AN EVALUATION OF FLUE GAS DESULFURIZATION GYPSUM FOR
ABANDONED MINE LAND RECLAMATION
THESIS
Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in
the Graduate School of The Ohio State University
By
Rachael A. Pasini, B.A
Civil Engineering Graduate Program
The Ohio State University
2009
Thesis Committee:
Professor Harold Walker, Advisor
Professor Linda Weavers
Dr. Tarunjit Butalia
ABSTRACT
The goal of this study is to understand the environmental impacts associated with using FGD gypsum for abandoned mine land reclamation (AMLR) in Ohio. There are over 200,000 acres of abandoned strip mines and over 600,000 acres of abandoned underground mines throughout Ohio that cause physical and chemical hazards to surrounding waterways, soil systems, and wildlife. Ten abandoned mine lands (AMLs) in eastern Ohio were reviewed in order to classify the different types of AMLs and build model scenarios. FGD gypsum samples from two different coal combustion power plants were tested in order to characterize the leaching behavior of the material under a variety of conditions. The USEPA Methods 1311 (TCLP) and 1312 (SPLP) were used alongside a three-tiered framework developed by Kosson et al. In order to assess the overall impacts of using FGD gypsum for AMLR, a Life Cycle Assessment (LCA) was conducted to calculate output emissions and energy consumption. A combination of conventional, ecological, and economic input-output LCA were used in this study.
The TCLP and SPLP results yielded concentrations of regulated constituents well below Ohio standards for beneficial use application. Thus, according to such standards the FGD gypsum samples can be classified as nonhazardous substances that present no significant impact on groundwater or risk to public health. The results from the three-
ii tiered Kosson et al. framework provided evidence that the leaching of Ca, S, and SO4 is not necessarily affected by pH, but is mainly dependent on the solubility of gypsum.
Some constituents, such as Mg and Mn, were dependent upon pH while other showed trends as a function of time. For instance, a majority of the available B was extracted in each experiment and concentration increased over 8 days, but did not change in response to varying pH. Less than 0.02% of Hg was available for leaching in each experiment and decreased over time, suggesting that long-term leaching of Hg is potentially not a concern. Overall, the leaching results demonstrate that the FGD gypsum samples do not present significant impacts on surrounding waters.
The LCA results showed that using FGD gypsum for AMLR is a better option than landfilling in terms of overall negative environmental impacts. In addition, it showed that there are only slight differences between reclaiming abandoned highwalls, highwall pits, and underground mine entranceways; though, due to high unit costs gob pile reclamation was shown to be the least favorable option. The results from this analysis do not include the positive impacts associated with reclamation, however, such as acid mine drainage abatement and reduced sedimentation runoff. Thus, stakeholders must weight the trade-offs between cost, benefits of reclamation, and overall negative impacts when developing AMLR plans.
iii ACKNOWLEDGEMENTS
First and foremost, I would like to thank my parents for always supporting my goals and encouraging my education. Next, I thank Dr. Harold Walker for providing the opportunity to study environmental engineering and gain valuable research and field experience. Thank you to my committee members, Dr. Tarunjit Butalia and Dr. Linda
Weavers, for supporting my research.
I would like to thank members of the environmental engineering lab group, especially Ruiyang Xiao, Yuan Gao, Mengling Li, Xuan Li, and Qing Le for their friendship and sharing their knowledge in the laboratory. I thank the process system engineering group, especially Bob Urban and Dr. Bhavik Bakshi, for assisting me and inviting me to learn about their research.
Thank you to Dr. John Lenhart, Dr. William Wolfe, Robert Baker (Baker
Consultants), Richard Warden (ODNR), and Bill Richardson (AEP) for providing instruction and information for my research. To Kevin Jewell at OARDC for performing many laboratory tests and answering my many questions.
Last, but not least, I thank my close friends and companions, especially Alex
Morison, that make me laugh and encourage me to grow.
iv VITA
March 1984……………………………… Born
June 2002………………………………… Magnificat High School
June 2007………………………………… B.S. Industrial and Systems Engineering,
The Ohio State University
June 2008 to present……………………… Graduate Research Associate, Department of
Civil and Environmental Engineering and
Geodetic Sciences, The Ohio State
University
Field of Study
Major Field: Civil Engineering
Environmental Engineering
v TABLE OF CONTENTS
Page
LIST OF TABLES ...... xi
LIST OF FIGURES ...... xiii
LIST OF ABBREVIATIONS...... xv
CHAPTER 1: INTRODUCTION ...... 1
1.1 Scope of Work...... 1
1.2 Changes in FGD Material Production ...... 2
1.3 Abandoned Mine Lands in Ohio...... 3
1.4 Leaching Experiments ...... 6
1.5 Life Cycle Assessment ...... 8
1.6 Outline of Chapters...... 9
CHAPTER 2: ABANDONED MINE LANDS AND FGD MATERIAL UTILIZATION
IN OHIO ...... 11
2.1 History of Mining and Legislation in Ohio ...... 11
2.2 History of FGD Material Production in Ohio...... 15
2.3 FGD Material Usage at Mine Sites...... 22
2.3.1 Otsego Gob Pile ...... 24 vi 2.3.2 Fleming Surface Mine...... 24
2.3.3 Freeport Gob Pile ...... 25
2.3.4 Broken Aro Remining ...... 26
2.3.5 Roberts-Dawson Injection Project ...... 26
2.3.6 Rehoboth Refuse Pile ...... 27
2.3.7 Rock Run Gob Pile...... 28
2.3.8 Central Ohio Coal Company Coal Refuse Impoundment ...... 28
2.3.9 Conseville Prep Plant Refuse Pile...... 29
2.3.10 Conesville Highwall...... 29
2.3.11 Summary of AML Sites...... 30
CHAPTER 3: CHEMICAL AND PHYSICAL PROPERTIES OF FGD GYSPUM ...... 32
3.1 Chemical Properties...... 32
3.1.2 Mercury...... 35
3.1.3 Boron...... 36
3.1.1 Elemental Analysis...... 37
3.2 Physical and Mechanical Properties ...... 40
3.3 Overview of Leaching Experiments ...... 41
3.4 Methodology ...... 41
3.4.1 Toxicity Characterization Leaching Procedure ...... 41
vii 3.4.2 Synthetic Precipitation Leaching Procedure ...... 42
3.4.3 Integrated Framework ...... 43
3.4.3.1 Tier 1 ...... 43
3.4.3.2 Tier 2 ...... 43
3.4.3.3 Tier 3 ...... 45
3.4.4 Chemical analysis ...... 47
3.4.5 Visual MINTEQ ...... 47
3.5 Results and Discussion ...... 47
3.5.1 TCLP and SPLP ...... 48
3.5.1.1 pH...... 50
3.5.1.2 Geochemical speciation ...... 51
3.5.1.3 Calcium and sulfur ...... 52
3.5.1.4 Mercury...... 53
3.5.1.5 Boron...... 53
3.5.2 Kossen et al. Integrated Framework Results...... 57
3.5.2.1 Tier 1 Results ...... 57
3.5.2.2 Tier 2 ...... 61
3.5.3.3 Tier 3 ...... 67
3.6 Conclusion ...... 73 viii CHAPTER 4: LIFE CYCLE ASSESSMENT OVERVIEW AND METHODOLOGY ....75
4.1 Overview of Life Cycle Assessment ...... 75
4.1.1 Types of LCAs ...... 76
4.1.1.1 Conventional LCA ...... 77
4.1.1.2 Economic Input-Output LCA ...... 77
4.1.1.3 Ecological LCA ...... 78
4.1.2 Limitations of LCAs ...... 78
4.2 Life Cycle Assessment Methodology...... 79
4.2.1 Goal Definition and Scope ...... 79
4.2.1.1 Goal definition ...... 79
4.2.1.2 Level of specificity ...... 80
4.2.1.3 Functional unit ...... 80
4.2.1.4 Scope and system boundaries ...... 81
4.2.2 Life Cycle Inventory...... 85
4.2.2.1 Flow Diagram...... 85
4.2.2.2 Data Quality Goals...... 87
4.2.3 Life Cycle Impact Assessment...... 87
4.2.4 Life Cycle Interpretation ...... 89
CHAPTER 5: LIFE CYCLE ASSESSMENT RESULTS AND DISCUSSION ...... 90
ix 5.1 Inventory Results ...... 90
5.1.1 Process Analysis Inventory Results ...... 90
5.1.2 Eco-LCA Inventory Results ...... 94
5.2 Impact Assessment Results...... 99
5.2.1 Impact Categories ...... 99
5.2.2 Direct Impacts ...... 102
5.2.3 Normalized Impacts ...... 106
5.3 Life Cycle Interpretation...... 110
5.4 Sensitivity Analysis ...... 112
CHAPTER 6: CONCLUSIONS ...... 116
APPENDIX A: RAW DATA FROM LEACHING TESTS ...... 118
APPENDIX B: SATURATION INDICES FROM VISUAL MINTEQ ...... 125
APPENDIX C: RAW DATA FROM LIFE CYCLE INVENTORY ...... 132
LIST OF REFERENCES ...... 142
x LIST OF TABLES
Page
Table 3.1: Range of elements present in FGD gypsum (Source: Pflughoeft-Hassett et al.,
2007) ...... 34
Table 3.2: Elemental Composition of FGD Gypsum ...... 39
Table 3.3: Quantities of acid/base added to each extraction bottle for Tier 2...... 45
Table 3.5: TCLP results compared to beneficial use and drinking water criteria...... 55
Table 3.6: SPLP results compared to beneficial use and drinking water criteria...... 56
Table 3.7: Tier 1 Results (“n.d.” means “not determined” - see Section 3.5.2.1) ...... 60
Table 3.8: Tier 2 Results ...... 64
Table 3.9: Tier 3 Results ...... 70
Table 4.1: Description of scenarios used for the LCA ...... 84
Table 4.2: Life cycle impact categories (Source: USEPA, 2006) ...... 88
Table 5.1: Average cost of each AMLR sub process based on 2004-2008 ODNR data. ...91
Table 5.2: Total cost of each scenario and cost per ton of FGD gypsum used...... 93
Table 5.3: Inventory results from NONROAD database...... 95
Table 5.4: Inventory results from Eco-LCA database...... 96
Table 5.5: Energy flow results from Eco-LCA database...... 97
xi Table 5.6: Land usage, water consumption, and carbon flow from the Eco-LCA database.
...... 98
Table 5.7: Top ten economic contributors to reclamation activities...... 103
Table 5.8: Comparison of total sulfur dioxide emissions (tons) from each scenario to normalized sulfur dioxide emissions (tons)...... 110
Table 5.9: Descriptions of sensitivity tests...... 114
Table 5.10: Sensitivity analysis results from economic and geographic variations...... 115
Table A.1: Raw data TCLP results ...... 118
Table A.2: Raw data from SPLP results...... 120
Table A.3: Raw data from Tier 1 results ...... 122
Table A.4: Raw data from Tier 2 results ...... 123
Table A.5: Raw data from Tier 3 results ...... 124
Table B.1: Saturation Indices for TCLP...... 125
Table B.2: Saturation Indices for SPLP...... 126
Table B.3: Saturation Indices for Tier 1...... 127
Table B.4: Saturation Indices for Tier 2...... 128
Table B.5: Saturation Indices for Tier 3...... 130
Table C.1: NONROAD2008 raw data, construction equipment...... 132
Table C.2: NONROAD2008 raw data, lawn and garden equipment ...... 134
Table C.3: Eco-LCA raw data ...... 138
Table C.4: EIO-LCA raw data...... 140
xii LIST OF FIGURES
Page
Figure 2.1: Common FGD wet scrubbing process (Source: fgdbyproducts.org) ...... 17
Figure 2.2: Absorber tank in a FGD wet scrubbing process (Source: babcock.com)...... 18
Figure 2.3: Common FGD dry scrubbing process (Source: USEPA; University of North
Carolina) ...... 19
Figure 2.4: Locations of coal combustion power plants, coal mines, and coal preparation plants (Source: Butalia and Wolfe, 2000) ...... 21
Figure 2.5: Total FGD production anticipated by 2012 (Source: Wolfe et al., 2009)...... 22
Figure 2.6: Tons of FGD material utilized at AMLR sites (Source: Wolfe et al., 2009) ...30
Figure 2.7: Cost per ton of FGD material utilized at AMLR sites (Source: Wolfe et al.,
2009) ...... 31
Figure 3.1: Set-up of Tier 3 procedure - (a) compacted sample in a 10-cm diameter cylindrical concrete test-mold and 10 cm in height, (b) sample and mold in extraction bottle containing 1000mL of deionized water, and (c) extraction bottle covered with air- tight lid...... 46
Figure 3.2: Barium concentration in leachate as a function of pH ...... 65
Figure 3.3: Cadmium concentration in leachate as a function of pH ...... 65
xiii Figure 3.4: Chromium concentration in leachate as a function of pH...... 65
Figure 3.5: Mercury concentration as a function of pH ...... 66
Figure 3.6: Calcium concentration in leachate as a function of pH ...... 66
Figure 3.7: Sulfate concentration in leachate as a function of pH...... 66
Figure 3.8: Concentration of barium (µg/mL) as a function of time (days) ...... 71
Figure 3.9: Concentration of chromium (µg/mL) as a function of time (days) ...... 71
Figure 3.10: Concentration of mercury (µg/mL) as a function of time (days) ...... 71
Figure 3.11: Concentration of calcium (µg/mL) as a function of time (days) ...... 72
Figure 3.12: Concentration of sulfate (µg/mL) as a function of time (days)...... 72
Figure 4.1: Major life cycle processes for AMLR with FGD gypsum ...... 82
Figure 4.2: Flow diagram of AMLR...... 86
Figure 5.1: Impact categories for each scenario...... 100
Figure 5.2: Conventional air pollutants emissions (metric tons) according to direct economic percentage...... 104
Figure 5.3: Total toxic releases and transfers (kg) according to direct economic percentage...... 104
Figure 5.4: Total energy consumption (Terajoules) according to direct economic percentage...... 105
Figure 5.5: Global warming potential (metric tons of CO2 equivalents) according to direct economic percentage...... 105
Figure 5.6: Impact categories normalized against U.S. flow...... 108
xiv LIST OF ABBREVIATIONS
AML Abandoned Mine Land
AMLR Abandoned Mine Land Reclamation
CCP Coal Combustion Products
DMRM Division of Mineral Resources Management
FGD Flue Gas Desulfuization
LCA Life Cycle Assessment
LCI Life Cycle Inventory
LCIA Life Cycle Impact Assessment
ODNR Ohio Department of Natural Resources
OEPA Ohio Environmental Protection Agency
ORC Ohio Revised Code
RCRA Resource Conservation and Recovery Act
SMCRA Surface Mining Control and Reclamation Act
SPLP Synthetic Precipitation Leaching Procedure
TCLP Toxicity Characteristic Leaching Procedure
USEPA United States Environmental Protection Agency
xv CHAPTER 1: INTRODUCTION
1.1 Scope of Work
The overall goal of this study was to understand the fundamental properties of
FGD gypsum and the impacts associated with utilizing FGD gypsum for abandoned mine land reclamation. Many past studies have evaluated the chemical and physical characteristics of FGD material for various applications, including mine land reclamation; but changes in FGD material production in Ohio necessitate further investigation. In this study, FGD gypsum samples from two coal combustion power plants were analyzed in a laboratory to determine their elemental composition and the leaching behavior of inorganic constituents under a variety of conditions. The Toxicity
Characteristic Leaching Procedure (TCLP) and Synthetic Precipitation Leaching
Procedure (SPLP) were used along with the Kosson et al. (2002) Integrated Framework to provide a data set and help understand leaching capability in a variety of disposal and reuse scenarios. Results from the TCLP and SPLP analyses were compared to the Ohio
Beneficial Use Criteria and USEPA’s Drinking Water Maximum Contaminant Levels.
In order to understand the broader environmental impacts of using FGD gypsum for abandoned mine land reclamation, such as overall greenhouse gas emissions, ozone depletion potential, and energy use, a Life Cycle Assessment was conducted in
1 accordance with USEPA’s guidelines, which follow ISO 14000 standards. Results from this assessment were used to interpret how the process of reclaiming an abandoned strip or underground mine land with FGD gypsum impacts the environment. The following sections describe in more detail the need for laboratory and life cycle analyses as well as the specific methods used.
1.2 Changes in FGD Material Production
The production of flue gas desulfurization (FGD) material from coal combustion power plants is undergoing changes in the state of Ohio, which may have implications on beneficial use applications, such as abandoned mine land reclamation (AMLR). Flue gas desulfurization technology has been implemented as a result of the Clean Air Act
Amendments of 1990 in order to remove sulfur dioxide emissions into the atmosphere.
The common FGD technology in Ohio involves a wet scrubbing process, which exposes the flue gas to a lime and water slurry before it exits the flue stack. The byproduct of this technology is a stream of calcium sulfite (CaSO3), which contains a high moisture content and must be dewatered and stabilized with fly ash and lime before storing and transporting. The resulting material is typically referred to as “stabilized
FGD”. Many AMLR projects utilize this calcium sulfite FGD material to recontour the original landscape of abandoned mine lands (AMLs), eliminate dangerous highwalls, abate acid mine drainage (AMD), and reduce sedimentation runoff clogging waterways.
In order to make storage and handling easier as well as to create a more marketable product, recent FGD scrubbing technologies are force oxidizing the calcium sulfite and
2 producing calcium sulfate dihydrate (CaSO4!2H2O), or FGD gypsum. The FGD gypsum is easily dewatered and can be marketable in the wallboard and agricultural industries.
Only two coal combustion power plants in Ohio currently produce FGD gypsum:
AEP’s Cardinal Plant in Jefferson County and Duke Energy’s Zimmer Plant in Clermont
County. AEP owns four other coal combustion power plants in Ohio - two of which produce the calcium sulfite material - that create approximately 3.6 million tons of FGD material each year (Butalia et al., 1999). By the end of 2009, forced oxidation scrubbers will be installed in four currently operating power plants with a projected FGD gypsum production of 5 million tons each year by 2012 (Wolfe et al., 2009). This means that 60% of all FGD byproduct production in Ohio will be FGD gypsum (Butalia et al., 1999).
Unfortunately, the wallboard manufacturing industry is relatively saturated in the United
States, so the opportunity to use this excess of FGD gypsum for AMLR projects is an attractive alternative. However, there is minimal data in the research community regarding the characterization of FGD gypsum and the potential environmental impacts associated with its beneficial use for AMLR.
1.3 Abandoned Mine Lands in Ohio
Prior to the late 1940s, coal companies were not required by law to reclaim active or abandoned strip or underground mines. That changed in 1947 when the Ohio government issued its first Strip Mining Coal Act that required coal companies to pay a bond of $100/acre in order to ensure reclamation would occur. Revisions of the act over the subsequent 25 years tightened restrictions and gave coal companies more accountability for their operations. In 1972, Ohio issued a revised Strip Mining Coal Act 3 that required reclamation of all active mine lands to be completed before bonds were issued to the coal companies. The first piece of federal legislation, called the Surface
Mining Control and Reclamation Act, was enforced in 1977 and placed restrictions upon surface mining operations, permitting, siting, and reclamation activities. Since then, the federal and Ohio governments have revised their statutes and dramatically influenced how coal companies operate and how reclamation occurs. However, despite these advancements there still remains many acres of unreclaimed mine lands in the United
States that were abandoned prior to legislative requirements.
Currently in Ohio, there are over 200,000 acres of abandoned strip mines and over
600,000 acres of abandoned underground mines (Wolfe et al., 2009). Such AMLs contribute to waterway sedimentation issues, acid mine drainage, and various physical and chemical hazards to humans and wildlife. Large quantities of sedimentation runoff have disrupted the natural flow of streams near AMLs and have thus impacted the streams’ ecology. Acid mine drainage that flows into streams also impacts aquatic ecology and presents hazards to nearby residents that use local streams and wells as a primary source of drinking water and for agriculture operations. In addition, acid mine drainage destroys vegetation and changes the surrounding soil ecology. There are some obvious physical hazards associated with AMLs, such as dangerous highwalls that range from 20 ft to 150 ft in height, highwall pits that collect acid runoff, vertical underground mine entrance ways, lateral underground mine entranceways that appear attractive to dangerous animals and curious adventurers, and exposed coal seams left unreclaimed.
4 Of the $351 million (2007 dollars) it will take to reclaim AMLs currently causing severe environmental impacts, approximately $139 million of reclamation activity has been completed and $6 million is currently being funded by federal grants, which leaves
$206 million of reclamation activity that remains unfunded (AMLIS, 2007). In order to keep AMLR a priority despite the lack of federal funding, collaboration with coal combustion power plants to supply beneficial use materials for reclamation is an appealing cost-saving option. However, there are environmental concerns about using coal combustion products (CCPs) for direct land application, including the leaching of mercury into groundwater. According the the federal government, beneficial use materials, such as FGD gypsum, fall under the RCRA Subtitle D classification and thus are considered nonhazardous wastes. Nonetheless, Ohio has set regulatory standards for the leachable concentrations of arsenic, barium, cadmium, chromium, mercury, lead, and selenium in beneficial use materials in order to protect surface and groundwater.
The Ohio Department of Natural Resources (ODNR) requires that thorough investigations be performed to characterize both the material intended for beneficial use and the AML site. The extent of investigation of AMLR projects have varied in the past, where some projects have no known records of geologic profiles, water quality data, or leaching behavior of the beneficial use material utilized. Current AMLR proposals are required to contain specific data regarding the chemical and physical properties of the beneficial use material, hydrologic and geologic characteristics of the reclamation site, and water quality monitoring programs. Though researchers have studied the calcium sulfite FGD material and its impacts on water quality at particular AMLR sites, it cannot
5 be assumed that FGD gypsum will behave similarly under such conditions. Thus, new investigations are needed specifically to characterize the leaching behavior of FGD gypsum and ensure it meets beneficial use criteria for AMLR applications.
1.4 Leaching Experiments
The Toxicity Characteristic Leaching Procedure (TCLP) is the method used to determine whether FGD gypsum is a hazardous material. According to the USEPA definition, toxicity refers to “the degree to which a chemical substance (or physical agent) elicits a deleterious or adverse effect upon the biological system of an organism exposed to the substance over a designated time period,” while a hazard is defined as “the likelihood that a substance will cause an injury or adverse effect under specified conditions” (USEPA, 2009a). Thus, toxicity refers to the relationship between a particular dose of a substance, exposure time, exposure pathway, and health response, while a hazard refers to the probability that an exposure to a substance at a particular dose, within a defined time or exposure frequency, and pathway will cause adverse human health effects (USEPA, 2009a) for which a government regulation has been established (Dawson and Mercer, 1986).
The TCLP regulatory limits that dictate whether a material is hazardous or non- hazardous were developed based on an exposure and risk assessment of potentially toxic substances (i.e., As, Ba, Cd, Cr, Pb, Hg, and Se). Potential direct exposure to such substances are compared to soil screening levels in order to determine whether a material is hazardous when directly inhaled, ingested, or absorbed in the skin (Townsend et al.,
2003). However, the potential indirect exposure to such substances (i.e., exposure to
6 contaminated groundwater) is compared to concentrations that account for dilution and attenuation factors. The regulatory limits used to compare TCLP results account for the dilution and attenuation of each constituent as it passes through soil layers before reaching groundwater (Townsend et al., 2003).
The premise of a TCLP is to expose the material to an extraction fluid of acetic acid (pH of 4.93 ± 0.5) and rotate end-over-end for 18 ± 2 hours. This method assumes equilibrium is reached after 18 hours and only tests the material at a single pH, which limits a more complete characterization. A similar approach, the Synthetic Precipitation
Leaching Procedure (SPLP) is the method used to determine whether FGD gypsum can be used for beneficial use applications in Ohio. It can be used in addition to the TCLP to estimate how the material will leach when exposed to rainwater. In the SPLP, the material is exposed to an extraction fluid of a 60/40 sulfuric/nitric acid mixture (pH 4.20
± 0.5) and rotated end-over-end for 18 ± 2 hours. Again, this test limits characterization to a single scenario based on a single pH and extraction fluid.
Both the TCLP and SPLP have been criticized by researchers as providing only a single-case scenario instead of a more complete representation of how a material behaves when exposed to liquid. The TCLP was developed to classify materials as hazardous or non-hazardous by exposing the materials to organic acids and simulating their leaching behavior in municipal solid waste landfills, assumed to contain 95% municipal solid waste and 5% industrial waste (Sorini, 1996). Research has shown that organic acids, such as the acetic acid used in the TCLP, may enhance extraction (Cheng et al., 2008;
Mohapatra et al., 2005), and therefore the TCLP may not represent the actual leaching
7 behavior that would occur at AML sites. The SPLP was originally created to determine the leaching behavior of contaminated soils and their subsequent effects on groundwater
(Sorini, 1996) where municipal solid waste is not present. It also tests how acid rain affects leaching behavior of materials , but is known to both overestimate and underestimate the risk of groundwater contamination (Townsend et al., 2006).
To alleviate such concerns, Kosson et al. have created a three-tiered framework to examine (i) the total availability of constituents over a pH range, (ii) equilibrium release at individual pH values, and (iii) time dependent release. For the purposes of AMLR, the
Kosson et al. framework in conjunction with the TCLP and SPLP is used in this study and provides valuable information for developing a more detailed leaching characterization model of FGD gypsum. However, despite the benefits of such procedures, these only consider the immediate impacts to surface and ground water quality within the area surrounding an AML and neglect other impacts associated with the entire process of reclamation.
1.5 Life Cycle Assessment
In order to assess a broader set of impacts that AMLR may have on the environment, it is necessary to analyze AMLR as a process that accepts multiple inputs
(e.g., FGD gypsum) and outputs various emissions (e.g., carbon dioxide emissions to the atmosphere). The method for conducting such an analysis is called a Life Cycle
Assessment (LCA). An LCA is a standardized methodology in which a product or system of interest is analyzed from raw material extraction to end-fate, also known as a
8 cradle-to-grave analysis. The USEPA’s standard guidelines, which follow the ISO 14000 standards, was used the reference methodology for this study (see USEPA, 2006).
There are many critiques regarding the validity of LCAs, mainly due to the availability, variation, and aggregation of data (Ayres, 1995). In addition, LCA results are subject to scrutiny simply because of variations of scale, system scope, and even the goals for which the analysis is conducted. Nonetheless, performing an LCA is a beneficial process because it forces researchers to exhibit a systems mindset that aims to uncover the broader implications of each material and energy flow in a process.
The basic structure of conducting an LCA consists of four parts: (1) defining the goals and scope of the analysis, (2) collecting information by conducting a Life Cycle
Inventory (LCI), (3) performing a Life Cycle Impact Assessment (LCIA) in which the
LCI results are analyzed, and (4) interpreting and evaluating the results. The results of this analysis were categorized according to type of impact (e.g., global climate change, eutrophication of major bodies of water, photochemical smog potential, etc.). A sensitivity analysis was performed at the end of the study to determine how responsive the results are to changes in different variables.
1.6 Outline of Chapters
The remainder of this document is divided into five chapters. Chapter 2 provides background information regarding the history of coal mining and legislation in Ohio, a history of FGD production in Ohio, details regarding FGD technologies, and an overview of previous AMLR sites that utilized FGD material. Chapter 3 discusses the properties of
FGD gypsum, including the chemical and physical characteristics. Chapter 3 also
9 explains the methodology followed for leaching experiments and provides a discussion on the resulting data. Chapter 4 introduces the LCA methodology, justifications and limitations associated with such an analysis, and the goals, scope, and assumptions. The results of the LCA are presented in Chapter 5 and include an impact assessment, interpretation, and sensitivity analysis. Chapter 6 summarizes this study’s results and conclusions. Raw data tables, Life Cycle Inventories, and all other relevant documentation can be found in the Appendices.
10 CHAPTER 2: ABANDONED MINE LANDS AND FGD MATERIAL
UTILIZATION IN OHIO
2.1 History of Mining and Legislation in Ohio
Coal mining began in Ohio around the early 1800’s and consisted mainly of underground mining. Strip mining production was first recorded in 1914 and accelerated during World War II due to rapid technological advancements and the loss of experienced workers to the war efforts (Crowell, 1995). During these times, reclamation of mine lands was not common nor favorable for any coal company, mainly because it was seen as an unnecessary cost. Instead, gob and spoil piles were left where they were dumped, coal seams remained exposed, and underground mine entrances were left wide open.
There are some accounts of attempted reclamation in the early 1900s, where a few coal operators were known to plant trees on old strip mines to revegetate the land or cover exposed surfaces with clover and shrubs from pine trees (Crowell, 1995). As a result, the Ohio Reclamation Association was created to take heed of this tree-planting trend. Such an attempt demonstrated that coal operators and the state of Ohio were interested in the surrounding ecology, but reclamation was not considered a high priority until government regulations were enacted.
11 In 1947 Ohio passed its first coal mining law, called the Strip Mining Coal Act, which required mine operators to have a state-issued license and pay a bond of $100/acre for post-mining reclamation and to revegetate the disturbed area. Two years later the
Division of Reclamation within the Department of Agriculture was formed, along with an updated version of the Strip Mining Coal Act. In the same year, the first underground mining law was passed and required operators to close or fence all surface openings abandoned after 1941. In 1955, the Division of Reclamation was moved from the
Department of Agriculture to the Ohio Department of Natural Resources (ODNR), and the reclamation bond rate increased to $220/acre. By 1965, coal operators were required by law to successfully revegetate and regrade the mine land (whereas prior to 1965 there was no statute requiring “successful” vegetation) and submit a post-mining reclamation plan along with their mining permit application (Crowell, 1995). Another revision of the
Strip Mining Coal Act was made in 1972 that required operators to recontour the original landscape of the mining area, replace topsoil, and revegetate the land before the State issued the reclamation bond
Five years later in 1977, the United States Congress passed its first piece of coal mining legislation – the Surface Mining Control and Reclamation Act (SMCRA). This legislation established standards for coal mining, mine land reclamation, permitting, performance, and inspection and began an AML fund by taxing coal operators with respect to production rates. SMCRA also established the creation of the federal
Department of the Interior’s Office of Surface Mining Reclamation and Enforcement
(OSM). In response to SMCRA, Ohio updated its strip mining legislation and
12 strengthened the 1949 underground mining law. Such updates were approved by the
Secretary of the Interior in 1982 and went into effect immediately after. Later in March of 1995, the Ohio Acid Mine Drainage Abatement and Treatment (AMDAT) fund was established to provided funding specifically for AMD-related projects.
Most of the federal and state legislation discussed so far has directly affected only the coal mine operators. However, in 1990 the Clean Air Act (CAA) began regulating air pollution from point sources, which includes the flue stacks of coal combustion power plants. The CAA restricts the quantities of particulate matter, volatile organic compounds
(VOCs), nitrogen oxides (NOx), sulfur dioxide (SO2), and other common air pollutants that can exit a flue stack. This required coal combustion power plants to install new technologies that capture air contaminants, such as desulfurization scrubbers. As a result, new material streams emerged as byproducts of coal combustion (e.g., fly ash and FGD material). Immediate concern arose over these auxiliary coal combustion products
(CCPs) since they contain potentially environmentally hazardous constituents (e.g., mercury, lead, cadmium, etc.).
A year before SMCRA was enacted, the Resource Conservation and Recovery Act
(RCRA) of 1976 was established with particular concern over hazardous waste management. RCRA requires that strict attention be paid to the generation, handling, storage, transportation, and disposal of hazardous waste (NRC, 2006). It was initially uncertain whether CCPs fell into the hazardous waste category, but after extensive investigation the USEPA dismissed the need to classify CCPs as hazardous waste; thus
CCPs fall under RCRA Subtitle D when they are disposed in landfills and surface
13 impoundments. CCPs are also considered “beneficial use” materials that can be recovered and sold to the market. Because of the this classification, CCPs used for AML projects do not necessarily fall under all RCRA Subtitle D regulations; rather, they fall under a set of state-specific criteria that correspond to regulations from the Clean Water
Act (CWA) and Safe Drinking Water Act.
Chapter 1513 of the Ohio Revised Code (ORC) dictates Ohio’s policy on coal mining, reclamation activities, and beneficial use criteria. The Ohio Environmental
Protection Agency (OEPA) created an Interim Alternative Waste Management Program
(IAWMP) in 1997 that categorized beneficial use applications into two “alternative disposal options”: engineered use and land application (OEPA, 1997). Engineered use means that a waste product can be used as an additive, fill, or liner for various construction projects, roads, ponds, landfill caps, etc. A land application means that the material can be used intentionally to improve soil or land conditions, such that it does not threaten the safety of humans and wildlife. ODNR created the Division of Mineral
Resources Management (DMRM) in 2000, which integrated the the Division of Mines and Reclamation and the Division of Oil and Gas. ODNR-DMRM is currently responsible for addressing safety at coal mines and AML reclamation activities.
Current events have stirred a new debate over whether to classify CCPs as hazardous waste. On December 22, 2008, a coal ash disposal cell in Tennessee’s
Kingston Power Plant collapsed and spilled 5.4 million cubic yards of fly ash and water into two major rivers and the surrounding land, which destroyed homes, disrupted utility lines, blocked roads, and killed fish in nearby tributaries (USEPA, 2009b). The EPA
14 immediately executed a monitoring program to determine whether the spill posed immediate health threats and found concentrations of arsenic, cobalt, iron, and thallium above residential soil screening levels (USEPA, 2009b). In addition, arsenic levels were greater than Removal Action Levels (RALs), which are used to determine whether an investigation and immediate removal of contamination are necessary. Surface waters used for residential drinking water supply contained levels of arsenic that exceeded the
Tennessee Water Quality Criteria; and levels of arsenic, cadmium, chromium, copper, lead, mercury, nickel, selenium, and zinc exceeded the National Recommended Ambient
Water Quality Criteria, which protects aquatic life. However, a series of TCLP tests conducted by the EPA resulted in concentrations below regulated criteria; thus, the spilt material was classified as non-hazardous.
The Tennessee ash spill has sparked an increase in social and environmental concerns over the storage and use of coal related materials and has reintroduced the debate on how to classify CCPs. If the USEPA changes current regulations and classifies some or all CCPs under RCRA Subtitle C, there will be significant changes in beneficial use options and criteria.
2.2 History of FGD Material Production in Ohio
As a result of the Clean Air Act, coal combustion power plants began installing sulfur dioxide scrubbing systems in the early 1990s. There are various scrubbing technologies used to remove sulfur dioxide from the flue gas, but the two most commonly used are termed “wet” and “dry” processes. The wet process consists of a lime or limestone and water slurry that transforms sulfur dioxide and calcium hydroxide
15 into calcium sulfite and water (Figures 2.1 and 2.2). The flue gas enters enters an absorber tank where the slurry is contained and chemically reacts with the lime or limestone and water to form particulate matter that settles to the bottom of the tank. The solid calcium sulfite particles are then removed, dewatered, and stored. More recent scrubbing technologies pump air or pure oxygen into the absorber tank to force oxidize the calcium sulfite to calcium sulfate. More details regarding the chemical transformation is provided in Section 3.1. The dry process injects either a lime or sodium based powder material into the flue gas stream to extract the sulfur dioxide and create a solid salt, which is removed with a particulate matter collection system (Figure 2.3). Wet systems are generally more efficient due to better sorbent utilization, and can remove more than 90% sulfur dioxide content (Henzel et al., 1982).
Besides the popular wet and dry scrubbers, there are other less common scrubbers found in Ohio, such as the lime injection multistage barrier (LIMB) process. In this method, a calcium-based sorbet is injected directly into the coal combustion boiler and reacts with sulfur dioxide and oxygen to form calcium sulfate that is extracted with fly ash (Kost et al., 2005). A similar process, the fluidized bed combustion (FBC) method, adds the calcium-based sorbent to the coal stream as it enters the boiler, reacts with sulfur dioxide and oxygen to form calcium sulfate, and is extracted along with the FBC bottom ash and fly ash (Kost et al., 2005).
16 Figure 2.1: Common FGD wet scrubbing process (Source: fgdbyproducts.org)
17 Figure 2.2: Absorber tank in a FGD wet scrubbing process (Source: babcock.com)
18 !"##$%&'& & ((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((( &
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
1+-=$(%'277$'!+2,%$88+'+$,'+$&%'6,%*$%6'=+$4$/%*"%+,'($6&+,-%!=$%87)$%-6&%=).+/+!"%:+0$05% '227+,-%)&+,-%36!$(%&#(6"<0%%K=$&$%/$4+'$&%=64$%*$$,%)&$/%2,%.$/+'67%36&!$%+,'+,$(6!2(&%6,/% 6%8$3%.),+'+#67%36&!$%'2.*)&!2(&0%
E,%%."/#$0"#("$/'%&"'("%5%!=$%87)$%-6&$&%6($%+,!(2/)'$/%+,!2%6,%6*&2(*+,-%!23$(%:/("$(<% 3=$($%!=$%-6&$&%6($%'2,!6'!$/%3+!=%6%8+,$7"%6!2.+B$/%67J67+,$%&7)(("0%%F'+/%-6&$&%6($%6*&2(*$/% *"%!=$%&7)(("%.+@!)($%6,/%($6'!%!2%82(.%&27+/%&67!&%3=+'=%6($%($.24$/%*"%!=$%#6(!+')76!$% '2,!(27%/$4+'$0%%K=$%=$6!%28%!=$%87)$%-6&%+&%)&$/%!2%$46#2(6!$%677%!=$%36!$(%/(2#7$!&5%7$64+,-%6% ,2,?&6!)(6!$/%87)$%-6&%!2%$@+!%!=$%6*&2(*$(%!23$(0%%D#(6"%/("$(&%6($%'6#6*7$%28%6'=+$4+,-%=+-=% :OPQR<%6'+/%-6&%($.2467%$88+'+$,'+$&0%%K=$&$%/$4+'$&%=64$%*$$,%)&$/%2,%+,/)&!(+67%6,/%)!+7+!"% *2+7$(&%6,/%.),+'+#67%36&!$%'2.*)&!2(&0%
+,-./0$!"1&$ 2/3$45/6078$,790:8,57$4:/.660/$43480;$
Figure 2.3: Common FGD dry scrubbing process (Source: USEPA; University of
!"#$ %&'"()*#$ North Carolina)
Ultimately, FGD material production relies on the amount of coal that is
combusted and the coal’s sulfur content. In essence, a higher sulfur content yields more
FGD material. Figure 2.4 displays the geographical distribution of coal combustion
power plants in Ohio. Two coal combustion power plants in Ohio currently produce FGD
gypsum – AEP’s Cardinal Plant (CAR) in Jefferson County and Duke Energy’s Zimmer
Plant (ZIM) in Clermont County. AEP owns four other coal combustion power plants,
two of which produce calcium sulfite FGD material – Conesville Plant (CON) in
Coshocton County and Gavin Plant (GAV) in Gallia County – that constitute 19 approximately 3.6 million tons of FGD material each year (Butalia et al., 1999). In 2009, forced oxidation scrubbers will be installed in four currently operating power plants –
AEP’s Conesville Plant, Muskingum Plant (MUS) in Muskingum County, and Kyger
Plant (KYG) in Gallia County and First Energy’s Sammis Plant (SMS) in Jefferson
County – with a projected FGD gypsum production of 5 million tons by 2012 (Wolfe et al., 2009). This means that 60% of all FGD byproduct production in Ohio will be FGD gypsum (Butalia et al., 1999). Given that the wallboard manufacturing industry is saturated, the opportunity to use FGD gypsum for AMLR projects is an attractive alternative. However, there is minimal data in the research community regarding the characterization of FGD gypsum the potential environmental impacts associated with its beneficial use for AMLR.
20 Figure 2.4: Locations of coal combustion power plants, coal mines, and coal preparation plants (Source: Butalia and Wolfe, 2000)
21 10,000,000
7,500,000
5,000,000 short tons
2,500,000
0 CAR CON GAV KYG MSK SMS Total
FGD gypsum future rate FGD gypsum current production rate Stabilized FGD material current production rate
Figure 2.5: Total FGD production anticipated by 2012 (Source: Wolfe et al., 2009)
2.3 FGD Material Usage at Mine Sites
The National Academic Press published a document in 2006 entitled “Managing
Coal Combustion Residues in Mines” that addresses the environmental, engineering, legislative, and social issues associated with using CCPs in mine lands (NRC, 2006). It discusses the importance of characterizing the mine site proposed for reclamation and the
CCP to be applied; provides design and construction guidelines for individual
22 reclamation activities; recommends pre- and post reclamation monitoring programs of soil, surface waters, and groundwater; and suggests how to conduct an overall performance review. The document makes a distinction among reclaiming active mine sites, abandoned mine sites, and re-mined abandoned sites and concludes that site and
CCP characterization, design and construction planning, and monitoring should be the same for each scenario. However, performance standards and reviews should take into account differences in each scenario’s design and construction, given the quality of original site (e.g., level of degradation).
In accordance with ORC Chapter 1513, ODNR requires a detailed chemical and physical analysis of all coal combustion byproducts intended for beneficial use applications, with particular concern for the material's leaching characteristics (ODNR,
2008). In addition, detailed hydrologic and geologic data is required to characterize the
AML site and long-term monitoring plans must be submitted along with a reclamation permit application. Specifically, AMLR proposals are required to contain data from the
Toxicity Characteristic Leaching Procedure (TCLP) that comply with EPA drinking water standards and a “Maximum Acceptable Leachate Concentration” criteria for beneficial use materials. In addition, ODNR requires detailed plans for sampling surface and ground water (e.g., sampling locations and frequency), geologic profiles of the AML site, surface and ground water flow pre-reclamation and anticipated flow post-reclamation, and a five-year monitoring plan.
In order to identify common applications of FGD material and general reclamation methods for this study, ten AMLR projects in eastern Ohio were reviewed
23 during the summer of 2008. Table 2.1 provides an overview of each project. Generally, the projects that were a part of a research study exhibit extensive laboratory data and detailed site characterization. As required by ODNR, the TCLP was the chosen method to determine the leaching characteristics of FGD material. In some cases, a modified
TCLP was performed by substituting the standardized extraction fluid with acid mine drainage or rain water specific to a site. Various geologic and hydrologic methods were performed to characterize surface and ground water flows and estimate potential impacts to the surrounding area. The following sub-sections provide a summary of each project, which is taken from Wolfe et al. (2009).
2.3.1 Otsego Gob Pile
The Otsego site is located in Monroe Township of Muskingum County. It consisted of a 9-acre unstable gob pile near an abandoned underground mine with severe
AMD and erosion. No formal studies were conducted on this site, so minimal data and information exist. The pile was reclaimed simply by placing and 8-inch layer of stabilized FGD material from Conesville Power Plant followed by a 6-inch layer of resoil material. Reclamation of occurred from October 1992 through October 1993 and cost approximately $80,080. About 1,600 tons of stabilized FGD was used for this project.
Revegetation was successful, where the stabilized FGD acted as a buffer to raise the pH of the gob pile.
2.3.2 Fleming Surface Mine
The Fleming site is located in Franklin Township of Tuscarawas, Ohio. Previous miners left about 25 acres of exposed underclay that eroded over time, 45 acres of
24 unreclaimed surface mine land, and 5 acres of exposed coal refuse. These site conditions resulted in heavy off-site sedimentation into nearby waterways, physical hazards (e.g.,
10-foot deep gullies and ditches), and AMD discharge. Field studies were conducted by
OSU and OARDC (see Dick, 2006). The dry FGD material used for this site came from a General Motors Plant in Pontiac, Michigan that burned Ohio coal. The TCLP was performed to characterize the FGD material’s leaching behavior prior to reclamation.
Three reclamation treatments were used: (1) a control area with limestone and 8 inches of resoil, (2) an area with dry FGD material (125 tons/acre), and (3) an area with dry FGD material (125 tons/acre) and yard waste (50 tons/acre). Reclamation began in
August 1993 and was completed in November 1995. The total construction cost of the project was $451,636 and the total amount of FGD material used was 6,250 tons. The three treatment areas were successfully revegetated, but 10 to 15 acres of spoil still requires stabilization.
2.3.3 Freeport Gob Pile
The Freeport site is located in Lafayette Township of Coshocton County and consisted of 6 acres of unstable, exposed coal refuse. A nearby stream received AMD from the site and suffered degradation for one-half mile stretch. Fly ash and wet FGD material from the Conesville Power Plant were used for reclamation. The pile was regraded to 4.5 acres and a 5-inch layer of fly ash, a 2-inch layer of wet FGD material, and an 8-inch layer of resoil were spread on top. Lime was added to the site at particular intervals: (1) after fly ash and FGD material were added (25 tons/acre), (2) after resoil material was spread (25 tons/acre), and (3) after revegetated grasses has grown (5 tons/
25 acre). Revegetation was successful with no signs of erosion or sedimentation runoff.
The project occurred from April 1995 through October 1995, cost $44,238 total, and used
4050 tons of fly ash and FDG material.
2.3.4 Broken Aro Remining
The Broken Aro underground mine was abandoned in the early 1920s and remined in 1996. There was an underground mine opening and a 8-acre gob pile with visible signs of erosion and AMD discharge into a nearby river. Research studies were performed by AEP and The Ohio State University. Fixated FGD material from
Conesville Power Plant were used to create a semi-impermeable underground mine seal to prevent AMD leakage and two ponds were created to collect and treat AMD. The FGD material was backfilled up on top the seal and against an adjacent highwall to recreate the original contour of the land. In addition, 18 inches of fixated FGD material and 6 inches of soil was placed over two coal refuse piles spanning 3.5 acres. Construction took place
November 1997 to March 1998, costed $208,119, and used 28,973 tons of fixated FGD material. Overall the mine seal successfully mitigated AMD discharge and the refuse pile was fully revegetated, but some AMD still discharging somewhere from the site.
2.3.5 Roberts-Dawson Injection Project
The Roberts-Dawson site was an abandoned 14.6-acre underground mine located in both Adams Township of Muskingum County and Franklin Township of Coshocton
County. The mine was discharging AMD into a stream that flows into a collection pond, which eventually drains into a nearby river. Extensive research studies were preformed by The Ohio State University (see Lamminen et al., 2001; Taerakul et al., 2004). A
26 special grout mixture of 1.25:1 FGD material and fly ash from Conesville Power Plant was used along with 5% lime. A lower strength grout mixture was injected into 317 boreholes, followed by a higher strength grout mixture, to plug the mine and prevent water from going in and seeping out. Construction began in October 1997 and ended
January 1998. The total cost of the site characterization studies, construction costs, and monitoring programs was over $1.6 million. Overall, approximately 23,806 tons of FGD material was injected into the mine. AMD still discharged post-reclamation due to potential leaks between the seal and underclay region.
2.3.6 Rehoboth Refuse Pile
The Rehoboth site is located in Clayton Township of Perry County and constisted of 65 acres of an unstable coal refuse pile that discharged AMD and sedimentation into a nearby river. FGD material was used as a pond liner, impermeable cap, buffer material, and soil amendment for this project. The 65 acre pile was regraded to 45 acres and a rock channel was installed to direct runoff. The refuse pile cap contained 2 inches of compacted FGD material, 2 inches of a FGD and coal refuse mixture, and a resoil layer containing a mixture of spoil material (3.5 inches), FGD material (3.5 inches), and composted yard waste (1 inch). Reclamation took place between February 1997 and
November 1999. The construction cost was $924,000 and 284,402 tons of FGD material was used. Boron toxicity from the coal refuse layers affected revegetation, so in 2004 lime was applied to the site at 25 tons/acre. This costed $79,860, which increased the total reclamation cost to slightly over $1 million. Many areas of the gob pile were
27 successfully vegetated with no visible signs of erosion; however, some areas remained eroded where runoff was highly concentrated.
2.3.7 Rock Run Gob Pile
The Rock Run site is located in Coal Township of Perry county. Two abandoned underground mines and a spoil pile affected about 17 acres of land and a nearby creek with AMD and sedimentation runoff. The piles were regraded to approximately 5.5 acres and was covered with a 2-foot layer of FGD material from Conesville Power Plant, an 8- inch layer of resoil, and lime. Another area of the coal refuse pile was simply covered with a 2-foot layer of resoil and lime. In addition, a passive water treatment system was installed to further mitigate the effects of AMD. Construction began in July 1998 and ended March 2000, with a total cost of $291,412 and a total of 26,959 tons of FGD material. A post-reclamation research study was conducted by Department of Geological
Sciences at Ohio University (Stoertz et al.). In addition, a market analysis proved that the total costs savings from using FGD material as a coal refuse cap was $12,600 per acre
(Butalia et al., 2000). Overall, the project was successfully revegetated and the FGD pile cap aided in preventing AMD discharge.
2.3.8 Central Ohio Coal Company Coal Refuse Impoundment
The Central Ohio Coal Company (COCCO) refuse site is located in Meigsville
Township of Morgan County. It functioned as an impoundment for excess coal refuse and other coal mining related wastes, which creates physical hazards and a toxic environment. Surrounding ditches and channels collected surface runoff and directed water to wastewater treatment plants. A 30-inch layer of FGD material from Conesville
28 Power Plant was placed atop the impoundment along with 18-inches of resoil. The project construction lasted from May 1999 to October 1999. Approximately 75,000 tons of FGD material was used, but no cost information is known. Revegetation was successful and there are currently no signs of erosion.
2.3.9 Conseville Prep Plant Refuse Pile
The Conesville Prep Plant Refuse Pile is located in Franklin Township of
Coshocton County. This site was originally an abandoned surface mine that turned into a coal refuse disposal area in 1984. AMD from the pile was discharging into nearby collection ponds, so a permit was issued to divert 587,000 tons of FGD material from
Conesville Power Plant over a 30-year period to cover the area. The cover consists of 1 to 3 feet of FGD material and 12 inches of resoiling material. As of July 2008, approximately 260,113 tons of FGD material has been placed and about 29 acres have been fully reclaimed and successfully revegetated.
2.3.10 Conesville Highwall
The Conesville Highwall is located in Franklin Township of Coshocton County and is approximately 140 feet in height and 1,800 feet in length. In addition to this physical hazard, AMD was found discharging from the coal seam. Reclamation on this site began in January 2004 with nearly 1.7 million tons of FGD material covered with a
12-inch layer of resoil. Thus far, about 4 of the 30 acres have been successfully reclaimed with 950,000 tons of FGD material.
29 2.3.11 Summary of AML Sites
The total costs for each project varied according to the geographical and geological characteristics of the AML sites, the quantity of FGD gypsum utilized, and revegetation design. Though total costs are important to consider, unit costs are necessary to examine in order to perform a cost comparison across different projects.
Figure 2.6 and 2.7 provide a summary of FGD material used for each AML site and the corresponding unit cost of construction. This type of information is helpful for evaluating the overall environmental impacts associated with various AMLR projects, which is discussed in more detail in Chapters 5 and 6 of this document.
Fleming 6,250
Rock Run Gob Pile 16,000
Roberts - Dawson UGM 24,139
Broken Aro 28,973
Freeport Gob Pile 1,704
Rehoboth Gob Pile 284,402
Otsego Gob Pile 1,600
0 75,000 150,000 225,000 300,000 Tons of FGD Material Utilized
Figure 2.6: Tons of FGD material utilized at AMLR sites (Source: Wolfe et al., 2009)
30 Fleming $2.40
Rock Run Gob Pile $1.08
Roberts - Dawson UGM $50.50
Broken Aro $1.03
Rehoboth Gob Pile $1.10
Otsego Gob Pile $4.50
$0 $15 $30 $45 $60 Cost per Ton of FGD Material Utilized
Figure 2.7: Cost per ton of FGD material utilized at AMLR sites (Source: Wolfe et al., 2009)
31 CHAPTER 3: CHEMICAL AND PHYSICAL PROPERTIES OF FGD GYSPUM
This chapter presents information regarding how FGD gypsum is created in a wet scrubbing process, its typical chemical composition, and the physical properties useful for beneficial use applications. In addition, this chapter discusses the methodology and results of leaching characterization procedures used for FGD gypsum samples.
3.1 Chemical Properties
As air flows through the FGD wet scrubber system, it enters a water slurry with either lime (Ca(OH)2) or limestone (CaCO3). The lime or limestone dissociates and sulfur dioxide binds to calcium atoms to form solid calcium sulfite (CaSO3), as shown in
Equation 3.1 and 3.2 (Pflughoeft-Hassett et al., 2007). This calcium sulfite is oxidized with oxygen gas to form calcium sulfate dihydrate (CaSO4!2H2O), a synthetic form of gypsum (Equation 3.3) (Pflughoeft-Hassett et al., 2007).
Ca(OH)2 (solid) + SO2 (gas) " CaSO3 (solid) + H2O (liquid) Equation 3.1
CaCO3 (solid) + SO2 (gas) " CaSO3 (solid) + CO2 (gas) Equation 3.2
CaSO3 (solid) + 1/2O2 (gas) + 2H2O " CaSO4!2H2O (solid) Equation 3.3
32 To clarify, there is a difference between gypsum, or hydrated calcium sulfate
(CaSO4!2H2O), and anhydrous calcium sulfate (CaSO4) or anhydrite. Gypsum is a softer material with a larger molecular weight and smaller density than anhydrite. Due to the differences in hardness and density, gypsum is a much more favorable product for agricultural applications because gypsum is more soluble than anhydrite. Greater solubility means that more calcium and sulfate ions are available for vegetation. In addition, wallboard manufacturing facilities require FGD material to be 95% pure calcium sulfate dihydrate, not anhydrite.
Forced oxidation FGD scrubbers do not solely produce gypsum. The byproduct stream is actually a mixture mainly composed of gypsum, but with smaller concentrations of calcium sulfite, fly ash, unreacted lime or limestone, and other compounds that were formed in the coal combustion boiler and travelled through the gas stream. AEP’s
Cardinal Plant performs extensive and frequent testing on their FGD gypsum in order to increase its market value for the gypsum manufacturing sector. They require that the
FGD by-product material is composed of 95% pure calcium sulfate dihydrate with a maximum of 0.5% calcium sulfite, 1.5% fly ash, 1.5% silica (SiO2), and 5.0% calcite
(CaCO3 and MgCO3). They have other qualifications including particle size, organic content, oil content, and metals.
In general, FGD gypsum contains a wide range of elements. Pflughoeft-Hassett et al. (2007) conducted a literature review and found ranges of many elements that span multiple orders of magnitude. For instance, Cr concentrations ranges from 0.56 to 42 percent by weight and Se concentrations range from 2.0 to 30.0 percent by weight (Table
33 3.1). Pflughoeft-Hassett et al. (2007) also concluded that the crystalline structure, major chemical composition, and trace element content of FGD gypsum was similar to that of natural gypsum. Though, the chemical composition of FGD gypsum greatly depends upon the chemical composition of the source coal, constituent removal technologies (e.g., fly ash removal systems), and any chemical reactions that occur during combustion and through the gas stream.
Element Concentration (% wt) Element Concentration (% wt) As 0.6 - 4.0 Mo 0.5 - 12.0 B <0.0003 - 0.0042 Na 0.023 - 5.8 Ba 33.8 - 77.2 Ni 0.73 - 20.1 Ca 18.5 - 24.3 P 0.0005 - 0.05 Cd 0.003 - 1.2 Pb 0.8 - 12.0 Cr 0.56 - 42.0 S 14.9 - 20.9 Cu 0.96 - 27.9 Sb 0.1 - 9.1 Fe 0.026 - 0.3 Se 2.0 - 30.0 Hg 0.004 - 1.4 Si 0.022 - 69.2 K <0.004 - 0.09 Tl 0.6 - 2.0 Li <9 - 48.8 V <1 - 73.2 Mg 0.010 - 0.74 Zn 3.4 - 102 Mn 1.37 - 52.1
Table 3.1: Range of elements present in FGD gypsum (Source: Pflughoeft-Hassett et al., 2007)
34 3.1.2 Mercury
The presence of mercury in coal combustion byproducts has become a great environmental concern. The average concentration of mercury in coal is 0.1 ppm
(Kairies et al., 2005). Most of the mercury from coal is collected in the fly ash system or the FGD scrubber, but some can escape the flue stack as an atmospheric emission. Wet
FGD scrubbers are known to co-remove 50% to 75% of the soluble oxidized mercury
(Pavlish et al., 2003; Senior et al., 2000; Srivastava et al., 2006; Withum et al., 2004).
According to a literature review performed by Pflughoeft-Hassett et al. (2007), a concentration range of 0.004-1.4 ppm was found in solid FGD gypsum samples.
Kairies et al. (2006) performed a study to examine the mercury concentration and leaching behavior in FGD gypsum using a continuous stirred-tank reactor (CSTX). The motivation of the study was to determine how much mercury was being transferred to wallboard manufacturing facilities and captured in drywall. They found a wide range of mercury concentrations spanning an order of magnitude in both the FGD gypsum samples from coal combustion power plants and completed drywall products. This provides evidence that mercury concentrations, along with the overall composition of
FGD gypsum, is highly dependent upon the chemical composition of the source coal and flue gas processes. They also found that no mercury actually leached from the FGD gypsum samples and less than 1% of the total Hg concentrated leached in the wallboard samples. Another study determined that mercury can be released during calcination processes at wallboard manufacturing facilities (Heebink and Hassett, 2003).
35 Schroeder et al (2006) exposed FGD gypsum samples to deionized water, acetic acid, hydroxylamine, hydrochloric acid, hydrogen peroxide, and ammonium acetate to extract particular species. Less than 2% of the original mercury content was present in the post-leaching residue, where most leachate contained between 0 and 2 x 10-5 ppm and the maximum concentration was 5 x 10-5 ppm. They found that mercury species were typically particulate-bound and that gypsum was not responsible for leaching behavior or mercury, rather 80% of mercury was retained when high concentrations of Al and Fe were present. However, they concluded that this retention behavior does not necessarily apply to AML scenarios where pH is low and conditions are anoxic.
Various studies have shown that higher temperatures increase mercury emissions in natural environments (Gustin et al., 1997; Gustin et al., 2006; Carpi et al., 1998) and that adding moisture to soil also increases emissions (Gustin et al., 2006; Wallschlager et al., 1999; Ericksen et al., 2005; Bahlmann et al., 2004a). Other studies have investigated the effects of photochemistry on mercury emissions (Gustin et al., 2006; Carpi et al.,
1998; Wallschlager et al., 1999; Xin et al., 2007; Bahlmann et al., 2004b; Gustin et al.,
2002). Pekney et al. (2008) researched the photochemical effects on FGD gypsum and fly ash samples. They concluded that UV-A light within the range 0.6 to 2.2 µW/cm and visible light within the range 430 to 570 lux did not necessarily affect mercury flux of
FGD gypsum.
3.1.3 Boron
Boron is another element of concern, especially for AMLR scenarios. Coal typically contains high concentrations of boron (Barth et al., 1984), which is oxidized
36 during combustion and collected mainly in fly ash (NRC, 2006). FGD gypsum is known to contain within <0.0003-0.0042 boron by percent weight (Pflughoeft-Hassett et al.,
2007). Boron is relatively non-toxic to humans, but can be toxic to plants at high concentrations. Plants need boron to maintain cell wall structure; however, the difference between boron deficiency and toxicity is very slight (Reid et al., 2004). Though plants species vary in their boron requirements, the general toxicity threshold is 100 mg/kg
(Adriano, 2001; Clark et al., 1999). This presents an issue when using CCPs for AMLR due to the sensitivity of vegetation on boron concentration. For example, the Rehoboth site discussed in Section 2.3.6 experienced boron toxicity that affected revegetation. To alleviate the issue an additional layer of lime was placed on the site, which increased total construction costs by $79,860.
3.1.1 Elemental Analysis
Samples of solid, dry FGD gypsum from AEP’s Cardinal Plant and First Energy’s
Bruce Mansfield Plant were analyzed at STARLab/OARDC to determine their elemental composition. Total sulfur content was digested with perchloric acid then analyzed with an Inductively Coupled Plasma Spectrophotometer (ICP) (Teledyne Leeman Labs
Prodigy Dual) (AOAC, 1988). The mercury concentration was determined by Cold
Vapor Atomic Fluorescence Analysis (CETAC M8000) (USEPA, 2001c). The remaining elemental composition was determined with an ICP (Jones et al., 1991).
Table 3.2 displays the results of this analysis. As anticipated, calcium and sulfur represented the majority of each sample’s composition. Arsenic, cadmium, chromium, selenium, and lead concentrations of both samples were below detection limits. Boron
37 concentrations were relatively low, but greater in the Cardinal Plant sample. In addition, the Cardinal Plant sample yielded a greater mercury concentration than the Bruce
Mansfield Plant sample, but was not found in significant quantities during leaching experimentation as described in later sections of this chapter.
38 Bruce Element Cardinal Mansfield (!g/g) Plant Plant Al 307 70.5 As <1.28 <1.28 B 4.31 <1.52 Ba 65.5 46.0 Be <0.091 <0.091 Ca 207000 192000 Cd <0.048 <0.048 Co <0.146 <0.146 Cr <0.194 <0.194 Cu <0.378 <0.378 Fe 750 349 Hg 0.355 <0.005 K <74.7 <74.7 Li 28.5 10.8 Mg 1460 <0.494 Mn <0.059 <0.059 Mo <0.225 <0.225 Ni <0.182 <0.182 P <0.975 <0.975 Pb <0.774 <0.774 S 136000 142000 Sb 6.12 5.29 Se <2.32 <2.32 Si 389 43.5 Sr 187 167 Tl <0.190 <0.190 V <0.297 <0.297 Zn 1.55 <0.552
Table 3.2: Elemental Composition of FGD Gypsum
39 3.2 Physical and Mechanical Properties
FGD gypsum is a powder-like substance that can be white, gray, or even a red- brown color. Its particle size can range from 0.025 mm to 0.25 mm in diameter depending upon the scrubbing technology used (Kost et al., 2005). Its specific gravity ranges from 2.31 to 2.33 and has a hardness of 1.5 to 2 (Mohs scale). The 28-day unconfined compressive strength of FGD gypsum can range from 12 MPa to 26 MPa
(Kovacs and Mulnar, 2003; Singh and Garg, 1997; Tazawa, 1998; Tesarek et al., 2007), yet other research has shown a range of 213 to 358 kPa (Chesner et al., 2008).
Studies have shown that soils containing FGD material are less susceptible to erosion and compaction and have a greater field capacity than those without FGD material (Korak et al., 1998; Norton et al., 1999). In fact, gypsum is a common material used to mitigate compaction of soils containing large quantities of sodium and improve water flow (Alcordo et al., 1995; USDA, 1954). This characteristic of FGD gypsum is not favorable for AMLR scenarios that involve capping a pile or sealing a mine entrance to prevent water from entering and AMD from exiting. When FGD material is fixated with fly ash and lime, it can create a semi-impermeable layer to cover gob piles and seal underground mine entranceways (e.g., Broken Aro project). However, it may be susceptible to weathering under harsh AMD conditions (Lamminen et al., 2001).
40 3.3 Overview of Leaching Experiments
FGD gypsum samples were taken from two different coal combustion plants:
American Electric Power’s Cardinal Plant in Brilliant, Ohio and First Energy’s Bruce
Mansfield Plant in Shippingport, Pennsylvania. Bruce Mansfield Plant samples were chosen because the scrubbing technology used is similar to that of First Energy’s Sammis
Plant in Stratton, Ohio. To characterize the leaching behavior of these samples, three standardized methods were implemented: the EPA Standard Method 1311, Toxicity
Leaching Characteristic Procedure (TCLP), the EPA Standard Method 1312, Synthetic
Precipitation Leaching Procedure (SPLP), and an Integrated Framework established by
Kosson et al. (2002).
Kosson et al. (2002) presents the argument that TCLP is a single, worst-case scenario test, and have thus created an alternative three-tiered framework that provides material-specific information regarding (a) total constituent concentrations available for leaching, (b) concentrations based on constituent solid-to-liquid equilibrium, and (c) concentrations based on mass transfer rate. The FGD gypsum from AEP’s Cardinal Plant was analyzed under each tier, as described in the following paragraphs.
3.4 Methodology
3.4.1 Toxicity Characterization Leaching Procedure
The TCLP was implemented to determine the mobility of inorganic constituents within the synthetic calcium sulfate at an initial pH of 4.93±0.05 with the use of acetic acid and LS ratio of 20:1. Three 100 g samples from AEP’s Cardinal Plant and two 100 g samples from First Energy’s Bruce Mansfield Plant were placed into high-density
41 polyethylene extraction vessels (Thermo Scientific #2124-0005), along with 2000 mL of extraction fluid. The extraction fluid was prepared by diluting 5.7 mL of glacial acetic acid (Fisher Chemical #A507-500) and 64.3 mL of sodium hydroxide (Fisher Chemical
#S318-500) with deionized water to 1000 mL. The extraction vessels and their contents were rotated end-over-end at a speed of 30±2 rpm for 18±2 hours. The liquid from each extraction vessel was then filtered through a 0.7 #m binder-free borosilicate glass fiber filters (Whatman #1825-047) and preserved with 2.5% trace metal grade nitric acid
(Fisher Chemical #A509-500) for analysis.
3.4.2 Synthetic Precipitation Leaching Procedure
The SPLP was implemented to determine the mobility of inorganic constituents exposed to synthetic rainwater at a pH of 4.90±0.05 with the use of a sulfuric acid and nitric acid mixture and an LS ratio of 20:1. Two 100 g samples from AEP’s Cardinal
Plant and three 100 g samples from First Energy’s Bruce Mansfield Plant were placed into high-density polyethylene extraction vessels (Thermo Scientific #2124-0005), along with 2000 mL of extraction fluid. Unlike the TCLP, the extraction fluid was prepared by diluting 60 g of concentrated trace mental grade sulfuric acid (Fisher Chemical
#A510-500) and 40 g of concentrated trace metal grade nitric acid (Fisher Chemical
#A509-500) with deionized water to 1000 mL. The extraction vessels were rotated end- over-end at a speed of 30±2 rpm for 18±2 hours. As described for the TCLP, the liquid from each extraction vessel was then filtered (Whatman #1825-047) and preserved with
2.5% trace metal grade nitric acid (Fisher Chemical #A509-500) for analysis.
42 3.4.3 Integrated Framework
3.4.3.1 Tier 1
The Tier 1 screening test was performed in accordance with the Netherlands
Environmental Agency Standard NEN 7371 (2004) to evaluate the total availability of each constituent present in FGD gypsum. One 16 g sample and 800 mL of deionized water were magnetically stirred until all FGD gypsum was in suspension, and the initial pH was recorded. After 10 minutes, the pH of the mixture was recorded again and the material was classified as “neutral reactive” (i.e., the initial pH was less than 10, but the pH after 10 minutes was greater than 7). Under this classification, 0.2 M nitric acid was added until the mixture reached a pH of 7.0. Periodically, 0.2 M nitric acid was added to the solution in order to sustain a pH of 7.0±0.5 for three hours. The liquid portion was then filtered through 0.45 #m membrane filters and preserved. The solid portion remaining on the filters was mixed with 800 mL of deionized water by a magnetic stirrer, and 1 M nitric acid was added to reach a pH of 3.0 . Periodically, 1 M nitric acid was added to the solution in order to sustain a pH of 3.0±0.5 for three hours. Again, the liquid was filtered through 0.45 #m membrane filter and added to the previous filtered liquid.
This combined leachate was preserved with 2.5% nitric acid for analysis.
3.4.3.2 Tier 2
The Tier 2 test chosen is entitled, “SR002.1 Alkalinity, Solubility, and Release as a Function of pH” (Kosson et al., 2000). Initially, a material-specific titration curve was created in accordance with “pH001.0 pH Titration Pretest” (Kossen et al., 2000) for the pH range 2 through 11. To create a titration curve for pH range 2 through 6, an 8 g
43 sample was mixed with 800 mL of deionized water, 500 #L aliquots of 2 N nitric acid were added, and then measuring the pH after 20 minutes. For pH range 6 through 11, another 8 g sample was mixed with 800 mL of deionized water and 100 #L aliquots of 1
N sodium hydroxide (for pH range 6 through 9). The equivalent volumes of nitric acid and sodium hydroxide were normalized against the 8 g sample in order to determine the required amounts needed to reach the target pH range. This information was used to create a schedule of acid, base, and deionized water additions needed to conduct the
SR002.1 procedure (Table 3.3). Nine 40 g of samples were added to nine different extraction bottles (Thermo Scientific #332189-0016) , along with appropriate amounts of acid, base and deionized water associated with each target pH value (includes one replicate and one blank) as shown in Table. The bottles and their contents were rotated end-over-end at a speed of 28±2 rpm for 48 hours. After a settling time of 15 minutes, the liquid from each bottle was filtered through a 0.45 #m polypropylene membrane and preserved with 2.5% nitric acid for analysis.
44 Volume of Volume of Target pH deionized acid/base (ml) water (ml)
2 5.32 394.68
3 2.72 397.28
4 1.80 398.20
5 1.4 398.60
6 0.88 399.12
7 0.00 400.00
9 0.08 399.92
Table 3.3: Quantities of acid/base added to each extraction bottle for Tier 2.
3.4.3.3 Tier 3
The Tier 3 test is entitled, “MT002.1 Mass Transfer Rate in Granular
Materials” (Kosson et al., 2000). A sample of FGD gypsum was compacted into a 10-cm diameter cylindrical concrete test-mold to a height of 10 cm (Figure 3.1a). The sample and mold were placed gently in an extraction bottle (Cole-Parmer #AP-06083-15) containing 1000 mL of deionized water according to procedure, such that the only the top surface of the sample was in direct contact with the water (Figure 3.1b). The extraction bottle was covered with an air-tight lid and left alone for 1 hour (Figure 3.1c). After the one hour period, the sample and mold were removed from the extraction bottle and
45 placed into a separate extraction bottle filled with 1000 mL of fresh deionized water. The sample and mold were then left alone for another hour. The leachate remaining in the original extraction bottle was filtered through a 0.45 #m polypropylene membrane and preserved with 2.5% nitric acid for analysis. This process continued until the leachate was collected for the following cumulative time intervals: 1, 2, 5, and 8 hours, 1, 2, 4, and
8 days.
(a) (b) (c)
Figure 3.1: Set-up of Tier 3 procedure - (a) compacted sample in a 10-cm diameter cylindrical concrete test-mold and 10 cm in height, (b) sample and mold in extraction bottle containing 1000mL of deionized water, and (c) extraction bottle covered with air-tight lid.
46 3.4.4 Chemical analysis
Constituent concentrations of each leachate sample were tested at STARLab in
Wooster, Ohio. Each leachate sample was analyzed for Al, As, B, Ba, Be, Ca, Bd, Co, Cr,
Cu, Fe, K, Li, Mg, Mn, Mo, Ni, P, Pb, S, Sb, Se, Si, SO4-S, Sr, Tl, V, and Zn with an ICP
(Teledyne Leeman Labs Prodigy Dual) after microwave digestion to determine constituent concentrations (Greenburg et al., 1992, Section 3120B). Sulfate (as sulfur) was measured with an Ion Chromatograph (Dionex DX 120) (Greenburg et al., 1992,
Section 4110C) and mercury was measured by Cold Vapor Atomic Fluorescence Analysis
(USEPA, 2001c).
3.4.5 Visual MINTEQ
The software program Visual MINTEQ version 2.60 was used to determine the geochemical speciation of leachate solution for all procedures. MINTEQA2 is the original version of this program created by the USEPA in 1999. In October 2007, Jon
Petter Gustafsson created Visual MINTEQ, a Windows-compatible version. The pH data and leachate concentrations (as mg/L) for each procedure were inputted into the program.
Ionic strength was set at “to be calculated” and temperature was 25˚C. After selecting
“Run MINTEQ”, the saturation indices were obtained and used to determine how mineral precipitation controlled leaching in each solution.
3.5 Results and Discussion
All sample analyses yielded concentrations of regulated constituents well below
Ohio Beneficial Use Criteria and Ohio EPA’s drinking water maximum contaminant
47 levels. TCLP and SPLP results are displayed in Tables 3.5 and 3.6, respectively, and the
Kosson et al. Integrated Framework results are presented in Tables 3.7 through 3.9.
3.5.1 TCLP and SPLP
The TCLP analysis of both samples yielded concentrations of arsenic, barium, cadmium, chromium, copper, lead, and selenium well below EPA drinking water standards and the Ohio non-toxic criteria. On average, concentrations of boron, iron, magnesium, manganese, strontium, and zinc were notably higher in Cardinal Plant samples. As anticipated, leachate from all FGD gypsum samples mostly contained concentrations of calcium, sulfur, and sulfate (as sulfur).
The SPLP analysis yielded concentrations of arsenic, barium, cadmium, chromium, copper, lead, and selenium well below EPA drinking water standards and the
Ohio EPA non-toxic criteria. On average, concentrations of boron, magnesium, and manganese were found to be greater in Cardinal Plant samples, while concentrations of iron and phosphorus were greater in Bruce Mansfield samples. Calcium, sulfur, and sulfate (as sulfur) were the most abundant constituents found in all FGD gypsum samples, as expected.
In addition to the Ohio non-toxic criteria and EPA drinking water standards, the results of the TCLP and SPLP were also compared to the Ohio Secondary Maximum
Contaminant Level/Drinking Water Standards. According to Chapter 3745-82 of the
ORC, this set of standards is defined as “he advisable maximum level of a contaminant in water which is delivered to the free-flowing outlet of the ultimate user of a public water system”. Essentially, these values represent the maximum concentrations that can leave a
48 public water system intended for public consumption and use. The results from both the
TCLP and SPLP show that concentrations of sulfate surpass the secondary criteria. The
TCLP results also show that concentrations of Fe from the Cardinal Plant sample surpass this criteria. In addition, the Cardinal Plant sample during the TCLP and SPLP yielded concentrations of Mn above the secondary criteria. Though, the criteria does not necessarily place any restriction on utilizing FGD gypsum for AMLR, it does indicate that source waters containing large quantities of sulfate, Fe, and Mn may require additional treatment at a public water utility.
Though the results of the TCLP provide justification for classifying FGD gypsum as a non-hazardous material, the results do not necessarily reveal information regarding the mechanisms that control leaching. A study performed by Cheng et al. (2008) investigated the effects of naturally occurring organic ligands on the leaching of fixated
FGD material. The study concluded that the organic ligands oxalate, citrate, maleate, and
Pahokee peat humic acid did effect the leaching behavior of fixated FGD material at particular pH values. Another study conducted by Mohapatra et al. (2005) concluded that oxalate, citrate, and acetate influence the leaching of arsenic, and that leaching in the presence of these organic species is pH dependent. Though the scope of this study is to understand the leaching behavior of FGD gypsum at AML sites and not at municipal solid waste landfills, such studies suggest that naturally occurring organic acids at AML sites may influence the leaching behavior of FGD gypsum with respect to pH. In addition, these studies suggest that the use of acetic acid in the TCLP test may have influenced the leaching behavior of the FGD gypsum samples in this study. Acetic acid is
49 considered an aggressive organic acid, which may explain why the TCLP results yielded higher constituent concentrations than the SPLP results.
3.5.1.1 pH
Calcium sulfite and hydrated calcium sulfate are typically neutral or alkaline materials. Research has shown that though FGD material is beneficial for vegetation due to high calcium and sulfur content, it does not necessarily act well as a buffer to amend acidic soil commonly found at AML sites (Clark et al., 2000). The initial pH of the
Cardinal Plant sample averaged 8.03 and the initial pH of the Bruce Mansfield Plant averaged 6.39 (Table 3.4).
The relatively low final pH values of the TCLP test suggest that neither FGD gypsum sample was able to substantially buffer the system in the presence of acetic acid.
In addition, the Bruce Mansfield sample was unable to substantially buffer the SPLP extraction fluid containing a sulfuric/nitric acid mixture. Though the acetic acid used in the TCLP test is a weak acid, it is considered more aggressive than the sulfuric and nitric acids used in the SPLP test. Given that the Cardinal Plant sample was more alkaline than the Bruce Mansfield sample, it performed well as a buffer in the presence of a sulfuric/ nitric acid mixture. Since alkalinity is potentially associated with calcium concentration and given that Cardinal Plant samples yielded higher concentrations of calcium in the
TCLP and SPLP tests, it may be inferred that the buffering capacity of Cardinal Plant samples is generally greater than that of the Bruce Mansfield Plant samples.
With respect to AMLR, raising the pH of an acidic soil system is not necessarily the sole focus; rather, the focus may be on finding an optimal pH that increases the
50 availability of nutrients to promote plant growth and reduce the availability of toxic substances (Clark et al., 2001). In general, pH is an important environmental factor that influences the speciation of chemical constituents and possibly the leaching behavior of
FGD gypsum. The characteristics of extraction fluids used in the TCLP and SPLP tests along with their respective pH values likely influenced the release of each constituent from samples.
Bruce Mansfield Cardinal Plant Plant Initial pH 8.03 6.39 TCLP final pH 5.92 5.21 (extraction fluid pH 4.93±0.05) SPLP final pH 7.91 5.24 (extraction fluid pH 4.20±0.05)
Table 3.4: Comparison of the initial and final pH values for TCLP and SPLP tests.
3.5.1.2 Geochemical speciation
A list of selected saturation indices can be found in Appendix B. Gypsum was near saturation (saturation index: -0.350), yet slightly undersaturated, as was the anhydrite species (saturation index: -0.600) in the TCLP testing of Cardinal Plant samples. The supersaturated species is this system were aluminum hydroxide (Al(OH)3
(soil)), aluminum oxide (Al2O3), Al4(OH)10SO4, alunite (KAl3(SO4)2(OH)6), barite 51 (BaSO4), boehmite (gamma-AlO(OH)), diaspore, gibbsite, hercynite, imogolite, kaolinite, and plumbgummite. Such solids may have been responsible for the leaching behavior of
Cardinal Plant samples in the TCLP system. The supersaturated species in the SPLP leachate were aluminum hydroxide, barite, boehmite, diaspore, gibbsite, hercynite, and hydroxyapatite. Gypsum and anhydrite were near saturation (saturation indices: -0.433 and -0.683, respectively).
The TCLP testing of the Bruce Mansfield Plant samples yielded a nearly saturated gypsum and anhydrite species (saturation indices: -0.328 and -0.578, respectively), and the supersaturated species consisted of alunite, barite, diaspore, and gibbsite. Gypsum and anhydrite were also near saturation in the SPLP leachate (saturation indices: -0.454 and -0.704, respectively), and the supersaturated species were Al(OH)3, alunite, barite, diaspore, and gibbsite.
3.5.1.3 Calcium and sulfur
Calcium and sulfur are essential nutrients that support plant growth. FGD gypsum is mainly composed of calcium and sulfur, which makes it a favorable material for the revegetation of AML sites; however, it is essential to know how much calcium and sulfur is available for leaching in order to assess its potential benefits. The Cardinal Plant samples yielded greater concentrations of calcium and sulfur than the Bruce Mansfield
Plant, most likely due to differences in source coal composition, combustion processes, and fly ash collection systems. In addition, Cardinal Plant intentionally produces an FGD byproduct with high gypsum purity in order to create a more marketable product for the wallboard industry.
52 The presence of calcium and sulfur at AML sites may reduce the availability of potentially toxic constituents, such as aluminum (Clark et al., 2001). Aluminum can be toxic to plants at low pH levels (Kinraide, 1991); but, high concentrations of calcium and sulfur may alleviate aluminum toxicity since calcium tends to readily exchange with aluminum on soil particles (Foy, 1992). Contrarily, such high concentrations of calcium and sulfur could result in excessive accumulation in plant tissues and deprive plants of other essential nutrients (Clark et al., 2001), specifically magnesium, potassium, and phosphorus (Korcak et al., 1998).
3.5.1.4 Mercury
As mentioned earlier in this chapter, the presence of mercury in coal combustion byproducts has become a great environmental concern. Most of the mercury contained in coal is collected as fly ash, scrubbed out during the FGD process, or emitted into the atmosphere. The elemental analysis of mercury in the Cardinal Plant sample yielded a concentration of 0.335 ppm, yet during the TCLP and SPLP tests only 1.8 x 10-5 ppm and
3.6 x 10-6 ppm leached from the material. However, this only represents the leaching of mercury at a singe pH value over a time period of 18 hours. The Kosson et al. integrated framework results discussed in Section 4.4 provide more information regarding the leaching behavior of mercury at different pH values and as a function of time.
3.5.1.5 Boron
Boron is not regulated under the Ohio Non-Toxic Criteria or EPA Drinking Water
MCLs, but as stated previously, it can be toxic to vegetation at high concentrations. The elemental analysis of the Cardinal Plant soil sample yielded a boron concentration of 4.31
53 µg/g, and in the TCLP and SPLP tests 0.137 µg/mL and 0.130 µg/mL leached from the sample, respectively. The boron concentration in the Bruce Mansfield Plant soil sample was below the detection limit 1.52 µg/g, and the TCLP and SPLP tests yielded concentrations of 0.0255 µg/mL and 0.0297 µg/mL, respectively. Given that generally,
100 mg/kg (100 µg/g) is the threshold for boron toxicity in plants (Adriano, 2001; Clark et al., 1999), both FGD gypsum samples contained and leached relatively insignificant quantities of boron.
54 Bruce Mansfield Cardinal Plant Limits Plant Element Ohio Ohio TCLP Standard TCLP Standard EPA (!g/ml) Non-Toxic Secondary Mean Deviation Mean Deviation MCL Criteria Levels Al <0.034 0.000 <0.035 0.000 0.2 As <0.006 0.000 0.0075 0.002 0.3 0.01 B 0.137 0.004 0.0255 0.001 Ba 0.373 0.024 0.270 0.007 60 2 Be <0.00046 0.000 <0.00046 0.000 Ca 848 2.60 668 9.84 Cd 0.00167 0.000 0.00055 0.000 0.15 0.005 Co 0.0015 0.000 <0.00073 0.000 Cr 0.00587 0.000 0.00575 0.001 3 0.1 Cu <0.001 0.000 <0.001 0.000 1.3 Fe 0.335 0.025 0.0218 0.002 0.3 Hg 1.8E-05 9.6E-06 2.6E-06 4.3E-07 0.06 0.002 K 2.01 0.128 1.76 0.053 Li 0.132 0.021 0.084 0.004 Mg 14.7 0.946 0.236 0.002 Mn 0.150 0.002 <0.0003 0.000 0.05 Mo 0.00213 0.000 0.00355 0.002 Ni 0.00753 0.001 <0.00091 0.000 P 0.017 0.004 0.0287 0.007 Pb <0.003 0.000 <0.003 0.000 1.5 0.015 S 570 14.9 689 19.8 Sb 0.0226 0.001 0.0237 0.002 Se 0.0123 0.003 <0.011 0.000 1 0.05 Si 0.432 0.010 0.195 0.004 SO4-S 615 2.43 729 2.44 250 Sr 1.13 0.004 0.693 0.009 Tl 0.00847 0.003 <0.007 0.000 V <0.001 0.000 0.002 0.001 Zn 0.170 0.004 0.0560 0.010 5
Table 3.5: TCLP results compared to beneficial use and drinking water criteria.
55 Cardinal Plant Bruce Mansfield Plant Limits
Ohio Ohio Element SPLP Standard SPLP Standard EPA Non-Toxic Secondary (!g/ml) Mean Deviation Mean Deviation MCL Criteria Levels Al <0.034 0.000 <0.034 0.000 0.2 As <0.006 0.000 <0.006 0.000 0.3 0.01 B 0.130 0.001 0.0297 0.003 Ba 0.101 0.007 0.0976 0.008 60 2 Be <0.00046 0.000 <0.00046 0.000 Ca 520 5.97 461 23.6 Cd 0.0015 0.002 0.0005 0.000 0.15 0.005 Co <0.00073 0.000 <0.00073 0.000 Cr 0.0044 0.000 0.0048 0.000 3 0.1 Cu <0.001 0.000 <0.001 0.000 1.3 Fe <0.00085 0.000 0.0157 0.015 0.3 Hg 3.6E-06 4.77E-07 2.72E-06 1.14E-06 0.06 0.002 K <0.373 0.000 <0.373 0.000 Li 0.077 0.012 0.0697 0.006 Mg 1.29 0.007 0.259 0.032 Mn 0.0681 0.005 0.00137 0.000 0.05 Mo 0.00275 0.000 0.00313 0.001 Ni 0.00185 0.000 0.00160 0.000 P <0.004 0.000 0.0328 0.005 Pb <0.003 0.000 <0.003 0.000 1.5 0.015 S 422 17.9 485 18.0 Sb 0.0263 0.001 0.0302 0.003 Se <0.011 0.000 <0.011 0.000 1 0.05 Si 0.134 0.024 0.178 0.027 SO4-S 482 6.82 484 2.97 250 Sr 0.616 0.017 0.447 0.026 Tl <0.007 0.000 <0.007 0.000 V 0.0014 0.001 <0.001 0.000 Zn 0.0464 0.008 0.0803 0.005 5
Table 3.6: SPLP results compared to beneficial use and drinking water criteria.
56 3.5.2 Kossen et al. Integrated Framework Results
3.5.2.1 Tier 1 Results
The goal of Tier 1 was to provide a preliminary screening of the maximum constituent concentrations available for leaching. The column titled “Leachate
Concentration” in Table 3.7 displays the results. Though the Kosson et al. Framework is not directly associated with any regulatory criteria, it is worth mentioning that concentrations of As, Ba, Cd, Cr, Cu, Hg, Pb, and Se were well below the Ohio non-toxic criteria and EPA drinking water limits.
The system was modeled using Visual MINTEQ, and it was discovered that the system was supersaturated with species Al(OH)3 (soil), Al4(OH)10SO4, alunite, barite, boehmite, diaspore, gibbsite, hercynite, imogolite, and kaolinite. These solids may have been responsible for the leaching behavior in this system. Gypsum and anhydrite were near saturation (saturation index -0.429 and -0.679, respectively), which explains why a relatively small percentage of calcium and sulfur were leached from the sample. See
Appendix B for a list of selected saturation indices.
In accordance with NEN 7371:2004, the availability of each constituent relative to the concentrations within the solid FGD gypsum sample (see Table 3.2) was determined and displayed as a percentage in Table 3.7. For concentrations below detection limits, the detection limits were used as the maximum potential concentration. Thus, relative availability of Al, As, Be, Co, K, P, Pb, Se, and Tl represent the maximum potential relative availability. The equipment used to determine the elemental composition of the solid FGD gypsum sample had higher detection limits than the equipment used to analyze
57 the leachate in Tier 1, which means there is uncertainty as to exactly how much of each element was present in the solid sample. The concentrations of Cd, Cr, Cu, Mn, Mo, Na,
Ni, Tl, and V present in Tier 1 leachate were greater than the solid sample detection limits, and the leachate concentration of Zn was greater than the solid sample concentration. Thus, the relative availability of these constituents was unable to be determined and is denoted by “n.d.” in Table 3.7. Such discrepancies may be the result of a non-homogenous mixture of elements within the solid FGD gypsum sample. Samples for each leaching procedure were drawn from a single container, yet within that container concentrations of elements were not necessarily evenly dispersed. Thus, it is possible that the solid sample analyzed was less concentrated with the elements listed above, relative to the sample used for Tier 1.
Nonetheless, the relative availability values provide a better understanding of how much Ca, S, Hg, and B leached from the solid sample. As mentioned previously, gypsum and anhydrite species were near saturation making only 28.3% and 28.8% of calcium and sulfur, respectively, available for leaching. The leachable concentration of mercury was approximately 0.0002% of the total concentration, a seemingly insignificant quantity.
Contrarily, almost all the boron present in the solid sample leached in solution (94.7%).
Since boron is very soluble in water, this high leachable concentration is important to consider when applying FGD gypsum to AMLR projects.
In addition to comparing the leaching results of Tier 1 to the elemental composition of the solid FGD gypsum sample, the Tier 1 results were compared with the
TCLP and SPLP concentrations. It was recognized that concentrations of Al, As, Be, Co,
58 K, P, Pb, Se, and Tl were below detection limits for both Tier 1 and SPLP. Tier 1 extracted more Ba, Ca, Cr, Cu, Fe, Hg, Mg, Mo, Ni, Si, V, and Zn than the SPLP; but, the
SPLP extracted more B, Li, Mn, S, Sb, and SO4 than Tier 1. Similar concentrations of Cd and Sr were found in both procedures, suggesting that the extraction fluids and pH had no effect on their leaching. Greater concentrations of S and SO4 in the SPLP most likely resulted because the extraction fluid consisted of a 60/40 sulfuric/nitric acid mixture, which likely increased the total S concentration in the system. Since the SPLP simulates the effects of rainwater on leaching, it may be inferred that boron will leach more in the presence of rainwater than solely in the presence of nitric acid. In addition, it may be inferred that mercury will behave contrarily, such that it will leach less when exposed to rainwater than solely to nitric acid.
Concentration of Al, As, and Be were below detection limits for both Tier 1 and the TCLP. Tier 1 extracted more Cr, Cu, Fe, Mo, Ni, and V than the TCLP; but the
TCLP extracted more B, Ba, Ca, Co, Hg, K, Li, Mg, Mn, P, S, Sb, Se, Si, SO4, Sr, Tl, and
Zn than Tier 1. Similar concentrations of Cd were found in both procedures, suggesting that the extraction fluids and pH had no effect on their leaching. As discussed previously, the extraction fluid in the TCLP is acetic acid, an organic acid that may enhance leaching.
This may explain why more leaching was observed in the TCLP compared to Tier 1.
Overall, it is not clear whether the Tier 1 test actually provided a screening of the maximum leaching potential of FGD gypsum. According to the results of this study, the
TCLP extracted the greatest leachable concentrations of most elements from FGD gypsum samples. Tier 1 ranked second and the SPLP extracted the least.
59 Leachate FGD Gypsum Relative Element Concentration Concentration Availability (!g/ml) (µg/g) (%) Al <0.034 307 <1.11 As <0.006 <1.28 <47.2 B 0.0405 4.31 94.7 Ba 0.161 65.5 24.7 Be <0.00046 <0.091 <50.9 Ca 582 207000 28.3 Cd 0.0017 <0.048 n.d. Co <0.00073 <0.146 <31.7 Cr 0.0076 <0.194 n.d. Cu 0.0059 <0.378 n.d. Fe 0.660 750 8.87 Hg 6.03E-06 0.355 0.000171 K <0.373 <74.7 <50.3 Li 0.0367 28.5 13.0 Mg 12.9 1460 89.0 Mn 0.0570 <0.059 n.d. Mo 0.004 <0.225 n.d. Na 0.0698 <13.04 n.d. Ni 0.0349 <0.182 n.d. P <0.004 <0.975 <41.3 Pb <0.003 <0.774 <39.0 S 388 136000 28.8 Sb 0.0198 6.12 32.5 Se <0.011 <2.32 <47.8 Si 0.201 389 5.19 SO4-S 473 - - Sr 0.668 187 36.0 Tl <0.007 <0.190 n.d. V 0.0028 <0.297 n.d. Zn 0.0750 1.55 n.d.
Table 3.7: Tier 1 Results (“n.d.” means “not determined” - see Section 3.5.2.1)
60 3.5.2.2 Tier 2
Tier 2 was performed in order to assess the leaching potential of each constituent as a function of pH. Resulting data is displayed in Table 3.8. Al, As, Co, Cu, K, P, Pb,
Se, and V concentrations were below detection limits at each pH value. B concentrations remained stable at approximately 0.22 µg/mL, and only slight fluctuations were seen in
Li, Mo, and Sb. Concentrations of Ba, Mg, Mn, and Ni decreased as pH increased, likely because species containing these elements complexed and/or approached saturation as pH increased. Cd decreased from pH 2-5 and mostly likely continued this trend from pH
6-9, though concentrations were below detection limits. Concentrations of Cr seemed to fluctuate, but considering that Cr typically exists in elemental form at low pH levels and forms bonds with (OH)- ions as pH increases (Nriagu and Nieboer, 1988), it is likely that the fluctuation was simply a result of the inherent variation of Cr in FGD gypsum samples and that the general trend was a decrease in Cr concentration.
Gypsum and anhydrite species were near saturation at all pH levels, but there was some fluctuation in Ca concentrations. At pH 2 gypsum was slightly less saturated
(saturation index -0.701), which may explain the relatively low Ca concentration; however, the concentration was greater at pH 3-5 since Ca usually exists in elemental form at low pH levels. Since solubility of gypsum is not affected by pH within the range
6.5-9.5 (Lehr, 2000), Ca remained stable with respect to gypsum; but, at pH 9 and 11 hydroxylapatite became supersaturated,which most likely decreased Ca availability.
Concentrations of S and SO4 increased slightly from pH 2-5, most likely due to a
61 decrease in bisulfate, and remained relatively stable from pH 6-9. This lack of change is likely a result of the relatively stable saturations of gypsum and barite.
Hg concentrations remained relatively stable, but spiked at pH 6 and 9. Previous studies state that insignificant concentrations of Hg are leachable from FGD gypsum
(Kairies et al., 2005), specifically at pH levels greater than 2 (Schroeder et al., 2006). In this study, the Hg concentration in the solid FGD gypsum sample was 0.355 ppm, which is within the typical range of 0.004-1.4 ppm (Pflughoeft-Hassett et al., 2007), and the maximum concentration in the Tier 2 leachate was 7.90 x 10-5 ppm. This leachate concentration is only 0.0223% of the total Hg content in the solid sample, which is well below the 1-2% leaching potential found in other studies (Kairies et al., 2005; Schroeder et al., 2006). According to the literature, it is likely that most of the Hg adsorbed to particles containing Al and/or Fe (Schroeder et al., 2006). Thus, the spikes observed at pH 6 and 9 were most likely a result of the inherent variation of mercury in the FGD gypsum sample.
According to Visual MINTEQ, it is possible that the supersaturated species of Al,
Ba, Fe, Pb, and Si were controlling leaching behavior at particular pH values. Barite was supersaturated at pH 2 though 9 and diaspore and gibbsite were supersaturated from pH
5 through 9. In addition to barite, diaspore, and gibbsite, at pH 6 and 7 the supersaturated species were Al(OH)3 (soil), Al2O3, Al4(OH)10SO4, boehmite (AlO(OH)), hercynite
(FeAl2O4), imogolite (, and kaolinite. At pH 9, however, the additional supersaturated species likely to have controlled leaching were hercynite, hydroxyapatite
(Ca5(PO4)3(OH)), Pb(OH)2), and tenorite (CuO). At pH 11 the supersaturated species
62 were barite, brucite, chrysolite, greenalite, hercynite, hydroxyapatite, Ni(OH)2, Pb(OH)2, and tenroite. See Appendix B for a list of selected saturation indices.
The results from Tier 2 within the pH range 3-7 were compared directly to Tier 1 results. Both Tier 2 and Tier 1 results yielded concentrations of Al, As, Be, K, P, Pb, and
Se below detection limits. Tier 1 extracted greater concentrations of Ca, Cd, Cr, Cu, Fe,
Mg, Ni, and Zn than Tier 2, which provides evidence that Tier 1 results may represent the maximum leaching potential for such elements under the specified conditions. However,
Tier 2 extracted greater concentrations of B, Hg, Mn, Mo, Na, Sb, Si, SO4, and Sr than
Tier 1. Ba, Li, Si, and Sr concentrations were similar at pH 7, but Tier 2 generally extracted more from pH 3-6. Results were similar for S and SO4 at pH 4 and 5, while
Tier 2 extracted more at pH 6 and 7 and Tier 1 extracted more at pH 3. Such differences in concentrations may be a result of the uneven distribution of such elements within the
FGD gypsum samples, or it may indicate that the Tier 1 test did not necessarily extract the maximum leachable quantity of each constituent.
Recall that the TCLP generally extracted more leachable concentrations than Tier
1, and that Tier 1 was generally able to extract more leachable concentrations than the
SPLP. The same pattern holds true when comparing Tier 2 results with the TCLP and
SPLP. In general, at pH 4 and 5 the Tier 2 test was able to extract more B, Ba, Ca, Hg,
Mg, Mn, Mo, Ni, Si, and Sr than the SPLP. The TCLP, however, was able to extract more of most constituents expect B and Mo. Sb was extracted in similar quantities between
Tier 2, SPLP, and TCLP suggesting that pH and extraction fluid did not influence leaching.
63 Element (!g/ml) pH 11 pH 9 pH 7 pH 6 pH 5 pH 4 pH 3 pH 2 Al <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 As <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 B 0.0077 0.217 0.227 0.222 0.222 0.220 0.220 0.224 Ba 0.0755 0.142 0.168 0.164 0.179 0.187 0.184 0.136 Be <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 0.0005 Ca 295 451 527 486 558 582 564 404 Cd <0.00024 <0.00024 <0.00024 <0.00024 0.0003 0.0004 0.0007 0.0015 Co <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 Cr 0.0047 0.0019 0.0043 0.0051 0.0047 0.0032 0.0037 0.0056 Cu <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 Fe 0.0022 <0.00085 <0.00085 0.205 0.0051 <0.00085 0.0011 <0.00085 Hg 6.92E-06 7.90E-05 9.55E-06 2.16E-05 8.11E-06 8.18E-06 7.32E-06 8.25E-06 K 0.6460 <0.373 <0.373 <0.373 <0.373 <0.373 <0.373 <0.373 Li <0.017 0.0270 0.0377 0.0420 0.0379 0.0413 0.0278 0.0313 Mg 0.0649 1.63 2.25 2.46 2.81 6.36 10.7 18.9 Mn 0.0058 0.0022 0.109 0.143 0.149 0.182 0.213 0.315 Mo 0.0016 0.0083 0.0049 0.0062 0.0050 0.0053 0.0096 0.0076 Na 0.0684 6.20 0.428 0.388 0.429 0.456 0.506 0.312 Ni 0.0012 <0.00091 0.0013 0.0017 0.0032 0.0038 0.0064 0.0165 P <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 Pb <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 S 238 397 414 406 391 382 374 376 Sb 0.0135 0.0283 0.0299 0.0295 0.0266 0.0263 0.0370 0.0244 Se <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 Si <0.011 0.275 0.210 0.309 0.250 0.350 0.485 1.07 SO4-S 322 495 499 493 487 472 455 422 Sr 0.339 0.593 0.698 0.644 0.745 0.782 0.771 0.578 Tl <0.007 <0.001 0.0077 0.0018 0.0051 <0.001 0.0022 0.0049 V <0.001 <0.015 <0.015 <0.015 <0.015 <0.015 <0.015 <0.015 Zn 0.0095 <0.002 <0.002 0.0077 0.0071 0.0166 0.0356 0.0345
Table 3.8: Tier 2 Results
64 0.20 0.15 0.10 µg/mL 0.05 0 0 2 4 6 8 10 12 pH
Figure 3.2: Barium concentration in leachate as a function of pH
0.0015 0.0011 0.0008
µg/mL 0.0004 0 0 2 4 6 8 10 12 pH
Figure 3.3: Cadmium concentration in leachate as a function of pH
0.0060 0.0045 0.0030 µg/mL 0.0015 0 0 2 4 6 8 10 12 pH
Figure 3.4: Chromium concentration in leachate as a function of pH
65 0.00008 0.00006 0.00004 µg/mL 0.00002 0 0 2 4 6 8 10 12 pH
Figure 3.5: Mercury concentration as a function of pH
600 450 300 µg/mL 150 0 0 2 4 6 8 10 12 pH
Figure 3.6: Calcium concentration in leachate as a function of pH
500 375 250 µg/mL 125 0 0 2 4 6 8 10 12 pH
Figure 3.7: Sulfate concentration in leachate as a function of pH
66 3.5.3.3 Tier 3
The Tier 3 test results demonstrate the leaching behavior of a monolithic sample of FGD gypsum over defined time intervals. Table 3.9 displays concentrations in the leachate after each time interval, as well as the cumulative concentration. For concentrations that fell below detection limits, the detection limit was used to calculate cumulative concentrations and therefore represents the maximum concentration. Both individual and cumulative values fall below regulatory limits for all constituents of concern.
The concentrations of As, Cd, Cu, Pb, and Se were below detection limits.
Generally, Mg, Mn, Mo, B, and Sr concentrations tended to increase with time. K and Ni likely followed the same trend, but concentrations were below detection limits at 2 and 5 hours on, respectively. Some fluctuation was seen in Ba and Hg concentrations. Barium leaching was most likely controlled by barite formation; and at times 5 hr, 8 hr, 1 day, and
4 days barite was undersaturated, which may explain why leachate concentrations were lower at such times. Maximum leaching of Hg occurred after one day, then decreased over the remaining seven days. The cumulative concentration of Hg (0.0001 ppm) is the largest amount found across all leaching procedures. Though it is an order of magnitude greater than values typically found in literature (e.g., Schroeder et al. (2006) found concentrations of 0 to 2 x 10-5 ppm with a maximum concentration of 5 x 10-5 ppm), it is only 0.0282% of the total Hg content in the original solid sample, which is consistent with the literature threshold of 1-2% (Kairies et al., 2006; Schroeder et al., 2006).
67 The super saturated species that likely controlled overall leaching concentrations were Al(OH)3, Al2O3, Al4(OH)10SO4, alunite, barite, boehmite, diaspore, gibbsite, hercynite, imogolite, and plumbgummite. Barite was undersaturated at 5 hr, 8 hr, 1 day, and 4 day, while kaolinite was supersaturated at times 5 hr, 4 days, and 8 days. Gypsum and anhydrite were near saturation at 1 hr (saturation indices: -0.697 and -0.947, respectively), but undersaturated throughout the remaining time intervals (saturation index: <-1.1). This likely explains why the maximum Ca, S, and SO4 concentrations were observed at time 1 hr and less was observed at subsequent times. Fluctuations in concentrations after 1 hr, however, were most likely due to fluctuations in other mineral saturations and changes in complexes formed.
The cumulative concentrations leached in Tier 3 were compared with that of previous leaching results. Concentrations of Al, As, Be, Cd, Co, P, Pb, Se, and Tl were below detection limits for both Tier 3 and Tier 1. Tier 3 accumulated greater concentrations of B, Ba, Ca, Cr, Hg, K, Mo, S, Sb, Si, SO4, and Sr than Tier 1, while Tier
1 extracted more Fe, Mg, Mn, Ni, and Zn. Tier 3 also extracted more B, Ca, Cr, Fe, Hg,
Mo, S, Sb, SO4, and Sr than the TCLP procedure. The TCLP extracted more Ba, Mg,
Mn, and Zn. At pH 6 and 7, Tier 2 was only able to extract greater concentrations of B,
Mg, Mn, and Na than Tier 3. Based on this comparison, it can be inferred that Ba, Mg,
Mn, and Zn are more pH dependent, while pH has no apparent effect on Ca, S, and SO4 leaching. This comparison may provide further evidence that Tier 1 did not extract the maximum leachable concentrations present in solid samples.
68 Overall, the Tier 3 results provide evidence that when FGD gypsum was exposed to deionized water for increased time intervals the availability of each constituent changed due to changes in mineral saturation. This may be representative of how availability might change at an AMLR site due to various intervals of water exposure. If
AMLR conditions remain within a pH range of 6.5-9.5, the solubility of gypsum will not likely be altered (Lehr, 2000) and leachable Ca, S, and SO4 concentrations will remain stable over time. B, Mg, and Mn concentrations may increase with time; but Mg and Mn concentrations will be more affected by changes in pH, whereas B availability will likely not depend on pH. This signifies that though a majority of B is expected to leach from the solid FGD gypsum, it will do so gradually and will be unaltered by changes in pH.
Mg and Mn concentrations are also expected to leach gradually from the solid FGD gypsum, but will be more available at lower pH levels. Hg concentrations peaked at one day during the Tier 3 test and decreased during subsequent time intervals, suggesting that long-term leaching of Hg at AMLR sites is not necessarily a concern likely due to adsorption onto Al and Fe particles.
69 Element (!g/ml) 1 hour 2 hours 5 hours 8 hours 1 day 2 days 4 days 8 days Cumulative pH 6.28 6.39 6.41 6.46 6.59 6.63 6.74 6.89 - Al <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 <0.272 As <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.048 B 0.0077 0.0082 0.0084 0.0077 0.0168 0.0256 0.0307 0.0506 0.1557 Ba 0.0755 0.0351 0.0170 0.0225 0.0265 0.0402 0.0280 0.0418 0.2866 Be <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00368 Ca 295 134 60.3 84.5 97.7 149 96.8 149 1066 Cd <0.00024 <0.00024 <0.00024 <0.00024 <0.00024 0.0003 <0.00024 <0.00024 <0.00198 Co <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00584 Cr 0.0047 0.0026 0.0017 0.0025 0.0025 0.0043 0.0041 0.0053 0.0277 Cu <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.008 Fe 0.0022 <0.00085 <0.00085 <0.00085 <0.00085 0.334 <0.00085 <0.00085 <0.340 Hg 6.92E-6 8.77E-6 6.54E-6 8.42E-6 1.27E-5 5.86E-6 4.90E-6 3.94E-6 5.81E-5 K 0.646 <0.373 <0.373 <0.373 <0.373 <0.373 <0.373 <0.373 <3.257 Li <0.017 <0.017 <0.017 <0.017 <0.017 <0.017 <0.017 <0.017 <0.136 Mg 0.0649 0.0580 0.0694 0.0620 0.128 0.181 0.349 0.676 1.589 Mn 0.0058 0.0030 0.0023 0.0016 0.0066 0.0069 0.0102 0.0132 0.0496 Mo 0.0016 0.0013 0.0014 <0.001 0.0013 0.0029 0.0028 0.0025 0.0148 Ni 0.0012 0.0018 <0.00091 <0.00091 <0.00091 <0.00091 <0.00091 <0.00091 <0.00846 P <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.032 Pb <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 0.024 S 238 111 51.8 72.9 83.7 127 82.3 125 891 Sb 0.0135 0.0124 0.0109 0.0119 0.0130 0.0130 0.0099 0.0139 0.0985 Se <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 <0.088 Si <0.011 <0.011 0.0363 <0.011 <0.011 0.0163 0.0871 0.181 0.365 SO4-S 322 143 59.1 86.7 102 162 104 165 1143 Sr 0.339 0.155 0.0689 0.0959 0.113 0.172 0.117 0.182 1.244 Tl <0.007 <0.007 <0.007 <0.007 <0.007 0.0107 <0.007 <0.007 0.0597 V <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.008 Zn 0.0095 0.0037 <0.002 <0.002 <0.002 <0.002 <0.002 <0.002 0.025
Table 3.9: Tier 3 Results
70 0.08 0.06 0.04 µg/mL 0.02 0 0 1 2 3 5 6 7 8 Time (days)
Figure 3.8: Concentration of barium (µg/mL) as a function of time (days)
0.006
0.004
µg/mL 0.002
0 0 1 2 3 5 6 7 8 Time (days)
Figure 3.9: Concentration of chromium (µg/mL) as a function of time (days)
0.000015 0.000012 0.000009 0.000006 µg/mL 0.000003 0 0 1 2 3 5 6 7 8 Time (days)
Figure 3.10: Concentration of mercury (µg/mL) as a function of time (days)
71 300 240 180 120 µg/mL 60 0 0 1 2 3 5 6 7 8 Time (days)
Figure 3.11: Concentration of calcium (µg/mL) as a function of time (days)
400 320 240 160 µg/mL 80 0 0 1 2 3 5 6 7 8 Time (days)
Figure 3.12: Concentration of sulfate (µg/mL) as a function of time (days)
72 3.6 Conclusion
The results from this study showed that in general, Tier 3 was able to extract more leachable concentrations of most constituents than the TCLP, which extracted more leachable concentrations than Tier 1, Tier 2, and the SPLP. Some of the variation in this data may be explained by the inherent non-homogenous mixture of constituents within the solid FGD gypsum sample. Tier 2 provides evidence that some constituents (e.g., Mg and Mn) depend upon pH, but that the presence of an organic acid may enhance leaching.
Tier 3 provides evidence that Ca, S, and SO4 are more depended on the solubility of gypsum and anhydrite. Comparisons between the three Tiers, the TCLP, and SPLP show that extraction fluid and mineral saturation may control the leaching of some constituents
(e.g., Ba).
The results from Tier 1, Tier 2, and Tier 3 provide useful information regarding the use of FGD gypsum at AML sites. Instead of simply providing a “leachate concentration (as measured in the TCLP), below which no significant impact to drinking water is anticipated” (Kosson et al., 2002), it evaluates the release of constituents in various conditions. This is helpful in determining whether FGD gypsum will be useful for AMLR projects based on the characteristics of the AML site. For instance, such results can aid in determining whether FGD gypsum will perform well as a buffer in acidic soils, leach toxic elements in AMD scenarios or increase background levels, and provide nutrition or toxicity to vegetation.
The Tier 2 results indicate that no significant concentrations of regulated constituents are expected to leach at low pH levels commonly found at AML sites.
73 Specifically, Hg concentrations are expected to be relatively low and unaltered at pH 2-5.
Mg and Mn, however, are likely to be more available for leaching at lower pH levels and are useful nutrients for revegetation. Ca, S, and SO4 concentrations rely more on the solubility of gypsum than pH, but will likely be relatively stable at lower pH levels. Tier
3 results suggest that a majority of the available concentration of each constituent is expected to leach during an initial short-term exposure to water. Ca, S, and SO4 concentrations are expected to remain relatively stable after a few short intervals of exposure, while B, Mg, and Mn are expected to increase gradually with subsequent exposures. However, the concentrations of regulated constituents are expected to be insignificant over time, which means that the long-term leaching of potentially hazardous constituents is not necessarily a concern for AMLR projects.
Overall, the FGD gypsum samples exhibited leaching behavior that is favorable for beneficial use. All regulated constituents were well below state standards, which makes FGD gypsum a non-hazardous substance suitable for environmental application.
In addition, the TCLP and SPLP experiments yielded results that complied with the
USEPA’s drinking water standards. The material has a relatively neutral pH, which makes it compatible with most soil ecologies, but does not necessarily act well as a buffer to raise the pH of an acidic environment. Due to its high concentration of sulfur, calcium, and magnesium, FGD gypsum is capable of supplying vegetation with essential nutrients to support growth (Clark et al., 2001).
74 CHAPTER 4: LIFE CYCLE ASSESSMENT OVERVIEW AND METHODOLOGY
4.1 Overview of Life Cycle Assessment
A Life Cycle Assessment (LCA) is a systems approach to understanding the broader implications of industrial practices (USEPA, 2006). It is an analytical method that takes into account all the materials, energy, resources, and processes directly related to the fabrication of a particular product and the subsequent environmental emissions, energy consumption, and waste streams that result. Such a method is termed a “cradle- to-grave” analysis, because it investigates material and energy streams from raw material extraction through waste disposal (USEPA, 2006). It investigates the impacts of a product’s immediate material and energy streams, as well as the material and energy streams that afford product fabrication (e.g., truck manufacturing).
The motivation for conducting an LCA in this study is to clarify overall environmental impacts associated with using FGD gypsum for abandoned mine land reclamation. LCAs have been used in previous studies regarding coal combustion and coal combustion byproducts; however, minimal research targets the actual utilization of coal combustion byproducts for specific beneficial use purposes. For example, Whitfield
(2006) performed a detailed LCA on the impacts of coal combustion, from raw material extraction (i.e., coal mining) to the disposal of combustion byproducts in Florida. The
75 study presents details regarding the total environmental emissions, energy use, etc. with different coal combustion byproduct disposal options; however, it does not specify the impacts associated with particular beneficial use options.
Babbitt and Lindner (2008) developed a set of CCP disposal scenarios and conducted an LCA to determine the environmental impacts and savings associated with beneficial use options in Florida. For example, they compared a scenario where 50% of all CCPs were disposed in landfills and 50% were utilized for beneficial use purposes with a scenario where all CCPs were disposed in landfills or surface impoundments.
They concluded that it was possible to reduce environmental impacts by utilizing CCPs; however, the was no data comparing individual beneficial use options.
Spath et al. (1999) conducted an LCA on the entire coal combustion process and considered the impacts of recycling CCPs versus disposal in landfills, but there was no data regarding specific CCP uses. Thorneloe (2002) took a life cycle approach to uncovering the path of mercury throughout the coal combustion process and investigated whether more mercury was available in disposal or reuse options. Since mercury was the focus of this study, there was no overall assessment regarding total environmental impacts.
4.1.1 Types of LCAs
Though the methodology of an LCA is standardized, there are different scales that an LCA can describe. A process-level analysis considers only the immediate materials, energy, and processes needed to perform a certain activity or produce a particular product, whereas an LCA includes a broader set of material and energy flows (e.g., the
76 fabrication of various materials used during raw material extraction and of transportation equipment used to haul materials). The following sections briefly describe different types of LCAs and how they are used in this study.
4.1.1.1 Conventional LCA
A conventional LCA broadens a product analysis by accounting for all energy and emissions that flow through relevant processes. To perform a detailed LCA from scratch requires a substantial amount of time and research in order to generate a suitable database
(USEPA, 2006). To alleviate this major constraint, efforts have been made to create national and global inventory databases that can be accessed through various software programs and internet websites. The data is aggregated by the North American Industry
Classification System (NAICS), which classifies all industrial processes into sectors according to their activities. This study uses some of these databases to model different
AMLR scenarios and assess AMLR at a process level.
4.1.1.2 Economic Input-Output LCA
Researchers at Carnegie-Mellon in Pittsburgh, Pennsylvania developed a broader scale analysis that normalizes sector data according to economic activity. This type of database is considered an Economic Input-Output LCA (EIO-LCA). It is used to determine the flows of emissions and energy through various sectors based on how sectors exchange money (Hendrickson et al., 2006). Essentially, the EIO-LCA provides valuable economic data and allows researchers to map the interactions between sectors.
It is used primarily in this study to distinguish between direct and indirect flows through sectors that support AMLR. Direct flows represent the the emissions and energy that are
77 a outputted due to AMLR, while indirect flows represent the emissions and energy that are outputted elsewhere in other sectors whether or not AMLR takes place (Hendrickson et al., 2006).
4.1.1.3 Ecological LCA
Researchers at the The Ohio State University have created the Ecological LCA
(Eco-LCA) database that accounts for natural capital, goods and services provided by nature, carbon sequestration, land and water usage, as well as the energy made available through sector activities (Zhang et al, 2008). The EIO-LCA model is the foundation of this database, such that all data is represented in terms of economic exchanges between sectors. It is used in this study to categorize overall environmental impacts of AMLR, such as global warming potential and ozone depletion potential.
4.1.2 Limitations of LCAs
LCAs are of course limited by their availability of data, precise and accurate data, and the goals of the researchers. Process-level studies are limited by their narrow scope, though they usually have more precise immediate data. Conventional LCAs are generally very costly and time consuming. The amount of time and effort needed to create a database and account for all the pertinent material and energy streams and their corresponding data is tremendous. Broader datasets, such that of the EIO-LCA and Eco-
LCA are useful for less detailed studies; however, such data is aggregated by NAICS sector and may not accurately represent all AMLR scenarios. All things considered,
LCAs are still very valuable studies because they can broaden stakeholder discussions and provide a basis for comparing seemingly unlike processes.
78 4.2 Life Cycle Assessment Methodology
A standard LCA consists of four major parts: goal definition and scope, inventory analysis, impact assessment, and interpretation of results (USEPA, 2006). The goal definition and scope component describes the overall purpose of the study and serves as a guide for subsequent tasks. Essentially, this step defines the system of interest and necessary assumptions. Once the overall system is defined, a life cycle inventory (LCI) plan is created in order to collect all necessary input and output data (e.g., energy, emissions, cost, etc.). This data can be collected from various sources, depending on the required levels of quality and specificity. Within the LCI component, more detailed system definitions and assumptions emerge. Next, a life cycle impact assessment (LCIA) is performed to evaluate the potential human and environmental health effects associated with the LCI data. This assessment follows an interpretation procedure to communicate and evaluate the results, as well as provide recommendations. The rest of Section 4.2 describes each of the LCA components in more detail.
4.2.1 Goal Definition and Scope
Goal definition is the first step of an LCA, and potentially the most important.
Clearly defining the goals provides guidance throughout the study to ensure that appropriate system boundaries are defined, relevant impacts are considered, and that the results adequately represent what is sought.
4.2.1.1 Goal definition
The overall goal of this study is to understand the trade-offs and environmental impacts associated with using FGD gypsum for AMLR. In addition, this study provides a
79 resource for decision-makers to consider the broader impacts of FGD gypsum application within different scenarios and hopefully instigates discussion for future detailed investigations.
4.2.1.2 Level of specificity
This study is representative of FGD gypsum application for AMLR in Eastern
Ohio. Most of the information and assumptions in this study are specific to AEP’s
Cardinal Plant because it is the only Eastern Ohio plant currently producing FGD gypsum. Therefore items such as FGD gypsum production rates, on-site land filling operations, trucking capacity, etc. are specific to Cardinal Plant; however, such items are considered similar across all coal combustion power plants in Eastern Ohio. All information regarding reclamation activities are specific to common AML scenarios that surround Cardinal Plant. Though reclamation activities are the same across scenarios, the occurrence and details of each scenario differs along Eastern Ohio. Aggregate data is used to calculate emissions from AML equipment, transportation, and cost. More information regarding these data sources can be found in Section 4.2.3 Life Cycle
Inventory.
4.2.1.3 Functional unit
A functional unit is chosen to normalize inventory data and provide a means of comparison across scenarios. The functional unit chosen for this study is one ton of FGD gypsum produced. An alternative functional unit, one ton of coal combusted, was considered so that the results of this study could more easily be compared to results of other CCP studies. However, since the focus of this study is utilizing FGD gypsum for
80 AMLR, we can disregard the means by which it is produced since it lies outside our scope. Therefore, one ton of FGD gypsum is the more appropriate functional unit.
Outputs in relation to one ton of coal combusted can be easily calculated if necessary based on FGD gypsum production rates and coal combustion rates.
4.2.1.4 Scope and system boundaries
The system of interest includes all processes necessary for AMLR, such as the produced FGD gypsum and all materials, equipment, energy, labor, etc. necessary to apply FGD gypsum to an AML site. Figure 4.1 provides an illustration of the system and its major components. There are four major categories in the life cycle of AMLR with
FGD gypsum: raw material acquisition, manufacturing, use, and waste management.
Given that coal combustion will take place and that FGD gypsum will be produced, we designate FGD gypsum as the primary raw material. All activities involved in storing, handling, and transporting FGD gypsum are included in the raw materials acquisition process. The manufacturing process simply involves the stabilization of FGD gypsum with fly ash and lime, if necessary. The use process involves all equipment, tools, ancillary materials, labor, and energy needed for the reclamation. Finally, all equipment, labor, and transportation needed for required monitoring activities is included in the waste management process.
81 System Boundary
Raw Material Acquisition FGD gypsum Handling Storage Raw " Emissions Transportation within Materials " to air site Labor $ Manufacturing Stabilization Lime Prepare for shipment Fly Ash " Emissions Energy " Trucks to water Storage and handling $ Use Reclamation Machines Equipment Hazardous Trucks " Solid Wastes " Tools Waste Ancillary materials Labor $ Waste Management Monitoring activities Labor " Other Leachate collection Materials Releases Transportation
Figure 4.1: Major life cycle processes for AMLR with FGD gypsum
82 There are different types of AMLs in eastern Ohio and different methods for which FGD gypsum can be disposed. Since the focus of our study is on the beneficial use of FGD gypsum, we decided to model end-use scenarios that are typical for a coal combustion power plants and use those scenarios as a basis of discussion and comparison. Six different scenarios were modeled based on AML site characteristics within a defined radius of AEP’s Cardinal Plant and the reclamation processes common across all AMLR activities. Table 4.1 describes these scenarios. The AML sites were determined by a student research group at The Ohio State University based on various mapping programs. ArcMap®, AutoCAD®, and Goggle Earth™ were used to determine the location of five coal combustion power plants in Ohio (Cardinal, Conesville,
Muskingum River, Sammis, and Garvin Plants) and surface mines within a 25-mile radius from each plant. Data from ODNR was used to locate underground mines surrounding these five plants.
83 Scenario Description
1 HWP, 5 Highwall pit within a 5-mile radius of plant Average volume: 401,382 cubic yards Average area: 3.1 acres
2 HWP, 10 Highwall pit within a 10-mile radius of plant Average volume: 226,730 cubic yards Average area: 2.1
3 HW, 5 Highwall within a 5-mile radius of plant Average volume: 4,618,248 cubic yards Average area: 19.1 acres
4 UM, 5 Underground mine seal and highwall within a 5-mile radius of plant Average volume of mine entrance: 30 cubic yards Average volume of highwall: 50,000 cubic yards Average area: 2.1 acres
5 GP Abandoned gob pile Average volume: 1842 cubic yards Average area: 10 acres
6 LF Subtitle D landfill within a 1-mile radius of plant Average volume capacity: 3,136,320 cubic yards Average area: 24 acres
Table 4.1: Description of scenarios used for the LCA
84 4.2.2 Life Cycle Inventory
The Life Cycle Inventory (LCI) is a database specifically generated to describe energy and material flows throughout the chosen system (USEPA, 2006). It includes all energy and material inputs, energy and material consumptions, output emissions to the atmosphere, water, and soil, and other waste streams. An LCI is performed by first developing a flow diagram to pin point all relevant input-output streams and processes.
Next, a data collection plan is constructed that explicitly defines the data quality goals, data sources, and preparing a data collection worksheet. Once the goals and procedures have been specified, data is then collected. Finally, the data is evaluated and comparisons of scenarios can be made. The following subsections describe the methodology for each
LCI step in more detail.
4.2.2.1 Flow Diagram
In Section 4.2.1.4, the system of interest was defined and a basic input-output diagram was provided. In this section, the system diagram is expanded to include more detailed processes and output streams of interest. Figure 4.2 presents the flow of energy and materials into the AMLR process along with some sectors that provide relevant inputs into the AMLR process.
85 Clearing, Grubbing, Grading Truck manufacturing
Impoundment Heat/energy Dewatering Heavy duty equipment manufacturing Earthwork Emissions
Diesel fuel production
Channeling Waste Limestone mining and production Revegetation
Laboratory instrument manufacturing Landscape Maintenance
Figure 4.2: Flow diagram of AMLR.
86 4.2.2.2 Data Quality Goals
Site-specific data was used as the basis for creating beneficial use scenarios.
AML and landfill site characteristics were developed with respect to the geological characteristics of the area surrounding Cardinal Plant. Details regarding AMLR activities were used from ODNR records and represent common practices, loads, and prices.
Process-level emissions data was queried through USEPA’s NONROAD2008 model, which represents equipment-specific data. All aggregated life cycle emissions and energy were taken from the Eco-LCA and EIO-LCA models. These models contain the same emissions databases, but the EIO-LCA provide more details regarding economic activities.
4.2.3 Life Cycle Impact Assessment
The purpose of a Life Cycle Impact Assessment (LCIA) is to evaluate overall environmental impacts by categorizing and normalizing the LCI data. There is an important distinction between LCIAs and other impact assessments (e.g., risk assessment), where traditional impact assessments tend to focus on the process-level and
LCIAs focus on finding linkages and mapping out flows within systems (USEPA, 2006).
In this study, we used process-level data to create typical AMLR scenarios, which we then used to gather life cycle data.
The Eco-LCA data in this study was divided into 5 impact categories according to convention. Table 4.2 displays examples of the specific LCI data included in each category as well as the unit of comparison. The EIO-LCA data was categorized according to total conventional air pollutants, total toxic releases, total energy
87 consumption, and total global warming potential. The Eco-LCA data was then normalized against all U.S. sector data to uncover the contribution AMLR makes to national resource and energy flows.
Impact Category Examples of LCI Data Characterization Factor
Global Warming Carbon dioxide, nitrogen Carbon dioxide equivalents dioxide, methane, CFCs, HCFCs, methyl bromide
Stratospheric Ozone CFCs, HCFCs, halons, Trichlorofluoromethane Depletion methyl bromide (CFC-11) equivalents
Acidification Sulfur oxides, nitrogen Hydrogen (H+) ion oxides, hydrochloric acid, equivalents hydrofluoric acid, ammonia
Eutrophication Phosphate, nitrogen oxide, Phosphate equivalents nitrogen dioxide, nitrates, ammonia
Photochemical Smog Non-methane hydrocarbon Ethane equivalents
Table 4.2: Life cycle impact categories (Source: USEPA, 2006)
88 4.2.4 Life Cycle Interpretation
The final step of an LCA is to provide an interpretation and evaluation of the study. This entails reviewing the goals, LCI techniques, and stating conclusions about the overall LCI and LCIA results. In addition, a sensitivity analysis is performed to assess how the results change to variations in the data. Details regarding this methodology are discussed in Section 5.4
89 CHAPTER 5: LIFE CYCLE ASSESSMENT RESULTS AND DISCUSSION
5.1 Inventory Results
The LCI results from both USEPA’s NONROAD model and the Eco-LCA model are presented in Sections 5.1.1 and 5.1.2, respectively. The data includes total cost estimates, unit costs (cost per ton of FGD gypsum), common pollutant emissions, carbon sequestration, and energy, land, and water utilization.
5.1.1 Process Analysis Inventory Results
Table 5.1 provides the estimated cost for each AMLR sub-process (Figure 4.2) based 2004-2008 ODNR construction data. The assumption made based on ODNR averages were that each project required 1,000,000 gallons dewatered and 1,000 tons of
Type C and D rock for channels. Mobilization/access averaged 5% of the total cost of the project. Other unit costs were multiplied by corresponding values in Table 4.1. To obtain a total project cost, the cost of each sub-process was totaled. For example, HWP,5 was calculated in the following manner:
Total Cost = Mobilization/Access + Clearing/Grubbing/Grading + Dewatering +
Earthwork + Rock Channeling + Revegetation + Landscaping Maintenance
Total Cost = (0.05 * Total Cost) + ($1775/acre * 3.1 acres) + ($0.005 gal * 106 gal) +
90 ($2/yd3 * 401382 yd3) + ($31/ton * 103 ton) + ($1468/acre * 3.1 acres) +
($400/acre * 3.1 acres)
Total Cost = (0.05 * Total Cost) + $855,057
Total Cost = $894,797
AMLR Sub-Process Unit Unit Cost Mobilization/Access 5% of total cost Clearing/Grubbing/Grading acre $1,775.00 Dewatering gallon $0.005 Earthwork cubic yards $2.00 Rock channeling ton $31.00 Revegetation acre $1,468.00 Landscaping maintenance acre $400.00
Table 5.1: Average cost of each AMLR sub process based on 2004-2008 ODNR data.
Landfilling costs (LF) were estimated based on data in the 2008 RSMeans Building
Construction Cost Data (Waier et al., 2008). The following unit costs from the RSMeans
Catalog were used to calculate total LF cost: hauling cost: $1.67/yd3, dozer backfill:
$1.26/yd3, and compacting backfill: $2.35/yd3. Total project costs are divided by the amount of FGD gypsum required at each scenario, which was estimated to be the average volume of each site. 91 Table 5.2 provides the estimated costs for each scenario. As mentioned in previous chapters, on-site landfills fall under RCRA Subtitle D, which require operators to follow a particular set of siting, design, operating, closing, and monitoring regulations.
The landfill scenario used in this study assumes a landfill is already in place and does not include initial capital costs (e.g., liner systems) or closing costs (e.g., capping layers and monitoring). Landfilling is the most expensive scenario overall, however, the cost per ton of FGD gypsum is not very different from the AMLR scenarios. This is because the landfill can accept almost 6 million tons of FGD gypsum, which decreases the unit cost.
Unlike on-site landfills, the AMLR sites do not fall under RCRA Subtitle D and thus do not require as extensive siting, design, and monitoring. Scenario 3 is the most expensive project because of its large geographical size, but yields the lowest unit cost since it utilizes a substantial amount of FGD gypsum. Contrarily, scenario 5 is the cheapest project because of its relatively small size, but its unit cost is significantly higher than all other scenarios due to its relatively small quantity of FGD gypsum. The distinction between total costs and unit costs plays an important role when assessing the overall impact of AMLR scenarios and is discussed further in Section 5.3.
92 Cost per ton Scenario Total cost FGD gypsum 1 HWP, 5 $894,797 $1.24 2 HWP, 10 $607,485 $1.26 3 HW, 5 $9,833,766 $1.18 4 UM, 5 $151,274 $1.68 5 GP $80,120 $24.10 6 LF $15,731,781 $2.78
Table 5.2: Total cost of each scenario and cost per ton of FGD gypsum used.
Table 5.3 displays the inventory results from USEPA’s NONROAD database. The data represents emissions produced during AMLR and does not account for the reduced effects of reclaiming AMLs (e.g., reduced AMD and sediment discharge). The pollutants listed are total hydrocarbons (THC), total organic gases (TOG), non-methane organic gases (NMOG), non-methane hydrocarbons (NMHC), volatile organic compounds
(VOC), nitrogen oxides (NOx), carbon monoxide (CO), carbon dioxide (CO2), sulfur dioxide (SO2), and particulate matter (PM10 and PM2.5). There is a general trend in the emissions data similar to that of the cost estimate data. Scenario 5 ranks highest in all emissions by at least an order of magnitude and consumes the most amount of fuel.
Scenario 2 ranks second in fuel consumption and emissions, but is still significantly less than Scenario 5. Scenarios 1, 3, and 4 have very similar results due to coincidental similarities among the ratios of their respective operating times and amount of FGD
93 gypsum utilized. Scenario 6 consumes the least amount of fuel and emits the least amount of pollutants.
5.1.2 Eco-LCA Inventory Results
The Eco-LCA inventory data displayed in Table 5.4 describes the quantity of each pollutant emitted as a function of economic activity. In addition, it represents the aggregation of life cycle data, including natural services, reported from the entire “waste management and remediation services” sector. The values listed in Table 5.4 are directly dependent upon the cost per ton of FGD gypsum used and thus exhibit a similar trend as the data in Table 5.2. Taking into account the life cycle flow of pollutant emissions, scenario 5 still remains the highest ranked, but this time scenario 6 is ranked second.
Scenario 4 ranks third and scenarios 1, 2, and 3 exhibit very similar results.
In addition to emission data, the Eco-LCA database models the energy flow, water consumption, land usage, and carbon flow throughout the life cycle of each scenario
(Tables 5.5 and 5.6). A majority of the energy used in each scenario comes from sunlight
- a natural service - along with coal, crude oil, and natural gas. The trends among the data in Table 5.5 are equivalent to the emission data described above. In terms of land usage and water consumption, each scenario relies heavily on timber land to support its activities, and the majority of water used comes from independent power plants. Carbon sequestration is minimal for each scenario, but occurs mostly due to forests surrounding the sites.
94 Scenario: 1 2 3 4 5 6
Category Unit HWP, 5 HWP, 10 HW, 5 UM, 5 GP LF
Total Emissions tons 1.19E-06 1.77E-06 1.19E-06 1.19E-06 1.77E-05 8.60E-07
Total Fuel Used gallons 0.088 0.131 0.088 0.088 1.32 0.057
Total THC tons 6.38E-10 9.70E-10 6.50E-10 6.50E-10 9.70E-09 4.32E-10
Total TOG tons 6.61E-10 1.04E-09 6.94E-10 6.94E-10 1.04E-08 4.62E-10
Total NMOG tons 6.83E-10 1.02E-09 6.83E-10 6.83E-10 1.02E-08 4.55E-10
Total NMHC tons 6.39E-10 9.54E-10 6.39E-10 6.39E-10 9.54E-09 4.25E-10
Total VOC tons 6.82E-10 1.02E-09 6.82E-10 6.82E-10 1.02E-08 4.55E-10
Total NOx tons 9.92E-09 1.48E-08 9.92E-09 9.92E-09 1.48E-07 7.14E-09
Total CO tons 3.90E-09 1.74E-06 3.90E-09 3.90E-09 5.82E-08 2.60E-09
Total CO2 tons 1.17E-06 1.74E-06 1.17E-06 1.17E-06 1.74E-05 8.46E-07
Total SOx tons 2.51E-10 3.75E-10 2.51E-10 2.51E-10 3.75E-09 1.82E-10
Total PM10 tons 5.90E-10 8.81E-10 5.90E-10 5.90E-10 8.81E-09 4.18E-10
Total PM2.5 tons 5.72E-10 8.54E-10 5.72E-10 5.72E-10 8.54E-09 4.05E-10
Table 5.3: Inventory results from NONROAD database.
95 Scenario: 1 2 3 4 5 6 Emission (kg) HWP, 5 HWP, 10 HW, 5 UM, 5 GP LF 1,1,1-Trichloroethane 2.39E-07 2.45E-07 2.29E-07 3.25E-07 4.67E-06 5.39E-07 Ammonia 4.10E-05 4.19E-05 3.92E-05 5.56E-05 8.00E-04 9.23E-05 Carbon dioxide 1.85E+00 1.89E+00 1.77E+00 2.51E+00 3.61E+01 4.17E+00 Carbon monoxide 1.02E-01 1.04E-01 9.72E-02 1.38E-01 1.99E+00 2.29E-01 Ethanol 8.05E-05 8.23E-05 7.69E-05 1.09E-04 1.57E-03 1.81E-04 Hydrofluorocarbons 6.47E-06 6.61E-06 6.18E-06 8.78E-06 1.26E-04 1.46E-05 Lead 2.53E-05 2.59E-05 2.42E-05 3.43E-05 4.94E-04 5.70E-05 Methane 3.10E-01 3.17E-01 2.96E-01 4.20E-01 6.05E+00 6.98E-01 Mono-nitrogen oxides 6.42E-03 6.56E-03 6.13E-03 8.71E-03 1.25E-01 1.44E-02 Nitrous oxide 7.08E-05 7.23E-05 6.76E-05 9.60E-05 1.38E-03 1.59E-04 PM10 1.48E-02 1.51E-02 1.42E-02 2.01E-02 2.89E-01 3.33E-02 Styrene 3.53E-06 3.61E-06 3.37E-06 4.79E-06 6.89E-05 7.95E-06 Sulfur dioxide 2.98E-03 3.04E-03 2.84E-03 4.04E-03 5.80E-02 6.69E-03 Toluene 1.35E-05 1.38E-05 1.29E-05 1.83E-05 2.63E-04 3.03E-05 VOC 1.70E-02 1.74E-02 1.63E-02 2.31E-02 3.32E-01 3.83E-02
Table 5.4: Inventory results from Eco-LCA database.
96 Scenario: 1 2 3 4 5 6 Energy Flow (J) HWP, 5 HWP, 10 HW, 5 UM, 5 GP LF Coal 2.63E+06 2.69E+06 2.51E+06 3.57E+06 5.13E+07 5.92E+06 Crude oil 1.80E+07 1.84E+07 1.72E+07 2.45E+07 3.52E+08 4.06E+07 Detrital matter 5.55E+04 5.67E+04 5.30E+04 7.52E+04 1.08E+06 1.25E+05 Fish 2.40E+03 2.45E+03 2.29E+03 3.25E+03 4.67E+04 5.39E+03 Geothermal 7.73E+03 7.90E+03 7.38E+03 1.05E+04 1.51E+05 1.74E+04 Grass 1.77E+04 1.80E+04 1.69E+04 2.40E+04 3.45E+05 3.97E+04 Hydropotential 1.86E+05 1.91E+05 1.78E+05 2.53E+05 3.64E+06 4.20E+05 Natural gas 4.77E+06 4.87E+06 4.55E+06 6.47E+06 9.30E+07 1.07E+07 Nuclear 3.30E+05 3.37E+05 3.15E+05 4.47E+05 6.43E+06 7.42E+05 Soil erosion 3.26E+02 3.33E+02 3.11E+02 4.42E+02 6.36E+03 7.34E+02 (construction) Soil erosion (farm) 9.21E+03 9.41E+03 8.80E+03 1.25E+04 1.80E+05 2.07E+04 Sunlight (farm) 1.12E+08 1.15E+08 1.07E+08 1.52E+08 2.19E+09 2.53E+08 Sunlight (forest) 3.93E+09 4.01E+09 3.75E+09 5.33E+09 7.66E+10 8.84E+09 Sunlight (ranch) 1.54E+08 1.58E+08 1.47E+08 2.09E+08 3.01E+09 3.47E+08 Wind 1.72E+03 1.75E+03 1.64E+03 2.33E+03 3.35E+04 3.86E+03 Wood (dry) 1.15E+06 1.18E+06 1.10E+06 1.56E+06 2.25E+07 2.60E+06
Table 5.5: Energy flow results from Eco-LCA database.
97 Scenario: 1 2 3 4 5 6 HWP, 5 HWP, 10 HW, 5 UM, 5 GP LF Land Usage (ha) Crop land 4.68E-06 4.78E-06 4.47E-06 6.35E-06 9.13E-05 1.05E-05 Range and pasture land 6.42E-06 6.56E-06 6.13E-06 8.71E-06 1.25E-04 1.45E-05 Timber land 1.64E-04 1.68E-04 1.57E-04 2.23E-04 3.20E-03 3.70E-04 Urban and industrial 1.21E-07 1.24E-07 1.16E-07 1.64E-07 2.37E-06 2.73E-07 land Water Usage (L) Water (agriculture & 2.53E+00 2.59E+00 2.42E+00 3.43E+00 4.94E+01 5.69E+00 livestock) Water (powerplant) 3.99E+01 4.07E+01 3.81E+01 5.41E+01 7.78E+02 8.97E+01 Water (public supply) 1.16E+01 1.19E+01 1.11E+01 1.58E+01 2.27E+02 2.61E+01 Carbon Sequestration (tons) Carbon dioxide emitted 2.04E-03 2.08E-03 1.95E-03 2.76E-03 3.98E-02 4.59E-03 CO2 sequestered (farm) -1.01E-05 -1.03E-05 -9.61E-06 -1.37E-05 -1.96E-04 -2.26E-05 CO2 sequestered -3.06E-04 -3.12E-04 -2.92E-04 -4.14E-04 -5.96E-03 -6.87E-04 (forest) CO2 sequestered (ranch) -1.75E-06 -1.79E-06 -1.67E-06 -2.38E-06 -3.42E-05 -3.94E-06
Table 5.6: Land usage, water consumption, and carbon flow from the Eco-LCA database.
98 5.2 Impact Assessment Results
The LCI data described above were grouped together according to common LCA impact categories: global warming, stratospheric ozone depletion, acidification, eutrophication, and photochemical smog potentials. The trends found in each of these categories are similar to those in Section 5.1.2. Figure 5.1 displays the impact results of each scenario. The EIO-LCA data was used to distinguish between the direct and indirect resource flows from each major contributing sector. Last, the impact data was normalized against U.S. flows to determine how much AMLR and landfilling of FGD gypsum contributes to the national impacts.
5.2.1 Impact Categories
In general, the gob pile reclamation causes the greatest environmental impact by an order of magnitude. This is simply because the amount of FGD gypsum utilized relative to the standard costs of reclamation create a large unit cost, which then translates into greater emissions according the LCA models. Landfilling causes the next greatest environmental impact, yet is substantially less that that of the gob pile reclamation. Since the landfilling unit cost is slightly higher than that of the four other reclamation scenarios, the landfilling option results in slighter greater impacts. The other four AMLR scenarios have comparable impacts in most categories, except the underground mine seal has relatively greater global warming, acidification, and eutrophication potential.
99 Global Warming Potential
HWP, 5 9.72 HWP, 10 9.93 HW, 5 9.28 UM, 5 13.18 (a) GP 189.63 LF 21.87 0 50 100 150 200 CO2 equivalents
Ozone Depletion Potential
HWP, 5 2.57E-09 HWP, 10 2.62E-09 HW, 5 2.45E-09 UM, 5 3.48E-09 (b) GP 5.01E-08 LF 5.78E-09 0E+00 2E-08 4E-08 6E-08 CFC-11 equivalents
Acidification Potential
HWP, 5 7.60E-03 HWP, 10 7.76E-03 HW, 5 7.26E-03 (c) UM, 5 1.03E-02 GP 1.48E-01 LF 1.71E-02 0E+00 4E-02 8E-02 1E-01 2E-01 Hydrogen ion equivalents Continued
Figure 5.1: Impact categories for each scenario.
100 Figure 5.1 continued
Eutrophication Potential
HWP, 5 8.48E-04 HWP, 10 8.67E-04 HW, 5 8.10E-04 (d) UM, 5 1.15E-03 GP 1.65E-02 LF 1.91E-03 0E+00 5E-03 1E-02 2E-02 2E-02 Phosphate equivalents
Smog Potential
HWP, 5 7.09E-03 HWP, 10 7.24E-03 HW, 5 6.77E-03 (e) UM, 5 9.61E-03 GP 1.38E-01 LF 1.59E-02 0E+00 4E-02 8E-02 1E-01 2E-01 Ethane equivalents
101 5.2.2 Direct Impacts
The EIO-LCA provides more detail into the economic activity within and between sectors, which is useful for determining how much of each emission is a direct and indirect result of reclamation. Table 5.7 displays the top ten sectors that contribute to the waste management and remediation services sector in order of “direct economic percentage”. Overall, 77.1% of all economic exchanges directly affect reclamation activities. Naturally, almost all of the economic activity of the waste management and remediation services sector contributes to its own operations, where the remaining 1.8% is distributed elsewhere. The greatest direct economic activity from other sectors comes mainly from maintenance and repairs of machinery, electronic equipment, tires, and automobiles; industrial process instruments, which includes water quality monitoring technology, processes control equipment, and laboratory instruments; and natural gas distribution and petroleum refineries.
The direct economic percentage for each sector is used to distinguish between direct and indirect flows; however, it does not necessarily imply that the sectors listed in
Table 5.7 are the top contributing sectors in terms of emissions and energy consumption.
Figures 5.2 through 5.5 show the top contributing sectors for total conventional air pollutant emissions (SO2, CO, NOx, VOC, Pb, and PM10), total toxic releases (air, land, water, and transfers), global warming potential (CO2, CH4, N2O, and CFCs), and energy consumption (coal, natural gas, electricity, LPG, etc.) per million dollars of economic activity.
102 Sector Direct Economic % Total for all sectors 77.1 Waste management and remediation services 98.2 Commercial machinery repair and maintenance 84 Electronic equipment repair and maintenance 82.1 Household goods repair and maintenance 81.9 Automotive repair and maintenance, except car washes 81.8 Industrial process variable instruments 81.3 Water, sewage and other systems 78.7 Natural gas distribution 71.3 Tire manufacturing 71.3 Petroleum refineries 70
Table 5.7: Top ten economic contributors to reclamation activities.
Evidently, each impact category in Figures 5.2-5.5 has a similar set of major contributing sectors that differ slightly from the list in Table 5.7. Overall, the power generation and supply sector is the greatest contributor of direct and indirect impact flows. Truck transportation naturally ranks highest for air emissions, but only about half of such emissions directly flow though reclamation activities. Petroleum refineries and oil and gas extraction are two other main impact contributors, but most of the impacts associated with oil and gas extraction are indirectly related to reclamation.
103 Truck transportation Power generation and supply Wholesale trade Support activities for transportation Petroleum refineries Couriers and messengers Commercial machinery repair and maintenance Natural gas distribution Rail transportation
0 0.75 1.50 2.25 3.00 Air Pollutant Emissions (metric tons) Direct Total Indirect Total
Figure 5.2: Conventional air pollutants emissions (metric tons) according to direct economic percentage.
Power generation and supply Petroleum refineries Other basic organic chemical manufacturing Iron and steel mills Other basic inorganic chemical manufacturing Industrial gas manufacturing Primary nonferrous metal, except copper and aluminum Petrochemical manufacturing Oil and gas extraction
0 15 30 45 60 Total Toxic Releases/Transfers (kg) Direct Indirect
Figure 5.3: Total toxic releases and transfers (kg) according to direct economic percentage.
104 Power generation and supply Petroleum refineries Truck transportation Air transportation Oil and gas extraction Pipeline transportation Industrial gas manufacturing State and local government electric utilities Other basic inorganic chemical manufacturing
0 0.8 1.5 2.3 3.0 Total Energy Consumption (TJ) Direct Indirect
Figure 5.4: Total energy consumption (Terajoules) according to direct economic percentage.
Power generation and supply Oil and gas extraction Petroleum refineries Truck transportation Pipeline transportation Industrial gas manufacturing Air transportation State and local government electric utilities Natural gas distribution Coal mining
0 75 150 225 300 Global Warming Potential (mt CO2 eq) Direct Indirect
Figure 5.5: Global warming potential (metric tons of CO2 equivalents) according to direct economic percentage. 105 5.2.3 Normalized Impacts
The Eco-LCA model calculates the ratio of the flow of each resource in the chosen sector to the total U.S. flow. Figure 5.6 (a)-(e) illustrates how each scenario contributes to each impact category relative to the total impacts by all U.S. sectors. The trends seen in these figures are identical to the trends in Figure 5.1 (a)-(e). However, the normalized values are relatively insignificant compared to total U.S. flows, which implies that reclamation activities do not negatively impact the environment on a national scale relative to other sectors. The only impact category that might be a subject of interest is the global warming potential. Relative to other sectors reclamation activities do not contribute much to global warming, but reclamation does require the use of heavy trucks and off-road machinery that emit carbon monoxide, carbon dioxide, and methane - key contributors to total greenhouse gas emissions. It is worth noting that the data in this
LCA do not account for the positive impacts AMLR has on local environments (e.g.,
AMD abatement). Thus, it is possible that AMLR may contribute more positive impacts than negative impacts at a local scale.
Since the reason FGD gypsum is available for AMLR is because of sulfur dioxide emission regulations and new scrubbing technologies, it is interesting to note the contribution of sulfur dioxide from reclamation activities relative to the U.S. flow. It would be ironic and counterproductive to find large quantities of sulfur dioxide emissions coming from reclamation activities. Evidently, the amount of sulfur dioxide emitted through the waste management and remediation services sector is insignificant compared to national emission values, as shown in Table 5.8. This may mean that there are some
106 other sectors emitting large quantities of sulfur dioxide that dramatically affect overall sulfur dioxide emissions, or it may mean that a culmination of many sectors emitting relatively small quantities of sulfur dioxide contribute to overall emissions. In the latter case, reduction of sulfur dioxide emissions from one sector may not dramatically affect overall emissions.
107 Normalized Global Warming Potential
HWP, 5 2.88E-10 HWP, 10 2.94E-10 HW, 5 2.75E-10 (a) UM, 5 3.91E-10 GP 5.62E-09 LF 6.49E-10 0E+00 2E-09 3E-09 5E-09 6E-09 CO2 equivalents
Normalized Ozone Depletion Potential
HWP, 5 1.88E-13 HWP, 10 1.92E-13 HW, 5 1.79E-13 UM, 5 2.54E-13 (b) GP 3.66E-12 LF 4.22E-13 0E+00 1E-12 2E-12 3E-12 4E-12 CFC-11 equivalents
Normalized Acidification Potential
HWP, 5 1.31E-12 HWP, 10 1.34E-12 HW, 5 1.25E-12 (c) UM, 5 1.78E-12 GP 2.56E-11 LF 2.95E-12 0E+00 8E-12 2E-11 2E-11 3E-11 Hydogen ion equivalents Continued
Figure 5.6: Impact categories normalized against U.S. flow.
108 Figure 5.6 continued
Normalized Eutrophication Potential
HWP, 5 1.98E-13 HWP, 10 2.02E-13 HW, 5 1.89E-13 (d) UM, 5 2.68E-13 GP 3.86E-12 LF 4.45E-13 0E+00 1E-12 2E-12 3E-12 4E-12 Phosphate equivalents
Normalized Smog Potential
HWP, 5 7.78E-13 HWP, 10 7.95E-13 HW, 5 7.43E-13 (e) UM, 5 1.06E-12 GP 1.52E-11 LF 1.75E-12 0E+00 5E-12 1E-11 2E-11 2E-11 Ethane equivalents
109 Scenario SO2 (tons) Normalized SO2 (tons) 1 HWP, 5 3.27E-06 2.20E-16 2 HWP, 10 3.34E-06 2.25E-16 3 HW, 5 3.13E-06 2.10E-16 4 UM, 5 4.44E-06 2.98E-16 5 GP 6.39E-05 4.29E-15 6 LF 7.36E-06 4.95E-16
Table 5.8: Comparison of total sulfur dioxide emissions (tons) from each scenario to normalized sulfur dioxide emissions (tons).
5.3 Life Cycle Interpretation
As stated in Chapter 5, the overall goal of this study is to understand the trade-offs and environmental impacts associated with using FGD gypsum for AMLR. At the process level, we find that on-site landfilling emits the least amount of emissions and consumes the least amount of energy compared to the AMLR scenarios. However, when we broaden the scale to include all life cycle flows, on-site landfilling has relatively greater environmental impacts than AMLR, with the exception of scenario 5.
Significant differences between most AMLR scenarios is less clear on both the process and life cycle scale. Recall that data only represents the emissions that are potentially harmful to the environment and neglects the positive impacts that AMLR may have. For example, an AML site may discharge AMD and heavy sediment loads into a nearby river, which disrupts soil and aquatic ecologies. As shown in past projects 110 (Chapter 2), reclaiming the AML may likely reduce such runoff and alleviate ecological damage. At the local scale, this seems like a beneficial course of action; however, there are environmental costs associated with reclamation, both local and global. The data in this study does not capture the net environmental costs since no corresponding data exists currently.
In Table 5.2, the estimated cost of each scenario is provided along with the cost per ton of FGD gypsum used. LCAs require the use of a functional unit to provide a basis of comparison across data sets, but this can create a dilemma when choosing an ideal scenario. Naturally, stakeholders prefer low-cost and low-impact options. Since there are no notable differences in environmental impacts it is tempting to choose the cheapest unit cost scenario in order to maximize efficiency and FGD gypsum utilization; however, this does not necessarily imply that such a scenario has the cheapest total cost.
Hence, decision-makers must balance the trade-offs between efficient utilization of FGD gypsum, overall environmental impact minimization, and total project costs.
In order to balance these trade-offs well, stakeholders must clearly define their short- and long-term goals. For instance, if stakeholders are willing to pay higher short- term costs to avoid landfilling and excessive environmental impacts, a gob pile reclamation may be favorable. Contrarily, during economic down times stakeholders may be more interested in lower short-term costs that do not compromise the environment. In which case, reclaiming a dangerous highwall pit may be a more favorable option.
111 Another way stakeholders can evaluate trade-offs between scenarios, especially during uncertain economic times, is by performing a sensitivity analysis. A sensitivity analysis is actually a data quality measure to examine how the LCA results are affected by variations in input. It is valuable for determining the individual processes in need of improvement and for anticipating changes in overall costs and impacts. The following section provides more details on this analysis and how the results of this study responded.
5.4 Sensitivity Analysis
The major purpose of a sensitivity analysis is to determine which variables have the greatest affect on the overall results. Since the models used in this study are primarily based on economic activity, changes were made to various cost data to uncover the individual process that have the greatest affect. The economic changes are representative of past and potential future cost fluctuations from ODNR records. In addition, some changes were made to the scenario assumptions representing common AMLR sites surrounding AEP’s Cardinal Plant. To test the impacts of geographical differences and
FGD gypsum requirements, each scenario was altered to represent AMLR sites surrounding AEP’s Conesville Plant in Coshocton County. Table 5.9 describes each independent test performed.
The results from each sensitivity test is displayed in Table 5.10. These values represent the percent change of each environmental impact category for each scenario, where negative values denote a decrease in impact. Scenarios 1, 2, 3, 4, and 6 are mostly sensitive to changes in earthwork costs, while scenario 5 is mostly sensitive to mobilization/access costs. Changes in dewatering costs have a substantial effect on
112 scenario 4 and a moderate effect on scenarios 1 and 2. The other economic changes do not seem to alter the overall results much.
In terms of the geographical variations, only scenarios 3, 4, and 5 are notably effected. The impacts associated with scenario 3 increased by approximately 50% because of the substantial decrease in AML volume and area, which increased the unit cost of FGD gypsum. Scenarios 4 and 5 experienced the opposite effect, where their associated impacts decreased due to an increase in AML volume and area and a subsequent decrease in unit cost. Scenario 6 remained unaffected by decreases in size and most economic variations.
113 Test Description Number
Economic variations
1 Increase cost of mobilization/access from 5% to 9% of total cost
2 Increase cost of clearing and grubbing from $1,775 to $2,228 per acre
3 Increase cost of dewatering from $0.01 to to $0.1 per gallon
4 Decrease cost of earthwork from $2.00 to $1.32 per cubic yard
5 Decrease cost of revegetation from $1,456 to $877 per acre
6 Decrease cost of maintenance from $400 to $274 per acre
Geographic variations
7 HWP, 5 - average volume: 268,711 cubic yards; average area: 2.1 acres
8 HWP, 10 - average volume: 267,041 cubic yards; average area: 2.1 acres
9 HW, 5 - average volume: 102,941 cubic yards; average area: 2.1 acres
10 US, 5 - two mine seals; average area: 10 acres
11 GP - average area: 25 acres
12 LF - average volume: 1,306,800 cubic yards; average area: 10 acres
Table 5.9: Descriptions of sensitivity tests.
114 Scenario: 1 2 3 4 5 6
Test HWP, 5 HWP, 10 HW, 5 UM, 5 GP LF Number
1 5.0% 5.3% 4.4% 8.0% 11.1% 0.0%
2 0.2% 0.2% 0.1% 0.7% 6.0% 0.0%
3 11.2% 16.5% 1.0% 66.1% 1.2% 0.0%
4 -31.5% -30.6% -33.6% -20.2% 4.9% -34.1%
5 0.4% 0.7% -0.1% 2.6% -1.2% 0.0%
6 0.5% 0.8% 0.0% 3.3% 4.8% 0.0%
7 3.0% 0.0% 0.0% 0.0% 0.0% 0.0%
8 0.0% 0.9% 0.0% 0.0% 0.0% 0.0%
9 0.0% 0.0% 51.9% 0.0% 0.0% 0.0%
10 0.0% 0.0% 0.0% -16.9% 0.0% 0.0%
11 0.0% 0.0% 0.0% 0.0% -25.9% 0.0%
12 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Table 5.10: Sensitivity analysis results from economic and geographic variations.
115 CHAPTER 6: CONCLUSIONS
Given that there are over 200,000 acres of abandoned strip mines and over
600,000 acres of abandoned underground mines in Ohio and approximately $206 million needed to reclaim these lands, taking advantage of beneficial use materials is an attractive option that can decrease operation and construction costs for coal companies, coal combustion power plants, the federal and state governments, and taxpayers. The goal of this study was to understand how using the beneficial use material, FGD gypsum, for
AMLR impacts surface and ground waters, air quality, energy utilization, and the ecology of AML sites.
Of the ten AML sites reclaimed with FGD material that were reviewed in this study, a majority of the projects were successfully revegetated and mitigated the effects of sedimentation runoff, erosion, and AMD. FGD gypsum does not necessarily act well as an acidic soil buffer, but it can act as a semi-impermeable liner to prevent water from entering an AML site and prevent AMD from discharging.
The elemental composition and production rate of FGD gypsum is directly dependent upon the elemental composition of the source coal and the rate of coal combustion. FGD material from forced oxidation scrubbers is primarily composed of
116 calcium sulfate dihydrate, but will contain traces of calcium sulfite, lime or limestone, fly ash, silica, and calcite.
The TCLP, SPLP, and Kossen et al. framework were useful for characterizing the leaching potential of hazardous constituents from FGD gypsum, and all three tests resulted in concentrations well below regulatory limits. Most of the leaching activity occurred within one hour, suggesting that long-term leaching of hazardous elements is not a concern for AMLR application.
The LCA provided valuable insight into the flow of hazardous pollutants throughout the AMLR process, both directly and indirectly. In addition, it showed how
AMRL contributes to the national flow of pollutants and energy through economic activity. The process-level emissions and energy consumption are primarily a results of the heavy-duty equipment used for clearing/grubbing, channeling, earthwork, and hauling. Most of the life cycle emissions were a result of truck transportation, power generation and supply, petroleum refineries, and oil and gas extraction. In general,
AMLR contributes an insignificant amount of emissions and consumes a relatively insignificant amount of energy compared to nation resource flows.
Overall, FGD gypsum is a favorable material for AMLR because of its neutral or moderately alkaline pH that is compatible with natural soil ecologies, abundant calcium and sulfur concentrations that are essential for vegetation growth, low leaching availability of mercury and other environmentally hazardous constituents, relatively inexpensive transportation and earthwork unit costs, and reliable production by coal combustion power plants throughout eastern Ohio.
117 APPENDIX A: RAW DATA FROM LEACHING TESTS
Element Cardinal Bruce Mansfield (!g/ml) Blank 1 2 3 1 2
pH 5.8840 5.9660 5.9460 5.2210 5.1940 5.3580
Al <0.034 <0.034 <0.034 <0.034 <0.034 <0.034
As <0.006 <0.006 <0.006 <0.006 0.0093 <0.006
B 0.1402 0.1332 0.1375 0.0245 0.0265 0.0205
Ba 0.4008 0.3616 0.3576 0.2752 0.2655 0.0702
Be <0.000 <0.000 <0.000 <0.000 <0.000 <0.000
Ca 850.3571 845.5102 849.5862 675.3641 661.4446 0.0833
Cd 0.0014 0.0017 0.0019 0.0005 0.0006 <0.000
Co 0.0017 0.0014 0.0014 <0.000 <0.000 <0.000
Cr 0.0055 0.0058 0.0063 0.0064 0.0051 <0.000
Cu <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
Fe 0.3574 0.3073 0.3404 0.0201 0.0234 <0.000
Hg 0.00000975 0.00001598 0.00002859 0.00000291 0.00000230 0.00000292
K 1.8596 2.0845 2.0791 1.7996 1.7251 1.6648
Li 0.1559 0.1227 0.1178 0.0865 0.0815 0.0609
Continued
Table A.1: Raw data TCLP results
118 Table A.1 continued
Element Cardinal Bruce Mansfield (!g/ml) Blank 1 2 3 1 2
Mg 15.6472 13.7657 14.5403 0.2342 0.2372 <0.002
Mn 0.1507 0.1470 0.1515 <0.000 <0.000 <0.000
Mo 0.0021 0.0017 0.0026 0.0047 0.0024 <0.001
Na 1155.3488 1187.5111 1182.2103 1138.5456 1107.2937 1103.5404
Ni 0.0072 0.0072 0.0082 <0.000 <0.000 <0.000
P <0.015 <0.015 0.0214 0.0336 0.0238 0.0171
Pb <0.003 <0.003 <0.003 <0.003 <0.003 <0.003
S 555.0182 570.2330 584.8506 675.1316 703.1588 0.0841
Sb 0.0226 0.0212 0.0240 0.0222 0.0251 <0.005
Se <0.011 0.0154 <0.011 <0.011 <0.011 <0.011
Si 0.4224 0.4416 0.4305 0.1918 0.1976 <0.011
SO4-S 611.9267 616.1867 616.0700 727.5467 730.9900 2.0478
Sr 1.1323 1.1260 1.1343 0.6988 0.6863 0.0015
Tl <0.007 <0.007 0.0114 <0.007 <0.007 <0.007
V <0.001 <0.001 <0.001 0.0024 0.0016 <0.001
Zn 0.1742 0.1668 0.1677 0.0489 0.0630 0.0519
119 Element Cardinal Bruce Mansfield (!g/ml) Blank 1 2 1 2 3
pH 7.8470 7.9650 5.5580 5.5470 5.1640 4.2300
Al <0.034 <0.034 <0.034 <0.034 <0.034 <0.034
As <0.006 <0.006 <0.006 <0.006 <0.006 <0.006
B 0.1292 0.1303 0.0308 0.0267 0.0317 0.0213
Ba 0.1059 0.0958 0.1062 0.0955 0.0910 0.0469
Be <0.000 <0.000 <0.000 <0.000 <0.000 <0.000
Ca 524.6544 516.2165 487.8672 448.5769 445.6853 0.0944
Cd <0.000 0.0003 0.0004 0.0004 0.0006 <0.000
Co <0.000 <0.000 <0.000 <0.000 <0.000 <0.000
Cr 0.0045 0.0043 0.0044 0.0049 0.0051 <0.000
Cu <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
Fe <0.000 <0.000 0.0040 0.0327 0.0105 0.0023
Hg 0.00000392 0.00000324 0.00000402 0.00000189 0.00000225 0.00000292
K <0.373 <0.373 <0.373 <0.373 <0.373 <0.373
Li 0.0855 0.0685 0.0768 0.0677 0.0646 0.0428
Mg 1.2855 1.2960 0.2708 0.2841 0.2234 0.0074
Mn 0.0648 0.0714 0.0012 0.0013 0.0016 0.0007
Mo 0.0026 0.0029 0.0026 0.0030 0.0038 <0.001
Continued
Table A.2: Raw data from SPLP results
120 Table A.2 continued
Element Cardinal Bruce Mansfield (!g/ml) Blank 1 2 1 2 3
Na 0.8895 0.6724 0.6765 0.5914 0.5086 0.3061
Ni 0.0015 0.0022 <0.000 <0.000 0.0016 <0.000
P <0.004 <0.004 0.0294 0.0361 <0.004 0.0082
Pb <0.003 <0.003 <0.003 <0.003 <0.003 <0.003
S 409.3090 434.5583 464.6563 491.9529 498.6293 0.6639
Sb 0.0271 0.0255 0.0282 0.0284 0.0339 <0.005
Se <0.011 <0.011 <0.011 <0.011 <0.011 <0.011
Si 0.1170 0.1511 0.1560 0.2079 0.1691 <0.011
SO4-S 476.7733 486.4233 486.1067 480.9533 486.0900 14.9566
Sr 0.6279 0.6039 0.4774 0.4345 0.4301 0.0006
Tl <0.007 <0.007 <0.007 <0.007 <0.007 <0.007
V 0.0018 <0.001 <0.001 <0.001 <0.001 <0.001
Zn 0.0523 0.0404 0.0815 0.0749 0.0845 0.1283
121 Leachate Element Concentration (!g/ml) Al <0.034 As <0.006 B 0.0405 Ba 0.161 Be <0.00046 Ca 582 Cd 0.0017 Co <0.00073 Cr 0.0076 Cu 0.0059 Fe 0.660 Hg 6.03E-06 K <0.373 Li 0.0367 Mg 12.9 Mn 0.0570 Mo 0.004 Na 0.0698 Ni 0.0349 P <0.004 Pb <0.003 S 388 Sb 0.0198 Se <0.011 Si 0.201 SO4-S 473 Sr 0.668 Tl <0.007 V 0.0028 Zn 0.0750
Table A.3: Raw data from Tier 1 results
122 Element (!g/ml) pH 11 pH 9 pH 7 pH 6 pH 5 pH 4 pH 3 pH 2 Al <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 As <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 B 0.0077 0.217 0.227 0.222 0.222 0.220 0.220 0.224 Ba 0.0755 0.142 0.168 0.164 0.179 0.187 0.184 0.136 Be <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 0.0005 Ca 295 451 527 486 558 582 564 404 Cd <0.00024 <0.00024 <0.00024 <0.00024 0.0003 0.0004 0.0007 0.0015 Co <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 Cr 0.0047 0.0019 0.0043 0.0051 0.0047 0.0032 0.0037 0.0056 Cu <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 Fe 0.0022 <0.00085 <0.00085 0.205 0.0051 <0.00085 0.0011 <0.00085 Hg 6.92E-06 7.90E-05 9.55E-06 2.16E-05 8.11E-06 8.18E-06 7.32E-06 8.25E-06 K 0.6460 <0.373 <0.373 <0.373 <0.373 <0.373 <0.373 <0.373 Li <0.017 0.0270 0.0377 0.0420 0.0379 0.0413 0.0278 0.0313 Mg 0.0649 1.63 2.25 2.46 2.81 6.36 10.7 18.9 Mn 0.0058 0.0022 0.109 0.143 0.149 0.182 0.213 0.315 Mo 0.0016 0.0083 0.0049 0.0062 0.0050 0.0053 0.0096 0.0076 Na 0.0684 6.20 0.428 0.388 0.429 0.456 0.506 0.312 Ni 0.0012 <0.00091 0.0013 0.0017 0.0032 0.0038 0.0064 0.0165 P <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 Pb <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 S 238 397 414 406 391 382 374 376 Sb 0.0135 0.0283 0.0299 0.0295 0.0266 0.0263 0.0370 0.0244 Se <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 Si <0.011 0.275 0.210 0.309 0.250 0.350 0.485 1.07 SO4-S 322 495 499 493 487 472 455 422 Sr 0.339 0.593 0.698 0.644 0.745 0.782 0.771 0.578 Tl <0.007 <0.001 0.0077 0.0018 0.0051 <0.001 0.0022 0.0049 V <0.001 <0.015 <0.015 <0.015 <0.015 <0.015 <0.015 <0.015 Zn 0.0095 <0.002 <0.002 0.0077 0.0071 0.0166 0.0356 0.0345
Table A.4: Raw data from Tier 2 results
123 Element (!g/ml) 1 hour 2 hours 5 hours 8 hours 1 day 2 days 4 days 8 days pH 6.28 6.39 6.41 6.46 6.59 6.63 6.74 6.89 Al <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 <0.034 As <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 B 0.0077 0.0082 0.0084 0.0077 0.0168 0.0256 0.0307 0.0506 Ba 0.0755 0.0351 0.0170 0.0225 0.0265 0.0402 0.0280 0.0418 Be <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 <0.00046 Ca 295 134 60.3 84.5 97.7 149 96.8 149 Cd <0.00024 <0.00024 <0.00024 <0.00024 <0.00024 0.0003 <0.00024 <0.00024 Co <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 <0.00073 Cr 0.0047 0.0026 0.0017 0.0025 0.0025 0.0043 0.0041 0.0053 Cu <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 Fe 0.0022 <0.00085 <0.00085 <0.00085 <0.00085 0.334 <0.00085 <0.00085 Hg 6.92E-6 8.77E-6 6.54E-6 8.42E-6 1.27E-5 5.86E-6 4.90E-6 3.94E-6 K 0.646 <0.373 <0.373 <0.373 <0.373 <0.373 <0.373 <0.373 Li <0.017 <0.017 <0.017 <0.017 <0.017 <0.017 <0.017 <0.017 Mg 0.0649 0.0580 0.0694 0.0620 0.128 0.181 0.349 0.676 Mn 0.0058 0.0030 0.0023 0.0016 0.0066 0.0069 0.0102 0.0132 Mo 0.0016 0.0013 0.0014 <0.001 0.0013 0.0029 0.0028 0.0025 Ni 0.0012 0.0018 <0.00091 <0.00091 <0.00091 <0.00091 <0.00091 <0.00091 P <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004 Pb <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 <0.003 S 238 111 51.8 72.9 83.7 127 82.3 125 Sb 0.0135 0.0124 0.0109 0.0119 0.0130 0.0130 0.0099 0.0139 Se <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 <0.011 Si <0.011 <0.011 0.0363 <0.011 <0.011 0.0163 0.0871 0.181 SO4-S 322 143 59.1 86.7 102 162 104 165 Sr 0.339 0.155 0.0689 0.0959 0.113 0.172 0.117 0.182 Tl <0.007 <0.007 <0.007 <0.007 <0.007 0.0107 <0.007 <0.007 V <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 Zn 0.0095 0.0037 <0.002 <0.002 <0.002 <0.002 <0.002 <0.002
Table A.5: Raw data from Tier 3 results
124 APPENDIX B: SATURATION INDICES FROM VISUAL MINTEQ
Plant Sample Mineral Sat. Index Al(OH)3 (Soil) 1.558 Al2O3 0.044 Al4(OH)10SO4 2.007 Alunite 3.104 Anhydrite -0.600 Barite 1.149 Boehmite 1.270 Cardinal Plant Diaspore 2.975 Gibbsite (C) 2.108 Gypsum -0.350 Hercynite 2.977 Imogolite 1.356 Kaolinite 1.580 Plumbgummite 0.783 Alunite 0.627 Anhydrite -0.578 Barite 1.142 Bruce Mansfield Plant Diaspore 1.373 Gibbsite (C) 0.506 Gypsum -0.328
Table B.1: Saturation Indices for TCLP.
125 Plant Sample Mineral Sat. Index Al(OH)3 (Soil) 0.822 Anhydrite -0.683 Barite 0.720 Boehmite 0.534 Cardinal Plant Diaspore 2.239 Gibbsite (C) 1.372 Gypsum -0.433 Hercynite 2.956 Hydroxyapatite 3.323 Al(OH)3 (Soil) 0.153 Alunite 0.360 Anhydrite -0.704 Bruce Mansfield Plant Barite 0.739 Diaspore 1.570 Gibbsite (C) 0.703 Gypsum -0.454
Table B.2: Saturation Indices for SPLP.
126 Mineral Sat. Index Al(OH)3 (Soil) 1.501 Al4(OH)10SO4 1.927 Alunite 2.450 Anhydrite -0.679 Barite 0.875 Boehmite 1.213 Diaspore 2.918 Gibbsite (C) 2.051 Gypsum -0.429 Hercynite 3.091 Imogolite 0.904 Kaolinite 0.792
Table B.3: Saturation Indices for Tier 1.
127 Mineral Sat. Index Anhydrite -0.951 pH 2 Barite 0.682 Gypsum -0.701 Anhydrite -0.713 pH 3 Barite 0.911 Gypsum -0.463 Anhydrite -0.676 pH 4 Barite 0.942 Gypsum -0.426 Anhydrite -0.668 Barite 0.952 pH 5 Diaspore 0.920 Gibbsite (C) 0.053 Gypsum -0.418 Al(OH)3 (Soil) 1.701 Al2O3 0.329 Al4(OH)10SO4 2.480 Alunite 2.715 Anhydrite -0.688 Barite 0.956 Boehmite 1.413 pH 6 Diaspore 3.118 Gibbsite (C) 2.251 Gypsum -0.438 Hercynite 3.298 Imogolite 1.490 Kaolinite 1.565 Plumbgummite 0.174
Continued
Table B.4: Saturation Indices for Tier 2.
128 Table B.4, continued
Al(OH)3 (Soil) 1.653 Al2O3 0.234 Al4(OH)10SO4 0.279 Anhydrite -0.668 Barite 0.950 Boehmite 1.365 pH 7 Diaspore 3.070 Gibbsite (C) 2.203 Gypsum -0.418 Hercynite 2.816 Imogolite 1.227 Kaolinite 1.133 Anhydrite -0.702 Barite 0.914 Diaspore 1.161 Gibbsite (C) 0.294 pH 9 Gypsum -0.452 Hercynite 2.921 Hydroxyapatite 6.432 Pb(OH)2 0.406 Tenorite(c) 0.555 Anhydrite -0.691 Barite 0.930 Brucite 0.291 Chrysotile 6.546 Greenalite 0.992 pH 11 Gypsum -0.441 Hercynite 0.914 Hydroxyapatite 9.072 Ni(OH)2 (c) 0.054 Pb(OH)2 0.761 Tenorite(c) 0.038
129 Time Mineral Sat. Index Time Mineral Sat. Index Al(OH)3 (Soil) 1.912 Al(OH)3 (Soil) 1.957 Al2O3 0.751 Al2O3 0.842 Al4(OH)10SO4 2.673 Al4(OH)10SO4 2.075 Alunite 2.574 Alunite 1.067 Anhydrite -0.947 Anhydrite -2.012 Barite 0.574 Boehmite 1.669 1hr Boehmite 1.624 5hr Diaspore 3.374 Diaspore 3.329 Gibbsite (C) 2.507 Gibbsite (C) 2.462 Gypsum -1.762 Gypsum -0.697 Hercynite 2.448 Hercynite 2.360 Imogolite 1.072 Imogolite 0.463 Kaolinite 0.214 Plumbgummite 0.844 Plumbgummite 1.310 Al(OH)3 (Soil) 1.946 Al(OH)3 (Soil) 1.950 Al2O3 0.820 Al2O3 0.828 Al4(OH)10SO4 2.362 Al4(OH)10SO4 2.074 Alunite 1.903 Alunite 1.148 Anhydrite -1.443 Anhydrite -1.762 Barite 0.080 Boehmite 1.662 8hr 2hr Boehmite 1.658 Diaspore 3.367 Diaspore 3.363 Gibbsite (C) 2.500 Gibbsite (C) 2.496 Gypsum -1.512 Gypsum -1.193 Hercynite 2.508 Hercynite 2.314 Imogolite 0.540 Imogolite 0.531 Plumbgummite 1.196 Plumbgummite 1.113
Continued
Table B.5: Saturation Indices for Tier 3.
130 Table B.5, continued.
Al(OH)3 (Soil) 1.921 Al(OH)3 (Soil) 1.854 Al2O3 0.769 Al2O3 0.635 Al4(OH)10SO4 1.753 Al4(OH)10SO4 1.186 Alunite 0.779 Alunite 0.135 Anhydrite -1.658 Anhydrite -1.654 Boehmite 1.633 Boehmite 1.566 1d Diaspore 3.338 4d Diaspore 3.271 Gibbsite (C) 2.471 Gibbsite (C) 2.404 Gypsum -1.408 Gypsum -1.404 Hercynite 2.692 Hercynite 2.865 Imogolite 0.480 Imogolite 1.245 Plumbgummite 0.997 Kaolinite 0.767 Plumbgummite 0.675 Al(OH)3 (Soil) 1.897 Al2O3 0.722 Al(OH)3 (Soil) 1.757 Al4(OH)10SO4 1.709 Al2O3 0.442 Alunite 0.848 Al4(OH)10SO4 0.649 Anhydrite -1.369 Anhydrite -1.362 Barite 0.169 Barite 0.192 2d Boehmite 1.609 Boehmite 1.469 Diaspore 3.314 8d Diaspore 3.174 Gibbsite (C) 2.447 Gibbsite (C) 2.307 Gypsum -1.119 Gypsum -1.112 Hercynite 5.297 Hercynite 2.926 Imogolite 0.605 Imogolite 1.369 Plumbgummite 0.774 Kaolinite 1.209 Plumbgummite 0.108
131 APPENDIX C: RAW DATA FROM LIFE CYCLE INVENTORY
SCC 2270002015 2270002048 2270002066 2270002078 2270002051
EQUIP Rollers Graders Tractors/ Dumpers/ Off- Loaders/ Tenders highway Backhoes Trucks
TOG exhaust 0.141717 0.302213 0.268137 0.069426 1.662304
NMOG exhaust 0.139598 0.297694 0.264127 0.068388 1.637447
NMHC exhaust 0.130327 0.277923 0.246586 0.063846 1.528698
VOC exhaust 0.139466 0.297412 0.263877 0.068323 1.635893
PM10 exhaust 0.147930 0.286783 0.193014 0.043773 1.562571
PM25 exhaust 0.143492 0.278179 0.187224 0.042459 1.515694
TOGCrankcase 0.002732 0.006043 0.005352 0.001372 0.033246
NMOGCrankcase 0.002691 0.005953 0.005272 0.001351 0.032749
NMHCCrankcase 0.002513 0.005557 0.004922 0.001261 0.030574
VOCCrankcase 0.002689 0.005947 0.005267 0.001350 0.032718
THC exhaust 0.132446 0.282442 0.250595 0.064884 1.553555
Continued
Table C.1: NONROAD2008 raw data, construction equipment
132 Table C.1 continued
SCC 2270002015 2270002048 2270002066 2270002078 2270002051
EQUIP Rollers Graders Tractors/ Dumpers/ Off-highway Loaders/ Tenders Trucks Backhoes
CO exhaust 0.866556 1.439376 1.210869 0.272003 10.119316
NOx exhaust 1.659671 4.070327 1.302786 0.245876 30.368755
CO2 exhaust 202.348272 545.491795 129.124213 23.363858 3559.176146
SO2 exhaust 0.043527 0.117341 0.027773 0.005025 0.765618
PM exhaust 0.147930 0.286783 0.193014 0.043773 1.562571
THCcrankcase 0.002554 0.005648 0.005002 0.001282 0.031071
133 SCC 2260004026 2265004051 2270004066
EQUIP Trimmers/ Shredders Chippers/ Edgers/ < 6 HP Stump Brush Grinders Cutter
TOG exhaust 0.076225 0.042094 0.134172
NMOG exhaust 0.075568 0.038058 0.132166
NMHC exhaust 0.072356 0.036323 0.123388
VOC exhaust 0.075495 0.037655 0.132040
PM10 exhaust 0.012450 0.000475 0.104589
PM25 exhaust 0.011454 0.000437 0.101452
TOGCrankcase 0.000000 0.000000 0.002676
NMOGCrankcase 0.000000 0.000000 0.002636
NMHCCrankcase 0.000000 0.000000 0.002461
VOCCrankcase 0.000000 0.000000 0.002633
TOG Diurnal 0.000357 0.000809 0.000000 Loss
NMOG Diurnal 0.000357 0.000809 0.000000 Loss
NMHC Diurnal 0.000357 0.000809 0.000000 Loss
VOC Diurnal 0.000357 0.000809 0.000000 Loss
Continued
Table C.2: NONROAD2008 raw data, lawn and garden equipment
134 Table C.2 continued
TOG Hot Soak 0.000167 0.000057 0.000000
NMOG Hot Soak 0.000167 0.000057 0.000000
NMHC Hot Soak 0.000167 0.000057 0.000000
VOC Hot Soak 0.000167 0.000057 0.000000
TOG Running 0.000668 0.001404 0.000000 Loss
NMOG Running 0.000668 0.001404 0.000000 Loss
NMHC Running 0.000668 0.001404 0.000000 Loss
VOC Running 0.000668 0.001404 0.000000 Loss
TOG Tank 0.000771 0.001346 0.000000 Permeation
NMOG Tank 0.000771 0.001346 0.000000 Permeation
NMHC Tank 0.000771 0.001346 0.000000 Permeation
VOC Tank 0.000771 0.001346 0.000000 Permeation
TOG Hose 0.000643 0.001640 0.000000 Permeation
NMOG Hose 0.000643 0.001640 0.000000 Permeation
NMHC Hose 0.000643 0.001640 0.000000 Permeation
VOC Hose 0.000643 0.001640 0.000000 Permeation
Continued
135 Table C.2 continued
TOG Vdisplace 0.000754 0.000740 0.000000
NMOG Vdisplace 0.000754 0.000740 0.000000
NMHC Vdisplace 0.000754 0.000740 0.000000
VOC Vdisplace 0.000754 0.000740 0.000000
TOG Spillage 0.017548 0.011113 0.000000
NMOG Spillage 0.017548 0.011113 0.000000
NMHC Spillage 0.017548 0.011113 0.000000
VOC Spillage 0.017548 0.011113 0.000000
THC exhaust 0.073013 0.040359 0.125395
CO exhaust 0.395811 0.653302 0.569346
NOx exhaust 0.003134 0.005467 1.389880
CO2 exhaust 1.606740 1.664627 133.082666
SO2 exhaust 0.000331 0.000343 0.026914
PM exhaust 0.012450 0.000475 0.104589
THCcrankcase 0.000000 0.000000 0.002501
MissingCrankcase 0.000000 0.000000 0.000000
THC Diurnal Loss 0.000357 0.000809 0.000000
THC Hot Soak 0.000167 0.000057 0.000000
THC Running 0.000668 0.001404 0.000000 Loss
THC Tank 0.000771 0.001346 0.000000 Permeation
THC Hose 0.000643 0.001640 0.000000 Permeation
Continued
136 Table C.2 continued
MissingHosePerm 0.000000 0.000000 0.000000
THC Vdisplace 0.000754 0.000740 0.000000
THC Spillage 0.017548 0.011113 0.000000
Pop 238.556614 38.194060 9.527860
Act 100.285294 5.859912 13.594822
DaysInRun 365.000000 365.000000 365.000000
Population x 87073.164110 13940.831900 3477.668776 DaysInRun
HPAvg x Activity 139.260505 19.734912 841.149666 x LF
137 Emission kilograms
1,1,1-Trichloroethane 1.94E-07
1,1,1-Trichloroethane (air) 2.08E-08
1,1,1-Trichloroethane (soil) 1.73E-07
1,1,1-Trichloroethane(water) 1.43E-10
Ammonia 3.32E-05
Ammonia (air) 4.80E-06
Ammonia (soil) 2.17E-05
Ammonia (water) 3.54E-07
Carbon dioxide 1.500
Carbon monoxide 8.24E-02
Ethanol 6.52E-05
Ethanol (air) 6.34E-06
Ethanol (soil) 3.81E-07
Ethanol (water) 1.24E-07
Hydrofluorocarbons 5.24E-06
Lead 2.05E-05
Methane 2.51E-01
Mono-nitrogen oxides 5.20E-03
Nitrous oxide 5.73E-05
PM10 1.20E-02
Continued
Table C.3: Eco-LCA raw data
138 Table C.3 continued
Styrene 2.86E-06
Styrene (air) 1.65E-06
Styrene (soil) 1.07E-06
Styrene (water) 2.43E-10
Sulfur dioxide 2.41E-03
Toluene 1.09E-05
Toluene (air) 3.86E-06
Toluene (soil) 1.03E-06
Toluene (water) 8.25E-09
VOC 1.38E-02
139 Total Value Direct Direct Sector Economic Added Economic Economic $mill $mill $mill %
Total for all sectors 1.92 1.00 1.48 77.1
Waste management and remediation services 1.148564 0.604284 1.128228 98.2
Commercial machinery repair and maintenance 0.047091 0.030083 0.039535 84
Electronic equipment repair and maintenance 0.025564 0.015766 0.020987 82.1
Household goods repair and maintenance 0.014689 0.00915 0.012035 81.9
Automotive repair and maintenance, except car 0.028036 0.014547 0.022944 81.8 washes
Industrial process variable instruments 0.002684 0.001151 0.002182 81.3
Water, sewage and other systems 0.000893 0.00058 0.000703 78.7
Natural gas distribution 0.010081 0.003312 0.007186 71.3
Tire manufacturing 0.003761 0.001272 0.002681 71.3
Petroleum refineries 0.051322 0.004931 0.035914 70
Industrial gas manufacturing 0.001802 0.000977 0.001252 69.5
Broom, brush, and mop manufacturing 0.00014 0.000063 0.000097 69.3
Sawmills 0.010877 0.002607 0.007489 68.9
Continued
Table C.4: EIO-LCA raw data
140 Table C.4 continued
Other State and local government enterprises 0.007927 0.003805 0.005381 67.9
Fitness and recreational sports centers 0.000466 0.000248 0.000314 67.4
Cut stock, resawing lumber, and planing 0.001355 0.000327 0.000896 66.1
Power generation and supply 0.025316 0.01581 0.016635 65.7
Other support services 0.004646 0.003228 0.003034 65.3
State and local government electric utilities 0.002481 0.00172 0.001598 64.4
Federal electric utilities 0.00093 0.000746 0.000599 64.4
141 LIST OF REFERENCES
Adriano, D.C. (2001) Trace elements in terrestrial environments. Springer-Verlag, New York, NY.
AOAC (1988) Official Method 975.03 Metals in Plants and Pet Foods Atomic Absorption Spectrophotometric Method, First Action 1975, Final Action 1988.
Association of Analytical Communitities (1998) Official Method 975.03, Metals in plants and pet foods, Atomic Absorption Spectrophotometric Method, First Action 1975, Final Action 1988.
Ayres, R.U. (1995) Life cycle analysis: A critique. Resources, Conservation and Recycling, 14:199-223.
Babbitt, C.W. and A.S. Lindner (2008) A life cycle comparison of disposal and beneficial use of coal combustion products in Florida, Part 2: impact assessment of disposal and beneficial use options. The International Journal of Life Cycle Assessment, 13:555-563.
Bahlmann E, Ebinhaus R, Ruck W. (2004a) The effect of soil moisture on the emissions of mercury from soils. RMZ – Mater Geoenviron, 51:791–4.
Bahlmann E, Ebinhaus R, Ruck W. (2004b) Influence of solar radiation on mercury emission fluxes from soils. RMZ – Mater Geoenviron, 51:787–90.
Bigham, J.M., Kosta, D.A., Stehouwera, R.C., Beeghlyc, J.H., Fowlerb, R., Trainad, S.J., Wolfe, W.E., and W.A. Dick (2005) Mineralogical and engineering characteristics of dry flue gas desulfurization products. Fuel, 84: 1839–1848.
Butalia, T.S., Dyer, P., Stwell, R., and W.E. Wolfe (1999) FGD by-product utilization at Ohio coal mine sites. The Ohio State Extention Fact Sheet. The Ohio State University, Columbus, Ohio.
142 Butalia,T.S. and W.E. Wolfe (2000) Ohio market opportunities for utilization of Ohio flue gas desulfurization (FGD) and other coal combustion products (CCPs). The Ohio State University, Department of Civil and Environmental Engineering, Columbus, Ohio.
Carnegie Mellon University Green Design Institute. (2008) Economic Input-Output Life Cycle Assessment (EIO-LCA), US 1997 Industry Benchmark model [Internet], Available from:
Carpi A, Lindberg SE. (1998) Application of a Teflon dynamic flux chamber for quantifying soil mercury flux: tests and results over background soil. Atmos Environ, 32:873–82.
Chesner, W., Collins, R., Mackay, M., and J. Emery (2008) The User Guidelines for Waste and Byproduct Material in Pavement Construction. Last updated 28 Jul 2008
Clark, R.B., S.K. Zeto, K.D. Ritchey, and V.C. Baligar. 1999. Boron accumulation by maize grown in acidic soil amended with coal combustion products. Fuel 78:179– 185.
Clark, R.B., Ritchey, K.D., and V.C. Baligar (2000) Benefits and constraints for use of FGD products on agricultural land. Fuel, 80: 821-828.
Crowell, D.L. (1995) History of the coal-mining industry in Ohio. Ohio Department of Natural Resources, Division of Geological Survey.
Dawson, G.W. and B.W. Mercer (1986) Hazardous waste management. John Wiley & Sons.
Dick, W. (2006) Flue gas desulfurization (FGD) product as a soil amendment for mine reclamation. The Ohio State University, School of Environment and Natural Resources, Wooster, Ohio.
Dontsova, K., Lee, Y.B., Slater, B.K., and J.M. Bigham (2005) Gypsum for agricultural use in Ohio-sources and quality of available products. The Ohio State University Fact Sheet, ANR-20-05. The Ohio State University School of Natural Resources, Columbus, Ohio.
Ericksen JA, Gustin MS, Lindberg SE, Olund SD, Krabbenhoft DP. (2005) Assessing the potential for re-emission of mercury deposited in precipitation from arid soils using a stable isotope. Eviron Sci Technol, 39:8001–7. 143 Kairies, C., Schroeder, K. and C. Cardone (2006) Mercury in gypsum produced from flue gas desulfurization. Fuel, 85: 2530–2536.
Korcak, R.F., In: Wright, R.J., Kemper, W.D., Milner P.D., Power, J.F., and R.F. Korcak, editors (1998) Agricultural uses of municipal, animal, and industrial byproducts, Conservation research report no. 44. Beltsville, MD: US Department of Agriculture, Agricultural Research Service.
Kosson, D.S., van der Sloot, H.A., Sanchez, F., and A.C. Garrabrants (2002) An integrated framework for evaluating leaching in waste management and utilization of secondary materials. Environmental Engineering Science, 19(3): 159-204.
Kost, D.A., Bigham, J.M., Stehouwer, R.C., Beeghly, J.H., Fowler, R., Traina, S.J., Wolfe, W.E., and W.A. Dick (2005) Chemical and physical properties of dry flue gas desulfurization products. Journal of Environmental Quality, 34: 676-686.
Forster, P., V. Ramaswamy, P. Artaxo, T. Berntsen, R. Betts, D.W. Fahey, J. Haywood, J. Lean, D.C. Lowe, G. Myhre, J. Nganga, R. Prinn, G. Raga, M. Schulz and R. Van Dorland (2007) Changes in atmospheric constituents and in radiative forcing. In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.
Greenberg, A.E., L.S. Clesceri and A.D. Eaton (1992) In: Standard Methods for the Examination of Water and Wastewater, 18th ed. American Public Health Association, American Water Works Association and Water Environment Federation. APHA-AWWA-WEF, Washington, DC.
Gustin, M.S., Taylor, G.E., and R.A. Maxey (1997) Effect of temperature and air movement on the flux of elemental mercury from substrate to atmosphere. J Geophys Res,102: 3891–8.
Gustin MS, Beister H, Kim CS. (2002) Investigation of the light-enhanced emission of mercury from naturally enriched substrates. Atmos Environ, 36:3241–54.
Gustin, M.S., Engle, M., Ericksen, J., Lyman, S., Stamenkovic, J., and M. Xin (2006) Mercury exchange between the atmosphere and low mercury containing substrates. Appl Geochem, 21:1913–23. 144 Heebink, L.V. and D.J. Hassett (2003) Mercury release from FGD. 2003 International Ash Utilization Symposium, Center for Applied Energy Research, University of Kentucky, Paper #75. Copyright is held by the Authors.
Hendrickson, C. T., Lave, L. B., Matthews, H. S. (2006). Environmental Life Cycle Assessment of Goods and Services: An Input-Output Approach. Resources for the Future Press.
Henzel, D.S., Laseke, B.A., Smith, E.O., and D.O. Swensen (1982) Handbook for flue gas desulfurization scrubbing with limestone. Pollution technology review, no. 94. Westwood, NJ: Noyes Data Corporation.
International Energy Agency (2004) Sensitivity and uncertainty, Annex 31, Energy- Related Environmental Impacts of Buildings. Canada Mortgage and Housing Corportation, Ottawa, Ontario.
Jones, J. B. Sr., B. Wolf and H. A. Mills (1991) Microwave digestion using CEM microwave digestion system. In: Plant Analysis Handbook. Micro-Macro Publishing. Athens, GA.
Kovacs, F. and J. Molnar (2003) Basic properties of flue gas desulfurization. Acta Montanistica Slovaca, 8:16-19.
Lamminen, M., Wood, J., Walker, H., Chin, Y.P., He, Y., and S.J. Traina (2001) Effect of flue gas desulfurization (FGD) by-product on water quality at an underground mine. Journal of Environmental Quality, 30: 1371-1381.
Lehr, J.K (2000) Standard handbook of environmental science, health, and technology. McGraw-Hill: New York, NY.
National Research Council (2006) Managing Coal Combustion Residues in Mines, The National Academic Press, Washington, D.C.
NEN 7341 (1994) Leaching characteristics of soil, construction materials and wastes- Determination of the availability of inorganic constituents for leaching from construction materials and waste materials. Delft, The Netherlands: NNI (Dutch Standardization Institute).
Norton, L.D., Shainberg, I., Cihacek, L., and J.H. Edwards, In: Wallace, A. and R.E., Terry, editors (1998) Handbook of soil conditioners, Substances that enhance the physical properties of soil. New York: Marcel Dekker. 145 Nriagu, J.O and E. Nieboer (1988) Chromium in the Natural and Human Environments. Wiley-Interscience.
Ohio Department of Natural Resources (2008) Reclamation Proposal for Beneficial Use of CCB. Division of Mineral Resources Management. DNR-744-8092.
Ohio Environmental Protection Agency (1997) Management Directive: Division of Labor Regarding Traditional Waste Management Practices and Alternative Waste Disposal Practices, Memorandum for Chiefs of Division of Surface Water and Division of Solid and Infectious Waste Management.
Pavlish, J.H., Sondreal, E.A., Mann, M.D., Olson, E.S., Galbreath, K.C., Laudal, D.L, et al. (2003) Status review of mercury control operations for coal-fired power plants. Fuel Process Technology, 82:89-165.
Pekney, N.J, Martello, D., Schroeder, K., and E Granite (2009) Environmental chamber measurements of mercury flux from coal utilization by-products. Fuel, 88:890-7.
Pflughoeft-Hassett, D.F., Hassett, D.J., and K. Eylands (2007) A Comparison of Properties of FGD & Natural Gypsum Products. Agricultural & Industrial Uses of FGD Gypsum Workshop. Atlanta, Georgia: October 23, 2007.
Prusinski, J.R., Cleveland, M.W. and D. Saylak (1995) Development and construction of road bases from flue gas desulfurization material blends. Proceedings of the Eleventh International Ash Utilization Symposium. Electric Power Research Institute, Report No. TR-104657, Volume 1, Palo Alto, California.
Schroeder, K., Beatty, L, Cardone, C., Hreha, D., Kairies, C., Rohar, P., Thompson, R., and F. White (2006) Leach testing of FGD material. DOE / NETL Mercury Control Technology Conference. December 11-13, 2006
Senior, C.L., Helble, J.J., and A.F. Sarofim (2000) Predicting the speciation of mercury emissions from coal-fired power plants. Proceedings of Air Quality II: Mercury, Trace Elements, and Particulate Matter Conference. McLean, VA, September 19-21, 2000. Paper A5-2.
Singh M. and M. Garg (1997) Retarding action of various chemicals on setting and hardening characteristics of gypsum plaster at di!erent pH. Cement Concrete Res, 27:947–50.
146 Smith, C. L. (1991) Case histories in full scale utilization of fly ash - fixated FGD sludge. Proceedings of the Ninth International Ash Utilization Symposium. Electric Power Research Institute, Report No. GS-7162, Volume II, Palo Alto, California.
Sorini, S.S. (1996) Leaching Tests: Commonly Used Methods, Examples of Applications to Coal Combustion By-Products, and Needs for the Next Generation; Session I: Coal Combustion By-Products Characterization.
Spath, P.L., Mann, M.K., and D.R. Kerr (1999) Life Cycle Assessment of Coal-fired Power Production. National Renewable Energy Laboratory Prepared under Task No. BP911030.
Srivastava, R.K., Hutson, N., Martin, B., Princiotta, F., and J. Staudt (2006) Control of mercury emissions from coal-fired electric utility boilers: An overview of the status of mercury control technologies. Environmental Science and Technology, 40:1385-1393.
Stoertz, M.W., Farley, M.E., Bullock, B., Pereira, P., and B. Laverty () FGD as an impermeable cap for coal waste. Ohio University, Department of Geological Sciences, Athens, Ohio.
Taerakul, P., Lamminen, M., He, Y., Walker, H.W., Traina, S.J., and E. Whitlatch (2004) Long-term behavior of fixated flue gas desulfurization material grout in mine drainage environments. Journal of Environmental Engineering, 130(7): 816-823.
Tazawa E. (1998) Effect of self-stress on flexural strength of gypsum–polymer composites. Adv Cement Based Mater, 7:1–7.
Tesarek, P., Drchalova, J., Kolisko, J., Rovnanikova, P., and R. Cerny (2007) Flue gas desulfurization gypsum: Study of basic mechanical, hydric and thermal properties. Construction and Building Materials, 21:1500-09.
Thorneloe, S. (2002) Life-Cycle Evaluation of Coal Combustion Residues. DOE workshop, 19 Mar 2002, Pittsburgh, PA.
Townsend, T., Dubey, B., and T. Tolaymat (2006) Interpretation of Synthetic Precipitation Leaching Procedure (SPLP) Results for Assessing Risk to Groundwater from Land-Applied Granular Waste. Environmental Engineering Science, 23:239-251.
U.S. Environmental Protection Agency (2009a) Waste and Cleanup Risk Assessment glossary. Last updated 10 July 2009
U.S. Environmental Protection Agency (2006) Life cycle assessment: principles and practice. Office of Research and Development, National Risk Management Research Laboratory. EPA/600/R-06-060.
U.S. Environmental Protection Agency (2001a) Method 245.7, Mercury in Water by Cold Vapor Atomic Fluorescence Spectrometry. EPA-821-R-01-008.
U.S. Environmental Protection Agency (2001b) Life cycle engineering guidelines. Office of Research and Development. EPA/600/R-01-101.
U.S. Environmental Protection Agency (2001c) Method 245.7, Mercury in Water by Cold Vapor Atomic Fluorescence Spectrometry. EPA-821-R-01-008. 2001.
U.S. Environmental Protection Agency (1997) Method 1311, Toxicity Characteristic Leaching Procedure, U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency (1994) Method 1312, Synthetic Precipitation Leaching Procedure, U.S. Environmental Protection Agency.
Waier, P.R. (editor), Babbitt, C., Baker, T., Balboni, B., and R.A. Bastoni (contributors) (2008) RSMeans Building Construction Cost Data 2008, 66th Edition. Means Building Construction Cost Data.
Wallschlager D, Turner RR, London J, Ebinghaus R, Kock HH, Sommar J, et al. (1999) Factors affecting the measurement of mercury emissions from soils with flux chambers. J Geophys Res, 104(21):859–71.
Whitfield, C. (2003) The production, disposal, and beneficial use of coal combustion products in Florida. Masters thesis, University of Florida.
Withum, J.A., Tseng, S.C., and J.E. Locke (2004) Evaluation of mercury emissions from coal-fired facilities with SCR and FGD systems. Topical Report No. 3, DE- FC26-02NT41589, CONSOL Energy, Inc., October 2004.
Wolfe, W., et al. (2009) FGD by-product utilization at Ohio coal mine sites: past, present, & future, Final Report.
148 Xin M, Gustin MS. (2007) Gaseous elemental mercury exchange with low mercury containing soils: investigation of controlling factors. Appl Geochem, 22:1451–66.
Zang, Y., Baral, A., and B.R. Bakshi (2008) Accounting for Ecological Resources in Life Cycle Assessment. Technical Report, The Ohio State University.
149