Citation
Filtration and life span analysis of a pervious concrete filter

Material Information

Title:
Filtration and life span analysis of a pervious concrete filter
Creator:
Majersky, Gregory Michael
Publication Date:
Language:
English
Physical Description:
ix, 81 leaves : illustrations ; 28 cm

Subjects

Subjects / Keywords:
Lightweight concrete ( lcsh )
Filters and filtration ( lcsh )
Pavements, Porous ( lcsh )
Filters and filtration ( fast )
Lightweight concrete ( fast )
Pavements, Porous ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 80-81).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Gregory Michael Majersky.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
262685212 ( OCLC )
ocn262685212
Classification:
LD1193.L547 2008m M34 ( lcc )

Full Text
FILTRATION AND LIFE SPAN ANALYSIS
OF A PERVIOUS CONCRETE FILTER
BS Environmental Science Slippery Rock University, 1998
A thesis submitted to the
University of Colorado Denver
In partial fulfillment
Of the requirements for the degree of
Master of Science
Environmental Engineering
2008
by
Gregory Michael Majersky


Copyright April 29, 2008, Gregory Michael Majersky


This thesis for the Master of Science
degree by
Gregory Michael Majersky
has been approved
by
Stephan A. Durham
Anuradha Ramaswami
Hhokvo^
David Mays


Majersky, Gregory Michael (M.S., Civil Engineering)
Filtration and Life Span Analysis of a Pervious Concrete
Thesis directed by Assistant Professor Stephan Durham
ABSTRACT
Pervious concrete has grown in popularity as an alternative infrastructure
medium to enhance the quality of surface water in developed geographical areas.
Previous studies have shown that pervious concrete structures and the subgrade
layer that supports them are effective at significantly reducing the concentrations
of common pollutants such as soap, motor oil, brake fluid, brake dust, and roof
shingle particles from rain water and snow melt. Studies thus far have shown
enough promise to encourage growing use of pervious concrete in infrastructure
categories such as storm channels, sidewalks and parking lots. This paper
examines the use of pervious concrete to remove metals and neutralize pH from
a synthetic solution simulating acid mine drainage (AMD). In addition,
destructive testing of cores taken from pre-filtration and post filtration filters were
used to evaluate the potential lifespan of a pervious concrete filter when
exposed to the prolonged flow of polluted water.


This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
---------2Z3------------
Stephan A. Durham


ACKNOWLEDGEMENTS
The University of Colorado Denver has completed this work funded in part by the
Environmental Protection Agency through their P3 student design competition.
Further, I would like to thank Dr. Linda Figueroa at the Colorado School of Mines
for providing site specific knowledge that aided in the laboratory replication of
acid mine drainage (AMD) conditions for the purpose of this research.
Most of all, Id like to thank Dr. Stephan Durham for his direction and support of
this research, Dr. David Mays and Dr. Ramaswami for their perspectives and
opinions and Dr. Randy Tagg for pointing this conceptual research in the right
direction and his continual follow up.


TABLE OF CONTENTS
List of Figures..................................................vi
List of Tables.................................................viii
Chapter
1 Introduction....................................................1
1.1 Overview....................................................1
1.2 Research Objective..........................................2
2 Background.....................................................4
2.1 EPA P3 Competition............................................4
2.2 EPA P3 Report Results.........................................6
2.2.1 Filter Alternatives.........................................7
2.2.2 Filter Cost Estimate........................................9
2.2.3 Research Objectives........................................10
2.2.4 Scope of Work..............................................10
2.2.5 Data Results and Findings..................................14
2.2.6 Discussions, Conclusions and Recommendations...............20
3. Literature Review.............................................23
3.1 Purpose, Research Objectives and Scope of Work...............23
3.1.1 Purpose....................................................23
3.1.1.1 The Formation of Acid Mine Drainage......................23
3.1.1.2 Active versus Passive Remediation........................26
v


3.1.1.3 Active or Traditional Remediation.......................26
3.1.1.4 Passive or Innovative Remediation.......................27
3.1.1.5 Passive Remediation Techniques...........................28
3.1.2 Research Objectives........................................31
3.1.2.1 Pervious Concrete as a PRB..............................31
3.1.2.2 Material Description of Pervious Concrete................34
3.1.3 Scope of Work..............................................36
3.1.3.1 Recovery of Metals In The Waste Stream..................36
4 Problem Statement.............................................38
5 Experimental Design and Procedure.............................40
5.1 Experimental Design......................................40
5.2 Experimental Procedure.......................................40
5.3 Experiment Equipment.........................................49
6 Results.......................................................57
6.1 AMD Metals Filtration Results................................57
6.2 Compressive Strength Test....................................58
6.3 Filter 1 and 2 Performance Comparisons.......................63
6.4 Conversion of Species from Dissolved to Suspended Particles.67
6.5 Compressive Strength Test Results............................72
7 Conclusions and Recommendations for Future Studies............74
Bibliography.....................................................79
vi


LIST OF FIGURES
Figure
2.1. Water Sample Poured onto the Surface of the Filter.............12
2.2. Concentration of Metals After Erosion Study....................14
2.3. Dissolved Oxygen and Temperature Measurements..................16
2.4. Percent Removal of Metals for Unsieved and Sieved Filters......18
3.1. a and 3.2b Pervious vs. Impervious Concrete Surfaces...........32
5.1. Sampling Bottles...............................................41
5.2. Distilled Water................................................41
5.3. Sodium Sulfate................................................ 41
5.4. Iron and Zinc Sulfate..........................................41
5.5. Experimental Flow Rate.........................................42
5.6. Approaching Maximum Accommodating Flow Rate....................42
5.7. AMD Filtration Testing In Progress.............................48
5.8. Laboratory Model of AMD Test Configuration.....................54
5.9. Wrapped Filter Supports and Sample Collection Tray.............55
5.10. AMD Test Apparatus Without the Filter........................55
5.11. Complete AMD Test Apparatus, Side View.......................56
5.12. Complete AMD Test Apparatus, Front View......................56
6.1. Filtrate Concentrations versus Time...........................62
6.2. Filtrate Concentrations versus Time...........................63
vii


6.3. Percent Concentration of Iron in Filters 1 and 2
64
6.4. Percent Concentration of Sodium in Filters 1 and 2..........65
6.5. Percent Concentration of Zinc in Filters 1 and 2.............65
6.6. Percent Concentration of Sulfate in Filters 1 and 2..........66
6.7. Percent pH modification by Filters 1 and 2...................66
6.8. Filter 1 Dissolved Species Concentrations....................68
6.9. Filter 1 Dissolved Species Concentrations....................69
6.10. P3 vs. AMD Filter 1 Performance Comparison.................71
6.11. P3 vs. AMD Filter 2 Performance Comparison.................71
6.12. Comparisons of Compressive Strength Test Results...........72
viii


LIST OF TABLES
Tables
2.1 Materials Cost for Pervious Concrete Filter......................... 9
2.2 Filter Cost..........................................................9
2.3 Bacterial Filtration Results...................................... 15
2.4 Bacterial Dissolved Oxygen Results..................................16
5.1 Silver Cycle mine AMD composition...................................46
6.1 Filter 1 Results....................................................60
6.2 Filter 2 Results....................................................61
6.3 EPA P3 Cation Filtration Results....................................70
6.4 Percent Difference in Compression Strength Tests....................73
IX


Chapter 1
Introduction
1.1 Overview
Acid mine drainage (AMD) begins when pyrite (FeS) is exposed to water
and oxygen, releasing hydrogen ions [EPA, 2007], In unexposed
subsurface areas this process is slow and can be naturally buffered by
surrounding water. Mining increases the amount of water that can absorb
oxygen, thus accelerating the process. Further hydrogen ion release and
the resulting lowering of pH occurs when ferric iron (Fe3+) precipitates out
as Fe(OH)3 or combines with more pyrite to accelerate the original
process. The increased acidity of the AMD also dissolves other minerals
out of the surrounding rock.
Pervious concrete is an innovative construction material that offers
numerous economical and environmental benefits. Pervious concretes,
like conventional concrete consists of portland cement, water, and
aggregates. The porous nature of this paving material is achieved by
eliminating the fine aggregates from the concrete mixture. By reducing
the amount of fines in the mixture, air voids are created in the concrete
allowing water to pass through the concrete pavement into the underlying
-1 -


soil layer. It is the presence of fine particles throughout the concrete
mixture that makes concrete impervious when applied to sidewalks,
driveways, or other such structures. This imperviousness is seen in the
runoff that occurs during precipitation or snowmelt events. Where
pervious concrete has been placed, runoff has been dramatically reduced
and evidence exists demonstrating that the water quality entering the
groundwater or underground drainage system has been improved due to
the removal of pollution deposited by parked cars, lawn fertilizer, soap, etc.
The objective of this study is to evaluate the effectiveness of pervious
concrete as a water filter for general water quality improvement of acid
mine drainage (AMD). Current AMD remediation techniques and
technologies are costly to install, require large areas of land and
installation and maintenance are energy intensive. Recent studies have
been undertaken to examine how to recover dissolved metals from AMD
remediation systems. However, these studies have shown that the
process of removing metals to be costly and energy intensive.
1.2 Research Objective
The objective of this thesis is to analyze the efficiency of pervious
concrete at removing dissolved metals, generating metal precipitates that
-2-


will settle in the filtered solution and measuring the rate of decay of the
filter when exposed to the flow of simulated AMD. This will be achieved
by:
1. Constructing one pervious concrete filter to be used solely for
compressive strength testing; two more filters were fabricated for
the filtration experimentation. One filter used in the filtration test
will be used for post-filtration compressive strength testing.
2. Formulate a synthetic solution with a pH level and ion
concentrations that simulate the concentrations found by analysis
of the National Tunnel Mine AMD by the Colorado School of Mines.
3. Test the pervious concrete filter by collecting the filtrate in one
beaker, then using a 0.25 micron glass filter to separate precipitate
from dissolved ions for analysis.
-3 -


Chapter 2
Background
2.1 EPA P3 Competition
Research on pervious concrete filtration began with the submission of a
grant proposal for the EPAs 2006-2007 People, Places and Prosperity
(P3) competition, which focuses on encouraging student research teams
to find methods and develop technologies that are commercially viable
and will improve the lives of people primarily in developing countries.
Once the EPA awarded the grant, an advisory panel of three faculty
members was formed. These faculty members represented specialties in
the filter material, fluid mechanics, chemistry and sustainability as it
applies to the developing world. The purpose of this panel was to advise
the primary researcher on what direction(s) to proceed to achieve the best
results for the P3 competition. The results of this experiment
demonstrated initial promise in the areas of dissolved oxygen
enhancement, pH modification, pathogen removal and neutralization as
well as the removal of dissolved ions from solution and encouraged further
research. Of special note is the filters apparent ability to remove
significant quantities of dissolved sodium from solution, though not enough
to qualify as desalinization to the point of making water potable.
-4-


The original purpose of this filtration project was to design a filter that
could economically meet the potable water needs of developing nations
and take advantage of existing infrastructure and local expertise.
Approximately 65 to 70% of the rural populations in developing areas of
the world do not have access to a safe source of water [Water Partners
International, 2006], In addition, more than five million people die from
water-related disease [Pacific Institute, 2002], These facts substantiate
the need to provide clean drinking water to a large portion of the worlds
population. In fact, the author of this thesis has personally witnessed
people drinking from sewage and industrial waste contaminated canals in
the Min Hang district of Shanghai, China. This observation reinforces the
need to examine possible methods of providing suitable drinking water in
developing communities of the world. The need to supply clean drinking
water can be emphasized by the following facts:
The World Health Organization estimates that 80% of all sickness
in the world can be contributed to non-potable water and sanitation
[Washington Post, 1997],
If no action is taken to provide suitable means of obtaining clean
drinking water, as many as 135 million people will die from water-
related diseases by 2020 [Pacific Institute, 2002],
-5 -


Data relating water, sanitation, and hygiene intervention has shown
a decrease in sickness from diarrhea by 25-33% [Esrey, et. al.
1991],
2.2 EPA P3 Report Results
The mission of the EPA People, Prosperity, and the Planet (P3) project is
to address environmental issues while providing economic opportunity and
maintaining economic growth. It is the hypothesis of this author that the
pervious concrete filter addresses the everyday lack of clean drinking
water in many countries and has the potential to provide drinking water in
flooded areas, where potable water infrastructure may be rendered
inadequate. This filter has the potential to help maintain economic growth
by improving overall public health and morale by providing improved
drinking water quality. Sickness and fatality could possibly be reduced by
this low cost solution to improving water quality. The target populations
for the P3 study were residents of rural and suburban areas without
sufficient water treatment facilities. The research team hypothesized that
this filter system could produce a sufficiently high quality level of drinking
water for developing communities throughout the world.
-6-


2.2.1 Filter Alternatives
Prior to selecting pervious concrete as a filtration material, a review of
alternative point-of-use water treatment technologies was conducted. The
benefits and disadvantages of each are presented in the following:
Chlorine is a widely available chemical available in tablet form that
dissolves easily in water. Small amounts of chlorine are effective
against most types of bacteria and in these concentrations chlorine
is generally not harmful to humans.
Using old fabric has shown to significantly reduce bacteria counts
and remove visibly unattractive sediments. Natural fabrics will
swell and absorb water as it is poured through the fabric, trapping
bacteria and sediments. The economic condition of people who
use this method is often very poor. Old clothing that is used to treat
water must be replaced by buying or making new clothing,
something that may not be economically feasible. Also, the cloth
filters will retain the trapped contaminants and must be washed
separately in clean water (which will become polluted) to maintain
usefulness.
-7-


Sand filtration has become popular due to the effectiveness of sand
in removing bacteria and other pollutants, as well as the general
availability of sand. However, rapid sand filters must be
backwashed everyday to maintain effectiveness, which limits their
use in areas lacking the electrical power necessary to run the
backwash pumps.
Each of these approaches has a number of advantages and
disadvantages, suggesting that a combination may provide a reliable,
cost-effective, and environmentally responsible approach to point-of-use
water treatment. The goal of this study is to investigate pervious concrete
as a complimentary approach that would take advantage of two distinct
features:
Pervious concrete filters are unsaturated, unlike cloth or sand
filtration. This introduces an air-water interface that may provide
treatment analogous to that in a trickling filter.
The cement paste itself may provide treatment through its
interaction with the trickling water, representing a second qualitative
difference with respect to other filter materials.
-8-


2.2.2 Filter Cost Estimate
As shown in Table 2.1, the cost to produce a pervious concrete filter is
relatively inexpensive. The cost of a single pervious concrete filter with
the dimensions of 10 inches x 10 inches x 18 inches is $2.45. This cost
includes only material expenses for cement, rock, and water. Cost
associated with mixing the concrete mixture was excluded since the
sample could easily be mixed by hand.
Table 2.1 Materials Cost for Pervious Concrete Filter
Sample Size 1.04 cf Amount ($US)
Cement 21.2 lb 1.27
Rock 130.7 lb 1.18
Water 7.2 lb 0.002
Pervious Concrete Filter Materials Cost = 2.45
Additionally, not all locations around the world will have the industrial
infrastructure necessary to produce chemical forms of water treatment.
Developing areas may not be able to produce the required amount of
potable water for an entire community. The necessary industrial
infrastructure required to mass produce pervious concrete filters is already
present in many of these areas. Table 2.2 provides cost estimates of the
pervious concrete filter and currently available filter materials [Existing
Water Filtration Methods, 2007]. In addition, the estimated life of the filter
is included.
-9-


Table 2.2 Filter Cost [Existing Water Filtration Methods, 2007]
Filter Method Cost ($US) Estimated Life of Filter
Pervious Concrete 2.45 2-4 Months
Activated Carbon Cartridge 15.00 1 Month
5-15 Micron Fabric 511.75 3-6 Months
Sari Cloth Fabric <1.00 < 30 Days
2.2.3 Research Objectives
The objectives of the P3 research were:
To demonstrate that pervious concrete filters can be fabricated as
regular concrete but without fine grained materials. The ease and
simplicity in fabrication of these samples does not require special
training and special fabrication materials.
To examine the pervious concrete filter effectiveness in improving
overall water quality for consumption or other uses.
To investigate the hypothesis that eliminating gravel larger than
0.25 inches (sieved) is more effective in bacteria and inorganic
compound removal than an unsorted composition of coarse grains
(unsieved) when the filter dimensions were identical.
2.2.4 Scope of Work
The scope of work for the P3 research included designing, fabricating,
testing, and analyzing pervious concrete filters for their effectiveness in
-10-


removing bacteria and other contaminants. The scope of work for this
study included:
Designing the shape and size of the pervious concrete filters
Designing the pervious concrete mixture
Fabricating the pervious filter
Testing the pervious filter by measuring the contaminants before
and after the test water passed through the sample
Analyze the test results to determine the filter effectiveness
The pervious concrete filter was selected to have a 10in x 10in. cross-
section and depth of 18in. to allow for adequate area for the water to flow
through the section. Pervious concrete typically has a water-to-cement
ratio (w/c) of approximately 0.30 and a total cementitious content of 500 to
600 Ib/cu.yd. The P3 team chose to use a mixture with a w/c equal to
0.30 and a total cement content of 550 Ib/cy. These values were
determined based on trial and error of previous mixtures. Wood forms
conforming to the size needed for the filter were constructed such that
reuse of the forms would be possible. Three pervious concrete samples
were fabricated. One specimen was used to examine the possibility of
physical erosion due to flowing water. This filter was unsieved, using all
sizes of the coarse aggregate meeting ASTM C33. Two additional filters
-11 -


were constructed to examine the impact coarse aggregate particle size
had on filter effectiveness. One of the filters was fabricated with sieved
coarse aggregate. The particle size of the coarse aggregate was reduced
to a maximum size of 0.25in. The other sample was not sieved. Both
filters (sieved and unsieved) were exposed to identical and increasing
concentrations of bacteria and inorganic solutions to examine each filters
effectiveness. All solutions used when testing water quality were prepared
using deionized water. In addition, the sample collection tray was
thoroughly cleaned using soap and hot water. Each collection container
was rinsed with deionized water between each sample collection period.
Figure 2.1 shows a water sample being poured onto the surface of the
filter.
The filters were tested by measuring water samples contaminated with
bacteria and metals before and after the water passed through the
pervious sample. The scope of this research was limited to the use of less
hazardous materials in order to reduce the risk of injury or impairment to
research personnel and minimize possible contamination of the laboratory
environment.
- 12-


Figure 2.1. Water Sample Poured onto the Surface of the Filter
The bacteria used for coliform research was Micrococcus luteus. This
species has a size range of 0.5 to 3.5 micrometers and average size of
2.0 micrometers [Landau, 2002], These size parameters allow for this
bacterium as a suitable replacement for potentially more hazardous E. coli
species. Diluted solutions of Micrococcus luteus were prepared at 25, 30,
35, 40 and 45 mg of stock/L of deionized water. These samples were
then measured for optical density as Total Suspended Solids (TSS). To
simulate inorganic contamination and measure desalinization potential,
solutions of increasing concentrations of sodium, iron, and copper were
prepared.
- 13 -


2.2.5 Data Results and Findings
The first test performed was an erosion test. This was to examine how
much, if any, cementitious material or aggregate would be physically
removed by the presence of flowing water. Initially, the first three tests
produced slightly turbid water with a small amount of pebbles. As a result
of this test, the research team began to flush the filter with cold tap water
for 70 minutes at a flow rate of 1 L/hr. Laboratory analyses of samples
taken from the erosion test were measured for metals concentrations and
pH at Evergreen Analytical in Lakewood, CO with an Inductively Coupled
Plasma (ICP) instrument. The erosion test samples were analyzed for
total recoverable metals, with the primary focus being iron, aluminum, and
magnesium. Samples 2, 3, and 4 experienced higher than expected
levels of magnesium. Results showed that both iron and aluminum
concentrations were within the Environmental Protection Agency (EPA)
secondary maximum concentration limits (MCLs) for drinking water [EPA,
2007]. Figure 2.3 shows the concentrations of iron, aluminum, and
magnesium for the seven samples taken during the erosion testing.
- 14-


Erosion Study Metals Concentration
1 2 3 4 5 6 7
Sample
Iron Aluminum Magnesium
Figure 2.2. Concentrations of Metals After Erosion Study
Bacterial testing was performed on the sieved and unsieved filters. These
tests were conducted to provide a comparison of filter performance
between the unsieved vs. sieved filters. The bacterial filtration results are
shown in Table 2.3.
Table 2.3 Bacterial Filtration Results
Stock Unsieved Filter Sieved Filter
Dilutions Pre-Filtration Post-Filtration Pre-Filtration Post-Filtration
mL/500 mL TSS (mg/L) TSS (mg/L) TSS (mg/L) TSS (mg/L)
25 100 0.000001 137 0
30 134 0.000002 175 0
35 141 0 191 0.000001
40 177 0.000001 251 0
45 174 0 265 0
-15 -


Pre-filtration and post-filtration dissolved oxygen (DO) measurements
were taken to measure oxygenation abilities of the unsieved and sieved
filters. The sieved filter produced an average of 0.21 mg/L increase in DO
levels over the unsieved filter. Percent removal of bacteria was calculated
on a mg of bacteria/L basis. One Micrococcus luteus has an average wet
weight of 0.6 picograms [Landau, 2002], The filter was successful in
removing bacteria from a concentration of about 10A8 bacterial per mL of
water to less than 1 per mL. The percentage of bacterial removal for both
filters was well in excess of the EPA primary MCLs for bacteria (99.9%).
Table 2.4 shows the bacterial dissolved oxygen results for the sieved and
unsieved filters.
Table 2.4 Bacterial Dissolved Oxygen Results
Filter with Unsieved Coarse Aggregate Filter with Sieved Coarse Aggregate
Stock dilutions mL/500 mL Bacterial TSS (mg/L) Pre- filtration DO Post- filtration D.O. DO difference Bacterial TSS (mg/L) Pre- filtration DO Post- filtration D.O. DO difference
25 100 3.87 5.30 1.43 137 3.54 4.78 1.24
30 134 3.27 4.99 1.72 175 2.70 4.48 1.78
35 141 3.07 5.19 2.12 191 2.38 4.57 2.19
40 177 2.26 4.70 2.44 251 2.15 4.81 2.66
45 174 3.57 4.97 1.40 265 2.13 4.41 2.28
Dissolved oxygen and temperature change measurements were also
taken during pre-filtration and post-filtration. The unsieved filter produced
-16-


a greater increase in D.O. and a greater decrease in water temperature.
See Figure 2.3.
Dissolved Oxygen and Temperature Measurments
Sample Number
0.00
-0.20
-0.40
-0.60
-0.80
-1.00
-1.20
-1.40
-1.60
D.O. Change (Unsieved) a D.O. Change (Sieved)
Temp. Difference (Unsieved) Temperature Difference (Sieved)
Figure 2.3. Dissolved Oxygen and Temperature Measurements
Dissolved metals analysis was performed to evaluate both filters ability to
remove particles smaller than a bacterium (viruses, molecules and atoms)
with the hope that this filtration system could provide overall water quality
improvements with respect to virus removal, removal of hazardous organic
compounds, removal of hazardous metallic elements and possibly partial
desalinization. The testing materials that were selected as contaminates
included sodium, iron and copper. These contaminates are of minimum
toxicity to provide a safe work environment and sufficiently small diameter
- 17-
Temperature
Difference (Celcius)


to allow the results to be extrapolated to more hazardous, larger diameter
compounds.
The results of the metals analysis found that iron was removed to less
than the minimum detection level of the ICP in all but the third trial. The
concentrations of iron in both cases were in excess of the EPAs
secondary drinking water standard of 0.3 mg/L [EPA, 2007]. The average
percent removal of iron by the unsieved filter was 0.15% greater than the
sieved filter. Sodium was added to simulate sea water at an average
concentration of 35,000 mg/L. The percentage of sodium removed by
both filters increased similarly with each successive trial. However, the
average percentage of sodium removed by the sieved filter was greater
than that of the unsieved filter by 1.13%. The rate of percent increase in
sodium removal was 0.074 % greater for the sieved filter than the
unsieved filter. Figures 2.4 and 2.5 shows the percent removal of metals
for both the unsieved and sieved filters, respectively.
-18-


Metals Removed By Unsieved Filter
CD
O)
re
*-*
c
cd
o
_
CD
0.
V
o'
12 3 4
Copper Removal
Iron Removal
Sodium Removal
Sam pie
Figure 2.4. Percent Removal of Metals for Unsieved Filter
Metals Removed By Sieved Filter
a>
05
re
c
CD
2
CD
0.
120.00
100.00
80.00
60.00
40.00
20.00
0.00
Copper Removal
Iron Removal
Sodium Removal
12 3 4
Sample
Figure 2.5. Percent Removal of Metals for Sieved Filter
- 19-


Lastly, pH was measured for each sample. The range of pH of the filtered
water was 11.45 to 12.52. This very high level of pH is of concern and
should be investigated in further studies.
2.2.6 Discussions. Conclusions and Recommendations
This study discussed a potentially new method of providing potable water
to people in developing communities without adequate drinking water.
Economic opportunities are created as a result of the need for fabrication
of these concrete filters in these areas. Concrete is a recyclable material
providing for both environmental and economical benefits. Infrastructure
is already in place around the world to produce concrete, thus little or no
additional effort is needed to produce pervious concrete filters. New
construction is not required in the production of these filters resulting in no
significant increase in energy consumption.
Preliminary results demonstrated that the filter was effective for filtering
bacteria; however, some other chemical pollutants may increase in
concentration, particularly pH. Additional testing is needed to evaluate
some of the unsuccessful factors discussed below.
The unsuccessful factors of this experiment were an unexplainable
increase in dissolved copper concentrations (from 2 to 9 mg/L) and a very
-20-


high pH (11 to 12). The high pH levels in the test samples are likely the
result of large concentrations of hydroxide compounds in portland cement,
making the water non-potable. The high pH of the water could potentially
be offset by the fact that polluted natural waters and industrial waste water
are typically acidic, with a pH level less than 6 [Grippo, R.S., 2006,
Lenntech, 1998]. Supplemental testing will determine the ability of the
pervious concrete filter to neutralize acidic waters.
The initial dissolved copper concentrations were designed to be 0.13, 1.3,
13 and 130 mg/L; however, analytical results determined that 13 mg/L of
copper was found in all four samples. In addition, dissolved copper
concentrations in the samples indicate an increase in copper passing
through the filter. Further testing should be conducted to evaluate copper
concentrations and removal using this type of filter.
Due to the omnipresent nature of concrete and early indications of a wide
variety of filtration capabilities, the potential impacts are broadly applicable
and transferable to various industrial sectors. Examples include pre-
treatment for desalinization plants to prolong the life of nanopore filters,
local authorities delivering pervious concrete filters to flood stricken areas,
and in-line treatment of acidic industrial waste.
-21 -


The impact of the project is cleaner water, which has an enormous impact
on the quality of life for all life on this planet. This study represented Phase
I of the P3 competition and focused on the application of an original and
innovative idea applied to an existing product, resulting in a new use for
pervious concrete.
The results of the P3 experiment were applied to the proposed thesis
research, this time with the focus of the research being on AMD
remediation. The focus of this research was on the removal of inorganic
pollutants, pH modification in the presence of low pH water, enhanced
oxidation of water and pre and post test compressive strength testing used
to determine the rate of decay of structural strength of the pervious
concrete filter.
-22-


Chapter 3
Literature Review
3.1 Purpose, Research Objectives and Scope of Work
3.1.1 Purpose
Many independent studies have been performed on various Acid Mine
Drainage (AMD) remediation techniques and technologies. From a top-
down viewpoint, AMD from hard rock mines (defined as non-coal, metal
mines) [EPA, 2007] can be remediated according to a standard list of
methods that can be applied to handle a list of common issues such as
sulfates, metals, oxygen and sulfate reducing bacteria [Wildeman, 1999],
The following research will examine a potentially new technology to
extend the life span of passive and perhaps active remediation systems,
enhance pH modification of AMD as well as provide a mechanism to
recover commercial metals from abandoned mines.
3.1.1.1 The Formation of Acid Mine Drainage
Once a mine is dug, the conditions for AMD have already been created.
Geological formations isolated from the outside environment are exposed
to oxygen and water, physical and chemical processes for removing the
rock, removing desired minerals from the rock and the disposal of waste
-23 -


rock enhance the mixing of oxygen, water, minerals and bacteria that
might be present in the rock. After a mine has been abandoned or closed,
water will continue to enter the mine from both surface and subsurface
sources and oxygen will be introduced to this mixture at some point inside
and outside of the mine.
According to the Bureau of Land Management and the Mineral Policy
Center, between 80,000 and 570,000 abandoned mines exist in the US
with most existing in the western US [EPA, 2007],
AMD discoloration is readily visible, starting out as iron II sulfide (aka
pyrite or fools gold) (Eq.1) and becoming iron III ions (Eq.2) which then
become iron III hydroxide (aka yellow boy) and providing a color
transition from yellow to orange/red as the iron III hydroxide precipitates
out in the presence of highly acidic waters (Eq.3)
4FeS2(S) + 1402(g) + 4H20(i) > 4Fe2+(aq) + 8S042'(aq) + 8H+(aq) (Eq. 1)
4Fe2+(aq) + 02(g) + 4H+(aq) > 4Fe3+(aq) + 2H20(i) (Eq. 2)
4Fe3+(aq) + 12 H20(i) -> 4Fe(OH)3(s) + 12H+(aq) (Eq. 3)
Summation: 4FeS2(S) + 1502(g) + 14H20(i) > 4Fe(OH)3(S) + 8S042'(aq)
+ 16H+(aq) [Wildeman, 1999]
-24-


These equations only account for iron and do not account for other metals
commonly found in mines such as copper, zinc, cadmium, lead and
arsenic sulfides [Wildeman, 1999], These metals will undergo similar
oxidation. Additionally, according to Christine Costello [EPA, 2007], metal
sufides present in the surrounding rock can contribute to the generation of
AMD and increase the concentrations of mineral pollution in the waste
stream.
Example reactions:
AI(+3) + 3H20 <-> AI(OH)3 + 3H(+)
Mn(+2) + 0.25 02(aq) + 2.5 H20 <-> Mn(OH) + 2H(+)
Sphalerite ZnS(s) + 202(aq) -> Zn(+2) + S04('2)
Galena PbS(s) + 202(aq) -> Pb(+2) + S04('2)
Millerite NiS(s) + 202(aq) -> Ni(+2) + S04(2)
Greenockite CdS(s) + 202(aq) -> Cd(+2) + S04('2)
Covellite CuS(s) + 202(aq) -> Cu(+2) + S04<2)
Chalcopyrite CuFeS2(s) + 402(aq) -> Cu(+2) + Fe(+2) +
so4<-2)
In addition, since these reactions take place in natural conditions, the
presence of acidophilic bacteria such as Thiobacillus ferroxidans speed up
the above mentioned chemical processes.
-25 -


3.1.1.2 Active versus Passive Remediation
Typically, all of the metals presented in the previous section can be
precipitated out of solution by increasing the pH, with some metals
requiring higher pH levels than others and under aerobic conditions to
form hydroxides and oxyhydroxides [Snoeyink, 1980, Sawyer, 2003].
However, the situation can be made complicated when two or more of
these metals are present in solution, as is the case with manganese and
iron (II) or in the presence of sulfates and anaerobic conditions [EPA,
2007], Low pH is typically handled through the addition of alkaline
substances but because of the presence of sulfates may require the use
of Sulfate Reducing Bacteria (SRB) contained in bioreactors. The SRBs
reduce soluble sulfate compounds under anaerobic conditions to soluble
sulfide compounds which then combine with metal ions to form metal
sulfide precipitates:
S042' + 2CH20 > H2S + 2HC03 H2S + M+2 MS + 2H+ [EPA,
2007],
3.1.1.3 Active or Traditional Remediation
The EPA defines active remediation as ...the improvement of water
quality by methods which require ongoing inputs of artificial energy and/or
(bio) chemical reagents with the most predominate one is ODAS -
-26-


oxidation, dosing with alkali, and sedimentation [EPA, 2007], The process
is similar to that of traditional wastewater treatment plants. Traditional or
active treatments common to wastewater treatment plants include:
sulfidization, biosedimentation, sorption and ion exchange, and membrane
processes like filtration and reverse osmosis. The waters are removed
from their course, treated and then discharged [EPA, 2007], Active
remediation is often complex and expensive to install and operate, but
may be the only available option especially if the site to be remediated is a
superfund site. Active/Traditional remediation can also involve the use
of slopes to reduce flow velocity and revegetation to absorb waste after
the active treatment has taken place and before the treated water joins the
surface water.
3.1.1.4 Passive or Innovative Remediation
The EPA defines passive or innovative remediation to be those that treat
waters or solids using enhanced natural processes, in-situ and require
minimal upkeep, which includes encapsulation of entire sites [EPA, 2007],
Over the past 30 years, research has shown that not plants, but other
biochemical and geochemical processes were able to improve water
quality and that SRB played a key part in the removal of minerals present
-27-


but the addition of alkaline substances to adjust pH is itself not considered
innovative though the methods of application can be [EPA, 2007],
3.1.1.5 Passive Remediation Techniques
Anoxic Limestone Drains: This passive technique involves sending
polluted waters through crushed limestone for pH stabilization and to
create a precipitate of metal hydroxides and oxyhydroxides. The main
possible drawback to this technique is that the pores in the crushed
limestone can be clogged with organic material and metallic pollutants can
bind to the highly alkaline surface of the limestone, in effect physically
limiting contact between the limestone and the AMD [EPA, 2007],
Constructed Wetlands: There are two types of constructed wetlands:
aerobic and anaerobic/compost. The overall goal of constructed
wetlands is to create a condition of net alkalinity, allowing the additional
H+ ions released during hydrolysis of metal ions to be buffered: Fe+3
+2H20 -> FeOOH + 3H+. A possible drawback to this technique is that
precipitate can accumulate and limit wetlands effectiveness and thus the
precipitate must be removed [EPA, 2007],
Sulfate Reducing Bioreactors (SRBs): Typically, bioreactors are lined pits
that can contain a wide variety of organic matter and an alkaline agent. In
-28 -


some instances, these pits may contain a sulfate reducing bacteria in
place of the alkaline agent. Bioreactors can also take the form of tank
reactors, however, tank reactors are not discussed by the literature cited
for the purpose of this research [EPA, 2007], Bioreactors have the
advantage of using natural process provided by bacteria in aerobic and
anaerobic conditions to reduce minerals to insoluble forms and reduce
sulfate to sulfide using natural processes over a wide range of
temperatures (possibly as low as 4 C). Additionally, specialized bacteria
can be introduced to reduce metals such as selenium or uranium all with a
minimal maintenance. SRBs can be open pits, buried pits and possibly
placed inside of mines depending on remediation and security needs.
Example reactions [Gusek, 2002]:
-2 2 2 2
S04 + 2 CH20 = S' + 2 HCO3 + 2 H+ Zn+ + S' = ZnS
3AI3+ + K+ + 6H20 + 2S042 = KAI3(0H)6(S04)2(Alunite) + 6H+
Possible drawbacks to such a system include sustained temperatures
below 0 C, age of the wetland and periods of very low water flow
[Figueroa, et al, 2007],
-29-


Successive Alkalinity Producing Systems (SAPS): These systems have
three major system elements used to remove metal from AMD and modify
the pH of the waste stream: the drainage system, organic mulch layer and
a limestone layer within water tight basins. Pipes are utilized for proper
AMD containment and transport to the basins, and to ensure that organic
and limestone layers remain submerged. Within the organic mulch,
aerobic bacteria remove dissolved oxygen (DO) and anaerobic bacteria
generate alkalinity by reducing sulfates. The thickness of the organic
mulch layer must be sufficient to remove DO to concentrations below
1mg/L to prevent armoring of the limestone layer. Often, a valve flushed
drain pipe is also installed such that insoluble iron and aluminum gel (often
called floe) can be flushed out of the limestone layer before the limestone
pores are clogged. SAPS are often more efficient than wetlands in overall
effectiveness, but regular maintenance is required (flushing of floe) and
the flow path to the SAPS ponds must be highly regulated so that flow
does not enter the settling pond The grade of the landscape must also be
tightly controlled.
Permeable Reactive Barriers (PRBs): The name describes the system, a
permeable barrier that allows water to flow through it while material within
the barrier reacts with the AMD. Often these PRBs contain organic
material to foster the growth of sulfate reducing bacteria and possibly
-30-


alkaline material to elevate pH [EPA, 2007], It has been suggested that
SAPS and PRBs can be classified as an SRB, however, the main purpose
of these systems is to add alkaline material and not to utilize bacteria as
the main mechanism for remediation [Gusek, 2002], PRBs are flexible in
definition and can be utilized above or below ground. However, surface
PRBs often risk increasing DO which would free metals into solution from
their less soluble sulfide form. The use of PRBs in uranium mines or
where AMD has high concentrations of uranium may require that special
materials such as bone char phosphate, zero-valance iron pellets or
foamed iron oxide pellets can be added to the PRB, system to stimulate
reduction and removal of the uranium waste from the AMD.
Biosolids and Phytoremediation: These two methods are used to
remediate tailings piles, sludge and other types of contaminated solids
and are not applicable to this research [EPA, 2002],
3.1.2 Research Objectives
3.1.2.1 Pervious Concrete as a Permeable Reactive Barrier
Pervious concrete is beginning to see extensive use in infrastructure
projects from parking lots to green roofs as a way to improve the quality
-31 -


of runoff before it enters the ground [Haselbach, et. al. 2002; Park, et. al.
2006], Typically, pervious concretes, like conventional concrete consists
of portland cement, water, and aggregates. The increased porosity of this
material is achieved by eliminating the sand from the concrete mixture. By
reducing the amount of sand in the mixture, air voids are created in the
concrete allowing water to pass through the concrete. Pervious concrete
has approximately a 15-25% void structure allowing between 3 to 8
gallons of water per minute to pass through one square foot section of
concrete [NRMCA, 2005], Pervious concrete may also be recycled at the
end of the designed life [Majersky, et. al. 2007],
As was mentioned in Section 3.1.1.5, a Permeable Reactive Barrier is one
that facilities a chemical reaction with AMD by elevating pH to facilitate the
formation of insoluble metal hydroxides and sulfides. Beyond the
presence of sulfate reducing bacteria, the presence of alkaline materials
are needed to drive forward the process of precipitating out metals from
AMD. The portland cement used in the making of both pervious and
conventional concrete is highly alkaline and when water passes through
pervious concrete, the resulting pH may be above 12 and decrease to
around 9 as the pervious concrete is continuously immersed in neutral pH
water up to period of 90 days [Park, et. al. 2006], This ability to raise the
-32-


pH of water by a factor of 2 or more plus the demonstrated ability of the
filter to remove up to 100% of iron as iron II nitrate and up to 78% of
sodium as sodium chloride when the average aggregate size of pervious
concrete is sieved down to 0.25 cm makes this an attractive material for
use in AMD remediation. This filter has also shown the ability to remove
or neutralize Micrococcus luteus bacteria concentrations by a factor of 9
[Majersky, et. al. 2007]. Micrococcus luteus has an average diameter of 2
micrometers versus Thiobacillus ferrooxidans which has an average
diameter of 0.75 micrometers [Wildeman, 1999].
With these previous studies into the filtration and chemical properties of
pervious concrete taken into consideration, it is conceivable that pervious
concrete could perform the function of a PRB.
The purpose of the P3 research was to introduce a simple and existing
technology that may be able to remove sufficient quantities of bacteria
from water in rural areas to enhance the health and quality of life for rural
populations around the world. In addition, the P3 research has shown that
a globally omnipresent and recyclable material in concrete has the
potential to be used to remove single celled organisms from water. The
author has chosen pervious concrete as the material used in this thesis.
-33-


The constituent materials for concrete exist all over the world. This thesis
takes advantage of the advances in pervious concrete technology and
applications to develop a technique for providing a readily available
material to produce cleaner water.
3.1.2,2 Material Descriptioh of Pervious Concrete
Pervious concrete is an innovative construction material that offers
numerous economical and environmental benefits. Pervious concretes,
like conventional concrete consists of Portland cement, water, and
aggregates. The increased porosity of this material is achieved by
eliminating the sand from the concrete mixture. By reducing the amount
of sand in the mixture, air voids are created in the concrete allowing water
to pass through the concrete. Pervious concrete has approximately a 15-
25% void structure allowing between 3 to 8 gallons of water per minute to
pass through one square foot section of concrete [NRMCA, 2005],
Pervious concrete may also be recycled at the end of the designed life.
The porous nature of pervious concrete when compared to conventional
concrete is shown in Figure 3a and 3b.
-34-


Figure 3.1a and 3.2b Pervious vs. Impervious Concrete Surfaces
This type of concrete has been applied in parking lots and green roofing
projects to improve the quality of rain water before it enters urban
drainage systems [Haselbach, et. al.] and to potentially decrease
suspended sediments in flowing bodies of water [Park, et. al.]. Pervious
concrete is a hardened mixture comprised of water, cement, and gravel.
The mixture has little to no sand. With the absence of sand, voids are
present to allow water to flow through the concrete structure. The
objective of this thesis is to design the shape, size, and effective gravel
size to successfully filter contaminates from AMD.
The author chose concrete due to its wide use and availability throughout
the world and has been used by mankind for construction for thousands of
years. Global expertise in working with this material is present and
sufficient supply and production infrastructure are already in place. No
-35-


additional energy is required to create systems to produce this traditional
material already used worldwide. As a result, no additional pollution is
entered into the environment.
3.1.3 Scope of Work
3.1.3.1 Economical Recovery of Metals In The Waste Stream
With the industrial rise of the BRICs (Brazil, Russia, India and China), the
metals industry has sought faster, more economical ways to bring metals
to market. Recycling scrap metal from junkyards has been a common
practice in the US and the pace of this recycling has quickened, mostly
due to Chinas and Indias expanding economies but also due to overall
global demand for ore.
Another source identified as having potentially significant recycling value
is AMD. Recovering industrial and precious metals from AMD not only
provides another source of non-mined raw material for the metals industry;
the income can offset the cost of AMD remediation. One example cited by
the Pennsylvania Department of Environmental Protection (DEP) states
the annual cost to state taxpayers for AMD remediation to be $23 million
dollars a year and the estimated state wide value of sludge from these
-36-


systems to contain millions of dollars in metals, yet it is handled as waste
[Rathbun, 2004],
Since 1991, some countries like Canada began programs to study the
feasibility of recovering metals from the slurry of large scale AMD
remediation projects involving such diverse recovery technologies as
chemical treatment, solvent extraction, ion exchange, bioadsorption and
electrowinning (MEND, 2007], Other applicable technologies include
roasting, air drying and wet gravity separation. These processes increase
the cost of extracting the metals. [Gusek, et. al., 2005],
-37-


Chapter 4
Problem Statement
The formation of AMD begins when ground water moves through the soil
into a mine and creates a magnification effect such that the increase in
acidity of the water, the more metals it dissolves from surrounding rock.
This results in an increasingly beneficial environment for acidophilic
bacteria, which contributes to increasing the acidity of the groundwater.
This ground water flows out of the mine either through man-made
openings or through existing crevices in the earths surface.
Typically, AMD water is highly acidic and contains a high concentration of
dissolved metals. The result is the water is often inhospitable to aquatic
life and during periods of increased hydrologic flow; a greater volume of
AMD may be produced. Often AMD water enters existing bodies of water
and can destabilize the environment with the scale of the damage
dependent on the dissolved metals concentration, the types of metals, the
volume of AMD entering other bodies of water and the pH.
Traditional AMD remediation systems focus on treating the acidity and
capture the dissolved metals only to protect the limestone that is used to
-38-


balance the pH of the AMD. These systems typically cover substantial
areas of land resulting in limitations on where these systems are placed.
The size of the system needed to handle the AMD volume may also
require construction that alters the surrounding environment.
This thesis will evaluate pervious concrete as an AMD remediation system
focusing primarily on dissolved metals removal instead of the traditional
approach of focusing on raising pH. The performance of this type of filter
was studied and included herein. Additionally, the conceptual use of this
filter would be to remove the dissolved metals first, with pH adjustment a
secondary goal. This is due to the belief that the dissolved metals in AMD
are more toxic to local biology than the low pH. An added incentive is that
recovering the dissolved metals could give this filter the potential to
provide considerable economic benefit to implementing parties as well as
create more sustainable mining industry practices.
-39-


Chapter 5
Experimental Design and Procedure
5.1 Experimental Design
Experts in the field of Acid Mine Drainage (AMD) have performed
substantial research that demonstrates the various forms of passive
remediation. Past research has been conducted to achieve the desired
result of transforming AMD into water quality that approaches the state of
natural waters with minimal impact on the local environment, minimal
expenditure, and maintenance. As cited in the literature review,
government and private organizations in the United States and
international have begun investigating ways to recover metals that are
removed from AMD waste stream. The recovery of these metals offers
significant potential financial returns as well as the promotion of a
sustainable mining industry.
5.2 Experimental Procedure
The primary objective of this research is to investigate the use of pervious
concrete as a method of removing metals from AMD. In addition, the high
pH levels of the AMD will be neutralized as a result of this study. These
objectives will be accomplished in the following tasks:
-40-


Task 1: Perform an extensive literature review on previously developed
AMD remediation efforts, AMD studies, and pervious concrete applications
in the US and internationally. The findings of this review will provide for a
better understanding of the contaminants found in AMD, AMD remediation
efforts, and additional information beneficial to this study.
Task 2: Develop and test a laboratory model to examine the effectiveness
of pervious concrete in removing metals and neutralizing pH levels of
simulated AMD water. The filter dimensions were reduced from the
dimensions 10 inches by 10 inches by 18 inches applied to the P3
competition based on the low flow calculated in Task 3. This was
performed because much of the surface area would not be needed and to
increase the level of personal safety when handling the filter.
The pH of the Silver Cycle Mine was measured at between 5 and 5.5 by
the Colorado School of Mines [Figueroa, et al 2007]. The water used for
this experiment was distilled water with an assumed pH of 7. After
calibration of the pH meter (pH during calibration was 7.11 and 7.12 for
trials 1 and 2, respectively), the measured pH of the distilled water
averaged 6.9. The amount of 12.1 Normal (for monoprotic acids, 1 N = 1
Molar (M)) hydrochloric acid (HCI) needed to reduce the pH from 7 to 5
-41 -


was calculated by subtracting 10A-7 mols H+/L from 10A-5 mols H+/L,
multiply by 38 liters and multiply by 1 L/12.1 M. The volume of HCI needed
to adjust the pH was 0.32 milliliters (mL). This volume was added before
the metal sulfate salts were weighed and added. The influent pH for trial 1
was 2.6 and the influent pH for trial 2 was 2.8.
The given concentrations of sulfate (S04), iron (as Fe2') and zinc (Zn2+)
did not include the anions and cations they need to be bound to exist in
solid form. Sodium sulfate (NaS04) was chosen as the source of sulfate
to continue evaluating the filters ability to remove sodium from water.
Zinc sulfate and iron II sulfate (ZnS04 and FeS04) were the sources of
zinc and iron. The mass of each salt added to the experimental water was
determined by dividing the mass of each species in the Silver Cycle Mine
by the mass of each metal sulfate [Figueroa, et. al., 2007], For example,
mass of zinc (65 g/mol) / mass of zinc sulfate (161 g/mol) = 0.4, divide the
Silver Cycle mine zinc concentration (75 mg/L) by 0.4 = 187.5 mg/L =
0.188 g/L 38L = 7.1 g. Thus, 7.1g of Zinc Sulfate was needed for this
experiment. The calculated masses of the required salts were weighed on
a scale one at a time and poured into the AMD solution after the HCI was
added. The influent concentrations for all species were as follows:
-42-


24 mg/L of iron II sulfate, 329 mg/L of sodium sulfate and 75 mg/L of zinc
sulfate.
The figures in the following pages illustrate components used in the
experiment. Figure 5.1 illustrates the plastic sample bottles, from left to
right, used to collect samples for anion, total metals and dissolved metals
analysis. Figure 5.2 illustrates some of the 22 gallons of distilled water
used to prepare the simulated AMD water and provide baseline
concentrations of anion, iron, zinc and sodium present in the water and in
the filtrate as well as provide rinse water between trials 1 and 2. Figures
5.3 and 5.4 illustrate the containers of metal salts used to prepare the
simulated AMD water. The filters were also washed with one gallon of
distilled water prior to testing with AMD water. The wash filtrate was
then collected and analyzed for total metals. Samples for hours 1-5
were analyzed for total dissolved metals, total metals and anions (as
sulfate). The total required sample volume was 625 ml, with 125 ml
collected for anion analysis and 250 ml for both total dissolved and total
metals analysis. These volumes are required per the sample standards
provided by Evergreen Laboratories and also allow sufficient volume for
retesting.
-43 -


Figure 5.2. Distilled Water
44


Figure 5.3. Sodium Sulfate
Figure 5.4. Iron and Zinc Sulfate
-45 -


Task 3: Simulate the flow rate at the Silver Cycle Mine entrance using the
laboratory model consisting of a valve of known diameter. A >2 inch PVC
ball valve was used for this purpose. The width of one section of adit (the
surface channel through which the AMD flows) was given at 1 foot and the
maximum flow was 10 gallons per minute as measured by a stop watch
and bucket test.. At the 1 foot width section, a depth of 1 foot was
assumed. After comparing the ratio of adit and the fully open valve
velocities (1.3 ft/min and 231 ft/min) and applying the ratio (231/1.3=178)
then dividing 1.3 E-3 ftA2 (the area of the ball valve) by 178 and converting
to square inches, it was determined that a valve area of 1.05 E-3 inA2
would be needed.
A flow rate of 0.013 gpm (77.4 min/gal) and a total time to empty the ten
gallon tank being 774 minutes (12.9 hours) are unrealistic in the provided
laboratory conditions. Human error and the manually operated PVC valve
make achieving the level of valve area accuracy not possible. It was
decided to attempt to reduce the flow to a manually achievable 38.7
min/gal. The valve diameter used to achieve this was found by trial and
error method using a stopwatch and repeatedly filling the ten gallon tank
with tap water. The hydrologic behavior of the AMD water in the filter is
-46-


illustrated in 5.5 through 5.7 a, b and c. This behavior is similar to the
hydrologic behavior observed in the P3 experiment.
Figure 5.5. Experimental Flow Rate
-47-


Figure 5.6. Approaching Maximum Flow Rate Without Splashing
The remaining filtrate was allowed to drain into a sink connected to the
public waste water system, with the sink faucet being allowed to run fully
open to dilute the concentration of metals and modify the pH of the filtrate.
This was possible since the metals present in the AMD solution
represented EPA secondary drinking water contaminants and with a fully
open sink faucet, it was believed that the concentrations should be
sufficiently diluted as to bring the concentrations below EPA secondary
contaminant regulations.
-48-


(b) (c)
Figure 5.7. AMD Filtration Testing In Progress
Task 4: Test the compressive strength of two pervious concrete filters,
with one of the filters not subjected to the synthetic AMD and one used
during the AMD filtration portion of the study. This was accomplished by
coring a 3 inch diameter cylinder from the pervious concrete samples.
Cores were extracted at 4.5 inches from the top and 4.5 inches bottom of
both filters. A concrete saw was used to cut the filters just below the core
samples taken near the top of each filter. A third core sample was then
taken 9 inches from the bottom of each filter. This provided samples near
-49-


the top, middle, and bottom of each filter. Each core sample was marked
top, middle and bottom, with the top and bottom of each core sample
made flush using the concrete saw.
The cores were measured for length and diameter then tested in a
compressive strength testing device. Correction factors were applied to
calculate the pounds per square inch (psi) strength of each core sample
that did not meet minimum length to diameter ratios per ASTM C39.
Task 5: The final tasks included analyzing the results and provide
recommendations for future research activities.
5.3 Experiment Equipment
The experiment represents an attempt to faithfully recreate the conditions
of the Silver Cycle Mine adit in Idaho Springs, CO. The entrance to the
mine is located on private property and is not accessible per the owners
request made to Dr. Linda Figueroa at the Colorado School of Mines, who
was conducting AMD research on this mine at the time of the owners
request. Additionally, laboratory conditions are required to continue
studying the potential capabilities of the pervious concrete water filter, with
-50-


the initial research results submitted to the EPA as part of its 2006-2007
P3 competition. The experiment is designed to incorporate the pH and
chemical composition of the Silver Cycle Mine while linearly reducing the
ratio of mine flow to AMD channel width to a laboratory scale model
incorporating a water tank and PVC piping. The following equipment and
procedures were used to recreate the AMD conditions as well as collect
and analyze the sample waters.
A 38 Liter (10 US gallon) water tank with a PVC pipe and 1/4 PVC inch tap
as well as two plastic wrap covered 2x4 wood beams were used to
simulate the AMD flow from the mine. The water tank held the synthetic
AMD solution, the PVC pipe and tap were used to provide a non-reactive
transport mechanism to bring the fluid to the filter with flow control. The
wood beams were used to position the filter over the sink and provide
sufficient working space for samples to be collected. The wax paper was
used to keep the AMD solution from seeping through the wood and into
the sample containers, as the wood may act as another layer of filtration.
Table 5.1 lists the concentrations of AMD contaminants that are
represented in the Silver Cycle Mine taken during a single sampling event.
A fraction of the mass of each metallic salt that was added (the mass of
-51 -


sulfur is 30% of sulfate and 21% of sodium sulfate, the mass of zinc is
40% of zinc sulfate, the mass of iron is 37% of iron sulfate (as iron II
sulfate)).
Table 5.1 Silver Cycle mine AMD composition
Species Concentration (mg/L)
Sulfate as Sulfur 700
Zinc 65-75
Iron 40
Aluminum <1.0
pH 5.0-5.5
Three pervious concrete filters were fabricated in the laboratory. One was
used for pre-filtration compression strength testing; the other two were
used for a side by side filtration performance comparison with one of those
two filters used for post-filtration compression testing.
A sufficient supply of sample containers, 34 in total, as defined by EPA
standard methods for anions (sulfate), zinc, iron, and sodium were used in
collecting post-filtration water samples. The sulfate, zinc and iron
represented the primary constituents of the Silver Cycle AMD water and
the sodium was utilized to reinforce past test results demonstrating the
ability of the filter to partially desalinate water.
-52-


Enough containers were used to provide a 5 point data curve in an
attempt to find an outlying anomaly. There is no minimal number of
samples needed to meet EPA standards. In addition, there were two
samples of water blanks and two samples of filtrate blanks to provide a
baseline of the distilled water mineral content and filtrate mineral content.
Tap water was left flowing to dilute the volume of filtrate not used for
sampling. This was performed to dilute the filtrate to below EPA
secondary water quality standards.
This experiment attempted to mimic both the chemical composition and
flow rate of the Silver Cycle Mine under laboratory conditions. The Silver
Cycle Mine is located in the town of Idaho Springs, Colorado, near
Colorado Boulevard. It is one of many old mines in the area and receives
ground water infiltration, which contributes to the formation of AMD. This
mine was chosen for this research because of previous AMD research
conducted on this mine by the Colorado School of Mines. This university
was unable to continue their research into possible remediation methods
due to the mine adit being located on private property and the owner not
wanting to have their property disturbed. The study performed by the
Colorado School of Mines included the National Tunnel mine, however the
-53-


Silver Cycle Mine was chosen due to its higher concentrations of metals
and lower pH in order to test the effectiveness of the pervious concrete
filter under the most stressful conditions.
The proposed experiment involved filling one 38 L (10 US gallon) tank with
synthetic AMD and letting the contaminated water pass out of the tank,
through the filter and into the laboratory sink at a flow rate of 0.013 gpm
through a 0.5 inch dia. PVC ball valve (0.001 sq. ft). This was based on
the maximum flow rate of 10 gpm through a 1 sq. ft section of the AMD
stream exiting Silver Cycle adit. Before each trial period, the ten gallon
tank was rinsed with hot tap water and room temperature distilled water to
remove contaminants. Neither soap nor physical wiping was utilized to
prevent chemical contamination that might alter the pH or mineral content
of the influent water. Refer to Figure 5.8.
-54-


<-
5x5x1 8 inch pervious
concrete filter
Figure 5.8. Laboratory Model of AMD Test Configuration
Measurements for pH and temperature were taken before and after the
water passed through the filter using an Oakton Acorn series pH and
temperature meter. Each post-filtration measurement was performed on
the water in the sample bottles. Measuring the concentration of dissolved
oxygen (DO) in the filtrate was not possible due to the lack of a readily
available DO meter. Post filtration samples were collected and sent to
Evergreen Laboratories for chemical analysis due to the guaranteed
turnaround time for the results of the analysis.
-55-


Figure 5.10. AMD Test Apparatus Without the Filter
-56-


Figure 5.11. Complete AMD Test Apparatus, Side View
Figure 5.12. Complete AMD Test Apparatus, Front View
-57-


Chapter 6
Experimental Results
6.1 Introduction
Three filters were fabricated for this experiment. One was tested for pre-
filtration compression strength testing and the other two were used for a
side by side comparison of their AMD simulated water filtration capabilities.
Two filters were used during the filtration portion of the study in order to
establish a consistent performance between two simultaneously fabricated
filters. The filtration tests for each filter took place over an approximately
6.5 hour period and 5 samples were taken every hour starting 30 minutes
after the test was initiated.
After allowing the filters to dry, one of the filters used during the AMD
filtration was selected for post filtration compression strength testing. The
results of the compression strength tests were used to estimate the
structural decay of the filter over time.
The filtration aspect of the experiment focused on determining the
percentage removal of iron, sodium and zinc anions via trickle filtration in
the pervious concrete media and the percentage of total and dissolved
-58-


cations remaining in the filtrate. Percent removal of sulfate anions was
also analyzed.
6.2 Overall Filter Performance
The distilled water samples showed undetectable results for all cations as
total metals. The filter blank samples for filter 1 had 0.130, 8.22 and 0.129
mg/L of iron, sodium and zinc, respectively. The filter blank samples for
filter 2 had 0.773, 49.3 and 0.160 mg/L of iron, sodium and zinc,
respectively. Erosion studies of the P3 filter, with a 75% larger cross
sectional area, show a significant reduction in the concentration of
magnesium after the initial sample was taken.
Filter 1 on average removed percent total iron, sodium, zinc and sulfate
concentrations of 85%, 61%, 74% and 63%, respectively while filter 2 on
average removed percent total iron, sodium, zinc and sulfate
concentrations of 71%, 34%, 70 and 29, respectively. Filter 1 also
removed 6% of sulfate present in the synthetic AMD water while filter 2
remove 29% of sulfate present in the experimental water. Tables 6.1 and
6.2 illustrate the data derived from laboratory analysis of the filtrate and
figures 6.1 and 6.2 illustrate the overall performance characteristics of
each filter. Initial concentrations of species in table 6.2 are in red due to
suspected contamination. This was first noticed in the laboratory results
-59-


when the 1 and 2 hour effluent samples were of a higher concentration
than the initial influent concentrations used for both trials.
To obtain the new initial concentrations, the linear slope of the percent
concentration in the effluent was calculated and subtracted from the first
hour results. The following is an example calculation:
rise/run = 9.5 therefore the 1 hour sample has 4.0% (13.5%-9.5%) S04 removal
and the 0 hour concentration is 1024.4 mg/L
Based on the P3 studies, the initial cation concentrations in the filter blank
water is not considered to contribute to the concentrations of cations in the
AMD filtrate and can be considered zero mg/L. This theory also applies
to the unexpectedly high concentration of sodium in the distilled water filter
blank sample of filter 2. The low pH water may have added some extra
sodium to the effluent concentrations, but this does not explain the
abnormal concentrations of sulfate. Contamination is still the most
plausible explanation for the abnormal concentrations of both sodium and
sulfate present in the filtrate of filter 2.
-60-


Table 6.1. Filter 1 Results
Filter 1
Sample hour
Parameter Influent 1 hour 2 hour 3 hour 4 hour 5 hour Avg
PH 2.6 5.9 6.0 5.8 5.75 5.9 6
Temp. (C) 25 25 25 25 25 25 25
Sulfate mg/L 884 562 239 203 223 400 325
% S04 removed by filter 36.43 72.96 77.04 74.77 54.75 63
Std. Dev 17
Variance 302

Dissolved Metals
Total Fe mg/L 24.00 4.68 0.64 0.00 0.00 0.16 1
Sodium mg/L 329.00 196.00 84.40 81.20 84.00 159.00 121
Zinc mg/L 75.00 29.50 20.80 10.50' 10.80 18.90 18
% Diss. Fe cone. 19.50 2.68 0.00 0.00 0.65 5
% Diss. Na cone. 59.57 25.65 24.68 25.53 48.33 37
% Diss. Zn cone. 39.33 27.73 14.00 14.40 25.20 24
Std Dev. Fe 8
Std Dev. Na 16
Std Dev. Zn 11

Total Metals Influent 1 hour 2 hour 3 hour 4 hour 5 hour Avg
Total Fe mg/L 24.00 8.13 3.23 1.57 1.82 3.66 4
Sodium mg/L 329.00 227.00 84.70 82.40 87.40 168.00 130
Zinc mg/L 75.00 31.50 21.00 10.80 12.30 20.20 19
total % Fe cone. 33.88 13.46 6.54 7.58 15.25 15
total % Na cone. 69.00 25.74 25.05 26.57 51.06 39
total % Zn cone. 42.00 28.00 14.40 16.40 26.93 26
Std Dev. Fe 11
Std Dev. Na 20
Std Dev. Zn 11
% Fe as dissolved 0.00 57.56 7.92 0.00 0.00 1.92 13
% Na as dissolved 0.00 86.34 37.18 35.77 37.00 70.04 53
% Zn as dissolved 0.00 93.65 66.03 33.33 34.29 60.00 57
% Fe removed by filter 0.00 66.13 86.54 93.46 92.42 84.75 85
% Na removed by filter 0.00 31.00 74.26 74.95 73.43 48.94 61
% Zn removed by filter 0.00 58.00 72.00 85.60 83.60 73.07 74
-61 -


Table 6.2. Filter 2 Results
Filter 2
Sample hour
Parameter Influent 1 hour 2 hour 3 hour 4 hour 5 hour Avg
PH 2.80 5.92 5.82 5.77 5.80 6.10 6
Temp. (C) 25.00 25.00 25.00 25.00 25.00 25.00 25
Sulfate mg/L 1024.00 985.00 852.00 724.00 579.00 476.00 658
% S04 removed bv filter 3.81 16.80 29.30 43.46 53.52 29
Std Dev. 20
Variance 398
Sum of Std Dev 37

Dissolved Metals
Total Fe mg/L 35.23 11.70 9.01 9.25 8.34 3.94 8
Sodium mg/L 440.50 437.00 382.00 310.00 252.00 176.00 311
Zinc mg/L 110.35 38.90 38.80 36.80 36.60 33.40 37
% Diss. Fe cone. 33.21 25.57 26.26 23.67 11.18 24
% Diss. Na cone. 99.21 86.72 70.37 57.21 39.95 71
% Diss. Zn cone. 35.25 35.16 33.35 33.17 30.27 33
Std Dev. Fe 8
Std Dev. Na 23
Std Dev. Zn 2

Total Metals Influent 1 hour 2 hour 3 hour 4 hour 5 hour Avg
Total Fe mg/L 37.55 14.50 12.20 10.90 9.43 6.53 11
Sodium mg/L 494.35 449.00 396.00 317.00 254.00 221.00 327
Zinc mg/L 125.66 39.30 39.80 37.90 37.50 36.50 38
total % Fe cone. 38.62 32.49 29.03 25.11 17.39 29
total % Na cone. 90.83 80.11 64.12 51.38 44.71 66
total % Zn cone. 31.27 31.67 30.16 29.84 29.05 30
Std Dev. Fe 8
Std Dev. Na 19
Std Dev. Zn 1
% Fe as dissolved 0.00 80.69 62.14 63.79 57.52 27.17 58
% Na as dissolved 0.00 97.33 85.08 69.04 56.12 39.20 69
% Zn as dissolved 0.00 98.98 98.73 93.64 93.13 84.99 94
% Fe removed bv filter 0.00 61.38 67.51 70.97 74.89 82.61 71
% Na removed by filter 0.00 9.17 19.89 35.88 48.62 55.29 34
% Zn removed by filter 0.00 68.73 68.33 69.84 70.16 70.95 70
Figure 6.1 shows an overall decrease in species concentration; however,
the last sample taken during the filtration test shows a slight increase in
species concentration. This pattern is similar to the results of the P3 filter
and is most likely due to an accumulation of a critical concentration of
chemical species and fluid that results in a flush out of those species.
-62-


Filter 1 Percent Total Species
Concentrations Present in Filtrate
Sulfate
Iron
a Sodium
X Zinc
Sample hour
Figure 6.1 Filtrate Concentrations versus Time
It was expected to see a similar pattern for filter 2. The percent
concentration versus time for filter 2 is shown in Figure 6.2. Instead, a
shallower slope was observed with a slight increase in percent
concentration in the last sample hour, which was consistent with both the
P3 filter and filter 1. Though the cross sectional dimensions between the
P3 filter and these filters were different by 75%, the length of the filters
remained the same at 18 inches. A possible reason for this is due to the
presence of residual species from the test of filter 1. As mentioned in
Chapter 5, only hot tap water and room temperature distilled water were
used to clean the ten gallon tank and PVC piping. No soap was used to
prevent chemical contamination.
-63 -


Rlter 2 Percent Total Species
Concentrations Present in Filtrate
Sulfate
Iron
A Sodium
Zinc
Sample hour
Figure 6.2 Filtrate Concentrations versus Time
6.3 Filter 1 and 2 Performance Comparisons
Figures 6.3 to 6.6 illustrate the percent concentration of each species
retained by each filter based on the results of the total metals analysis of
the filtrate. This percentage was based on subtracting the total species
concentration (consisting of both dissolved and suspended particles
present in the filtrate) from the beginning concentration and dividing by the
beginning concentration.
The graph lines for both filters show slopes similar to the respective slopes
in Figures 6.1 and 6.2. For the purposes of performance comparison, the
difference in percent removal of iron and zinc were 14% and 4%,
-64-


respectively while the difference in percent removal of sodium and sulfate
was 27% and 34%, respectively. It should be noted that the percent
concentration curves in figures 6.1 to 6.6 and 6.9 to 6.10 show the
breakthrough time for filter 1 but not for filter 2, despite a greater initial
concentration of sulfates and sodium. Considering the similar dimensions
and portland cement content demonstrated by the similar pH adjustment
results, a possible difference between the two filters that would explain the
lack of similar breakthrough times would be increased void space in filter 2.
Percent Removal of Iron by Filters 1 and 2
O
M
(0
+4
c
O
c
o
o
a?
Sample hour
100.00
1 hour 2 hour 3 hour 4 hour 5 hour
Figure 6.3. Percent Concentration of Iron in Filters 1 and 2
-65 -


Percent Removal of Sodium by Filters 1 and 2
0>
o
o
O
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
Filter 1
Filter 2
1 hour 2 hour 3 hour 4 hour 5 hour
Sample hour
Figure 6.4. Percent Concentration of Sodium in Filters 1 and 2
Percent Removal of Zinc by Filters 1 and 2
c
o
E
c
0)
o
c
o
O
90.00
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
Filter 1
Filter 2
1 hour 2 hour 3 hour 4 hour 5 hour
Sample hour
Figure 6.5. Percent Concentration of Zinc in Filters 1 and 2
-66-


Percentage of Sulfate
Retained by the Filters 1 and 2
90.00
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
1 hour 2 hour 3 hour 4 hour 5 hour
1st Filter
2nd Filter
Sample hour
Figure 6.6. Percent Concentration of Sulfate in Filters 1 and 2
1 hour 2 hour 3 hour 4 hour 5 hour
Q 1st Filter
2nd Filter
Sam pie hour
Figure 6.7. Percent pH modification by Filters 1 and 2
The consistent overlapping of confidence intervals in Figures 6.3, 6.4 and
6.7 suggest consistent performance between filters 1 and 2 in removing
-67-


iron and sodium as well as modifying the pH. By percentage, filters 1 and
2 show relatively similar performance in removing iron and zinc while
exhibiting a dissimilar performance in removing sodium and sulfate.
Also note that filter 1 consistently retained higher percentages of each
species than filter 2. Whether this discrepancy is due to the higher
concentrations of species in the AMD solution used in the second trial or
inconsistencies in the internal structure of the filter is not known at this
time and should be further investigated. The very strong correlation in pH
modification between filters 1 and 2 suggests an equal concentration of
Portland cement in each filter, ruling this factor out as a potential variable
affecting the ability to remove various chemicals. The inverted curve of
the series of samples of both filters suggests hydraulic breakthrough
pushing accumulated higher pH water out of the pore space.
6.4 Conversion of Species from Dissolved to Suspended Particles
Another quantifiable aspect of the pervious concrete filters effectiveness
is what percentage of the species influent pollution can be converted from
a dissolved state to suspended particles through reactions with the
Portland cement. In Figure 6.9, there is an increase in the percent
concentration of dissolved species up to the 5th sample hour while in
Figure 6.10 there is no increase in the percent concentration in the 5th
-68-


sample hour. If both filters showed a similar pattern of increase of
dissolve species concentration, the results may suggest the flushing of
dissolved species within the filter after a certain time period, the
accumulation of hydrologic pressure within the filter or the onset the of
exhaustion of the concentration of portland cement within both filters.
Since there is not a similar pattern in each filter and the pH modification
suggests a similar concentration of portland cement in both filters, an
excess of sulfate within the filter combined with the hydrologic load may
be a contributing factor.
Filter 1 Percent Dissolved Species
Concentrations Present in Filtrate
o
o
Sample hour
Iron
A Sodium
Zinc
Figure 6.8. Filter 1 Dissolved Species Concentrations
-69-


Filter 2 Percent Dissolved Species Concentrations
Present in Filtrate
Iron
A Sodium
Zinc
1 hour 2 hour 3 hour 4 hour 5 hour
Sample hour
Figure 6.9. Filter 2 Dissolved Species Concentrations
Figure 6.11 and figure 6.12 illustrates a performance comparison between
AMD filter 1,2 and the P3 filter tested in 2007. Between filter 1 and the P3
filter, iron, sodium and zinc/copper had a comparable difference of 15%,
17% and 14%, respectively. Of note, despite AMD filter 1 having a 75%
smaller cross sectional area than the P3 filter, 14% more zinc was
removed than copper. Between filter 2 and the P3 filter, the percent
difference in removal of iron, sodium, and zinc/copper was 29%, 44% and
10%. For the purposes of comparing the physical properties of each
cation, zinc and copper are adjacent to each other on the periodic table,
with zinc having a larger atomic radius. Sodium has roughly 1/2 of the
electronegativity of the other metals and roughly 1/2 to 1/3 of the atomic
mass of the other metals in the study. Sodium and zinc posses one
electron in their outermost shell while iron and copper posses 2 electrons
-70-


in their outermost shell. Sodium, copper and zinc have Van der Waals
(intermolecular forces) radii of 227, 140 and 139 picometers (pm),
respectively and iron did not have a radii given. Calculated atomic radii for
sodium, iron, copper and zinc are 190, 156, 142 and 145 pm respectively.
The discussion of physical chemistry as it relates to the filters
performance is outside the scope of this study but may be useful in
customizing future filter designs to target the removal of specific metals,
including the addition of substances not native to pervious concrete.
Table 6.3. EPA P3 Cation Filtration Results
P3 cation filtration results
Sample Final Final Final Percent Removal Percent Removal Percent Removal
Cu(2+) Fe(2+) Na(1+) Cu(2+) Fe(2+) Na(1+)
1 2.98 0 10600 77.08 100.00 69.71
2 4.02 0 9790 69.08 100.00 72.03
3 6.57 0.542 6950 49.46 99.46 80.14
4 7.47 0 3840 42.54 100.00 89.03
Average 5.26 0.14 7795.00 59.54 99.86 77.73
-71 -


P3 vs. AMD Filter 1 Performance
c
o
c
0)
o
c
o
O
AMD Iron
P3 Iron
AMD Sodium
P3 Sodium
AMD Zinc
P3 Copper
Sample Number
Figure 6.10. P3 vs. AMD Filter 1 Performance Comparison
P3 vs. AMD Filter 2 Performance
c
o
c
0)
o
c
o
o
AMD Iron
P3 Iron
AMD Sodium
P3 Sodium
AMD Zinc
P3Copper
Sample Number
Figure 6.11 P3 vs. AMD Filter 2 Performance Comparison
-72-


6.5 Compressive Strength Test Results
Figure 6.11 illustrates the results of the compression strength testing
performed on cores taken from the pre-filtration filter and post-filtration
filter. There were outlying data in both tests and both occurred in the core
samples taken at the middle of the filter, this may be due to a reduction in
void space located in the middle of the filter. The standard deviation and
variation of each of the cores from each filter show a much wider variation
of compressive strength in the pre-filtration filter than in the post-filtration
filter. This may be due to the corrosive effects of the AMD solution acting
throughout the post-filtration filter. Figure 6.11 shows that the AMD
solution reduced the compressive strength of the top section of the post-
filtration filter by about 36%, where the AMD solution would have the
lowest pH and the smallest area of flow. The difference of compressive
strength in the middle is the smallest difference, 9%, and may be due to
reduced pore space. The difference in the compressive strength of 13%
between the cores from the bottom of each filter may be due to the
dispersion of AMD water as it reaches the bottom of the filter. The overall
reduction in compressive strength of the post-filtration versus pre-filtration
unit was about 28%. This is possibly a result of 6.5 hours of exposure to
an average of pH 2.7 AMD water. Based on the authors experience with
materials testing within the city of Aurora, Colorado, the compressive
-73 -


strength of this concrete typically may not meet the compressive strength
requirements of sidewalk or curb and gutter concrete. This low
compressive strength may be due to the intentional lowering of the cement
to aggregate ratio for the purpose of maximizing pore space. This filter is
not intended to support forces beyond its own weight and hydraulic forces.
Filter Compression Strength
Top Middle Bottom
Core location
Figure 6.12. Comparisons of Compressive Strength Test Results
Table 6.4 Percent Difference in Compression Strength Test Results
Pre-filtration (psi) Post-filtration (psi) % difference
Top 1979 1261 36
Middle 2101 1910 9
Bottom 1106 958 13
Average 1543 1110 28
Std. Dev. 543 486 15
Variance 294506 236552 214
-74-


Chapter 7
Conclusions and Recommendations
Filter 1 on average removed percent total iron, sodium, zinc and sulfate
concentrations of 85%, 61%, 74% and 63%, respectively while filter 2 on
average removed percent total iron, sodium, zinc and sulfate
concentrations of 71%, 34%, 69 and 29, respectively. Filter 1 also
removed 6% of sulfate present in the synthetic AMD water while filter 2
remove 29% of sulfate present in the experimental water. Filter 1s
percent removal of species are closer to the P3s average removed
percent total sodium, iron and copper concentrations of 55.54%, 99.86%
and 77.73%, respectively.
Based on the percent difference in filtration of species between AMD filter
1 and the P3 filter and considering AMD filters 1 and 2 had 75% less
surface area than the P3 filter, AMD filter 1 performed similarly in the
areas of iron, sodium and zinc removal but did not have similar
performance in the removal of sulfate. AMD filter 2 performed similarly to
the P3 filter in zinc removal however the percent differences in the
removal of iron, sodium and sulfate are much greater than AMD filter 1.
The potential rate of conversion of sulfate to sulfite is also not known at
this time.
-75 -


Comparisons of pH modification between the AMD test filters and the P3
filter has a difference of a pH of 1 and may be due to the highly acidic
water initially present in the AMD trials versus the neutral water used for
the P3 trial (the P3 filter increase pH from 7 to an average of 11.5 while
the AMD filters increase pH from approximately 2.7 to 6). It is also worth
noting that both filters used in this thesis reduced the iron concentration
below the EPA secondary drinking water standards and the AMD filters
approached the minimum pH standard. Sulfate concentration was
reduced by over 50% in filter 1, though not to levels approaching EPA
standards.
Regarding filter breakthrough both in terms of filter saturation by species
and hydraulic breakthrough, figure 6.7 shows an increase in the pH of the
filtrate in both samples. This may indicate that hydraulic breakthrough is
occurring because the pH adjustment occurs when the low pH AMD water
reacts with the high pH portland cement. The hydraulic breakthrough may
be occurring due to a certain volume of water being present in the filter,
which would increase the downward force on water at the lower end of the
filter and push it into the filtrate collection tray. The filter may continue
working past the 5 hour mark though the concentrations of species in the
-76-


filtrate would remain higher than in samples taken at earlier times. This
pattern may occur until either the portland cement content in the filter is
exhausted or the concentration of species in the filter reaches its own
breakthrough level. At this time a definitive time period cannot be given
without further study focusing on hydrological characteristics.
Further studies would be needed to determine the maximum influent rate
that a filter of specific dimensions could accommodate without reaching
hydraulic breakthrough, with cross sectional area and length potentially
being considered independent of each other. It should be mentioned that
one filter would never be considered sufficient in any AMD scenario. The
dimensions of the filter allow for ease of handling by manual labor or the
use of machinery that can be transported by standard off-road vehicles.
The recommendations for future study are as follows:
Repeat this experiment in an attempt to verify these results, using
one tank for each trial to avoid contamination and if possible, not
using plastic tanks with non-smooth surfaces.
Repeat this experiment with an emphasis on including copper as a
cation.
-77-


Conduct a study of the filters hydraulic properties to determine
breakthrough times based on a series of influent flow rates as well
as whether the breakthrough concentrations increase or level off to
a fairly constant concentration in the effluent over time.
Conduct a study of the filters ability to remove and precipitate high
value metals such as gold, silver, titanium, platinum, etc.
Conduct a study of the effectiveness of this filter in removing and
chemically neutralizing phosphate based fertilizers, mercury and
arsenic in low and neutral pH waters.
Filtrate analysis should also include a comparison of sulfate and
sulfide concentrations (or nitrate and nitrite) as the conversion of
sulfate to sulfite should aid in the precipitation of metals out of the
solution instead of being recombined into metal sulfates.
Comparisons of water with high sulfate concentrations to low
sulfate concentrations and the effects of sulfate attack on the filter
through compressive strength and sulfate testing should be
performed.
Analyzing the ability of the filter in removing the above mentioned
contaminants in high turbidity waters.
-78-


The effects of including calcium carbonate in combination with and
as a replacement for standard aggregate in removing all of the
above mentioned contaminants.
Establish the maximum influent rate that the filter could
accommodate before reaching hydraulic breakthrough using the
cross sectional areas and lengths of both the P3 and AMD filters as
initial baseline dimensions.
Test a series of filters using the maximum influent rate established
in the previous bullet point for both hydraulic loading and filtration
studies.
A field test of three filters of the same dimensions placed side by
side in an AMD adit of an abandoned mine during a period of
typically high surface flow (such as late spring in the US Rocky
Mountains). Pre-filtration and post-filtration water samples should
be taken. In addition, a single filter should be returned to the lab for
compression strength testing with the remaining filters moved into a
narrow section of the adit to ensure that maximum filtration of the
AMD stream is taking place.
-79-


References
EPA. (2003), Online publication,
http://clu-in.org/download/studentpapers/costello amd.pdf (June 6,
2007)
L. Figueroa, A. Miller, M. Zaluski, and D. Bless (2007). Evaluation of a
Two-Stage Passive Treatment Approach for Mining Influenced
Waters".
Gusek, James J. (2002). Sulfate-Reducing Bioreactor Design And
Operating Issues: Is This The Passive Treatment Technology For
Your Mine Drainage?. Presented at the 2002 National Association
of Abandoned Mine Land Programs Annual Conference,
September 15-18, 2002, in Park City, Utah.
Gusek, J., Clarke-Whistler, K. (2005). Where Does The Recover of Metal
Resources From Passive Treatment Systems Fit In Sustainable
Development Initiatives Associated With Large Mining Projects?
Haselbach, L.M. and. Freeman, R.M. (2006) Vertical Porosity
Distributions in Pervious Concrete Pavement. Materials Journal
Vol.103, No.6, Nov-Dec. 2006, pg.452-459
G. Majersky, S. Durham, A. Ramaswami. (2007). Sustainable Concrete
Filtration System for Developing Communities, PHASE I.
University of Colorado at Denver. Presented at the 2007 EPA P3
Conference, April 24-25, 2007 in Washington, DC
MEND Project. (2007). Study on Metals Recovery/Recycling from Acid
Mine Drainage. Online publication. Executive Summary, Natural
Resources Canada.
http://www.nrcan-rncan.qc.ca/mms/canmet-mtb/mmsl-
Imsm/mend/default e.htm. (November 19, 2007)
National Ready-Mixed Concrete Association, NRMCA, (2005). Pervious
Concrete: When it Rains, it Drains. National Ready-Mixed
Concrete Association, Silver Spring, Maryland.
-80-


Park, S., (2002). Chungnam National University, Daejeon, Republic of
Korea; Mang Tia, University of Florida. An experimental study on
the water-purification properties of porous concrete, USA; 19
November 2002
Rathbun, T. (2004). DEP Promotes Recovery of Metals from Acid Mine
Drainage as Economic Driver. Online posting. Commonwealth of
Pennsylvania Official News Release. November 17, 2004.
http://www.ahs.dep.state.pa.us/newsreleases/default.asp?ID=3200
(November 19, 2007)
Sawyer, C., McCarty, P., Parkin, G. Chemistry for Environmental
Engineering and Science. Singapore: McGraw-Hill, 2003, pp. 3-27,
114-190, 542-569, 659-663, 670-675.
Snoeyink, V., Jenkins, D. Water Chemistry. New York: John Wiley & Sons,
1980, pp 86-412.
Virginia Tech University (2005). Successive Alkalinity Producing Systems
for Renovation of Acid Mine Drainage.
http://www.cses.vt.edu/PRP/Research Results/SAPS.html (June 6,
2007)
Wildeman, T. (1999) Acid Mine Drainage Experiments Online posting
http://www.mines.edu/fs home/ihoran/ch126/index.htm (June 6,
2007)
- 81 -