REMEDIATION OF HIGH-STRENGTH MINE IMPACTED WATER …

The Pennsylvania State University

The Graduate School

College of Engineering

REMEDIATION OF HIGH-STRENGTH MINE IMPACTED WATER WITH CRAB

SHELL SUBSTRATE MIXTURES: LABORATORY COLUMN AND FIELD PILOT TESTS

A Thesis in

Environmental Engineering

by

Jessica A. Grembi

2011 Jessica A. Grembi

Submitted in Partial Fulfillment of the Requirements

for the Degree of

Master of Science

May 2011

The thesis of Jessica A. Grembi was reviewed and approved* by the following:

Rachel A. Brennan Assistant Professor of Environmental Engineering Thesis Advisor

Brian A. Dempsey Professor of Environmental Engineering

William D. Burgos Professor of Civil Engineering Department of Civil and Environmental Engineering

Peggy A. Johnson Professor of Civil Engineering Head of the Department of Civil and Environmental Engineering

*Signatures are on file in the Graduate School

ABSTRACT

Anaerobic passive treatment systems remediating high-strength mine impacted water

(MIW) have not displayed consistent success. For example, the high iron (140 mg/L) and acidity

(380 mg/L as CaCO3) of the Klondike-1 discharge near Ashville, PA, caused premature clogging

of a vertical flow pond which was filled with a traditional spent mushroom compost (SMC) and

limestone substrate. In this study, continuous-flow columns and pilot-scale field reactors were

used to evaluate if treatment of high-strength MIW can be improved using crab shell as a

substrate amendment.

For the lab study, continuous-flow columns containing 50– 100% crab shell (with the

balance SMC) were compared to a sand control and a column containing the traditional 90%

SMC and 10% limestone. MIW for the column study was obtained from the Klondike-1 site and

pumped at a flow rate of 0.25 mL/min to maintain a 16 h hydraulic retention time within each

column. After determining the best performing substrate mixture (70% crab shell + 30% SMC)

in the column test, a pilot-scale field study was initiated, in which 1000-gallon tanks were filled

with a limestone underdrain and an upper substrate layer of: 1) 100% crab shell; 2) 70% crab

shell + 30% SMC; or 3) 90% SMC + 10% limestone. A fourth tank containing a sandstone

underdrain with a 70% crab shell + 30% SMC substrate layer was installed to determine if similar

performance could be achieved without the limestone underdrain. Aqueous samples were

collected from the columns/reactors and analyzed for pH, ORP, ammonia, acidity, alkalinity,

DOC, anions, and metals. Additional samples taken after passive aeration were also monitored.

In the column study, the 70% crab shell + 30% SMC column treated double the volume

of MIW, removed more than twice the mass of metals, and sustained pH above 5.0 for almost

twice as long as the traditional SMC and limestone substrate. To date, the field study results

mirror the laboratory findings.

iii

TABLE OF CONTENTS

1. Introduction ...................................................................................................................... 1

1.1 Mine Impacted Water ............................................................................................... 1 1.2 Anaerobic Sulfate-Reducing Passive Treatment ...................................................... 3 1.3 Composition of Organic Substrate Layer ................................................................. 7 1.4 Crab shell as an alternative substrate for VFPs ........................................................ 8 1.5 Background of the Klondike-1 Site .......................................................................... 10 1.6 Objectives of Current Study ..................................................................................... 11

2. Materials and Methods ..................................................................................................... 13

2.1 Water Source ............................................................................................................ 13 2.2 Substrates ................................................................................................................. 14 2.3 Continuous-Flow Column Setup .............................................................................. 18 2.4 Analytical Methods .................................................................................................. 22 2.5 Conservative Tracer Tests ........................................................................................ 23

3. Continuous-flow column laboratory experiment ............................................................. 25

3.1 Source Water ............................................................................................................ 25 3.2 Conservative Tracer Tests ........................................................................................ 26 3.3 pH, Alkalinity and Acidity ....................................................................................... 27 3.4 Metals Removal........................................................................................................ 31

3.4.1 Primary Metals ............................................................................................... 31 3.4.2 Trace Metals ................................................................................................... 36

3.5 Sulfate Reduction ..................................................................................................... 39 3.6 Carbon and Nitrogen Species ................................................................................... 40 3.7 Other Cations ........................................................................................................... 42

4. Discussion ........................................................................................................................ 45

4.1 Alkalinity .................................................................................................................. 45 4.2 Metals Removal........................................................................................................ 49

4.2.1 Al .................................................................................................................... 50 4.2.2 Fe Removal Within Treatment Column ......................................................... 51 4.2.3 Fe Removal After Passive Aeration and Settling ........................................... 53 4.2.4 Mn .................................................................................................................. 53 4.2.5 Trace Metals ................................................................................................... 54

4.3 Carbon Species ......................................................................................................... 59 4.4 Cations ...................................................................................................................... 62 4.5 Longevity of Treatment ............................................................................................ 65

5. Field Pilot System ............................................................................................................ 68

5.1 System Concept ........................................................................................................ 68 5.2 System Design .......................................................................................................... 70

iv

5.3 Construction, Incubation, and Field Sampling ......................................................... 73 5.4 Results ...................................................................................................................... 76

6. Conclusions ...................................................................................................................... 78

6.1 Treatment Longevity for Engineering Designs ........................................................ 78 6.2 Potential Concerns.................................................................................................... 78

7. Future Work ..................................................................................................................... 81

References ................................................................................................................................ 84

Appendix A Conservative Tracer Tests .................................................................................. 90

Appendix B In-line pH and ORP Probes ................................................................................ 92

Appendix C Metals Removal After Passive Aeration and Settling ........................................ 97

Appendix D Sulfate Data ........................................................................................................ 104

Appendix E Cation Data Plots ................................................................................................ 106

Appendix F Metals Mass Balance Calculations ..................................................................... 111

Appendix G Organic Carbon Mass Balance Calculations ...................................................... 116

Appendix H Visual MINTEQ Geochemical Modeling .......................................................... 120

Appendix I Treatment Scale-up using a 1:1 Crab Shell to Proppant Ratio ............................ 123

Appendix J Field Pilot System Installation and Sampling Photos .......................................... 128

v

LIST OF FIGURES

Figure 1-1. Cross-sectional schematic of a vertical flow pond (VFP). .................................... 4

Figure 1-2. Risk classification categories established for passive treatment systems

remediating net acidic discharges in the 2009 PA DEP Program Implementation

Guidelines for the Bureau of Abandoned Mine Reclamation Acid Mine Drainage

Set-Aside Program which guides funding for remediation projects (taken from PA

DEP, 2009). ...................................................................................................................... 6

Figure 1-3. Relative distribution of chitin, protein, and CaCO3 for various species

(Muzzarelli, 1977). ........................................................................................................... 9

Figure 2-1. Laboratory continuous-flow columns used to treat Klondike-1 MIW. ................ 19

Figure 2-2. Schematic of continuous-flow column experimental setup.................................. 20

Figure 2-3. Passive aeration and settling were accomplished in bins subsequent to the

continuous-flow columns. Sample cells were used to collect water exiting the

settling bins to monitor increased metals removal from this additional

oxidation/precipitation step after anaerobic treatment. Photo taken on day 36 of the

experiment. ....................................................................................................................... 21

Figure 3-1. Alkalinity generation and acidity data from continuous-flow columns treating

MIW from the Klondike-1 site. ........................................................................................ 30

Figure 3-2. Breakthrough curves for dissolved Al (A) and pH measurements (B) taken

after continuous-flow columns treating Klondike-1 MIW. .............................................. 32

Figure 3-3. Breakthrough curves for dissolved Fe measured after continuous-flow

columns treating Klondike-1 MIW (A) and after subsequent passive aeration and

settling (B)........................................................................................................................ 34

Figure 3-4. Breakthrough curves for dissolved Mn measured after continuous-flow

columns treating Klondike-1 MIW. ................................................................................. 35

Figure 3-5. Breakthrough curves for dissolved cobalt (A) and zinc (B) measured after

continuous-flow columns treating Klondike-1 MIW. ...................................................... 37

Figure 3-6. Breakthrough curves for dissolved nickel measured after continuous-flow

columns treating Klondike-1 MIW. ................................................................................. 38

Figure 3-7. Experimental columns photographed after 84 days of continuous-flow

conditions with Klondike-1 MIW. Note black precipitates which formed in all

except the sand control column. The remaining four treatment columns (not shown)

also displayed the formation of black precipitates. .......................................................... 39

vi

Figure 3-8. Dissolved organic carbon measured in column effluent during continuous-

flow column test treating MIW from the Klondike-1 site. Inset graph shows

maximum values achieved at beginning of experiment. .................................................. 41

Figure 3-9. Ammonium measured from column effluent during continuous-flow column

test treating MIW from the Klondike-1 site. .................................................................... 42

Figure 4-1. CaCO3 calculated from experimental alkalinity and acidity data versus

theoretical CaCO3 data for substrate mixtures. ................................................................ 47

Figure 4-2. Percent of total trace metal loading retained (after breakthrough) within

treatment columns at completion of continuous-flow column experiment treating

Klondike-1 MIW. ............................................................................................................. 56

Figure 4-3. Substrate exhaustion with respect to DOC and alkalinity generation. +

symbol above column indicates the value presented is the PV when the experiment

ended, thus the potential exists for additional DOC generation until complete

substrate exhaustion. ........................................................................................................ 60

Figure 4-4. Total C remaining in each treatment column at completion of continuous-

flow column experiment. ................................................................................................. 61

Figure 4-5. Speciation of relevant cations in the 100% crab shell column after 10 PV

based on geochemical modeling with Visual MINTEQ and an assumed 250 mg/L

sulfate reduction. Results for the 70% CS + 30% SMC column were identical for

this time point and those for the traditional 90% SMC + 10% LS column varied by

no more than 1%. ............................................................................................................. 63

Figure 5-1. Schematic of pilot-scale VFPs installed at Klondike-1 field site. ........................ 69

Figure 5-2. Schematic of pre-existing full-scale treatment system at the Klondike-1 site

with the location of the pilot system indicated. ................................................................ 70

Figure 5-3. Limestone (A) and sandstone (B) rocks used in underdrains for field pilot-

scale VFPs treating MIW at the Klondike-1 site.............................................................. 72

Figure 5-4. Photo of microbial tea-bag style sample pouches filled with organic substrate

and buried within each pilot-scale reactor at the Klondike-1 site. ................................... 74

Figure 5-5. Pilot-scale VFPs and subsequent aerobic settling ponds installed to treat

MIW at the Klondike-1 field site. .................................................................................... 75

Figure 5-6. pH values of MIW influent and pilot-scale reactor effluent from initial 90

days of monitoring. .......................................................................................................... 77

Figure 5-7. Alkalinity generated from pilot-scale reactors during initial 90 days of the

field test. ........................................................................................................................... 77

vii

Figure A-1. Conservative tracer test response curves for continuous-flow columns.............. 91

Figure B-1. Comparison of pH readings taken during continuous-flow column test from

bench top electrode and electrodes mounted in flow-through cells. Symbols

connected by a line indicate bench top electrode readings; unconnected symbols

indicate in-line electrode readings. .................................................................................. 94

Figure B-2. Comparison of pH readings taken during continuous-flow columns test from

bench top electrode and electrodes mounted in flow-through cells. Symbols

connected by a line indicate bench top electrode readings; unconnected symbols

indicate in-line electrode readings. .................................................................................. 95

Figure B-3. ORP measured with in-line electrodes in effluent from continuous-flow

columns treating MIW from the Klondike-1 site. ............................................................ 96

Figure C-1. Breakthrough curves for dissolved Al measured in continuous-flow columns

treating Klondike-1 MIW (A) and after subsequent passive aeration and settling (B). ... 98

Figure C-2. Breakthrough curves for dissolved Fe measured in continuous-flow columns

treating Klondike-1 MIW (A) and after subsequent passive aeration and settling (B). ... 99

Figure C-3. Breakthrough curves for dissolved Mn measured in continuous-flow


settling (B)........................................................................................................................ 100

Figure C-4. Breakthrough curves for dissolved cobalt measured in continuous-flow


settling (B)........................................................................................................................ 101

Figure C-5. Breakthrough curves for dissolved nickel measured in continuous-flow


settling (B)........................................................................................................................ 102

Figure C-6. Breakthrough curves for dissolved zinc measured in continuous-flow


settling (B)........................................................................................................................ 103

Figure D-1. Sulfate data for continuous-flow columns treating Klondike-1 MIW. ................ 105

Figure E-1. Dissolved Ca measured in continuous-flow columns treating Klondike-1

MIW. Inset graph shows maximum values achieved at beginning of experiment;

axes have same units as large plot. ................................................................................... 106

Figure E-2. Dissolved K measured in continuous-flow columns treating Klondike-1



viii

Figure E-3. Dissolved Mg measured in continuous-flow columns treating Klondike-1

MIW. ................................................................................................................................ 108

Figure E-4. Dissolved Na measured in continuous-flow columns treating Klondike-1



Figure E-5. Dissolved PO43—

P measured in continuous-flow columns treating Klondike-

1 MIW. ............................................................................................................................. 110

Figure F-1. Percent of each metal retained within columns treating Klondike-1 MIW at

completion of experiment (after 181 days of continuous-flow conditions). .................... 113

Figure G-1. Scenarios 1 (initial values after incubation), 2 (average after 10 PV), and 3

(pH=5) using the SO42-

:HS- ratio from iteration A. Results reveal no considerable

difference in solubility of metal species related to variations in cation and carbonate

loadings in the 100% CS column within the pH range encountered during the

continuous-flow columns experiment (pH 2.5-7.5). In fact, iteration A for all 9

scenarios produced similar results, indicating the SO42-

:HS- ratio dominates

solubility of metals within each of the systems under the given circumstances. ............. 118

Figure G-2. Effect on solubility/saturation of total dissolved Fe as SO42-

:HS- ratios are

increased (increased SO42-

:HS- ratio indicates limited or no sulfate reduction is

occurring). ........................................................................................................................ 119

Figure H-1. Correlation between amount of crab shell within treatment system and

amount of original organic carbon remaining within system at completion of

continuous-flow experiment treating Klondike-1 MIW. .................................................. 122

Figure J-1. MIW at the Klondike-1 site. ................................................................................. 128

Figure J-2. Tank piping modifications and installation of underdrain piping network, July

26, 2010. ........................................................................................................................... 129

Figure J-3. Placement of septic tanks used to simulate pilot-scale VFPs to treat MIW at

the Klondike-1 site. .......................................................................................................... 129

Figure J-4. Placement of rock underdrains into tanks, done manually to avoid damage to

underdrain piping system! ................................................................................................ 130

Figure J-5. Completed installation of limestone rock underdrain system ............................... 131

Figure J-6. 1,000 pound super sack of crab shell unloaded into back of dump truck to be

mixed with sand proppant and SMC. ............................................................................... 131

Figure J-7. Filling of organic substrate mixtures into pilot-scale VFPs to treat MIW at the

Klondike-1 site ................................................................................................................. 132

ix

Figure J-8. Placement of microbial tea-bag style sampling pouches 8-10 inches into

organic substrate material. ............................................................................................... 133

Figure J-9. A layer of pea gravel was added to the top of each reactor to prevent

loss/disturbance of organic substrate. .............................................................................. 134

Figure J-10. Installation of influent piping system. Pipes were emplaced to gravity feed

from an oxidation pond of the existing full-scale treatment system at the Klondike-1

site. A dock was built to facilitate maintenance of influent hose lines. .......................... 135

Figure J-11. Individual influent hoses attach to the buried PVC piping approximately 12

inches below the water surface and feed water to each pilot-scale VFP. Water enters

through ¼ inch holes drilled into the final 2 feet of flexible tubing, which is covered

with mesh to discourage iron precipitates from entering the system. .............................. 136

Figure J-12. Piping network leading from oxidation pond of current full-scale treatment

system to feed pilot-scale VFPs. ...................................................................................... 136

Figure J-13. View of the four pilot-scale VFPs and aerobic settling ponds. .......................... 137

Figure J-14. Earl Smithmyer, President of the CCWA, assisted tremendously with the

pilot-system installation, specifically with the piping networks. ..................................... 137

Figure J-15. Water was added to the pilot-scale VFPs on August 2, 2010. Some

overflow problems were encountered with the settling ponds, as they were not

properly leveled. ............................................................................................................... 138

Figure J-16. Orifices were created to maintain a flow rate of 0.2 gallons per minute

throughout the pilot-scale study. ...................................................................................... 138

Figure J-17. Aeration of the tank effluent was encouraged via a miniature cascade

constructed with corrugated piping. ................................................................................. 139

Figure J-18. The system was flushed and then left to incubate for a week prior to

initiation of continuous-flow operations. ......................................................................... 140

Figure J-19. Sampling event at the pilot-scale VFPs during Fall 2010. ................................. 140

Figure J-20. As expected, the pilot systems froze over the winter months. ............................ 141

Figure J-21. Many thanks to Shan Lin and Sara Goots, whose help made the pilot system

install go quickly, and who have conducted analysis of the data from the field pilot-

scale study! ....................................................................................................................... 142

Figure J-22. Thanks to Duke, for cheerfully spending his summer at the Klondike-1 site

with me! ........................................................................................................................... 142

x

LIST OF TABLES

Table 2-1. Field measurements and dissolved metals analysis of water collected from the

Klondike-1 site at various times to supply the continuous-flow column test. ................. 14

Table 2-2. Particle size distribution of crab shell (CS) and SMC used as packing materials

in the continuous-flow column study. .............................................................................. 15

Table 2-3. Mass of solid packing materials used in each continuous-flow column. ............... 16

Table 2-4. Extractible metals and compost analysis of the continuous-flow column

packing materials. ............................................................................................................ 17

Table 3-1. Average water quality parameters (taken weekly for the duration of the

experiment) of continuous-flow column influent............................................................. 26

Table 3-2. Flow characteristics of continuous-flow columns treating Klondike-1 MIW,

measured using tracer tests at the completion of the 181-day experiment. ...................... 27

Table 3-3. Maximum pH and duration of neutralization capacity achieved using different

substrates to treat Klondike-1 MIW in the continuous-flow column experiment. ........... 28

Table 3-4. Average concentration after 10 pore volumes (PV) and maximum

concentrations of Ca, K, Mg, Na, and P noted in influent water and effluent from

continuous-flow columns treating Klondike-1 MIW with substrates containing

mixtures of crab shell (CS), spent mushroom compost (SMC), and/or limestone

(LS). ................................................................................................................................. 44

Table 4-1. Metal treatment capacity for each continuous-flow column utilizing 40 g

substrate mixtures to treat Klondike-1 MIW. .................................................................. 49

Table 4-2. Fe retained within treatment columns (after breakthrough) at completion of

continuous-flow column test treating Klondike-1 MIW. ................................................. 52

Table 4-3. Tolerance limits and analytic, free ion, and active free ion average

concentrations (after 10 PV) for cations of interest from continuous-flow columns

treating Klondike-1 MIW. ................................................................................................ 64

Table 4-4. Iterative calculations used to determine the theoretical masses and volumes of

crab shell and sand needed if a 1:1 packing ratio (by mass) were used in the

continuous-flow column study. Bolded row indicates the mass required to fill a

~700 mL column, as used in this study. ........................................................................... 66

Table 4-5. Experimental and theoretical treatment longevity of crab shell substrate

mixtures for treating high-strength MIW. Experimental longevity was determined in

the column study using a 1:12 (by mass) substrate to proppant ratio (40 g total

substrate). Theoretical longevity was estimated by extrapolating the results to a 1:1

(by mass) crab shell to sand proppant ratio that would be used in the field. ................... 66

xi

Table 4-6. Theoretical total metals removal, volume of MIW treated in each column, and

substrate loading factor. Values were calculated based experimental vales from

Table 4-1 and the scale-up factor to account for a 1:1 (by mass) crab shell to sand

proppant ratio that would be used in the field. ................................................................. 67

Table 5-1. Actual and designed mixture of the organic substrate layer in each pilot-scale

VFP installed to treat MIW at the Klondike-1 site........................................................... 71

Table F-1. Metals mass balance for continuous-flow columns conducted at completion of

experiment (after 181 days of operation). ........................................................................ 114

Table G-1. Visual MINTEQ geochemical modeling scenarios, consisting of 6 iterations

of SO42-

:HS- ratios each .................................................................................................... 117

Table G-2. Carbonate, cation, and dissolved metals concentrations for Visual MINTEQ

geochemical modeling scenarios. ..................................................................................... 117

Table H-1. Organic carbon mass balance for continuous-flow columns treating

Klondike-1 MIW (performed at completion of experiment, after 181 days of

operation). ........................................................................................................................ 121

Table I-1. Theoretical total mass of crab shell and sand able to fit into a 100% crab shell

column (~700 mL) assuming a 1:1 crab shell to sand proppant mass ratio. .................... 123

Table I-2. Theoretical total mass of crab shell, SMC, and sand able to fit into a 90% crab

shell + 10% SMC column (~700 mL) assuming a 1:1 crab shell to sand proppant

mass ratio. ........................................................................................................................ 124



mass ratio. ........................................................................................................................ 124



mass ratio. ........................................................................................................................ 125



mass ratio. ........................................................................................................................ 125



mass ratio. ........................................................................................................................ 126

Table I-7. Theoretical total mass of SMC and limestone able to fit into a traditional 90%

SMC + 10% limestone column (~700 mL). ..................................................................... 126

xii

Table I-8. Calculated scale-up factors based on theoretical total mass of substrate

required to fill ~700 mL volume and actual mass used in the experiment. ...................... 127

xiii

ACKNOWLEDGEMENTS

First, I would like to thank the three undergraduate students, Brad Sick, Sara Goots and

Shan Lin, who have joined in this research in differing capacities ranging from collaboration

during the laboratory continuous-flow column test to installation and monitoring of the pilot-scale

field system. Their company and shared learning has greatly enhanced my graduate experience.

I would like to thank the faculty and staff within the Department of Civil and Environmental

Engineering, most especially Dr. Rachel Brennan for her mentorship and encouragement, and Dr.

Brian Dempsey and Dr. Bill Burgos for serving on my committee. The graduate students within

the department, specifically those in the Brennan Research Group, have helped me significantly

along the way and are acknowledged for their guidance and friendship. Also, the support of my

family and the three little bears (Duke, Henry, and Mya) is gratefully recognized for reminding

me that life awaits outside the confines of the Sackett building. The assistance of Earl

Smithmeyer, the Clearfield Creek Watershed Association, and the Foundation for Pennsylvania

Watersheds is also gratefully acknowledged for their role in the facilitating the success of the

pilot-scale system installation.

Pergamon Press is acknowledged for their permission to reproduce copyrighted

material in Figure 1.3. This research is supported in part by the National Science Foundation

CAREER Award No. CBET-0644983. Any opinions, findings, and conclusions or

recommendations expressed in this material are those of the authors and do not necessarily reflect

the views of the National Science Foundation.

1

1. Introduction

The EPA estimates that there are over 200,000 abandoned mine sites which impact over

10,000 miles of streams in the United States (U.S) (EPA, 1997). In Pennsylvania alone, a study

completed in 2010 found that mine impacted water (MIW) affected more than 5,500 miles of

streams (PA DEP, 2010). However, this is not a problem unique to the U.S. and can be found on

every continent except Antarctica, where mining has been banned by an international treaty

(Romero et al., 2010; Baruah and Khare, 2010; Dinelli et al., 1999; Schippers et al., 2007; Tutu et

al., 2008; Edraki et al., 2005). MIW varies greatly based on the geologic makeup of the bedrock

being mined, but typically contains high concentrations of metals, sulfur species, and acidity

which can harm aquatic life and threaten the quality of potable water supplies.

1.1 Mine Impacted Water

The term MIW has replaced acid mine drainage, acid rock discharge, and other phrases

because it better describes the polluted effluent which drains from sites where the mining of coal

or metal ores has exposed the reactive surface of rocks and minerals. Although the discussion

below will show that acidity is typically produced in MIW, that is not always the case. In

geologic strata containing limestone, natural buffering of the MIW can occur, resulting in alkaline

discharges. A common mineral found in geologic formations associated with coal mining

operations is pyrite (FeS2). The exposure of FeS2 to oxygen and water during and after mining

results in the formation of ferrous iron, sulfate (SO42-

), and increased acidity according to the

following equation:

2

2FeS2(s) + 7O2 + 2H2O 2Fe2+

+ 4SO42-

+ 4H+

Eq. 1-1

Initially, solutions of MIW have circum-neutral pH values, at which time abiotic

oxidation of ferrous iron (Fe2+

) occurs slowly. As additional FeS2 is oxidized and pH begins to

drop, biotic Fe2+

oxidation becomes the prominent mechanism. As the energy gained from the

oxidation of Fe2+

is relatively low, a single organism must carry out the reaction multiple times to

achieve the energy required for cell functions and growth (Maier et al., 2009). Thus, the biotic

oxidation of Fe2+

occurs much more rapidly than abiotic oxidation. Both processes are achieved

via the following reaction pathway:

4Fe2+

+ O2 + 4H+ 4Fe

3+ + 2H2O Eq. 1-2

Ferric iron (Fe3+

) produced by the above pathway also reacts spontaneously with

additional pyrite and propagates the oxidation process via the following reaction:

FeS2(s) + 14Fe3+

+ 8H2O 15Fe2+

+ 2SO42-

+ 16H+

Eq. 1-3

Thus, once pyrite is exposed to air and water, the combination of microbial and

chemically mediated reactions creates a rapid cycle that promotes further oxidation of the

remaining pyrite.

The highly acidic environment also results in the dissolution of accompanying minerals,

causing the release of numerous other metals in addition to Fe. Primary metals of concern in

MIW of the mid-Atlantic, bituminous coal region are aluminum (Al), iron (Fe), and manganese

(Mn) (Cravotta, 2008). Trace metals found in this region, and also found in considerably higher

concentrations elsewhere, include cobalt (Co), nickel (Ni), and zinc (Zn), among others (Cravotta,

2008; Schippers et al., 2007; Baruah et al., 2010; Dinelli et al., 2001 ).

3

The final resulting step, which produces the characteristic Fe precipitation known as

―yellow-boy‖ is expressed by Eq. 1-4:

Fe3+

+ 3H2O Fe(OH)3 + 3H+

Eq. 1-4

MIW can affect local waterways in numerous facets. First, the low pH often associated

with MIW is an extreme environment which most organisms are not suited to survive in.

Secondly, the high metal concentrations can be toxic to fish and other biota. Finally, the

precipitated minerals that naturally form when conditions change (i.e., increased pH and oxygen

concentrations when MIW enters a stream and undergoes mixing) can coat the bottom of the

channel, sealing off food and oxygen sources. The least tolerant species are driven off or die

initially, followed by impacts throughout the food chain until the entire ecosystem is altered.

Water quality of surface water supplies is also a matter of human health, as it often supplies local

drinking water facilities. Increased metals concentrations can require expensive treatment beyond

that typically provided, driving up costs for potable water.

1.2 Anaerobic Sulfate-Reducing Passive Treatment

Although active treatment systems are often used to treat effluent water from ongoing mining

operations, they are expensive and require daily maintenance. The most cost-effective method to

remediate MIW at abandoned sites, where restoration funding is extremely limited, is passive

treatment systems which do not require frequent attention. The only passive systems capable of

removing all components of MIW (metals, sulfur species, and acidity) are anaerobic biologically-

based systems, which utilize sulfate reducing bacteria (SRB) to facilitate the precipitation of

metal sulfides while simultaneously generating alkalinity.

4

Vertical flow ponds (VFPs) are a typical approach to anaerobic passive treatment which

allows water to flow vertically downward through an anoxic layer of organic substrate, and then

through a layer of limestone rock to produce additional alkalinity before exiting through a

network of under drains (Figure 1-1). These system designs typically include an additional

treatment cell subsequent to the VFP in the form of a settling pond were aeration is encouraged

and additional metal (hydr)oxides precipitate from the neutralized water (Doshi, 2006).

Figure 1-1. Cross-sectional schematic of a vertical flow pond (VFP).

MIW rich in SO42-

provides an optimal source of electron acceptors for SRB, which are

among the most ubiquitous organisms on the planet (Faulwetter et al., 2009). When provided

with an acceptable electron donor, these microorganisms will reduce sulfate to sulfide and

generate alkalinity in the form of bicarbonate (HCO3-) according to the following reaction:

SO42-

+ 2CH2O H2S + 2HCO3- Eq. 1-5

The H2S formed promotes the precipitation of metal ions in the water as metal sulfides:

Me2+

+ H2S MeS(s) + 2H+

Eq. 1-6

The utilization of this process allows for the controlled precipitation of pure metal sulfides, which

are desirable for several reasons. First, metal sulfides have low solubility over a wide pH range

and are highly stable (Jandová et al., 2005). This reduces the likelihood that metal sulfides will

5

be released when conditions within the treatment system fluctuate over varied seasonal

conditions. Secondly, metal sulfides have been shown to produce extremely small particle sizes,

resulting in a considerably lower volume than hydroxide precipitates (Lewis, 2010). This allows

for the precipitation of more metals within a given volume, and can reduce the overall footprint of

a treatment system. Finally, if metals can be precipitated under optimized conditions, high purity

is achievable which allows for collection and beneficial reuse as pigmenting agents for paint and

other products (Lewis, 2010).

Although conventional use of such passive systems has proven both economical and

effective for numerous low-strength MIW (low acidity and low metals concentrations)

applications, these systems have failed to consistently treat sources with high metals loads and

flow rates. It has been shown that anaerobic passive treatment systems can experience reduced

reactivity and permeability well before the expected exhaustion of the treatment substrate. Loss

of reactivity is caused by coating of the substrate materials with precipitates (armoring) and

permeability is limited by clogging when precipitates fill the pore space of the system (e.g., Rees

et al., 2001; Watzlaf et al., 2002, 2004; Ziemkiewicz et al., 2003; Rose et al., 2004; Simon et al.,

2005). In addition, passive treatment systems are often installed in remote forested locations

where the surface area for large treatment ponds is frequently limited due to site constraints. In

addition to remoteness, many mines are located on steep hillsides or in narrow valleys, such that

construction of passive systems may again be restricted by the availability of flat areas. Thus,

system construction is not always done according to optimal design, but instead with respect to

the footprint available onsite (Matthies et al., 2010).

A comprehensive study conducted by the U.S. Office of Surface Mining and the

Pennsylvania Department of Environmental Protection (PA DEP, 2008) surveyed over 250

passive treatment systems constructed in PA between 1990 and 2008 to evaluate the performance

6

of the systems, to identify systems requiring additional maintenance or rehabilitation, and to

better define appropriate technologies for different classifications of discharges. The results of

this survey led to the establishment of risk classification categories for MIW discharges based on

flow rate and metals concentrations (Figure 1-2). Waters containing > 50 mg/L combined metals

(Al + Fe) were automatically designated high-risk regardless of the f low rate. Discharges with

lower concentrations of Al + Fe, but with higher flow rates, can also be assessed as high-risk. In

2009, the PA DEP released new guidelines for the funding of MIW remediation systems, which

virtually eliminated the ability of high-risk discharges to receive funding for passive treatment

systems (PA DEP, 2009). Due to the belief that the discharges categorized as high-risk are not

necessarily high-risk if the treatment system is designed properly, the term high-strength MIW

will be used henceforth in this thesis to describe MIW with high metals concentrations and or

flow rates.

Figure 1-2. Risk classification categories established for passive treatment systems remediating

net acidic discharges in the 2009 PA DEP Program Implementation Guidelines for

the Bureau of Abandoned Mine Reclamation Acid Mine Drainage Set-Aside

Program which guides funding for remediation projects (taken from PA DEP, 2009).

7

1.3 Composition of Organic Substrate Layer

It has also been hypothesized that short lifetime and low performance associated with

some passive biological systems could be related to the efficiency of the organic substrate to

sustain SRB activity, in addition to the permeability and reactivity problems discussed above

(Pruden et al., 2007). The composition of the organic substrate layer is critical to the success of

anaerobic passive treatment systems as it maintains conditions for a complex microbial

community which supports the growth and metabolism of SRB. The most common organic

substrates are composed of cellulosic materials. However SRB can only utilize short chain

organic acids and alcohols as their carbon (C) and electron donor sources, thus they require a

healthy population of fermentative and cellulose degrading organisms. These upstream

symbiants hydrolyze cellulose and other polymeric compounds and transform the degradation

products into simpler organic compounds that can be utilized by the SRB (Logan et al., 2005).

In addition to providing a C and electron donor source, the organic substrate must also be

able to provide other components essential for SRB and their supporting microbial community.

Namely, the provision of nutrients such as nitrogen (N) and phosphorus (P) is required. It has

been reported that insufficient N could possibly be limiting SRB activity in passive anaerobic

treatment systems (Waybrant et al., 2002). SRB also require an optimal pH between 5 and 8

(Willow and Cohen, 2003; Jong and Parry, 2006), which is typically achieved with the

supplementation of limestone chips within the substrate or is achieved completely via the

limestone underdrain. Finally, the substrate layer should possess the ability to degrade slowly,

allowing the system to last for longer periods without the requirement for substrate replacement.

The porosity of the substrate should also be considered to maintain flow through the

system. Due to decomposition of the substrate over time, it has been suggested that non-reactive

materials, such as pea gravel, be mixed with the organic substrate layer (Doshi, 2006; Rötting et

8

al., 2008). In addition to maintaining permeability in the long term, inert proppants can induce

large pore spaces which will not be clogged by precipitates as easily.

Typical substrate layer composition in the central Pennsylvania region has been a mixture

of spent mushroom compost (SMC) and limestone chips due to local availability and low cost.

However, to overcome the difficulties mentioned, various mixtures of organic materials have

been evaluated with respect to their ability to provide an adequate environment for sustained

sulfate reduction including: manure and SMC (Nicromat et al., 2006); municipal leaf compost

mixed with cattle/horse manure and sewage sludge (Morales et al., 2005); a mixture of beech

wood chips, pulverized alfalfa, and pine shavings (Pereyra et al., 2008); sawdust and manure

(Hallberg and Johnson, 2005); a mixture of pine wood chips and sawdust, alfalfa hay, kiln dust,

and dairy cow manure (Hiibel et al., 2008); leaves and compost (Viggi et al., 2009); grass cuttings

(Matshusa-Masithi et al., 2009); and mushroom compost and straw (Dann et al., 2009).

Although these substrate mixtures have shown to provide treatment and sustain SRB in the

laboratory environment to varying degrees, the underlying factors controlling success have not

been definitively determined. Thus, their suitability for application at full-scale field sites is not

guaranteed.

1.4 Crab shell as an alternative substrate for VFPs

The world’s market for seafood crustaceans, particularly shrimp, crab, and lobster, is

several million tons per year, of which 50% is discarded as shell waste (Gerente et al., 2007).

Crab shells contain carbon, nitrogen, and alkalinity in a complex matrix of chitin, protein, and

calcium carbonate, which could be utilized as an organic substrate for use within anaerobic

passive MIW treatment systems. Chitin, the world’s second most abundant naturally occurring

polysaccharide, and its deacetylated derivative, chitosan, have been explored for use in a variety

9

of areas, including: pulp and paper mill waste treatment, medical bandages which accelerate

healing, and recently remediation of MIW, among others (Hayes et al., 2008). The amount of

chitin, protein, and calcium carbonate associated with different crustacean exoskeletons varies by

species (Figure 1-3). A portion of the chitinous material within the crab shell is a naturally

deacetylated form of chitosan, which has additionally been proven to remove metal ions from

solution via adsorption. This property is likely to play a role in MIW treatment (Robinson-Lora

and Brennan, 2010b). The diversion of this resource from the waste stream and its subsequent

utilization for MIW remediation systems embodies the concept of sustainability and yields a win-

win situation for both the seafood and mining industries.

Figure 1-3. Relative distribution of chitin, protein, and CaCO3 for various species (Muzzarelli,

1977).

Research investigating crab shell has shown simultaneous biological, chemical, and

physical remediation of low-strength MIW. In previous laboratory and field studies, crab-shell

has out-performed other substrates by rapidly removing Fe and Al, as well as Mn, something

10

which other substrates have been unable to accomplish at circumneutral pH (Daubert and

Brennan, 2007; Venot et al., 2008; Robinson-Lora and Brennan, 2009; Robinson-Lora and

Brennan, 2010; Newcombe and Brennan, 2010). In addition, the integrated source of CaCO3 has

been proposed to negate the need for supplemental alkalinity sources, which could potentially

lower overall system costs. Most recently, it has been shown that crab shell mixed with SMC can

further decrease costs, as crab shell ($0.75/lb) is more expensive than other typical substrates

($0.025/lb) (Newcombe and Brennan, 2010). These results led to the suggestion that crab shell

could be utilized as a substrate amendment within the VFP of sites suffering from clogging due to

excessive metals concentrations and limited footprints.

1.5 Background of the Klondike-1 Site

Acidic discharges from the abandoned Klondike mine, located near Ashville, Pennsylvania, have

contributed to the impairment of the Little Laurel Run. This waterway is a tributary of Clearfield

Creek, whose waters eventually flow into the Chesapeake Bay. Two discrete discharges,

Klondike-1 and Klondike-2, were identified at the site, and through the efforts of the Clearfield

Creek Watershed Association (CCWA) funds were obtained to design and build a treatment

system for each discharge. The Klondike-1 discharge (located at 40° 33.117 N, 78° 29.798 W) is

located at the site of the original mine shaft entrance. This area was later strip-mined, and it is

believed that a section of the strip intersects a portion of the underground mine near this location.

Thus the water is likely a combination of water seeping from the original mine shaft as well as

seep from mine tailings left during the strip mining operations. Monitoring of the Klondike-1

discharge by the CCWA indicated an average Fe concentration of 141 mg/L, acidity of 417 mg/L

as CaCO3, Mn of 30 mg/L, pH 3.3, and flow rate of 15 gallons per minute (gpm) (Rose, 2008).

Sulfate concentrations for this discharge range from 700-1400 mg/L.

11

The treatment system designed for the Klondike-1 discharge consisted of a primary

oxidation pond, VFP, an aerobic settling pond, and then a constructed wetland as a final polishing

step. Unfortunately, the depth of the limestone and compost layers were reduced from original

design parameters due to budget constraints. Construction was completed in November 2007, but

within 9 months the VFP had clogged due to a layer of Fe precipitates (orange layer in Figure

1-1) which formed on top of the organic substrate layer. The Fe precipitates were removed from

the VFP and two additional oxidation ponds were constructed at the beginning of the treatment

system to facilitate additional low-pH biological Fe oxidation. With the new oxidation pond

treatment cells, the system still does not meet effluent requirements for iron and acidity. Even

during optimal operating conditions the system was never able to sufficiently remove Mn from

the water, and thus was not capable of completely neutralizing acidity. In addition, the system has

the potential to clog again if these cells do not operate as designed.

1.6 Objectives of Current Study

The overarching goal of this research was to determine if high-strength MIW could successfully

be treated via anaerobic, biologically-mediated, passive treatment systems utilizing crab shell

substrate mixtures. Previous work investigating crab shell for the remediation of low-strength

MIW had only evaluated 100% crab shell and fractions of ≤ 50% (by mass) crab shell mixed with

SMC. Thus, the first objective of this research was to evaluate substrate mixtures containing 50-

100% crab shell to determine the optimal ratio of crab shell to SMC. As subsequent aeration had

not been investigated in coordination with crab shell substrates mixtures, the second objective of

this study was to determine the additional treatment efficiency afforded by passive aeration and

settling after the simulated VFP.

12

Due to a very active watershed group with ties to the university, the Klondike-1 site was

identified as a local, high-strength MIW discharge. A secondary goal of this project became to

assist in solving a local problem by helping to provide data which could be used in guiding

decisions for a possible retro-fit of the existing failed system at the Klondike-1 site with an

organic substrate amendment in the near future. The main objective of this portion of the project

was to design and build a pilot-scale reactor to determine the effects, if any, of scale-up on the

treatment system while at the same time optimizing treatment conditions for the water quality

characteristics present at the Klondike-1 site.

13

2. Materials and Methods

2.1 Water Source

All water used for the continuous-flow column test was collected from the Klondike-1

site. Water was collected approximately 250 yards downstream from the point of emergence in a

non-stagnant, deep channel section of the discharge stream (40° 33.102 N, 78° 29.838 W). Water

was collected from this location six times throughout the duration of the continuous-flow test to

allow for a realistic fluctuation of water quality over varying environmental conditions

(temperature, rainfall, etc.) experienced at the site. When required, ice on the surface of the

channel was broken to obtain access to water beneath. Field measurements for temperature, pH,

conductivity, and oxidation reduction potential (ORP) were taken onsite (Table 2-1). ORP was

measured using an Oakton® Waterproof ORPTestr 10, and temperature, pH, and conductivity

were measured using an Oakton® Multi-Parameter Tester 35. Samples were also collected for

dissolved metals analysis (results provided in Table 2-1), which upon return to the laboratory

were subsequently filtered, preserved, and analyzed via the method described in section 2.4.

Flexible plastic tubing (1-inch diameter) and a hand pump were used to transfer the water

into high-density polyethylene containers (20 L jerry cans or 50 L carboys), which were capped

with minimal headspace. Immediately upon return to the lab, all water storage containers were

continuously purged with argon gas to maintain an anoxic environment and minimize Fe

oxidation. The influent water reservoirs were covered with opaque black plastic to prevent the

growth of phototrophic organisms (e.g., algae) that could potentially produce oxygen within the

system.

14

Table 2-1. Field measurements and dissolved metals analysis of water collected from the

Klondike-1 site at various times to supply the continuous-flow column test.

Batch of water collected from the Klondike-1 site Average

1 2 3 4 5 6

Date Collected 10/19/09 11/30/09 1/6/10 1/30/10 3/5/10 4/1/10 N/A

Date Introduced to

Columns

10/29/09 11/30/09 1/7/10 1/30/10 3/5/10 4/9/10 N/A

Water temp. (°C) 2.4 4.2 - 10.5 0.5 15.7 6.7

pH 2.77 3.11 2.96 3.42 3.72 3.93 3.3

Conductivity (µS/cm) 1520 1370 - 848 1300 1520 1312

ORP (RmV) 390 675 - 603 395 409 494

Al

(mg/L)

2.93 2.70 2.26 3.59 2.12 3.26 2.81

Fe 133 107 111 98.4 111 68.2 105

Mn 43.4 40.0 33.5 30.3 35.4 29.4 35.3

Co 0.48 0.45 0.41 0.37 0.40 0.37 0.41

Ni 0.41 1.16 1.18 0.98 0.70 0.68 0.9

Zn 0.27 0.25 0.25 0.31 0.25 0.32 0.28

SO42-

1183 1171 1054 886 985 725 1000

- No field measurement taken

2.2 Substrates

The following substrates were used to promote the remediation of the collected water in the

laboratory column tests: ChitoRem® Chitin Complex (grade SC-20, JRW Bioremediation,

Lenexa, KS); SMC (Mushroom Test Demonstration Facility, The Pennsylvania State University);

and limestone (0.420-0.841 mm, 88.89% CaCO3, New Enterprise Stone and Lime Company,

Tyrone, PA). The ChitoRem® Chitin Complex, here forth referred to as crab shell (CS), is a

product derived from Dungeness crab shell and contains ~10% chitin, ~12% protein, and ~78%

mineral matter (68% as CaCO3) (Robison-Lora and Brennan, 2009b). A particle size distribution

of the organic substrates is provided in Table 2-2.

15

Table 2-2. Particle size distribution of crab shell (CS) and SMC used as packing materials in the

continuous-flow column study.

Particle Size % of substrate (based on dry mass)

(mm) CS SMC

>4.75 0 10.6

2.36-4.75 2.9 15.3

1.2-2.36 27.1 24.7

0.85-1.2 16.2 33.3

0.297-0.85 34.1 13.1

0.15-0.297 10.8 1.1

0.075-0.15 4.8 0.1

<0.075 3.2 0

Losses 0.9 1.8

TOTAL 100 100

Silica sand (0.85-2.36 mm, Badger Mining Corporation, Taylor, WI) was also used as an

inert packing material within the columns to achieve two separate goals. The first was to serve as

a proppant to increase the permeability within the column. The second goal was to ensure an

adequate hydraulic retention time (HRT) while also providing for substrate exhaustion within a

reasonable laboratory time-scale. The sand was washed overnight in 0.25 M nitric acid, rinsed in

de-ionized water, and completely dried (105°C) prior to use to prevent potential metal leaching

from the particles during the column test. Based on previous column studies and microcosm

experiments (Newcombe and Brennan, 2010), incremental fractions of crab shell between 50%

and 100% (by mass) were mixed with SMC (Table 2-3. Mass of solid packing materials used in

each continuous-flow column.). A 100% sand column was used as an experimental control, and a

column filled with the traditional 90% SMC and 10% limestone chip substrate was also used for

comparison purposes.

16

Table 2-3. Mass of solid packing materials used in each continuous-flow column.

Treatment Column Column Contents (g)

CS SMC Limestone (LS) *Sand

Sand Control 0 0 0 666

100% CS 40 0 0 475

90% CS + 10%SMC 36 4 0 431

80% CS + 20%SMC 32 8 0 465

70% CS + 30%SMC 28 12 0 412

60% CS + 40%SMC 24 16 0 507

50% CS + 50%SMC 20 20 0 459

Traditional 90% SMC + 10% LS 0 36 4 545

*Includes bottom plug + amount mixed as a proppant with substrate + top plug

Each column was packed with a total of 40 g of substrate mixed with 360 g of sand as a

proppant (1:9 substrate to sand ratio by mass) to fill the majority of the column (~700 mL) while

maintaining hydraulic conductivity. Sand plugs were also used at the top and bottom ends of the

column to prevent loss of substrate through the influent and effluent ports.

Column packing materials were analyzed to determine extractable metals which could

potentially leach off during the experiment and contributed to metals concentrations determined

in subsequent analysis. Additionally, the mass of total carbon and nitrogen were determined in

order to obtain the carbon to ratio (C:N), which could be useful in understanding microbial

activity. Substrate analyses for total carbon and nitrogen (combustion method) and extractable

metals (Mehlich 3 method) were performed by the Agricultural Analytical Services Laboratory at

The Pennsylvania State University and are provided in Table 2-4.

17

Table 2-4. Extractible metals and compost analysis of the continuous-flow column packing

materials.

Analyte Crab

Shell

SMC Limestone Sand

Extractable Metals – Mehlich 3 method (reported as mg/kg)

Al BDL 4.48 BDL 9.42

Ca 28,300 24,600 36,600 176

Co BDL BDL BDL BDL

Fe 8.13 70.1 24.5 3.39

K 3020 10,900 26.2 11.0

Mg 2,120 1,700 325 14.6

Mn 44.1 49.1 5.31 1.89

Na 10,800 732 38.0 46.5

P 2,530 907 3.18 4.92

S 1,220 4,500 12.5 5.13

Zn 38.8 28.9 0.533 0.309

Compost Analysis – combustion method (reported on "as is" basis)

pH 8.5 7.7 - -

Organic Matter (%) 42.1 19.8 - -

Nitrogen (%) 4.7 0.6 - -

Carbon (%) 23.9 11.7 - -

Carbon: Nitrogen Ratio 5 21 - -

Calcium Carbonate Equivalence (%) 35.9 10.1 88.9 -

BDL – Below detection limit

Sediment inoculants were deemed unnecessary based off previous work indicating that the

substrate alone provides sufficient bacteria to initiate growth of a microbial consortium diverse

enough to support sulfate reduction (Christensen et al., 1996; Newcombe and Brennan, 2010).

Instead, columns were packed in a non-sterile environment and provided an 8-day incubation

period prior to the initiation of continuous-flow conditions to promote establishment of the

indigenous microbial community.

18

2.3 Continuous-Flow Column Setup

Continuous-flow columns were used to simulate the flow through a VFP containing different

substrates. Columns were constructed using 2 foot long, 1.5 inch diameter polyvinyl chloride

(PVC) pipe (Harvel Clear™ Schedule 40 PVC pipe and fittings, United States Plastic Corp.) with

end-caps of the same material (Figure 2-1). Three holes (1/2 inch diameter) were drilled into the

side of the column (1 inch above the bottom end cap, in the center of the column, and 1 inch

below the top end cap) to facilitate extraction of the substrate material for molecular analysis of

the microbial community at the completion of the experiment. Each hole was filled with a butyl

rubber stopper during packing and continuous-flow conditions. Materials were not washed or

sterilized in any way to simulate realistic conditions expected in construction of a field system.

Columns were flushed with argon gas during packing to remove oxygen and allow for anoxic

packing conditions. Solids were wet-packed into the column in approximately 1 inch lifts with

free-standing source water, beginning with 30 g of sand, followed by the substrate/sand mixture.

An additional sand plug was added to the effluent end of the column to completely fill the

remaining volume. A second end cap was then affixed to the top of each column with PVC

cement. In the same manner as described for the influent water reservoirs, the columns were

covered with opaque black plastic for the duration of the test to prevent the growth of

phototrophic organisms that could potentially produce oxygen within the system.

19

Figure 2-1. Laboratory continuous-flow columns used to treat Klondike-1 MIW.

Source water was pumped from a 50 L reservoir vertically upward through each column

to provide a consistent flow. Water was diverted from the reservoir to 8 separate lines

(Masterflex Tygon® lab L/S® 13, Cole-Parmer) where it was dispensed to each column by a

peristaltic pump consisting of a digital drive and 4-roller cartridge head (Masterflex L/S, Cole-

Parmer) at a set rate of 0.25 mL/min to produce a 16 h HRT.

Two flow-through cells (Cole-Parmer® Universal Flow Through Adapter) were mounted

directly above each column for in-line measurement of pH and ORP. Sampling cells 7.5 inches

in length (Harvel Clear™ Schedule 40 PVC pipe and fittings, ¾ inch, United States Plastic Corp.)

were placed at the effluent end of the columns, subsequent to the flow-through cells, to facilitate

sample collection for analysis (Figure 2-2).

20

Figure 2-2. Schematic of continuous-flow column experimental setup.

Effluent was routed from the sampling cell into an open-topped bin where it was

passively aerated and allowed to settle in a simulated aeration pond with a 45 h HRT, which is

typically used following VFPs to oxidize and remove metals (. The simulated aerobic settling

pond drained into a second sampling cell, identical to the first. Flow from the second sampling

cell drained into a graduated cylinder which recorded the total effluent volume released from the

column over time. Volumes were recorded every 2-5 days and used to determine the actual

volume of water flowing through each column due to slight variations in flow rates throughout

the course of the test.

21

Figure 2-3. Passive aeration and settling were accomplished in bins subsequent to the

continuous-flow columns. Sample cells were used to collect water exiting the

settling bins to monitor increased metals removal from this additional

oxidation/precipitation step after anaerobic treatment. Photo taken on day 36 of the

experiment.

Columns were sampled every 1-7 days during continuous-flow conditions, depending on

the observed rate of changes in water quality. Samples collected from the first sampling cells

(before the aerobic settling pond) were measured immediately for pH, ORP, acidity, alkalinity,

and ammonium, and samples were preserved for later analysis of dissolved metals, Fe speciation,

anions, and dissolved organic carbon (DOC). Samples were concurrently collected from the

second sampling cells and preserved for dissolved metals analysis and Fe speciation. Dissolved

oxygen (DO) measurements were also taken from the aeration bins each time samples were

collected.

22

The columns were designed based on previous research which indicated that 1g of crab

shell had the capacity to treat 1 L of MIW (Robinson-Lora, 2009). Based off this assumption, the

columns were designed to last 116 days before alkalinity exhaustion. However, the substrate

lasted longer than anticipated and the experiment was run for 181 days to ensure complete

exhaustion.

The 6-month column study was conducted in collaboration with another student (Bradley

Sick, Undergraduate Honors Student) and weekly sample analysis was conducted jointly.

Bradley Sick collected and analyzed samples from the sand control, 90% CS + 10% SMC, 70%

CS + 30% SMC, and 50% CS + 50% SMC columns. I collected and analyzed samples from the

100% CS, 80% CS + 20% SMC, 60% CS + 40% SMC, and 90% SMC + 10% limestone columns.

Analysis of anions, DOC, and Fe speciation was also completed by the author.

2.4 Analytical Methods

pH was measured on samples extracted from the sampling cell using a bench-top electrode

(Thermo-ORION) connected to a pH/mV meter (Accumet® Basic AB15, Fisher Scientific). The

pH electrode was calibrated using standard 4.0, 7.0, and 10.0 buffers. Ammonium was also

measured using an electrode (ISE ORION 9512) and the same pH/mV meter, and compared to 1

mg/L and 10 mg/L ammonium standards. DO was measured with an Accumet® Research AR40

meter and a self-stirring BOD probe (Fisher Scientific). Acidity and alkalinity were measured

using titrations as described in Standard Methods for the Examination of Water of Wastewater

(Methods 2310 and 2320; APHA 1998). Endpoints used for these titrations were pH 4.5 for

alkalinity and pH 8.3 for acidity. pH, ammonium, acidity, and alkalinity were all measured

within 4 hours of sample collection.

23

Samples were prepared for dissolved metals analysis by filtering with a 0.45µm filter,

acidifying to pH < 2 with 60-70% HNO3, and sparging with lab air through a 25 gauge needle for

5 minutes (to drive off hydrogen sulfide). These samples were sent to the Pennsylvania State

University Materials Characterization Laboratory to be measured using inductively coupled

plasma-atomic emission spectroscopy (ICP-AES; Leeman Labs PS300UV).

DOC was analyzed using a total organic carbon analyzer (TOC-V CSN, Shimadzu).

Samples for DOC analysis were pretreated with a 0.45µm filter and diluted when necessary to

achieve a minimum sample volume or when salts were determined to exceed the instrument

threshold. Due to the expected high concentrations of volatile fatty acids, the most accurate

method (non-purgeable organic carbon method) for the instrument could not be used. Instead,

total carbon and inorganic carbon were measured separately and the organic carbon was

calculated as the difference between these two measurements. Inorganic carbon analysis was

conducted using a setting of 1.5% acid (2N hydrochloric acid) with a sparge time of 1.0 minutes.

2.5 Conservative Tracer Tests

Conservative (non-partitioning) tracer tests were performed on each column using sodium

chloride to determine the pore volume and also the HRT. A minimum mass for the chloride slug

(mtracer, mg) was calculated using Equation 2-1, where tr is the estimated HRT (min) of the

column, QL is the column flow rate (L/min), and MDL is the method detection limit (mg/L) of

chloride on the IC.

Eq. 2-1

Using an anticipated retention time of 16 hr, a nominal flow rate of 0.0002 L/min, and a

MDL for chloride of 1 mg/L, the calculated minimum tracer mass was 19.2 mg chloride. The

tracer solution was prepared by dissolving 412 mg sodium chloride in 20 mL deionized water. A

24

volume of 2 mL of this tracer solution (25 mg chloride) was injected through the influent port of

each column.

25

3. Continuous-flow column laboratory experiment

The continuous-flow column experiment was conducted in a controlled laboratory

setting from October 2009 until May 2010 (181 days).

3.1 Source Water

In addition to field measurements, water quality parameters were measured regularly

from the Klondike-1 water as it was supplied to the continuous-flow columns. At each sampling

point throughout the test, analysis for pH, alkalinity, acidity, DOC, sulfate, and dissolved metals

were measured. Averages over the course of the experiment were calculated and can be found in

Table 3-1. The average pH and dissolved Fe concentrations are noticeably lower than that

observed in the field when source water was collected (Table 2-1). This is likely caused by the

precipitation of Fe3+

species present in the water. Recalling Eq. 1-4, the precipitation of ―yellow-

boy‖ reduces dissolved Fe and also lowers the pH. Orange-yellow precipitates were noted in the

bottom of the source water storage containers in the laboratory, providing further evidence for

this hypothesis. In addition, it was noted that dissolved Fe concentrations of the influent water

decreased over time as it was maintained within water storage containers in the laboratory. This

indicates the possible continued oxidation of Fe2+

, which could be a result abiotic heterogeneous

Fe2+

oxidation.

26

Table 3-1. Average water quality parameters (taken weekly for the duration of the experiment) of

continuous-flow column influent.

Parameter Influent water average

pH 2.54 ± 0.14

Alkalinity (mg/L as CaCO3) 0

Acidity (mg/L as CaCO3) 330 ± 76

Al (mg/L) 2.73 ± 0.42

Fe (mg/L) 62.3 ± 21

Mn (mg/L) 36.2 ± 5.6

Co (mg/L) 0.42 ± 0.05

Ni (mg/L) 0.92 ± 0.38

Zn (mg/L) 0.26 ± 0.03

SO42-

(mg/L) 994 ± 180

Dissolved organic carbon (mg/L) 2.75 ± 1.4

3.2 Conservative Tracer Tests

Tracer tests were performed near the end of the experiment in order not to disturb the microbial

community and potentially influence treatment performance. Between April 17, 2010, and May

3, 2010, tracer tests were completed on all columns, except the sand control which was completed

approximately one month earlier.

Tracer test results were used to estimate the HRT, which was then used to calculate the

dispersion number and the effective pore volume of each experimental column. HRT and

effective PV varied considerably between columns (Table 3-2). Based on the calculated

dispersion numbers, all columns exhibited low dispersion (d<0.05) flow characteristics except the

sand control, which was classified as moderate dispersion. Additional detail on the calculations

used and tracer curves can be found in Appendix A. Tracer tests were used to present the data

more accurately, by allowing a comparison among columns based upon the volume of water

27

treated as opposed to comparing systems at different time points, when flow might not have been

consistent.

Table 3-2. Flow characteristics of continuous-flow columns treating Klondike-1 MIW, measured

using tracer tests at the completion of the 181-day experiment.

Treatment Column

Dispersion

number

Calculated

effective

pore volume

Flow rate

during

tracer test

Hydraulic

retention time

(mL) (mL/min) (hr)

Sand Control 0.065 190.5 0.300 10.6

100% CS 0.041 275.9 0.261 17.6

90% CS + 10% SMC 0.035 260.4 0.252 17.2

80% CS + 20% SMC 0.039 296.8 0.269 18.4

70% CS + 30% SMC 0.026 246.8 0.253 16.3

60% CS + 40% SMC 0.045 216.7 0.282 12.8

50% CS + 50% SMC 0.027 223.6 0.273 13.6

Traditional 90% SMC + 10% LS 0.022 221.1 0.284 13.0

3.3 pH, Alkalinity and Acidity

Measurements for pH were taken using a bench top electrode on samples collected from the

effluent sample cell, and also using in-line electrodes mounted in flow-through cells near the

column effluent port (Figure 2-2). As expected, measurements from the in-line electrodes were

slightly lower, but mirrored trends noted with the bench top electrode, with the exception of the

100% CS column (see Appendix B). Increased pH was expected in samples extracted from the

anoxic column environment when biologically-produced CO2 was partitioned into the air from

the effluent water. As bench-top analysis was conducted quickly upon effluent water extraction

from the sample cell, a large difference was not noted. Results provided below are from the

bench top electrode only.

All treatments increased the pH of the water from influent values (average pH 2.54) to

above 6.0. Columns containing any fraction of crab shell maintained pH above 5.0 longer than

28

the traditional substrate (90% SMC + 10% LS), which was only able to sustain this level of

treatment for 70 days (80 PV). The 100%, 90%, and 70% crab shell columns all maintained pH

above 5.0 for almost twice as long as the traditional substrate (Table 3-3). Results in Table 3-3

also indicate reduced performance of the 60% and 80% crab shell columns. Within the first four

weeks of the experiment, both of these columns experienced oxygen intrusions within the

column. This occurred due to three-way valves being left in the wrong position after sampling

which led to a pressure build-up within the column. The pressure was eventually released by the

ejection of a butyl rubber stopper from the side of the column. On both instances, this occurred

over night and was not found until the next morning. However, over the course of the night,

water had drained from the column above the point of the hole and had also been pumping out of

the open hole in the side of the column. This likely led to the disruption of the anaerobic

microbial consortium, which affected subsequent decreases in treatment capacity. Due to these

circumstances, the 60% and 80% crab shell columns are not considered to adequately portray

expected treatment efficiencies under normal conditions.

Table 3-3. Maximum pH and duration of neutralization capacity achieved using different

substrates to treat Klondike-1 MIW in the continuous-flow column experiment.

Treatment Column Max.

pH

pH sustained

above 5.0

pH returned to

influent value

PV (days) PV (days)

Sand Control 2.64 0 (0) 0

100% CS 7.62 141 (146) 176+ (180+)

90% CS + 10% SMC 7.26 131 (117) 199+ (180+)

80% CS + 20% SMC 7.26 95 (104) 137 (153)

70% CS + 30% SMC 7.16 137 (124) 195+ (180+)

60% CS + 40% SMC 7.34 108 (90) 196 (160)

50% CS + 50% SMC 7.15 114 (96) 161 (132)

Traditional 90% SMC + 10% LS 6.38 79 (70) 116 (97)

+ indicates value when the experiment ended, thus the potential for additional treatment

capacity

29

Sufficient alkalinity was generated in all treatments to completely neutralize the acidity

present in the influent water (average 329 mg/L as CaCO3) from the onset of continuous-flow

conditions (Figure 3-1). The column containing the traditional substrate mixture achieved a

maximum alkalinity generation of 260 mg/L as CaCO3 at the first sampling point following an

eight day incubation period whereas columns containing any fraction of crab shell substrate

produced significantly higher alkalinity initially, with maximum concentrations near 5,800 mg/L

(not shown). Columns containing crab shell were able to maintain net alkaline conditions for

longer than the traditional treatment substrate, nearly twice as long for the 100%, 90%, and 70%

crab shell columns.

30

Figure 3-1. Alkalinity generation and acidity data from continuous-flow columns treating MIW

from the Klondike-1 site.

-1000

-800

-600

-400

-200

0

200

400

0 50 100 150 200

Ac

idit

y (m

g/L

as

Ca

CO

3)

Pore Volumes

0

200

400

600

800

1000

0 50 100 150 200

Alk

alin

ity (m

g/L

as

Ca

CO

3)

Inf luent

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC

Traditional 90% SMC + 10% LS

B

A

31

3.4 Metals Removal

Dissolved metals were monitored at two locations: in the effluent of the continuous-flow column

and also after passive aeration and settling. The following breakthrough curves render data in the

form of normalized concentrations, defined as the measured concentration divided by the inlet

concentration (

, plotted against PV of water treated. Normalized concentrations are used to

negate the effects of fluctuations in influent water quality, which occurred due to seasonal

changes at the Klondike-1 site and also due to the precipitation of Fe in the source water

container. In addition, normalized curves provide an easy assessment of treatment efficiency as

a normalized concentration of zero represents 100% treatment efficiency (100% removal) with a

decrease in treatment efficiency realized as normalized concentrations approach 1.0 (0%

removal). Normalized concentrations greater than one indicate that the columns were adding

elements to the water. This was likely caused by desorption and re-suspension of adsorbed metal

precipitates once the pH dropped below that corresponding to minimum solubility for the

respective precipitate. For the results and discussion to follow, breakthrough is defined as

effluent concentration exceeding 50% of the influent concentration (corresponding to 0.5 on the

normalized concentration plot).

3.4.1 Primary Metals

Data for dissolved Al (average influent concentration of 2.67 mg/L) presents very clear

breakthrough curves, which appear to be dependent on pH (Figure 3-2). Dissolved Al was

completely removed by all treatment columns until ~70 PV when the treatment capacity of the

traditional substrate became exhausted. Figure 3-2A reveals that columns containing 70%, 90%

and 100% crab shell lasted twice as long as the traditional substrate before exhaustion of Al

32

treatment capacity. Dissolved Al concentrations measured after passive aeration and settling

were very similar to those measured in the column effluent, indicating that contact with oxygen in

the settling pond does not affect the removal of Al (see Appendix C). In addition, results from

the sand control column (normalized concentration =1) suggest that any Al removal was caused

by the treatment substrate mixture and not other experimental conditions.

Figure 3-2. Breakthrough curves for dissolved Al (A) and pH measurements (B) taken after

continuous-flow columns treating Klondike-1 MIW.

2

3

4

5

6

7

8

0 50 100 150 200

pH

Pore Volumes

0

1

2

3

4

5

6

7

8

9

10

0 50 100 150 200

Al (C

/C0)

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


B

A

33

Results from the column effluent show that Fe (average influent concentration of 62.6

mg/L) was partially removed within the column substrate for all treatment columns (Figure

3-3A). Some substrate mixtures achieved complete removal (all except Sand Control, 100% CS,

and Traditional 90% SMC + 10% LS), however this trend is not consistent past 25 PV in any of

the columns (Figure 3-3A). In comparison, Fe removal subsequent to passive aeration and

settling was sustained significantly longer. Breakthrough (to 50% of influent concentration) of

dissolved Fe after the settling ponds occurred at approximately 95 PV for the traditional substrate,

and did not occur until ~150 PV for the majority of the columns containing any amount of crab

shell (Figure 3-3B). Figure 3-3A also reveals that the sand column was affording some removal

of dissolved Fe (normalized concentration <1), indicative of experimental conditions playing a

role in Fe removal.

34

Figure 3-3. Breakthrough curves for dissolved Fe measured after continuous-flow columns

treating Klondike-1 MIW (A) and after subsequent passive aeration and settling (B).

0.0

0.5

1.0

1.5

2.0

2.5

0 50 100 150 200

Fe

(C

/C0)

Pore Volumes

B

0.0

0.5

1.0

1.5

2.0

2.5

0 50 100 150 200

Fe

(C

/C0)

Sand Control 100% CS90% CS + 10% SMC 80% CS + 20% SMC70% CS + 30% SMC 60% CS + 40% SMC50% CS + 50% SMC Traditional 90% SMC + 10% LS

A

35

Removal of dissolved Mn (average influent concentration of 36.1 mg/L) occurred over a

much shorter time. The traditional substrate reached breakthrough (to 50% of influent

concentrations) within 5 PV as compared to the columns containing any amount of crab shell

which lasted from 10-17 PV. The column containing 100% crab shell was able to sustain partial

removal (normalized concentration <1.0) of Mn for up to 54 PV, over five times longer than the

traditional substrate which exhausted within 10 PV. Dissolved Mn concentrations measured after

passive aeration and settling were very similar to those measured in the column effluent, with a

minor shift to the right (increased PV) likely caused by the time required for the water to pass

from the first sampling cell to the second at the given flow rate (see Appendix C). In addition,

results from the sand control column (normalized concentration =1) suggest that any Mn removal

was truly caused by the treatment substrate mixture, not experimental conditions.

Figure 3-4. Breakthrough curves for dissolved Mn measured after continuous-flow columns

treating Klondike-1 MIW.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0 50 100 150 200

Mn

(C

/C0)

Pore Volumes

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


36

3.4.2 Trace Metals

The breakthrough curves for cobalt (average influent concentration of 0.42 mg/L) and zinc are

similar to those presented for Al (Figure 3-5 and Figure 3-2) with respect to the time (PV) at

which breakthrough occurred. Each column appears to reach complete breakthrough at similar

PVs in the three plots. From the breakthrough curves presented, it appears that dissolved zinc is

not well removed within the treatment columns. However, the appearance of low removal

efficiency (normalized concentration >0.5) is caused by low concentrations of zinc in the source

water (average concentration of 0.26 mg/L) coupled with low detection limits for the analytical

method (0.20 mg/L). All raw data indicating concentrations below the detection limit were

reported as 0.20 mg/L, thus the lowest achievable normalized concentration would be 0.77

(calculated as

.

37

Figure 3-5. Breakthrough curves for dissolved cobalt (A) and zinc (B) measured after


Complete nickel removal was not achieved during the continuous-flow column

experiment. However, approximately 35% removal efficiency was maintained for a considerable

period in all of the treatments, with a decrease in removal efficiency occurring first in the

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

0 50 100 150 200

Zn

(C

/C0)

Pore Volumes

Sand Control100% CS90% CS + 10% SMC80% CS + 20% SMC70% CS + 30% SMC60% CS + 40% SMC50% CS + 50% SMCTraditional 90% SMC + 10% LS

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 50 100 150 200

Co

(C

/C0)

B

A

38

traditional substrate at around 50 PV (Figure 3-6). Treatment substrates containing any fraction

of crab shell maintained 35% removal efficiency for 100 PV (80% CS + 20% SMC, 60% CS +

40% SMC, and 50% CS + 50% SMC columns) up to 150 PV (100% CS, 90% CS + 10% SMC,

and 70% CS + 30% SMC columns).

Figure 3-6. Breakthrough curves for dissolved nickel measured after continuous-flow columns

treating Klondike-1 MIW.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 50 100 150 200

Ni (C

/C0)

Pore Volumes

Sand Control 100% CS

90% CS + 10% SMC 80% CS + 20% SMC

70% CS + 30% SMC 60% CS + 40% SMC

50% CS + 50% SMC Traditional 90% SMC + 10% LS

39

3.5 Sulfate Reduction

Sulfate was monitored via two separate analytical methods, IC and ICP, which yielded similar

results, indicating almost no sulfate reduction was occurring within the treatment columns.

However, within days of initiation of continuous-flow conditions, visual inspection of the

columns showed black precipitates in all except the sand control column. These black

precipitates remained within the columns for the duration of the experiment (Figure 3-7).

Figure 3-7. Experimental columns photographed after 84 days of continuous-flow conditions

with Klondike-1 MIW. Note black precipitates which formed in all except the sand

control column. The remaining four treatment columns (not shown) also displayed

the formation of black precipitates.

100% CS 90% CS 80% CSSand

40

Black precipitates can be indicative of metal sulfide precipitation, and thus imply a

reduced sulfur species could have been present within the columns and also in the effluent water.

Sulfate data, and a discussion on possible complications leading to inaccurate results, is included

in Appendix D.

3.6 Carbon and Nitrogen Species

Carbon and nitrogen are among a few key nutrients required by microorganisms to sustain sulfate

reduction. If excess polymeric substrate (chitin or cellulose) is hydrolyzed and fermented, carbon

and nitrogen can be released from the system into the effluent water and potentially negatively

affect downstream water quality. DOC was measured at each sampling point (Figure 3-8).

Maximum DOC concentrations for each column were observed immediately following a week-

long incubation period (Figure 3-8 inset). While the average DOC in the influent water was 2.75

mg/L, the traditional SMC-limestone substrate reached a maximum of 330mg/L, and the 100%

CS column reached a maximum of 5300 mg/L.

41

Figure 3-8. Dissolved organic carbon measured in column effluent during continuous-flow

column test treating MIW from the Klondike-1 site. Inset graph shows maximum

values achieved at beginning of experiment.

Similar to the trend noted with DOC, maximum ammonium concentrations were also

measured immediately following incubation for each column (Figure 3-9). The crab shell

columns achieved ammonium concentrations (28.4-32.9 mg/L NH4+-N) nearly three times as high

as that achieved in the traditional SMC and limestone substrate column (11.9 mg/L NH4+-N).

0

5

10

15

20

25

30

35

40

45

50

0 50 100 150 200

DO

C (

mg

/L)

Pore Volumes

Influent

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


0

1,000

2,000

3,000

4,000

5,000

0 5 10

DO

C (

mg

/L)

Pore Volumes

42

Figure 3-9. Ammonium measured from column effluent during continuous-flow column test

treating MIW from the Klondike-1 site.

3.7 Other Cations

Crab shell substrate sold commercially for remediation purposes contains a considerable amount

of fines as shown in the particle distribution analysis (Table 2-2). These fines are beneficial as

they provide a high surface area for increased dissolution of calcium carbonate and other nutrients

which initiate several processes in newly installed systems, including the stimulation of the

microbial community. However, during initial start-up of treatment systems containing crab

shell, high concentrations of cations have been noted (Robinson-Lora and Brennan, 2010a). After

the fines are dissolved and/or flushed out of the system, a steady-state condition is achieved

within the treatment system and concentrations stabilize (data plots provided in Appendix E).

High concentrations of some minerals or mineral salts could be of concern to aquatic ecosystems

when being released into natural waterways. In order to determine the risk associated with these

0

5

10

15

20

25

30

35

0 50 100 150 200

Am

mo

niu

m (

mg

/L N

H4+-

N )

Pore Volumes

Influent

Sand Control

100% CS

90% CS + 10%SMC

80% CS + 20%SMC

70% CS + 30%SMC

60% CS + 40%SMC

50% CS + 50%SMC

Traditional 90%SMC + 10% LS

43

mineral salts, calcium (Ca), potassium (K), magnesium (Mg), sodium (Na), and phosphorous (P)

were monitored during the course of the continuous-flow column experiment. All maximum

concentrations were noted on day 0 of the experiment, after the one week incubation period.

After ~10 PV, concentrations reached a stable concentration (Table 3-4). While maximum

concentrations from the crab shell-containing columns were nearly 3 times larger for Ca and Mg,

and 10 times larger for Na than the traditional substrate mixture, average values stabilized in

close approximation to each other for all of the columns. Initially K values were higher in the

traditional 90% SMC + 10% LS substrate than any other treatment column, but these values also

stabilized to similar concentrations for all treatment columns. P released from the traditional

substrate was comparable with the crab shell-containing mixtures initially, but quickly dropped to

considerably lower levels.

44

Table 3-4. Average concentration after 10 pore volumes (PV) and maximum concentrations of

Ca, K, Mg, Na, and P noted in influent water and effluent from continuous-flow

columns treating Klondike-1 MIW with substrates containing mixtures of crab shell

(CS), spent mushroom compost (SMC), and/or limestone (LS).

Average concentrationa (mg/L) after 10 PV

[max. concentration]

Treatment Column Ca K Mg Na P

Influent 109

[126]

4.65

[5.49]

102

[122]

5.78

[6.87]

0.27

[0.95]

Sand Control 109

[127]

7.14

[27.3]

100

[126]

5.84

[6.81]

0.24

[0.98]

100% CS 187

[1240]

7.46

[144]

104

[216]

6.77

[637]

7.69

[17.8]

90% CS + 10% SMC 183

[1440]

9.37

[300]

103

[272]

6.57

[859]

7.47

[32.5]

80% CS + 20% SMC 170

[1095]

7.37

[225]

103

[207]

6.49

[681]

5.82

[24.5]

70% CS + 30% SMC 194

[1590]

7.90

[358]

105

[277]

6.65

[770]

7.18

[25.3]

60% CS + 40% SMC 175

[906]

9.33

[239]

105

[184]

6.50

[560]

4.71

[15.6]

50% CS + 50% SMC 173

[1170]

8.51

[289]

103

[231]

6.46

[492]

4.15

[24.2]

Traditional 90% SMC + 10% LS 153

[414]

9.03

[370]

104

[127]

6.0

[43]

0.33

[21.1]

Recommended Tolerance Limit for Fish

Cultureb (mg/L)

160 5 15 75 n/a

a Measured by inductively coupled plasma-atomic emission spectroscopy (ICP-AES)

b Meade (1989)

45

4. Discussion

4.1 Alkalinity

Alkalinity generation in systems containing crab shell has been attributed to three mechanisms:

dissolution of calcium carbonate contained within the shell matrix, generation of VFAs during

fermentation, and bicarbonate produced during sulfate reduction (Robinson-Lora and Brennan,

2009b; Robinson-Lora and Brennan, 2010a). Research conducted previously in our lab has

shown that the surface area of crab shell is over 14 times greater than limestone chips (0.106-0.85

mm diameter) (Robinson-Lora and Brennan, 2009b; Robinson-Lora and Brennan, 2011). In

addition, it has been proposed that the biogenic character of the crab shell-associated carbonates

might be more reactive than other forms of calcium carbonate. These differences likely account

for the increased dissolution of crab shell as compared to the limestone contained in the

traditional treatment column in the current experiment. A strong correlation (R2=0.99) was noted

previously (Newcombe and Brennan, 2010) between alkalinity generation and the amount of crab

shell within the treatment system. The same relationship was noted during this experiment,

however with a lower correlation (R2=0.97) due to the high total alkalinity noted within the 70%

CS +30% SMC column.

The data presented here further supports previous results indicating that a portion of the

alkalinity was produced biotically in columns containing crab shell. Theoretical mass of

carbonate supplied by the substrate material in each column was calculated (Eq. 4-1) using the

mass of each substrate and calcium carbonate equivalence (CCE) data.

46

Eq. 4-1

Where is the theoretical mass of carbonate (in mg) supplied to the system

m is mass (in mg) of substrate (from Table 2-3)

CCE (%) is the calcium carbonate equivalence of the substrate (from Table 2-4)

In addition, estimates for the total CaCO3 released from each treatment column were computed by

integrating the area under the curves for experimental alkalinity and acidity data (Eq. 4-2).

Eq. 4-2

Where m is mass (in mg)

n is the sampling iteration

is alkalinity production (in mg/L as CaCO3)

is acidity neutralized (in mg/L as CaCO3)

is the PV of MIW treated between the current sampling iteration

(n) and the previous sampling iteration (n-1)

is the effective pore volume (L) for the column (found in Table 3-2)

Assumption: The alkalinity and acidity (mg/L as CaCO3) were assumed to be

constant between sampling iterations

Figure 4-1 shows that additional alkalinity was produced (above theoretical mass based

on CCE data) for all of the columns containing crab shell. Assuming that the entire mineral

portion of the crab shell matrix disassociated within the treatment columns, it appears as though

12-25% (depending on the specific column) of the alkalinity generated was above the calculated

47

theoretical mass and may therefore be attributed to biotic activity. The assumption that complete

dissolution occurred is not necessarily valid, however. Most likely some of the carbonate

contained within the crab shell matrix would not be accessible until complete degradation of the

matrix. After completion of the experiment, substrate material was extracted from each column

for future microbial analysis. During that time, it was noted that substrate materials could be

visibly identified as CS, SMC and LS, respectively, within the substrate mixtures. Thus, it is not

suggested that complete degradation occurred during this experiment, so it is likely that even

greater portions of the alkalinity produced in the crab shell columns can be attributed to

biological activity.

Figure 4-1. CaCO3 calculated from experimental alkalinity and acidity data versus theoretical

CaCO3 data for substrate mixtures.

0

5,000

10,000

15,000

20,000

CaC

O3

(mg)

Experimental Alkalinity and Acidity Measurements

Theoretical CaCO3 within Substrate Matrix

48

Previous research (Robinson-Lora and Brennan, 2009b) determined that fermentation and

sulfate reduction by-products each accounted for 25% of the alkalinity within crab shell MIW

treatment systems, with the remaining 50% being a result of dissolution of crab shell-associated

carbonate. These results were from a static microcosm test, and it was noted that alkalinity

production was likely limited by exposure of the surface area of particles. Thus, higher alkalinity

generation from the dissolution of carbonate materials is expected under continuous-flow

conditions. The data reported here is likely more representative of the fraction of alkalinity

generated by biotic and aboitic mechanisms.

Figure 4-1also illustrates that the theoretical alkalinity generating capacity of the

traditional SMC + LS substrate mixture was not exhausted during the experiment, although the

experimental data (Figure 3-1) suggest that it was. DOC data from this treatment column imply

fermentation was occurring, thus it can be assumed that some of the calculated alkalinity

generation was caused by that mechanism in addition to dissolution of calcite. It is possible that

the limestone within this column became armored with metal precipitates and was thus

inaccessible for further dissolution, leading to a portion of the alkalinity generating potential

remaining within the column even after apparent exhaustion. This has been documented in the

literature as a main problem for MIW treatment systems containing limestone (Ziemkiewicz et al.

1997). It is also important to note that the limestone used in this experiment was smaller than the

limestone chips typically mixed with SMC in field treatment systems (0.420-0.841 mm vs.

roughly 9 mm). The lower surface area of the limestone used for field systems could lead to even

slower dissolution for full-scale treatment systems, making crab shell substrates even more

advantageous.

49

4.2 Metals Removal

The ability of each column to remove metals of interest expired at different times throughout the

experiment. The treatment capacity of each substrate mixture was calculated as total metals

removed (Table 4-1), with treatment defined as the period when pH5. These results indicate

that mixtures containing any amount of crab shell can remove a minimum of 1.5 times the total

amount of metals from the Klondike-1 MIW as compared to the traditional substrate.

Table 4-1. Metal treatment capacity for each continuous-flow column utilizing 40 g substrate

mixtures to treat Klondike-1 MIW.

Treatment Column

Cumulative Metals Treateda until pH<5

(mg removed)

MIW

Treateda

(L) Al Co Feb Mn Ni Zn Total

100% CS 102 15 2,140 5 16 10 2,290 39

90% CS + 10% SMC 89 14 1,920 57 14 9 2,100 33

80% CS + 20% SMC 74 12 1,810 3 12 8 1,920 28

70% CS + 30% SMC 94 14 2,070 7 15 9 2,210 34

60% CS + 40% SMC 57 10 1,480 26 6 6 1,590 23

50% CS + 50% SMC 65 10 1,530 3 7 7 1,620 26

Traditional 90% SMC + 10% LS 41 4 930 0 4 2 974 18

ausing 40 g of substrate

bCalculated after passive aeration and settling, all others calculated directly from column effluent

To further aid in understanding metal removal mechanisms, a mass balance was

conducted to determine the amount, if any, of each dissolved metal retained within the columns at

the end of the experiment (Appendix F). The results of the mass balance indicate that no

considerable amounts of Al or Mn were retained within any of the columns but that Fe, Co, Ni,

and Zn were retained to varying degrees, including the sand control for Fe (discussed in detail in

the following sections). Removal of Al, Fe, Mn, and trace metals within the 70% CS +30% SMC

column achieved similar rates to those noted within the 100% CS and 90% CS + 10% SMC

columns, with total metals removed exceeding all other treatment columns. In addition, the 70%

50

CS + 30% SMC column did not experience negative effects noted in the columns containing

higher concentrations of crab shell during the attempted oxidation of Fe (discussed below). Thus,

the 70% CS + 30% SMC substrate mixture was selected as the best combination of those tested

for the Klondike-1 water quality conditions. A discussion on potential removal mechanisms for

specific metals is provided below.

4.2.1 Al

Although this research was aimed to determine the effectiveness of crab shell mixtures for

treating high-strength MIW, previous crab shell experiments with low-strength MIW were

actually conducted with higher initial concentrations of dissolved Al (10-14 mg/L) (Daubert and

Brennan, 2007; Robinson-Lora, 2009b, 2010a; Newcombe, 2010). The removal of Al in this

experiment was expected to display similar results. Indeed, as previously noted, a direct

correlation between pH and Al removal was observed, with breakthrough occurring in the

columns when pH was no longer maintained above 4.0-4.25, which corresponds with the

minimum solubility for Al(OH)3 for the given conditions (Kso=10-33

, Snoeyink and Jenkins,

1980). Although the data, and previous publications regarding similar systems, indicate likely

removal of Al as a hydroxide, a geochemical model (Visual MINTEQ ver. 2.53) was used to

predict saturation indices for the Al species present given the conditions encountered during the

present experiment and diaspore (AlOOH) was the only reported precipitate. Regardless of the

actual Al oxy(hydroxide) species, it is evident that Al was removed from the treatment columns

based on the solubility of the precipitate, and that pH will dictate the longevity of treatment.

Additional research should be conducted to determine removal efficiency at concentrations >

10mg/L dissolved Al.

51

4.2.2 Fe Removal Within Treatment Column

It is presumed that oxidation of Fe within the storage container resulted in ferric iron species,

including oxides/hydroxides, entering the columns. It is also feasible that nanoparticles of Fe

(oxy)hydroxides (Silvester et al., 2005; Cravotta, 2008) were measured as a portion of the

dissolved Fe ( if they were smaller than the 0.45µm filter opening size). Figure 3-7 shows a

noted orange discoloration near the influent port of the sand column implying that Fe

oxides/hydroxide precipitates were either physically retained (filtered) by the solid packing

materials or some sort of sorption mechanism occurred between the precipitates and the surface

of the packing materials. Despite the fact that the orange Fe (oxy)hydroxide precipitation was

only evident within the sand column, it likely occurred within the other columns as well, with the

orange precipitates not visible due to blackening of the columns.

A mass balance of the dissolved Fe entering and exiting the system indicated that Fe was

retained within all columns (Table 4-2). In fact, the data indicates better overall retention (after

breakthrough) of Fe within the sand and traditional columns than the crab shell-containing

columns. It is most likely that the portion of Fe retained within the sand column is in the form of

the ferric iron (oxy)hydroxides, as no other mechanisms for metal removal were noted. However,

it is postulated that within the other columns a portion of the oxidized Fe was subsequently

reduced to ferrous iron. ORP data indicates reducing conditions were present in all of the

treatment columns. Work by Roden and Urrutia (1999) indicate that biological reduction of ferric

iron on the surface of the precipitates is likely under anaerobic conditions, resulting in the release

of ferrous species back into solution. This could account for the lower amount of Fe retained in

columns containing crab shell as reducing conditions were noted to have lasted longer within

those systems (Appendix B, ORP data), although no direct correlation can be made.

52

Table 4-2. Fe retained within treatment columns (after breakthrough) at completion of

continuous-flow column test treating Klondike-1 MIW.

Treatment Column Fe (all m are in mg)

min mout mretained % retained

Sand Control 2,870 1,770 1,100 38%

100% CS 2,900 2,090 810 28%

90% CS + 10% SMC 3,020 2,090 930 31%

80% CS + 20% SMC 2,910 2,260 650 22%

70% CS + 30% SMC 2,890 1,950 940 33%

60% CS + 40% SMC 2,852 2,250 602 21%

50% CS + 50% SMC 2,880 2,060 820 28%

Traditional 90% SMC + 10% LS 2,880 1,790 1,090 38%

Removal of dissolved Fe within the actual treatment columns could have been attained by

precipitation with sulfides, sorption onto Al or Fe (oxy)hydroxides (Urrutia et al., 1999; Jeon et

al., 2003; Silvester et al., 2005; Larese-Casanova and Scherer, 2007) or to the surface of metal

sulfides (Jong and Parry, 2004). Visual MINTEQ predicted the formation of mackinawite (FeS)

for conditions where sulfide was present. Each of these mechanisms is expected to cease or slow

considerably by pH 5. As a drop is noted in pH to 5 or below, the following occur

simultaneously: Al precipitates are dissolving, releasing any adsorbed Fe; the sorption edge is

reached for Fe on Fe (oxy)hydroxide and FeS surfaces (Jong and Parry, 2004; Lerese-Casanova

and Scherer, 2007); and FeS precipitates are also expected to undergo dissolution (based on

modeling with Visual MINTEQ, Appendix G). In addition, optimal conditions for SRB include

pH between 6-8. Below pH 4 sulfate reduction is expected to slow considerably from rates

achieved at circum-neutral pH (Jong and Parry, 2006), leaving little sulfide present in the pore

water. As mentioned before, SRB can only utilize short chain organic acids and alcohols, which

require upstream cellulose and chitin degradation by other microorganisms. Cellulolysis is most

effective at pH 6 and above (Logan et al., 2005) and biological chitin hydrolysis has also been

optimized around pH 5-6, depending on the organism (Kapat et al., 1996; Roy et al., 2003;

53

Ramírez-Coutiño et al, 2006). Thus, around pH 5, cellulose and chitin degradation has slowed

and could potentially be limiting the carbon source for SRB.

4.2.3 Fe Removal After Passive Aeration and Settling

The full-scale treatment system at the Klondike-1 site, utilizing the traditional 90% SMC and

10% LS substrate mixture, experienced complications when a layer of Fe oxides formed on the

top of the organic substrate layer and clogged the VFP. One benefit noted previously for the crab

shell-containing systems is that the reducing conditions achieved within the anaerobic zone are so

strong as to prevent precipitation of Fe oxides. Although the reducing conditions are desirable

when water is within the VFP, our research shows it can cause problems upon attempted

oxidation. The dissolved Fe results after settling (Figure 3-3B) Figure 3-3. Breakthrough curves

for dissolved Fe measured after continuous-flow columns treating Klondike-1 MIW (A) and after

subsequent passive aeration and settling (B). indicate inefficient treatment within the 100% and

90% CS columns for the first 75-85 PV. This is hypothesized to be a result of the extremely

reducing conditions achieved within columns containing large fractions of crab shell. This

phenomenon was not noticed with fractions of crab shell 80%. This is an important implication

as there appears to be a point above which crab shell addition can inhibit the oxidation of Fe in

subsequent aeration steps.

4.2.4 Mn

Complete removal of dissolved Mn occurred for only a short period near the beginning of the

experiment, when the pH remained above 7. At near-neutral pH, saturation of rhodocrosite

(MnCO3) can be expected (Cravotta, 2008) as well as the adsorption onto Al or Fe

54

(oxy)hydroxide precipitates (reviewed in Cravotta and Trahan, 1999). Geochemical modeling

with Visual MINTEQ indicated saturation with rhodocrosite and MnHPO4 at pH from 5.5-7.0,

which is in agreement with previous work (Robinson-Lora and Brennan, 2011). Recent studies

have indicated sorption and/or (co)precipitation as the primary removal mechanisms for dissolved

Mn in anaerobic treatment systems containing crab shell (Robinson-Lora and Brennan, 2010b,

2011). In that investigation, however, interferences in analytical methods due to high calcite

content prevented a conclusive identification of the exact mechanism for Mn removal. The

results of this experiment indicate any of the mechanisms mentioned could be responsible, and no

evidence presented here provides additional insight into the likelihood of one process over

another.

4.2.5 Trace Metals

Several mechanisms could explain the removal of Co, Ni, and Zn within the treatment columns.

Research has shown that surface interactions with precipitated Al, Fe, and Mn (oxy)hydroxides

can lead to the sorption or co-precipitation of trace metals in AMD treatment systems (Lee et al.,

2002; Kairies et al., 2005; Arai, 2008; Peltier et al., 2010; Miller et al., 2011). Adsorption is

regulated by the surface charge of the individual (oxy)hydroxide precipitate and can be affected

by crystallinity, the pH of the associated water, and the presence of other ionic species such as

SO42-

and Cl- (Micera et al., 1986). Due to the short duration of sustained Mn removal recorded

within the treatment columns, surface interactions with this metal were not expected to play a

large role in the removal of trace metals. Precipitation of trace metals as sulfides or even in other

mineral forms are potential removal mechanisms, given appropriate conditions. In fact, it has

been shown that sulfides have a higher affinity for precipitation with trace metals over Fe

(Machemer and Wildeman, 1992; Jong and Parry, 2004), and that adsorption and/or co-

55

precipitation of trace metals with metal sulfide precipitates is likely (Jong and Parry, 2004;

Charriau, 2011).

As noted previously, a mass balance was conducted to determine the amount, if any, of

each dissolved metal retained within the columns at the end of the experiment (Appendix F).

Figure 4-2 shows that all treatment columns retained some Co and Ni after breakthrough,

indicating another removal mechanism in addition to surface interactions. In order to help

determine possible precipitates formed, Visual MINTEQ was used to model the environmental

conditions and predict saturation indices. Results for Co and Ni species at various sulfate/sulfide

ratios (to postulate different sulfate reduction rates) indicated CoS and NiS as the only

precipitates for any of the sulfide concentrations investigated (Appendix G). These metal sulfides

are not expected to become soluble until pH falls below 2 (Jandová, 2005; and Visual MINTEQ

models), which never occurred in the treatment columns as the influent water was maintained

around pH 2.5, thus they would be expected to be retained within the treatment columns. An

assumption can be made with respect to the amount of metal removed via precipitation as a metal

sulfide based on the % retained within the column (Figure 4-2). Columns containing crab shell

retained 41-61% of the total Co, 10-31% of the total Ni, and from 0-66% of the total Zn;

compared to 15%, 6%, and 0% for Co, Ni, and Zn, respectively, retained within the traditional

SMC and limestone treatment column.

56

Figure 4-2. Percent of total trace metal loading retained (after breakthrough) within treatment

columns at completion of continuous-flow column experiment treating Klondike-1

MIW.

Co and Ni findings also imply that a minimum of one other removal mechanism was

occurring within the treatment columns: sorption to metal (oxy)hydroxides. The breakthrough of

Co occurred as the pH approached and fell below pH 5, and Ni breakthrough curves follow

shortly after, as the pH continued to drop (Figure 3-2,Figure 3-5, and Figure 3-6) This

corresponds directly to the time when dissolved Al was first detected in the effluent of each

column (just prior to Al breakthrough). However, pH 5 also corresponds generally to the

adsorption edges for these trace metals associated with Fe (oxy)hydroxides (Spark et al., 1995;

Kairies et al., 2005, Arai et al., 2008). Thus, the desportion from Al and/or Fe (oxy)hydroxides

could also be involved in the release of Co and Ni from the treatment systems at breakthrough. A

small increase in the normalized concentration of Ni after ~15 PV was noted to coincide with Mn

breakthrough for each of the treatment columns. Ni removal after 15 PV was noticeably lower

(larger normalized concentrations) than that achieved within the first 15 PV for all columns. This

-10%

0%

10%

20%

30%

40%

50%

60%

70%

80%

Co Ni Zn

% R

eta

ined

in

Co

lum

n

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


57

implies that surface interactions or co-precipitation could also be a factor responsible for Ni

removal as long as Mn removal is sustained.

If additional sulfide were available in solution when Co and Ni were desorbed from Al,

Fe, and Mn (oxy)hydroxides, metal sulfides could likely have formed. The evidence of

breakthrough (dissolved Co and Ni exiting the column) around pH 5 implies that adequate sulfide

was not present to consume the desorbed metals released. As discussed previously, SRB activity

becomes limited below pH 5-6 as cellulose and chitin degradation ceases and SRB begin to slow

their metabolism. These very distinct removal mechanisms for trace metals, which all cease as

the pH of the system drops below 5, make it difficult to determine which mechanism was

dominant in these systems.

The percent of Co retained displays a linear correlation (R2 = 0.960) to the amount of crab

shell contained in the column, except for the 70% column which retained the same percentage as

the 100% column (~62%) (Figure 4-2). The same general trend of increasing mass of crab shell

leading to higher trace metal retention was also observed with respect to Ni. If these metals are

indeed retained as metal sulfides, one could indirectly infer a trend relating mass of crab shell to

sulfide production/sulfate reduction. It is important also to note that the 100% and 70% columns

achieved similar retention of Co and Ni, and thus potentially similar sulfate reduction, despite the

difference in mass of crab shell. As crab shell is more expensive than SMC, the implications of

using a lower CS:SMC ratio for cost savings are important.

Although influent concentrations of dissolved Ni were very low (0.85 ±0.31 mg/L),

complete Ni removal was not achieved at any time throughout the experiment in any of the

treatment columns (Figure 3-6). Regardless of the mechanism, incomplete removal indicates a

limiting condition within the treatment system. Constraints to Ni removal could be related to

limited adsorption sites due to a higher affinity for other metals at those sites. Indeed, Jeon et al.

58

(2003) found the sorption affinity of divalent metal ions onto hematite to be in the order of Fe

Zn > Co Ni. However, limited sulfide production could also constrain Ni removal if other

metals outcompete Ni for formation of metal sulfides. Most likely, a combination of these two

limitations reduced the efficiency of Ni removal in these tests.

Zinc was removed completely in each treatment column until breakthrough, which

occurred over a range of pH (3-5), depending on the column. Modeling with Visual MINTEQ

predicted only one Zn precipitate for the given conditions, sphalerite (ZnS), which would only

form in the presence of sulfide. Literature indicates, however, that some portion of the dissolved

Zn would also adsorb onto Fe and Al (oxy)hydroxides to varying extents within the continuous-

flow columns. Similar to the other trace metals, it was expected that any portion of adsorbed Zn

would be released around pH 5 when Al (oxy)hydroxide species dissolved and as the sorption

edge for Zn-Fe (oxy)hydroxides was approached. However this relationship was not noticed for

any of the crab shell-containing columns.

Solubility of ZnS can be dependent on several factors, including the concentration of

dissolved sulfide within solution. Modeling in Visual MINTEQ allowed for the investigation of

the solubility of ZnS over a range of HS- concentrations (Appendix G). The results indicate that

at high sulfide concentrations, sphalerite would form as low as pH 2. However at sulfide

concentrations < 2 mg/L, the pH at which sphalerite became soluble rose to over 3.

Although dissolved Zn breakthrough occurred near pH 3.5 for all of the crab shell-

containing columns, no Zn was retained within the 50% CS + 50% SMC column, and the 80%

CS + 20% SMC column showed only minimal retention of Zn after breakthrough (~2%). Both of

these columns experienced breakthrough sooner than the other crab shell columns and the

breakthrough curves eventually returned to a normalized concentration of zero (indicating a

return to influent Zn concentrations, Figure 3-5). The experiment was concluded before the other

59

crab shell columns saw effluent concentrations return to influent concentrations. Thus, it is

postulated that if the experiment were continued, the remaining Zn within the other crab shell

treatment columns would have been released.

Breakthrough of Zn in the traditional 90% SMC + 10% LS column did not follow the

same trend as the crab-shell containing columns, with breakthrough noted around pH 5. This

occurred simultaneously with Al breakthrough and also the adsorption edge for Fe

(oxy)hydroxides, so adsorption is postulated as the primary mechanism involved in Zn removal

within this column.

4.3 Carbon Species

The question has been raised as to whether the provision of organic carbon or alkalinity

generation exhausts first in MIW remediation systems utilizing crab shell substrates. The data

from this experiment show that alkalinity generation is depleted well before organic carbon

becomes limiting (Figure 4-3).

60

Figure 4-3. Substrate exhaustion with respect to DOC and alkalinity generation. + symbol above

column indicates the value presented is the PV when the experiment ended, thus the

potential exists for additional DOC generation until complete substrate exhaustion.

DOC is primarily composed of volatile fatty acids produced during the biological

breakdown of the organic substrate within the treatment columns (Newcombe and Brennan, 2010,

Robinson-Lora and Brennan, 2009b). In columns containing crab shell, easily fermentable

proteins within the crab shell matrix are utilized first and then biodegradation of the chitin

polymer occurs. Based on Figure 4-3 alone, one might assume that the substrate is depleted of

organic C for the 50% CS + 50% SMC and traditional 90% SMC + 10% LS treatment columns

which reached DOC exhaustion; however, an organic C mass balance indicates that a large

portion of the original C remains (Figure 4-4, Appendix H).

0

50

100

150

200

250

PV

to

Su

bst

rate

Exh

aust

ion

DOC

Alkalinity

++

+

+

+

61

Figure 4-4. Total C remaining in each treatment column at completion of continuous-flow

column experiment.

Exhaustion of DOC before depletion of the C source indicates a constraint on

fermentation, cellulose degradation, and other biotic processes responsible for the breakdown of

the substrate, likely caused by the reduced pH as discussed above. Columns containing more

crab shell lost more C over the course of the continuous-flow experiment, most likely because pH

was sustained above 5 for a longer period. However, none were depleted below 50% of the

starting mass, further indicating additional treatment capacity with respect to provision of C

source for microbial activity. This indicates that the longevity of systems utilizing crab shell can

0%

20%

40%

60%

80%

100%

% o

f to

tal C

rem

ain

ing i

n s

yst

em

62

be increased by a supplemental alkalinity source such as limestone after depletion of the shell-

associated carbonates.

4.4 Cations

Based on the data collected during the continuous-flow column study, it appears as though

cations such as Ca, K, Mg, and Na could be of concern for fish in local waterways due to the

extremely high concentrations reached during start-up (Appendix E). However, since the

formation of soluble metal–ligand complexes can significantly impact the effect of aqueous phase

species on fish, Viadero et al. (2004), suggested it was more realistic to compare the average

active free ion concentration against fish toxicity standards. Visual MINTEQ geochemical

modeling software was used to predict the speciation of Ca, K, Mg, and Na under three different

scenarios for the 100% CS column, the 70% CS + 30% SMC column, and the traditional 90%

SMC + 10% LS column (Appendix G). Results for all conditions indicated almost the entire

amount of K and Na were in the form of free ions (>98%). Free ion concentrations of Ca and Mg

were dependent on the estimated sulfate reduction occurring. Higher sulfate reduction led to

greater speciation in the free ion form generally; the % of Ca as a free ion species ranged between

73-87% (100% CS column), 82-92% (70% Cs + 30% SMC column), and 77-90% (traditional

substrate column) for the sulfate reduction ranges evaluated. Results for Mg free ion species

were 1-4% greater than Ca for all cases. Average concentrations of these cations stabilized after

the initial 10 PV were flushed through the system. Thus, the stabilized concentrations were of

particular interest to determine potential long-term fish toxicity. Results for Ca and Mg

speciation are provided in Figure 4-5.

63

Figure 4-5. Speciation of relevant cations in the 100% crab shell column after 10 PV based on

geochemical modeling with Visual MINTEQ and an assumed 250 mg/L sulfate

reduction. Results for the 70% CS + 30% SMC column were identical for this time

point and those for the traditional 90% SMC + 10% LS column varied by no more

than 1%.

From the results in Figure 4-5, average free ion concentrations and average active free

ion concentrations of each cation were calculated (Table 4-3). These results indicate that after

approximately 10 PV, free ion Ca concentrations are low enough to be considered safe for fish,

but Mg concentrations are still a concern for the system analyzed in this study. However, it

should be noted that the sustained high levels of Mg were a result of high influent concentrations,

and not due to the treatment system itself.

Similar calculations were conducted to determine if the maximum concentrations, noted

immediately following the incubation period, were of concern when measured as the active free

ion concentrations. These values still exceeded recommended limits. The consequences of acute

exposure for fish to Mg concentrations above recommended levels should be considered for

specific species present in local waterways prior to evaluation of use of crab shell mixtures in

treatment systems at those locations.

Ca+2

73%

CaSO4 (aq)

26%

Ca

Mg+2

78%

MgSO4 (aq)

22%

Mg

64

Table 4-3. Tolerance limits and analytic, free ion, and active free ion average concentrations (after 10 PV) for cations of interest from


mg/L after 10 PV

Average analytic concentration Average free ion

concentrationa

Average active free ion

concentrationb

Treatment Column Ca K Mg Na P Ca K Mg Na P Ca K Mg Na P

Influent 109 4.65 102 5.78 0.27 80 4.4 80 5.7 0.26 26 3.4 26 4.4 0.02

Sand Control 109 7.14 100 5.84 0.24 80 6.8 78 5.7 0.24 26 5.2 26 4.4 0.02

100% CS 187 7.46 104 6.77 7.69 137 7.1 81 6.6 7.5 45 5.5 27 5.1 0.60

90% CS + 10% SMC 183 9.37 103 6.57 7.47 134 8.9 80 6.4 7.3 44 6.9 27 5.0 0.59

80% CS + 20% SMC 170 7.37 103 6.49 5.82 124 7.0 80 6.4 5.7 41 5.4 27 4.9 0.46

70% CS + 30% SMC 194 7.90 105 6.65 7.18 142 7.5 82 6.5 7.0 47 5.8 27 5.0 0.56

60% CS + 40% SMC 175 9.33 105 6.50 4.71 128 8.9 82 6.4 4.6 42 6.8 27 4.9 0.37

50% CS + 50% SMC 173 8.51 103 6.46 4.15 126 8.1 80 6.3 4.1 42 6.2 27 4.9 0.33

Traditional

90% SMC + 10% LS 153 9.03 104 6.00 0.33

112 8.6 81 5.9 0.32

37 6.6 27 4.5 0.03

Recommended tolerance

limit for fishc (mg/L)

160 5 15 75 n/a

160 5 15 75 n/a 160 5 15 75 n/a

a Determined from Figure 4-5.

b Activity coefficient for mono-, di-, & trivalent ions were assumed to be 0.77, 0.33, & 0.08 respectively, based on an ionic strength of 0.1

c Meade (1989)

65

4.5 Longevity of Treatment

As noted in Section 2.2, the majority of each column was filled with an inert packing material

(Table 2-3) to allow for the required HRT while still exhausting within a reasonable laboratory

time-scale. It has been proposed that in field applications, crab shell should be mixed in a ratio of

1:1 (by mass) with an inert proppant to maximize longevity while ensuring adequate permeability

(Starr and Lebow, 2005). In order to estimate the longevity of crab shell treatment systems with

respect to the adjusted ratio (approximately 1:12) of crab shell to proppant used in the laboratory

column study, a multiplying factor was calculated.

The total volume of the experimental continuous-flow columns were 694 mL. However

the endcaps were slightly convex in design, allowing an additional volume to be held within the

column. Thus, the total column volume was assumed to be 700 mL. In order to determine the

mass of crab shell and sand required to fill the column completely using a 1:1 mass ratio,

iterations were completed in an Excel spreadsheet using the bulk density of crab shell (0.45

g/mL) and sand (1.55 g/mL) determined previously in the laboratory. Results in Table 4-4

indicate an estimated crab shell mass of 246 g to fill the column on a 1:1 mass ratio with sand.

Compared to the mass of substrate used in the 100% crab shell column (40 g), this indicates just

over a 6 fold increase in the mass of crab shell within the system. The same calculation was

performed for all of the treatment columns to determine the scale-up factor if the columns were

designed using the 1:1 crab shell to proppant ratio (Appendix I). The traditional substrate ratio

was calculated in the same way, only excluding the proppant material altogether, as a proppant is

not typically used with this substrate.

66

Table 4-4. Iterative calculations used to determine the theoretical masses and volumes of crab

shell and sand needed if a 1:1 packing ratio (by mass) were used in the continuous-

flow column study. Bolded row indicates the mass required to fill a ~700 mL

column, as used in this study.

Crab Shell (CS) Sand CS + Sand

Volume (mL) Mass (g) Volume (mL) Mass (g) Volume (mL)

284 630 284 183 813

268 595 268 173 768

252 560 252 163 723

249 553 249 161 714

246 546 246 159 705

243 539 243 156 695

189 420 189 122 542

173 385 173 112 497

Bulk density of CS = 0.45 g/mL; Bulk density of sand = 1.55 g/mL

The prior discussion (Section 4.1-4.2) has shown that efficient treatment is sustained

within the system as long as pH is maintained above 5.0. Using the information in Table 4-4, the

values from Table 3-3 were recalculated to estimate the theoretical system longevity based on the

use of a 1:1 (by mass) crab shell to proppant ratio ( Table 4-5).

Table 4-5. Experimental and theoretical treatment longevity of crab shell substrate mixtures for

treating high-strength MIW. Experimental longevity was determined in the column

study using a 1:12 (by mass) substrate to proppant ratio (40 g total substrate).

Theoretical longevity was estimated by extrapolating the results to a 1:1 (by mass)

crab shell to sand proppant ratio that would be used in the field.

Treatment Column

Experimental Scale-

up

Factorb

Theoretical

Substrate

(g)

Longevitya

PV (days)

Substrate

(g)

Longevitya

PV (days)

100% CS 40 141 (146) 6.2 246 874 (905)

90% CS + 10% SMC 40 131 (117) 6.1 242 799 (714)

80% CS + 20% SMC 40 95 (104) 6.0 238 570 (624)

70% CS + 30% SMC 40 137 (124) 5.9 235 808 (732)

60% CS + 40% SMC 40 108 (90) 5.7 226 616 (513)

50% CS + 50% SMC 40 114 (96) 5.3 213 604 (509)

Traditional 90% SMC + 10% LS 40 79 (70) 5.7 229 450 (399) a Longevity of treatment system defined as sustainment of pH above 5.0

b Factor to account for scale-up of the experimental 1:12 substrate to proppant ratio to the

proposed 1:1 crab shell to proppant ratio suggested (no proppant suggested for traditional

substrate). All ratios are by mass.

67

In addition, Table 4-1 was also re-evaluated with respect to total metals removed and

total volume of MIW treated based on the scale-up factor. The estimated treatment capacity of

each continuous-flow column can be found in Table 4-6.

Table 4-6. Theoretical total metals removal, volume of MIW treated in each column, and

substrate loading factor. Values were calculated based experimental vales from

Table 4-1 and the scale-up factor to account for a 1:1 (by mass) crab shell to sand

proppant ratio that would be used in the field.

Treatment Column

Theoretical Substrate Loading

Factor

Substrate

(g)

Metals

Removed

(mg)

MIW

Treated

(L)

(g substrate / L MIW)

100% CS 246 14,000 240 1.0

90% CS + 10% SMC 242 13,000 200 1.2

80% CS + 20% SMC 238 11,000 170 1.4

70% CS + 30% SMC 235 13,000 200 1.2

60% CS + 40% SMC 226 9,000 130 1.7

50% CS + 50% SMC 213 8,600 140 1.5

Traditional 90% SMC + 10% LS 229 5,600 100 2.3

68

5. Field Pilot System

In the spring of 2010, the Foundation for Pennsylvania Watersheds (FPW) awarded a $15,000

grant to the Clearfield Creek Watershed Association (CCWA) to conduct a pilot-scale test of a

proposed crab shell treatment system at the Klondike-1 site. The objective of the pilot study was

to confirm Penn State laboratory data indicating that crab shell and SMC mixtures would be

effective at treating this discharge, and to determine the effects of scale-up and environmental

field conditions on efficiency of the treatment system. The grant application was written by Dr.

Rachel Brennan in collaboration with Dr. Art Rose with the intent that researchers within Dr.

Brennan’s group would design, install, and monitor the pilot system. After the funds were

secured, the design of the system began in June 2010.

5.1 System Concept

Based on data collected in the continuous-flow column study reported here, an optimal ratio of

crab shell to SMC was determined to be 2.33:1 (70% CS to 30% SMC by mass). Funding from

FPW allowed for four replicate treatment reactors so that comparisons could be made between

different organic substrate mixtures and also different underdrain materials at the pilot-scale. The

reactors consisted of one of the following organic substrate layers: 1) 100% CS, 2) 70% CS +

30% SMC, 3) traditional 90% SMC + 10% LS, and either a limestone or sandstone (SS)

underdrain. Each treatment consisted of a 1000-gallon tank reactor fitted with an underdrain

piping network (to simulate a VFP) and two subsequent 350 gallon aerobic settling ponds

arranged in series. The pilot-VFPs were constructed using polyethylene septic tanks with two

manholes (20.125 inch diameter) at the top to allow access during packing and later substrate

69

sampling. The pre-existing inlet port of the tank was retrofitted with an orifice pipe to control the

influent flow rate (design discussion follows in Section 5.2). The pre-existing outlet of the tank

was retrofitted to attach the underdrain piping system to enable effluent flow. Tanks were then

filled with a 2’ rock layer (covering the underdrain piping network), a 3’ substrate layer and a 3‖

layer of pea gravel to hold down the substrate (Figure 5-1). The reactors were designed to have

standing water on the surface and remain open to the air in a manner comparable to a VFP.

Figure 5-1. Schematic of pilot-scale VFPs installed at Klondike-1 field site.

As mentioned previously, a full-scale treatment system is already constructed at the

Klondike-1 site which limited the footprint of the pilot system. The layout of the existing site

however, afforded a convenient location for the pilot system adjacent to the precipitation pond

(Figure 5-2). This location was ideal because it was slightly down-gradient, which allowed for

the siphoning and gravity flow of water to the pilot-scale system instead of the use of a pump.

70

Figure 5-2. Schematic of pre-existing full-scale treatment system at the Klondike-1 site with the

location of the pilot system indicated.

5.2 System Design

Water to feed the pilot system was drawn directly from the precipitation pond at a distance of

approximately 16‖ below the water surface through individual piping networks for each reactor.

Hydraulic calculations were conducted to design an appropriate orifice opening to maintain flow

at 0.2 gpm. Major and minor head losses associated with the designed piping network were taken

into account, and an orifice opening with a 3 mm diameter was selected.

As funding allowed for monitoring of the pilot-scale system for a minimum of one year,

substrate design calculations were conducted to allow for system exhaustion in no less than one

year. A 17.5 hr HRT, flow rate of 0.2 gpm, a 1:1 crab shell to sand ratio (by mass), and the

assumption that 1 g of crab shell treats 1L of MIW from this site, were used to calculate the time

161.5 ft

VERTICAL FLOW

POND

OXIDATION

DITCHES

PRECIP.

POND

AEROBIC

SETTLING

POND

WETLAND

DISCHARGE TO

LITTLE LAUREL RUN

Direction of water flow

Pilot-scale system

Source of MIW

71

to exhaustion. The organic substrate layer of the reactors consisted of various mixtures (Table

5-1) of the following: SC-20 crab shell; SMC (available onsite from the construction of the

previously installed treatment system); white silica sand (Seymore Brothers, Inc., Altoona, PA),

limestone chips (91% CaCO3, available onsite from the construction of the previously installed

treatment system), and pea gravel (Somogyi’s Route 22 Supply, Ebensburg, PA).

Table 5-1. Actual and designed mixture of the organic substrate layer in each pilot-scale VFP

installed to treat MIW at the Klondike-1 site.

The underdrains were constructed approximately 2’ deep with 3-inch rocks: limestone

rock (AASHTO#1, 99.3% CaCO3, New Enterprise Stone and Lime Company, Tyrone, PA) or

sandstone (SS) rock (#4 Sandstone, 0% CaCO3, Kinkead Aggregates, Homer City, PA) (Figure

5-3).

Reactor Name

Reactor Organic Substrate Layer Components

(kg)

Actual [Design]

Crab Shell Sand SMC LS chips

100% CS + LS Underdrain 685 [570] 680 [570] 0 [0] 0 [0]

70% CS + 30% SMC + LS Underdrain 577 [500] 576 [500] 247 [214] 0 [0]

70% CS + 30% SMC + SS Underdrain 577 [500] 576 [500] 247 [214] 0 [0]

90% SMC + 10% LS + LS Underdrain 0 [0] 0 [0] 549 [476] 61 [53]

72

Figure 5-3. Limestone (A) and sandstone (B) rocks used in underdrains for field pilot-scale VFPs

treating MIW at the Klondike-1 site.

The tanks were partially buried to assist with insulation and better imitate an actual VFP.

Full-scale systems have the benefit of faster flow rates and large volumes of water, thus it was

expected that the system would freeze over the winter months, but it was anticipated that burying

the reactors would help mitigate this effect to some degree.

Passive effluent aeration was designed for the system in the form of a miniature cascade

(corrugated pipe) leading from the tank outlet down a vertical distance of two feet to the aeration

ponds. Initial designs included one large settling pond; however, cost restrictions associated with

ponds of the appropriate size required the use of two smaller ponds arranged in series. Water

exiting the second pond of each treatment set-up was connected into a main effluent drainage

pipe which diverted flow to the wetland of the full-scale treatment system already on site.

A

B

73

5.3 Construction, Incubation, and Field Sampling

Construction on the pilot-scale system began at the end of July, 2010, and lasted approximately

two weeks (Appendix J contains a photo-documentation of the installation). Volunteers from the

CCWA and members of the Brennan Research Group completed all aspects of the installation,

with the exception of excavation which was accomplished by a hired contractor (John Slovikosky

Excavating).

As per design, the bottom two feet of the reactors were filled with either LS or SS rocks.

Because no physical barrier is included between the underdrain rock layer and the upper organic

substrate layer in VFPs, some substrate naturally settles into the rock layer during construction.

This was accounted for in the design of the pilot reactors, however the amount of substrate to fall

through during loading was more than assumed. Due to the availability of additional substrate

on-site, more was added to attain the desired 3 ft substrate layer. The design versus actual

substrate mass for each reactor can be found in Table 5-1.

In order to facilitate future analysis and investigation of the microbial community present

within the systems, special tea-bag style sample pouches were created and buried within the

organic substrate layer of each reactor. 3-inch X 3-inch bags were constructed from extruded

nylon mesh and sewn together with 10lb. Triline fishing line (Figure 5-4). Each bag contained

10 g of organic substrate, taken from the mixture within the reactor. Bags were attached to a

piece of polypropylene twine and were positioned approximately 8-10 inches below the surface

of the organic substrate layer. A total of 32 bags were placed into each reactor, with 16 bags

located directly under each of the two manholes; the tea-bags near the influent end of the tank

were numbered 1-16 and those near the effluent end were numbered 17-32. The twine was

intended to facilitate removal of the bags and was secured to the outside of the manhole. A

74

random number generator was then used to establish a schedule of which bag would be pulled in

conjunction with each sampling event.

Figure 5-4. Photo of microbial tea-bag style sample pouches filled with organic substrate and

buried within each pilot-scale reactor at the Klondike-1 site.

Once construction was complete, the reactors were filled with water from the Klondike-1

discharge and were flushed with approximately 4,300 gallons of water to assist in removing fines.

An attempt was made to conduct tracer tests using sodium chloride; however the background salt

coming off the reactors confounded the tests. After flushing, the reactors were incubated for one

week to allow the development of the microbial consortium.

75

Figure 5-5. Pilot-scale VFPs and subsequent aerobic settling ponds installed to treat MIW at the

Klondike-1 field site.

On August 27, 2010, regular operation of the reactors was initiated for all except the

reactor with the sandstone underdrain. This reactor displayed reduced flow rates during the

flushing phase and became completely clogged during the incubation period for unknown

reasons. The reactor was eventually unclogged one week later by pumping water into the outlet,

and no further problems were encountered. Thus, this reactor is on a 7-day lag behind the other

reactors.

Reactors were sampled every week for the first month, biweekly for another month, and

then monthly thereafter. Monitoring will continue for a minimum of one year. Field probes ae

used onsite to measure pH, ORP, conductivity, total dissolved solids (TDS), and temperature of

the reactor effluent and of the final aerobic settling pond. Samples are collected from the reactor

effluent, transported on ice to the laboratory and measured within 4 hours for pH, acidity,

alkalinity, and ammonium and a sample is preserved for later analysis of dissolved metals,

anions, DOC, and Fe speciation. Samples are concurrently collected from the second aerobic

76

settling pond, preserved upon return to the lab, and measured for dissolved metals analysis and Fe

speciation according to the methods described in section 2.4.

5.4 Results

Within the pilot system pH levels reached within the first 90 days of operation were slightly

higher than those achieved in the continuous-flow columns. All four reactors have maintained

pH above 6.5 for the 90 days of monitoring conducted thus far (Figure 5-6). Alkalinity

generation within the traditional treatment reactor for the pilot system reached a maximum of 260

mg/L as CaCO3 after incubation, identical to that achieved within this substrate during the

continuous-flow experiment (Figure 5-7). In general, alkalinity and acidity from the field

reactors are following the same trend as the continuous-flow columns, and all reactors are

maintaining a consistent net alkaline effluent.

Complete removal of dissolved iron and aluminum has been maintained thus far within

all pilot-scale field reactors, and partial manganese removal has also been observed (not shown).

77

Figure 5-6. pH values of MIW influent and pilot-scale reactor effluent from initial 90 days of

monitoring.

Figure 5-7. Alkalinity generated from pilot-scale reactors during initial 90 days of the field test.

2

3

4

5

6

7

8

9

0 20 40 60 80

pH

Time (days)

Influent100% CS + LS Underdrain70%CS/30%SMC + LS Underdrain70%CS/30%SMC + SS Underdrain90%SMC/10%LS + LS Underdrain

0

1000

2000

3000

4000

5000

0 20 40 60 80

Alk

ali

nit

y (

mg

/L a

s C

aC

O3)

Time (d)

Influent100% CS + LS Underdrain70%CS/30%SMC + LS Underdrain70%CS/30%SMC + SS Underdrain90%SMC/10%LS + LS Underdrain

78

6. Conclusions

Based on results obtained in the continuous-flow column experiment, several conclusions can be

made regarding the use of mixed crab shell and SMC substrates for the treatment of MIW,

specifically with respect to longevity and design parameters as well as potential areas of concern.

6.1 Treatment Longevity for Engineering Designs

Mixed crab shell and SMC substrates are effective at treating high-strength MIW from

the Klondike-1 site, maintaining the pH > 5.0 for nearly twice as long as the traditional

substrate. Average influent values of Al (2.81 mg/L), Fe (62 mg/L), and Mn (38.2

mg/L), as well as trace metals, were removed for up to 160 PV in some cases.

The data obtained from the continuous-flow column experiment indicates that a substrate

mixture of 70% crab shell and 30% SMC is the most effective at treating high-strength

MIW, with only 40 g of substrate removing approximately 2,200 mg of metals from 34 L

of MIW.

Treatment columns contained almost a 1:12 ratio of substrate to inert packing material,

therefore estimations of treatment capacity/longevity were made which indicated an

average 6-fold increase in the treatment capacity. The 70% crab shell and 30% SMC

column was estimated to be capable of maintaining pH > 5.0 for over 800 PV, and

treating over 200 L of MIW from the Klondike-1 site.

Substrate loading factors (g substrate / L MIW) were calculated for each substrate ratio

and ranged from 1.0 for 100% crab shell to 2.3 for the traditional 90% SMC and 10%

79

limestone substrate. The 70% crab shell and 30% SMC loading factor was determined to

be of 1.2 g substrate / L MIW.

Alkalinity generation is the limiting factor in longevity of mixed crab shell treatment

systems. The ability of the substrate to maintain pH above 5 within the system drives a

majority of the proposed metals removal mechanisms. Despite the fact that the organic

substrates are not exhausted with respect to organic C source, several key physiochemical

and biological processes cease to function below pH 5, as discussed previously. Thus,

the ability of the system to treat is a direct function of pH. Although crab shell has

received favorable attention due to the integral source of CaCO3, these carbonates are not

adequate to sustain alkalinity generation long enough to achieve complete biodegradation

of the organic substrate. Addition of an external alkalinity source should be considered

as either a periodic supplement to crab shell substrates or a portion of the overall system

design (such as a limestone underdrain).

Passive aeration after anoxic treatment increases the removal of Fe by up to 50% in crab

shell-containing columns. High percentages of crab shell in the substrate mixture

displayed an inferior capacity to oxidize Fe upon passive aeration. It is proposed that

conditions are still extremely reducing as water exits the simulated anaerobic wetland.

This did not occur with fractions 80% crab shell. Thus, substrate mixtures containing

80% CS will experience both reduced costs, and improved subsequent aerobic metals

removal.

The pilot system at the Klondike-1 site will need to be monitored until exhaustion of

alkalinity generation (until pH falls below 5.0) in order to provide valuable scale-up and

design information to enable the successful implementation of crab shell systems for

future remediation of MIW.

80

6.2 Potential Concerns

Substrates containing more than 50% crab shell (by mass) produce higher levels of

NH4+, Ca, K, Mg, Na, and P than the traditional 90% SMC + 10% LS substrate,

especially at early times. The effects of these high concentrations should be considered

for the first several pore volumes of water through a treatment system, specifically with

respect to fish toxicity and the potential for eutrophication.

As pH falls below 5.0, metals breakthrough begins to occur (except for Mn which

experiences breakthrough much sooner). Systems should be monitored for downward

trends in pH as an indication that action needs to be taken. If no additional buffering

capacity is added (via fresh substrate, a limestone amendment, or some other addition)

metals can be lost from the system. Many of the precipitates are in the form of

(oxy)hydroxides and can be resolubilized when pH falls below 5.0. As a significant

mass of metals could be held within the treatment system, a release of them all at once

could result in extremely high concentrations (considerably higher than the influent MIW

water). This would be true in any treatment system utilizing a neutralizing material such

as CaCO3, however.

Due to the compromised performance of the 80% and 60% crab shell columns, it is not

absolutely certain that 70% crab shell + 30% SMC is the optimal ratio. There is still the

possibility that the 60% crab shell + 40% SMC column could have performed similarly to

the 70% crab shell column if the oxygen intrusion had not occurred.

81

7. Future Work

Further research could be undertaken to further elucidate topics related to the use of treatment

systems utilizing mixed crab shell and SMC substrates and ensure appropriate design. The

following list provides subject matter suggested for additional investigation.

Although one of the goals of this test was to monitor the treatment efficiency of crab shell

substrates in remediating ―high-strength‖ MIW, the concentration of Al and several trace

metals were low. Future research should focus on the ability of crab shell substrate

mixtures to treat concentrations of Al higher than 16 mg/L, as well as higher

concentrations of Ag, As, Co, Cu, Ni, Pb, and U.

Although some work has been done with respect to Mn removal in crab shell treatment

systems, additional determination of the contribution to the removal of other metals by

surface interactions on the crab shell material (adsorption, co-precipitation, and surface

precipitation) could help to illuminate the removal mechanisms of specific metals and

lead to a greater understanding of potential benefits and limitations related to the use of

this substrate.

Samples for Fe speciation were collected from the continuous-flow columns at two

locations during each sampling event: from the column effluent and from the simulated

aerobic settling pond. Processing of those samples would lead to a better understanding

of the actual Fe species present and allow for further analysis into the likely removal

pathways associated with Fe.

Tracer tests were not run until the end of the current experiment. However, it is

recommended that a series of tracer tests be run to determine the changes in porosity as

82

substrate is degraded, precipitates are formed, and settling of the packing materials

occurs. Loss of permeability has been noted to be a cause of failure in anaerobic passive

treatment systems, so the ability of crab shell substrate mixtures to maintain permeability

over time would be a distinct advantage. Also, the study should determine the need for

inclusion of inert packing materials as a proppant within crab shell substrate mixtures. If

mixtures containing SMC can retain permeability without the use of an additional

proppant material, considerable cost savings could be realized for full-scale systems.

Additional sources of alkalinity should be evaluated in conjunction with crab shell

substrate mixtures to increase the longevity of the system. The current field pilot-scale

system has incorporated limestone underdrains to increase alkalinity generation within

the systems. This should sustain biodegradation of the organic substrate materials for an

extended period beyond what would be accomplished with crab shell alone.

At the completion of the continuous-flow column study, samples of the packing materials

from within each column were preserved for microbial analysis. Cloning work has been

started to evaluate the community established within each substrate mixture in order to

determine the differences in species, if any. Evaluation of the microbial community will

also be conducted on samples from the field-scale reactors over the course of the pilot

test to monitor changes within the community over time (start up, steady state conditions,

decline) and over varying environmental conditions (seasonal temperature and flow

fluctuations).

Continued assessment of the pilot-scale study will allow a more realistic evaluation of

actual treatment capacity achievable in systems utilizing mixtures of crab shell substrates,

as well as providing valuable information on the challenges, if any, related to the scale-up

of mixed crab shell and SMC treatment systems.

83

In light of the oxygen intrusion into the 60% crab shell + 40% SMC column and given

the fact that this substrate ratio would be more economically advantageous, further

evaluation should be conducted to determine if the performance is significantly different

from the 70% crab shell + 30% SMC substrate.

Future collaboration with the PA Cooperative Fish and Wildlife Research Unit in Penn

State’s School of Forest Resources could help to estimate the risk to aquatic creatures of

increased cations released during start-up of full-scale treatment systems containing crab

shell mixtures.

84

References

APHA, AWWA, and WEF. ―Standard Methods for the Examination of Water and Wastewater:

20th Edition.‖ Eds. Clesceri, L.S., Greenberg, A.E., and Eaton, A.D. Baltimore: APHA,

1998.

Arai, Y. (2008) ―Spectroscopic evidence for Ni(II) surface speciation at the iron oxyhydroxides-

water interface.‖ Environmental Science and Technology, 1151-1156.

Baruah, B. P., and Khare, P. (2010) ―Mobility of trace and potentially harmful elements in the

environment from high sulfur Indian cola mines.‖ Applied Geochemistry, 25: 1621-1631.

Charriau, A., Lesven, L., Gao, Y., Leermakers, M., Baeyens, W., Ouddane, B., and Billon, G.

(2011) ―Trace metal behavior in riverine sediments: role of organic matter and sulfides.‖

Applied Geochemisty, 26: 80-90.

Christensen, B., Morten, L. and Lien, T. (1996) ―Treatment of acid mine water by sulfate-

reducing bacteria; results from a bench scale experiment.‖ Water Research, 30(7): 1617-

1624.

Cravotta, C. A. (2008) ―Dissolved metals and associated constituients in abandoned coal-mine

discharges, Pennsylvania, USA. Part 2: Geochemical controls on constiituient

concentrations.‖ Applied Geochemistry, 23: 203-226.

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.

Dann, A.L., Cooper, R.S., Bowman, J.P. (2009) ―Investigation and optimization of a passively

operated compost-based system for remediation of acidic, highly iron- and sulfate-rich

industrial waste water.‖ Water Research, 43: 2302-2316.

Daubert, L. N. and Brennan, R. A (2007) ―Passive remediation of acid mine drainage using crab

shell chitin.‖ Environmental Engineering Science, 24(10): 1475-1479.

Dinelli, E., Lucchini, F., Fabbri, M., and Cortecci, G. (1999) ―Metal distribution and

environmental problems related to sulfide oxidation in the Libiola copper mine area

(Ligurian Apennines, Italy).‖ Journal of Geochemical Exploration, 74: 141-152.

Doshi, S. M. (2006) ―Bioremediation of acid mine drainage using sulfate-reducing bacteria.‖

Prepared for U.S. Environmental Protection Agency, Office of Solid Waste and

Emergency Response, Office of Superfund Remediation and Technology Innovation,

Washington, DC.

85

Edraki, M., Golding, S. D., Baublys, K. A., and Lawrence, M. G. (2005) ―Hydrochemistry,

mineralology, and sulfur isotope geochemistry of acid mine drainage at the Mt. Morgan

mine environment, Queensland, Australia.‖ Applied Geochemistry, 20: 789-805.

Environmental Protection Agency (EPA) (1997) National Hardrock Mining Framework: United

States Environmental Protection Agency Office of Water (4023). Washington, DC, USA.

Faulwetter, J. L., Gagnon, V., Sundberg, C., Chazarenc, F., Burr, M. D., Brisson, J., Camper, A.

K., Stein, O. R. (2009) ―Microbial processes influencing performance of treatment

wetlands: A review.‖ Ecological Engineering, 35: 987-1004.

Gerente, C., Lee, V. K. C., Le Cloirec, P., and McKay, G. (2007) ―Application of chitosan for the

removal of metals from wastewaters by adsorption – mechanisms and model reviews.‖

Critical Reviews in Environmental Science and Technology, 37: 41-127.

Hallberg, K. B., and Johnson, D. B. (2005) ―Microbiology of a wetland ecosystem constructed to

remediate mine drainage from a heavy metal mine.‖ Science of the Total Environment,

338: 53–66.

Hayes, M., Carney, B., Slater, J., and Bruck, W. (2008) ―Mining marine shellfish wastes for

bioactive molecules: Chitin and chitosan – Part B: Applications.‖ Biotechnology Journal,

3: 878-889.

Hiibel, A. R., Pereyra, L. P., Inman, L. Y., Tischer, A., Reisman, D. J., Reardon, K. F., Pruden,

A. (2008) ―Microbial communit analysis of two field-scale sulfate-reducing bioreactors

treating mine drainage.” Environmental Microbiology, 10: 2087-2097.

Jandová, J., Lisá, K., Vu, H., and Vranka, F. (2005) ―Separation of copper and cobalt-nickel

sulphide concentrates during processing of manganese deep ocean nodules.‖

Hydrometallurgy, 77: 75-79.

Jong, T., and Parry, D. L. (2004) ―Adsorption of Pb(II), Cu(II), Cd(II), Zn(II), Ni(II), Fe(II), and

As(V) on bacterially produced metal sulfides.‖ Journal of Colloid and Interface Science,

275: 61-71.

Jong, T., and Parry, D. L. (2006) ―Microbial sulfate reduction under sequentially acidic

conditions in an upflow anaerobic packed bed bioreactor.‖ Water Research, 40(13):

2561-2571.

Kairies, C. L., Capo, R. C., and Watzlaf, G. R. (2002) ―Chemical and physical properties of iron

hydroxide precipitates associated with passively treated coal mine drainage in the

Bituminous Region of Pennsylvania and Maryland.‖ Applied Geochemistry, 20: 1445-

1460.

Kapat, A., Rakshit, S. K., and Panda, T. (1996) ―Parameters optimization of chitin hydrolysis by

Trichoderma harzianum chitinase under assay conditions.‖ Bioprocess Engineering, 14:

275-279.

86

Larese-Casanova, P., and Scherer, M. M. (2007) ―Fe(II) sorption onto hematite: new insights

based on spectroscopic measurements.‖ Environmental Science and Technology 41: 471-

477.

Lee, G., Bigham, J. M., and Faure, G. (2002) ―Removal of trace metals by coprecipitation with

Fe, Al and Mn from natural waters contaminated with acid mine drainage from

Ducktown Mining District, Tennessee.‖ Applied Geochemistry, 17: 569-581.

Lewis, A. E. (2010) ―Review of metal sulphide precipitation.‖ Hydrometallurgy, 104: 222-234.

Logan, M.V., Reardon, K.F., Figueroa, L.A., McLain, J.E.T., Ahmann, D.M. (2005) ―Microbial

community activities during establishment, performance, and decline of bench-scale

passive treatment systems for mine draingage.‖ Water Research, 39: 4537-4551.

Machemer, S. D. and Wildeman, T. R. (1992) ―Adsorption compared with sulfide precipitation

as metal removal processes from acid mine drainage in a constructed wetland.‖ Journal

of Contaminant Hydrology, 9: 115-131.

Maier, R. M., Pepper, I. L., and Gerba, C. P. Environmental Microbiology, 2nd

ed.; Academic

Press, Burlington, MA 2009; p 323-324.

Matshusa-Masithi, M.P., Ogola, J.S., Chimuka, L. (2009) ―Use of compost bacteria to degrade

cellulose from grass clippings in biological removal of sulphate from acid mine

drainage.‖ Water South Africa, 35: 111-116.

Matthies, R., Aplin, A. C., and Jarvis, A. P. (2010) ―Performance of a passive treatment system

for net-acidic coal mine drainage over five years of operation.‖ Science of the Total

Environment, 408: 4877-4885.

McMahon, J. A., and Snodgrass, W. J. (1996) ―Chemical equilibrium data bases and tools for

calculating zinc sulfide geochemistry in freshwater sediments and its toxicity.‖ Water

Quality Resources Journal of Canada, 31 (3): 577-591.

Meade, J. W. Aquaculture Management; Van Nostrand Reinhold, New York, NY 1989; p 9-10.

Metcalf and Eddy. "Wastewater Engineering Treatment and Reuse: 4th

Edition." Eds.

Tchbanoglous, G., Burton, F. L., and Stensel, H. D. Boston: McGraw Hill, 2003.

Micera. G., Gessa, C., Melis. P., Premoli, A., Dallocchio, R. and Deiana, S. (1986) ―Zinc (II)

Adsorption on Aluminum Hydroxide.‖ Colloids and Surfaces, 17(4): 389-394.

Miller, A., Figueroa, L., and Wildeman, T. (2011) ―Zinc and nickel removal in simulated

limestone treatment of mining influenced water.‖ Applied Geochemistry, 26: 125-132.

Morales, T. A., Dopson, M., Athar, R., and Herbert, R. B. J. (2005) ―Analysis of bacterial

diversity in acidic pond water and compost after treatment of artificial acid mine drainage

for metal removal.‖ Biotechnology and Bioengineering, 90: 543–551.

87

Muzzarelli, R.A.A. Chitin; Pergamon Press, New York, 1977.

Newcombe, C. E., and Brennan, R. A. (2010) "Improved passive treatment of acid mine drainage

in mushroom compost amended with crab-shell chitin." Journal of Environmental

Engineering, 136 (6): 616 - 626.

Nicomrat, D., Dick,W.A., Tuovinen, O.H. (2006) ―Assessment of the microbial community in a

constructed wetland that receives acid coal mine drainage.‖ Microbial Ecology, 51: 83-

89.

Peltier, E., Van Der Lelie, D., and Sparks, D. L. (2010) ―Formation and stability of Ni-Al

hydroxide phases in soils.‖ Environmental Science and Technology, 44: 302-308.

Pennsylvania Department of Environmental Protection (PA DEP) and U.S. Office of Surface

Mining Reclamation and Enforcement. (2008) ―Mine Drainage Treatability and Project

Selection Guidelines- June 10, 2008.‖ p. 3.

PA DEP (2009) ―Program Implementation Guidelines, Revised Draft – July 19, 2009.‖ Bureau

of Abandoned Mine Reclamation Acid Mine Drainage Set-Aside Program. p. 19.

PA DEP (2010) ―2010 Pennsylvania Integrated Water Quality Monitoring and Assessment

Report: Clean Water Act Section 305(b) Report and 303(d) List‖ p. 38.

Pereyra, L. P., Hiibel, S. R., Pruden, A., Reardon, K. F. (2008) ―Comparison of microbial

community composition and activity in sulfate-reducing batch systems remediating mine

drainage.‖ Biotechnology and Bioengineering, 101: 702-713.

Ramírez-Coutiño, L., Marin-Cervantes, M. C., Huerta, S., Revah, S., and Shirai, K. (2006)

―Enzymatic hydrolysis of chitin in the production of ogliosaccharides using Lecanicillium

fungicola chitinases.‖ Process Biochemistry, 41: 1106-1110.

Robinson-Lora, M. A. (2009) ―Evaluation of chitinous material as a multifunctional substrate for

the remediation of mine impacted water [dissertation].‖ University Park, PA:

Pennsylvania State University.

Robinson-Lora, M. A., and Brennan, R. A. (2009a) "The use of crab-shell chitin for biological

denitrification: batch and column tests." Bioresource Technology, 100 (2): 534 - 541.

Robinson-Lora, M. A., and Brennan, R.A. (2009b) "Efficient metal removal and neutralization of

acid mine drainage by crab-shell chitin under batch and continuous-flow conditions."

Bioresource Technology, 100 (21): 5063 - 5071.

Robinson-Lora, M. A., and Brennan, R.A. (2010a) "Chitin complex for the remediation of mine

impacted water: geochemistry of metal removal and comparison with other common

substrates." Applied Geochemistry, 25 (3): 336 - 344.

http://www.engr.psu.edu/ce/enve/brennan/publications/Newcombe%20&%20Brennan_JEE_2010.pdf

http://www.engr.psu.edu/ce/enve/brennan/publications/Newcombe%20&%20Brennan_JEE_2010.pdf

http://www.engr.psu.edu/ce/enve/brennan/publications/Robinson-Lora%20and%20Brennan_Denitrification%20with%20chitin_Bioresource%20Technol_2009.pdf

http://www.engr.psu.edu/ce/enve/brennan/publications/Robinson-Lora%20and%20Brennan_BioresourceTechnol_2009.pdf

http://www.engr.psu.edu/ce/enve/brennan/publications/Robinson-Lora%20and%20Brennan_BioresourceTechnol_2009.pdf

http://dx.doi.org/10.1016/j.apgeochem.2009.11.016

88

Robinson-Lora, M. A., and Brennan, R. A. (2010b) "Biosorption of manganese onto chitin and

associated proteins during the treatment of mine impacted water." Chemical Engineering

Journal, 162: 565 - 572.

Robinson-Lora, M. A., and Brennan, R. A. (2011) "Anaerobic precipitation of manganese and co-

existing metals in mine impacted water treated by crab shell-associated minerals."

Applied Geochemistry, 26: 853-862.

Roden, E. E. and Urrutia, M. M. (1999) ―Ferrous iron removal promotes microbial reduction of

crystalline iron(III) oxides.‖ Environmental Science and Technology, 33:1847-1853.

Romero, F. M., Prol-Ledesma, R. M., Canet, C., Alvarez, L. N., and Pérez-Vásquez, R. (2010)

―Acid drainage at the inactive Santa Lucia mine, Western Cuba: Natural attenuation of

arsenic, barium, and lead, and geochemical behavior of rare earth elements.‖ Applied

Geochemistry, 25: 716-727.

Rose, A. W. (2008) ―Klondike AMD Treatment Project Final Report.‖ Prepared for the Clearfield

Creek Watershed Association, Ashville, PA.

Rötting, T. S., Thomas, R. C., Ayora, C., and Carrera, J. (2008) ―Passive treatment of acid mine

drainage with high metal concentrations using dispersed alkaline substrate‖ Journal of

Environmental Quality, 37: 1741-1751.

Roy, I., Sardar, M., and Gupta, M. N. (2003) ―Hydrolysis of chitin by Pectinex™.‖ Enzyme and

Microbial Technology, 32: 582-588.

Schippers, A., Kock, D., Schwartz, M.,, Böttcher, M. E., Vogel, H., and Hagger, M. (2007)

―Geomicrobiological and geochemical investigation of a pyrrhotite-containing mine

waste tailings dam near Selebi-Phikwe in Botswana.‖ Journal of Geochemical

Exploration, 92: 151-158.

Silvester, E., Charlet, L., Tournassat, C., Gehin, A., Greneche, J. M., and Liger, E. (2005)

―Redox potential measurements and Mossbauer spectrometry of Fe(II) adsorbed onto

Fe(III) (oxyhydr)oxides.‖ Geochimica Et Cosmochimica Acta, 69: 4801-4815.

Spark, K. M., Johnson, B. B., and. Wells, J. D. (1995) ―Characterizing heavy-metal adsorption on

oxides and oxyhydroxides.‖ European Journal of Soil Science, 46:621-631.

Starr, B. and Lebow, P. (2005) ―Phase II SBIR final report: bioremediation of chlorinated

solvents in saturated, low permeability soils." NSF SBIR program, 2005.

Tutu, H., McCarthy, T. S., and Cukrowska, E. (2008) ―The chemical characteristics of acid mine

drainage with particular reference to sources, distribution, and remediation: The

Witwatersrand Basin, South Africa as a case study.‖ Applied Geochemistry, 23: 3666-

3684.

http://www.engr.psu.edu/ce/enve/brennan/publications/Robinson-Lora%20and%20Brennan_Biosorption%20of%20Mn_CEJ_2010.pdf



http://www.elsevier.com/wps/find/journaldescription.cws_home/768/description#description

89

Venot, C., Figueroa, L., Brennan, R. A., Wildeman, T. R., Risemen, D., Sieczkowski, M. (2008)

―Comparing chitin and organic substrates on the national tunnel waters in Blackhawk,

Colorado for manganese removal.‖ In: Barnhisel, R. I. (Ed.) Proceedings of the American

Society of Mining and Reclamation Richmand, VA, June 14-19, 2008. ASMR,

Lexington, KY, pp. 1232-1351.

Viadero, R. C., Tierney, A. E., Semmesn, K. J. (2004) ―Use of treated mine water for rainbow

trout (Oncorhynchus mykiss) culture: a production scale assessment.‖ Aquaculture

Engineering, 31, 319-336.

Viggi, C. C., Pagnanelli, F. Toro, L. (2009) ―Sulphate bioreduction for the treatment of polluted

wates: solid versus liquid organic substrates.‖ Journal of Chemical Technology and

Biotechnology, 84: 859-863.

Waybrant, K. R., Ptacek, C. J., and Blowes, D. W. (2002) ―Treatment of mine drainage using

permeable reactive barriers: column experiments.‖ Environmental Science and

Technology, 36: 1349-1356.

Willow, M. A., and Cohen, R. R. H. (2003) ―pH, dissolved oxygen, and adsorption effects on

metal removal in anaerobic bioreactors.‖ Journal of Environmental Quality, 32: 1212-

1221.

Ziemkiewicz, P. F., Skousen, J. G., Brant, D. L., Sterner, P. L. and Lovett, R. J. (1997) ―Acid

mine drainage treatment with armored limestone in open limestone channels.‖ Journal of

Environmental Quality, 26: 1017-1024.

Ziemkiewicz, P. F., Skousen, J. G., and Simmons, J. (2003) ―Long-term performance of passive

acid mine drainage treatment systems. ― Mine Water and the Environment, 22: 118-129.

90

Appendix A

Conservative Tracer Tests

The HRT, variance, and dispersion number were calculated from the tracer test data

using Equations A.1-A.4 (Metcalf and Eddy, 2003). The dispersion number is the ratio of

mass transport due to dispersion and advection. The calculated HRT (t) was used as the

estimated retention time (τ) to calculate the dispersion number. The HRT and measured flow

rate were then used to calculate pore volume.

Equation A-1

Equation A-2

Equation A-3

Equation A-4

Where:

91

Figure A-1. Conservative tracer test response curves for continuous-flow columns.

0

50

100

150

200

250

0 5 10 15 20 25 30 35

Ch

lori

de

(m

g/L

)

Time (hr)

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


92

Appendix B

In-line pH and ORP Probes

pH and ORP electrodes were utilized in flow-though cells positioned to receive water

directly from the column effluent port (just prior to the sample cell). Electrodes were maintained

in the flow-through cells continuously and during each sampling event, the bench top meter

(Accumet® Basic AB15, Fisher Scientific) was used in conjunction with the respective in-line

electrode to monitor pH and ORP.

In addition to short chain organic acids and alcohols, CO2 is a byproduct of fermentation.

In solution, CO2 is typically present as carbonic acid (H2CO3), which will lower the pH of the

solution. It was thought that in-line readings would provide a more accurate measure of system

pH, due to the possible partitioning of CO2 out of the water during sample handling. Thus, the in-

line pH electrodes were expected to provide pH readings slightly lower than the bench top pH

readings, and allow insight into the potential effects of air exposure to samples during transport

from the field site for the subsequent field study. As expected, the in-line electrodes consistently

measured lower pH than the bench top electrode for all except the 100% CS and 60% CS + 40%

SMC columns (Figure B-1)Figure B-1. Comparison of pH readings taken during continuous-

flow column test from bench top electrode and electrodes mounted in flow-through cells.

Symbols connected by a line indicate bench top electrode readings; unconnected symbols indicate

in-line electrode readings..

The 100% CS and 60% CS + 40% SMC columns exhibited different trends (Figure B-2).

The electrode associated with the 60% CS + 40% SMC column initially measured pH higher than

the bench top electrode, but after ~125 PV, exhibited the same trend as the majority of the in-line

93

electrodes (slightly lower than the bench top electrode). It is suspected that anaerobic biological

processes (i.e. fermentation) were not occurring to the same degree within this column at the

beginning of the experiment due to the introduction of O2 during a blow-out of a sampling port.

The electrode associated with the 100% CS column provided results which were

consistent with the bench top electrode for the first ~75 PV. From that point forward, results

fluctuated between pH 5.5 and 6.5, but did not deviate from this range. Upon inspection at the

conclusion of the experiment, it appeared that a biofilm had developed on the electrode surface

within the flow-through cell. Consequently, the microenvironment within the biofilm was being

maintained with a pH between 5.5 and 6.5, skewing results for the pH of the actual column

effluent water.

94

Figure B-1. Comparison of pH readings taken during continuous-flow column test from bench

top electrode and electrodes mounted in flow-through cells. Symbols connected by

a line indicate bench top electrode readings; unconnected symbols indicate in-line

electrode readings.

1

2

3

4

5

6

7

8

0 50 100 150 200

pH

Pore Volumes

Sand Control

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

50% CS + 50% SMC


95

Figure B-2. Comparison of pH readings taken during continuous-flow columns test from bench

top electrode and electrodes mounted in flow-through cells. Symbols connected by

a line indicate bench top electrode readings; unconnected symbols indicate in-line

electrode readings.

ORP electrodes were also found to have developed varying degrees of biofilms when

they were inspected near the end of the experiment. Although the trend for the crab shell

columns (with the exception of the 80% CS + 20% SMC column) indicated ORP slowly

increased from an initial value near -400 mV as expected, it is unknown if the rate of increase

was affected by the biofilm. It is also unknown if the results from the 80% CS + 20% SMC

column are more realistic, as that electrode also had biofilm growth. Thus, the data presented in

Figure B-3 cannot be presumed to be an accurate reflection of actual effluent water quality.

2

3

4

5

6

7

8

0 50 100 150 200

pH

Pore Volumes

100% CS

60% CS + 40% SMC

96

Figure B-3. ORP measured with in-line electrodes in effluent from continuous-flow columns

treating MIW from the Klondike-1 site.

-800

-600

-400

-200

0

200

400

600

800

0 50 100 150 200

OR

P (

mV

)

Pore Volumes


97

Appendix C

Metals Removal After Passive Aeration and Settling

Chapter 3.5 provides data for dissolved Al, Fe, Mn, cobalt, nickel, and zinc. For each

continuous-flow column, metals were measured at two points in the treatment scheme: in the

column effluent and also after passive aeration and settling of the column effluent. Some metals,

such as dissolved Fe, experienced enhanced removal after passive aeration and are discussed in

the aforementioned chapter. All metals data, both from the column effluent and after passive

aeration and settling is provided here for completeness and also for comparison purposes.

98

Figure C-1. Breakthrough curves for dissolved Al measured in continuous-flow columns treating

Klondike-1 MIW (A) and after subsequent passive aeration and settling (B).

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

0 50 100 150 200

Al (C

/C0)

Pore Volumes

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

0 50 100 150 200

Al (C

/C0) Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


B

A

99

Figure C-2. Breakthrough curves for dissolved Fe measured in continuous-flow columns treating

Klondike-1 MIW (A) and after subsequent passive aeration and settling (B).

0.0

0.5

1.0

1.5

2.0

2.5

0 50 100 150 200

Fe

(C

/C0)

Pore Volumes

B

0.0

0.5

1.0

1.5

2.0

2.5

0 50 100 150 200

Fe

(C

/C0)

Sand Control 100% CS90% CS + 10% SMC 80% CS + 20% SMC70% CS + 30% SMC 60% CS + 40% SMC50% CS + 50% SMC Traditional 90% SMC + 10% LS

A

100

Figure C-3. Breakthrough curves for dissolved Mn measured in continuous-flow columns


0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0 50 100 150 200

Mn

(C/C

0)


90% CS + 10% SMC 80% CS + 20% SMC

70% CS + 30% SMC 60% CS + 40% SMC


0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0 50 100 150 200

Mn

(C/C

0)

Pore Volumes

B

A

101

Figure C-4. Breakthrough curves for dissolved cobalt measured in continuous-flow columns


0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 50 100 150 200

Co

(C

/C0)

Pore Volumes

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 50 100 150 200

Co

(C

/C0)


B

A

102

Figure C-5. Breakthrough curves for dissolved nickel measured in continuous-flow columns


0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0 50 100 150 200

Ni (C

/C0)

Pore Volumes

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 50 100 150 200

Ni (C

/C0)


90% CS + 10% SMC 80% CS + 20% SMC

70% CS + 30% SMC 60% CS + 40% SMC


B

A

103

Figure C-6. Breakthrough curves for dissolved zinc measured in continuous-flow columns


0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

0 50 100 150 200

Zn

(C

/C0)

Pore Volumes

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

0 50 100 150 200

Zn

(C

/C0)


B

A

104

Appendix D

Sulfate Data

Chapter 3.5 includes a brief presentation of findings indicating likely sulfate (SO42-

)

reduction. SO42-

data was obtained via two analytical methods, IC and ICP, which yielded similar

results, indicating almost no SO42-

reduction was occurring within the treatment columns. ICP

data (Figure D-1) was originally reported as the element sulfur (S). Sample pretreatment and

preservation methods (Robinson-Lora and Brennan, 2010a) required acidification to pH < 2 and

sparging with lab air to drive off hydrogen sulfide from the sample prior to analysis so that all S

reported could be assumed to exist as SO42-

species. As such, SO42-

concentrations were

calculated from the reported S value according to the molecular weight ratio of SO42-

:S (2.996:1).

The data presented in Figure D-1 is after conversion to SO42-

, and lines are a running average of

interpolated data. Data obtained from the IC is not reported.

SO42-

data would be expected to display a negative correlation with metals removal and

DOC production (as sulfate goes down, metals removal and DOC should increase). However,

these correlations were not noted. In addition, the maximum sulfate removal appears to happen

around ~125 PV for all of the columns, which corresponds to pH below 5, where SRB are known

to show reduced activity. Considering these findings and the visual indications of potential

sulfate removal (Chapter 3.5), this data is not considered to be a valid assessment of sulfate

within the system.

105

Figure D-1. Sulfate data for continuous-flow columns treating Klondike-1 MIW.

600

800

1,000

1,200

1,400

1,600

1,800

0 50 100 150 200

SO4

2- (

mg

/L)

Pore Volumes

Influent

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


106

Appendix E

Cation Data Plots

Specific cations of interest (Ca, K, Mg, Na, and P) are discussed in Chapters 3 and 4.

Although some data points are referenced or included in tables in those chapters, the full data sets

are included in this Appendix.

Figure E-1. Dissolved Ca measured in continuous-flow columns treating Klondike-1 MIW. Inset

graph shows maximum values achieved at beginning of experiment; axes have same

units as large plot.

50

100

150

200

250

300

350

400

450

500

0 50 100 150 200

Ca

(m

g/L

)

Pore Volumes

Inf luent

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


0

200

400

600

800

1,000

1,200

1,400

1,600

0 5 10 15 20

107

Figure E-2. Dissolved K measured in continuous-flow columns treating Klondike-1 MIW. Inset



0

50

100

150

200

250

300

350

400

0 50 100 150 200

K (m

g/L

)

Pore Volumes

Inf luent

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


0

100

200

300

400

0 1 2 3 4 5

108

Figure E-3. Dissolved Mg measured in continuous-flow columns treating Klondike-1 MIW.

0

50

100

150

200

250

300

0 50 100 150 200

Mg

(m

g/L

)

Pore Volumes

Influent

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


109

Figure E-4. Dissolved Na measured in continuous-flow columns treating Klondike-1 MIW. Inset



4

5

6

7

8

9

10

11

12

13

14

0 50 100 150 200

Na

(m

g/L

)

Pore Volumes

Influent

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


0

100

200

300

400

500

600

700

800

900

0 1 2 3 4 5

110

Figure E-5. Dissolved PO43—

P measured in continuous-flow columns treating Klondike-1 MIW.

0

5

10

15

20

25

30

35

0 50 100 150 200

PO

43

- -P

(m

g/L

)

Pore Volumes

Influent

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


111

Appendix F

Metals Mass Balance Calculations

Mass balances were conducted on each column to determine the amount, if any, of each

metal (Al, Co, Fe, Mn, Ni, and Zn) retained within the column packing materials at the

completion of the experiment (after 181 days of continuous-flow operations).

The total mass entering each column was calculated using the formula:

Where is the calculated mass (mg) entering the column


is the influent concentration of the metal (mg/L) at sampling iteration (n)

is the PV of MIW treated between the current sampling iteration (n) and the

previous sampling iteration (n-1)


The following assumptions were made:

1). The concentration (mg/L) at each sampling iteration was assumed to be constant over

the entire volume of MIW treated since the previous sampling iteration ( ).

2). was assumed to be 1 for the first sampling iteration (immediately

following incubation) as the column was filled with 1 PV of MIW during incubation.

112

The total mass of Al, Co, Fe, Mn, and Ni exiting each column was calculated using the

formula:

Where is the calculated mass (mg) exiting the column

nEx.is the sampling iteration at exhaustion (defined as breakthrough to 50% of the influent

concentration)

is the effluent concentration of the metal (mg/L) at sampling iteration (n)

The assumptions stated above were also taken into account.

The results using this method for Zn did not produce logical results. Calculations

indicated mout of the columns exceeding min by significant amounts (> 50% for some columns).

Extractable metals results from the column packing materials (Table 2-4) were considered as a

possible source of the additional Zn. After any potential Zn released from the packing materials

was accounted for, masses still exceeded reasonable results. It is hypothesized that low influent

concentrations of Zn (average 0.26 mg/L) coupled with a high detection limit (0.20 mg/L)

resulted in an inadequate portrayal of mass retained within the system using the equations

described above. Effluent concentrations below the detection limit (BDL) were reported as the

detection limit (0.20 mg/L). During times of complete removal, this resulted in a small portion of

Zn seemingly being removed. For example, if the influent concentration was 0.28 mg/L and the

effluent was BDL, the mout of the column would have been calculated using a concentration of

0.20 mg/L. This would lead to an overestimation of the mass exiting the column, and a resulting

mout value which was significantly larger than the mass inputted into the system.

In order to better estimate the amount of Zn retained within the column, an additional

assumption was made: When concentration of the effluent was BDL, it was assumed that

113

complete removal occurred, and an effluent concentration of 0.0 mg/L was used. This resulted in

realistic numbers which are further discussed in section 4.2.4.

Calculations were accomplished in Microsoft Excel, and results are provided in Figure F-1 and

Table F-1.

Figure F-1. Percent of each metal retained within columns treating Klondike-1 MIW at

completion of experiment (after 181 days of continuous-flow conditions).

-10%

0%

10%

20%

30%

40%

50%

60%

70%

Al Co Fe Mn Ni Zn

% R

eta

ined

in

Co

lum

n

Sand Control

100% CS

90% CS + 10% SMC

80% CS + 20% SMC

70% CS + 30% SMC

60% CS + 40% SMC

50% CS + 50% SMC


114

Table F-1. Metals mass balance for continuous-flow columns conducted at completion of experiment (after 181 days of operation).

Treatment Column

Metals mass balance for continuous-flow columns (all m are in mg)

Al Co

min mout mretained % retained min mout mretained % retained

Sand Control 131 134 -2.5 -2% 20.0 20 -0.2 -1%

100% CS 134 133 1.2 1% 20.4 8.0 12 61%

90% CS + 10% SMC 138 143 -4.5 -3% 20.9 10 11 51%

80% CS + 20% SMC 136 140 -3.5 -3% 20.6 11 10 49%

70% CS + 30% SMC 134 129 4.9 4% 20.1 7.9 12 61%

60% CS + 40% SMC 132 124 7.9 6% 20.0 11 9.5 47%

50% CS + 50% SMC 135 133 1.8 1% 20.4 12 8.3 41%

Traditional 90% SMC + 10% LS 133 142 -8.9 -7% 20.1 17 2.9 15%

Fe Mn


Sand Control 2870 1770 1100 38% 1704 1730 -26 -2%

100% CS 2900 2090 810 28% 1740 1750 -10 -1%

90% CS + 10% SMC 3020 2090 930 31% 1780 1760 20 1%

80% CS + 20% SMC 2910 2260 650 22% 1760 1780 -20 -1%

70% CS + 30% SMC 2890 1950 940 33% 1710 1720 -10 -1%

60% CS + 40% SMC 2852 2250 602 21% 1710 1730 -20 -1%

50% CS + 50% SMC 2880 2060 820 28% 1730 1750 -20 -1%

Traditional 90% SMC + 10% LS 2880 1790 1090 38% 1720 1760 -40 -2%

115

Table F-1. (continued) Metals mass balance for continuous-flow columns conducted at completion of experiment (after 181 days

of operation).

Treatment Column

Metals mass balance for continuous-flow columns (all m are in mg)

Ni Zn


Sand Control 46 50 -4.3 -10% 13 13 -0.2 -1%

100% CS 47 32 14 31% 13 4.3 8.6 66%

90% CS + 10% SMC 49 37 12 25% 13 10 3.2 24%

80% CS + 20% SMC 47 37 10 22% 13 13 0.3 2%

70% CS + 30% SMC 47 33 14 30% 13 7.1 5.8 45%

60% CS + 40% SMC 46 39 6.5 14% 13 7.1 5.6 44%

50% CS + 50% SMC 47 42 4.8 10% 13 14 -0.9 -7%

Traditional 90% SMC + 10% LS 47 44 2.7 6% 13 14 -0.8 -7%

116

Appendix G

Visual MINTEQ Geochemical Modeling

Visual MINTEQ was used to determine solubility of different metal species over a range

of pH values. In addition, models attempted to determine differences in speciation and saturation

indices of metals and other cations under a range of sulfate/sulfide ratios. 9 scenarios were

assessed, 3 each for the 100% CS column, 70% CS + 30% SMC column, and traditional 90%

SMC + 10% LS column. Each scenario contained 6 iterations of different sulfate/sulfide ratios.

The scenarios and iterations are described in more detail in Table G-1 and Table G-2. Selected

results are presented in Figure G-1 and Figure G-2. In addition, these results were used to

determine free ion concentrations of cations in Chapter 4.4.

117

Table G-1. Visual MINTEQ geochemical modeling scenarios, consisting of 6 iterations of SO42-

:HS- ratios each

Scenario Description Iteration SO4

2-

(mg/L)

HS-

(mg/L)

1 100% CS column, Cation concentrations immediately following incubation A 75% sulfate removal 250 250

2 100% CS column, Cation concentration average after 10 PV B 50% sulfate removal 500 170

3 100% CS column, Cation concentration when pH =5 C 25% sulfate removal 750 83

4 70% CS column, Cation concentration immediately following incubation D 15% sulfate removal 850 50

5 70% CS column, Cation concentration average after 10 PV E 10% sulfate removal 900 33

6 70% CS column, Cation concentration when pH =5 F 1% sulfate removal 990 3.3

7 Traditional column, Cation concentration immediately following incubation

8 Traditional column, Cation concentration average after 10 PV

9 Traditional column, Cation concentration when pH =5

Table G-2. Carbonate, cation, and dissolved metals concentrations for Visual MINTEQ geochemical modeling scenarios.

Variable Carbonate/Cation Concentrations (mg/L) Constant Dissolved Metals Concentrations

(mg/L)

Scenario CO32-

Ca K Mg Na P Al Co Fe Mn Ni Zn

1 2316 1235 144 215 637 17.8 2.81 0.42 105 36.2 0.92 0.26

2 247 187 7.46 104 6.77 7.69 2.81 0.42 105 36.2 0.92 0.26

3 0 126 7.05 104 5.82 11.3 2.81 0.42 105 36.2 0.92 0.26

4 2290 1590 358 276 770 15.7 2.81 0.42 105 36.2 0.92 0.26

5 284 194 7.9 105 6.5 7.18 2.81 0.42 105 36.2 0.92 0.26

6 0 120 5.27 101 5.3 2.9 2.81 0.42 105 36.2 0.92 0.26

7 434 414 369 127 43 21.1 2.81 0.42 105 36.2 0.92 0.26

8 16.5 153 9.03 104 6 0.033 2.81 0.42 105 36.2 0.92 0.26

9 0 149 7.14 113 6.26 1.02 2.81 0.42 105 36.2 0.92 0.26

118

Figure G-1. Scenarios 1 (initial values after incubation), 2 (average after

10 PV), and 3 (pH=5) using the SO42-

:HS- ratio from iteration A. Results

reveal no considerable difference in solubility of metal species related to

variations in cation and carbonate loadings in the 100% CS column within

the pH range encountered during the continuous-flow columns

experiment (pH 2.5-7.5). In fact, iteration A for all 9 scenarios produced

similar results, indicating the SO42-

:HS- ratio dominates solubility of

metals within each of the systems under the given circumstances.

1.E-18

1.E-16

1.E-14

1.E-12

1.E-10

1.E-08

1.E-06

1.E-04

1.E-02

1.E+00

-1 4 9 14

log

[Me

] (M

)

pH

Al

Co

Fe

Mn

Ni

Zn

1A

1.E-18

1.E-16

1.E-14

1.E-12

1.E-10

1.E-08

1.E-06

1.E-04

1.E-02

1.E+00

-1 4 9 14

log

[Me

] (M

)

pH

Al

Co

Fe

Mn

Ni

Zn

2A

1.E-18

1.E-16

1.E-14

1.E-12

1.E-10

1.E-08

1.E-06

1.E-04

1.E-02

1.E+00

-1 4 9 14

log

[Me

] (M

)

pH

Al

Co

Fe

Mn

Ni

Zn

3A

119

Figure G-2. Effect on solubility/saturation of total dissolved Fe as

SO42-

:HS- ratios are increased (increased SO4

2-:HS

- ratio indicates limited

or no sulfate reduction is occurring).

1.E-18

1.E-16

1.E-14

1.E-12

1.E-10

1.E-08

1.E-06

1.E-04

1.E-02

1.E+00

-1 4 9 14

log

[Me

] (M

)

pH

Al

Co

Fe

Mn

Ni

Zn

3A

1.E-17

1.E-15

1.E-13

1.E-11

1.E-09

1.E-07

1.E-05

1.E-03

1.E-01-1 4 9 14

log

[Me

] (M

)

pH

Al

Co

Fe

Mn

Ni

Zn

3C

1.E-16

1.E-14

1.E-12

1.E-10

1.E-08

1.E-06

1.E-04

1.E-02

1.E+00

-1 4 9 14

log

[Me

] (M

)

pH

Al

Co

Fe

Mn

Ni

Zn

3E

120

Appendix H

Organic Carbon Mass Balance Calculations

An organic C mass balance was conducted on each column to determine the amount, if

any, of organic C remaining within the substrate materials at completion of the experiment (after

181 days of continuous flow operations). The stoichiometric ratio of carbon within the crab shell

structure was determined to be

The starting mass of organic C within each column was calculated using the formula:

Where is the calculated mass (mg) of organic C within the column.

is the mass (mg) of crab shell in the substrate mixture.

is the mass (mg) of spent mushroom compost in the substrate mixture.

is the carbon content of crab shell (%).

is the carbon content of spent mushroom compost (%).

is the % of the total carbon which is organic carbon. Calculated from the

stoichiometric carbon content of chitin, protein, and mineral portions of crab

shells and found to be 58%.

is the % of the total carbon which is organic carbon. Determined to be 100%

for SMC as there are no known sources of inorganic carbon within SMC.

Values for and can be found in Table 2-3and and can be found in Table

2-4.

The total mass of organic C exiting each column was calculated using the formula:

121

Where is the calculated mass (mg) of organic C exiting the column


is the effluent DOC concentration (mg/L) at sampling iteration (n)

is the PV of MIW treated between the current sampling iteration (n) and

the previous sampling iteration (n-1)


The following assumptions were made:

1). The concentration (mg/L) at each sampling iteration was assumed to be constant over

the entire volume of MIW treated since the previous sampling iteration (

).

2). was assumed to be 1 for the first sampling iteration (immediately

following incubation) as the column was filled with 1 PV of MIW during incubation.

Calculations were accomplished in Microsoft Excel, and results are provided in Table H-1.

Table H-1. Organic carbon mass balance for continuous-flow columns treating Klondike-1 MIW

(performed at completion of experiment, after 181 days of operation).

Treatment Column mOC in mOC out mOCremaining % OC

remaining

Sand Control 126 132 -6 -5%

100% CS 5694 3840 1854 33%

90% CS + 10% SMC 5617 2594 3023 54%

80% CS + 20% SMC 5540 2048 3493 63%

70% CS + 30% SMC 5464 2078 3385 62%

60% CS + 40% SMC 5387 1671 3716 69%

50% CS + 50% SMC 5310 1503 3807 72%

Traditional 90% SMC + 10% LS 4446 250 4196 94%

122

Results indicated a general correlation (R2=.88) between the amount of crab shell within

the system and the % OC utilized (Figure H-1).

Figure H-1. Correlation between amount of crab shell within treatment system and amount of

original organic carbon remaining within system at completion of continuous-flow

experiment treating Klondike-1 MIW.

R² = 0.8838

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 20 40 60 80 100

% O

rgan

ic C

Rem

ain

ing

% Crab Shell in System

123

Appendix I

Treatment Scale-up using a 1:1 Crab Shell to Proppant Ratio

Chapter 4.5 discussed the longevity estimations for columns packed under ideal usage

conditions with a crab shell to proppant ratio of 1:1 by mass. The following calculations were

used to determine the actual mass of substrate materials that could have theoretically fit into the

~700 mL volume of the continuous-flow column. Bolded lines indicate the closest value to 700

mL. Once total theoretical masses were calculated, they were divided by the actual mass of

substrate used in each treatment column (40 g) to determine a scale-up factor (Table I-8)

Table I-1. Theoretical total mass of crab shell and sand able to fit into a 100% crab shell column

(~700 mL) assuming a 1:1 crab shell to sand proppant mass ratio.

100% Crab Shell

Crab Shell (CS) Sand CS + Sand

Volume

(mL)

Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

315 700 315 203 903

299 665 299 193 858

284 630 284 183 813

268 595 268 173 768

252 560 252 163 723

249 553 249 161 714

246 546 246 159 705

243 539 243 156 695

189 420 189 122 542

Bulk density of CS = 0.45 g/mL; Bulk density of sand = 1.55 g/mL

124

Table I-2. Theoretical total mass of crab shell, SMC, and sand able to fit into a 90% crab shell +

10% SMC column (~700 mL) assuming a 1:1 crab shell to sand proppant mass ratio.

90% Crab Shell & 10% SMC

Crab Shell (CS) SMC Sand CS + SMC + Sand

Volume

(mL)

Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

248 550 28 92 248 160 801

243 539 27 90 243 156 785

238 528 26 88 238 153 769

233 517 26 86 233 150 753

228 506 25 84 228 147 737

223 495 25 83 223 144 721

218 484 24 81 218 141 705

213 473 24 79 213 137 689

Bulk density of CS = 0.45 g/mL; Bulk density of SMC = 0.3 g/mL; Bulk density of sand = 1.55 g/mL





Volume

(mL) Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

216 480 54 180 216 139 799

212 470 53 176 212 137 783

207 461 52 173 207 134 767

203 451 51 169 203 131 751

199 442 50 166 199 128 735

194 432 49 162 194 125 719

190 422 48 158 190 123 703

186 413 46 155 186 120 687

181 403 45 151 181 117 671


125





Volume

(mL) Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

191 425 82 273 191 123 822

187 417 80 268 187 121 805

184 408 79 262 184 118 789

180 400 77 257 180 116 772

176 391 75 251 176 114 756

172 383 74 246 172 111 739

168 374 72 240 168 109 723

164 366 70 235 164 106 707

161 357 69 230 161 104 690






Volume

(mL) Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

158 350 105 350 158 102 802

154 343 103 343 154 100 786

151 336 101 336 151 98 770

148 329 99 329 148 96 754

145 322 97 322 145 93 737

142 315 95 315 142 91 721

139 308 92 308 139 89 705

135 301 90 301 135 87 689

132 294 88 294 132 85 673


126





Volume

(mL)

Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

124 275 124 413 124 80 767

121 270 121 404 121 78 752

119 264 119 396 119 77 737

116 259 116 388 116 75 721

114 253 114 380 114 73 706

111 248 111 371 111 72 691

109 242 109 363 109 70 675

106 237 106 355 106 69 660


Table I-7. Theoretical total mass of SMC and limestone able to fit into a traditional 90% SMC +

10% limestone column (~700 mL).

90% SMC & 10% Limestone Chips

SMC Limestone SMC + Limestone

Volume

(mL)

Mass

(g)

Volume

(mL)

Mass

(g)

Volume

(mL)

210 700 23 18 718

206 686 23 18 704

202 672 22 18 690

197 658 22 17 675

193 644 21 17 661

189 630 21 16 646

185 616 21 16 632

Bulk density of SMC = 0.3 g/mL; Bulk density of limestone = 1.28 g/mL

127

Table I-8. Calculated scale-up factors based on theoretical total mass of substrate required to fill

~700 mL volume and actual mass used in the experiment.

Treatment Column

Theoretical

Mass Substrate

Required (g)

Total Mass

Substrate in

Experiment (g)

Scale-up

Factor

100% CS 246 40 6.2

90% CS + 10% SMC 242 40 6.1

80% CS + 20% SMC 238 40 6.0

70% CS + 30% SMC 235 40 5.9

60% CS + 40% SMC 226 40 5.7

50% CS + 50% SMC 213 40 5.3

Traditional 90% SMC + 10% LS 229 40 5.7

128

Appendix J

Field Pilot System Installation and Sampling Photos

Figure J-1. MIW at the Klondike-1 site.

Figure J-2. Tank piping modifications and installation of underdrain piping network, July 26,

2010.

Figure J-3. Placement of septic tanks used to simulate pilot-scale VFPs to treat MIW at the

Klondike-1 site.

Figure J-4. Placement of rock underdrains into tanks, done manually to avoid damage to

underdrain piping system!

Figure J-5. Completed installation of limestone rock underdrain system

Figure J-6. 1,000 pound super sack of crab shell unloaded into back of dump truck to be mixed

with sand proppant and SMC.

Figure J-7. Filling of organic substrate mixtures into pilot-scale VFPs to treat MIW at the

Klondike-1 site

Figure J-8. Placement of microbial tea-bag style sampling pouches 8-10 inches into organic

substrate material.

Figure J-9. A layer of pea gravel was added to the top of each reactor to prevent loss/disturbance

of organic substrate.

Figure J-10. Installation of influent piping system. Pipes were emplaced to gravity feed from an

oxidation pond of the existing full-scale treatment system at the Klondike-1 site. A

dock was built to facilitate maintenance of influent hose lines.

Figure J-11. Individual influent hoses attach to the buried PVC piping approximately 12 inches

below the water surface and feed water to each pilot-scale VFP. Water enters

through ¼ inch holes drilled into the final 2 feet of flexible tubing, which is covered

with mesh to discourage iron precipitates from entering the system.

Figure J-12. Piping network leading from oxidation pond of current full-scale treatment system

to feed pilot-scale VFPs.

Figure J-13. View of the four pilot-scale VFPs and aerobic settling ponds.

Figure J-14. Earl Smithmyer, President of the CCWA, assisted tremendously with the pilot-

system installation, specifically with the piping networks.

Figure J-15. Water was added to the pilot-scale VFPs on August 2, 2010. Some overflow

problems were encountered with the settling ponds, as they were not properly

leveled.

Figure J-16. Orifices were created to maintain a flow rate of 0.2 gallons per minute throughout

the pilot-scale study.

Figure J-17. Aeration of the tank effluent was encouraged via a miniature cascade constructed

with corrugated piping.

Figure J-18. The system was flushed and then left to incubate for a week prior to initiation of

continuous-flow operations.

Figure J-19. Sampling event at the pilot-scale VFPs during Fall 2010.

Figure J-20. As expected, the pilot systems froze over the winter months.

Figure J-21. Many thanks to Shan Lin and Sara Goots, whose help made the pilot system install

go quickly, and who have conducted analysis of the data from the field pilot-scale

study!

Figure J-22. Thanks to Duke, for cheerfully spending his summer at the Klondike-1 site with me!

Date post:	08-Dec-2021
Category:	Documents
Upload:	others
View:	3 times
Download:	0 times

REMEDIATION OF HIGH-STRENGTH MINE IMPACTED WATER …

Documents