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.

50 Background Foundry sand is a uniformly graded, high-quality sand byproduct from the ferrous and nonferrous metal casting industry (FIRST 2004a). The metal casting industry uses the foundry sand in two ways. The first is as a molding material to form the external shape of the cast part. The second is used as a core material to fill the internal void space in products such as engine blocks. Because sand grains do not naturally adhere to each other to hold the desired mold shape, binders are added to the sand. Spent (recycled) foundry sand (RFS) can include other materials from foundry processes such as cleaning and grinding operations, slag, and dust collector equipment (i.e., bag houses) (Partridge and Alleman 1998). Binder systems can be either clay-bonded systems (green sand) or chemically bonded systems (resin sands) (FIRST 2004a). Partridge and Alleman (1998) summarized the types of binders used in various types of casting operations (Table 64). Green sands are used to produce about 90% of the casting volume in the United States and consist of 85% to 95% silica, 4% to 10% bentonite clay, 2% to 10% carbonaceous additive, and 2% to 5% water. The carbon content gives the sand a black color. Resin sands are used in core making, where high strengths are needed to with- stand the heat of the molten metal and in mold making. Most of the chemical binders consist of an organic binder (e.g., oil, cereal, and wood proteins; Hughes 2002) that is activated by a catalyst, although some systems use an inorganic binder such as portland cement or sodium silicate. The most common chemical binder systems are phenolic-urethanes, epoxy- resins, furfyl alcohol, and sodium silicates (FIRST 2004a). The resin sands tend to be somewhat coarser in texture than the green sands. Additional information can be found at the following websites: • American Foundry Society: www.afsinc.org • Recycled Materials Resource Center (RMRC): www. rmrc.unh.edu/ • Turner–Fairbanks Highway Research Center (TFHRC): http://www.fhwa.dot.gov/research/tfhrc/. Most foundries have two sand systems. One system is for external modeling lines and a second one for feeding the internal core lines. After the metal is poured and the cast product is cooled, the green sand is shaken off of the part, recovered, and reconditioned for reuse in the molding process. Used cores are reclaimed during the cooling and shaking pro- cesses. The reclaimed material is crushed and reintroduced into the green sand systems to replace a portion of the sand lost in the process. Broken and/or excess cores or those that do not break down when crushed are discarded. The flow chart for a typical foundry is shown in Figure 12. Examples of amount of typical individual byproducts by the molding type are shown in Table 65. It may be important to separate the sand streams at the foundry because of the different material characteristics needed for external and core molding. These sands may be contami- nated with metal and/or large chunks of burned cores, referred to as core butts, which will need further crushing, separation, and screening before recycling. costs Bhat and Lovell (1997) estimated the in-place cost of a flow- able fill at around $40/yd3, which was considerably higher than the cost of conventional soil backfill. Assuming the cost of cementitious material (portland cement and fly ash) to be $60/ton, the cost of sand to be $4/ton, the combined cost of the fill would be $8.64/ton. If the clean sand is replaced by RFS at a cost of $1.50/ton but requires about 50% more cement, the cost of the fill could be reduced by 25% to $6.44/ton. Transportation costs were generally the highest cost factor in recycling foundry sands (FIRST 2004a). Hughes (2002) noted that RFS consortiums generated significant cost savings for small to medium sized foundries. For example, a recovery facility was established in 1985 to manage the RFS from 33 iron, brass, and steel foundries in Pennsylvania. This con- sortium accepted nonhazardous foundry sand, slag, refracto- ries, and dust from foundries within 100 miles of its land fill. The facility charged a fee for each haul that was considerably lower than the average statewide tipping fee of $35 per ton. The use of a recycling facility was estimated to save the found- ries more than $15 million in tipping fees since the monofill began operation in 1990. About 75% of the RFS was recycled for use in HMA. The remaining 25% of the waste stream was comprised of metal, refractory, core butts, and slag, which are diverted to other recycling markets. chapter four Foundry sands

Type Name Binders and Additives Inorganic Option Green sand Clays, water, starch, and sea coral Alumina phosphate Aluminum phosphate resin and metal oxide hardeners Cold Set/No-Bake Options Furan Furfural alcohol resins, urea, phenol, and aryl sulfonic acids Phenolic urethane Phenol formaldehyde resin, isocyanates, and liquid amines Sodium silicate Liquid sodium silicate and liquid organic ester Cold Box Options Phenolic urethane Phenol formaldehyde resin, polymeric isocyanate, and gaseous amine Silicate-CO2 Liquid sodium silicate, coal dust, clays, and CO2 gas Heat Activated Options Hot box Furfural alcohol or phenolic resin, urea, formaldehyde, and acid catalyst Shell molding Phenol formaldehyde resins, calcium stearate, Vinsol, iron oxide, and hexamethylene tetra-amine Air set Various oil resins Core oil Unsaturated oil resins, oxygen sources, and solvents Partridge and Alleman (1998). TABLE 64 SUMMARy OF FOUndRy SAnd BIndER TyPES AS A FUnCTIOn OF THE TyPE OF CASTInG TyPE Mold Production Core Production (in-house) Scrap Metal Storage Casting Melting Finishing Shake Out Recycled Sand Recycled Scrap Metal Virgin Sands Clays Water Organic Additives (Sea coal, cellulose, starch) Externally Produced Cores Virgin Sands Graphite Wash Organic Additives (phenolics, isocyanates, petroleum distillates, amines, formaldehydes, etc.) Scrap Metal Fines, Core Butts, and Excess Sand Slag Finishing Waste Finished Castings Foundry Byproducts FIGURE 12 Flow chart for the generation of foundry sand byproducts (after Partridge and Alleman 1998). Waste Type Foundry Type Malleable Ductile iron Gray iron Steel Aluminum Brass and bronze Refractories 40 50 80 140 20 40 System Sand 1,250 2,190 670 2,790 280 100 Core Sand 310 100 30 550 1,370 140 Cleaning Room Waste 60 90 80 270 20 30 Slag 100 400 220 350 — — Coke Ash — 60 — — Dust Collector Discharge 20 — 190 30 — — Miscellaneous 2 — 110 5 5 5 Totals 1,785 2,890 1,380 4,135 1,695 315 After Bhat and Lovell (1997). TABLE 65 ESTIMATEd POUndS OF FOUndRy WASTER PER TOn OF METAL CASTInG

52 ranging from 0% to 12%. Foundry sands with 6% to 10% clay typically have a liquid limit greater than 20% and a plastic index (PI) greater than 2. Typical physical properties of RFS and natural sands were reported by Bhat and Lovell (1997) in Table 68. The RFS median particle size (d50) was about half of that for river sand used in the study and the fineness modulus was between 40% and 50% smaller. The water content varied substantially. The LOI was also variable and significantly larger for the RFS compared with the pre-consumer foundry sand. The ion concentration in the leachate was much greater in the RFS compared with either the river sand or the pre-consumer foundry sand. The RFS occasionally had a higher maximum void ratio (emax) and lower minimum void ratio (emin) than the river sand used for comparison. The maximum and minimum dry unit weight and specific gravities were somewhat lower for the RFS than for the control river sand. The water absorption of the RFS can exhibit a wide range of characteristics; this was related to the amount of moisture held by the material. Engineering Properties Foundry sands (pre-consumer) without clay typically have internal friction angles between 30° and 35° as determined with direct shear testing (FIRST 2004b; Tables 69 and 70). usagE and Production There are approximately 2,300 foundries spread throughout the United States; each state has some form of foundry industry (IRC 2010). Figure 13 shows the top ten foundry production states in the United States. Most of the foundry production is centered around the Great Lakes and in both Texas and California (FIRST 2004a). Foundries typically send about 28.2% of their spent foundry sand to beneficial reuse programs (IRC 2010). The uses for different spent foundry sands are shown in Table 66. Physical and chemical Properties The original, pre-consumer foundry sand properties are shown in Table 67 (FIRST 2004b). Pre-consumer foundry sand can meet the requirements of the Unified Soil Classification System as SP, SM, or SP-SM, and by the AASHTO classification system for A-3, A-2, or A-2-4. Foundry sands are typically nonplastic or low plasticity sand with a silt or clay content FIGURE 13 The ten states with major sources of recycled foundry sand (dark shading) (after FIRST 2004a). After IRC (2010). Primary Material Primary Application Secondary Application Ferrous Foundry Sands Structural fill Granular base Aluminum Foundry Sands Cement manufacturing Granular base TABLE 66 TyPICAL REUSES FOR RFS FIRST (2004b); RMRC (2009); TFHRC (2010). Property Pre-Consumer Foundry Sand (FIRST 2004a) Spent Foundry Sand (TFHRC 2010) ASTM standard Foundry sand with clay (5%) FS#1 Foundry sand without clay FS#2 Bulk Density, lb/ft3 C29 60–70 80–90 — Moisture Content, % D2216 3–5 0.5–2% 0.1 to 10.1 Specific Gravity D854 2.5–2.7 2.6–2.8 2.30 to 2.55 Dry Density, pcf D698 Standard Proctor 110 to 115 100 to 110 — Water Capacity, absorption ASTM C128 — — 0.45 Optimum Moisture Content, % D69 8 to 12 8 to 10 — Permeability Coefficient, cm/s D2434 AASHTO T215 10-3 to 10-7 10-2 to 10-6 10-3 to 10-6 Plastic Limit/Plastic Index ASTM D2434 AASHTO T90 — — Non-plastic TABLE 67 TyPICAL PHySICAL PROPERTIES OF FOUndRy SAnd

53 Property Material River sand Pre-consumer foundry sand RFS Source 1 Source 2 Source 3 d 50 , mm 0.75 0.32 0.36 0.39 0.26 Coefficient of Uniformity 2.83 2.24 2.41 6.3 5.0 Fineness Modulus 2.98 1.50 1.57 1.78 1.38 Water Content, % 0.5 0.25 0.9 10.4 1.3 Loss on Ignition, % 6.0 0.1 3.8 7.8 2.1 Ion Concentration in Leachate, mg/L Cl -1 NA NA 17.6 60 9.4 SO 4 -2 NA NA 63.9 340.2 120.9 e max 0.69 — 0.91 1.78 1.01 e min 0.45 — 0.73 1.06 0.67 γ dmax , lb/ft 3 118 — 94 75 95 γ dmin , lb/ft 3 101 — 85 55 79 Specific Gravity 2.69 2.66 2.53 2.42 2.50 Bulk Specific Gravity, SSD 2.62 2.64 2.48 2.25 2.45 Water Absorption, % 1.6 0.5 1.5 5.5 1.6 After Bhat and Lovell (1997). NA = not applicable. — = indicates no data. TABLE 68 PHySICAL PROPERTIES OF RFS Materials Internal Friction,o Cohesion, psi Permeability, cm/sec Loose Dense Loose Dense Green Sand with Clay (6%–12%) 32°–34° 37°–41° 0.60–0.75 1.44–1.82 2.8 x 10 -5 to 2.6 x 10-6 Clean Green Sand Without Clay 30° 35° — — 3 x 10 -3 to 5 x 10-3 Chemically Bonded Sand — — 0.06 1.04 4.5 x 10 -3 to 5.9 x 10-4 Natural Sand 29°–30° 36°–41° — — 10 -3 to 10-4 After FIRST (2004b). TABLE 69 TyPICAL PRE-COnSUMER FOUndRy SAnd EnGInEERInG PROPERTIES TABLE 70 TyPICAL SPEnd FOUndRy SAnd PROPERTIES Property Test Method Results Micro-Deval Abrasion Loss, % — <2 Magnesium Sulfate Soundness Loss, % ASTM C88 5 to 47 Friction Angle — 33o to 40o California Bearing Ratio, % ASTM D1883 4 to 20 TFHRC (2010). For pre-consumer foundry sands with sands, the angle of internal friction is from 32° to 41° with cohesion values from 0.6 to 1.82 psi. Post-consumer foundry sand engineering properties reported on the TFHRC website (2010) show simi- lar, but somewhat different, ranges of values. TFHRC reports a very good resistance to abrasion but a wide range of resistance to freeze/thaw damage (MgSO4 soundness). Environmentally related Properties One or more of four leachate tests commonly used to evalu- ate RFS leachate characteristics were the extraction proce- dure (EP) toxicity method, Toxicity Characteristic Leaching Procedure (TCLP) method, AFS method, and ASTM method. Metals of interest include arsenic, barium, cadmium, chromium, lead, mercury, selenium, and silver. Organics are a concern because many of these compounds are added to the foundry sands (Bhat and Lovell 1997). Partridge and Alleman (1998) provided a summary of pre- viously reported research that assessed the leaching potential of organics from nine binder and core making processes. The major casting processes represented in the study were phenol formaldehyde, phenolic urethane, furan hot box, furan no-bake, phenolic ester, core oil, phenolic isocyanate, and furan warm box (Table 71). Organic chemicals detected for any one of four regulatory EPA lists were reported. The four EPA lists of compounds were the priority pollutant list identifying chemicals, which were environmentally hazard- ous and found in water (88 compounds, excluding pesti- cides and PCBs), the TCLP chemical list (38 compounds, excluding pesticides), the drinking water standards (dWS),

54 and the proposed solid waste disposal facility criteria under Subtitle d of the RCRA act. The only compounds exceeding any of these standards were benzene, which was higher than the dWS maximum contaminating levels for three of the nine RFS. Tetrachloroethene concentrations were at the trigger level used when dWS have not been established. The authors noted these concentrations in leachate testing would likely never exceed limits in actual field testing. Both the core oil and phenolic urethane binder systems leached the greatest number of organic chemicals. The literature review summarized by Partridge and Alleman (1998) provided information from an early 1990s Wisconsin– Madison research study that evaluated RFS from three found- ries. none of the sands were hazardous according to the RCRA criteria. A top priority parameter of greatest concern was iron with concentrations higher than dWS in TCLP leachates when compared with results for virgin soil. Other parameters of inter- est included fluoride, pH, and total dissolved solids. Second- ary priority parameters included arsenic, chromium, copper, manganese, zinc, phenolics, and sulfates. The Industrial Resources Council (2010) reported on studies of RFS byproducts that showed little uptake of trace metals from iron and aluminum foundry sands. However, there were some concerns with trace metals in RFS from brass and bronze foundries. Trace metal concentrations in most clay- bonded iron and aluminum RFS were similar to those found in naturally occurring soils, while leachate may contain trace metal concentrations that may exceed drinking water stan- dards, the leachate results were found to be similar to those of other construction materials such as native soils or gravels (IRC 2010). This information led to the EPA endorsing the use of properly managed ferrous and aluminum RFS as a construction material. Hughes (2002) also noted that TCLP testing indicated RFS was nonhazardous. Synthetic Precipitation Leaching Proce- dure (SPLP) testing showed that leachate from HMA design blends without RFS had higher concentrations than when RFS was included. HMA plant emissions testing when using a mix with 10% RFS was found to not be statistically different for HMA mixes without RFS. aPPlications Partridge and Alleman (1998) reported the results for a survey on the use of recycled foundry sand in highway applications of ten states with foundry production operations (Table 72). The common applications were reported as concrete, HMA, and road base. applications—Bound Cement FIRST (n.d.) provided a case study for the use of RFS in the manufacture of cement in Mason City, Iowa. In this case, the portland cement producer received approximately 75,000 tons per year of the RFS from a waste management firm in Michi- gan, which was supplied by eight regional foundries located in Iowa, Minnesota, Wisconsin, and Illinois. Costs for the RFS recycling facility included transportation to the plant, testing for chemical and physical properties, and crushing and pro- cessing sand. Flowable Fill Bhat and Lovell (1997) investigated the use of RFS, along with Type F fly ash, in flowable fill in Indiana. desirable properties Chemical Compound Quantization Limit, ppb Maximum Concentration, ppb 002 001 enotecA 11 2 enezneB Benzoic Acid ND 400 2,4-Dimethylphenol 20 120 Ethylbenzene 0.4 24 94 2 enahteorolhcirT-1,1,1 084 1 enelahthpaN 023 1 enelahthpanlyhteM-2 045 03 lonehP 16 04 etalahthplyhtemiD 83 03 enerhtnanehP 7 2 enehteorolhcarteT 16 5.0 eneuloT 051 03 sloserC 041 4.0 senelyX After Partridge and Alleman (1998). ND = not determined. TABLE 71 MEASURABLE ORGAnIC COMPOUndS In RFS FROM nInE CASTInG PROCESSES

55 of the flowable fill were local availability, easily delivered and placed, weather resistant, self-compacting, and an unconfined strength below 150 psi so that it can be easily excavated with a backhoe in case repairs or reconstruction were needed. The American Concrete Institute modified flow test was used to evaluate flowability. This test involved placing a 3-in.-diameter by 5-in.-tall open-ended cylinder on a smooth, level, surface, then filling with the material, quickly lifting the cylinder and measuring the diameter of the circular section formed. A spread of 8 to 9 in. or greater was considered to be self- leveling. The angularity of the fine aggregate was evaluated with the flow cone sand test. This test showed that for a given flowability, the use of the RFS significantly increased the water demand needed to achieve a given flow compared with either the river sand or pre-consumer foundry sand mixes. The presence of the fly ash in the mix was needed to improve the flowability. At low fly ash contents, a lubricant effect was achieved as a result of the round shape of the fly ash. At high fly ash concentrations, the fly ash tended to flocculate and the viscous forces appeared to dominate. Other testing included an evaluation of penetration resis- tance (ASTM C403), bleeding (measured by collecting the free water on the surface of a sample), and the surround- ing drainage conditions. The factors that influenced the early strength of the mixes were cement content, environmental curing conditions, the nature of the fly ash, and the drainage conditions surrounding the flowable fill materials. Higher cement content mixes had higher early strengths, and flowable fill material with drainage around the perimeter gained strength more quickly than undrained conditions. The Bhat and Lovell research also evaluated the constant head permeability (ASTM d5045), pH, and toxicity of the flowable fill. The permeability was similar for all of the mixes, ranging between 2.6 × 10-6 and 1.2 × 10-5. The pH of both the bleed water and pore solution ranged between 10.0 and 11.7. Toxicity testing, using MicroTox™ showed mixed results that led to a conclusion by the researchers that more testing was needed before a conclusion could be drawn. FHWA (2003) described the Abrams Creek Improve- ment project in Cleveland, Ohio, which used RFS to encase four 10-ft-diameter concrete pipes. The pipes were placed only 1 ft apart during a major extension of a runway at the Cleveland Hopkins International Airport. Once placed, the pipes and fill had 65 ft of fill placed over the top of the drain- age system. The close proximity of the pipes as well as the deep fill required the strong bedding for the pipe. RFS was delivered to the concrete ready mix plant where it was mixed with 150 lb/yd3 of cement and 60 gal/yd3 of water to produce flowable fill with strength of between 125 and 300 psi at a cost of $30/yd3. HMA Hughes (2002) noted that the RFS properties differed based on the type of original foundry sand (i.e., green, resin). The chemically bonded RFS was drier and had lower fines content than green RFS. Important information required by the HMA producer were identification of the type of RFS and how the sand streams were separated, comingled, etc. Hughes noted there were a number of independent recycling operators who addressed the majority of these concerns in their operations. Regardless of who collected the RFS, post-processing was needed prior to use in HMA applications. The RFS needed to be reprocessed into a consistent, high-quality product com- parable to virgin sand. There were three steps needed in post-processing RFS: 1. Removal of general refuse and other contaminates 2. Removal of metals 3. Processing and sizing. This required the installation of reprocessing systems to provide contaminate-free, and screened to provide market- able gradations. Most foundries that post-process their RFS invested in screening, crushing, and magnetic separation units. Application IA IL IN MI MN NJ NY OH PA WI x x x x revoC yliaD llifdnaL x x x tnemknabmE yawhgiH Roadway Subbase x x x x x x x x x x x x esabbuS toL gnikraP Concrete and Asphalt x x x x x x x x x x x lliF edargbuS noitadnuoF x x x lliF elbawolF x x x x lliF lareneG x x x x rehtO Partridge and Alleman (1998). TABLE 72 SUMMARy OF RFS USE In THE UnITEd STATES

56 Resizing green RFS resulted in a an excess of minus 0.075 mm fines that needed to be monitored so that the maximum percent 0.075 mm was not exceeded for HMA applications. This also required the post-processor to address fugitive dust controls. Hughes noted that small to medium size foundries typically did not have the capital to invest in post-processing operations. A cost-effective method of recycling for these facilities was to have a partner handling multiple spent sand streams from a collective of foundries. A sufficient incoming supply of material was needed for the HMA producer to maintain a consistent rate of production. FIRST (2003) provided a case study for gray iron RFS from a foundry in Michigan City, Indiana, that was used in HMA. The foundry paid the haul costs and the contractor to use the RFS. About 4,000 tons of RFS was used in the project, with a cost savings of 75% (about $50,000 savings for the foundry) over the typical tipping fee costs. The RFS made up about 10% by weight of the HMA aggregate. The RFS represented four types of sand: green sand (<10%), no-bake (<25%), cold box (about 25%), and warm box (about 40%). Comments from the contractor suggested the RFS improved the smoothness of the finished surface. Reported RFS properties are shown in Table 73. applications—unbound Embankment and Fill Partridge and Alleman (1998) evaluated the performance of foundry sand from a gray iron foundry that used a phe- nolic urethane binder as the core binder in embankments in Indiana. Both RFS and control embankments were placed with geotechnical and environmental monitoring instru- mentation. Evaluations of the geotechnical properties of the RFS embankment showed that although RFS performed as a natural sand structural fill with acceptable strength and deformation characteristics, it was not as freely draining as the natural sand. Environmental information was collected using Microtox™ and nitrotox bioassay, ion chromatography, and inductively coupled plasma testing for metals. Bioassay results indicated the RFS did not have a higher toxicity than those expected from natural sands. Ion migration was measured in the RFS lysimeters, but the concentrations were below reuse regulatory criteria. Metal concentrations were generally below Indiana regulatory reuse Type III criteria and typically below drinking water standards. When metal concentrations in the monitoring wells exceeded the criteria, excess concentrations occurred in both up- and down-gradient wells. This was interpreted by the researchers as background metal concentrations rather than contributions from the RFS. Partridge and Alleman noted barriers that included a lack of decision-based scientific tools (e.g., life-cycle cost analysis or risk-based analyses), liability exposure from state and federal regulations from using a regulated byproduct, lack of foundries commitment to reuse, post-processing needs and QC, and a formal marketing strategy. FHWA (2003) reported that the Ohio dOT used RFS to repair an embankment on Ohio SR-271. The RFS was selected because it was deemed to have the necessary strength and permeability properties. The RFS was successfully com- pacted at the optimum moisture content. Independent testing laboratory results for the RFS are shown in Table 74. After FIRST (2003). Property Value Density, lb/ft3 100–110 Bulk Specific Gravity 2.61 Sodium Sulfate Soundness, % 6.9 Water Absorption, % 0.4 Uncompacted Void Content,% 33.2 TABLE 73 RFS PROPERTIES Property AASHTO Results Spec. Criteria Grain Size Analysis T 11-91 T 27-93 Pass ODOT 203 Sodium Sulfate Soundness T 104-94 3% NA AN AN 69-98 T timiL diuqiL Plastic Limit & Plasticity Index T 90-96 Nonplastic NA Moisture-Density Relationship T 99-95 γd = 107.0 pcf >120 pcf Direct Shear Test T 236-92 Ν = 350 >340 006,5 19-882 T ytivitsiseR Ω-cm NA AN 2.9 19-982 T Hp AN 8.78 59-092 T leveL etafluS AN 53 49-192 T edirolhC FHWA (2003). NA = not available; ODOT = Ohio DOT. TABLE 74 TESTInG And PROPERTIES OF RFS EMBAnkMEnT In OHIO

57 In another application, the Ohio dOT used RFS as a fill for the extension of Oak Tree Boulevard in the city of Indepen- dence. Fill was needed for an approximately 2,000 ft long by 600 ft wide by a depth of up to 50 ft ravine. About half of the fill for this project was the RFS. The primary use of the RFS was because it was immediately and locally available. A clay- like dike at the open end of the embankment was constructed to contain the fill and a 3 ft clay cap was placed over the completed embankment. Proper compaction was obtained as long as the moisture content was maintained. Base and Subbase FIRST (2005) reported a case study using about 23,000 cubic yards of RFS as subbase fill for an airport runway constructed in Shawano, Wisconsin, in 2005. The original assumption on the part of the recycler was that the cores would break down during compaction. during construction of the sub- base this did not occur and large fragments were found in the byproduct supply. The recycler ended up screening the RFS before placement. This additional work resulted in an in-place cost per cubic yard of about $5. Compared with a typical borrow cost in the area of $8 per cubic yard, this still resulted in a cost savings to the owner. during construction, the RFS needed to be pre-wet before compaction to meet the required compaction requirements (Table 75). A significant amount of water was needed and it was important that compaction be done correctly the first time, because previously compacted RFS did not behave like the loose RFS. A performance review of the project after the first winter showed no differential heaving, a problem in the area, and no excessive joint movement. Specifications Partridge and Alleman (1998) summarized state industrial waste classification systems and environmental test procedures after a survey of 14 major foundry states’ environmental regu- latory criteria (Table 76). Most of the surveyed states classified RFS as either a solid or residual waste. Six of the 14 states used multiple tiers within a classification category. Thirteen of the 14 states used the TCLP method for determining leachate and seven of these states also require additional chemical testing beyond TCLP. FHWA (2003) reported that the Ohio dOT used a general specification for embankment construction using recycled materials that contains specific guidance on the use of foundry sand. Guidance included directions to place the RFS on a prepared foundation in horizontal loose lifts not to exceed 8 in. and compact the lifts to a stable, durable con- dition with at least eight passes of a vibratory steel wheel roller with a minimum weight of 10 tons or centrifugal equivalent. The compaction of the lifts needed to achieve 98% of the maximum density. The sides and top of the RFS needed to be covered with natural soil with a minimum vertical cover of 3 ft, measured from the subgrade elevation, and a minimum horizontal cover of 8 ft, measured from the final slope line. The Indiana dOT (2007) developed a specification for RFS use in one of two borrowed specifications upon approval from the Geotechnical section. This specification required the contractor to submit an MSdS and a copy of the Indiana department of Environmental Management waste classifica- tion certification for Type III or IV residual sands prior to use. An example of the RFS certification form is shown in Figure 14 and their Indemnification Clause in Figure 15. The specification put limits on the placement of the RFS near water sources at not closer than: • 100 ft horizontally of a stream, river, lake, reservoir, wetland, or other protected environmental resource area. • 150 ft horizontally of a well, spring, or other ground source of potable water. Other restrictions on placement locations not to be used were: • Adjacent to metallic pipes or other metallic structures. • Encasement material. • Mechanically stabilized earth wall applications with metal reinforcement. Safety considerations limited the release of fugitive dust and loss of material during transporting. Spraying with water, limewater, or other sealing type sprays could be used for dust control. The RFS needed to be encased in the same work day as it was placed. Barriers The Foundation Industry (2007) conducted a survey of foundries to determine the extent of their recycling programs and to identify barriers to the increased use of RFS recy- cling programs. The most commonly cited reason for limited recycling programs was the lack of a local market for spent (post-consumer) sand (Table 77). Quantities generated and storage-related issues were the next two most cited reasons After FIRST (2005). Compaction Requirements, % Depth Required, in. 81–0 001 81–8 59 23–81 09 44–23 58 TABLE 75 COMPACTIOn REQUIREMEnTS SET FOR SHAWAnO, WISCOnSIn, AIRPORT RUnWAy

TABLE 76 SUMMARy OF STATE IndUSTRIAL WASTE CLASSIFICATIOn SySTEMS And BEnEFICIAL REUSE TESTInG REQUIREMEnTS State Industrial Waste Classifi a serudecorP tseT metsyS noitac nd Standards for Beneficial Reuse Solid waste category Special waste category Residual waste category Multiple tiers within category State hazardous classification TCLP or acid- based test ASTM D3987 or neutral based test Use of individual state test systems Identification of specific test levels Additional chemical testing beyond TCLP or neutral leachate tests Bulk analysis x x x x amabalA x x x x x x ainrofilaC x x x aigroeG x x x x x sionillI x x x x x x x anaidnI x x x awoI x x x x sttesuhcassaM x x x x x x nagihciM x x x kroY weN x x x x x x oihO x x x x x ainavlysnneP x x x x eessenneT x x x x x x x saxeT x x x x x x x nisnocsiW After Partridge and Alleman (1998).

59 RECYCLED FOUNDRY SAND (RFS) SOURCE CERTIFICATION This is to certify recycled foundry sand (RFS) stockpiles geographically located as follows: RFS ___________________________________________________________________ _______________________________________________________________________ RFS was produced by the ________________________ __________________ Company located in ____________________(City), and _________________ (State) and was shipped for use on Indiana Department of Transportation projects is Type _________________ (III or IV) material according to IDEM's restricted waste criteria and that the material has passed Microtox™ (ITM 215) test criteria. If any metal concentration exceeds 80% of the allowable limits for a Type III the foundry shall provide the Department with an acceptable indemnification clause. The _________________ RFS source also agree that processes and stockpiles associated with the production of such RFS may be inspected and sampled at regular intervals by properly identified representatives of the Department or a duly assigned representative. ________________ (Date of Signing) ________________________________ (RFS Producer) ______________________________(Title) ______________________________(Signature) State of _____________________________) SS: County of _____________________) Subscribed and sworn to before me by ___________________________________ of the firm of ________________________________ this __________ day of ________________ 20__. _______________________________ Notary Public My Commission Expires: ______________________ This certification has been reviewed and approved by: _____________________________Date________________ (Materials and Tests Division representative) FIGURE 14 Example of Indiana’s RFS source certification form (Indiana DOT 2007). RECYCLED FOUNDRY SAND (RFS) INDEMNIFICATION CLAUSE ______________ RFS producer shall indemnify, defend, exculpate, and hold harmless the State of Indiana, it officials, and employees from any liability of the State of Indiana for loss, damage, injury, or other casualty of whatever kind or to whomever caused, arising out of or resulting from a violation of the federal or Indiana Occupational Safety and Health Acts (OSHA), the Resource Conservation and Recovery Act (RCRA), the Comprehensive Environmental Response, Compensation and Liability Act (CERCLA), or any other environmental law, regulation, ordinance, order or decree (collectively referred to hereinafter as “Environmental Laws”), as a result of the supply, testing, and application of residual sand or other materials supplied under this Contract by ________________ source, whether due in whole or in part of the negligent acts or omissions of: (1) _________________ Foundry, its agents, officers, or employees, or other persons engaged in the performance of the contract; or (2) the joint negligence of them and the State Of Indiana, its officials, agents, or employees. This contract shall include, but not be limited to, indemnification from: (1) any environmental contamination liability due to the supply, testing, and application of residual sand in road base, embankments, or other projects designated by the Department as agreed to by the parties, and (2) any liability for the clean up or removal of residual sand, or materials incorporating such sand, pursuant to any Environmental Law. The RFS producer also agrees to defend any such action on behalf of the State of Indiana, to pay all reasonable expenses and attorneys fees for such defense, and shall have the right to settle all such claims. Provided, however, that no liability shall arise for any such fees or expenses incurred prior to the time that ______________ Foundry shall have first received actual and timely written notice of any claim against the State which is covered by this Indemnification Agreement. If timely written notice of any claim hereunder is not received by _______________ Foundry, and _________________ Foundry is thereby prejudiced in its ability to defend or indemnify, then to the extent of such prejudice, this Indemnification Agreement shall be void. This Indemnification Agreement does not create any rights in any third party, and is solely for the benefit of the State Of Indiana and its agents, officials, and employees. FIGURE 15 Indiana’s RFS indemnification clause (2007). stnednopseR fo tnecreP gnisueR toN rof nosaeR 1.04 dnas tneps rof tekram lacol oN 2.42 tcejorp rof hguone etareneg ton seoD 9.22 seititnauq tneiciffus ni dnas erots ot ecalp oN 3.51 smialc ytilibail erutuf ksir ot tnaw ton oD 0.41 esuer tneverp scitsiretcarahC State regulation and paperwork make it too time-consuming or state regulations do not allow reuse 9.6 4.6 desu yllaicifeneb eb dluoc ti wonk ton diD Foundry Industry (2007). TABLE 77 REASOnS FOR nOT REUSInG FOUndRy SAnd

60 for limited recycling followed by liability concerns and undesirable material properties. Another limitation identified in an Indiana state foundry evaluation indicated that intra- foundry competition could also limit recycling. For example, the larger foundries in Indiana had a lock on supplying foundry sand to cement kilns, which prevented the smaller foundries from marketing to this industry. Reuse Programs In 2006, the EPA published a toolkit for developing reuse programs for foundry sand. The structure, state review, and approval for beneficial use should include information for waste classification, case-by-case determinations, and hybrid applications. Ongoing testing guidance should be given for periodic and event-base, or event-based-only cases. This toolkit provided the following standardized definitions to be used for reuse program development. • Approval: A state agency’s endorsement of proposed beneficial reuse activities. This state endorsement may be in written format, although some states endorse pro- posed activities without a formal written response to generators or end-users. • Case-by-case determinations: states review proposed reuse activities on an individual basis. • Waste classification: States establish categories that are defined by ranges of contaminant thresholds for specific reuses and/or waste types. In general, byproducts with low concentrations of constituents of concern are less restricted in their reuse activities. Conversely, byproducts with higher concentrations are more restricted. These categories standardize the review process for proposed reuse activities and streamline the approval process. • Event-based testing: Establishes the frequency of sampling and testing to confirm that the foundry sand’s composition has not changed. In this case, generators or end-users must test the sand when a specific incident occurs, such as a change in the foundry process generat- ing the waste sand. Prior to starting the reuse program design, the types of foundries from which the sands originate needed to be identi- fied. Iron, steel, and aluminum foundries were the most com- mon, but some regions of the country had other foundries for copper, brass, bronze, beryllium, cobalt zinc, lead tin nickel, magnesium, and titanium casting. The level of agency and contractor involvement in the approval process needed to be identified, as this had an impact on work load and responsi- bilities (Table 78). nearby industries that could potentially use RFS needed to be identified. The main activities were commonly stabi- lized or bound materials (e.g., portland cement, asphalt, and concrete products), geotechnical (confined) applications (e.g., road bases, structural fills, and embankments), and other products such as soil amendments, manufactured soil, and top dressing. Sampling and test method requirements needed to be defined as well as the constituents to test QC and QA pro- grams developed. The frequency of testing needed will be a function of the consistency of the byproduct. Byproducts with consistent physical and chemical properties required less frequent testing. Once this information had been assembled, six steps were defined for the development of a reuse program: 1. define program structure 2. Identify siting or location restrictions: a. define siting standards that need to be established for reuse activities (e.g., environmental resources to be protected such as ground water, wetlands, etc.; minimum distances, and bans) b. demand on agency resources. 3. Identify state reviews needed to initiate projects 4. Obtain state approval for beneficial uses 5. develop initial sampling and testing results 6. develop a program for periodic and/or event-based sampling, testing, and reporting. The type(s) of program structures were considered in Step 1. The choices made in Step 1 carried over to Step 3. The type of program structure defined the extent of the required initial and ongoing level of resource commitment to reuse programs. A waste classification structure established reuse standards that varied by the type of byproduct. This type of program required more agency and industry resources up front, but minimized the ongoing resource commitments. These types of classifications were usually based on constituent levels that were stringent enough to ensure environmental safety for all possible uses. For example, Illinois and Indiana set the maxi- mum allowable leaching concentrations for arsenic, but used different category designations and levels (Table 79). A case-by-case program structure typically created a basic set of standards that all of the byproducts must meet to be eligible for beneficial reuse. The hybrid structure combined the waste classification categories with the case-by-case reviews to help streamline the process for reuse applications within the waste classification categories while being flexible enough to consider other applications individually. Table 80 provides an overview of the wide range of reuse programs used by state agencies. The range of types of programs and responsibilities highlight the reason both agencies and contractors list regu- lations as a barrier to increased use. When trying to create a market for byproducts, reclaimers were faced with a number of classification systems and regulations that needed to be met within a given market area.

61 agEncy survEy rEsults Three agencies had used recycled foundry sands in embank- ments, two in flowable fill, and one in drainage materials (Table 81). The only use for recycled sand blasting waste was in HMA applications. Table 82 shows only five states reporting that they had experience using recycled foundry sand in highway applications. no states were currently using sands from sand blasting operations (Figure 16). Only two states provided comments about their experi- ence with foundry sands (Table 83). Alabama noted it is allowed for soil aggregate mixtures as long as their exist- ing specification could be met. Utah, that did not use RFS in highway applications, highlighted properties that made foundry sand high quality for casting, were not always con- sidered high quality for highway applications. The uniform size and round shape made it difficult to meet well-graded aggregate specifications and angular fines requirements for bases and HMA. summary oF Foundry sand inFormation list of Byproducts The list of the most commonly researched and used byprod- ucts include green sands and core sands. The foundry sand byproducts could be separated by their use in the casting process, which could alter physical and chemical properties. These differences were a function of the Activity Responsibility Agency Burden Industry Burden None Low Med. High None Low Med. High Review of Proposed Reuse Activity Industry required to keep records of sampling and testing results x x Agency reviews initial sampling and testing x x Agency reviews sampling and testing on a case-by-case basis for environmental impacts x x Agency reviews sampling and testing plus additional information regarding environmental impacts (i.e., ground water contamination, off-site releases, air pollution, etc.) x x Written Submissions of Approval No written approval to generators needed x x Agency submits written approval to generators or end-uses for some, but not all reuse applications (dependent on volume used) x x Agency submits written approval to generators or end-uses for all reuse activities x x Ongoing Oversight Industry conducts periodic and event- based sampling and testing but does not report the results to the agency unless a significant change occurs x x Industry conducts event-based sampling and testing and reports the results to the agency x x Industry conducts periodic and event- based sampling and testing and reports the results to the agency x x In addition to periodic and event- based sampling and testing and reporting, industry reports additional information regarding reuse activities x x After EPA (2006). TABLE 78 LEVELS OF RESPOnSIBILITIES And IMPACTS On RESOURCES FOR AGEnCIES And IndUSTRy TABLE 79 ExAMPLES OF dIFFEREnT WASTE CLASSIFICATIOnS FOR ARSEnIC In BEnEFICIAL REUSE APPLICATIOnS After EPA (2002). Illinois Indiana Category Threshold Category Threshold Beneficially Usable 0.05 mg/L Type IV 0.05 mg/L Potentially Usable 0.1 mg/L Type III 0.50 mg/L Low Risk 0.25 mg/L Type II 1.3 mg/L Chemical Waste >0.35 mg/L Type I 5.0 mg/L

State Classification System Allowable Reuses Sampling and Testing Requirements Other Alabama Single-tiered waste classification Applications prohibited Generators certify waste quarterly or when process changes Generators maintain records California Approvals and designation determined by Department of Toxic Substances Control (DTCS), Water Boards, and, California Integrated Waste Management Board RCRA and state- determined hazardous wastes standards used Leachate testing required Depends on application Illinois 4-tiered classification beneficial reuse Depends on classification Generator-provided leachate testing from each waste stream Generator certification Indiana 4-tiered classification beneficial reuse Depends on classification Generator-provided leachate testing to classify Additional testing may be required; stockpile site restrictions; dust controls Iowa Concentration criteria for intended reuse application Dependent on concentration levels Leachate and pH testing RFS use does not require a permit; a foundry sand management plan needs to be submitted Louisiana Source of byproduct and types of facilities (5 levels) Solid waste regulations define uses Applicant must conduct a detailed analysis for specific compounds Third party certification; completion of application for use; site location information Maine Application for use of secondary materials and special waste Defined by regulation and on case-by-case basis Applicant needs to submit sampling and analytical work plan Meet or exceed requirements for materials that will be replaced with byproduct; no use in environmentally sensitive areas; annual report of activities; licensee submission of application Massachusetts Volume of material in application Applicant needs to describe benefits and demonstrate safe handling, storage, use, and end products Application needs to contain physical and chemical properties Draft beneficial use regulations in progress Michigan Petition for classification as either inert material or low-hazard industrial waste Depends on classification Applicant needs to submit TLCP, SPLP, or other test results as required with annual re-tests submitted Submit description of material, schematic of processes and raw materials used, maximum and annual amounts generated monthly and annually, documentation supporting non-hazardous classification, and description of proposed use ssergorp ni snoitaluger esu laicifeneb tfarD — esac-yb-esaC esac-yb-esaC atosenniM New York 16 materials identified in regulation; foundry sand not on list Also allow case-by-case Defined by regulation Not allowed if decontamination or special handling/processing before use is required Leachate testing and sampling plan Case-by-case requires a description of the byproduct and proposed use; demonstration of safety; byproduct control plan Ohio 4-tier waste classification Depends on concentration thresholds Applicant submission of leachate testing; annual tests submitted to agency Minimization of byproduct volume preferable; cannot create a nuisance; storage subject to nuisance and erosion regulations; cannot place in environmentally sensitive areas TABLE 80 SUMMARy OF STATE REUSE PROGRAMS FOR SPEnT FOUndRy SAnd AS OF 2002

Pennsylvania General permits issued on either regional or statewide basis Applicant applies for one of three permit types: beneficial use as pipe bedding; beneficial use in concrete or asphalt applications, or as a beneficial use in road bed construction Applicant submission of byproduct characterization and leachate concentrations Applicant submission of annual report, demonstration of beneficial contribution, notification to local agency of intent to use, acceptable storage and use of byproduct, protection of water quality, and maintain on-site records for 5 years Rhode Island Applicant requests variance from Solid Waste Regulations; variances expire after 1 year; positive results may allow a renewal for a period of 3 years Case-by-case Applicant submits testing plan Applicant must minimize environmental hazards, demonstrate reuse is a viable substitute for raw materials, demonstrate no adverse impact on health and natural resources, assess market extent, describe in-place controls, demonstrate reuse is not simply an alternative method of disposal, and describe any post-processing Tennessee Contaminate thresholds for nontoxic designation; division approval for others Depends on designation — Generator maintains records, byproducts approved by division need to designate generator and proposed use, estimated volume of byproduct to be used, proposed silt/runoff control and site specifics Texas 3-tier waste classification Depends on classification None if classified as non- waste; non-hazardous requires leachate testing, analysis of hydrocarbons, and verification of absence of PCBs Generators maintain on-site records West Virginia Guidance on beneficial use; application process Guidelines contain reuse applications Submission of sampling and analysis plan Required plan approval from Solid Waste Management Wisconsin 5-tiered waste classification system Depends on classification Leachate testing; frequency depends on quantities Applicant submission of initial and annual certifications, public notification (quantity- based), written notification to DNR (quantity-based), leachate monitoring for certain transportation facilities After EPA (2007). TLCP = toxicity characteristic leaching procedure; SPLP = synthetic precipitation leaching procedure; PCB = polychlorinated byphenols; DNR = Department of Natural Resources.

64 type of additive used with the original foundry sand, the type of metal being cast, and the specific casting process used. test methods The test methods in Table 84 have been used to evaluate foundry sand and highway application products that use foundry sand. Both AASHTO and ASTM standards are shown in this table. materials Preparation and Byproduct Quality control Material preparation and QC found in the literature and agency surveys included: • The casting cores have been hardened by additives such as epoxies, resins, organic binders (e.g., portland cement and wood proteins) to form the inside of the part. This component of spent foundry sand was used to form the inside shapes of the part and needed further crushing, separation, and screening before recycling. • Post-processing needed to include the removal of gen- eral refuse and other contaminates, metals, and sizing. • Green sands were used to form the external modeling lines and were reclaimed and reused by the foundry until they failed to meet foundry sand requirements materials Handling concerns no specific handling concerns were noted in the literature or agency survey. design adaptations The following design adaptations were found when using RFS: • Spent foundry sands typically required higher portland cement contents that needed to be addressed during the mix design. • Fly ash was needed in PCC mix designs to compensate for a loss of workability as a result of the RFS. • Fill designs needed to account for less freely draining material. • When constructing embankments with RFS, the follow- ing recommendations were made by the FHWA (2003): – Place the byproduct on a prepared foundation in hor- izontal loose lifts not to exceed 8 in. – Compact the lifts to a stable, durable condition with at least eight passes of a vibratory steel wheel roller with a minimum weight of 10 tons or centrifugal equiva- lent. The compaction of the lifts needs to achieve 98% of the maximum density. – Cover the sides and top of the RFS with natural soil with a minimum vertical cover of 3 ft, measured from Question: Manufacturing or Misc. Construction Byproducts: Is your state using, or has ever used, these byproducts in highway applications? * Sand blasting waste: sand along with finishing materials after resurfacing * Sand, foundry: high quality sand recycled after metal castings of products Type of Byproduct Asphalt Cements or Emulsions Crack Sealants Drainage Materials Embank. Flowable Fill HMA Pavement Surface Treatment (non- structural) PCC Soil Stability Sand Blasting Waste 0 0 0 0 0 1 0 0 0 Sand, Foundry 0 0 1 3 2 0 0 0 0 Embank. = embankment. TABLE 81 RESULTS FOR AGEnCy SURVEy FOR FOUndRy SAnd ByPROdUCTS USEd In HIGHWAy APPLICATIOnS Number of Applications States Sand blasting waste Sand, foundry 2 — WI 1 NC IA, IN, OH, PA TABLE 82 STATES USInG FOUndRy SAnd ByPROdUCTS In HIGHWAy APPLICATIOnS In 2009 FIGURE 16 States using foundry sand in highway applications. 1 2 1 1 1 1 2009 Foundry Sands Byproducts

65 the subgrade elevation and a minimum horizontal cover of 8 ft, measured from the final slope line. construction concerns Construction concerns that need to be considered when using RFS include: • Additional crushing and compaction efforts that may be needed if the spent foundry sand cores were not crushed prior to use in base applications. • Proper moisture content was needed to achieve the desired in-place density in unbound applications. A significant amount of water may be needed so that compaction is achieved the first time around. RFS was difficult to re-wet because of the clay additive. Failures, causes, and lessons learned The lessons learned when using RFS in demonstration projects include: • When used in unbound applications, the RFS needed to be pre-wet and at optimum moisture content on the first round of compaction as the clay additive content tends to prohibit further compaction after re-wetting. • RFS cores needed to be crushed before use in unbound applications as it was difficult to crush them during compaction. Barriers Barriers noted included: • Lack of decision-based scientific tools (e.g., life-cycle cost analysis or risk-based analyses) • Liability exposure from state and federal regulations from using a regulated byproduct • Lack of foundries committed to reuse • Post-processing needs • Lack of QC for the RFS • no formal marketing strategy • A number of classification systems and regulations that need to be met within a given market area. costs Regional recycling facilities reduced the cost of the byprod- ucts. A recycling facility provided a single disposal location for smaller foundry operations, post-processing operations for useable byproducts with consistent properties, and adequate quantities for a given application product. tnemmoC etatS AL Foundry sand is allowed in granular soil materials (Section 821 of ALDOT Standard Specifications) and in soil aggregate materials (Section 823) provided it meets the testing/specification requirements. Foundry sand stockpiles are required to be inspected/approved on an individual basis. Issues with consistency of foundry sand properties and its potential for chemical reactions is also a concern. UT Foundry sand is not “High Quality Sand” in the construction world. In its native gradation the single sizes often do not provide a stable material for compaction purposes. It could be if intermixed with other sizes, but typically the single size gradations and shape are not stable. Therefore, to place and use, the gradation has to be modified. This costs money and doesn’t make it always as attractive in a low-bid environment. Again, it gets down to the issue of suppliers wanting us to use their product as is rather than modifying their product and/or quality control to provide us the product we want. TABLE 83 SUMMARy OF COMMEnTS On USInG FOUndRy SAnd In HIGHWAy APPLICATIOnS TABLE 84 TEST METHOdS USEd TO EVALUATE ByPROdUCTS And HIGHWAy APPLICATIOn PROdUCTS Test Methods Title AASHTO Methods T112 Standard method of test for lightweight pieces in aggregate T215 Standard method of test for permeability of granular soils T90 Standard method of test for determining the plastic limit and plasticity index of soils ASTM Methods C128 Standard Test Method for Density, Relative Density (Specific Gravity), and Absorption of Fine Aggregates C142 Standard Test Method for Clay Lumps and Friable Particles in Aggregates C29 Standard Test Method for Potential Alkali Reactivity of Cement-Aggregate Combinations (Mortar-Bar Method) C403 Standard Test Method for Time of Setting of Concrete Mixtures by Penetration Resistance C88 Standard Test Method for Soundness of Aggregates by Use of Sodium Sulfate or Magnesium Sulfate D1883 Standard Test Method for CBR of Laboratory Compacted Soils D2216 Standard Test Method for Laboratory Determination of Water (Moisture) Content of Soil and Rock by Mass D2434 Standard Test Method for Permeability of Granular Soils (Constant Head) D3987 Standard Test Method for Shake Extraction of Solid Waste with Water D69 Standard Test Methods for Friction Tapes D698 Standard Test Methods for Laboratory Compaction Characteristics of Soils Using Standard Effort D854 Standard Test Method for Specific Gravity of Soil Solids by Water Pycnometer