United States
Environmental Protection
Agency
Solid Waste and
Emergency Response
(5305W)
EPA53O R-96-O 7 S
August 1996
4
Study of Selected Petroleum
Refining Residuals
\
Industry Study
Photocopied on recycled paper that contains at least 20 % post consumer fiber
-------
STUDY OF SELECTED
PETROLEUM REFINING RESIDUALS
INDUSTRY STUDY
Parti
August 1996
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Solid Waste
Hazardous Waste Identification Division
401 M Street, SW
Washington, DC 20460
-------
TABLE OF CONTENTS Page Number
1.0 INTRODUCTION 1
1.1 BACKGROUND 1
1.2 OTHER EPA REGULATORY PROGRAMS IMPACTING THE
PETROLEUM REFINING INDUSTRY 2
1.3 INDUSTRY STUDY FINDINGS 3
2.0 INDUSTRY DESCRIPTION 8
2.1 PETROLEUM REFINING INDUSTRY PROFILE 8
2.2 INDUSTRY STUDY 10
2.2.1 Site Selection 11
2.2.2 Engineering Site Visits 13
2.2.3 RCRA §3007 Questionnaire 13
2.2.4 Familiarization Sampling 14
2.2.5 Record Sampling 15
2.2.6 Split Samples Analyzed by API 20
2.2.7 Synthesis 20
3.0 PROCESS AND WASTE DESCRIPTIONS 21
3.1 REFINERY PROCESS OVERVIEW 21
3.2 CRUDE OIL DESALTING 25
3.2.1 Process Description 25
3.2.2 Desalting Sludge 26
3.3 HYDROCRACKING 34
3.3.1 Process Description 34
3.3.2 Spent Hydrocracking Catalyst 35
3.4 ISOMERIZATION 43
3.4.1 Isomerization Process Description 43
3.4.2 Isomerization Catalyst 46
3.4.3 Isomerization Treating Clay 54
3.5 EXTRACTION 58
3.5.1 Extraction Process Description 58
3.5.2 Extraction Treating Clay 59
3.6 ALKYLATION 66
3.6.1 Sulfuric Acid Alkylation Process Description 66
3.6.2 Hydrofluoric Acid Alkylation Process Description 67
3.6.3 Spent Treating Clay from Alkylation 69
3.6.4 Catalyst from Hydrofluoric Acid Alkylation 75
3.6.5 Acid Soluble Oil from Hydrofluoric Acid Alkylation 77
3.7 POLYMERIZATION 82
3.7.1 Process Descriptions 82
3.7.2 Spent Phosphoric Acid Polymerization Catalyst 84
3.7.3 Spent Dimersol Polymerization Catalyst 87
3.8 RESIDUAL UPGRADING 92
3.8.1 Process Descriptions 92
3.8.2 Off-specification Product from Residual Upgrading 96
Petroleum Refining Industry Study 2 August 1996
-------
3.8.3 Process Sludge from Residual Upgrading 97
i.9 LUBE OIL PROCESSING 103
3.9.1 Process Descriptions 103
3.9.2 Treating Clay from Lube Oil Processing 107
i.10 H2S REMOVAL AND SULFUR COMPLEX 112
3.10.1 Process Description 112
3.10.2 Off-Specification Product from Sulfur Complex and H2S
Removal Facilities 117
3.10.3 Off-Specification Treating Solution from Sulfur Complex and
H2S Removal Facilities 124
i.ll CLAY FILTERING 132
3.11.1 Process Description 132
3.11.2 Treating Clay from Clay Filtering 134
i.12 RESIDUAL OIL TANK STORAGE 142
3.12.1 Residual Oil Storage Tank Sludge 142
Petroleum Refining Industry Study 3 August 1996
-------
LIST OF TABLES
Page Number
Table 1.1. Petroleum Refining Residuals Identified in the EDF/EPA
Consent Decree
Table 1.2. Overview of 15 Study Residuals of Concern as Managed in 1992
Table 2.1. Engineering Site Visit Facilities
Table 2.2. Study Residuals Volume Statistics
Table 2.3. Residuals Collected for Record Analysis
Table 2.4. Descriptions of Samples Collected for Record Analysis
Table 3.2.1. Generation Statistics for Desalting Sludge, 1992
Table 3.2.2. Desalter Sludge: Physical Properties
Table 3.2.3. Desalting Sludge Record Sampling Locations
Table 3.2.4. Desalting Sludge Characterization
Table 3.3.1. Generation Statistics for Hydrocracking Catalyst, 1992
Table 3.3.2. Hydrocracking Catalyst Physical Properties
Table 3.3.3. Spent Hydrocracking Catalyst Record Sampling Locations
Table 3.3.4. Spent Hydrocracking Catalyst Characterization
Table 3.4.1. Generation Statistics for Catalyst from Isomerization, 1992 ....
Table 3.4.2. Catalyst from Isomerization: Physical Properties
Table 3.4.3. Spent Isomerization Catalyst Record Sampling Locations
Table 3.4.4. Residual Characterization Data for Spent Isomerization Catalyst
Table 3.4.5. Generation Statistics for Treating Clay from Isomerization, 1992
Table 3.4.6. Treating Clay from Isomerization: Physical Properties
Table 3.4.7. Isomerization Spent Sorbent Record Sampling Locations
Table 3.5.1. Generation Statistics for Treating Clay from Extraction, 1992 . .
Table 3.5.2. Treating Clay from Extraction: Physical Properties
Table 3.5.3. Extraction Spent Sorbent Record Sampling Locations
Table 3.5.4. Residual Characterization Data for Spent Treating Clay from
Extraction/Isomerization
Table 3.6.1. Generation Statistics for Treating Clay from Alkylation, 1992 . .
Table 3.6.2. Treating Clay from Alkylation: Physical Properties
Table 3.6.3. Alkylation Treating Clay Record Sampling Locations
Table 3.6.4. Alkylation Treating Clay Characterization
Table 3.6.5. Generation Statistics for Catalyst from HF Alkylation, 1992 ....
Table 3.6.6. Catalyst from HF Alkylation: Physical Properties
Table 3.6.7. Generation Statistics for Acid Soluble Oil, 1992
Table 3.6.8. Acid Soluble Oil: Physical Properties
Table 3.6.9. Acid Soluble Oil Record Sampling Locations
Table 3.6.10. Acid Soluble Oil Characterization
Table 3.7.1. Generation Statistics for Phosphoric Acid Catalyst from
Polymerization, 1992
Table 3.7.2. Phosphoric Acid Catalyst from Polymerization: Physical
Properties
Table 3.7.3. Phosphoric Acid Polymerization Catalyst Record Sampling
Locations
Table 3.7.4. Generation Statistics for Spent Dimersol Polymerization
Catalyst, 1992
2
5
12
15
16
17
27
30
30
31
37
38
39
40
48
49
49
51
55
57
57
61
63
63
64
70
72
72
73
76
76
78
79
79
80
85
86
86
88
Petroleum Refining Industry Study
August 1996
-------
Table 3.7.5. Spent Dimersol Polymerization Catalyst Physical Properties 89
Table 3.7.6. Dimersol Polymerization Catalyst Record Sampling Locations 89
Table 3.7.7. Polymerization Catalyst Characterization 90
Table 3.8.1. Generation Statistics for Off-Specification Product from Residual
Upgrading, 1992 96
Table 3.8.2. Off-Specification Product from Residual Upgrading: Physical
Properties 97
Table 3.8.3. Generation Statistics for Process Sludge from Residual
Upgrading, 1992 98
Table 3.8.4. Process Sludge from Residual Upgrading: Physical Properties 100
Table 3.8.5. Process Sludge from Residual Upgrading Record Sampling
Locations 100
Table 3.8.6. Process Sludge from Residual Upgrading Characterization 101
Table 3.9.1. Generation Statistics for Treating Clay from Lube Oil, 1992 108
Table 3.9.2. Treating Clay from Lube Oil: Physical Properties 109
Table 3.9.3. Treating Clay from Lube Oil Processing Record Sampling
Locations 109
Table 3.9.4. Treating Clay from Lube Oil Processing Characterization 110
Table 3.10.1. Sulfur Removal Technologies Reported in RCRA §3007
Questionnaire 112
Table 3.10.2. Generation Statistics for Off-Spec Sulfur, 1992 119
Table 3.10.3. Off-Specification Sulfur: Physical Properties 121
Table 3.10.4. Off-Specification Sulfur Record Sampling Locations 121
Table 3.10.5. Residual Characterization Data for Off-Specification Sulfur 122
Table 3.10.6. Generation Statistics for Spent Amine for H2S Removal, 1992 125
Table 3.10.7. Generation Statistics for Stretford Solution for H2S Removal,
1992 126
Table 3.10.8. Spent Amine: Physical Properties 127
Table 3.10.9. Spent Stretford Solution: Physical Properties 128
Table 3.10.10. Off-Specification Treating Solution Record Sampling Locations 128
Table 3.10.11. Characterization Data for Off-Specification Treating Solution
from Sulfur Complex and H2S Removal 130
Table 3.11.1. Generation Statistics for Treating Clay from Clay Filtering, 1992 136
Table 3.11.2. Treating Clay from Clay Filtering: Physical Properties 137
Table 3.11.3. Treating Clay Record Sampling Locations 137
Table 3.11.4. Residual Characterization Data for Treating Clay 139
Table 3.12.1. Generation Statistics for Residual Oil Tank Sludge, 1992 144
Table 3.12.2. Residual Oil Tank Sludge: Physical Properties 146
Table 3.12.3. Residual Oil Tank Sludge Record Sampling Locations 146
Table 3.12.4. Residual Oil Tank Sludge Characterization 147
Petroleum Refining Industry Study
August 1996
-------
LIST OF FIGURES Page Number
Figure 2.1. Geographical Distribution of U.S. Refineries 9
Figure 3.1. Simplified Refinery Process Flow Diagram 22
Figure 3.2.1. Desalting Process Flow Diagram 25
Figure 3.3.1. Hydrocracking Process Flow Diagram 35
Figure 3.4.1. Isomerization Process Flow Diagram 43
Figure 3.5.1. Extraction Process Flow Diagram 58
Figure 3.6.1. H2SO4 Alkylation Process Flow Diagram 66
Figure 3.6.2. FTP Alkylation Process Flow Diagram 68
Figure 3.7.1. Process Flow Diagram for Phosphoric Acid Polymerization
Process 83
Figure 3.7.2. Dimersol Polymerization Process Flow Diagram 84
Figure 3.8.1. Solvent Deasphalting Process Flow Diagram 93
Figure 3.8.2. Asphalt Oxidation Process Flow Diagram 94
Figure 3.8.3. Supercritical Extraction Process Flow Diagram 95
Figure 3.9.1. Lube Oil Processing Flow Diagram 104
Figure 3.10.1. Amine Sulfur Removal Process Flow Diagram 113
Figure 3.10.2. Claus Sulfur Recovery Process Flow Diagram 114
Figure 3.10.3. SCOT® Tail Gas Sulfur Removal Process Flow Diagram 115
Petroleum Refining Industry Study 6 August 1996
-------
1.0 INTRODUCTION
1.1 BACKGROUND
The U.S. Environmental Protection Agency (EPA) is directed in section 3001(e)(2) of
the Resource Conservation and Recovery Act (RCRA) (42 U.S.C. §6921 (e)(2)) to determine
whether to list as hazardous wastes a number of different wastes including those of the
petroleum refining industry. A lawsuit by the Environmental Defense Fund (EDF) in 1989
resulted in a consent decree approved by the court, that sets out an extensive series of deadlines
for making the listing determinations required by Section 3001 (e)(2). The deadlines include
those for making final listing determinations as well as for concluding various related studies or
reports on the industries of concern. With respect to the refining industry, the consent decree
identifies 14 specific residuals for which the Agency must make listing determinations and an
additional 15 residuals for which the Agency must conduct a study. These 29 residuals,
subsequently referred to as the Residuals of Concern (RCs), are listed in Table 1.1. As a result
of the consent decree, the Agency embarked on a project to determine whether these 29 RCs
pose a threat to human health and the environment and to develop a basis for making such a
determination. As a result of the preliminary evaluation of the waste subject to the listing
determination, EPA proposed a rule in which eleven wastes were not to be listed and three
wastes were to be listed as hazardous wastes: K169, K170, and K171 (clarified slurry oil storage
tank sediments and/or filter/separation solids from catalytic cracking, catalyst from
hydrotreating, and catalyst from hydrorefming, respectively) (60 FR 57747, November 20,
1995). The final determination will be issued under the applicable terms of the consent decree.
This report is the result of the Agency's study of the remaining 15 residuals.
The Petroleum Refining Industry was previously studied by OSW in the 1980s. This
original effort involved sampling and analysis of a number of residuals at 19 sites, distribution of
a RCRA §3007 questionnaire to 180 refineries (characterizing the industry as of 1983), and,
ultimately, a listing determination effort focused on wastewater treatment sludges, culminating
in the promulgation of hazardous waste listings F037 and F038 (respectively, primary and
secondary oil/water/solids separation sludges from petroleum refining).
As part of the Agency's current investigation of residuals from petroleum refining, the
Agency conducted engineering site visits at 20 refineries to gain an understanding of the present
state of the industry. These 20 refineries were randomly selected from the 185 refineries
operating in the continental United States in 1992. Familiarization samples of various residuals
were collected at 3 of the 20 refineries to obtain data on the nature of the RCs and to identify
potential problems with respect to future analysis. The Agency then conducted record sampling
and analysis of the RCs. During the record sampling timeframe, an additional 6 facilities were
randomly selected to increase sample availability. Approximately 100 record samples were
collected and analyzed. Concurrently, the Agency developed, distributed and evaluated a RCRA
§3007 survey to the 180 refineries in the U.S.
Petroleum Refining Industry Study 1 August 1996
-------
Table 1.1. Petroleum Refining Residuals Identified in the EDF/EPA Consent Decree
Listing Residuals
Clarified slurry oil sludge from catalytic cracking
Unleaded storage tank sludge
Crude storage tank sludge
Process sludge from sulfur complex and H2S removal facilities (sulfur complex sludge)
Sludge from HF alkylation
Sludge from H2SO4 alkylation
Catalyst from catalytic hydrotreating
Catalyst from catalytic reforming
Catalyst and fines from catalytic cracking (FCC catalyst and FCC fines)
Catalyst from catalytic hydrorefining
Catalyst from H2SO4 alkylation
Catalyst from sulfur complex and H2S removal facilities (Glaus and tail gas treating catalysts)
Off-spec product and fines from thermal processes (Off-spec coke and fines)
Spent caustic from liquid treating
Study Residuals
Desalting sludge from crude desalting
Residual oil storage tank sludge
Process sludge from residual upgrading
Catalyst from extraction/isomerization processes*
Catalyst from catalytic hydrocracking
Catalyst from polymerization
Catalyst from HF alkylation
Off-spec product and fines from residual upgrading
Off-spec product from sulfur complex and H2S removal facilities (Off-spec sulfur)
Off-spec treating solution from sulfur complex and H2S removal facilities (Spent amine and spent
Stretford solution)
Acid-soluble oil from HF alkylation (ASO)
Treating clay from clay filtering
Treating clay from lube oil processing
Treating clay from the extraction/isomerization process
Treating clay from alkylation
*As described in Section 3.5 Extraction, catalyst used for extraction does not exist. The Agency believes it has been
classified as a residual of concern inappropriately based on erroneous old data. Therefore, only catalyst from
isomerization will be discussed in this study.
1.2 OTHER EPA REGULATORY PROGRAMS IMPACTING THE PETROLEUM
REFINING INDUSTRY
Each of EPA's major program offices has long-standing regulatory controls tailored to
the petroleum refining industry. Some of the more significant programs with some relevance to
OSW's listing determinations and industry study include:
• The Clean Air Act's Benzene National Emissions Standards for Hazardous Air
Pollutants (NESHAPS), designed to control benzene releases from process and waste
management units.
Petroleum Refining Industry Study 2 August 1996
-------
• The Clean Air Act's National Ambient Air Quality Standards (NAAQS), which
prescribe limits for sulfur oxides (SOx), carbon monoxide (CO), particulates, nitrogen
oxides (NOx), volatile organic compounds (VOCs), and ozone.
• The Clean Air Act's NESHAPs for Petroleum Refineries (40 CFR Part 63, Subpart
CC, see 60 FR 43244, August 18, 1995), designed to control hazardous air pollutants
(HAPs).
• The Clean Water Act sets specific technology-based limits and water quality-based
standards for discharges to surface waters and publically-owned treatment works
(POTWs) including standards designed specifically for discharges from the petroleum
refining industry.
• The Toxicity Characteristic, particularly for benzene, in combination with the F037/
F038 sludge listings, has had a significant impact on the industry's wastewater
treatment operations, forcing closure of many impoundments and redesign of tank-
based treatment systems.
• The Land Disposal Restrictions (LDR) Program, including the ongoing Phase III and
IV development work.
1.3 INDUSTRY STUDY FINDINGS
This document describes EPA's approach to conducting the industry study required by
the EDF/EPA consent decree. The consent decree requires that EPA "fully characterize" the
study residuals and how they are managed. "The report shall include a discussion of the
concentration of toxic constituents in each waste, the volume of each waste generated, and the
management practices for each waste (including plausible mismanagement practices)."
The statutory definition of "hazardous waste" is waste that may cause harm or pose a
hazard to human health or the environment "when improperly treated, stored, transported, or
disposed of, or otherwise managed."
To implement this section of the statute, EPA considers available information on current
management practices, and also exercises judgment as to plausible ways the waste could be
managed in addition to those practices actually reported. EPA then judges which management
practices have the potential to pose the greatest risk to human health or the environment and
those practices would be assessed in a risk assessment.
As EPA explained in the preamble to the dyes and pigments proposed listing [59 FR
66072], EPA generally assumes that placement in an unlined landfill is a reasonably plausible
management scenario for solids that potentially poses significant risks and thus would be
"mismanagement" that should be examined by further risk assessment. For liquid wastes,
unlined surface impoundments are such a presumptive mismanagement scenario. In past risk
assessment work, EPA has found that these two scenarios are generally the scenarios most likely
to pose a risk to ground water and thus would be mismanagement scenarios for a hazardous
waste. In some cases, EPA has also found it appropriate to examine waste piles for solids prone
Petroleum Refining Industry Study 3 August 1996
-------
to transport by wind or erosion and to look at an aerated tank for volatile hazardous constituents
in waste waters.
EPA also considers other scenarios, such as land application without Federal regulatory
controls, as possible mismanagement scenarios and, where there is evidence that such practices
occur for a particular waste stream, would consider whether further evaluation is appropriate. If
EPA determines that a presumptive mismanagement scenario, such as disposal in an unlined
surface impoundment, does not occur and would not reasonably be expected to occur, EPA may
consider it implausible and instead use a more likely scenario as the plausible mismanagement
scenario for subsequent analysis.
In the recent proposal to list petroleum residuals, EPA found the following waste
management practices to pose the greatest risk and be the basis for judging whether these wastes
posed a potential risk to human health or the environment when mismanaged:
• Unlined landfills
• Unlined surface impoundments
• Land application units not subject to Federal regulations
With respect to the residuals in this study, EPA found that the following management
practices and their associated residuals (see Table 1.2) were reported and thus would be
mismanagement scenarios EPA would further evaluate to ascertain if there were a potential risk:
• Unlined landfills
- Residual oil storage tank sludge
- Process sludge from residual upgrading
- Catalyst from catalytic hydrocracking
- Catalyst from polymerization
- Off-spec product from sulfur complex and H2S removal facilities (off-spec sulfur)
- Off-spec treating solution from sulfur complex and H2S removal facilities (spent
amine and spent Stretford solution)
Petroleum Refining Industry Study 4 August 1996
-------
Table 1.2. Overview of 15 Study Residuals of Concern as Managed in 1992
Management Practice
Residuals of Concern: Study Residuals
ASO
rnt
Isom
Catalyst!
rnt
HF
Catalyst
rnt
Polymer
Catalyst
mt
Desalting
Sludge
mt
Hydro-
Cracking
Catalyst
mt
Off-spec
Prod. Resid
Upgrading
mt
Off-spec
Sulfur
Product
mt
Sludge
Resid
Upgrad
mt
Resid Oil
Tank
Sludge
mt
Off-spec
Treating
Solution
mt
Treating
Clay
Alkylation
mt
Treating
Clay
Clay Filter
mt
Treat Clay
Isom/
Extract
mt
Treating
Clay from
Lube Oil
mt
TOTALS
MT
Percent
of Total
DISPOSAL
Disposal offsite Subtitle D landfill
Disposal offsite Subtitle C landfill
Disposal onsite Subtitle C landfill
Disposal onsite Subtitle D landfill
Disposal onsite or offsite underground injection
Storage or disposal onsite surface impoundment
Other disposal onsite/roadbed mixing
Jse as cover in onsite landfill
Use as cap for onsite landfarm, fill material, or vent
TOTAL DISPOSED
0
0
44
44
0
1,429
65
349
256
2,099
29
221
102
2
354
1,593
992
2,584
0
5,043
3,576
289
226
9,133
138
0
62
7
0
207
6,458
622
4
30
132
7
7,254
200
39
711
673
1
1,624
634
24
67
626
4
1,355
3,641
1,735
52
1,032
16
20
6,497
937
516
58
496
138
2,145
37
79
5
120
20,138
7,913
886
3,485
675
133
158
7
20
33,417
16.8%
6.6%
0.7%
2.9%
0.6%
0.1%
0.1%
0.0%
0.0%
27.9%
DISCHARGED
Discharge to onsite wastewater treatment facility
Discharge to POTW
Discharge to surface water under NPDES
Discharge to offsite POTW
TOTAL DISCHARGED
1,258
3,600
3,600
0
152
152
0
128
647
1,266
1,913
0
0
0
3
1
1
47
0
205
0
6,849
1,566
8,415
0
0
7
507
507
0
0
1,648
648
12,374
1,566
14,588
1.4%
0.5%
10.3%
1.3%
12.2%
RECOVERED, RECYCLED, REUSED, REGENERATED
Metals Reclamation
Transfer metal catalyst for reclamation or regeneration
Recycle to Processes
Recovery onsite via distillation, coker, or cat cracker
Onsite reuse
Other recycling, reclamation or reuse/sulfur recov. unit
Recovery onsite in catalytic cracker
Recovery onsite in coker
Other recovery onsite/alky
Other recovery onsite/hydroprocessing
Other recovery onsite/reuse in extraction process
Miscellaneous On-site Recycling
Reuse onsite as replacement catalyst for another unit
Other recovery onsite
Other recycling, reclamation or reuse/offsite reuse
Other recycling, reclamation or reuse/cement plant
TOTAL RECYCLED
50
3,641
1,019
1,300
510
370
6,890
293
293
0
749
749
52
52
13,185
159
13,345
800
800
0
2
2
16
16
310
0
0
310
5,127
13
1,150
6,290
91
30
771
892
89
20
20
38
161
329
33
28
62
354
249
603
18,819
376
20
15
4,791
1,840
1,300
510
800
159
724
68
1,210
30,633
15.7%
0.3%
0.0%
0.0%
4.0%
1.5%
1.1%
0.4%
0.7%
0.1%
0.6%
0.1%
1.0%
25.6%
STORAGE
Storage in pile
TOTAL STORED (interim)
0
0
0
0
0
0
0
0
0
0
0
0
30
30
128
128
20
20
0
178
178
0.1%
0.1%
ft
C
g
C3
OQ
OQ
(3
-------
Table 1.2. Overview of 15 Study Residuals of Concern as Managed in 1992 (continued)
Management Practice
Residuals of Concern: Study Residuals
ASO
rnt
Isom
Catalyst!
rnt
HF
Catalyst
rnt
Polymer
Catalyst
mt
Desalting
Sludge
mt
Hydro-
Cracking
Catalyst
mt
Off-spec
Prod. Resid
Upgrading
mt
Off-spec
Sulfur
Product
mt
Sludge
Resid
Upgrad
mt
Resid Oil
Tank
Sludge
mt
Off-spec
Treating
Solution
mt
Treating
Clay
Alkylation
mt
Treating
Clay
Clay Filter
mt
Treat Clay
Isom/
Extract
mt
Treating
Clay from
Lube Oil
mt
TOTALS
MT
Percent
of Total
TRANSFER
Transfer of acid or caustic for recycle, reuse, reclamation
Transfer for use as ingredient in products placed on land
Transfer to N.O.S. offsite entity and final management
Transfer to another petroleum refinery
Transfer for direct use as a fuel or to make a fuel
Transfer with coke product or other refinery product
Transfer to other offsite entity/carbon regeneration
Transfer to other offsite entity/amine reclaimer
Transfer to other offsite entity/alumina manufacturer
Transfer to other offsite entity/smelter
Transfer to other offsite entity/used as a raw material feed
TOTAL TRANSFERRED
741
3,731
4,472
0
0
543
543
1,938
1,938
2,100
2,100
0
15
0
7
488
509
0
5
5
35
927
962
2,475
166
2,641
405
155
560
176
95
5
54
329
14
14
0
768
14
3,027
2,773
3,747
54
166
405
155
488
14,073
0.6%
0.0%
2.5%
2.3%
3.1%
0.0%
0.1%
0.3%
0.1%
0.4%
11.8%
TREATMENT
Evaporation*
Bioremediation*
\leutralization
Offsite incineration, stabilization, or reuse
Onsite boiler
Onsite industrial furnace
Onsite land treatment
Offsite land treatment
TOTAL TREATED (interim)
GRAND TOTAL
11,388
0
2,610
3,274
17,272
33,493
28.0%
0
0
337
0.3%
0
152
0.1%
0
0
728
728
4,119
3.4%
56
346
53
455
4,841
4.0%
0
18,029
15.1%
0
800
0.7%
1
1
2
9,647
8.1%
9
9
242
0.2%
530
4
534
9,107
7.6%
0
9
9
23,881
19.9%
59
59
2,895
2.4%
8
21
42
923
198
1,193
8,990
7.5%
231
231
2,471
2.1%
10
10
733
0.6%
8
21
11,388
108
2,619
3,274
2,827
256
20,502
119,738
0
0
9.5%
0.1%
2.2%
2.7%
2.4%
0.2%
17.1%
ft
C
g
C3
OQ
* To avoid double counting, these intermediate steps were not included in the total.
qg
(3
-------
- Treating clay from clay filtering
- Treating clay from lube oil processing
- Treating clay from the extraction/isomerization process
- Treating clay from alkylation
• Unlined surface impoundments
- Residual oil storage tank sludge
- Off-spec treating solution from sulfur complex and H2S removal facilities (spent
amine and spent Stretford solution)
• Land application not subject to Federal regulations
- Residual oil storage tank sludge
- Catalyst from polymerization
- Off-spec product from sulfur complex and H2S removal facilities (off-spec sulfur)
- Treating clay from clay filtering
- Treating clay from lube oil processing
- Treating clay from the extraction/isomerization process
- Treating clay from alkylation
In addition, EPA found that the management practice of mixing of treating clays with
roadbed materials for onsite use was reported and would merit evaluation as a potential
mismanagement scenario.
Section 2.0 provides an overview of the petroleum refining industry and EPA's approach
to this study. The fifteen study residuals identified in the consent decree accounted for
approximately 120,000 metric tons in 1992, compared to over 3.1 million metric tons of listing
residuals generated in 1992. Table 1.2 provides a description of the 15 study residuals by
management practice and volume generated. The Agency believes that the management
practices reported consist of virtually all of the plausible management practices to which the
residuals may be subjected. Section 3.0 describes the refinery processes associated with
generating the consent decree residuals of concern and detailed characterization of each of the
study residuals as required by the consent decree.
Petroleum Refining Industry Study 7 August 1996
-------
2.0 INDUSTRY DESCRIPTION
2.1 PETROLEUM REFINING INDUSTRY PROFILE
In 19921, the U.S. petroleum refining industry consisted of 185 refineries (of which 171
were fully active during the year) owned by 91 corporations. Atmospheric crude oil distillation
capacity totaled 15,120,630 barrels per calendar day (bpcd) (DOE, 1993). As of January 1,
1996, U.S. capacity totaled 15,341,000 bpcd, showing little change in the Nation's refining
capacity since the Agency's baseline year. Figure 2.1 illustrates the distribution of refineries
across the country. Refineries can be classified in terms of size and complexity of operations.
Forty-four percent of the refineries operating in 1992 processed less than 50,000 barrels per day
of crude, while the 20 largest companies account for 77 percent of the nation's total refining
capacity.
The simplest refineries use distillation to separate gasoline or lube oil fractions from
crude, leaving the further refining of their residuum to other refineries or for use in asphalt.
Approximately 18 percent of the U.S.'s refineries are these simple topping, asphalt, or lube oil
refineries. More sophisticated refineries will have thermal and/or catalytic cracking capabilities,
allowing them to extract a greater fraction of gasoline blending stocks from their crude. The
largest refineries are often integrated with chemical plants, and utilize the full range of catalytic
cracking, hydroprocessing, alkylation and thermal processes to optimize their crude utilization.
Section 3.1 describes the major unit operations typically found in refining operations.
The refining industry has undergone significant restructuring over the past 15 years.
Much of this restructuring has been in response to the price allocation programs of the 1970s and
industry deregulation in the 1980s. While the total national refining capacity dropped 17 percent
since 1980 to 15 million barrels per day, the number of refineries dropped 45 percent from 311
in 1980 to approximately 171 active in 1992 (and 169 as of 1/1/96). Refinery utilization rates
over the 1980 to 1992 period rose from 75 percent to 90 percent. (API, 1993). Very few new
refineries have been constructed in the past decade; the industry instead tends to focus on
expansions of existing plants.
The facilities closed tended to be smaller, inefficient refineries. Larger existing facilities
with capacities over 100,000 bbl/day have increased production to off-set the facility closings.
The data presented above indicates that the petroleum refining industry has been going
through a consolidation, which has resulted in a large decrease in the number of refineries in the
United States, but only a slight decrease in production. It is expected that this trend will
'The Agency conducted its industry-wide survey in 1993-1994, characterizing residual generation in 1992.
Thus, 1992 was considered the Agency's baseline year. The Agency has no reason to conclude that 1992 was not
representative of industry management practices. EPA's risk assessment modeling used as input the 1992 data for the
RCs as a "snap shot" of the industry's management practices. However, information for years other than 1992 is
provided in the pertinent sections of the study.
Petroleum Refining Industry Study 8 August 1996
-------
.—
'•Z
4*
^a
4*
o
=
s
.c
OJTj
O
a>
O
0/j
Petroleum Refining Industry Study
August 1996
-------
continue, with refineries continuing to close, but expansions occurring at others, keeping the
total refinery capacity in line with demand for refinery products.
In addition to restructuring, the industry is adding and changing production operations.
Many of these process changes are being implemented as a result of two factors: (1) today's
crudes tend to be heavier and contain higher levels of sulfur and metals, requiring process
modifications, and (2) a series of important pollution control regulations have been
implemented, including new gasoline reformulation rules designed to reduce the amount of
volatile components in gasoline, and new regulations requiring low-sulfur diesel fuels. These
heavier crudes and new rules have caused refineries to make process modifications to their
gasoline production units such as catalytic cracker units, installing additional sulfur removal
hydrotreaters, and constructing unit processes to manufacture additives such as oxygenates.
Many of the process modifications in response to the reformulated gasoline and low
sulfur diesel fuels have been implemented since 1992. The Oil and Gas Journal (December,
1993, 1994, and 1995) reports the following major processing capacity changes from year end
1992 to year end 1995:
• 5.5 percent capacity increase in thermal operations (forecast to further increase by new
construction scheduled to be completed in 1996)
• 8.7 percent capacity increase in hydrocracking operations
• 9.8 percent capacity increase in combined hydrorefming and hydrotreating operations
(there was a 16 percent increase in hydrotreating capacity offset by a 12 percent
decrease in hydrorefming capacity).
• 13.8 percent increase in aromatic and isomerization unit capacity.
• 5.6 percent increase in alkylation capacity
• 11.3 percent increase in lube production capacity
• 7.7 percent decrease in asphalt production
• Small capacity increases for crude distillation, reforming, and catalytic cracking
(increases of 0.9, 0.7, and 1.6 percent, respectively).
2.2 INDUSTRY STUDY
OSW's current listing determination and industry study for the petroleum refining
industry has been underway since 1992 and can be characterized in terms of two major avenues
for information collection: field work and survey evaluation. As part of the Agency's field
work, site selection, engineering site visits, familiarization sampling, and record sampling were
conducted. The survey effort included the development, distribution, and assessment of an
extensive industry-wide RCRA §3007 survey. Each of these elements is described further
below, reflecting the relative order in which these activities were conducted.
Petroleum Refining Industry Study 10 August 1996
-------
2.2.1 Site Selection
EPA's field work activities were focussed on a limited number of refineries, allowing the
Agency to establish strong lines of communication with the selected facilities, and maximizing
efficiency of information collection. After considering logistical and budgetary constraints, the
Agency determined that it would limit its field work to 20 refineries.
The Agency defined a site selection procedure that was used in selecting the 20 site visits
from the population of 185 domestic refineries in the continental U.S.. The objectives of the
selection procedure were:
• to ensure that the characterization data obtained from residuals at the 20 selected
facilities could be used to make valid, meaningful statements about those residuals
industry-wide.
• to give the Agency first-hand exposure to both large and small refineries.
• to be fair to all domestic refineries.
The Agency chose to select facilities randomly rather than purposefully. Although a
randomly selected group of refineries did not offer as many sampling opportunities as a hand-
picked group (e.g., focusing on those larger refineries that generate most of the RCs), the
Agency favored random selection because it did not require subjective input, and also because it
lends itself to statistical analysis, which is useful in making general statements about the
population of residuals.
The Agency broke the industry into two strata based on atmospheric distillation capacity
and made random selections from each stratum independently. The high-capacity stratum (those
with a crude capacity of 100,000 bpcd or greater) contains the top 30 percent of refineries, which
together account for 70 percent of the refining industry's capacity. The stratification enables the
Agency to weigh the selection toward the larger facilities on the basis that they produce larger
volumes of residuals, and that they offer a larger number of residual streams per site visit. The
Agency chose to select 12 of the 20 site visits, 60 percent, from the high-capacity stratum. The
smaller facilities had a lower chance of being selected, but not as low as they would have if the
likelihood of selection was based strictly on size. The selected facilities are presented in Table
2.12.
2Upon initial contact with several of the randomly selected refineries, it was determined that they were
inappropriate candidates for site visits because they had stopped operation and were not generating any residuals of
interest to the Agency. Replacement facilities were then selected randomly from the same stratum.
The list of refineries slated for field investigations was expanded in June, 1994 to allow the Agency to fill out
certain categories of samples that proved to be difficult to find in the field. The final list presented in Table 2.1
represents those refineries at which site visits actually occurred.
Petroleum Refining Industry Study 11 August 1996
-------
Table 2.1. Engineering Site Visit Facilities
Refinery
Amoco Oil
Arco
Ashland
Ashland
BPOil
BPOil
Chevron (purchased by Clark)1
Chevron1
Conoco 1
Exxon
Koch
Little America
Marathon
Murphy
Pennzoil
Phibro Energy1
Rock Island (purchased by Marathon)
Shell
Shell
Shell
Star Enterprise1
Star Enterprise1
Sun
Texaco
Total
Young
Location
Texas City, Texas
Ferndale, Washington
Canton, Ohio
Catlettsburg, Kentucky
Belle Chasse, Louisiana
Toledo, Ohio
Port Arthur, Texas
Salt Lake City, Utah
Commerce City, Colorado
Billings, Montana
St. Paul, Minnesota
Evansville, Wyoming
Garyville, Louisiana
Superior, Wisconsin
Shreveport, Louisiana
Houston, Texas
Indianapolis, Indiana
Deer Park, Texas
Norco, Louisiana
Wood River, Illinois
Convent, Louisiana
Port Arthur, Texas
Philadelphia, Pennsylvania
Anacortes, Washington
Ardmore, Oklahoma
Douglasville, Georgia
Initial Site Visit Date
March 29, 1993
June 9, 1993
May 24, 1993
March 22, 1993
May3, 1993
May 26, 1993
August 31, 1994
February 21, 1995
To be determined
June 9, 1993
May 19, 1993
JuneS, 1993
April 22, 1993
May 17, 1993
May5, 1993
April 20, 1995
April 26, 1993
March 31, 1993
April 20, 1993
May 28, 1993
August 30, 1994
September 21, 1994
May 12, 1993
June 10, 1993
June 23, 1993
June 21, 1993
1Refinery selected to augment record sample availability.
Petroleum Refining Industry Study
12
August 1996
-------
2.2.2 Engineering Site Visits
The field activities were initiated with a series of engineering site visits to the selected
facilities. The purpose of these trips was to:
• Develop a firm understanding of the processes associated with the RCs
• Understand how, when, why, and where each residual is generated and managed
• Establish a schedule of sampling opportunities
• Establish a dialogue with the refinery personnel to ensure optimal sampling and
collection of representative samples.
An engineering site visit report was developed for each of the trips; these are available in
the CBI and non-CBI dockets, as appropriate. For the later site visits conducted in 1994 and
1995, the engineering site visit reports were combined with the analytical data reports prepared
for each facility. The site visit reports included the following elements:
• Purpose of the site visit
• Refinery summary, including general information gathered during the site visit, as well
as data gleaned from telephone conversations and reviews of EPA files, the refinery's
process flow diagram, and expected residual availability
• A discussion of the processes used at the refinery generating the residuals of concern
• Source reduction and recycling techniques employed by the refinery
• A description of onsite residual management facilities
• A chronology of the site visit.
2.2.3 RCRA §3007 Questionnaire
EPA developed an extensive questionnaire under the authority of §3007 of RCRA for
distribution to the petroleum refining industry. A blank copy of the survey instrument is
provided in the RCRA docket. The questionnaire was organized into the following areas:
I. Corporate and facility information
II. Crude oil and product information
III. Facility process flow diagram
IV. Process units: general information
V. Process units: flow diagrams and process descriptions
VI. Residual generation and management
VII. Residual and contaminated soil and debris characterization
VIII. Residual management units: unit-specific characterization
IX. Unit-specific media characterization
Petroleum Refining Industry Study 13 August 1996
-------
X. General facility characterization (focusing on exposure pathway characterization)
XII. Source reduction efforts
XIII. Certification.
The survey was distributed in August 1993 to all refineries identified as active in 1992 in
the DOE Petroleum Supply Annual. Of the 185 surveys distributed, completed responses were
obtained for 172 refineries. Thirteen refineries notified EPA that they had stopped operations at
some point in or after 1992 and thus were unable to complete the survey due to no staffing or
inaccessible or unavailable data.
The survey responses were reviewed by SAIC chemical engineers for completeness and
then entered into a relational data base known as the 1992 Petroleum Refining Data Base
(PRDB). The entries were subjected to a series of automated quality assurance programs to
identify inappropriate entries and missing data links. An exhaustive engineering review of each
facility's response was then conducted, resulting in follow-up letters to most of the industry
seeking clarifications, corrections, and additional data where needed. The responses to the
followup letters were entered into the data base. A wide variety of additional quality assurance
checks were run on the data to ensure that the residuals of concern were characterized as
completely and accurately as possible. Follow-up telephone interviews were conducted as
necessary to address remaining data issues. After extensive review, the Agency believes that the
data are reliable and represent the industry's current residual generation and management
practices.
Table 2.2 describes the survey results for each of the study residuals of concern, sorted
by total volume generated in metric tons (MT).
2.2.4 Familiarization Sampling
The early phases of the analytical phase of this listing determination consisted of the
development of a Quality Assurance Project Plan (QAPjP) for sampling and analysis, followed
by the collection and analysis of six "familiarization" samples (five listing residuals and one
study residual). The purpose of collecting these samples was to assess the effectiveness of the
methods identified in the QAPjP for the analysis of the actual residuals of concern. Due to the
high hydrocarbon content of many of the RCs, there was concern at the outset of the project that
analytical interferences would prevent the contracted laboratory from achieving adequate
quantitation limits; familiarization analysis allowed the laboratories to experiment with the
analytical methods and waste matrices and optimize operating procedures.
In addition, the first version of the QAPjP identified a list of target analytes that was
derived from previous Agency efforts to characterize refinery residuals. These included the
Delisting Program's list of analytes of concern for refinery residuals, the "Skinner List", an
evaluation of compounds detected in the sampling and analysis program for listing refinery
residuals in the 1980s, and the judgment of EPA and SAIC chemists who evaluated the process
chemistry of the residuals of concern. During familiarization sample analysis, particular
attention was paid to the tentatively identified compounds to determine whether they should be
added to the target analyte list.
Petroleum Refining Industry Study 14 August 1996
-------
Table 2.2. Study Residuals Volume Statistics
Study Residual Description
Acid Soluble Oil
Hydrocracking Catalyst
Off-specification Product from Sulfur Complex and H,S Removal
Residual Oil Tank Sludge
Treating Clay from Clay Filtering
Desalting Sludge
Off-specification Treating Solution from Sulfur Complex and H2S
Removal (spent amine and spent Stretford solution)
Catalyst from Polymerization (phosphoric acid and Dimersol)
Treating Clay from Alkylation
Treating Clay from Isomerization/Extraction
Off-specification Product from Residual Upgrading
Treating Clay from Lube Oil
Catalyst from Isomerization
Sludge from Residual Upgrading
Catalyst from HF Alkylation
Total
#of
Reported
Residuals
80
83
93
62
244
141
76
42
88
43
3
19
21
34
3
1,061
Total Volume
(MT)
33,493
18,029
9,647
9,107
8,990
4,841
23,881
4,119
2,895
2,472
800
733
337
242
152
119,738
Samples of five listing residuals were collected for familiarization analysis: crude oil
tank sediments, hydrotreating catalyst, sulfur complex sludge, H2SO4 alkylation catalyst, and
spent caustic. One study residual, acid soluble oil, was analyzed under this program. The results
of the familiarization effort essentially confirmed the techniques identified in the QAPjP and
indicated that the laboratories generally would be able to achieve adequate quantitation of the
target analytes. The familiarization and final QAPjPs are provided in the docket to the
November 20, 1995 proposed rulemaking.
2.2.5 Record Sampling
Upon completion of the familiarization sampling and analysis effort, the Agency initiated
record sampling and analysis of the listing and study residuals. Given budgetary constraints, the
Agency set a goal of collecting 4-6 samples of each of the listing residuals, and 2-4 samples of
the study residuals for a total of 134 samples3. Table 2.3 shows the 103 samples that were
actually collected. The numbers in the darkened boxes refer to Table 2.4 which lists each of the
sample numbers, sample dates, facility names, and other information describing the residual
samples.
3The Agency determined that one listing residual, catalyst from sulfuric acid alkylation, would not be sampled
due to the existing regulatory exemption for sulfuric acid destined for reclamation, and that one study residual,
catalyst from HF alkylation, could not be sampled due to its extremely rare generation.
Petroleum Refining Industry Study
15
August 1996
-------
Table 2.3. Residuals Collected for Record Analysis
Listing Residuals
Crude oil tank sludge
Unleaded gasoline tank sludge
CSO sludge
FCC catalyst and fines
Catalyst from hydrotreating
Catalyst from hydrorefining
Catalyst from reforming
Sulfuric acid alkylation sludge
HF alkylation sludge
Sulfur complex sludge
Catalyst from sulfur complex
Off-spec product & fines/thermal process
Spent caustic
Study Residuals
Residual oil tank sludge
Desalting sludge
Hydrocracking catalyst
Catalyst from isomerization
Treating clay from isomerization/extraction
Catalyst from polymerization
Treating clay, alkylation (HF and H2S04)
ASO
Off-spec sulfur product
Spent treating solution (amine)
Process sludge from residual upgrading
Off-spec product, residual upgrading
Treating clay from lube oil
Treating clay from clay filtering
Record Samples
1 I 2 I 3 I
33 67 73
34 42 65
14 49 72
21 36 85
3 22 37
1 12 13 26 27 28
6 44 55 83 94 69
79 75
19 47 51 74 96
10 25 29 80 70
9 15 23 24 52 54
30 45 59 63 81 84
16 17 32 62 64 95
Familiarization
Samples
Notes: Sulfuric Acid Alkylation catalyst is not presented in this figure. One familiarization sample of sulfuric
acid catalyst was captured and analyzed. HF catalyst is constant boiling mixture (CBM) and is not
shown in this figure.
The sampling team maintained monthly phone contact with the targeted refineries to
maintain an optimized sampling schedule. Despite careful coordination with the refineries and
best efforts to identify and collect all available samples, there were several categories of study
residuals for which the targeted minimum number of samples could not be collected:
• Two samples of residual oil tank sludge were collected. This residual is available only
for a brief period during tank turnarounds, which may occur only every 10 years. In
several cases, refineries mixed their residual oil and clarified slurry oil (CSO) in the
same tank.
Petroleum Refining Industry Study
16
August 1996
-------
Table 2.4. Descriptions of
Collected for Record
Count
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Residual Name
FCC catalyst and fines
Off-spec sulfur
Catalyst from reforming
Catalyst from hydrocracking
Desalting sludge
Catalyst from hydrotreating
Treating clay
Off-spec sulfur
Catalyst from sulfur complex
Sulfur complex sludge
Process sludge from residual upgrading
FCC catalyst and fines
FCC catalyst and fines
CSO sludge
Catalyst from sulfur complex
Spent caustic
Spent caustic
ASO
HFalkylation sludge
Treating clay from alkylation
Catalyst from hydrorefining
Catalyst from reforming
Catalyst from sulfur complex
Catalyst from sulfur complex
Sulfur complex sludge
FCC catalyst and fines
FCC catalyst and fines
FCC catalyst and fines
Sulfur complex sludge
Off-spec product & fines from thermal process
Treating clay
Spent caustic
Crude oil tank sludge
Unleaded gasoline tank sludge
Catalyst from polymerization
Catalyst from hydrorefining
Catalyst from reforming
ASO
Sample
Number
R2-FC-01
R2-SP-01
R2-CR-01
R2-CC-02
R1-DS-01
R1-TC-01
R1-CF-01
R1-SP-01
R1-SC-01
R1-ME-01
R1-RU-01
R4-FC-01
R4-FC-02
R4-SO-01
R4-SC-01
R3-LT-01
R3-LT-02
R3-AS-01
R3-HS-01
R3-CA-01
R5-TC-01
R5-CR-01
R5-SC-01
R5-SC-02
R5-ME-02.03
R5-FC-02
R6-FC-01
R6-FC-02
R6-ME-01
R6-TP-01
R6-CF-01
R6-LT-01
R6B-CS-01
R6B-US-01
R6B-PC-01
R7B-RC-01
R7B-CR-01
R5B-AS-01
Sample
Date
30-Sep-93
30-Sep-93
01-Oct-93
04-Oct-93
26-Oct-93
26-Oct-93
27-Oct-93
27-Oct-93
27-Oct-93
27-Oct-93
27-Oct-93
16-NOV-93
16-NOV-93
16-Nov-93
16-Nov-93
18-Nov-93
18-Nov-93
18-Nov-93
18-Nov-93
18-Nov-93
07-Feb-94
07-Feb-94
07-Feb-94
07-Feb-94
07-Feb-94
07-Feb-94
09-Feb-94
09-Feb-94
09-Feb-94
09-Feb-94
09-Feb-94
09-Feb-94
15-Mar-94
31-Mar-94
15-Mar-94
14-Mar-94
14-Mar-94
16-Mar-94
Notes
ESP Fines.
Taken from low spots on the unit.
Platinum catalyst.
2nd stage, Ni/W.
Removed from vessel.
Naphtha reformer pretreat, CoMo.
Kerosene.
From product tank.
AI203.
MEA reclaimer bottoms.
ROSE butane surge tank sludge.
Equilibrium cat. from hopper.
ESP fines, truck trailer comp.
Tank sludge from pad.
Claus unit alumina, super sack comp.
Tank samp. Cresylic, concentrated.
Tank samp. Sulfidic, concentrated.
Non-neutralized, separator drum sample
Not dewatered. Dredge from pit.
HF. Propane treater. Drum composite.
Heavy Gas Oil, CoMo
CCR fines, Pt
Claus
Tail gas, CoMo
Refinery MDEA filter cartridge
Wet Scrubber Fines
Equil. from unit
Wet scrubber fines
Refinery DEA filter cartridqe
Coke fines.
Kerosene
Naph. Comb. Gas oil & Kero
Mix of centrifuqe and uncentrifuqed
Water washed solids, collected by refinery
Dimersol. filter
Diesel hydrorefiner
Platinum
Acid reqen settler bottoms, not neutralized
Refinery
Shell, Wood River, Illinois
Shell, Wood River, Illinois
Shell, Wood River, Illinois
Shell, Wood River, Illinois
Marathon, Indianapolis
Marathon, Indianapolis
Marathon, Indianapolis
Marathon, Indianapolis
Marathon, Indianapolis
Marathon, Indianapolis
Marathon, Indianapolis
Little America, Evansville, Wy
Little America, Evansville, Wy
Little America, Evansville, Wy
Little America, Evansville, Wy
Exxon, Billinqs, Montana
Exxon, Billinqs, Montana
Exxon, Billinqs, Montana
Exxon, Billinqs, Montana
Exxon, Billinqs, Montana
Marathon, Garyville, LA
Marathon, Garyville, LA
Marathon, Garyville, LA
Marathon, Garyville, LA
Marathon, Garyville, LA
Marathon, Garyville, LA
Shell, Norco, LA
Shell, Norco, LA
Shell, Norco, LA
Shell, Norco, LA
Shell, Norco, LA
Shell, Norco, LA
Shell, Norco, LA
Shell, Norco, LA
Shell, Norco, LA
BP, Belle Chase, LA
BP, Belle Chase, LA
Marathon. Garvville. LA
ft
I
K
B'
OQ
-------
Table 2.4. Descriptions of Samples Collected for Record Analysis
Count
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
Residual Name
Catalyst from isomerization
Off-spec sulfur
Residual oil tank sludqe
Unleaded qasoline tank sludqe
Catalyst from hydrocrackinq
Catalyst from hydrotreatinq
Off-spec product & fines from thermal processes
H2S04 alkylation sludqe
HFalkylation sludqe
Catalyst from isomerization
CSO sludqe
Desaltinq sludqe
HFalkylation sludqe
Catalyst from sulfur complex
Crude oil tank sludqe
Catalyst from sulfur complex
Catalyst from hydrotreatinq
Catalyst from reforminq
Treatinq clay
Spent amine
Off-spec product & fines from thermal processes
Treatinq clay from lube oil
Spent amine
Spent caustic
Off-spec product & fines from thermal processes
Spent caustic
Unleaded qasoline tank sludqe
Catalyst from polymerization
Crude oil tank sludqe
Treatinq clay from extraction
Catalyst from hydrotreatinq
Sulfur complex sludqe
Catalyst from isomerization
CSO sludqe
Crude oil tank sludqe
HFalkylation sludqe
Catalyst from reforminq
Treatinq clay from alkvlation
Sample
Number
R5B-1C-01
R7B-SP-01
R8A-RS-01
R8A-US-01
R8A-CC-01
R8A-TC-01
R8A-TP-01
R8B-SS-01
R8B-HS-01
R8B-IC-01
R9-SO-01.02
R9-DS-01
R9-HS-01
R7B-SC-01
R10-CS-01
R11-SC-01
R11-TC-01
R11-CR-01
R11-CF-01
R11-SA-01
R11-TP-01
R13-CL-01
R13-SA-01
R13-LT-01
R12-TP-01
R12-LT-01
R16-US-01
R16-PC-01.02
R8C-CS-01
R8D-CI-01
R18-TC-01
R18-ME-01
R18-IC-01
R1B-CS-01
R4B-CS-01
R15-HS-01
R15-CR-01
R15-CA-01
Sample
Date
16-Mar-94
14-Mar-94
30-Apr-94
14-Apr-94
30-Mar-94
30-Mar-94
30-Mar-94
30-Apr-94
30-Apr-94
30-Apr-94
17-May-94
17-May-94
17-May-94
14-Mar-94
26-Auq-94
10-May-94
10-May-94
10-May-94
10-May-94
10-May-94
10-May-94
30-Apr-94
30-Apr-94
30-Apr-94
12-May-94
12-May-94
03-Auq-94
03-Auq-94
01-Jul-94
15-Nov-96
20-Oct-94
14-Oct-94
20-Oct-94
26-Auq-94
26-Auq-94
02-Auq-94
02-Auq-94
02-Auq-94
Notes
Butamer, platinum
From cleaned out tank
CSO and Resid.
Collected by refinery
Hydroproc., 1st staqe cracker, CoMo
NiMo, landfilled
Fines, F&K processed
From Froq pond, not dewatered
Notdewatered.dredqed
Butamer, Pt
Filters (and blank)
SCOT catalyst
SCOT, CoMo
NiMo, naphtha treater
Pt/Rh
Reformer sulfur trap
DEA
Coke fines
Clay dust
DEA
Sulfidic
Coke fines, from trap
Cresylic
2 catalysts from Dimersol and H2P04
collected by refinery from tank bottom
collected by refinery
naptha
MEA sludqe, collected by refinery
Penex
mixed CSO/resid
Filter cake sludqe
Dredqed from pit
CCR fines
Butane
Refinery
Marathon, Garyville, LA
BP, Belle Chase, LA
Amoco, Texas City
Amoco, Texas City
Amoco, Texas City
Amoco, Texas City
Amoco, Texas City
Amoco, Texas City
Amoco, Texas City
Amoco, Texas City
Murphy, Superior, Wl
Murphy, Superior, Wl
Murphy, Superior, Wl
BP, Belle Chase, LA
Ashland, Catletsburq, KY
ARCO, Ferndale, WA
ARCO, Ferndale, WA
ARCO, Ferndale, WA
ARCO, Ferndale, WA
ARCO, Ferndale, WA
ARCO, Ferndale, WA
Shell, Deer Park, TX
Shell, Deer Park, TX
Shell, Deer Park, TX
Texaco, Anacortes, WA
Texaco, Anacortes, WA
Koch
Koch
Amoco, Texas City
Amoco, Texas City
Ashland, Canton, OH
Ashland, Canton, OH
Ashland, Canton, OH
Marathon, Indianapolis
Little America
Total, Ardmore, OK
Total, Ardmore
Total. Ardmore
ft
I
K
B'
OQ
-------
Table 2.4. Descriptions of Samples Collected for Record Analysis
Count
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
Residual Name
ASO
Spent amine
Catalyst from reforminq
Sulfur complex sludqe
Off-spec product & fines from thermal processes
Spent amine
Catalyst from hydrotreatinq
Off-spec product & fines from thermal processes
Catalyst from hydrorefininq
Treatinq clay from alkylation
Catalyst from hydrocrackinq
CSO sludqe
Crude oil tank sludqe
Desaltinq sludqe
Crude oil tank sludqe
Residual oil tank sludqe
ASO
Catalyst from hydrotreatinq
Spent caustic
HF alkylation sludqe
Catalyst from isomerization
Treatinq clay from isomerization
Treatinq clay from alkylation
Off-spec sulfur
Treatinq clay from clay filterinq
Desaltinq sludqe
Sample
Number
R15-AS-01
R15-SA-01
R14-CR-01
R14-ME-01
R14-TP-01
R14-SA-01
R3B-TC-01
R3B-TP-01
R21-RC-01
R21-CA-01
R20-CC-01
R20-SO-01
R19-CS-01
R11B-DS-01
R22-CS-01
R22-RS-01
R7C-AS-01
R22-TC-01
R22B-LT-01
R7C-HS-01
R23B-CI-01
R23B-IC-01
R23-CA-01
R23-SP-01
R23-CF-01
R24-DS-01
Sample
Date
02-Auq-94
02-Auq-94
07-Jun-94
07-Jun-94
07-Jun-94
07-Jun-94
12-Jul-94
12-Jul-94
31-Auq-94
31-Auq-94
30-Auq-94
30-Auq-94
12-Oct-96
01-Sep-94
21-Sep-94
21-Sep-94
12-Oct-96
21-Sep-94
11-Oct-96
12-Oct-96
19-Apr-95
19-Apr-95
19-Jan-95
19-Jan-95
19-Jan-95
20-Apr-95
Notes
Neut., skimmed from pit
MDEA
Cyclic Pt reformer
DEAdiatomaceous earth
Delayed cokinq fines
DEA from sump
Naptha treater
Fluid coker chunky coke
H-Oil unit, movinq bed
collected by refinery
caustic from H2S04 alky, sulfidic
Filter press
Pt catalyst
Mole sieve, butamer feed treater
propane treater
diesel washed
Sludqe from Lakos separator
Refinery
Total, Ardmore, OK
Total, Ardmore, OK
BP, Toledo, OH
BP, Toledo, OH
BP, Toledo, OH
BP, Toledo, OH
Exxon, Billings, MT
Exxon, Billings, MT
Chevron, Port Arthur, TX
Chevron, Port Arthur, TX
Star, Convent, LA
Star, Convent, LA
Pennzoil, Shreveport, LA
ARCO, Ferndale, WA
Star, Port Arthur, TX
Star, Port Arthur, TX
BP, Belle Chase, LA
Star, Port Arthur, TX
Star, Port Arthur, TX
BP, Belle Chase, LA
Chevron, Salt Lake City
Chevron, Salt Lake City
Chevron, Salt Lake City
Chevron, Salt Lake City
Chevron, Salt Lake City
Phibro, Houston, TX
Familiarization Samples
F1
F2
F3
F4
F5
F6
Spent Caustic
Catalyst from hydrotreatinq
Sulfur complex sludqe
ASO
Crude oil tank sludqe
Sulfuric Acid Catalyst
A-SC-01
A-HC-01
C-SS-01
C-AS-01
B-TS-01
B-SA-01
08-May-93
10-May-93
23-Jun-93
23-Jun-93
15-May-93
15-May-93
Comminqled.
Cobalt molybdenum.
MEA Reclaimer sludqe.
Neutralized.
Filter cake.
Spent from third unit.
Marathon, Garyville
Marathon, Garyville
Amoco, Texas City
Amoco, Texas City
Sun, Philadelphia
Sun, Philadelphia
ft
I
K
B'
OQ
>
I
-------
• Two samples of treating clay from isomerization/extraction were collected. This
residual is available only for a brief period during unit turnarounds, which may occur
only every 3-5 years. This residual was not readily available from the set of facilities
selected for sampling.
• One sample of treating clay from lube oil processes was collected. Due to the
specialty of the processes, a limited number of refineries produce lube oils and not all
of these facilities use clay filtering. This residual is not readily available, and was
extremely difficult to find from the facilities randomly selected for sampling.
• One sample of residual upgrading sludge was collected. This residual is not readily
available from the set of facilities selected for sampling.
• No samples of off-specification product from residual upgrading were collected. As is
discussed further in Section 3.7.2, the Agency believes that this residual was
inappropriately classified as a residual due to the evaluation of inaccurate old data.
This residual was reported as being generated by only one facility in the 1992 §3007
Survey.
Each of the samples collected was analyzed for the total and Toxicity Characteristics
Leaching Procedure (TCLP) concentrations of the target analytes identified in the QAPjP. In
addition, certain residuals were tested for different characteristics based on the Agency's
understanding of the residuals developed during the engineering site visits. Each sample was
also analyzed for the ten most abundant nontarget volatile and the 20 most abundant nontarget
semi-volatile organics in each sample. These tentatively identified compounds (TICs) were not
subjected to QA/QC evaluation (e.g., MS/MSD analyses) and thus were considered tentative.
2.2.6 Split Samples Analyzed by API
The American Petroleum Institute (API) accompanied the EPA contractor (SAIC) on
virtually all sampling trips and collected split samples of many of the record samples. API's
analytical results for a number of the samples were made available to EPA for comparison
purposes. In general, the Agency found that the API and EPA split sample analyses had very
good agreement. Appendix B of the Listing Background Document, available in the RCRA
docket for the 11/20/95 proposal, presents the Agency's comparison of the split sample results.
2.2.7 Synthesis
The results of the Agency's four year investigation have been synthesized in this report
and in the Listing Background document for the November 20, 1995 proposed rulemaking.
Additional supporting documents are available in the docket for that rulemaking.
Petroleum Refining Industry Study 20 August 1996
-------
3.0 PROCESS AND WASTE DESCRIPTIONS
3.1 REFINERY PROCESS OVERVIEW
Refineries in the United States vary in size and complexity and are generally geared to a
particular crude slate and, to a certain degree, reflect the demand for specific products in the
general vicinity of the refinery. Figure 3.1 depicts a process flow diagram for a hypothetical
refinery that employs the major, classic unit operations used in the refinery industry. These unit
operations are described briefly below, and in more detail in the remainder of this section. Each
subsection is devoted to a major unit operation that generates one or more of the study residuals
of concern and provides information related to the process, a description of the residual and how
and why it is generated, management practices used by the industry for each residual, the results
of the Agency's characterization of each residual, and summary information regarding source
reduction opportunities and achievements.
Storage Facilities: Large storage capacities are needed for refinery feed and products.
Sediments from corrosion and impurities accumulate in these storage tanks. The consent decree
identifies sludges from the storage of crude oil, clarified slurry oil, and unleaded gasoline for
consideration as listed wastes. Residual oil storage tank sludge was identified as a study
residual.
Crude Desalting: Clay, salt, and other suspended solids must be removed from the
crude prior to distillation to prevent corrosion and deposits. These materials are removed by
water washing and electrostatic separation. Desalting sludge is a study residual.
Distillation: After being desalted, the crude is subjected to atmospheric distillation,
separating the crude by boiling point into light ends, naphtha, middle distillate (light and heavy
gas oil), and a bottoms fraction. The bottoms fraction is frequently subjected to further distilla-
tion under vacuum to increase gas oil yield. No residuals from distillation are under
investigation.
Catalytic Cracking: Catalytic cracking converts heavy distillate to compounds with
lower boiling points (e.g., naphthas), which are fractionated. Cracking is typically conducted in
a fluidized bed reactor with a regenerator to continuously reactivate the catalyst. Cracking
catalysts are typically zeolites. The flue gas from the regenerator typically passes through dry or
wet fines removal equipment and carbon monoxide oxidation prior to being released to the
atmosphere. Catalyst and fines, as well as sediments from storage of and solids removal from
clarified slurry oil (the bottoms fraction from catalytic cracking), are listing residuals of concern.
Hydroprocessing: Hydroprocessing includes (1) hydrotreating and hydrorefining (or
hydrodesulfurization), which improve the quality of various products (e.g., by removing sulfur,
nitrogen, oxygen, metals, and waxes and by converting olefms to saturated compounds); and (2)
hydrocracking, which cracks heavy materials, creating lower-boiling, more valuable products.
Hydrotreating is typically less severe than hydrorefining and is applied to lighter cuts. Hydro-
cracking is a more severe operation than hydrorefining, using higher temperature and longer
contact time, resulting in significant reduction in feed molecular size. Hydroprocessing catalysts
Petroleum Refining Industry Study 21 August 1996
-------
M)
o
4*
w
o
PH
=
•5
•a
a>
S
"a
8
S
0/j
Petroleum Refining Industry Study
22
August 1996
-------
are typically some combination of nickel, molybdenum, and cobalt. Typical applications of
hydroprocessing include treating distillate to produce low-sulfur diesel fuel, treating naphtha
reformer feed to remove catalyst poisons, and treating catalytic cracking unit feed to reduce
catalyst deactivation. Hydrotreating and hydrorefming catalysts are listing residuals, while
hydrocracking catalyst is a study residual.
Thermal Processes: Thermal cracking uses the application of heat to reduce high-
boiling compounds to lower-boiling products. Delayed (batch) or fluid (continuous) coking is
essentially high-severity thermal cracking and is used on very heavy residuum (e.g., vacuum
bottoms) to obtain lower-boiling cracked products. (Residuum feeds are not amenable to
catalytic processes because of fouling and deactivation.) Products are olefmic and include gas,
naphtha, gas oils, and coke. Visbreaking is also thermal cracking; its purpose is to decrease the
viscosity of heavy fuel oil so that it can be atomized and burned at lower temperatures than
would otherwise be necessary. Other processes conducting thermal cracking also would be
designated as thermal processes. Off-spec product and fines is a listing residual from these
processes.
Catalytic Reforming: Straight run naphtha is upgraded via reforming to improve octane
for use as motor gasoline. Reforming reactions consist of (1) dehydrogenation of cycloparaffins
to form aromatics and (2) cyclization and dehydrogenation of straight chain aliphatics to form
aromatics. Feeds are hydrotreated to prevent catalyst poisoning. Operations may be
semiregenerative (cyclic), fully-regenerative, or continuous (moving bed) catalyst systems.
Precious metal catalysts are used in this process. Spent reforming catalyst is a listing residual.
Polymerization: Polymerization units convert olefms (e.g., propylene) into higher
octane polymers. Two principal types of polymerization units include fixed-bed reactors, which
typically use solid-supported phosphoric acid as the catalyst, and Dimersol® units, which
typically use liquid organometallic compounds as the catalyst. Spent polymerization catalyst is a
study residual.
Alkylation: Olefms of 3 to 5 carbon atoms (e.g., from catalytic cracking and coking)
react with isobutane (e.g., from catalytic cracking) to give high octane products. Sulfuric
(H2SO4) or hydrofluoric (HF) acid act as catalysts. Spent sulfuric acid, sulfuric acid alkylation
sludges, and FTP sludges are listing residuals, while spent FTP acid, acid soluble oil and treating
clays are study residuals.
Isomerization: Isomerization converts straight chain paraffins in gasoline stocks into
higher octane isomers. Isomer and normal paraffins are separated; normal paraffins are then
catalytically isomerized. Precious metal catalysts are used in this process. Spent catalysts and
treating clays are study residuals from this process.
Extraction: Extraction is a separation process using differences in solubility to separate,
or extract, a specific group of compounds. A common application of extraction is the separation
of benzene from reformate. Treating clay is a study residual from this process.
Lube Oil Processing: Vacuum distillates are treated and refined to produce a variety of
lubricants. Wax, aromatics, and asphalts are removed by unit operations such as solvent extrac-
Petroleum Refining Industry Study 23 August 1996
-------
tion and hydroprocessing; clay may also be used. Various additives are used to meet product
specifications for thermal stability, oxidation resistances, viscosity, pour point, etc. Treating
clay is a study residual from this process.
Residual Upgrading: Vacuum tower distillation bottoms and other residuum feeds can
be upgraded to higher value products such as higher grade asphalt or feed to catalytic cracking
processes. Residual upgrading includes processes where asphalt components are separated from
gas oil components by the use of a solvent. It also includes processes where the asphalt value of
the residuum is upgraded (e.g., by oxidation) prior to sale. Off-spec product and fines, as well
as process sludges, are study residuals from this category.
Blending and Treating: Various petroleum components and additives are blended to
different product (e.g., gasoline) specifications. Clay and caustic may be used to remove sulfur,
improve color, and improve other product qualities. Spent caustic is a listing residual, while
treating clay is a study residual.
Sulfur Recovery: Some types of crude typically contain high levels of sulfur, which
must be removed at various points of the refining process. Sulfur compounds are converted to
H2S and are removed by amine scrubbing. The H2S often is converted to pure sulfur in a Claus
plant. Off-gases from the Claus plant typically are subject to tail gas treating in a unit such as a
SCOT® treater for additional sulfur recovery. Process sludges and spent catalysts are listing
residuals; off-spec product and off-spec treating solutions are study residuals.
Light Ends (Vapor) Recovery: Valuable light ends from various processes are
recovered and separated. Fractionation can produce light olefms and isobutane for alkylation, n-
butane for gasoline, and propane for liquid petroleum gas (LPG). Caustic may be used to
remove sulfur compounds. Spent caustic is a listing residual of concern.
Petroleum Refining Industry Study 24 August 1996
-------
3.2 CRUDE OIL DESALTING
Crude oil removed from the ground is contaminated with a variety of substances,
including gases, water, and various minerals (dirt). Cleanup of the crude oil is achieved in two
ways. First, field separation, located near the site of the oil wells, provides for gravity separation
of the three phases: gases, water (with entrained dirt), and crude oil. The second cleanup
operation is crude oil desalting conducted at the refinery. Crude oil desalting is a water-washing
operation prior to atmospheric distillation which achieves additional crude oil cleanup. Water
washing removes much of the water-soluble minerals and suspended solids from the crude. If
these contaminants were not removed, they would cause a variety of operating problems
throughout the refinery including the blockage of equipment, the corrosion of equipment, and
the deactivation of catalysts.
3.2.1 Process Description
To operate efficiently and effectively the crude oil desalter must achieve an intimate
mixing of the water wash and crude, and then separate the phases so that water will not enter
downstream unit operations. The crude oil entering a desalting unit is typically heated to 100 -
300°F to achieve reduced viscosity for better mixing. In addition, the desalter operates at
pressures of at least 40 lb/in2 gauge to reduce vaporization. Intimate mixing is achieved through
a throttling valve or emulsifier orifice and the oil-water emulsion is then introduced into a
gravity settler. The settler utilizes a high-voltage electrostatic field to agglomerate water
droplets for easier separation. Following separation, the water phase is discharged from the unit,
carrying salt, minerals, dirt, and other water-soluble materials with it.
Desalting efficiency can be increased by the addition of multiple stages, and in some
cases acids, caustic, or other chemicals may be added to promote additional treatment. A
simplified process flow diagram for crude oil desalting is shown in Figure 3.2.1.
Figure 3.2.1. Desalting Process Flow Diagram
-M **
Crad*
Dasaltad
Crnda Oil
TJ_
toWWTP
Petroleum Refining Industry Study 25 August 1996
-------
3.2.2 Desalting Sludge
3.2.2.1 Description
Desalting sludge is continuously separated from the crude oil and settles to the bottom of
the desalter with the water wash. The majority of the sludge is removed from the desalter with
the water wash and is discharged to the facility's wastewater treatment plant. The sludge then
becomes part of the wastewater treatment sludges. On a regular basis (e.g., weekly), water jets
at the bottom of the desalter are activated, stirring up sludge that has built up on the bottom of
the unit and flushing it to wastewater treatment. This process is known as "mud washing" and
allows the units to continue to operate without shutting down for manual sludge removal.
Desalting sludge is removed from the unit during unit turnarounds, often associated with
turnarounds of the distillation column. These turnarounds are infrequent (e.g., every several
years). Some refineries operate enough desalters in parallel to allow for turnarounds while the
distillation columns continue to operate.
At turnaround, the sludge can be removed in several different ways. Based on the results
of the questionnaire, approximately half of the total number of desalting sludge waste streams
are removed from the desalter using a vacuum truck, permanent or portable piping, or other
similar means where the sludge is removed in a slurry state. Another 25 percent of the sludges
are removed manually by maintenance workers while the removal method for the remaining 25
percent of the sludges was not clear. The questionnaire data further indicated that half of the
desalting sludge streams are further piped or stored in tanker trucks following removal, while the
remaining half are stored in drums or a dumpster.
As with some tank sludges, some facilities remove their desalting sludge using a vacuum
truck or similar slurring device, then centrifuge the material and store the solids in a drum or
dumpster. Such procedures would explain the apparent discrepancy between the number of
streams removed as solid and the number of streams stored in containers (presumably also as
solid). Questionnaire data indicate that approximately 10 percent of the streams generated in
1992 underwent dewatering or a similar volume reduction procedure.
3.2.2.2 Generation and Management
Eighty facilities reported generating a total quantity of 4,841 MT of desalting sludge in
1992, according to the 1992 RCRA §3007 Survey. Desalting sludge includes material generated
from turnaround operations; materials continuously flushed to wastewater treatment are
generally omitted. The survey contained a residual identification code for "desalter sludge". All
residuals assigned this code, and any misidentified residual determined to be desalter sludge
generated from a process assigned process code for "desalting" were considered "desalter
sludge" residuals. This corresponds to residual code 02-A in Section VII.2 of the survey and
process code 01-A, 01-B, 01-C, and 01-D in Section IV-l.C. Quality assurance was conducted
to ensure that all desalting sludge residuals were correctly identified and coded.
Petroleum Refining Industry Study 26 August 1996
-------
Based on the results of the survey, 148 facilities use desalting units and are thus likely to
generate desalting sludge. Due to the infrequent generation of this residual, not all of these 148
facilities generated desalting sludge in 1992. In addition, some facilities do not generate
desalting sludge at all because they do not conduct unit turnarounds, or do not find any settled
sludge when conducting maintenance. However, there was no reason to expect that 1992 would
not be a typical year with regard to desalting sludge generation and management. Table 3.2.1
provides a description of the quantity generated, number of streams reported, number of streams
not reporting volumes (data requested was unavailable and facilities were not required to
generate it), total and average volumes.
Table 3.2.1. Generation Statistics for Desalting Sludge, 1992
Final Management
Discharge to onsite wastewater
treatment facility
Disposal in onsite or offsite
underground injection
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite Subtitle D landfill
Offsite incineration
Offsite land treatment
Onsite land treatment
Recovery onsite in a coker
Transfer for direct use as a fuel or to
make a fuel
TOTAL
#of
Streams
25
1
14
15
2
8
4
8
3
17
97
# of Streams
w/ Unreported
Volume
9
0
1
5
0
1
0
0
3
1
20
Total Volume
(MT)
2,041.62
2.00
28.80
221.40
102.00
56.00
53.20
345.76
52.40
1,937.60
4,840.78
Average
Volume (MT)
81.66
2.00
2.06
14.76
51.00
7.00
13.30
43.22
17.47
113.98
49.90
Note that 42 percent of desalting sludge volumes are discharged to onsite wastewater
treatment. During engineering site and sampling visits, it was observed that refineries would
simply flush the sludge to wastewater treatment during desalter turnarounds in a manner similar
to mud washing.
Over half of the desalting sludge residuals (48) were reported to be managed as
characteristically hazardous (most commonly DO 18), accounting for 40 percent of the sludge
volume.4 Twenty seven of these streams were managed with F or K listed wastes, reflecting
their frequent management in wastewater treatment systems.
4These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., managed in WWTP, Subtitle C
landfill, transfer as a fuel, etc.).
Petroleum Refining Industry Study
27
August 1996
-------
3.2.2.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.2.1. The Agency
gathered information suggesting other management practices had been used in other years
including: "disposal in onsite Subtitle C landfill" (86 MT), "disposal in onsite surface
impoundment" (1 MT), and "recovery onsite via distillation" (0.5 MT). These non-1992
practices are generally comparable to practices reported in 1992 (i.e., off-site Subtitle C
landfilling and recovery in a coker). The very small volume reported to have been disposed in a
surface impoundment reflects the management of this residual with the refinery's wastewater in
a zero discharge wastewater treatment facility with a final evaporation pond; this management
practice is comparable to the 1992 reported practice of "disposal in onsite wastewater treatment
facility". EPA also compared management practices reported for desalting sludge to those
reported for crude oil tank sediment because of expected similarities in composition and
management. Similar land disposal practices were reported for both residuals.
3.2.2.4 Characterization
Two sources of residual characterization data were developed during the industry study:
• Table 3.2.2 summarizes the physical properties of desalting sludge as reported in
Section VILA of the RCRA §3007 survey.
• Four record samples of desalting sludge were collected and analyzed by EPA. These
sludges represent the various types of desalting operations and sludge generation
methods typically used by the industry and are summarized in Table 3.2.3. The
samples represent sludges generated during turnaround operations (the most common
way desalting sludge is generated), and also represents sludges both with and without
undergoing interim deoiling or dewatering steps.
Table 3.2.4 provides a summary of the characterization data collected under this
sampling effort. The record samples are believed to be representative of desalting sludge as
typically generated by the industry. All four record samples were analyzed for total and TCLP
levels of volatiles, semivolatiles, and metals. Two of three samples analyzed for TCLP Benzene
exhibited the toxicity characteristic for benzene (i.e., the level of benzene in these samples'
TCLP extracts exceeded the corresponding regulatory level). Only constituents detected in at
least one sample are shown in Table 3.2.4.
3.2.2.5 Source Reduction
The electrostatic desalter removes most of the solids, salts and water present in the crude
oil. Minimizing the introduction or recycling of solids to the crude unit will assist the reduction
of desalting sludge, since solids attract oil and produce emulsions.
Petroleum Refining Industry Study 28 August 1996
-------
The amount of desalting sludge formed is a function of the efficiency of the desalter but
more fundamentally is a characteristic of the crude oil. Methods of managing desalting sludge
center on increasing the efficiency of the desalter and de-emulsifiers which increase the
capability of separating the oil, water and solid phases.
Reference
"New Process Effectively Recovers Oil From Refinery
Waste Streams." Oil & Gas Journal. August 15, 1994.
"Filtration Method Efficiently Desalts Crude in
Commercial Test." Oil & Gas Journal. May 17, 1993.
D.T. Cindric, B. Klein, A.R. Gentry and H.M. Gomaa.
"Reduce Crude Unit Pollution With These Technologies."
Hydrocarbon Processing. August, 1 993.
"Waste Minimization in the Petroleum Industry: A
Compendium of Practices." API. November, 1991.
Waste Minimization/Management Methods
Enhanced separation of oil, water and solids.
Alternative process: single-stage filtration.
Includes topic of more effective separation of
phases in desalter.
Practices described: 1 . Shear mixing used
to mix desalter wash water and crude. 2.
Turbulence avoided by using lower pressure
water to prevent emulsion formation.
Petroleum Refining Industry Study
29
August 1996
-------
Table 3.2.2. Desalter Sludge: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Vapor Pressure, mm Hg
Vapor Pressure Temperature, °C
Viscosity, Ib/ft-sec
Viscosity Temperature, °C
Specific Gravity
BTU Content, BTU/lb
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Other, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 1 0-1 00 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
118
60
67
73
103
47
14
9
3
5
69
56
157
151
170
111
10
9
12
9
8
7
#of
Unreported
Values1
144
202
195
189
159
215
248
253
259
257
193
206
105
111
92
151
252
253
250
253
254
255
10th%
6.10
0.15
0.80
43.89
5.00
1.00
0.00
20.00
0.00
0.00
0.90
270.00
0.00
0.00
9.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
50th %
7.00
1.00
82.00
60.00
16.00
15.00
10.50
30.00
0.00
30.00
1.10
3,590.00
30.00
15.00
45.00
0.00
0.00
90.00
10.00
0.00
0.00
200.00
90th %
8.40
250.00
500.00
94.44
70.00
35.00
150.00
40.00
1500.00
50.00
1.70
10,000.00
78.00
50.00
100.00
30.00
50.00
100.00
100.00
100.00
0.00
2,000.00
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
Table 3.2.3. Desalting Sludge Record Sampling Locations
Sample No.
R1-DS-01
R9-DS-01
R11-DS-01
R24-DS-01
Facility
Marathon, Indianapolis, IN
Murphy, Superior, Wl
ARCO, Ferndale, WA
Phibro, Houston, TX
Description
From electrostatic precipitator turnaround.
Sludge/slurry removed directly from unit
Turnaround sludge/slurry taken from drums
Dewatered sludge from turnaround taken from
bins
Continuously generated "solids" from brine
separator; sample mostly aqueous
Petroleum Refining Industry Study
30
August 1996
-------
Table 3.2.4. Desalting Sludge Characterization
Volatile Organics - Method 8260A ug/kg
Acetone
Benzene
n-Butylbenzene
sec-Butylbenzene
Ethylbenzene
Isopropylbenzene
p-lsopropyltoluene
Methylene chloride
Methyl ethyl ketone
n-Propylbenzene
Toluene
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
o-Xylene
m,p-Xylenes
Naphthalene
CAS No.
67641
71432
104518
135988
100414
98828
99876
75092
78933
103651
108883
95636
108678
95476
1 08383 /
106423
91203
R1-DS-01
200,000
230,000
< 62,500
< 62,500
180,000
< 62,500
< 62,500
J 49,000
< 62,500
< 62,500
660,000
350,000
140,000
290,000
950,000
< 62,500
R9-DS-01
< 625
22,000
42,000
24,000
150,000
36,000
25,000
< 625
< 625
74,000
220,000
230,000
85,000
190,000
380,000
55,000
R11B-DS-01
< 1 ,250
28,000
31 ,000
19,000
48,000
27,000
18,000
< 1 ,250
< 1 ,250
44,000
61 ,000
68,000
34,000
54,000
67,000
54,000
( ug/L >
R24-DS-01
160
36
< 5
< 5
J 7
< 5
< 5
< 5
41
< 5
77
35
12
38
70
32
Average
Cone
67,292
93,333
36,500
21 ,500
126,000
31 ,500
21 ,500
16,958
NA
60,167
313,667
216,000
86,333
178,000
465,667
54,500
Maximum Cone
200,000
230,000
42,000
24,000
180,000
36,000
25,000
49,000
NA
74,000
660,000
350,000
140,000
290,000
950,000
55,000
Comments
1
1
1
1
1
TCLP Volatile Organics - Methods 1311 and 8260A ug/L
Acetone
Benzene
Ethylbenzene
Toluene
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
Methylene chloride
o-Xylene
m,p-Xylene
Naphthalene
CAS No.
67641
71432
100414
108883
95636
108678
75092
95476
1 08383 /
106423
91203
R1-DS-01
770
5,200
550
5,200
< 250
< 250
1,000
1,100
2,400
< 250
R9-DS-01
< 50
1,700
340
2,000
190
J 54
1,200
540
1,100
< 50
R11B-DS-01
B 260
280
120
760
J 69
J 22
J 23
200
490
JB 52
Semivolatile Organics - Method 8270B ug/kg
Benzo(a)pyrene
Carbazole
Chrysene
Dibenzofuran
2,4-Dimethylphenol
Fluorene
CAS No.
50328
86748
218019
132649
105679
86737
R1-DS-01
J 4,300
< 13,200
< 6,600
< 6,600
< 6,600
J 6.000
R9-DS-01
J 5,600
< 20,625
< 10,313
12,000
< 10,313
24.000
R11B-DS-01
< 10,000
< 20,000
J 13,000
J 16,000
< 10,000
26.000
R24-DS-01
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
( ug/L >
R24-DS-01
< 5
43
< 5
< 5
190
< 5
Average
Cone
360
2,393
337
2,653
130
38
741
613
1,330
51
Maximum Cone
770
5,200
550
5,200
190
54
1,200
1,100
2,400
52
Comments
1
1
1
Average
Cone
4,950
NA
9,971
1 1 ,533
NA
18.667
Maximum Cone
5,600
NA
13,000
16,000
NA
26.000
Comments
1
ft
I
K
B'
OQ
-------
Table
Sludge Characterization
ciT
1
g,
B'
OQ
1
CC
1-
U)
K>
^
Semivolatile Orqanics - Method 8270B ug/kg (continued)
Phenanthrene
Phenol
Pyrene
1 -Methylnaphthalene
2-Methylnaphthalene
2-Methylchrysene
2-Methylphenol
3/4-Methylphenol
Naphthalene
CAS No.
85018
108952
129000
90120
91576
3351324
95487
NA
91203
TCLP Semivolatile Org
Benzo(a)pyrene
Bis(2-ethylhexyl)phthalate
Di-n-butyl phthalate
2,4-Dimethylphenol
1 -Methylnaphthalene
2-Methylnaphthalene
2-Methylphenol
3/4-Methylphenol
Naphthalene
Phenol
CAS No.
50328
117817
84742
105679
90120
91576
95487
NA
91203
108952
R1-DS-01
J 12,000
< 6,600
< 6,600
48,000
66,000
< 13,200
< 6,600
< 6,600
33,000
R9-DS-01
61 ,000
< 10,313
J 10,000
220,000
330,000
J 13,000
< 10,313
< 10,313
110,000
R11B-DS-01
68,000
< 10,000
< 10,000
180,000
240,000
< 20,000
< 10,000
< 10,000
130,000
( ug/L >
R24-DS-01
26
900
< 5
81
130
< 10
340
530
110
Average
Cone
47,000
NA
8,867
149,333
212,000
13,000
NA
NA
91 ,000
Maximum Cone
68,000
NA
10,000
220,000
330,000
13,000
NA
NA
130,000
Comments
1
anics - Methods 1311 and 8270B ug/L
R1-DS-01
JB 16
< 50
< 50
J 26
J 32
J 34
J 48
J 68
J 86
200
R9-DS-01
< 50
B 500
< 50
< 50
J 50
J 60
J 25
J 40
J 61
< 50
R11B-DS-01
< 50
< 50
J 20
J 73
J 71
J 92
J 43
J 49
120
J 54
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg
Aluminum
Antimony
Arsenic
Barium
Beryllium
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
CAS No.
7429905
7440360
7440382
7440393
7440417
7440439
7440702
7440473
7440484
7440508
7439896
R1-DS-01
2,600
16.0
16.0
2,200
< 0.50
2.90
16,000
110
27.0
680
71 .000
R9-DS-01
3,700
14.0
34.0
1,700
< 0.50
1.80
5,300
76.0
16.0
340
55.000
R11B-DS-01
7,500
< 6.00
16.0
1,400
1.40
3.40
3,300
150
13.0
430
77.000
R24-DS-01
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
(mg/L)
R24-DS-01
11.0
0.28
0.05
1.80
< 0.0025
< 0.0025
230
0.17
< 0.025
1.20
200
Average
Cone
16
200
20
50
51
62
39
52
89
101
Maximum Cone
16
500
20
73
71
92
48
68
120
200
Comments
1
1
Average
Cone
4,600
12.0
22.0
1,767
0.80
2.70
8,200
112
18.7
483
67.667
Maximum Cone
7,500
16.0
34.0
2,200
1.40
3.40
16,000
150
27.0
680
77.000
Comments
-------
Table
Sludge
ciT
1
g,
B'
OQ
1
CC
4"
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg (continued)
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Potassium
Selenium
Sodium
Thallium
Vanadium
Zinc
CAS No.
7439921
7439954
7439965
7439976
7439987
7440020
7440097
7782492
7440235
7440280
7440622
7440666
R1-DS-01
1,100
2,200
310
41.0
17.0
76.0
< 500
140
< 500
< 1.00
36.0
1,300
R9-DS-01
390
3,200
250
4.40
19.0
100
< 500
22.0
< 500
7.00
37.0
1,900
R11B-DS-01
160
3,300
450
39.0
16.0
110
< 500
75.0
< 500
< 1.00
120
5,400
(mg/L)
R24-DS-01
0.36
68.0
1.60
0.0085
< 0.034
0.48
41.0
< 0.0025
830
< 0.005
0.12
2.20
Average
Cone
550
2,900
337
28.1
17.3
95.3
NA
79.0
NA
3.00
64.3
2,867
Maximum Cone
1,100
3,300
450
41.0
19.0
110
NA
140
NA
7.00
120
5,400
Comments
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Aluminum
Barium
Calcium
Chromium
Iron
Magnesium
Manganese
Nickel
Zinc
CAS No.
7429905
7440393
7440702
7440473
7439896
7439954
7439965
7440020
7440666
R1-DS-01
7.60
2.60
580
0.87
210
71.0
6.00
< 0.20
2.00
R9-DS-01
< 1.00
< 1.00
150.00
< 0.05
24.00
< 25.0
1.60
< 0.20
2.90
R11B-DS-01
< 1.00
3.50
54.0
0.12
190
< 25.0
4.60
0.52
57.00
R24-DS-01
NA
NA
NA
NA
NA
NA
NA
NA
NA
Average
Cone
3.20
2.37
261
0.35
141
40.3
4.07
0.31
20.6
Maximum Cone
7.60
3.50
580
0.87
210
71.0
6.00
0.52
57.0
Comments
Comments:
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound that meets the identification criteria for which the result is less than the
laboratory detection limit, but greater than zero.
NA Not Applicable.
-------
3.3 HYDROCRACKING
Petroleum refining hydroprocessing techniques include hydrocracking, hydrorefming,
and hydrotreating. Hydrorefming and hydrotreating processes and their respective catalyst
residuals are described in the Listing Background Document for the November 20, 1995
proposed rule. Hydrocracking processes are similar to hydrotreating and hydrorefming
processes in that they remove organic sulfur and nitrogen from the process feeds, but differ in
that they also serve to break heavier fraction feeds into lighter fractions. As refinery crudes have
become heavier, hydrocracking, a more recent process development compared to long-
established conversion processes such as thermal cracking, has become more widely used. The
current trend to heavier feeds and lighter high-quality feeds causes hydrocracking to offer
advantages to future refining operations.
In addition, hydrocracking is a versatile process, and under mild conditions can be
utilized for hydrotreating (typically fractions that need to be saturated to give good burning
quality) and under more severe conditions can be utilized as a cracker (typically feeds that are
too heavy or too contaminant-laden for catalytic cracking). As a result of this flexibility,
hydrocracking processes can appear in refinery operations in a number of different places.
3.3.1 Process Description
The process flow for hydrocracking is similar to that for hydrotreating: the feed is mixed
with a hydrogen-rich gas, pumped to operating pressure and heated, and fed to one or more
catalytic reactors in series. Hydrocracking units are typically designed with two stages: the first
uses a hydrotreating catalyst to remove nitrogen and heavy aromatics, while the second stage
conducts cracking. The catalysts for each stage are held in separate vessels. Organic sulfur and
nitrogen are converted to H2S and NH3, and some unsaturated olefms or aromatics are saturated
or cracked to form lighter compounds. In addition, heavy metal contaminants are adsorbed onto
the catalyst. Following the reactor, the effluent is separated via stabilization and fractionation
steps into its various fractions. There are two major differences between hydrocracking and
hydrotreating: 1) operating pressures are much higher, in the range from 2,000 - 3,000 lb/in2
gauge, and 2) hydrogen consumption is much higher, in the range from (1,200 - 1,600
SCF/barrel of feed), dependent on the feed. The feed is generally a heavy gas oil or heavier
stream.
Catalysts employed in hydrocracking reactors have multiple functions. First, the catalyst
has a metallic component (cobalt, nickel, tungsten, vanadium, molybdenum, platinum,
palladium, or a combination of these metals) responsible for the catalysis of the hydrogenation
and desulfurization/denitrification reactions. In addition, these metals are supported on a highly
acidic support (silica-alumina, acid-treated clays, acid-metal phosphates, or alumina) responsible
for the cracking reactions. A simplified process flow diagram is shown in Figure 3.3.1.
Petroleum Refining Industry Study 34 August 1996
-------
Figure 3.3.1. Hydrocracking Process Flow Diagram
CokerGasGil
Cat dadoed
Light Gas Cil
H2
^^-—-^ Flash Gas
c
Flash Drum I
Fiacttonator
-Naphtha
Intermediate
^ Naphtha
*> Distillate
3.3.2 Spent Hydrocracking Catalyst
3.3.2.1 Description
Metal deposition acts to deactivate, or poison, the hydrocracking catalyst. In addition,
carbon from the cracking reactions deactivates the catalyst. The catalyst's life is dependent on
the severity of cracking and metal deposition and is changed out every 6 months to 8 years. The
catalyst closest to the entrance (top) of the reactor becomes deactivated first, and for this reason
is sometimes replaced more frequently than the entire reactor contents (a "topping" operation).
When catalyst activity is unacceptable, the reactor is taken out of service and typically undergoes
a hydrogen sweep to burn residual hydrocarbon, then a nitrogen sweep to cool the reactor and
remove occupational hazards such as hydrogen sulfide and benzene. Such procedures were
reported by most facilities. The following additional procedures were reported to be employed
by fewer facilities, typically only one or two:
• Oxidation (to burn residual hydrocarbon)
• Cat nap technology or diesel wash (to lower vapor pressure of hazardous volatiles)
• Wet dump, water wash, or soda ash wash (to neutralize sulfides and remove volatiles)
• Steam stripping (to remove volatiles)
• Evacuation (a technique possibly similar to nitrogen sweep)
• Some facilities report using no pretreatment methods prior to catalyst removal.
Petroleum Refining Industry Study
35
August 1996
-------
In some processes, a moving bed of catalyst is used instead of a fixed bed. In this process,
catalyst is continuously and slowly moved countercurrent to the hydrocarbon flow. Spent
catalyst is generated almost continuously and fresh catalyst added as needed for makeup. This
configuration differs significantly from the fixed bed design with respect to spent catalyst
generation frequency.
Unlike hydrotreating and hydrorefining catalysts (discussed in Listing Background
Document), both precious metal and nonprecious metal catalysts are used in hydrocracking
processes. Based on a total of 46 facilities reporting spent hydrocracking catalyst generation, 34
(74%) reported using nickel/molybdenum, 11 (24%) reported using nickel/tungsten, and 11
(24%) reported using palladium. An additional 16 facilities (35%) reported using other metals
in their catalyst such as cobalt, copper, magnesium, monometallic nickel, phosphorus, tin, and
zinc. As stated in Section 3.3.1, many hydrocracking units are constructed as a hydrorefining
stage followed by a cracking stage. In reporting catalysis use, refineries may not have
differentiated between hydrorefining and cracking functions in their response. In this section,
data for both pretreatment (hydrorefining function) and hydrocracking catalysts are presented.
Approximately 2,500 MT of the hydrocracking catalyst generated in 1992 was identified
as displaying hazardous characteristics.5 This is approximately 15 percent of the total volume
managed. The most commonly displayed hazardous waste codes were D001 (ignitable), D003
(reactive), D004 (TC arsenic) and DO 18 (TC benzene).
3.3.2.2 Generation and Management
During reactor change-outs, spent hydrocracking catalysts are removed from the reactors
using a variety of techniques including gravity dumping and water drilling. Upon removal from
the catalyst bed, the catalyst may be screened to remove fines or catalyst support media. The
catalyst is typically stored in covered bins pending shipment off site for disposal or recovery.
Twenty-eight facilities reported generating a total quantity of 18,000 MT of this residual
in 1992, according to the 1992 RCRA §3007 Questionnaire. Residuals were assigned to be
"spent hydrocracking catalyst" if they were assigned a residual identification code of "spent
solid catalyst" or "spent catalyst fines" and were generated from a process identified as a
hydrocracking unit. These correspond to residual code 03-A in Section VII.2 of the
questionnaire and process code 05 in Section IV-l.C of the questionnaire. Quality assurance
was conducted by ensuring that all hydrocracking catalysts previously identified in the
questionnaire (i.e., in Section V.B) were assigned in Section VII.2.
5These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., Subtitle C landfill, transfer for
metals reclamation, etc.).
Petroleum Refining Industry Study 36 August 1996
-------
Based on the results of the questionnaire, 47 facilities use hydrocracking units and are
thus likely to generate spent hydrocracking catalyst. Due to the infrequent generation of this
residual, not all of these 47 facilities generated spent catalyst in 1992. However, there was no
reason to expect that 1992 would not be a typical year with regard to hydrocracking catalyst
generation and management. Table 3.3.1 provides a description of the quantity generated,
number of streams reported, number of streams not reporting volumes (data requested was
unavailable and facilities were not required to generate it), total and average volumes.
Table 3.3.1. Generation Statistics for Hydrocracking Catalyst, 1992
Final Management
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Reuse onsite as replacement catalyst for
another unit
Transfer metal catalyst for reclamation or
regeneration
Transfer to another petroleum refinery
TOTAL
#of
Streams
7
8
1
45
14
75
# of Streams w/
Unreported Volume
0
0
0
2
0
2
Total Volume
(MT)
1,592.70
991.50
159.40
13,185.56
2,100.00
18,029.16
Average
Volume (MT)
227.53
123.94
159.40
293.01
150.00
295.56
3.3.2.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.3.1. The Agency
gathered information suggesting other management practices had been used in other years
including: "disposal in onsite Subtitle D landfill" (8 MT) and "other recycling, reclamation, or
reuse: cement plant" (320 MT). These non-1992 practices are comparable to 1992 practices
(i.e., off-site Subtitle D landfilling) or to typical practices for alumina-based catalysts (e.g.,
cement plants).
The Agency has no other data to suggest other management practices are used for
hydrocracking catalysts due to the physical characteristics and chemical composition of the
waste. EPA compared the management practice reported for hydrocracking catalysts to those
reported for hydrotreating and hydrorefining catalysts based on expected similarities. Similar
land disposal practices were reported for all three residuals.
3.3.2.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.3.2 summarizes the physical properties of the spent catalyst as reported in
Section VILA of the §3007 survey.
• Three record samples of spent hydrocracking catalyst were collected and analyzed by
EPA and are summarized in Table 3.3.3. The record samples represent the most
frequently used catalysts (i.e., nickel/tungsten and nickel/molybdenum, together used
Petroleum Refining Industry Study
37
August 1996
-------
by well over half of the refineries with hydrocracking processes. In addition, heavy
gas oil or similar distillate/residual feed is the most common application of
hydrocracking reactors, according to the questionnaire. Therefore, the record samples
are expected to represent most of the spent catalyst generated in the industry.
However, another frequently used catalyst (palladium) is not represented, and catalysts
employing feeds other than heavy gas oil (e.g., lube oil) may not have the same
characteristics when spent.
Table 3.3.4 provides a summary of the characterization data collected under this
sampling effort. All three record samples were analyzed for total and TCLP levels of volatiles,
semivolatiles, metals and ignitability. One of three samples exhibited the ignitability
characteristic. Only constituents detected in at least one sample are shown in Table 3.3.4.
Table 3.3.2. Hydrocracking Catalyst Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Specific Gravity
Specific Gravity Temperature, °C
BTU Content, BTU/lb
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Other, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 10-100 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
39
21
38
36
17
14
54
10
4
64
63
101
62
28
42
31
27
27
13
# of Unreported
Values1
102
120
103
105
124
127
87
131
137
77
78
40
79
113
99
110
114
114
128
10th%
5.00
0.30
1.00
60.00
0.00
0.00
0.80
17.80
0.00
0.00
0.00
100.00
0.00
0.00
95.00
0.00
0.00
0.00
0.00
50th %
6.80
3.20
12.50
157.50
0.36
0.63
1.74
20.00
0.00
0.00
0.00
100.00
0.00
0.00
99.00
1.00
0.00
0.00
1,600.00
90th %
9.14
10.00
9,500.00
200.00
9.00
8.00
3.15
25.00
7,485.00
0.00
0.00
100.00
0.00
0.00
100.00
5.00
0.00
0.00
2,000.00
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgement.
Petroleum Refining Industry Study
38
August 1996
-------
Table 3.3.3. Spent Hydrocracking Catalyst Record Sampling Locations
Sample No.
R2-CC-01
R8A-CC-01
R20-CC-01
Facility
Shell, Wood River, IL
Amoco, Texas City, TX
Star, Convent, LA
Description
Nickel/tungsten catalyst, fixed bed, heavy gas oil feed
Nickel/molybdenum catalyst, moving bed, heavy gas oil feed
Mixed nickel/tungsten and nickel/molybdenum catalyst, moving
bed, heavy gas oil feed
3.3.2.5 Source Reduction
There is little that can be done to reduce the quantity of these generated catalyst since, by
design, they must be periodically replaced with fresh catalyst. As a result, the greatest
opportunity for waste minimization arises from sending these materials offsite for metals
regeneration, reclamation, or other reuse.
Refinery hydrocracking catalysts generally consist of cobalt and molybdenum or nickel
and molybdenum on an alumina support. Typically, the catalysts are regenerated after use.
However, industry is interested in finding more specific, long-lasting catalysts. Extensive
research is performed in producing new catalysts. Information on hydrotreating and
hydrorefining catalysts are also presented below because some of this information may be
relevant to hydrocracking catalysts.
Reference
Monticello, D.J. "Biocatalytic Desulfurization."
Hydrocarbon Processing. February, 1994.
"NPRA Q&A 1 : Refiners Focus on FCC,
Hydroprocessing, and Alkylation Catalyst." Oil & Gas
Journal. March 28, 1994.
Gorra, F., Scribano, G., Christensen, P., Anderson, K.V.,
and Corsaro, O.G. "New Catalyst, Improve Presulfiding
Result in 4+ Year Hydrotreater Run." Oil & Gas Journal.
August 23, 1993.
"Petroleum-derived Additive Reduces Coke on
Hydrotreating Catalyst." Oil & Gas Journal. December
27,1993.
"Waste Minimization in the Petroleum Industry: A
Compendium of Practices." API. November, 1991.
Waste Minimization/Management Methods
An alternative to metal catalysts is the
development of microorganisms that can
catalyze the reaction.
Methods in improving catalyst life and
performance.
Material substitution to extend catalyst life.
Process modification extends life of catalyst.
Practices listed: 1 . Metals reclamation, 2.
Recycling to cement, 3. Recycling to
fertilizer plants.
Petroleum Refining Industry Study
39
August 1996
-------
Table 3.3.4.
Hydrocracking
Characterization
Volatile Organics - Method 8260A ug/kg
Acetone
Acrolein
Benzene
n-Butylbenzene
sec-Butylbenzene
Ethyl benzene
Isopropyl benzene
p-lsopropyltoluene
Naphthalene
n-Propylbenzene
Toluene
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
o-Xylene
m,p-Xylenes
CAS No.
67641
107028
71432
104518
135988
100414
98828
99876
91203
103651
108883
95636
108678
95476
108383/106423
R2-CC-02
5,300
2,500
370,000
10,000
< 1,250
35,000
< 1,250
< 1,250
< 1,250
5,000
300,000
25,000
8,000
23,000
60,000
TCLP Volatile Ore
Benzene
Ethyl benzene
Methylene chloride
Toluene
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
o-Xylene
m,p-Xylene
CAS No.
71432
100414
75092
108883
95636
108678
95476
108383/106423
R2-CC-02
10,000
470
< 50
6,600
J 94
< 50
290
750
R8A-CC-01
< 6,250
< 6,250
15,000
40,000
18,000
95,000
34,000
28,000
64,000
49,000
120,000
170,000
48,000
120,000
250,000
R20-CC-01
< 1,250
< 1,250
< 1,250
12,000
13,000
< 1,250
3,700
8,500
7,600
< 1,250
< 1,250
21,000
4,200
3,400
7,100
Average Cone
3,275
1,875
128,750
20,667
10,750
43,750
17,625
12,583
24,283
27,000
140,417
72,000
20,067
48,800
105,700
Maximum Cone
5,300
2,500
370,000
40,000
18,000
95,000
34,000
28,000
64,000
49,000
300,000
170,000
48,000
120,000
250,000
Comments
1
1
anics - Methods 1311 and 8260A ug/L
R8A-CC-01
230
180
250
640
120
< 50
270
410
R20-CC-01
< 50
< 50
< 50
< 50
J 44
J 61
< 50
< 50
Average Cone
3,427
233
117
2,430
86
54
203
403
Maximum Cone
10,000
470
250
6,600
120
61
290
750
Comments
Semivolatile Organics - Method 8270B ug/kg
Acenaphthene
Benz(a)anthracene
Benzofluoranthene (total)
Benzo(q,h,i)perylene
Benzo(a)pyrene
Carbazole
4-Chlorophenyl phenyl ether
Chrysene
Dibenzofuran
7.1 2-Dimethvlbenz(a)anthracene
CAS No.
83329
56553
NA
191242
50328
86748
7005723
218019
132649
57976
R2-CC-02
< 165
< 165
< 165
< 165
< 165
< 330
< 165
< 165
1,200
< 165
R8A-CC-01
20,000
J 6,900
J 5,000
28,000
J 3,100
74,000
< 4,125
17,000
9,700
< 4.125
R20-CC-01
32,000
< 10,313
31,000
42,000
29,000
J 24,000
83,000
68,000
J 13,000
45.000
Average Cone
17,388
3,533
12,055
23,388
10,755
32,777
29,097
28,388
7,967
16.430
Maximum Cone
32,000
6,900
31,000
42,000
29,000
74,000
83,000
68,000
13,000
45.000
Comments
1
ft
I
K
B'
OQ
-------
Table
Spent Hydrocracking Catalyst Characterization
Semivolatile Organics - Method 8270B ug/kg (continued)
Fluoranthene
Fluorene
lndeno(1,2,3-cd)pyrene
3-Methylcholanthrene
2-Methylchrysene
1-Methylnaphthalene
2-Methylnaphthalene
2-Methylphenol
Naphthalene
Phenanthrene
Pyrene
CAS No.
206440
86737
193395
56495
3351324
90120
91576
95487
91203
85018
129000
R2-CC-02
< 165
2,800
< 165
< 165
< 330
< 330
< 165
< 165
< 165
1,200
1,600
R8A-CC-01
20,000
40,000
J 4,600
< 4,125
J 13,000
56,000
110,000
< 4,125
43,000
180,000
430,000
R20-CC-01
25,000
82,000
< 10,313
23,000
64,000
230,000
390,000
J 7,000
45,000
160,000
680,000
Average Cone
15,055
41,600
2,383
9,097
25,777
95,443
166,722
3,763
29,388
113,733
370,533
Maximum Cone
25,000
82,000
4,600
23,000
64,000
230,000
390,000
7,000
45,000
180,000
680,000
Comments
1
TCLP Semivolatile Organics - Methods 1311 and 8270B ug/L
Carbazole
2,4-Dimethylphenol
1-Methylnaphthalene
2-Methylnaphthalene
2-Methylphenol
3/4-Methylphenol (total)
Naphthalene
Phenol
Phenanthrene
Pyrene
CAS No.
86748
105679
90120
91576
95487
NA
91203
108952
85018
129000
R2-CC-02
< 100
< 50
< 100
< 50
J 66
J 76
< 50
J 53
< 50
< 50
R8A-CC-01
J 78
J 23
J 24
J 46
J 70
J 49
J 44
< 50
J 23
J 42
R20-CC-01
< 100
J 44
J 41
J 59
J 25
J 17
J 26
J 63
< 50
< 50
Average Cone
78
34
33
52
54
47
35
55
23
42
Maximum Cone
78
44
41
59
70
76
44
63
23
42
Comments
1
1
1
1
1
1
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg
Aluminum
Antimony
Arsenic
Beryllium
Chromium
Cobalt
Copper
Iron
Lead
Manganese
Molybdenum
Nickel
CAS No.
7429905
7440360
7440382
7440417
7440473
7440484
7440508
7439896
7439921
7439965
7439987
7440020
R2-CC-02
120,000
< 6.0
12.0
< 0.5
130
24.0
55.0
52,000
< 0.3
390
< 6.5
19,000
R8A-CC-01
53,000
220
29.0
160
68.0
440
35.0
2,200
15.0
16.0
5,400
28,000
R20-CC-01
110,000
< 6.0
< 5.0
18.0
< 1.0
< 5.0
< 2.5
570
1.6
< 1.5
17,000
27,000
Average Cone
94,333
77.3
15.3
59.5
66.3
156
30.8
18,257
5.6
136
7,469
24,667
Maximum Cone
120,000
220
29.0
160
130
440
55.0
52,000
15.0
390
17,000
28,000
Comments
g
OQ
-------
Petroleum Refining Industry Study
i auie J.J.+. ^pem nyuruiraiKiiig ^ aiaiysi ^ iiaracieri/;aiiuii ^uiiimueuj
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg (continued)
Selenium
Sodium
Vanadium
Zinc
CAS No.
7782492
7440235
7440622
7440666
R2-CC-02
< 0.5
1,200
37.0
82.0
R8A-CC-01
4.0
2,000
140,000
110
R20-CC-01
< 0.5
< 500
49,000
< 2.0
Average Cone
1.7
1,233
63,012
64.7
Maximum Cone
4.0
2,000
140,000
110
Comments
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Aluminum
Chromium
Iron
Manganese
Nickel
Vanadium
Zinc
CAS No.
7429905
7440473
7439896
7439965
7440020
7440622
7440666
R2-CC-02
26.0
0.35
130
10.0
110
< 0.25
0.58
R8A-CC-01
< 1.00
< 0.05
< 0.50
< 0.08
3.60
4.70
< 0.10
R20-CC-01
< 1.00
< 0.05
< 0.50
< 0.08
0.43
< 0.25
< 0.10
Average Cone
9.33
0.15
43.7
3.38
38.0
1.73
0.26
Maximum Cone
26.00
0.35
130
10.0
110
4.70
0.58
Comments
Miscellaneous Characterization
Iqnitability (oF)
R2-CC-02
138
R8A-CC-01
145
R20-CC-01
NA
Comments:
>
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound that meets the identification criteria for which the result is less than the laboratory detection limit, but greater than
zero.
ND Not Detected.
ND Not Applicable.
-------
Petroleum Refining Industry Study 43 August 1996
-------
STUDY OF SELECTED
PETROLEUM REFINING RESIDUALS
INDUSTRY STUDY
Part 2
August 1996
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Solid Waste
Hazardous Waste Identification Division
401 M Street, SW
Washington, DC 20460
-------
3.4 ISOMERIZATION
The purpose of isomerization is to increase the refinery's production of high octane, low
aromatic gasoline. Gasoline with low benzene and aromatics is newly specified in the California
market and is expected to be adopted by other states in the future (Oil & Gas Journal, 1995)1.
3.4.1 Isomerization Process Description
Principal applications of isomerization at refineries are naphtha isomerization, which
produces a gasoline blending component, and butane isomerization, which produces isobutane
feed for the alkylation unit. Figure 3.4.1 depicts a generic process flow diagram for
isomerization. Based on the results of the RCRA §3007 questionnaire, 65 facilities reported
having isomerization units, distributed as follows (some facilities have more than one type of
isomerization unit):
• 47 facilities have naphtha isomerization units
• 15 facilities have butane isomerization units
• 7 facilities have other types of isomerization units.
Figure 1.1.1. Isomerization Process Flow Diagram
C4 +
Feed
FhelGas
Drying
H2
Spent
MoLeeuhr
Sieve
r
Spent
Catalyst
Reactor
a
Caustic
Wash
Caustic
Iso nierate
Make-up
Caustic
lOil & Gas Journal, "Deadline Looming for California Refineries to Supply Phase IIRFG," December 11,
1995, pages 21-25.
Petroleum Refining Industry Study
49
August 1996
-------
3.4.1.1 Naphtha Isomerization
Gasoline, or naphtha, is generated throughout the refinery and consists of a mix of C5 and
higher hydrocarbons in straight, branched, or ring configuration. Naphtha isomerization
converts the straight chains to branched, significantly raising their octane number. A common
source of such "low grade" naphtha is light straight run, which consists of the lighter fraction
(C5/C6) of naphtha from atmospheric crude distillation. The reduction of lead in gasoline in the
1970s increased the demand for isomerization technology; prior to that time naphtha
isomerization was not widely used (Meyers, 1986).
As found from the RCRA §3007 questionnaire results, the most common naphtha
isomerization processes presently used in the industry are UOP's Penex process and Union
Carbide's Total Isomerization Process (TIP). Other licensed processes used include the Union
Carbide Hysomer process and the BP Isomerization process. In these four processes, naphtha is
combined with hydrogen and flows through one or two fixed bed reactors in series; the catalyst
consists of a precious metal catalyst on a support (non-precious metal catalysts are rarely, if
ever, used for naphtha isomerization). The reactor effluent is sent to a series of columns where
hydrogen and fuel gas are separated from the isomerate product. The isomerate, having a
significantly higher octane number than the light straight run feed, is charged to the gasoline
blending pool. Although the isomerization reaction is not a net consumer or producer of
hydrogen, the presence of hydrogen prevents coking and subsequent deactivation of the catalyst
(Meyers, 1986).
From a solid waste generation perspective, the principal differences between the various
processes relate to the catalyst used; this will in turn affect the feed pretreatment steps and spent
catalyst characterization. The two principal types of catalyst identified in the industry are: (1)
platinum on zeolite, which operates at temperatures above 200°C, and (2) platinum chloride on
alumina, which operates at temperatures below 200°C. The higher temperatures are
characteristic of the TIP and Hysomer processes, while the lower temperatures are characteristic
of the Penex process and the BP process. The effect of these two different precious metal
catalysts on the process are as follows:
• Dioxin formation. To maintain an environment of hydrogen chloride in the
reactor required for catalyst activity, the platinum chloride catalyst requires a
small but continuous addition of a chlorinated organic compound (e.g., carbon
tetrachloride) to the feed. Although no oxygen is present during operating
conditions, the conditions encountered during unit turnaround and catalyst
removal (see Section 3.4.3) could result in dioxin formation. During sampling
and analysis, the Agency tested for dioxin and the results are presented in Table
3.4.4.
Unlike reforming unit catalyst (a platinum catalyst discussed in the Listing
Background Document), the isomerization unit catalyst apparently does
not undergo in situ regeneration. One refinery stated that they do not
conduct regeneration because coke does not form and contaminate the
catalyst (making regeneration unnecessary), and design information for
these units does not mention in situ regeneration.
Petroleum Refining Industry Study 50 August 1996
-------
• Feed pretreatment. The platinum chloride catalyst, operating at the lower
temperatures, provides better conversion of paraffins to isomers. However, this
catalyst is susceptible to water, sulfur, and nitrogen as catalyst poisons (Meyers,
1986). To combat these contaminants, the feed is commonly desulfurized over a
cobalt/molybdenum or similar catalyst and generated H2S is removed prior to the
isomerization reactor. To further protect against sulfur poisoning, some processes
include a guard column between the hydrodesulfurization reactor and the
isomerization reactor to remove additional sulfur-containing compounds. Rather
than consisting of Co/Mo (like many hydrotreating catalysts), this guard column
often consists of zinc oxide, nickel on alumina, or copper oxide.
To remove water from the desulfurized naphtha, the hydrocarbon feed is typically
dried using molecular sieve. When the molecular sieve is saturated, it is taken
off-line for water desorption while the hydrocarbon is rerouted to a parallel
molecular sieve vessel. In a similar way, water is removed from the hydrogen
feed. Certain molecular sieves can remove both sulfur compounds and water
from hydrogen or hydrocarbon feeds.
The platinum on zeolite catalyst is less susceptible to poisoning by these
contaminants and reportedly requires none, or significantly less, pretreatment
(Meyers, 1986).
Another difference in operating practices found among individual refineries is product
stream recycling to increase yield and octane. These qualities can be increased by (1) recycling
the paraffins to the reactor following their separation from the isomerized product, or (2)
separating (and effectively concentrating) low octane paraffins from other high octane feed
components such as isomers and aromatics. These steps can be performed using either
conventional fractionation or an adsorbent. In the latter case, the normal paraffins are adsorbed
onto zeolite or another adsorbent while the isomers pass through. The paraffins are desorbed
and introduced as isomerization reactor feed, while the isomers bypass the isomerization reactor
and are introduced to a post reactor stabilizer. Not all refineries conduct such separation,
although separation of the feed or product using molecular sieve is integral to the Union Carbide
Total Isomerization Process.
3.4.1.2 Butane Isomerization
The purpose of butane isomerization is to generate feed material for a facility's alkylation
or MTBE production unit; alkylation unit feed includes isobutane and olefms, while the raw
materials used in making MTBE are isobutylene and methanol. Butane isomerization is a much
older process than naphtha isomerization, having been used in refineries since World War II.
Presently, the most prevalent method of producing isobutane from n-butane is the UOP Butamer
process, similar in many ways to the isomerization of naphtha over platinum chloride catalyst.
In the Butamer process, normal butane, generated from throughout the refinery and separated
from other butanes by distillation, is combined with hydrogen and a chlorinated organic
compound. The hydrogen is used to suppress the polymerization of olefm intermediates, while
the chlorine source is used to maintain catalyst activity. The feed flows through one or two
fixed bed reactors in series, containing platinum chloride on alumina catalyst. The isobutane
Petroleum Refining Industry Study 51 August 1996
-------
product is recovered and used as alkylation unit feed. Butane isomerization takes place at lower
temperatures than naphtha isomerization.
Like platinum chloride catalyst used in naphtha isomerization, the Butamer catalyst is
poisoned by water and sulfur, as well as fluoride (Meyers, 1986). These compounds are
removed from the hydrogen and hydrocarbon feed by molecular sieve.
Although the Butamer process and others using platinum chloride on alumina as a
catalyst dominate the industry, other technologies are also used. Three facilities conducting
butane isomerization do not use platinum catalysts. Instead, the catalyst is aluminum
chloride/hydrochloric acid and generates an almost continuous spent catalyst waste stream in
slurry or sludge form.
3.4.1.3 Other Isomerization Processes
Seven facilities reported using isomerization for purposes other than naphtha or butane
isomerization. Such applications demonstrate the integration of petroleum refining and chemical
production at many refineries. Some of these processes more closely represent petrochemical
production than refining processes because they are not widely reported by refineries as a
refining step, are not used for fuel production, and produce commodity chemicals. The
processes reported by these seven facilities can be classified into three areas:
• Xylene Isomerization. Four facilities report processes to convert xylene isomers
(e.g., from an extraction process) to p- and/or o-xylene. Unlike the naphtha and
butane isomerization units described above, the catalyst is not precious metal.
The xylene products are sold.
• Cyclohexane Isomerization. Two facilities produce cyclohexane from raw
materials that include benzene and hydrogen. Unlike the naphtha and butane
isomerization units described above, the catalyst is not precious metal.
• Butylene Isomerization. Two facilities produce butylene from various C4
olefms. Butylene is used for feed to the alkylation unit. A precious metal
(palladium) catalyst is used.
3.4.2 Isomerization Catalyst
3.4.2.1 Description
As discussed in Section 3.4.1, the most prevalent catalyst used for both butane and
naphtha isomerization is platinum or platinum chloride on alumina or zeolite. When the catalyst
loses activity, it is removed from the reactor and replaced with fresh catalyst. Prior to removal,
the reactor may be swept to remove hydrocarbons from the catalyst. These preparation steps can
include one or more of the following:
• Nitrogen sweep (to remove hydrocarbon)
• Oxygen sweep (to burn hydrocarbon)
Petroleum Refining Industry Study 52 August 1996
-------
• Steam stripping (to remove hydrocarbon).
This procedure of catalyst preparation, removal, and replacement is relatively lengthy (typically
one week or more) and requires the unit, or at least the reactor, to be shut down such that no
hydrocarbon is processed during the time of catalyst replacement.
There are a handful of isomerization processes used at domestic refineries that do not use
platinum or platinum chloride catalyst. At these facilities, spent catalyst is generated in one of
the following two methods:
• A method similar to the generation of spent platinum/platinum chloride catalyst
described above. Fixed-bed processes are used in both palladium and non-
precious metal catalyst applications and spent solid catalyst is infrequently
removed.
• A method where catalyst is removed from the fixed-bed reactor frequently (up to
once a day) in liquid/semi-solid form, presumably with little to no disruption of
the process. This method is used only for one process which uses aluminum
chloride/hydrochloric acid catalyst.
Another type of catalyst seen in conjunction with an isomerization unit is desulfurization
catalyst. In many naphtha isomerization processes, the feed typically contains high levels of
mercaptans which are converted to H2S over a non-precious metal catalyst, such as
cobalt/molybdenum. Such catalysts were discussed in the Listing Background Document under
the broad name of "hydrotreating catalysts" and will not be discussed here.
3.4.2.2 Generation and Management
The spent catalyst is vacuumed or gravity dumped from the reactors. Based on
information from site visits, most refineries place the material directly into closed containers
such as 55-gallon drums, flow-bins, or 1 cubic yard "supersacks." The frequency of generation
is typically between 2 and 10 years, with a small number of facilities generating a slurry/sludge
continuously. In 1992, only one facility reported classifying this residual as RCRA hazardous
(this facility classified 2 MT as D001).2 In other years, some facilities reported that this residual
carried a RCRA hazardous waste code of DO 18 (TCLP benzene).
Eighteen facilities reported generating a total quantity of 337 MT of this residual in
1992, according to the 1992 RCRA §3007 Questionnaire. The questionnaire reported that 65
facilities have isomerization units and thus are likely to generate spent isomerization catalyst at
some time. Due to the infrequent generation of this residual, not all of these 65 facilities
generated spent catalyst in 1992. However, there was no reason to expect that 1992 would not
be a typical year with regard to this residual's generation and management.
2These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., Subtitle C landfill, transfer for
metals reclamation, etc.).
Petroleum Refining Industry Study 53 August 1996
-------
Petroleum Refining Industry Study 54 August 1996
-------
Residuals were assigned to be "spent isomerization catalyst" if they were assigned a
residual identification code of "spent solid catalyst" and were generated from a process
identified as an isomerization unit. These correspond to residual code 03-A in Section VILA of
the questionnaire and process code 10 in Section IV. C of the questionnaire. The small volume
of continuously generated residuals (discussed in Section 3.4.3.1) were typically omitted from
these statistics, because they were most often characterized as sludges. However, as stated in
Section 3.4.1, some nonprecious metal catalysts are also used in fixed bed processes and are
included in these statistics. Table 3.4.1 provides a description of the 1992 management
practices, quantity generated, number of streams reported, number of streams not reporting
volumes (data requested was unavailable and facilities were not required to generate it), total and
average volumes.
Table 3.4.1. Generation Statistics for Catalyst from Isomerization, 1992
Final Management
Disposal in offsite Subtitle C landfill
Transfer metal catalyst for
reclamation or regeneration
TOTAL
#of
Streams
3
17
20
# of Streams w/
Unreported Volume
0
0
0
Total
Volume (MT)
43.79
293.40
337.19
Average
Volume (MT)
14.60
17.26
16.86
J.Z.Z.J
Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.4.1. The Agency
assessed information reported for other years but no additional management practices were
reported for this residual. In addition, EPA compared the management practice reported for
isomerization catalysts to those reported for reforming catalysts (a listing residual described in
the Listing Background Document) based on expected similarities. The vast majority of both
wastes are reclaimed due to their precious metal content.
3.4.2.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.4.2 summarizes the physical properties of the spent catalyst as reported in
Section VILA of the §3007 survey.
• Four record samples of spent isomerization catalyst were collected and analyzed
by EPA. These spent catalysts represent the majority of processes used by the
industry. Sampling information is summarized in Table 3.4.3.
Petroleum Refining Industry Study
55
August 1996
-------
Table 3.4.2. Catalyst from Isomerization: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Specific Gravity
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Other, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 10-100 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
20
14
16
15
14
13
23
32
32
52
29
11
22
13
13
11
9
# of Unreported
Values1
51
57
55
56
56
58
48
39
39
19
42
60
49
58
58
60
62
10th%
3.35
0.04
0.90
60.00
0.00
0.00
0.65
0.00
0.00
95.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
50th %
4.50
1.00
3.00
100.00
0.50
0.20
1.08
0.00
0.00
100.00
0.00
0.00
100.00
0.00
0.00
0.00
1,590.00
90th %
7.75
11.60
100.00
200.00
1.00
3.00
3.00
0.00
1.00
100.00
0.00
100.00
100.00
1.00
5.00
0.00
2,000.00
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgement.
Table 3.4.3. Spent Isomerization Catalyst Record Sampling Locations
Sample number
R5B-IC-01
R8B-IC-01
R18-IC-01
R23B-IC-01
Facility
Marathon, Garyville LA
Amoco, Texas City TX
Ashland, Canton OH
Chevron, Salt Lake
City, UT
Description: Process Name/Catalyst Type
Butane isomerization (UOP Butamer process),
platinum chloride catalyst
Naphtha isomerization (UOP Penex process),
platinum chloride catalyst
Naphtha isomerization (UOP Penex process),
platinum chloride catalyst
Butane isomerization (UOP Butamer process),
platinum chloride catalyst
Petroleum Refining Industry Study
56
August 1996
-------
The collected samples are expected to be representative of processes using platinum
chloride catalyst. Other processes use platinum catalyst or (rarely) non-precious metal catalysts.
Because similar feeds are processed by most isomerization processes, these spent catalysts are
expected to display similar characteristics, with the following exceptions: (1) spent platinum
chloride catalysts (and possibly aluminum chloride/hydrogen chloride catalysts) are the only
catalysts expected to contain dioxins, because of the presence of chlorine in the process, (2)
platinum chloride catalysts require cleaner feed (i.e., water and sulfur are catalyst poisons), and
thus concentrations of some contaminants may be greater in spent catalysts from processes not
using platinum chloride catalysts.
All four record samples were analyzed for total and TCLP levels of volatiles,
semivolatiles, metals, and reactivity (pyrophoricity). Three samples were analyzed for total
levels of dioxins/furans. Three of the four samples were found to exhibit the toxicity
characteristic for benzene (i.e., the level of benzene in these samples' TCLP extracts exceeded
the corresponding regulatory level). A summary of the analytical results is presented in Table
3.4.4. Only constituents detected in at least one sample are shown in this table.
3.4.2.5
Source Reduction
As in the case of the hydrocracking catalyst, source reduction methods are those that
extend the life of the catalyst. Currently, recycling of the spent catalyst by sending to metals
reclamation is a common practice since the catalyst is platinum.
Reference
J. Liers, J. Mensinger, A. Mosch, W. Reschefilowski.
"Reforming Using Erionite Catalysts." Hydrocarbon
Processing. Aug. 1993.
Waste Minimization Methods
The platinum catalyst together with erionite
increases isomerization.
Petroleum Refining Industry Study
57
August 1996
-------
Petroleum Refining Industry Study 58
i iiuic j.-t.-t. ixir^iiiuai ^ IKII
I
-------
Petroleum Refining Industry Study 59
i auie J.J.+. rvcMuuai ^ nai aciei izauun uuiu lui jpeni ASUHICI izauun ^aiaiysi ^uiuiiuieu;
Semivolatile Organics - Method 8270B tig/kg
Bis(2-ethylhexyl) phthalate
7,1 2-Dimethylbenz(a)anthracene
Isophorone
2,4-Dimethylphenol
2-Methylphenol
3/4-Methylphenol
Phenol
CAS No.
117817
57976
78591
105679
95487
NA
108952
R5B-IC-01
< 165
] 73
1,200
< 165
< 165
< 165
< 165
R8B-IC-01
] 410
] 600
15,000
< 413
< 413
< 413
< 413
R18-IC-01
710
< 165
] 220
1,000
640
1,500
1,700
R23B-IC-01
< 165
< 165
2,700
< 165
< 165
< 165
< 165
Average Cone
363
251
4,780
436
346
561
611
Maximum Cone
710
600
15,000
1,000
640
1,500
1,700
Comments
TCLP Semivolatile Organics - Methods 1311 and 8270B \iglL
Bis(2-ethylhexyl) phthalate
Di-n-butyl phthalate
2,4-Dimethylphenol
2-Methylphenol
3/4-Methylphenol (total)
Phenol
Isophorone
CAS No.
117817
84742
105679
95487
NA
108952
78591
R5B-IC-01
< 50
] 31
< 50
< 50
< 50
< 50
< 50
R8B-IC-01
< 25,000
< 25,000
320,000
140,000
870,000
< 25,000
] 6,700
R18-IC-01
] 18
] 50
] 53
140
240
840
< 50
R23B-IC-01
< 50
< 50
< 50
< 50
< 50
< 50
< 50
Average Cone
18
44
51
80
113
313
NA
Maximum Cone
18
50
53
140
240
840
NA
Comments
1,2
2
2
2
2
2
2
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg
Aluminum
Arsenic
Chromium
Copper
Iron
Nickel
Zinc
CAS No.
7429905
7440382
7440473
7440508
7439896
7440020
7440666
R5B-IC-01
460,000
< 1.00
20.0
< 2.50
< 10.0
14.0
< 2.00
R8B-IC-01
130,000
< 1.00
17.0
< 2.50
54.0
10.0
< 2.00
R18-IC-01
260,000
26.0
17.0
< 2.50
190
< 4.00
< 2.00
R23B-IC-01
230,000
< 1.00
17.0
5.50
73.0
< 4.00
9.2
Average Cone
270,000
7.25
17.8
3.25
81.8
8.00
3.80
Maximum Cone
460,000
26.0
20.0
5.50
190
14.0
9.20
Comments
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Aluminum
Chromium
Iron
Lead
Manganese
Zinc
CAS No.
7429905
7440473
7439896
7439921
7439965
7440666
R5B-IC-01
620
< 0.05
2.40
< 0.015
< 0.08
B 0.45
R8B-IC-01
560
< 0.05
< 0.50
< 0.015
< 0.08
B 0.41
R18-IC-01
380
0.13
7.60
0.045
0.42
B 0.70
R23B-IC-01
450
< 0.05
< 0.50
< 0.015
< 0.08
B 0.53
Average Cone
503
0.07
2.75
0.023
0.16
0.52
Maximum Cone
620
0.13
7.60
0.045
0.42
0.70
Comments
>
I
-------
Petroleum Refining Industry Study
i auie J.J.+. rvesiuuai ^ iiaracieri/;aiiuii uaia lur ^peiii isuiiierizauuii ^ aiaiysi ^uiuiiiueuj
Dioxins/Furans - Method 8290 ng/kg
2,3,7,8-TCDF
Total TCDF
2,3,4,6,7,8-HxCDF
Total HxCDF
1,2,3,4,6,7,8-HpCDF
Total HpCDF
1,2,3,4,6,7,8-HpCDD
Total HpCDD
OCDF
OCDD
2,3,7,8-TCDD Equivalence
CAS No.
51207319
55722275
60851345
55684941
67562394
38998753
35822469
37871004
39001020
3268879
1746016
R5B-IC-01
< 0.13
< 0.13
0.32
B 0.32
0.42
2.10
3.00
B 3.00
3.70
B 43.0
0.11
R8B-IC-01
< 0.20
< 0.20
< 0.34
< 0.34
< 0.26
< 0.26
< 0.60
< 0.60
< 0.80
B 1.70
0.0017
R18-IC-01
B 0.69
B 0.69
< 0.50
< 0.50
< 0.37
< 0.37
< 0.50
< 0.50
< 0.55
B 1.70
0.071
R23B-IC-01
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Average Cone
0.34
0.34
0.32
0.32
0.35
0.91
1.37
1.37
1.68
15.47
0.06
Maximum Cone
0.69
0.69
0.32
0.32
0.42
2.10
3.00
3.00
3.70
43.00
0.11
Comments
1
1
Comments:
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
2 TCLP Semivolatile Organic results for sample R8B-IC-01 are excluded from the calculations.
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound that meets the identification criteria for which the result is less than the laboratory detection limit, but greater than
zero.
NA Not Applicable.
-------
3.4.3 Isomerization Treating Clay
3.4.3.1 Description
Not all facilities with isomerization units use "treating clay," or adsorbents. However,
solid adsorbents can be used in three places in the isomerization process:
• Hydrocarbon feed purification. Processes using platinum chloride catalysts require
a purified feed. Both spent molecular sieve (for drying) and spent metal-alumina (for
sulfur removal) are generated.
• Hydrogen feed purification. Processes using platinum chloride catalysts require dry
hydrogen gas. Spent molecular sieve is generated.
• Paraffin separation of the feed or product. Various types of processes use
adsorbents for paraffin separation. Molecular sieve is the most common adsorbent for
this application.
All of these adsorbents go through adsorption/desorption cycles. Over time, the adsorbent loses
its capacity or efficiency and is removed from the vessel and replaced with fresh adsorbent.
Prior to removal, the vessel can be swept to remove light hydrocarbons and hydrogen sulfide
from the vessel. Typically, processes use adsorbent beds in parallel so that one bed can be on-
line (adsorption mode) while the second is off-line for desorption or replacement.
3.4.3.2 Generation and Management
When spent, adsorbents from isomerization are vacuumed or gravity dumped from the
vessels. Interim storage can include 55-gallon drums, flow-bins, dumpsters, or piles. The
frequency of generation is highly dependent on the generating process: isomerization adsorbents
are typically generated approximately every 5 years, while extraction clay is typically generated
once per year or less. According to questionnaire results, 6 facilities reported classifying 39.5
MT of this residual as RCRA hazardous in 1992 (most typically as D018, D001, and D006).3
This is consistent with reporting for other years.
Twenty-two facilities reported generating a total quantity of approximately 597 MT of
this residual in 1992, according to the 1992 RCRA §3007 Questionnaire. The questionnaire
reported that 65 facilities have isomerization units. However, not all of these facilities use clay,
molecular sieve, or other adsorbents in their process; 25 percent of facilities with isomerization
units did not report generating any clay residual for their process in any year, indicating either
that clay is either not used, has not yet been replaced, or is generated so infrequently that
respondents could not recall when, if ever, the clay was last replaced. In addition, these
adsorbents may be replaced less often than once per year or not in 1992, particularly those
3These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., Subtitle C landfill, transfer for
metals reclamation, etc.).
Petroleum Refining Industry Study 61 August 1996
-------
associated with the isomerization process. However, there was no reason to expect that 1992
would not be a typical year with regard to this residual's generation and management.
Residuals were assigned to be "spent clay from isomerization" if they were assigned a
residual identification code of "spent sorbent" and were generated from a process identified as
an isomerization or extraction unit. These correspond to residual code 07 in Section VILA of
the questionnaire and process code 10 in Section IV.C of the questionnaire. Table 3.4.5 provides
a description of the 1992 management practices, quantity generated, number of streams reported,
number of streams not reporting volumes (data requested was unavailable and facilities were not
required to generate it), total and average volumes.
Table 3.4.5. Generation Statistics for Treating Clay from Isomerization, 1992
Final Management
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite Subtitle C landfill
Disposal in onsite Subtitle D landfill
Other discharge or disposal offsite:
broker
Other recycling, or reuse: cement
plant
Other recycling, or reuse: onsite road
material
Storage in pile
Transfer metal catalyst for reclamation
or regeneration
TOTAL
#of
Streams
14
6
1
2
2
2
4
7
5
43
# of Streams w/
Unreported Volume
0
0
0
0
0
0
0
0
0
0
Total Volume
(MT)
202
140
18
46.8
14
2.5
138
19.7
15
596
Average
Volume (MT)
14.4
23.3
18
23.4
7
1.25
34.5
2.8
3
13.8
3.4.3.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.4.5. The Agency
gathered information from other years but no additional management practices were reported for
this residual. In addition, EPA compared the management practice reported for isomerization
treating clay to those reported for treating clays from extraction, alkylation, and lube oil4 based
on expected similarities. Land treatment was reported for these other types of treating clays,
therefore it is likely that land treatment is a plausible management practice for clays from
isomerization.
4EPA did not compare these management practices to those reported for the broader category of "treating clay
from clay filtering" due to the diverse types of materials included in this miscellaneous category.
Petroleum Refining Industry Study
62
August 1996
-------
3.4.3.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.4.6 summarizes the physical properties of the spent adsorbents as reported in
Section VILA of the §3007 survey.
• One record sample of spent adsorbents from isomerization were collected and
analyzed by EPA. The isomerization treating clay was categorized with the extraction
clay in the consent decree, therefore, the sampling information is summarized with the
extraction clay in Table 3.4.7.
The one record sample was analyzed for total and TCLP levels of volatiles,
semivolatiles, and metals, and ignitability. The sample was not found to exhibit a hazardous
waste characteristic. A summary of the results is presented in Table 3.4.7. Only constituents
detected in at least one sample are shown in this table.
3.4.3.5 Source Reduction
Treating clay for isomerization is generally used as a method of prolonging the life of the
catalyst or for product polishing. Because they are used as a source reduction technique for
other residuals, no source reduction methods for the clays were found.
Petroleum Refining Industry Study 63 August 1996
-------
Table 3.4.6. Treating Clay from Isomerization: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Specific Gravity
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 10-100 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
37
22
27
20
20
18
31
50
51
75
22
32
23
20
20
9
# of Unreported
Values1
71
86
81
88
85
87
77
58
57
33
86
76
85
88
88
98
10th%
5.9
0
0
19.17
0
0
0.8
0
0
97.5
0
0
0
0
0
0
50th %
7
1
1
60
0.75
0.18
1.2
0
0
100
0
100
0
0
0
2
90th %
9.4
10
100
131.7
1.5
2
2.2
3.5
0.1
100
100
100
7.5
0
0
3000
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgement.
Table 3.4.7. Isomerization Spent Sorbent Record Sampling Locations
Sample number
R23B-CI-01
Facility
Chevron, Salt Lake City UT
Description
Molecular sieve, drying butane feed prior to
isomerization
Petroleum Refining Industry Study
64
August 1996
-------
3.5 EXTRACTION
Extraction processes separate more valuable chemical mixtures from a mixed aromatic
and paraffinic stream. At refineries, extraction processes most commonly fall into two types: (1)
"heavy end" extraction, commonly used in lube oil manufacture and deasphalting operations to
upgrade and further process gas oils, and (2) gasoline component extraction, commonly used to
separate some of the more valuable aromatics from naphtha. "Heavy end" extraction is
discussed with other residual upgrading technologies in Section 3.8 of this document. The
gasoline component extraction processes are discussed here.
3.5.1 Extraction Process Description
Thirty facilities reported using gasoline component extraction processes in their
refineries. By far the most common type of gasoline component extraction process conducted at
refineries, according to the RCRA §3007 questionnaire, is the recovery of benzene, toluene, and
mixed xylenes from reformate (i.e., the product from a catalytic reforming unit) for sales or
further processing. Most extraction units actually consist of two sections in series: an extraction
section, which separates aromatics from non-aromatics using continuous liquid-liquid extraction,
and a distillation section, which separates the various aromatic compounds from each other in a
series of fractionation towers. Figure 3.5.1 depicts a generic extraction process flow diagram.
Figure 1.2.1. Extraction Process Flow Diagram
Rafl irate
In the extraction section, the charge is countercurrently contacted with a solvent. The
solvent is most commonly sulfolane, C4H8SO2, or tetraethylene glycol,
O(CH2CH2OCH2CH2OH)2, although a small number of facilities use diglycol amine,
O(CCO)CCN. The raffmate is separated from the aromatic-rich solvent in a tower. The
aromatic-poor raffmate is water-washed to remove solvent and used elsewhere in the refinery.
The aromatic-rich extract is also water-washed to remove solvent and the aromatics sent to the
distillation section for separation into benzene, toluene, and xylenes.
Petroleum Refining Industry Study
65
August 1996
-------
In the distillation section, the aromatic extract is distilled to remove benzene from the top
of the column; the bottoms are sent to the next column. In successive columns, toluene and
finally xylene are removed. The bottoms from the xylene tower (C9 aromatics) are sent to
gasoline blending. Some facilities omit the distillation section altogether, using their extraction
unit simply to produce low and high octane blending stocks.
To decrease the unit's loading, the feed can be separated prior to extraction so that only
the most desirable fractions, such as C6 to C8, are upgraded. This eliminates a final distillation
step and eliminates a heavy aromatic stream as a product from the benzene-toluene-xylene
separation.
Several other gasoline component extraction processes are each reported by only 1 or 2
refineries in the industry. Other refineries may use these processes, but did not report them
because of their resemblance to petrochemical operations of solvent manufacture, etc., which
some refineries considered out of the survey scope. As a result, the database may not accurately
reflect the incidence of these processes. These processes are as follows:
• The UOP Parex process separates p-xylene from mixed C8 aromatics. C8 feed is
injected countercurrently to a bed of solid adsorbent, which adsorbs p-xylene. The bed
is then desorbed and the p-xylene is recovered in the extract for use in petrochemical
production. This process is typically associated with a xylene isomerization process
(Meyers, 1986). This arrangement differs from the overwhelming majority of
extraction processes, which are associated with reforming processes.
• The Union Carbide IsoSiv process separates normal C6-C8 paraffins from the other
branched and ring compounds present in light straight run. In this process, the
paraffins are adsorbed onto a fixed bed of molecular sieve. The paraffins are desorbed
and used as petrochemical feedstock, solvents, etc., while the branched and ring
compounds are used for gasoline blending (Meyers, 1986).
• One facility uses a process similar to the gasoline component extraction process
described above, but with a slightly heavier feed.
• Heavy naphtha is fed to a fixed bed of silica gel. Aromatics are adsorbed while
paraffins pass through. When saturated, the bed is desorbed with benzene and the
product distilled to form various solvents. No other adsorbents are used in the process.
3.5.2 Extraction Treating Clay
3.5.2.1. Description
Wastes generated from the reformate extraction processes include the following:
• "Fuel side." Treating clay is used to remove impurities from the hydrocarbon
following extraction; the most common application is the filtering of the aromatic
fraction prior to benzene distillation (to keep impurities out of the downstream
fractions), although a small number of facilities use the clay to filter the benzene
Petroleum Refining Industry Study 66 August 1996
-------
product stream only. The purpose of the clay is to remove olefins, suspended solids,
and trace amount of solvent by a combination of adsorption and catalytic processes. A
few facilities also treat the raffinate (non-aromatic) stream with clay. Many facilities
did not report a clay treating step anywhere in their reformate extraction process. For
these facilities, clay treating is evidently not required to achieve the target product
limits.
• "Solvent side." Various treatment methods are used to remove impurities such as
polymers and salts from the lean solvent. A slip stream of lean solvent is processed
using ion-exchange, sock filters, carbon adsorption, or regeneration. This is similar to
the methods used to treat amine in sulfur-removal systems. An intermittent stream of
spent solvent can sometimes be generated.
Only the "fuel side" residuals are discussed and evaluated in Section 3.4.4. The "solvent side"
residuals are generally classified as miscellaneous sludges in the database and their volumes
were not tabulated in Table 3.5.1 (below).
As stated above, reformate extraction is the most common type of gasoline component
extraction process, but the small number of other processes also generate spent adsorbents.
These processes are unlike reformate extraction because the adsorbent is used for aromatic
separation (in reformate extraction, clay treatment occurs following aromatic extraction). In
these processes, spent adsorbent is also periodically generated, although generally less frequently
so than in the reformate extraction process. These materials were included in the statistics
presented in Table 3.5.1.
3.5.2.2 Generation and Management
When spent, adsorbents from extraction are vacuumed or gravity dumped from the
vessels. Interim storage can include 55-gallon drums, flow-bins, dumpsters, or piles. The
frequency of generation is highly dependent on the generating process: extraction clay is
typically generated once per year or less. According to questionnaire results, 2 facilities reported
classifying 81.3 MT of this residual as RCRA hazardous in 1992 (as D018).5 This is consistent
with reporting for other years.
Fifteen facilities reported generating a total quantity of approximately 1900 MT of this
residual in 1992, according to the 1992 RCRA §3007 Questionnaire. The questionnaire reported
that 30 facilities have extraction units. However, not all of these facilities use clay, molecular
sieve, or other adsorbents in their process; 33 percent of facilities with extraction units did not
report generating any clay residual for their process in any year, indicating either that clay is
either not used, has not yet bee replaced, or is generated so infrequently that respondents could
not recall when, if ever, the clay was last replaced. However, there was no reason to expect that
1992 would not be a typical year with regard to this residual's generation and management.
Extraction clays are generated more frequently and in greater quantity than isomerization clays.
5These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., Subtitle C landfill, transfer for
metals reclamation, etc.).
Petroleum Refining Industry Study 67 August 1996
-------
Residuals were assigned to be "spent clay from extraction" if they were assigned a
residual identification code of "spent sorbent" and were generated from a process identified as
an isomerization or extraction unit. These correspond to residual code 07 in Section VILA and
process codes 12 in Section IV.C of the survey. Table 3.5.1 provides a description of the 1992
management practices, quantity generated, number of streams reported, number of streams not
reporting volumes (data requested was unavailable and facilities were not required to generate
it), total and average volumes.
Table 3.5.1. Generation Statistics for Treating Clay from Extraction, 1992
Final Management
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite Subtitle C landfill
Disposal in onsite Subtitle D landfill
Onsite land treatment
Other recycling, or reuse: cement
plant
Transfer metal catalyst for
reclamation or regeneration
TOTAL
#of
Streams
10
4
1
2
3
1
1
22
# of Streams w/
Unreported Volume
0
0
0
0
0
0
0
0
Total Volume
(MT)
734.8
376.3
40
448.8
231
26
18
1875
Average
Volume (MT)
88.4
94
40
224.4
78
26
18
85.2
J. J.Z.J
Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.5.1. The Agency
gathered information suggesting that "offsite land treatment" (95 MT) was used in other years.
This practice is comparable to the practice reported for 1992 (i.e., onsite land treatment). In
addition, EPA compared the management practice reported for extraction treating clay to those
reported for treating clays from isomerization, alkylation, and lube oil6 based on expected
similarities. No additional management practices were reported.
6EPA did not compare these management practices to those reported for the broader category of "treating clay
from clay filtering" due to the diverse types of materials included in this miscellaneous category.
Petroleum Refining Industry Study
68
August 1996
-------
3.5.2.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.5.2 summarizes the physical properties of the spent adsorbents as reported in
Section VILA of the §3007 survey.
• One record sample of spent adsorbents from extraction was collected and analyzed by
EPA. The sampling information is summarized in Table 3.5.3.
The record sample was analyzed for total and TCLP levels of volatiles, semivolatiles,
and metals, and ignitability. It was not found to exhibit a hazardous waste characteristic. A
summary of the results is presented in Table 3.5.4. Only constituents detected in at least one
sample are shown in this table. This residual was categorized with isomerization clay in the
consent decree, and the characterization information for both residuals is presented in Table
3.5.4.
3.5.2.5 Source Reduction
Treating clay for extraction is generally used as a method of prolonging the life of the
catalyst or for product polishing. Because they are used as a source reduction technique for
other residuals, no source reduction methods for the clays were found.
Petroleum Refining Industry Study 69 August 1996
-------
Table 3.5.2. Treating Clay from Extraction: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Specific Gravity
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 1 0-1 00 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
12
14
13
10
6
5
9
20
19
25
10
11
9
8
9
1
# of Unreported
Values1
17
15
16
19
23
24
20
9
10
4
19
18
20
21
20
28
10th%
4.28
0
0
37.78
0
0
0.9
0
0
98
0
0
0
0
0
10
50th %
6.65
0.5
1
71.1
0.85
0.34
1
0
0
100
0
85
0
0
0
10
90th %
7.5
250
100
96.1
1
100
2
11
1
100
100
100
20
20
100
10
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgement.
Table 3.5.3. Extraction Spent Sorbent Record Sampling Locations
Sample number
R8D-CI-01
Facility
Amoco, Texas City, TX
Description
Clay from aromatic extraction unit (reformate
feed)
Petroleum Refining Industry Study
70
August 1996
-------
Table 3.5.4. Residual Characterization Data for
Spent Treating Clay from Extraction/Isomerization
Volatile Orqanics - Method 8260A ug/kg
Acetone
Benzene
Isopropylbenzene
Toluene
Naphthalene
CAS No.
67641
71432
98828
108883
91203
R8D-CI-01
< 600
2,500
J 650
36,000
< 600
R23B-CI-01
940,000
< 62,500
< 62,500
< 62,500
J 29,000
Average Cone
470,300
2,500
650
36,000
14,800
Maximum Cone
940,000
2,500
650
36,000
29,000
Comments
1
1
1
TCLP Volatile Orqanics - Methods 1311 and 8260A ug/L
Acetone
Benzene
4-Methyl-2-pentanone
Methyl ethyl ketone
Toluene
1,2,4-Trimethylbenzene
1,3,5-Trimethylbenzene
m,p-Xylene
Naphthalene
CAS No.
67641
71432
108101
78933
108883
95636
95476
108383/106423
91203
R8D-CI-01
120
< 50
< 50
< 50
< 50
< 50
< 50
< 50
< 50
R23B-CI-01
32,000
J 45
6,100
3,800
110
250
630
J 62
J 30
Semivolatile Orqanics - Method 8270B
Fluoranthene
Fluorene
Isophorone
2,4-Dimethylphenol
3/4-Methylphenol
Naphthalene
1 -Methylnaphthalene
2-Methylnaphthalene
CAS No.
206440
86737
78591
105679
NA
91203
90120
91576
R8D-CI-01
J 130
< 165
< 165
< 165
< 165
J 280
J 220
520
R23B-CI-01
< 165
J 220
130,000
J 2,800
J 150
< 165
J 650
J 310
Average Cone
16,060
45
3,075
1,925
80
150
340
56
30
Maximum Cone
32,000
45
6,100
3,800
110
250
630
62
30
Comments
1
1
w/kg
Average Cone
130
193
65,083
1,483
150
223
435
415
Maximum Cone
130
220
130,000
2,800
150
280
650
520
Comments
1
1
TCLP Semivolatile Orqanics - Methods 1311 and 8270B ug/L
Isophorone
2-Methylphenol
3/4-Methylphenol (total)
CAS No.
78591
95487
NA
R8D-CI-01
< 50
< 50
< 50
R23B-CI-01
7,300
J 34
J 99
Total Metals - Methods 6010, 7060, 7421, 7470, 7471
Aluminum
Barium
Calcium
Chromium
Iron
Lead
CAS No.
7429905
7440393
7440702
7440473
7439896
7439921
R8D-CI-01
8,300
250
4,700
< 1.00
1,800
13.0
R23B-CI-01
110,000
< 20.0
4,500
14.0
3,000
1.60
Average Cone
3,675
34
75
Maximum Cone
7,300
34
99
Comments
1
and 7841 mg/kg
Average Cone
59,150
135
4,600
7.50
2,400
7.30
Maximum Cone
110,000
250
4,700
14.0
3,000
13.0
Comments
Petroleum Refining Industry Study
71
August 1996
-------
Table 3.5.4. Residual Characterization Data for
Spent Treating Clay from Extraction/Isomerization (continued)
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mq/kq (continued)
Magnesium
Manganese
Potassium
Sodium
Vanadium
Zinc
CAS No.
7439954
7439965
7440097
7440235
7440622
7440666
R8D-CI-01
4,300
350
< 500
< 500
20.0
8.80
R23B-CI-01
9,600
43.0
1,300
81,000
10.0
28.0
Average Cone
6,950
197
900
40,750
15.0
18.4
Maximum Cone
9,600
350
1,300
81,000
20.0
28.0
Comments
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Aluminum
Calcium
Iron
Lead
Magnesium
Manganese
CAS No.
7429905
7440702
7439896
7439921
7439954
7439965
R8D-CI-01
< 1.00
160
18.0
0.04
82.0
12.0
R23B-CI-01
17.0
< 25.0
1.30
< 0.015
50.0
< 0.08
Average Cone
9.00
92.5
9.65
0.03
66.0
6.04
Maximum Cone
17.0
160.0
18.0
0.04
82.0
12.0
Comments
Comments:
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
Notes:
B Analyte also detected in the associated method blank.
C Compound's concentration is estimated. Mass spectral data indicate the presence of a compound that meets the identification criteria for
which the result is less than the laboratory detection limit, but greater than zero.
ND Not Detected.
NA Not Applicable.
Petroleum Refining Industry Study
72
August 1996
-------
3.6 ALKYLATION
The petroleum refining industry uses both hydrofluoric and sulfuric acid catalyzed
alkylation processes to form high octane products. DOE reported that 103 facilities operated
almost 1.1 million BPSD of alkylation capacity; 49 facilities used sulfuric acid and 59 used HF.
While the general chemistry of these processes is the same, the HF process includes a closed
loop and integral recycling step for the HF acid, while the sulfuric acid process requires a
separate acid regeneration process, which generally occurs off site. Study residuals are
generated from both alkylation processes.
3.6.1 Sulfuric Acid Alkylation Process Description
In the sulfuric acid alkylation process, olefin and isobutane gases are contacted over
concentrated sulfuric acid (H2SO4) catalyst to synthesize alkylates for octane-boosting. The
reaction products are separated by distillation and scrubbed with caustic. Alkylate product has a
Research Octane Number in the range of 92 to 99. Figure 3.6.1 provides a generic process flow
diagram for H2SO4 alkylation.
Figure 1.3.1. H2SO4 Alkylation Process Flow Diagram
Cracked Gas
Isobutane
Era pane
Isobutane
Make-up
catalyst
n-Butane-*—
Altylate-
Lime
Water to
WWTP "
ait Acid
Neutralization
Pit
OEQ-Q6
Neutralisation
Sludge
The olefin stream is mixed with isobutane and H2SO4 in the reactor. To prevent
polymerization and to obtain a higher quality yield, temperatures for the H2SO4 catalyzed
reaction are kept between 40 and 50°F (McKetta, 1992). Since the reactions are carried out
below atmospheric temperatures during most of the year, refrigeration is required. Pressures are
maintained so all reaction streams are in their liquid form. The streams are mixed well during
their long residence time in the reactor to allow optimum reaction to occur.
The hydrocarbon/acid mixture then moves to the acid separator, where it is allowed to
settle and separate. The hydrocarbons are drawn off the top and sent to a caustic wash to
Petroleum Refining Industry Study
73
August 1996
-------
neutralize any remaining trace acid. The acid is drawn from the bottom and recycled back to the
reactor. A portion of the acid catalyst is continuously bled and replaced with fresh acid to
maintain the reactor's acid concentration at around 90 percent. This spent H2SO4 was a listing
residual of concern.
In the fractionator, the hydrocarbon streams are separated into the alkylate and saturated
gases. The isobutane is recycled back into the reactor as feed. Light end products may be
filtered with sorbents to remove trace H2SO4 acid, caustic or water. The sorbents (e.g., treating
clays) are study residuals of concern.
Some facilities have neutralization tanks (in and above ground), referred to as pits, which
neutralize spent caustic and any acid generated from spills prior to discharge to the WWTP,
serving as surge tanks. Neutralizing agents (sodium, calcium, potassium hydroxides) are
selected by the refineries. If necessary, the influent to the pit is neutralized and, depending on
the neutralizing agent, the precipitated salts form a sludge. This sludge was also a listing
residual of concern. Sludge may also be generated in process line junction boxes, in the spent
H2SO4 holding tank, and during turnaround. However, due to the aqueous solubility of sodium,
calcium, and potassium sulfates, sludge generation rates are relatively low and the majority of
neutralization salts (e.g., sodium sulfate) are solubilized and discharged to the WWTP.
3.6.2 Hydrofluoric Acid Alkylation Process Description
Hydrofluoric acid alkylation is very similar to the H2SO4 alkylation process. In the
hydrofluoric acid alkylation process, olefin and isobutane gases are contacted over hydrofluoric
acid (HF) catalyst to synthesize alkylates for octane-boosting. The reaction products are
separated by distillation and scrubbed with caustic. Alkylate product has a research octane
number (RON) in the range of 92 to 99. Because it is clean burning and contributes to reduced
emissions, alkylate is a highly valued component in premium and reformulated gasolines. The
HF process differs from the H2SO4 alkylation in that the HF catalyst is managed in a closed-loop
process, never leaving the unit for replacement or regeneration. Figure 3.6.2 provides a generic
process flow diagram for HF alkylation.
The olefin stream is mixed with the isobutane and HF in the reactor. To prevent
polymerization and to receive a higher quality yield, temperatures for the HF catalyzed reaction
are maintained at approximately 100°F. Pressures are kept so all reaction streams are in their
liquid form (usually 85 to 120 psi). The streams are mixed well in the reactor to allow optimum
reaction to occur.
The hydrocarbon/acid mixture then moves to the settler, where it is allowed to settle and
phase separate. The hydrocarbons are drawn off the top and sent to a fractionator. The acid is
drawn from the bottom and recycled back to the reactor. A slip stream of acid is sent to an acid
regenerator where distillation separates the HF acid from by-product contaminants. The HF acid
from the regenerator is recycled back to the reactor. Fresh acid is added to replace acid losses at
a rate of about 500 pounds per day per 5,000 BPSD alkylation unit capacity (a small to medium
size unit).
Petroleum Refining Industry Study 74 August 1996
-------
Figure 3.6.2. HF Alkylation Process Flow Diagram
HFAcid
Reactor
/ \
Settler
Acid
r
-
-
^
Acii
Acid
PeteiEKitor
ASO
^
Da; ant
\fessel
Acid Retuin
Catalyst
Sip Stream
ASO I
Lima
Spills and Equipment Drains
CEM
Spent
Caustic
Neutralisation
Add
•Soluble
Oil
Witerto
•WWIF
Neutralisation
Sludge
A residual of high molecular-weight reaction by-products dissolves in the HF acid
catalyst and lowers its effectiveness. To maintain the catalyst activity, a slip stream of catalyst is
distilled, leaving the by-product, acid soluble oil (ASO), as a residue. The ASO is charged to a
decanting vessel where an aqueous phase settles out. The aqueous phase, an azeotropic mixture
of HF acid and water, is referred to as constant boiling mixture (CBM). The ASO is scrubbed
with potassium hydroxide (KOH) to remove trace amounts of HF and either recycled, sold as
product (e.g., residual fuel), or burned in the unit's boiler. The CBM is sent to the neutralization
tank. In some cases, the ASO from the regenerator is sent directly to the neutralization tank.
The ASO is a residual of concern for the petroleum refining study.
A series of fractionators distills the product streams from the reactor into the alkylate,
saturated gases, and HF acid. Isobutane and HF are recycled back into the reactor as feed.
The main fractionator overhead is charged to the depropanizer and debutanizer, where
high-purity propane and butane are produced. The propane and butane are then passed through
the alumina treater for HF removal. Once catalytically defluorinated, they are KOH-treated and
sent to LPG storage.
As HF is neutralized by aqueous KOH, soluble potassium fluoride (KF) is produced and
the caustic is eventually depleted. Some facilities employ KOH regeneration. Periodically some
of the KF-containing neutralizing solution is withdrawn to the KOH regenerator. In this vessel
Petroleum Refining Industry Study
75
August 1996
-------
KF reacts with a lime slurry to produce insoluble calcium fluoride (CaF2) and thereby
regenerates KF to KOH. The regenerated KOH is then returned to the system, and the solid
CaF2 is routed to the neutralizing tank. The KF, at facilities that do not have a regenerator, is
sent directly to the neutralizing tank, where it is reacted with lime to form a sludge.
Spent caustic, KOH scrubbers, acidic waters from acid sewers and, in some cases, CBM
are charged to in-ground neutralization tanks (referred to by industry as pits), which neutralize
effluent to the WWTP. Neutralizing controls fluoride levels to the WWTP. Neutralizing agents
(sodium, calcium, and potassium hydroxide) are selected based on the refineries' WWTP
permits. Effluent to the pit is neutralized, generally with lime, which forms a sludge (calcium
fluoride) that collects on the bottom of the tank. This sludge was a listing residual of concern.
FTP acid is an extremely corrosive and toxic chemical. Refineries go to great lengths to
protect their personnel from HF contact. Prior to entrance to an FTP alkylation unit, personnel
must have special training and wear various levels of personal protective clothing (depending
upon the work to be performed). The unit is generally cordoned off and marked as an FTP hazard
area. Valves, flanges, and any place where leaks can occur are painted with a special paint that
will change colors when contacted with FTP. The units are continuously monitored and alarms
are activated if an FTP leak is detected.
3.6.3 Spent Treating Clay from Alkylation
3.6.3.1 Description
Treating clay from alkylation predominantly includes (1) molecular sieves used for
drying feed and (2) alumina used for removing fluorinated compounds from the product. Both
are applications in FTP alkylation; clays are little used in sulfuric acid alkylation. Specifically,
the industry reported 83 treating clay residuals from alkylation in 1992, accounting for 2,890
metric tons of residuals. Only 7 of these residuals (143 metric tons) were from sulfuric acid
alkylation processes.
After fractionation, products may be passed through a filter filled with sorbents (referred
to as treating clay) to remove trace amounts of acid, caustic, or water. Sorbents typically used in
this service include alumina, molecular sieve, sand, and salt.
Treating clay becomes spent when breakthrough of H2SO4 or FTP acid, caustic, or water
occurs. Depending on the type of clay and the type of service, breakthrough can occur anywhere
between 2 months and 5 years (e.g., alumina in FTP service is typically 2 months and salt treaters
can be as long as 5 years). Prior to removal the clay may undergo one of the following in situ
treatments:
• Nitrogen sweep
• Propane sweep
• Steam stripping
• Methane sweep
Petroleum Refining Industry Study 76 August 1996
-------
Following removal, the spent clay is placed in closed containers and is typically sent to
an offsite landfill. Certain types of treating clay, such as alumina, are more amenable to
recycling and may be sent offsite to a smelter or a cement kiln to be used as process feeds.
In 1992, less than 2 percent of the volume of spent treating clay from alkylation was
managed as hazardous, with one residual reported to be D004, and three others reported
generically to be managed as hazardous (i.e., no specific codes were reported).7
3.6.3.2 Generation and Management
The RCRA §3007 Survey responses indicated 2,895 MT of spent treating clay were
generated in 1992. Residuals were assigned to be "treating clay from alkylation" if they were
assigned a residual identification code of "spent sorbent" and was generated from a process
identified as a sulfuric acid or HF alkylation unit. This corresponds to residual code "07" in
Section VII. 1 and process codes "09-A" or "09-B" in Section IV-l.C of the questionnaire. Due
to the frequent generation of this residual, not all 103 facilities generated spent treating clay in
1992. However, there was no reason to expect that 1992 would not be a typical year with regard
to this residual's generation and management. Table 3.6.1 provides a description of the total
quantity generated, number of streams not reporting volumes (data requested was unavailable
and facilities were not required to generate it), total and average volumes.
Table 3.6.1. Generation Statistics for Treating Clay from Alkylation, 1992
Final Management
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite Subtitle C landfill
Disposal in onsite Subtitle D landfill
Disposal in onsite wastewater
treatment facility
Onsite land treatment
Other recycling, or reuse: cement plant
Other recycling, or reuse: onsite road
material
Storage in pile
Transfer to offsite entity: alumina
manufacturer, smelter, or other
unspecified recycle
TOTAL
#of
Streams
28
4
3
18
0
4
4
1
6
15
83
# of Streams w/
Unreported Volume
2
1
0
0
2
0
0
0
0
0
5
Total Volume
(MT)
633.7
23.9
67.0
626.3
-
59.2
770.5
3.6
30.0
680.4
2,894.6
Average
Volume (MT)
22.6
6
22.3
34.8
-
14.8
154.1
3.6
5.0
45.4
34.9
'These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., managed in WWTP, Subtitle C
landfill, transfer to offsite entity, etc.).
Petroleum Refining Industry Study
77
August 1996
-------
3.6.3.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.6.1. The Agency
gathered information suggesting other management practices have been used in other years
including: "disposal onsite in surface impoundment" (38.4 MT), "other recycling, reclamation,
or reuse: offsite fluoride recovery" (23.6 MT), and "offsite incineration" (3.6 MT). The very
small volume reported to have been disposed in a surface impoundment was placed in the
surface impoundment the year it was closed, suggesting the inert material was used as fill. The
refinery reported the future management of the spent clay would be sent offsite to a cement kiln
for reuse. Similarly, the very small volume reported for offsite fluoride recovery was a
management practice seen as a trend for fluoride containing residuals during the engineering site
visits. The very small volume reported for offsite incineration are comparable to the 1992
practices for other treating clay residuals (e.g., clay filtering)
3.6.3.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.6.2 summarizes the physical properties of the alkylation sorbents as reported
in Section VILA of the §3007 survey.
• Four record samples of actual treating clay were collected and analyzed by EPA.
These spent clays are all from HF processes and represent the various types of spent
sorbents typically used by the industry as summarized in Table 3.6.3.
The four record samples were analyzed for total and TCLP levels of volatiles,
semivolatiles, metals, fluorides, reactivity and ignitability. None of the samples were found to
exhibit any of the hazardous waste characteristics. A summary of the results is presented in
Table 3.6.4. Only constituents detected in at least one sample are shown in this table.
3.6.3.5 Source Reduction
Several solid-acid catalysts used for alkylation are being tested in pilot plants. The solid-
catalyst reactor systems are different from the current liquid-acid systems, but for one solid-
catalyst operation, the other process equipment is compatible. The three types of new solid
catalyst include aluminum chloride, alumina/zirconium halide, and antimony pentafluoride (a
slurry system). It is unclear whether these processes will generate more or less treating clays
than current processes. Theoretically, these processes would not require filtering for acid and
water removal.
The February 1, 1993 issue of the Oil & Gas Journal reported that Conoco's Ponca City,
Oklahoma refinery sold reclaimed fluorinated alumina to Kaiser Aluminum & Chemical
Corporation's plant in Mead, Washington. The fluorinated alumina is substituted for aluminum
fluoride, a "bath" chemical used in aluminum manufacturing.
Petroleum Refining Industry Study 78 August 1996
-------
Table 3.6.2. Treating Clay from Alkylation: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Specific Gravity
Specific Gravity Temperature, °C
BTU Content, BTU/lb
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Other, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 1 0-1 00 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
60
39
45
43
43
25
54
27
12
89
85
123
77
41
53
37
38
36
21
#of
Unreported
Values1
91
112
106
108
108
126
97
124
139
62
66
28
74
110
98
114
113
115
130
10th%
2.91
0.00
0.00
60.00
0.00
0.00
0.70
15.00
0.00
0.00
0.00
96.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
50th %
7.00
0.25
4.00
93.33
0.05
0.00
1.24
15.60
0.00
0.00
0.00
100.00
0.00
0.00
100.00
0.00
0.00
0.00
1,200.00
90th %
9.00
250.00
170.00
100.00
1.00
1.00
2.24
25.00
500.00
8.00
1.00
100.00
0.00
100.00
100.00
50.00
3.00
0.00
9,525.00
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgement.
Table 3.6.3. Alkylation Treating Clay Record Sampling Locations
Sample Number
R3-CA-01
R15-CA-01
R21-CA-01
R23-CA-01
Location
Exxon, Billings, MT
Total, Ardmore, OK
Chevron, Pt. Arthur, TX
Chevron, Salt Lake City, UT
Description
Alumina propane product treater1
Alumina butane product treater1
Alumina propane or butane product treater1
Alumina propane product treater1
1HF process
Petroleum Refining Industry Study
79
August 1996
-------
Petroleum Refining Industry Study 80
Volatile Organics - Method 8260A tig/kg
Acetone
Benzene
sec-Butylbenzene
p-lsopropyltoluene
Methyl ethyl ketone
Toluene
1,2,4-Trimethylbenzene
o-Xylene
m,p-Xylenes
Naphthalene
CAS No.
67641
71432
135988
99876
78933
108883
95636
95476
108383/106423
91203
R3-CA-01
42,000
J 67
< 625
< 625
< 625
J 67
J 112
< 625
J 136
< 625
R15-CA-01
680
< 25
< 25
< 25
290
< 25
< 25
< 25
< 25
< 25
R21-CA-01
J 880
< 625
J 1,200
J 800
< 625
< 625
2,100
J 530
1,300
J 1,100
TCLP Volatile Organics - Methods 1311 and 8260A |
Acetone
Toluene
Methyl ethyl ketone
m,p-Xylene
CAS No.
67641
108883
78933
108383/106423
R3-CA-01
1,500
JB 11
J 95
JB 10
R15-CA-01
< 50
< 50
< 50
< 50
R21-CA-01
280
< 50
210
< 50
R23-CA-01
13,000
< 650
< 650
< 650
1,300
< 650
< 650
< 650
< 650
< 650
Average Cone
14,140
46
625
525
710
46
722
278
528
600
Maximum Cone
42,000
67
1,200
800
1,300
67
2,100
530
1,300
1,100
Comments
1
1
1
ig/L
R23-CA-01
B 1,100
< 50
250
< 50
Average Cone
733
11
151
10
Maximum Cone
1,500
11
250
10
Comments
1
1
Semivolatile Organics - Method 8270B tig/kg
Di-n-butyl phthalate
Phenanthrene
CAS No.
84742
85018
R3-CA-01
< 165
< 165
R15-CA-01
< 165
J 160
R21-CA-01
J 200
< 165
R23-CA-01
< 165
< 165
Average Cone
174
160
Maximum Cone
200
160
Comments
1
TCLP Semivolatile Organics - Methods 1311 and 8270B ug/L
Bis(2-ethylhexyl)phthalate
CAS No.
117817
R3-CA-01
J 10
R15-CA-01
< 50
R21-CA-01
< 50
R23-CA-01
< 50
Average Cone
10
Maximum Cone
10
Comments
1
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg
Aluminum
Arsenic
Beryllium
Iron
Manganese
Sodium
Zinc
CAS No.
7429905
7440382
7440417
7439896
7439965
7440235
7440666
R3-CA-01
240,000
26.0
2.20
23.0
< 1.5
2,200
23.0
R15-CA-01
170,000
13.0
1.70
< 5.0
4.70
2,000
33.0
R21-CA-01
210,000
< 5.0
2.00
< 5.0
6.50
8,000
40.0
R23-CA-01
240,000
< 5.0
2.20
52.0
6.90
7,700
39.0
Average Cone
215,000
12.3
2.03
21.3
4.90
4,975
33.8
Maximum Cone
240,000
26.0
2.20
52.0
6.90
8,000
40.0
Comments
-------
Petroleum Refining Industry Study
i auie o.o.t. /AiKyiauuii i reauiig ^ ui y ^ iiaracieri/;aiiuii ^uiiimueuj
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Aluminum
Beryllium
Iron
Manganese
Zinc
CAS No.
7429905
7440417
7439896
7439965
7440666
R3-CA-01
5,300
0.05
1.60
< 0.08
B 1.10
R15-CA-01
1,300
< 0.025
< 0.50
< 0.08
B 0.60
R21-CA-01
4,100
< 0.025
1.10
0.18
B 0.82
R23-CA-01
4,100
< 0.025
1.00
0.17
B 0.85
Average Cone
3,700
0.031
1.05
0.13
0.84
Maximum Cone
5,300
0.050
1.60
0.18
1.10
Comments
Miscellaneous Characterization
Total Fluorine (mq/kq)
R3-CA-01
39,000
R15-CA-01
4,500
R21-CA-01
NA
R23-CA-01
NA
Average Cone
21,750
Maximum Cone
39,000
Comments
Comments:
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound that meets the identification criteria for which the result is less than the laboratory detection limit, but greater than
zero.
ND Not Detected.
NA Not Applicable.
-------
At the Ponca City refinery, Conoco uses activated alumina in one of the alkylation units
to extract fluorides from propane and butane products. In the process, activated alumina is
converted to aluminum fluoride. Activated alumina reaches the end of its useful life when 60-
80% of the material is converted to aluminum fluoride. That is when it become an additive for
aluminum manufacturers.
During EPA's site visits, one facility used distillation to dry its feed to the HF acid
alkylation unit. Most facilities use a molecular sieve treating clay for this step, therefore this
process configuration eliminates the need for molecular sieve infrequently generating an RC.
Some refineries are experimenting with additives to the HF acid catalyst. The purpose of
these additives is to reduce the risk from an accidental leak of FTP acid to the atmosphere.
Although the technology is principally developed in reaction to safety concerns, it is likely that
such additives would be present in some of the study residuals such as acid soluble oil. The
identity of those additives were not reported (Oil and Gas Journal, August 22, 1994).
3.6.4 Catalyst from Hydrofluoric Acid Alkylation
3.6.4.1 Description
The consent decree which identifies the residuals to be examined in this study specified
"catalyst from FTP alkylation". However, the analysis used to identify the residuals of concern in
the consent decree contained some flaws and erroneously identified this alkylation catalyst as
being generated in significant quantities. Upon further review of the data used to characterize
this residual (derived from EPA's 1983 survey of the petroleum refining industry), it was
determined that several large volume residuals were inappropriately identified as spent catalyst
and instead should have been classified as acid soluble oil (ASO). After adjusting the data to
remove these mischaracterized residuals, the remaining residuals classified as spent FTP catalyst
accounted for small volumes which are on the order of magnitude observed in the Agency's 1992
data.
A residual of high molecular-weight reaction by-products dissolves in the FTP acid
catalyst and lowers its effectiveness. To maintain catalyst activity, a slip stream of FTP acid is
sent to an acid regenerator where distillations separates the FTP acid from by-product
contaminants, called acid soluble oil. The FTP acid from the regenerators is recycled back to the
reactor. Fresh acid is added to replace acid losses at a rate of about 500 pounds per day
depending on unit capacity.
ASO is charged to a decanting vessel where an aqueous phase settles out. The aqueous
phase, an azeotropic mixture of FTP acid and water, is referred to as constant boiling mixture
(CBM). CBM is charged to the neutralization tank which neutralize effluent to the WWTP. The
neutralization sludge was examined in the listing proposal and Background Document. The
effluent from the neutralization tanks are reported to go to the WWTP. The Agency has no data
suggesting that it can be handled in any other way.
As stated above, FTP acid is an extremely corrosive and toxic chemical. Refineries go to
great lengths to protect their personnel from coming into direct contact with FTP acid.
Petroleum Refining Industry Study 82 August 1996
-------
3.6.4.2 Generation and Management
The refineries reported generating approximately 152 MT of HF alkylation catalyst in
1992. Residuals were assigned to be "HF alkylation catalyst" if they were assigned a residual
identification code of "liquid catalyst" and was generated from a process identified as an FTP acid
alkylation unit. This corresponds to residual code 03-B in Section VIF2 of the questionnaire and
process code 09-B in Section IV-l.C of the questionnaire. Table 3.6.5 provides a description of
the quantity generated, number of streams reported, and number of unreported volumes.
Catalyst from FTP alkylation includes spills and removed acid from the FTP alkylation process.
Table 3.6.5. Generation Statistics for Catalyst from HF Alkylation, 1992
Final Management
Discharge to WWTP
#of
Streams
3
# of Streams w/
Unreported Volume
0
Total Volume
(MT)
151.94
Average
Volume (MT)
50.65
3.6.4.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.6.5. No data were
available to the Agency suggesting any other management practices.
3.6.4.4 Characterization
Only one source of residual characterization is available from the industry study,
reflecting the fact that this residual is not generated for management:
• Table 3.6.6 summarizes the physical properties of the FTP catalyst as reported in
Section VIFA of the §3007 survey.
Due to the rareness of the generation of this residual, no samples of this residual were
available for collection and analysis during record sampling.
Table 3.6.6. Catalyst from HF Alkylation: Physical Properties
Properties
PH
Vapor Pressure, mm Hg
Specific Gravity
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Other, %
#of
Values
2
1
1
2
2
2
2
# of Unreported
Values
1
2
2
1
1
1
1
10th%
2.00
775.00
1.00
0.00
0.00
0.00
100.00
50th %
2.00
775.00
1.00
0.00
0.00
0.00
100.00
90th %
2.00
775.00
1.00
0.00
0.00
0.00
100.00
Petroleum Refining Industry Study
83
August 1996
-------
3.6.4.5 Source Reduction
As described in the spent treating clay alkylation in Section 3.6.3.5, several solid-acid
catalysts used for alkylation are being tested in pilot plants. The reactor systems are different
from the current liquid-acid systems, but for one system the other equipment is compatible.
Three types of the new solid catalyst include aluminum chloride, alumina/zirconium halide, and
antimony pentafluoride (a slurry system).
In general, additional source reduction is not possible because of the closed loop recycle
process and the strict controls placed on this material due to the severe health hazards associated
with contact and inhalation.
3.6.5 Acid Soluble Oil from Hydrofluoric Acid Alkylation
3.6.5.1 Description
A residual of high molecular-weight reaction by-products dissolves in the HF acid
catalyst and lowers its effectiveness. To maintain the catalyst activity, a slip stream of catalyst is
distilled, leaving the by-product, acid soluble oil (ASO), as a residue. The ASO is charged to a
decanting vessel where an aqueous phase settles out. The ASO is scrubbed with potassium
hydroxide (KOH) to remove trace amounts of HF and is either recycled, sold as product (e.g.,
residual fuel), or burned in the unit's boiler. In some cases, the ASO from the regenerator is sent
directly to the neutralization tanks. Effluent from the neutralization tanks is sent to the WWTP.
Neutralization tank sludges were examined under the listing proposal and Background
Document.
ASO is generated exclusively from the HF process. The sulfuric acid alkylation process
does not generate ASO.
Eight residuals of ASO, accounting for 25 percent of this category's volume, was
reported as being managed as either D001, D002, or D008.8
3.6.5.2 Generation and Management
The refineries reported generating approximately 33,493 MT of ASO in 1992. Residuals
were assigned to be "ASO" if they were assigned a residual identification code of "alkylation
acid regeneration tars" and were generated from a process identified as an HF acid alkylation
unit. This corresponds to residual code 08 in Section VII. 1 and process code 09-B in Section
IV-l.C of the questionnaire. Note that sludges generated from neutralization of acid soluble oil
were examined under the proposal and the Background Document and are not included here.
Table 3.6.7 provides a description of the quantity generated, number of streams reported, and
number of unreported volumes.
8These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., managed in WWTP, transfer as a
fuel, offsite incineration, etc.).
Petroleum Refining Industry Study 84 August 1996
-------
Table 3.6.7. Generation Statistics for Acid Soluble Oil, 1992
Final Management
Discharge to onsite wastewater treatment facility
Neutralization
Offsite incineration
Onsite boiler
Onsite industrial furnace
Other recovery onsite: alkylation or hydrotreating/
hydrorefining process or unknown
Recovery onsite in a catalytic coker
Recovery onsite in a coker
Recovery onsite via distillation
Transfer for direct use as a fuel or to make a fuel
Transfer with coke product or other refinery product
TOTAL
#of
Streams
6
15
2
3
10
3
5
1
2
2
4
53
#of
Streams w/
Unreported
Volume
0
14
0
0
1
1
0
0
3
0
1
20
Total
Volume
(MT)
4,858.8
11,387.9
0.2
2,610.3
3,274
2,180
3,641.3
1,019
50
740.6
3,731
33,493
Average
Volume
(MT)
809.8
759.2
0.1
870.1
327.4
726.7
728.3
1,019
25
370.3
932.8
631.9
3.6.5.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized in Table 3.6.7. The Agency gathered
information suggesting that "disposal in industrial Subtitle D landfill" (1 MT) was used in other
years. Upon closer examination of this residual, EPA determined that the facility neutralized its
ASO and landfilled the sludge. This management practice is consistent with the practices
reported above.
3.6.5.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.6.8 summarizes the physical properties of the ASO as reported in Section
VILA of the §3007 survey.
• Four record samples of actual ASO were collected and analyzed by EPA. The ASO
represent the various types of interim management practices typically used by the
industry (i.e., with and without neutralization) and are summarized in Table 3.6.9.
The four record samples were analyzed for total and TCLP levels of volatiles,
semivolatiles, and metals, as well as ignitability. Three of the samples were found to exhibit the
hazardous waste characteristic of ignitability. A summary of the results is presented in Table
3.6.10. Only constituents detected in at least one sample are shown in this table.
Petroleum Refining Industry Study
85
August 1996
-------
Table 3.6.8. Acid Soluble Oil: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Vapor Pressure, mm Hg
Vapor Pressure Temperature, °C
Viscosity, Ib/ft-sec
Viscosity Temperature, °C
Specific Gravity
Specific Gravity Temperature, °C
BTU Content, BTU/lb
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Other, %
#of
Values
30
12
14
27
26
16
10
9
11
6
34
12
15
47
56
32
27
#of
Unreported
Values1
59
77
75
62
63
73
79
80
78
83
55
77
74
42
33
57
62
10th%
2.00
0.00
0.00
25.00
15.00
30.00
3.00
20.00
0.00
15.00
0.80
15.00
750.00
0.00
50.00
0.00
0.00
50th %
6.50
0.13
5.00
60.00
90.00
77.00
135.00
25.00
0.01
17.50
0.90
15.00
15,000.00
10.00
98.00
0.00
0.00
90th %
10.75
50.00
200.00
93.33
100.00
100.00
575.00
38.00
0.40
37.80
1.00
15.60
19,000.00
75.00
100.00
30.00
0.05
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
Table 3.6.9. Acid Soluble Oil Record Sampling Locations
Sample Number
R3-AS-01
R5B-AS-01
R15-AS-01
R7C-AS-01
Location
Exxon, Billings, MT
Marathon, Garyville, LA
Total, Ardmore, OK
BP, Belle Chasse, LA
Description
Un-neutralized separator drum sample
Acid regenerator settler bottoms, not neutralized
Neutralized, skimmed from pit
Neutralized from storage tank
3.6.5.5 Source Reduction
As described in previous sections, several solid-acid catalysts used for alkylation are
being tested in pilot plants. The reactor systems are different from the current liquid-acid
systems, but for one system the other equipment is compatible. Three types of the new solid
catalyst include aluminum chloride, alumina/zirconium halide, and antimony pentafluoride (a
slurry system).
It is likely that ASO will not be generated in a solid catalyst system.
Petroleum Refining Industry Study
August 1996
-------
Table 3.6.10. Acid Soluble Oil Characterization
Volatile Organics - Method 8260A
Acetone
Acrolein
Benzene
n-Butylbenzene
sec-Butylbenzene
tert-Butylbenzene
Carbon disulfide
trans-1 ,3-Dichloropropene
Ethylbenzene
Isopropylbenzene
p-lsopropyltoluene
Methyl ethyl ketone
4-Methyl-2-pentanone
n-Propylbenzene
Toluene
1,2,4-Trimethylbenzene
1,3,5-Trimethylbenzene
o-Xylene
m,p-Xylenes
Naphthalene
CAS No.
67641
107028
71432
104518
135988
98066
75150
10061026
100414
98828
99876
78933
108101
103651
108883
95636
108678
95476
108383/106423
91203
TCLP Volatile Org
Acetone
Isopropylbenzene
Methyl ethyl ketone
CAS No.
67641
98828
78933
R3-AS-01
49,000
< 6250
< 6250
< 6250
< 6250
< 6250
< 6250
< 6250
< 6250
< 6250
< 6250
< 6250
< 6250
< 6250
< 6250
18,000
< 6250
< 6250
16,000
< 6250
\jql\_
R5B-AS-01
< 625
< 625
< 625
< 625
< 625
< 625
< 625
< 625
< 625
< 625
< 625
< 625
< 625
< 625
< 625
7,400
< 625
< 625
2,100
< 625
R7C-AS-01
B 40,000
25,000
30,000
J 9,500
J 2,600
J 7,200
J 1,800
J 1,600
37,000
J 3,100
J 6,600
27,000
26,000
J 8,200
41,000
110,000
27,000
20,000
55,000
30,000
(ug/kg)
R15-AS-01
3,000
< 1,250
< 1,250
< 1,250
J 488
J 1,350
< 1,250
< 1,250
< 1,250
< 1,250
< 1,250
< 1,250
< 1,250
< 1,250
< 1,250
3,300
J 1,260
< 1,250
< 1,250
< 1,250
Average Cone
23,156
8,281
9,531
4,406
1,238
3,856
1,225
1,158
11,281
1,658
3,681
8,781
8,531
4,081
12,281
34,675
8,784
7,031
18,588
9,531
Maximum Cone
49,000
25,000
30,000
9,500
2,600
7,200
1,800
1,600
37,000
3,100
6,600
27,000
26,000
8,200
41,000
110,000
27,000
20,000
55,000
30,000
Comments
1
1
1
1
anics - Methods 1311 and 8260A ug/L
R3-AS-01
NA
NA
NA
R5B-AS-01
NA
NA
NA
R7C-AS-01
NA
NA
NA
Semivolatile Organics - Method 8270B ug/L
Methyl ethyl ketone
1-Methylnaphthalene
2-Methylnaphthalene
Naphthalene
2-Methylnaphthalene
Naphthalene
CAS No
78933
90120
91576
91203
91576
91203
R3-AS-01
NA
< 250,000
< 250,000
< 250,000
< 250,000
< 250,000
R5B-AS-01
NA
< 46,000
< 46,000
< 46,000
< 46,000
< 46,000
R7C-AS-01
NA
100,000
180,000
79,000
180,000
79,000
R15-AS-01
B 350
J 32
J 80
(ug/kg)
R15-AS-01
J 80
< 12,375
< 12,375
< 12,375
< 12,375
< 12,375
Average Cone
350
32
80
Maximum Cone
350
32
80
Comments
Average Cone
80
73,000
113,000
62,500
113,000
62,500
Maximum Cone
80
100,000
180,000
79,000
180,000
79,000
Comments
1
1
1
1
1
g
OQ
-------
Petroleum Refining Industry Study 88
i auie j.o.iu. /vciu csuiuuie \jn ^ iiaracieri/;aiiuii ^uiiimueuj
TCLP Semivolatile Organics - Methods 1311 and 8270B ug/L
Aniline
CAS No.
62553
R3-AS-01
NA
R5B-AS-01
NA
R7C-AS-01
NA
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg
Aluminum
Calcium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Potassium
Sodium
Zinc
CAS No.
7429905
7440702
7440508
7439896
7439921
7439965
7439976
7440020
7440097
7440235
7440666
R3-AS-01
< 0.10
< 2.50
1.00
< 0.50
0.64
< 0.015
< 0.01
< 0.04
< 2.50
< 2.50
0.27
R5B-AS-01
< 0.10
< 2.50
< 0.13
< 0.50
< 0.015
< 0.015
0.022
< 0.04
< 2.50
< 2.50
< 0.10
R7C-AS-01
< 0.10
< 2.50
< 0.13
< 0.50
< 0.015
< 0.015
< 0.01
< 0.04
< 2.50
< 2.50
< 0.10
R15-AS-01
J 20
(mg/kg)
R15-AS-01
290
29,000
37.0
120
< 0.30
5.00
< 0.05
15.0
5,900
1,300
< 2.00
Average Cone
20
Maximum Cone
20
Comments
Average Cone
NA
NA
0.42
NA
0.22
NA
0.014
NA
NA
NA
0.16
Maximum Cone
NA
NA
1.00
NA
0.64
NA
0.022
NA
NA
NA
0.27
Comments
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Potassium
Zinc
CAS No.
7440097
7440666
R3-AS-01
NA
NA
R5B-AS-01
NA
NA
R7C-AS-01
NA
NA
R15-AS-01
140
B 0.24
Average Cone
140
0.24
Maximum Cone
140
0.24
Comments
Miscellaneous Characterization
Total Fluorine (mq/L)
Iqnitability (oF)
Corrosivity ( pH )
Heat of Combustion ( BTU/lb )
R3-AS-01
450
132
3
18,700
R5B-AS-01
110
57
5
19,245
R7C-AS-01
19.0
97
7
19,000
R15-AS-01
9,300 mq/kq
> 158
10.8
14,000
Average Cone
193
NA
NA
17,736
Maximum Cone
450
NA
NA
19,245
Comments
Comments:
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound that meets the identification criteria for which the result is less than the laboratory detection limit, but greater than
zero.
ND Not Detected.
NA Not Applicable.
-------
3.7 POLYMERIZATION
Polymerization is a process utilized for the conversion of propane/propylene and/or
butane/butene feeds from other operations into a low molecular weight, higher-octane, polymer
product, referred to as dimate. Dimate is used as a high octane gasoline blending component of
unleaded gasolines.
Almost 12 percent of the industry's polymerization catalyst (Dimersol and phosphoric
acid) volume was reported to be managed as a hazardous waste ("as hazardous", D002 and
D007).9
3.7.1 Process Descriptions
There are primarily two polymerization processes utilized by the petroleum refining
industry: phosphoric acid polymerization and the Dimersol process, licensed by IFF (Institute
Francais du Petrole, or the French Petroleum Institute). Process descriptions for each of these
two processes are provided in the following sections.
3.7.1.1 Phosphoric Acid Polymerization
Phosphoric acid polymerization units produce marginal octane gasoline from propylene
feeds from other operating units (i.e., the FCC unit, coking, etc). Phosphoric acid
polymerization is more widely used by industry than the Dimersol process, representing 80
percent of all polymerization units in the United States. Phosphoric acid polymerization unit
capacities range from 400 to 8,000 barrels per stream day, with the majority of units ranging
between 2,200 and 3,000 barrels per stream day (as reported in the §3007 survey).
Phosphoric acid polymerization utilizes a catalyst consisting of an alumina substrate
impregnated with phosphoric acid. A typical phosphoric acid polymerization unit contains one
or more reactors consisting of a series of tubes coming off of a single header. The reactor feed is
charged to the header and flows through the tubes. The tubes are packed with the phosphoric
acid catalyst. The reaction conditions are controlled to stop the polymerization at the desired C6
or C9 product. The polymerization reaction is highly exothermic and boiler feed water is fed
through the reactor (on the shell side of the tubes) to recover the heat for use as steam. Over
time, the catalyst's acid sites become blocked and the catalyst is slated for change-out.
After leaving the reactor, the reactor effluent is fractionated to give the desired products.
A simplified process flow diagram for a typical phosphoric acid polymerization unit is shown in
Figure 3.7.1.
9These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., Subtitle C landfill, recovery in
coker, etc.).
Petroleum Refining Industry Study 89 August 1996
-------
Figure 3.7.1. Process Flow Diagram for Phosphoric Acid Polymerization Process
Spent
Phosphoric
Acid
Catalyst
Reactor
LPG
Flash
Drum
Stabilizer
Polymer
. Gasoline
3.7.1.2 Dimersol Polymerization
As stated above, Dimersol polymerization units represent only 20 percent of the existing
polymerization units in the United States. The capacity of Dimersol units range from 1,000 to
5,500 barrels per stream day, with an average capacity of approximately 3,200 barrels per stream
day (as reported in the §3007 survey).
The Dimersol process is used to dimerize light olefins such as ethylene, propylene and
butylene. The process typically begins with the pretreatment of the propane/propylene or
butane/butene feed prior to entering the reactor section of the process. Pretreatment can include
the use of molecular sieve dryers, sand filters, etc. to remove water and/or H2S. Water in the
feed stream can deactivate the catalysts used in the Dimersol process. After drying the feed is
combined with a liquid nickel carboxylate/ethyl aluminum dichloride (EADC) catalyst prior to
entering the first of a series of three reactors. The first two are continuous stirred batch reactors
and the third is a plug-flow tubular reactor. The reactor feed is converted to the process product,
dimate, primarily in the first reactor, and additional conversion is achieved in the last two
reactors. The final reactor effluent consists of dimate product, unreacted C3/C4s, and liquid
catalyst. Immediately following the last reactor, the liquid catalyst is removed from the reactor
effluent by treating the reactor effluent with caustic, subsequent water washing, and filtering to
remove solids. Spent caustic residuals are typically reused or reclaimed on- or off-site, and as a
result, do not constitute solid wastes. After filtering, the product stream enters a "Dimersol
stabilizer," a distillation unit that removes unreacted LPG from the dimate product. In some
cases, the product stream is also further treated by drying. LPG from the stabilizer overhead is
typically sent to another unit of the refinery for further processing. The dimate product from the
bottom of the stabilizer is sent to storage or product blending.
Petroleum Refining Industry Study
90
August 1996
-------
A simplified process flow diagram for a typical Dimersol polymerization unit is shown in
Figure 3. 7.2.
Figure 3.7.2. Dimersol Polymerization Process Flow Diagram
Fesd
Catalyst Remover
LPG
W3CM1
Spent Caustr:
Stabiliser
Fresh. Caustic
3.7.2 Spent Phosphoric Acid Polymerization Catalyst
3.7.2.1 Description
Spent phosphoric acid polymerization catalyst is generated after the solid catalyst active
sites have become blocked and lost their reactivity.
3.7.2.2 Generation and Management
During reactor change-outs, spent phosphoric acid catalysts are flushed or water drilled
from the shell-and-tube reactors.
Twenty-two facilities reported generating a total quantity of 3,358 MT of this residual in
1992, according to the 1992 RCRA §3007 Questionnaire. Residuals were assigned to be "spent
phosphoric acid polymerization catalyst" if they were assigned a residual identification code of
"spent solid catalyst" or "spent catalyst fines" and were generated from a process identified as a
phosphoric acid polymerization unit. These correspond to residual codes 03-A and 03-B in
Section VII.2 of the questionnaire and process code 11-A in Section IV-l.C of the questionnaire.
Quality assurance was conducted by ensuring that all phosphoric acid polymerization catalysts
previously identified in the questionnaire (i.e., in Section V.B) were assigned in Section VII.2.
Based on the results of the questionnaire, 25 facilities use phosphoric acid polymerization
units and are thus likely to generate spent phosphoric acid polymerization catalyst. Due to the
infrequent generation of this residual, not all of these 25 facilities generated spent catalyst in
1992. However, there was no reason to expect that 1992 would not be a typical year with regard
to this residual's generation and management. Table 3.7.1 provides a description of the quantity
generated, number of streams reported, number of streams not reporting volumes (data requested
was unavailable and facilities were not required to generate it), total and average volumes.
Petroleum Refining Industry Study
91
August 1996
-------
Table 3.7.1. Generation Statistics for Phosphoric Acid Catalyst
from Polymerization, 1992
Final Management
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite Subtitle C landfill
Disposal in onsite Subtitle D landfill
Onsite land treatment
Transfer for use as an ingredient in
products placed on the land
TOTAL
#of
Streams
12
3
2
6
3
7
33
# of Streams
w/ Unreported
Volume
0
0
0
0
0
0
0
Total Volume
(MT)
1,429.5
62
349
246.8
728
542.5
3357.8
Average
Volume (MT)
119
20.7
174.5
41
242.7
77.5
101.7
» ^7 O f
>. 1.2..
Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.7.1. No data were
available to the Agency suggesting any other management practices.
3.7.2.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.7.2 summarizes the physical properties of the spent catalyst as reported in
Section VILA of the §3007 survey.
• One record sample of phosphoric acid polymerization catalyst was collected and
analyzed by EPA. The sample is representative of typical phosphoric acid
polymerization catalyst used by the industry and is summarized in Table 3.7.3.
The record sample was analyzed for total and TCLP levels of volatiles, semivolatiles,
and metals, reactivity (pyrophoricity) and corrosivity. The sample was found to exhibit the
hazardous waste characteristic of corrosivity. Dimersol and phosphoric acid catalysts were
categorized together in the consent decree, therefore, a summary of the results for both residuals
is presented in Table 3.7.7. Only constituents detected in at least one sample are shown in this
table.
Petroleum Refining Industry Study
92
August 1996
-------
Table 3.7.2. Phosphoric Acid Catalyst from Polymerization: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Specific Gravity
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 10-100 urn, %
Particle <10 urn, %
Median Particle Diameter,
microns
#of
Values
21
12
12
14
16
15
20
29
28
35
16
17
16
16
16
10
# of Unreported
Values1
21
30
30
28
26
27
22
13
14
7
26
25
26
26
26
31
10th%
1.4
0.01
1
60
0
0
0.85
0
0
50
0
0
0
0
0
5030
50th %
4.7
7
10
93.3
0
0
0.96
0
0
100
0
95
5
0
0
12,000
90th %
7
40
40
200
25.5
16.6
1.4
50
1
100
0
95
5
100
0
12,000
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
Table 3.7.3. Phosphoric Acid Polymerization Catalyst Record Sampling Locations
Sample Number
R16-PC-01
Location
Koch, St. Paul, MN
Description
Phosphoric acid catalyst
3.7.2.5 Source Reduction
No source reduction techniques were reported by industry or found in the literature
search for this residual.
Petroleum Refining Industry Study
93
August 1996
-------
3.7.3 Spent Dimersol Polymerization Catalyst
3.7.3.1 Description
Dimersol catalyst is added to the reactor feed stream and exits the final reactor as part of
the reactor effluent. The liquid catalyst is then removed from the reactor effluent by
neutralization (contact with caustic). Spent caustic streams, containing the spent dimersol
catalyst, are commonly reused on-site or sent off-site for metals reclamation or caustic recovery,
and as a result are typically not solid wastes. Spent catalyst also may be generated in two other
points in the process. First, during routine shutdowns spent catalyst may be generated as a
component of any reactor sludge removed from the reactors. Second, certain Dimersol
processes contain filters following caustic neutralization and water washing to remove entrained
residual nickel from the dimate product. The filters are removed and disposed periodically.
3.7.3.2 Generation and Management
Dimersol catalysts are generated as solid wastes in the form of reactor sludges generated
during reactor clean-outs and as spent nickel filters.
Four facilities reported generating a total quantity of 761.5 MT of this residual as a
reactor sludge in 1992, according to the 1992 RCRA §3007 Survey. Residuals were assigned to
be "spent dimersol catalyst" if they were assigned a residual identification code of "spent solid
catalyst" or "spent catalyst fines" or "other process sludge" and were generated from a process
identified as a Dimersol polymerization unit. These correspond to residual codes "03-A," "03-
B" and "02-D" in Section VII.2 and process code "11-B" in Section IV-l.C of the questionnaire.
Quality assurance was conducted by ensuring that all dimersol catalysts previously identified in
the questionnaire (i.e., in Section V.B) were assigned in Section VII.2.
Based on the results of the survey, 7 facilities use Dimersol polymerization units and may
generate spent dimersol catalyst. Due to the continuous generation of this residual, 1992 is
expected to be a typical year in regard to catalyst generation volume and management. There
was no reason to expect that 1992 would not be a typical year with regard to this residual's
generation and management. Table 3.7.4 provides a description of the quantity generated,
number of streams reported, number of streams not reporting volumes (data requested was
unavailable and facilities were not required to generate it), total and average volumes.
3.7.3.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.7.4. No data were
available to the Agency suggesting any other management practices. Unlike with phosphoric
acid polymerization catalyst, EPA does not expect spent Dimersol catalyst to be land treated due
to the physical nature of the filters.
Petroleum Refining Industry Study 94 August 1996
-------
Table 3.7.4. Generation Statistics for Spent Dimersol Polymerization Catalyst, 1992
Final Management
Disposal Offsite Subtitle C Landfill
Disposal Onsite Subtitle D Landfill
Offsite incineration
Recover onsite in a coker
TOTAL
#of
Streams
1
1
1
1
4
# of Streams w/
Unreported
Volume
0
0
1
0
1
Total
Volume (MT)
3.4
8.8
0.3
749
761.5
Average
Volume (MT)
3.4
8.8
0.3
749
190.4
3.7.3.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.7.5 summarizes the physical properties of the spent catalyst as reported in
Section VILA of the §3007 survey.
• Two record samples of Dimersol polymerization catalyst were collected and analyzed
by EPA. The samples represent typical Dimersol polymerization catalyst used by the
industry and are summarized in Table 3.7.6.
The two record samples were analyzed for total and TCLP levels of volatiles,
semivolatiles, and metals, and pyrophoricity and corrosivity. None of the samples were found to
exhibit a hazardous waste characteristic. Dimersol and phosphoric acid catalysts were
categorized together in the consent decree, therefore, a summary of the results for both residuals
is presented in Table 3.7.7. Only constituents detected in at least one sample are shown in this
table.
3.7.3.5 Source Reduction
No source reduction techniques were reported by industry or found in the literature
search for this residual.
Petroleum Refining Industry Study
95
August 1996
-------
Table 3.7.5. Spent Dimersol Polymerization Catalyst Physical Properties
Properties
PH
Flash Point, C
Oil and Grease, vol%
Total Organic Carbon, vol%
Specific Gravity
Aqueous Liquid, %
Organic Liquid, %
Solid, %
#of
Values
7
4
3
3
6
11
11
11
#of
Unreported
Values1
4
7
8
8
5
0
0
0
10th%
3.8
93.3
2.6
0.08
0.7
0
0
20
Mean
5.5
93.3
5.3
4.1
1.2
0
0
100
90th %
9
100
6.4
9.5
1.4
70
60
100
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
Table 3.7.6. Dimersol Polymerization Catalyst Record Sampling Locations
Sample Number
R6B-PC-01
R16-PC-02
Location
Shell, Norco, LA
Koch, St. Paul, MN
Description
Dimersol filter
Dimersol filter
Petroleum Refining Industry Study
96
August 1996
-------
Petroleum Refining Industry Study 97
1
I
-------
Petroleum Refining Industry Study 98
i auie j.i.i. ruiymei i£
-------
STUDY OF SELECTED
PETROLEUM REFINING RESIDUALS
INDUSTRY STUDY
August 1996
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Solid Waste
Hazardous Waste Identification Division
401 M Street, SW
Washington, DC 20460
-------
3.8 RESIDUAL UPGRADING
After vacuum distillation, there are still some valuable oils left in the vacuum-reduced
crude. Vacuum tower distillation bottoms and other residuum feeds can be upgraded to higher
value products such as higher grade asphalt or feed to catalytic cracking processes. Residual
upgrading includes processes where asphalt components are separated from gas oil components
by the use of a solvent. It also includes processes where the asphalt value of the residuum is
upgraded (e.g., by oxidation) prior to sale. Off-spec product and fines, as well as process
sludges, are study residuals from this category.
3.8.1 Process Descriptions
A total of 47 refineries reported using residual upgrading units. Four types of residual
upgrading processes were reported in the 1992 RCRA §3007 Petroleum Refining Survey:
• Solvent Deasphalting
• Asphalt Oxidation
• Supercritical Extraction
• Asphalt Emulsion
Asphalt uses are typically divided into use as road oils, cutback asphalts, asphalt
emulsions, and solid asphalts. These asphalt products are used in paving roads, roofing, paints,
varnishes, insulating, rust-protective compositions, battery boxes, and compounding materials
that go into rubber products, brake linings, and fuel briquettes (REF).
3.8.1.1 Solvent Deasphalting
Residuum from vacuum distillation is separated into asphalt components and gas oil
components by solvent deasphalting. Figure 3.8.1 provides a simplified process flow diagram.
The hydrocarbon solvent is compressed and contacted with the residuum feed. The extract
contains the paraffinic fractions (deasphalted oil or DAO), and the raffmate contains the
asphaltic components. The extract and raffmate streams are sent to separate solvent recovery
systems to reclaim the solvent. The DAO may be further refined or processed, used as catalytic
cracking feed, sent to lube oil processing/blending, or sold as finished product. The following
types of solvents are typically used for the following residual upgrading processes:
• Propane is the best choice for lube oil production due to its ability to extract only
paraffinic hydrocarbons and to reject most of the carbon residue. (McKetta)
• A mixture of propane and butane is valuable for preparing feedstocks for catalytic
cracking processes due to its ability to remove metal-bearing components.
(McKetta)
• Pentane deasphalting, plus hydrodesulfurization, can produce more feed for
catalytic cracking or low sulfur fuel oil. (McKetta)
Petroleum Refining Industry Study 107 August 1996
-------
Figure 3.8.1. Solvent Deasphalting Process Flow Diagram
Solvent Recycle
Residuum
Liquid-
Liquid
Extractor
' «—
^•~
Solvent
Recovery
Section
Sludge
Solvent
Recovery
Section
Asphalt
• One facility reported using propane and phenol solvents for deasphalting
residuum. The DAO is sent to lube oil processing and the asphalt fraction is sent
to delayed coking or fuel oil blending.
During process upsets, heavy hydrocarbons may become entrained in the solvent
recovery systems, and off-specification product may be generated. The entrained hydrocarbons
are periodically removed from the unit as a process sludge and typically disposed in an industrial
landfill. The off-specification product are returned to the process for re-processing.
3.8.1.2
Asphalt Oxidation (Asphalt Blowing")
Residuum from the vacuum tower or from solvent deasphalting is upgraded by oxidation
with air. Figure 3.8.2 provides a simplified process flow diagram. Air is blown through the
asphalt that is heated to about 500°F, starting an exothermic reaction. The temperature is
controlled by regulating the amount of air and by circulating oil or water through cooling coils
within the oxidizer. The oxygen in the air reacts with hydrogen in the residuum to form water,
and the reaction also couples smaller molecules of asphalt into larger molecules to create a
heavier product. These reactions changes the characteristics of the asphalt to a product with the
desired properties.
During this process, coke will form on the oxidizer walls and the air sparger. The coke is
removed periodically (1 to 2 years) and sent to the coke pad for sale, mixed with asphalt for use
as road material, stored, or disposed. The off-gases from the process are scrubbed to remove
hydrocarbons prior to burning in an thermal unit such as an incinerator or furnace.
Petroleum Refining Industry Study
108
August 1996
-------
Figure 3.8.2. Asphalt Oxidation Process Flow Diagram
GasOi
HaatSE
Residuum.
Racyda
Hash
Chamber
/•\
Elcwsr
033 Q-03
LPG
Gasolins
Naphtha
Gas Oil
•^r-
-fc
Asphalt
1 ^
h.
CutBacls
(Naphtha or Ksmssns)
Blown. Asphalt
Asphalt
Emulsions
Supercritical Extraction
The Residuum Oil Supercritical Extraction (ROSE) process is not, in a strict sense, a
supercritical fluid extraction process. The primary extraction step is not carried out at
supercritical conditions, but at liquid conditions that take advantage of the variable solvent
power of a near-critical liquid. A simplified process flow diagram is provided in Figure 3.8.3.
The first stage of the ROSE process consists of mixing residuum with compressed liquid butane
or pentane and precipitating the undesired asphaltene fraction. Butane is used for its higher
solvent power for heavy hydrocarbons. If an intermediate resin fraction is desired, another
separator and stripper system would be used directly after the asphaltene separator. To recover a
resin fraction, the overhead from the asphaltene separator is heated to near the critical
temperature of the butane. At the elevated, near-critical temperature, the solvent power of the
compressed liquid butane decreases and the resins precipitate from solution. The remaining
fraction would consist of deasphalted light oils dissolved in butane. The butane is typically
recovered using steam.
Petroleum Refining Industry Study
109
August 1996
-------
Figure 3.8.3. Supercritical Extraction Process Flow Diagram
OBQ-Q-I
The DAO may be sent to FCC, blended into lubricating oil, or sold as finished product.
The asphaltene and resins are reported to be blended into No. 6 fuel oil. The solvent and steam
are condensed and collected in a surge drum where the solvent is recycled back to the process.
This surge drum accumulates sludges during process upsets that are removed during routine
process turnarounds and disposed as nonhazardous wastes.
Asphalt Emulsion
Residuals from the vacuum tower may be upgraded to an asphalt emulsion by milling
soap (or shear mixing) with the asphalt. These emulsions are used for road oils, where good
adhesion is required.
This process generated residuals from the cleanout of the soap tanks and from the
generation of off-spec emulsions. The soap tank cleanout residuals are typically sent to the
wastewater treatment plant, and the off-spec emulsions are sent to a pit where heat is applied to
break the emulsion. The soap fraction is sent the wastewater treatment system and the oil
fraction is recycled back to the coker feed.
Petroleum Refining Industry Study
110
August 1996
-------
3.8.2 Off-specification Product from Residual Upgrading
3.8.2.1 Description
This residual was identified in the consent decree based on an incorrect characterization
of data in a supporting document generated from 1983 PRDB data. After conducting a review
of the underlying data, it was determined that volumes associated with the category of "off-
specification product from residual upgrading" were actually process sludges generated during
process upset conditions. The Agency's finding regarding this category was corroborated during
its field investigation where this residual category was not identified and in the §3007 survey
results. Generally, refineries re-work any residuum that does not initially meet product
specifications within the upgrading process and rarely (one reported in 1992 in the §3007
survey) generate off-specification product for disposal.
3.8.2.2
Generation and Management
Off-spec product from residual upgrading includes material generated from asphalt
oxidation, solvent deasphalting, and other upgrading processes. Residuals were assigned to be
"off-specification product from residual upgrading" if they were assigned a residual
identification code of "off-specification product" or "fines" and were generated from a process
identified as a residual upgrading unit. These correspond to residual codes "05" and "06" in
Section VII.2 of the questionnaire and process code "13" in Section IV-l.C of the questionnaire.
Based on the results of the questionnaire, 47 facilities use residual upgrading processes
and thus could potentially generate off-specification product from residual upgrading. Only one
facility reported this residual, generating 800 MT that was recovered within the process. The
base year, 1992, was expected to be a typical year for residual upgrading processes and the
survey results are in keeping with the Agency's understanding of this process. Table 3.8.1
provides a description of the quantity generated and number of reporting facilities.
Table 3.8.1. Generation Statistics for Off-Specification Product
from Residual Upgrading, 1992
Final Management
Other recovery onsite: reuse in
extraction process
#of
Streams
1
# of Streams
w/ Unreported
Volume
0
Total Volume
(MT)
800
Average
Volume (MT)
800
j.o.z.j
Plausible Management
The Agency does not find it necessary to consider other management practices because
off-spec product from residual upgrading had been classified as a residual of concern based on
erroneous old data and in fact is not generated for disposal.
Petroleum Refining Industry Study
111
August 1996
-------
3.8.2.4
Characterization
study:
Only one source of residual characterization data were developed during the industry
Table 3.8.2 summarizes the physical properties of the off-specification product as
reported in Section VILA of the §3007 survey.
Because it is rarely generated, no record samples of this residual were available during
record sampling for analysis.
Table 3.8.2. Off-Specification Product from Residual Upgrading: Physical Properties
Properties
Flash Point, °C
Specific Gravity
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Other, %
#of
Values
1
1
1
1
1
1
#of
Unreported
Values1
2
2
2
2
2
2
10th%
99.00
1.02
40.00
60.00
100.00
100.00
50th %
99.00
1.02
40.00
60.00
100.00
100.00
90th %
99.00
1.02
40.00
60.00
100.00
100.00
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
3.8.2.5
Source Reduction
No source reduction techniques were reported by industry or found in the literature
search for this residual.
3.8.3 Process Sludge from Residual Upgrading
3.8.3.1 Description
Process sludge is generated from miscellaneous parts of the various residual upgrading
processes. This category is neither uniform nor routinely generated. Solvent deasphalting may
generate a sludge due to hydrocarbon carryover in the solvent recovery system. Similarly, the
ROSE process may generate sludges due to process upsets in the solvent condensate collection
system. Additional sludges may be generated during unit turnarounds and in surge drums and
condensate knockout drums.
Petroleum Refining Industry Study
112
August 1996
-------
Three residuals were reported to be managed "as hazardous", accounting for 25 percent
of the volume of this category generated in 1992.1
. J .
Generation and Management
Twenty-one facilities reported generating a total quantity of 241 MT of this residual in
1992, according to the 1992 survey. Residuals were assigned to be "process sludge from
residual upgrading" if they were assigned a residual identification code of "process sludge" and
were generated from a process identified as a "residual upgrading" unit. These correspond to
residual code "02-D" in Section VII.2 of the questionnaire and process code "13" in Section IV-
l.C of the questionnaire.
Based on the results of the questionnaire, 47 facilities use residual upgrading units and
thus may generate process sludge from residual upgrading. Due to the infrequent generation of
this residual, not all of these 47 facilities generated sludge in 1992. However, there was no
reason to expect that 1992 would not be a typical year with regard to this residual's generation
and management. Table 3.8.3 provides a description of the quantity generated, number of
streams reported, number of streams not reporting volumes (data requested was unavailable and
facilities were not required to generate it), total and average volumes.
Table 3.8.3. Generation Statistics for Process Sludge from Residual Upgrading, 1992
Final Management
Discharge to onsite wastewater
treatment facility
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite Subtitle C landfill
Disposal in onsite Subtitle D landfill
Offsite incineration
Other recycling, reclamation, or reuse:
onsite road material
Recovery onsite via distillation
Transfer with coke product or other
refinery product
TOTAL
#of
Streams
3
12
1
4
2
1
4
1
4
32
# of Streams
w/ Unreported
Volume
0
0
0
0
0
0
0
0
0
0
Total Volume
(MT)
3.94
137.56
0.10
62.00
7.30
9.00
0.22
16.00
5.44
241.56
Average
Volume (MT)
1.31
11.46
0.10
15.50
3.65
9.00
0.06
16.00
1.36
7.55
'These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., managed in WWTP, Subtitle C
landfill, etc.).
Petroleum Refining Industry Study
113
August 1996
-------
3.8.3.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.8.3. The Agency
gathered information suggesting that "recovery onsite in an asphalt production unit" (3.6 MT)
and "transfer to offsite entity: unspecified" (unreported quantity) were used in other years. This
non-1992 management practice is comparable with other recovery practices reported in 1992.
3.8.3.4 Characterization
Two sources of residual characterization data were developed during the industry study:
• Table 3.8.4 summarizes the physical properties of the sludge as reported in
Section VILA of the §3007 survey.
• One record sample of process sludge from residual upgrading was collected and
analyzed by EPA. This sample is summarized in Table 3.8.5.
The sample was analyzed for total and TCLP levels of volatiles, semivolatiles, metals,
and ignitability. The sample was found to exhibit the toxicity characteristic for benzene. A
summary of the results is presented in Table 3.8.6. Only constituents detected in the sample are
shown in this table.
3.8.3.5 Source Reduction
Source reduction techniques were reported to be process modifications and better
housekeeping. This residual is generated infrequently and in very small quantities, therefore
limited information was expected.
Petroleum Refining Industry Study 114 August 1996
-------
Table 3.8.4. Process Sludge from Residual Upgrading: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Specific Gravity
BTU Content, BTU/lb
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Other, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 10-100 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
11
8
7
14
7
16
12
3
23
23
34
18
12
9
5
1
1
1
#of
Unreported
Values1
38
41
42
35
42
33
37
46
26
26
15
31
37
40
44
48
48
48
10th%
5.50
0.01
0.01
82.22
0.10
50.00
0.90
11.00
0.00
0.00
10.00
0.00
20.00
1.00
0.00
0.00
0.00
60.00
50th %
6.30
0.74
15.00
94.17
9.00
98.50
1.08
5,000.00
0.00
5.00
99.00
0.00
50.00
49.00
1.00
0.00
0.00
60.00
90th %
7.60
50.00
4400.00
315.56
100.00
100.00
1.85
10,000.00
25.00
90.00
100.00
2.00
100.00
80.00
1.00
0.00
0.00
60.00
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
Table 3.8.5. Process Sludge from Residual Upgrading Record Sampling Locations
Sample Number
R1-RU-01
Location
Marathon, Indianapolis, IN
Description
ROSE unit scale/sludge
Petroleum Refining Industry Study
115
August 1996
-------
Table 3.8.6. Process Sludge from Residual Upgrading Characterization
Volatile Organics - Method 8260A M9/kg
Acetone
Benzene
Ethylbenzene
Methylene chloride
4-Methyl-2-pentanone
n-Propylbenzene
Toluene
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
o-Xylene
m,p-Xylenes
Naphthalene
CAS No.
67641
71432
100414
75092
108101
103651
108883
95636
108678
95476
108383/106423
91203
R1-RU-01
B 120,000
73,000
130,000
64,000
63,000
65,000
310,000
570,000
150,000
230,000
690,000
160,000
Comments
TCLP Volatile Organics - Methods 1311 and 8260A M9/L
Benzene
Ethylbenzene
Toluene
1 ,2,4-Trimethylbenzene
o-Xylene
m,p-Xylene
CAS No.
71432
100414
108883
95636
95476
108383/106423
R1-RU-01
2,600
570
4,100
990
1,300
2,800
Comments
Semivolatile Organics - Method 8270B M9/kg
Acenaphthene
Anthracene
Dibenzofuran
Fluorene
Phenanthrene
Pyrene
1 -Methylnaphthalene
2-Methylnaphthalene
Naphthalene
CAS No
83329
120127
132649
86737
85018
129000
90120
91576
91203
R1-RU-01
J 38,000
J 13,000
J 13,000
J 39,000
120,000
J 19,000
390,000
570,000
190,000
Comments
TCLP Semivolatile Organics - Methods 1311 and 8270B |jg/L
Bis(2-ethylhexyl)phthalate
2,4-Dimethylphenol
Indene
1 -Methylnaphthalene
2-Methylnaphthalene
CAS No.
117817
105679
95136
90120
91576
R1-RU-01
J 30
J 52
J 16
J 96
130
Comments
Petroleum Refining Industry Study
116
August 1996
-------
Table 3.8.6. Process Sludge from Residual Upgrading Characterization (continued)
TCLP Semivolatile Organics - Methods 1311 and 8270B ug/L (continued)
2-Methylphenol
3/4-Methylphenol
Naphthalene
Phenol
CAS No.
95487
NA
91203
108952
Total Metals - Methods 6010, 7060, 7421
Aluminum
Antimony
Arsenic
Barium
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Vanadium
Zinc
CAS No.
7429905
7440360
7440382
7440393
7440439
7440702
7440473
7440484
7440508
7439896
7439921
7439954
7439965
7439976
7439987
7440020
7440622
7440666
R1-RU-01
J 65
J 85
190
J 57
Comments
7470, 7471, and 7841 mg/kg
R1-RU-01
150
14.0
43.0
41.0
1.10
15,000
86.0
13.0
92.0
200,000
20.0
6,500
770
0.11
24.0
90.0
100
40.0
Comments
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Calcium
Iron
Manganese
Zinc
CAS No.
7440702
7439896
7439965
7440666
R1-RU-01
130
120
3.90
0.24
Comments
Miscellaneous Characterization
Ignitability (oF)
R1-RU-01
199
Comments
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound
that meets the identification criteria for which the result is less than the laboratory detection limit, but
greater than zero.
Petroleum Refining Industry Study
117
August 1996
-------
3.9 LUBE OIL PROCESSING
Vacuum distillates are treated and refined to produce a variety of lubricants. Wax,
aromatics, and asphalts are removed by unit operations such as solvent extraction and
hydroprocessing; clay may also be used. Various additives are used to meet product
specifications for thermal stability, oxidation resistances, viscosity, pour point, etc.
3.9.1 Process Descriptions
The manufacture of lubricating oil base stocks consists of five basic steps:
1) Distillation
2) Deasphalting to prepare the feedstocks
3) Solvent or hydrogen refining to improve viscosity index and quality
4) Solvent or catalytic dewaxing to remove wax and improve low temperature properties
of paraffin!c lubes
5) Clay or hydrogen finishing to improve color, stability, and quality of the lube base
stock.
Based on results of the 1992 survey, 22 facilities reported conducting lube oil processing.
The finished lube stocks are blended with each other and additives using batch and continuous
methods to produce formulated lubricants. The most common route to finishing lube feedstocks
consists of solvent refining, solvent dewaxing, and hydrogen finishing. The solvent and clay
processing route or the hydrogen refining and solvent dewaxing route are also used. The all-
hydrogen processing (lube hydrocracking-catalytic dewaxing-hydrorefming) route is used by
two refiners for the manufacture of a limited number of paraffmic base oils. Figure 3.9.1
provides a general process flow diagram for lube oil processing.
Lube Distillation
Lube processing may be the primary production process at some facilities, while at others
it is only one of many operations. The initial step is to separate the crude into the fractions
which are the raw stocks for the various products to be produced. The basic process consists of
an atmospheric distillation unit and a vacuum distillation unit. The majority of the lube stocks
boil in the range between 580°F and 1000°F and are distilled in the vacuum unit to the proper
viscosity and flash specifications. Caustic solutions are sometimes introduced to the feed to
neutralize organic acids present in some crude oils. This practice reduces or eliminates corrosion
in downstream processing units, and improves color, stability, and refining response of lube
distillates.
Petroleum Refining Industry Study 118 August 1996
-------
Hs -*•
Figure 3.9.1. Lube Oil Processing Flow Diagram
Railed Hydrogen
Feed
QuerchGas
Heater
be
Ab
Lean
Amine
Gas
Pith P^overy
ifi
!
Light Gas &
""*" Furnace Col
. Light
Neutral
,. Medium
Neutral
, Heavy
Neutral
*• Wax
Lube Deasphalting
Other facilities incorporate lube deasphalting to process vacuum residuum into lube oil
base stocks. Propane deasphalting is most commonly used to remove asphaltenes and resins
which contribute an undesirable dark color to the lube base stocks. This process typically uses
baffle towers or rotating disk contactors to mix the propane with the feed. Solvent recovery is
accomplished with evaporators, and supercritical solvent recovery processes are also used in
some deasphalting units. Another deasphalting process is the Duo-Sol Process that is used to
both deasphalt and extract lubricating oil feedstocks. Propane is used as the deasphalting solvent
and a mixture of phenol and cresylic acids are used as the extraction solvent. The extraction is
conducted in a series of batch extractors followed by solvent recovery in multistage flash
distillation and stripping towers. See the section on Residual Upgrading for additional
discussion on these processes.
Lube Refining Processes
Chemical, solvent, and hydrogen refining processes have been developed and are used to
remove aromatics and other undesirable constituents, and to improve the viscosity index and
quality of lube base stocks. Traditional chemical processes that use sulfuric acid and clay
refining have been replaced by solvent extraction/refining and hydrotreating which are more
effective, cost efficient, and environmentally more acceptable. Chemical refining is used most
often for the reclamation of used lubricating oils or in combination with solvent or hydrogen
refining processes for the manufacture of specialty lubricating oils and by-products.
Chemical Refining Processes: Acid-alkali refining, also called "wet refining", is a
process where lubricating oils are contacted with sulfuric acid followed by neutralization with
Petroleum Refining Industry Study
119
August 1996
-------
alkali. Oil and acid are mixed and an acid sludge is allowed to coagulate. The sludge is
removed or the oil is decanted after settling, and more acid is added and the process repeated.
Acid-clay refining, also called "dry refining" is similar to acid-alkali refining with the
exception that clay and a neutralizing agent are used for neutralization. This process is used for
oils that form emulsions during neutralization.
Neutralization with aqueous and alcoholic caustic, soda ash lime, and other neutralizing
agents is used to remove organic acids from some feedstocks. This process is conducted to
reduce organic acid corrosion in downstream units or to improve the refining response and color
stability of lube feedstocks.
Hydrogen Refining Processes: Hydrogen refining, also called hydrotreating, has since
been replaced with solvent refining processes which are more cost effective. Hydrotreating
consists of lube hydrocracking as an alternative to solvent extraction, and hydrorefming to
prepare specialty products or to stabilize hydrocracked base stocks. Hydrocracking catalysts are
proprietary to the licensors and consist of mixtures of cobalt, nickel, molybdenum, and tungsten
on an alumina or silica-alumina-based carrier. Hydrorefming catalysts are proprietary but
usually consist of nickel-molybdenum on alumina.
Lube hydrocracking are used to remove nitrogen, oxygen, and sulfur, and convert the
undesirable polynuclear aromatics and polynuclear naphthenes to mononuclear naphthenes,
aromatics, and isoparaffms which are typically desired in lube base stocks. Feedstocks consist of
unrefined distillates and deasphalted oils, solvent extracted distillates and deasphalted oils, cycle
oils, hydrogen refined oils, and mixtures of these hydrocarbon fractions.
Lube hydrorefming processes are used to stabilize or improve the quality of lube base
stocks from lube hydrocracking processes and for manufacture of specialty oils. Feedstocks are
dependent on the nature of the crude source but generally consist of waxy or dewaxed-sol vent-
extracted or hydrogen-refined paraffmic oils and refined or unrefined naphthenic and paraffmic
oils from some selected crudes.
Solvent Refining Processes: Feedstocks from solvent refining processes consist of
paraffmic and naphthenic distillates, deasphalted oils, hydrogen refined distillates and
deasphalted oils, cycle oils, and dewaxed oils. The products are refined oils destined for further
processing or finished lube base stocks. The by-products are aromatic extracts which are used in
the manufacture of rubber, carbon black, petrochemicals, FCCU feed, fuel oil, or asphalt. The
major solvents used today are N-methyl-2-pyrolidone (NMP) and furfural, with phenol and
liquid sulfur dioxide used to a lesser extent.
The solvents are typically recovered in a series of flash towers. Steam or inert gas
strippers are used to remove traces of solvent, and a solvent purification system is used to
remove water and other impurities from the recovered solvent.
Petroleum Refining Industry Study 120 August 1996
-------
Lube Dewaxing Processes
Lube feedstocks typically contain increased wax content resulting from deasphalting and
refining processes. These waxes are normally solid at ambient temperatures and must be
removed to manufacture lube oil products with the necessary low temperature properties.
Catalytic dewaxing and solvent dewaxing (the most prevalent) are processes currently in use;
older technologies include cold settling, pressure filtration, and centrifuge dewaxing.
Catalytic Dewaxing: Because solvent dewaxing is relatively expensive for the
production of low pour point oils, various catalytic dewaxing (selective hydrocracking)
processes have been developed for the manufacture of lube oil base stocks. The basic process
consists of a reactor containing a proprietary dewaxing catalyst followed by a second reactor
containing a hydrogen finishing catalyst to saturate olefins created by the dewaxing reaction and
to improve stability, color and demulsibility of the finished lube oil.
Solvent Dewaxing: Solvent dewaxing consists of the following steps: crystallization,
filtration, and solvent recovery. In the crystallization step, the feedstock is diluted with the
solvent and chilled, solidifying the wax components. The filtration step removes the wax from
the solution of dewaxed oil and solvent. Solvent recovery removes the solvent from the wax
cake and filtrate for recycle by flash distillation and stripping. The major processes in use today
are the ketone dewaxing processes. Other processes that are used to a lesser degree include the
Di/Me Process and the propane dewaxing process.
The most widely used ketone processes are the Texaco Solvent Dewaxing Process and
the Exxon Dilchill Process. Both processes consist of diluting the waxy feedstock with solvent
while chilling at a controlled rate to produce a slurry. The slurry is filtered using rotary vacuum
filters and the wax cake is washed with cold solvent. The filtrate is used to prechill the
feedstock and solvent mixture. The primary wax cake is diluted with additional solvent and
filtered again to reduce the oil content in the wax. The solvent recovered from the dewaxed oil
and wax cake by flash vaporization and recycled back into the process. The Texaco Solvent
Dewaxing Process (also called the MEK process) uses a mixture of MEK and toluene as the
dewaxing solvent, and sometimes uses mixtures of other ketones and aromatic solvents. The
Exxon Dilchill Dewaxing Process uses a direct cold solvent dilution-chilling process in a special
crystallizer in place of the scraped surface exchangers used in the Texaco process.
The Di/Me Dewaxing Process uses a mixture of dichloroethane and methylene dichloride
as the dewaxing solvent. This process is used by a few refineries in Europe. The Propane
Dewaxing Process is essentially the same as the ketone process except for the following:
propane is used as the dewaxing solvent and higher pressure equipment is required, and chilling
is done in evaporative chillers by vaporizing a portion of the dewaxing solvent. Although this
process generates a better product and does not require crystallizers, the temperature differential
between the dewaxed oil and the filtration temperature is higher than for the ketone processes
(higher energy costs), and dewaxing aids are required to get good filtration rates.
Petroleum Refining Industry Study 121 August 1996
-------
Lube Oil Finishing Processes
Today, hydrogen finishing processes (also referred to as hydrorefming) have largely
replaced the more costly acid and clay finishing processes. Hydrogen finishing processes are
mild hydrogenation processes used to improve the color, odor, thermal, and oxidative stability,
and demulsibility of lube base stocks. The process consists of fixed bed catalytic reactors that
typically use a nickel-molybdenum catalyst to neutralize, desulfurize, and denitrify lube base
stocks. These processes do not saturate aromatics or break carbon-carbon bonds as in other
hydrogen finishing processes. Sulfuric acid treating is still used by some refiners for the
manufacture of specialty oils and the reclamation of used oils. This process is typically
conducted in batch or continuous processes similar to the chemical refining processes discussed
earlier, with the exception that the amount of acid used is much lower that used in acid refining.
Clay contacting involves mixing the oil with fine bleaching clay at elevated temperature
followed by separation of the oil and clay. This process improves color and chemical, thermal,
and color stability of the lube base stock, and is often combined with acid finishing. Clay
percolation is a static bed absorption process used to purify, decolorize, and finish lube stocks
and waxes. It is still used in the manufacture of refrigeration oils, transformer oils, turbine oils,
white oils, and waxes.
3.9.2 Treating Clay from Lube Oil Processing
3.9.2.1 Description
The majority of treating clays (including other sorbents) generated from lube oil
processing are from acid-clay treating in refining or lube oil finishing. The average volume is
approximately 40 metric tons.
3.9.2.2 Generation and Management
The spent clay is vacuumed or gravity dumped from the vessels into piles or into
containers such as drums and roll-off bins. Only one residual was reported to be managed "as
hazardous" from this category in 1992.
Seven facilities reported generating a total quantify of approximately 733 metric tons of
this residual in 1992, according to the 1992 RCRA §3007 Questionnaire. Residual were
assigned to be "treating clay from lube oil processes" if they were assigned a residual
identification code of "spent sorbent" and were generated from a lube oil process. These
correspond to residual code "05" in Section VILA of the questionnaire and process code "17" in
Section IV.C of the questionnaire. Table 3.9.1 provides a description of the 1992 management
practices, quantity generated, number of streams reported, number of streams not reporting
volumes (data requested was unavailable and facilities were not required to generate it), total and
average volumes.
Petroleum Refining Industry Study 122 August 1996
-------
Table 3.9.1. Generation Statistics for Treating Clay from Lube Oil, 1992
Final Management
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite Subtitle C landfill
Onsite land treatment
Other recycling, reclamation, or reuse:
cement plant
Other recycling, reclamation, or reuse:
onsite regeneration
TOTAL
#of
Streams
1
2
1
1
1
12
18
# of Streams
w/ Unreported
Volume
1
0
0
0
0
0
1
Total Volume
(MT)
36.7
78.7
5
9.8
249.2
354
733.4
Average
Volume (MT)
36.7
39.4
5
9.8
249.2
29.5
40.7
3.9.2.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.9.1. No data were
available to the Agency suggesting any other management practices. In addition, EPA compared
the management practice reported for lube oil treating clay to those reported for treating clays
from extraction, alkylation, and isomerization2 based on expected similarities. No additional
management practices were reported.
3.9.2.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.9.2 summarizes the physical and chemical properties of treating clay from lube
oil processes as reported in Section VILA of the §3007 survey.
• One record sample of treating clay from lube oil processes was collected and analyzed
by EPA. Sampling information is summarized in Table 3.9.3.
The collected sample is expected to be generally representative of treating clay from lube
oil processes. The sample was analyzed for total and TCLP levels of volatiles, semi-volatiles,
and metals. The sample did not exhibit any of the hazardous waste characteristics. A summary
of the analytical results is presented in Table 3.9.4. Only constituents detected in the sample are
reported.
2EPA did not compare these management practices to those reported for the broader category of "treating clay from
clay filtering" due to the diverse types of materials included in this miscellaneous category.
Petroleum Refining Industry Study
123
August 1996
-------
Table 3.9.2. Treating Clay from Lube Oil: Physical Properties
Properties
PH
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Specific Gravity
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 10-100 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
3
2
12
12
15
4
4
7
2
2
2
4
2
2
#of
Unreported
Values
17
18
8
8
5
16
16
13
18
18
18
16
18
18
10th%
3.80
95.00
1.00
1.00
0.90
0.00
0.00
100.00
0.00
0.00
8.40
0.00
0.00
0.00
50th %
7.40
95.00
1.00
1.00
3.20
0.00
0.00
100.00
0.00
45.80
54.20
50.00
0.00
400.00
90th %
7.40
95.00
1.00
1.00
3.20
0.00
0.00
100.00
0.00
91.60
100.00
100.00
0.00
800.00
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
Table 3.9.3. Treating Clay from Lube Oil Processing Record Sampling Locations
Sample Number
R13-CL-01
Location
Shell, Deer Park, TX
Description
Pellets from wax treating
3.9.3.5 Source Reduction
This residual is generated infrequently and in very small quantities. Treating clays use
for product polishing in lube oil manufacturing are being phased out by industry. No source
reduction methods were reported by industry or found in the literature search.
Petroleum Refining Industry Study
124
August 1996
-------
Table 3.9.4. Treating Clay from Lube Oil Processing Characterization
Volatile Organics - Method 8260A M9/kg
Benzene
Ethylbenzene
Methylene chloride
n-Propylbenzene
Toluene
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
o-Xylene
m,p-Xylenes
CAS No.
71432
100414
75092
103651
108883
95636
108678
95476
108383/106423
R13-CL-01
11
J 8
24
J 8
31
78
34
18
52
Comments
TCLP Volatile Organics - Methods 1311 and 8260A |jg/L
Methylene chloride
CAS No.
75092
R13-CL-01
B 2,600
Comments
Semivolatile Organics - Method 8270B M9/kg
Bis(2-ethylhexyl)phthalate
Di-n-butyl phthalate
N-Nitrosodiphenylamine
CAS No
117817
84742
86306
R13-CL-01
38,000
J 390
J 470
Comments
TCLP Semivolatile Organics - Methods 1311 and 8270B |jg/L
2-Methylphenol
3/4-Methylphenol
CAS No.
95487
NA
R13-CL-01
J 18
J 18
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kc
Aluminum
Barium
Calcium
Chromium
Copper
Iron
Lead
Manganese
Vanadium
Zinc
CAS No.
7429905
7440393
7440702
7440473
7440508
7439896
7439921
7439965
7440622
7440666
R13-CL-01
140,000
53.0
1,300
100
260
19,000
36.0
180
130
120
Comments
i
Comments
Petroleum Refining Industry Study
125
August 1996
-------
Table 3.9.4. Treating Clay from Lube Oil Processing Characterization (continued)
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Aluminum
Copper
Manganese
Zinc
CAS No.
7429905
7440508
7439965
7440666
R13-CL-01
12.0
0.90
1.50
B 0.94
Comments
Miscellaneous Characterization
Ignitability (oF)
R13-CL-01
NA
Comments
Comments:
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound
that meets the identification criteria for which the result is less than the laboratory detection limit, but
greater than zero.
ND Not Detected.
NA Not Applicable.
Petroleum Refining Industry Study
126
August 1996
-------
3.10 H2S REMOVAL AND SULFUR COMPLEX
3.10.1 Process Description
All crude oil contains sulfur, which must be removed at various points of the refining
process. The predominant technique for treating light petroleum gases is (1) amine scrubbing
followed by (2) recovery of elemental sulfur in a Claus unit followed by (3) final sulfur removal
in a tail gas unit. This dominance is shown in Table 3.10.1, which presents the sulfur
complex/removal processes reported in the RCRA §3007 Survey.
Table 3.10.1. Sulfur Removal Technologies Reported in RCRA §3007 Questionnaire
Technique
Amine-based sulfur removal
Claus sulfur recovery2
Other sulfur removal or recovery
SCOT®-type tail gas unit3
Other tail gas treating unit4
Number of
Facilities
106
101
16
50
19
Percentage of
Facilities1
86
82
13
41
15
Percentage of the 123 facilities reporting any sulfur removal/complex technique.
2Note that more facilities perform sulfur removal than perform sulfur recovery. Some refineries transfer their H2S-
containing amine offsite to another nearby refinery.
3Shell and other companies license similar technologies. All are included here as "SCOT®-type."
414 facilities use the Beavon-Stretford process for tail gas treating.
Caustic or water is often used in conjunction with, or instead of, amine solution to
remove sulfur, particularly for liquid petroleum fractions. These processes, however, are
generally not considered sulfur removal processes because either (1) the sulfur is not further
complexed from these solutions (i.e., is not removed from the solution), or (2) if removed, it
occurs in a sour water stripper which is in the domain of the facility's wastewater treatment
system. Such processes are considered to be liquid treating with caustic, which was discussed in
the Listing Background Document.
The dominant sulfur removal/complex train, amine scrubbing followed by Claus unit
followed by SCOT®-type tail gas treating, is discussed below. In addition, the second-most
popular tail gas system, the Beavon-Stretford system, is discussed. Finally, other processes
reported in the questionnaires are discussed.
Petroleum Refining Industry Study
127
August 1996
-------
3.10.1.1 Amine Scrubbin
As shown in Table 3.10.1, amine scrubbing is used by most facilities, with 106 refineries
reporting this process in the questionnaire. A typical process flow diagram for an amine
scrubbing system is shown in Figure 3.10.1. The purpose of the unit is to remove H2S from
refinery fuel gas for economical downstream recovery. Fuel gas from the refinery is fed to a
countercurrent absorber with a 25 to 30 percent aqueous solution of amine such as
monoethanolamine (MEA), diethanolamine (DEA), or methyldiethanolamine (MDEA). The
H2S reacts with the amine solution to form a complex, "rich" amine. Typically, a refinery will
have several absorbers located throughout the refinery depending on the location of service.
These "rich" streams are combined and sent to a common location at the sulfur plant where the
H2S is stripped from the amine in the reverse reaction. The "lean" amine is recycled back to the
absorbers.
Figure 3.10.1. Amine Sulfur Removal Process Flow Diagram
Sweetened Gas
Lean Amine to otha
units in the i efinery
H^toClaus
unit
3ouuc FuelGas
Ahs rabei
RichAmnine
Heat Stable
Salt Sludge
y EichDEA faom othei
units in the i efenery
Lean Amine
Slip Sbcearn
3.10.1.2 Claus Unit
The H2S from the sulfur removal unit is most often recovered in a Claus system as
elemental sulfur. Table 3.10.1 shows that 101 refineries reported this process in the
questionnaire. A typical process flow diagram for a Claus unit is shown in Figure 3.10.2. In a
Claus unit, the H2S is partially combusted with air to form a mixture of SO2 and H2S. It then
passes through a reactor containing activated alumina catalyst to form sulfur by the following
endothermic reaction:
2 H2S + SO2 --> 3 S + 2 H2O
The reaction is typically conducted at atmospheric pressure. The resulting sulfur is condensed to
its molten state, drained to a storage pit, and reheated. The typical Claus unit consists of three
such reactor/condenser/reheaters to achieve an overall sulfur removal yield of 90 to 95 percent.
Petroleum Refining Industry Study
128
August 1996
-------
At this point the tail gas can be (1) combusted and released to the atmosphere, or (2) sent to a
tail gas unit to achieve greater sulfur reduction.
Figure 3.10.2. Claus Sulfur Recovery Process Flow Diagram
BF\V
HiS
Air
Furnace
i
Steam
OBQ-H+
Elemental
Sulfur
Reactor
J
Spent
Catalyst
1
Steam
Sulfur
Condenser
Tail Gas to
TailGasUni
Off-spec
product sulfur
Product
Sulfur
3.10.1.3 SCOT® Tailgas Unit
The most common type of tail gas unit uses a hydrotreating reactor followed by amine
scrubbing to recover and recycle sulfur, in the form of H2S, to the Claus unit. Shell licenses this
technology as the Shell Claus Offgas Treating (SCOT®) unit; many other refineries reported
using similar designs licensed by other vendors. All can be represented by the generalized
process flow diagram shown in Figure 3.10.3.
Tail gas (containing H2S and SO2) is contacted with H2 and reduced in a hydrotreating
reactor to form H2S and H2O. The catalyst is typically cobalt/molybdenum on alumina. The gas
is then cooled in a water contractor. The water circulates in the column and requires periodic
purging due to impurity buildup; filters may be used to control levels of particulates or
impurities in the circulating water.
The H2S-containing gas enters an amine absorber which is typically in a system
segregated from the other refinery amine systems discussed above. The purpose of segregation
is two-fold: (1) the tail gas treater frequently uses a different amine than the rest of the plant,
such as MDEA or diisopropyl amine (DIPA), and (2) the tail gas is frequently cleaner than the
refinery fuel gas (in regard to contaminants) and segregation of the systems reduces maintenance
requirements for the SCOT® unit. Amines chosen for use in the tail gas system tend to be more
selective for H2S and are not affected by the high levels of CO2 in the offgas.
Petroleum Refining Industry Study
129
August 1996
-------
The "rich" amine generated from this step is desorbed in a stripper; the lean amine is
recirculated while the liberated H2S is sent to the Claus unit. Particulate filters are sometimes
used to remove contaminants from lean amine.
Figure 3.10.3. SCOT® Tail Gas Sulfur Removal Process Flow Diagram
'tent Gas
H2S
Tail Gas
k
H2 t
R&Attor
^
Cf
Spent
Catatyst
OBQ-H3
3.10.1.4 Beavon-Stretford Tail Gas Unit
This system was reported to be used by 14 facilities. A hydrotreating reactor converts
SO2 in the offgas to H2S. The generated H2S is contacted with Stretford solution (a mixture of
vanadium salt, anthraquinone disulfonic acid (ADA), sodium carbonate, and sodium hydroxide)
in a liquid-gas absorber. The H2S reacts stepwise with sodium carbonate and ADA to produce
elemental sulfur, with vanadium serving as a catalyst. The solution proceeds to a tank where
oxygen is added to regenerate the reactants. One or more froth or slurry tanks are used to skim
the product sulfur from the solution, which is recirculated to the absorber.
3.10.1.5 Other Processes
Although the amine/Claus train followed by a SCOT® or Beavon-Stretford tail gas unit is
the dominant system used in the industry, it is not exclusive. Some refineries, mostly small
asphalt plants, do not require sulfur removal processes at all, while others use alternative
technologies. Each of these processes are used by less than five refineries, and most often are
used by only one or two facilities. In decreasing order of usage, these other processes are as
follows:
Petroleum Refining Industry Study
130
August 1996
-------
Sulfur Removal/Recovery Processes
Sodium Hydrosulfide: Fuel gas containing H2S is contacted with sodium hydroxide in
an absorption column. The resulting liquid is product sodium hydrosulfide (NaHS).
Iron Chelate: Fuel gas containing H2S is contacted with iron chelate catalyst dissolved
in solution. H2S is converted to elemental sulfur, which is recovered.
Stretford: Similar to iron chelate, except Stretford solution is used instead of iron
chelate solution.
Ammonium Thiosulfate: In this process, H2S is contacted with air to form SO2. The
SO2 is contacted with ammonia in a series of absorption column to produce ammonium
thiosulfate for offsite sale. (Kirk-Othmer, 1983)
Hyperion: Fuel gas is contacted over a solid catalyst to form elemental sulfur. The
sulfur is collected and sold. The catalyst is comprised of iron and naphthoquinonsulfonic acid.
Sulfatreat: The Sulfatreat material is a black granular solid powder; the H2S forms a
chemical bond with the solid. When the bed reaches capacity, the Sulfatreat solids are removed
and replaced with fresh material. The sulfur is not recovered.
A few facilities report sour water stripping, which was not part of the scope of the
survey. The actual number of sour water strippers is likely to be much greater than reported in
the questionnaire.
Hysulf: This process is under development by Marathon Oil Company and was not
reported by any facilities in the questionnaire. Hydrogen sulfide is contacted with a liquid
quinone in an organic solvent such as n-methyl-2-pyrolidone (NMP), forming sulfur. The sulfur
is removed and the quinone reacted to its original state, producing hydrogen gas (The National
EnvironmentalJournal, March/April 1995).
Tail Gas Processes
Caustic Scrubbing: An incinerator converts trace sulfur compounds in the offgas to
SO2. The gas is contacted with caustic which is sent to the wastewater treatment system.
Polyethylene Glycol: Offgas from the Claus unit is contacted with this solution to
generate an elemental sulfur product. Unlike the Beavon Stretford process, no hydrogenation
reactor is used to convert SO2 to H2S. (Kirk-Othmer, 1983)
Selectox: A hydrogenation reactor converts SO2 in the offgas to H2S. A solid catalyst in
a fixed bed reactor converts the H2S to elemental sulfur. The elemental sulfur is recovered and
sold. (Hydrocarbon Processing, April 1994).
Sulfite/Bisulfite Tail Gas Treating Unit: Following Claus reactors, an incinerator
converts trace sulfur compounds to SO2. The gas is contacted with sulfite solution in an
Petroleum Refining Industry Study 131 August 1996
-------
absorber, where SO2 reacts with the sulfite to produce a bisulfite solution. The gas is then
emitted to the stack. The bisulfite is regenerated and liberated SO2 is sent to the Claus units for
recovery. (Kirk-Othmer, 1983)
3.10.2 Off-Specification Product from Sulfur Complex and H2S Removal Facilities
3.10.2.1 Description
Elemental sulfur is the most common product from sulfur complex and H2S removal
facilities, although a small number of facilities generate product sodium hydrosulfide or
ammonium thiosulfate, as discussed in Section 3.10.1.5. Like other refinery products, sulfur
must meet certain customer specifications such as color and impurity levels. The failure of the
refinery to meet these requirements causes the sulfur to be "off-spec."
Stretford System
Although the Beavon-Stretford system is used by only 14 refineries, off-spec sulfur
generated from this process accounts for 2/3 of the refinery-wide 1992 generation of off-spec
sulfur. Sources of this volume are as follows:
• Product sulfur: Some refineries routinely dispose of their continuously generated
product sulfur rather than sell it. Presumably, these refineries have operational
difficulties making "on-spec" sulfur from the vanadium-catalyzed process. The small
number of refineries managing sulfur this way account for most of the quantity of off-
spec sulfur generated industry-wide. Other refineries sell all or most of their product
sulfur and only dispose of sulfur generated from spills, etc.
• Filtered solids from spent Stretford solution: As discussed further in Section
3.10.3, many refineries report that a portion of the circulating Stretford solution must
be purged to remove impurities in the system. After purging, some refineries filter out
the solids prior to further managing the spent solution.
• Turnaround sludge (sediment): Every few years, the process units are thoroughly
cleaned as preparation for maintenance. The principal source of this turnaround
sludge is the froth (slurry) tank.
• Miscellaneous sludges (sediments): Other solids build up in the system, including
tank sludges and process drain pit sludge. They are removed intermittently.
Every residual generated by the Stretford process contains elemental (product) sulfur
because sulfur is a reaction product. Most refineries designated the above materials as off-spec
product in their questionnaire response, and these residuals are included in statistics discussed
later in this Section.
Petroleum Refining Industry Study 132 August 1996
-------
Claus System
Based on database responses, many Claus units generate off-spec sulfur at frequencies
ranging from 2 months to 2 years. Sources of such sulfur are spills, process upsets, turnarounds,
or maintenance operations. Some refineries generate off-spec sulfur more frequently; one
refinery reports that certain spots are drained daily to ensure proper operation.
Other Systems
The amine scrubbing and SCOT® units do not generate off-spec sulfur because they do
not generate product sulfur (their product is H2S, an intermediate for the Claus sulfur recovery
unit). Other systems generating elemental sulfur or product sulfur compounds can generate off-
spec sulfur for the same reasons described above for Claus and Stretford processes.
3.10.2.2 Generation and Management
Most off-spec sulfur from Claus units is solid with little water content. The off-spec
sulfur residuals described above from the Stretford process contain varying levels of solution
which would give the residual a solid, sludge, or slurry form. Some refineries report filtering
this material to generate off-spec sulfur with higher solids levels.
Based on the questionnaire responses, most refineries (regardless of process) reported
storing off-spec sulfur onsite in a drum, in a dumpster, or in a pile prior to its final destination.
In 1992, five facilities reported classifying this residual as RCRA hazardous (a total quantity of
2,551 MT were reported), however, the hazard waste code was generally not reported.3
Sixty facilities reported generating a total quantity of almost 9,650 MT of this residual in
1992, according to the 1992 RCRA §3007 Survey. As stated in Section 3.10.1, 123 facilities
reported sulfur complex/removal processes. The remaining 63 facilities either report never
generating this residual, or reported generation in years other than 1992 (due to intermittent
generation). There was no reason to expect that 1992 would not be a typical year with regard to
this residual's generation and management. Because most of the generation quantity is
concentrated at a small number of facilities using the Stretford process, however, future
operational changes at those sites could greatly impact the industry-wide residual generation
rate.
Residuals were assigned to be "off-spec sulfur" if they were assigned a residual
identification code of "off-spec product" and were generated from a process identified as a sulfur
removal or complex unit. These correspond to residual code 05 in Section VILA of the
questionnaire and process code 15 in Section IV.C of the questionnaire. Table 3.10.2 provides a
description of the 1992 management practices, quantity generated, number of streams reported,
number of streams not reporting volumes (data requested was unavailable and facilities were not
required to generate it), total and average volumes.
3These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., Subtitle C landfill, transfer to offsite
entity, etc.).
Petroleum Refining Industry Study 133 August 1996
-------
Table 3.10.2. Generation Statistics for Off-Spec Sulfur, 1992
Final Management
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite Subtitle C landfill
Disposal in onsite Subtitle D landfill
Other disposal offsite (anticipated to be
Subtitle C landfill)
Offsite incineration
Offsite land treatment
Other recovery onsite: sulfur plant
Transfer for use as an ingredient in
products placed on the land
Transfer to other offsite entity
Transfer with coke product or other
refinery product
TOTAL
#of
Streams
41
6
3
10
1
1
1
1
1
1
4
70
# of Streams
w/ Unreported
Volume
10
2
0
3
0
0
0
1
0
2
0
21
Total Volume
(MT)
5,043.53
3,575.50
289.07
225.50
0.10
0.70
0.95
2.00
15.00
487.80
6.52
9,646.57
Average
Volume (MT)
123.01
510.79
96.36
22.55
0.10
0.70
0.95
2.00
15.00
487.80
1.63
137.8
3.10.2.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.10.2. No data were
available to the Agency suggesting any other management practices.
3.10.2.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.10.3 summarizes the physical and chemical properties of off-spec sulfur as
reported in Section VILA of the §3007 survey.
• Four record samples of off-spec sulfur were collected and analyzed by EPA. All of
these were collected from the Claus process. Sampling information is summarized in
Table 3.10.4.
The collected samples are expected to be representative of off-spec sulfur generated from
Claus units, the sulfur recovery process used by most refineries. They are not expected to
represent off-spec sulfur from the Stretford process because vanadium would be present in off-
spec sulfur from this process at levels higher than those found in off-spec sulfur from Claus
units. Concentrations of other contaminants may also differ.
Petroleum Refining Industry Study
134
August 1996
-------
All four record samples were analyzed for total and TCLP levels of volatiles,
semivolatiles and metals. None of the samples were found to exhibit a hazardous waste
characteristic. A summary of the analytical results is presented in Table 3.10.5. Only
constituents detected in at least one sample are shown in this table.
3.10.2.5 Source Reduction
During EPA's site visit, one facility was observed to generate "off-spec" sulfur product
daily. Portions of the sulfur plant are being replaced with a newer design. As a result, waste
sulfur residual from equipment "low points" will no longer be generated.
Petroleum Refining Industry Study 135 August 1996
-------
Table 3.10.3. Off-Specification Sulfur: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Vapor Pressure, mm Hg
Vapor Pressure Temperature, °C
Specific Gravity
Specific Gravity Temperature, °C
BTU Content, BTU/lb
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 10-100 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
45
20
35
30
28
12
9
9
35
11
15
46
44
82
28
24
23
14
14
7
#of
Unreported
Values1
62
87
72
77
78
95
98
98
72
96
92
61
63
25
79
83
84
93
93
100
10th%
2.80
0.00
0.00
60.00
0.00
0.00
0.00
20.00
0.80
4.00
0.00
0.00
0.00
60.00
0.00
0.00
0.00
0.00
0.00
0.00
50th %
5.50
0.25
1.23
93.33
0.54
0.00
0.10
140.00
1.36
15.60
4,606.00
0.00
0.00
100.00
80.00
22.50
0.00
0.00
0.00
0.00
90th %
9.00
20.85
92.00
187.78
13.10
1.00
11.00
284.00
2.07
21.10
4,606.00
5.00
100.00
100.00
100.00
100.00
100.00
0.00
0.00
200.00
facilities were not required to do additional testing,
data or engineering judgment.
therefore information provided was based on previously collected
Table 3.10.4. Off-Specification Sulfur Record Sampling Locations
Sample number
R1-SP-01
R2-SP-01
R7B-SP-01
R23-SP-01
Facility
Marathon, Indianapolis, IN
Shell, Wood River, IL
BP, Belle Chase, LA
Chevron, Salt Lake City,
UT
Description
Glaus unit: contents of product tank destined for
disposal
Glaus unit: generated daily from unit "low spots"
Glaus unit: from cleaning and turnaround of
product tank
Glaus unit: from loading spills, connection
leaks, and sumps
Petroleum Refining Industry Study
136
August 1996
-------
Petroleum Refining Industry Study 137
Volatile Organics - Method 8260A tig/kg
Acetone
CAS No.
67641
R1-SP-01
< 25
R2-SP-01
< 25
R7B-SP-01
< 5
R23-SP-01
2,000
Average Cone
514
Maximum Cone
2,000
Comments
TCLP Volatile Organics - Methods 1311 and 8260A \iglL
Acetone
CAS No.
67641
R1-SP-01
B 2,300
R2-SP-01
< 50
R7B-SP-01
< 50
R23-SP-01
B 160
Average Cone
640
Maximum Cone
2,300
Comments
Semivolatile Organics - Method 8270B tig/kg
Bis(2-ethylhexyl) phthalate
Benzo(a)pyrene
Benzo(q,h,i) perylene
Chrysene
Di-n-butyl phthalate
Di-n-octyl phthalate
Pyridine
Fluorene
2-Methylchrysene
1-Methylnaphthalene
2-Methylnaphthalene
Phenanthrene
CAS No.
117817
50328
191242
218019
84742
117840
110861
86737
3351324
90120
91576
85018
R1-SP-01
J 75
< 165
< 165
< 165
< 165
< 165
< 165
< 165
< 330
< 330
< 165
< 165
R2-SP-01
< 165
< 165
< 165
< 165
< 165
< 165
J 160
< 165
< 330
< 330
< 165
< 165
R7B-SP-01
880
< 165
< 165
< 165
J 140
J 180
< 165
J 280
< 330
680
760
J 140
R23-SP-01
460
J 110
J 130
J 270
< 165
< 165
< 165
< 165
J 230
< 330
< 165
< 165
Average Cone
395
110
130
191
140
169
160
194
230
418
314
140
Maximum Cone
880
110
130
270
140
180
160
280
230
680
760
140
Comments
1
1
1
1
1
1
TCLP Semivolatile Organics - Methods 1311 and 8270B ug/L
Bis(2-ethvlhexvl) phthalate
CAS No.
117817
R1-SP-01
< 50
R2-SP-01
J 11
R7B-SP-01
< 50
R23-SP-01
< 50
Average Cone
11
Maximum Cone
11
Comments
1
OQ
(3
-------
Petroleum Refining Industry Study 138
laoie o.iu.3. itesiuuai ^naracierizauon uaia lor uii-apeeiiicauon aunur icommueuj
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg
Aluminum
Barium
Calcium
Chromium
Copper
Iron
Lead
Manganese
Molybdenum
Nickel
Zinc
CAS No.
7429905
7440393
7440702
7440473
7440508
7439896
7439921
7439965
7439987
7440020
7440666
R1-SP-01
< 20
< 20
< 500
2.70
< 2.50
62.0
< 0.30
< 1.50
< 6.50
< 4.00
< 2.00
R2-SP-01
< 20
< 20
< 500
< 1.00
< 2.50
610
0.83
< 1.50
< 6.50
< 4.00
< 2.00
R7B-SP-01
780
90.0
3,400
62.0
68.0
22,000
4.30
91.0
15.0
21.0
140
R23-SP-01
350
< 20
< 500
4.70
8.40
710
3.40
3.20
< 6.50
< 4.00
34.0
Average Cone
293
37.5
1,225
17.6
20.4
5,846
2.21
24.3
8.63
8.25
44.5
Maximum Cone
780
90.0
3,400
62.0
68.0
22,000
4.30
91.0
15.0
21.0
140
Comments
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Aluminum
Calcium
Chromium
Iron
Manganese
Zinc
CAS No.
7429905
7440702
7440473
7439896
7439965
7440666
R1-SP-01
< 1.00
< 25.0
< 0.05
< 0.50
< 0.08
0.31
R2-SP-01
< 1.00
< 25.0
< 0.05
16.0
0.26
< 0.10
R7B-SP-01
5.90
62.0
0.43
44.0
0.77
B 2.90
R23-SP-01
< 1.00
< 25.0
< 0.05
1.50
< 0.08
B 0.87
Average Cone
2.23
34.3
0.15
15.5
0.30
1.05
Maximum Cone
5.90
62.0
0.43
44.0
0.77
2.90
Comments
Comments:
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound that meets the identification criteria for which the result is less than the laboratory detection limit, but greater than
zero.
qg
(3
-------
3.10.3 Off-Specification Treating Solution from Sulfur Complex and H2S Removal
Facilities
3.10.3.1 Description
All treating solutions used in refinery sulfur removal systems are regenerative, meaning
the solution is used over and over in a closed system (for example, amines use multiple
absorption/desorption cycles, while Stretford solution undergoes multiple reversible reactions).
In the following instances the treating solution becomes "off-spec" and cannot be reused:
• Amine systems. At most refineries, amine continuously leaves the closed system
through entrainment in overhead gas, leaks, and other routes. The amine is collected
in various locations such as sumps and either returned to the process or discharged to
the refinery's wastewater treatment (possibly due to purity constraints).
At some refineries, the circulating amine must be replaced in whole or in part due to
contamination or process upset. Rarely, a refinery may change from one amine to
another and completely remove the existing amine from the system prior to
introducing the new solution.
• Stretford systems. Many refineries report that a portion of the circulating Stretford
solution must be purged to remove impurities in the system. After purging, some
refineries filter out the solids prior to further managing the spent solution. Stretford
systems are used at a smaller number (15) of facilities. Unlike amine systems,
Stretford solution is generally used only in tail gas treating.
During operation, the treating solution alternatively becomes "rich" (i.e., containing H2S) and
"lean" (i.e., containing low levels or no H2S). In all observed cases, a refinery will generate off-
spec treating solution when it is "lean."
Approximately 800 MT of off-spec treating solution generated in 1992 was identified by
6 facilities as displaying hazardous characteristics.4 The facilities designated the wastes with
hazardous waste codes D002 (corrosive), D003 (reactive), DO 10 (TC selenium), and DO 18 (TC
benzene). No single hazardous waste code was reported by more than one facility.
3.10.3.2 Generation and Management
Spent Amine Solution
As discussed in Section 3.10.1, the amine sulfur removal process is the dominant sulfur
removal process for gas streams used in the industry. Amine solutions are aqueous and are
typically stored in covered sumps, tanks, etc. In the 1992 questionnaire, most facilities did not
4These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., managed in WWTP, Subtitle C
landfill, transfer for reclamation, etc.).
Petroleum Refining Industry Study 139 August 1996
-------
report how their off-spec treating solution is stored prior to final management; those that did
indicated storage in a tank (most common), storage in a container, or storage in a sump.
Forty-four facilities reported generating a total quantity of 4,627 MT of spent amine in
1992, according to the 1992 RCRA §3007 Questionnaire. Residuals were assigned to be "off-
spec treating solution (spent amine)" if they were assigned a residual identification code of
"treating solution" and were generated from a sulfur complex or H2S removal process. These
correspond to residual codes of "04-B" or "04-C" in Section VILA and process code "15-A" and
"15-D" in Section IV-l.C of the questionnaire. Based on the results of the questionnaire,
approximately 123 facilities employ some type of sulfur removal system (most of these systems
employ treating solution). Many facilities generate this residual on an intermittent basis, or only
during unusual circumstances such as upsets. Therefore, not all of these 123 facilities are
expected to generate off-spec treating solution.
Table 3.10.6 provides a description of the 1992 management practices, quantity
generated, number of streams reported, number of streams not reporting volumes (data requested
was unavailable and facilities were not required to generate it), total and average volumes.
Table 3.10.6. Generation Statistics for Spent Amine for H2S Removal, 1992
Final Management
Discharge to onsite wastewater treatment
facility
Discharge to offsite privately-owned WWT
facility
Disposal in onsite or offsite underground
injection
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite surface impoundment
Neutralization
Onsite boiler
Other recovery onsite: recycle to the
process
Recovery onsite in catalytic cracker
Transfer to other offsite entity/amine
reclaimer
TOTAL
#of
Streams
40
1
4
1
1
3
1
1
3
1
3
59
# of Streams
w/ Unreported
Volume
16
0
0
0
0
0
0
0
4
0
0
20
Total Volume
(MT)
1 ,224.2
152
673.3
200
39
0.8
0.2
9.1
12.8
1,150
166
4,627.4
Average
Volume (MT)
30.6
152
168.3
200
39
0.3
0.2
9.1
4.27
1,150
55.3
78.4
Petroleum Refining Industry Study
140
August 1996
-------
Spent Stretford Solution
The second most frequently used process is the Stretford sulfur removal/complex
process. Stretford solutions are aqueous and are typically stored in covered sumps, tanks, etc.
Twelve facilities reported generating a total quantity of 19,254.5 MT of spent Stretford
solution in 1992, according to the 1992 RCRA §3007 Questionnaire. Residuals were assigned to
be "spent Stretford solution" if they were assigned a residual identification code of "treating
solution" and were generated from a sulfur complex or H2S removal process. These correspond
to residual codes of "04-B" or "04-C" in Section VILA and process code "15-B" and "15-E" in
Section IV-l.C of the questionnaire.
Table 3.10.7 provides a description of the 1992 management practices, quantity
generated, number of streams reported, number of streams not reporting volumes (data requested
was unavailable and facilities were not required to generate it), total and average volumes.
Table 3.10.7. Generation Statistics for Stretford Solution for H2S Removal, 1992
Final Management
Discharge to onsite wastewater treatment
facility
Discharge to offsite privately-owned
WWT facility
Disposal in onsite Subtitle D landfill
Transfer metal catalyst for reclamation or
regeneration
Transfer of acid or caustic for
reclamation, regeneration, or recovery
TOTAL
#of
Streams
4
3
1
2
3
13
# of Streams
w/ Unreported
Volume
2
0
0
0
0
2
Total
Volume (MT)
4,830
6,111.5
711
5,127
2,475
19,254.5
Average
Volume (MT)
1,207.5
2,037.2
711
2563.5
825
1,481
3.10.3.3 Plausible Management
Spent Amine
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.10.6. The Agency
gathered information suggesting other management practices have been used in other years
including: "onsite Subtitle D landfill" (200 MT) and "offsite incineration" (120 MT). These
non-1992 practices are generally comparable to practices reported in 1992 (i.e., off-site Subtitle
D landfilling and on-site boiler, respectively).
Petroleum Refining Industry Study
141
August 1996
-------
Spent Stretford Solution
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.10.7. Even though
spent Stretford solution has different properties, it is possible that the solution could be managed
as the spent amine in Table 3.10.6.
3.10.3.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Tables 3.10.8 and 3.10.9 summarize the physical properties of spent amine and spent
Stretford solution as reported in Section VILA of the §3007 survey.
• Four record samples of spent amine solution were collected and analyzed by EPA.
The sample locations are summarized in Table 3.10.10.
• No samples of spent Stretford solution were available from the randomly selected
facilities during record sampling.
Table 3.10.8. Spent Amine: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Vapor Pressure, mm Hg
Vapor Pressure Temperature, °C
Viscosity, Ib/ft-sec
Specific Gravity
Specific Gravity Temperature, °C
Aqueous Liquid, %
Organic Liquid, %
Solid, %
#of
Values
36
5
10
16
11
16
12
13
10
34
16
61
43
36
#of
Unreported
Values1
67
98
93
87
92
87
91
90
93
69
87
42
60
67
10th%
4.5
0
1.41
-10
0
0
1
15
0
1
15
0
0
0
50th %
9.1
5
280
90.6
0.1
10
30
25
0
1.1
17.5
100
0
0
90th %
11.8
12
7,500
168.9
1
15
300
50
10
1.1
38
100
100
20
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
Petroleum Refining Industry Study
142
August 1996
-------
Table 3.10.9. Spent Stretford Solution: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Oil and Grease, vol%
Total Organic Carbon, vol%
Vapor Pressure, mm Hg
Specific Gravity
COD, mg/L
Aqueous Liquid, %
Organic Liquid, %
Solid, %
#of
Values
10
2
2
1
4
3
8
4
9
3
10
#of
Unreported
Values1
12
19
19
20
17
18
14
17
13
19
12
10th%
8.3
1
0.1
1
0
1.5
1
100
0
0
0.5
50th %
8.8
1.35
3,190
1
0
10
1.1
6,930
90
0
10
90th %
9.7
1.7
6,380
1
1
20
1.5
6,930
100
0
100
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
Table 3.10.10. Off-Specification Treating Solution Record Sampling Locations
Sample Number
R11-SA-01
R13-SA-01
R14-SA-01
R15-SA-01
Facility
ARCO, Ferndale, WA
Shell, Deer Park, TX
BP, Toledo, OH
Total, Ardmore, OK
Description
Refinery DEA system: circulating amine
Refinery DEA system: circulating amine
Refinery DEA system: from sump
collecting knock-out pot liquid, etc, prior to
its exiting the system
Refinery MDEA system: circulating amine
All of the samples were taken from refinery amine systems and are believed to represent
the various types of spent amine generated by refineries. No samples from the tail gas system
units were collected. Tail gas residuals are expected to be cleaner because the feeds are cleaner.
Therefore, the tail gas treating residuals are expected to exhibit levels of contaminants no higher
than those found in the sampled residuals. No samples of Stretford solution were taken.
Stretford systems were not used by the facilities randomly selected by the Agency for record
sampling. Samples of Stretford solution are expected to exhibit higher levels of vanadium than
amine solution because vanadium is present in new Stretford solution; levels of some organic
contaminants may be lower because most refineries use their Stretford system to treat low-
organic Claus unit tail gas.
Several of the samples were taken from the process line (i.e., at the time of sampling, the
refinery had no immediate plans to remove the sampled treating solution from the system).
Petroleum Refining Industry Study
143
August 1996
-------
However, these refineries indicated they do remove all or part of their circulating amine on an
infrequent basis due to process upset or excessive contaminant levels. The sampled amine is
expected to have contaminant concentrations at least as high as when the circulating amine is
removed from the system. Physical properties such as pH and flash point are expected to be
similar as well.
All four samples were analyzed for total and TCLP levels of volatiles, semivolatiles, and
metals, pH, total amines, and ignitability. Two samples were also analyzed for reactive sulfides.
One sample exhibited the characteristic of ignitability. A summary of the results is presented in
Table 3.10.11. Only constituents detected in at least one sample are shown in this table.
3.10.3.5 Source Reduction
Source reduction of amine involves modifying the process. During the site visits,
information was gathered that several facilities capture the amine for recycling. Two facilities
replaced the cloth filter at the sulfur recovery unit with an etched metal mechanical filter. The
new filter requires less maintenance, and also eliminates amine discharges to the wastewater
treatment plant due to filter change-outs. Another two facilities have installed sumps at the
sulfur complex. The sumps capture amine that is drained from the filters during bag change-outs
and recycle it to the amine system. Without the sumps, the amine drained from the filters is
discharged to the wastewater treatment plant.
Reference
Stewart, E. J. and Lanning, R. A. "Reduce Amine Plant
Solvent Losses, Part 2." Hydrocarbon Processing. June,
1994.
"Liquid Catalyst Efficiently Removes H2S From Liquid
Sulfur." Oil & Gas Journal. July 17, 1989.
Stewart, E. J. and Lanning, R. A. "Reduce Amine Plant
Solvent Losses, Part 1." Hydrocarbon Processing. May,
1994.
Waste Minimization/Management Methods
Process modification.
Lower catalyst quantities needed to remove H2S in the
sulfur degassing process.
Process modification.
Petroleum Refining Industry Study
144
August 1996
-------
Table 3.10.11. Characterization Data for Off-
Treating Solution from Sulfur
and H?S Removal
Volatile Organics - Method 8260A ug/L
Acetone
Benzene
Toluene
o-Xylene
m,p-Xylenes
Naphthalene
CAS No.
67641
71432
108883
95476
108383/106423
91203
R11-SA-01
< 25
< 25
< 25
< 25
< 25
< 25
R13-SA-01
< 50
< 50
< 50
< 50
< 50
< 50
R14-SA-01
< 25
88
220
J 24
69
J 32
R15-SA-01
10
< 5
< 5
< 5
< 5
< 5
Average Cone
10
42
75
15
37
19
Maximum Cone
10
88
220
24
69
32
Comments
1
1
1
Semivolatile Organics - Method 8270B ug/L
Acenaphthene
Anthracene
Aniline
Benz(a)anthracene
Bis(2-ethylhexyl)phthalate
Carbazole
Chrysene
Dibenzofuran
2,4-Dimethylphenol
Fluoranthene
Fluorene
2-Methylchrysene
1-Methylnaphthalene
2-Methylnaphthalene
2-Methylphenol
3/4-Methylphenol
Phenanthrene
Phenol
Pyrene
1-Naphthylamine
Naphthalene
CAS No.
83329
120127
62553
56553
117817
86748
218019
132649
105679
206440
86737
3351324
90120
91576
95487
NA
108952
134327
91203
R11-SA-01
< 50
J 18
< 50
< 50
JB 26
J 80
< 50
< 50
110
J 17
< 50
< 100
< 100
< 50
360
1,200
J 50
4,400
J 25
< 50
< 50
R13-SA-01
< 545
< 545
J 540
< 545
< 545
< 1,090
< 545
< 545
< 545
< 545
< 545
< 1,090
< 1,090
< 545
< 545
< 545
< 545
< 545
< 545
< 545
< 545
R14-SA-01
180
250
< 50
J 34
J 17
< 100
J 71
160
J 86
< 50
1,100
J 84
2,500
3,400
210
1,000
3,000
3,100
430
< 50
150
R15-SA-01
< 575
< 575
< 575
< 575
< 575
< 1,150
< 575
< 575
< 575
< 575
< 575
< 1,150
< 1,150
< 575
< 575
< 575
< 575
< 575
< 575
J 230
< 575
Average Cone
115
134
213
34
22
80
61
105
98
17
568
84
1,210
1,143
285
830
1,043
2,155
228
110
100
Maximum Cone
180
250
540
34
26
80
71
160
110
17
1,100
84
2,500
3,400
360
1,200
3,000
4,400
430
230
150
Comments
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
g
OQ
>
I
-------
Table 3.10.11. Characterization Data for Off-Specification Treating Solution from Sulfur Complex and H2S Removal
(continued)
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Aluminum
Antimony
Cadmium
Chromium
Cobalt
Copper
Iron
Manganese
Potassium
Selenium
Sodium
Zinc
CAS No.
7429905
7440360
7440439
7440473
7440484
7440508
7439896
7439965
7440097
7782492
7440235
7440666
R11-SA-01
0.39
0.81
0.035
0.26
0.11
< 0.013
39.0
0.31
21.0
0.031
8.40
< 0.01
R13-SA-01
< 0.10
< 0.03
< 0.003
0.99
< 0.025
< 0.013
14.0
2.30
< 2.50
0.61
< 2.50
< 0.01
R14-SA-01
< 0.10
< 0.03
< 0.003
0.021
< 0.025
0.034
1.10
0.043
< 2.50
0.038
< 2.50
0.039
R15-SA-01
< 0.10
0.62
0.025
0.031
0.099
< 0.013
0.11
< 0.008
22.0
0.99
2,300
< 0.01
Average Cone
0.17
0.37
0.016
0.326
0.065
0.018
13.6
0.67
12.0
0.42
578
0.017
Maximum Cone
0.39
0.81
0.035
0.990
0.110
0.034
39.0
2.30
22.0
0.99
2,300
0.039
Comments
Miscellaneous Characterization
Ignitability (oF)
Corrosivity (pH units)
Reactivity - Total ReleasableH2S (mg/L)
Amines - Methyldiethanolamine (mg/L)
Amines - Ethanolamine (mg/L)
Amines - Diethanolamine (mg/L)
R11-SA-01
> 211
10
< 20
ND
4,400
330,000
R13-SA-01
NA
10
NA
ND
4,500
280,000
R14-SA-01
> 210
8.9
48
ND
ND
41,300
R15-SA-01
90
11.5
NA
36,000
ND
ND
Average Cone
NA
NA
NA
36,000
4,450
217,100
Maximum Cone
NA
NA
NA
36,000
4,500
330,000
Comments
g,
B'
OQ
Comments:
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
TCLP was not performed because these were liquid samples
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound that meets the identification criteria for which the result is less than the laboratory detection limit, but greater than
zero.
ND Not Detected.
NA Not Applicable.
qg
(3
-------
3.11 CLAY FILTERING
Clay belongs to a broad class of materials designed to remove impurities via adsorption.
Examples of clay include Fullers earth, natural clay, and acid treated clay. However, similar
materials such as bauxite are also available and used to impart similar qualities to the product.
In addition, materials such as sand, salt, molecular sieve, and activated carbon are used for
removing impurities by adsorption or other physical mechanisms. All solid materials discussed
in Section 3.11.1 are termed as "solid sorbents" for the purposes of defining this residual
category.
3.11.1 Process Description
Clay or other adsorbents are used to remove impurities from many hydrocarbon streams.
Some of these applications are associated with isomerization, extraction, alkylation, and lube oil
processing; such processes are discussed in the respective sections of this document. Other solid
media remove impurities from amine solutions used in hydrogen sulfide removal systems; such
media were discussed in the Listing Background Document. Solid media used in all other
refinery processes are summarized and discussed in this section. The principal applications are
described below.
Kerosene Clay Filtering: Clay treatment removes diolefins, asphaltic materials, resins,
and acids; this improves the color of the product and removes gum-forming impurities (Speight,
1991). The RCRA §3007 Survey indicates that approximately 90 facilities use this process;
some facilities have multiple treaters or treat different streams, so that an estimated 150
processes exist. Most clay treatment is conducted as a fixed bed. A typical clay volume is 2,000
ft3, distributed in 1 or more vessels. Alternatively to the fixed bed process, the clay can be
mixed with the hydrocarbon and filtered in a belt press. In addition to kerosene, some facilities
identify filtering furnace oils through clay and generating spent clay in a similar manner.
Catalyst Support in Merox and Minalk Systems: The Merox and Minalk caustic
treatment systems convert mercaptans to disulfides using oxygen and an organometallic catalyst
in an alkaline environment. Depending on the process configuration, the disulfides can remain
in the hydrocarbon product (a "sweetening" process) or the disulfides can be removed by settling
(an "extractive" process). These treatment processes are commonly applied to gasoline, but
refinery streams ranging from propane to diesel undergo this treatment.
The catalyst can either be dissolved in the caustic or can be supported on a fixed bed.
Either activated carbon, coal, or charcoal are typically used as support material for solid
supported catalyst (the hydrocarbon passes over the catalyst, where reaction occurs). These
materials provide contact area for reaction when the catalyst is dissolved in the caustic. The
RCRA §3007 survey indicates that approximately 25 facilities (using 40 processes) reported
generating spent carbon, coal, or charcoal from these processes; additional facilities likely
generate this residual but did not report generation in the questionnaire because the residual is
typically generated infrequently.
Drying: Water is removed from many hydrocarbon streams ranging from diesel fuel to
propane. Water must be removed for reasons including: (1) product specifications (e.g., jet fuel
Petroleum Refining Industry Study 147 August 1996
-------
has low tolerances for water content), and (2) reactor feed preparation (e.g., precious metal
catalysts are often poisoned by water). Salt and sand are commonly used for the first
application, while molecular sieve is commonly used for the second application.
When hydrocarbon is passed through a fixed bed of sand, the moisture collects on the
sand particles and eventually settles to the bottom of the vessel, where the water is removed. In
a salt drier, water in the stream dissolves salt (e.g., sodium chloride) which then collects in the
vessel bottom and is periodically removed. As a result, the vessel requires periodic topping with
solid salt.
Salt and sand treaters can be found throughout the refinery to treat hydrocarbons ranging
from diesel to propane. They are commonly found following aqueous treatments such as caustic
washing, water washing, or Merox caustic treatment. In these treatments, the hydrocarbon is
contacted with the aqueous stream; the hydrocarbon then passes through salt or sand to remove
residual moisture. The RCRA §3007 questionnaire indicates that approximately 60 facilities
(using 150 processes) reported generating spent salt or sand from these processes; additional
facilities likely generate this residual but did not report generation in the questionnaire because it
was not generated in 1992.
Molecular sieves are most commonly used to selectively adsorb water and sulfur
compounds from light hydrocarbon fractions such as propane and propylene. The hydrocarbon
passes through a fixed bed of molecular sieve. After the bed is saturated, water is desorbed by
passing heated fuel gas over the bed to release the adsorbed water and sulfur compounds into the
regeneration gas stream, which is commonly sent to a flare stack. Molecular sieves are often
used for drying feed to the isomerization unit and HF acid alkylation unit, applications that are
discussed in Sections 3.4 and 3.5, respectively, of this document. Other applications include
drying propane or propylene prior to entering the Dimersol unit, drying naphtha entering the
reformer, and feed preparation for other reaction units. Molecular sieves are also used to dry
light-end product streams from the hydrocracker, catalytic reformer, and light-ends recovery
unit. Less common uses also exist for molecular sieves including the separation of light-end
fractions such as methanol, butane, and butylene. In total, the RCRA §3007 questionnaire
indicates that approximately 70 facilities (using 150 processes) reported generating spent
molecular sieve; this includes the applications of HF acid alkylation and isomerization that are
discussed elsewhere in this document, but excludes additional facilities that are likely generate
this residual but did not report 1992 generation in the questionnaire.
Sulfur and Chloride Guards in Catalytic Reforming: As discussed in the Listing
Background Document, catalytic reforming units require a platinum catalyst; this catalyst is
readily poisoned by sulfur compounds. To prolong catalyst life, many refineries install sulfur
traps to remove sulfur compounds prior to the reforming catalyst bed. This material can consist
of granular or pelletized metal oxides, such as copper or magnesium. These materials (1)
remove H2S, (2) convert mercaptans to H2S and organic sulfides, and (3) remove generated H2S.
The material can be desorbed, reactivated, and reused (Perry's, 1950). Alumina also is used to
treat light naphtha prior to isomerization (which also uses precious metal catalyst). The RCRA
§3007 questionnaire indicates that approximately 20 facilities reported generating spent sulfur
guards from 35 applications, most often as guards for reforming and isomerization reactors.
Petroleum Refining Industry Study 148 August 1996
-------
Additional facilities may employ sulfur guards but did not report generation in the questionnaire
because the residual is typically generated infrequently.
Alumina beds may be used to remove chlorides from the hydrogen produced from the
reforming process. The hydrogen is then used throughout the refinery. The alumina bed is
expected to last for 24-30 months prior to chloride breakthrough, when replacement of the
alumina is required. Reformate from the reformer may also be passed through alumina to
remove chloride. The RCRA §3007 questionnaire indicates that approximately 15 facilities
reported generating spent chloride guards from 25 applications, most often in the reforming
process.
Propane Treating by Alumina: An activated alumina bed is used to de-fluorinate
propane generated from a propane stripper. The propane then is dried in a sand tower and a
drier which also contains alumina. Both the defluorinator and drier periodically generate spent
alumina.
Particulate Filters: Entrained solids can be removed by in-line cartridge filters. These
cartridges are commonly used for finishing kerosene, diesel fuel, etc., prior to sale.
Approximately 10 facilities reported generating spent cartridges from 20 applications, according
to the questionnaire results.
In most of the applications discussed above, the use of solid media such as clay, sand,
etc. are not the only options refineries have in imparting the desired properties on a product. For
example, drying can be conducted by simple distillation. Hydrotreating and caustic treating are
common alternatives to the clay treatment of jet fuel by removing undesirable contaminants
from the kerosene/jet fuel fraction. And, as discussed above, the Merox process can be
conducted with or without solid supported catalyst.
3.11.2 Treating Clay from Clay Filtering
3.11.2.1 Description
Generated at many places in the refinery, spent solid sorbents have liquid contents
ranging from very low (e.g., for molecular sieves treating light hydrocarbons) to oil-saturated
material (e.g., for clay used for treating kerosene). The substrate is either inorganic (such as
alumina, zeolite, or clay) or organic (such as activated carbon). Most applications are fixed bed,
where the material is charged to vessels and the hydrocarbon passed through the fixed bed of
solid sorption media. The fixed bed can remain in service for a period of time ranging from
several months to 10 years, depending on the application. At the end of service, the vessel is
opened, the "spent" material removed, and the vessel recharged.
Petroleum Refining Industry Study 149 August 1996
-------
3.11.2.2 Generation and Management
The spent clay is vacuumed or gravity dumped from the vessels into piles or into
containers such as drums and roll-off bins. The RCRA §3007 questionnaire and site visits
indicate that very few other interim storage methods are used.
In 1992, approximately 30 facilities reported that 1,700 MT of this residual was managed
as hazardous. The most commonly designated waste codes were D001 (ignitable), D008 (TC
lead), and D018 (TC benzene).5 This is consistent with how the residual was reported to be
managed in other years.
One hundred facilities reported generating a total quantity of approximately 9,000 MT of
this residual in 1992, according to the 1992 RCRA §3007 Questionnaire. There was no reason
to expect that 1992 would not be a typical year with regard to this residual's generation and
management. Residuals were assigned to be "treating clay from clay filtering" if they were
assigned a residual identification code of "spent sorbent" (residual coded "07") and were not
generated from a process identified as an alkylation, isomerization, extraction, sulfur removal, or
lube oil unit (process codes "09," "10," "12," "15," and "17," respectively) (sorbents from these
units are discussed elsewhere in this document or in the Listing Background Document). The
frequency of generation is highly variable as discussed in Section 3.11.1. Table 3.11.1 provides
a description of the 1992 management practices, quantity generated, number of streams reported,
number of streams not reporting volumes (data requested was unavailable and facilities were not
required to generate it), total and average volumes.
The wide array of management methods reflect the numerous applications of sorbents.
For example, disposed salt from salt driers can be managed in onsite wastewater treatment
plants, cement plants can accept spent alumina, and catalyst reclaimers can accept sulfur sorbers
having recoverable metals. The large quantity disposed, however, demonstrates that for most
applications and refineries the spent clay is seen as a low value solid waste.
3.11.2.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.11.1. The Agency
gathered information suggesting other management practices have been used in other years
including: "other recycling, reclamation, or reuse: unknown" (1 MT), "other recycling,
reclamation, or reuse: onsite road material" (13.5 MT) and "reuse as a replacement catalyst for
another unit" (5 MT). These non-1992 very small management practices are comparable to the
recycling practices reported in 1992.
5These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., managed in WWTP, Subtitle C
landfill, transfer as a fuel, etc.).
Petroleum Refining Industry Study 150 August 1996
-------
Table 3.11.1. Generation Statistics for Treating Clay from Clay Filtering, 1992
Final Management
Discharge to onsite wastewater
treatment facility
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite Subtitle C landfill
Disposal in onsite Subtitle D landfill
Evaporation
Offsite incineration
Offsite land treatment
Onsite land treatment
Other disposal onsite:
bioremediation, fill material, or onsite
berms
Other recovery onsite: recycle to
process
Other recycling, reclamation, or
reuse: cement plant
Offsite filter recycling
Storage in pile
Recovery in coker
Transfer for direct use as a fuel or to
make a fuel
Transfer for use as an ingredient in
products placed on the land
Transfer metal catalyst for
reclamation or regeneration
Transfer to other offsite entity/carbon
regeneration
Transfer with coke product or other
refinery product
TOTAL
#of
Streams
14
91
42
1
15
1
7
9
16
5
1
5
2
2
1
1
6
10
2
1
232
# of Streams w/
Unreported Volume
3
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4
Total
Volume (MT)
514
3,642.1
1,735
52.4
1,031.9
7.9
42.1
198.3
923.1
57.4
20.1
161.4
38
128
20
95
175.8
89.4
53.6
4.5
8,990
Average
Volume (MT)
36.7
40
41.3
52.4
68.8
7.9
6
22
57.7
11.5
20.1
32.3
19
64
20
95
29.3
8.9
26.8
4.5
38.8
3.11.2.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.11.2 summarizes the physical properties of the spent clay as reported in
Section VILA of the §3007 survey.
• Four record samples of spent clay were collected and analyzed by EPA. These spent
clays represent some of the various types of applications used by the industry.
Sampling information is summarized in Table 3.11.3.
Petroleum Refining Industry Study
151
August 1996
-------
Table 3.11.2. Treating Clay from Clay Filtering: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Specific Gravity
Specific Gravity Temperature, °C
BTU Content, BTU/lb
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 10-100 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
171
100
106
132
94
50
167
50
31
230
240
346
59
91
70
54
49
48
#of
Unreported
Values1
334
405
399
373
411
455
338
455
474
275
265
159
446
414
435
451
456
457
10th%
4.6
0
0
57.2
0
0
0.7
15
0
0
0
89.0
0
0
0
0
0
0
50th %
7.6
0.5
10
93.3
1
1
1.3
20
2,000
0
0
100
0
100
10
0
0
1,000
90th %
10.4
50
125
200
17.5
55
2.6
25
13,500
10.3
5
100
100
100
100
20
0
3,000
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
Table 3.11.3. Treating Clay Record Sampling Locations
Sample #
R1-CF-01
R6-CF-01
R11-CF-01
R23-CF-01
Facility
Marathon Indianapolis, IN
Shell Norco, LA
ARCO Ferndale, WA
Chevron, Salt Lake City, UT
Description
kerosene/jet treating clay (fixed bed process)
kerosene/jet treating clay (bag filter process,
generated daily)
reformer unit sulfur trap
kerosene/jet treating clay
Petroleum Refining Industry Study
152
August 1996
-------
The collected samples are expected to be representative of treating clay from kerosene
treatment. Section 3.11.1 shows that kerosene clay treatment represents the highest single use of
sorbents in refineries (outside of the sulfur recovery, isomerization, and alkylation processes that
are not included in the scope of this study residual). In addition, a cursory review of the 1992
generation data presented in Section 3.11.2.2 shows that the 1992 generation rate of spent
kerosene treating clay represents at least half of the total 1992 quantity from all sources
identified in Section 3.11.1.
One of the samples is representative of a sulfur guard bed. Other applications of spent
sorbents (discussed in Section 3.11.1) are not well represented by the record sampling.
Specifically:
• Spent activated carbon from Merox treatment, salt and sand from product drying,
particulate filters, and chloride removal beds are not expected to resemble these
materials.
• Spent molecular sieves and alumina are not represented by the collected record
samples. However, they may be represented by the record samples of isomerization
treating clay and alkylation treating clay, discussed in Sections 3.4 and 3.5,
respectively.
All four record samples were analyzed for total and TCLP levels of volatiles,
semivolatiles, and metals. Two samples were analyzed for ignitability and all were analyzed for
reactivity (pyrophoricity). One of the samples was found to exhibit the ignitability
characteristic. High manganese concentrations in one sample result from the adsorbent make-
up. A summary of the results is presented in Table 3.11.4. Only constituents detected in at least
one sample are shown in this table.
3.11.2.5 Source Reduction
One facility reported that its jet fuel treating clay is regenerated once by back-washing
the clay bed with jet fuel to "fluff the clay and alleviate the pressure drop.
Petroleum Refining Industry Study 153 August 1996
-------
Table 3.11.4. Residual Characterization Data for Treating Clay
Volatile Organics - Method 8260A ug/kg
Acetone
Benzene
n-Butylbenzene
sec-Butylbenzene
Ethylbenzene
Isopropylbenzene
p-lsopropyltoluene
n-Propylbenzene
Methylene chloride
Toluene
1,2,4-Trimethylbenzene
1,3,5-Trimethylbenzene
o-Xylene
m,p-Xylenes
Naphthalene
CAS No.
67641
71432
104518
135988
100414
98828
99876
103651
75092
108883
95636
108678
95476
108383/106423
91203
R1-CF-01
260,000
< 125,000
< 125,000
< 125,000
< 125,000
< 125,000
< 125,000
< 125,000
< 125,000
< 125,000
580,000
< 125,000
< 125,000
300,000
310,000
R6-CF-01
< 565
8,500
94,000
54,000
76,000
44,000
59,000
70,000
< 565
140,000
620,000
210,000
180,000
380,000
350,000
R11-CF-01
< 25
540
< 25
< 25
J 28
< 25
< 25
< 25
100
340
< 25
< 25
89
130
< 25
R23-CF-01
< 1,250
< 1,250
< 1,250
< 1,250
2,800
< 1,250
< 1,250
< 1,250
< 1,250
3,600
32,000
13,000
7,200
23,000
9,800
Average Cone
65,460
3,430
31,758
18,425
26,276
15,092
20,092
23,758
100
67,235
308,006
87,006
78,072
175,783
167,456
Maximum Cone
260,000
8,500
94,000
54,000
76,000
44,000
59,000
70,000
100
140,000
620,000
210,000
180,000
380,000
350,000
Comments
1
1
1
1
1
1
1
1
TCLP Volatile Organics - Methods 1311 and 8260A ug/L
Acetone
Benzene
Ethylbenzene
Methylene chloride
Toluene
1,2,4-Trimethylbenzene
1,3,5-Trimethylbenzene
o-Xylene
m,p-Xylene
Naphthalene
CAS No.
67641
71432
100414
75092
108883
95636
108678
95476
108383/106423
91203
R1-CF-01
43,000
< 1,250
< 1,250
2,600
< 1,250
4,900
< 1,250
< 1,250
< 1,250
< 1,250
R6-CF-01
< 50
100
190
< 50
850
840
270
610
1,200
650
R11-CF-01
< 50
J 44
< 50
1,700
210
< 50
< 50
< 50
< 50
< 50
R23-CF-01
B 100
< 50
< 50
< 50
< 50
J 62
< 50
J 44
110
J 71
Average Cone
10,800
65
97
1,100
370
1,463
123
235
453
257
Maximum Cone
43,000
100
190
2,600
850
4,900
270
610
1,200
650
Comments
1
1
1
1
1
1
1
Semivolatile Organics - Method 8270B ug/kg
Bis(2-ethylhexyl) phthalate
Carbazole
Di-n-butyl phthalate
Dibenzofuran
Fluorene
2.4-Dimethvlphenol
CAS No.
117817
86748
57976
132649
86737
105679
R1-CF-01
< 6,600
< 13,200
< 6,600
< 6,600
< 6,600
< 6,600
R6-CF-01
< 4,125
< 8,250
< 4,125
J 24,000
< 4,125
< 4,125
R11-CF-01
J 100
< 330
420
< 165
< 165
2,500
R23-CF-01
< 4,150
J 6,000
< 4,150
< 4,150
20,000
< 4,150
Average Cone
100
3,165
420
8,729
7,723
2,500
Maximum Cone
100
6,000
420
24,000
20,000
2,500
Comments
1
1
1
1
g
OQ
C
qg
s3
-------
Table
Residual Characterization Data for Treating
Semivolatile Organics - Method 8270B ug/kg (continued)
2-Methylphenol
3/4-Methylphenol
1-Methylnaphthalene
2-Methylnaphthalene
Naphthalene
Phenanthrene
Phenol
CAS No.
95487
NA
90120
91576
91203
85018
108952
R1-CF-01
< 6,600
< 6,600
980,000
150,000
120,000
< 6,600
< 6,600
R6-CF-01
< 4,125
< 4,125
890,000
1,200,000
740,000
J 4,800
< 4,125
R11-CF-01
9,000
30,000
< 165
< 165
< 165
< 165
20,000
R23-CF-01
< 4,150
< 4,150
78,000
92,000
43,000
25,000
< 4,150
Average Cone
5,969
11,219
487,041
360,541
225,791
9,141
8,719
Maximum Cone
9,000
30,000
980,000
1,200,000
740,000
25,000
20,000
Comments
TCLP Semivolatile Organics - Methods 1311 and 8270B ug/L
Bis(2-ethylhexyl) phthalate
Dibenzofuran
Di-n-butyl phthalate
2,4-Dimethylphenol
Fluorene
1-Methylnaphthalene
2-Methylnaphthalene
Naphthalene
2-Methylphenol
3/4-Methylphenol (total)
Phenol
CAS No.
117817
132649
84742
105679
86737
90120
91576
91203
95487
NA
108952
R1-CF-01
290
< 50
< 50
350
< 50
J 190
220
600
310
580
< 50
R6-CF-01
J 16
J 17
JB 19
J 73
J 41
550
780
700
< 50
< 50
< 50
R11-CF-01
< 250
< 250
< 250
1,400
< 250
< 250
< 500
< 250
7,800
6,300
2,300
R23-CF-01
< 50
< 50
< 50
< 50
< 50
J 130
120
140
< 50
< 50
< 50
Average Cone
152
17
19
468
41
280
405
423
2,053
1,745
613
Maximum Cone
290
17
19
1,400
41
550
780
700
7,800
6,300
2,300
Comments
1
1
1
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg
Aluminum
Arsenic
Barium
Beryllium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
CAS No.
7429905
7440382
7440393
7440417
7440702
7440473
7440484
7440508
7439896
7439921
7439954
7439965
7439976
R1-CF-01
12,000
3.20
78.0
3.80
4,500
37.0
12.0
< 2.50
9,400
4.80
9,400
130
< 0.05
R6-CF-01
6,800
< 1.00
< 20.0
< 0.50
16,000
24.0
< 5.00
< 2.50
3,800
1.90
10,000
140
< 0.05
R11-CF-01
110,000
14.0
< 20.0
< 0.50
< 500
34.0
34.0
5.30
97.0
2.70
< 500
150,000
< 0.05
R23-CF-01
13,000
16.0
59.0
2.50
4,400
39.0
11.0
620
9,800
6.00
9,300
120
0.26
Average Cone
35,450
8.55
44.3
1.83
6,350
33.5
15.5
158
5,774
3.85
7,300
37,598
0.10
Maximum Cone
110,000
16.0
78.0
3.80
16,000
39.0
34.0
620
9,800
6.00
10,000
150,000
0.26
Comments
g
OQ
-------
Petroleum Refining Industry Study 1 56
i auie j.Li.t. rvesiuuai ^ iiaraneri/;aiiuii uaia lur i reauiig ^ la y ^uiiimueuj
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg (continued)
Molybdenum
Nickel
Potassium
Selenium
Silver
Sodium
Vanadium
Zinc
CAS No.
7439987
7440020
7440097
7782492
7440224
7440235
7440622
7440666
R1-CF-01
< 6.50
16.0
1,400
< 0.50
< 1.00
34,000
37.0
47.0
R6-CF-01
< 6.50
< 4.00
< 500
< 0.50
< 1.00
< 500
21.0
19.0
R11-CF-01
14.0
< 4.00
< 500
22.0
70.0
< 500
34.0
< 2.00
R23-CF-01
< 6.50
31.0
1,300
< 0.50
< 1.00
< 500
35.0
55.0
Average Cone
8.38
13.8
925
5.88
18.3
8,875
31.8
30.8
Maximum Cone
14.0
31.0
1,400
22.0
70.0
34,000
37.0
55.0
Comments
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Aluminum
Arsenic
Calcium
Copper
Iron
Magnesium
Manganese
Silver
Zinc
CAS No.
7429905
7440382
7440702
7440508
7439896
7439954
7439965
7440224
7440666
R1-CF-01
< 1.00
< 0.05
54
< 0.13
< 0.50
< 25.0
< 0.08
< 0.05
< 0.10
R6-CF-01
< 1.00
< 0.05
590
< 0.13
< 0.50
91
2.60
< 0.05
B 0.76
R11-CF-01
< 1.00
< 0.05
< 25.0
< 0.13
< 0.50
< 25.0
1,400
0.10
< 0.10
R23-CF-01
3.90
0.13
60.0
0.89
1.00
< 25.0
0.85
< 0.05
B 0.27
Average Cone
1.73
0.07
182
0.32
0.63
41.5
351
0.06
0.31
Maximum Cone
3.90
0.13
590
0.89
1.00
91.0
1,400
0.10
0.76
Comments
Miscellaneous Characterization
Ignitability (oF)
R1-CF-01
185
R6-CF-01
131
R11-CF-01
NA
R23-CF-01
NA
Average Cone
NA
Maximum Cone
NA
Comments
Comments:
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound that meets the identification criteria for which the result is less than the laboratory detection limit, but greater than
zero.
ND Not Detected.
NA Not Applicable.
OQ
(3
-------
3.12 RESIDUAL OIL TANK STORAGE
Almost every refinery stores its feed and products in tanks onsite. Occasionally (every
10 to 20 years), tanks require sediment removal due to maintenance, inspection, or sediment
buildup. These tank bottoms are removed by techniques ranging from manual shoveling to
robotics and filtration. Residual oil tank sludge is a study residual of concern.
Residual oil is generally considered to be equivalent to No. 6 fuel oil which is a heavy
residue oil sometimes called Bunker C when used to fuel ocean-going vessels. Preheating is
required for both handling and burning. It is typically produced from units such as atmospheric
and vacuum distillation, hydrocracking, delayed coking, and visbreaking. The fluid catalytic
cracking unit also contributes to the refinery's heavy oil pool, but EPA terms this material
"clarified slurry oil," or CSO, and discussed this product separately in the Listing Background
Document (October 31, 1995).
According to DOE's Petroleum Supply Annual, approximately 400 million barrels of
"residual oil" was domestically used in 1992 (including imports and exports). The use profile in
1994 was as follows (DOE's Fuel Oil and Kerosene Sales 1994}:
Sector 1990 Consumption of Residual Fuel Oil
Electric Utility 40%
Shipping 35%
Industrial 15%
Commercial and Other 10%
The larger utilities often have their own specifications when purchasing residual fuel oil. These
can include sulfur, nitrogen, ash, and vanadium. The current ASTM standard for No. 6 oil (D-
396) specifies only three parameters: minimum flash point (of 150°F), maximum water and
sediment (of 2 percent), and a viscosity range (Bonnet, 1994). Thus, the characteristics of
residual oil, and the generated tank sludge, can vary greatly depending on the buyer and the
refinery.
3.12.1 Residual Oil Storage Tank Sludge
In 1992, 125 U.S. refineries reported approximately 717 residual oil storage tanks. From
the survey, tank volume was reported for about 10 percent (73) of these tanks (excluding
outliers); the average tank volume was approximately 77,000 barrels. DOE's Petroleum Supply
Annual 1992 reported that refineries produced about 327 million barrels of No. 6 fuel oil or
residual oil or approximately 900,000 barrels per day (this likely includes CSO).
3.12.1.1 Description
Residual oil tank sludge consists of heavy hydrocarbons, rust and scale from process
pipes and reactors, and entrapped oil that settles to the bottom of the tank. It can be manually re-
moved directly from the tank after drainage of the residual oil or, commonly, removed using a
variety of oil recovery techniques. The recovered oil is returned generally to slop oil storage
while the remaining solids are collected and discarded as waste.
Petroleum Refining Industry Study 157 August 1996
-------
Once a tank is taken out of service, many refineries use in situ and ex situ oil recovery
techniques. Common in situ oil recovery techniques include hot distillate washing, and steam
stripping. This allows entrapped oil to float to the top of the sediment layer and be recovered
prior to removal of the sediment from the tank. Ex situ recovery methods are usually performed
by a contractor at the tank site and include filtration, centrifuging, and settling. Separated oil is
recycled back to the process or sent to the slop oil tanks, and the water phase is sent to the
wastewater treatment plant (WWTP). The solids are managed in a variety of ways, but
primarily are disposed of in Subtitle C and D landfills (78 percent in 1992).
Many refineries reduce tank bottom buildup with in-tank mixers. Mixers keep the
sediments or solids continuously in suspension so that they travel with the residual oil.
In 1992, less than one percent of the volume of residual oil tank bottom sludge was
reported to be managed as hazardous.6 Of the few refineries that reported a hazardous waste
designation for this residual in 1992, only one reported a hazardous waste code (the others
specified handling the sludge as hazardous without designating a code).
3.12.1.2 Generation and Management
The refineries reported generating 9,107 MT of residual oil tank bottom sludge in 1992.
Residual oil tank sludge includes sludges from No. 6 oil and similar product tanks. Sludges
from tanks identified as containing a mixture of residual oil and clarified slurry oil were
included in the scope of K170 and are omitted here. Residuals were assigned to be "residual oil
tank sludge" if they were assigned a residual identification code of "residual oil tank sediment,"
corresponding to residual code "01-B" in Section VII. 1 of the questionnaire. Process
wastewaters, decantates, and recovered oils (e.g., from deoiling or dewatering operations) were
eliminated from the analysis. These correspond to residual codes "09," "10," and "13" (newly
added "recovered oil") in the questionnaire. Quality assurance was conducted by ensuring that
all residual oil tank sludges previously identified in the questionnaire (i.e., in Section V.D) were
assigned in Section VII. 1. Table 3.12.1 provides a description of the 1992 management
practices, quantity generated, number of streams reported, number of streams not reporting
volumes, and average volumes.
When cleaning a tank, it is common for refineries to use some type of in situ treatment,
such as washing with lighter fuel, to recover oil from the top layers of sludge where there is a
high percentage of free oil. However, treatment or recovery practices after this depend on the
refinery's planned final management method. If land disposed (as most residual oil tank sludge
was in 1992), low free liquid must be achieved; such levels can be achieved by sludge
deoiling/dewatering or stabilization. A refinery may conduct this
"These percentages do not match up directly with any one of the management scenarios because the number of
streams and the volume are a combination of several management scenarios (i.e., managed in WWTP, Subtitle C
landfill, recovery onsite in coker, etc.).
Petroleum Refining Industry Study 158 August 1996
-------
Table 3.12.1. Generation Statistics for Residual Oil Tank Sludge, 1992
Final Management
Discharge to onsite wastewater
treatment facility
Disposal in offsite Subtitle D landfill
Disposal in offsite Subtitle C landfill
Disposal in onsite Subtitle C landfill
Disposal in onsite Subtitle D landfill
Disposal in onsite surface impoundment
Offsite land treatment
Onsite land treatment
Other recycling, reclamation, or reuse:
cover for onsite landfill
Recovery onsite via distillation
Transfer for use as an ingredient in
products placed on the land
Transfer to another petroleum refinery
TOTAL
#of
Streams
1
13
8
2
3
1
1
2
1
1
1
1
35
# of Streams
w/ Unreported
Volume
0
4
0
0
0
0
1
0
0
3
0
0
8
Total Volume
(MT)
47
6,458
622
4
30.4
132
4
530.4
7.2
310
35
927
9,107
Average
Volume (MT)
47
496.8
77.8
2
10.1
132
4
265.2
7.2
310
35
927
260.2
treatment for only some of the waste (e.g., the top layers); in the deeper sections of sludge where
free liquid levels are lower no treatment may be performed. In addition to lower liquid levels,
treatment or deoiling may be used to achieve lower levels of benzene or other hazardous
properties.
3.12.1.3 Plausible Management
EPA believes that most of the plausible management practices for this residual were
reported in the 1992 RCRA §3007 Survey, as summarized above in Table 3.12.1. The Agency
gathered information suggesting other management practices have been used in other years
including: "recovery onsite in an asphalt production unit" (9.2 MT), "transfer for direct use as a
fuel or to make a fuel" (380.8 MT), "transfer with coke product or other refinery product" (5
MT), "onsite industrial furnace" (39 MT), "recycle to process" (unknown quantity), "recovery in
coker" (unknown quantity), and "recovery in a catalytic cracker" (unknown quantity). These
non-1992 management practices are generally comparable to the recycling practices reported in
1992.
Petroleum Refining Industry Study
159
August 1996
-------
3.12.1.4 Characterization
Two sources of residual characterization were developed during the industry study:
• Table 3.12.2 summarizes the physical properties of residual oil tank sludges as
reported in Section VILA of the §3007 survey.
• Two record samples of actual residual oil sludge were collected and analyzed by EPA.
These sludges represent the various types of treatment typically used by the industry
and are summarized in Table 3.12.3.
Table 3.12.4 provides a summary of the characterization data collected under this
sampling effort. The record samples collected are believed to be representative of residual oil
tank sludges generated by the industry.
The samples collected of the composite of oily and de-oiled sediment are representative
of industry treatment practices. As reported in the RCRA 3007 questionnaires, 10 of the 34
residual oil tank sludges (30 percent) that were ultimately managed in a land treatment or landfill
in 1992 were deoiled in some manner, most often by filtration or centrifuge. This management
resulted in volume reduction averaging 55 percent. Another 7 (20 percent) were stabilized,
resulting in the volume increasing by an average of 55 percent. The remaining 17 residuals (50
percent) were not reported to be treated ex situ in any manner. The sampled refineries represent
two alternative interim management procedures: free liquid reduction using stabilization
(Amoco), and ex situ deoiling (Star). Therefore, the record samples represent the various types
of ex situ treatment typically performed for residual oil tank sludge, but may not represent cases
in which no treatment is performed. However, the same contaminants will be present in all three
types of sludge (i.e., deoiled. stabilized, and untreated), but their levels may differ.
As illustrated in Table 3.12.4, none of the record samples exhibited a hazardous waste
characteristic. Only constituents detected in at least one sample are shown in this table.
3.12.1.5 Source Reduction
Only a small quantity of sludge was reported to be deoiled in 1992, as reported in the
§3007 survey. Of the 34 residuals disposed in landfills or land treatment units in 1992, 10
residuals, totaling approximately 1,000 MT. The remaining 24 residuals, totaling approximately
7,600 MT, were reported to be untreated or underwent volume addition treatment (such as
stabilization. As stated in Section 3.12.1.3, the average volume reduction achieved by deoiling
was 55 percent (as calculated from those facilities providing sludge quantities prior to and
following deoiling in 1992).
Petroleum Refining Industry Study 160 August 1996
-------
Table 3.12.2. Residual Oil Tank Sludge: Physical Properties
Properties
PH
Reactive CN, ppm
Reactive S, ppm
Flash Point, °C
Oil and Grease, vol%
Total Organic Carbon, vol%
Vapor Pressure, mm Hg
Vapor Pressure Temperature, °C
Viscosity, Ib/ft-sec
Specific Gravity
BTU Content, BTU/lb
Aqueous Liquid, %
Organic Liquid, %
Solid, %
Other, %
Particle >60 mm, %
Particle 1-60 mm, %
Particle 100 um-1 mm, %
Particle 10-100 urn, %
Particle <10 urn, %
Median Particle Diameter, microns
#of
Values
39
27
27
42
36
20
11
9
6
30
16
78
82
91
65
4
6
5
4
4
3
# of Unreported
Values1
87
99
99
84
90
106
115
117
120
96
110
48
44
35
61
122
120
121
122
122
123
10th%
5.5
0
0
60
9
3.5
0
25
0.01
0.9
600
0
0
1
0
0
0
0
0
0
0
Mean
7
0.3
2.5
93.3
34.1
51
0.1
37.8
50.2
1.2
5,000
0
18
60
0
0
50
50
0
0
0
90th %
8.5
5
15
140
99
85.3
10
38
500
2.4
20,000
50
86
100
0
0
100
100
1
0
15,000
facilities were not required to do additional testing, therefore information provided was based on previously collected
data or engineering judgment.
Table 3.12.3. Residual Oil Tank Sludge Record Sampling Locations
Sample No.
R8B-RS-01
R22-RS-01
Facility
Amoco, Texas City, TX
Star, Port Arthur, TX
Description:
Residual oil and CSO mixed.1 Cleaning
procedure: pumped down, mixed with
diatomaceous earth, removed with backhoe.
Residual oil.2 Cleaning procedure: washed
with lighter oil, centrifuged to generate cake.
1The refinery has a fluid catalytic cracking unit and generates CSO. An unknown quantity of CSO was stored in the
sampled tank.
2The refinery has a fluid catalytic cracking unit and generates CSO. It is unknown if, or to what extent, CSO was
stored in the sampled tank.
Petroleum Refining Industry Study
161
August 1996
-------
Table 3.12.4. Residual Oil Tank Sludge Characterization
Volatile Organics - Method 8260A tig/kg
n-Butylbenzene
Ethylbenzene
p-lsopropyltoluene
n-Propylbenzene
Toluene
1,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
o-Xylene
m,p-Xylenes
Naphthalene
CAS No.
104518
100414
99876
103651
108883
95636
108678
95476
108383/106423
91203
R8B-RS-01
< 6,250
13,000
< 6,250
J 6,850
26,000
43,000
J 11,000
19,000
51,000
64,000
R22-RS-01
3,600
J 1,600
J 470
J 1,600
< 1,250
18,000
4,200
J 1,800
7,400
19,000
Average Cone
3,600
7,300
470
4,225
13,625
30,500
7,600
10,400
29,200
41,500
Maximum Cone
3,600
13,000
470
6,850
26,000
43,000
11,000
19,000
51,000
64,000
Comments
1
1
TCLP Volatile Organics - Methods 1311 and 8260A ug/L
Benzene
Ethylbenzene
Toluene
1,2,4-Trimethylbenzene
Methylene chloride
o-Xylene
m,p-Xylene
Naphthalene
CAS No.
71432
100414
108883
95636
75092
95476
108383/106423
91203
R8B-RS-01
110
J 55
690
J 79
B 1,200
J 96
220
J 91
R22-RS-01
< 50
< 50
< 50
< 50
< 50
< 50
JB 28
J 46
Average Cone
80
53
370
65
625
73
124
69
Maximum Cone
110
55
690
79
1,200
96
220
91
Comments
Semivolatile Organics - Method 8270B tig/kg
Acenaphthene
Anthracene
Benz(a)anthracene
Benzofluoranthene (total)
Benzo(g,h,i)perylene
Benzo(a)pyrene
Bis(2-ethylhexyl)phthalate
Carbazole
Chrysene
Dibenzofuran
Dibenz(a,h)anthracene
3,3'-Dichlorobenzidine
Fluoranthene
Fluorene
lndeno(1 ,2,3-cd)pyrene
Phenanthrene
Pyrene
1-Methylnaphthalene
2-Methylnaphthalene
2-Methylchrysene
Naphthalene
CAS No
83329
120127
56553
NA
191242
50328
117817
86748
218019
132649
53703
91941
206440
86737
193395
85018
129000
90120
91576
3351324
91203
R8B-RS-01
60,000
150,000
480,000
130,000
450,000
250,000
< 10,313
< 20,625
800,000
25,000
65,000
< 10,313
120,000
160,000
58,000
1,000,000
3,500,000
500,000
650,000
380,000
230,000
R22-RS-01
27,000
< 4,125
9,200
34,000
36,000
87,000
10,000
J 16,000
170,000
8,700
J 8,000
87,000
< 4,125
38,000
< 4,125
220,000
46,000
250,000
410,000
< 8,250
110,000
Average Cone
43,500
77,063
244,600
82,000
243,000
168,500
10,000
16,000
485,000
16,850
36,500
48,656
62,063
99,000
31,063
610,000
1,773,000
375,000
530,000
194,125
170,000
Maximum Cone
60,000
150,000
480,000
130,000
450,000
250,000
10,000
16,000
800,000
25,000
65,000
87,000
120,000
160,000
58,000
1,000,000
3,500,000
500,000
650,000
380,000
230,000
Comments
1
1
Petroleum Refining Industry Study
162
August 1996
-------
Table 3.12.4. Residual Oil Tank Sludge Characterization (continued)
TCLP Semivolatile Organics - Methods 1311 and 8270B \iglL
Di-n-butylphthalate
1-Methylnaphthalene
2-Methylnaphthalene
Naphthalene
CAS No.
84742
90120
91576
91203
R8B-RS-01
< 50
J 28
J 37
J 37
R22-RS-01
JB 24
J 54
J 74
J 73
Average Cone
24
41
56
55
Maximum Cone
24
54
74
73
Comments
1
Total Metals - Methods 6010, 7060, 7421, 7470, 7471, and 7841 mg/kg
Aluminum
Arsenic
Barium
Beryllium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Sodium
Vanadium
Zinc
CAS No.
7429905
7440382
7440393
7440417
7440702
7440473
7440484
7440508
7439896
7439921
7439954
7439965
7439976
7439987
7440020
7440235
7440622
7440666
R8B-RS-01
9,100
3.00
< 20.0
1.80
< 500
11.0
130
7.40
1,600
6.50
< 500
12.0
1.50
330
410
< 500
1,400
75.0
R22-RS-01
38,000
< 1.00
230
< 0.50
1,400
31.0
< 5.00
110
11,000
84.0
4,300
67.0
< 0.05
18.0
83.0
3,200
480
200
Average Cone
23,550
2.00
125
1.15
950
21.0
67.5
58.7
6,300
45.3
2,400
39.5
0.78
174
247
1,850
940
138
Maximum Cone
38,000
3.00
230
1.80
1,400
31.0
130
110
11,000
84.0
4,300
67.0
1.50
330
410
3,200
1,400
200
Comments
TCLP Metals - Methods 1311, 6010, 7060, 7421, 7470, 7471, and 7841 mg/L
Aluminum
Iron
Manganese
Zinc
CAS No.
7429905
7439896
7439965
7440666
R8B-RS-01
< 1.00
< 0.50
< 0.08
B 0.26
R22-RS-01
3.70
10.0
1.10
1.20
Average Cone
2.35
5.25
0.59
0.73
Maximum Cone
3.70
10.0
1.10
1.20
Comments
Comments:
1 Detection limits greater than the highest detected concentration are excluded from the calculations.
Notes:
B Analyte also detected in the associated method blank.
J Compound's concentration is estimated. Mass spectral data indicate the presence of a compound that meets the identification criteria for which the
result is less than the laboratory detection limit, but greater than zero.
ND Not Detected.
NA Not Applicable.
Petroleum Refining Industry Study
163
August 1996
-------
In situ oil recovery techniques can greatly reduce the total amount of residual oil tank
sludge to be disposed as well as reduce volatile constituents such as benzene. As discussed
above, recovery methods include distillate washing, nonpetroleum solvent washing, water wash
with surfactant, and steam stripping. These operations allow entrapped oil to float to the top of
the sediment layer and be recovered prior to removal from the tank. Separated oil is recycled
back to the process or sent to the slop oil tanks, and the water phase is sent to the WWTP.
Oily sludges are emulsions formed due to a surface attraction among oily droplets, water
droplets, and solid particles. If the solids are large and dense, the resultant material will settle
and become a sludge. The surface charge interactions between the solid particles and oil
droplets cause the sludge to become stable and difficult to separate. However, the sludge can be
separated into its individual components by mechanically removing the solids or by neutralizing
the surface charge on the solids and oil droplets.
The predominant method of minimizing the formation of tank sludge is the use of mixers
to keep the sludges continuously in suspension. A common mixer configuration is a sweeping
mixer that automatically oscillates to produce a sweeping motion over the floor of the tank,
keeping the heavy oil and particles suspended.
Of the twenty facilities that EPA visited, eight listed methods in recovering oil from tank
sludges. Several facilities wash the tanks with light oils and water, whereas another facility
washes with a surfactant followed by pressure filtration.
Reference
"Re-refiner Fluidizes Tank Residue Using Portable Mixer."
Oil & Gas Journal. September 5, 1994.
Kuriakose, A.P., Manjooran, S. JochuBaby.
"Utilization of Refinery Sludge for Lighter Oils and Industrial
Bitumen." Energy & Fuels, vol.8, no. 3. May-June, 1994.
"Environmental Processes '93: Challenge in the '90s."
Hydrocarbon Processing. August, 1993.
"Waste Minimization in the Petroleum Industry : A
Compendium of Practices." API. November, 1991.
Waste Minimization/Management Methods
A portable mixer was used to cut lighter oil into the
partially gelled residue.
Utilizing waste sludge.
A variety of technologies described, such as
bioslurry treatment of oily wastes, oily -waste
recovery, and evaporation/solvent extraction.
Sludge formation can be minimized by mixing
contents of tank.
Petroleum Refining Industry Study
164
August 1996
-------
BIBLIOGRAPHY
1992 RCRA §3007 Survey of the Petroleum Refining Industry Database.
Department of Energy, Energy Information Administration. Petroleum Supply Annual 1992,
Volume 1. May 1993.
Donald Bonett, "ASTM D-396 Specification for No. 6 Fuel Oil," in Proceedings, 1993 Fuel Oil
Utilization Workshop, Electric Power Research Institute, August 1994 (page 3-101).
Fuel Oil and Kerosene Sales 1994, U.S. Department of Energy, September 1995 (DOE/EIA-
0340(92)71).
Hydrocarbon Processing. "Refining Processes'94." November 1994.
Hydrocarbon Processing. "Gas Processing'94." April 1994.
Kirk-Othmer. Encyclopedia of Chemical Technology. Third Edition, Volume 22. 1983.
McKetta, John J. Petroleum Processing Handbook. Marcel Dekker, Inc. 1992.
Meyers, Robert A. Handbook of Petroleum Refining Processes. McGraw-Hill Book Company.
1986.
Perry's, 1950. John H. Perry, ed. Chemical Engineer's Handbook. McGraw-Hill, New York.
Third edition, 1950.
Speight, 1991. James Speight. The Chemistry and Technology of Petroleum. Marcel Dekker,
New York. Second edition, 1991.
Petroleum Refining Industry Study 165 August 1996
------- |