CHARACTERIZATION OF MINERAL PROCESSING WASTES AND MATERIALS ENVIRONMENTAL PROTECTION AGENCY

Transcription

1 HARATERIZATION OF MINERAL PROESSING WASTES AND MATERIALS ENVIRONMENTAL PROTETION AGENY 1997

2 1. Introduction In its January 25, 1996 Supplemental Proposed Rule, the Agency assumed that landbased storage units historically have been a significant part of the production process of the mining and mineral processing industries, primarily because of the large volumes of materials managed by the industry. EPA believed that the quantities of secondary materials were too large to be managed in anything other than land-based units. Based on new data and further analysis, however, EPA has found that generation rates of wastes from the mineral processing industry are similar to other industrial wastes currently regulated under RRA. Further, many of the newly identified wastes have concentrations of hazardous constituents that are similar to wastes currently regulated under the RRA Subtitle program. The Agency is presenting information indicating that land-based storage units are not necessarily an integral to the mineral processing industry. The Agency s information indicates that tanks, containers, and buildings, which provide greater environmental protection, can be used to store mineral processing secondary materials prior to recycling. In the Supplemental Proposed Rule, the Agency also raised the issue of whether to allow mineral processing secondary materials to be recycled in units generating Bevill-exempt wastes. The Agency has found many cases in which environmental damages were caused by these Bevill-exempt wastes, including several cases in which non-bevill feedstocks were being added to units generating the exempt waste. (See Damage ases and Environmental Releases, EPA, 1997). Because of these cases, the Agency is concerned about the contribution of contaminants from non-bevill feedstocks. Therefore, the Agency has compared desirable and undesirable contaminants in virgin Bevill unit feedstocks with secondary materials that might be used as alternative feedstocks to these units, and found that these secondary materials often have higher contaminant concentrations than the virgin feedstock. 2. omparison of Waste Stream and Toxicity Data This section first presents current waste stream data and then compares these data with the data used in the RIA accompanying the January 1996 Supplemental Proposed Rule. Finally, this section compares mineral processing waste streams with currently listed RRA Subtitle wastes in terms of generation volumes and toxicity.. January 1996 Data vs Data Provided in Public omments The Agency received many comments on its January 1996 proposed mineral processing rule that addressed the amounts and types of secondary materials generated. Based on a review of those comments, the Agency found that a number of waste streams either are no longer generated or are generated in different quantities than previously 2

3 believed. In addition, one commenter suggested new waste streams should be added to the analysis. Exhibit 1 lists all of the waste streams that: 1) have dropped out of the analysis; 2) have been added to the analysis; or 3) have been assigned a new average facility generation rate based on the review of new data. As seen in the exhibit, 32 waste streams have dropped out of the analysis because in most cases the Agency determined that these wastes are either non-hazardous or not stored in land-based units. Two waste streams, elemental phosphorous andersen filter media and furnace building washdown were added to the analysis because the Agency received comments suggesting they were hazardous. Two more waste streams, elemental phosphorous AFM rinsate and furnace scrubber blowdown were assigned new generation rates. Please note that average facility generation rates, rather than total sector generation rates, were used because the Bevill exclusion defines high volume streams based on average facility generation rates. Exhibits 2 and 3 graphically present the old and new average facility generation rates for mineral processing secondary materials for those cases in which the current generation rates differ from those rates used in the December 1995 RIA supporting the January 1996 Supplemental Proposed Rule. In most cases, changes in generation rates resulted from the removal of waste streams from the analysis (hence the large numbers of 0s in the exhibits). Exhibit 2 shows that facility generation rates for phosphorous andersen media filter and zinc WWTP solids (both solid wastes) rose because of their recent inclusion in the analysis. Exhibit 3 demonstrates that the facility generation rates for three liquid wastes, phosphorous furnace building washdown and furnace scrubber blowdown and AFM rinsate rose between December In two of the cases, wastes were added for the January 1997 analysis, while in the third case, commenters reported a change in generation rate. In the January 1996 Supplemental Proposed Rule, 148 mineral processing secondary materials were identified that could be affected by the Phase IV LDRs. However, due to new information contained in public comment on the proposed rule, as well as other additional information, the Agency now believes 118 mineral processing secondary materials will be affected by the Phase IV LDRs. Of these streams, 61 are solid and 57 are liquid. urrent Volumes Exhibit 4 presents average and maximum facility generation rates for all solid waste streams in the current data set in ascending order, while Exhibit 5 presents this information for 1 liquid waste streams. These figures were obtained from the mineral processing cost RIA cost model. Only three waste streams are generated in volumes above the high volume criterion (45,000 mt/yr for solid materials or 1,000,000 mt/yr for liquid materials) used in the 1990 Report 1 Please note that these generation rates have not been corrected to account for uncertainty in hazard characteristics, as was done in the cost model for the RIA. 3

4 Exhibit 1 Waste Streams Status hanges Since December 1995 Sector Waste Stream Waste Type Action Reason Beryllium Bertrandite thickener slurry Liquid Dropped Out of Analysis Public comment indicate previous agency decision on beneficiation processing line Beryllium Beryl thickener slurry Liquid Dropped Out of Analysis Public comment indicate previous agency decision on beneficiation processing line Beryllium Spent barren filtrate streams Liquid Dropped Out of Analysis Public comment indicate previous agency decision on beneficiation processing line Beryllium Spent raffinate Liquid Dropped Out of Analysis Public comment indicate previous agency decision on beneficiation processing line Boron Waste liquor Liquid Dropped Out of Analysis Determined to be not-hazardous opper AP dust/sludge Solid Dropped Out of Analysis Not land stored opper Process wastewaters Liquid Dropped Out of Analysis Not land stored opper Scrubber blowdown Liquid Dropped Out of Analysis Believed to be same as acid plant blowdown, removed to prevent double counting opper Spent bleed electrolyte Liquid Dropped Out of Analysis Not land stored opper Surface impoundment waste Liquid Dropped Out of Analysis Double counted (same as process liquids wastewaters) opper Tankhouse slimes Solid Dropped Out of Analysis Not land stored opper Waste contact cooling water Liquid Dropped Out of Analysis Not land stored Elemental Phosphorous AFM rinsate Liquid hanged Generation Rate ommenter provided data Elemental Phosphorous Andersen Filter Media Solid Added to Analysis ommenter indicated this material is hazardous Elemental Phosphorous Dust Solid Dropped Out of Analysis ommenter provided data Elemental Phosphorous Furnace Building Washdown Liquid Added to Analysis ommenter provided data 4

5 Elemental Phosphorous Furnace offgas solids Solid Dropped Out of Analysis Not toxic Elemental Phosphorous Furnace scrubber blowdown Liquid hanged Generation Rate ommenter provided data Exhibit 1 (ontinued) Waste Streams Status hanges Since December 1995 Sector Waste Stream Waste Type Action Reason Gold and Silver Refining wastes Solid Dropped Out of Analysis Generated at secondary smelter only Gold and Silver Slag Solid Dropped Out of Analysis Not land stored Gold and Silver Spent Furnace Dust Solid Dropped Out of Analysis Not land stored Gold and Silver Wastewater Liquid Dropped Out of Analysis Generated at secondary smelter only Gold and Silver Wastewater treatment sludge Solid Dropped Out of Analysis Generated at secondary smelter only Lead Acid plant blowdown Liquid Dropped Out of Analysis Fully recycled, not land stored Lead Baghouse dust Solid Dropped Out of Analysis Fully recycled, not land stored Lead Process wastewater Liquid Dropped Out of Analysis Fully recycled, not land stored Lead Surface impoundment waste Liquid Dropped Out of Analysis No longer generated liquids Molybdenum Molybdic oxide refining wastes Solid Dropped Out of Analysis No longer generated Rare Earths Spent lead filter cake Solid Dropped Out of Analysis Fully recycled, not land stored Rare Earths Waste solvent Liquid Dropped Out of Analysis Fully recycled, not land stored Rare Earths Waste zinc contaminated with Solid Dropped Out of Analysis No longer generated mercury Titanium and Titanium Scrap detergent wash water Liquid Dropped Out of Analysis Not hazardous Dioxide Titanium and Titanium Waste acids (hloride process) Liquid Dropped Out of Analysis Fully recycled/treated, not land Dioxide stored Titanium and Titanium Waste ferric chloride Liquid Dropped Out of Analysis Same as Wastes acids (chloride Dioxide process) Zinc Spent surface impoundment Solid Dropped Out of Analysis No longer generated solids Zinc Zinc-lean slag Solid Dropped Out of Analysis This is a special waste Zinc WWTP solids Solid Added to Analysis New information on management practices 5

6 6

7 Exhibit 2 Average Facility Generation Rates December 1995 vs. January 1997 (Solid Wastes - Expected Value ase) 1200 Dec Jan. 97 Generation Rate (mt/yr) Phosphorous Andersen Filter Media Zinc WWTP Solids Rare Earths Spent lead filter cake Zinc Spent surface impoundment solids opper Tankhouse slimes Molybdenum, Ferromolybdenum, and Ammonium Molybdate Elemental Phosphorus Dust Rare Earths Waste zinc contaminated with mercury Waste Stream 7

8 Exhibit 2 (ontinued) Average Facility Generation Rates December 1995 vs. January 1997 (Solid Wastes - Expected Value ase) Dec. 95 Jan Generation Rate (mt/yr) Zinc Zinc-lean slag opper AP dust/sludge Gold and Silver Refining wastes Gold and Silver Slag Gold and Silver Spent Furnace Dust Gold and Silver Wastewater treatment sludge Elemental Phosphorus Furnace offgas solids Lead Baghouse dust Waste Stream 8

9 Exhibit Average Facility Generation Rates December 1995 vs. January 1997 (Liquid Wastes - Expected Value ase) Dec. 95 Jan. 97 Generation Rate (mt/yr) Phosphorous Furnace Building Washdown Phosphorous Furnace scrubber blowdown opper Waste contact cooling water Phosphorous AFM rinsate Beryllium Beryl thickener slurry 9 Titanium and Titanium Dioxide Waste ferric chloride Waste Stream Titanium and Titanium Dioxide Waste acids (hloride process) Boron Waste liquor Rare Earths Waste solvent Gold and Silver Wastewater

10 Exhibit 3 (ontinued) Average Facility Generation Rates December 1995 vs. January 1997 (Liquid Wastes - Expected Value ase) Dec. 95 Jan Generation Rate (mt/yr) opper Scrubber blowdown opper Spent bleed electrolyte opper Surface impoundment waste liquids Beryllium Spent barren filtrate streams Beryllium Bertrandite thickener slurry Lead Acid plant blowdown Titanium and Titanium Dioxide Scrap detergent wash water Beryllium Spent raffinate opper Process wastewaters Lead Surface impoundment waste liquids Lead Process wastewater Waste Stream 10

11 Exhibit 4 Average Facility Waste Generation Rates (mt/yr) Solid Wastes ommodity Waste Stream Expected Maximum Mercury Dust 1 1 Germanium Leach residues 2 3 Mercury Furnace residue 6 11 Platinum Slag Bismuth Electrolytic slimes alcium Dust with quicklime Zinc Spent cloths, bags, and filters Germanium hlorinator wet air pollution control sludge Germanium Hydrolysis filtrate Germanium Waste still liquor Antimony Stripped anolyte solids Lead Solid residues Uranium Uranium chips from ingot production Selenium Spent filter cake 85 1,700 Selenium Slag 85 1,700 Selenium Tellurium slime wastes 85 1,700 Selenium Waste solids 85 1,700 Zinc WWTP Solids Phosphorous Andersen Filter Media Uranium Slag 250 1,000 opper WWTP sludge Lead Spent furnace brick Rare Earths Electrolytic cell caustic wet AP sludge 350 7,000 Magnesium ast house dust 380 7,600 admium opper and lead sulfate filter cakes 475 9,500 admium opper removal filter cake 475 9,500 admium Iron containing impurities 475 9,500 admium Lead sulfate waste 475 9,500 admium Post-leach filter cake 475 9,500 admium Zinc precipitates 475 9,500 Bismuth Slag ,000 Tantalum Digester sludge Tellurium Slag 500 4,500 Tellurium Solid waste residues 500 4,500 Zinc Discarded refractory brick 500 1,000 Aluminum ast house dust Lead Baghouse incinerator ash 1,000 10,000 Tantalum Spent raffinate solids 1,000 1,000 Rare Earths Solvent extraction crud 1,150 4,500 Aluminum Electrolysis waste 1,250 2,500 11

12 Exhibit 4 Average Facility Waste Generation Rates (mt/yr) Solid Wastes ommodity Waste Stream Expected Maximum Bismuth Alloy residues 1,500 6,000 Bismuth Lead and zinc chlorides 1,500 6,000 Bismuth Metal chloride residues 1,500 3,000 Antimony Slag and furnace residue 1,750 3,500 Lead Slurried AP Dust 2,300 2,300 Lead Acid plant sludge 2,350 4,700 Titanium Spent surface impoundments solids 2,550 5,100 Zinc Spent goethite and leach cake residues 5,000 5,000 Zinc Spent synthetic gypsum 5,300 5,300 Zinc Waste ferrosilicon 8,500 17,000 Titanium Smut from Mg recovery 11,000 23,000 Beryllium Filtration discard 11,500 45,000 Molybdenum Flue dust/gases 11,500 45,000 Pyrobitumens Still bottoms 11,500 45,000 Magnesium Smut 13,000 13,000 Synthetic Rutile AP dust/sludges 15,000 30,000 Lead Stockpiled miscellaneous plant waste 22,000 45,000 Rhenium Spent rhenium raffinate 22,000 44,000 Synthetic Rutile Spent iron oxide slurry 22,500 45,000 Titanium WWTP sludge/solids 60,000 60,000 Lead WWTP sludges/solids 95,000 95,000 12

13 Exhibit 5 Average Facility Waste Generation Rates (mt/yr) Liquid Wastes ommodity Waste Stream Expected Maximum oal Gas Multiple effects evaporator concentrate - 65,000 Tungsten Spent acid and rinse water - 3,500 Zirconium Spent acid leachate from Zr alloy prod ,000 Zirconium Spent acid leachate from Zr metal prod ,000 Rhenium Spent barren scrubber liquor Bismuth Waste acids Uranium Waste nitric acid from UO2 production Zinc TA tower blowdown Titanium Spent surface impoundment liquids Molybdenum Liquid residues Germanium Waste acid wash and rinse water 275 1,000 Germanium Spent acid/leachate 275 1,000 Uranium Vaporizer condensate 275 1,000 Uranium Superheater condensate 275 1,000 Scandium Spent acids 280 1,000 Scandium Spent solvents from solvent extraction 280 1,000 Platinum Spent acids 285 1,000 Platinum Spent solvents 285 1,000 Tungsten Process wastewater 370 1,500 Titanium Pickle liquor and wash water 450 1,100 admium austic washwater 475 9,500 admium Spent leach solution 475 9,500 admium Spent purification solution 475 9,500 admium Scrubber wastewater 475 9,500 admium Spent electrolyte 475 9,500 Tellurium Waste electrolyte ,000 Phosphorous AFM rinsate 2,000 2,000 Antimony Autoclave filtrate 2,250 9,000 Fluorspar Off-spec fluosilicic acid 2,500 15,000 Pyrobitumens Waste catalysts 2,500 10,000 Titanium Scrap milling scrubber water 2,500 6,000 Bismuth Spent caustic soda 3,050 12,000 Bismuth Spent electrolyte 3,050 12,000 Bismuth Spent soda solution 3,050 12,000 Bismuth Waste acid solutions 3,050 12,000 Mercury Quench water 5,500 60,000 Rare Earths Process wastewater 7,000 7,000 Tellurium Wastewater 10,000 20,000 Zirconium Leaching rinse water from Zr alloy prod. 10,500 26,000 Rare Earths Spent ammonium nitrate processing solution 14,000 14,000 Synthetic Rutile Spent acid solution 15,000 30,000 13

14 Exhibit 5 Average Facility Waste Generation Rates (mt/yr) Liquid Wastes ommodity Waste Stream Expected Maximum Titanium Waste acids (Sulfate process) 20,000 39,000 Beryllium hip treatment wastewater 25,000 1,000,000 Selenium Plant process wastewater 33,000 33,000 Tantalum Process wastewater 75,000 75,000 Zinc Acid plant blowdown 130, ,000 Phosphorous Furnace scrubber blowdown 210, ,000 Titanium Leach liquor and sponge wash water 240, ,000 Rare Earths Wastewater from caustic wet AP 250,000 1,000,000 Zirconium Leaching rinse water from Zr metal prod. 250,000 1,000,000 Phosphorous Furnace Building Washdown 350, ,000 Zinc Wastewater treatment plant liquid effluent 435, ,000 Rare Earths Spent scrubber liquor 500,000 1,000,000 opper Acid plant blowdown 530, ,000 Zinc Spent surface impoundment liquids 630, ,000 Lead WWTP liquid effluent 880, ,000 Zinc Process wastewater 1,700,000 1,700,000 14

15 to ongress on Special Wastes from Mineral Processing in the expected volume case. These streams, which are described in more detail below, are: Wastewater treatment plant sludges and solids from the titanium sector, Wastewater treatment plant sludges and solids from the lead sector, and Process wastewater from the zinc sector. Exhibit 6 is a histogram of average facility solid waste generation rates for all streams presently in the analysis. As can be seen in this exhibit, 48 of the 61 wastes streams have average facility generation rates at or below 5000 mt/yr. Exhibit 7 provides a more detailed look at the distribution of these 48 lower volume waste streams. Of these 48 low volume waste streams, 35 are generated at rates at or below 500 mt/yr. Exhibit 8 presents a histogram of average facility liquid waste generation rates. As can be seen in this exhibit, 31 of the 51 waste streams, are generated at average rates of less than 5,000 mt/yr. Exhibit 9 presents a more detailed look at these wastes. Of these 35 low volume waste streams, 22 are generated at rates at or below 500 mt/yr. In summary, the Agency found that of the 118 waste streams, 115 (97 percent) were generated in quantities lower than the respective high volume Bevill cutoffs for solids and liquids. Even more demonstrative is that 79 (48 solid wastes and 31 liquid wastes) of these 118 wastes steams (67 percent) are generated in quantities less than 5000 tons per year. High Volume Streams The three wastes streams that exceed the high volume thresholds are described below, 2 because they may require special consideration in determining appropriate storage practices. These streams exceed the thresholds in all three cases, because they are generated by commingling numerous other waste streams, either directly in the case of process wastewater from the zinc sector or as a result of treatment operations (i.e., the two WWTP sludge streams). Titanium - Wastewater Treatment Plant Sludge/Solids Wastewater treatment plant (WWTP) sludge/solids, a post-mineral processing waste, consists of sludges and solids resulting from the treatment of the wastewater treatment plant liquid effluent. Sludge/solids are disposed in on- or off-site landfills. Approximately 420,000 metric tons are generated annually by the entire sector. Titanium waste may exhibit the characteristics of toxicity (chromium). Lead - WWTP Sludges/Solids Wastewater treatment sludges and solids consist of solid materials that settle following lime neutralization of influent wastewaters. The sludges and solids typically are recycled to the sinter feed preparation operation. For example, at the Doe Run Herculaneum facility, a thickener serves as the final collection point for solids in the WWTP. Thickener solids are dewatered using a filter press and 2 US EPA, Identification and Descriptions of Mineral Processing Sectors and Waste Streams, December 1995, pp. 401, 711, and

16 then shipped by rail car to the sinter plant. Approximately 380,000 metric tons of WWTP sludges and solids 16

17 Exhibit Histogram Distribution of Average Facility Generation Rates (All Solid Wastes - Expected Value ase) Frequency < = 5,000 < = 10,000 < = 15,000 < = 20,000 < = 25,000 < = 40,000 < = 45,000 < = 50,000 > 50,000 Range (Generation Rate in mt/yr) 17

18 Exhibit Histogram Distribution of Average Facility Generation Rates (Low Volume Solid Wastes Only - Expected Value ase) Frequency < = 500 < = 1000 < = 1500 < = 2000 < = 2500 < = 3500 < = 4000 < = 4500 < = 5000 Range (Generation Rate in mt/yr) 18

19 Exhibit 8 19

20 Exhibit 9 20

21 are generated annually by the entire sector. The waste generation rate per facility is greater than 45,000 metric tons/yr due to commingling of numerous waste streams. The Newly identified mineral 25 Histogram Distribution of Average Facility Generation Rates (Low Volume Liquid Wastes Only - Expected Value ase) Frequency < = 500 < = 1000 < = 1500 < = 2000 < = 2500 < = 3500 < = 4000 < = 4500 < = 5000 Range (Generation Rate inmt/yr) processing Waste haracterization Data Set contains data indicating that this waste stream may exhibit a hazardous characteristic. The lead waste stream may exhibit the characteristic of toxicity (cadmium and lead). The waste stream is fully recycled and is classified as a sludge. In April 1991, SAI conducted a study that contains data on samples of clarifier underflow and filter press solids collected from the wastewater treatment system (WWTP-1) at Doe Run's Herculaneum, Missouri facility. The clarifier underflow sample, which is derived from plant washdown and acid plant blowdown, exhibited the toxicity characteristic for cadmium (8.51 mg/l). 21

22 The filter press solids, which are derived from thickened clarifier underflow and sinter plant blowdown, exhibited the toxicity characteristic for lead (185 mg/l) and cadmium (98.8 mg/l). The Doe Run samples were not analyzed for any organic compounds. (SAI, 1991b, pp. 13, 15) Zinc - Process Wastewater Process wastewater is generated at all four of the operating zinc processing plants. Approximately 6.6 million metric tons of process wastewater are generated annually at the four U.S. primary zinc facilities. (EPA, August 1992) (The excessive generation rate for this wastewater [i.e., greater than one million metric tons/yr] is due to commingling of numerous wastestreams.) Zinc process wastewater may be recycled and may exhibit the characteristic of toxicity for arsenic, cadmium, chromium, lead, selenium, and silver. It may also exhibit the corrosivity characteristic. The waste is classified as a spent material. At ZA's electrolytic refinery in Bartlesville, Oklahoma process wastewaters consist of small streams from the roasting, purification, electrowinning, and zinc secondaries processes. Process wastewater and plant runoff collect in two large, clay-lined surface impoundments and are pumped to the wastewater treatment plant for neutralization. At ZA's Monaca, Pennsylvania smelter, wastewaters include plant runoff as well as process wastewater from the blue powder impoundments and the zinc sulfate circuit. These wastewaters collect in a lined equalization basin and are treated in a two-stage neutralization process. omparison with Similar Listed Wastes This section compares the volumes, toxicities, and management of mineral processing secondary materials with several similar industrial wastes currently regulated under RRA. The purpose of this comparison is to demonstrate the similarity between mineral processing secondary materials and RRA Subtitle hazardous waste. Generation Rates Many mineral processing waste streams are generated in quantities similar to other industrial wastes that are managed in tanks, containers, and buildings. To compare generation rates, EPA identified 8 listed waste streams that are similar to mineral processing wastes, or are generated by industries similar to the primary mineral processing industry. Generation rates for each of these waste streams were obtained from the Biennial Reporting System (BRS). An average generation rate for each of the listed wastes was then calculated by dividing total waste generated by the number of facilities that reported the waste stream in 3 BRS. These average rates are presented in Exhibit 10. Exhibit 11 compares the generation rates of the solid listed wastes in Exhibit 10 (F006, F019, K060, K061, K069, K088, and K141) with all of the solid newly identified mineral processing wastes. Exhibit 12 provides 3 BRS data include information on wastes managed at commercial treatment, storage, and disposal facilities (TSDFs), as well as generators. Prior to calculating average facility generation rates, IF deleted from the BRS data any entries for known commercial TSDFs (e.g., hemical Waste Management, USPI, lean Harbors, etc.) 22

23 more detail on the lower volume (i.e., less than 5,000 mt/yr) mineral processing wastes. Exhibit 13 compares the generation rates of the liquid listed wastes in Exhibit 10 (K062) with all of the liquid newly identified mineral processing wastes. It is evident from Exhibits 11 through 13 that the 14 listed waste streams often are generated at higher rates than wastes generated in the mineral processing industry. For example, K069 (Emission control dust/sludge from secondary lead smelting) has the smallest average generation rate (710 mt/yr) of the solid listed wastes outlined in Exhibit 10. However, this generation rate is still larger than average generation rates for 35 of the 61 newly identified solid waste streams in the analysis. Twenty-one of the newly identified solid wastes are generated in quantities similar to the average generation rates of the solid listed wastes in Exhibit 10, while only five waste streams are generated in larger quantities. Only two of those five waste streams (lead WWTP sludges/solids and titanium WWTP sludges/solids) are generated in sufficient quantities to be above the high-volume threshold for Bevill status (45,000 mt/yr for solid wastes). K062 (spent pickle liquor generated by steel finishing operations), has the smallest average generation rate (31,335 mt/yr) of the liquid listed wastes in Exhibit 10. This generation rate is greater than 39 of the 53 newly identified liquid waste streams in the analysis. One of the 53 liquid wastes is generated in quantities similar to the average generation rates of the liquid wastes in Exhibit 10, and thirteen liquid wastes are generated in larger quantities than the liquid listed wastes. Only one newly identified liquid waste stream, zinc process wastewater, is generated in sufficient quantity to be above the high volume threshold of the analysis (1,000,000 mt/yr for liquid wastes). Toxicities Many of the newly identified mineral processing waste streams pose threats similar to those posed by RRA listed wastes. Of the current 118 newly identified waste streams in the analysis, 23 are known to be hazardous, and the rest are suspected to be hazardous. Exhibit 14 summarizes the T metal and corrosivity characteristic data for these 23 waste streams. This exhibit lists each known hazardous waste stream, the metals for which it fails T levels, and the ph level if the waste is corrosive. Using the RRA Subtitle Identification and Listing of Hazardous Waste Background Document (hereafter referred to as the listing background document) as a source, EPA identified leachate or total constituent concentration data for seven listed wastes. The listing background document, however, provided concentration data for only three of the eight T metals: chromium, cadmium, and lead. In addition, some of the leachate results presented in the listing background document were obtained using water extraction tests rather than the EP test. Although water extraction tests and the EP test will not likely produce the same leachate concentration due to the difference in leaching agents (distilled water vs. acetic acid), water extraction data can be compared to EP data because: wastes may leach harmful concentrations of lead, cadmium, and hexavalent chromium even under relatively mild environmental conditions. If these wastes are exposed to more acidic disposal environments, for example disposal environments subject to acid rainfall, these metals would most likely be 23

24 solubilized to a considerable extent, since lead, and cadmium (including their oxides), as well as most chromium compounds, are more soluble in acid than in distilled water. 4 Exhibit 15 presents a summary of the toxicity characteristic data for the seven listed wastes identified from the listing background document. In order to easily compare the listed waste leachate concentrations with the leachate concentrations of the newly identified mineral processing wastes, a combined mean and maximum range of chromium, cadmium, and lead concentrations for the seven listed wastes were calculated using the data in Exhibit 15. The mean leachate concentrations for chromium, cadmium, and lead range from 6.03 mg/l to mg/l, <0.01 mg/l to mg/l, and 1.47 mg/l to mg/l, respectively. Likewise, the maximum leachate concentrations for chromium, cadmium, and lead range from 12 mg/l to 4250 mg/l, <0.01 mg/l to 268 mg/l, and 2.10 mg/l to 1550 mg/l, respectively. Exhibit 16 presents a comparison of the ranges in constituent concentrations exhibited by the listed wastes and the newly identified mineral processing wastes. As can be seen in Exhibit 16, 15 of the 23 mineral processing wastes exhibit leachate concentrations of chromium, cadmium, and lead at levels that are equal to or greater than those levels exhibited by the seven listed wastes summarized in Exhibit 15. These fifteen mineral processing wastes, arranged in alphabetical order, are as follows: Aluminum ast House Dust; opper Acid Plant Blowdown; Elemental Phosphorous AFM Rinsate; Elemental Phosphorous Furnace Scrubber Blowdown; Lead Baghouse Incinerator Ash; Lead Slurried AP Dust; Lead Spent Furnace Brick; Lead Stockpile Miscellaneous Plant Waste; Rare Earths Process Wastewater; Selenium Plant Process Wastewater; Titanium and Titanium Dioxide Waste Acids (Sulfate Process); Zinc Acid Plant Blowdown; Zinc Process Wastewater; Zinc Spent Goethite and Leach ake Residues; and Zinc Spent Synthetic Gypsum. A comparison of the remaining eight newly identified mineral processing waste streams could not be made because the listing background document did not provide data 4 U.S. Environmental Protection Agency, Office of Solid Waste. Background Document, Resource onservation and Recovery Act, Subtitle - Identification and Listing of Hazardous Waste, Washington D..,

25 for arsenic, barium, mercury, selenium, silver, or ph. However, it is important to note that at least one sample from each of these waste streams exhibits the RRA characteristics of a hazardous waste (i.e., exceeds the T or the characteristic for corrosivity). onclusions Based on the comments received and further evaluation of new data, the Agency has found the volumes of hazardous secondary materials from mineral processing to be much lower than earlier believed. Additional Agency analyses also suggest that mineral processing wastes contain concentrations of contaminants similar to those found in RRA listed wastes. Because the newly identified mineral processing wastes exhibit characteristics that are similar to other RRA listed metal-bearing waste streams, the Agency believes that the newly identified mineral processing wastes should be subject to the same storage requirements (and prohibitions) faced by all other RRA hazardous wastes (listed and characteristic). These requirements are described briefly below. As stipulated in 40 FR (a), generators of RRA hazardous wastes (listed and characteristic) may accumulate hazardous waste on-site for 90 days or less without a permit or without having interim status, provided that the waste is placed: (I) In containers and the generator complies with subpart I of 40 FR part 265; and/or (ii) (iii) (iv) In tanks and the generator complies with subpart J of 40 FR part 265, except and ; and/or On drip pads and the generator complies with subpart W of part 265 and maintains the following records at the facility... The waste is placed in containment buildings and the generator complies with subpart DD of 40 FR part RRA hazardous wastes (listed and characteristic) that are stored in excess of 90 days (unless an extension is granted) must be stored in Part B permitted storage units that meet minimum technical construction requirements. Furthermore, as stipulated in 40 FR , the land-based storage of wastes that exhibit concentrations of hazardous constituents in excess of the treatment standards set under the Land Disposal Restrictions program, is prohibited 25

26 Exhibit 10. Average Facility Generation of Listed Wastes Waste Waste Description Average Number Generation (Metric Tons/Year) F006 Wastewater treatment sludges from electroplating operations 3,528 F019 Wastewater treatment sludges from the aluminum coating 3,207 K060 Ammonia still lime sludge from coking operations 17,978 K061 Emission control dust/sludg e from the primary production of steel in electric 13,975 furnaces K062 Spent pickle liquor generated by steel finishing operations 31,335 K069 Emission control dust/sludge from secondary lead smelting 710 K088 Spent potliners from primary aluminum reduction 4,613 K141 Process residues from the recovery of coal tar, including, but not limited to, 4,221 collecting sump residues from the production of coke from coal or th e recovery of coke by-products produced form coal. This listing does no t include K087 (decanter tank 26

27 Exhibit 11 Histogram Distribution of Average Facility Generation Rates (All Solid Wastes - Expected Value ase) Frequency K K < = 5,000 < = 10,000 < = 15,000 < = 20,000 < = 25,000 < = 40,000 < = 45,000 < = 50,000 > 50,000 Range (Generation Rate in mt/yr) F006, F019, K069, K088, K141 27

28 Exhibit 12 28

29 Exhibit 13 Histogram Distribution of Average Facility Generation Rates (All Liquid Wastes - Expected Value ase) Frequency K < = 5,000 < = 10,000 < = 15,000 < = 20,000 < = 25,000 < = 40,000 < = 45,000 < = 50,000 > 50,000 Range (Generation Rate in mt/yr) 29

30 EXHIBIT 14. SUMMARY OF EP ANALYSIS RESULTS FOR MINERAL PROESSING WASTE Sector/Waste Stream onstituent T Limits Min Mean Max Number Numbe Number (mg/l) of r of above Sample detect T s s Aluminum ast House Dust admium Mercury oal Gas MEE oncentrate Arsenic Selenium opper Acid Plant Blowdown Arsenic admium hromium Lead Mercury Selenium Silver ph 2<pH> Elemental Phosphorous AFM Rinsate admium Selenium Elemental Phosphorous Furnace Scrubbe r Blowdown admium Lead Baghouse Incinerator Ash admium Lead Lead Slurried AP Dust admium

31 EXHIBIT 14. SUMMARY OF EP ANALYSIS RESULTS FOR MINERAL PROESSING WASTE Sector/Waste Stream onstituent T Limits Min Mean Max Number Numbe Number (mg/l) of r of above Sample detect T s s Lead Lead Spent Furnace Brick Lead Lead Stockpile Miscellaneous Plant Waste admium Lead Lead WWTP Liquid Effluent ph 2<pH> Lead WWTP Sludges/Solids ph 2<pH> Magnesium & Magnesia Smut Barium Rare Earths Spent Ammonium Nitrate Processin g Solution ph 2<pH> Rare Earths Process Wastewater Lead Selenium Plant Process Wastewater Lead ph 2<pH> Tantalum Process Wastewater ph 2<pH> Titanium and Titanium Dioxide Waste Acids (Sulfate Process) Arsenic hromium Selenium

32 EXHIBIT 14. SUMMARY OF EP ANALYSIS RESULTS FOR MINERAL PROESSING WASTE Sector/Waste Stream onstituent T Limits Min Mean Max Number Numbe Number (mg/l) of r of above Sample detect T s s Titanium and Titanium Dioxide Leach Liquor & Sponge Wash Water Silver ph 2<pH> ph 2<pH> Zinc Acid Plant Blowdown Arsenic admium hromium Selenium Silver ph 2<pH> Zinc Process Wastewater Arsenic admium hromium Lead Selenium Silver ph 2<pH> Zinc Spent Goethite & Leach ake Residues Arsenic admium hromium Selenium Silver Zinc Spent Surface Impoundment Liquids ph 2<pH>

33 EXHIBIT 14. SUMMARY OF EP ANALYSIS RESULTS FOR MINERAL PROESSING Sector/Waste Stream onstituent T Limits Min Mean Max N (mg/l) S Zinc Spent Synthetic Gypsum admium Note: Gray shading indicates detection limit may have been equal to or higher than T limit. Exhibit 15 Listed Waste Leachate oncentration Data (Source: EPA Listing Background Document) Listed Waste onstituent T Limits Min Mean Max (mg/l) F006 1 hromium 5 < admium 1 < Lead K061 1 hromium 5 < admium Lead 5 < K062 2 hromium admium Lead K064 3 hromium admium Lead K065 2 hromium admium Lead K066 hromium admium 1 <0.01 <0.01 <0.01 Lead K069 hromium admium Lead Incorporates both EP and water test data. 2 This listed waste is a liquid; therefore, total concentration data is presented. 3 Based on one data point. 33

34 believed. to more then are The for Elemental ongress generated filter shipped arsenic, detail solubilized press Phosphorous In by on annually barium, addition, solids, rail the Special EXHIBIT car lower to which by mercury, a to one considerable Wastes the volume are commenter entire 14. sinter derived selenium, SUMMARY from sector. plant. (i.e., extent, Mineral from less suggested Approximately The silver, thickened than waste since Exhibit OF Processing 5,000 EP or new generation lead, ph. 45 clarifier ANALYSIS 380,000 Exhibit mt/yr) waste and However, in underflow cadmium Exhibit rate the streams metric mineral 23 expected per RESULTS (ontinued) it facility tons is should and processing 1(including important of sinter volume is WWTP FOR greater be plant added their MINERAL wastes. to case. sludges than note blowdown, to These 45,000 that the and PROESSING WASTE Average Facility Waste Generation Rates (mt/yr) Liquid Solid Wastes Sector/Waste Stream onstituent T Limits Min Mean Max Number Numbe Number ommodity Waste Waste Stream Stream Expected Expected (mg/l) Maximum Maximum of r of above Titanium Sample detect T Bismuth s s

35 EXHIBIT 16. OMPARISON OF LISTED WASTE AND MINERAL PROESSING WASTE ONENTRATION DATA (mg/l) Sector/Waste Stream onstituent T Limits Min Mean Max Range of Listed Waste oncentrations Mean Max Aluminum ast House Dust admium < < Mercury oal Gas MEE oncentrate Arsenic Selenium opper Acid Plant Blowdown Arsenic admium < < hromium Lead Mercury Selenium Silver ph 2<pH> Elemental Phosphorous AFM Rinsate admium < < Selenium Elemental Phosphorous Furnace Scrubbe r Blowdown admium < <

36 EXHIBIT 16. OMPARISON OF LISTED WASTE AND MINERAL PROESSING WASTE ONENTRATION DATA (mg/l) Sector/Waste Stream onstituent T Limits Min Mean Max Range of Listed Waste oncentrations Mean Max Lead Baghouse Incinerator Ash admium < < Lead Lead Slurried AP Dust admium < < Lead Lead Spent Furnace Brick Lead Lead Stockpile Miscellaneous Plant Waste admium < < Lead Lead WWTP Liquid Effluent ph 2<pH> Lead WWTP Sludges/Solids ph 2<pH> Magnesium & Magnesia Smut Barium Rare Earths Spent Ammonium Nitrate Processin g Solution ph 2<pH> Rare Earths Process Wastewater Lead Selenium Plant Process Wastewater Lead ph 2<pH>

37 EXHIBIT 16. OMPARISON OF LISTED WASTE AND MINERAL PROESSING WASTE ONENTRATION DATA (mg/l) Sector/Waste Stream onstituent T Limits Min Mean Max Range of Listed Waste oncentrations Mean Max Tantalum Process Wastewater ph 2<pH> Titanium and Titanium Dioxide Waste Acids (Sulfate Process) Arsenic hromium Selenium Silver ph 2<pH> Titanium and Titanium Dioxide Leach Liquor & Sponge Wash Water ph 2<pH> Zinc Acid Plant Blowdown Arsenic admium < < hromium Selenium Silver ph 2<pH> Zinc Process Wastewater Arsenic admium < < hromium Lead Selenium Silver ph 2<pH>

38 EXHIBIT 16. OMPARISON OF LISTED WASTE AND MINERAL PROESSING WASTE ONENTRATION DATA (mg/l) Sector/Waste Stream onstituent T Limits Min Mean Max Range of Listed Waste oncentrations Mean Max Zinc Spent Goethite & Leach ake Residues Arsenic admium < < hromium Selenium Silver Zinc Spent Surface Impoundment Liquids ph 2<pH> Zinc Spent Synthetic Gypsum admium < <

39 unless the wastes are stored in tanks, containers, or containment buildings on-site solely for the purpose of the accumulation of such quantities of hazardous waste as necessary to facilitate proper recovery, treatment, or disposal and the generator complies with the requirement in and parts 264 and omparison with Bevill feedstock In the Supplemental Proposed Rule, the Agency raised the issue of whether to allow mineral processing secondary materials to be recycled in units generating Bevill-exempt wastes. The Agency has conducted further research on this issue, and found many cases in which environmental damages were caused by these Bevill-exempt wastes, including several cases in which non-bevill feedstocks were being added to the unit generating the exempt waste (See Damage ases and Environmental Releases, EPA, 1997). To assist in determining whether to allow alternative (non-virgin) feedstocks to be added to these Bevill units, the Agency has compared desirable and undesirable constituents of virgin Bevill unit feedstocks with those of secondary materials that might be used as alternative feedstocks to these units. This section begins with a brief description of the minerals purification process. Typical concentrations of desired and undesirable constituents are then discussed. Next, an example from the copper sector is used to show how data from mineral processing operations could be compared with the virgin feedstocks. Finally, we present conclusions regarding the comparison. General Review of the Minerals Purification Process Several stages are involved in the production of valuable products from ore. First, overburden (the consolidated or unconsolidated material that overlies a deposit of useful ore) must be remove to expose the ore. The ores are then extracted (mined) by a variety of surface and underground procedures. Surface mining methods include open-pit mining, open-cut mining, open-cast mining, dredging, and strip mining. Underground mining creates adits (horizontal passages) or shafts by room-and-pillar, block caving, timbered stope, open stope, and other methods. Waste rock, the portion of the ore body that is barren or submarginal rock or ore that has been mined but is not of sufficient value to warrant treatment, must be separated from the ore containing value. The ore containing the value must then be 5 beneficiated (concentrated or dressed). As defined in 40 FR 261.4(b)(7), beneficiation operations include: crushing; grinding; washing; dissolution; crystallization; filtration; sorting; sizing; drying; sintering; pelletizing; briquetting; calcining (to remove water and/or carbon 5 U.S. EPA, 1985, Report to ongress: Wastes from the Extraction and Beneficiation of Metallic Ores, Phosphate Rock, Asbestos, Overburden from Uranium Mining, and Oil Shale, pp

40 6 dioxide); roasting, autoclaving and/or chlorination in preparation for leaching ; gravity concentration; magnetic separation; electrostatic separation; flotation; ion exchange; solvent extraction; electrowinning; precipitation; amalgamation; and heap, dump, vat, tank, and in situ leaching. Beneficiation typically yields an intermediate product that often is further purified in mineral processing operations (such as smelting or refining) to produce pure metal and metal products. oncentrations of Desired onstituents Ores typically contain fairly low concentrations of the metal(s) of interest. Exhibit 17 presents estimates of indicates typical metal concentrations in ores. As ores move through extraction, beneficiation, and processing operations, the percentage of the metals of interest increases. For example, in pyrometallurgical processing of copper ores (sulfide ores), the ore (which has an initial concentration of less than one percent) is sent to milling and flotation. The concentrate from flotation has a copper concentration of 20 to 30 percent. The concentrate is then sent to smelting, which raises the copper concentration to 50 to 75 percent. The product of smelting (matte) is sent to the converter, which produces blister copper and further purifies the copper concentrations to 98 to 99 percent. The blister copper is sent to fire refining and then electrorefining to produce copper that is percent pure. Exhibit 17. Estimated Percentage Metal in Ore Mining Industry Segment Typical Percentage of Metal in Ore opper Gold Iron Lead Molybdenum Silver Tungsten Zinc Source: Report to ongress: Wastes from the Extraction and Beneficiation of Metallic Ores, Phosphate Rock, Asbestos, Overburden from Uranium Mining, and Oil Shale, December 1985, p oncentrations of Impurities 6 Except where the roasting (and/or autoclaving and/or chlorination)/leaching sequence produces a final or intermediate product that does not undergo further beneficiation or processing. 39

41 Tracking impurities as the ore moves through the production process is more difficult, because impurities are removed in almost every step of the purification process. Some of the impurities are removed early in the beneficiation operations, while others that are physically or chemically similar to or bound with the metal of interest may be carried along the production process for most of the processing steps. The copper sector again provides an example of how the concentrations of impurities increase or decrease as the ore is purified. Because concentration levels of impurities in the intermediate products were not available, EPA evaluated constituent concentrations in wastes generated at various points along the production process. These concentrations are summarized in Exhibits 18 and 19. There are a number of limitations associated with these data, which were obtained from several facilities over a number of years. These factors add potential for error because concentrations of impurities vary greatly within and between ore deposits, each facility is likely to mine a different ore deposit and use a different production process. In addition, increasing environmental regulation in the last ten years and changing economic markets have also caused facilities to make changes in their production processes and ore selection. Moreover, comparing the constituent concentrations of beneficiation wastes with mineral processing wastes may not adequately address EPA s underlying question about the appropriateness of using alternative feedstocks in Bevill units. EPA has examined the constituent concentrations of both beneficiation wastes and processing wastes in determining whether any of these wastes should retain their Bevill exempt status in two Reports to ongress. The comparison between extraction and beneficiation waste concentrations and mineral processing waste concentrations can yield only an insight into when in the process impurities are removed. For a complete understanding of how impurities are generated in the production process, waste volume should be considered in addition to concentration levels. Several noteworthy trends can be discerned in Table 18, which a compares concentrations in solid waste from beneficiation and mineral processing operations. The first trend is that concentrations of both the desired product and impurities are higher in the mineral processing wastes than in the beneficiation wastes. The next trend is that among the mineral processing wastes, constituent concentrations vary significantly from one waste to another. For example, the waste with the highest copper concentration (converter flue dust) has the lowest arsenic concentration. Table 19 displays contaminant concentrations for liquid wastes from beneficiation and mineral processing operations. As was seen in the data for solid waste, the mineral processing wastes have higher concentrations of almost all constituents than the beneficiation waste. The only exceptions are for chromium in WWTP liquid effluent and molybdenum in both WWTP liquid effluent and scrubber blowdown. The variability in constituent concentrations seen in the solid wastes is also seen in the liquid wastes. The highest levels of the 10 constituents are almost evenly divided between spent bleed electrolyte and acid plant blowdown. However, scrubber blowdown, which has the lowest concentration of copper has the highest concentration of mercury. 40