FINAL
     BEST DEMONSTRATED AVAILABLE TECHNOLOGY (BDAT)
BACKGROUND DOCUMENT (ADDENDUM) FOR ALL NONWASTEWATER
 FORMS OF K061 AND ALTERNATIVE BDAT TREATMENT STANDARDS
             FOR F006 AND K062 NONWASTEWATERS
                     Larry Rosengrant, Chief
                   Treatment Technology Section
                      Laura Lopez/Lisa Jones
                        Project Managers
                U.S. Environmental Protection Agency
                       Office of Solid Waste
                       401 M Street, S.W.
                     Washington, D.C. 20460
                           July 1992.

-------
                             ACKNOWLEDGMENTS
      This document was prepared for the U.S. Environmental Protection Agency, Office of
Solid Waste, by Versar Inc. under Contract No. 68-W9-0068.  Mr. Larry Rosengrant, Chief,
Treatment Technology Section, Waste Treatment Branch, served as the EPA Program Manager
during the preparation of this document. The technical project officers were Ms. Laura Lopez
and Ms. Lisa Jones.  (Mr. Steven Silverman served as legal advisor.)

      Versar personnel involved in the preparation of this document included Mr. Jerome
Strauss, Program Manager; Mr. Stephen Schwartz, Assistant Program Manager; Mr. Stanley
Moore,  Principal Investigator and Author; Mr. Jeremy Flint, Ms. Kathryn Jones, and Mr.
Amanjit Paintal, Engineering Team; Ms. Justine Alchowiak, Quality Assurance Officer; Dr.
Fauod Moumen, Senior Statistician; Ms. Janeice Zeaman, Technical Editor; and Ms. Christine
Thompson, Secretary.

-------
                           TABLE OF CONTENTS



Section                                                               2agfi

1.     INTRODUCTION AND SUMMARY	         1-1

2.     INDUSTRY AFFECTED AND WASTE CHARACTERIZATION ...         2-1

      2.1    Industry Affected  	         2'1
      2.2    Waste Characterization	         2-2

3.     APPLICABLE AND DEMONSTRATED METALS RECOVERY
      TECHNOLOGIES AND IDENTIFICATION OF BEST
      DEMONSTRATED AVAILABLE TECHNOLOGY  	         3-1

      3.1    Applicable Recovery Technologies	         3-1

            3.1.1  Recycling	         3-2
            3.1.2  Pyrometallurgical Recovery Processes	         3-2
            3.1.3  Hydrometallurgical Recovery Processes	        3-13

      3.2    Demonstrated Treatment Technologies  	        3-16
      3.3    Identification of Best Demonstrated Available
            Technology (BDAT)  	        3-17

4.    PERFORMANCE DATA	         4'!

      4.1    Data Reviewed and Selected to Develop the
            Treatment Standards Based on HTMR	         4'1
      4.2   Data Reviewed But Not Selected to Develop
            Treatment Standards Based on HTMR	         4'3

            4.2.1  Horsehead Resource Development Co., Inc.
                  (HRD) Data	        4'3
            4.2.2  International Metals Reclamation Company
                  (INMETCO) Data  	        4'8
            4.2.3  Sumitomo Corporation of America Data	         4-9
            4.2.4  International Mill Service (IMS) Data  	         4~9
            4.2.5  St. Joe Company Data	        4-l°
            4.2.6  Beckett Technologies Data	        4-10
            4.2.7  SKF Plasma Technologies Data	        4-1°

-------
                       TABLE OF CONTENTS (Continued)



Section                                                                 Eagfi

5.     SELECTION OF REGULATED CONSTITUENTS	         5-1

      5.1   Constituents Identified in the Waste as Generated	         5-1

            5.1.1  Organics  	         5-2
            5.1.2  Cyanide	         5-2
            5.1.3  Metals	         5-3

      5.2   Constituents Identified in the HTMR Residuals	         5-5
      5.3   Constituents Selected for Regulation	         5-5

6.     DEVELOPMENT OF BOAT TREATMENT STANDARDS	         6-1

      6.1   Treatment Standards Calculations Based on HTMR
            Treatment Performance Data from SKF Plasma
            Technologies	         6-6
      6.2   Treatment Standards Calculations Based on HTMR
            Treatment Performance Data from International
            Mill Service (IMS)	         6-8
      6.3   Treatment Standards Calculations Based on
            HTMR Treatment Performance Data from
            International Metals Reclamation Company
            (INMETCO) 	        6-11
      6.4   Treatment Standards Calculations Based on HTMR
            Treatment Performance Data from Horsehead
            Resource Development Co., Inc. (HRD)	        6-13
      6.5   Treatment Standards Calculation for
            Cyanide	       6-16
      6.6   Treatment Standards Calculations for Metals
            (i.e., Highest Standard from the Four Sets of
            Treatment Performance Data)	        6-18

7.    GENERIC EXCLUSION FOR K061, F006, AND K062
      NONWASTEWATER RESIDUES (SUCH AS SLAG) GENERATED
      FROM HTMR PROCESSES AND RELATED ISSUES	        7-1

      7.1   Development of Health-Based Generic Exclusion
            Levels and Calculation of Allowable Concentrations .......         7-6

                                       U

-------
                       TABLE OF CONTENTS (Continued)
Section                                                                     Page

      7.2   Product Uses of HTMR Residues  	        7-15
      7.3   Tracking Requirements ....'.	        7-16
      7.4   Testing Requirements	        7-17

            7.4.1  Operating Conditions (Condition 1)	        7-18
            7.4.2  Testing (Condition 2)  	        7-19
            7.4.3  Waste Holding and Handling (Condition 3)	        7-20
            7.4.4  Exclusion Levels (Condition 4)  	        7-21
            7.4.5  Data Submittals (Condition 5)	        7-21

      7.5   Applicability of Generic Exclusion to Other
            Treatment Residues	        7-21
      7.6   Regulatory Status of Nonwastewater Residues from
            HTMR That Do Not Meet the Generic Exclusion
            Levels	        7-23

            7.6.1  Application of the Variance from Solid
                   Waste Classification	        7-24
            7.6.2  Application of the Derived-From Rule	        7-27

8.    REFERENCES	          8-1


-------
                               LIST OF TABLES
Table 1-1    BDAT Treatment Standards for K061 (All
            Nonwastewater Forms), and Alternative Treatment
            Standards for K062 and F006 Nonwastewaters  	

Table 1-2    Generic Exclusion Levels for K061 (All Nonwastewater
            Forms), K062 and F006 HTMR Residues (Nonwastewaters)

Table 2-1    BDAT List Constituents Composition for Untreated
            K061 Nonwastewaters  	

Table 2-2    BDAT List Constituents Composition for Untreated
            F006 Nonwastewaters	

Table 2-3    BDAT List Constituents Composition for Untreated
            K062 Nonwastewaters  	

Table 3-1    Comparison Characterization Data and Treatment
            Performance Data for Inputs to an HTMR Process
            Versus Inputs to Stabilization	

Table 4-1    Performance Data for HTMR of Zinc (Series of
            Waelz  Kilns) of High  Zinc Subcategory K061
            Nonwastewaters	

Table 4-2    Performance Data (1988d) for HTMR of Zinc (Series of
            Waelz Kilns) of High and Low Zinc Subcategory
            K061 Nonwastewaters   	

Table 4-3    Performance Data for HTMR of Zinc (Plasma Arc
           , Reactor) of High Zinc  Subcategory K061
            Nonwastewaters	

Table 4-4    Performance Data for HTMR of Zinc (Plasma Arc
            Reactor) of High Zinc  Subcategory K061
            Nonwastewaters	
Page



  1-4


  1-5


  2-3


  2-6


  2-7



 3-19



 4-11



 4-12



 4-13



 4-14
                                        IV

-------
                                LIST OF TABLES
                                                                           Page
Table 4-5     Performance Data for HTMR (Rotary Hearth Furnace/
             Electric Furnace) of K061 (High and Low Zinc),
             K062, F006, and Characteristic Wastes Containing
             Metals	

Table 4-6     Performance Data for Incineration (Fluidized Bed)
             of Cyanide in K048 and K052 Nonwastewaters	

Table 4-7     Performance Data for HTMR of Zinc (Series of Waelz
             Kilns) of Low Zinc Subcategory K061 Nonwastewaters

Table 4-8     Performance Data for HTMR of Zinc (Flame Reactor)
             of High and Low Zinc Subcategory K061
             Nonwastewaters	

Table 4-9     Performance Data for HTMR of Zinc (Flame Reactor)
             of K061 Nonwastewaters	,

Table 4-10   Performance Data for HTMR (Rotary Hearth Furnace/
             Electric Furnace) of K061, K062, F006, and
             Characteristic Wastes Containing Metals	

Table 4-11   Performance Data for HTMR (Rotary Hearth Furnace/
             Electric Furnace) of K061, K062, F006, and
             Characteristic Wastes Containing Metals	

Table 4-12   Performance Data for HTMR (Rotary Hearth Furnace/
             Electric Furnace) of K061, K062, F006, and
             Characteristic Wastes Containing Metals	

Table 4-13   Performance Data for HTMR (Rotary Hearth Furnace/
             Electric Furnace) of K061, K062, F006, and
             Characteristic Wastes Containing Metals	,

Table 4-14   Performance Data for HTMR of Zinc (Molten Slag
             Reactor System) of High Zinc Subcategory K061
             Nonwastewaters	
4-15


4-16


4-17



4-18


4-20



4-21



4-22



4-23



4-24



4-25

-------
                                LIST OF TABLES
Table 4-IS   Performance Data for HTMR of Zinc (Plasma Arc
            Furnace) of High Zinc Subcategory K061
            Nonwastewaters	.'	

Table 4-16   Performance Data for HTMR of Zinc (Flame Reactor)
            of High Zinc Subcategory K061 Nonwastewaters  ....

Table 4-17   Performance Data for HTMR of Zinc (Plasma Reactor)
            of High Zinc Subcategory K061 Nonwastewaters

Table 4-18   Performance Data for HTMR of Zinc (Plasma Arc
            Reactor) of Low Zinc Subcategory K061
            Nonwastewaters	

Table 5-1    Possible Sources of Metals in Scrap Used in
            Steelmaking	

Table 5-2    Chemical and Physical Properties for 13 Metals
            in HTMR Processes	

Table 6-1    Corrected Data Used in  the Calculation of Treatment
            Standards Based on Performance of HTMR	

Table 6-2    Calculation of Treatment Standards Based on HTMR
            Performance Data from  SKF Plasma Technologies . . .

Table 6-3    Calculation of Treatment Standards Based on HTMR
            Performance Data from  International Mill Service . . .

Table 6-4    Calculation of Treatment Standards Based on HTMR
            Performance Data from  International Metals
            Reclamation Company  	
Page



4-26


4-27


4-28



4-29


  5-4


  5-6


  6-4


  6-7


6-10



6-12
                                        VI

-------
                                 LIST OF TABLES
                                                                            Page
Table 6-5     Calculation of Treatment Standards Based on HTMR
             Performance Data from Horsehead Resource Development
             Company	

Table 6-6     Calculation of Treatment Standards for Cyanide in
             F006 Based on Incineration Treatment Performance
             Data for Cyanide in K048 and K052 Nonwastewaters  . . .

Table 6-7     Calculation of Treatment Standards for Metals from
             the Four Sets of HTMR Performance Data	

Table 7-1     EPACML-Derived Dilution Attenuation Factors for
             Landfills	

Table 7-2     Data Sources for Subtitle D Landfill Survey	

Table 7-3     Generic Exclusion Levels for K061, K062,  and F006
             HTMR Residues (Nonwastewaters)  	

Table 7-4     Health-Based Levels (HBL) and Maximum Contaminant
             Levels (MCL) for Constituents of Concern	

Table 7-5     Treatment Performance Data (TCLP Analysis) for Residues
             (Dross), i.e., Oxidized Zinc Material (OZM) from
             HTMR of K061 High Zinc Subcategory Nonwastewater . .
6-15



6-17


6-19


 7-4

 7-7


 7-8


7-13



7-26
                                        vu

-------
                        1. INTRODUCTION AND SUMMARY

       This background document is an addendum for K061, K062, and F006 nonwastewaters*
The  purpose of this  background  document is to present the United  States Environmental
Protection Agency's (EPA's) rationale and the supporting technical information for removing
the existing  subcategories for  K061  nonwastewaters  (i.e.,  low  zinc and  high zinc)  and
establishing one set of treatment standards that will apply to all nonwastewater forms of K061.
This background document also presents the Agency's decision for establishing alternative
treatment standards for F006 and K062 nonwastewaters.  The treatment standards  for all
nonwastewater forms of K061  and  the alternative treatment standards for  F006 and K062
nonwastewaters are based on  a transfer of High Temperature Metals  Recovery  (HTMR)
treatment performance  used to develop  the final  treatment  standards for  K061 high  zinc
subcategory nonwastewaters.

       In the First Third Rule (August  1988), EPA established BDAT treatment standards for
the listed waste identified in Tide 40, Code of Federal Regulations, Part 261.32 as K061  i.e.,
emission control dust/sludge from the primary production of steel in electric furnaces.  In that
rulemaking, EPA established two subcategories for K061 nonwastewaters based on treatment
technology information, i.e., the high zinc subcategory (equal to or greater than IS percent  total
zinc concentration) and  the low zinc subcategory (less than IS percent total zinc concentration).
The  Agency determined that zinc could be recovered on a routine basis  from K061 wastes
containing equal to or  greater than  IS percent total zinc using a process identified  as High
Temperature Metals Recovery (HTMR).  Hence, EPA determined that 15 percent represented
    *  For the purpose of the Land Disposal Restrictions, nonwastewaters are defined as those wastes containing
       greater than 1 percent (weight basis) filterable solids or greater than 1 percent (weight basis) total organic
       carbon (TOC). Wastes not meeting this definition are classified as wastewaters.  (Treatment standards for
       K061  wastewaters are presented in the Third Third Rule 55 FR 22599 and  in the May  1990 K061
       Addendum, which can be found in the administrative record for that rulemaking. Treatment standards for
       F006 nonwastewaters are presented in the First Third Rule 53 FR 31153 and in the August 1988 Final
       BDAT Background Document for F006.  Treatment standards for K062 are presented in the First Third
       Rule 53 FR 31165 and in the Final BDAT Background Document for K062.
 25254108.01\secl                             1-1

-------
a reasonable cutoff concentration for routine recovery of zinc.  The Agency determined HTMR
to be Best  Demonstrated  Available Technology  (BDAT) for  K061 high zinc subcategory
nonwastewaters and stabilization to be BDAT for K061 low zinc subcategory nonwastewaters.

      Based on recent information,  EPA believes that the cutoff level based on zinc content
makes little technical sense, and that HTMR is the "best" treatment technology for both high and
low K061 nonwastewaters subcategories. This determination is based on the fact that HTMR
decreases the amount of material sent for  land disposal, recovers valuable resources, and
incorporates metals that are not recovered into  an extremely stable slag matrix.  Furthermore,
data for HTMR of low zinc subcategory K061 nonwastewaters  show that HTMR of low zinc
K061 achieves the same level of treatment performance for the slag residuals as HTMR of high
zinc subcategory K061 nonwastewaters. Hence, the Agency is promulgating HTMR as BDAT
.for all K061 nonwastewaters.  Because the Agency believes HTMR to be BDAT for all K061
nonwastewaters, EPA is transferring the performance of HTMR  of high zinc subcategory K061
nonwastewaters to low zinc subcategory K061  nonwastewaters.  EPA notes, however, that in
transferring the performance of HTMR to low zinc K061 nonwastewaters, it is not requiring that
this technology be used; rather, any technology that can meet the revised treatment standards can
be used, including stabilization.

       The Agency is also promulgating alternative treatment  standards based on HTMR as
BDAT for F006 and K062 nonwastewaters and is placing them in a new regulatory section for
alternative standards, i.e., 40  CFR 268.45.   These alternative standards are being promulgated
in order to achieve the same goal of treatment using BDAT,  but generally are designed to
provide alternative means  of  compliance with  the promulgated standards. The  Agency is not
promulgating these treatment  standards as a replacement of the existing standards for F006 and
K062 wastes, but rather as alternatives since it is not known if all F006 and K062 wastes are
amenable to metals recovery,  and the Agency has not been able to define the universe of those
wastes that are recoverable.  By developing treatment standards and generic exclusion levels
 (shown in Tables 1-1 and 1-2) based on HTMR as alternative BDAT, EPA hopes to encourage

 25254108.0I\secl                           1-2

-------
recovery of metals from F006 and K062 wastes .that are amenable to recovery technologies.
F006 is  listed in 40 CFR Part 261  as  wastewater treatment  sludges  from electroplating
operations.  K062 is listed as spent pickle liquor generated by  steel finishing operations at
facilities within the iron and steel industry (SIC codes 331 and 332).  Treatment standards for
both wastes codes were originally established in the First Third rulemaking.  (See £E at 31152
and 31164, August 17, 1988.)

       In this final rule,  the Agency is not promulgating a BDAT treatment standard  nor a
generic exclusion level for vanadium.  This decision  is based on the fact that the Agency
determined that its data base for vanadium is insufficient to fully characterize the performance
of HTMR processes for vanadium.

       The Agency is promulgating generic  exclusion  levels for F006 and K062 HTMR
nonwastewater residues. The generic exclusion levels include all the 40 CFR Appendix VHI and
indicator metals that might reasonably be expected to be present in the HTMR nonwastewater
residues  from processing F006, K061, and K062 wastes  by  HTMR.  (This is consistent with
RCRA section 3001(f) requiring EPA to evaluate whether toxic constituents in addition to those
for which a  waste is listed could make a waste hazardous.)  A detailed discussion of the
Agency's rationale for establishing the generic exclusion levels is presented in Section 7 of this
document.
252S4108.01\aecl                             1-3

-------
                   Table 1-1 BOAT Treatment Standards for K061
                 (All Nonwastewater Forms), and Alternative Treatment
                     Standards for K062 and F006 Nonwastewaters
Regulated constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium (Total)
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Zinc

Cyanide* (total)
Maximum for any
sinBlfi_Cpmposite sample
TCLP
(mg/1)
2.1
0.055
7.6
0.014
0.19
0.33
0.37
0.0090
5.0
0.16
0.30
0.078
5.3
Maximum for any
single composite sample
Total Concentration
(mg/kg)
1.8
aThe treatment standard for cyanide applies only to F006 nonwastewaters.
25254108.0 l\secl
1-4

-------
                    Table 1-2  Generic Exclusion Levels for K061

                    (All Nonwastewater Forms), K062, and F006

                         HTMR Residues (Nonwastewaters)
                                                     Maximum for any

                                                   Dingle composite sample

                                                        TCLP (mg/1)
Regulated constituent
 Chromium (total)
 ^^^^^•««

 Lead
 Mercury
 M^^^VHMM


 Nickel
 Selenium
 ^^^•••^"

 Silver
 Thallium
 ^^^^••0

 Zinc
                                                       Maximum for any

                                                          composite sample
                                                      Total Concentration

                                                           (mg/kg)
  Cyanide8 (total)
 s^^^^sss^^^s^s


*The exclusion level for cyanide applies only to F006 nonwastewaters.
 25254108.01\secl
                                          1-5

-------
          2. INDUSTRY AFFECTED AND WASTE CHARACTERIZATION

2.1          Industry Affected

       K061 waste is generated by the iron and steel industry and is defined as emission control
dust/sludge from the primary production of steel in electric furnaces. The Agency has no new
data that would change the description of the iron and steel industry contained in the Final Best
Demonstrated Available Technology (BDAT) Background Document for K061 (USEPA 1988a).

       F006 waste is generated  as  the  wastewater  treatment sludges from the  following
processes:  (1) common and precious  metals electroplating, except tin, zinc (segregated basis) ,
aluminum, and zinc-aluminum  plating on  carbon steel; (2) anodizing, except sulfuric acid
anodizing of aluminum; (3) chemical etching and milling, except when performed on aluminum;
and (4) cleaning and stripping, except when associated with tin, zinc, and aluminum plating on
carbon steel. Additional information  on industries affected pertaining to F006 may be found in
the final BDAT Background Document for F006 waste. The Agency has no new data that would
change the description of industries generating F006 wastes other than that presented in the final
BDAT Background Document for F006.

       K062 waste is a product of the steel industry's steel finishing operations and is defined
as the spent pickle liquor generated from these operations. Further details on industries affected
with respect to K062 is contained in the Final BDAT Background Document for K062, August
1988. Additional information on industries affected pertaining to K062 may be found in the final
BDAT Background Document for K062 wastes. The Agency has no new data that would change
    *  'Zinc plating (segregated basis)' refers to noncyanidic zinc plating processes.  For example, wastewater
       treatment sludges from zinc plating using baths formulated from zinc oxide and/or sodium hydroxide would
       be excluded from the listing while sludges from baths from zinc cyanide and/or sodium cyanide would not
       be excluded. Where both cyanidic and noncyanidic baths are used, the exclusion applies to sludges from
       the noncyanidic processes as long as they are segregated from sludges that result from cyanidic plating
 25254108.0 I\sec2                            2-1

-------
  the description of the industry generating K062 wastes other than that presented in the final
  BDAT Background Document for K062.
  2-2          Waste
        Waste characterization data (showing the ranges of the concentrations for the untreated
  wastes) for nonwastewater forms of K061, F006, and K062 wastes are presented in Tables 2-1
  through 2-3.  Additional pertinent information with respect to data values for the untreated waste
  and treatment performance data are presented in the Data Document for Characterization and
  HTMR Treatment Performance Data for K061, K062, and F006 Nonwastewaters.  This data
  document may be found in the Administrative Record for this rulemaking.
25254108.01\sec2                            2-2

-------
                     Table 2-1  BOAT List Constituents Composition for Untreated K061 Nonwastewaters

BOAT List
Constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
Cyanide
(«
Total
Concentration
(mg/kg)
5.03-294
10.2-400
24-400
<0.5-8.08
1.35-4,988
< 0.05- 106,000
1.29-139,000
0.0002-41
< 10-22,000
0.068-600
2.5-71
0.75-50
24-475
3,900-320,000
-
i)
TCLP
(mg/l)
.
<0.010-<0.095
0.130-1.58
-
<0.015-33.2
<0.007-9.05
< 0.300-6 1.2
<0.002-0.0047
•*
<0.005-0.193
0.021-<0.025
-
-
-
-
(t
Total
Concentration*
(mg/kg)
-
14-120
<0.01-690
-
13-17,900
3,300-156.300
1,600-46,600
<4-35
1,000-113.700
9.3-260
<2-150


3,200-405.100
0.5-2.3
>)
TCLP
(mg/l)
-
-
- -
-
<0.1-17.1
< 0.1-99.5
0.2-76
" ,
0.2-45
-
-
-
-
3.2-1,475
-
(c
Total
Concentration
(mg/kg)
284-298
24-35.4
< 12-33.2
< 10-0.7
213-231
14,900-17,500
10,300-14.500
1.8-5.04
11.600-15,400
20.6-27.7
158-192
• <3-1.8
389-656
46,600-62,000
0.8-1.67
)
TCLP
(mg/l)
2.31-3.18
< 0.004- < 0.02
0.019-0.433
0.017-0.029
2.7-0.122
122-183
.0-1
0.0129-0.0279
0.89-11.3
0.154-0.242
0.046-0.099
< 0.006
2.7-5.95
2.43-221
-
-no data.
Source (a)
Source (b)
Source (c)
Final BOAT Background Document for K061, August 1988 (represents 17 data sources).
INMETCO characterization data for K061 waste treated in 1988.
INMETCO characterization data for the June 1991 HTMR test for K061 and other metal-bearing waste streams.

-------
                                                       Table 2-1  (continued)

BOAT List
Constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium

T ntl ^O IHI WH
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
(
Total
Concentration
(mg/kg)
-
-
-
-
800-1,800

25.000-102,600
6.300*48,000

9,200-24,900
-


.
i
83,400-311,300
•I)
TCLP
(mg/I) j
-
-
-
1


.
•
-
_
•
i.
_
_
-
(«
Total
Concentration
(mg/kg)
90-1,010
20-2,510
200-470
-
450-2000

1,90049,500
10,300-28,500
.
300-15,000
-
70-180 .
-
100-830
47,700-205,000 j
>)
TCLP
(mg/l)

-
-
-


-
-
-
-
'
.
-
-
-
-no data.
Source (d)
Source (e)
1NMETCO characterization data (March 4, 1991) for K061 nonwastewaters for the years 1989, 1990. and 1991.
Horsehead Resource Development Co.,  Inc. (HRD) submitted to EPA during the comment period for the proposed
nilemaking for K061 high zinc subcategory nonwastewaters.  (These data show the lowest and highest averages of the
concentration compiled from the HRD data listed in five tables.)

-------
                                                       Table 2-1  (continued)

BOAT List
Constituent
Antimony
Arsenic




Lead
Mercury
Nickel

Silver
Thallium

Zinc
(1
Total
Concentration
(rag/kg)
.
.
-
.
1.100
.
22,500
_

.
.
-
-
138.400
)
TCLP
(mg/1)
-
-


•
-
-
-
-
-
-
-
-
-
(g)
Total !
Concentration j TCLP
(mg/kg) j (mg/1)
-
9-80
40-510
-
100-1.600
380-6,800
6,500-72,500
1.-29
20-600
0-17
10-168
-
-
50,000-508,000 )
> no data.
Source (0      International Mill Service (IMS) data dated February 2, 1990.
Source (g)      Report for the Center for Metals Production by Honehead Resource Development Co., Inc. (HRD), August 1988.

-------
                       Table 2-2  BOAT List Constituents Composition for Untreated F006 Nonwastewaters

BOAT List
Constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium •
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
Cyanide (total)
(
Total
Concentration
(mg/kg)
< 10-22.4
<0.4-S
0.74-85.5
<0.1-<97.6
0.003-22,000
<0.002-290,000
< 0.001-24.500
<0.2-<1
0.06-170,000
<0.03-<23
0.51-38.9 i
<10-<20 !
1.26 i
8.86-90,200
<0.02S-1,970
a)
TCLP
(mg/I)
-
-
-
-
-
-
-
_
-
•

-
•
-
-
(
Total
Concentration
(ngfcg) .
<6-31
3-30
< 1-9.4
•
-
1,500-136,000
500-2,900
< 0.04-0.3
53,400416,000 i
' 
-------
                      Table 2-3  BOAT List Constituents Composition for Untreated K062 Nonwastewaters

BOAT List
Constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
Cyanide (total)
(a)
Total
Concentration TCLP
(mg/kg) (mg/l)
.
<0. 1-3
< 10
-
<5
2-12,400
0.12-1.550
.
4-100,310
-
-
-
-
<0.4-9
-
(b)
Total
Concentration TCLP
(mg/kg) • (mg/l)
-
<2-184
<0.01-<100
-
<2-4.7 0.01-0.07
45,500-171,000 0.6-47.9
500-5,900 0.1-0.3
-
26,600-85,000 1.9-669
-
-
-
-
190-22,000 0.1-3.7
0.7-1.7 J
(c
Total
Concentration
(mg/kg)
423-1, 170 ,
16-22.1
< 16-236
<0.3-<13
11-11.8
3,570-19,400
196-200
0.79-1.3
9,790-153,000
<2-7.9
5.4-6.8
<4
149-169
572-401
3.4-7.73
)
TCLP
(mg/l)
0.08-0.41
< 0.004- < 0.008
< 0.098-0. 112
< 0.005
0.018-1.14
0.541-23
0-0.12 .
< 0.0002-0.0003
1-10.5
<0.003-<0.015
0.022-0.024
< 0.006
< 0.01 1-0.29
0.854-1.08
.
-no data.
Source (a)
Source (b)
Source (c)
Final BOAT Background Document for K062, August 1988.
INMETCO characterization data for K062 waste treated in 1988.
INMETCO characterization data for the June 1991 HTMR test for K062 and other metal-bearing waste streams.

-------
 3.  APPLICABLE AND DEMONSTRATED METALS RECOVERY TECHNOLOGIES
                AND IDENTIFICATION OF BEST DEMONSTRATED
                            AVAILABLE TECHNOLOGY
       This section presents the discussion of the applicable and demonstrated metals recovery
technologies for K061. These technologies are also applicable to some forms of F006 and K062
nonwastewaters.  (The Agency is promulgating alternative treatment standards as opposed to
revised standards because it is not known if all F006 and K062 nonwastewaters are amenable
to metals recovery.) This section also discusses the Agency's determination of HTMR as BDAT
for K061 nonwastewaters (all nonwastewater forms) and EPA's decision to promulgate HTMR
as an alternative BDAT for F006 and K062 nonwastewaters.

3.1    Applicable Recovery Technologies

       The metals recovery technologies applicable for treatment of K061 nonwastewaters and
some forms of F006 and K062 nonwastewaters are those that reduce the concentration of the
BDAT list metal constituents* in the treated residuals and which can generate residuals that are
resistant to leaching of metals.  The following recovery processes have been determined to be
applicable for recovery of metal constituents present in K061 nonwastewaters and in some forms
of F006 and K062 nonwastewaters.
       Although copper is not one of the metals being regulated in K061, K062, or F006 wastes, the Agency notes
       that there are numerous treatment processes available for the recovery of copper from metal-bearing waste
       streams (e.g., copper waste streams generated from operations such as electroplating and circuit board
       manufacturing). HTMR is one technology that has been used to recover copper from metal-bearing waste
       streams. Information available to the Agency indirattn mat the St. Joe Company's HTMR flame reactor
       process (now operated by Horsehead Resource Development Company, Inc.) has been successful at
       recovering a salable copper-aickel-cobalt alloy from one of its internal waste streams. For copper-bearing
       wastewater streams, there are numerous conventional treatment technologies that include the following:
       evaporation, electrowinning, electrodialysis, reverse «gmo«i«, and ion exchange. All of these technologies
       generally operate on the basic principle  that they concentrate the dragged-out plating solution (from
       electroplating waste streams, for example) from the rinse water to a degree that the solution can be returned
       to the plating bath.
25254108.01\aec3                             3-1

-------
3.1.1        Recycling

       Recycling is applicable to K061 nonwastewaters.  Recycling of K061 directly back into
the electric furnace from which it was originally produced facilitates the recovery of the metals
for steelmaking while reducing or eliminating the material to be land disposed.

3.1.2        Pyrometallurgical Recovery Processes

       The pyrometallurgical processes discwwd are generally relative to their applicability to
K061 nonwastewaters.  These processes are also applicable to some forms of F006 and K062
metal-bearing nonwastewaters.  Pyrometallurgical recovery processes  employ physical and
chemical reactions at elevated temperatures for the extraction/separation of metals from ores and
other materials.  Pyrometallurgical recovery process axe referred to as High Temperature Metals
Recovery (HTMR) for purposes of the land disposal restrictions program.

       HTMR is similar to recycling in that it is used to recover metals from K061, F006, and
K062 nonwastewaters for reuse, and it reduces the concentration, teachability, and volume of
waste to be land  disposed.  Some examples of HTMR systems include rotary kilns, flame
reactors,  electric  furnaces, plasma arc  furnaces, slag  reactors, and  rotary hearth/electric
furnaces.  The following are descriptions of specific HTMR processes that may be  used to
recover metal constituents from K061, F006, and K062 nonwastewaters.
          Daw McKe    '
          The Davy McKee Hi-Plas Furnace is a unique design that features a proprietary
          sleeve reactor that surrounds a long direct current (dc) transferred arc. In the Davy
          McKee Hi-Plas furnace, an argon stabilized dc transferred arc is struck between the
          nonconsumable water-cooled cathode gun and the molten pool of reactants in the
          furnace hearth in contact with the anode in the bottom of the furnace.  The plasma
          column passes down a water-cooled reaction sleeve or cyclone reactor that surrounds
          a substantial length of the arc column.  Reactants are injected  tangentially into the
 25254108.01\MC3                            3-2

-------
         sleeve at a number of points above the plasma gun tip and with a sufficiently high
         velocity to form a uniform covering to the inner wall of the reaction sleeve. Radiant
         and convected heat from the arc melts the reactants to form a film of molten material
         flowing down the inner wall of the sleeve. The molten material then drops into the
         furnace hearth region where the reaction is completed in the  molten bath heated by
         the impingement of the plasma column, which is used in the melting and smelting
         reactions.  Radiation to the furnace walls and roof is minimized.  The process is
         controlled by adjusting the arc power and the material feed rate.

         Since the plasma torch is remotely located from the hot furnace reactants, the torch
         is not exposed to damage by splashing and hot fumes.   The resulting long arc is
         stabilized by the vortex action of the  injection and  reaction gases  in the sleeve
         reactor.   Arc instability caused by the turbulent gases in -the smelting furnace is
         thereby avoided.

         In the process, the electric arc furnace (EAF) dust is mixed with coke and flux before
         being pneumatically fed into the furnace where smelting  occurs.  The metal oxide
         content of the dust is reduced at the melting temperature of about 2912°F (1600°C).
         A slag and  the nonvolatile metals are recovered in the furnace hearth and are
         intermittently tapped from the furnace.  The metal is recycled to the EAF,  and the
         slag is disposed of in the same way as EAF slag.

         Volatile metals such as zinc and lead  are recovered from the furnace gas in a zinc
         splash condenser (a  refractory-lined box containing  a pool of molten zinc).  In the
         condenser, the impeller of a vertical rotor is immersed in the zinc and when rotated
         creates a spray of fine zinc droplets.  Zinc and lead vapors passing through the
         chamber  condense on the -fine droplets.  Proper temperature of the zinc bath is
         maintained via an immersed  water-cooled coil. The remaining gases  leave the
         condenser and are combusted and cooled with excess air prior to being cleaned in a
         baghouse.

         If an EAF scrubber slurry is being processed, it is dried prior to treatment in the
         furnace.  In this case, the slurry is normally dried using the furnace gas from the Hi-
         Plas Furnace.  Paniculate carryover from the furnace to the zinc condenser reduces
         both the condenser's efficiency and the recovery of zinc and  lead: Minimizing dust
         carryover is therefore  very important.  The advantage of the Hi-Plas Furnace over
         other systems is said to be the use of the sleeve reactor.  Because the dust enters at
         the top of the sleeve, it is forced  to the energy from the arc.  Injecting the dust
         tangentially into the sleeve forces the dust to the inner wall where it will stick to the
         molten product running down the walls. This molten product drops off the bottom
         of the sleeve to form a felling curtain of material around the arc.  The process gases
25254108.01\Mc3                             3-3

-------
         have to pass through the curtain to exit the furnace, and in doing so there is a further
         opportunity for dust removal.
                  Electric Furnace Process

         Elkem's EAF dust processing basically consists of four principal areas that are
         interconnected to form  a complete .stand-alone plant.   These areas  are (1) feed
         receiving storage and preparation, (2) thermal processing, (3) zinc condensing and
         casting, and (4) off-gas treatment and recycling.

         In the Elkem process, EAF dust collected from Elkem's steel mill is prepared as
         briquettes for feed to Elkem's "Multi-Purpose Furnace.*  Elkem's Multi-Purpose
         Furnace is equipped with a self-replacing  frozen slag liner  which is the central
         component of the EAF  dust processing system.  The furnace is an airtight, three-
         electrode, slag resistance furnace, circular in design and equipped with an automatic
         slag-tapping flow control module. A short off-gas outlet flue is provided in the upper
         part of the furnace body immediately beneath the roof.  Furnace off-gases, including
         metallic fumes, exit the  furnace through the flue and enter the zinc condenser.  The
         furnace  operating  temperatures  are as follows:  slag 2642°F (1450°C), off-gas
         2012°F (1100°C), and molten metal 2732°F (1500°Q after carburization.

         The major thermal movement within the bath is an upward direction from  the
         electrodes, across the upper surface of the melt to the frozen slag side walls, and
         downward to the metal accumulation in  the furnace hearth.   This movement
         constantly melts and erodes the bottom of the feed layer into the slag and provides
         an active slag homogenizing action through thermal stirring, ensuring that nonvolatile
         elements remain encapsulated in the slag in a silica matrix.  The furnace off-gas
         contains virtually no dust carryover from the raw feed rwfgrial  Reacted components
         of the off-gas are volatilized metallic fume (zinc, lead, and cadmium), together with
         carbon monoxide and carbon dioxide. Small* quantities of halides are also in the
         fume, mainly potassium chloride and sodium chloride.

         The  next  principal area of the  Elkem  process involves  zinc  condensing and
         processing. Off-gas from the furnace operating at a temperature of 2012°F (1100°C)
         enters the condenser, where zinc  is condensed.  Lead volatilized in the furnace is also
         condensed in the zinc spray. It leaves the condenser, together with the molten zinc,
         through an underflow at the end of the condenser  and enters a  skimming sump.
         Periodically, zinc is transferred from the sump to a "liquation" bath. In the liquation
         bath, lead separates from the zinc into a lower layer.  The zinc then overflows into
         a casting bath, where its temperature is increased to 932°F (500°C), which is suitable
          for tapping and casting into shapes as may be required for sale or in-plant use. Lead
25254108.01\aec3                             3-4

-------
          accumulating in the liquation bath is pumped out periodically and cast into blocks for
          sale.  Any iron entering the condenser dissolves in the zinc and is separated in the
          liquation as an intermediate alloy, "hard metal." It is periodically removed manually
          and recycled backed through the plant feed systems.

          The next principal area of the Elkem process is the handling of the off-gas.  Any
          uncaptured  zinc, along with other components of the gas, leaves the condenser and
          enters the gas washing tower. Here most particulates are removed and collected in
          a sludge, which together with water is pumped into a clarifier.  Gases leaving the
          washing tower, mostly carbon monoxide and carbon dioxide with minor quantities of
          remaining paniculate, go to a Venturi scrubber.  The  gaseous discharge from the
          Venturi scrubber is then sent to a demister, where water reclaimed from the clarifier
          overflow is used for the second stage of scrubbing.  The resultant Venturi scrubber
          and demister water is collected, and the captured particulates are pumped back to the
          gas' washing tower.  Remaining gases are passed through  a thermal oxidizer to
          convert the carbon monoxide to carbon dioxide, then to a quench tower for cooling,
          and then to  a baghouse. Liquid effluent from the quench tower is sent to the clarifier
          overflow sump and pumped back to the clarifier. A bleed stream from the clarifier
          overflow is periodically pumped to a water treatment facility before final discharge.

          The clarifier underflow,  containing mostly zinc and lead oxides, is filtered, and the
          filter cake is periodically returned to the plant feed system for drying and recycling.
          Any liquids are returned to the clarifier.
          Enviroscience Company

          Enviroscience Company uses an  Cupola furnace in its high temperature metals
          recovery process.  The Cupola furnace operates at 2800°F and produces a fully
          'molten slag. Their HTMR process consist of blending liquids; semi solid; and solid
          F, D and or K series wastes to produce a chemical formation that can be smelted into
          metal alloys, metal oxides, and mineral wool of commercial value. The mineral wool
          can be used as a substitute for fiber glass.  Also,  zinc oxide is collected from the
          baghouse and sent to a refiner.  Additional information on the Enviroscience HTMR
          process is included in the Administrative Record.

          Horsehgfld  Resource Development Company. Inc. (HRD^ Series of Wael7f ffilns
          Process

          The waelz kiln is a type of rotary kiln. The term waelz is derived from the German
          word  "waelzan,"  which means to trundle or roll-an accurate description of the
          movement of the charges through the rotating kiln.  A series of waelz kilns is used
25254108.01\wc3                            3-5

-------
           by Horsehead Resource Development Company, Inc. (HRD) to process electric arc
           furnace dust (averaging 18 to 19 percent zinc).  Waelzing of the  K061 dust is
           accomplished by mixing coal and appropriate amounts of limestone and silica "fluxes"
           to maintain the desired kiln conditions.

           In the waelz kiln process, the first kiln is heated to 2372°F (1300°C). Excess carbon
           in the kiln bed ensures that 95 percent of the zinc is reduced to metal and volatilized
           along with substantial amounts of the cadmium and lead present.  The volatilized
           metals are reoxidized in the gas stream above the bed and collected as crude oxide
           in a baghouse.  (These materials also are stored in open piles before being calcined.)
           The iron-rich residual materials (IRM) are discharged continuously from the opposite
           end of the first kiln as granular, porous, nonvitrified slag. HRD then further purifies
           the  crude zinc oxide by processing it in a calcining kiln to selectively volatilize
           cadmium, lead, chlorine, and fluorine, .separating them from the zinc oxide.  The
           calcining kilns used by HRD can be physically identical to the waelz  kilns.  When
           a kiln  is operated as a calcining loin, it is fed directly with the crude zinc  oxide
           without coal or flux additions.  No reduction takes place in the second (calcining)
           kiln, and all heat is provided by a natural gas or oil burner to heat the  kiln to  1292-
           1832°F (700-1000°Q.  The lead and cadmium are volatilized as oxides,  sulfides,
           sulfates, and/or chlorides in an oxidizing atmosphere.  Apparently, much of the zinc
           chloride  present or  formed in the kiln reacts  with lead  and caH"him oxides,
           contributing to the efficiency of die refining step. If a purer zinc oxide product is
           required, the temperature and/or residence time of the kiln can be increased.  This
           calcined material is then sent to Zinc Corporation of America (ZCA) to be  smelted.
           It is  normally not capable of being used as a product as is,  but rather  requires
           smelting to recover usable zinc.

           The volatilized product of the calcining kiln, often referred to as the lead or lead
           cadmium concentrate, is collected in a baghouse.  This material is carefully bagged
           in containers and is currently sent to an HRD facility in Bartlesville, Oklahoma, for
           processing and recovery of both the lead and cadmium values.


           Horsehead Resource Development Company. Inc. fHRDI Flame Reactor Process

           Horsehead Resource Development Company,  Inc. (HRD) uses a flame reactor
           process that is  a  two-stage, carbon-fueled, flash-smelting system that efficiently
           recovers zinc, lead,  and cadmium as a recyclable crude oxide and produces a slag
           that HRD states is nonhazardous.  In summary, the process generally involves the
           following steps.
25254108.01\Mc3                             3-6

-------
          Fine coal, coke breeze, or natural gas is pneumatically injected into a water-cooled
          burner (first stage of the reactor) and intensively mixed and reacted with oxygen-
          enriched air (40 to 70 percent oxygen) under fuel-rich conditions at a temperature
          greater than 3632°F (2000°C) to produce a hot reducing gas.  Then, the fine, dry
          metallurgical feed is pneumatically injected into the hot reducing gas stream in a
          second water-cooled stage to reduce zinc, lead, and cadmium compounds to metal
          vapors at 2912°F (1600°C) in a high-velocity particle-flame suspension.

          Lead- and cadmium-free molten slag flows along with the combustion gases through
          the reactor.  A portion of the slag freezes on the water-cooled reactor walls to form
          a protective layer on which the molten slag flows down into a horizontal gas/liquid
          separator, where it is  separated from the reactor gases.  The slag is men tapped,
          cooled, and sold as an iron-rich aggregate.  Zinc, lead, and cadmium are recovered
          as crude, heavy metal oxide, marketable as a feedstock to industry.

          According to the report from the Center for Metals Production for the "Flame
          Reactor Process for Electric Arc Furnace Dust" by HRD (August 1988), the crude
          zinc oxides produced from EAF dust processed in die flame reactor would probably
          not  meet  the feed specification for  the zinc smelter of Zinc Corporation of America
          (ZCA).  Thus, in the case of the flame reactor's zinc oxide, an upgrading step must
          be used to achieve the desired specifications for smelting. The report indicated that
          a caustic  leaching process is HRD's preferred method-for upgrading the crude zinc
          oxide.
          International Medals RcctoflUlfltiPP Company (INMETfTfl jpofafy Hearth Furnace
          Followed bv Electric Furnace

          INMETCO (FJlwood City, Pennsylvania) uses a rotary hearth furnace and an electric
          furnace in its HTMR processing of EAF dust (K061) and other metal-bearing waste
          streams. INMETCO uses these waste streams concentrated in nickel, chromium, and
          iron as feedstocks to produce chromium/nickel/iron remelt alloy which is used as a
          feedstock to produce stainless steel.  This process is most effective in recovering
          wastes that ordinarily contain 1.5 percent or greater chromium/nickel combination.

          In general, the pyrometallurgical process consists of the following four basic steps:
          (1) feed preparation, (2) reduction, (3) smelting, and (4)  metal casting.  The first step
          of the process involves blending determined quantities of nickel, chromium, and iron
          wastes with a reducing agent  (i.e., coke or coal) and additives that are used to
          develop  the slag matrix (i.e., lime and magnesia). All feed materials have been
          pretreated to ensure a uniform  size.  The mixture of feed materials is pelletized to
25254108.01\aec3                            3-7

-------
           produce pellets strong enough to resist disintegration in the subsequent thermal
           operations.

           The second step in the process involves partial reduction of the metal oxides in a
           rotary hearth furnace operating at 2300'F.  During this stage, portions of the zinc,
           cadmium, and lead volatilize  (approximately SO percent of those metals that are
           present in the feed) and are discharged to a baghouse. Next, the sintered pellets are
           melted in an electric  arc smelting furnace with an average metal temperature of
           2720°F and an average slag temperature of 2940°F.  Lime, silica, alumina, and
           magnesia separate  to form the liquid slag,  which floats on the surface of the arc
           furnace; the remaining zinc, cadmium, and  lead volatilize; the gas containing these
           volatilized materials is discharged to a wet scrubber.  The metal remaining after the
           slag is removed is poured into a refractory-lined ladle from which it is cast into
           "pigs,* which are sold to steel mills. Top wastes'generated from the treatment of the
           off-gas streams are sent  as  K061 hazardous waste (i.e.,  manifested as K061) to
           another facility to recover the zinc, lead, and cadmium.  The slag is air cooled as it
           is poured down a sand ramp and then is  used as road aggregate.
          International Mill SSTYI06 flMS) Plasni8 Fy,,,_--

          International Mill Service (IMS) Incorporated uses a plasma furnace to process K061
          EAF dust.  In the IMS process, the electric arc furnace dust (K061) feedstock is
          transported through a totally enclosed conveyance system from the IMS  electric
          furnace baghouse (from the production of steel).  The EAF dust is then blended with
          a coke breeze or coal fines reductant and metered into the plasma furnace vessel by
          two conveying screws.   The metallurgic endothermic  reactions occur at high
          temperatures. The energy required is supplied by a hollow, direct current graphite
          electrode.  Argon or nitrogen gas is blown through  the middle of the electrode to
          provide a path for electrical energy transfer, as well as for arc  stabilization.  The
          current passes from the electrode to the anodes fixed in the bottom of the furnace
          hearth.

          The resulting bath is operated within a temperature range of 2522° to 2732°F (1400°
          to 1500°C) to ensure that the reduction reaction occurs.  Within the metallurgical
          process, iron is only partially reduced in the furnace and will remain in the hearth in
          its oxide state as a slag.  In this temperature range, the zinc, cadmium, and lead will
          vaporize off the bath. All remaining elements remain in the furnace as a slag and are
          tapped on an  as-required basis.  The metal vapors, as well as the mix of emitted
          carbon monoxide and carbon dioxide gases, pass through a hole in the top of the
          furnace roof to the "zinc splash condenser' through a refractory-lined duct.
252S4108.01NMC3                            3-8

-------
          The metal  vapors and carbon monoxide/carbon dioxide gas mixture  enter  the
          condenser, where they are rapidly cooled to ensure condensation.  As the metallic
          vapors are condensed into their molten states, the zinc and lead reportedly exist as
          two separate liquid phases. Because of the higher density of the lead, it sinks to the
          bottom of the condenser into a "liquation" vessel, where controlled cooling occurs.
          The molten zinc is cooled and held at 968° to 851 °F to facilitate final separation of
          the zinc and lead for quality purposes.  The zinc is poured into molds, and the zinc-
          rich lead at the bottom of the vessel is cast into separate molds as a secondary
          product.

          The carbon monoxide and carbon dioxide gas mixture and any remaining metallic
          vapors enter an afterburner upon exiting the condenser main chamber.  They  are
          mixed with outside air, the remaining  metallic vapors are oxidized, and the carbon
          monoxide is burned to its stable state of carbon dioxide.  Dilution air is  added
          downstream of the carbon monoxide  burner to cool the gas mixture prior to its
          entering a baghouse.  The cooled  streams  from the afterburner (less than 270°F
          (132°C)) enter the baghouse.  The particulate metallic oxides formed in  the
          afterburner are collected in the baghouse and recycled back to the furnace for further
          recovery-of metallics.

          SKF Plasmadust Process

          The Plasmadust process was developed by SKF Plasma Technologies AB in Sweden
          for processing electric arc furnace baghouse dust (K061) and similar waste oxides
          from the steel industry. The dust is decomposed in the process, and the metals in it
          are recovered. According to SKF Plasma Technologies, the Plasmadust process  has
          the capability to effectively treat dusts from both carbon and alloy steelmaking.

          In general, in the SKF Plasmadust process, the dust is mixed with coal powder, sand,
          and internally recycled process wastes.• This mix  is then injected pneumatically into
          the lower part of a coke-filled shaft furnace provided with three plasma generators
          that supply  the thermal energy  needed to reduce the metal oxides in  the dust.
          Consequently, there are three separate reaction  zones in the coke column.   The
          injected material mix is heated by the plasma gas generated by the plasma generator,
          which converts the electrical energy  to highly concentrated thermal energy in the
          form of gas. The material mix is melted and reduced in the three reaction zones.
          Molten iron and slag drop down through the coke column and collect in the hearth.
          The iron is tapped at a temperature of about 2552°F (1400°C); the slag, which floats
          on the iron, is tapped at a somewhat higher temperature. The molten iron is cast into
          pigs or other shapes before shipping back to steel mills for resmelting. Zinc, lead,
          and cadmium from the original furnace feed dust leave the furnace as metal vapors
25254108.01\wc3                             3-9

-------
         in the off-gases. The slag is well reduced and has a very low content of volatile
         metals.
Sumitomo Molten Slag lector Pmce^




   23?S5SS»fc^2^!!i;K



      by an oil burner. Heavy metals are dissolved simultaneously with the
      ig and a modifier (Fukenite) and then fixed within the granulated slag  Less
      ictals become incorporated into the slap that i« fnrm~i  \*~^'^-
              metals become incorporated into the slag that is formed.  MorTvofctite
        ^pereent^61^ *" * baghouse' The zinc c00*0* o{ *« baghouse dust is around
        Ticron CnrnnraHnn
                          S??1*1 HaSma ^ ^^ !»«»« for processing EAF

        «»i-   n        ^ b«ring waste streams. Ticron states
        can be mstalled on-site and be an extension of the principal steel
                   -
        its process can yield metals of commercial grade.


        The Ticron process operates by employing a standard AC three phase electric arc
        steel making furnace which has been scaled up to a one mega waTuril  A

                              Whfch
                                    ' "* ^"^ te-P«     of 1500 to
                                          zone is more than a meter in
              and generates uniform heat transfer that is diffused throughout the furna
25254108.01\Mc3
                             3-10

-------
          ZIA Technologies

          The HTMR processing of K061 by ZIA Technologies is performed in an Inclined
          Rotary Reduction System (IRRS). In the HTMR JKRS process, electric arc furnace
          dust is mixed with a small amount of binder and pulverized coal to provide sufficient
          reductant (carbon) for the reduction of metal oxides. The resulting mixture is fed
          onto  a standard pelletizing disk to. produce 3/8-  to 1/2-inch pellets.  Before the
          mixture is fed  to the pelletizer,  the  dust, pulverized coal, and binder are all
          transported pneumatically or in screw conveyors to prevent dust from escaping to the
          atmosphere.  The "greenbalT pellets are next conveyed by belt directly to the IRRS
          furnace, which is essentially a specially designed kiln fired by an oxy-fuel burner.
          Under controlled conditions of temperature and  atmosphere, the pellets are heated,
          the moisture and volatile matter contained in the coal are driven off, and the pellet
          temperature is raised  to the level needed for the reduction reactions to occur.  The
          pellets are heated in the rotating IRRS furnace to a maximum temperature of 1150°F
          (2102°C).  (Exceeding 2102°F  would result in melting of the pellet surface and
          sticking of pellets into  lumps or  formation of  rings  in the  kiln.)   When the
          temperatures are kept below 2102°F, the formation of zinc chloride is prevented.
          Zinc chloride can be detrimental to the ultimate condensation of zinc metal.

          At approximately  1652°F (900°Q, reduction of the metallic oxides of lead, zinc,
          cadmium, and iron begins to occur.  At this temperature, the reduced zinc, lead, and
          cadmium are vaporized and carried as a metallic vapor in the off-gas stream.  The
          iron, contained as oxide in the pellet, is reduced to metallic form and remains in the
          pellet, together with various slag-forming compounds that came from the electric arc
          furnace as part of the dust The pellet is now a form of direct reduction iron (DRI)
          from which virtually all the "other metals have been removed, and which also contains
          a significant portion of the slagmaking materials needed to refine steel.  The  DRI
          pellets, which are self-fluxing, are discharged from the IRRS furnace directly into a
          water bath for quenching to a temperature below the  ignition  point to avoid
          reoxidation.   The  water  quench bath  also  acts as an atmospheric seal  enabling
          continuous withdrawal of the iron-containing pellets while avoiding the entry of large
          quantities of air to the zinc vapor-containing exhaust gas stream.

          ZIA states that the reduced iron pellets are then suitable for charging directly  back
          into an electric arc furnace to reclaim those iron units that were previously lost with
          the waste dust. Copper and sulfur impurities contained in the charged pellets remain
          with the pellets.

          The off-gases from the IRRS furnace are comprised of zinc, lead, and cadmium metal
          vapors, evaporated pellet moisture, volatile matter driven from the coal in the pellets,
          and the products of combustion.  The off-gases exit the IRRS furnace directly into
25254108.01\Mc3                            3-11

-------
        the afterburner and retort section of the system. This off-gas stream may also contain
        a small amount of product dust which results from minor pellet degradation caused
        by the rotary action of the furnace.   Air is admitted to the afterburner system to
        reoxidize the metal vapors and to combust any carbon monoxide gas and volatile
        matter from the coal. The amount of air is also adjusted to control the afterburner
        temperature to a maximum of 2012°F (1100°C). The hot gases circulate around the
        retort  and provide the energy needed to drive the secondary reduction process
        (described below).  The hot gases then proceed through a section of duct where they
        are cooled by radiation before entering cartridge-type filter units. It is in these filter
        units that the metal oxides and any product dust are collected  prior to discharge of
        the products of combustion to the atmosphere.  Cartridge filters were chosen over a
        more typical baghouse because of the inherently higher  efficiency.  Dilution air is
        also introduced into the off-gas stream following the retort section to further assist
        in reducing the waste-gas temperature.

        After  all of the volatile  metals  removed from the electric  arc furnace dust and
        concentrated as an enriched zinc oxide dust, the next stage is the production of Prime
        Western Grade zinc. Except for trace carryover amounts, any undesirable elements
         to this operation (chlorides and iron) are retainedin the fluxed iron pellets.   Rather
         than combining reduction and condensation in one process, two separate operations
         are employed.  By allowing the IRRS furnace to reduce the oxides and then collect
         the concentrated reoxidized metals, the concentrated material can then be repelletized
         with coal (carbon) as the reductant together with a binder. These pellets are then
         reduced in a vertical retort to form a concentrated metal vapor stream.  By using a
         vertical retort technology with its low waste gas volume, no iron or copper dust is
         carried over to contaminate the product from the condensation process. The off-gas
         stream from the vertical  retort is essentially pure zinc, lead, and cadmium vapors.
         To make this two-step technology economically viable, a retort was built inside the
         final stage of the afterburner where an excess of high temperature energy existed that
         had to be removed prior to admitting the IRRS furnace off-gas stream to the filters.
         To eliminate the need to preheat the oxide pellets and precoke them prior to entering
         the retort, a retort with two outlets was developed.  The upper outlet vents the
         moisture driven off from the "greenbaU' pellets, as well as the volatile matter
         evolved from the  coal during the drying and pre-heating stage.  These gases are
         ducted back into the afterburner, where they are combusted.

         A second outlet is provided at the point where pellet temperature inside the retort
         reaches approximately 1652°F (900°C). Metal vapors exit the retort from the second
         outlet and are ducted to a standard Imperial  Smelting Process (ISP) zinc splash
         condenser.   If some heavy metal vapors form above this second outlet, they will
         condense on the colder material  above and will be refluxed back down the retort to
         the higher  vaporizing temperature zone.  It is important that no water  vapor is
25254108.01\Mc3
                                         3-12

-------
          allowed to reach the vapor outlet because water vapor would permit the reoxidation
          of metal vapors and reduce the overall efficiency of condensation. Metal vapors that
          pass  through the splash condenser without condensing  are  ducted back  to  the
          afterburner for reoxidation and collection. Because of the closed-loop design, all zinc
          and lead eventually have to exit the process as a condensed material. The- residue
          remaining in the retort after reduction exits the bottom of the retort column, is cooled
          and then sent back to the beginning of the process  where it  is pulverized and
          reintroduced as part of the primary pellet feed. In this way, any leachable metals that
          were not reduced and removed in the vertical retort are circulated back through the
          system until 100 percent of the leachable metals are recovered. The remainder of the
          residue ends up as the slag constituent of the DRI pellets and returns to the electric
          arc furnace, where it is removed as furnace slag.  The slag component of the pellets
          is the same as the material used for making electric arc furnace slag.  It can be used
          to replace that amount of material normally added to the furnace so that furnace slag
          volume remains unchanged.
3.1.3        Hydrometalhirgical Recovery Processes


       Hydrometallurgical recovery processes  extract and recover metals from materials by
processes  in which solutions play a predominant role.   Some hydrometallurgical processes
include chemical precipitation, leaching, ion exchange, solvent extraction, and electrowinning.


       The Agency has limited information indicating that some facilities are using a series of

technologies, which include chemical precipitation, ion exchange, and electrowinning, to recover
from waste materials the same  metals present in K061 wastes.   These hydrometallurgical

technologies may also be applicable for recovering metals from other metal-bearing waste

streams such  as  F006  and  K062  wastes.    Some of these facilities  claim  that  these
hydrometallurgical  processes,  unlike other processes, could generate  no residues  for  land

disposal.  The Agency notes  that the  concentration-based treatment standards are based on

BDAT, i.e., HTMR;  however, non-HTMR recovery processes are not  precluded from being

used to achieve the concentration-based treatment standards, provided  the  standards are not

achieved through the use of impermissible dilution.
25254108.01\aec3                            3-13

-------
         Encvcle Texas
        Jn Encycle's hydrometallurgical process there are two major options for processing K061
        dust: (I) by a sulfuric acid-based process and (2) by a caustic leach process.

        Acid Process: In the acid-based process, the metals are solubilized at a low pH (near 1)
               are then removed from solution by chemical precipitation, crystallization, and/or
                     . For ^ampje, the iron is primarily present in the ferric state and can be
                             ftom soludon * a PH of about 3- Tte P^ty of this floe can be
                                                     The ferric hydroxide is then removed
                             separation techniques such as filtration.
        The recovered ferric hydroxide can then be used directly in such applications as paint
        pigments or it .can be made into a variety of products.  If further refining is reqiiiredfthe
        feme hydroxide can be solubilized  in sulfuric acid  and then recrystallized as ferric
        sulrate.   The crystallization is controlled  to improve product quality and  minimize
        impurities.  Alternatively,  the recovered  ferric hydroxide is blended with a reductant
                °J!    J!*^' .ind  fluxin8 *geots'  »«*« tone, and then  agglomerated, for
        The lead and zinc remaining in solution are then removed by chemical precipitation as
        LSJZ^ "^ i^"""- «»*•  <*>•) or as a carbonate with either sodium
        carbonate or carbon dioxide, as a sulfide  using sodium sulfide or hydrogen sulfide  or
        as a sulfate using an evaporator/crystallizer.                             sumac, or

        The nickel and/or chrome solubilized in the acid leach is then extracted.  Avarietyof
        technologies are available for this  processing.   Nickel is  removed by  standard
        commercially available, ion-exchange resins in which the solution is passed through these
        resins. The resins initially adsorb both zinc and nickel. However, as the resins become
        fuUy loaded, nickel crowds out the zinc. The resins are staged in sequential order based
        on the level of loading.  Once a resin is fully loaded, it is stripped of the nickel by a
        strong acid wash, usually sulfuric acid.  The resulting nickel solution can be processed
        to produce a wide variety of mckri producte mdudir« mckel siilfsto, lu'ckel metal  nickel
       hydroxide, and nickel chloride. Each of these end products produces various amounts
       of residuals depending on the purity of the product and process employed.

       The  chrome in solution  can be isolated from the other metals by oxidizing it  to the
       hexavalent state and then raising the pH of the solution to around 9 to remove the other
       metals present (via filtration).  The resulting hexavalent chrome can then be reduced to
       produce chrome hydroxide, a chrome-ore substitute, or it can be converted into a variety
       of chrome chemicals such as sodium dichromate.
25254108.01W3                            3.14

-------
      Caustic Leach Process:  The other basic technique for processing K061 dust is to perform
      a caustic leach on the dust to solubilize the zinc as a zincate. Zinc is removed from this
      solution as metal by electrowinning (or as a hydroxide by adjusting the pH downward
      to about 8.5 with an acid).

      The iron,  nickel, and chrome left behind in the solid is then processed by techniques
      similar to  those in the acid leach process.
             Recycling Technologies Corporation (MRTO Process

      This process involves placing K061  and potentially other metal-bearing waste systems
      into a heated ammonium chloride solution. The metallic oxides, with the exception of
      iron oxide and a small amount of sand, are soluble in the ammonium chloride solution.
      The iron and sand are filtered into a reusable cake form, which can be xesmelted by the
      steel manufacturers.  The filtrate solution  contains metallic oxides. Zinc dust is added
      to the filtrate solution, cementing out all metals except zinc oxide. The remaining zinc
      oxide is crystallized out and the filtrate liquor is reused in the next batch.

      Approximately 95 percent of the original ammonium chloride solution remains following
      the process  and is reused in the batch process.   The remaining ammonium chloride
      solution binds to the zinc oxide crystals and the iron  filter cake.  It is then removed by
      washing; MRTC expects that the resultant wash water can then be reused in the recycling
      process.
      Recontek

      In the first stage of Recontek's processing, industrial wastes are classified into four
      groups: acids, cyanides, alkalines, and solids.  The wastes are separated and stored in
      four different storage areas based on waste type.  Thermal decomposition is then used
      for the destruction of cyanide.  At this point, the treated cyanide wastes are combined
      with acid, alkaline, and solid waste streams to form a metal-rich solution.  The types and
      concentration of metals in the solution are controlled by mixing and matching the storage
      tanks from which the ingredients are selected. In the next stage, water is distilled from
      the solution to concentrate the precious and base metals, and the solution is then sent to
      the precious metals refinery for final purification into  gold, silver, platinum, and
      palladium.

      Next, any iron is precipitated from the solution and converted into ferrous sulfate for use
      in the water purification or fertilizer industries. An electrowin cell house is then used
      to recover copper and tin metal  from the solution. After electrowinning, the solution is
25254108.01\Mc3                            3-15

-------
        concentrated and nickel is crystallized as nickel sulfate. The nickel sulfate can then be
        reused by the plating industry.
        Chromium is then precipitated from the remaining solution by controlling both the pH
        and the oxidation-reduction potential (ORP) of the solution.  The precipitated chromium
        is filtered, washed, and dried* The chromium can then be sold as an additive for the
        stainless steel industry. A "zinc cementation' process is used on the remaining solution
        for extraction of cadmium, which can then be sold to cadmium refiners.  Zinc remains
        in solution and is then electrowinned in a manner similar to that used for copper.  The
        recovered zinc can be sent to zinc smelters for sale.  "Magnesium cementation" is used
        to recover trace metals, and the residual solution is converted back to usable acid and
        sodium hydroxide, which are reused in the process or sold.

 3.2    Demonstrated Treatment

        EPA considers a technology demonstrated if it is or has been used on a commercial basis.
 Based on available information, waelz kilns, flame reactors, plasma furnaces, electric furnaces,
 and  combination  rotary  hearth/electric  furnaces have  been  demonstrated for  K061
 nonwastewaters and for some F006 and K062 nonwastewater streams.

       With respect to the HTMR processes (i.e., pyrometallurgical processes), the Agency does
 not have sufficient treatment performance data from the other applicable HTMR processes (e.g.,
 molten slag  reactor, the  Enviroscience HTMR process, or the  Etkem pyrometallurgical
 processes) to determine whether they are effective for treating K061 nonwastewaters or F006
 and K062 nonwastewaters. Treatment performance data for the Davy McKee ffi-Plas Furnace
 for K061 nonwastewaters were submitted during the comment period (for the high zinc K061
 subcategory nonwastewaters rulemaking). However, since these data were only preliminary, a
 full assessment of this HTMR process could not be made.   The Agency has not received
 treatment performance data for the Inclined Rotary Reduction HTMR System; thus, it could not
be evaluated to determine whether it is demonstrated for HTMR of K061 nonwastewaters or
F006 and K062 nonwastewaters.
25254108.01\Mc3                           3.16

-------
      Regarding the hydrometallurgical processes, Recontek states that its hydrometallurgical
process was developed targeting waste streams F006, K061, K062, D002,  D008, D009,  and
D011.  However, the Agency notes limitations to the types of wastes that are accepted by
Recontek (albeit some were stated to be economical).  Encycle provided the Agency with
information on its process but did not provide any treatment performance data for its process
residuals in order for the Agency to evaluate process efficiency.  MRTC provided information
and treatment performance data for its hydrometallurgical process. The treatment performance
data (toxicity characteristic  leaching procedure (TCLP) leachate analysis) appear promising;
however, the Agency notes that data for some metals show high leachate levels with respect to
the BOAT treatment standards. The Agency notes that it has not received information showing
that this process has been demonstrated on a full-scale basis for treatment of K061, F006, or
K062 nonwastewaters.

3.3    Identification of Best Demonstrate^ Avft'lflMc Tfylrnol°gy fBDA'D

       In the First Third Rule, EPA determined HTMR to be BOAT for K061 nonwastewaters
containing  15 percent or greater zinc content.   The  Agency believes HTMR, rather  than
stabilization, to be the BDAT for low zinc K061 because HTMR decreases the amount of
material sent for land disposal, recovers valuable resources, and incorporates metals that axe not •
recovered into an extremely stable slag matrix.

       Data indicate that residuals for land disposal generated from HTMR units generally leach
concentrations of metals that are comparable to (and, for most metals, lower than) those residues
from stabilization of K061 wastes. These comparable teachability results for slag (i.e., residues
for land disposal) from HTMR processes are occurring in spite of the fact that HTMR slag
residues are generally more concentrated in toxic metals. As a result, it appears that the HTMR
processes are achieving stabilization of metals through chemical reactions with materials such
as lime and silica at elevated  temperatures.  This seems  to indicate  that an overall, more


25254108.01VMC3                            3-17

-------
effective stabilization is being achieved at the high temperatures.  In some instances, HTMR
residues may leach at higher levels than stabilized matrices, but this is typically related directly
to the high concentrations of metals in the influent materials.  For instance, performance data
for stabilization of F006 nonwastewaters (used  to develop nickel  and chromium  treatment
standards for low  zinc subcategory K061  nonwastewaters) represented  treatment  of less
concentrated  wastes than the HTMR data shown in Table 3-1 (i.e., representing inputs to the
INMETCO HTMR processes for 1988). The average concentration of nickel in F006 (untreated
stabilization data) was $,449 mg/kg compared to 180,400 mg/kg (untreated HTMR data for
INMETCO).

       The use of HTMR is also consistent with the national policy, identified in the Hazardous
and  Solid Waste Amendments (HSWA) of the Resource Conservation and Recovery Act
(RCRA), to  reduce the quantity of hazardous  constituents disposed.  Since  HTMR is  a
technology that recovers valuable constituents from waste materials, there is typically no increase
in the volume of the waste  residuals resulting from recovery treatment  For example, percent
metal recovery data for some HTMR processes show that HTMR processes can recover the
following:  99 percent of the nickel, 97 percent of the cadmium, 87 percent of the chromium,
86 percent of the  lead, and 99  percent of the iron.   This is in contrast to nonrecovery
technologies  such as stabilization,  which is not intended to reduce the total metal concentration
or waste volume. In addition, because metals are being recovered instead of land disposed, they
do not have to be processed from ore concentrate; this saves energy and pollution of another
source.
25254108.01\aec3                           3-18

-------
              Table 3-1  Comparison Characterization Data and Treatment Performance Data for Inputs to an HTMR Process
                                                      Versus Inputs to Stabilization

BOAT Lot
Constituent
Antimony
Anenic
Barium
Beryllium
v7Bo in iiiiH
Chfouuuin
Lead
Mercury
Nickel
Selenium
Silver
ThalHum
Vanadium
Zinc
Cyanide .
Untreated P006 (Stabilization)
Avenge Value
Total (ing/kg)
.
Average Value
TCLPmg/1
.
1 <0.01
24.7
.
IS
6,176
2.SS6
.
6.449
.
11.2
0.47
.
2.8
43
78
<0.001
173
<0.01
0.2S
I
1
17,626

498

Untreated F006 (HTMR)
Average Value
Total (mg/kg)
10.3
Average Value
TCLPmg/1
<0.06
IS ! <0.004
S.9 |
I
i
<4 0.007
38,100
1,400
0.17
180,400
<1
1.3
0.19
<0.0002
800
<0.004
2.1 { 0.02S
"

I
i
23,600 j 375
70.4 1 0.17
Untreated K061 (HTMR)
Average Value
Total (mg/kg)
.
Average Value
TCLPmg/1
.
47 !
148

j
2,204
84.200
5.4
14.1
11.900 { 6.3
IS
1.000-113.700
-
9.0
S7 j
36

1
i
51, 000
4S6
0.94 1
Untreated K062 (HTMR)
Avenge Value
Total (mg/kg)
.
69
Avenge Value
TCLPmg/1
.
-
1.1
-
2.4
85.000
1,600
-
46,200

-------
                                                             Table 3-1 (continued)

BDAT List
Constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium

_.
(Jbromtuffl
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
Cyanide
Treated F006 (StaMliation)
Qm^mtntian Tnlal
(mgftg)

(mg/I)
j
i <0.01-<0.02



-
-
-
-
-
-
0.04-1.18

<0.01-3.23
0.03-1.21
0.20-2.39
<0.001
0.02-16.5
<0.01-0.20
<0.01-0.15


-
-
< 0.01 -36.9
-
Treated F006 (HTMR), K061, K062, and
Characteristic Wastes Containing Metals

wonoeninmon IOHU
(mg/kg)
< 3-391
<0.44).7
Concentration TCLP
(mg/1)
<0.06
<0.04-<0.005
3.8-285 I 0.44-3.17
<0.5-6.7 j < 0.005-0. 135
<0.5-17.8
0.006-0.104
930-32,500 } <0.03-2.17
< 4-86.1
<0.1
< 1.6-289 j
< 1.5-53.2 i
<0.5-9.75 j
i
< 0.6- < 1.2 i
< 1.1-190 1
<0.4-217 <
<0.05
< 0.04-0.38
<0.0002
<0.04-1.93
< 0.003- <0.015
0.005-0.048
<0.006
<0.01 1-0.07
0.098-1.25

Source: Final BOAT Background Document for F006 (August 1988) and data from INMETCO, submitted to EPA, October 1991 for the HTMR test for K061 and other
       metal-bearing waste streams.
-No data
•Includes data for different binder-to-waste ratio (i.e., 0.2-1.5).

-------
                            4. PERFORMANCE DATA

      The following paragraphs discuss the performance data used to develop the treatment
standards.  The treatment standards being promulgated in this  final rule are based on the
performance of HTMR and were originally developed  for K061 high  zinc nonwastewaters;
however, much of the performance data used was representative of mixed influent waste streams.
Data used to develop the HTMR treatment standards consisted of mixtures of K061 (both high
and low zinc subcategories), K062, F006, and characteristic wastes containing metals such as
D001, D002, D006, and D007.  As long as K061 high zinc subcategory wastes  were present in
the feed influent, the Agency used the HTMR performance data to develop treatment standards.

      The Agency is transferring the  treatment performance of HTMR for K061 high zinc
subcategory to other metal wastes, since  this  treatment performance was actually based on
recovery of metals from different wastes. Additionally, pyrometallurgical principals indicate that
metals will consistently partition in accordance with their metal properties at high temperatures.
As long as the waste material contains high concentrations  of metals  and does not contain
constituents that can adversely  affect  metal product specifications, metals will  predictably
partition regardless of the waste matrix.  (See Section 5 for more information about metal
properties as related to partitioning.) The HTMR data presented below explain why they were
or were not  used to  develop concentration-based treatment standards for K061  high zinc
subcategory nonwastewaters.

4.1   Data Reviewed and Selected to Develop the Treatment Standards Based on HTMR

      When EPA developed the treatment standards based on HTMR, the Agency considered
only data which met the following criteria:  (1) treatment data for well-designed and well-
operated HTMR processes; (2) data for treatment of high zinc subcategory K061  nonwastewaters
(containing equal to or greater than IS percent zinc) or data from HTMR processes treating a
mixture of high zinc subcategory nonwastewaters and other metal wastes; and (3) TCLP leachate

25254108.01\sec4                            4-1

-------
 data.  Accordingly, for the final rulemaking for K061  high zinc subcategory nonwastewaters,
 the data selected (which met the above criteria) consisted of four data sets. With respect to the
 "four" data sets  (discussed  below) used to develop treatment standards, the term data set
 represents a set of data from a particular HTMR facility; some facilities may have more than
 one set of data.  The first set of data (Table 4-1, at the end of this section) consists of 14 data
 points (including some data points that were below detection limits) for most of the BDAT. list
 metal constituents in HTMR slag.  This data set was submitted to EPA (prior to the proposed
 rule for high zinc subcategory K061 nonwastewaters)  by Horsehead Resource Development
 Company, Inc. (HRD), from HTMR units (waelz kilns) operated at its Calumet, Illinois,  and
 Palmerton, Pennsylvania, facilities.  Another data set (Table 4-2) from HRD, which consisted
 of three data points for all  the metal constituents (for its waelz kiln HTMR process), was
 collected by EPA during the First Third rulemaking. This data set is included as part of the first
 data set, i.e., HRD data. A  second data set (Table 4-3) was submitted to the Agency by SKF
 Plasma Technologies during  the First Third rulemaking.  This  data set consisted of one data
 point for most metal constituents and represented treatment of electric arc furnace (EAF) flue
 dust (K061) in a plasma arc  reactor.  The third data set (Table 4-4),  from International Mill
 Service (IMS), was submitted to the Agency after the comment period for the proposed rule for
 K061 high zinc subcategory nonwastewaters.  These data consisted of 16 data points for each
 of the metal constituents for HTMR of high zinc subcategory K061 nonwastewaters in a plasma
 furnace.  A fourth  data set  (Table 4-5), from International Metals  Reclamation Company
 (INMETCO), consisted  of three dau points  for each of the metal constituents representing
 treatment of K061, K062, F006, and several characteristic wastes in an HTMR system made up
 of a rotary hearth furnace followed by an electric furnace.

       Note, because cyanide is a common constituent of F006 and the Agency has no treatment
 performance data for destruction of cyanide in an HTMR system, EPA is transferring the K048-
 K052 treatment performance of incineration of cyanide. These data are presented in Table 4-6.
The Agency believes HTMR will achieve a level of destruction similar to incineration because
HTMR operates at longer residence times and occurs at higher temperatures than incineration

25254108.01\sec4                            4-2

-------
(i.e., most HTMR units operate at approximately 1,200-1,600°C while incinerators typically
operate at less than 1,100°C).

4.2   Data Reviewed But Not Selected to Develop Treatment Standards Based on HTMR

     . The Agency also reviewed other HTMR treatment performance data; however, these data
were not used to develop the treatment standards based on HTMR for the reasons discussed in
the following sections.   Additional data has been  received since the development of the final
K061 high zinc subcategory rulemaking.  (See Section 4.2.2.) The Agency believes that it is not
always necessary to revise treatment standards when  new data are received,  especially in
instances where the existing standards are being met. The Agency has reviewed the new data
and believes that although some data may show higher leachate levels than treatment standards,
in the past, many facilities operated in a manner that would produce a slag leaching metal levels
below the toxicity characteristic (TC) levels. Now, it seems that facilities are exploring slag
chemistry options and  process operation alternatives to improve  the  performance of their
technologies to achieve the treatment standards and generic exclusion levels. Thus, the Agency
believes that the standards are achievable by a wide variety of HTMR processes.

4.2.1  Horsehead Resource Development Co., Inc. (HRD) Data

4.2.1.1      WaelzKiln

       Four sets of treatment performance data, submitted during the First Third rulemaking,
were available demonstrating recovery  of zinc from K061 nonwastewaters using  a series of
waelz kilns (rotary kilns). These data were determined (in the First Third Rule) to be from an
HTMR system that was not well-operated.  These data are presented  in Table 4-7.   HRD
submitted comments disputing  the Agency's earlier findings that these data did not represent a
well-operated HTMR system. The Agency remained convinced of its earlier findings; however,
it did not use these data for setting final treatment standards.  The Agency responded to  HRD's

25254108.01\sec4                             4-3

-------
objections to EPA's determination of the fact that four of HRD's data sets were not well-
operated in the First Third Rule.   Additional information on the Agency's response to HRD's
objections not to use some of their data is contained in the First Third Rule's administrative
record for K061.

       The Agency remained convinced of its findings and was not persuaded to reverse them.
In summary, EPA determined that these four data sets were from HTMR processes that were
not well-operated. The reasons for this determination (made in the First Third Rule) were that
the Agency found that these processes were operating at insufficient operating temperatures
and/or had  deficient calcium to silica  ratios.   EPA notes  that some of HRD's operating
parameters are Confidential  Business Information (CBI) and, thus, are not presented here.
However, this CBI is located in the CBI portion of the administrative record for the K061 First
Third rulemaking.  The Agency also notes that HRD acknowledged that sample set number 5
(sample set number 3 in Table 4-7 of this background document), "is properly excluded due to
an upset condition in the kiln which caused a ring accretion to form requiring that the kiln be
shut down during the collection of sample 5 in order to remove the accretion and restore proper
kiln operation" (Comment Reference:  LDR7 L009).

       These HRD data (cited in Table 4-7 of this background document) show generally higher
total concentrations for zinc in the treated wastes,  i.e., 6,710, 23,600,24,300, and 27,400 ppm.
Conversely, the values for the three sets of data from  the well-operated processes (shown in
Table 4-2) indicate generally lower total zinc levels, i.e., 4,550, 4,680, and 11,200 ppm.  The
Agency noted that this variation in treated waste total levels for zinc (from the processes that
were not well-operated) occurred even though the untreated zinc total concentrations from all
seven data sets (Tables 4-2 and 4-7) were fairly stable, ranging from 129,000 to 155,000 ppm.
As noted above, the Agency concluded that poor operating conditions, i.e., improper operating
temperature and/or insufficient calcium to silica  ratios, were related to the poor performance
with respect to these data.
25254108.01\sec4                            4-4

-------
      In addition to other operating parameters, the Agency was convinced that zinc (TCLP
leachate levels)  is a good indicator of how effectively the HTMR process is recovering zinc.
Poor zinc recovery seems to be related to poor maintenance of proper operating parameters, as
discussed above. This can cause more metals to be found in the slag, resulting in both greater
slag volumes and the potential for more  metals to leach into the environment. EPA concluded
that  improper  removal  of zinc can likewise relate to less   immobilization  of hazardous
constituents.

      As previously mentioned, the Agency determined in the First Third Final Rule (and has
not changed the determination) that the treatment performance data from the  four data sets
referenced above were not from well-operated processes. EPA further noted that the HRD data
submitted prior  to the proposed rule for K061 high zinc subcategory nonwastewaters (Table 4-1)
show relatively consistent TCLP  leachate values for all the metals compared to the values
observed in the First Third data, notably those from the processes that were not well-operated
(shown  in'Table 4-7). This finding showed that the process could be better operated.

4.2.1.2  Flame Reactor HTMR Process

      HRD also submitted treatment performance data for  its flame  reactor HTMR process
(much of the data came from the  company's earlier filed delisting petitions).  These data are
shown in Tables 4-8 and 4-9.  The Agency reviewed these data and  believes that the flame
reactor process  may not have been properly operated for optimum metals recovery.  The Agency
also reviewed the recent treatment performance data submitted  by the commenter for its flame
reactor  process for HTMR of K061.   EPA considered using  these data but determined, as
explained below,  that these data were from an HTMR process that could not have been well-
operated. Thus, EPA decided not to use either the earlier or the recent flame reactor data based
on the reasons  discussed below.  The Agency reviewed  the report, Flame Reactor Process for
Hectric Arc Furnace Dust, prepared by HRD for the Center for Metals Production (CMP Report
No. 88-1, August 1988). EPA believes that some problems with the flame reactor discussed in

25254108.01\sec4                             4-5

-------
the report confirm the same concerns it noted when evaluating the current flame reactor data
submitted during the comment period.

       The report provided information on all aspects of using the flame reactor process for
EAF dust. In particular, the source test involved testing dust from different sources covering
a wide range of compositions to demonstrate the reactor's  flexibility.  This test  was also
performed to show the effects of feed ratios and combustion air composition on the  flame
reactor's performance. The Agency noted that the report concluded that "specific questions were
left unanswered, including those regarding the suitability of the process for variations in dust
composition" (p. 4-1 of the report). In addition, the report pointed out that the test results must
be evaluated in light of several deviations from the  original operating plans (p. 4-7 of the
report).

       In the source test, lead was the only element (zinc was not analyzed since it was not an
EP toxic metal) that exceeded the delisting limits (according to the report (pp. 1-3 and 4-19),
the delisting criteria are 6.3 times the Federal drinking water standards) in a few of the tests.
In fact, the report (and also in the commenter's recent  TCLP data, showed lead levels to be
much higher than delisting limits in the EP leachates.  The report concluded that in some cases
where the lead delisting levels were exceeded, these instances occurred in tests having "high feed
rates  and high levels of oxygen enrichment" (p. 4-19 of the report).  The conclusion of the
report further stated that "proper operating conditions will eliminate this problem* (emphasis
added) (p. 4-19 of the report).  Hence, EPA might conclude that where lead leachate levels are
high, operating conditions (such as temperature,  carbon monoxide  to carbon dioxide ratios
(CO/CC>2),  fuel, combustion  air rates, and combustion air composition) must be improper.
Additionally, in the discussion of zinc recovery,  the report found that there was considerable
variability over a wide range depending on EAF dust composition, dust feed rate, and the level
of oxygen enrichment.  Zinc recovery decreased as both dust feed rate and oxygen enrichment
increased; the effects were more significant for dust containing high levels of zinc.  With respect
to coke usage, the report concluded that "kinetic factors become increasingly dominant as  the

2525410S.01\sec4                             4-6

-------
zinc content of the EAF dust increases, and tend to limit the reactor's capacity for zinc fuming"
(p. 4-14 of the report).  In other words, it  appears  to be  more difficult to recover high
percentages of zinc from high zinc dust.  Thus, for high zinc dust more zinc is left in the slag
and,  therefore, is more likely to leach.

       The report further suggested that operating conditions (i.e., proper coke and combustion
usage) affect the volume of combustion gas generated, thus controlling the reaction kinetics and,
ultimately, affect the reactor's performance; hence, zinc recovery.  According to the report, "the
reactor could easily be scaled  based upon the combustion gas volume needed to achieve a
specific zinc recovery (emphasis added) and production rate.  Scaling calculations  would also
be needed to consider such factors as retention time and gas velocities. Higher temperatures and
higher carbon monoxide/carbon dioxide ratios may also enhance reaction kinetics and lessen the
importance of combustion gas volume" (p. 4-14 of the report).  Hence, EPA might conclude that
adjustments in operating conditions, such as retention time, could improve zinc recovery and
probably reduce zinc concentration in the slag  leachate.

       The Agency remained convinced that zinc residuals (TCLP leachate levels) are a good
indicator of how  effectively the system is recovering zinc.  Poor zinc recovery seems to be
related to poor maintenance of proper operating temperatures and other parameters.  These can
lead to more metals in the slag,  causing greater slag volumes and the potential for more metals
to leach into the environment. EPA concluded  that improper removal of zinc can likewise relate
to less immobilization of hazardous constituents.

       Noting that the report described the experimental flame reactor program,  the Agency
concluded that there is a relationship between the experimental flame reactor program described
in the report and the current flame  reactor's operation, since problems were observed with
respect to lead and zinc. Thus, the treatment of other metals may also have not been optimized.
EPA believed that the flame reactor process was not designed to optimize the performance with
respect to BOAT levels promulgated  in the final rule for K061 high zinc subcategory. In light

25254108.01\sec4                           4-7

-------
 of the aforementioned reasons, the Agency chose not to use the current flame reactor data. The
 Agency noted that the data in the commenter's delisting petition showed better performance than
 the data that HRD wanted EPA to use in the final rule.  EPA believed that this difference in
 results indicates that there were problems with the flame reactor process.
       •

        Further, the Agency noted the recent performance of HRD's flame reactor under EPA's
 Superfund Innovative Technology Evaluation (SITE) program.  (See the administrative record
 for the Superfund Innovative Technology Evaluation Demonstration Bulletin, Flame Reactor,
 Horsehead Resources Development Company, Inc.)  Tests conducted under this program to
 demonstrate the flame reactor's suitability for treating secondary lead smelters slag show treated
 TCLP slag analysis for lead averaging 0.33 mg/1 for 18 data points. The Agency acknowledged
 that these data were from a different waste matrix; however, these data show relatively low and
 consistent levels for lead in the TCLP extract, while noting that the total concentration of lead
 for the untreated waste averaged 54,066 mg/kg.

 4.2.2  International Metals Reclamation Company  (INMETCO) Data

       INMETCO submitted three data sets during the First Third rulemaking and one data set
 dated July 11, 1990.. Three sets of treatment performance data were available demonstrating the
 recovery of chromium/nickel and the separation of zinc using a rotary hearth furnace/electric
 furnace system (Table 4-10). These data were not used because the Agency had information
 concerning   INMETCO's  process  which  indicated  that  these  data  would  represent
 chromium/nickel recovery of low zinc subcategory K061 nonwastewaters mixed with F006,
 K062, and several characteristic wastes.  Insufficient information was provided to determine
 whether high zinc  subcategory K061 nonwastewaters were also being treated along  with these
 wastes. (The Agency notes that although some low zinc K061 nonwastewater data  were used
 to develop the final treatment standards for K061  high zinc subcategory, they were used only
 when the Agency was certain that the HTMR process was treating a mixture of low and high
zinc subcategory nonwastewaters.)

2S254108.01\sec4                            4_g

-------
      INMETCO also submitted additional TCLP slag analysis data from HTMR units. These
data were submitted to the Agency prior to the proposed rule  for K061 high zinc subcategory
nonwastewaters.   These data (Table 4-11) consisted of  four data sets, but there  was no
information to determine whether high zinc K061 nonwastewaters or a mixture of low and high
zinc  K061  wastes  were being treated  in this process at that  time.  Further,  the  Agency
determined that these data indicated that the treatment performance of the process would achieve
the treatment standards.

      Additionally, INMETCO recently submitted total concentration and TCLP slag  analysis
data from a HTMR test for K0$l conducted at INMETCO in June 1991. These data (Table 4-
12) consisted of 16 data sets. The preliminary indication is that some of the data points for
chromium may  not meet the proposed BDAT levels.  However, these data are  from grab
sampling as opposed to composite sampling from which the proposed standards for K061 (all
nonwastewaters) and alternative standards for K062 and F006  nonwastewaters are based. The
Agency notes,  however, that the most recent composite sampling data for INMETCO's slag
analyses for the year 1991 indicate that the treatment standards are achievable.  These total
concentration data and TCLP leachate slag analysis data (shown in Table 4-13) were submitted
from INMETCO for composite sampling of their slag for the years 1990 and 1991.

4.2.3  Sumitomo Corporation, of America Data

       During  the  First Third  rulemaking, Sumitomo  submitted two  sets  of treatment
performance data demonstrating recovery of zinc from K061 using a molten slag reactor system.
However, these data, presented in Table 4-14, did not provide treated TCLP leachate values.

4.2.4  International Mill Service Inc. (IMS) Data

       IMS submitted to the Agency (December 21,1990) 10 sets of treatment performance data
demonstrating recovery of zinc from K061 electric arc furnace (EAF) dust (K061).  In the

25254108.01\sec4                            4-9

-------
 proposed rule for the K061 high zinc subcategory, EPA reviewed process information from IMS
 indicating that the data were not for treatment of high zinc subcategory K061 nonwastewaters.
 During the comment period to the proposed rule, the Agency was informed that these data did,
 in fact, represent high zinc subcategory K061 waste.  However, IMS later informed EPA that
 these data were not from TCLP analysis but, rather, were from the Synthetic Precipitation
 Leaching Procedure (SPLP or Method 1312), which is not appropriate for setting treatment
 standards in accordance with the Land Disposal Restrictions BDAT methodology. The Agency
 noted that the TCLP analysis is used in accordance with the BDAT methodology because it is
 designed to reflect leaching in a landfill scenario.  These data are presented in Table 4-15.

 4.2.5  St. Joe Company Data

       One set of treatment performance data demonstrating recovery of zinc from K061 using
 a flame reactor was submitted during the First Third rulemaking.  These data, presented in
 Table 4-16, did not provide any treated TCLP leachate values.

 4.2.6  Beckett Technologies Data

       Heckett Technologies  submitted  data (shown in Table 4-17)  that represented the
 preliminary results of their HTMR process, i.e., Davy McKeeHi Plas Furnace.  The Agency
 considered these data. Because these data were only  preliminary, however, EPA chose not to
 use them since other data were available that were not preliminary and represented well-designed
 and well-operated HTMR processes.

 4.2.7 SKF Plasma  Technologies Data

      One set of treatment performance data, submitted during the First Third rulemaking, was
 available demonstrating  recovery of zinc  from K061 in a plasma arc reactor. . These data,
presented in Table 4-18, were not used because no TCLP leachate values were given.

25254108.01\sec4                           4.10

-------
                                  Table 4-1  Performance Data for HTMR of Zinc (Series of Waelz Kilns)
                                             of High Zinc Subcategory K061  Nonwastewaters
Constituent

Antimony


Beryllium
Cadmium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
K061
untreated
total
concen-
tration
(mg/kg)

— —
_
_^
	
	
..
__
__
__
_
_
	
__
170.000-
185.000
Treated slag concentration (mg/1. TCLP leachate)
sample sets
1
	
< 0.005
0.77
__
<0.01
0.01
<0.1
<0.001
<0.02
0.01
<0.02
_.
_
-
2
__
<0.005
0.42
..
<0.01
<0.01
<0.1
<0.001
<0.02
0.01
<0.02
_
_
-
3
__
<0.005
2.15
_
<0.01
<0.01
<0.1
<0.001
<0.02
<0.01
<0.02
_
_
-
4
_
<0.005
1.4
_
<0.01
<0.01

-------
                      Table 4-2  Performance Data (1988d) for HTMR of Zinc (Series of Waelz Kilns)
                                of High and Low Zinc Subcategory K061 Nonwastewaters
Constituent

Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
Total untreated
K061 concentration
(mg/kg)
Sample set
1
73
56
184
0.18
394
1,190
15,500
1.0
449
5.2
23
1.5
37
145,000
2
80
65
190
1.5
808
903
20,800
1.6
261
8.2
29
1.3
25
135,000
3
78
127
204
<0.5
290
1,080
6,400
1.1
295
20
44
<1.0
33
155,000
Treated slag concentration
(mg/kg)
Sample set
1
162
75
346
1.9
15
748
1,940
<0.1
579
4.2
42
<1.0
32
11,200
2
405
113
467
4.0

-------
        Table 4-3  Performance Data for HTMR of Zinc (Plasma Arc Reactor) of
                     High Zinc Subcategory K061 Nonwastewaters

Constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
BOAT constituents detected
Untreated total waste
concentration
(mg/kg)
50-150
< 100-400
—
—
200-900
400-5,000
1,500-2,800
24,000-50,000
7-41
1,000-3,000
_
_
_
_
150,000-320,000
Treated total waste
concentration (slag)
(mg/kg)
<20
<4-13
< 3,000
—
< 10-500
2,000-12,000
10-1,500
50-1,500
<5
200-1,000
—
—
—
—
50-2,000
Treated waste (slag)
TCLP
(mg/1)
—
< 0.005
2.5
—
< 0.005
0.013
—
<0.05
< 0.0002
0.22
<0.05
0.014
—
<0.02
• -
- = No data.

Source: SKF Plasma Technologies data submitted to EPA in the First Third rulemaking (USEPA 1988a).
2S2S4108.01\sec4
4-13

-------
                                          Table 4-4  Performance Data for HTMR of Zinc (Plasma Arc Reactor)
                                                    of High Zinc Subcategory of K061 Nonwastewaters


Conniinem
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chioiiuum
Lead
Mercury
Nfckrf
Selenium
Silver
Thallium
Vinadiutn
Zinc


Unirealed
total
wane
concen-
tration
(mg/kg)
-
-
-
-
-
•
•
-
-
-
•
-
-
•
Treated wade (tlag)
TCLP(mg/l)

1
0.057
0.017
I JO
<0.001
<0.005

-------
       Table 4-5  Performance Data for HTMR (Rotary Hearth Furnace/Electric Furnace)
                 of K061  (High and Low Zinc), K062, F006, and Characteristic
                                   Wastes Containing Metals
Constituent
Untreated
total waste
concentration
(nag/kg)

i
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
_
_
—
_
_
_
—
_
_
_
_
—
—
*
Treated slag TCLP leachate concentration
(mg/1)
Sample sets
1
<0.3
< 0.008
0.540
0.010
<0.010
<0.03
0.0114
< 0.0004
3.41
<0.03
< 0.010
< 0.012
<0.04
0.114
2 >
<0.3
< 0.008
0.282
<0.006
< 0.010
<0.03
0.0086
< 0.0004
. 2.18
<0.03
< 0.010
< 0.024
<0.04
0.240
3
<0.3
< 0.008
0.180
< 0.006
< 0.010
<0.03
0.0036
< 0.0004
2.04
<0.03
<0.010
< 0.012
<0.04
0.206
"Information provided in INMETCO comments states that these data represent high zinc K061 subcategory.
- No data.
Source: INMETCO presented to EPA during the comment period for high zinc K061 nonwastewatera proposed rule.
25254108.01\sec4
4-15

-------
                  Table 4-6  Performance Data for Incineration (Fluidized Bed) of Cyanide in K048 and K052 Nonwastewaters
Constituent


Cyanide (total)
Untreated Wade fag/kg)

1
0.7/0,8*
Sample teu
2
<0.1/OJ»
3
<0.1/<0.I*
4
1/1.4*

5
<0.l/<0.l»

6
0.9/0.6*
Treated W«te (Kg/kg)
Sample KU
1
<0.l
2
0.4
3
<0.1
4
0.5
5
<0.1
6
o.s
Source: Final BOAT Background Document for K048-KQS2 (A0gu« 1988).



•Rm value it foe KW4 wane and the aeeond value w for KOSI.

-------
Table 4-7  Performance Data for HTMR of Zinc (Series of Waelz Kilns) of Low Zinc Subcategory K061 Nonwastewaters
Constituent


Antimony
Anenie
Barium
Beryllium
Cadmium
Chioiiuuin
Copper
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
Untreated total
waaie concentration
(mg/kg)
Sample teu
1
89
59
169
0.55
737
905
2.080
14,400
1.4
184
13
30
2.7
24
129,400
2
65
55
164

-------
Table 4-8  Performance Data for HTMR of Zinc (Flame Reactor) of High and
              Low Zinc Subcategory K061 Nonwastewaters
                  Treated Slag Residuals TCLP (mg/I)
Sample
Ml*
,
2
3
4
5
<
7
8
9
10
II
12
13
14
IS
16
17
It
19
20
21
32
P
Sb
^
_
_
_
_
—
_
_
_
<0.10
_
<0.10
_
-
_
..
<0.10
_
_
_
_
0.10
-
Ai
0.01
0.01
<0.01
<0.01
0.02
0.01
0.02
0.02

-------
                                                                 Table 4-8 (c       led)
Sample
•el*
24
25
26
27
28
29
30
31
32
Sb

-------
Table 4-9  Performance Data for HTMR of Zinc (Flame Reactor)
                     of K061 Nonwastewaters
                     (Treated Slag Residuals)

Simple let
1
2
3
4
5
6
1
8
9
10
II
12
13
EPToxkhy(mg/l)
Pb
<0.2S
0.09
<.Q25
0.13
<0.02
<0.02
<0.02
<0.002
<0.027
<0.048
<0.02
<0.02
<0.02
Cd
<.004
0.004
<.004
«.004

-------
                                 Table 4-10  Performance Data for HTMR (Rotary Hearth Furnace/Electric Furnace) of
                                             K061, K062, F006, and Characteristic Wastes Containing Metals

Constituent



Lead
Zinc
Simple set 11
Untreated waste
TCLP
(mg/1)
—
256
213
mm
-
Tinted wade
(•tag)
TCLP
(mg/1)
_
0.65
0.62
_

Sample Ml 12
Untreated wane
TCLP
(mg/l)
—
6.8
5.4
0.39
-
Treated waste
(•kg)
TCLP
(mg/1)
-
0.40
0.28
0.35
-
Sample aet 13
Untreated waste
TCLP
(mg/1)
56
<0.10
-
365
4.973
Treated waste
(slag)
TCLP
(mg/1)
0.05
<0.01
-
0.38
0.94
£*       - = No data.
         Commenta on data-  Unknown zinc concentration; however, other information presented to the Agency concerning DiMETCO's HTMR process indicates that these data do not represent metals
         recovery from waste in the high zinc KM I nonwastewalers. Further, there is no information to suggest that these data were derived from K061 in the low zinc subcategory that were mixed with K061
         wastes in the high zinc •ibcalcgory.
         Source: Final BDAT Background Document for K061. August 1988. from INMETCO, Inc. (USEPA 1988a).

-------
   Table 4-11  Performance Data for HTMR (Rotary Hearth Furnace/Electric Furnace) of

             K061, K062, F006, and Characteristic Wastes Containing Metals
                                                Treated slag concentration

                                                        TCLP

                                                        (mg/1)
                 Untreated

                   waste

                concentration
 Constituent
 =^^=^

Antimony
 Chromium
 •BU.^^^H


 Lead
                                                                               0.0018
                                                                                 ^^•M^


                                                                                 0.37
 Nickel
 •H^^^^MM


 Selenium
 ^•i^^^w


 Silver
                                                                                <0.04
                                                                                •HW1^*«


                                                                                <0.01
<0.04
^M^^^^^M


<0.01
<0.04
^^^•M^^H

<0.01
 Thallium
 •^v—•—•

 Vanadium
- = No data.
                                                                                        data
are representative of low zinc K061 that was mixed with high zinc K051.


Source: INMETCO, dated July 11. 1990.
 25254108.01\sec4
                                          4-22

-------
Table
Performance Data for HTMR (Rotary Hearth Furnace/Electric Fi
                                                   S|i| Toul Ce
of K061, K062, F006, and Characteristic Wastes Contain^- Metals















Th».






*W 	 1 	

Mi in lift

» • •



Ztoo
SOUND: WMl
SMI




3.9
4S
IJ
<0j6
•

-------
      Table 4-13 Performance Data for HTMR (Rotary Hearth/Electric Furnace) of
            K061, K062, F006, and Characteristic Wastes Containing Metals
1
BOAT List
Constituent
Antimony
f Arsenic
	
Barium
Beryllium
Cadmium
1 Chromium
_ 	
Lead
Mercury
Nickel
Selenium
Silver
Thallium
]( Vanadium
1 Zinc
11 	 ^^^_^__ — ^—
Sample set SI
Treated Waste
Total
(mg/kg)
ND
ND
ND
_
300-500
33,000-39,000
1,500-2,700
ND
800-1,100
38-41
ND
_
_
100-400
I
TCLP
(mg/1)
ND
ND
0.02-0.7
ND
0.03-0.04
0.12-0.76
0.02-0.2
ND
0.35-1.7
ND
ND
—
—
0.24-0.26
ssssss^^ss^^ss^s^s
Sample set #2
Treated Waste
Total
(mg/kg)
<500
—
100-400
<500
<100
30,000-37,000
< 100-220
-
200-400
<500
<500

—
130-300
±5========
TCLP
(mg/1)
======
<0.06
< 0.008
0.18-0.8
0.01- < 0.006
< 0.01-0.06
< 0.03-0. 14
0.004-0.18
< 0.0004
0.88-3.41
<0.03 1
< 0.01-0.02
—
—
0.04-0.240 |
Source- INMCTCO monthly composites of slag samples for 1990 and 1991, for chromium and nickel-
      ' bearing feed stocks, i.e., K061, K062, F006, D001, D007, and D002/D006.

- = No data.

ND = Not detected.

Set 1 - 1990 performance data, Set 2 - 1991 performance data.
 25254108.01\sec4
                                          4-24

-------
      Table 4-14 Performance Data for HTMR of Zinc (Molten Slag Reactor System)
                     of High Zinc Subcategory K061 Nonwastewaters

Constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Vanadiuip
Zinc
Sample set #1
Total
untreated waste
concentration •
(mg/kg)
_
trace
_
—
600
3,900
4,500
_
_
_
_
_
188,200
Treated waste
concentration
(mg/kg)
—
trace
_
—
trace
6,500
200
—
_
—
—
—
12,200
Sample set #2
Total
untreated waste
concentration
(mg/kg)
—
—
—
—
20.2-30.0
0.7-1.4
348-SS6
-
-
-
—
—
—
Treated waste
concentration
(mg/kg)
e —
—
—
—
0.01-0.07
0.04-0.3
. 0.05-0.80
-
-
—
—
—
—
- = No data.

Comments on data: No TCLP leachate values.

Source: Final BOAT Background Document for K061, August 1988, from Sumitomo Corporation (USEPA 1988a).
25254108.01\sec4
                                          4-25

-------
                 Table 4-15  Performance Data for HTMR of Zinc (Plasma Arc Furnace) of High Zinc Subcategory KUfti wonwastewaters

•
Antimony
Arsenic
Barium
BeiylHum
Cidnrium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc

Untreated
WMU con»mnik»
. QMS later indicated to the Agency that

-------
           Table 4-16 Performance Data for HTMR of Zinc (Flame Reactor)
                  of High Zinc Subcategory K061 Nonwastewaters
Constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
f
Untreated total
waste concentration
(mg/kg)
—
—
—
— .
1,000 .
8,000
30,000
—
—
—
—
_
—
220,000
Treated total
waste concentration
(mg/kg)
—
— _
—
—
50
13,000
2,000
—
—
—
—
—
—
40,000
- = No data.

Comments on data:  No TCLP leachate values.

Source: Final BDAT Background Document for K061, August 1988, from St. Joe Company.
       (Note:  St. Joe Company Flame Reactor is currently owned by Horsehead Resource
       Development Company, (USEPA 1988a).)
25254108.01\sec4
4-27

-------
           Table 4-17  Performance Data for HTMR of Zinc (Plasma Reactor)
                   of High Zinc Subcategory K061 Nonwastewaters

Constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
Untreated Waste
(mg/kg)
Sample
set#l
-
.
_
-
180
34,200
13,000
_
3,100
-
-
-
.
93,400
Sample
set #2
-
-
-
-
-
84,100
4,200
-
8,800
-
-
-
-
173,500
Sample
set #3
-
-
-
-
-
84,100
4,200
-
8,800
-
-
-
-
173,500
Treated slag
TCLP (mg/1)
Sample
set#l
0.011
0.0012
3.07
0.033
0.01
<0.05
0.24
< 0.001
0.13
< 0.0005
<0.01
<0.10
<0.05
0.100
Sample
set #2
0.013
0.0009
0.59
0.033
0.01
1.25
0.1
<0.001
0.24
< 0.0005
0.01
<0.10
<0.05
0.374
Sample
set #3
0.013
0.0024
0.35
0.033
<0.01
0.68
<0.10
< 0.001
3.54
<0.0005
<0.01
<0.10
<0.05
0.708
- = No data.

Comments on data: These data were preliminary data from the Davy McKee Hi-Plas Furnace.

Source: Heckett Technologies, submitted to EPA during the comment period for K061 high zinc
       subcategory rulemaking.
25254108.01\sec4
4-28

-------
         Table 4-18  Performance Data for HTMR of Zinc (Plasma Arc Reactor)
                   of Low Zinc Subcategory K061  Nonwastewaters
Constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
Untreated
total waste
concentration
(mg/kg)
—
—
—
—
100-600
60,000-100,000
600-14,000
0.7-16
15,000-22,000
—
~
—
—
22,000-53,000
Treated slag
concentration
(mg/kg)
20
2.1
<200
—
<2
40,000-170,000
<5
<1
300-2,200
—
—
—
—
50-98
— = No data.

Comments on data: No TCLP leachate data for treated residuals.

Source:  SKF Plasma Technologies data submitted to EPA in August 1987.
25254108.01\sec4
4-29

-------
                5.  SELECTION OF REGULATED CONSTITUENTS

      This section presents EPA's rationale for selection of the regulated constituents in HTMR
residuals generated  from the  recovery  of  metals  present  in  K061,  K062, and  F006
nonwastewaters.  EPA is promulgating standards for 13 (as discussed earlier, the Agency is not
promulgating a treatment standard for vanadium) metals in all nonwastewater forms of K061 and
in the alternative standards for F006 and K062 nonwastewaters. EPA believes this is a valid
approach because all 13 metals have the potential to be present in K061, K062,  and  F006 as
generated, and it is a common industry practice to combine K061, F006, K062, and other metal
bearing  waste streams as feed sources to the HTMR process.  As a result, all 13 metals can
potentially be present at treatable levels. Furthermore, as some metals are removed in HTMR
processes, others will be concentrated in the treated residues (i.e., slag, baghouse dust).  As a
result, these concentrated metals may have the potential to be present at elevated levels (above
the untreated levels) in the treated residuals.

5.1   Constituents Identified in the Waste as Generated

      Generally, the  constituents  to be  regulated  are selected  from  a list of hazardous
constituents known as the BDAT list constituents.  (See methodology document for developing
BOAT treatment standards, USEPA 1988a.) This list is divided into the following categories:
volatile organics, semivolatile organics, metals,  inorganics other than metals, organochlorine
pesticides, phenoxyacetic acid herbicides, organophosphorous insecticides, PCBs, and dioxins
and furans.  EPA may revise this list as additional data and information become available.

      To identify potential constituents for regulation, the Agency reviews the characteristics
of the waste as generated and  those of the residuals,  as well as the elevated potential for
constituents to be present based on the waste generating process. Furthermore, the Agency may
 25254108.01\sec5                            5-1

-------
choose to regulate BOAT  list waste constituents that control the operation of the treatment
process.

5.1.1         Organics

       The available characterization data show no BDAT list organic constituents present at
detectable levels for K061 and K062 nonwastewaters. For F006, the available characterization
data do not show organic  constituents present at detectable levels.  However, based on the
Agency's knowledge of the generating process for F006, EPA believes that it is highly probable
that organics from solvents may be present in F006 wastes. The Agency, however, believes that
it is not necessary to regulate any organic constituents that may be present in K061, K062, or
F006 nonwastewaters since it is believed that they will be destroyed to nondetectable levels due
to the high temperatures and long residence times at which HTMR processes operate.

5.1.2         Cyanide

       The available characterization data for F006 indicate that cyanide is present at detectable
 levels. Hence, the Agency is regulating cyanide because it is a common constituent in most
 F006 wastes. The majority of the characterization data for K061 and K062 do not show cyanide
 at detectable levels; however, characterization data from INMETCO show some low levels of
 cyanide (ranging from 0.5-7.7 mg/kg) in the untreated waste.  EPA, however, believes that any
 cyanide that may be present in F006, K061, or K062 nonwastewaters will be destroyed to
 nondetectable  levels by the  high temperatures and long residence times  at  which HTMR
 processes operate.
 2S254108.01\sec5                            5-2

-------
5.1.3        Metals

       Based on the available characterization data, all metals on the BDAT list are present in
the K061 untreated waste (raw waste) at detectable levels and have the potential to be present
in high concentrations depending on the metal concentration  in the scrap materials smelted to
produce steel. Table 5-1 shows the potential sources of metals that may be present in K061 as
a result of the smelting of scrap metals.

       The available characterization data for K062 nonwastewaters show detectable levels of
all 13 metals except beryllium and thallium.  K062 wastes are generated by steel finishing
operations.  Since pickling of steel removes oxides that form on the steel's surface, these spent
pickling agents (e.g., acid solutions) may contain all 13 metals because the steel that is pickled
can be derived from scrap metals and produce many different grades of steel containing varying
metal concentrations.  EPA has determined that scrap metals and  the different grades of steel
may potentially contain all 13 metals; hence, it is possible that all  13 metals have the potential
to be present in K062 nonwastewaters.

       The available characterization data  for F006 nonwastewaters show  detectable  levels of
all 13 metals with the exception of beryllium and selenium.  The metal processing operations
that generate F006 consist of electroplating, anodizing, chemical etching and milling, and metal
cleaning and stripping.  These processes are performed on various types of metals and steels
(e.g., carbon steel, stainless steel, etc.) using various other metals and chemicals. For example,
the electroplating of metals  involves the process in which ferrous or nonferrous base  materials
are electroplated with cadmium, copper, nickel, chromium, brass, bronze, silver, gold, zinc, tin,
lead, iron, aluminum, etc.  As a result,  the Agency believes  that all 13 metals may potentially
be present in F006 nonwastewaters.
 25254108.01\sec5                             5-3

-------
                 Table 5-1  Possible Sources  of Metals in Scrap Used in Stedmaking
Constituent
Antimony
Anenic
Barium
Beryllium
Cadmium
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Zinc
Source
Antimony may be found in older, non-maintenance free batieriea, cast iron and ductile iron.
bearing*, gears, and ammunition.
Some potential aourcea of araenic are lead acid batieriea, lead ammunition, automotive body
solder, locomotive fireboxes, auybolla, straps, and plate* for heat exchanger* and condenaer
tube*.
Barium may be found in lead acid batieriea. lubricanta in vacuum X-ray tubea. and
television picture tubea.
Beryllium may be found in electrical contact*, (witches, clips, springs, die caata for plastic*,
and aonaparkiag safety tools.
Cadmium mav bft found in nieJcel-cadfiiiiifli hattAfu*. anlar nhf^m/nltair ^11* rarfmiiim
plating application*, and cadmium pigment*.
Chromium i* uaed in electroplating proceue*. flame and arc welding, and auialeia ateel.
Lead may be found in battery plates, drosses, ikimming*, and induauial acrap such aa
•olden, babbitt, cable sheathing, etc.

device*, ennfro) instrument*. »"^ WSJ*** •«"• sludffes ffeiwrmled in lihnntnriri •«!
electrolytic refilling plant*.
Nickel acrap may be found in old and/or new sheet, plate, bar, lube, and other wrought
nickel acrap solid*, copper-nickel peelings, plating racks, hanger*, and Maialeat and
specialty steel alloya.
Selenium ia added in trace amount* to various grade* of atainlea* steel*, indicating thai it
may be found at higher level* in dust* with high chromium and nickel content*.
Silver-bearing nuteiuli include jewelry, e|ectri
-------
5.2    Constituents Identified in the HTMR Residuals

       Based on the principles of the pyrometallurgical processes, different metals partition to
different HTMR residues (or products) at different concentrations depending on the design and
operating conditions of the HTMR processes along with the chemical and physical properties of
the metals. In essence, treatment of metals when using HTMR is directly related to partitioning
of the metals based on their volatility and their compounds as the waste is exposed to the high
temperatures of the HTMR process.  (See Table 5-2.)  Low-boiling point metals are volatilized
and subsequently recovered. High-boiling point metals are either reduced to form an immiscible
layer of molten metal (slag layer floats on top of the metal layer) or thermo-chemically stabilized
in HTMR residues such as slags.  The relative stability may depend on the solubility of the metal
and/or metal compound in the slag matrix.  This thermo-chemical stabilization of the relatively
nonvolatile  metals  occurs  because of  high temperatures,  the  relatively  efficient  mixing
conditions, the oxidation-reduction conditions in the primary furnace, and the presence of other
inorganic constituents that act, in effect, as stabilization reagents.  In fact,  many  of the same
conventional cementitious stabilization  reagents, such as calcium, silica, and alumina, are also
used as additives in some HTMR processes to achieve desirable HTMR operating conditions,
as well as to enhance desirable slag  properties.

5.3    Constituents Selected for Regulation

       For K061 in general, the Agency is regulating 13 metals (vanadium is not being regulated
in this final rule) for several reasons. The original standards for K061 (promulgated in the First
Thirds Rule) were considered interim standards,  53 FR  31164,  based  on  stabilization
performance data and not the performance of HTMR which is BOAT for K061 nonwastewaters.
These interim standards were set until HTMR capacity could come on-line, enabling the Agency
to better examine the performance  of HTMR units.   Further, at that time, the  Agency was
unaware  of the wide variety of metals composition in K061 wastes; hence,  the Agency did not
 25254108.01\sec5                            5-5

-------
Table 5-2 Chemical and Physical Properties for 13 Metals in HTMR Processes
Metal
Antimony
Arsenic
Beryllium
Barium
Cadmium
Chromium
Lead
Ease of Oxidation
Metal, eaiily oxidized to SbjOj.
Metal, eaiily oxidized to AsjOj.
Metal, easily oxidized to BeO.
Metal, very easily converted to BaO.
Easily oxidized to CdO at elevated
temperatures.
Will react win oxygen above 200*C to
formCrjO).
t
Eaiily convened to PbO.
Ease of Oxide Reduction
Reducible whh carbon at higher
temperatures {over 600'Q
Reducible at higher temperatures with
carbon.
Very difficult to reduce.
Very difficult to reduce.
Easily reduced with coke (C) to the .
metal at elevated temperature*.
CrjOj can be reduced with coke to
CrO.
CrO difficult to reduce.
CrOj decomposes to CrjOj and
oxygen above I96*C.
All oxides eaiily reduced with carbon
to the metal.
Boiling
Point of
MeUl('C)
631
813
2,970
1,640
765
2.172
1.740
BoUng
Point of
Oxide (*C)
1.550
193
3,900
2.000
1.559
CrO,
decomposes
196
CrjO,
decomposes
4000
CrO
decomposes
<2800
PbO-M«
PbOj
decomposes
290
FbjQf
decomposes
500
Solubility of Oxide in Stag
Low; requires good mixing to enure
solution. Otherwise, St^Oj will
remain at a separate phase.
Limited; requires good mixing to
ensure solution. Otherwise, AajOj
will remain ai a separate phase; will
react with alkaline slags to form
calcium arsenate, which is slag
soluble.
High; BeO will react with molten
silicates to form beryllium silkate.
High; BaO reacts with molten silicates
to form barium silicate.
Oxide ii somewhat soluble after
conversion to the silicate.
Low, except at high temperatures.
Somewhat soluble is lead silicate.
Expected Disproportionition of Metal
Compounds in the HTMR Processes
Some will be reduced to Sb metal.
volatilized and converted to SbjOj in gas
phase. Most will exit in the gas phase.
Most will volatilize as As^j and become
incorporated into the crude ZnO produced.
Almost totally incorporated into the slag
unless high reducing conditions are present.
Beryllium silicate is a natural mineral.
Almost totally incorporated into slag.
Moil oxide will be reduced to the metal.
The metal will volatilize and become
oxidized and collected. Remainder will be
incorporated as slag.
Cr2Oj will partially dissolve in (he stag and
partially be present as a aepante mollen
metal. Volatilization as metal or oxide will
be negligible.
Most oxide will be converted to metal.
The metal will volatilize and be oxidized
and recovered as oxide. A minority will be
dissolved as lead silicate in the slag.

-------
                                                                         Table 5-2 (continued)
Meul
Mercury
Nickel
Silver
Selenium
Thallium
Zinc
Eue of Oxidation
Above SOO°C, oxide reverti to
elemental metal.
Above lOOO'C. Ibe meul it readily
oxidizei lo NK).
Oxide icveiu to meul at 230*C.
Easily burnt lo SejO.
At 900'C, meul easily oxidized lo
TljO.
Meul easily oxidized lo ZnO.
Ease of Oxide Reduction
Easily reduced.
Above 400°C, the reduction with
carbon will occur.
Easily reduced lo meul.
SeOj is • gas al elevated temperatures.
TIjOj readily reduced. TljO difficult
lo reduce completely.
Reduced lo meul at process

Boiling
Point of
Meul (°C)
351
2,732
2,212
685
1.457
907
Boiling
Point of
Oxide (*Q
Decom-
poses to
meul and
oxygen at
SOO'C
>2.000
Decom-
poses al 230
350
1120-1,865
TI2Oj
at 875
1.975
Solubility of Oxide in Slag
Mercury metal will volatilize al these1
temperatures.
Nickel oxide will react to form nickel
silicate, which will dissolve in slag.
Oxide decomposes at 230*C.
SeO2 it somewhat soluble in silicate
matrices.
TIjO dissolves in slag to form lhallous
silicate.
Dissolves in silicate slags as zinc
silicate.
Expected Disproponionalion of Metal
Compounds in the HTMR Processes
Mercury vaporizes and is recovered as
melal particulars.
Nickel will primarily dissolve in the slag,
unless furnace conditions are highly
reducing, it will form an inmisicible molten
meul layer below die tlag layer.
Most metal will volatilize and be recovered
as paniculate.
Most selenium will volatilize and be
recovered at telenile-conUining
particulars. Small amounts will be
incorporated in tlag.
.Most thallium will volatilize as meul. be
convened to and collected as an oxide
impurity in the crude zinc oxide produced
A small amount will be incorporated into
tlag.
Mostly volatilizes aa metal, converts lo and
it collected a* the oxide. Some zinc will
become incorporated into alag.
Source: O.V. Samsonov, 1973. Hie Oxide Handbook, Plenum Press, NY.

-------
establish stabilization standards  for all 13 metals.   Information now available to the Agency
suggests that all 13 metals have a reasonably high potential for being present in any given K061
waste because of the nature of the steel manufacturing process from  which K061 is generated.
Data on  the  composition of K061  indicate that  these  13 metals are present  at varying
concentrations in K061 wastes from different generating facilities.  This appears to be related
to the types of scrap materials smelted in the electric furnace, the metals added to make certain
types of steel alloys, and/or the grade of steel produced. (Table 5-1 presents information on the
possible sources of metals that may be present in scrap metals.)

        With respect to F006 and K062 nonwastewaters, the Agency believes that all 13 metals
also have the  potential to be present in these wastes.  EPA believes this because the processes
generating  these wastes involve metal processing such as electroplating, etching,  anodizing,
 milling, pickling, and metals finishing that produce residues that may contain components of the
 substrate steel. Since steel is typically made from scrap metals and can be produced in different
 grades with varying metal concentrations, all 13 metals have the potential to be present. (See
 Section 5.1.3.)

        Since  all 13 metals have the potential to be present in K061,  K062, and F006, they also
 have the potential to be in the HTMR residues depending upon where the metals partition in the
 recovery process.  The Agency believes that improper operation of the HTMR process could
 result  in shifts in partitioning of certain metals to products  (e.g., metal alloys), intermediates,
 slag, or  other nonwastewater residues.  Hence, the metal distribution in  the HTMR process  is
 highly dependent upon parameters such as the operating temperature of the heat  zones, the
 composition  of metals and other elements in the feed, zone  residence  times,  flow rates,
 oxidation/reduction conditions, and mixing. The Agency believes that there is also an inherent
 metallurgical interdependency between certain metals, based on their atomic structure. Based
 on these factors, the Agency concludes that all metal-bearing materials (nonhazardous as well
 as hazardous) placed into HTMR processes could affect the ultimate composition and teachability
 of metals from HTMR nonwastewater residues.  Hence, the Agency believes that regulation  of
  25254108.01\sec5
                                            5-8

-------
13 metals (vanadium is not being regulated) will provide a means of ensuring that the HTMR
processes, when used to treat K061, K062, and/or F006 wastes, are well-designed  and well-
operated (i.e., truly BDAT) with due consideration of all feed materials.  Since all 13 metals are
potentially present in the treatment residues and may be hazardous to either human health or the
environment, EPA is promulgating treatment standards that will ensure control of the teachability
of all 13 metals.
25254l08.01\sec5                             5-9

-------
            6.  DEVELOPMENT OF BDAT TREATMENT STANDARDS

      The final treatment standards, based on the performance of HTMR, were calculated from
data that are representative of properly designed and properly operated HTMR processes.  The
HTMR  processes  have  been  demonstrated. to recover  metals  from  high  zinc  K061
nonwastewaters or mixtures of K061 nonwastewaters containing high and low  zinc subcategory
K061 nonwastewaters. Some of these data also represented the treatment performance of F006
and K062 nonwastewaters since these wastes are often mixed with K061 to achieve a desired
feed mixture. Data that meet these requirements include:  (1)  3 TCLP leachate analyses for all
13 metals and 13 TCLP leachate analyses for the 9 toxicity characteristic (TC) metals in the slag
(i.e., IRM) generated by the HRD waelz kilns process (Tables 4-1 and 4-2);  (2) 16 TCLP
leachate analyses for all 13 metals in the slag generated by the IMS plasma furnace process
(Table 4-4); (3) 1 TCLP leachate analysis for 10 metals in the slag generated by the SKF plasma
furnace process (Table 4-3); and (4) 3 TCLP leachate analyses for all 13 metals in the slag
generated by the INMETCO electric furnace process (Table 4-5). Performance data are discussed
in Section 4 of this document.

       Given that all of these technologies are capable of achieving  substantial immobilization
of hazardous constituents (though not identical levels of performance), EPA believes it to be
appropriate  to combine  the performance achievability when developing treatment standards.
EPA further notes that certain apparent differences in performance result from different reported
detection limits.  For many of the metals, all reported data show nondetectable levels of metals
in the HTMR slag, but different limits of detection due to different slag matrices (or perhaps due
to differing levels of performance by analytical laboratories).  In these cases, EPA uses the
highest analytical detection limits in order to accommodate performance of as  many of the well-
operated HTMR technologies as possible.
 252S4108.01\sec6                           6-1

-------
       As a result, the final treatment standards, based on the performance of HTMR, have been
calculated using the following BOAT methodology. First, treatment standards were determined
individually for each process. Then, the four sets of standards were compared to one another.
Based  on  this-comparison, the Agency selected the highest standard for each metal from each
of the four processes to allow for process variability and detection limit difficulties.  This
approach  derives limits achievable by all of  the major" HTMR technologies (and  probably
achievable by stabilization as well) since, properly  operated,  these technologies all appear
capable of substantially reducing the mobility of metals in HTMR slags.

       The following discussion details the specific methodology* used to calculate the final
treatment standards  based on the performance of HTMR as BDAT.   Before the  treatment
standards were calculated, the treated data were corrected for analytical accuracy by using the
available matrix spike percent recovery values to calculate accuracy correction factors (ACFs).
These corrected data are shown in Table 6-1.  The Agency had matrix  spike percent recovery
data from two HTMR process residuals (i.e., slag) that were considered from welWesigned and
 well-operated HTMR processes.  These data were (1) matrix spike percent recovery  data from
 Horsehead Resource Development Company, Inc. (HRD) obtained by EPA in the First Third
 Rule and (2) matrix spike percent recovery data submitted from International Mill Service (IMS)
 during the comment period to the proposed rule for K061 high zinc subcategory nonwastewaters.
 The HRD  matrix spike percent recovery data were used to correct HRD's data,  shown  in
 Table 6-5. The IMS matrix spike percent recovery data were used to correct its data,  shown
 on Table 6-3.  Because no matrix spike percent recovery data were available for the data from
 SKF  Plasma Technologies and INMETCO, the Agency used the matrix spike percent recovery
 data from IMS to correct the data from SKF Plasma Technologies and INMETCO. The Agency
 considered it appropriate to use the matrix spike percent recovery data from IMS because the
 slag from IMS's technology was more similar to SKF Plasma Technologies' and INMETCO's
      * For more information on the methodology for calculating BDAT treatment standards, see Methodology for
        Developing BDAT Treatment Standards (USEPA 1989).
  2S254108.01\sec6
                                            6-2

-------
slag than the slag from HRD.  HRD's technology does not form a true molten slag, whereas,
technologies from SKF, INMETCO, and IMS do form a molten slag.

       Next, the Agency calculated the Accuracy Correction Factors (ACFs), i.e., the reciprocal
of the lowest  matrix  spike  percent recovery value for each constituent.   The  ACFs were
multiplied by the treated values to yield the corrected data.  The accuracy-corrected data used
to calculate the final treatment standards are shown in Table 6-1.  The ACFs, percent recoveries,
and calculation of the final treatment standards are also shown in Tables 6-2 to 6-7.

       As discussed earlier,  in determining the treatment standards, the Agency used four sets
(representing the performance of four different HTMR  processes) that were considered
representative of well-designed and well-operated HTMR processes. The first step in developing
the treatment standards, based on the different HTMR processes, was to calculate treatment
standards from each of the four sets of treatment performance data. The next step was to select
the highest treatment standard for each metal derived from the four individual data sets.

       The calculation of final treatment standards for each of the four data sets is discussed
below.  Each  discussion is immediately followed by a table showing the corrected data, the
descriptive summary statistics for these data, and the specific methodology used to calculate the
treatment standard for each metal constituent and for cyanide in F006.
25254108.01\sec6                            6-3

-------
Table 6-1  Corrected Data Used in the Calculation of Treatment Standards Based on Performance of HTMR
Corrected
data

J
2
3
4
5
6
7
8
9
10
II
12
13
14
15
16
17
Antimony
Anenic
Barium
Beryllium
CTwiiTllllin
dirorniufn
Lead
Mcpcuiy
Honehead Resource Development Co. (HRD)
0.8382
0.929
0.3749
.
.
.
-
-
.
.
.
-
.
.
.
-
.
0.013

-------
Table 6-1 (continued)
Corrected
data

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Antimony
Arsenic
Barium

O.OS7
0.077
<0.074
0.052
0.0323
0.024
<0.010
0.064
0.042
0.013
0.030
<0.010
<0.010
< 0.010
<0.010
<0.010
0.0194
0.031
0.027
0.023
0.0085
0.0095
<0.0057
0.0319
0.0216
0.00798
0.0091
<0.0057
<0.0057
<0.0057
0.0068
<0.0057
1.404
2.19
1.48
1.43
1.99
1.68
0.51
1.03
1.37
0.875
4.08
0.378
0.443
0.454
2.19
0.475
IMS' mnplei 2-6 an SO gam iimplei.
IMS' Minplei 1 and 7-16 in 100 gnm samplei.
1
2
3
	
<.3
<.3
<.3-
<0.0091
<0.0091
<0.0091
0.583
0.305
0.194
Beryllium
Cadmium
Chronuum
Lead
Mercury
International MOb Senfce (IMS) Data
<0.00l
< 0.003
<0.003
<0.001
<0.003
<0.003
<0.003
<0.003
<0.003
<0.003
<0.003
<0.003
<0.003
<0.003
<0.003
<0.003
<0.0053
<0.0052
<0.0052
<0.0052
<0.0052
<0.0052
<0.0042
<0.0042
<0.0042
0.0095
0.0064
0.0127
<0.0042

-------
6.1
             SKF Play™ Technologies (See Table 6-2.)
       Only one data point was available for 9 metal constituents.  The Agency considered this
HTMR process  to  represent a well-designed and well-operated HTMR  process  and,  thus,
calculated treatment standards for 9 metals using the corrected data multiplied by a variability
factor of 2.8.  These  calculations are shown on Table 6-2.  The Agency reiterates that the
 methodology was modified wherein the highest treatment standard was selected for each metal
 constituent from four sets of treatment performance data. Hence, the Agency has not relied on
 just one datum point in the consideration of the treatment standard.
    25254108.01\sec6
                                              6-6

-------
                                Table 6-2  Calculation of Treatment Standards Based on HTMR Performance Data
                                                             From SKF Plasma Technologies


Percent recovery
(PRXft)1
Accuracy correction
factor (ACF)3
Mean of corrected
data (CD)
Variability factor (VF)
Fonnula for
calculating treatment

Treatment
standard fTS) (mg/1)
Antimony

—

-


—



.

Anenic
1
88

1.136
<0.0057

2.8
CDx
2.8

0.016

Barium
1
93

1.015
2.70

2.8
CDx
2.8

7.6'

Beryllium
m
.

-
m

.
m


.

Cadmium
1
94

1.064
<0.0053

2.8
CDx
2.8

0.015

Chromium
i
98

1.021
0.0133
•
2.8
CDx
2.8

0.037

Lead
1
97

1.031
<0.052

2.8
CDx
2.8

0.15

Mercury
1
9S.22

1.05
<0.0002
1
2.8
CDx
2.8

0.00059

Nickel
1
95

1.052
0.231

2.8
CDx
2.8

.0.65

Selenium
1
90

1.11
<0.056

2.8
CDx
2.8

0.16

Silver
1
100

1.0
0.014

2.8
CDx
2.8

0.039

Thallium
-
-

-
-

-
-


-

Zinc
-
-

-
-

-
-


-

CD - Corrected data.
<  - Detection limit value.
    • No data.
1 Matrix tpike data were Iraiufentd from the IMS plasma ftlmace matrix tpike •lag data (100 gram aamplea)- No matrix apiked data were available for Iheae data. Thua. the Agency decided it waa appropriate
to uae  the matrix apike data from another HTMR plasma furnace proceu that hat a true molten alag.
2 The matrix apike recovery value ia the average of the matrix apike data for all the melala. lince no value for thia metal waa available from a molten alag matrix.
3 ACF • loom.

-------
  6.2    Treatment Standards Calculations Based OF» HTMR Treatment Performance
         from International Mill Service (See Table 6-3.)
         Antimony and Arsenic
         There  were detection limit values and detected values for these two  metals.  For
         antimony, there were detected values above and below the highest detection limit; for
         arsenic, none of the detected values were below the highest detection limit.  Thus, the
         treatment standards for  these two metals were calculated using  the standard BOAT
         formula, i.e., Treatment Standard 
-------
          Cadmium. Chromium. Lead. Thallium, and Zinc

          For these five metals, there were detected values and detection limit values. Since all
          of the detected values were equal to or greater than the highest detection limit, the
          treatment standards were calculated using the standard BDAT formula, i.e.,

          TS  =  Exp (mean  of logtransformed data  +  2.33  (standard  deviation  of the
          logtransformed data)).

          Mercury and Selenium

          For these two metals, there were detection limit values and detected values. Since the
          detected values were below the highest detection limit value and none were above it, the
          Agency selected the highest detection limit and multiplied it times a 2.8 variability factor.
25254108.01\sec6                                 6-9

-------
                             Table 6-3  Calculation of Treatment Standards Based on HTMR Performance Data from
                                                                  International Mill Service

Number of
tamplei
Ibn-.!* mrmnni
(«)(%)1
Accuracy
(ACT)4
Mean of
(CD) or HDL
Variabyhy factor
(VF)
Formate for
cilntlitim
treatment
•tandaid
Treatment
etandai4(IS)
(rag/1)
A-uno^
16
too
98.63
1.0
1.01
0.0321
(Men)
5.2
TS-Exp
Bxp
(y+2.33
(V))
0,011
Chromium
16
98
100
, 1.021
1.0
0.016
(Mew)
2.5
TS-Exp
(y+2.33
(V))
0.041
Lead
16
97
100
1.031
1.0
0.1179
(Man)
1.4
TS-Bxp
(y+2J3
(*7»
0.17
Mercury
16
95.23
98.6^
1.05
1.01
<0.0032
(HDL)
2.8*
TS-
HDU2.
8
0.0090
Nickel
16
95
100
1.052
1.0
<0.027
(HDL)
2.8>
TS-
HDLx
2.8
0.076
Selenium
16
90
100
1.11
1.0
<0.05
(HDL)
2.8'
TS-
HDLx2.
8
0.14
Silver
16
100
100
1.0
1.0
0.015
(Mean)
1.5
TS«Exp
(y+2.33
(»y))
0.023
Thallium
16
86
93
1.16
1.075
0.0055
(Mean)
1.2
TS=Ejtp
(y+2.33
(»y))
0.0065
Zinc
16
98
100
1.021
1.0
0.615
(Mean)
8.6
TS=Exp
(y+2.33
(•y»
5.3
HDL  -  Hifbert detection limit.
,TS   -  Treatment ataadanl.
y     -  Meuofkiiiuwfoimtdecfrecteddatt.
•y    -  Standaido^viatiM of loftiuiafonned corrected data.
 <    -  Detection limit value.
1  -Hr«i«cowiyvahHWM (mine 100 f^iaiivlea fern ra
wai comidered     to be juct 100 percent.)
2  - Tto variability factor rixm for tow coittmieM ia 2^
ahown       were derived by dividing the treatment Aandaid by the mean of the conccled data.
3  - Matrix apifcevahie it me average of (be matrix apike data for all the mdali tince no value wai given for thit metal.
4  = ACP • 100/FR.
IMS data.  (SeeTabk6-l.) (Any value over 100 percent recovery

     (from tfaia data tet) i.e., HDL x 2.8; other variability faclon

-------
   6.3 Treatment Standards Calculations Based on HTMR Treatment Performance Data from
       International Metals Reclamation Company ONMETCO) (See Table 6-4.)
       Antimony. Arsenic. Cadmium. Chromium. Mercury. Selenium. Silver, and Thallium


       All values for these eight metals were nondetected values; hence, the treatment standards
       were calculated by selecting the highest detection limit and  multiplying it times a 2.8
       variability factor.


       Barium. Nickel. Lead, and Zinc


       For these four metals, all the values were detected values; hence, the  treatment standards
       were calculated using the standard BDAT formula, i.e.,


       TS =  Exp  (mean of the logtransformerf  data  +  2.33  (standard deviation  of the
       logtransformed data)).


       Beryllium


       There were detection limits and one detected value for beryllium; however, this detected
       value was not below the highest detection unit. Thus, the Agency used the standard BDAT
       formula to calculate the treatment standard, i.e.,


       TS  =  Exp  (mean  of the logtransformed  data  +  2.33  (standard deviation of the
       logtransformed data)).
25254108.01\sec6                               6-11

-------
                                        Table 6-4 Calculation of Treatment Standards Based on HTMR Performance Data
                                                             from  International Metals Reclamation Company
.
NWIIDGT Of MinpMO

(ft)1
JlMumnr .ymACtlna fnvAw
(ACF?*
Me*n of corrected An*
(CD) or (HDL)
VnUrilin; Actor 
(y+2.33(i»»
0.030
Mefcwy
3
95.2>
1.05
<0.00042
(HDL)
2.#
TS=
HDLx2.8
0.0012
Nickel
3
95
1.052
2.67
(Men)
1.9
TS"Exp
(y+2.33(iy)}
5.0
Selenium
3
90
1.11
<0.033
(HDL)
2.t>
TS-
HDU2.B
0.092
Silver
3
100
1.0
< 0.010
(HDL)
2»'
TS=
HDLx2.B
0.028 '
•niallium
3
U
1.16
00271
(HDL)
2S1
TS=
HDU2.8
0078
Zinc
3
98
1.021
01903
(Man)
2.4
TS=Exp
(y+2.33(iy»
046
HDL -  Highert detection limit.
TS   •  Treiunert Mimbrd.
j    "  MeanoflogtnuitfonnedcoRecleddrti.
ay   -  Standard deviation of loftranafoimed corrected data.
<   •  Detection limit value.
1    - Matrix apike data were Iraoafemd from the IMS' plaima furnace matrix apike alag data (100 gram aamplee).  No matrix apike data were available for theae data, n the Agency decided it wa§ appropriate to uie
    other DMlrix^ked^rromuiodierHTMRproceJtlhrth^ a true inoheniUg0.e., IMS).
2    "The variability factor drawn (or Ihit cotictitueol ie 2.1, the formula uaed for calculating die treatment nandarde (or mil cotutituent from this data act, i.e., HDL x 2.1; ether variability (acton were derived by
    dividing the treatment atandard by (he mean of the corrected data.
3    - Matrix «pike recovery value U the average of me matrix apike data for all the raclale lince no value waa available from a mohea alag matrix for Ihia metal.
4  • ACF • 100YPR.

-------
   Horsehead Resource Development Co.. Inc. (HRD) (See Table 6-5.)
    Antimony. Barium, and Zinc


    For these three metals, all the values were detected values; hence, the formula used for
    calculating the treatment standards was the standard BDAT formula, i.e.,
    TS  =  Exp (mean of  the  logtransformed  data  + 2.33  (standard  deviation of  the
    logtransformed data)).
    Arsenic
    For arsenic, the detected values were above and below the highest detection limit; hence,
    the standard BDAT formula was used to calculate the treatment standards, i.e.,
    TS  =  Exp (mean of  the  logtransformed  data + 2.33  (standard  deviation of the
    logtransformed data)).
    Beryllium. Mercury. Nickel, and Selenium


    For these four metals, there were detection limits and detected values.  Since none of the
    detected values were below the highest detection limit, the standard BDAT  formula was
    used to calculate the treatment standards, i.e.,
    TS  = Exp (mean of  the logtransformed data +  2.33 (standard  deviation  of the
    logtransformed data)).
25254108.01\sec6                            6-13

-------
  Cadmium. Chromium. I-ead. and Silver
  There were detection limits and detected values for all four metals.  Since the detected
  values were only below and not above the highest detection limit, the treatment standard
  calculation was derived by taking the highest detection limit and multiplying it times a 2.8
  variability factor.
   Thallium
   All values for thallium were nondetected values; hence, the Agency selected the highest
   detection limit and multiplied it times a variability factor of 2.8.
25254108.01\sec6
                                          6-14

-------
                                   Table 6-5  Calculation of Treatment Standard       id on HTMR Performance Data from
                                                         Horsehead Resource Development Company, Inc.

Number of
samples
Percent
recovery PR
(*)
Accuracy
correction
factor (ACP)
Mem of
corroded dttft
or HDL
Variability
factor (VF)
Formula for
calculating
treatment
standard
Treatment
standard (TS)
(mg/1)
Antimony
3
92
1.09
0.714
(Mean)
2.9
TS=Exp
(y+2.33
(•y»
2.1
Arsenic
17
100
1.0
0.006
(Mean)
2.0
TS-=>Exp
(y+2.33
(•y))
0.012
Barium
17
90
1.11
1.528
(Mean)
4.4
TS=Exp
(y+2.33
(>y))
6.7
Beryllium
3
90
1.11
<0.0013
(Mean)
2.1
TS=Exp
(y 4-2.33
(•y))
0.0030
Cadmium
17
87
1.15
<0.0689
(HDL)
2.81
TS-
HDLx2.8
0.19
Chromium
17
68
1.47
< 0.1 176
(HDL)
2.8>
TS=
HDLx2.8
0.33
Lead
17
76
1.32
< 0.132
(HDL)
2.8>
TS=
HDLX2.8
0.37
Mercury
17
90
1.11
0.0011
(Mean)
3.6
TS=Exp
(y+2.33
(ay))
0.0041
Nickel
17
93
1.08
0.0427
(Mean)
3.7
•TS=Exp
(y +2.33
fry))
0.16
Selenium
17
48
2.08
0.0202
(Mean)
2.1
TS=Exp
(y+2.33
(ay))
0.042
Silver
17
76
1.32
< 0.105
(HDL)
2.8»
TS=
HDLx2.8
0.30
Thallium
17
96
1.04
<0.0104
(HDL)
2.8'
TS=
HDLx2.8
0.029
Zinc
3
98
1.02
0.1527
(Mean)
3.3
TS=Exp
(y-f2.33
(ay))
0.50
HDL 
-------
6.5 Treatment Standards Calculation for Cyanide (See Table 6-6.)

       The alternative treatment standard for cyanide in F006 is based on the incineration
treatment performance data used to develop the K048-K052 cyanide treatment standard.  (See
Table 4-6.)  The Agency has no data on the treatment performance of HTMR for destruction
of cyanide but believes that HTMR will achieve a level of destruction for cyanide similar to
incineration.  EPA believes the levels will be similar, since HTMR occurs at higher temperatures
than incineration  (i.e.,  most HTMR  units  operate at  l,220°C-lf600eC while incineration
typically operates at  less than  1,100°C).   Further,  HTMR typically operates with  longer
residence time.
25254108.06\sec6                            6-16

-------
                          Table 6-6  Calculation of Treatment Standards for Cyanide in F006 Based on Incineration
                                Treatment Performance Data for Cyanide in K048 and K052 Nonwastewaters
Regulated Constituent
Cyanide
UnsUbilized
Incinerator Ash
from Plant A (rag/kg)
0.5-1.4
Arithmetic
Average of Corrected
Treatment Value (mgSkg)
0.274
Variability Factor (VF)
6.37
Treatment Standard*
Average K VF (Dig/kg)
1.8
Source: Final BOAT Background Document for K048-K052 (August I9ti8).

*Thc values shown on this table for treatment standards have been rounded to snow significant figures only.

-------
6.6    Treatment Standards Calculations for Metals fi.e.. Highest Standard f™n ^ f ffir
       Sets of Treatment Performance Dafq) (See Table 6-7.)


       Table 6-7 presents the treatment  standards and summary data for the standards that

resulted from the highest value selected among the four sets of standards (as previously discussed
in the Introduction to Section 6).
2S2S4108.01\aec6                            6-18

-------
            Table 6-7 Calculation of Treatment Standards for M       from the Four Sets of HTMR  Performance Data
=========
=====
Number of
Source*
recovery
(PR) (ft)
Accuracy
correction
factor
(ACF)
Men of
corrected
data or HDL
or CD
Variability
factor (VF)
Formula for
treatment
•andard
Treatment
^_______»_
=====
Antimony
=====
3
HRD
92
1.09
0.714
(Mean)
2.9
TS**E«p
(y+2.33
2.1
___«^«
=====
Anenic
=====
16
IMS
88
1.14
0.014
(Mean)
3.9
TS-Exp
(y+2.33
(•y»
O.OSS
========
Banum
=====
1
SKF
93
1.08
2.70
(CD)
2.8>
TS-
CD«2.8
7.6
=====
Beryllium
=====
3
INMETCO
100
'1.0
0.0073
(Mean)
1.9
TS-Eip
(y+2.33
0.014
=====
Cadmium
=====
17
HRD
87
1.15
<0.0689
(HDL)
2.8'
TSi
HDU2.8
0.19
Chromium
=====
17
HRD
68
1.47
< O.I 176
(HDL)
2.8'
TS-
HDU2.8
0.33
Lead
=====
17
HRD
76
1.32
 bclut ata»a fat *•
          MM *tm« fcf **i*^ *• MMM
          %•• 1 ak*» » 1 iu » t !••
                                    I 8. dua h. tkc fc«nuU »*td U cakulatinf the Irealmtol aundard for Ihia vecific conUituent from Ihil daU Ml. i c.. HDL I 2.8; other variability hciora ahirwn
                                   » *• wwi of rt
                            •*•«• r«ov«ry value* »CM ua» fnim

-------
   7.  GENERIC EXCLUSION FOR K061, F006, AND K062 NONWASTEWATER
                RESIDUES (SUCH AS SLAG) GENERATED FROM
                   HTMR PROCESSES AND RELATED ISSUES
      This section discusses the development of the generic exclusion  levels for HTMR
nonwastewater residues.  The Agency concluded in the final rulemaking for K061 high zinc
subcategory nonwastewaters that K061 (both'low and high zinc) HTMR nonwastewater residues
that meet generic exclusion levels for all constituents and that exhibit no characteristics of a
hazardous waste will  not be hazardous.  The decision to generically exclude nonwastewater
HTMR residues is based on the fact that the treatment process is well-defined and, thus, does
not require an in-depth evaluation of each facility's processes.  The Agency believes that the
"derived-from" rule's presumption of hazardousness no longer should  apply to HTMR residues
which have treated toxic metals to specific levels. The Agency has made this determination after
considering the protectiveness factors in § 300 l(f) and after satisfying the underlying philosophy
of the delisting provisions.

       The generic exclusion levels include all the Appendix Vm and indicator metals that might
reasonably be expected to be present  in the HTMR nonwastewater residues from processing
F006, K061, and/or K062 wastes by HTMR.  (This is consistent with RCRA section 300l(f)
which  requires EPA  to evaluate whether toxic constituents, in addition to those for which a
waste is listed, could  make a waste hazardous.) The Agency notes that it is not setting exclusion
levels for organic constituents that could be common in F006 wastes. The Agency does not have
specific performance data on the destruction of organics in HTMR units.  However,  since
HTMR units operate at higher temperatures and longer residence times than incinerators, the
 Agency believes  that HTMR will  achieve a level of destruction similar to  or better than
 incineration. Consequently,  the Agency believes that the regulation of organic constituents  in
 HTMR residues is not necessary since any organic constituent will be destroyed to nondetectable
 levels in HTMR units, and the regulation of 13 metals (vanadium is not being regulated in the
 final rule) will ensure the proper operation of HTMR systems.

 252S410S.01\ttc7                           7-1

-------
       Residues from HTMR of K061 (and with this rulemaking residues from F006 and K062
wastes) wastes in units identified as rotary kilns, flame reactors, electric furnaces, plasma arc
furnaces, slag reactors, and rotary hearth furnace/electric furnace combinations or industrial
furnaces (as defined in 40 CFR 260.10(6), (7), and (12)) are excluded from the hazardous waste
regulations when  disposed of in a Subtitle D landfill. The exclusions are provisional upon the
residues meeting  the generic exclusion levels for all constituents, and that they do not exhibit
one or more of the hazardous waste characteristics.  The reasons for specifying HTMR for the
exclusion are provided in  Section 7.5 of this document  titled,  'Applicability of  Generic
Exclusion to Other Treatment Residues." In addition, the residues will be subject to the tracking
and testing requirements described in Sections 7.3 and 7.4.

       The Agency evaluated the treatment standard levels using the EPA Composite Model for
Landfills (EPACML),  which predicts the potential for ground-water contamination from wastes
that are land disposed.  The EPACML simulates the movement of contaminants for a Subtitle D
waste management unit and migration through the subsurface to a potential drinking water well.
This  model estimates a dilution-attenuation factor (DAF) for contaminants, which represents the
 reduction in concentration expected to occur during migration. The Agency used the EPACML
 previously in establishing the Toxicity Characteristic (TC) (see 55 ZB H326; March 29, 1990),
 and recently adapted it for use in evaluation of petitions to exclude ("delist") hazardous wastes
 under 40 CFR 260.22  (see 56 FJfc 67197, December 30,1991 and 56 ER 32993, July 18,1991).

       The EPACML uses a Monte Carlo simulation technique to account for the wide range
 of hydrogeologic settings found at municipal waste landfills, as well as the uncertainty in the
 data. The Monte Carlo analysis generates a distribution of DAFs which can be represented by
 a  cumulative frequency distribution (i.e., probability) curve. DAFs along this curve decrease
 as the cumulative frequency increases (high DAFs) to "worst case" situations at high cumulative
 frequencies (low DAFs).  See the above cited TC rule for more details on the development of
 the EPACML.
  252S410«.01\«c7                            7-2

-------
      For use in delisting evaluations, the Agency modified the model somewhat to allow the
use of the volume of the petitioned waste to be used as a fixed input value.  Delistings are
facility specific,  and the volume of waste (e.g., annual generation volume) is defined in the
petition.  Furthermore, EPA also selected the 95th  percentile as being the most appropriate
cumulative frequency for use in delisting.  The modified use of the EPACML is described in
detail in a recent proposal (56 F_B 32993; July 18,  1991).

      The Agency believes that a DAF of  10 is  most appropriate for use in evaluating the
treatment standard levels, based on the past  DAFs calculated for the TC rule and for use in
delisting.  A DAF of 10 corresponds to approximately the 95th percentile level generated from
EPACML simulations used to support the TC rule (see 55 ER 11826; March 29, 1990).  The
Agency used a DAF of 100 for establishing the TC regulatory levels for wastes that are "clearly
hazardous", and noted that it was appropriate to set the level on  an "order-of-magnitude"
precision. A DAF of 100 corresponds to a cumulative frequency close to the 85th percentile.
An exclusion establishes regulatory levels below which the waste may reasonably be expected
to be nonhazardous, therefore, the Agency believes it  is  appropriate that the cumulative
frequency used be somewhat higher than the percentile used to establish the clearly hazardous
TC levels.

       The EPACML as modified for delisting generates DAFs that vary from a maximum of
 100 for relatively small volume waste generated (1,000 cubic yards/yr.) to DAFs approaching
 10 for larger volume generators) (300-400 cubic yards/yr.).  A table for DAFs for different rates
 of waste generation is given in  Table 7-1. Note that this table is applicable to the delisting of
 an ongoing waste generation process.  To account for the total amount of waste generated and
 ultimately land-disposed, the annual waste volume has been multiplied by a factor of 20, based
 25254108.01\«ec7                            7-3

-------
             Table 7-1  EPACML-Derived Dilution and Attenuation Factors
                                    for Landfills

	
Waste Volume
1 (cubic yards per year)
1,000
1,250
1,500

2,000
2,500
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
12,500
15,000
20,000
25,000
. 30,000
40,000
I 50,000
95th Percentile DAF
EPACML Landfill
100*
96
90
84
79
74
68
57
54
48
45
43
40
36
33
29
27
24
23
20
19
252S410S,Ol\sec7
7-4

-------
                                Table 7-1 (continued)

Waste Volume
(cubic yards per year)
60,000
80,000
90,000
100,000
150,000
200,000
250,000
300,000
400,000
95th Percentile DAF
EPACML Landfill
17
17
16
15
14
13
12
12
10
•DAF maximum is 100 corresponding to the Toxicity Characteristic.
 25254108.01\sec7
                                          7-5

-------
 on a 20-year active lifetime of a Subtitle D unit. Therefore, the table reflects DAFs calculated
- for landfills sized to be 20 times the annual generation rate shown in the table (e.g., the DAF
 for an annual waste volume of 10,000 cubic yards was calculated for a landfill unit containing
 200,000 cubic yards of wastes).

        A DAF of about 10 also results from the use of the EPACML DAFs generated for use
 in delisting.  For the purpose of establishing a generic exclusion level, rather than a facility-
 specific delisting,  the volume of waste to be excluded is not well defined.  Therefore, the
 Agency believes it is appropriate to assume a reasonable worst case landfill size, i.e.,  a landfill
 corresponding to the 95th percentile in size for the Subtitle D landfill units contained in the EPA -
 database  The data contained in Table 7-2 shows  the distribution in the size of active landfills,
 and indicates that the 95th percentile size landfill would be on the order of 8 million cubic yards
  (or in terms of the units in Table 7-2, about 6 million cubic meters).  A landfill of this size is
  equivalent to an annual generation rate of approximately 400,000 cubic yards, which would have
  a corresponding DAF of about 10 (see Table 7-1).

         Under a DAF  of 10, and the appropriate health-based  level (e.g., drinking water
   standard) EPA evaluated the BDAT-based levels and established generic exclusion levels which
   it considers safe to human health and the environment. The following section (i.e., Section 7.1)
   provides details on the health-based levels shown on Table 7-3.

   7.!   Tyranny"* .rIMJ,-imni«—* «-**» T ™* «"d Mutation of Allowable
          Tte Delisting Section of EPVs Office of Solid Waste, in its review of delisting penaons
   and generic  exclustoni, evataaes levels  of  carcinogens and systemic toxicants listed  «
   Appendices vnivm.a»dIXof40CFRPan261. The expose assumption used to assess me
   hazard of a petitioned waste i, ingestion of contaminated ground w«er. leacha*. or was«wa«.
    For bo* carcinogens and systemic toxicants, *e water inote as»mption i, 2-Iittrs per^lay for
    2S254108.01\sec7

-------
Table 7-2 Data Source:  Subtitle D Landfill Survey
                  Active Units
                   Univariate
Variable «= LFA4 To(J|1 Vo|ume rf ^.^ LF ^ (C{J ^

H
Mean
StdDev
Skewness
USS
CV
T:Meaa«0
SgoRank
Num-0


Moments
1180
171 1939
7586576
13.9915
7.I32E+I6
443.157
7.75146
34839S
1180



SuroWgts
Sum
Variance
Kurtosis
CSS
Sid Mean
Prob>|T|
Prab>|S|


KsaaMOMBOB:

1180
2020088436
5.756E-H3
271.13
4.786E+16
220854
0.0001
0.001



Quanliles (DEF = 4)
100% Max
75% Q3
SO%M«d
25% Ql
0%Min

Range
Q3-QI
Mode
177471293
861551
201029
38416.1
189.541

177471103
823135
28718.4
Missing Value Count
% Count/Nobs
gaaamamsaac;
99%
95%
90%
10%
5%
1%



116
8.95
aaaoBBBBa
25754902
6296676
3100205
10396
4439.97
681.774

*


=3EESOB*S^
Exlnemes
Lowest
189.541
191.264
236.065
251.573
305.564



^•^••••^^^^^M.
awraosass
^BSSS^BSSS^
Highest
60308590
61216090
63180427
88124060
177471293


•^•^——^—^—^•^••i^^
^««<-i-^M«>H.^_M_
=3=5===

-------
            Table 7-3 Generic Exclusion Levels for K061, K062, and F006
                          HTMR Residues (Nonwastewaters)
Regulated 'constituent
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium (total)
Lead
| Mercury
Nickel
Selenium
Silver
Thallium
Zinc

Cyanide* (total)
Maximum for any single cofliposite sample
TCLP leacbate
(mg/1)
0.01
0.055
7.6
0.01
0.05
0.33
0.15
0.0090
1.0
0.16
0.30
0.02
70
Maximum for any yngle composite sample
Total Concentration (mg/kg)
1.8
•The level for cyanide applies only to F006 nonwastewaters.
25254l08.01\Mc7
                                         7-8

-------
an average 70-kilogram (kg) adult over a 70-year lifetime.  The use of a 70-year lifetime
considers the effects of carcinogens as a function of cumulative doses, rather than doses received
by any small subsection of the population. In addition, in cases where constituents are both
carcinogens and systemic toxicants, the more conservative carcinogenic slope  factor takes
precedence over the toxicant reference dose.

       The following equation is used to calculate the delisting and exclusion health-based levels
for carcinogens:
                        Dc = (R x W x LT) / (CSF x I x A x ED)

       where:        Dc    = delisting  health-based level for
                                  carcinogen (mg/1)
                     R     = assumed  risk level = 10"6
                     W    - body weight - 70 kg
                     LT    - assumed lifetime = 70 yean
                     CSF  = carcinogenic slope factor =*
                              experimental potency (mg/kg/day)'1
                     I      — intake assumption = 2 I/day
                     A     = absorption factor = 1
                     ED    - exposure duration « 70 years

        The following equation is used to calculate the delisting health-based levels for systemic
  toxicants:

                                Dfi =  (RfD x W) / a  x A)
  25254108.01\wc7
                                           7-9

-------
       where:       Ds     = delisting health-based level for
                                systemic toxicant (mg/1)
                    RfD    = reference dose (mg/kg/day)
                    W.     = body weight = 70 kg
                    I       = intake assumption =  2  I/day
                    A      = absorption factor = 1

       Constituents that have verified health-based levels are listed in EPA's Integrated Risk
Information System  (IRIS), which is maintained by  the Office of Health and Environmental
Assessment in the Office of Research and Development.  The information listed  in IRIS is
designed to be a guide for the evaluation of potential health' problems and is included in IRIS
only after an intra-office work group of EPA lexicologists and other scientists have reviewed the
facts.   IRIS provides  verified information  for oral and/or inhalation reference doses, risk
estimates for carcinogen!city, drinking water health advisories, risk management summaries, and
other supplemental data. (IRIS provides the carcinogenic slope factors and the reference doses
that are needed in the above equations.) IRIS is currently available to EPA staff through the
EPA electronic mail  system as well as to the general public, who can access the system through
various on-line networks such as DIALCOM, Inc., the  Public Health Network (PHN), and the
National Library of Medicine's TOXNET.  The general public should contact any of the latter
networks to obtain an IRIS account.  Hard copies of IRIS information for all constituents with
verified delisting health-based levels is provided by the Delisting Section upon request.

       Some constituents used in delisting are not in IRIS. In these cases, other references, such
as Health and Environmental Effects Documents (HEEDs), Office of Drinking Water (ODW)
health advisories, Carcinogen Assessment Group (CAG) recommendations, and various chemical
files are used and will be provided by the Delisting Section upon request.  The same equations
presented above are  used to calculate delisting health-based levels.

25254108.01\sec7                           7-10

-------
       During a delisting determination, EPA often uses appropriate fate and transport models
and waste-specific information (e.g., waste volume and constituent concentration data) to predict
the impact of a petitioned waste on human health and the environment.  In selecting appropriate
models, the Agency chooses a reasonable worst-case  management scenario for the petitioned
waste and considers plausible exposure routes for hazardous constituents present in the waste.
The Agency believes a reasonable worst-case scenario is appropriate when evaluating whether
a waste should be relieved of the protective management constraints of RCRA Subtitle C.

       Under a landfill disposal scenario, the plausible exposure route of concern for hazardous
constituents is ingestion of contaminated ground water.  The EPACML model approximates the
transport processes likely to occur in a drinking water aquifer below a waste disposal site. The
waste-specific parameters used in the EPACML model  are  the leachate concentrations  of
constituents of concern and the volume of the waste generated annually. Typically, the leachate
concentrations are provided in a delisting petition by an  appropriate leaching test (e.g., the
Toxicity Characteristic Leaching  Procedure)  on a  constituent-specific basis.   The result
calculated by the model for a given constituent of concern should be less than or equal to the
respective drinking water standard, or other EPA  health-based level (HBL) as appropriate, to
pass a delisting evaluation.
       The maximum allowable teachable concentration (C^) for a delisting constituent of
 concern can be calculated by  multiplying the level of regulatory concern (i.e., HBL) by the
 EPACML model dilution factor (Dp). The dilution factor, which is dependent on waste volume,
 decreases  from a  maximum of 100 for small waste  volumes (1,000-cubic yards) to a value
 approaching 10 as waste volumes approach 300,000-  to 400,000-cubic yards.

        In cases where the estimated waste volume approaches 300,000 to 400,000 cubic yards,
 a dilution  factor of 10 is used in the following equation:
 2S254108.01\wc7                            7-11

-------
                                 (HBL)(DF)      DF   «  10

       For example,  using the current HBL for cadmium of 0.005  mg/l,  the  maximum
constituent leachate concentration is calculated as:

                                 «  (0.005 mg/1) (10)

                                 -  0.05 mg/l

       If the maximum leachate concentration reported for a constituent of concern exceeds the
calculated  maximum allowable concentration, the Agency may conclude that the waste could
present a threat to either human health or the environment.

       When available,  HBLs  are Maximum  Contaminant Levels (MCLs),  which can  be
obtained from EPA's IRIS. MCLs are promulgated under the Safe Drinking Water Act (SDWA)
of 1974, as last amended in 1988, and consider technology and economic feasibility as well as
health effects. Finalized MCLs are used as HBLs for delisting for carcinogens and systemic
 toxicants when available.  Proposed MCLs are used as HBLs for delisting for carcinogens and
 systemic toxicants when  Finalized MCLs are not available. Table 7-4 summarizes the regulatory
 levels of concern for certain inorganics and the maximum allowable concentrations for a solid
 waste currently used in delisting and exclusion determinations.

        In the absence of formal MCLs, the Agency has also used other appropriate HBLS to
 establish delisting levels.  In the absence  of a new MCL for lead, the Agency believes that
 prudence  requires that the exclusion level be established using the more conservative action level
 of 0.015 mg/l. EPA established the new treatment standard for lead instead of an MCL because.
 as EPA  concluded in the  preamble to the final rule for K061 high  zinc subcategory
 nonwastewater,  there is no apparent threshold for various health effects associated with lead.
  2525410S.01\sec7
                                           7-12

-------
             Table 7-4  Health-Based Levels (HBL) and Maximum Contaminant
                         Levels  (MCL) for Constituents of Concern
CAS No.
7440-36-0
7440-38-2
7440-39-3
7440-41-7
7440-43-9
7440-47-3
57-12-5
7439-92-1
7439-97-6
7440-02-0
7782-49-2
7440-22-4
7440-28-0
7440-66-6
Compound
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cyanide
Lead
Mercury
Nickel
Selenium
Stiver
Thallium
Zinc
HBL
(mg/1)
1 x 1(T2
5xl(T2
1
1 xUT3
5 x 10T3
1 x 1QT1
2 x KT1
1.5 x 10T2
2xl(T3
IxUT1
5xHT2
5x UT2
2x i(r3
7
HBL
Source
1
2
2
1
3
3
1
5
3
. 1
3
2
1
6
Maximum
allowable
concentration
(mg/1)
0.1
0.5
10
0.01
0.05
1.0
2.0
0.15
0.02
1.0
0.5
0.5
0.02
70
'Assumes a OAF of 10.
Sources:
                                                             i:  Proposed Rule. 55 ££ 30370-30448.
1.    1990 National Primary and Secondary Drinking Water Regulati
     USEPA (July 25).
2.    Maximum r/«t.»iM>n Levels for Organic and Inorganic Chemicals. 45 fB 57332.  USEPA (August 27)
3.    199 1 National Primary and Secondary Drinking Water Regulation*: Final Rule. 56 fB 3526-3597. USEPA
     (January 30).
4.    EPA's Integrated Risk Information System (IRIS) (verified health-based levels).
5.    Maximum r****™**** Level Goals and National Primary Drinking Water Regulation for Lead and Copper:
     Final Rule 56 fB 26460.
6.    1990 Health Effects Assessment Summary Table. Third Quarter OERR, 9200.b-303-(90-3).
 25254108.01\sec7
                                             7-13

-------
Given the Agency's goal is to minimize lead exposure among sensitive populations, the treatment
standard with an action level was established. While the action level is not a formal MCL, EPA
stated in the preamble to the lead rule that the level of 0.015 mg/1 is "associated with substantial
public health protection." (See 56 FR 26477.)

     Although the existing lead MCL of 0.05 mg/1 will remain in effect until November 9,1992,
the Agency believes the use of this  level in setting the exclusion level is inappropriate.  The
effective date  for the action level and accompanying treatment standard for lead were delayed
in order to allow public drinking water systems sufficient time to comply with this new rule.
The Agency believes that establishing exclusion levels using an old MCL that will  soon  be
superseded by a more stringent standard  is not sufficiently protective of public health.

       As previously stated, the BDAT standard and EPACML-based levels are not identical,
since each set was calculated for a different purpose:  the BDAT standards are technology-based
levels, while the EPACML results are derived  from health-based modeling.  In order to  be
eligible for the generic exclusion, the residues  must meet the concentration levels shown  in
Table 7-3.

       For five of these metal constituents (barium, chromium, mercury, selenium, and silver),
the technology-based treatment standards are slightly lower than the health-based levels.  EPA
does not regard these values as significantly different; however, the difference ranges from 0.011
ppm (mercury) to 2.4 ppm for barium.   In addition, since HTMR is a destruction technology
for cyanide, the Agency has chosen to regulate total cyanide instead of leachable cyanide as the
generic exclusion level.  Given that the Agency is excluding these wastes generically, rather than
after a more individualized examination as part of a facility-specific delisting, EPA believes that
it is appropriate to  use the slightly lower value for this exclusion.  It should be noted that this
action is consistent  with the  Agency's position in the Third Third rule, where it maintained that
land disposal prohibitions can apply to wastes that are hazardous when they are generated, even

 *
2S254108.01\sec7                            7-14

-------
if they are not hazardous  after disposal.   (See 55  FR 22652-22653.)  However, EPA is not
invoking that principle to justify its decision here,  given that the exclusion is generic and the
values practically equivalent in any case. Because of the large discrepancy (> tenfold) between
the health-based and technology-based levels for zinc (i.e., 70 mg/1 opposed  to 5.3 mg/1), the
Agency has decided to designate the health-based level as the generic exclusion requirement.

7.2  Product Uses  of HTMR Residues

       The generic exclusion of K061, K062, and F006 HTMR residues applies only to residues
which are disposed of in  Subtitle D units (i.e., landfills or piles).  However, the majority of
these slags are not landfUled, but rather are used in a  manner constituting disposal as road base
material or (less often) as an anti-skid  material (56 FR 15024).  The Agency has decided that
its regulatory tools for evaluating road base and anti-skid uses are too uncertain to determine
whether residues used as  road base or anti-skid material  should be excluded. The EPACML
model evaluates possible risks posed by landfill disposal.  It may also be suitable for evaluating
residue used as a road base material, since this situation may be viewed as similar to (or more
protective than) a  capped landfill.  The  Agency  has not had time to make a full technical
 assessment of this  point.  The EPACML model alone may  not be fully suitable for evaluating
 the safety of slag used as an anti-skid material, because  this apparently  uncontrolled use may
 present exposure pathways (i.e., airborne inhalation and surface runoff) that the model does not
 consider.  Thus, the exclusion levels apply  only  for those modes  of management that EPA
 currently feels confident in  evaluating with the EPACML  model, namely, disposal in a land
 disposal unit.

        Under current regulations, if a hazardous waste is used in a manner constituting disposal,
 it is exempt from further regulation, provided  it undergoes a chemical reaction so as to be
 inseparable by physical means and provided it meets the  land disposal restrictions treatment
 standards for each hazardous constituent that it contains  (40 CFR 266.20).  Thus, in this final
  25254108.01\ttc7
                                           7-15

-------
  rule, such practices as the use of the KTMR residue for road base or anti-skid material are not
  immediately prohibited (provided the residue meets the treaunent standard and any easting state
  requirements).

         This case differs from other delistings in that the HTMR K061 residuals would be used
  in a manner constituting disposal, raising the concern that a ground-water model alone may not
  be adequate to simulates a worst-case scenario. (EPA notes that it has considered air-blown dust
  exposure pathways in past delistings, but views the situation presented in this rulemaking as
  different.  Previous situations involved possible exposures  from air-bom losses under land
  disposal conditions, whereas, today's action potentially involves continual deposit of waste over
  a wide expanse of road systems.)
 7.3
        The generic exclusion for K061 , F006, and K062 HTMR residues that meet the exclusion
 levels (in Part 261) and that do not exhibit any hazardous characteristics is limited, as already
 discussed, to such waste that is disposed of in Subtitle D units. Because K061 , K062, and F006
 HTMR residues are still defined as hazardous at the point of initial generation, EPA believes that
 tracking and certification are needed to ensure proper handling.  This final rule is amending 40
 CFR Section 268.9(d) and 261.3(c)(2)(ii)(C) to require that the generator or treater prepare the
 notification and certification for the initial shipment only, place one copy in the generator's or
 treater's own files, and send another copy to the appropriate EPA region or authorized state.
 The documentation must be retained by the generator or treater for at least five years.   The
 notification and certification must be updated if the process or operation generating the waste
 changes  and/or if the subtitle D facility receiving the waste changes.   The document must
 include the name and address of the subtitle D facility receiving the waste, a waste description,
 applicble treatment standards, and a certification that the standards have been met. For K061,
25254108.01\«ec7
                                         7-16

-------
K062,  and F006 high residues from  high temperature metal recovery, the recordkeeping
requirements  in 40 CFR 261.3(c)(2)(ii)(C) supersede those in 268.7(a)(6).

7.4  Testing Requirements

       The land disposal restriction program imposes site-specific testing requirements in order
to verify that regulatory requirements have been satisfied. Treatment facilities which wish to
meet the exclusion requirements must test treated wastes at a  frequency specified in their waste
analysis  plan in order to determine  whether they have met the  exclusion levels.   (See
40 CFR 268.7(b) and 55  FR  22669.)  In the case  where treatment is performed at the
generator's site in a way not requiring  a permit, testing is required at a frequency specified in
the self-implementing waste analysis plan required by 40 CFR 268.7(a)(4).  However, at a
minimum, a  facility's waste analysis plan (or a generator's self-implementing waste analysis
plan) must specify that composite samples of the K061, K062, and F006 HTMR  slag residues
be collected and analyzed quarterly and/or when the process or operation  changes.  (See 40 CFR
264.13(a)(3)  and 265.13 (a)(3).)  The Agency believes  that it is appropriate  to allow the
frequency of testing, beyond the quarterly minimum, to be determined in the waste analysis plan,
taking into account facility-specific factors such as waste types, waste variability, quantity, batch
size, and type of treatment unit.  The  Agency believes that permit writers will consider  these
factors when establishing testing conditions in  the waste analysis plans.

       The following sections, beginning with Section 7.4.1, present examples of testing criteria
that could be followed for residues to be considered "generally excluded" or nonhazardous that
are  generated from:  (1) the thermal treatment of electric arc furnace dust  (EPA Hazardous
Waste No. K061), originating from  the primary production of steel; (2) the thermal treatment
of wastewater  treatment  sludges (EPA  Hazardous Waste No.  F006)  from the  following
 processes: (a) common and precious metals electroplating except tin, zinc (segregated basis),
 aluminum, and zinc-aluminum plating on  carbon steel,  (b)  anodizing, except sulfuric and


 25254108.01\»c7                            7-17

-------
anodizing of aluminum, (c) chemical etching and milling, except when performed on aluminum,

and (d) cleaning and stripping, except when associated with tin, zinc, and aluminum plating on

carbon steel; and (3) the thermal treatment of spent pickle liquor (EPA Hazardous Waste No.

K062).   The exclusion for these wastes is conditioned upon  the receipt and review of data

obtained from the facility's HTMR unit after it is established as an operational treatment furnace

and from each additional HTMR unit that may be established in the future. To ensure that each

HTMR unit operates properly and that hazardous constituents are not present in the generated

residual  slag at  levels  of regulatory  concern once the HTMR unit  has been established, the

facility  must implement a testing  program for each HTMR  unit.   The following example

provides one approach  in establishing these testing conditions.  Depending on the specific case

(e.g., expected variability  in waste feed, waste volume treated), testing frequency could be
reduced, but must be completed on at least a quarterly basis.


7.4.1  Operating Conditions (Condition 1)


       The  facility would submit information to the Agency pertaining to the design and
operation of the  HTMR unit as stated below. This requirement is referred to as  Condition 1.


       (A)   Initial Verification Testing:  During the first 20 days of operation of an on-line,
             full-scale HTMR  unit, (as an operational treatment furnace), the  facility  must
             monitor and submit to EPA the reclamation furnace design operating conditions
             (including, but not limited to: temperature range of the furnace; EAF dust feed
             rate and  composition; carbon source feed rate; oxygen/air feed rate; target feed
             compositions and feed rates; information concerning composition and feed rate of
             other materials added' to the system; detailed  information about the destiny of all
             other residuals (i.e., where are they sent for further recovery, time frame, and
             storage/handling procedures); and reclamation furnace reaction time of the raw
             materials).   This information on operating conditions should  encompass all
             conditions used for preliminary testing runs and those anticipated for subsequent
             waste processing.  During initial verification testing, the  petitioner must also
             demonstrate to EPA how  the range of operating conditions could affect the
             process (i.e., submit analyses of representative grab samples, as specified under
             Condition 2, of the residual slag generated under the expected range of operanng


2525410S.01\Mc7                           7-18

-------
             conditions).  The facility must submit the information specified in this condition
             and obtained during this initial period no later than 90 days after the first full-
             scale treatment of untreated EAF dust (K061).

       (B)    Subsequent Verification Testing:   During subsequent verification  testing, the
             facility must monitor the performance of the HTMR unit at all times to ensure
             that it falls within the range of operating conditions demonstrated during initial
             verification testing to be adequate to maintain the levels of hazardous constituents
             below the delisting  levels  specified in .Condition 4.   Records of the operating
             conditions of the reclamation furnace (including, but not limited to:  temperature
             range of the furnace; waste feed rate and composition; carbon source feed rate;
             oxygen/air feed rate;  target feed compositions and  feed rates;  information
             concerning composition and feed  rate of other materials  added  to  the system;
             detailed information about  the destiny .of all other residuals (i.e.,  where are they
             sent for further recovery, time frame,  and storage/handling  procedures);  and
             reclamation furnace reaction time of the raw materials should be maintained onsite
             for a minimum of 3  years.  This information must be furnished upon request and
             made available  for inspection by any employee or representative of EPA or the
             State where the HTMR unit is located.
7.4.2  Testing (Condition 2)


       Sample collection and  analyses (including quality control  (QQ procedures)  must be

performed according to SW-846 and BDAT methodologies.  These requirements are discussed

below and are referred to as Condition 2.
       (A)   Initial Inorganics Verification Testing:  During the first 20 days of operation of
             an on-line, full-scale HTMR unit (as an operational treatment furnace), the facility
             must collect and analyze daily composites of residual slag. Daily composites must
             be composed of representative grab samples collected every 3 hours during each
             24-hour furnace operating cycle.  The residual slag samples must be analyzed,
             prior to the disposal of the residual slag, for the inorganic constituents listed in
             Condition  4.  The facility must report the analytical test data,  including QC
             information, obtained during this initial period no later than 90 days after the first
             full-scale treatment of untreated EAF dust (K061).

       (B)   Subsequent Verification Testing: Following the initial 20-day testing period and
             provided that the results of initial  verification testing support an  exclusion, the
 25254108.01\«c7                            7-19

-------
              facility must collect and analyze weekly composites of residual slag.  Weekly
              composites must be composed of representative grab samples collected every 8
              hours during a 24-hour furnace operating cycle.  These samples must be analyzed,
              prior to the disposal of the residual slag, for the teachable concentrations of the
              inorganic constituents listed in Condition 4.  The analytical data, including QC
              information,  must  be compiled, summarized, and maintained  onsite for  a
              minimum of 5  years.  These data must be furnished  upon request  and made
              available for inspection by any employee or representative of EPA or the State
              where the HTMR unit is located.
              Changes in Operating  Conditions:  If after completing the initial verification test
              period in Conditions 1(A) and 2(A), the facility changes the operating conditions
              developed under Condition 1(A), then the facility must repeat the testing required
              in Conditions 1(A) and 2(A) with the new conditions. Following this period, the
              facility must collect and analyze weekly samples according to Condition 2(B).
7.4.3  Waste Holding and Handling (Condition 3)

       The facility will be required to store, as hazardous, all HTMR residual slag generated
until it has completed and compared initial verification testing (for daily composites as specified
in Condition 2(A)) or subsequent analyses (for weekly composites as specified in Condition
2(B»,  as appropriate, with the delisting levels set forth in Condition 4.  If the  levels of
hazardous constituents measured in the samples of HTMR residual generated do not exceed the
levels set forth in Condition 4, then the HTMR residual is nonhazardous for purpose of disposal
in a Subtitle D landfill and may be managed and disposed of in accordance with all applicable
solid waste regulations.  If hazardous constituent levels in any sample exceed any of the generic
exclusion levels set in Condition 4, the HTMR residual generated during the  time  period
corresponding to this sample must be retreated until it meets these levels (analyses must be
repeated) or managed and disposed of  in accordance with Subtitle C  of RCRA.   Residual
generated for which the required analysis  is not complete or valid must be managed and disposed
of in accordance with Subtitle C of RCRA until valid analysis demonstrates that Condition 4 is
satisfied.
25254108.0l\*ec7                            7-20

-------
7.4.4  Exclusion Levels (Condition 4)

       For K061, K062, and F006 HTMR residues, the TCLP leachate concentrations for the
metals may not exceed the following levels (mg/1):  antimony 0.10; arsenic 0.055; barium 7.6;
beryllium 0.010;  cadmium 0.05; chromium 0.33; lead 0.15; mercury  0.0090; nickel 1.0;
selenium 0.16; silver 0.30; thallium 0.02; zinc 70; and cyanide 1.8. Metal concentrations must
be measured in the waste leachate by the TCLP method, except for cyanide for which analysis
must be based on total concentration (mg/kg).

7.4.5  Data Submittals (Condition 5)

       At least 6 months prior to operation of a new HTMR unit, the facility must notify EPA
when the HTMR unit is scheduled to be on-line.  Relevant information including, but not limited
to, design and proposed operation parameters, projected average  annual waste  generation
volume,  and specific areas where the new HTMR unit differs from the facility's other HTMR
unit(s) on file must be submitted to EPA within the time period specified.  At EPA's request,
the facility must submit any other analytical data obtained through Conditions  1(B) and 2(B)
within the time period specified. Failure to submit the required data within the specified time
period or to maintain the required records onsite for the specified time will be considered by the
Agency, at its discretion, sufficient basis, to revoke the exclusion or exclusion amendment to the
extent directed by EPA.   All data  must be accompanied  by a signed copy  certifying  all
 statements attest to the truth and accuracy of the data submitted.

 7.5 Apniicabiiitv of c^ngrte Exclusion to Other Treatment Residues
        The generic exclusion applies only to those nonwastewater residues generated by HTMR
 processes and not to others, such as hydrometallurgical processes or stabilization. The Agency
 has insufficient data to fully evaluate the residues from hydrometallurgical processes.  However,
 2S2S4108.01\Mc7                           7-21

-------
the limited available information indicates a high leachability.  Moreover, given the Agency's
current paucity of information, EPA has no idea what an appropriate testing regime for residues
from hydrometallurgical processes would be, even assuming that these residues could meet the
exclusion levels. EPA, thus, believes it unwarranted to make residues from hydrometallurgical
recovery processes eligible for this generic exclusion at this time.

       There are several reasons for not excluding stabilized residues generically. The HTMR
residues demonstrate  consistent  leaching behavior, whereas, stabilised  matrices  are  quite
variable.  The chemical bonding  that occurs in the high temperature and oxidation/reduction
conditions within the HTMR units is inherently different than the bonding that forms the basis
of cementitious and  pozzolanic stabilization. In addition, the kinetics of the reaction forming
the bonds in these HTMR processes are superior  to the  kinetics of bond  formation  in
cementitious reactions. (Cement is  not typically considered set until after a minimum of 72
hours and often not considered fully cured until after 28 days.)  Stabilization has  also been
documented as a process that is highly matrix-dependent and prone to chemical interference.
(Data in support of this conclusion are located in the background documents to the First, Second,
and Third Third Rules.)  Most commercial stabilization facilities have to develop special mixes
to control curing time  and/or product integrity (often measured by comprehensive strength).

        Another reason for  not allowing stabilized residues to be  generically excluded is the
possibility of impermissible dilution, which must be considered on a case-by-case basis with
stabilization, but not with HTMR.  Hence, facility-specific delistings are preferred for stabilized
wastes  so  that the Agency can evaluate waste-to-binder and waste-to-waste ratios and make a
determination about treatment versus dilution.  Finally, the Agency believes that HTMR is a
preferred  technique for  managing the  K061, K062,  and  F006  waste over stabilization
technologies, in light of its resource recovery potential  and the differences in volumes of treated
wastes.  Stabilization  generally increases volume, while HTMR generally decreases volume.
 25254108.01\sec7                            7-22

-------
Thus, the Agency does not believe it is warranted to develop a somewhat technically sketchy
generic exclusion for stabilization.

       EPA notes that it is not precluding the use of stabilization as treatment for K061, K062,
and F006 wastes, and that facility-specific delisting remains an option for stabilized K061, K062,
and F006 wastes.  However, due to the inherent differences between HTMR and stabilization
stated above and the fact that insufficient data currently exist to promulgated a generic exclusion
for stabilized K061, K062, and F006 wastes,  the Agency has determined that the generic
exclusion levels are not applicable to stabilized K061 residues.  The Agency believes that more
individualized consideration of stabilization is warranted before residues from the process are
delisted.

     The Agency, however, wishes to note that the issue of uniform exclusion levels is presently
under consideration by the Agency as part of the Hazardous Waste Identification Rule (HWIR).
(See 57 FR 21450, May 20, 1992).
7.6  Regulatory Status ftf NpnwMtewater Residues From HTMR Tfrgf PQ ftfot Meet the
     Generic Exclusion Levels
       Under the Federal regulations, hazardous wastes destinffd  for reclamation  remain
 classified as solid and hazardous wastes until reclamation is complete.  Reclamation is normally
 incomplete until the end product of the process is fully recovered (50 FR 15 633,634, and 655).
 The line the Agency has traditionally drawn between partially and fully reclaimed material when
 thermal  mercury recovery is involved is that secondary material* remain wastes until smelting
 is completed.  Id- it 634 (recovered metals only needing to be refined [the processing step
 following smelting] are products, not wastes). This interpretation is consistent with RCRA's
 cradle-to-grave mandate by retaining authority until a usable metal is recovered (Cf. API v EPA,
 906 F.2d at 741).
 25254108.01\Mc7                           7-23

-------
 7.6.1  Application of the Variance From Solid Waste Classification

        The rules also provide for a variance from solid waste classification for materials that
 have been partially but not fully reclaimed (40 CFR 26i.30(c)). Criteria for granting a variance
 include the  degree of processing that  the material has undergone, the degree  of further
 processing required, the value of the material after it has been reclaimed, the degree of which
 the initially reclaimed material is like an analogous raw material, the extent to which an end
 market for the material is guaranteed, and (perhaps most important) the extent to  which the
 initially reclaimed material is handled to  minimize loss (40 CFR 260.31(c)).

        Applying these rules to the dross  from HTMR splash condensers, EPA has decided to
 amend its rules by excluding from Subtitle C jurisdiction the splash condenser dross residue
 (hereafter referred to  SCDR) generated by  certain HTMR  processes.  This material is
 specifically generated  as the nonproduct skimming from the splash condenser along  with
 recovered zinc and  lead meeting Western  grade zinc metal spedfu^CMO.e., 98 percent pure
 metals), which are products under the rules (40 CFR 261.3(c)(2) final sentence).  The dross is
 presently  a solid waste because it is partially but not fully reclaimed (i.e., it still requires
 smelting or other recovery before a usable metal is extracted) and, mus, wmildreinain a K061,
 K062, and F006 waste  until it is excluded from the rules. (See 40 CFR 261.2(a)(l) and 56 FR
 at 7144.)  Based on pubUc  wnunent aiid conoboratiiig iiiform^
 today's rule, the SCDR is collected directly from the splash ccfldenser and drummed.  It is then
 stored for short periods (not exceeding 2 weeks) and sold to a thermal zinc processing facility
 where it is used  as a  source of zinc or  reused onsite in the HTMR process.   CH« SCDR
 normally  contains SO to 60 percent zinc.) At  the thermal processing facility, the drums are
 stored indoors in a secure manner (on concrete flooring  and with controls against  airborne
 migration). The material is then processed for recovery by crushing and, in combination with
other feedstocks, grinding, and by thermal recovery of zinc.
25254108.01\*c7                            7-24

-------
       The SCDR stream is small  in volume.   In addition,  most of the toxic metals that
originate in the K061 do not partition to the SCDR:  approximately 90 percent partition to zinc
and lead products-or to baghouse dusts.  Those toxic metals remaining in the SCDR have
reduced mobility from the original K061, (and EPA believes is similarity reduced in K062, and
F006).  The SCDR does not exhibit characteristics of hazardous waste.  In the case of K061
nonwastewaters, the (EAF dust) SCDR is also changed in physical form from the original K061;
it is no longer a dust but rather a solidified matrix.

       The Agency evaluated the material against the criteria for determining whether a waste
that is partially but not fully reclaimed  should still be classified as a solid waste (40 CFR
260.31(c)). Although these criteria were established for a variance determination, EPA believes
that they  are relevant  in determining whether  this material  should be considered to be
•discarded" within the  meaning of Section 260.31(c)(l). The Agency has received adequate
information in this case to exclude the material by rule.  Table 7-5 presents TCLP leachate
analysis data  for International Mill Service Oxidized Zinc Material (OZM), i.e., dross residue
from its zinc splash condenser.  In  particular, the Agency finds that the SCDR results from
substantial processing (as shown by the volume reduction, partitioning of toxic metals to other
outputs of the process,  change in physical form, and reduction in mobility of toxic metals) (40
CFR 260.31(c)(l)); that the material Is sold for the value (or reprocessed onsite to recover high
concentrations of zinc) (40 CFR 260.31(c)(l)); that the material contains zinc concentrations
comparable to those of other nonwaste secondary sources of zinc  (and more zinc than natural
ones) (40 CFR 260.31(c)(3)); mat an end market for the material appears assured (40 CFR
260.31(c)(4)); and that it is handled safely up  to  the point  of final reclamation (40 CFR
260.31(c)(5)).
 25254108.01\Mc7                           7-25

-------
                               Table 7-5  Treatment Performance Data (TCLP Analysis) for Residues (Dross),

                       i.e., Oxidized Zinc  Material (OZM) from HTMR of K061 High Zinc Subcategory Nonwastewater

feu
1
2
9
4

Aa
<0.01S
<0.025

o\

-------
    Based on these facts, the Agency has decided to exclude the K061, F006, and K062 SCDR
from RCRA jurisdiction when it is utilized as a source of zinc in the zinc recovery operations,
provided it is shipped in drums  (if it is sent offsite),  and that there is no land disposal of the
material before it is  recycled.  Thus, for example, the material remains a solid waste if it is
stored in piles on the land.  In such a case, it would be "part of the waste disposal problem,'
and discarded,  rf American Mining Congress v. EPAV 907 F.2d at 1186.] In addition, in order
for this exclusion to  be implementable and to serve as a check against mishandling,  EPA is
interpreting "current rules to require that the HTMR facility maintain a one-time notice in its
operating record or other files stating that the SCDR is generated, then excluded, and what its
disposition is. (See 40 CFR 268.7(a)(6), 56 FR 3878.)

7.6.2  Application of the Derived-From Rule

       The Agency is, hereby, classifying the application of the derived-from rule to all residues
from HTMR that do  not meet the generic exclusion levels or qualify for a variance from solid
                                                       s
waste classification.  In general, two categories of HTMR materials have been identified by the
Agency.  These categories are:  (1)  nonwastewater residues (including slag) into which a
significant portion of the relatively nonvolatile hazardous metals have partitioned and  that are
generated from a furnace from which zinc and other  relatively volatile hazardous metals have
been separated,  and  (2) residues generated by removing particulates from the furnace off-gas
such as baghouse dusts and scrubber sludges into which volatile hazardous constituents (e.g.,
lead and cadmium) have partitioned.

        Application of the  derived-from rule to residues such as slag generated during HTMR
of K061, K062, and F006 is soundly based because toxic nonvolatile metals (such as chromium,
nickel, and vanadium) partition into these residues. Because of the high operating temperatures
of the HTMR furnaces (approximately 1,600°Q, low boiling-point inorganic constituents present
 in K061, K062, and F006,  such as zinc, lead, and cadmium volatilize and are subsequently
 recovered, while high-boiling point constituents such as chromium, nickel, barium, iron, and

 25254108.01\Mc7                           7-27

-------
silica remain in the molten mixture which is typically considered slag.  (See Section 5, Table
5-2 for details about the behavior of the 13  metal constituents in HTMR processes.) Since the
latter residues are typically land disposed or used on the land (e.g., as road aggregate or as road
slippage material) and since the hazardous metals contained in them have the potential to leach
from these residues, the Agency believes these residues are  appropriately considered K061,
K062, and/or F006 unless they meet the generic exclusion levels for HTTMR nonwastewater
residues.

        Typically, all HTMR units use some form of Air Pollution Control Device (APCD) to
capture paniculate matter present in the off-gases.  Two of the  most popular devices, baghouses
and wet scrubbers, generate residuals.  These HTMR residues contain the low-boiling point
metals (e.g., baghouse dusts and scrubber sludges) and would also be considered listed wastes
by virtue of the derived-from rule.  These residues,  although rich in desirable metals, often
contain contaminant constituents that must be removed by pretreatment processes before the
 material can be sent for refinement or used as a feedstock.
  25254108.01\Mc7
                                           7-28

-------
                                 8.  REFERENCES
American Iron and Sted Institute.  September 1986.  Study to evaluate impact of EAF dust land
disposal on human health and the environment.

Bethlehem Steel Corporation. May 1985. Final report, project number RF-2570-1-2.  Electric
arc furnace dust-disposal, recycle and recovery.

Center for Metals Production. (CMP Report No. 88-1). "Flame Reactor Process for Electric
Arc Furnace Dust" prepared by Horsehead Resource Development Company Inc.  August 1988.

Iron and Steel Engineer. Kotraba, Norman L.  Inclined Rotary Reduction System for Recycling
Electric Arc Furnace Baghouse Dust, August 1991.

Institute of Scrap Recycling Industries, Inc. Scrap Specifications Circular 1990. Guidelines for
Ferrous Scrap Nonferrous Scrap Paper Stock. 1990

Kirk-Qthmer Encyclopedia of Chemical Technology. Third Edition,  1978.  John Wiley and
Sons, NY.

N.H.  Analytical Software.  "STATISTK" (an interactive statistical analysis  program for
minicomputers).

Recontek Company.  1990.  Technical  information submitted to EPA concerning recycling,
alternative treatment processes for F006 and D002 wastes..

Recycled Metals of the 1980*8. by the National Association of Recycling Industries, 1982.

Samsonov, G. C., "The Oxide Handbook/ 1973. Plenum Press, NY

 SKF Plasma Technologies.  1987. Key data for the Scandust Plant for treating EAF flue dust
 (K061). August 1987 (industry-submitted data).

 Squire, Sanders, and Dempsey.  1990.  Information and treatment performance data submitted
 to EPA concerning Horsehead Resource Development Co. Inc.'s recycling of K061 waste
 stream.

 st~i ratings Handbook. Fifth Edition,  Peter F. Weiser, Editor, by the Steel Founders' Society
 of America, 1980.
  25254108.01\iec8

-------
Steel Products Manual, by the American Iron and Steel Institute, 1982.  Swidler and Berlin.
1990. Information and treatment performance data concerning International Mill Service, Inc.,
for K061 waste stream.

USDOC.  1982.  U.S. Department of Commerce.  Characterization, recovery and recycling of
electric arc furnace dust.  Prepared by Lehigh University.

USEPA.  N.D. U.S. Environmental Protection Agency.  Office of Health and Environmental
Assessment in the Office of Research and Development.  The Integrated Risk Information
System (IRIS). IRIS is accessible on-line through DIALCOM, Inc., the Public Health Network
(PHN), and the National Library of Medicine's TOXNET.

USEPA.   1980.   U.S.  Environmental Protection Agency.   Background listing document,
November 14, 1980. Washington, D.C.: U.S. Environmental Protection Agency.

USEPA.  1985.  U.S. Environmental Protection Agency.  Characterization of waste streams
listed in 40 CFR Section 261 waste profiles. Volume H.  Prepared by Environ Corporation for
Waste Identification Branch, Characterization as Assessment Division, U.S. Environmental
Protection Agency.

USEPA.  1986a. U.S. Environmental Protection Agency. National survey of hazardous waste
treatment, storage, disposal, and recycling facilities. Washington, D.C.: U.S. Environmental
Protection Agency.

 USEPA.   I986b.   U.S. Environmental  Protection Agency.   Office  of Solid Waste  and
 Emergency Response.  Test methods for evaluating solid wastes: physical/chemical methods.
 SW-846.  3rd Ed.  November 1986.  Washington, D.C.:  U.S. Environmental Protection
 Agency.

 USEPA.  1986c.   U.S. Environmental Protection Agency.  Onsite engineering  report of
 treatment technology performance and operation for Envirite Corporation. Prepared by Versar
 for Office of Solid Waste, U.S. Environmental Protection Agency, under Contract No. 68-01-
 7053.  December  1986.

 USEPA. 1986. U.S. Environmental Protection Agency.  Onsite engineering report of treatment
 technology performance and operation for Envirite Corporation. Prepared by V«r ftjfte
 Office of Solid Waste, U.S. Environmental Protection Agency under Contract No. 68-01-7053.
 December 1986.

 USEPA.  1987a.   U.S. Environmental Protection Agency.   Engineering  analysis of the
 production of electric arc furnace steel. Draft report.
  252S4108.01\sec8                           8-2

-------
USEPA.   1987b.  U.S.  Environmental Protection Agency.  Onsite engineering report for
Horsehead Resource Development Company Inc. for K061.  Draft report.

USEPA.   1987c.  U.S.  Environmental Protection Agency.  Onsite engineering report for
Waterways Experiment Station for K061.  Draft Report.

USEPA.   1988a. U.S. Environmental Protection Agency.  Final best demonstrated available
technology (BOAT) background document for K061.  EPA/530-SW-88-031D. Washington,
D.C.:  U.S. Environmental Protection Agency.

USEPA.   1988b. U.S. Environmental Protection Agency.  Final Administrative Rulemaking
Record for K061.

USEPA.   1988c.  U.S. Environmental Protection Agency, Office of Solid Waste.   Generic
quality assurance plan for Land Disposal Restrictions Program ("BOAT"). Washington, D.C.:
U.S. Environmental Protection Agency.

USEPA.   1988d.   U.S. Environmental Protection Agency.  Onsite engineering report for
Horsehead Resource Development Co., Inc. for K061. Draft report.  Prepared by Versar Inc.
for the Office of Solid Waste, U.S. Environmental Protection Agency, under Contract No. 68-
01-7053.  Washington, D.C.: U.S. Environmental Protection Agency.

USEPA.  1988e.   U.S. Environmental Protection Agency.  Onsite engineering  report for
Waterways Experiment Station for K061. Draft report.

USEPA.  1988f.  U.S. Environmental Protection Agency.  Final best demonstrated available
 technology (BOAT) background document for F006. EPA/530 SW-88-031L.

 USEPA.  1988g.  U.S. Environmental Protection Agency.  Final best demonstrated available
 technology (BOAT) background document for K062. EPA/530 SW-88-031.

 USEPA.  1989. U.S. Environmental Protection Agency, Office of Solid Waste. Methodology
 for developing BOAT treatment standards. Washington, D.C.: U.S. Environmental Protection
 Agency.

 USEPA.  199Qa,  U.S. Environmental Protection Agency,  Office of Solid Waste.  Final best
 demonstrated available technology (BOAT) background  document for organic U and  P and
 multi-source leachate (F039), Volumes A, B, and C. Washington, D.C.: U.S. Environmental
 Protection Agency.

 USEPA.  1990b. U.S. Environmental Protection Agency, Office of Solid Waste.   Final best
 demonstrated available technology (BOAT) background document for K061 (Addendum).
  25254108.01\iec8                          8-3

-------
        1991   U S. Environmental Protection Agency.  SHE Superfund Innovative
         Evaluation   Demonstration Bulletin,  Flame Reactor,  Horseheao Resource
Developmcait Company, Inc.
                                               assa-
(July 25).
 (January 30).
   25254l08.01\»c8

-------