Recycled Materials and Byproducts in Highway Applications—Manufacturing and Construction Byproducts, Volume 8 (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

84 Sulfur ByproductS Background A major byproduct from the oil and gas industries is brim- stone, which is essentially elemental sulfur (Shell 2010). Sul- fur, in the form of sulfuric acid, is also a byproduct of ferrous and nonferrous metal smelting. The use of sulfur as a binder to produce a construction material has been explored for more than a century (McBee et al. 1985). These early efforts used the sulfur as the binder in mortars and concretes to produce acid-resistance mixes with good strength. Research in the mid- 1930s discovered that thermal properties of the sulfur mixes could be improved by adding an olefin polysulfide, marketed under the name of Thiokol. In the 1940s, sample preparation and specifications for sulfur polymer concrete were standard- ized by ASTM as ASTM C1312 and C1159. Sulfur was first used in asphalt cements in the early 19th century as a product that was minimally sensitive to temperature changes and weathered well. The original use fell out of favor when air-blown asphalts began to be mar- keted. The substitution of sulfur for a portion of the asphalt cement was investigated in the late 1930s; however, addi- tional development of sulfur-extended asphalts did not come along until the mid-1970s when the oil embargo increased the cost of crude oil and the availability of asphalt cement was limited. Uses of sulfur as an extender (i.e., replace- ment) for asphalt cement in HMA were researched. Highway applications for sulfur include sulfur extended asphalt (SEA) and SC. Sulfur, a naturally occurring compo- nent in asphalt, can be substituted for the more expensive portland or asphalt cement. Sulfur was most commonly com- bined with polymers and aggregates to produce sulfur poly- mer concrete starting in the early 1990s. The main uses were as a rapid repair mix and to encapsulate hazardous materials (Mattus and Mattus 1994). Additional information can be found at the following website: Sulfur Institute: http://www. sulphurinstitute.org/. Initial evaluations of the physical properties of sulfur showed the internal structural characteristic transition upon heating at about 212°F (100°C), which results in a decrease in volume (Mattus and Mattus 1994). When subjected to ther- mal cycling, the sulfur tended to disintegrate if some form of stabilizing admixture was not used. Shrinkage problems were overcome by using 5% of dicyclopentadiene and oligomers of cyclopentadiene (used in equal amounts to make the 5% admixture). This combination of sulfur and polymers was des- ignated as sulfur polymer cement (SPC). When the SPC was combined with aggregates, the material was referred to as SC. The physical propertied of sulfur polymer concrete at one day are comparable to those of conventional PCC at 28 days (Table 108). Although the air voids for both types of con- crete are similar, the air voids in the sulfur polymer concrete are not interconnected as they are in conventional PCC. This property makes the sulfur mix impermeable. Engineering properties Thermal coefficient of expansion for sulfur was reported as 46 µin./in.-°C between 25°C and 95°C and increased to 1,000 µin./in.-°C between 95°C and 108°C (McBee et al. 1985). Environmentally related properties Research was conducted to evaluate the potential for biological degradation as some bacteria attack elemental sulfur (Mattus and Mattus 1994). Tests at the Brookfield National Laboratory found no bacterial (e.g., thiobacilly bacteria) or fungi activity after 21 days, incubation at 95°F to 99°F (35°C to 37°C), rela- tive humidity greater than 85%, and in the presence of nutrient agar to sustain growth. However, under certain specific circum- stances such as cooling towers and in the presence of pyrite containing rock, sulfur polymer concrete was susceptible to bacterial attack. The bacterial attack resulted in the formation of sulfuric acid (metabolic byproduct). use and production Petroleum refining increased an average of 1.08% per year for 1971 through 2000. During the same period, the techno- logical advances in emissions controls at refineries resulted in a 6.89% increase in recovered sulfur (Ober 2002; Schneider 2006). The trend in production of recovered sulfur is shown in Figure 21. costs As of 2006 there was at least 15 million tonnes of stored sul- fur from oil-sand oil upgraders, oil refineries, and natural gas processing plants (Schneider 2006). The storage of sulfur was an expense to these industries because of the cost of treating and neutralizing acidic water runoff. The cost of storage was as much as $3.00 per tonne in 2006 (Schneider 2006). chapter six Sulfur and SulfatE WaStE

85 Disadvantages to using sulfur polymer concrete were: • Material will not burn on its own but still meets the criteria for the U.S.DOT regulations as flammable material. • Viscosity rises sharply when the temperature is more than about 320°F (160°C). The additional polymeriza- tion in the sulfur makes the liquid that is “gummy and unpourable.” • Above 320°F (160°C) hydrogen sulfide gas or sulfur dioxide forms are poisonous and flammable. Mattus and Mattus (1994) investigated sulfur to encap- sulate low specific gravity materials such as fly ash. These researchers noted that if the sulfur and fly ash was kept hot literature review Bound Applications—Sulfur Concrete Benefits to using sulfur in concrete (Micropowder 2010) were identified as: • Sulfur polymer concrete: – Gains strength rapidly (about 80% within a few hours of placement) – Resistant to acids such as sulfuric acid, hydrochloric acid, and nitric acid – Durable in corrosive environments – High density – Resists cracking – Resists plastic deformation. FIGURE 21 Production of sulfur from refinery processes (after Ober 2002). 1972 1982 1992 1999 Su lfu r, M ill io n To ns /Y r Petroleum , M illion Barrels/D ay 18 16 14 12 10 8 6 4 2 0 9 8 7 6 5 4 3 2 1 0 Recovered Sulfur Petroleum Property Sulfur Polymer Concrete Conventional Portland Cement Concrete Mix Component Sulfur Polymer Cement, %wt 14 to 18 0 9 ot 6 0 tw% ,retaW 0 9 ot 6 tw% ,relliF lareniM 81 ot 21 0 tw% ,tnemeC dnaltroP 03 24 ot 83 tw% ,dnaS 54 73 ot 33 tw% ,etagerggA esraoC Hardened Properties syad 82 ta seitreporP yad 1 ta seitreporP Compressive Strength, psi 7,000 to 10,000 3,500 to 5,000 005 005,1 ot 000,1 isp ,htgnertS elisneT 535 000,2 ot 053,1 isp ,htgnertS laruxelF Elastic Modulus, 106 0.4 0.4 isp 0.3 ot 3.0 1.0 ot 0.0 % ,noitprosbA erutsioM 0.4 0.6 ot 0.3 % ,tnetnoC dioV riA Impact Strength, compressive, ft-lb 100 to 119 81 Impact Strength, flexural, ft-lb 0.3 to 0.5 0.2 01.0 ot 60.0 21.0 ot 80.0 % ,egaknirhS raeniL Coefficient of Thermal Expansion, µin./in. 14.0 to 14.7 5.2 5.2 ot 4.2 ytivarG cificepS After Mattus and Mattus (1994). TABLE 108 REPORTED PROPERTIES OF SULFUR POLyMER CONCRETE COMPARED WITH CONVENTIONAL PCC

86 The results showed the combination of fly ash filler and sulfur modified with crumb rubber using the wet process pro- duced the highest strength of any of the mixes (Figure 22). Regardless of the method of adding the rubber or type of filler the strength decreased with increased rubber content. Adeh et al. (2008) evaluated the use of sulfur, additives, mineral fillers, and aggregates in the development of a SC mixture. The research studied the influence of blending tem- peratures on mortar strengths (Table 109). Three different approaches were used to blend the sulfur and olefinic addi- tive. The first of the three methods blended the materials at 130°C to 140°C and mixed for 3 h. The second method blended the two materials at 150°C to 160°C for 3 h. As of 2010, Shell (2010c) was marketing a proprietary product, Thiocrete. Shell noted that Thiocrete had a lower carbon footprint when compared with conventional portland cement products. Even though Thiocrete was mixed at high temperatures during production, the total process used less energy than conventional cement production. (i.e., liquid) for a period of time, the fly ash floated to the top of the hot sulfur. Preferred materials in need of immobilization were inorganic, low solubility compounds (e.g., sludges, pre- cipitates, and incinerator ash). Combining sulfur with organic materials ion exchangers and highly soluble compounds was not recommended. Construction concerns noted the need for a mixer with a heated mixer and paddle. A weir was needed for flow control as valve control was not successful. Mixing speed needed to be controlled and slow enough to prevent air entrainment. One concern was too rapid cooling of the mix under field conditions as the impermeability of the sulfur polymer con- crete was compromised as a result of void formation. The nonsulfur solids needed to have very low moisture contents for proper mixing and hardened properties. Xi et al. (2004) explored the use of crumb rubber and sul- fur to prepare sulfur rubber concrete (SRC). Preparation of the SRC preheated all of the materials to 130°C to 146°C (i.e., sulfur, natural aggregates, rubber particles, and mineral fillers). The rubber was used to replace a portion of the finer natural aggregates. Two general sizes of crumb rubber were used with the large size having an average size of 4.12 mm. The small particles had an average size of 1.85 mm. The per- cent of replacement ranged from zero to 50%. Two types of mineral filler were evaluated (fly ash and portland cement). Materials were combined at different mixing temperatures and processing techniques (wet and dry). The wet process for preparing SRC involved mixing the sulfur with rubber particles and holding the blend at mixing temperature for a period of time. The dry process mixed the rubber particles with the aggregates before mixing. St re ng th , M Pa Rubber, % by Volume 60 20 30 40 50 0 10 302010 050 40 No filler & dry process Fly ash & wet process Portland cement (10%) & dry process Portland cement (14%) & dry process FIGURE 22 Influence of crumb rubber and sulfur as binder and aggregate replacements in conventional PCC (after Xi et al. 2004). TABLE 109 INFLUENCE OF BLENDING TIME ON COMPRESSIVE STRENGTHS OF SULFUR MODIFIED MORTARS Compressive Strength (3 days) of Mortars with Different Blending Temperatures Used for Blending Sulfur and Olefinic Additive, kg/m3 140o 061 C oC 401 302 211 802 28 642 621 273 023 egarevA 601 egarevA 862 After Adeh et al. (2008).

87 Bound Applications—SEA Mahoney et al. (1982) reported the results from placing SEA mixes at the Washington State University test track at Pullman, Washington. The test track used a circular layout where the loading wheel revolves around the center point of the circle. The test track was capable of accommodating 12 test sections. This study used two thicknesses, one surface mix, three types of base top lift, two types of base bottom lift, and two types of subgrades. Mix designs were conducted for three ratios of sulfur to asphalt (0:100, 30:70, and 50:50). Based on an equivalent volume basis, the optimum binder contents were established as 5.5% for the 0:100 ratio (i.e., control mix), and 6.5% and 7.4% for the 30:70 and 50:50 ratios, respectively. Researchers noted the kneading compactor used in the Hveem mix design method resulted in lower air voids when compared with Marshall compaction at the perceived optimum binder content. This would result in a lower optimum asphalt content set for the SEA mixes. The results from the track testing and data analysis showed that the control and SEA mixes had similar fatigue character- istics, but the SEA mixes could accommodate higher bending strains at low levels of repetitions. SEA has been used periodically over the last century (McBee et al. 1985). In 1938, SEA was used to produce a sta- ble mix using 25% sulfur as an asphalt replacement. This early investigation did not become popular because it was not cost- effective until the oil embargo in the 1970s when the increased cost and decreased availability of asphalt cement became a concern. At this time, two approaches were explored for using sulfur as a paving material. The first approach used sulfur as a replacement for the asphalt cement, while the second eval- uated its use as a structuring agent that would allow the use of lower quality aggregates. Societe Nationale Elf-Aquaitaine in Canada developed a proprietary method for pre-blending asphalt and sulfur before introduction into the HMA plant. The U.S. Bureau of Mines developed a process that used a shearing action within the HMA plant. In the first decade of 2000, Shell started to market sulfur pellets for blending SEA binders. Three critical temperatures to be considered when using Thiopave in mixes were 240°F, which was the temperature at which the pellets melted; 300°F, which was the temperature at which H2S generation was likely; and 265°F, which was considered the ideal discharge temperature. For safety reasons, the sulfur extended asphalt mix (SEAM) required the use of warm mix technology that was provided by the use of the wax product. Benefits to using sulfur in asphalt cement mixes (Shell 2010a; Mattus and Mattus 1994) included: • Increased stiffness without becoming brittle at cold temperatures – Use of softer, lower viscosity asphalt cements in cold climates while minimizing rutting problems during hot summer seasons. • Better performance than conventional HMA in extremely hot or cold climates. • Improved overall structural capacity of the pavement system. • Can be reheated since the hardening process is thermo- setting. • Cleaning operations are limited to making sure sul- fur polymer concrete is not contaminated with other materials. McBee et al. (1985) evaluated a number of SEA mixes. One asphalt cement (AR 2000) was modified using six levels of percent volume of sulfur (0%, 15%, 25%, 35%, 50%, and 75%). The original viscosity versus temperature relationship was decreased significantly when the asphalt was replaced with the sulfur. However, the amount of the decrease was dependent on the volume of sulfur. As the volume of sulfur increased, the difference decreased between the unmodified and modified asphalts at any given temperature. In the production of asphalt cement concrete, the amount of each component was proportioned into the plant using weights of each component. The weight of the SEA binder to be added to achieve the optimum binder content needed to be adjusted to account for the sulfur specific gravity with the following equation: % Weight of SEA = − − ( )  A RR S R G 100 100 Where: A = asphalt content in conventional design, %wt; R = sulfur to asphalt substitution ratio; S = sulfur to be used in SEA binder, %wt; and G = specific gravity of asphalt cement. Testing of the SEA mixtures showed significant changes to the properties of the paving material. When the volume of sul- fur in the compacted sample increased, the specific gravity, air voids, Marshall stability, and dynamic modulus increased. The economic advantage for using sulfur as a replacement for asphalt cement was lost in the 1980s and remained lost until about 2008. There has been a resurgence of interest in using sulfur that has led to more advanced methods for intro- ducing the byproduct into the HMA plant. Stuart (1990) documented and compared the performance of SEA roadways to conventional HMA pavements that had been constructed from 3 to 7 years before. Cores were obtained from 18 projects. The sulfur was added to the mix in one of three ways (colloid mill preblending, in-line liq- uid blending, and direct liquid feed). Cores were taken from

88 each of the projects. In some cases, more than one location was cored. A pavement condition survey was completed so that the pavement condition index (PCI) could be calculated (Table 110). The PCI implements the distress deduct values using a standard formulation. The PCI values varied by proj- ect and location but were not statistically different owing to the replacement of the asphalt cement with the sulfur. Laboratory testing included the determination of core prop- erties for diametral resilient modulus, diametral creep, mois- ture susceptibility, and stress-controlled-repeated load fatigue cracking (diametral) testing. The results showed no significant difference because of the sulfur on the resilient modulus prop- erties of the mixes. There was also no significant difference in the creep modulus at 41°F (5°C); however, at 77°F (25°C) the sulfur significantly reduced the permanent deformation mea- sured during creep modulus testing. At 104°F (40°C), sul- fur slightly decreased creep modulus at short loading times and slightly increased the stiffness at longer loading times (i.e., better rut resistance). Sulfur decreased both the tensile strength ratio and the resilient modulus ratio in the older pave- ment cores. The values were 79.8% and 79.1% for the retained tensile strength and retained resilient modulus, respectively, for the conventional mixes, but only 67.4% and 54.9% for the sulfur mixes. The fatigue testing showed that results were similar for about 50% of the mixes and when sulfur did impact the fatigue life the sulfur decreased the fatigue life. The composition of the mixtures was validated using sol- vent extraction to recover both the binder and the aggregates. Extraction could be accomplished using trichloroethylene (TCE) with the reflux method or TCE with a centrifuge if the solvent was heated to 150°F (65.5°C). The Abson recovery method was used to obtain binder samples from the cores. This recovery method significantly softened the SEA binder. Binder testing showed that the sulfur in the SEA tended to settle out while being reheated for test sample preparation. When vacuum viscosities were determined, the sulfur left a film on the sides of the glass tubes, which was difficult to remove. The SEA binder results were too variable to draw all but one conclusion. The viscosity of the 40% sulfur SEA was initially softer than the conventional asphalt by about one specification grade. In 2007, a sulfur extended asphalt mix using 40% of Thiopave by weight of binder was placed in Qatar A conven- tional and HMA pavement was placed as the control section Shell (2010a). Indirect tensile stiffness testing was used to evaluate the change in stiffness over a range of temperatures. Below 86°F (30°C) the surface mixes were close in stiff- ness while the base course mix was about 25% higher than the control. At or above this temperature the stiffness of the base course SEA mix was about 75% higher than the control and the surface course SEA mix was about 25% higher than the control. Information reported by Shell (Palmer 2010) State Age Blending Method PCI Deduct Values for Pavement Distresses Rutting Combined cracking Bleeding Potholes AC SEA AC SEA AC SEA AC SEA AC SEA CA 4.3 C 100 100 0 0 0 0 0 0 0 0 CB-1 3.2 B 100 100 0 0 0 0 0 0 0 0 CB-2 B 100 100 0 0 0 0 0 0 0 0 DE 6.4 B, C 90 85 29 0 29 47 0 0 0 0 GA 4.6 C 87 90 16 0 16 10 0 0 0 4 ID-1 4.0 B 100 100 0 0 0 0 0 0 0 0 ID-2 95 100 0 0 0 0 0 0 0 0 KS 5.0 C 0 0 — — — — — — — — LA 6.0/7.2 B 90 87 0 0 5 11 0 0 0 12 ME–Benton 1 4.1 C — — — — — — — — — — ME–Benton 2 4.1 C 87 92 0 0 47 28 0 0 0 0 ME–Benton 3 4.1 C 87 84 0 0 47 44 0 0 0 0 ME–Crystal 6.2 C 88 80 11 37 8 0 0 0 0 0 MN 7.0 C 49 79 0 0 51 61 72 0 0 26 MS 4.4 C 100 100 0 0 0 0 0 0 0 0 ND-1 4.4 C 82 80 20 14 6 4 0 0 0 0 ND-2 5.2 B 85 83 4 19 11 8 0 0 0 0 NM 3.7 B 95 100 0 0 6 0 0 0 0 0 TX–College Station 7.4 A 57 80 31 9 58 24 0 0 37 16 TX–Pecos 4.2 B 100 100 0 0 0 0 0 0 0 0 TX– Nocogdoches 5.2 C 80 85 0 0 13 24 24 0 0 0 WI 3.6 B 47 83 48 17 5 0 0 0 0 0 WY 3.7 C 82 80 15 0 49 49 0 0 0 0 After Stuart (1990). A = colloid mill preblending. B = in-line blending (liquid). C = direct feed (liquid). TABLE 110 SUMMARy OF PAVEMENT CONDITION FOR SEA AND CONVENTIONAL HMA PAVEMENTS

89 showed that Marshall stability of the conventional HMA was just over 1,800 lb at either 1 or 14 days. The Thiopave SEAM had a stability that was more than 10% greater than the con- ventional mix at 1 day and more than 80% higher at 14 days. Laboratory research in France in 2007 evaluated the low temperature sulfur extended asphalt mix properties (Shell 2010a). The study used three penetration grades without and with 40% Thiopave. The thermal stress restrained specimen test (TSRST) was used to show that the sulfur modifier did not significantly alter the cold temperature properties of the base asphalt. Additional research with two asphalts used in the construction of Chinese test sections (Shell 2010d) showed that the stiffness modulus ratio increased with increased tem- perature (Table 111). Additional research on Thiopave in asphalt concrete at the National Center for Asphalt Technology (NCAT) used stiffness modulus master curves to show the influence of the sulfur additive on material properties. The results of the testing showed that Thiopave provided more resistance to rutting than conventional HMA at slower loads and/or higher tem- peratures (Shell 2010d). Transportation Research Laboratory (TRL) in the United Kingdom evaluated the impact of Thiopave on pavement deflections and design. The analysis showed the Thiopave pavement structure could potentially extend the pavement life by up to 40%. The data and analysis was also used to esti- mate the reduction in layer thickness that could be achieved while still maintaining a stiffness and pavement life similar to conventional mixes. A computer program, BISAR, was used to calculate expected stresses, strains, and deflection in any layer or position. The BISAR inputs were elastic modu- lus, Poisson’s ratio, layer thickness, and loading characteris- tics (i.e., imposed stress, radius of loaded area, and coordi- nates of loads). The results showed the HMA layer could be reduced by 10% while still achieving similar load carrying capabilities to the conventional HMA. Environmental issues about the recyclability of SEA mix- tures were evaluated at Tonghi University in China. The research used cold recycling of Thiopave RAP with emulsions to produce the recycled mix. The recycled SEAM was con- sidered similar in properties to that of a conventionally used cold mix (Shell 2010e). No work using hot recycling was assessed for potential environmental and worker safety con- cerns. Shell noted that the use of Thiopave reduced green- house gas production (Shell 2010c) since it was produced at lower temperatures than conventional HMA. Al-Mehthel et al. (2010) reported on the results from three test sections during construction in Saudi Arabia. The first section was placed on the Khursaniyah access road (0.33 km long, two lane road, 30:70 blend, and conventional), the second on the Shedgum–Hofuf road (0.25 km, two lanes, SEA full depth, SEA wear course, and conventional), and the third on Dhahran–Jubai expressway (500 m long, one lane wide, and 30:70 blend wear course). The Khursaniyah access road was constructed in March 2006 and pavement condition surveys were conducted in September 2006, June 2007, October 2007, September 2008, January 2009, and June 2009. No signs of distresses were observed and the PCI was consistently around 95. The Shedgum–Hofuf road was opened to traffic in early 2009 and the first pavement condition survey was conducted in December 2009. Only minor rutting in the wheel paths was noted from heavily loaded trucks. The Dhahran–Jubail expressway, a heavily trafficked roadway (2,679,465 annual vehicles), had a PCI of 98 in December 2009. Air monitoring at the Khusaniyah construction site is reviewed in Table 112. The sulfur dioxide concentrations ranged from 0 to 8 ppm close to the source (auger), but were lower and acceptable at either the driver and foreman Test Temperature, oC aPG ,52CA aPG ,ssenffitS 02 CA Thiopave Control Stiffness ratio Thiopave Control Stiffness ratio 10 6.4 6.0 1.07 6.5 6.2 1.05 20 6.0 5.0 1.20 6.0 5.2 1.15 30 5.0 3.2 1.56 5.0 3.3 1.52 After Shell (2010d). TABLE 111 INFLUENCE OF THIOPAVE ON MIX PROPERTIES Location of Probe SO2, ppm H2S, ppm Remarks Max. Mean Min. Max. Mean Min. Probe 29 to 49 cm over Auger 3.118 0.56 0.156 3.17 2.00 0.26 450°C H2S/SO2 analyzer 8.0 1.89 0.0 — — — S710 analyzer Probe at Elevated Levels 74.0 93.0 Probe at driver level (2.5 m) 15.0 404.0 Probe at foreman level (18 m) After Al-Mehthel et al. (2010). TABLE 112 RESULTS FROM AIR EMISSIONS TESTING

90 locations. The construction temperatures ranged from 255°F to 297°F (124°C to 147°C). Shell Sulphur Solutions (Palmer 2010) developed a new product, Thiopave, which can be used to replace about 20% to 25% with sulfur. The percent of replacement depended on the mix design and type of project. Thiopave was described as a proprietary pelletized sulfur form that was added to the HMA plant. An organic compaction agent (wax) was also used to improve workability at lower temperatures. The recommended order of addition of materials into the plant was as follows: 1. Hot aggregate 2. RAP 3. Virgin asphalt cement 4. Wax 5. Baghouse dust 6. Thiopave pellets. SulfatE WaStE Sulfate rich byproducts, fluorogypsum and phosphogyp- sum, are the result of the production of hydrofluoric and phosphoric acid. The fluorogypsum byproduct (RMRC 2008; TFHRC 2010) is the result of combining fluorspar and sulfuric acid and is discharged in a slurry that solidi- fies over time in the holding ponds, which then needs to be crushed and separated if the byproduct is to be used. The resulting byproduct is sulfate rich with a mostly well- graded sand silt particle size (Table 113). Phosphogysum (RMRC 2008) is a solid byproduct from phosphoric acid production and is a byproduct from a wet process which used hydrochloric acid to treat phosphate rock. The process is outlined in Figure 23. fluorogypSum ByproductS About 100,000 tons of byproduct is produced annually from locations in Delaware, New Jersey, Louisiana, and Texas. The deposits of byproduct are hard and require extraction using typical quarrying processes. The end result is a byprod- uct with a top size of about 1/5 in. and fines that are comprised primarily of sand-sized calcium sulfate particles. The spe- cific gravities are similar to typical construction aggregates (between 2.06 and 2.50); however, the moisture content FIGURE 23 Schematic of phosphate process (after Deshpande 2003). Phosphate Ore Phosphatic Clay Slimes Beneficiation Sand Tailings Wet Process (Reaction with H2SO4) Elemental Phosphorus Thermal Process (Electric Arc Furnace) Phosphate Rock FertilizersPhosphoric Acid ScaleFerro-Phosphorus R226 U238 Slag Phosphogypsum R226 U238 R226 U238 R226 U238 R226 U238 R226 U238 R226 U238 Pb210 Po210 After TFHRC (2009). Constituents and Properties Coarse Sulfate Fine Sulfate % by Weight Sulfate (CaSO4) 71 65.6 Fluoride (F) 1.6 2.5 Free Water 8.6 10.4 Combined Water 14.9 15.2 Acidity 6.4 5.4 Hp TABLE 113 TyPICAL CHEMICAL COMPOSITION OF FLUOROGyPSUM

91 varies widely. The coarser fractions have a moisture content range of from 6% to 9%, whereas the fine fraction moisture ranges from 6% to 20%. In 1996, Vipulanandan et al. (1996) reported the internal angle of friction, cohesion, and unit weight of this byproduct as 40°, 14 psi, and 96 lb/ft3, respectively. Gradation param- eters for D15, D50, D85, Cu, and Cc were 0.012, 0.045, 0.20, 6.9, and 0.87, respectively. The material was described as angular and well graded. The Atterberg limits were 39 for the plastic limit and 47 for the liquid limit giving a plasticity index of 8. The unconfined compressive strength was 64 psi with a wet density of 117 lb/ft3. Table 114 shows results from the TCLP leachate testing. phoSphogypSum ByproductS Phosphate ore is comprised of one-third each of quartz sands, clay mineral, and phosphate particles. Fourteen phosphate rock producing mines were active in 2001 and were located in Idaho (3), Florida (8), and one each in Utah and North Carolina. Florida produces approximately 30 million tons of phosphogypsum annually, most of which is stored. In 1989, the EPA prohibited the use of this byproduct for any purpose unless the proposed use would be at least as protective of human health as leaving it in the stack (Rush et al. 2005). phySical and chEmical propErtiES Phosphoric acid is produced from finely ground phosphate rock that contains relatively high concentrations of naturally occurring radioactive impurities of radium226 and uranium238 (Deshpande 2003). Deshpande investigated the possibility of stabilizing the phosphogypsum by binding it in blended cement. This research focused on defining the appropriate proportions of phosphogypsum, class C fly ash, and portland cement Type II combinations for PCC in marine applica- tions. The specific objectives of the research were to develop blended cement proportions with acceptable physical and engineering properties, minimize dissolution of Ca, SO4, Ra, and toxic metal concentrations in saltwater, and be econo- mical. Blends with the following proportions were evalu- ated: 73:25:2, 67:30:3, 63:35:3 of phosphogypsum:fly ash: cement. The 73:25:2 blend proved to be the most economical at an estimated cost of $10.62 per ton (2001 year basis). The radionuclide concentrations are shown in Table 115 and the oxides of the raw phosphogypsum are shown in Table 116; values for fly ash and cement are included in this table for comparison and the physical properties provided in Table 117. Trace metals were below the EPA standards (Table 118). Leachate concentrations in the raw phosphogypsum exceeded the TCLP limits, but were below the limits when bound in cement (Tables 119 and 120). No information was provided about the leaching potential of recycled (crushed) blended cements. applicationS—Bound Blended cement Guo et al. (2001) reported on the results of blended phos- phogypsum, fly ash, and cement blocks used in marine envi- ronment after 1.5 years of submersion. All of the composite blocks survived with no signs of degradation. SEM, wave- length dispersive microprobe, and XRD suggests a reaction between the composites and the saltwater result in precipi- tation of calcite on the block surface. This deposit provides encapsulation of the composites that helps protect the blocks from saltwater attack and dissolution. Deshpande (2003) conducted a literature review that iden- tified a number of research projects designed to use phospho- gypsum in highway applications. Applications identified in this report included roadway bases, embankments (Thimmegowda 1994), flowable fill (Gandham 1995), cement stabilized soils (Joshi 1997), synthetic lightweight fill (Holmstrom and Swan 1999). The Deshpande research focused on using the blended cements to form fill replacement materials to minimize coastal erosion. hot mix asphalt Tao and Zhang (2006) presented findings using blended calcium sulfate (BCS), the fluorogypsum byproduct in After Vipulanandan (1996). Constituents and Properties Values TCLP Ba, mg/L 0.09 Cr, mg/L 0.11 Pb, mg/L 1.56 Ca, mg/L 422 Acidity pH 4.6 TABLE 114 TCLP RESULTS FOR FLUOROGyPSUM Radionuclide Concentration (pCi/g) Half-Life (Yr) U238 6 4.9 × 109 U234 6.2 2.4 × 105 Th230 13 8.0 × 104 Pb210 26 2.2 × 101 Ra226 33 1.622 × 103 Po210 26 3.78 × 101 After USEPA (1993); Deshpande (2003). TABLE 115 RADIONUCLIDE CONCENTRATIONS IN PHOSPHOGyPSUM

92 Constituent Phosphogypsum Components for Blended Cement Louisiana Texas Florida Fly ash (Type C) Cement (Type II) % by Weight CaO 29–31 32.5 25–31 27.24 63.85 SO4 50–53 53.1 55–58 — — SiO2 5–10 2.5 3–18 34.46 21.43 Al2O3 0.1–0.3 0.1 0.1–0.3 17.83 4.34 Fe2O3 0.1–0.2 0.1 0.2 6.58 5.14 P2O5 0.7–1.3 0.65 0.5–4.0 — — MgO — — — 6.07 0.9 Fe 0.3–1.0 1.2 0.2–0.8 — — Acidity pH 2.8–5.0 2.6–5.2 2.5–6.0 12.2 — After Taha and Seals (1992); Deshpande (2003). TABLE 116 CONSTITUENTS OF PHOSPHOGyPSUM AND OTHER BLENDED CEMENT MATERIALS Trace Element Concentration (mg/L) Arsenic (As) 1.0–5.0 Barium (Ba) 50 Cadmium (Cd) 0.3–0.4 Chromium (Cr) 2.0–5.0 Lead (Pb) 2.0–10.0 Mercury (Hg) 0.02–0.05 Selenium (Se) 1 Silver (Ag) 0.1–0.2 U3O8 0.0–0.5 After Deshpande (2003). TABLE 117 TyPICAL TRACE METALS IN PHOSPHOGyPSUM seulaV seitreporP 53.2 ot 23.2 ytivarG cificepS 81 ot 1 % ,erutsioM eerF Fineness (passing the 0.075 mm) 74% to 75% yticitsalp on ot elttiL yticitsalP Maximum Dry Density, lb/ft3 91.7 to 104.3 Unified Soil Classification System Silty soil (ML) After Deshpande (2003). TABLE 118 PHySICAL PROPERTIES OF RAW PHOSPHOGyPSUM PG:Class C Fly Ash:Portland Type II Cement Mean Metal Conc. in the TCLP Leachate (mg/L) ± Stand. Dev. for n = 3 Cr Cu Zn Fe Pb Cd 73%:25%:02% 0.073 ± 0.023 0.188 ± 0.147 0.045 ± 0.012 0.986 ± 0.370 0.292 ± 0.173 0.044 ± 0.013 67%:30%:03% 0.069 ± 0.021 0.177 ± 0.097 0.044 ± 0.014 0.991 ± 0.348 0.281 ± 0.136 0.042 ± 0.011 63%:35%:02% 0.063 ± 0.019 0.173 ± 0.115 0.043 ± 0.012 0.930 ± 0.376 0.278 ± 0.126 0.043 ± 0.009 62%:35%:03% 0.078 ± 0.021 0.211 ± 0.128 0.048 ± 0.009 0.972 ± 0.348 0.339 ± 0.141 0.046 ± 0.009 EPA Toxicity Limits1 5.0 — — — 5.0 1.0 After Deshpande (2003). 1: 40 CRF 261.24 (USEPA 1999). — = not applicable. TABLE 119 SUMMARy OF TRACE METALS IN BOUND PHOSPHOGyPSUM, FLy ASH, AND CEMENT SAMPLES cementitious blends, in pavements. The maximum dry unit weight of the BCS was 109 lb/ft3 at an optimum moisture content of 12%. The BCS was stabilized with GGBFS (Grade 120) at 10% by weight. Curing conditions for samples to be tested required that the samples be wrapped in plastic and placed in a 100% humidity, 70°F curing room. The samples without the GGBFS showed high initial strength (dry), but decreased substantially when exposed to moisture. Construction problems with particle degradation can be expected. The original BCS gradation was initially similar to the control limestone gradation, but became signifi- cantly finer after compaction. The unconfined compressive strength of the GGBFS/BCS samples had unsoaked and soaked values of approximately 1,000 and 850 psi, respectively, at 28 days. Durability testing showed a fairly consistent rate of mass loss per cycle of about 5 grams per cycle and a volumetric strain (expansion) of less than 0.8% after 100 days.

93 In 2009, Zhong et al. published research that reported on the use of BCS in roadway applications. The authors noted that this material has been used for more than 10 years as a base layer in Louisiana pavements. The major concern when using this material in bases was the moisture sensitivity of the byproduct. High moisture contents could result in construction problems with achieving the desired in situ densities and with long-term performance as support for the pavement structure. Three test sections were evaluated in an accelerated pavement testing facility. The testing indicated that the GGBFS stabi- lized BCS section significantly outperformed both the other BCS and control sections. Falling Weight Deflectometer test- ing indicated that the HMA layer thickness could be reduced, thereby resulting in a cost savings. Summary of SulfatE Byproduct information Sulfate byproducts include fluorogypsum and phospho- gypsum. Only a limited amount of information was found for these byproducts. No specific test methods were found in this information. Louisiana was the only state that has evaluated blended calcium sulfate, the fluorogypsum byproduct in cementitious blends, as a base material. These byproducts needed to be bound to minimize unde- sirable leachates. No additional information was available with regard to materials handling, QC, design changes, or construction guidelines. PG:Class C Fly Ash:Portland Type II Cement TCLP Leachate Concentration (mg/L) As Cd Pb Se 73%:25%:02% 67%:30%:03% 63%:35%:02% 62%:35%:03% 0.24 0.22 0.18 0.16 0.16 0.12 0.10 0.14 0.16 0.28 0.16 0.20 0.32 0.30 0.14 0.17 Raw Phosphogypsum 1.0–5.0 0.3–0.4 2.0–10.0 1.0 EPA Regulatory Limits1 5.0 1.0 5.0 1.0 MCL in Drinking Water 0.05 0.01 0.05 0.05 After Deshpande (2003). 140 CRF 261.24 (USEPA 1999). MCL = maximum contaminate level (USEPA Safe Water Drinking Act, revised in 1999). TABLE 120 SUMMARy OF TRACE METALS IN BOUND PHOSPHOGyPSUM, FLy ASH, AND CEMENT SAMPLES