This item has the following downloads:
1 INVESTIGATION OF AMMONIA AND TRACE ELEMENTS LEACHING FROM CEMENT AND CONCRETE PRODUCTS AMENDED WITH CEMENT KILN AND COAL BOILER COMBUSTION BY PRODUCTS By JOSHUA BRADLEY HAYES A THESIS PRESENTED TO THE GRADUATE SCHOOL OF TH E UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQU IREMENTS FOR THE DEGREE OF MASTE R OF ENGINEERING UNIVERSITY OF FLORIDA 2013
2 2013 Joshua Bradley Hayes
3 To my beautiful fianc e Tamara and my Mother
4 ACKNOWLEDGMENTS I would like to thank my committee Dr. Christopher Ferraro and Dr. Chang Yu Wu, for their guidance and instruction I add special thanks to my committee chair Dr. Tim Townsend for his patience and expertise. I must thank my mother and lovely fiance Tamara Smith for the ir support Support for this research originated with the Florida Department of Transportation, Tallahassee, Florida. This was followed by support from the Hinkley Center for Solid and Hazardous Waste Management. In addition, t he support and assistance o f the management and operations staff of Separations Technologies Inc., Progress Energy and Tampa Electric Company (TECO) and CEMEX Brooksville are acknowledged and sincerely appreciated. Finally, I would like to thank my colleagues and fellow graduate st udents for their help: Lin Shou, Jun Wang, Wesley Oehmig, Max Krause, James Lloyd, and Saraya Pleasant.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 Background and Problem Statement ................................ ................................ ...... 11 Research Objectives ................................ ................................ ............................... 13 Research Approach ................................ ................................ ................................ 13 Organization of Thesis ................................ ................................ ............................ 14 2 LITERATURE REVIEW ................................ ................................ .......................... 16 Coal Fly Ash B ackground ................................ ................................ ....................... 16 Fly Ash Classifications ................................ ................................ ...................... 17 Fly Ash Production and Use ................................ ................................ ............. 18 Ammoniated Fly Ash Production ................................ ................................ ...... 19 Changes to NOx Controls and Ammonia Adsorption on Fly Ash ...................... 20 Leaching Characteristics of Ammon iated Fly Ash ................................ ................... 21 Conclusions ................................ ................................ ................................ ............ 21 Filter Dust Shuttling ................................ ................................ ................................ 23 Generation and Management of Cement Kiln Dust ................................ ................ 26 EPA Regulatory Status and History of CKD ................................ ............................ 27 Trace Metals in CKD, Cement and Concrete ................................ .......................... 28 Leaching of CKD, Cement and Concrete ................................ ................................ 32 Semi infinite model of diffusion ................................ ................................ ............... 38 Conclusions ................................ ................................ ................................ ............ 40 3 LEACHING OF AMMONIA FROM CONCRETE AMENDED WITH AMMONIATED COAL FLY ASH ................................ ................................ ............. 42 Introduction ................................ ................................ ................................ ............. 42 Materials and Methods ................................ ................................ ............................ 44 Sample Collection and Processing ................................ ................................ ... 44 Characterization of Amm oniated Fly Ash ................................ ......................... 46 Leaching Methodology ................................ ................................ ..................... 47 Data Analysis and Risk Assessment ................................ ................................ 48
6 Results and Discussion ................................ ................................ ........................... 51 Fly Ash Characterization ................................ ................................ .................. 51 Monolithic Leaching ................................ ................................ .......................... 53 SPLP Results ................................ ................................ ................................ ... 59 Conclusions ................................ ................................ ................................ ............ 62 4 LEACHING OF TRACE METALS FROM CONCRETE AMENDED WITH CEMENT KILN BAGHOUSE F ILTER DUST ................................ .......................... 63 Introduction ................................ ................................ ................................ ............. 63 Material and Methods ................................ ................................ ............................. 65 Sample Collec tion and Processing ................................ ................................ ... 65 Elemental Analysis ................................ ................................ ........................... 68 Leaching Methodology ................................ ................................ ..................... 69 Risk Assessment ................................ ................................ .............................. 72 Results and Discussion ................................ ................................ ........................... 72 Elemental Analysis ................................ ................................ ........................... 72 Monolith Leaching ................................ ................................ ............................ 75 Synthetic Precipitation Procedure ................................ ................................ .... 78 pH Dependent Leaching Procedure ................................ ................................ 80 Multiple Extraction Procedure ................................ ................................ ........... 85 Conclusions ................................ ................................ ................................ ............ 88 5 SUMMARY AND CONCLUSIONS ................................ ................................ .......... 90 Summary of Work Conducted ................................ ................................ ................. 90 Major Conclusions ................................ ................................ ................................ .. 90 Additional Research Needs ................................ ................................ .................... 91 LITERATURE CITED ................................ ................................ ................................ .... 93 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 97
7 LIST OF TABLES Table page 2 1 Composition of coal fly ashes by coal source ................................ ..................... 17 2 2 List of researchers previously and currently researching AFA ............................ 23 2 3 Elemental and anion variation in U.S. cement kiln dust ................................ ...... 33 2 4 Maximum concentration of contaminants for toxicity characteristic metals ......... 34 2 5 Total concentration of trace metals in CKD and cement ................................ ..... 35 2 6 TCLP results from cement and CKD ................................ ................................ .. 35 3 1 Collected fly ash measured ammonium content, LOI and moisture content ....... 52 3 2 Mill report on chemical composition of fly ash collected from Crystal River power complex ................................ ................................ ................................ ... 53 3 3 SPLP analysis of crushed concrete amended with AFA ................................ ..... 60 3 4 Influence of pH on the SPLP unionized ammonia (NH 3 ) concentrations from the crushed concrete samples amended with AFA ................................ ............. 61 4 1 Energy dispersive spectroscopy of BFD and cement samples ........................... 73 4 2 Mercury co ncentration and speciation of concrete components ......................... 74 4 3 Total trace metals concentration in BFD, cement and aggregate ....................... 75 4 4 SP LP results for BFD, cement, and cement/BFD blend samples ....................... 79 4 5 SPLP of crushed concrete samples amended with BFD sample D .................... 79
8 LIST OF FIGURES Figure page 2 1 Diagram of a Cement Kiln with an Alkali Bypass ................................ ................ 30 3 1 Sequential Extraction to Determine Ammonium Content of Fly Ash Samples .... 51 3 2 Mass Fraction of Ammonium Released from Concrete Monoliths Amended with AFA ................................ ................................ ................................ ............. 54 3 3 Mass Release of Ammonium from Concrete Monoliths Amended with AFA as a Function of Leaching Time ................................ ................................ .............. 56 3 4 Leachability Index of Ammonium in Concrete Monoliths Amended with AFA as a Function of Leaching Time ................................ ................................ .......... 57 3 5 Maximum Concentration of Ammonia Released from Concrete Monoliths Amended with AFA and Derived Leachability Threshold (LH) ............................ 59 4 1 Cumulati ve Aqueous Phase Release of Barium from Concrete Monoliths as a Function of Leaching Time ................................ ................................ .............. 77 4 2 pH Dependent Leaching Procedure Results on BFD Sample D and Comparison to GCTL ................................ ................................ .......................... 83 4 3 pH Dependent Leaching Procedure Results on Concrete Amended with BFD Sample D and Comparison to GCTL ................................ ................................ .. 84 4 4 pH Dependent Leaching Test Re sults on Concrete Amended with BFD Sample D and Comparison to GCTL ................................ ................................ .. 85 4 5 Multiple Extraction Procedure Results on BFD Sample D and Comparison to GCTL ................................ ................................ ................................ .................. 87 4 6 Multiple Extraction Procedure Results on Concrete Amended with BFD Sample D and Comparison to GCTL ................................ ................................ .. 88
9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering INVESTIGATION OF AMMONIA AND TRACE ELEMENTS LEACHING FROM CEMENT AND CONCRETE PRODUCTS AMENDED WITH CEMENT KILN AND COAL BOILER COMBUSTION BY PRODUCTS By Joshua B radley Hayes August 2013 Chair: Timothy Townsend Major: Environmental Engineering Sciences C oal fired power plants are more frequently utilizing air pollution control technologies that result in fly ash with elevated concentrations of ammonia ; referred to as ammoniated fly ash (AFA) Additionally, recent harmonization and changes to the American Society for Testing and Materials (ASTM) and American Association of State Highway and Transportation Officials (AASHTO) standards for inorganic processing add itions ( ASTM C150, ASTM C465 and AASHTO M3 27 ) has allowed the addition of calcined byproducts, e.g. baghouse filter dust (BFD) or cement kiln dust (CKD), to the final cement product at cement kilns. T he objectives of this thesis were to identify potential environmental health concerns of using ammoniated fly ash (AFA) and baghouse filter dust (BFD) in cement/ concrete products due to leaching of ammonia and trace metals. Leaching tests on concrete monoliths amended with AFA showed a substantial propor tion o f initial ammonia content, 18 37%, leached from concrete; while batch leaching of crushed concrete and fly ash showed most or nearly all ammonia leached, 58 68% and ~100%, respectively. Based on risk based regulatory thresholds (Florida
10 Surface Water C leanup Target Levels), a leachability risk estimate for the protection of surface waters was made for AFA used in concrete applications at 350 mg NH 4 + /kg AFA. Analysis of the elemental composition of BFD showed similar composition of bulk elements to cemen t collected and historical ranges for trace metals in CKD. Since only 5% by mass of BFD is allowed in cement, it is unlikely that BFD additions will affect the leaching or composition of cement/concrete products significantly.
11 CHAPTER 1 INTRODUCTION Background a nd Problem Statement It has been a standard practice to utilize coal fly ash in cement and concrete products since the 1930 s (Kosmatka et al., 2002). The addition of coal fly ash to Portland cement concrete (PCC) improves its workability, reduces segregat ion, bleeding, heat evolution and permeability, inhibits alkali aggregate reactions and enhances sulfate resistance as well as reducing costs and increasing beneficial reuse (US DOT, 2011). Additionally, recent harmonization and changes to the American So ciety for Testing and Materials (ASTM) and American Association of State Highway and Transportation Officials (AASHTO) standards (ASTM C150, ASTM C465, and AASHTO M327) for inorganic processing additio ns has allowed the addition of cement kiln d ust (CKD) o r similar calcined byproducts ( e.g. baghouse filter dust (BFD) ) to the final cement product at cement kilns. Both materials, coal fly ash and cement kiln dust, have been exempted as hazardous wastes by the United States Environmental Protection Agency (EPA contentious issue in some cases. In recent years, coal fired power plants are more frequently utilizing air pollution control technologies that result in fly ash with elevated concentrations of ammo nia (referred to herein as ammoniated fly ash, AFA). To meet regulatory mandates for nitrogen oxide (NO x ) control, coal fired power plants often employ a technology known as selective catalytic reduction (SCR ) to mitigate NO x emissions. In the SCR proces s, ammonia is injected into the combustion exhaust prior to a catalytic reactor where the
12 ammonia reacts with the NO x to form nitrogen gas. Usually, when the SCR unit is staged upstream of the particulate control devices (e.g., baghouses or ESPs), excess unreacted ammonia will be captured with the fly ash in the particulate control device. Additionally, ammonia injection is commonly used to improve particulate collection in air pollution control devices. This process is referred to as flue gas conditioning and has an added benefit of reducing plume opacity. Both practices can lead to elevated levels of ammonium salts adsorbed to the fly ash (FDEP, 2011; Kosson et al., 2009; Rathbone and Robl, 2003) Release of adsorbed ammonia from fly ash utilized in concr ete mixtures has raised concern for environmental and public health and a n evaluation of potential release is needed. Cement kiln operators employ a variety of air pollution control techniques to limit emissions from the facility and to capture particulat e materials for return to the manufacturing process. Baghouse s or other particulate control devices such as electrostatic precipitators (ESPs), are used at various locations in the manufacturing train to collect dust particles produced in the process. In some cases, these materials, referred to herein as baghouse filter dust (BFD) are extracted from the system and are recycled back into the cement production process. However, at times they may be added and mixed with the final cement product, a technique herein referred to as dust shuttling. Dust shuttling has been proposed as a strategy for limiting the atmospheric emissions of mercury from the cement calcining process. Concerns over mercury in cement manufacturing have been raised because coal combusti on facilities have in recent years been required to remove more mercury from their emissions, and thus
13 mercury levels become elevated in the byproducts such as coal fly ash, which in turn is increasingly used as a feed mineral in cement kilns. The process of dust shuttling allows mercury to be removed from the system by transferring the CKD to the final cement product, thus reducing emissions to the atmosphere. Given that a major intent of dust shuttling is to decrease mercury emissions from the stack and as a result, this practice increases mercury output with the cement product, questions have been raised regarding the potential implication of the increased mercury in the concrete used in construction projects. Additionally, given that the dust may cont ain elevated concentrations of other trace metals (Ag, As, Ba, Cd, Cr, Hg, Pb, Se, V, and Zn) of environmental concern, research on possible impacts to cement products containing BFD merits evaluation. Research Objectives The objectives of the research pre sented in this thesis are: To determine leaching of ammonia (NH 3 ) and selected trace metals (Ag, As, Ba, Cd, Cr, Hg, Pb, Se, V, and Zn) from concrete monoliths amended with AFA or BFD. To determine the leaching of ammonia (NH 3 ) and selected trace metals (A g, As, Ba, Cd, Cr, Hg, Pb, Se, V, and Zn) from crushed concrete amended with AFA or BFD. Assess the risk associated with the leaching of ammonia and trace metals from concrete and crushed concrete amended with AFA and BFD. Research Approach Samples of AFA and BFD were collected from the Crystal River Power Complex and Brooksville cement kiln. The AFA or BFD were assessed for risk by utilizing a research approach that looked at three aspects of the combustion byproducts ; composition, use and reuse. This was accomplished by an initial characterization of the combustion byproducts, investigation into an intended use scenarios in the built
14 environment (concrete monoliths), and a reuse scenario (crushed concrete). The analysis results f or total trace metals comp osition (EPA Method 3050b) of the BFD and total ammonia content of AFA, and leaching assessments of concrete amended with combustion byproducts and crushed concrete amended with AFA or BFD were compared to risk based regulatory thresholds, i.e. Florida Cle anup Target Levels (CTL) to assess risk to human health and the environment due to leaching The total ammonia content of fly ash collected was determined by sequential extractions using deionized water in a zero headspace extractor (ZHE). Bulk elemental composition of BFD was determined by Scanning Electron Microscopy Energy Dispersive X Ray Spectroscopy (SEM EDS), while trace element composition of BFD was determined by a strong acid digestion (EPA Method 3050b) and Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP AES). The i ntended use scenario (concrete monoliths) leaching potential was assessed by utilizing a monolith leaching procedure (EPA Method 1315). A reuse scenario was leaching potential was assessed by utilizing batch leaching t ests under equilibrium conditions. O rganization of Thesis Chapter 2 of this thesis will present information from material publi shed and consider s poli cies and practices with regard to CKD /BFD and AFA use (Chapter 2 Literature Review) Additionally, t he c urrent regulatory limits wi th respect to ammonia/ammonium and trace metals were reviewed and summarized In Chapter 3 the experimental methodology and results from the study of use of AFA in concrete mixes are pre sented and discussed. While in Chapter 4 the methodology and results from the study of CKD use as an inorganic processing addition are presented and
15 discussed. Chapter 5 presents a summary of findings from both studies and conclusions drawn. The final section include s a biographical sketch of the author.
16 CHAPTER 2 LITERATURE REVIEW C hapter 2 provides an overview of exi s ting literature regarding the production, disposal and management, composition, and leaching of ammoniated coal fly as h and cement kiln dust. Review of existing research of leaching of trac e metals from concrete products is also included. Chapter 2 begins with a review of literature regarding coal fly ash and ammoniated coal fly ash (AFA) and conclusions that can be drawn from existing literature. Next, existing pertinent literature regardin g cement kiln dust (CKD) is present ed. Coal Fly Ash Background Coal fly ash is produced from the combustion of pulverized coal for electricity generation. The composition of coal fly ash by coal type can be seen in Table 2 1 (US EPA, 2011a) Fly ash is car ried with the flue gas from the boiler to where it is collected in the baghouse or electrostatic precipitator. Fly ash is composed mostly of silica with nearly all particles spherical in shape. Fly ash is a pozzolan, a siliceous material which in the prese nce of water will react with calcium hydroxide at room temperature to produce cementitious compounds, e.g. see Equation 2 1. ( 2 1) Because of its spherical shape and pozzolanic properties, fly ash is useful in cement and concrete applications. Fly ash has been used in the past as an additive, feedstock or compone nt of concrete products, grout, cement, fill er material for structural applications and embankments, waste stabilization and/or solidification, soil modification
17 and/or stabilization, flowable fill, road bases, sub bases, pavement, and mineral fill in asph alt (US EPA, 2011a). Table 2 1 Composition of coal f ly a shes by c oal s ource Component Bituminous Sub bituminous Lignite SiO 2 20 60 40 60 15 45 Al 2 O 3 5 35 20 30 10 25 Fe 2 O 3 10 40 4 10 4 1 5 CaO 1 12 5 30 15 40 MgO 0 5 1 6 3 10 SO 3 0 4 0 2 0 10 Fly Ash Classifications By the American Society for Testing and Materials (ASTM) definitions, specifically ASTM C618, there are two classes of fly ash, Class C and Class F, whic h can be used as additive s in the production of Portland cement or as a partial replacement of Portland cement in concrete mixtures (ASTM, 2011). These classes of fly ash differ in silica, alumina, and iron oxide content. Class F fly ash should have a comb ined silica, alumina, and iron oxide content equal to or greater than 70% wt. while Class C fly ash should have a combined silica, alumina, and iron oxide content between 50 70% wt. (ASTM, 2011). These constituent concentrations are largely influenced by the chemical composition of the coal, i.e. bituminous or sub bituminous coal, burned to produce them. The state of Florida produces Class F fly ash in abundance and this is the primary class of fly ash added to concrete mixtures in the state; however, som e Class C fly ash is utilized as well. Additionally, to meet these standards the fly ash must have a fineness of 45 m or less and have a loss on ignition (LOI) of less than 6%. Class F fly ash is typically produced from bituminous coal and rarely exhibit s
18 cementitious properties when mixed alone with water. Conversely, Class C fly ash is typically produced from sub bituminous coal and exhibits pozzolanic activity when mixed with water. Class F fly ash produces a more stable and durable concrete product th an Class C fly ash. It should be noted that not all fly ashes meet the ASTM C618 definitions and these definitions do not contain environmental regulations. Fly Ash Production and Use In 2011, approximately 11.8 million tons of coal fly ash was used in th e production of Port land cement concrete (ACAA, 2011 ). The addition of coal fly ash to Portland cement concrete (PCC) improves its workability, reduces segregation, bleeding, heat evolution and permeability, inhibits alkali aggregate reactions and enhances sulfate resistance as well as reducing costs and increasing beneficial reuse (US DOT, 2011). Fly ash improves workability because of the spherical shape of the fly cement paste to improve the workability of the concrete product. Segregation and bleeding of binder materials from aggregate in concrete is reduced with the addition of fly ash because of its fineness and uniformity. It should be noted that the original mo tivation for including fly ash in cement wa s because of its slower reaction rate and therefore slower heat evolution. This slower heat evolution is desirable when a large concrete structure is built, e.g. Hoover dam, because internal heating can become dan gerous to the structural integrity of the concrete. The addition of Class F fly ash can reduce the prevalence of alkali aggregate reactions ; thereby creating a more stable and durable concrete product However, care must be taken with the addition of Class C fly ash since it can sometimes contain alkali metals and can actually increase the alkali aggregate reaction. The alkali aggregate reaction occurs over time in concrete between
19 the highly alkaline cement paste and reactive non crystalline ( amorphous ) si lica which is found in many common aggregates The alkali aggregate reaction is the same as the Pozzolanic reaction but in the opposite direction of Equation 2 1 (US DOT, 2011). By forcing the p ozzolanic reaction in reverse, a swelling gel of H 2 SiO 4 2 and alkali metals is produced which causes cracking in the concrete. Sulfate resistance refers to the resistance of the concrete to the transport of sulfate through its pore structure. This influx of sulfate can cause the in situ production of ettringite whic h swells inside the concrete structure. Resistance to the flux of sulfate will result in reducing the cracking due to this expansion. This is similar to the alkali aggregate reaction previously mentioned (US DOT, 2011). Again, it should be noted that fly a sh characteristics vary between different sources. Ammoniated Fly Ash Production Recent changes to regulations, in the form of the 2005 Clean Air Interstate Rule (CAIR; 40 CFR Parts 51, 72, 73, 74, 77, 78 and 96), of acid rain precursors, i.e. NO x and SO x have affected the quality and composition of coal fly ash (FDEP, 2011; Kosson et al., 2009; Rathbone and Robl, 2003). Typical changes to fly ash quality are an increase in ammonia content and carbon content. Carbon content corresponding to an LOI of grea ter than 6% should not be used in the production of Portland cement (ASTM, 2011). The increased ammonia concentration has led to a concern over the health and safety of workers using cement/concrete amended with ammoniated fly ash (Rathbone and Robl, 2003 ). One ash processing company (Separation Technologies, Inc.) states that, as an industry standard, fly ash with ammonia concentrations greater than 100 mg/kg (ppm) is unmarketable (Bittner et al., 2009). Meanwhile, the Rathbone and Robl
20 (2003) study indic ated that ammonia concentrations released to the air from ammoniated fly ash could be as high as 200 ppm if there was adequate ventilation. Changes to NOx Controls and Ammonia Adsorption on Fly Ash Low NO x burners (LNB), Selective Catalytic Reduction syste ms (SCR) and Selective Non Catalytic Reduction systems (SNCR) have been installed at many coal power plants to meet the NO x reduction required in the CAIR standards. These changes have resulted in higher concentration s of ammonia adsorbed to the fly ash an d/or lower boiler temperatures which result in residual unburned carbon in the fly ash (FDEP, 2011; Kosson et al., 2009; Rathbone et al., 2003). The elevated ammonia and carbon levels in the fly ash can affect the quality of the fly ash This could lead t o a shortage of appropriate fly ash for use as a pozzolan in cement production (Bittner et al., 2009). SCR and/or SNCR systems used for post combustion NO x control can increase the amount of ammonia in the flue gas due to ammonia slip, i.e. unreacted ammon ia from the NO x reduction reaction. Additionally, ammonia can be deposited on fly ash from ammonia injection for flue gas conditioning (FGC) prior to the particulate controls. The degree to which ammonia is deposited is dependent on the SO 3 content, fly as h sulfur content, alkalinity of the fly ash, the ammonia concentration in the flue gas and the ash loading in the flue gas (Bittner et al., 2009). The addition of ammonia for NO x controls can result in fly ashes with 200 2500 ppm concentrations (Bittner et al., 2009). The ratio of the concentration of ammonia adsorbed to the fly ash and the concentration in the flue gas has been reported to be approximately 50:1 (Bittner et al., 2009). This result means that flue gas ammonia concentrations, i.e. ammonia slip, higher than 2 ppm would likely result in an
21 unmarketable fly ash, i.e. > 100 ppm ammonia concentration in the fly ash. The maximum permissible limit of ammonia slip in most SCR systems, i.e. 2 ppm, is below the typical levels seen in all of the alter native NO x post combustion controls: FGC (50 ppm in flue gas/2500 ppm in fly ash), SNCR (5 20 ppm in flue gas/250 1000 ppm in fly ash) and SCR (0 5 ppm in flue gas/0 250 ppm in fly ash) (Bittner, et. al., 2009). In another study conducted by PMI, I nc., the concentration of ammonia i n fly ash from SCR was generally found to be 60 ppm, while AFA from SNCR wa s in the 230~735 ppm range (Giampa, 1999). This result agrees closely with the predictions made by Bittner et al. (2009). Leaching Characteristics of Ammoniated Fly Ash One study found that ammonia was adsorbed on fly ash as one of two salts, ammonium bisulfate and ammonium sulfate, when sulfur is present in the flue gas. Additionally, it was found that there was no significant difference between the leaching behaviors of these two salts. Ammoniated fly ash was exposed to deionized water for various intervals and the extracted ammonia mass was measured after each exposure interval. It was shown that approximately 85% of the initial ammonia was leac hed in the first ten minutes and 99% of the initial ammonia was leached after one hour of exposure to deionized water (Wang et al., 2002). It is clear that the ammoni um salts absorbed onto the surface of the fly ash is highly soluble and can be reliable le ached from the fly Conclusions Based on previou s data and studies, the following conclusions can be drawn ( Bittner et al., 2009 ; Giampa, 1999 ; FDEP, 2011; Kosson et al., 2009; Rathbone et al., 2003 ) Followi ng the installation of NO x control technology at coal fired power plants,
22 there has been an increase of ammonia adsorbed onto the fly ash captured in the baghouse s or electrostatic precipitator s (ESP). The concentration of ammonia adsorbed onto the fly ash is variable between facilities, air pollution control technology and management practices but generally ranges between 50 3000 ppm. It is unclear if these concentrations of ammonia adsorption onto fly ash are at levels that could potentially expose worker s, the public or the environment to unsafe levels of ammonia. Based on the literature review a list of researchers and institutes previously and currently researching the production, characterization, disposal and reuse of AFA has been compiled in Table 2 2 O nly the works by Rathbone et al. constitute studies directly related to the reuse of AFA in cement and/or concrete and potential exposure to ammonia. However, no studies were found which study the leaching of ammonia from concrete amended with AFA. It is necessary to quantify the release of ammonia to an aqueous environment due to leaching from concrete amended with AFA. Ammonia release should be compared to regulatory limits to determine if risk thresholds are being exceeded. In summary, it is not clea r if levels of ammonia leaching from concrete amended with AFA will reach unsafe levels. A clear relationship should be drawn between ammonia adsorbed on to fly ash and possible release since ammonia levels i n fly ash will vary between facilities. Also, it is clear that the concrete mix proportion s, e.g. fly ash replacement ratio, water to cement ratio, will affect the ammonia content of the concrete product. Finally, determining the extent of release from concrete over extended periods of time should be qua ntified.
23 Table 2 2 List of researchers previously and currently r esearching AFA Name Institute/Company Research Robert Rathbone University of Kentucky Characterization Robert Hurt Brown University Charact erization Jay R. Turner Washington University, St. Louis Adsorption mechanism Lamar Larrimore Southern C ompany Characterization Hao Wang University of Alabama Leaching Carol Cardone Department of Energy Characterization Henry Liu Freight Pipeline Comp any Leaching Jianming Wang University of Missouri Rolla, Leaching Vincent Giampa Progress Materials, Inc. Control D. Kosson Vanderbilt University Leaching Anthony Palumbo Oak Ridge National Laboratory Leaching Filter Dust Shuttling Filter dust shuttl ing is the practice of removing a certain percentage of the dust collected in the final pyroprocessing filter (usually a baghouse) prior to exhaust of the cement kiln and adding this material as an inorganic process addition to the finishing mill. This mat erial, referred to as baghouse filter dust (BFD), differs from the more commonly known waste product cement kiln dust ( CKD ) since it is a desired product for use as an inorganic processing addition and not a waste product like CKD. However, it should be no ted that BFD is produced in a similar manner as CKD and all literature found on CKD can be used as a bench mark for BFD since the latter is a novel material and little literature is available. Dust shuttling is done to reduce emissions of mercury and other trace metals, as well as reducing alkali salts build up within process equipment, by incorporating those metals and salts adsorbed to the baghouse filter dust into the final product (FDEP, 2009; Linero and Read, 2010). I t is acceptable to shuttle BFD int o the finishing mill to a maximum of 5% by mass of cement as long as the filter dust meets the ASTM compositio n specifications, i.e.
24 ASTM C150 Additionally, in the state of Florida, with generally low levels of alkali in the final product, the shuttling o f filter dust may actually enhance the performance of the cement, especially at early ages of curing. Therefore, it is now possible to implement the dust removal practice without necessarily violating ASTM C150 or equivalent standards (Taylor, 2008) This same phenomenon of volatile or semi volatile metals being adsorbed to raw materials, lime, or commercial adsorbents, has been documented in studies in Germany, Canada, Oregon, Maryland, Norway, and Switzerland where filter dust shuttling has been practiced (FDEP, 2009). The shuttling process is expected to significantly reduce the air emissions of mercury by incorporating and stabilizing mercury and other trace metals in the cement and concrete being amended with filter dust. One facility in Maryland has s een reductions of mercury emissions up to 40% from former levels (FDEP, 2009). It is assumed that most of the reduction s in air emissions are due to mercury being adsorbed to particulate collected in the baghouse and then shuttled to the final product. It is important to consider the fate of the mercury and other trace metals transferred from the air emissions into the cement and concrete via the dust shuttling process and its potential exposure to the public and environment (FDEP, 2009). The FDEP conclusio ns, with respect to the material property effects from inorganic process additions, were similarly stated in a study by the National Cooperative Highway Research Program (NCHRP) Report 607 ( Taylor 2008 ) announcing the harmonization of AASHTO M85 and ASTM C150. The ASTM standard C150 states that provisions are to be included for use of up to 5% by mass of inorganic process additions, based on information published in the NCHRP report Qualification testing is
25 required for amounts over 1% via a revised ASTM C465 or the new AASHTO specification, M327 (equivalent to the revised ASTM C465). Related changes to potential Bogue phase calculations are required to account for the use of process additions and limestone additions in the cement (PCA, 2013). Additionall y, the NCHRP study states that p rocessing additions are materials interground to aid in the manufacture or handling, or both, of a Portland cement. Previously, processing additions were limited to 1% by mass in AASHTO M85 and both ASTM C150 and AASHTO M85 required qualification of processing additions by ASTM C465. While the first processing additions used in Portland cement were organic grinding aids, inorganic processing additions, such as granulated blast furnace slag and fly ash, have since been used i n the manufacture of Portland cements to improve efficiency of manufacturing. Within the NCHRP study, laboratory tests w ere conducted on inorganic process addition dosages up to and just above maximum limits imposed by on ignition and insoluble residue. These would appear to limit maximum dosages of the materials tested to be between 3 and 8%, except for slag, which was not limited by this approach. A maximum dosage of any processing addition of 5% by mass of cement has been selected based on global practice (Taylor, 2008). To summarize, the harmonization of AASHTO M85 and AST M C150 has resulted in changes to the allowed mass percent of inorganic process additions allowed in Type I Portland cement, i.e. 5% by mass of cement. Cement manufacturers must ensure that the BFD they are using as an inorganic process addition meets the specifications of ASTM C465 or equivalent. According to the NCHRP study this would preclude the use
26 of greater than 8% by mass of cement addition of BF D due to Loss On Ignition (LOI) and insoluble residue limitation of BF D. This conclusion is an important distinction for later development of experimental methodology described within this report. Generation and Management of Cement Kiln Dust Cement kiln dust (CKD) is the fine grained, solid, highly alkaline waste removed from cement kiln exhaust gas by air pollution control devices and disposed of. CKD is removed from the production process due to its higher content of chloride and sulfate salts. This practice improves the quality of the final product cement and avoids unwanted buildup of alkali salts in the preheater or precalciner. However, some dust does not have high salt characteristics and can be considered part of the raw feed if returned to the pyro process (FDEP, 2009). It is common to classify CKD as the material that is caught in the baghouses (BHs ) or electrostatic precipitators (ESPs) and recycled into the kiln to produce more cement. Some would argue that the waste dust disposed of or removed from the production process differs from the dust reused in the production process as a raw material or a s an inorganic processing addition ( IPA ) in the finished product ( FDEP, 2009; Linero and Read, 2010 ) CKD that does not meet compositional specifications listed in ASTM C465 10 are not used as an IPA (ASTM, 2013 a ). By definition, CKD is wasted filter dust that is not returned to the production process and is typically disposed of in land based disposal units (i.e., landfills, waste piles, or surface impoundments), although some is also sold for beneficial reuse (EPA, 2013). CKD is removed from the producti on process when high alkali, chloride or sulfate concentrations build up within the CKD over time The alkali concentrations can reduce the strength and workability of concrete utilizing cement amended with CKD making it
27 unsuitable for most construction a pplications; therefore, CKD is removed from the cement production process. At some facilities an alkali bypass is installed at the back end of the kiln, see Figure 2 1, to remove volatilize d alkali salts. Historically, the majority of CKD generated was s tockpiled or landfilled at great expense to the manufacturer. CKD has beneficial uses as a soil amendment, as media for wastewater treatment, and as landfill cover soil or backfill ( Adask a, 2008 ) Some newly retrofitted cement plants are currently utilizing material which had been landfilled or stockpiled and are reintroducing it back into the kiln. However, CKD storage has not been in practice for many years and there are no known stor age piles at many cement plants ( Adaska, 2008 ) EPA Regulatory Status and History of CKD CKD is categorized by EPA as a "special waste" and has been temporarily exempted from federal haz ardous waste regulations under Subtitle C of the Resource Conservation and Recovery Act (RCRA) since 198 0 This ruling means that CKD was deemed exempt from being categorized as a hazardous waste for purposes of RCRA regulations. As required by RCRA, EPA s tudied the adverse effects on human health and the environment from the disposal of cement kiln dust. EPA found that some environmental harm results from CKD waste, and in 1993, reported these and other findings to Congress. Subsequently, Congress require d EPA to determine the appropriate regulatory framework for managing CKD waste. In 1995, EPA determined that some additional control of CKD was needed. Although current disposal practices cause some environmental damage, the EPA found that regulating ceme nt kiln dust as a hazardous waste was not appropriate.
28 classified as a non hazardous waste provided that the specific management standards are met. CKD not managed in compliance with the standards was proposed to be a standards ( 2013 ) In 2002, EPA published a notice of data availability (NODA) in the Federal Register (67 FR 48648) that it was considering a new approach to CKD management as RCRA Subtitle D (non hazardous, solid waste) and temporarily suspended the proposed RCRA Subtitle C (hazardous waste) portion of the 1999 rule for 3 to 5 years to assess how CKD management practices and state regulatory programs would evolve. Based upon its assessment EPA will either formally withdraw or will promulgate that portion of the 1999 proposed rule. As to date, no further ruling has been propos ed by the EPA with regard to decision making on CKD regulated as a hazardous (Subtitle C) or non hazardous waste (Subtitle D). T race Metals in CKD, Cement and Concrete Concentratio ns of trace metals in CKD range widely from g/kg to a few hundred mg/kg, depending on the species and location (Haynes and Kramer, 1982). A trace metal of particular concern, due to its toxicity and mobility, is mercury. A study completed by the Institute of Energy in Norway attempted to determine the distribution and fate of mercury in cement kilns and cement plants. This was done by injecting 203 Hg, a radioactive isotope tracer, into three different injection sites and then its residence time and concentration in flue gases was measured as well as the
29 concentration adsorbed t o the filter dust. Of the three injections, one was injected into the raw milled limestone, one was injected into the precalciner or preheater, and the final injection site was directly into the rotary kiln, see Figure 2 1 ( Karstensen et al., 2010 ) The a mount of the 203 Hg tracer was then measured in exhaust gas at the kiln and clinker cooler stacks. When the 203 Hg was injected into the raw milled limestone there was an immediate peak in concentration out of the stack (5% of total 203 Hg) and then a much lo wer release rate for many hours afterwards; only after two days had all the 203 Hg left the system. A similar result was seen from the two other injections but at different concentrations. Similarly, the measurements in the filter d ust follow this same curv e. The results indicate that the injected 203 Hg adsorbed onto the limestone and other raw materials in the raw mill and then was slowly released over time. Additionally, the 203 Hg was completely distributed through the system and setup an internal cycle. A dsorption of 203 Hg onto raw materials, especially limestone, is a function of temperature, i.e. higher temperature adsorbing less 203 Hg onto the raw materials due to the volatile nature of mercury (Eriksen et al., 2006).
30 Figure 2 1 Diagram of a cement k iln with an alkali b ypass The results from the Norwegian report are further echoed in a set of FDEP comments in Docket No. OAR 2002 00 51 (FDEP, 2009). Mercury is trapped by building an internal cycle withi n the kiln process, and when the raw mill is stopped, this mercury is captured in the baghouse and diverted to t he finish mills where it is incorporated in the final product. This process will also control heavy metal emissions by capturing them and return ing them to the kiln. The control technology is to manage operating temperatures, remove the dust collected in the baghouse, and introduce it to the cement (FDEP, 2009). These comments cite the earlier statements by Eriksen et al. (2006) and introduce the concept of filter dust shuttling which will be discussed in greater detail in a later section of this report. The FDEP comments went further to cite a paper by the
31 Portland Cement Association. Concentrations of mercury in the stack are sensitive to the te mperature of the particulate control device and whether the raw mill is on line, i.e. whether the flue gases pass through the raw mill before exitin g (PCA, 1992) While a large portion of the mercury and other trace metals are introduced to the kiln via ra w materials, it should be realized that the materials in the raw mill have intimate contact with the exhaust gases from the kiln which results in adsorption of volatile species onto the raw materials. Due to this, the fuel combusted can increase the concen tration of these species in the raw materials being introduced to the kiln (Linero and Read, 2010). This increase will continue until the concentration entering the process via the raw materials and fuel is equal to that emitted from the stack on a long te rm basis (e.g. annual average). It should be noted, the concentration of Hg and other trace metals entering the raw material or fuel cycle vary with resource location. The observation by those in the cement industry that volatile and semi volatile metals adsorb to filter dust has been referred to for nearly two decades For example, in a 1992 study by the PCA, it was concluded that the volatile trace metals released from the stack at a particular cement plant can be reduced by 50% by discarding twice as mu ch CKD (PCA, 1992). The r eport further stated that the single most important factor influencing heavy metal concentration in CKD is the extent to which CKD is recirculated into the pyro process; the level of heavy metals in the fuels and raw materials is l argely unimportant in determining CKD metal content on a short term basis (e.g., time less than several cycles of the raw mill).
32 Leaching of CKD, Cement and Concrete Several past studies have been performed to characterize the concentration of constitue nt elements in CKD, cement, and in concrete. A 1982 study by the U. S. Bureau of Mines concluded that CKD is a large volume material and a potential resource as a substitute for lime. Any environmental considerations are minor; it has been shown that U. S. CKD is not a hazardous waste as defined by current regulations under RCRA (Haynes and Kramer, 1982 ; PCA, 1992; Duchesne and Reardon, 1998; Shivley et al. 1986; Hillier et al. 1999; Poon et al. 1985 ). This study was the first nationwide survey and charact erization of CKD. The elemental and anion variations in U.S. CKD are presented below in Table 2 3 Major elements, i.e. Al, Si, K, Ca, Ti, and Fe, were measured using a fused disk x ray spectrographic technique. All other elements were determined through a cid digestion (HNO 3 /HF) and analysis by flame atomic absorption spectroscopy (AAS). It should be noted that the acid digestion method used in this study constituted a total digestion of all minerals, including siliceous mineral which are typically consider ed inert and do not release their bound trace constituents. The CKD from 72% of the existing U.S. cement plants was assessed using the EPA Extraction Procedure (EP) toxicity test for its hazardous waste potential, with respect to concentrations of the eigh t metals listed in Table 2 4 All but one sample was in compliance with the EP toxicity limits (see Table 2 4 below); the non complying sample slightly exceeded the EP toxicity test criterion for lead. While this analysis is incomplete with respect to le achability, it does provide a bench mark of CKD composition for comparison of current samples of CKD. In a 1992 PCA study, the characterization, production and current composition of CKD was updated from the 1982 study (PCA, 1992). Similar to the 1982 rep ort,
33 approximately 70% of cement plants were surveyed and tested for both total (acid soluble) and leachable heavy metal concentration (EPA Method 1311: Toxicity Characteristic Leaching Procedure (TCLP)) in CKD and cement. The results showed no consistent correlation between total metals and leachable metals. That is, samples containing the highest total metals concentration did not necessarily produce the highest TCLP results (PCA, 1992). Moreover, samples containing the highest level of total metals leach ed only moderate amounts of metals. It was concluded that none of the cement samples exceeded the RCRA limits, i.e. Toxicity Characteristic (TC), for any of the metals. However, TCLP results for CKD from one facility produced samples exceeding TC limits fo r selenium and lead. The report stated that this result can be explained by the fact that the facility in question recycles 100% of their CKD back into the pyro process. Table 2 3 Elemental and anion v ariat ion in U.S. c ement k iln d ust Element or Anion Range mg/kg Mean mg/kg Median mg/kg Ag < 3 17 5.4 4.8 Al 9900 50200 23200 23100 As 1.3 518 24 9.3 Ba < 55 < 55 < 55 Be < 2 < 2 < 2 Bi < 50 < 50 < 50 Ca 106000 367000 295000 305000 Cd < 1.5 352 21 7.3 Co < 10 < 10 < 10 Cr 11 172 41 34 Cu 7 206 30 24 Fe 1000 44400 14700 14100 Hg < 0.13 1.0 < 0.13 < 0.13 K 3400 232000 36600 26800 Li < 4 76 18 16 Mg 1980 19100 7820 6820 Mn 63 2410 383 280 Mo < 50 < 50 < 50 Na 495 27700 4700 3190
34 Table 2 3. Continued Element or Anion Range mg/kg Mean mg/kg Median mg/kg Ni < 12 91 22 29 Pb 17 1750 253 148 Sb < 1.6 70 3.2 < 1.6 Si 26900 111000 63500 65100 Sn < 100 < 100 < 100 Sr 62 8750 670 430 Ti 500 2900 3530 1100 Tl < 60 185 < 60 < 60 V < 100 < 100 < 100 Zn 32 8660 462 167 Br < 200 < 200 < 200 Cl < 100 123000 6900 4900 F 100 3600 1300 1000 NO 3 200 16700 < 200 < 200 PO 4 200 1600 < 200 < 200 SO 4 4100 316000 77800 68600 Table 2 4 Maximum concentration of contaminants for toxicity characteristic m etals EPA Hazardous Waste Number* Contaminant Maximum Concentration mg/L D004 Arsenic 5.0 D005 Barium 100.0 D006 Cadmium 1.0 D007 Chromium 5.0 D 008 Lead 5.0 D009 Mercury 0.2 D010 Selenium 1.0 D011 Silver 5.0 *Pursuant to 40 CFR 261.30: Lists of Hazardous Wastes In the interim, between the report by Haynes and Kramer and the PCA study, changes in raw materials, fuels, processes, and test techn iques occurred. Surprisingly, the PCA report states that there is no significant change in concentrations of metals in CKD tested. Additionally it was stated that lead and chromium proved to be of greater interest than other metals because they are present in CKD and cement in measureable levels. Many of the other metals, while present, are at levels that are unlikely to cause a
35 hazardous material classification by TC. Totals and TCLP results can be seen in Table 2 5 and Table 2 6 respectively (PCA, 1992) The cement did not leach constituent metal to the degree to which the CKD did. While constituent concentrations of elements in CKD, cement, and concrete will vary with location the average concentration listed in the 1992 and 1982 reports give a useful re ference point. Table 2 5 Total concentration of trace metals in CKD and c ement Element Total in Cement, mg/kg Total in CKD, mg/kg Min. Avg. Max. Min. Avg. Max. Mercury < 0.001 0.014 0.039 0.004 0.66 25.5 0 Selenium 0.62 NM 2.23 2.68 28.14 307.00 Thallium 0.01 1.08 2.68 1.40 43.24 776.00 Cadmium 0.03 0.34 1.12 0.1 10.3 59.60 Lead 1.0 12.0 75.0 34.0 434.0 7390.0 Antimony 0.7 NM 4.0 0.3 NM 3.4 Silver 6.75 9.20 19.90 4.80 10.53 40.70 Arsenic 5.0 19.0 71 .0 2.0 18.0 159.0 Nickel 10.0 31.0 129.0 1.0 22.0 60.0 Barium 91.0 280.0 1402.0 35.0 172.0 767.0 Beryllium 0.32 1.13 3.05 0.13 0.65 3.54 Chromium 25.0 76.0 422.0 8.0 41.0 293.0 Table 2 6 TCLP r esults f rom c ement and CKD E lement TCLP Cement, mg/L TCLP CKD, mg/L Min. Avg. Max. Min. Avg. Max. Mercury 0.0001 0.0006 0.005 0.0002 0.0018 0.0223 Selenium 0.001 0.011 0.025 0.006 0.152 1.711 Thallium 0.002 0.01 0.028 0.01 0.38 4.50 Cadmium 0.0003 0.0019 0.0 123 0.0001 0.0288 0.22 Lead 0.002 0.009 0.029 0.002 0.349 9.718 Antimony 0.003 NM 0.063 0.003 0.012 0.031 Silver 0.003 0.07 0.12 0.03 0.07 0.17 Arsenic 0.005 0.027 0.084 0.003 0.066 0.636 Nickel 0.06 NM 0.17 0.06 0.13 0.32 Barium 0.49 1.35 4.27 0.12 1.04 9.19 Beryllium 0.0001 0.0005 0.003 0.0001 0.0004 0.0029 Chromium 0.07 0.54 1.54 0.01 0.1 1.29 A complete understanding of the chemistry of hydration of cement is not available due to the complex nature of the process. However, several studies have
36 indicated that cement and CKD leachability is influenced strongly by solubility of mineral structure containing trace metals; while concrete leachability is more controlled by diffusion of metal through the pore structure of the cement matrix (Van der Slo ot, 2002; Duchesne and Reardon, 1998; Shivley et al. 1986; Hillier et al. 1999; Poon et al. 1985; Serclerat et al. 2002). In CKD, trace element mobility is correlated to the solubility of minerals during the hydration process (Duchesne and Reardon, 1998; Serclerat et al. 1999). In the study by Duchesne and Reardon, it was stated that CKD is composed of mostly oxidized, anhydrous phases. These phases include lime (CaO), arcanite (K 2 SO 4 ), and sylvite (KCl) and are unstable or highly soluble in water at stan dard temperature and pressure (Duchesne and Reardon, 1998). When CKD contacts water, these phases will either completely dissolve or more stable and less soluble phases will precipitate. The concentration of some constituent elements in CKD leachates is co ntrolled by the solubility of secondary precipitates while other elements are controlled by their availability to the leachate solution and their diffusive flux into solution from the leaching of primary phases over time (Duchesne and Reardon, 1998). To di fferentiate between these two classes of leaching behavior, leaching procedures of several liquid to solid ratios of CKD were conducted. It was pointed out that if the liquid to solid ratio is halved and the element concentration in the leachate does not d ouble then there must be a solid phase control on the elements mobility. The test results indicated that there are no solubility controls on Na, Cl, Cr, Mo, Se, and, in most instances, K. However, evidence for solubility control was seen for Si, Ca, SO 4 2 Mg, Al, Zn, Ti, Sr and Ba concentrations (Duchesne and Reardon, 1998). These results strongly suggest the dependence of
37 elemental mobility on the liquid to solid ratio. This is important to cement and concretes because differing water additions to mixes d uring hydration can affect both the mobility of trace metals and permeability of the hardened state. In a different study the effects of varying pH were observed on the leachability of three trace metals, Zn, Cr, and Pb, in mortar bars. Mortar bars are ma de from a mixture of cement, water, and sand that is then allowed to set. The study found that most of the metals were retained when exposed to deionized water. However, metal concentrations differed for some metals as the pH was varied, i.e. Zn, Cr 3+ and Pb (Serclerat et al. 1999). These results are significant but a more complete analysis of all trace metals should be done wit h varying pH values Another paper cited several sources stating that silicate mineral structures are well documented in the adso rption of amphoteric metals on to their surface (Shively et al. 1986). Additionally, amorphous iron oxy hydroxide surfaces have been shown to provide multiple adsorption sites for trace metals. Leaching tests with varying pH were also addressed in the stud y by Shively et al. Acid attacks the cement through the permeation of the pore structure and dissolution of ions back through the chemically altered surface layer into solution. This reaction is similar to the natural weathering of silica rich minerals in the environment (Shively et al. 1986). It has been reported that multiple extractions of leachate showed a decrease in metals concentration and lowering of the pH in the leachate (Shively et al. 1986; Poon et al. 1985). These results show a very differen t scenario than a single extraction technique. In the Shively et al. report it was shown that none of the samples failed the
38 TC for heavy metals despite being doped with concentrations much higher than those seen in typical cement. In the Poon et al. stu dy it was stated that the interaction of many metals with the microstructure of the hydrated cement occurs in the early stage of hardening (Poon et al. 1985). However, the mechanism of fixation, i.e. chemical reaction or physical encapsulation, has still n ot been clearly revealed and may vary for different elements. For example, the immobilization mechanism for cesium in cement is thought to be purely physical encapsulation since cesium does not form an insoluble precipitate under the alkaline conditions of cement solutions. It should be realized that elements that do not form insoluble precipitate or are not chemically bound to the cement matrix may be more mobile in cement or concretes with higher porosity, e.g. cement treated with higher portions of air e ntrainment agent (Poon et al. 1985). Two different studies, Hillier et al. (1998) and Marion et al. (2005 ) looked at the long term leachability of concrete monolithic structures. However, both studies used deionized water as their leachate in tanks witho ut multiple extractions of leachate. This is significant because as previously stated in Shively et al., the acid cement reaction in lower pH leachates reflects natural degradation of the cement matrix. Additionally, concrete structures in the environment are rarely isolated and contained in the same water for an extended period. This would cause the leachate to become highly alkaline and this could significantly affect the mobility of constituent elements. Semi infinite model of diffusion Leaching can be affected by a number of factors, including pH, temperature, chemical and physical encapsulation of analytes, and leaching solution chemistry. However, with respect to leaching of analytes from monoliths, three main mechanisms
39 that generally control the lea ching process are surface wash off, diffusion transport, and surface dissolution (Torras et al., 2011). Surface wash off is a process by which the surface concentration of the analyte on the monolith is quickly solubilized into the bulk leaching solution. Diffusion transport is the migration of analytes by random motion from higher to lower concentration through a medium (i.e., concrete). Surface dissolution is the physical solubilization of the medium into the leaching solution. A semi dynamic monolith lea ching test (e.g., Method 1315) can be used to determine the main leaching mechanism, mass transfer rate, and changes in the leaching mechanism over time (Torras et al., 2011). Calculation of the interface diffusion coefficient between the concrete and wa ter is based on the semi infinite diffusion model (Glicksman, 2000), which is applied to a diffusion process from a source (e.g., a cylinder, a plate or a slab) to an infinite bath. This model can be applied to ammonia releasing from a concrete slab into w ater. A relationship can be developed between the mass release of ammonium and the infinite diffusion model ( Kosson et al., 2009; Glicksman, 2000). (2 2 ) Where D i is the calculated diffusion coefficient for leaching interval i (m 2 /s), t i is the leaching interval time (s), M ti is the measured mass release for interval i (mg/m 2 ), is the concrete density (kg dry/m 3 ) and C o (mg/kg) is the initial concentration of ammonium in the concrete.
40 Based on experimental measurements of mass released over time, Equation 2 2 can be used to calculate the diffusion coefficient as a function of time. The measured mass release was determined by the relationship seen in Equation 2 3 ( 2 3 ) Where c meas is the measured concentration of ammonium in the extraction fluid (mg/L), v i is the volume of extraction fluid employed for leaching interval i (L) and A is the surface area of the monolith (m 2 ). Assuming the semi infinite diffusion model, it is appropriate to visualize the concrete monolith used in this study as a cylinder diffusing radially into an infinite bath of deionized water. The diffusion coefficient for ammonium can be used to predict release rates of ammonium in hardened concrete under different scenarios, e.g. concrete of different dimensions and ammonium concentrations (Kosson et al., 2009; Torras et al., 2011). Conclusions From previous studies it is clear that CKD can contain elevated levels of trace metals ( H aynes and Kramer, 1982 ; PCA, 1992) It is also clear that ATSM and AASHTO harmonization has allowed several materials to be used as inorganic processing additions i.e. gypsum, coal fly ash, blast furnace slag, CKD, or other partially calcined materials ; h owever, no environmental or public health considerations were part of the decision making. While significant work has been done with regard to leaching from cement, concrete, and CKD; clear predictions of release of trace metals cannot be made due to uncer tainty of the hydration reaction and dissolution of various solid phases ( (Van der Sloot, 2002; Duchesne and Reardon, 1998; Shivley et al. 1986; Hillier
41 et al. 1999; Poon et al. 1985; Serclerat et al. 2002 ) Additionally, no studies were found that directl y addressed the leaching of trace metals from concrete amended with CKD. Taking these considerations in mind, it is necessary to address leaching from CKD, and concrete amended with CKD.
42 CHAPTER 3 LEACHING OF AMMONIA FROM CONCRETE AMENDE D WITH AMMONIATED COAL FLY A SH Introduction In recent years, coal fired power plants have increasingly utilized air pollution control technologies that result in fly ash with elevated concentrations of ammonia, referred to herein as ammoniated fly ash, or AFA (Wang et al., 2002). T o meet regulatory mandates for nitrogen oxide (NO x ) control, coal fired power plants often employ selective catalytic reduction (SCR) to mitigate NO x emissions. In the SCR process, ammonia is injected into the combustion exhaust prior to a catalytic react or where ammonia reacts with NO x to form nitrogen gas. Usually, when the SCR unit is staged upstream of the particulate control devices (e.g., baghouses or electrostatic precipitators), excess unreacted ammonia will be captured with the fly ash in the par ticulate control device. Additionally, ammonia injection is used to improve particulate collection in air pollution control devi ces and to reduce plume opacity, a process referred to as flue gas conditioning. These processes can result in fly ash with hig h levels of ammonium salts (200 2500 ppm) adsorbed to the fly ash surface (Wang et al., 2002; Rathbone, R. and Robl, T. 200 3; Cardone, et al., 2005; Bittner et al., 2009 ). Because of i ts pozzolanic properties, fly ash is useful in cement and concrete a pplications (EPA, 2011). By the American Society for Testing and Materials (ASTM) definitions, specifically ASTM C618, there are two classes of fly ash, Class C and Class F, which can be used as an additive in the production of Portland cement or as a part ial replacement of Portland cement in concrete mixtures (ASTM, 2013b ). These classes of fly ash differ in silica, alumina, and iron oxide content. In 2011, approximately 11.8
43 million tons of coal fly ash was used in the production of Portland cement concre te (ACAA, 2011). The addition of coal fly ash to Portland cement concrete (PCC) improves its workability, reduces segregation, bleeding, heat evolution and permeability, inhibits alkali aggregate reactions and enhances sulfate resistance as well as reducin g costs and increasing beneficial reuse (EPA, 2011). However, due to adsorption of highly soluble ammonium salts concrete amended with fly ash can rele ase, i.e. leaching or volatilizion ammonia to the environment ( Rathbone, R. and Robl, T. 200 3; Cardone, et al., 2005; Wang, et al., 2002) Limited research has been done on the use of AFA in concrete blends and the subsequent leaching of ammonia/ammonium from crushed concrete and concrete monoliths. Previous studies have focused on the release of ammonia t o the ambient air or the release of ammonia from the fly ash alone ( Rathbone, R. and Robl, T. 200 3; Cardone, et al., 2005; Wang, et al., 2002 ) Previous studies reported that ammonia is in the form of ammonium sulfate and bisulfate salts and therefore hig hly soluble (Cardone, et al., 2005; Wang, et al., 2002). However, it is not clear what mass of ammonia/ammonium salts will be retained in hardened concrete. Due to this uncertainty, direct comparison to regulatory limits on ammonia lea ching cannot be made. T he parameters of most interest in concrete mixes with regard to ammonia release are the amount of cementitious material added to the mix (CM), the percent of cementitious material replaced with fly ash (FA), water content (WC), and the water to cementiti ous material ratio (W/CM). These parameters determine the initial ammonia content of a concrete mix (Rathbone, R. and Robl, T., 2003).
44 This study focused on the leaching of ammonia from ammoniated fly ash (AFA) and hardened concrete amended with AFA. Amm onia levels of collected fly ash samples were determined and solubility of the ammonium salts was characterized To address the lack of literature with regard to ammonia leaching from concrete, batch leaching procedures and monolithic leaching were investi gated The degree to which ammonia was encapsulated in crushed or monolithic con crete was investigated. Leaching tests on concrete monoliths were used as a risk assessment for AFA applications in concrete products used in the built environment, e.g. bridge pilings, sidewalks. The batch leaching tests were used to simulate leaching from the reuse of concrete amended with AFA, e.g. road base, concrete aggregate. The leached concentration of ammonia was compared to regulatory limits for ammonia to provide a ri sk assessment of the reuse of AFA in concrete applications. Material s and Methods Sample Collection and Processing Fly ash samples originated from the Crystal River Power Complex located in Crystal River, FL. The facility operates four coal fired boiler u nits (units 1, 2, 4, and 5) and one nuclear unit (unit 3) at this location. Units 4 and 5 were the only units which were practicing NO x control and flue gas conditioning by ammonia injection; therefore, all fly ash collected came from these units. Samples were collected from baghouse collection hoppers and placed in sealed borosilicate glass jars or HDPE containers; approximately 10 gallons of fly ash was collected from both units 4 (fly ash A) and 5 (fly ash B) during a single sampling event. A blend of fl y ash A and fly ash B was made to provide another composite sample with differing ammonia content (fly ash C).
45 Concrete mixes were designed to maximize the cementitious content (CM = 446.50 kg CM/m 3 ), fly ash replacement ratio (FA = 0.5), water content (W C = 236.65 kg H 2 O/m 3 ), and water to cementitious content ratio (W/CM = 0.53) while still remaining within acceptable material specifications set forth in the state of Florida; all these parameters remained constant for all concrete samples created. The amo unt of cementitious material added to mixes was at a level typically reserved for mass concrete, i.e. monolithic concrete structures cast in place (FDOT, 2010).The fine and coarse aggregates remained constant in all batches at 30% and 40% by weight, respec tively. Three concrete mixes were made with different fly ash samples; sets 1 (fly ash B), set 2 (fly ash C) and set 3 (fly ash A) were produced in triplicate for both the batch leaching tests and monolith leaching tests. Concrete samples were mixed and pl aced in cylindrical molds; samples and molds were wrapped in plastic sheeting to ensure a high humidity environment for the initial 48 hour curing period. After samples had been cured for 48 hours, the samples destined for batch leaching tests were removed mesh sieve. Portland cement (Quikrete ASTM C150 Type I) was purchased for use in all concrete mixes (ASTM, 2013c). The coarse and fine aggregates for the monolith leaching tests were received from the FDOT State Materials Lab, Gainesville, FL. Fine aggregate consisted of siliceous sand graded to ASTM C33 specification (ASTM, 2013d). Coarse aggregate consisted of limestone graded to ASTM C33 specifications; coarse aggregate was sa turated surface dry upon use in concrete mixes (ASTM,
46 2013d). All concrete aggregate and cement was tested to ensure no measurable ammonia was present before use. Characterization of Ammoniated Fly Ash For ammonia content analysis, a zero headspace extrac tion vessel (Analytical Testing Company, Model C 102) was used to produce sequential extractions of a fly ash sample using deionized (DI) water at a liquid to solid (LS) ratio of 10:1, i.e. 200 mL DI water to 20 g of AFA (Wang et al., 2002). Each ZHE (Zer o Headspace Extractor) was mixed in an end over end fashion for one hour between extractions. The extraction fluid was analyzed by ion chromatography (Dionex IC 20); the corresponding ammonium concentration of the extraction fluid was used to directly cal culate the ammonium concentration absorbed onto the fly ash sample. It was assumed that all of the ammonium in the fly ash is in the form of ammonium sulfate and bisulfate salts that are easily solubilized into water (Wang et al., 2002). The relationship between the ammonium concentration measured in the extraction fluid and the ammonium concentration absorbed onto the fly ash was calculated using the following expression : (3 1) w here C AFA, Total is the total concentration of ammonium in fly ash (mg NH 4 + /kg AFA) after three sequential extractions, C IC is the measured concentration in extraction fluid by ion chromatography (mg NH 4 + /L), V E is the volume of extraction fluid used (L), and m AFA is the mass of fly ash used (kg AFA). The fly ash was exposed to sequential extractions until ammonium concentrations in the extraction fluid were sufficiently low,
47 ge nerally less than one percent of the concentration in the first extraction (Wang et al., 2002). During each extraction 75% of the initial extraction fluid, i.e. 150 mL, was removed and analyzed for ammonium content. The fraction of extraction fluid remov ed for analysis was replaced with new ammonium free DI water and the extraction and mixing process was repeated. The cumulative concentration of ammonium in the extraction fluid was calculated to determine the total ammonium content of the fly ash. The Loss On Ignition (LOI) of each fly ash collected was measured in accordance with ASTM D7 348/ASTM C311 (ASTM, 2013e; ASTM 2013f ). The fly ash (approximately 4 hours) to burn off any remnant volatile solids (e.g., carbon content, water content, SO 3 content, or ammonium content) in the fly ash. Samples were we ighed intermittently until a constant weight was attained. The moisture content was measured in a similar manner but at a lower temperature period (~24 hours) following ASTM C 311/ASTM D 7348 (ASTM, 2013e; ASTM 2013f ). Leach ing Methodology Concrete samples retained in their monolithic form were suspended from the lid of a Gamma Seal bucket. Gamma Seals are air tight container lids that can be placed on many 3.5 7 gallon HDPE buckets. Samples were than ready for extractio n by EPA Method 1315: Mass Transfer Rates of Constituents in Monolithic or Compacted Granular Materials Using a Semi Dynamic Tank Leaching Test (EPA, 2009). EPA Method 1315 is designed to provide the release rates of inorganic analytes contained in a monol ithic or compacted granular material as a function of leaching time under diffusion controlled release conditions. The method comprises leaching of continuously water saturated monolithic material in an eluent filled tank with periodic (total of nine
48 leac hing intervals, i ) renewal of the leaching solution. Sa mples are contacted with DI water at a specified Liquid to Surface Area (L/A) ratio, i.e. 9 1 mL/cm 2 The leaching solution wa s exchanged with fresh ammonium free water at nine pre determined interva ls. All faces of the monolith were exposed to the leaching solution and care was taken to minimize headspace and to create an air tight seal to reduce any fugitive ammonia release into the vessel headspace or outside of the vessel. These results were used to determine the extent of release of encapsulated ammonia and the leaching mechanism of the ammonia release. The EPA Synthetic Precipitation Leaching Procedure (SPLP) was used to determine the mobility of ammonia in crushed concrete samples under weather ing conditions, i.e. acid rain precipitation (EPA, 1994). This test would be useful in simulating the leaching of crushed concrete amended with AFA in a reuse application or in a scenario where it was improperly disposed of. The crushed concrete samples, s ize equal to 20 times the weight of the solid phase. The extraction fluid employed was DI water for volatile analytes; the solid phase and solution were placed in ZHE and mixed in an end over end fashion for approximately 18 hours. Following extraction, the liquid extract was separated from the solid phase by filtration through a 0.45m glass fiber filter. This liquid extract was then stored at pH 2 by an addition of nitric acid for later analysis by ion chromatography (Dionex, IC 20). Data Analysis and Risk Assessment The results from the SPLP analysis were used to determine the fraction of total ammonia released from the crushed concrete. This release was compared to regu latory limits on ammonia, namely the Florida Groundwater Cleanup Target Level (GCTL) and
49 Surface Water Cleanup Target Level (SWCTL), as part of a risk assessment for reuse of crushed concrete amended with AFA. The GCTLs are standards set to protect human h ealth from contaminants in groundwater. The GCTL for ammonia is a minimum criterion concerning health considerations and aesthetic factors. The SWCTL for ammonia is listed as a Florida Surface Water Quality Standard (F.A.C., Chapter 62 302), as such, the i ntent of the standard is to provide surface water concentration limits on ammonia to be protective of both human health and the environment. The regulatory thresholds for ammonia for the GCTL and SWCTL are 2.8 mg NH 3 N /L and 0.02 mg NH 3 /L, respectively. L eaching can be affected by a number of factors, including pH, temperature, chemical and physical encapsulation of analytes, and leaching solution chemistry. However, with respect to leaching of analytes from monoliths, three main mechanisms that generally control the leaching process are surface wash off, diffusion transport, and surface dissolution (Torras et al., 2011). Surface wash off is a process by which the surface concentration of the analyte on the monolith is quickly solubilized into the bulk leac hing solution. Diffusion transport is the migration of analytes from higher to lower concentration through a medium (i.e., concrete). Surface dissolution is the physical solubilization of the medium into the leaching solution. A semi dynamic monolith leach ing test (e.g., Method 1315) can be used to determine the main leaching mechanism, mass transfer rate, and changes in the leaching mechanism over time (Torras et al., 2011). Calculation of the interface diffusion coefficient between the concrete and wate r wa s based on the semi infinite diffusion model (Glicksman, 2000), which is applied to a
50 diffusion process from a source (e.g., a cylinder, a plate or a slab) to an infinite bath. This model can be applied to ammonia releasing from a concrete slab into th e atmosphere or leaching into water. A relationship can be developed between the mass the assumptions of the semi infinite diffusion model, see Equation 3 2 ( Kosson et al., 2009; Glicksman, 2000). (3 2) w here D i is the calculated diffusion coefficient for leaching interval i (m 2 /s), t i is the leaching interval time (s), M ti is the measured mass release per area for interval i (mg/m 2 ), is the concrete density (kg dry/m 3 ) and C o (mg/kg) is the initial concentration of ammonium in the concrete. Additionally, it is sometimes convenient to express the calculated diffusivity as a leachability index (LI), see Equation 3 3 (3 3) Based on experimental measurements of mass released over time, Equation 2 can be used to calculate the diffusion coefficient as a function of time. The measured mass re lease M ti was determined by the relationship seen in Equation 3 4. (3 4) w here c meas is the measured concentration of ammonium in the extraction fluid (mg/L), v i is t he volume of extraction fluid employed for leaching interval i (L) and A is the surface area of the monolith (m 2 ). Assuming the semi infinite diffusion model, it is appropriate to visualize the concrete monolith used in this study as a cylinder diffusing radially into an infinite bath
51 of deionized water. The diffusion coefficient for ammonium can be used to predict release rates of ammonium in hardened concrete under different scenarios, e.g. concrete of different dimensions and ammonium concentrations (K osson et al., 2009; Torras et al., 2011). Results and Discussion Fly Ash Characterization As can be seen in Figure 3 1 t he majority of ammonia (> 95%) wa s extracted during the first extraction of fly a sh; this trend was observed for all fly ash samples collected. These results mirror those found in another study using a similar extraction method (Wang et al., 2002). Therefore, only two extractions were conducted on fly ash samples. The mass release of a mmonia was similar to results seen in Wang et al., 2002. The observed concentrations for all fly ash samples can be seen in Table 3 1 Figure 3 1 Sequential extraction to determine ammonium content of fl y ash s amples 0 500 1,000 1,500 2,000 2,500 3,000 3,500 1 2 Ammonia Concentration (mg/kg) Extraction Fly Ash A Fly Ash B Fly Ash C
52 The LOI measurements for all fly ash samples were consistent. The LOI measurement was higher than the specification for the use of fly ash in concrete in accordance with ASTM C311; however, the fly ash used in experiments and subsequent con crete mixes were collected before any beneficiation practices were utilized to produce a more marketable fly ash. T he fly ash was used in experiments as received from the facility At the time of sample collection the facility was no longer selling their f ly ash and no fly ash with elevated levels of ammonia but still in LOI limits was being produced at the facility. The moisture content was consistent for all samples and very low; ammonia volatilization from unconditioned (un wetted) fly ash is unlikely. T he moisture content and LOI can be seen in Table 3 1 Table 3 1 Collected fly ash measured ammonium content, LOI and moisture c ontent Sample Identification Measured NH 4 + Concentration (ppm) LOI (%) Moisture Content (%) A 65 11.2 0.21 B 3211 11.2 0.22 C 1378 11.2 0.22 The chemical composition of the fly ash can be estimated from a mill report created two months prior to collection, see Table 3 2 From the data it can be seen that the fly ash collected wa s Class F fly ash. It should be noted that the LOI is lower than the reported values in Table 3 1 ; this observation can be explained by the fact that Crystal River stopped marketing their fly ash soon after this final mill report was created.
53 Table 3 2 Mill report on chemical composition of fly ash collected from Crystal River power c omplex Chemical Tests Results Silicon Dioxide (SiO 2 ), % 52.99 Aluminum Oxide (Al 2 O 3 ), % 26.78 Iron Oxide (Fe 2 O 3 ), % 6.91 Sum of SiO 2 Al 2 O 3 Fe 2 O 3 % 86.68 Calcium Oxide (CaO), % 3.62 Magnesium Oxide (MgO), % 1.28 Sulfur Trioxide (SO 3 ), % 0.40 Sodium Oxide (Na 2 O), % 0.26 Potassium (K 2 O), % 2.14 Total Alkalies (as Na 2 O), % 1.67 Moisture Content, % 0.13 Loss on Ignitio n, % 3.66 Specific Gravity 2.32 Monolithic Leaching Ammonium was continually released from all the monoliths during the leaching intervals up to collection of leachate on day 49; the final leaching interval was not used since ammonia was below detectio n limit of the IC (< 10 g/L). This release is presented as a cumulative mass fraction released in Figure 3 2 with approximately 18 37 % of initial ammonia released from the monoliths The monoliths with the smallest initial ammonia proportion (Set 3) rele ased the largest fraction of the initial ammonia content; while the opposite is true for the monoliths with the largest ammonia proportion (Set 1). This result can be explained by the fact that as the concrete cured the permeability of the concrete decreas ed. The decrease in concrete permeability would account for decrease in ammonia release in later periods. During the initial adsorption of water a larger mass release can be observed; concrete with less initial ammonia content would release a larger propor tion in the initial adsorption phase.
54 Figure 3 2 Mass f raction of a mmonium r eleased from c oncrete m onoliths a mended with AFA Concrete amended with fly ash with higher levels of ammonia will release ammo nia for longer periods of time and at higher mass release rates, see Figure 3 3 This is not a surprising result since the higher initial ammonium content of samples increases the diffusion gradient and source depletion is less likely The observed diffusi vity (represented as a leachability index, refer to Equation 3 3) of ammonium for each set of samples as a function of leaching time can be seen in Figure 3 4 The LIs incr eased with leaching time, due to source depletion and decreasing concrete permeabili ty. The decrease in hydraulic permeability of concrete as a function of curing age is well documented (Kosmatka et al., 2008). The diffusivity of ammonia in concrete is not constant; it decrease s with time. The first reason is due to physical and chemical changes of the concrete. The hydration process takes place during the entire mixing 0 5 10 15 20 25 30 35 40 0 10 20 30 40 50 60 Mass Fraction Released (%) Time (days) Set 1 Set 2 Set 3
55 and curing process. Calcium silicate hydrates formed during chemical reactions fill the pore space of the concrete which leaves fewer pathways for ammonia to diffuse out, t hereby increasing tortuosity of the diffusion pathways. In addition, the moisture content of the concrete will decrease as the concrete ages, which also prevents ammonium ions from diffusing into the ambient environment (Kosmatka et al., 2008; Cheng and Bi shop, 1990 ). The logarithmic representation of cumulative ammonium release (Log [ M t ]) and cumulative leaching time (Log [ t]) can be utilized to determine the dominant leaching mechanism during leaching intervals, see Figure 3 3 (Barna et al., 1997; Cheng and Bishop, 1990; Torras et al., 2011). If the data can be represented as a st raight line with a slope > 0.65, the dominant leaching mechanism is surface dissolution; however, if the data have a straight line slope between 0.35 and 0.65, diffusion is the dominant mechanism. Depletion is indicated if the slope of the line is less tha n 0.35 in later leaching periods (Torras et al., 2011). As can be seen in Figure 3 3 the dominant leaching mechanism during the latter leaching intervals is diffusion ; however, the larger release seen during the first week of leaching can be explained by an initial adsorption of water and thereby release of ammonia at higher rates T he observed diffusivity is a useful measure of the ability of concrete to encapsulate the ammonia adsorbed to the fly ash.
56 Figure 3 3 Mass r elease of a mmonium from c oncrete m onoliths a mended with AFA as a function of leaching t ime y = 0.4525x + 2.5037 R = 0.9567 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 -1.5 -0.5 0.5 1.5 2.5 t ) [mg/m 2 ] [days] Set 1 y = 0.3638x + 2.4033 R = 0.9804 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 -1.5 -0.5 0.5 1.5 2.5 t ) [mg/m 2 ] [days] Set 2 y = 0.4353x + 1.1026 R = 0.9731 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 -1.5 -0.5 0.5 1.5 2.5 t ) [mg/m 2 ] [days] Set 3
57 A common form of representing the ability of a material to encapsulate a contaminant is the leachability index (LI); the LI is the negative logar ithmic representation of the diffusivity of the species of interest. Generally, if the LI < 11 the species is highly mobile, if 11 < LI < 12.5 the species has an average mobility, and if the species has a LI > 12.5 then the species has low mobility (Bobiri ca et al., 2 010). As can be seen in Figure 3 3 the LI for ammonia during the initial leaching periods (< 7 days) was 11 12 indicating that the ammonia was initially mobile. However, in later leaching periods that LI rose above 12.5 to 13 14 indicating the remaining ammonia was better encapsulated. These results indicate that after the first week of curing the ammonia is far less mobile. This is explained by the decrease in concrete permeability over curing time. Figure 3 4 Leachability i ndex of a mmonium in c oncrete m onoliths a mended with AFA as a f unction of leaching t ime 8 9 10 11 12 13 14 15 0 10 20 30 40 50 60 Leachability Index Leaching Time (days) Set 1 Set 2 Set 3
58 The cumulative release of ammonia from the monolith leaching procedure can be used to predict the maximum concentration of ammonia under conservative conditions. If it is assumed that a concrete monolith, with the same concrete mix and L/SA ratio utilized in leaching tests, is submerged in a natural water body with no mixing and at a pH value of 7 the maximum ammonia concentration can be calculated from the cumulative ammonia release data of the monolith leaching procedure. Converting this maximum concentration to unionized ammonia a comparison can then be made to the Florida Surface Water Cleanup Target Level (SWCTL = 0.02 mg NH 3 /L). As can b e seen in Figure 3 5 below a least squares line can be used to predict a leachability threshold at 35 mg NH 3 /kg concrete. This leachability threshold gives guidance for a limit on ammonia content in concrete mixes for the protection of surface wate rs ; the leachability threshold can be extended to the fly ash by a mass balance calculation on the concrete mix (leachability threshold (fly ash) = 350 mg NH 3 /kg fly ash). It should be noted that this leachability threshold for the fly ash is specific to t his concrete mix design. However, since concrete mix was chosen to maximize ammonia content it is unlikely that more common concrete mixes would present a leachability risk for ammonia if the 350 mg NH 3 / kg fly ash is utilized.
59 Figure 3 5 Maximum c oncentration of a mmonia r eleased from c oncrete m onoliths a mended with AFA and d erived l eachability t hreshold (L T ) The monolith leaching results demonstrate that some risk is associated with the use of AFA in the built environment, e.g. bridge pilings, sidewalks; however, it is necessary to determine risk of concrete amended with AFA in a reuse scenario. For example, concrete from construction and demolition projects is sometimes reused as road base or concrete ag gregate to produce more concrete. To reuse the concrete the monolithic concrete is size reduced (crushed); batch leaching can be used to determine the leaching risk associated with reuse scenarios. SPLP Results The results from the SPLP test can be seen i n Table 3 3 It can be see n that all concrete sets leached the majority of their initial ammonia content (58 68 % ) It is noted that the concrete mix with the highest initial ammonia content (Set 1) leached the R = 0.9565 0 0.05 0.1 0 100 200 300 400 Max. Concentration (mg/L) NH 3 Conc. in Concrete (mg/kg) SWCTL = 0.02 mg NH 3 /L LT = 35 mg NH 3 /kg concrete
60 largest proportion of ammonia, i.e. approxim ately 68% of the initial ammonia, w hile the concrete mix with the lowest initial ammonia content (Set 3) leached the smallest proportion of ammonia, i.e. approximately 58% of the initial ammonia. This is not a surprising result if one takes into account th e much higher concentration gradient of ammonia in Set 1 as opposed to Set 3. The concentration gradient was likely responsible for more ammonia leaching from the crushed concrete. Table 3 3 SPLP a nalysis o f c rushed c oncrete a mended with AFA Sample Set NH 4 + Concentration (mg/L) Total NH 3 Released (%) Set 1 11.01 1.06 67.7 Set 2 4.64 0.230 66.3 Set 3 0.193 0.0097 58.4 When the ammonia release from the SPLP results is compared to regulatory limits for ammonia in the Florida GCTL it is obvious that there exist an upper limit for the concentration of ammonia in the fly ash that can be used without some potential impact to the environment (FDEP, 2009). For example, the leaching of ammonia from both S et s 1 and 2 were above the GCTL (GCTL = 2.8 mg NH 3 N /L). A simple statistical treatment (least squares fitting) of the SPLP analysis data and the total ammonia content of the concrete mixes can be compared to the GCTL to produce a leachability threshold esti mate (FDEP, 2009). The leachability threshold estimate for the crushed concrete samples (Set s 1, 2, and 3) was determined to be 85 mg NH 4 + /kg concrete. The SPLP results can be compared to the Florida SWCTL, at 0.02 mg NH 3 /L (unionized total ammonia) taki ng into account that the pH of most natural waters falls into the 6 to 9 range due to bicarbonate buffering (IFAS, 2011). It is obvious under some pH conditions all concrete samples could exceed the SWCTL, see Table 3 4 It should
61 be noted that temperature and salinity both affect the degree to which ionized ammonia is converted to unionized ammonia; however, these have a lesser degree of effect than pH. These comparisons are made to illustrate the possibility of a threat to environmental health due to eutr ophication or aquatic life toxicity However, ful l characterization of the fate of the ammonia leached is beyond the scope of this study and field stud ies should be conducted. Nevertheless these results do indicate that care should be taken when disposing or reusing crushed concrete that has been amended with highly ammoniated fly ash. Practices that result in concrete with ammonia concentration greater than 85 mg NH 4 + /kg may represent a threat to the environment due to the release of ammonia into surround ing waters. Table 3 4 Influence of pH on the SPLP u nionized a mmonia (NH 3 ) c oncentrations from the c rushed c oncrete s amples a mended with AFA Set 1 (mg NH 3 /L) Set 2 (mg NH 3 /L) Set 3 (mg NH 3 /L) 4 < 0.001 < 0.001 < 0.001 5 < 0.001 < 0.001 < 0.001 6 0.0076 0.0032 < 0.001 7 0.075 0.03 < 0.001 8 0.72 0.30 0.012 9 4.83 2.04 0.083 10 11.01 4.64 0.193 The SPLP analysis does not addr ess the loss of ammonia from concrete in a monolithic form. It is well known that the tortuosity of the concrete matrix encapsulates inorganic analytes, e.g. ammonia or trace metals, or retards their leaching ; this fact can be observed in the smaller relea se of initial ammonia mass from the monoliths in this study However, t he SPLP analysis does represent the leaching seen from concrete in a reuse scenario
62 C onclusions The experimental results suggest the dominant leaching mechanism for ammonia is diffus ion for later periods (> one week) of curing and initial adsorption of water during earlier periods (< one week) for concrete monoliths amended with AFA. The loss of ammonia mass is dependent on the initial ammonia mass present in the concrete matrix. The mass fraction of ammonia leached from samples varied between 15 35%. The diffusivity of ammonium in concrete decreases over time due to physical changes in the concrete matrix, such as, decreased hydraulic permeability and encapsulation. The ammonia diff usivity decreased precipitously after one week and the ammonia was well encapsulated, i.e. LI > 13. However, from the monolith studies it can be concluded that not only do concrete monoliths amended with higher concentrations of ammonia release more ammoni a they also release ammonia for a longer period. Leaching of initial ammonium content in crushed concrete was increased (~64%) when compared to the leaching from the monoliths. This is likely due to the loss of the monolithic concrete matrix, i.e. loss of tortuosity, and increased surface area of concrete particles. Due to this rapid loss of originally encapsulated ammonium there is a possibility of regulatory significant release of ammonium from crushed concrete. A leachability threshold estimate for the crushed concrete samples was determined to be 85 mg NH 4 + /kg concrete. Further investigation is needed into the loss of ammonia from different concrete mix designs; in addition to pilot scale tests to determine the fate of leached ammonium due to environmen tal factors that cannot be simulated in the laboratory. Additionally, effects of commonly used admixtures on ammonia loss were not investigated in this research.
63 CHAPTER 4 LEACHING OF TRACE METALS FROM CONCRETE AMENDED WITH CEMENT KILN BAGHOUSE FILTER DUST Intro duction Cement kiln operators employ a variety of air pollution control techniques to limit emissions and to capture particulate materials for return to the manufacturing process. Baghouses, as well as other particulate control devices, are used at various locations in the manufacturing process to collect dust particles produced in the process. In some cases a material commonly referred to as cement kiln dust (CKD) is extracted from the system, as a waste product, to remove chemicals unfavorable to the fin al cement product (e.g., alkali salts) (EPA, 2012). More often than not, however, recovered particulate materials are recycled back into the cement production process; in some cases the kiln is bypassed and the particulate material is incorporated into the final cement product, a technique referred to as dust shuttling. Such materials are considered to be Inorganic Processing Additions (IPA), and may be allowed as additions if the applicable ASTM and AASHTO standards for final cement products are met (ASTM C465 and/or AASHTO M327). Materials most likely to be used as inorganic processing additions are limestone, fly ash, bottom ash, slag, cement kiln feed, cement kiln dust, and calcined byproducts; for specificity the partially calcined byproducts collected in the cement kiln baghouse will be referred to baghouse filter dust (BFD) in this study (NCHRP, 2008). As a means of controlling mercury (Hg) stack emissions at cement kiln operations, some facilities have proposed or have instituted the practice known a s dust shuttling, where BFD is routed to be blended with the final cement product instead of returning it to the kiln. This technique allows mercury to leave the facility with the final cement
64 product rather than through stack emissions. However, this may cause a potential release of mercury to aqueous environments during the storage and handling of BFD as well as the potential releases from the concrete product amended with BFD. Furthermore, it has been well established that other trace metals (As, Ba, Cd, Cr, Pb, Se, V, Zn) can be found in particulate materials produced at cement kilns, e.g. CKD; these metals may pose a risk from direct exposure to cement or concrete amended with BFD or risk of leaching into the aqueous environment (Haynes and Kramer, 1982 ; PCA, 1992; Duchesne and Reardon, 1998; Serclerat et al. 1999; Van der Sloot, 2002; Shivley et al. 1986; Hillier et al. 1999; Poon et al. 1985). Two nationwide (U.S.) surveys (~70% of kilns surveyed in each study) have been done in the past to characteriz e the concentration of constituent elements in CKD and cement (Haynes and Kramer, 1982; PCA, 1992). Both studies concluded that while many trace metals are present they are at levels that are unlikely to cause a hazardous material classification by Toxicit y Characteristic (TC) definitions. However, it should be noted that CKD differs from BFD; BFD is a material sought for reuse as an IPA while CKD is defined as a waste. It is unclear if compositional differences exist between the two materials since no publ ished literature can be found on the novel BFD. As a bench mark the composition and leaching of CKD can be utilized. In CKD, trace element mobility is correlated to the solubility of minerals during the hydration process (Duchesne and Reardon, 1998; Sercle rat et al. 1999; Van der Sloot, 2002; Shivley et al. 1986; Hillier et al. 1999; Poon et al. 1985). When CKD contacts water, these soluble phases will either completely dissolve or more stable and less soluble phases will precipitate thereby releasing trace metals to a mobile phase.
65 Additionally, it is sometimes desired to simulate release of trace elements due to weathering. This can be accomplished by exposure to acidic solutions or multiple extractions of the same leaching fluid (Shively et al. 1986; Poon et al. 1985). While the release of trace metals from monolithic materials, e.g. concrete, is a diffusion controlled process (Hillier et al. 1999; Poon et al. 1985). A study was conducted to address the lack of literature with regard to leaching of trace m etals from BFD, cement/concrete products amended with BFD, and to compare the composition of BFD to that of CKD. Leaching tests on concrete monoliths were u sed as a risk assessment for BFD applications in concrete products used in the built environment, e. g. bridge pilings, sidewalks. The batch leaching tests were used to simulate leaching from the re use of concrete amended with BFD e.g. road base, concrete aggregate. Additionally, the extent of leaching of trace metals due to the use of BFD as an IPA shou ld be compared to environmental regulatory limits as a bench mark of environmental impact, e.g. Florida Soil Cleanup Target Levels, or Florida Groundwater Cleanup Target Levels. Finally, the leaching of trace metals should be evaluated under a variety of l eaching conditions, e.g. pH dependent, weathering conditions, monolithic leaching. Material and Methods Sample Collection and Processing One cement facility participated in the study, namely, the Brooksville cement facility. The Brooksville cement facility located in Brooksville, FL, is an integrated facility that includes a Portland cement manufacturing plant having one preheater dry process kiln system (Kiln 1), initial operation in the mid 1980s, one precalciner dry process kiln system (Kiln 2), initial operation in 2009, and a power plant. Kiln 1, along with the in line
66 kiln/raw mill and clinker cooler 1 share a baghouse fabric filter system and exhaust stack with the power plant. However, Kiln 2, the related in line kiln/raw mill, and clinker cooler 2, share a common baghouse fabric filter system and stack not connected to either kiln 1 or the power plant boiler. All baghouse filter dust which was collected from this facility for the study came from the Kiln 2 baghouse. Kiln 2 is designed for 156 TPH of cement clinker production (FDEP, 2013). The kiln is allowed to fire coal, petroleum coke, natural gas, coal fly ash, propane, distillate fuel oil, on specification oil, whole tires and alternative fuels. The Brooksville facility is the only facility in the state of Florida currently practicing baghouse dust shuttling but other facilities will likely be using this technique within less than two years given the upcoming revised NESHAP, subpart LLL (40 CFR 63) requirements to limit Hg emissions to 55 lb/mi llion ton of clinker on a 30 day average for existing kilns (FDEP, 2013). Samples of BFD or cement collected from the Brooksville facility were stored in five gallon High Density Polyethylene (HDPE) buckets. Four grab samples of BFD and one cement sample were collected; BFD Sample A (collected 10 28 2011), BFD Sample B (collected 12 15 2011), BFD Sample C (collected 02 28 2012), BFD Sample D (collect 08 24 2012), and Cement Sample A (collected 08 24 2012). Prior to use of any BFD or cement in experiments the material was homogenized by rotating the storage vessels in an end over end fashion for approximately 2 hours. Additionally, attempts were made, when removing a sample from the sample storage vessel for use in experiments, to mix the bulk sample and se lect the aliquot randomly.
67 Coarse and fine aggregate (Quikrete ASTM C33 specification) was purchased for use in all concrete mixes. Fine aggregate consisted of siliceous sand graded to ASTM C33 specification. Coarse aggregate consisted of limestone grad ed to ASTM C33 specifications; coarse aggregate was saturated surface dry upon use in concrete mixes ( Kosmatka, et al., 2002) All concrete aggregate and cement was characterized for trace metals before use. As previously mentioned, recent harmonization o f industrial standards (AASHTO M85 and ASTM C150), have led to an allowed addition of IPAs of 5% by mass to Type I Portland cement. Additionally, a National Cooperative Highway Research Program (NCHRP) study precluded the use of greater than 8% by mass add ition of dust collect at the kiln baghouse, i.e. baghouse filter dust (BFD), due to Loss On Ignition (LOI) and insoluble residue limitation (NCHRP, 2008). Taking this into account, two cement and BFD blends were created, at 5% BFD to 95% cement by mass and 10% BFD to 90% cement by mass. The blends were mixed in an end over end fashion for 24 hours before use in experiments. Concrete samples were mixed and placed in rectangular molds; samples and molds were wrapped in plastic sheeting to ensure a high humid ity environment for the initial 24 hour curing period; the concrete sample sets created contained the 10% BFD/90% cement blend (Set A) and the 5% BFD/95% cement blend (Set B) and a control that contained no BFD. The sets were made in triplicate. All concr ete constituent amounts were kept constant with the exception of the BFD mass additions. The cementitious content, water to cementitious content ratio, coarse and fine aggregate additions were kept constant for all sets at 2.05 kg, 0.53, 3.63 kg, and 2.90 kg,
68 respectively. These concrete mixes were used for crushed concrete leaching and monolithic leaching. The monolith samples were kept whole, while the crushed concrete Elemental Analy sis An initial analysis of cement and BFD samples collected from the Brooksville facility was performed via Scanning Electron Microscopy Energy Dispersive X Ray Spectroscopy (SEM EDS) and X Ray Diffraction Spectroscopy (XRD). The bulk elemental analysis and mineralogical analysis were performed via SEM EDS and XRD at the UF Major Analytical Instrumentation Center (MAIC) on the UF campus. Trace metals analysis was performed on all BFD and cement samples collected from participating facilities and other con crete constituents used in accordance with EPA Method 3050b Acid Digestion of Sediments, Sludges, and Soils (EPA, 1996). A mass of representative solid sample was weighed out (1 2 grams) and exposed to numerous additions of nitric acid, hydrogen peroxide and hydrochloric acid under refluxing conditions for up to 5 hours or until no visible changes in the sample can be observed due to nitric acid additions or hydrogen peroxide additions. The sample is then diluted to a predetermined volume and stored for later analysis via Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP AES). It should be noted that this method is not a total digestion technique for most materials. It is a strong acid digestion that will dissolve almost all elements that c Elements bound in silicate structures are not normally dissolved by this procedure; however, this should not be a concern since silicate bound compounds are not considered mobile in the environment.
69 The digestion p rocedure to determine mercury speciation was derived based on the sequential extraction procedure described in EPA Method 3200 (EPA, 2005), while the analysis of total mercury concentration was based on EPA Method 7474 (EPA, 2007). For each type of materia l, i.e. cement, BFD, fine aggregate, or coarse aggregate, five replicates of each 0.5 grams were weighed using an analytical scale (Sartorius MC210S, Goettingen, German) with a precision of 0.01 mg. The solid samples were then placed in thread capped poly tetrafluoroethylene (PTFE) tubes; 10 mL of the appropriate extraction solution for different mercury species were added to the PTFE tube. The PTFE tubes were heated up to about 100 C for 30 minutes in a microwave digestion system (CEM MDS 81D, Matthews, N C). The digests were diluted with deionized water to a pre determined level, to accommodate the linear dynamic range of the analytical instruments. The diluted solution was transferred into a 50 mL centrifuge tube and analyzed by a Hydride Generation Ato mic Fluorescence Spectrometer (HG AFS) (Aurora Biomed 3300, Vancouver, BC, Canada). The detection limit of the HG AFS was around 1 ng/mL. The mercury concentration in each type of samples was averaged from the five replicates. Leaching Methodology To addre ss monolithic leaching under diffusional release; concrete monoliths were created. After the concrete samples had been cured for 24 hours, they were removed from the molds and were placed in 54 gallon HDPE sealed containers. Samples were then ready for ext raction by the monolith leaching method (EPA Method 1315). The monolith leaching test was designed to provide the mass transfer rates (release rates) of inorganic analytes contained in a monolithic material, under diffusion controlled release conditions, a s a function of leaching time. The method comprises leaching of
70 continuously water saturated monolithic material in an eluent filled tank with periodic renewal of the leaching solution. Samples are contacted with deionized water at a specified Liquid to Su rface Area (LSa) ratio, i.e. 9 mL/cm2. All concrete samples, except the control, were amended with BFD Sample D (EPA, 2009). The widely utilized Synthetic Precipitation Leaching Procedure (SPLP) was performed on all BFD samples, cement/BFD blends, and cru shed concrete samples. The SPLP was used in determining the mobility of trace metals in the solid samples under weathering conditions, i.e. acid rain precipitation. For the solid phase samples, e.g. BFD or concrete, the sample was extracted with an amount of extraction fluid equal to 20 times the weight of the solid phase. The extraction fluid employed is deionized water adjusted to a pH of 4.22 0.05 with a mixture of 60/40 by mass of sulfuric acid and nitric acid, respectively. The solid phase and soluti on were sealed in HDPE bottles and mixed in an end over end fashion for 18 2 hours. Following extraction, the liquid extract was separated from the solid phase by filtration through a 0.45 m glass fiber filter by vacuum. This liquid extract was then sto red at pH 2, by nitric acid addition, for later analysis with ICP AES and CV AFS (EPA, 1994). An important parameter of nearly all leaching procedures is pH. The pH of the aqueous environment is of crucial importance in determining the long term leachin g behavior of cement and concrete. To address this, a pH dependent leaching test was utilized, EPA Method 1313. This method is designed to provide aqueous extracts representing the liquid solid partitioning (LSP) curve, that is, the extent to which an anal yte leaves the solid phase and partitions to the liquid extract, as a function of pH for inorganic constituents (e.g., metals) in solid materials. The LSP curve is evaluated as a
71 function of final extract pH at a liquid to solid ratio (L/S) of 10 mL extra ctant/g dry sample (g dry) and conditions that approach liquid solid chemical equilibrium. The pH of sample leaching batches is controlled with additions of nitric acid to attain a range of eluate pHs between 4 and 13. This method also yields the acid/base titration and buffering capacity of the tested material at an L/S of 10 mL extractant/g dry sample. Following extraction, the liquid extract was separated from the solid phase by filtration through a 0.45 m glass fiber filter. This liquid extract was the n stored at pH 2 for later analysis with ICP AES and CV AFS. The pH dependence test was performed on the BFD Sample D only since the trace metals concentrations were similar for all BFD samples and crushed concrete amended with BFD Sample D (EPA, 2010). The Multiple Extraction Procedure (MEP EPA Method 1320) is designed to simulate weathering via the leaching that a waste will undergo from repetitive precipitation of acid rain on an improperly designed sanitary landfill. The repetitive extractions reve al the highest concentration of each constituent that is likely to leach in a natural environment. Waste samples are extracted with deionized water which is maintained at a pH of 5.0 0.2, with acetic acid. Following extraction, the liquid extract was se parated from the solid phase by filtration through a 0.45 m glass fiber filter. This liquid extract was then stored at a pH of 2 for later analysis with ICP AES and CV AFS. Then the solid portions of the samples that remain after application of the firs t extraction is re extracted eight times using synthetic acid rain extraction fluid; i.e. deionized water adjusted to a pH of 3.0 0.2 with a mixture of 60/40 by mass of sulfuric acid and nitric acid, respectively. Following each extraction, the liquid ex tract was separated from the solid phase by filtration through a 0.45 m glass fiber filter. This
72 liquid extract was then stored at a pH of 2 for later analysis with ICP AES and CV AFS. The MEP test was performed on the BFD Sample D and crushed concrete amended with BFD Sample D (EPA, 1986). Risk Assessment The total trace metals concentration for the BFD and cement collected from the participating facilities were compared to Florida Soil Cleanup Target Levels (SCTL). Florida SCTLs were developed based on direct human contact with the material of concern and in soil acting as a source of groundwater or surface water contamination (i.e., leachability). The SCTLs for metals are used in this study as a bench mark to compare concentration levels in the BFD to a regulatory guideline and to provide a measure of potential environmental impact. Additionally, the Florida Groundwater Cleanup Target Levels (GCTLs) were used in a similar manner to compare concentrations of trace metals in the extraction fluid of leachi ng assessments, e.g. SPLP, MEP, to a regulatory threshold. The GCTLs are standards set to protect human health from contaminants in groundwater. The GCTLs are either primary or secondary standards, based on the Florida Drinking Water Standards (Chapter 62 550 F.A.C.), or minimum criteria concerning health considerations and aesthetic factors but are not listed in Chapter 62 550 F.A.C. Many of the GCTLs are comparable to national standards, i.e. EPA Maximum Contaminant Levels (MCLs). Results and Discussion E lemental Analysis The initial characterization of the BFD included an analysis of bulk and trace elements and a mineralogical analysis. The results of the bulk elemental analysis on the BFD samples and cement can be seen in Table 4 1 The results from the bulk elemental
73 analysis show that the BFD collected for this study fall within the ranges seen in historical data for CKD, i.e. Haynes and Kramer, 1982. The minerals present in the samples initially collected at the Brooksville facility include: calcium ca rbonate, titanium oxide, titanium hydride, carbon sulfide, silicon sulfide, graphite, silicon dioxide, ferric oxide, and calcium oxide. Based on the relative abundance of each element and the peak intensity of the XRD analysis, it was determined that the m ajor crystalline phase in the baghouse filter dust is calcium carbonate, likely from the raw materials, i.e. limestone, other mineral forms are minor or trace constituents. Table 4 1 Energy dispersive s pect r oscopy of BFD and cement s amples Element Element Concentration (%wt.) BFD Sample A BFD Sample B BFD Sample C BFD Sample D Cement Sample A Al 1.56 1.75 1.12 1.31 2.20 C 7.38 14.70 22.57 21.80 15.30 Ca 56.76 27.55 24.35 24.18 35.74 Cl 0.28 < 0.1* 0.0 9 0.11 < 0.1* Fe 1.65 0.73 0.41 0.60 2.71 K 0.77 0.35 0.23 0.25 0.36 Mg 0.17 0.23 0.19 0.20 0.28 O 27.65 52.10 48.59 48.96 37.55 S 0.16 < 0.1* 0.07 < 0.1* 0.63 Si 3.68 2.57 2.37 2.44 5.23 Ti 0.27 < 0.1* < 0.1* 0.17 <0.1* *Non detectable Table 4 2 l ists the mercury (Hg) concentration and speciation in the BFD, c ement, and other constituents in the concrete samples. Alkyl Hg was below the detection limit in all the samples, and therefore was eliminated from following studies. The high temperature and combustion condition in the cement kiln likely decomposes any organic phase compounds. The total Hg concentration in the BFD samples ranged from 0.91~1.52 mg/kg (ppm); below both the residential and industrial Florida SCTL for mercury, 3 mg/kg and 17 mg.kg respectively. Soluble inorganic mercury (SI Hg)
74 accounted for 61.54~73.43% of total Hg in the samples, while the rest was in a non soluble inorganic mercury (NSI Hg) phase. This is significant since the soluble form of mercury is more mobile and toxic t han the insoluble fraction. The total Hg concentration in the BFD was higher than the average concentration seen in previous national surveys of cement kilns (Haynes and Kramer, 1982; PCA, 1992), but still fell within observed ranges. The total Hg concentr ation in the Brooksville cement was 74.51 g/kg. The concentration in the cement was also higher than the available studies (PCA, 1992; Pistilli and Majko, 1984); the latte r showed Hg in the cement below 39 g/kg. Table 4 2 Mercury concentration and speciation of concrete c omponents Material Total Hg (g/kg) Soluble Inorganic Hg (g/kg) Percentage of Soluble Inorganic Hg (%) BFD Sample A 910 60 560 20 61.5 BFD Sample B 1430 70 1050 40 73.4 BFD S ample C 1520 90 990 100 65.4 BFD Sample D 1440 23 1030 100 71.5 Cement Sample A 74.51 6.24 47.02 8.94 63.1 Coarse aggregate 4.32 1.52 2.63 1.16 60.8 Fine aggregate 0.44 0.06 0.33 0.04 73.2 The total trace metals concentration fo r the BFD and other concrete constituents can be seen in Table 4 3. Comparison of trace metals concentrations in cement and BFD samples collected for this study, with those from the national surveys conducted in 1982 (Haynes and Kramer, 1982) and 1992 (PCA 1992), shows that the total metals concentrations fall close to the expected ranges from these surveys, with the one
75 exception being silver. For both the cement and BFD collected, the total silver measured was significantly lower than the 1982 and 1992 s urveys. The only trace metal that exceeded the Florida SCTL was arsenic; however, the arsenic levels were still within the historic levels previously observed. Table 4 3 Total trace m etals c oncentration in BFD, c ement and a ggregate Element Cement Sample A* (mg/kg) BFD* (mg/kg) CKD** (mg/kg) Coarse Aggregate* (mg/kg) Fine Aggregate* (mg/kg) SCTL, (mg/kg) Residential/ Industrial Ag 0.149 0.322 (0.216 0.411) 10.5 (4.80 40.7) < 0.1 <0.1 410/1200 As 50.2 20.8 (17.5 27.6) 18.0 (2.00 159) 0.647 0.113 2.1/12 Ba 190 67.8 (42.3 75.6) 172 (35.0 767) 5.07 1.17 120/ 130000 Cd 4.85 1.56 (1.43 1.68) 10.3 (0.1 59.6) 0.135 0.121 82/1700 Cr 53.4 19.81 (16.9 23.6) 41.0 (8.00 293) 4.84 0.954 210/470 Pb 90. 4 28.48 (17.1 52.7) 434 (34.0 7390) 0.823 1.35 400/1400 Se < 1.0 4.36 (3.22 5.63) 28.1 (2.68 307) < 1.0 < 1.0 440/11000 V 136 50.35 (41.0 59.9) Not tested 3.00 1.26 69/10000 Zn 635 70.94 (51.2 99.6) Not tested 4.46 15.0 26000/ 630000 *All v alues are replicate average ** (PCA, 1992) Monolith Leaching The release of the trace metals in the monolith leaching test was minimal. Indeed, only barium and zinc were above the detection limit of the ICP AES, while the mercury concentration, determined via CV AFS, in the extractions was not significantly different than that found in deionized water, i.e. < 0.2 ng/L. The mass release rate of barium as a function of time can be seen in Figure 4 1 It should also be noted that the
76 concentrations in the grap h of Figure 4 1 have been normalized with respect to the LSa ratio of the monolith test. The leaching of barium and zinc never exceeded the GCTLs for the respective elements during any of the extraction periods. The release of zinc was recorded only during the first 48 hours of leaching after which the concentration of zinc in the extraction fluid fell below the detection limit of the ICP AES. This result leads to the conclusion that zinc was released in an initial wash off period and did not release via di ffusion. The minimal release of all trace metals leads to the conclusion that these constituents are encapsulated and immobilized in the concrete matrix. The logarithmic representation of cumulative ammonium release (Log [ M t ]) and cumulative leaching ti me (Log [ t]) can be utilized to determine the dominant leaching mechanism during leaching intervals, see Figure 4 1 (Barna et al., 1997; Cheng and Bishop, 1990; Torras et al., 2011). If the data can be represented as a straight line with a slope > 0.65, the dominant leaching mechanism is surface dissolution; however, if the data have a straight line slope between 0.35 and 0.65, diffusion is the dominant mechanism. Depletion is indicated if the slope of the line is less than 0.35 in later leaching periods (Torras et al., 2011). As can be seen in Figure 4 1 the dominant leaching mechanism for barium during the leaching intervals is diffusion.
77 Figure 4 1 Cumulative a queous p hase r elease of b arium from c oncrete m onoliths as a function of leaching t ime y = 0.4692x + 0.6262 R = 0.9858 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 -2 -1 0 1 2 t ) [mg/m 2 ] Log (t) [days] Set A y = 0.4817x + 0.5476 R = 0.9795 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 -2 -1 0 1 2 t ) [mg/m 2 ] [days] Set B y = 0.4549x + 0.594 R = 0.9897 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 -2 -1 0 1 2 t ) [mg/m 2 ] [days] Control
78 Synthetic Precipitation Procedure The results of SPLP analysis can be seen in Table 4 4 and 4 5 for the BFD, cement, cement/BFD blends, and the crushed concrete samples, respectively. The final pH of SPL P extracts did not significantly deviate from the natural pH of the BFD, cement or concrete. This demonstrates that the release of trace metals in the SPLP analyses in largely due to the dissolution of oxidized, anhydrous phases they were initially bound t o (Duchesne and Reardon, 1998). The following elements were below the detection limit of the analytical instrument used for all samples analyzed; Ag, As, Cd, and Pb. Additionally, selenium and vanadium were below detection limits in the crushed concrete sa mples. Note that measurements for mercury are in units of ng/L. The calculation of the total percent released of each element showed, barium (42.3% released) and selenium (29.3% released) were the most mobile elements in the BFD; while the other trace elem ents released <7% of their total mass from the BFD. Barium was the constituent of most interest since it was highly mobile and had significant concentrations in both the BFD and cement. It should be noted that only ~2% of the total mercury leached from the BFD samples. Despite arsenic being above the Florida SCTL, the leaching of arsenic was minimal and below GCTL and detection limits for all samples. However, selenium and vanadium both leached above the Florida GCTL for the BFD samples.
79 Table 4 4 SPLP r esults for BFD, c ement, and c ement/BFD b lend s amples Element Cement Sample A* (mg/L) BFD* (mg/L) 5% BFD Blend* (mg/L) 10% BFD Blend* (mg/L) GCTL (mg/L) Ag < 0.0025 <0.0025 <0.0025 <0.0025 0.1 As < 0.0096 < 0.0096 < 0.0096 < 0.0096 0.01 Ba 1.22 0.440 (0.350 0.630) 1.47 1.60 2 Cd < 0.0006 < 0.0006 < 0.0006 < 0.0006 0.005 Cr 0.278 0.0254 0.172 0.149 0.1 Hg** (ng/L) 7.06 1070 (677 1575) 7.22 8.94 2000 Pb < 0.0085 < 0.0085 < 0.0085 < 0.0085 0.015 Se < 0 .0155 0.0638 (0.0251 0.0859) 0.0253 0.0232 0.05 V < 0.0014 0.147 (0.0832 0.215) < 0.0014 < 0.0014 0.049 Zn 0.057 0.240 (0.0356 0.500) 0.0552 0.0586 5 *All values are replicate average; **Note mercury in ng/L Table 4 5 SPLP of c rushed c oncrete s amples a mended with BFD s ample D Element Control* (mg/L) Set A* (mg/L) Set B* (mg/L) GCTL (mg/L) Ag < 0.0025 < 0.0025 < 0.0025 0.1 As < 0.0096 < 0.0096 < 0.0096 0.01 Ba 0.718 0.0327 0.848 0.0191 0.798 0.0145 2 Cd < 0.0006 < 0.0006 < 0.0006 0.005 Cr 0.0205 0.0005 0.0238 0.0003 0.0226 0.0007 0.1 Hg** (ng/L) 36.0 6.9 48.0 5.9 32.0 2.2 2000 Pb < 0.0085 < 0.0085 < 0.0085 0.015 Se < 0.0155 < 0.0155 < 0.0155 0.05 V < 0.0014 < 0.0014 < 0.0014 0.049 Zn 0.0318 0.0064 0.0288 0.0079 0.0321 0.0073 5 *All values are replicate average; **Note mercury in ng/L The results of the SPLP analysis on the BFD/cement blends can be seen in Table 4 4 Only chromium was above the Florida GCTL; while bo th selenium and vanadium were now below the GCTL, likely due to a dilution of both elements in the blends. It can
80 also be concluded that for the BFD/cement blends it is likely that the chromium released in the SPLP results is largely due to the contributio n from the higher initial chromium concentration of the cement and not the BFD, see Table 4 4 Additionally, with the exception of selenium, mercury and barium, all concentrations of trace metals are lower for the blends than the cement. These results supp ort the conclusion that, at the additions of BFD currently allowed by ASTM or AASHTO standards, the amended cement represents no increased risk to worker or environmental health as compared to cement alone. The results of the SPLP on the crushed concrete s amples showed that all elemental concentrations were lower than those SPLP results seen in the BFD Sample D or Cement Sample A used in the concrete mix, see Table 4 5 Comparison of the SPLP results on the control and BFD amended samples shows little diff erence on leached trace metal concentrations. This result strongly supports the conclusion that little to no effect can be seen on the leachability of metals due to the addition of BFD in the amounts currently allowed. Furthermore, it can be concluded, tha t the trace metals that do leach are largely from the cement and not the small mass addition of BFD. pH Dependent Leaching Procedure As previously mentioned, it is important to go beyond the commonly utilized leaching protocols, i.e. TCLP or SPLP, and inve stigate leaching under pH dependent scenarios and weathering conditions. Acid attacks the cement through the permeation of the pore structure and dissolution of ions back through the chemically altered surface layer into solution. This reaction is similar to the natural weathering of silica rich minerals in the environment (Shively et al. 1986). These conditions can be simulated by the pH dependence leaching test (EPA Method 1313). The maximum mass of
81 constituent released over the range of method pH conditi considered an estimate of the maximum mass of the constituent leachable under field leaching conditions for intermediate time frames and the domain of the laboratory test pHs (EPA, 2010). Although, it should be noted that, the relationships between eluate concentrations observed from this method and field leachate must be considered in the context of the material being tested and the field scenario being evaluated. The results of the pH dependence test on BFD Sample D can be s een in Figure 4 2 As can be seen in Figure 4 2 metals leached in varying concentrations at different eluate pHs. Barium and zinc released largely in the more acidic regime (pH = 4 6); this is indicative of cationic species (Garrabrants et al., 2010). Vanadium and selenium displayed an amphoteric nature with highest release at low and high pH values. Release of chromium dominated in the basic regime; this indicates the mobility of chromium hydroxides and formation/dissolution of chromium salts (e.g. Cr O 4 Cr 2 O 7 ) (Garrabrants et al., 2010). Selenium, vanadium and barium were all above the Florida GCTL at some pH regimes. Both selenium and vanadium were above GCTL at natural pH values (pH = 11.5), i.e. similar to the SPLP analysis, while barium was abov e the GCTL only in the highly acidic regime (pH ~ 2). Mercury showed no identifiable trend of release; w hile silver, arsenic, cadmium, and lead were below detection limits of the instrument. The results from the pH variable leaching test (Method 1313) on t he crushed concrete can be seen in Figures 4 3 and 4 4 The following elements released in a manner characteristic of cationic release: As, Ba, Cd, Pb, Se, and Zn. Chromium and vanadium are observed having some amphoteric nature; while the increased releas e of
82 mercury at neutral pH values can be explained by the formation of oxyanion complexes. Furthermore, the release of trace metals at different pH regimes in the concrete as compared to the BFD samples can be explained by the influence of the siliceous ma terial contained in the concrete aggregate (Shively et al. 1986). The following elements were above the GCTL; arsenic, barium, cadmium, chromium, lead, vanadium and zinc. However, the elements above the GCTLs were only above the regulatory threshold in the acidic regime (pH < 6). Such acidic conditions are unlikely to occur in field conditions, since this would cause dissolution of the concrete. It should be noted that there is no observable trend of increased leaching from concrete samples amended with BFD This fact may be due to the limited amount of BFD allowed as an addition. This leads to the conclusion that the leaching of trace elements seen in Figures 4 3 and 4 4 are largely due to the presence of cement in the concrete mix and not the BFD.
83 Figure 4 2 pH d ependent l eaching p rocedure r esults on BFD s ample D and c omparison to GCTL 0 0.5 1 1.5 2 2.5 3 3 4 5 6 7 8 9 10 11 12 Ba, mg/L pH GCTL = 2 mg/L 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 3 4 5 6 7 8 9 10 11 12 Hg g/L pH GCTL = 2 g/L 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 3 4 5 6 7 8 9 10 11 12 Cr, mg/L pH GCTL = 0.1 mg/L 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 3 4 5 6 7 8 9 10 11 12 Se, mg/L pH GCTL = 0.05 mg/L 0 0.05 0.1 0.15 0.2 0.25 0.3 3 4 5 6 7 8 9 10 11 12 V, mg/L pH GCTL = 0.049 mg/L 0 0.2 0.4 0.6 0.8 1 1.2 3 4 5 6 7 8 9 10 11 12 Zn, mg/L pH GCTL = 5 mg/L
84 Figure 4 3 pH d ependent l eaching p rocedure r esults on c onc rete a mended with BFD s ample D and c omparison to GCTL 0.001 0.01 0.1 1 10 0 2 4 6 8 10 12 14 As (mg/L) pH Set A Set B Control GCTL = 0.01 mg/L 0.1 1 10 0 2 4 6 8 10 12 14 Ba (mg/L) pH Set A Set B Control GCTL = 2 mg/L 0.0001 0.001 0.01 0.1 1 0 2 4 6 8 10 12 14 Cd (mg/L) pH Set A Set B Control GCTL = 0.005 mg/L 0.01 0.1 1 10 0 2 4 6 8 10 12 14 Cr (mg/L) pH Set A Set B Control GCTL = 0.1 1 10 100 1000 0 2 4 6 8 10 12 14 Hg (ng/L) pH Set A Set B Control GCTL = 2000 ng/L
85 Figure 4 4 pH d ependent l eaching t est r esults on c oncrete a mended with BFD s ample D and c omparison to GCTL Multiple Extraction Procedure The MEP can account for solid liquid phase equilibrium (if it exists) and weathering of solid samples from exposure to multiple extractions of acidic leachate. The results of the MEP on the BFD Sample D can be seen in Figure 4 5 Silver, cadmium, chromium, and lea d below the detection limit of the instrument. After each sequential extraction, concentrations in the extraction fluid decreased with the exception of arsenic; which was initially below the detection limit but after two sequential extractions the concentr ation of arsenic was recorded above the detection limit and the GCTL. This result can be explained by the weathering of the solid phase by acid attack from the extraction fluid 0.001 0.01 0.1 1 10 100 0 2 4 6 8 10 12 14 Zn (mg/L) pH Set A Set B Control GCTL = 5 mg/L 0.001 0.01 0.1 1 10 0 2 4 6 8 10 12 14 Pb (mg/L) pH Set A Set B Control GCTL = 0.015 mg/L 0.001 0.01 0.1 1 0 2 4 6 8 10 12 14 Se (mg/L) pH Set A Set B Control GCTL = 0.05 mg/L 0.0001 0.001 0.01 0.1 1 10 0 2 4 6 8 10 12 14 V (mg/L) pH Set A Set B Control GCTL = 0.049 mg/L
86 allowing additional bounded arsenic to be released (Poon, et al., 1985). Additi onally, selenium and vanadium were above the GCTL for some of the extractions; this is unsurprising when taking into account the release of these elements during the SPLP analysis. The results from the multiple extraction procedure performed on the crushed concrete samples can be seen in Figure 4 6 Only barium, chromium, zinc and vanadium were above the detection limits of the ICP AES. As can be observed from the graphs, chromium and barium followed a predictable pattern of decreasing release after each ex traction likely due to depletion of the initial barium and chromium found in the concrete samples. The release of zinc was variable and showed no discernible pattern. Release of vanadium exhibited an interesting trend, with increasing release over several extractions. A possible explanation is that the solubility of vanadium was controlled by the dissolution of another mineral or compound it was complexing with. Chromium was above the GCTL for the first extraction but fell below the regulatory threshold for all other sequential extractions.
87 Figure 4 5 Multiple e xtraction p rocedure r esults on BFD s ample D and c omparison to GCTL 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 1 2 3 4 5 6 7 8 9 Ba, mg/L Sequential Extraction GCTL = 2 mg/L 0 0.005 0.01 0.015 0.02 0 1 2 3 4 5 6 7 8 9 Hg g/L Sequential Extraction GCTL = 2 g/L 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 1 2 3 4 5 6 7 8 9 Se, mg/L Sequential Extraction GCTL = 0.05 mg/L 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0 1 2 3 4 5 6 7 8 9 V, mg/L Sequential Extraction GCTL = 0.049 mg/L 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 1 2 3 4 5 6 7 8 9 Zn, mg/L Sequential Extraction GCTL = 5 mg/L 0 0.01 0.02 0.03 0.04 0.05 0.06 0 1 2 3 4 5 6 7 8 9 As, mg/L Sequential Extraction GCTL = 0.01 mg/L
88 Figure 4 6 Mult iple e xtraction p rocedure r esults on c oncrete a mended with BFD s ample D and c omparison to GCTL Conclusions The total (mg/kg) and leachable (mg/L) concentrations of trace inorganic elements in BFD (a partially calcined byproduct of cement clinker production ) were evaluated and compared to regulatory threshold levels, e.g. Florida SCTLs and GCTL, to assess the potential for increased environmental risk associated with the use of the BFD. The total concentrations of elements in the BFD were assessed as part o f a risk screening evaluation; they were compared to concentrations in CKD, cement, and in a final concrete product. While the BFD collected was of similar composition to CKD, mercury was greater in concentration in the BFD than the historical values foun d for CKD. While 0 0.02 0.04 0.06 0.08 0 1 2 3 4 5 6 7 8 9 V (mg/L) Sequential Extraction Set A Set B Control GCTL = 0.049 mg/L 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 1 2 3 4 5 6 7 8 9 Ba (mg/L) Sequential Extractions Set A Set B Control GCTL = 2 mg/L 0 0.05 0.1 0.15 0.2 0 1 2 3 4 5 6 7 8 9 Cr (mg/L) Sequential Extraction Set A Set B Control GCTL = 0.1 mg/L 0 0.01 0.02 0.03 0.04 0.05 0 1 2 3 4 5 6 7 8 9 Zn (mg/L) Sequential Extraction Set A Set B Control GCTL = 5 mg/L
89 the BFD did contain arsenic above Florida SCTLs, the leaching of arsenic was minimal. The BFD did release selenium and vanadium above GCTLs when evaluated using the release if managed properly. Leaching of trace metals from cement or concrete products amended with BFD was minimal and was only above regulatory thresholds in highly acidic weathering conditions. The experimental results from the samples collected in thi s study, along with an estimated final concrete composition based on individual measurements, suggest that the final concrete product made from similar materials should not be expected to differ dramatically in overall composition from that of concrete mad e without BFD. The comparison of results of the different leaching assays to regulatory threshold levels do not suggest that leaching of trace metals from concrete products manufactured with BFD of similar composition as tested in this study and at similar mix designs will pose additional risk beyond any normal (without BFD) concrete production use.
90 CHAPTER 5 S UMMARY AND CONCLUSIO NS Summary of Work Conducted Recent changes in air pollution control at coal power plants and cement kilns has resulted in altered combu stion residues. Two combustion residues, ammoniated coal fly ash (AFA) and baghouse filter dust (BFD), are currently being reused as cement or concrete additive. It is necessary to investigate the environmental impact of reusing these novel combustion resi dues. The present study investigated the potential leaching of ammonia or selected trace metals ( Ag, As, Ba, Cd, Cr, Hg, Pb, Se, V, and Zn ) from cement/concrete products amended with AFA or BFD. The assessment of leaching was conducted with three separate goals in mind; namely, determination of ammonia content in the AFA and elemental composition in the BFD, leaching of ammonia or trace metals from concrete in monolithic form amended with AFA or BFD, and leaching of ammonia or trace metals from crushed conc rete amended with AFA or BFD. The leaching assessment of concrete monoliths simulates leaching of contaminants from scenarios in which AFA or BFD are used in a construction project or as part of the built environment, e.g. bridge pilings, sidewalks. The le aching assessment of crushed concrete simulates the reuse of concrete structures by size reduction, e.g. road base, aggregate in new concrete. Risks from leaching of contaminants in both leaching assessments were determined by comparison to appropriate Flo rida CTLs. Major Conclusions From t he experimental results the following conclusions can be made: Ammonia readily leaches from fly ash nearly 100% and crushed concrete, 58 6 8 %, whil e monolith leaching of ammonia is less, 18 37%. Use of fly ash with e levated levels of ammonia in concrete products does pose a risk to the environment; however, a leachability threshold
91 concentration for the protection of surface waters can be derived from the monolith leaching of ammonia and was shown to be 350 mg NH 3 /kg fly ash or 35 mg NH 3 /kg concrete. The largest mass release of ammonia from concrete monoliths occurred during the first week of leaching due to initial adsorption of water; after one week the leaching was due to diffusional release of ammonia and continuou sly decreased over leaching intervals due to decreased concrete permeability. Trace metals in concrete amended with BFD were well encapsulated (or too dil ute) in monoliths and size reduced concrete samples to raise concern due to leaching and potential ris k to human health and the environment. The BFD bulk composition was comparable to cement collected from the same facility and trace composition comparable to historical ranges for CKD (Haynes and Kramer, 1982; PCA, 1992). Due to small addition of BFD to th e final cement product at cement kilns (5% by mass; ASTM C465) it is unlikely that BFD will alter the composition of cement/concrete significantly or will threaten health or the environment under most concrete applications and reuse scenarios Additional Research Needs The use of admixtures in concrete mixes, e.g. air entraining agents and water reducing agents, may significantly affect the permeability and thereby the loss of ammonia and trace metals from concrete. Additional research needs to be directe d at the use of these admixtures and the loss of ammonia or trace metals from concrete. It is known that AFA will be used in some concrete mixes; the effect of ammonia on trace metals leaching from concrete warrants investigation. The long term fate (> two months) of ammonia encapsulated in concrete amended with AFA is not known. It is not clear that if under reuse scenarios of aged concrete amended with AFA presents residual release of encapsulated ammonia at regulatory significant levels. While it is unli kely the levels of trace metals found in concrete amended with BFD used in this study will not present significant threat to human health and the environment; the levels of trace
92 metals in BFD will vary upon location, time, and fuel type used in kilns. A b roader survey of trace metal levels found in BFD from other facilities is needed.
93 LITERATURE CITED American Coal A sh Association, 2011. 2011 Coal Combustion Product (CCP) Production & Use Survey Report. ACAA, Farmington Hills, MI. American Society of Test in g and Materials, 2013a. C465 10 Standard Specification for Processing Additions for Use in the Manufacture of Hydraulic Cements, Vol. 04.01. ASTM International, West Conshohocken, PA American Society of Testing and Materials, 2013b C618 12a Standard S pecification for Coal Fly Ash and Raw or Calcined Natural Pozzolan for Use in Concrete, Vol. 04.02. ASTM International, West Conshohocken, PA. American Society of Testing and Materials, 2013c. C150/C150M 12 Standard Sp ecification for Portland Cement Vol. 04.01. ASTM International, West Conshohocken, PA. American Society of Testing and Materials, 2013d. C33/C33M 13 Standard Speci fication for Concrete Aggregates Vol. 04.02. ASTM International, West Conshohocken, PA. American Society of Testing and Materi als, 2013 e D7348 08e1 Standard Test Methods for Loss on Ignition (LO I) of Solid Combustion Residues Vol. 05.06. ASTM International, West Conshohocken, PA. American So ciety of Testing and Materials, 2013 f C311/C311M 13 Standard Test Methods for Sampling and Testing Fly Ash or Natural Pozzolans for Use in Portland Cement Concrete, Vol. 04.02. ASTM International, West Conshohocken, PA. Barna, R., Sanchez, F., Mosz kowicz, P. and Mehu, J., 1997. Leaching Behavior of Pollutants in Stabilized/Solidi fied Waste s Journal of Hazardous Materials, 52, 287 310. Bobirica, C., Bobirica, L., Stanescu, R., and Constantinescu, I., 2010. Leaching Behavior of Cement Based Solidified Wastes Containing Hexavalent Chromium U.P.B. Sci. Bulletin, Vol. 72, Series B, Iss. 1. C ardone, C., Kim, A., Schroeder, K. 2005. Release of Ammonia from SCR/SNCR Fly Ashes. 2005 World of Coal Ash Conference Lexington Kentucky Cheng, K. and Bishop, P., 1990. Developing a Kinetic Leaching Model for Solidified/Stabilized Hazardous Wastes. J ournal of Hazardous Materials 24, 213 224.
94 Duch esne, J., Reardon, E.J., 1998. Determining Controls on Elemental Concentrations in Cement Kiln Dust Leachate. Waste Management, 18, 339 350. Florida Department of Environmental Protection 2005. Technical Report: Development of Cleanup Target Levels (CTLs) for Chapter 62 777, F.A.C. Tallahassee, F lorida Florida Department of E nvironmental Protection, 2009. Guidance for Determining Leachability by Analysis of SPLP Results. FDEP Bureau of Waste Cleanup. T allahassee, FL. Florida Department of En vironmental Protection (2013). CEMEX Construction Materials Florida, LLC Brooksville South Cement Plant, Title V Air Operation Permit Renewal (Permit No. 0530021 047 AV). Tallahassee, FL. Florida Department of Envir onmental Protection. Florida Department of Transportation, 2010. Standards for Specifications for Road and Bridge Construction, Tallahassee, Florida. Garrabrants, A., Kosson, D., Van der Sloot, A., Sanchez, F., Hjelmar, O., 2010. Background Information f or the Leaching Environmental Assessment Framework (LEAF) Test Methods, Research Triangle Park, North Carolina. Glicksman, M. E. (2000). Diffusion in Solids: Field theory, Solid State Principles, and A pplications. John Wiley & Sons, New York City, New Yor k. Ha ynes, B. and Kramer, G., 1982. An Analysis of Selected Trace Metals in Cement and Cement Kiln Dust. United States Bureau of Mines, Washington, D.C. Hillier, S.R., Sangha, C.M., Plunk ett, B.A., Walden, P.J., 1999. Long Term Leaching of Toxic Trace Me tals from Portland Cement Concrete. Cement and Concrete Research, 515 521. Kosmatka, S., Kerkhoff, B., Panarese, W., 2008. Design and Control of Concrete Mixtures, fourteenth ed. Portland Cement Association, Skokie, Illinois. Kosson, D., Sanchez, F., K ariher, P., Turner, L.H., Delapp, R., Seignette, P. 2009. Characterization of Coal Combustion Residues from Electric Utilities Leaching and Characterization Data Washington, D.C. Pistilli, M.; Majko, R., 1984. Optimizing the Amount of Class C Fly As h in Concrete Mixtures Cement and Concrete Aggregates, 6, (2).
95 Poon, C.S., Peters, C.J., Perry, R., 1985. Mechanism of Heavy Metal Stabilization by Cement Based Fixation Processes. The Science of the Total Environment, 55 71. Port land Cement Association, 1992. An Analysis of Selected Trace Metals in Cement and Kiln Dust. Portland Cement Association. Skokie, Illinois. Rathbone, R. and Robl, T ., 200 3 A Study of the Effects of Post Combustion Ammonia Injection on Fly Ash Quality: Characterizat ion of Ammonia Release from Concrete and Mortars Containing Fly Ash as a Pozzolanic Admixture Washington, D.C. Serclerat, I., Mos zkowicz, P., Pollet, B., 2000. Retention Mechanisms in Mortars of the Trace Metals Contained in Portland Cement Clinkers. Was te Management, 259 264. Shively, W., Bishop, P ., Gress, D., Brown, T., 1986. Leaching Tests of Heavy Metals Stabilized with Portland Cement. Water Pollution Control Federation, 234 241. Taylor, P., 2008. ts on Processing Additions in Cement Manufacturing. Washington, D.C.: National Cooperative Highway Research Program. Torras, J., Buj, I., Rovira, M., Pablo, J., 2011. Semi Dynamic L eaching Tests of Nickel ium Potassium Phosphate C ements. Journal of Hazardous Materials 186, 1954 1960. United States Environme ntal Protection Agency (1986). Method 1320 Multiple Extraction Procedure. Washington, D.C. U.S. Environmental Protection Agency. United States Enviro nme ntal Protection Agency (1994). Method 1312 Synthetic Precipitation Leaching Procedure. Washington, D.C. U.S. Environmental Protection Agency. United States Environmental Protection Agency (1996). Method 3050b Acid Digestion of Sediments, Sludges, an d Soils. Washington, D.C. U.S. Environmental Protection Agency. United States Environmental Protection Agency (2005). Method 3200 Mercury Species Fractionation and Quantification by Microwave Assisted Extraction, Selective Solvent Extraction and/or Solid Phase Extraction. Washington, D.C. U.S. Environmental Protection Agency.
96 United States Environmental Protection Agency (2007). Method 7474 Mercury in Sediment and Tissue Samples by Atomic Fluorescence S pectrometry Washington, D.C. U.S. Environmental Pr otection Agency. United States Environmental Protection Agency (2009). Preliminary Version of Method 1315 Mass Transfer Rates of Constituents in Monolithic or Compacted Granular Materials Using a Semi Dynamic Tank Leaching Procedure. Washington, D.C. U.S Environmental Protection Agency. United States Environme ntal Protection Agency (2010). Preliminary Version of Method 1313 Liquid Solid Partitioning as a Function of Extract pH in Solid Materials Using a Parallel Batch Procedure. Washington, D.C. U.S. E nvironmental Protection Agency. United States Environmen tal Protection Agency. (2011). EPA Materials Characterization Paper: Final Rulemaking: Identification of Nonhazardous Secondary Materials That Are Solid Waste Coal Combustion Residuals Coal Fly Ash, Bottom Ash, and Boiler Slag. Washington, D.C.: United States Environmental Protection Agency. United States Environmen tal Protection Agency. (2012). Cement Kiln Dust Waste. U.S. Environmental Protection Agency. Retrieved June, 21, 2013. From http://www.ep a.gov/osw/nonhaz/industrial/special/ckd/. Van der Sloot, H.A., 2002. Characterization of the Leaching Behavior of Concrete Mortars and of Cement Stabilized Wastes with Different Loadings for Long Term Environmental Assessment. Waste Management 181 186 Wang, H., Ban, H., Golden, D., Ladwig, K., 2002. Ammonia Release Characteristics from Coal Combustion Fly Ash. Fuel Chemistry Division 47(2), 836. Wilson, C., 2011. Water Quality Notes: pH. University of Florida: Institute of Food and Agricultural Scie nces Extension (IFAS). Gainesville, FL.
97 BIOGRAPHICAL SKETCH Joshua Bradley Hayes was born in Palatka, Florida in the United States of America. After graduating from Palatka High School in August 2001 and still unsure of his career path; he began taking (SJRCC) and working part time. In May 2007, he graduated from SJRCC with an Associate of Arts degree with a focus on pre engineering curriculum. In August 2007, Joshua Hayes began classes at the University of Florida, Gainesville. He chose to major in environmental engineering; focusing on solid waste management, environmental impact assessment and mitigation, water supply and treatment, and air pollution management. In May 2011, he graduated from the Universit y of Florida with a Bachelor of Science degree in environmental engineering. In August 2011, Joshua Hayes was admitted to graduate school at the University of Florida. He focused on solid and hazardous waste management and was mentored by Dr. Timothy Townsend, Ph.D., P.E. Joshua Hayes was involved in several research projects but was primarily concerned with the reuse of combustion by product, e.g. coal fly ash and cement kiln dust, in concrete applications. This thesis is a partial fulfillment of the requirements for the degree of master of engineering from the University of Florida.