25. The original
conditions refer to periods with no modifications to the sediment reactors and with N2
purging of the feed tank. The values were averaged from the duplicate set of sediment
gutter reactors. The minimum Fe(II) percent remaining occurred at a residence time of
74
10 hours and corresponded to 3% remaining. This Fe(II) percent remaining gradually
increased to 75%, at a residence time of 1 hour. These original condition periods during
the carbon dioxide (CD) and coconut fiber (Coir) experiments demonstrated a decrease in
Fe(II) percent remaining at 1 and 2 hours. RT1 and RT2 demonstrated almost identical
values of Fe(II) remaining at 5 and 2 hours.
0
20
40
60
80
100
0 2 4 6 8Hydraulic Residence Time (hrs)
Fe(II
) Rem
aini
ng (%
10
)
RT1 CD
Coir RT2
Figure 4-25. Dissolved Fe(II) percent remaining at pseudo-steady state for varying hydraulic residence times during times of no modifications to the sediment gutter reactors.
The Fe(II) oxidation efficiency of the reactors under original conditions improved
over the course of the experiment. Throughout the RT1 CD and Coir experiments, the
initial Hughes Borehole sediment in the sediment gutter reactors was not completely
replaced. Occasionally 1 gram of sediment was removed for microbial analysis, but the
same volume of fresh sediment was added as compensation. Complete fresh sediment
75
was collected for the repeat of the variable-residence time (RT2) experiment from the
same location of the mound.
The approximate age of the sediments, calculated in days from the time the gutter
reactors were started in the lab, is presented in Table 4-8. The Coir experiment contained
the oldest sediment and had been in use for around 100 days. RT2 had slightly younger
sediment than RT1, but both had been in use for around 20-30 days. The Fe(II) percent
remaining compared to the experimental age of the sediments demonstrates that Fe(II)
oxidation increased with time, but reached a peak at around 90 days for the 1 and 2 hour
times (Figure 4-26).
Table 4-8. Approximate experimental age of sediments, in days, at time of pseudo-steady state for varying hydraulic residence times during times of no modifications for the sediment gutter reactors.
10 hour 5 hour 2 hour 1 hourResidence Time 1 13 27 31 33
Carbon Dioxide 88 90Coconut Fiber 100 102
Residence Time 2 19 21
Residence Time
76
0
20
40
60
80
100
0 20 40 60 80 100 12
Age of sediment (days)
Fe(II
) Rem
aini
ng (%
0
)5 hour 2 hour 1 hour
Figure 4-26. Dissolved Fe(II) percent remaining for the age of sediments during times of no modifications to the sediment gutter reactors.
The Fe(II) percent remaining was compared to the total Fe percent remaining and
is presented in Figure 4-27. Since both variable-residence time experiments had similar
aged sediment, values for those residence times were averaged together. The 1 and 2
hour periods under original conditions for the Coir and CO2 experiments were also
averaged together. Despite the fairly high oxidation of dissolved Fe(II) during most
residence times, total dissolved Fe was not as effectively removed. Each residence time,
with the exception of 1 hour during RT1 and RT2, and 1 hour with the coir mat, had
approximately 25% less total Fe removal as compared to Fe(II) oxidation.
77
0
20
40
60
80
100
10 hour 5 hour 2 hour 1 hour 2 hour 1 hour 1 hour*w/coir mat
Fe R
emai
ning
(%)
Fe(II) Total Fe
Coir & CD Variable residence time RT1 & RT2
Figure 4-27. Dissolved Fe(II) and total dissolved Fe percent remaining for the gutter reactor experiments. All reported values are under “original” conditions, except for the 1 hour w/coir mat. Both variable residence time experiments contained similarly aged sediments whereas the Coir and CO2 experiments contained older sediments.
A comparison of the actual time it took for the fresh sediments to stabilize for 10
hour time in RT1 and the 5 hour time RT2 is presented in Figure 4-28. In RT1, a pseudo-
steady state condition was reached after ca. 300 hours, whereas in RT2 a pseudo-steady
state condition was reached after ca. 400 hours. The 10 hour time in RT1 achieved a
slightly better Fe(II) oxidation efficiency than the 5 hour time in RT2.
78
Figure 4-28. Dissolved Fe(II) oxidation efficiency for initial residence time of both
.5 Modeling Results for Fe(OH)3 Solubility
The solubility of ferric hydroxide, Fe(OH)3, was plotted versus pH for varying
levels of SO4 (Figure 4-29). Fe(III) concentrations with corresponding pH values from
the transects at Hughes Borehole and the first variable-residence time experiment were
also plotted on the graph. The concentration of SO4 in the AMD from Hughes Borehole
was approximately 600 mg/L, which is compatible with the middle orange line in Figure
4-29. The solubility of Fe(OH)3 increases with increasing SO4 , especially as the pH
decreases.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 50 100 150 200 250 300 350 400 450
Time (hrs)
Fe(II
)_ou
t / F
e(II)
_in
10 hr from RT1
5 hr from RT2
residence time experiments, RT1 and RT2. RT1 began with a 10 hour time and RT2began with a 5 hour. The red dashed line indicates no change in Fe(II)out / Fe(II)in.
4
2-
2-
2-
79
0
1
10
100
1,000
10,000
100,000
2.5 3 3.5 4 4.5
pH
Fe(O
H)3
mg/
Lno sulfate
600 mg/L sulfate
6000 mg/L sulfate
Hughes Borehole
RT1
Figure 4-29. Fe(OH)3 solubility versus pH for varying levels of SO4
2-. The figure was created with equilibrium equations and pKa values in Microsoft Excel. Fe(III) concentrations with corresponding pH values are also plotted on the figure. Hughes Borehole refers to fence and toe data, and RT1 refers to the first variable residence time experiment.
4.6 Modeling Results for Fe(II) Oxidation Kinetics
Zero-order Fe(II) oxidation rates were calculated for the control and sediment
gutter reactors during pseudo-steady state of the first variable-residence time (RT1)
experiment and the coir (Coir) experiment (see Section 3.5). The rates are presented in
Table 4-9, and ranged from 8.57 x 10-8 to 3.24x 10-7 mol/L-s in the sediment reactors and
9.24 x 10-10 to 3.07 x 10-8 mol/L-s in the control reactors. The highest Fe(II) oxidation
rate occurred during the coir mat insertion of the Coir experiment. The sediment reactors
had higher Fe(II) oxidation rates than the control reactors for every residence time.
However, the rates for the control reactors at 1 and 2 hour residence times were similar to
the oxidation rates for the sediment reactors at 5 and 10 hour residence times.
80
Table 4-9. Zero-order Fe(II) oxidation rates during pseudo-steady state of the gutter reactors from the first-variable residence time experiment and the insertion of the coir mat in the coir experiment.
Res. Time Sediment Control
10 hr 4.67E-08 3.52E-09
5 hr 8.57E-08 9.24E-10
2 hr 1.71E-07 2.48E-08
1 hr 1.12E-07 3.07E-08
Coir 1 hr 3.24E-07 1.26E-09
Fe(II) Oxidation Rate (mol/L-S)
RT1
81
5. DISCUSSION
The experimental data obtained from this study can be used to help develop
design parameters for a low-pH passive treatment system for AMD impacted areas.
Actual Fe(II) oxidation rates have a high variability, especially in field settings, and
depend on a multitude of factors, with some including dissolved Fe(II) concentration,
dissolved O2 concentration, bacterial concentrations, temperature, pH, organic nutrient
concentrations, and incident sunlight (Barry et al., 1994; Kirby et al., 1999). All of these
factors would have to be considered before a treatment system based on laboratory
experiments could be implemented in a real field setting. Since AMD and iron mound
sediment for the laboratory experiments were collected directly from Hughes Borehole,
many of the factors, such as Fe(II) concentration, bacterial concentrations, and organic
nutrient concentrations, can be assumed to be the same between the laboratory and the
field. However, some conditions such as the temperature and incident sunlight varied
considerably between the laboratory and field conditions. Nevertheless, useful
information was obtained from these experiments that can be utilized to create
remediation strategies for AMD sites.
A common feature of the Hughes Borehole transects and long-term fence and toe
monitoring was the drop in pH across the mound, which is indicative of the formation
and precipitation of Fe(III) hydroxides. Schwertmannite and goethite were present along
the flow path and these Fe(III) minerals are commonly found at iron-rich, high sulfur
mine drainage systems with a pH of 2.8 – 4.5 (Bingham et al., 1996). Since biological
Fe(II) oxidation rates are much greater than abiotic Fe(II) oxidation at pH < 4, it is likely
that biological oxidation is the controlling factor at Hughes Borehole (Williamson et al,
82
2006). Batch reactor data in Figure 5-1 shows that live aerobic reactors had much greater
Fe(II) oxidation rates than live anaerobic reactors and sterile reactors. This further gives
evidence that aerobic biological activity is responsible for much of the Fe(II) oxidation at
Hughes Borehole. Therefore, a passive treatment system focused around promoting
biological treatment instead of abiotic treatment is desirable.
0
10
20
30
40
0 5 10 15 20 25 30Time (hrs)
Dis
solv
ed F
e(II)
(mg/
L)
Filter Sterilized Aerobic Live
1% v/v Formaldehyde Anaerobic Live
Figure 5-1. Batch reactor data for sterile and live reactors with no iron mound sediment. The filter sterilized and 1% v/v formaldehyde reactors were also under aerobic conditions.
The variable-residence time gutter reactor experiments demonstrated that
residence time was an important parameter that controlled the Fe(II) oxidation efficiency.
In the early stages of the experiments, a residence time of at least 5 hours was needed for
almost complete, 90-97%, oxidation of the dissolved Fe(II). The Fe(II) oxidation
efficiencies decreased with decreasing residence time and the 1 hour time only oxidized
about 30% of the influent Fe(II). Furthermore, the age of the sediments in the reactors
seemed to influence the Fe(II) oxidation efficiencies. As the sediments in the channels
83
aged, better Fe(II) oxidation was observed until the sediments reached an age of
approximately 100 days (see Figure 4-26). The oxidation efficiencies for the 1 and 2
hour residence times under original conditions for the Coir and CO2 experiments, which
were conducted with older sediments than RT1 and RT2, showed this correlation of
improved Fe(II) oxidation efficiency versus sediment age. This implies that the biological
communities continued to develop over time and therefore were more capable of
oxidizing Fe(II) in the later stages of the experiments.
In addition, visual observations of the channels showed that metal hydroxide
precipitates formed on the control reactors and side walls of the sediment reactors. As a
result of this, the control reactors were cleaned out in between experiments to remove the
precipitates in an attempt to maintain the controls under original conditions. Since the
control reactors were not cleaned until after the first variable-residence time experiment,
this could explain the better Fe(II) oxidation rates of the control reactors at 1 and 2 hours
during RT1 (see Table 4-9). A passive treatment system where Fe(II) oxidation improves
over time would be favorable because it allows the system to be challenged with shorter
residence times and still have similar Fe(II) oxidation efficiencies. However, the short
acclimation period that followed each subsequent change in residence time would have to
be accounted for in the large-scale passive treatment reactor.
A comparison of the dissolved Fe(II) and total dissolved Fe percent remaining for
the pseudo-steady states for the varying residence times and conditions from the gutter
reactor was developed and is presented in Figure 5-2. For every residence time, the total
Fe percent remaining was higher than the Fe(II) percent remaining. With the exception
of the 1 hour residence time during RT1, all total dissolved Fe percent remaining values
84
were 20-30% greater than the corresponding dissolved Fe(II) percent remaining. A
similar trend of 25% greater total Fe remaining as compared to Fe(II) remaining was
evident for the fence and toe locations at Hughes Borehole. This suggests that much of
the Fe(III) formed from Fe(II) oxidation was not precipitated as ferric hydroxide species.
The high levels of sulfate in the AMD (~600 mg/L) have been shown to increase the
solubility of Fe(OH)3 and could be preventing more precipitates from forming (see
Figure 4-29). The coir mat insertion for the 1 hour residence time showed that only 50%
total Fe remained even though merely 5% of the Fe(II) remained. The coir mat may have
hindered the precipitation of more Fe(OH)3. The longer residence times did not seem to
allow for more Fe(OH)3 precipitation because the 2 hour time during the CO2
experiment had very similar Fe(II) and total Fe percent remaining values as the 5 hour
time during RT1 and RT2, even though there was a 2.5 times difference in residence
time. This is important for a treatment system because it shows that approximately 25%
greater total dissolved Fe than dissolved Fe(II) will be retained in the effluent of the
system at residence times of 10 hours or less.
85
0
20
40
60
80
100
Fence /Toe
10 hour 5 hour 2 hour 1 hour 2 hour 1 hour 1 hour*w/coir mat
Fe R
emai
ning
(%)
Fe(II) Total Fe
Coir & CD Variable residence time RT1 & RT2
Figure 5-2. Dissolved Fe(II) and total dissolved Fe percent remaining for select experiments. The fence/toe data was from August 14, 2008 to September 18, 2008. All gutter reactor measurements are from the sediment reactors at pseudo-steady states and are under original conditions of no modifications and N2 purging of the feed tank, with the exception of the 1 hour w/coir mat.
The two variable-residence time experiments indicate that an acclimation period
was needed for the biological communities to develop before the reactors reached a
steady state of Fe(II) oxidation. The difference in the time of the acclimation period, 300
hours vs. 450 hours, for acclimation periods with 10 hour and 5 hour residence times,
respectively, suggests that the residence time determines how much time is needed for the
reactors to reach steady state. Despite the differences in acclimation period, the initial
and repeated 5 hour and 2 hour residence times in experiments RT1 and RT2, had almost
identical Fe(II) oxidation efficiencies. In addition, both of these time conditions had
similarly aged sediments. This shows that once the biological communities developed,
the residence time controlled the Fe(II) oxidation efficiency for similarly aged sediments.
The design for a passive treatment reactor would need to allow for an acclimation period
86
for the biological communities to develop before the reactors would efficiently oxidize
Fe(II).
The large difference in Fe(II) oxidation efficiency for the control and sediment
gutter reactors shows the importance and necessity of utilizing the existing iron mound
sediments for treatment of AMD. Throughout all the gutter reactor experiments the
control reactors had very low Fe(II) oxidation efficiencies (0-10%) as compared to the
sediment reactors(25-97%). Consequently, a passive treatment reactor that used iron
mound sediments should display much greater Fe(II) oxidation efficiencies than a reactor
without sediment. However, since only one ratio of sediment to AMD, 100 g: 125 mL,
and one water column depth, ¼ in, were investigated with the gutter reactors, the effects
of these two parameters on Fe(II) oxidation efficiency is not known.
For the on-mound channel reactors in the field, residence times of approximately
12-24 minutes during period II in the treatment reactors did not oxidize dissolved Fe(II)
more effectively than residence times of 2-4 minutes in the control reactors when no
other physical or chemical modifications were being carried out (see Figures 4-8 and 4-
11). The physical treatment period with the plastic media, period III, and the chemical
treatment with the coir netting, channels C and D in period IV, showed that even with a
residence time of only 27-30 minutes, approximately 40% oxidation of the dissolved
Fe(II) could be accomplished. Additionally, the Fe(II) oxidation efficiency in the
sediment gutter reactors was greatly improved with the insertion of the coir mat (see
Figure 4-20). Due to the fact that dissolved nutrients, such as NPOC and TN, did not
significantly increase with the coir mats or coir netting, and carbon and nitrogen
concentrations do not seem to be limiting factors, the reason for this increase in
87
efficiency is not readily apparent. However, coupled with the physical modification
channel data, one implication is that an attachment surface could improve the Fe(II)
oxidation efficiency, even at relatively short residence times.
The average Fe(II) percent remaining for the on-mound channel sets (A-B, C-D,
E-F, G-H) were calculated for the four treatment periods and compared to the average
residence time of the corresponding period (Figure 5-3). Since no pseudo-steady state
was reached in each period for the on-mound channel reactors, an average of the Fe(II)
concentrations of all sampling events during each period was calculated. Additionally,
only the residence time was taken into consideration for this comparison. The overall
data trend presents that increased residence decreased the Fe(II) percent remaining
regardless of the type of treatment. Furthermore, at times of 10-35 minutes,
approximately 70 percent of the Fe(II) remained in the effluent of the reactors. This is
similar to the Fe(II) percent remaining at a 1 hour residence time during RT1 of the
laboratory gutter reactors, which demonstrated that more efficient Fe(II) oxidation
occurred in the on-mound reactors than in the laboratory reactors for the young
sediments.
88
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30 35
Hydraulic Residence Time (min)
Fe(
II) re
mai
ning
(%)
Control (G, H )
Treatment (A-F)
Figure 5-3. Dissolved Fe(II) percent remaining for the control and treatment channels from on-mound channel reactors. The average Fe(II) percent remaining for each channel set (A-B, C-D, E-F, G-H) were averaged for each period and graphed against the mean residence from the corresponding period.
Some factors that could have created discrepancies between the field and
laboratory reactors were the differences in temperature and incident sunlight between the
two. The gutter reactor experiment had a temperature of approximately 23oC (room
temperature of the lab), whereas the temperature of the Hughes emergent was 12.7oC, and
only reached about 15oC at the effluent end in the warmest months. Additionally,
Hughes Borehole iron mound received much sunlight due to the open canopy and the
laboratory experiments were covered and only received intermittent light from laboratory
lights. As temperature and incident sunlight increase, Fe(II) oxidation rates have been
shown to increase (Barry et al. 1994, Kirby et al., 1999).
The zero order Fe(II) oxidation rates for the sediment gutter reactors in RT1 and
the Coir experiments at pseudo-steady states ranged from 4.67 x 10-8 M/s to 3.24 x 10-7
89
M/s, or 0.16 mg/L-min to 1.1 mg/L-min . The rates for the sediment reactors were
similar to those calculated by Pesic et al., for biological oxidation by Thiobacillus at 9
parts per million (ppm) oxygen (O2) concentration (Figure 5-4). The O2 concentrations at
the effluent end of the sediment gutter reactors ranged from 4-8 ppm. This was similar to
the O2 concentrations at the effluent end of the on-mound channel reactors. Therefore,
passive treatment reactors constructed in a comparable manner to the laboratory gutter
reactors with sufficient oxygen could have analogous Fe(II) oxidation rates.
Figure 5-4. Biological and abiotic rates of iron(II) oxidation of acid mine drainage (adapted from Williamson et al., 2006). Squares represent the sediment reactors from RT1, and the triangle represents the sediment reactors from the Coir experiment. Red is 10 hour, blue is 5 hour, green is 2 hour, and black is 1 hour residence times. The two lines with O are taken from Pesic et al., 1989 and the circles are from various published oxidation rates from field studies.
2
The long-term monitoring of the fence and toe locations at Hughes Borehole (see
Figure 4-5) showed that from July to September 2008 considerable amount of Fe(II)
90
oxidation occurred across the entire mound without manmade modifications. However,
little to no Fe(II) oxidation occurred at all other sampling events. Also, the flow rate
from the emergent discharge varied by almost a factor of ten over the past 2 ½ years.
Furthermore, the transects showed that variations in Fe(II) oxidation and pH occurred at
similar distances during different times of the year. These findings suggest there is
variability in both the quantity and quality of the AMD emerging from the borehole. This
is important for passive treatment systems that utilize the existing mound because it
shows that there is seasonal and yearly variation that can affect the consistency of the
system. A treatment reactor would have to account for the flow fluctuations and water
quality variations of a field system.
The findings of this study are promising for AMD remediation because they
imply that efficient and adequate biological Fe(II) oxidation does occur at Hughes
Borehole and can be enhanced by physical modifications to the existing iron mound. The
alkalinity of the low-pH effluent from biological passive treatment systems would have to
be increased before the AMD is discharged into the environment, but removal of the
majority of metal load prior to alkaline additions can greatly increase the efficiency and
reliability of conventional treatment systems. Systems that utilize the combination of
biological and chemical treatment methods could be applied to many similar low-pH
mine-impacted areas.
91
6. CONCLUSIONS
The major conclusions from the research presented in this thesis which can be
applied to constructing a passive treatment reactor are summarized in the following list:
Biological Fe(II) oxidation occurs at Hughes Borehole and can be
enhanced by modifications to the existing iron mound. However, there is
variability in both the quantity and quality of AMD from the emergent
discharge and variability in the natural biological and chemical processes
occurring across the iron mound.
An acclimation period of 300-450 hours was necessary for the treatment
reactors to reach steady state. After this period, residence times of at least
5 and 10 hours were needed to oxidize 90% and 97% of the influent
Fe(II), respectively.
The dissolved Fe(II) oxidation efficiency of reactors with iron mound
sediments improved with age of the sediments until approximately 90
days. Fe(II) oxidation efficiencies for 1 and 2 hour residence times ranged
from 25-60%, and 60-80%, respectively, as the sediments aged. This
suggests that biological communities continue to develop over time and
can improve the efficiency of the reactors as the age of the sediments
increase.
Total Fe percent remaining were consistently 20-30% greater than Fe(II)
percent remaining for each residence time. Even though more Fe(OH)3
precipitated during longer residence times the relative percentage of Fe(II)
to total Fe stayed the same for each residence time.
92
The effluent end of the gutter reactors had pH values of 2.5 to 3.0, with
influent pH values of 3.0 to 4.0, even after the majority of Fe(II) was
oxidized. The effluent pH values stayed reasonably constant for specific
residence times despite the variation of Fe(II) oxidation efficiency during
the acclimation of each residence time.
The gutter reactors that contained iron mound sediment had much greater
Fe(II) oxidation efficiencies (25-97%) than reactors that only contained
AMD (0-10%). This suggests that iron mound sediments greatly improve
the Fe(II) oxidation efficiency.
No other dissolved metals besides Fe were removed from the AMD at
Hughes Borehole or in the laboratory-scale gutter reactors. The iron
mound sediment consisted of approximately 65-75% iron oxide solids and
25-35% volatile solids.
Additional surface area, (i.e. plastic media or coconut fiber), increased the
Fe(II) oxidation efficiency of the reactors even at relatively short residence
times of 10-60 minutes.
Regardless of the type of treatment in the on-mound reactors, increased
residence time resulted in increased Fe(II) oxidation efficiency.
Furthermore, the on-mound reactors were more effective than the
laboratory reactors at oxidizing dissolved Fe(II) at short residence times.
93
Fe(II) oxidation rates from the gutter reactors with iron mound sediments
ranged from 4.67 x 10-8 M/s to 3.24 x 10-7 M/s, or 0.16 mg/L-min to 1.1
mg/L-min. These values are very similar to published biological Fe(II)
oxidation rates.
Future Research
Additional experiments should examine the effects of other parameters in conjunction
with the optimal residence times. Experiments that test parameters such as the effects of
water column depth, varying levels of DO, and methods to allow for the precipitation of
more Fe(OH)3 would be very useful for the design of passive treatment systems.
Additionally, studies on which biological communities are most responsible for Fe(II)
oxidation, whether they are autotrophic or heterotrophic, could help determine the most
efficient design for passive treatment systems. Lastly, iron mound sediments and AMD
from other acidic mine-impacted sites could be tested with the same conditions from this
study to better understand how different sites react to similar treatment methods.
94
BIBLIOGRAPHY
Baker B. and Banfield J. (2003) Microbial communities in acid mine drainage. FEMS Microbiology Ecology, 44(2): 139-152.
Barry, R.C., Schnoor, J.L., Sulzberger, B., Sigg, L., and Stumm, W. (1994) IRON OXIDATION-KINETICS IN AN ACIDIC ALPINE LAKE. Water Research 28: 323-333.
Bigham J.M., Schwertmann U., Traina S.J., Winland R.L., and Wolf M. (1996). Schwertmannite and the chemical modeling of iron in acid sulfate waters. Geochmica et Cosmochimica Acta, 60(12): 2111-2121. Boyer, J., Sarnoski, B., 1995. 1995 progress report—statement of mutual intent strategic plan for the restoration and protection of streams and watersheds polluted by acid mine drainage from abandoned coal mines. Philadelphia, Pa., U.S. Environmental Protection Agency, appendix (http://www.epa.gov/reg3giss/libraryp.htm). Brady, K.B.C. (1998) Groundwater Chemistry from Previously Mined Areas as a Mine Drainage Prediction Tool. Chapter 9 In: Coal Mine Drainage Prediction and Pollution Prevention in Pennsylvania. Harrisburg: Pennsylvania Department of Environmental Protection. pp. 9-1 to 9-21. Cornell, R.M, and Schwertmann, U. (1996) The Iron Oxides; Structure, Properties, Reactions, Occurrence, and Uses. VCH, New York, 36-42.
Cravotta, C.A., and Trahan, M.K. (1999) Limestone drains to increase pH and remove dissolved metals from acidic mine drainage. Applied Geochemistry 14: 581-606.
Cravotta, C.A. (2008) Dissolved metals and associated constituents in abandoned coal-mine discharges, Pennsylvania, USA. Part 1: Constituent quantities and correlations. Applied Geochemistry 23: 166-202.
Cravotta, C.A. (2008) Dissolved metals and associated constituents in abandoned coal-mine discharges, Pennsylvania, USA. Part 2: Geochemical controls on constituent concentrations. Applied Geochemistry 23: 203-226.
Department of Conservation and Natural Resources (DCNR), Bureau of Topographic and Geologic Survey. (1992) http://www.dcnr.state.pa.us/topogeo/gismaps/index.aspx Demchak J, Mcdonald L, Skousen J (2001) Water quality from underground coal mines in northern West Virginia (1968-2000). Chambers Environmental Group and West Virginia Univ, Morgantown, WV, 22 p
95
Emerson, D., and Moyer, C. (1997) Isolation and characterization of novel iron-oxidizing bacteria that grow at circumneutral pH. Applied and Environmental Microbiology 63: 4784-4792.
GAI Consultants (2007). Final Report: Phase I SRB Low Mine Storage and Treatment Project Evaluation. Volume 1 of 2.
Gibert, O., de Pablo, J., Cortina, J.L., and Ayora, C. (2004) Chemical characterisation of natural organic substrates for biological mitigation of acid mine drainage. Water Research 38: 4186-4196.
Hurst, Christon, L et al., Manual of Environmental Microbiology, 2nd ed. ASM Press, Washington DC, 2002: pg 511.
Johnson, D.B. (1998) Biodiversity and ecology of acidophilic microorganisms. Fems Microbiology Ecology 27: 307-317.
Johnson, D.B., and Hallberg, K.B. (2003) The microbiology of acidic mine waters. Research in Microbiology 154: 466-473.
Johnson D.B., Rolfe S., Hallberg K.B., and Iversen E. (2001) Isolation and phylogenetic characterization of acidophilic microorganisms indigenous to acidic drainage waters at an abandoned Norwegian copper mine. Environmental Microbiology, 3(10): 630-637.
Kirby, C.S., and Cravotta, C.A. (2005) Net alkalinity and net acidity 1: Theoretical considerations. Applied Geochemistry 20: 1920-1940.
Kirby, C.S., Thomas, H.M., Southam, G., and Donald, R. (1999) Relative contributions of abiotic and biological factors in Fe(II) oxidation in mine drainage (vol 14, pg 511, 1999). Applied Geochemistry 14: 1101-1101.
Lucas, M. (2008). A comparison of two acid mine drainage sites in Central Pennsylvania: Field site characterizations and batch reactor experiments. Master of Science Thesis, Penn State University.
Luu, Y., Ramsay, B.A., and Ramsay, J.A. (2003) Nitrilotriacetate stimulation of anaerobic Fe(III) respiration by mobilization of humic materials in soil. Applied and Environmental Microbiology 69: 5255-5262.
Macdonal.Dg, and Clark, R.H. (1970) Oxidation of aqueous ferrous sulphate by Thiobacillus Ferrooxidans. Canadian Journal of Chemical Engineering 48: 669-&.
Malhotra et al., 2002 S. Malhotra, A.S. Tankhiwale, A.S. Rajvaidya and R.A. Pandey, Optimal conditions for bio-oxidation of ferrous ions to ferric ions using Thiobacillus ferrooxidans, Bioresource Technology 85 (2002), pp. 225–234
96
Nemati M., Harrison S.T.L., Hansford G.S., and Webb C. (1998) Biological oxidation of ferrous sulphate by Thiobacillus ferrooxidans: a review on the kinetic aspects. Biochemical Engineering Journal, 1: 171-190
Nengovhela NR, de Beer M, Greben HA, Maree JP, Strydom CA (2002) Iron (II) oxidation to support limestone neutralization in acid mine water. Proc, Wisa Conf, Durban, SA, 19-23 May 2002
Nordstrom, D.K. (1985): The rate of ferrous iron oxidation in a stream receiving acid mine effluent. U.S. Geol. Surv. Water-Supply Paper 2270, 113-119.
Noike, T., Nakamura, K., and Matsumoto, J.I. (1983) Oxidation of ferrous iron by acidophilic iron-oxidizing bacteria from a stream receiving acid mine drainage. Water Research 17: 21-27. Pennsylania Department of Environmental Protection 2003. Status Report: The environmental legacy of coal mining in Pennsylvania.
Pesic, B., Oliver, D.J., and Wichlacz, P. (1989) An electrochemical method of measuring the oxidation rate of ferrous to ferric iron with oxygen in the presence of Thiobacillus-Ferrooxidans. Biotechnology and Bioengineering 33: 428-439.
Rossman W., Wytovich E., and Seif J.M. (1997). Abandoned Mines – Pennsylvania’s single biggest water pollution problem. Pennsylvania Department of Environmental Protection.
Rowe, O.F., and Johnson, D.B. (2008) Comparison of ferric iron generation by different species of acidophilic bacteria immobilized in packed-bed reactors. Systematic and Applied Microbiology 31: 68-77.
Senko, J.M., Wanjugi, P., Lucas, M., Bruns, M.A., and Burgos, W.D. (2008) Characterization of Fe(II) oxidizing bacterial activities and communities at two acidic Appalachian coalmine drainage-impacted sites. Isme Journal 2: 1134-1145.
Singer P. and Stumm W. (1970) Acidic Mine Drainage: The Rate-Determining Step. Science, 167 (3921): 1121-1123.
Smith, J.R., Luthy, R.G., and Middleton, A.C. (1988) MICROBIAL FERROUS IRON OXIDATION IN ACIDIC SOLUTION. Journal Water Pollution Control Federation 60: 518-530.
Stook, K.; Tolaymat, T.; Ward, M.; Dubey, B.; Townsend, T.; Solo-Gabriele, H.; Bitton, G., Relative leaching and aquatic toxicity of pressure-treated wood products using batch leaching tests. Environmental Science & Technology 2005, 39, (1), 155-163.
97
Stumm W. and Morgan J.J. (1981) Aquatic Chemistry, 2nd edition. John Wiley & Sons, New York. Stumm W. and Morgan J.J. (1996) Aquatic Chemistry, 3rd edition. John Wiley & Sons, New York, 683-691. Williamson, M. A., Kirby, C. S., and Rimstidt, J. D., 2006. Iron dynamics in acid mine drainage7th International Conference on Acid Rock Drainage (ICARD). American Society of Mining and Reclamation (AMSR), Lexington, KY. Zinc T., Wolfe A., and Curley K. (2005) Restoring the Wealth of the Mountains: Cleaning up Appalachia’s Abandoned Mines. Trout Unlimited.
98
1.E+05
1.E+06
1.E+07
Feed Tank Ctrl Gutter1
Ctrl Gutter2
Sed.Gutter 1
Sed Gutter2
Aqueous Effluent Sample Location
CFU
/ m
L
End 10 hr End 5 hr End 2 hr
1.E+07
1.E+08
1.E+09
1.E+10
Sed. Gutter 1in
Sed. Gutter 2in
Sed. Gutter 1out
Sed. Gutter 2out
Sediment Sample Location
CFU
/ g
Initial End 10 hr End 5 hr End 2 hr
Appendix A: Microbial populations for the first variable-residence time experiment with the gutter reactors
Initial End 10 End 5 End 2 Mean S.D Mean S.D Mean S.D Mean S.D Feed Tank 3.66E+05 1.05E+05 3.62E+05 9.56E+04 6.03E+06 8.15E+05 3.45E+06 1.06E+06 Ctrl Gutter 1 4.62E+05 1.21E+05 5.09E+06 1.34E+06 3.24E+06 1.36E+06
Aqueous Ctrl Gutter 2 4.06E+05 1.15E+05 4.76E+06 1.05E+06 2.59E+06 1.14E+06 CFU / mL Sed. Gutter 1 4.62E+05 1.38E+05 4.74E+06 7.21E+05
Sed Gutter 2 6.72E+05 2.24E+05 4.32E+06 1.17E+06 3.64E+06 1.05E+06 Sed. Gutter 1 in 1.07E+08 1.15E+07 1.16E+08 3.56E+07 1.05E+09 2.70E+08 8.93E+08 1.54E+08 Iron Mound
Sediment Sed. Gutter 2 in 9.39E+07 1.30E+07 2.93E+08 8.57E+07 1.02E+09 2.60E+08 3.97E+08 8.49E+07 CFU / g Sed. Gutter 1 out 6.23E+07 7.25E+06 1.70E+08 4.69E+07 1.58E+09 3.74E+08 9.01E+08 1.26E+08
Sed. Gutter 2 out 5.08E+07 6.22E+06 1.49E+08 3.75E+07 8.51E+08 6.52E+08 1.50E+08 1.67E+08
99
Appendix B Tabulated Data for Hughes Borehole Chemistry
Table B:1
Data Corresponding to Figure 4-2
LocationDistance from
source pH Temp DO Fe(II)
(feet) (oC) (mg/L) (mg/L)8/21/2007 A 0.0 4.1 13.6 0.50 102.56
B1 27 4.09 12.9 1.1 102.33B2 36 4.06 13.3 1.39 99.05B3 46 4.06 13.1 0.69 99.76B4 60 4.06 13 2.61B5 65 3.95 13.3 2.9 64.25C1 54 3.88 13.8 3.77 96.72C2 62 3.84 13.9 4.89 96.48C3 75 3.81 13.8 5.19 97.19C4 93 3.45 16.4 6.82C5 101 3.38 16.5 6.41 99.29
12/7/2007 A 0 4.1 12.4 0.71 105.89B1 27 4.06 12.3 1.27 102.91B2 36 3.62 12.1 1.88 99.23B3 46 3.82 12 5.82 97.92B4 60 3.63 11.3 4.26 90.78B5 65 3.6 11.08 5.4 87.69C1 54 3.74 9.2 5.88 95.18C2 62 3.63 8.9 10.68 89.47C3 75 3.81 5.3 99.11C4 93 3.8 11.5 6.34 96.49C5 101 3.5 9.7 7.14 71.74
5/22/2009 A 0 3.83 12.6 1.36 63.33B6 16 3.85 12.4 1.64B7 46 3.86 12.5 1.96 60.69B8 76 3.86 12.5 2.32 59.45D1 97 3.84 12.7 3.86 59.96D2 106 3.85 13 3.4 59.60D3 153 3.77 13.8 7.17 59.03D4 161 3.64 15.4 8.46 56.50D5 201 3.61 19.5 10.94 54.64
100
Table B:2 Data corresponding to Figures 4-5 and 4-6
Fe(II) (mg/L) Total Fe (mg/L)
Date Fence Toe Fence Toe 7/3/08 109.6 100.4 7/8/08 96.6 92.8 7/17/08 87.2 56.2 7/22/08 107.5 100.4 101.98 99.8 8/7/08 109.9 34.7 99.49 70.2 8/14/08 99.8 33.3 101.74 59.3 8/21/08 108.6 42.7 8/28/08 116.1 47.5 9/4/08 102.2 39.8 100.20 65.7 9/12/08 114.4 9/18/08 100.2 39.2 101.09 57.7 9/25/08 100.6 45.6 100.85 69.0 10/3/08 99.5 78.2 99.91 90.6 10/9/08 100.4 16.5 102.04 52.4 11/7/08 104.3 105.18 11/13/08 102.4 12/5/08 102.2 86.9 106.89 95.9 1/19/09 100.0 101.0 100.97 106.8 2/20/09 98.0 87.5 102.16 92.8 3/20/09 91.0 70.9 4/3/09 88.7 74.6 87.5 80.6 5/6/09 67.7 62.1 5/22/09 63.2 54.8 6/23/09 66.4 56.3 66.70 60.5 7/3/09 67.3 59.4 71.36 62.98
101
Appendix C Tabulated Data for the On-Mound Channel Reactors
Table C:1
Data corresponding to Figure 4-8
Treatment Channels Control Channels Fe(II) (mg/L) Fe(II) (mg/L) Mean S.D. Mean S.D.
7/3/08 98.6 0.4 95.9 1.0 7/8/08 95.2 0.4 93.0 2.7 7/17/08 96.7 1.0 93.4 0.0 7/23/08 104.3 1.6 104.8 1.9 8/7/08 102.5 2.5 97.3 5.0 8/14/08 97.1 4.0 95.0 10.5 8/21/08 91.4 2.7 90.7 2.8 8/28/08 89.5 0.3 84.5 2.3 9/4/08 91.6 3.6 84.2 9/12/08 96.9 1.7 9/18/08 72.1 19.0 72.3 20.3 9/25/08 93.9 3.6 86.1 8.9 10/3/08 77.8 5.0 88.1 3.4 10/9/08 87.8 15.4 93.5 3.2 11/7/08 91.4 6.0 98.1 2.0 11/13/08 68.2 6.4 96.1 6.0 12/5/08 90.0 1.7 109.5 10.1 3/20/09 59.9 3.9 88.5 2.6 4/3/09 44.2 7.2 75.9 1.8 5/6/09 62.6 4.7 65.1 2.0 5/22/09 28.7 14.3 53.9 0.4 6/23/09 44.5 7.0 56.3 2.1 7/3/09 43.2 13.4 61.3 0.9
102
Table C:2 Data corresponding to Figure 4-9
Channel A,B C,D E,F G,H Fe(II) out / Fe(II) in Fe(II) out / Fe(II) in Fe(II) out / Fe(II) in Fe(II) out / Fe(II) in
Mean S.D. Mean S.D. Mean S.D. Mean S.D. 5/6/09 0.82 0.07 0.94 0.01 0.93 0.03 0.94 0.03 5/22/09 0.21 0.03 0.61 0.07 0.67 0.00 0.94 0.01 6/23/09 0.73 0.12 0.71 0.93 0.02 1.00 0.04 7/3/09 0.62 0.02 0.47 0.04 0.87 0.01 0.93 0.01
Table C:3 Data corresponding to Figure 4-10
5/6/2009 7/3/2009
NPOC (mg/L) Mean S.D Mean S.D Influent 0.76 0.49
Treatment 1.92 0.37 0.45 0.07 Control 0.78 0.11 0.36 0.05
TN (mg/L) Mean S.D Mean S.D Influent 0.91 0.87
Treatment 1.05 0.05 0.78 0.04 Control 0.89 0.01 0.82 0.00
Table C:4
Data corresponding to Figures 4-8 and 5-3
Period I II III IV Time (min) 2.6 2.3 4.4 1.9 G, H Fe(II) remain. (%) 94.0 81.0 90.1 93.7 S.D. 6.4 6.7 6.6 4.8 Time (min) 2.4 14.6 27.6 6.8 E, F Fe(II) remain. (%) 96.1 77.7 68.0 75.8 S.D. 6.4 12.2 9.8 17.5 Time (min) 2.5 24.3 32.5 18.2 C, D Fe(II) remain. (%) 96.8 82.5 74.6 59.0 S.D. 4.9 7.1 9.4 9.9 Time (min) 2.5 10.7 26.9 12.1 A, B Fe(II) remain. (%) 96.4 86.2 78.1 49.5 S.D. 4.8 5.5 14.6 22.8
103
Appendix D Tabulated Data for the Laboratory-Scale Gutter Reactors
Table D:1
Data corresponding to Figures 4-12, 4-13, and 4-16 Time Time(hrs) PVs Ctrl in S.D. Ctrl out S.D. Sed in S.D. Sed out S.D (hrs) PVs Ctrl in S.D. Ctrl out S.D. Sed in S.D.
10 hr 0 0 10 hr 0 010 1 106.0 0.7 105.4 0.4 105.0 0.3 54.3 4.5 10 1 106.0 0.7 99.5 3.9 105.0 0.331 3 103.1 2.5 97.8 6.3 103.3 0.7 103.3 7.2 31 3 103.7 3.5 104.3 0.8 103.9 0.952 5 107.8 1.8 107.6 0.3 104.4 0.9 120.0 5.5 52 5 101.9 2.3 108.4 1.1 109.9 0.773 7 110.5 5.7 104.1 4.8 105.2 3.3 110.8 25.2 73 7 108.9 16.0 104.8 0.1 104.797 10 106.9 0.8 106.5 2.5 106.3 0.4 78.7 51.1 97 10129 13 109.7 3.0 106.7 1.0 108.1 1.8 84.7 51.1 129 13 124.6 2.5 106.8 0.7 107.6 1.7146 15 104.7 0.7 103.8 0.5 107.4 2.7 81.7 35.9 146 15 112.8 11.2 107.8 1.1 114.9 10.3176 18 93.4 2.6 97.6 3.1 93.4 0.3 39.0 47.5 176 18 93.6 1.6 93.9 0.1 91.9 0.0191 19 99.6 0.1 92.7 3.6 102.6 1.3 27.1 32.6 191 19 100.0 2.1 89.5 1.6 103.8 1.6214 21 99.1 0.7 92.5 3.2 98.9 0.2 17.0 20.0 214 21 87.6 1.5 93.8 0.6 89.2 1.5262 26 102.2 0.2 98.5 2.4 102.7 1.5 2.6 0.1 262 26 104.8 2.2 96.6 0.2 102.2 0.8314 31 107.8 0.3 96.9 3.6 102.9 7.7 3.2 0.2 314 31 103.1 6.4 95.2 2.6 100.8 1.1
5 hr 334 33 5 hr 334 33336 34 96.5 2.5 91.4 8.2 92.9 2.8 5.4 0.2 336 34 95.5 1.5 89.7 3.9 92.4 0.3340 35 96.4 0.7 96.4 10.2 98.7 4.1 12.1 1.4 340 35 94.2 1.7 97.0 12.5 99.7 1.5345 36 93.4 1.4 80.6 1.9 93.5 0.7 12.9 0.7 345 36 94.9 5.5 86.6 3.6 87.4 10.1356 38 87.6 0.6 81.5 6.6 86.1 2.2 10.0 1.4 356 38 91.6 1.9 82.5 0.7 91.7 1.9366 40 96.5 5.9 81.8 2.4 93.8 2.8 12.2 2.9 366 40381 43 90.8 1.4 78.6 2.5 92.5 0.2 8.6 1.5 381 43 96.3 6.6 85.4 0.2 92.1 0.9440 55 75.3 2.3 64.0 0.6 75.3 0.2 3.0 0.6 440 55 79.8 0.3 68.6 0.9 78.9 3.7482 63 72.9 0.5 69.8 3.4 71.7 2.1 6.8 3.9 482 63 78.2 1.1 69.4 1.5 77.3 0.7629 92 104.7 1.5 104.9 0.4 110.9 8.8 7.3 2.8 629 92 103.7 0.7 110.1 3.0 100.0 33.2652 97 99.7 0.8 99.8 10.5 102.1 4.3 12.5 1.5 652 97 97.4 1.7 100.4 12.9 103.2 1.6
2 hr 670 101 2 hr 670 101672 101 107.8 1.1 99.3 8.9 118.9 5.4 13.1 8.4 672 101 105.5 13.9 132.4 11.8 180.0 45.3674 102 102.0 0.7 112.9 8.0 102.2 0.9 74.6 7.5 674 102 103.2 3.7 119.8 27.5 100.3 1.5676 104 103.6 0.5 102.4 1.0 100.9 1.4 63.4 12.7 676 104 114.5 2.2 94.5 0.0 105.8 0.8681 106 101.5 1.4 90.3 0.8 99.0 2.5 51.5 6.2 681 106 89.9 3.6 92.7 1.5 90.0 0.6695 113 102.0 3.7 103.3 0.6 104.0 5.2 40.0 9.5 695 113 101.6 1.1 102.2 0.5 104.1 3.7702 117 98.4 1.9 87.3 12.7 90.8 1.3 39.6 3.7 702 117 95.7 0.5 90.3 0.8 93.1 1.0720 126 99.6 4.5 88.9 2.0 102.9 0.1 36.9 1.5 720 126 95.0 10.6 105.1 1.0 103.1 5.3738 134 97.9 0.4 88.6 1.8 101.1 0.6 28.3 4.3 738 134 102.2 0.7 101.8 2.8 104.3 2.1755 143 96.4 1.1 86.5 7.1 95.4 0.2 28.5 4.0 755 143 96.4 1.1 86.5 7.1 95.4 0.2
1 hr 757 144 1 hr 757 144758 145 97.9 3.6 96.1 3.3 96.4 1.0 46.6 10.3 758 145 98.3 1.3 91.1 1.3 100.6 0.1759 146 94.1 0.4 82.6 10.5 92.5 1.8 55.3 0.6 759 146 100.6 1.1 93.5 6.8 99.9 6.3763 150 87.7 0.5 79.0 6.3 88.9 2.4 65.7 1.7 763 150 91.6 1.3 94.7 4.0 93.5 8.2776 163 90.4 0.8 89.6 5.7 89.0 0.5 67.2 9.3 776 163 100.0 1.8 99.5 12.5 98.5 8.6789 176 95.2 2.0 86.0 6.1 93.0 1.0 70.3 14.0 789 176 91.7 0.6 89.6 2.2 108.0 1.2
Fe(II) (mg/L) Total Fe (mg/L)
104
Table D:2 Data corresponding to Figures 4-14 and 4-15
Time Time
(hrs) PVs Ctrl in S.D. Ctrl out S.D. Sed in S.D. Sed out S.D (hrs) PVs Ctrl in S.D. Ctrl out S.D. Sed in S.D. Sed out10 hr 0 0 10 hr 0 0
10 1 3.79 0.02 4.20 0.00 3.81 0.00 2.86 0.03 10 1 2.5 1.4 5.8 0.1 3.3 0.3 4.831 3 3.83 0.01 4.02 0.05 3.76 0.06 2.85 0.02 31 3 3.8 0.4 5.8 0.0 3.4 0.3 4.652 5 3.83 0.01 3.77 0.11 3.71 0.16 2.83 0.07 52 5 3.9 0.1 7.9 0.0 3.4 0.4 6.573 7 3.84 0.00 3.74 0.16 3.83 0.00 2.84 0.07 73 7 4.9 1.1 7.9 0.5 4.1 0.1 6.397 10 3.85 0.00 3.78 0.05 3.82 0.01 2.79 0.06 97 10 3.8 0.3 7.5 0.2 3.8 0.0 6.2129 13 3.83 0.01 3.81 0.01 3.74 0.08 2.74 0.05 129 13 3.8 0.2 8.1 0.5 4.2 0.7 6.6146 15 3.81 0.02 3.73 0.06 3.82 0.02 2.38 0.44 146 15 3.8 0.0 7.9 0.3 4.1 0.3 6.0176 18 3.77 0.01 3.70 0.06 3.78 0.02 2.73 0.09 176 18 3.4 0.0 8.1 0.0 4.1 0.3 6.0191 19 3.79 0.00 3.72 0.03 3.75 0.04 2.72 0.04 191 19 3.0 0.3 7.3 0.5 3.4 0.3 5.3214 21 3.78 0.00 3.62 0.09 3.79 0.01 2.74 0.09 214 21 3.3 0.1 6.9 0.0 3.2 0.2 5.4262 26 3.71 0.00 3.51 0.05 3.69 0.01 2.64 0.04 262 26 3.4 0.1 8.4 0.1 3.9 0.4 6.1314 31 3.66 0.00 3.41 0.08 3.64 0.01 2.60 0.00 314 31 3.5 0.5 6.8 0.1 3.3 0.3 5.8
5 hr 334 33 5 hr 334 33336 34 3.49 0.00 3.36 0.05 3.49 0.01 2.66 0.04 336 34 3.4 0.1 8.1 0.1 3.6 0.2 6.8340 35 3.40 0.07 3.29 0.06 3.34 0.13 2.66 0.01 340 35 3.6 0.0 7.7 0.2 3.6 0.0 5.9345 36 3.40 0.05 3.30 0.01 3.46 0.01 2.72 0.05 345 36 3.3 0.1 9.0 0.1 3.3 0.1 6.6356 38 3.43 0.00 3.24 0.02 3.43 0.01 2.70 0.01 356 38 2.7 0.4 7.8 0.5 3.0 0.4 6.1366 40 3.39 0.02 3.24 0.02 3.42 0.00 2.69 0.02 366 40381 43 3.41 0.01 3.14 0.15 3.41 0.00 2.69 0.01 381 43 3.4 0.4 7.4 0.2 3.1 0.2 7.2440 55 3.22 0.03 3.14 0.03 3.25 0.02 2.69 0.01 440 55 3.3 0.4 7.9 0.2 4.2 0.3 7.0482 63 3.25 0.04 3.13 0.03 3.23 0.01 2.69 0.02 482 63 2.4 0.0 9.4 0.1 3.0 0.3 6.6629 92 3.18 0.01 3.16 0.01 3.18 0.01 2.81 0.01 629 92 2.4 0.2 9.7 0.2 2.8 0.6 7.8652 97 3.13 0.04 3.10 0.03 3.14 0.02 2.77 0.01 652 97 2.4 0.2 7.3 0.3 2.8 0.1 6.6
2 hr 670 101 2 hr 670 101672 101 672 101674 102 3.74 0.30 3.28 0.12 3.82 0.08 2.95 0.02 674 102676 104 3.95 0.01 3.83 0.02 3.92 0.00 2.95 0.01 676 104 2.0 0.2 6.7 0.1 2.1 0.5 5.0681 106 3.93 0.00 3.88 0.05 3.91 0.00 2.99 0.07 681 106 1.5 0.2 7.3 0.0 1.3 0.0 6.1695 113 3.92 0.00 3.90 0.01 3.82 0.12 3.01 0.00 695 113702 117 3.89 0.01 3.56 0.14 3.85 0.01 3.00 0.00 702 117 2.1 0.2 7.5 0.4 2.0 0.1 5.9720 126 3.86 0.06 3.82 0.02 3.96 0.01 3.09 0.04 720 126738 134 3.75 0.01 3.73 0.02 3.73 0.03 3.02 0.05 738 134 1.9 0.1 9.2 0.1 2.1 0.3 6.1755 143 755 143
1 hr 757 144 1 hr 757 144758 145 3.48 0.00 3.56 0.05 3.41 0.12 3.10 0.01 758 145 1.1 0.0 4.5 0.2 1.0 0.1 3.8759 146 3.42 0.09 3.47 0.02 3.50 0.01 3.12 0.03 759 146763 150 3.49 0.00 3.47 0.01 3.45 0.02 3.19 0.01 763 150 1.2 0.2 6.3 0.6 1.3 0.3 5.3776 163.1 3.49 0.00 3.48 0.00 3.48 0.00 3.22 0.08 776 163
Dissolved Oxygen (mg/L)pH
105
PV Ctrl 1 Ctrl 2 Sed 1 Sed 2Mean S.D. Mean S.D. Mean S.D. Mean S.D.
5 hr 00.8 0.96 1.06 0.83 0.89 3.29 0.01 3.27 0.04 3.30 0.01 2.81 0.013.1 1.01 0.99 0.86 0.92 3.30 0.01 3.28 0.02 3.28 0.00 2.85 0.055 0.92 0.93 0.83 0.88 3.21 0.00 3.18 0.01 3.21 0.00 2.73 0.01
12.9 1.00 0.97 0.90 0.93 3.22 0.00 3.22 0.02 3.22 0.01 2.80 0.0118.3 0.94 0.96 0.93 0.96 3.21 0.00 3.22 0.01 3.20 0.00 2.75 0.0233.7 1.07 1.02 0.67 0.9338.6 1.02 0.95 0.43 0.86 3.10 0.00 3.10 0.01 3.10 0.00 2.79 0.0649.2 0.99 0.96 0.37 0.87 3.09 0.01 3.05 0.06 3.10 0.01 2.78 0.0652.8 1.00 0.98 0.22 0.78 3.07 0.00 3.07 0.01 3.08 0.00 2.70 0.0561.6 0.97 0.99 0.16 0.83 3.11 0.00 3.11 0.01 3.12 0.01 2.81 0.1171 0.95 0.98 0.10 0.5276 0.93 0.92 0.07 0.29 3.08 0.00 3.06 0.02 3.09 0.00 2.73 0.04
86.2 0.89 0.91 0.07 0.13 3.05 0.01 2.99 0.05 3.07 0.00 2.70 0.022 hr 90.2 0.86 0.81 0.06 0.13 3.11 0.00 3.08 0.01 3.13 0.00 2.75 0.03
91.2 0.95 0.93 0.32 0.36 3.16 0.01 3.14 0.03 3.16 0.00 2.76 0.0493.45 0.96 0.95 0.33 0.38 3.14 0.00 3.11 0.01 3.14 0.00 2.76 0.01105.7 0.99 0.95 0.32 0.37 3.11 0.00 3.12 0.01 3.11 0.01 2.81 0.01128.2 0.93 0.32 0.28 3.11 3.09 3.10 0.00 2.82 0.00162.2 0.91 0.97 0.33 0.34 3.09 0.00 3.09 0.01 3.10 0.00 2.82 0.01
Ctrl in Ctrl out Sed in Sed outpHFe(II) out / Fe(II) in
Table D:3 Data corresponding to Figures 4-17 and 4-18
Table D:4 Data corresponding to Figure 4-23
Fe(II) out / Fe(II) inTime(hrs) Mean stdev Mean stdev Mean stdev Mean stdev Mean stdev
0 1.00 0.00 1.00 0.00 1.00 0.00 1.00 0.00 1.00 0.002 0.85 0.00 0.89 0.01 0.95 0.01 1.02 0.00 0.88 0.084 0.79 0.01 0.83 0.00 0.93 0.01 1.00 0.01 0.89 0.079 0.73 0.04 0.73 0.01 0.90 0.01 0.97 0.01 0.88 0.1024 0.38 0.03 0.33 0.00 0.57 0.04 0.70 0.01 0.85 0.16
SterileLive Coconut Fiber Straw Mulch
106
Table D:5 Data corresponding to Figure 4-20
Fe(II) out / Fe(II) in Total Fe out / Total Fe in
PV Ctrl S.D Coir S.D Ctrl S.D Coir S.D 0 2 1.02 0.01 0.45 0.02 5 0.86 0.05 0.39 0.02 8 0.92 0.03 0.37 25 0.99 0.02 0.52 0.12 27 0.90 0.05 0.44 0.03 29 31 0.93 0.06 0.47 0.00 1.02 0.17 0.49 0.14 33 0.97 0.00 0.62 0.03 0.94 0.01 0.63 0.03 47 0.93 0.02 0.45 0.02 0.75 0.13 0.58 0.15 73 0.96 0.02 0.23 0.03 0.97 0.03 0.56 0.10 101 0.95 0.03 0.09 0.00 0.99 0.02 0.51 0.10 122 0.99 0.05 0.03 0.01 1.01 0.03 0.43 122 123 0.93 0.03 0.32 0.22 0.95 0.05 0.72 0.29 125 0.93 0.00 0.40 0.10 130 0.91 0.08 0.45 0.14 132 0.98 0.01 0.34 0.14 139 0.98 0.01 0.30 0.14 0.97 0.05 0.58 0.04 153 0.95 0.05 0.37 0.05 173 0.99 0.04 0.41 0.09 1.08 0.01 0.53 0.05 174 0.96 0.00 0.48 0.08 176 0.96 0.02 0.51 0.06 181 0.96 0.09 0.50 0.04 196 0.92 0.01 0.35 0.07 0.96 0.67 0.28 225 0.93 0.08 0.36 0.14 243 0.78 0.09 0.09 0.02 0.85 0.02 0.43 0.08 250 0.98 0.02 0.03 0.03 271 0.81 0.12 0.01 0.01 0.79 0.04 0.37 0.04 274 0.85 0.01 0.42 0.13 278 0.85 0.01 0.34 0.10
107
eed
1.4
1.20.9
1.2
1.00.80.31.2
0.6
1.0
0.7
1.1
0.8
0.5
PV Ctrl in S.D. Ctrl out S.D. Coir in S.D. Coir out S.D. Ctrl in S.D. Ctrl out S.D. Coir in S.D. Coir out S.D. F02 3.38 0.01 3.36 0.01 3.36 0.02 2.89 0.02 4.5 0.3 8.8 0.3 5.3 0.3 5.7 0.25 3.36 0.05 3.42 0.01 3.35 2.89 0.038 3.35 0.11 3.42 0.02 3.30 0.19 2.81 0.08 3.8 0.1 7.6 0.1 4.0 0.5 6.2 0.5
25 3.26 0.02 3.25 0.03 3.27 0.03 2.92 0.03 4.7 0.8 9.2 0.7 5.3 0.2 6.8 0.22931 3.27 0.00 3.24 0.00 3.27 0.01 3.18 0.03 4.8 0.3 8.4 0.2 4.4 0.1 7.2 0.433 3.20 0.00 3.19 0.01 3.19 0.01 3.01 0.0147 3.19 0.00 3.17 0.01 3.18 0.01 2.91 0.03 4.8 1.3 9.0 0.9 4.7 0.2 6.2 0.073 3.07 0.00 3.06 0.01 3.06 0.01 2.83 0.04 4.1 0.4 8.0 0.3 3.8 0.6 4.5 0.5
101 3.47 0.05 3.30 0.01 3.45 0.02 2.73 0.03 2.5 0.1 7.3 0.3 2.3 0.3 3.5122 3.42 0.06 3.39 0.01 3.41 0.01 2.82 0.06 8.4 0.2 4.1 0.8122123 3.50 0.00 3.44 0.02 3.41 0.05 2.95 0.00 3.6 0.2 7.3 0.4 3.7 0.0 5.3 1.0125 3.48 0.02 3.37 0.01 3.43 0.00 2.91 0.01130 3.49 0.00 3.31 0.05 3.47 0.04 2.95 0.00132 3.29 0.01 3.23 0.03 3.28 0.01 2.76 0.00139 3.21 0.01 3.20 0.01 3.22 0.01 2.90 0.01 3.9 0.3 8.0 0.2 2.7 0.1 4.8 0.1153 3.55 0.00 3.37 0.01 3.53 0.09 2.90 0.00174 3.45 0.00 3.41 0.02 3.44 0.02 2.94 0.02 3.9 9.2 0.9 2.8 0.7 5.5 0.4176 3.45 0.00 3.40 0.03 3.43 0.03 2.89 0.03196 3.25 0.01 3.21 0.02 3.25 0.02 2.93 0.03 4.8 0.8 10.1 2.3 4.9 0.1 5.8 1.0225 3.41 0.03 3.33 0.05 3.37 0.03 2.93 0.04243 3.25 0.02 3.13 0.05 3.21 0.02 2.91 0.02 4.2 1.1 9.5 0.7 3.4 0.0 3.6250 3.05 0.01 3.01 0.01 3.05 0.01 2.76 0.01271 3.04 0.00 2.95 0.00 3.03 0.01 2.77 0.00 4.5 0.6 8.6 0.2 4.1 0.4 4.4 1.2274 3.03 0.00 2.97 0.01 3.01 0.03 2.82 0.02278 2.99 0.01 2.95 0.02 2.97 0.02 2.79 0.01 3.1 0.0 6.9 0.2 4.0 0.4 5.2 0.1
pH Dissolved Oxygen (mg/L)
Table D:6 Data corresponding to Figures 4-21 and 4-22
108
Table D:7 Data corresponding to Figure 4-23
NPOC TN Mean S.D Mean S.D
Ctrl 1 0.40 0.00 0.91 0.01 Ctrl 2 0.48 0.04 0.92 0.03 Coir 1 1.40 0.06 0.99 0.01 Coir 2 1.48 0.02 0.89 0.02 Feed 1.29 0.03 0.95 0.00
Table D:8 Data corresponding to Figure 4-29
pC Fe(OH)3
pH 2 2.5 3 3.5 4 4.5 Sulfate (mg/L)
0 7.26E+03 4.49E+02 4.18E+01 5.90E+00 1.22E+00 3.24E-01 600 1.77E+05 8.05E+03 3.19E+02 1.51E+01 1.51E+00 3.34E-01 6000 1.71E+06 7.64E+04 2.81E+03 9.81E+01 4.18E+00 4.18E-01
Appendix E Tabulated Data for the Discussion
Table E:1
Data corresponding to Figure 5-1
Time (hrs) Fe(II) (mg/L) 0 2 4 8 24
Filter Sterilized 33.4 32.1 30.0 32.6 29.7 Aerobic Live 33.4 30.8 27.5 24.3 7.3
1% v/v Formaldehyde 33.4 32.1 30.1 31.6 29.4 Anaerobic Live 33.4 33.8 31.0 30.2 29.1
Related Documents
