PB83-164749
Environmental Assessment Data Base for
Petroleum Refining Wastewa>ers and Residuals
Tulsa Univ., OK
Prepared for
Robert S. Kerr Environmental Research Lab.
Ada, OK
Feb 83
I
J
taputincflt 01 Commerce
ion Service
-------
EPA-600/2-83-C10
February 1983
ENVIRONMENTAL ASSESSMENT D;€>A BASE FOR PETROLEUM.
REFINING WASTEWATERS AND RESIDUALS
i
.... . ._,j by '
Francis S. Manning
Eric H. Snider
Department of Chemical Engineering
The University of Tulsa
Tulsa, Oklahoma 74104
EPA Grant No. R 805099
Project Officer
Fred M. Pfeff*r
Source Management Branch
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ADA, OKLAHOMA 74820
-------
TECHNICAL REPORT DATA
IPItosi rrad Inuncnons on the ttvene be/on complerirtft
REPORT NO.
EPA-600/2-83-010
3. RECIPIENT'S ACCESSION-NO.
PBE
1 6 ^ 7 A 9
4. TITLE AMD SUBTITLE '
Environmental Assessment Data Base for Petroleum
Refining Wastewaters and Residuals"
5. REPORT DATE
February 1983
6. PERFORMING ORGANIZATION CODE
7 AUTMORISI
Francis S. Manning and Eric rl. Snider
8. PERFORMING ORGANIZATION REPORT NO.
9 PERFORMING ORGANIZATION NAME ANO ADORES?
The University of Tulsa
600 South College
Tulsa, OK 74104
10. PROGRAM ELEMENT NO.
CBGB1C
R805099
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Robert's. Kerr Environmental Research Laboratory
P. 0. Box 1198
Ada, OK 74820
13. TYPE OF REPORT AND PERIOD COVERED
Fina 1/6/78-6/81
14. SPONSORING AGENCY CODE
600/15
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The objectives of this project were to develop an environmental assessment
data base for characterizing and treatment of petroleum refinery wastewaters and
residual sludges, and recommendation of further research needed to improve the
data base.
The project was conducted in three phases. Phase 1 was the establishment of a
Peer-Group Review Committee to provide direction to the project and to ensure that a
diversity of viewpoints was considered. Six eminent experts in the waste treatment
field were chosen to serve on the committee.
Phase 2 involved the preparation of four1comprehensive state-of-the-art reviews,
by outside consultants, to provide the environmental .assessment data base on refinery
wastewaters and residual sludges.
Phase 3 included a critical examination of the four individual s^ate-of-the-art
reviewsi selection of ei^ht areas where further research was considered to be needed
to improve the data base, and preparation of this report.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Croup
Activated carbon treatment, activated
sludge, process, industrial waste treatment,
oxidation ponds, petroleum refining, resi-
dues, sludge, toxicity, wastewater.
Environmental assessment
Priority pollutants
Research Needs
Residual Sludges
WasteWater characterlzat
Wastewater treatment
6F
7A
138
on
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
2O. SECURITY CLASS (This pott)
UNCLASSIFIED
22. PRICE
EPA Po.m 22:0-1 (»-7J)
-------
DISCLAIMER
Although the research described in this article has been funded wholly
or in part by the United States Environmental Protection Agency through
Grant No. R-805099 to The University of Tolsa, it has not been subjected to
the Agency's required peer and policy review and therefore does not necessar-
ily reflect the views of the Agency and no official endorsement should be
inferred.
11
-------
FOREWORD
The Environmental Protection Agency is charged by Congress
to protect the Nation's land, air and water systems. Under a
mandate of national environmental laws focused on air and water
quality, solid waste management and the control of toxic sub-
stances, pesticides, noise, and radiation, the Agency strives to
formulate and implement actions which lead to a compatible bal-
ance between human activities and the ability of natural systems
to support and nurture life. In partial response to these man-
dates, the Robert S. Kerr Environmental Research Laboratory,
Ada,, Oklahoma, is charged with the mission to manage research
programs, to investigate the nature, transport,' fate, and manage-
ment of pollutants in ground water and to develop and demonstrate
technologies for treating•wastewaters with soils arid Other nat-
ural systems; for controlling pollution from irrigated crop and
animal production agricultural activities; for developing and
demonstrating cost-effective land treatment systems for the
environmentally safe disposal of solid and hazardous wastes.
In a coordinated research effort with the industrial and
the academic communities, the EPA has generated a large data base
from the application of treatment and control technologies to
petroleum refinery wastewater and residuals. This report
summarizes the available information as a function of specific , •
technologies and projects the remaining research needs.
Clinton W. Hall, Director
Robert S. Kerr Environmental
Research Laboratory
iii
-------
ABSTRACT
The objectives of this project were to develop on environmentol ossessment data
base for characterizing and treatment of petroleum refine./ wastewaters anH residual
sludges, and recommendation of further research needed to improve the data base.
The project was conducted in three phases. Phase One was the establishment of
a Peer-Group Review Committee to provide direction to the project and to ensure that a
diversity of viewpoints were considered. Six eminent experts in the waste treatment
field were chosen to serve on the committee.
'.-' Phase Two involved the preparation of four comprehensive state-of-the-art
reviews, by outside consultants, to provide the environmental assessment data base on
refinery wastewaters and residual sludges.
Phase Three included a critical examination of the four individual state-of-the-
art reviews, selection of eight areas where further research was considered to be needed
to improve the data base, and preparation of this report.
This report was submitted in fulfillment of Contract R805099010 by the University
of Tulsa under sponsorship of the U. S. Environmental Protection Agency.' The report
covers the period from June 1978 to June 1981, and work was completed as of June 1981.
iv
-------
CONTENTS
Foreword., ii1
Abstract '. . . • > iv
Figures ...... vi
Tables viii
Acknowledgments xiii
1. Introduction 1
2. Conclusions and Recommendations 3
3. Characterization of the Petroleum Refining Industry and
Petroleum Refinery Wastewaters by M. R. Beychox 6
4. A Review of Pollutants in Petroleum Refinery .Wastewaters and
Effect on Aquatic Organisms by S. Burks . 68
5. An Evaluation of Existing and Emerging Control Technology
for the Treatment of Petroleum Refinery Wastewaters and
Sludges by 0. L. Ford 88
6. Evaluation of Emission Levels of Petroleum Refinery
Wastewaters and Residuals by E. H. Snider and F. S. Manning. 165
7. Recommended Research Needs 203
-------
FIGURES
Nt-mbar Poge
1 Crude oil statistics 14
2 Petroleum Administration for Defense (PAD) districts 16
3 Example of Category A Refinery 46
4 Example of Category B Refinery 48
5 Example of Category C Refinery 51
6 Example of Category D Refinery 53
7 Example of Category E Refinery 54
8 Wastewater sources and wastewater management in a refinery 56
9 Experimental observations concerning oil and grease reduction with
coagulant addition ' 98
10 Polyelectrolytes evaluated as coagulant aids in conjunction with alum
as the primary coagulant 100
11 Impact of influent oil and grease concentration on treatment performance
of conventional oil-water separators 103 ,
12 Impact of peck hydraulic, loading on API separatoi effluent quality .... 105
13 Treatment performance of pilot-scale titled-plate separator. 106
14 Variation in DAF treatment performance for oil and grsase removal -
Condition A 107
, 15 Variation in DAF treatment performance for oil and grease removal -
Condition B . . 108
. 16 Air flotation efficiency as a function of influent oil and grease
concentration 112
17 Variation in IAF treatment performance for oil and grease removal . . . : 113
18 Variation in IPS and IAF, treatment performance for suspended solids
removal , 116
19 Relative performance of oil-water separators for treatment of dispersed
" and emulsified oils. 117
vi
-------
FIGURES (conti.iued)
1 i
Number • Page
20 Relative oil-water separator performance when treat-ng low levels of
-- oil in wastewater. 1>9
21 Relationship between suspended solids and oil and grease concentrations
i in biological treatment plant effluent 129
22 Effluent oil and grease from biological unit processes for the treatment
of refinery wastewdters 1.30
23 Variation in RBS treatment performance for, COD and oil and grease
removal 133,
24 Variation in RBS treatment performance for phenol removal 135
25 Variation in GAC treatment performance for TOC removal 140
' " 26 Effluent COD attainable from activated carbon system 141
27 Variation in GAC treatment performance for oil and grease removal from
API separator effluent 143
28 Variation in GAC treatment performance for oil and grease removal from
asrated lagoon effluent •.....' 144
29 Effluent oi! and grease from GAC unit processes for the treatment of
refinery wastewaters 145
• 30 Oil and grease removal as a function of landfarm residence time 153
i 31 Petroleum refinery wastewater treatment plant schematic 158
• 32 Schematic of soiids handling and disposal alternatives 159
33 Tertiary carbon adsorption addition to petroleum refinery wastewater
treatment plant : . . . 1 160
vii
-------
TABLES
Number Page
' t
1 Sulfur Content of Crude Oils. , . 10
2 API Gravity of Crude Oils 12
3 Crude Oil Statistics for the Refining Industry . 13
4 ' Wastewater Sources From Refinery Unit Processes 43
5 , Refinery Classifications as Defined by Unit Processes in Each
Refinery Category 45.
6 Median Dry-Weather Wastewater Flows 58
7 Refinery Pollutant Loads and the Equivalent Wastewater
Pollutant Concentrations 64
8 Partial List of Compounds Identified in Extracts From Oil >
Refinery Wastewaters 75
9 Partial List of Compounds Identified in Neutral Extracts
From Oil Refinery Wastewaters , 76
10 List of Organic Compounds Identified or Detected in Neutral
Extracts From the Effluents From a Dissolved Air Flotation
(DAF), in Activated Sludge Final Clarifier (FC), or a Pilot-
Scale Mixed-Media Filter/Activated Carbon (MMF/AC)
Treatment Systems > r 78
11 Polyefectrolytes as Primary Coagulants for Crude Oil Dis-
persions °9
12 Evaluation of Dissolved Air Flotation Treatment Performance .... 110
13 Dissolved Air Flotation Design Parameters Ill
14 Evaluation of Induced Air Flotation Treatment Performance 115
15 Process Comparison for Oil and Grease Removal 120
16 Biological Treatment Coefficients for Petroleum Refinery
Wastewaters . . . . 125
17 Effluent Quality From a Modified Aerated Lagoon Treating
Petroleum Refinery Wastewaters - Monthly Average .Data 127
viii
-------
TABLES (Continued)
Number Poge
18 Pilot-Scale RBS Effluent Quality for Treatment of Petroleum
Refinery Wastewaters. 132
19 Compaiison of Upflow Sand and Deep-Bed Tertiary Filtration
Facilities for the Treatment of Refinery Wastewaters 137
20 Design Characteristics for Terticry Filtration Facilities . .138
21 Characterization of Oil Contaminated API Separator Bottoms 147
22 Characterization of Oil and Sludge Phases From Tilted-Plate
Separator 148
23 Pilot-Scale Filter Prefs Unit for Dewatering Oily Solid Wastes . . . .150
24 Pilot-Scale Vertical Solid Bowl Centrifuge Unit for Dewatering
- * Oily Solid Wos.es - EPA Category E Refinery . .151
25 Analysis of Soil From Landfarm Si*es for the Disposal of Oily
, Solid Wastes 155
1 i
26 Storm Water Runoff and Leachate Characteristics From a Land-
farm Site for the Disposal of Oily Solid Wastes . .156
27 Daily Variability Factors for Treatment of Petrcleum Refinery
Wastewaters ..;.... 162
28 Monthly Variability Factors for Treatment of Petroleum
Refinery Wastewaters •. 163
29 Median Raw Waste Loadings for Refinery Subcategories /
Through E as Reported in API-EPA Survey 166
30 Average Intake and Wastewater Effluent Concentrations After
Various Treatment Processes in a Class B Refinery k . 168
31 Concentration of n-Alkanes in the Neutral Fraction of the DAF ;
Effluent From a Class B Refinery and Percent Removal by rhe
Activated-Sludge and Activated-Carbon Units ....... i ... .169
: 32 Concentration of Cycloalkanes and Alkanes Other Than n-
; Alkanes in the Neutral Fraction of the DAF Effluent From a
Class B Refinery and Percent Removal by the Activated-Sludge
; and Activated-Carbon Units ......;.....„ .170
33 Concentration of Alkylated Benzenes in the Neutral Fraction
: of the DAF Effluent From a Class B Refinery and Percent
.Removal by the Activated-Sludge and Activated-Carbon Units . . ..171
-------
TAPLE5 (Continued)
Number , ' Poqe
34 Concentration of Indan and Tetralin and Related Compounds and
Their Alkylated Derivatives in the Neutral Fraction of the
Effluent From a Class 8 Refinery and Percent Removal by the
Activated-Sludge and Activated-Carbon Umts 172
35 Concentration of Naphthalene and Alkylated Naphthalenes in
the Neutral Fraction of the DAF Effluent From a Closs B
Refinery and Percent Removal by fhe Activated-Sludge and
Activated-Carbon Units 173,
36 Concentration of Alkylated Benzothiophenes and DibenzothioT
phenes in the Neutral Fraction of the DAF Effluent From a Class
B Refinery and Percent Removal by the Activated-Sludgt and
Activated-Carbon Units ' ',74
37 PNAs and Alkylated PNAs Other Than Naphthalenes in the
Neutral Fraction of the DAF Effluent From a Closs B Refinery
and Percent Removal by the Activated-Sludge and Activated-
, Carbon Units 175
38 Phenols in the Acid Fraction of the DAF Effluent From o Class
B Refinery ond Percent Removal by the Activated-Sludge and
Activated-Carbon Units , 176
39 Alkylated Pyridines in the Base Fraction of the DAF Effluent
From a Class B Refinery and Percent Removal by Activated-
Sludge and Activated-Carbon Units 177
40 Alkylated Quinolines in the B'ise Fraction of the DAF Effluent
From a Class B Refinery and Pe-cent Removal by Activated-
Sludge and Activated-Carbon L'nits 178 .
41 Alkyloted Anilines in Base Fraction of the DAF Effluent From a
Class B Refinery ond Percent Remqval by Activated-Sludge and
Activated-Carbon Units , . . . 179
42 Concentration Ranges of Priority Pollutants - Volatile Organics
Category Reported in Afl-EPA Priority Pollutants Survey 180
i
43 Concentration Ranges of Priority Pollutants - Semi-Volatile
Extractables Category Reported in API-EPA Priority Pollutants
Survey ' 181
44 Concentration Ranges of Priority Pollutants - Total Metals
Category Reported in API-EPA Priority Pollutants Survey 182
-------
. TABLES (Continued)
Number Poge
45 Concentrction of Pesticides and Asbestos Categories of Priority
Pollutant- Observed in Data Obtained During API-EPA Priority
Pollutants Survey 163
46 Median Performance of Activated Sludge Treatment Plants for , •
Three Class B Refineries and Two Class C Refineries ....... .184
47 Concentration Ranges'of Classical Parameters and the Priority
Pollutant Total Metals Observed in Pilot Scale Granular
Activated Carbon Treatment of Three Class B Refineries .186
48 Concentration Ranges of Classical Parameters and the Priority
Pollutant Total Metals Observed in Pilot Scale Granular
Activated Carbon Treatment of Two Class C Refineries ...... .187
49 Concentration Ranges of Classical Parameters and the Pitoriry
Pollutant Total Metals Observed in Pilot Scale Granular
Activated Carbon Treatment of a Refinery of Unknown Classi-
fication , 188
50 Concentration Ranges of Classical Parameters and the Priority
Pollutant Total Metals Observed in Pilot Scale Powdered
Activated Carbon Treatment of Refinery Wastewaters 189
51 Case History Long-Term Effluent Concentrations for Plants
Conforming to Abatement Levels I and II 190
52 Case History Activated Carbon Applications in Petroleum
Refinery Wastewater Treatment 191
53 Analytical Data for Priority Pollutants From One Refinery 193 ,
54 Air-Stripper Sampler Analysis Results From One Refinery 194
55 Properties of Simulated Oily Sludges Used in Study of Disposal
by Soil Cultivation 195
56 . Analytical Variability of Wastewater Parameters 197
57 Analytical Variability of Metal Concentrations in Refinery
Wastewaters 199
XI
-------
LIST OF ABBREVIATIONS
AC ~ activated carbon IR
API — American Pet-oleum Institute LPG
AS -- activated sludge MEK
BAT — best available pollution MIBK
control technology MLVSS
BCT -- best conventional pollution
control technology MMF/AC—
BMP's ~ best mcnogement practices
BOD — biological oxygen demand MS —
BPCTCA — best practicable control tech- NhiP
nology currently available NSPS ~~
BPT — best practical pollution
control technology PAC
RTX —benzene, toluene, xylenes PONA
CAM — carbon adsorption method
COD — chemical oxygen demand POTW —
CPI — corrugated plate interceptor
DAF — dissolved air flotation PPI
DEA —• diethonol omine p.t.b. —
DOE — Deportment of Energy RBS
EPA — Environmental Protection RCRA
Agency
ES — Engineering-Science, Inc. SDA
FC — final clarifier SPCC
FCC — fluid catalytic crocking
GAC a- granular activated carbon , TDS
GC —gas chromatogroph TOC
GLC -- gas liquid chroma tog rrjphy TOD
IAF — induced air flotation TON
TSS
- infrared
- liquefied petroleum gas
- methyl ethyl ketone
- methyl isobutyl ketone
- mixed liqucr volatile suspenued
solids
mixed media filter/activated
ccrbon
moss spectrometer
nnmethy I -2-pyrrol idone
New Source Performance
Standards ,
powdered activated carbon
paraffin-olefin-naphthene
aromatic
Publicly Owned Treatment
Works
parallel-plate interceptor
pounds per thousand barrels
rotating biological surface
Resource Conservation and
Recovery Act
solvent de-asphalting
spill prevention and counter**
measure
total dissolved solids
total organic carbon
total oxygen demand
threshold odor number
total suspended solids
XII
-------
ACKNOWLEDGMENTS
The outhors wish to express apprecioHon to the members of the Peer-Group
Review Committee for their participation and valuable contributions: Dr. Karim Ahmed,
Natural Resources Defense Council; Dr. John Caims of the Virginia Polytechnic Institute
and State University; Professor W. W. Eckenfelder J Vanderbilt University; Mr. James
Grutsch of The Standard Oil Company (Indiana); Dr. Martha Sager of American
University; and Ms., Judith Shaw of the American Petroleum Institute.
In addition, the state-of-the-art review authors - Mr. Milton R. Beychok, Dr.
Sterling (Bud) Burks, and Mr. Davis.L. Ford - provided much helpful input which is
gratefully acknowledged. , ,..
The continued assistance and interest of Mr. Fred M. Pfeffer, Project Officer,
is greatly uppreciated.
Finally, the able assistance of Mrs. Neida Whipple in organizing and typing the
manuscript is much appreciated.
XIII
-------
SECTION 1
INTRODUCTION ,
The goals of the present work were to develop a complete environmental assess-
ment data uase concerning the treatment of petroleum refining wastewatcrs and residuals,
and the recommendation of areas where further research is needed to improve the data
base. The goals were achieved by establishing a Peer-Group Review Committee,
developing state-of-the-art reviews by outside consultants, compilation of the sate~of-
the-art reviews, and selection of those areas where further'research could be most useful.
A Peer-Group Review Committee of six people, chosen for their expertise in
refinery wastewater treatment, was selected by the Project Manager in consultation with
the Project Officer and the Robert S. Kerr Environmental Research Laboratory. The
committee provided oversight direction to the project and included acknowledged
experts, most of whom had actual experience in petroleum refinery wastewatar manage-
ment. The members of the committee were chosen so as to represent a brocd spectrum of
diverse viewpoints from industry, universities, the American Petroleum Institute (AP!) and
the consulting field.
To develop the environmental assessment data base, state-of-the-art reviews
were prepared on these four topics:
1,. Characterization of the petroleum refining industry and refinery wastewaters.
The parameters affecting the generation of wastewater pollutants are discussed,
including crude pil compositions, refinery technologies and classifications,
wastewater sources and their pollutants. Fpture trends for the refining industry
are included.
2. A pollutant discussion and rationale for characterizing the wastewaters from
petroleum refining and their toxicity effects upon aquatic organisms. This
rationale includes a review of the analytical procedures used to measure, define
and assess the effects of the various pollutants. The list of pollutants considered
includes the current permit parameters (chemical oxygen demand, COD; bio-
logical oxygen demand, BOD; etc.) as well as the priority pollutants.
3. A comprehensive discussion and evaluation of existing and emerging wastewater
treatment and control technology. Data on the performance capabilities of the
various technologies are included.
'"•' . 1
-------
4. A complication of do to on the discharge levels of refinery wastewoter pollutants
including their avenues of discharge.' The data include estimates of accuracy,
precision, variance and causes of variance whenever possible.
Based upon a critical examination of the individual state~of~the-art reviews,
areas within the existing data base were selected as meriting further research to improve ,
the environmental assessment data base. Each of the specific areas selected ore discussed
so as fo explain why further research might be fruitful and to highlight the benefits that
might be expected of such research. The recommended research needs are summarized in
the Conclusions and Recommendations section of this report.
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
Existing refinery technology is very complnx and the development of new tech-
nology is an on-going process. Within the 20-year period of 1940-1960, a number of
technological "breakthroughs" occurred which included the development of fluid
catalytic cracking and processes involving catalysis in u hydrogen atmosphere (catalytic
reforming, hydrotreating and hydrocracking). With the current body of knowledge on
catalytic processes, the future development of new refinery technology is expected to be
more evolutionary in nature. Another period of rapid technological breakthrough is not
expected.
Many refineries already practice the in-plant reuse of treated wastewaters to
some extent and air-cooling has replaced much water-cooling. Th^re is little likelihood
that more intensive emphasis on wastewater resse will dramatically reduce wastewater
volumes.
In general, the current technology for refinery wastewater treating (consisting '
of in'plant reduction of wastewater generation, primary removal of oil and suspended
solids, and secondary treatment via biological oxidation) can satisfy the regulatory
criteria for control of the conventional pollutants such as oil, suspended solids, BOD,
COD, phenols, sulfides, ammonia, etc. There is much evidence that the current waste-
water treatment technology essentially removes or degrades those pollutants which cause
lethal, short-rerm toxicity to fish.
Eight areas have been identified in which further research is needed to improve
the environmental assessment data base for characterizing and treating refinery waste-
waters and residual sludges. Those areas of research needs are briefly listed below:
ACTUAL PLANT DATA COLLECTION AND CORRELATION
1. Correlation of the key design factors in activated sludge biotreatment, such as
mixing horsepower and reaction basin retention time, with pollutant removal
efficiency.
.2. Correlation of actual case history data on the use and effectiveness of chemical
oxidants (hydrogen peroxide, chlorine and ozone) in treating refinery waste-
waters. ,
-------
LABORATORY-SCALE RESEARCH AND DEVELOPMENT
3. identifying which specific refinery wastewater pollutants exhibit long-term fish
toxicify (lethal or sub-lethal). Development of techniques for the rapid deter-
mination of lor.g-term fish toxicity.
4.- - Determination of which analytical test methods ore the least reliable and their,
contribution to overall effluent quality variability. Development of new, more
reliable test methods, if possible.
PILOT-AND DEMONSTRATION-SCALE RESEARCH
5. Development of an economic method for rtscoverying and regenerating the
spent powdered activated carbon (PAC) used to enhance the performance of
activated sludge biotreaters.
6. Funding the demonstration and operation of the granular activated carbon (GAC)
process (for the tertiary treatment or secondary treatment of refinery wasfe"
woters0) in a unit capable of treatmenr of 200-'*00 gpm of refiner/ wostuwater.
i
7. For the londfarrning of oily sludges, determinution cf the relationship between
sludge vapor pressure and problems of odor control and air emissions. Develop
methods for mitigating such problems. Also investigate methods of resolving
problems with landfarming vis-a-vis Resource Conservation and Recovery Act
(RCRA) regulations if such problems exist.
FEASIBILITY AND GUIDANCE STUDY
8. Development of a comprehensive feasibility and guidance study regarding the
methods of producing a concentrated, residual pollutant brine (reverse osmosis
and evaporation) and the ultimate disposal of residual wastewater pollutants
in evaporation-percolation pone's, subsurface injection wells and remote dis-
posal dumps. Such a study should include a realistic assessment of capital costs
and energy usages as well a*, a realistic assessment of the benefits to be expected
in terms of the magnitude of improvement in the quality of the nation's waters.
The study should also asseis the cost-benefit ratio of the Environmental
Protection Agency's (EPA) New Source Performcnce Standards (NSPS) "no dis-
charge" requirement for refineries relative to control of other pollutant sources
such as non-point sources.
,' The above order of listing for the recommended research needs is not intended to
,be a prioritized listing. The listing is merely categorized so as to present the least costly
0 A description of the successful use of the GAC process for the secondary treatment of
. wastewater in a Japanese refiner/ (on a full-plant scale) has been published in the
- Oil & Gas Journal of May 11, 1981
-------
research recommendations first and the most costly recommendations last. A mo. 3 •
detailed rationale for the recommended research needs is presented in Section 9 of .this
report.
-------
SECTION 3
CHARACTERIZATION O'F PETROLEUM REFINING INDUSTRY
AND PETROLEUM REFINERY WASTEWATERS
M. R. Beychok
Environmental Consultant
, California
INTRODUCTION
i i
Most petroleum refineries are similar in that they process a crude oil raw
material so as to produce a variety of endproducts. But beyond that similarity, the
refining industry encompasses a broad range of refinery types consisting of an even
broader range of refining technologies or unit processes and producing a wide spectrum of
motor gasolines, heating and industrial fuel oils, lubricating oils and petrochemicals.
The purpose of this section is to provide a brief discussion nnd explanation of petroleum
refineries and the unit processes involved in the design and operation of refineries.
TYPES OF REFINERIES
Refineries differ considerably in terms of their range of endproducts. Some
refineries maximize the production of motor gasolines and some maximize the production
of heating oils and industrial fuel oils. Other refineries emphasize the production of
lubricating oils, and still others are designed to produce petrochemical endproducts or
intermediate product: for sale as feedstocks to petrochemical plants.
In some cases, refineries shift their endproduct emphasis on a seasonal basis.
For example, motor gasoline production will be maximized for the spring and summer,
and heating and fuel oil production will be maximized for the fail and winter.
i * . '
Many so-colled combinati. i refineries produce a product slate (endproduct
range) that includes the entire spectrum of gasolines, heating oils, industrial fuels, lubri-
cating oils and petrochemicals.
' In general, most petroleum refineries can be categorized as being within one or
more of the following types in terms of their product slate:
^ o Motor gasoline refineries
-------
o Heating and fuel oil refineries
o 'Lubricating oil refineries
o Petrochemical refineries
o Combination refineries
As will be discussed later herein, in terms of wastewater generation and composition,
refineries are probably better categorized by their complexity as defined by the types of
unit processes or technologies used within the refineries.
THE COMPOSITION OF PETROLEUM CRUDE OIL
The key factors involved in selecting the unit processes or technologies used
within a petroleum refinery are the desired product slate and the composition of the
refinery's raw materiel crude oils. Since the amount and composition of wastewaters
generated by o refinery are largely dependent upon the processes used in the refinery, it
becomes evident that the composition of the raw material crude oils is one of the factors
which determines the generation and composition of refinery wastewaters. Thus, it is
Important to have some understanding of the nature of petroleum crude oils and, 'in •
particular, to understand that crude oils obtained from different geographic oil field
sources can and do have significantly different compositions.
Petroleum crude oil >is primarily a mixture of various "hydrocarbons", which are
chemical molecules composed of hydrogen and carbon atoms. Those molecules may be
• quite simple or quite complex in the structural arrangement or linking of their component
atoms. The carbon atoms may be linked together in short or long straight chains with or
without branched "side-chains", or they may be lir.ked together in cyclic or ring arrange*
ments. The simpler hydrocarbons found in crude oils are "paraffins" or "saturated hydro-
carbons" in which,each carbon atom is linked with the maximum possible number of
hydrogen atoms in accordance with the generic formula of CnH2n+2- ^or example, a
paraffin molecule with 10 atoms of carbon (C) would be linked with 22 atoms of
hydrogen (H). The saturated or paraffinic hydrocarbons may be "normal" paraffins in
which carbon atoms are linked in straight chains, or "isoparoffins" in which the carbon
atom linkage includes branched chains. For example, normal pentane which is ""
and isopentane which is i~C5H^2 both have the same number of carbon and hydrogen
atoms in accordance wifh the generic paraffin formula of ^n^2r,+2' ^°v* normal pentane
has five carbon atoms in o straight chain while isopentane has o branched linkage of five
carbon atoms. .
Hydrocarbons with straight or branched carbon atom chains and containing less
than the maximum of four hydrogen atoms per carbon atom are called "Onsotu rated" or
"olefinic". For example, pentene (C5Hjn.) is unsoturated or mono-olefinic, and
pentadiene (C5Hg) is even more unsoturated or di-olefin?c.
The carbon atoms in hydrocarbon molecules may also be linked together to form
closed cycles or rings. When the carbon atoms in cyclic hydrocarbons are linked to the
maximum number of hydrogan atoms possible for such cyclic arrangements, the molecules
-------
are "saturated naphthenic hydrocarbons" with the generic formula of CnH2n (which i$
the same as for the unsoturated monc-olefins discussed above). The nophthenic hydro-
carbons are usually named with the prefix of "cyclo", as in cyclo-pentone
Naphthenic hydrocarbons may also be unsaturated as in cyclo-pentene
When hydrocarbon molecules have six carbon atoms linked together !n a ring
containing three unsoru rated links, the molecules are colled aromatics. For example,
benzene is an aromatic with the formula of COHO. The aromatic ring may also be linked
to a saturated side-chain as in toluene (€7^) and xylene (CgHjg), or to an unsaturcted
side-chain as in styrene (CgHg), all of which fail within the category of aromatics.
> i
Petroleum crude oils contain literally hundreds of different hydrocarbons, many
of which are very much more complex in their molecular structure than those discussed
above. However, the analysis of a crude oil. in terms of the relative quantities of the
four molecular categories discussed above provides a valjable insight as to the unit pro-
cesses that will be required in a refinery designed to produce a specific product slate from
that crude oil. Such analyses ore called PONA's which is an acronym for paraffin-
olefln-naphthene-aromatic analyses".
Crude oils also contain various amounts of organic sulfur and organic nitrogen
compounds. Some of the unit processes used in many refineries are designed to cbta-
lytically "desulfurize" certain of the intermediate products within the refinery. Those
desulfurization processes convert organic sulfur and nitrogen into gaseous hydrogen
sulfide (H2S) and ammonia (NH3) which ore then removed in large part by stripping or
distillation. However, a small part of the H2$ and NH3 almost invariably appears in
some of the refinery wastewaters in the form of ammonium hydrosulfide (N^SH) in
amounts ranging from about 100 ppm to 10,000 ppm or more^ . Refinery wastewaters con-
taining ammonium hydrosulfide or free hydrogen sulfide are commonly referred to as
'"sour waters" which may also be generated from unit processes other than catalytic
desulfurizers. The amount of hydrogen sulfide and ammonia contained in refinery sour
waters (as NH4$H) is directly related to the organic sulfur and organic nitrogen contents
of the raw material etude oils as well as to the type of unit processes used in a specific
refinery . '
• As discussed above, all crude oils are a complex mixture of hydrocarbons.
However, the amount of "light" or low -boiling hydrocarbons relative to "heavy" or high-
boiling hydrocarbons varies very considerably from crude oil to crude oil. In very
generalized terms, the hydrocarbons in crude- oil can be classified by their boiling range
and number of carbon atoms. Storting with the lightest or lowest-boiling hydrocarbons
(which are gases) and proceeding up th« boiling range scale to the heavy fuel oils,
yields:
° Crude oils rarely contain any olefinic hydrocarbons. However, olefins are produced
1 ~ by subsequent refinery processing.
'"" * ,-:•-.
8
-------
Number of Atmospheric
carbon atoms boiling range (°F)
Methane, ethane (gases) 1,2 -259, -128
Propane, butane 3,4 -44, 31
Naphthas (gasolines) 5 to 11 100 to 400 °
Diesel and heating oils 10 to 15 350 to 600 a
Industrial fuel oils , 16 plus 600 plus
aDesig.-!Otes range of initial boiling point to final boiling point.
Crude oils such as those from Qatar in the Middle East with a specific gravity
of 0.82 (an API gravity of 41°) and containing about 38 volume percent of naphthas are
referred to as "light" crude oils because of their relatively low specific gravity and
relatively high rontent of hydrocarbons in the naphtha boiling range of 100 to 400 °F.
Crude oils such as those from Bachaquern in South America with a'specific
gravity of 0.95 (an API gravity of 17°) and containing about 10 volume percent of
naphthas are referred to as "heavy" crude oils because of their relatively high specific
gravity qnd relatively low content of hydrocarbons in the naphtha boiling range.
> The amounts of the various boiling range materials in a crude oil and the
physical properties of the individual boiling range materials (specific gravity, sulfur and
nitrogen contents, PONA analyses, etc.) are determined by a complete crude oil
analysis usually called a "crude oil assay". The assay of the raw material crude oils is a
key factor in the selection of the unit processes required to achieve a desired refinery
product slate. As stated previously, *hat means that the generation and composition of
the refinery's wastewater depends upon the refinery's crude oil assay to a large extent.
To summarize, the key characteristics of a crude oil are:
o Its boiling range components
o Irs sulfur and nitrogen content
o Its PONA hydrocarbon analysis.
Those characteristics largely determine the unit processes required in refining the crude
oil, and they vary quite considerably with the geographic location of the crude oil
source.
*
Table 1 presents data on the sulfur content of crude oils relative to their geo-
graphic source on a worldwide basis. As shown in Table 1 (for the year 1974), 85 per-
cent of the crude oils from Africa and 100 percent of the crude oils from Australasia are
"low-sulfur" oils containing no more than 0.5 weight percent sulfur. In contrast, 47
percent of Middle Eastern and 71 percent of South American crude oils are "high-sulfur"
oils containing more than 2.0 weight percent sulfur. On an overall weighted basis:
-------
TABLE 1. SULFUR CONTENT OF CRUDE OILS
AFRICA
MIDDLE EAST
AUSTRALASIA a
SVROPE b
S. AMERICA
II . AMERICA C
TOTAL SOURCES
I a? ^f >
1974
Oil Production
(bbls/dau}
5,362,000
21,453,000
1,773,000
195,000
3,425,000
10,956,000
43,169,000
PERCENTAGE OF OIL PRODUCED
0.0-0.5
vt* S
05.0
0.0
100.0
57.7
4.2
57.8
29.9
0.51-1.0
fcrtt S
12.0
5.5
0.0
42.3
3.4
7.5
6.6
1.01-2.0
wt% S
3.0
47.6
0.0
0.0
21.7
26.3
32.4
> 2.0
hrt* 5
0.0
46.9
0.0
O.C
70.7
8.4
31.1
* Does not include China
Does not include Russia
g
Includes Canada, Mexico and U.S.A.
SOURCE: Thompson,C.J. et al, "Sulfur in World Crudes", Hydrocarbon Processing,
February 1376
10
-------
29.9 percent of the world's crude oils contain no more than 0.5 weight percent
sulfur.
6.6 percent of the world's crude oils contain between 0.5 and 1 .0 weight
percent sulfur.
32.4 percent of the world's crude oils contain between' 1 .0 and 2.0 weight
percent sulfur. • - '
31 .1 percent of the world's crude oils contain more than 2.0 weight percent
sulfur.
i
Table 2 (for the year 1978) presents similar data on the API gravity0 of crude •
oils. About 50 percent of South American crude oils are heavier than 30° API and, in
contrast, over 90 percent of Middle Eastern oils are lighter than 30° API. On an overall
weighted basis:
1 .9 percent of the world's crude oils are heavier than 20° API .
13.8 percent of the world's crude oils are between 20 and 30 API.
71 .4 percent of the world's dude oils are between 30 and 40 API .
12.9 percent of the world's crude oils are lighter than 40 API.
i '
Tables 1 and 2 illustrate the very wide diversity and range of composition found among
the world's major sources of crude oil. ,
A perspective on the amount of crude oil processed through U.S. refineries is
provided m Table 3 and graphically presented in Figure 1 . During the past six years:
o The number of U.S. refineries in operation has increased from 247 to 289.
o The national crude oil refining capacity has increased from about 14 million
to about 17 million barrels per day.
o The total amount of crude oil actually processed has ranged between 12
mi Dior, and 15 million barrels per day, representing 83 to 88 percent of the
national refining capacity. In other words, there has been 12 to 17 percent
of excess or unused refining capacity.
o The amount of imported crude processed in U.S. refineries has increased from
26 percent of the total oil processed to slightly over 40 percent.
The dependence of U.S. refineries on imported crude oil for a substantial part of
their throughput will most probably continue for some time to come. Since a major part
of those imports are from the Middle East, the sulfur content of the crude oils processed in
° API gravity is inversely prooortional to specific gravity. The higher the numerical
y"'ue °' API gravity, the lighter is the specific gravity.
11
-------
TABLE 2. API GRAVITY OF CRUDE OILS
AFRICA
MIDDLE EAST
AUSTRALASIA a
EUROPE b
S. AMERICA
H. AMERICA c
TOTAL SOURCES
1978
Oil Production
, (bbls/day)
5,850,000
20,618,000
2,698,000
1,615,000
3, 399, 000
21,043,000
45,223,000
PERCENTAGE OF OIL PRODUCED
0-19.9
"API
0.5
0.2
0.5
1.7
13.8
2.4
1.9
20-30
•API
17.1
7.J
11.5
7.3
46.5
16.2
13.8
30.1-40
"API
49.0
91.5 ,
57.9
67.2
32.1
61.6
71.4
> 40
•API
33.4
1.2
30.1
23.8
7.6
19.8
12.9
Does not include China
b
Does not,include Russia
c
Includes Canada, Mexico and U.S.A.
SOURCE: "Hcrldvide Production", Oil and Gas Journal, December 25, 1978
12
-------
~*s
TABLE 3. CRUDE OIL STATISTICS FOR THE REFINING INDUSTRY
(for th> yuan of 1973-1978, inclusive)
CJ
During the
year of:
1973
1974
1975
1976
1977
1976
-
SV.....J.A — , J 1
- ujuucr UJ.JL
" refining
Number of capacity.
refineries b/d
247
259
256
266
285
289
14,216,000
14,845,000
15,075,000
16,170,000
16,849,000
17,170,000
Total amount of
crude. oil processed
Amount of domestic Amount of Imported
crude oil processed crude oi.) processed
Percent of Percent of Percent of
refining crude oil crude oil
b/d capacity
12,452,000
12,241,000
12,480,000
13,406,000
14,727,000
14,788,000
87.6
82.5
82.8
82.9
87.4
86.1
b/d
9,208,000
. 8,764,000
8,375,000
8,119,000
8,179,000
8,718,000
processed - b/d
73.9
71.6
67.1
60.6
55.5
59.0
3,244,000
3,477,000
4,105,000
5,287,000
6,548,000
6,070,000
processed
26.1
28.4
32.9
39.4
~ 44.5
41.0
Note: All values of b/d are in average barrels per calendar day
SOURCES: (a) "Forecast and Review", Oil and Gas Journal, January 29, 1979
(b) Annual Kefining Issues of the Oil and Gas Journal for the years 1974 through 1979.
-------
18 r-
, 16
14 \-
2
vo
o
12 I-
10 h-
8 \—
270 ««
260 £2 '
250 |uj
240 •
13.9 %
excess
capacity
.41.0 %
imported
1973 1974 1975 1976 1977 1978
FIGURE1!. Crude oil statistics
U
-------
U.S. refineries will trend upward. As that occurs, U.S. refineries will have to install
additional desulfurizotion facilities which will increase the generation of sour waste-
waters. Additional hydrogen production facilities will also be needed to supply the need
for hydrogen in the desulfurizers. Since hydrogen production units require large amounts
of steam, there wiH be on increase in the fresh water demand of trie refineries as well as
an increase in wastewater steam blowdown and boiler feedwater demineralization wastes.
2
A study of the refining industry by the National Commission on Water Quality
in 1975 reported these average crude oil characteristics for refineries located in the
districts defined in Figure 2:
District Number of Weight API Gravity '
refineries* percent sulfur
1 29 1.16 34.3
2 70 0.68 36.3
3 104 0.52 35.3
4 30 0.97 35.3
5 56 1.07 32.6
(* As of January 1, 1979)
i
Although the above regional refining data on crude oil characteristics have probably
changed somewhat since 1975, they serve to illustrate again the diversity of crude oils
processed in U.S. refineries.
REFINING UNIT PROCESSES AND TECHNOLOGIES
A wide variety of,unit processes and technologies are used in petroleum
refineries. In general, they can be classified into four major groups:
o Distillation is used to separate the crude oil into various boiling range
fractions or "cuts". The crude oil is usually distilled in two steps,1 one at
essentially atmospheric pressure and one at vacuum or reduced pressure con-
ditions.
o Crocking is used to convert long-chain hydrocarbon molecules (which are
heavy, high-boiling oils) into shorter-chain hydrocarbons to increase the
yield of lighter, lower-boiling products. There are three basic types of
cracking:'
(a) Thermal cracking which occurs under controlled conditions of high.
temperature and moderate pressure.
(b) Fluid catalytic cracking (FCC) which occurs in the presence of a
fluidired catalyst and under controlled conditions of high temperature
and pressure slightly above atmospheric.
. (c) Hydrocracking which occurs in the presence of a fixed bed catalyst in a
15
-------
* ALASKA
HAWAII
PUERTO
RICO
FIGURE 2. Petroleum Administration for Defense (PAD) districts
-------
hydrogen-rich vapor phase environment under controlled conditions of
high temperature and high pressure.
i
o Molecular Rearrangement Processes of many kinds are used to upgrade inter-
mediate refining products by reshaping their molecular structures'and to
increase the yield of gasoline components by recombining hydrocarbons of
three and four carbon atom chains into eight or nine carbon atom chains.
The molecular rearrangement processes include catalytic reforming, catalytic
alkylation, catalytic polymerization and catalytic isomerization.. .all of
which will be discussed in more detail later in this section.
o Desulfurization Processes are used to remove sulfur compounds from inter-
mediate and endproducts. There are two basic types of desulfurization:
i ' '
(o) Catalytic hydrodesulfurization which converts organic sulfur compounds
into hydrogen sulfide gas for subsequent removal by distillation or
stripping. Catalyt?', hydrodesulfurization occurs in the presence of a
fixed bed, catolysi in a hydrogen-rich vapor phase environment under
controlled conditions of temperature and pressure. This process also
converts organic nitrogen compounds into gaseous ammonia which is sub-
sequently removed along with the hydrogen sulfide, although the degree
of nitrogen removal is usually not as high as that of sulfur removal.
Catalytic hydrodesulfurization is sometimes referred to as "hydrotreating".
(b) Non-catalytic scrubbing or washing of gases or hydrocarbon liquids with
alkaline solutions for the removal of hydrogen sulfide and mercaptor-,, or
for the conversion of mercaptans into less undesirable forms of sulfur.
These so-called "treating" processes are used for scrubbing refinery gases
as well as for liquids. The alkaline scrubbing of liquids also removes
organic acids such as phenols, cresols and naphthenic acids which nay be
present.
The following discussions provide brief summaries of unit processes in each of
the above groups and identifies their wastewater sources.
Atmospheric Crude Unit
The crude oil entering o refinery contains a small amount of emulsified brine
ranging from Q.I to 2.0 volume percent on the crude oil, and the brine may contain up
to 25 weight percent salt (mostly sodium chloride) . The salt content of the brine
associated with the crude oil is usually expressed as pounds per thousand barrels of oil
(p.t.b.) and the p.r.b. value ranges from 10 to 250 (as sodium chloride, NaCI). The
sources of the emulsified brine are the naturally occurring br'nes associated with the
original oil field and, ,in the case of oil transported by .ocean -going tankers, contami-
nation by sea water ballast in those tankers which are not provided with segregated
ballast holds. ,
17
-------
Most refineries include crude oil desalters (which ore either chemical or
electrostatic de-emulsifiers) to remove the brine from the crude oil as it enters the
otmospheric crude unit for processing. The crude oil is heated to about 250-300 °F and
mixed with wash water (about 5 volume percent on the crude oil) to assist the desalting.
process. Typically, some caustic (NaOH) is also added to maintain a slightly alkaline
pH level during the desalting. .
The desalted crude is then further heated to about 700 F and distilled into
various boiling range fractions or "cuts1". The distillation is done at essentially a+mos-
pheric pressure in what is commonly referred to as the "atmospheric crude unit" or, more
simply, the "crude unit". The boiling range fractions produced ,in a crude unit are
called "virgin" cuts since they are the natural ly occurring components of the crude oil.
The variety of virgin cuts that may be produced In a crude unit is specific to each
refinery and its desired slate of endproduct*. However, as a broad generality, the
typical range of virgin cuts from a crude unit include:
o Propane and butane
o Light ond heavy naphthas (raw gasoline components)
o Jet fuels and kerosines
o Diesel oils
o Light and heavy gas oils
o Heavy residual oil
Some of the row virgin cuts must be stripped by injection of live steam in so-called "side-
cut strippers" to reduce their ignition flash points. The stripping steam is subsequently
' condensed and removed from the overhead reflux drum of the crude unit's main distil-
lation tower.
The two major sources of wastewater in a crude unit are the effluent brine from
, the crude oil desalter ond the condensed stripping steam from the side-cut strippers.
Vacuum Unit
The heavy residual oil from the atmospheric crude unit is reheated to about 750
°F and is then further distilled under vacuum conditions so as to yield additional virgin
bciling range cuts called vacuum gas oils.
t
Steam jet ejectors are usually used to achieve the required vacuum conditions.
To prevent thermal degradation or coking of the heavy residual oil when it is heated to
750 F, live steam is usually injected into the heater tubes as a diluent. Stripping steam
is also injected into the bottom of the bacuum distillation tower to maximize the removal
of volatiles. The steam used in the vacuum producing jet ejectors,' the heater tubes and
the bottom of the vacuum tower is subsequently condensed and removed from the overhead
product.drum of the tower. That condensed steam constitutes the major source of waste-
water generated in a vacuum unit.
18
-------
The residual cil from the vacuum distillation unit is a heavy tar which typically
has an atmospheric boiling point in excess of 300-1000 °F. In other words, all the virgin
hydrocarbons in the crude oi! with boiling points below 800-1000 °F are usually removed
in the two distillation steps represented by the atmospheric crude unit and the vacuum
unit. . .
Fluid Catalytic Crocking
Virgin gas oils produced from crude and vacuum'distillation units have boiling
ranges of about 600-1100 °F. Their molecules contain perhaps 18 or more carbon atoms .
and their molecular weights are in the range of 300 to 400. These gas oils.can be pro-
cessed and blended for endproduct use as industrial fuel oils. However, in many cases,
it is more desirable to crack at least some of the virgin gas oils into lower-boiling
molecules to increase the refinery yield of gasoline blending components.
Fluid catalytic cracking is one of the processes widely used to crack gas oils.
The conversion0 of virgin gas oils into lower-boiling materials by catalytic cracking is
typically within the range of 75-90 percent. Some typical yields from virgin gas oil
cracking are:
Gas 2-5 weight percent
Propane and propylene 8-12 volume percent
Butanes and butylenes 13-18 volume percent
Naphtha (gasoline) 60-65 volume percent
Gas oils 10-25 volume percent
Coke 4-8 weight percent
(the above yields are as percentages of the feed)
The catalytic cracking of gas oils occurs in a fluidized bed reactor under con-
trolled conditions of temperature and flow f,! pressures slightly above atmospheric and in
the presence of a fluidized catalyst. The coke yield deposits on the catalyst which is
continuously cfrr.u!c.'ed to a separate regenerator vessel where the coke is burned off of
the catalyst. The heot of coke combustion is particlly absorbed by the regenerated
catalyst which returns to 'he fluidized bed reactor and provides a major portion of the
required cracking reaction heat. The remainder of the heat requirement is supplied by a
fuel-fired feedstock preheater.
t
The reaction yield mixture, which is in the vapor phase, then flows through a
distillation tower where the yield mixture is cooled and distilled into its various boiling
I range cuts (see above 'rypical yields). The hydrocarbon cracking reaction also cracks
organic sulfur and nitrogen compounds contained in the virgin gas oil feedstock. Thus,
the gas yield typically includes hydrogen sulfide as well as some ammonia, cyanides and
thiocyanates.
*° Defined as the disappearance of gas oil boiling range material.
19
-------
Steam is used in the fluidized reactor to strip volatile* from the spent catalyst
end some stripping steam may also be used in the reaction yield distillation tower. The
steam is subsequently condensed and removed from the overhead drum of the distillation
tower. Since that condensation occurs in intimate contact with the gas yield, the con-
densed steam contains some hydrogen sulfide, ammonia, cyanides and. thiocyanates.
- The cracking reaction also produces monohydric and dihydrfc phenols and fh!o~
phenols. Since some of those pheno'ics boil in the same temperature range as the yield
naphtha, they are condensed along with the naphtha, gas and steam in the distillation
tower overhead drum. And since ths phenolic; are water soluble, a portion of the
phenols appears in the steam condensate from the overhead drum';
In summary, the steam condensate from the fluid catalytic cracking distillation
tower overhead drum is the major wastewater source from a fluid catalytic cracking unit
and it will usually contain:
o Hydrogen sulfide and ammonia in the form of ammonium hydrosulfide
o Cyanides and thiocyanates
o Phenols and thiophenols.
i
Hydrocro eking
Hydrocracking process units are used for essentially the same purpose as fluid
catalytic crackers, and that is to upgrade high-boiling gas oils by cracking them into
lower-boiling material*. However, hydrocrocking accomplishes both cracking and hydro-
ganotion. Thus, th& hydrocracking yield products are essentially saturated hydrocarbons
with high concentrations of isoparaffins and naphthenes.. .whereas the fluid catalytic
cracking products are essentially unsaturoted with high concentrations of olefins. "
The gas oil feedstock tc a hydrocracker is heated to 500-900 F and passed over
fixed beds of catalyst at pressures ranging from 700 to 3000 psig in a hydrogen-rich vapor
environment. By selection of the proper cctalyst and the proper operating conditions,
hydrocrackers may be designed so as to operate in either of these two optional product
slate modes (or to have rha flexibility of shifting from one mode to the other):
t .
o' A mode which maximizes the yield of naphthas (gasoline components within
the boiling range of 100-400 °D.
o. A mode which 'maximizes the production of middle distillates such as jet fuels,
1 kerosene, d>esel oil and light gas oil.
t . . .
:The vapor phase reaction products from the fixed-bed hydrocracking reactors are cooled
land partially condensed by exchanging heat with the incoming feedstock followed by air
or water-cooled exchangers. The reaction products are then reduced in pressure by
stage-wise flashing. Part.of the flashed, hydrogen-rich vapor is recompressed and
.recycled to the reactors. Makeup hydrogen is provided to replace the hydrogen
—, , ^
20
-------
consumed by the hydrogenation reactions. The remainder of the flashed vapor is subse-
quently processed to recover propane and butane. Water is injected continuously into
the reaction product heat exchange and cooling train to wash out any buildup of salts or
other foulants.
The liquid phase from the final flash stage is distilled in so-called stabilizers to
remove residual propone and butane for subsequent recovery. The stabilized liquid pro~
duct is then further distilled to provide the boiling range cuts consistent with the selected
product slate mods of operation (see above).
Because of the hydrogenation which occurs in hydrocracking, a high percentage
(as much as 85 percent or more) of the organic sulfur and nitrogen compounds in the
hydrocracker feedstock gas oils is converted to gaseous hydrogen sulfide and ammonia.
Under the high operating pressures.existing in hydrocrackers, the partial pressure of the
hydrogen sulfide and ammonia are such that they are largely absorbed into the wash
water used in the reaction product heat exchanger train. Thus, the sour wash water with-
drawn from the recctfon product flash stages constitutes the major source of wasfewater
from a typical hydrocracker. The ammonium hydrosulfide content of that wastewater may
be as high as 1-3 percent (10,000 to 30,000 ppm). That contrasts with the wastewaters
from a fluid catalytic cracking unit which rarely contain more than 3000-5000 ppm of
ammonium hydrosulfide. However, because of the hydrogenation reactions and because
very little if any free oxygen is present in a hydrocracker, the hydrocracker wostewoter
should not contain any phenols.. .nor is there any evidence of cyanides or thiocyanates
being present. ,
Thermal Crocking — Delayed Coking
Thermal cracking processes are thos« in which cracking is induced simply by the
proper selection of temperature and pressure -conditions without using any catqlyst or
hydrogen. Three types of thermal crocking ore in active use today: delayed coking,
fluid bed coking and visbreaking.
Coking is a more severe form of cracking than visbreaking. The feedstock is
usually a residual oil from a vacuum distillation unit and the purpose of coking is to con-
vert that oil completely into cracked vapor and solid coke. In the delayed coking pro-
cess, the residual oil feedstock is heated to about 900-950 °F under a pressure of about
20-60 psig and in the presence of some steam injected into the heater coils. Part of the
thermal cracking occurs in the heater coils and the remainder occurs when the partially
vaporized oil from the heater passes into very large vertical "coking drums". The com-
plete conversion of the oil into cracked vapor and coke occurs through successive '
cracking and polymerization in the coking drums. The coke deposits itself on the walls of
the coking drums. A delayed coker utilizes at least two drums, with one in dnstream
coking service while the other is being decoked by high-pressure water jets. The amount
of coke produced from some delayed cokers in U.S. refineries is as much as 1500 to 3000
tons per day, although there are many smaller units as well.
21
-------
The cracked vapor flows from the coking drums to a distillation tower where it
. is cooled and distilled into various boiling.range cuts (very similar to the distillation
tower in a fluid catalytic cracker). As with the fluid catalytic cracker, the coker pro-
duct gas and liquids are unsaturated with a nigh percentage of olefins.
i ( '
The steam injected into the feedstock heater as well as stripping steam used in
"the distil lot ion tower,(and any associated tide-cut strippers) is condensed and removed
from the distillation tower overhead drum. That steam condensate is one of the major
wastewdter sources from a delayed coker and it contains the same contaminants as found
in the wastewaters from a fluid catalytic unit: ammonium hydrosulfide, phenols and
. thiophenols, and cyanides and thiocyanates.
The periodic.decoking of a coke drum involves a depressuring or "blowdown"
step prior to opening the coke drum. The blowdown voppr is usually routed first into the
coker distillation tower and then into c circulating quench tower for cooling the vapor.
Depending upon the specific unit design, the circulating quench system may involve
water which might result in another source of oily, sour and phenolic wastewater. How-
ever, it is very difficult to generalize tKe various designs employed for quenching coke
drum blowdown vapors.
The high-pressure jet water used to drill the product coke from the coking drums
' is usually recovered for reuse as decoking water. If it is not completely recovered and
reused, it constitutes a major, potential source of wastewater contaminated with coke
particles, heavy oil, .tar and probably phenols.
Thermal Crocking — Fluid Bed Coking ; <
I Fluid bed coking serves the same function as delayed coking, which is to con-
vert the heavy residual oil from vacuum distillation into cracked vapor and solid coke.
I In a fluid bed coker, the residual oil feedstock is heated to about 950-1000 °r ,
(much the same as for o delayed coker). and then erters a reactor which has a bed of hot,
'fluidized coke. In the reactor, the residual oil feedstock is completely cracked into
; vapor and product coke. The product coke becomes part of the fluidized coke bed which
•Is continuously circulated to a coke burner vessel in which about 20-30 percent of the
.product coke is burned. The heat from the coke combustion is partially absorbed by the
,circulating, fluidized coke which returns to the reactor and provides most of the heat
'required by the cracking reactions. The net product coke is withdrawn continuously from
the coke burner vesvsI.
i ' '
' The cracked vapor from the coking reactor flows through a quench section to a
; distil lotion tower where it is cooled and dhtilled into various boiling range cuts (very
similar to the distillation tower in a fluid catalytic cracker). As with the fluid catalytic
•cracker or the delayed coker, the product gas and liquids from a fluid coker are unsatu-
rated and hove a high percentage of olefinic hydrocarbons.
_}
i .
22
-------
Steam is injected into the bottom of the reocijr to strip volatile? from the
circulating, fluidizea coke. That steam as well as any stripping steam u;ed,in the distil-
lation tower (and any associated side-cut strippers) is condensed and removed from the
distillation tower overhead drum. The condensed steam is the major source of wastewater
from a fluid bed coker and it contains the sane contaminants as found in the wastewaters
from fluid catalytic cracking or delayed coking: ammonium hydrosulfide, phenols and
rhiopHenols, aid cyanides and thiocyanates.
There is o variant of the fluid coking process, known as Flexicoking, which
utilizes an integrated coke gasifier to produce refinery fuel gas from the net coke pro-
duct withdrawn from the coke burner. Very few, if any, Flexicoking units have been
installed in U.S. refineries as yet.
As a matter of interest, all of the delayed cokers in U.S. refineries are pro-
cessing o total of about 940,000 barrels per day of residual oil and producing about
40,000 tons per day of coke. In contrast, fluid bed cokers in U.S. refineries ore pro-
cessing a total of about 70,000 barrels psr day of residual oil and producing about 2500
tons per day of coke .
i '
Thermal Crocking — Visbreoking ,
i
Visbreaking is a relatively mild thermal cracking process. The purpose of vis-
breaking is, to convert high viscosity residual oils (from either atmospheric or vacuum
distillation units) into a lower viscosity fuel oil while minimizing the amount of naphtha
formed.
The residual oil feedstock is heated to the range of 850-950 F and mildly
cracked in a fired heater. The heater effluent is quenched and distilled into the desired
boiling range cuts. A typical range of cuts includes gas, naphtha,. light fuel oil and
heavy fuel oil.
Steam used in the cracking heater and the distillation tower side-cut strippers is
condensed and removed from the distillation tower overhead drum. That steam conden-
se te is a wastewater source containing sulfides and phenols.
In some cases, the heavy fuel oil from the bottom of the distillation tower is
vacuum distilled to recover additional light fuel oil. In those cases, there may be
another wostcwoter source originating with the required vacuum-producing steam jet
ejectors. '
There are less than a dozen visbrealcers still operating in U.S. refineries and
many of those are in, relatively small refineries.
> i
Steam Cracking (Olefins Production)
Steam crocking of hydrocarbons to produce ethylene and other olefins is a
1 i, ,
- i' • 23
-------
thermal cracking process which can be operated across a wide range of cracking severity
• on feedstocks ranging from gaseous ethane to heavy gas oils from vacuum distillation.
Steam cracking produces ethylene, propylene, butadiene, aroma tics and other by-
products, all of which are used as precursor feedstocks in the manufacture of a broad
spectrum of petrochemicals, plastics, synthetic rubbers and polymers. Since the primary
product is considered to be erhylene, steam crackers are often referred to as "ethylene
crackers". - -• - "
Steam cracking of light hydrocarbon feedstocks such as ethane, propane end
butane is less costly than steam cracking heavier feeds such as gas oils. The lighter feeds
also provide a higher yield of the primary product ethylene. For example, steam
cracking of ethane provides about an 85 percent ethylene yield whereas steam cracking of
gas oils provides about a 22-26 percent yield of ethylene. For that reason, the largest
part of the U.S. ethylene production is from light feedstocks. However, the availability
of light hydrocarbon feeds for ethylene production is becoming increasingly difficult and
• gas oils can be expected to assume a larger percentage of the feedstock roie in the
future.
As a brief perspective of the U.S. ethylene production picture (as of 1976):
Percent of total U.S. Number of
ethylene production . steam crackers
Distribution by industry:
, . 13 chemical companies 53 33
10 petroleum companies '47 18
-. Distribution by feedstock type;
Butane and lighter hydrocarbons 81.3
j Naphthas . ] 7.8
j Gas oils j 10.9 ,
| ! , . .
I As of January 1979, there were steam crackers in 14 U.S. refineries using gas oil feed- .
'stocks totalling to about 150,000 barrels per d-ry. Assuming that those units were
achieving 22-26 percent yields of ethylene, their total annual ethylene production would
be about four billion pounds per year which amounts to perhaps 12-13 percent of the
•current national annual ethylene production. ,
i • !• • ' i
j The feedstock fo a steam cracker is heated and cracked, in the presence of a
1 considerable amount of diluent steam, In tubular pyrolysis furnaces. The cracked pro-
ducts exit the furnaces at temperatures in the range of 1500-1700 °F and are rapidly
.quenched to prevent excessive cracking. The quenched vapors are further cooled in a
i distillation tower which removes the product fuel oil from the cracked vapors. The
•cooled vapors are then comp id to about 500-600 psig in multistage compressors with
'interstage cooling. Liquids comdensed interstage are removed and processed for recovery
of liquid products. The compressed gases are treated for removal of acid gases (hydropin
sutfide and carbon dioxide), dried, refriaerated nnd fractionated for recovery of aas
—*»»i i • n
1
24
-------
products.
A fairly typical product yield distribution when steam cracking gas oils is:
weight percent
on feed
• - - • Hydrogen-rich gas 1.5
Methane-rich fuel gas 10.5
Total gas 12.0
Ethylene ' 26.2
Propylene 14.6
Butanes and butylenes 5.2
Butadiene 5.1
Total C4's 10.3
Benzene-toluene-xylene 11.1
Non-arornatics , 7.8
Total gasoline 18.9
Fuel oil 18.0
Total products 100.0
i
The design configuration, cracking severity and feedstock type vary quite con-
siderably from unit to unit. It is therefore very difficult to be specific about the waste-
water sources from steam cracking. However, as with any thermal cracking process in
the presence of steam, there will be a sour, phenolic wastewater generated. In some
designs, that wastewater may be reused (at least to some extent) as boiler feedwater for '
steam generation. There may also be these additional sources of wastewater generation:
(a) spent caustic solution from the caustic washing of the cracked vapors at some inter-
, stage point in the compression train, and (b) an alkaline wastewater from the acid gas
removal system. ,
Interim Summary
i The unit processes discussed thus far are all within the first rwo major groups of
; refining processes and technologies, namely distillation and cracking processes.
. Functionally, those two groups involve: - < '
[' o Distillation of a refinery's crude oil feedstock to separate the crude oil into
i its component fractions or bo'ling range cuts.
o Cracking the long-chain crude oil fractions (those which have boiling ranges
higher than that of diesel oil) into shorter-chain gases and liquids.
However,'the gases and liquids derived from distillation and cracking are only inter-
...,mediate products which must undergo further processing to optimize the yield and quality
25
-------
of a refinery's endproduct slate. For example:
o Most of the intermediate products must be desulfurized either to meet end-
product sulfur specifications or to make them suitable for further processing.
o Many of the intermediate naphthas can be improved by reshaping their
molecular strjcture to increase their gasoline octane rating.
o Ths aromatics content (benzene, toluene, xylene) of many of the intermedi-
, ate naphthas can also be increased by molecular reshaping, which increases
the refinery's potential for producing petrochemicals.
o The short-chain propylene and butylene products from cracking can be con-
verted to longer-chain,'high octane gasoline components.
All of the processes summarize^ in the remainder of .this section are either desulfurization
or molecular rearrangement processes.. .in essence, processes which upgrade the inter-
mediate products from distillation and cracking.
- i
Catalytic .Reforming
i
The spark-ignited, internal combustion engine used in most automobiles performs
best with a gasoline fuel which, bums smoothly rather than exploding. If the fuel
explodes, the engine is heard to "knock" or "ping" and the engine performance deterio-
rates. The molecular structure of a gasoline determines its burning characteristics. Gas-
oline molecules of branched paraffins (isoparaffins) or. saturated rings (oromatics) bum
more smoothly than straight-chain paraffins (normal paraffins) or unsaturated rings
(naphthenes). Thus, the anti-knock quality of gasoline is increased by increasing its
isoparaffinicity and its aromaticity. The physical scale of antiknock quality is
measured by the "octane number" which is based on the iso-octane molecule having on
.octane number of 100. Thus, a gasoline with a 95 octane number rating has an anti-
knock quality that is 95 percent as good as the anti-knock quality of iso-octane.
Hydrocarbon molecules in the gasoline Lolling range of 100-400 F have five to
eight carbon atoms. As an example of the octane number of such molecules:
Carbon Octane
Structure atoms . number*
Pentane normal paraffin 5 62
Cyclopentane naphthene 5 85
2-methylbutane isoparaffin ,5 90
Hexane normal paraffin 6 26
Cyclohexane naphthene . 6 77 '
2,3-dimethylbutane isoparaffin 6 94
Benzene aromatic 6 115
26
-------
Corbon Octone
Structure otoms number*
Heptane normal paraffin 7 0
3,3-dimethylpentane isoparaffin , 7 87
Toluene aromat':- 7 104
2,2,4-trimethylpentone** isoporoffin 8 100
o-Xylene aromatic > 8 120 <
(* motor method; ** iso~octane, the octane number base)
As con be seen cbove, the octane numbers increase as the molecular structure for a given
number of carbon atoms changes from pq/offinic or ;iaphthenic to isoparaffinic or
'aromatic.
The octane numbers of the virgin naphthas from a crude unit depend upon their
PON A analyses and will vary from one crude source to another. However, in most cases,
their octane numbers are not high enough to satisfy today's gasoline requirements. The
octane numbers of catalyticallv cracked or thermally crocked naphthas are relatively
good, but they can still be upgraded significantly. The same is true for the heavier
portion of hydrocracked naphthas.
In most refineries, the primary purpose of the catalytic reforming process is to
upgrade the octane numbers of virgin and cracked naphthas by increasing their iso-
paraffinicity and oromaficify. In some cases, catalytic reforming is used to increase the
yield of aromotics for petrochemical usags as well as for gasoline octane improvement. '
The naphtha feedstock to a catalytic reformer must first be desulfurized to very
low levels, a few pom or less, -to avoid deactivation or "poisoning" of the reforming
catalyst. The naphtha is then vaporized and heated to the; range of 900-1000 F under
pressures of 150-500 psig and passed through fixed catalyst beds in a hydrogen-rich vapor
environment. Although a number of reactions occur during reforming, the predominant
ones are the dehydrogenation of naphthenes, to form aromotics and the branching of
paraffins to form isoparaffins. The dehydrogenation reactions make the reformer a net
producer of hydrogen.. .in other words, a source of hydrogen for other refining processes
which consume hydrogen. . '
The catalytic reaction products are cooled, partially condensed and separated
into gas and liquid. A portion of the hydrogen-rich gas is recompressed and recycled to
the reactors. The remainder enters the refinery's hydrogen supply system for use in other
process frits. The reaction product liquid is then fractionated to remove propane and
butane for subsequent recovery. The final "reformate" gasoline 5s sent to storage as a
blending component of the refinery's endproduct gasolines.
Since the naphtha feedstock to a catalytic reformer is desulfurized and water-
free, and since no steam injection or water washing is used in a reformer, there is
*
27
-------
•usually no source of wastewater in a refonner.
Catclytic Hydrodesulfurizotion ~~ Naphtha
Catalytic hydrodesulfurization is used to remove so I fur from virgin and cracked
naphthas0 which ore to be catolyticolly reformed. The catalytic hydrodesulfurization ,
process convert* organic sulfur into gaseous hydrogen sulfide which is then removed by
stripping cr distillation.
'The naphtha feedstock to a catalytic hydrodesulfurizotion unit is vaporized and
heated to *he range of 400-700 °F under pressures of 200-500 °F and passed through
fixed catalyst beds in a hydrogen-rich vapor environment. Hydrogenation is the predom-
inant reaction that occurs. Thus, organic sulfur and nitrogen in the naphtha feedstock are
converted to gaseous hydrogen sulfide and ammonia. Olefiriic hydrocarbons in the
naphtha feedstock are also hydrogenated or saturated to some extent. The hydrogen con-
sumed by rhe hydrogenation reactions is usually supplied by the excess hydrogen produced
in the subsequent catalytic reforming of the naphtha. ,
The reaction products are cooled, partially 'condensed and separated into gas
and liquid. The hydrogen-rich gas is recompressed and recycled to the reactors, along
with makeup hydrogen (usually from a reformer) to replace the hydrogen consumed in the
hydrodesulfurizatioo reactions. The reaction product liquid Is then stripped *o remove
hydrogen sulfide, ammonia and other gases. After distillation to remove propane and
butane for subsequent recovery, the hydrodesulfurized naphtha is sent to a catalytic
'reformer.
As noted earlier herein, the primary purpose of a naphtha hydrodesulfurization
unit is to remove sulfur down to levels of no more than a few pom so as to prevent
catalyst poisoning in the subsequent catalytic reforming of the naphtha. The same is true
of water in that the naphtha feed to reforming must be essentially completely free of
water. Any water in the desulfurized naphtha after stripping would be withdrawn from the
overhead drum of the distillation tower which removes propane and butane. That over-
head water is a source of sour wastewater from most naphtha hydrodesulfurization units.
If wash water is injected into the reaction product heat exchange train (similar to the
operation described earlier herein for ca'rc'/tic hydrocrackers), then that water will
absorb hydrogen sulfide and ammonia and constitutes another source of sour wastewater
from the hydrodesulfurization process. If the hydrodesulfurization feedstock includes
catalytic or thermally cracked naphthas, there is some possibility (although slight) that
° .With the exception of hydrocracked naphthas which are thoroughly desulfurized during
their formation in the hydrocracker.
° Virgin naphthas would not contain olefins since crude oils rarely,include olefinic
hydrocarbons. Nor would hydrocracked naphtha have any olefins since they would
have been hydrogenated in the hydrocracker. Catalytic and thermally crocked
naphthas would have high olefin contents.
28
-------
the hydrodesulfurizotion soor wastewater may contain some phenols.
Cotolytic Hydrodesulfurization — Distillates
Catalytic hydrodesulfurization is often used to remove sulfur and nitrcgen from a
variety of virgin and cracked distillate oils in the boil inn range of jet fuels, kerosines,
diesel oils ond light fuel oils in order to satisfy the pertinent market specifications for
those products. If the distillate oil feedstock contains crocked, olefinic material, then
catalytic hydrodesulfurization can also be used to hydrogenate or saturate a portion of the
olefins to improve the color, odor and oxidation stability of the product.
The catalytic hydrodesulfurization process for distillate oils is very similar to
that for naphthas except that, in general, the distillate oil process requires a higher
reaction temperature and pressure and consumes more hydrogen.
The sour wastewater sources from a distil lota hydrodesulfurization are essentially
the some as from a naphtha hydrodesulfurization unit.
Cotolytic Hydrodesulfurizotion — Heavy Oils
Catalytic hydrodesulfurization units may also be used in refineries to
desulfurize and denitrify heavy oils such as:
o Gas oil feedstocks to fluid catalytic crackers, which is considered to improve
the performance of the catalytic cracking unit.
o Gas oil feedstocks to hydrocrackers, to provide a prior removal of sulfur and
. . nitrogen.
o Heavy gas oils and residual oils (from both atmospheric and vacuum distil-
lation units) to produce low-su!fur industrial fuel oils.
The catalytic hydrodesulfurization process for heavy oils is similar to that for naphthas
and for distillate oils except that the heavy oil process, in general, requires a higher
reaction temperature and pressure and it consumes more hydrogen., In fact, the catalytic
hydrodesulfurization process for very heavy oils requires temperatures and pressures as
high or higher than those required in catalytic hydrocracking (500-900 °F and 700-3000
psig).
The sour wastewater source* from a heavy oil hydrodesulfurizotion unit are the
same as from a naphtha or distillate oil hydrodesulfurization unit.
All of the various catalytic hydrodesulfurizatior processes for naphthas, distil-
lates ond heavy oils are ofen referred to as "hydrotreoters".
29
-------
Gos Recovery Plants
Almost all of the refinery unit processes yield a byproduct gas. In general,
olefin-rich gases are produced by thermal and catalytic cracking while saturated
(paraffinic) gases are produced by hydrocracking, catalytic reforming and hydrotreating.
The typical byproduct gas components and their sources are:
Hydrogen -- ,
Catalytic reformers and hydrogen synthesis units are the major sources'of
hydrogen in most refineries. Other sources of hydrogen production are thermal and
catalytic cracking.0
Hydrogen sulfide — • .
Hydrogen sulfide, derived from the cracking and hydrogenation of organic
sulfur contained in crude oil, appears in the byproduct gases from thermal cracking, ,
catalytic cracking and hydrotreating. A small amount of free hydrogen sulfide may also
be distilled from the crude oil in an atmospheric crude unit.
Ammonia ~
Ammonia is derived from the cracking and hydrogeiation of organic nitrogen ,
contained in crude oil. Because of its very high solubility in water, a large part of the
ammonia appears in the wastewoters from the various process units rather than their by-
product gases. If both ammonia and hydrogen sulfide are present, the ammonia appears
in the wastewater as ammonium hydrosulfide.
Carbon dioxide —
A small amount of carbon dioxide appears in the byproduct gas from fluid
catalytic cracking, derived by entrainment of carbon dioxide in the fluidized catalyst
circulated from the fluid catalytic cracking coke~burr,!ng regenerator to the fluid
< catalytic cracking reactor. Carbon dioxide also appears in the byproduct gas from fluid
coking for much the same reason.
Methane and ethane —
The byproduct gases from atmospheric and vacuum distillation cf crude oil
contain small amounts of methane and ethane, some of which is virgin gas distilled from
the crude oil and some of which results because of a slight amount of thermal cracking
which occurs in the distillation heaters. All of the cracking processes (thermal,
catalytic and hydrocracking) cs well as catalytic reforming yield methane and ethane in
varying amounts.
° Hydrogen also appears in the byproduct gases from hydrocracking and hydrotreating.
However, those processes ore not hydrogen producers. The hydrogen appears only be-
cause it is brought in to maintain the required hydrogen-rich environment for those
" processes.
? ' V 30
-------
Ethylene —
Ethylene is on olefin and is yielded by all of tr<3 various thermal cracking pro-
cesses (delayed and fluid coking, visbreaking, sfeam cracking) and by fluid catalytic
cracking.
Propane, normal butane and isobutcne —
Any virgin propane and butanes in the crude oil appear in th« byproduct gos
from an atmospheric crude unit. Thermal cracking, catalytic cracking and hydrocracking
all yieid propane and butanes in varying amounts, as does catalytic reforming. In
general, the propane yield from the various processes appears almost completely in their
byproduct gases. However, sotr>e of the butane yield may be retained !n the naphtha
products from the processes. In orSer words, the stabilization (distillation) of the liquid
naphthas to remove propane and butane usually removes all of the propane but leaves a
small amount of butanes in the naphthas.
Propylene, normal butylene, !sobuty!«nes and butadiene —
These are all olsfins and they are yielded by all of the cracking processes,
either thermol or catalytic. None of the hydrogen-environment processes (catalytic
reforming, hydrotreating, hydrocracking) produce any o'lefinic gases. [Thus, although
thermal and catalytic cracking produces both olefinic and saturated gases, the hydrogen-
environment, processes produce only saturated gases.] In general, the propylene yield
from the various processes appears almost completely in their byproduct gases. However,
some of the butylene and butadiene yield may be retained in the naphtha products from
the processes depending upon the degree to which the liquid naphthas are stabilized.
Normal pentane, isopentanes and pentylenes —
These are light hydrocarbons with five carbon atom chains and boiling within the
range of 70-80 °F at atmospheric pressure. Under pressures slightly above atmospheric
and temperatures of about 100 °F, these hydrocarbons are largely recovered as part of the
various refinery naphthas. Small amounts of virgin normal 2nd isopentane may appear in
the byproduct gas from crude oil distillation units depending upon rhe operating con-
ditions in the distillation tower overhead drum. Some normal and isopentane may also
appear in the byproduct gas frorrt most of the refining processes discussed herein depending
upon their operating conditions for gasHiquid separation and upon their naphtha distil-
lation conditions and sequence. The same is true of the olefinic pentylenes except that
they will appear only in the byproduct gases from cracking processes, either thermal or
fluid catalytic. Psntylenes are not produced by any of the hydrogen-environment pro-
cesses.
The purpose of refinery gas recovery plants is to collect the byproduct gases
frotr. all of the unit processes and to separate them into various products. For example:
7 31
-------
Separated products from
gas recovery plant
Acid gases (H2$, CO2)
Refinery fuel gas_(H2, Cl, C2)
C3 LPG (C3, C3~)*
C4, LPGJIC4, C4, iC4~, C4=,
C4-)*
Light naphtha (iC5, C5,
Components of
combined byproduct
gases from refinery
unit processes
Name Symbol
Hydrogen H2
Hydrogen sulfide ^S
Amrionio , KH3
, Carbon dioxide CGi
Methane Cl
Ethane C2
.Ethylene C1T .
Propane C3_
Propylene C3~
Butane C4
Isobutane , iC4_
Butylenes C4~
Isobutylenes iC4~
Butadiene C4==
Pen tone C5
Isopentane 5C5
Pentylenes C5=
(* LPG is on acronym for "liquefied petroleum gas". The LPG cuts
from a gas recovery plant may be marketed as an endproduct or
further processed to produce naphtha.)
In a typical gas recovery plant, the byproduct gases from the various unit pro-
cesses are compressed and then processed through an absorption system followed by a
sequence of distillation towers.
The sour gas from the absorption system (the refinery fuel gas containing
hydrogen sulfide and carbon dioxide) is then treated or scrubbed with an amine solution
which absorbs and removes the acid gases (hydrogen sulfide and carbon dioxide) from the
refir.sry fuel gas. The acid gases are subsequently distilled and removed from rhe amine
solution, and the regenerated solution is recirculated for reuse in scrubbing additional,
sour pasJ The C3 and C4 LPG cuts from the distillation towers are also scrubbed with
amine for hydrogen sulfide removal. The omine treating of hydrocarbon gases and liquids
for removal of hy'drogen sulfide is often referred to as "sweetening". The acid gases from
the amine regeneration (distillation) are subsequently processed for conversion into by-
product sulfur. , .
The sweetened refinery fuel gas is used as fuel in the refinery process heaters and
steam-generating boilers. The LPG cuts may be marketed as endproduct fuels or further
processed for conversion into high-octane gasoline components (as discussed later herein).
32
-------
After being further treated by alkaline scrubbing for removal of organic sulfur compounds
known as mercaptans (as discussed later herein), the light naphtha is used as a blending
component of the refinery's endproduct gasolines.
Many refineries include two gas plants, one of which handles predominantly
saturated or poraffinic gases and one of which handles the predominantly olefinic gases.
This permits the refinery to segregate its LPG cuts and light naphthas into those which are
saturated and those which are olefinic or unsaturated.
In general, the wastewater sources from gas recovery plants fall into three
categories:
o Water condensed from the incoming gases by compression interstage and after-
stage cooling. The water from saturate gas plants will contain hydrogen
sulfide and ammcnia in the form of ammonium hydrosulfide. The water from
unsaturote gas plants will contain cyanides, thiocyanates and perhaps some
phenols in addition to ammonium hydrosulfide.
o Water removed from the overhead drum of the depropanizing tower in the
distillation sequence. The water will contain hydrogen sulfide and perhaps
traces of mercaptans.
o Water discharged from the amine scrubbing and regeneration system which will
contain'hydrogen sulfide and some of'the amine. Typically, the amine in
refinery usage is DEA (diethanol amine). The amount of wastewater from this
source varies widely with the design of the amine treater, bit it is generally
fairly small.
Merox Treating
Mercaptans are a form of organic sulfur (denoted as RSH) found in many refinery
streams. Merox treating is one of the many processes used for either removing
mercaptans from liquid hydrocarbons or for converting them to disulfides (denoted as RSSR).
Many of the treating processes use a caustic solution. Merox is a widely practised
treating process which uses caustic, mo thy I alcohol and air in the presence of a catalyst.
The Merox process may be designed for mercaptan extraction (removal) or for sweetening
which is the conversion of the mercaptons to disulfides.
The wastewater sources from a Merox unit are the spent caustic discharged as
well as spent wash water (when a final water wash is included in the design). Those
wostcwaters will contain sodium mercaptides (NaSH) derived from any hydrogen sulfide
which might be present in tne Mere* feedstocks. The wastewaters from Merox treating of
olefinic naphthas (from thermal crackers or fluid catalytic cracking units) will also conr
fain sodium phenolates.
33
-------
' Alkylotion Process
The olkylotioo process combines isobutane with propyleno and/or butylenes to
produce a very high-octane gasoline component known as ''alkylate". The alkylation
reaction occurs in the liquid phase and in the presence of either a sulfuric acid (H2$O4)
or hydrofluoric acid (HF) catalyst.
i»~ '
One of the reasons for segregating the production of C3 and C4 cuts by using
separate saturate and unsaturate gas plants is to facilitate the subsequent alkylation of the
unsaturated propylene and butylenes with saturated isobutone.
The feedstock C3 and C4 cuts to an> alkylation unit must first be treated for
removal of hydrogen sulfide and mercaptans (by amme treating followed by caustic or
Merox treating). In addition, the unsaturated C3 and C4 is usually processed through a
catalytic hydrogenation unit to convert diolefinic butadiene into mono-olefinic butylene
so as to avoid excessive consumption of the acid catalyst in the alkylation reaction. The
pretreated liquid feedstocks are cooled and thoroughly mixed with liquid acid (a sulfuric
acid alkylation reaction requires a refrigerated reaction zone). The reaction effluent is
separated from the acid and distilled to provide these products:
o Unreocted isobutane which is recycled to the reaction .zone.
; o Unreacted propone and normal butane which constitute an LPG endproduct.
o High-octane alky late gasoline.
The reaction products from both sulfuric acid and hydrofluoric acid alkylation
require the removal of entrained, residual acid. The exact mode of removal varies
between the two processes, but they both generate wastewaters which contain traces of
acid or acid salts (if caustic is used to neutralize the acid).
t , : ' ';
' The spent acid from alkylation is carefully segregated and made available for
regeneration in separate facilities. Normally, the spent acid does not enter the refinery
wastewoter system.
' A hydrofluoric acid alkylation unit must be very carefully designed to prevent
any escape of the extremely hazardous hydrofluoric acid. This is not to say that a
sulfuric acid alkylation unit does not require careful design as well, but hydrofluoric acid
is relatively much more hazardous than sulfuric acid.
Catalytic Polymerization (
i The catalytic polymerization process converts propylenes or propylenes and
butylenes into C6 or C7 dimers (isohexenes and isoheptenes) which are within the gasoline
boiling range and have good octane ratings. Some trimerization to C9 isononene*. also
occurs. The process is sometimes referred to as "dirnerizotion", and the endproduct
''"gasoline is referred to as "cat poly genc'Ine" or "dimer gaoline". The catalyst usually
i ~ ' . " ~ " ' . . . . ^
-34 ,
-------
•used in the process is phosphoric acid.
The sources of wastewater from a catalytic polymerization unit are spent caustic
and spent wash water which are used to remove entrained acid from the reaction effluent.
The wostewaters wit! contain some acid or acid salts.
Isomerizotion — Butane
i
Butane isomerization is a process which converts normal butane to isobutane. It
is used to supply additional isobutane for those refineries which lack sufficient isobutane
to maximize the alkyla'ion of their available propylena and butylenes.
Butane isomerization occurs over fixed catalyst beds at temperatures of 200-400
°F under pressures of 200-500 psig and in a hydroge.-.-rich vapor environment. The
reaction effluent is cooled, partially condensed and separated into gas and liquid. The
hydrogen-rich gas is recompressed and recycled to the reaction zone along with makeup
hydrogen. The liquid is distilled to remove dissolved hydrogen as byproduct gas and to
recover unreacted normal butane for recycle to the reaction zone.. The product isobutane
is sent to storage or directly to an alkylatfon unit.
i
Isomerization — Pentane and Hexone
Catalytic isomerization is also used to upgrade the octane rating of light,
paraffinic naphmas which are rich in normal pentane and normal hexane by converting
them into isopentane and isohexane. The process is almost identical to that described
above for the isomerization of butane. The endproduct isopentane and isohexone is often
referred to as "isom gasoline" or "isomerate".
Neither the butane nor the pentane and hexane isomerization processes produce
any wastewaters under normal operation.
i
I
Hydrogen Synthesis
As noted earlier herein, the two major sources of hydrogen in most refineries are:
(1) the catalytic reformers used for upgrading the octane numbers of naphthas and (2)
hydrogen synthesis units. Of the two sources, catalytic reforming is by far the largest:
,As of January 1979 , there were 36 refineries with hydrogen
synthesis units producing on aggregate total of about 1.7
billion SCFD (standard cubic feet per day) of hydrogen. At
the sama time, there were 204 refineries with catalytic
reformers processing an aggregate total of 3.8 million barrels
per day of naphthas. Assuming an average hydrogen yield of
900 SCF per barrel of naphtha processed, the aggregate total ,
hydrogen produced by catalytic reforming was 3.4 billion SCFD
...or twice that produced by hydrogen synthesis units.
i ' •
* .' • 35 .
-------
The various hydrogen synthesis processes in refinery usage include steam re~
forming of methane, stearr. .forming of naphtha and partial oxidation of residual oils. As
of January 1979^, steam reforming of methane accounted for the largest percentage of
hydrogen synthesis:
Steam reforming of methane 81 .2 percent
Steam reforming of naphtha 7.0 percent
Partial oxidation 6.0 percent
Others 5.8 percent
i •
The steam reforming of methane-rich refinery gas or natural gas to produce
hydrogen involves four steps:
o K^ form ing of methane and steam into hydrogen and oxides of carbon at
temperatures of 900-1600 °F under pressures of 300-600, psig within the
catalyst-filled' tubes of a fired heater. The main reaction which occurs is:
. , CH . + H-0 * 3H, + CO : . .. .
A t. £.
o Catalytic conversion or "shifting" of the carbon monoxide (yielded from the
reforming reaction) into carbon dioxide as in this reaction;
CO + HJD -» CO. + H0
1 i i . i
o Cooling the shift reaction effluent gas and removing the carbon dioxide by
scrubbing with monoethanol amine (MEA) or an equivalent acid gas
absorbent.
o Catalytic conversion or "methonotion" of any residual oxides of carbon into
methane:
CO + 3H2 •* CH4 + HjO : !
CO2 + 4H2 -» CH4 + 2H20
, Theoretically, the combination of the reforming step and the "shift" conversion
step requires two moles of steam for each four moles of hydrogen produced:
That amounts to a theoretical consumption of about 24,000 pounds of steam per million
SCF of hydrogen produced. In actual practice, the steam fed to the reforming reaction is
'about twice the theoretical consumption. When the shift reaction effluent gas is cooled
just ahead of the monoethanol amine scrubbing, the excess steam is condensed and
removed. In o steam-methane reforming unit producing 50 million SCFD of hydrogen, the
.condensate removed is about 100 gpm. The condensate will be saturated with carbon
dioxide and, if the feedstock contains nitrogen, some ammonia and cyanide may appear
ln the condensate. Many hydrogen synthesis plants are designed to reuse the condensate •
i
36
-------
as boiler feedwater for steam generation in the plants' waste heat recovery systems. If
not reused, the condensate constitutes the major wastewater source in a hydrogen
synthesis plant. ;
Aromatics Extraction
As disai-i&d previously, the naphtha produceo by catalytic reforming has a high
content of cromati'ci. In some refineries, the catalytic reformers are operated at con-
ditions designed specifically to increase the yield of benzene,, toluene and xylenes (which
are the aroma tics commonly referred to as BTX) for extraction as endproduct petro*'
chemicals. In fact,'as of January 1979 , 33 refineries in the U.S. included extraction
units producing an aggregate total of about 196,000 barrels per day of BTX.'
There aria mony solvent extraction and extractive distillation processes for the
removal and recovery of BTX from catalytic reformate. Typical BTX recoveries achieved
in such processes or« 99.9 percent for benzene, 99.5 percent for toluene ar.d 98 percent
for xylenes. Some of the solvents used for BTX extraction are:
o Tetramethylene sulfone (referred to as sulfolane)
o Tetraethylene glycol
o Morpholine derivatives such as formyl-morpholine. , ,
The non-aromarics or "raffinate" from BTX extraction units is used either as an endproduct
gasoline component or as feedstock for steam crackers (olefins production plants). In some
cases, the roffinate may be marketed as a feedstock for gasification plants producing
synthetic natural gas (SNG).
* Xylene has three isomeric forms known as ortho-, meta- and paro-xylene. Most
of the endproducr market is for ortho~ and para-xylene. There is little market, if any,
.for meta-xylene. Thus, many BTX extraction units are coupled with xylene isorr.erization
units which are catalytic processes designed to isomerize meta-xylene and increase the
yield of ortho- and para-Scylene.
'Many of the BTX extraction processes include washing of the BTX extract und the
raffinate streams with water to recover entrained solvent. The solvent-rich water is then
reprocessed to recover both the solvent and the water for reuse. Thus', in normal bpei—
at ion, BTX extraction does not generate a wastewater. Any liquid losses from an
extraction unit (sample drains, pump and valve drips, maloperation, etc.) would con-
taminate the refinery wastewater sewers with the extraction solvent ond/or the BTX
aromatics.
: I
i '
Petrochemicals ;
Many refinerios produce a range of petrochemicals among their endproducts. As
'discussed earlier, some refineries include the steam cracking of gas oils which produces
37
-------
petrochemicals such as ethylene, propylene, butadiene and BTX aromatics. And some
refineries extract BTX aromatics from catalytic reformate. Many other petrochemicals are
also produced in some refineries, with the more common ones being:
. o Ethylbenzene — produced by the catalytic reaction of benzene and ethyler.e.
o Styrene monomer — produced by the catalytic dehydrogenation of ethyl-
benzene.
i o Cumene'-- produced by the catalytic reaction of propylene and benzene.
o Cyclohexone — produced by the catalytic hydrogenotion of benzene.
In addition to those above, a very broad spectrum of oiher petrochemicals are manu-
factured from feedstocks derived by the refining of petrolsum crude oil. Almost every
major oil company now has «.• petrochemical manufacturing division. Mary oil companies
have also expanded into the large-scale manufacture of ammonia and other agricultural
chemicals. In many cases, the oil company's petrochemical plants are adjacent to their
, major refineries and it is very difficult to define the exact line of demarcation between
their petroleum refining operations and their petrochemical manufacturing.
There are far too many permutations of products and processes to generalize as to
the sources or composition of wastewaters from petrochemicals production. Each
individual refinery's petrochemical operations m*j$t be evaluated on a specific case-by-
case basis.
[ The reader who wishes to gain a perspective on the range and extent of the
petrochemical industry should read the excellent series of articles by Hatch and Motor .]
Lubricating Oils ,
' About 40 U.S. refineries are currently producing lubricating oils with an aggre-
gate output of approximately 230,000 barrels per day. The specifications for lubricating
oils vary considerably and include endproducts for:
o Automotive uses:
(a) Engine oils and transmission fluids
(b) Brake and shock absorber fluids
o Marine and heavy-duty industrial engines
o Special uses: • '
(o) Transformer oi Is
(b) Refrigeration system oils '
(c) Heat transfer oils . .
(d) Metal-working oils \
(e) Agricultural spray oils ; ,
The key properties of a lubricating oil which must be "tailored" to produce the
38
-------
•^specifications needed for each of the above endproduct uses are:
', o The fluidity of the oil as defined by its viscosity.
o The effect of temperature on the oil's viscosity as defined by its viscosity
index (VI) (the less the viscosity changes wfth temperature, the higher or
better is the viscosity index). •
o The lowtemperatc're'usefulness of the oil as measured by its pour point
temperature which is a function o' iti wax content (the higher the wax
content, the higher the pour point temperature ct which the oil congeals and
will not pour).
i , In general, the bosestocks from which a refinery produces the lubricating oils
are: • '
1 o Vacuum gas oils (with bciling ranges of 650-1000 F) produced from the
, refinery's vacuum distillation unit.
"f. - ----- - ,—.-.- . •.; .
o Vacuum residual oils (boiling above 1000 F) from which the asphalteiies
; must first be removed by using propane or other hydrocarbon solvents. The
removal of asphaltenes by propa.,3 or other solvents is called "solvent de-
i asphalting" and the de-asphalted oil is referred to as "SDA oil".
t
The viscosity index of the vacuum gas oil and SDA oil basestocks may be improved by
i solvent extraction of the aromatics and resins in the basestocks using any of these
solvents; :
i . ',
o Furfural •
o .Phenol
o Liquid sulfur dioxide (SO2)
o N-methyl-2-pyrrolidone (NMP)
Alternatively, catalytic hydrogenation may be used fo saturate the aromatics and resins
instead of solvent extraction, i ' ,,
1 . . !
The lubricating oils may be further processed for lowering of their pour points
(to improve their low-temperature usefulness) by solvent extraction of waxes using any of
these solvents;
o Propane
o Methyl ethyl ketone (MEK)
o Urea
o Propylene and acetone mixtures:
o Dichloroethane and methylene chloride mixtures
o MEK and methyl isobutyl ketont (MIBK) or toluene mixtures
! ' i • '
I
39
-------
• Catalytic hydrogenation may also be used as an alternative to the solvent extraction of
waxes.
Basestocks derived from naphthenic,crude oils usually require very little pro-
cessing to produce lubricating oils with little wax and low pour points but relatively
poor viscosity indexes. A typical naphthenic lubricating CM has these properties:
Viscosity at TOO °F ' 500 SUS (Saybolt Universal Seconds)
Viscosiry at 210 °F 54-56 SUS
Viscosiry Index 25-50
Pour Point -15 °F , •
Catalytic hydrogenation may be Osed to improve the viscosity index of a naphthenic
lubricating oil.
Paraffinic basestocks produce lubricating oils with good viscosity indexes but
high wax contents and high pour points. After solvent extraction of aromatics and reiins
followed by solvent de-waxing, typical paraffinic lubricating oils have these, properties:
100 vis oil 300 vis oil
Viscosity at 100 °F 100 SUS 300 SUS
Viscosiry at 210 °F 39.5 SUS , 52 SUS
Viscosity Index 95 95
Pour Point • 0 °F 0 °F
t
The lubricating oils produced from SDA vacuum residual oil are called "bright
i stocks" or "cylinder stocks". After solvent extraction of aromatics and resins followed
by solvrnt de-waxing, a typical bring stock has these properties: :
' 0 '
; Viscosity at 100 F 2650 SUS
; Viscosity at 210 °F 155 SUS
r Viscosity Index ; 95
i
i The variety of process options and sequences practiced in lubricating oil pro-
1 duction rrjakts it very difficult to generalize cs to the source and composition of their
! wastewaters. However, there is an obvious potential for any of these solvents to enter
: the refinery's wastewater sewers: furfural, MEK, MIBK, toluene, phenol, urea, NMP,
i acetone and others. i
: . ! I •' , '
! Asphalt Production I
, i
The petroleum refining industry produces a variety of asphalts or bitumens. The
primary end use for asphalt is in the paving of roads. Some specialty uses (roofing tar,
flooring tiles, etc.) account for a relatively small amount of the asphalt market. About
95 U.S. refineries currently produce an aggregate total of about 800,000 barrels per day
»of asphalt. . l_ '
40
-------
Asphalt Is produced by the oxidation of aspholtic, vacuum residual oils. The
oxidation is accomplished by blowing air through the heated residual oil. The hot gases
from the air blowing are cooled by direct quenching with steam and water, which pro~
duces a very oily wastewater. If any cracking occurs during the air blowing, the waste-
water will probably contain some heavy phenolic materials.
Sulfur Recovery
Many refineries convert their hydrogen sulfide-rich gases into byproduct solfur
by using a catalytic process commonly known as the Clans process. As discussed earlier
herein, hydrogen sylfide is removed from refinery gases and from C3 and C4 cuts by
absorption in an airine solution. The subsequent regeneration of the spent amine for
reuse (by distilling the spent amine) yields a concentrated hydrogen sulfide gas. That
hydrogen sulfide-rich gas from amine regeneration typically constitutes the major part of
the hydrogen sulfide fed to a Claus process.
The Claus process usually consists of a combustion step followed by two or three
fTxed-bed catalyst stages. About one-third of the incoming hydrogen sulfide is burned
to form sulfur dioxide in the combustion step:
The sulfur dioxide then reacts with the remainder of the hydrogen sulfide in the fixed-bed
catalyst stages to yield byproduct sulfur:
2H2S + SO2 -> 3S + 2H2O
The reaction effluent is cooled to the melting point of sulfur and the condensed, mclten
sulfur is withdrawn and sent to storage. The residual gases are either incinerated or
further processed (in so-colled "tail gas" units) for additional sulfur recovery.
Normally, the Claus process does not generate any wastewater. The water
vapor in the reaction effluent (see above reactions) remains in the toil gases which are
vented to the atmosphere after incineration or further tail gas processing.
Sour Water Strippers
The term "sour wastewater" refers to any wastewater which contains either:
o Dissolved hydrogen sulfide, or '
o Hydrogen sulfide and ammonia combined in the form of ammonium hydro-
! sulfide
. As discussed in the foregoing description of refinery unit processes, sour waters are
usually generated by these unit processes:
41
-------
o Crude oil distil lotion units, when processing crude oils which contain some
dissolved hydrogen sulfide ,
o Vacuum distillation units, when processing sour crude oils and some cracking
occurs in the distillation heater
o Fluid catalytic crackers processing sour gas oils (those containing organic
sulfur)
o Thermal crackers, either cokers or visbreakers, processing sour feedstocks
o Hydroaackers and hydrotreaters, which almost invariably process sour feed-
stocks
o Gas recovery plants handling sour gases
o Steam crackers processing sour feedstocks
Most refineries segregate and collect all of their sour waters for processing through a
• sour water stripper '"*, wherein about 95-99 percent of the hydrogen sulfide and 80~95
percent of the ammonia are removed by distillation. The overhead gas from (he stripper,
containing hydrogen sulfide and ammonia, may be sent to a refinery's Clous unit for
conversion into byproduct sulfur. The stripped water, containing o few ppm of hydrogen
su I fide and perhaps 50-200 ppm of ammonia, is then routed through a refinery's waste-
water treatment sequence.
Summary •
Table 4 summaries the possible wastewaters from the refinery unit processes dis-
cussed in this section.
! REFINERY CLASSIFICATION
I
Refineries may be classified in any number of ways. As discussed earlier herein,
• refineries might be classified on the basis of the endproducts they yield.. .such as motor
, fuel refineries, fuel oil refineries, lubricating oil refineries, efc. However, for the
j purposes of this report, refineries will b«s classified in terms of their processing com-
plexity as defined by the types of unit processes used in the refineries. In essence, the
refinery classifications in this report will be the same as used by the U.S. EPA°:
o Category A (Topping Refineries) ~ refineries which include atmospheric
crude oil distillation and catalytic reforming with or without any other pro-
cesses (excluding cracking processes). •
o Category B (Cracking Refineries) — refineries which include atmospheric
crude oil distillation, catalytic reforming and cracking with or without any '
> other processes (excluding lubricating oil and petrochemical production pro-
cesses).
42
-------
TABLE 4. WASTEWATcR SOURCES FROM REFINERY UNIT PROCESSES
Processes
Vaste-
waters?
Crude oil desalting ' yes
Atmospheric distillation yes
Vacuum distillation yes
Fluid catalytic cracking yes
Coking (delayed or fluid) yes
Vi streaking yes
Steap cracking (gas oils) yes
Catalytic hydrocracking
Catalytic reforming ' wo
Naphtha HDS yes
Distillate HDS yes
Heavy oil HDS yes
Gas recovery plants:
Unsaturates yes
Saturates ' yes
Merox treaters' ' yes
JlJtylation yes
Isomerization , no
Hydrogen synthesis (yes)
Arow»tics extraction (yes)
Petrochemicals (yes)
Lubricating oil yes
Asphalt , yes
Sulfur recovery no
Pollutants typically e-tpected in wastewater
Inorganic chlorides, HC, (B2S, phenols)
HC, H2S, (XH3, phenols)
HC, (NH3, phenols)
HC, H2S, NH3, CN, phenols
HC, H2S, NH3, CN, phenols
HC, HjS, NH3, CN, phenols
HC, H^S, NH3, CN, phenols
yes H2S, HH3, (HC)
H2S, NH3, HC, (phenols)
H2St NH3, HC, (phenols)
, NH3, HC, (phenols)
NH3, RSH, CN, aatine, (HC, phenols)
H2S, NHj, RSH, amine, (HC)
NaSH, NaSR, sodium phenolates, (HC)
Sulfuric or hydrofluoric acid or acid salts
(CO2, CN, HH3, amine)
(solvents, aromatic HC)
(va,rio'.is)
Solvents ard various others
HC, (phenol sj
Aiireviations: HC —
K2S —
NH3 —
CN —
hydrocarbon liquids or oils RSH
hydrogen sulfide SaSR
ammonia , NaSH
cyaiudes and thiocyanates CO2
— mercaptans<
"• sodium mercaptides
— sodium hydrosulfide
•~carbon dioxide
The pollutants enclosed in ( ) indicate those which may not be present in all
cases.
43
-------
o Category C (Petrochemical Refineries) — refineries which include atmospheric
crude oil distillation, catalytic reforming, cracking and petrochemical pro-
duction processes with or without any other processes (excluding lubricating
oil production processes).
o Category D (Lubricating Oil Refineries) ~ refineries which include atmos-
pheric crude oil distillation processes, catalytic reforming, cracking and
lubricating oil production processes with or without any other processes
(excluding petrochemical production processes).
o Category E (Integrated Refineries) -- refineries which include atmospheric
crude oil distillation, catalytic reforming, cracking, lubricating oil pro-
duction processes and petrochemical production processes with or without any
other processes. ;
; Table 5 lists the various unit processes which are included or may be included within each
of the refinery classification categories defined by the EPA. The EPA's classifications
ore fairly straight-forward except for the definition of petrochemical operations. The
CPA's definition includes "first generation" petrochemicals (such as 3TX and olefins ^
which ore produced directly from refinery intermediate streams) as well as "second gener-
ation" petrochemicals (such as cumene and styrene which are produced from the first
generation petrochemicals). And the EPA further stipulates that:
o Any production of second-generation petrochemicals constitutes petro-
i chemical operations or processes.
o But production of first-generation petrochemicals must amount to 15 percent
j or more of refinery pioJoction to constitute petrochemical operations.
| Thus, a smell steam cracker in a large refinery might be classified as a cracker, whereas
: a large steam cracker in a small refinery might be classified as a petrochemical unit...
i even though the steam cracker produces first-generation petrochemicals (ethy ene and
, propylene) in either case. The same situation applies to the extraction of BTX from
catalytic reformate.. .it might or might not be classified as a petrochemical unit
; depending upon its sire relative to the refinery size. As footnoted in Table 5, for the
', purposes of this report, steam crackers and oromatics extraction units are taken to be
: petrochemical processes on the basis that their output of first-generation petrochemicals
, exceeds 15 percent of the refinery production. But the readers of this report should ;
; understand that such may not always be the case in terms of the EPA's detinition. >
i ' *
Topping Refineries (Category A) ! •
s~^ j ,
Figure 3 is a flow diagram of a typical topping refinery in which:
o The feedstock crude oil is desalted and then distilled in an atmospheric
distil lot-on unit. I
o The virgin heavy naphtha from the atmospheric distillation unit is hydrode-
._)._.,
'44 ::-'
-------
TABLE 5. REFINERY CLASSIFICATIONS AS DEFINED BY UNIT PROCESSES IN EACH
REFINERY CATEGORY .
, CATEGORY
PROCESSES
Crude oil distillation:
Atmospheric
Vacuum
Fluid catalytic cracking
Catalytic hydrocracking
Delayed coking
Fluid bed coking
Visbreaking
Steam cracking (of gas oils)5
Catalytic informing
Naphtha HDS
'Distillate HDS
Heavy oil HDS
Gas recovery plants:
Unsaturates
Saturates
Alkylation
Catalytic polymerization
Butane isomerization
Pentane~Hexane isomerization
Hydrogen synthesis
Arontics extraction3
Petrochemicals
Lubricating oil
Asphalt
Sulfur recovery
x • • •
X • • •
X • • • •
X • • •
X • ' • •
• X • X
•
9
•
na
na
na
na
X
X
X
X
• These processes are Included or nay be included by the category definition
X , These processes are excluded by the category definition
na NI~+ applicable for refineries without cracking units
a Defined as petrochemical jreducing processes for this listing
CATEGORY A — Topping Refineries
CATEGORY 8 — Cracking Refineries
CATEGORY C — Petrochemical Refineries
CATEGORY D -- Lubricating Oil Rafineries
CATEGORY Z — Integrated Refineries
-------
Pipeline
Crude oil
TahKer
Crade Oil
CRUDE OIL
STORAC2
CRUDE OIL
DESALTING
Tanker
Hater
BALLAST
HATER
TREATMENT
Treated
Ballast Hater
Gas
ATMOSPHERIC
OISTILLATIOM
Lt
Naphtha H
Hvy Naphtha , f JESULFUHI7 ,'NC \
DISTILLATION
Vacuum
Can Oil
Reaiduun
SULFUM
RECOVEHV
• Sulfur
Fuel gas
». Propane. Butane
„ 1 Lt Virgin
Naphtha
-^-i
'
;:ATALYT1C I Heforaate
REFOKH1NG
Keroaine, Dieael
Casolin
Fuel oil
H2S
Sour
Hater.
Strlppml
""»'
FIGURE 3. An Example of a Category A Refinery (Topping Refinery)
-------
sulfOrized and caralytically reformed to yield a sulfur-free, high-octane
gasoline component. '.
o The virgin kerosine and diesel f.xim the atmospheric distillation unit are '.
hydrotreated for sulfur removal. ;
o The residual oil from the boftori of the atmospheric distillation unit is vacuum
— - distilled to yield additional gas oil. — - -- - -
o The combined' virgin ga«. oils from the atmospheric and vacuum distillation
units are hydrotreated for sulfur removal and then blended with the vacuum
distillation residual oiJ to produce fuel oil.
o Hydrogen for the desulfurizer and hydrotreaters is produced by the reforming
unit.
• o The various process unit gcses are routed through a gas recovery plant to
recover sweet refinery fuel gas, propone and butane, LPG, and a light virgin
naphtha.
" o~ The light virgin naphtha is Merox treated for removal of mercaptans and then
blended with the catalytic reformare to produce the refinery's endproduct
gasoline.
o The hydrogen sulfide from the amine scrubbing in the gas recovery plant and
from the refinery's sour water stripper is converted into byproduct sulfur in o
Clous sulfur recovery unit.
I The topping refinery depicted in Figure 3 is a fairly simple one. As shown in Table 5,
topping refineries may be quite a bit more complex. For example, if the catalytic
• reformer hydrogen yield is not sufficient to supply the hydrotreaters, then a hydrogen
synthesis unit would be required. In some ccses, the light virgin naphtha (pentanes and
, hexanes) may be catalytically isomerized for ocrarie rating improvement. Topping
] refineries may also include cromatics (BTX) extraction from the catalytic reformate,
' lubricating oil production and asphalt production. Thus, Figure 3 depicts only one of
: many processing configurations that could bs classified as a topping refinery. However,
, it contains the primary or fundamental process units typical of most topping refineries.
! | •' • .
\ Crocking Refineries (Category B) ' I
.! ' l
'. Figure 4 is a flow diagram of a cracking refinery in which:
•I '..- |
o The desalted crude oil feedstock is c'istilled (by atmospheric and vacuum
distillation) to yield gas, virgin naphthas, virgin kerosine and diesel, and
virgin gas oils. i
o The virgin heavy naphtha is hydiodesulfurized and catalytically reformed. .
The iulfur-free, high-octane catalytic reformate is sent to endproduct gas-
oline blending.
47 • r
-------
HjS
Sulful
-------
o The virgin kerosine, and diesel are hydrotreated for sulfur removal. —!
1 ' ' T
o To minimize fuel oil production and maximize gasoline production, tlie light
virgin' gas oils are cracked in a fluic catalytic cracking unit and the heavy '•
virgin gas oil is catalytically hydrocrocked. :
o Part of the vacuum distillation residual oil is coked (to further maximize ~
gasoline production), part ts oil—blown to produce asphalt and the remainder
is blended into fuel oil. .' > '.
o To maximize gasoline production even further, the gas oil yielded from the
coker is fed to the fluid catalytic cracking unit for cracking. '<
< i
o The heavy naphtha from the hydrocracker is sent to the catalytic reformer for
octane upgrading. •
.! ! ' ' •'
o The gases and light naphthas from the fluid catalytic crocking unit and the :
coker (with a high content of unsaturated olefinic components) are routed
through an unsaturate gas recovery plant.
t —
o" The isoporaffinlc, light hydrocracked naphtha, which has a very good octane
rating, is sent directly to endproduct gasoline blending.,
o- The other gases and light naphthas from the atvnosphfiric distillation unit, the
catalytic re Former, the hydrocracker and the hydrotreaters (all of which are
primarily saturated paraffins) ar« routed thrcogh a saturate gas, recover/ plant.
o The naphthas recovered from the two gas recovery plants ore Merox treated
for mercaptan removal and then sent to endproduct gasoline blending.
o An alkylation unit produces a very hljh'octane gasoline component by com-
bining isobutanr in the saturate gas plant C3/C4 cut with the propylene and
butylenss In the unsoturate gas plant C^T/C4~ cut.
o The refinery's endproduct gasolines are produced by blending:
Light virgin naphtha, catalytic reformate, alkylate, light and heavy
fluid catalytic cracking and coker naphthas, and light hydrocracked
naphtha.
o The refinery's distillate oils (virgin kerosine and diesel, fluid catalytic
cracking gas oil and hydrocracked diesel) are blended to produce kerosine,
jet fuel and diesel oil. Some of the heavier distillate oils are blended with
vacuum residual oil to produce endproduct fuel oil as well as refinery fuel
• oil. •; j
; t ' '
o Hydrogen sulfide from the amine .scrubbing in the gas recovery plants and
from the sour water stripper is converted to byproduct sulfur in a Claus sulfur
recovery unit.
s that
I
i A comparison of Figures 3 and 4 illustrates that a cracking refinery is very much more
rcomplex than a topping refinery. And, in rriany cases, it may be even more complex.
-------
2r example: J
i o The butylenes fed to alkylotion often require catalytic hydrogenation to con->
vert butadiene into butylenes. j
I I !
o If the isobutane in the saturate,gas plant, C3/C4, is not sufficient to alky-
late all of the refinery's propylene and burylene, then o butane isomerizatior*
unit is often added to convert the normal butane into isobutane.
o If the hydrogen yielded by the reformer is insufficient to supply the hydro-
tredters and the hydrocracker (which has a very high hydrogen demand),
then i hydrogen synthesis planr would be required.
if In some cases, a cracking refinery alkylates its butylenes,but catalytically
polymerizes its propylene. j • ' • .
o In some cases, the light virgin naphtha (pentanes and hexanes) may be
catalytically isomerized for o:tane-rating improvement.
i ' i ;
j"Tfie "cracking refinery depicted in Figure 4 has both fluid catalytic crocking and ' T'.
I catalytic hydrocrocking. That is not always the case. Some cracking refineries have
'only fluid catalytic crackers and some hove only hydrocrackers. Other cracking
1 refineries have no coker, and yet others may have yisbreakers (thermal crackers). Thus,
: a cracking refinery may also be considerably less complex than the one shown in Figure 4
i .. .but in general they are still much more complex than a topping refinery.
! * ' ' '
I Petrochemical Refinery (Category C) j
Figure 5 is a flow diagram of a petrochemical refinery. It is identical with the
cracking refinery in Figure 4 except for these additional process units:
o An oromatics production unit which extracts BTX from a portion of the
refinery's catalytic 'eformate.
o Second-generation petrochemical production processes:
i
(o) Ethylbenzene production by the catalytic reaction of benzene and
ethylene. .
(b) Sfyrene production by the catalytic dehydrogenation of ethylbenzene.
(c) Gumene production by the catalytic reaction of benzene and propylene.
1
The ethylene and propylene feedstocks (C2/C3T) for the second-generation i
petrochemical units ore obtained from the refinery unsaturate gas recovery
plant. The benzene feedstock it obtained from the BTX extraction unit. |
.INF
In terms of producing motor fuels, distillate oils and fuel oil, the petrochemical refinery
in Figure 5 uses the same configuration of process units as the cracking refinery in Figure
~4~r As w?th the cracking refinery, there may1 be many other petrochemical refinery —
50;:
-------
HJ»
•
. Cm-tie oil
t
Crude Oil
[CiUDI
I STOI
; OIL"!
IACK 1
(CRUDE OIL I
DESALTING]
• L_
Tanker
'. ••Heat Hater
; ~"
r~
CM . .'""_"
. • GAS KECOVE
A'.tCSPKExIC
DISTILLATION
VACUUM
DISTILLATION
Naphtha | *
• ' i \-« i ni4 111^
( ^] REFORHINC
.( HYDIOTKEATING ..f*'."""*!.
1 , ,.- Die««l
I 1
-„ ...».,- <. uNiAIUKAIl
*»i " GAS RECOVE
041 . » AND SHEETCN
i ' i Naphtha 1 T
CATALYTIC 1 1
r PRACKlwn At * l FCC Can Oil
-» Fuel gae J
^•3/*-"4 I
i i a •• ». I
RY 11
.J jj
1 * I
J ' j
i * "
— 8F«l I
I
C 3/^4 f
:D T
RV *
INC
1B,.-
| Coker Gas Oil * * Hydrocracked Diesel ri6 —
~Gaa T 1 I Renlduuai
"" r»> r^1 1 It
^^* HYP^t- ***^ Naphtha J
Gf» Oil_ CRACRINO
Caa
Reelduu. -T^dh ""f^t t
" 1- U —
,.,,,_ n , i
!• «u _
L-Jr
]
Sour
Hatere
1SU1.
REOO
OH |4| 1
. • Sulfur
VERY J
.
i t|ene3. j Lt Virgin Naphtha , (
-T 1
AROHATICS
"*• EXTRACTION
Benzene
L
LH PtTHO-
.1 CHEMICAL!
— •!_
r~
{A! KYLATIQk _<
i MERO)I 1 Lt F<
>. Xylen«
• Toluene
. P> Benzene
, , i -
»• Propane, Butane
AlVylate
X 1 CoK«r Naphtha^
j (^iftjx 1 Hvy FCC (. Colter Naphtha
«• Keroaine
» Jet Fuel •
•• Diesel oil
• Fuel Oil
—» Refinery Fuel Oil
Lt Hydrocra?Ked Naphtha
"2s
1
SOUK
-» HATER -
STR1PPKR
Stripped
^^ Hater
| ' GASOLINE K.KNB1KG WOL
BALLAST j - ' .-, 1 -
TREATMENT! toeiduuB " Cat*
Treated
^•1 1 »_• Ma
ft
MldUuei ASPHALT 1 . t ..
- a^»(iNG | * "" *""~ " -
FIGURE 5. An Exomple of o Category C Refinery (Petrochemical Refinery)
-------
Ycqnfigurations which are more complex:
: i " ,
o A steam cracker might be added, using part of the refinery's virgin gas oil as
feedstock, to produce ethylene, propylene, butadiene and aromatics as well
as byproduct gasoline and fuel oil.
o Cyclohexane might be produced by the catalytic hydrogenatinn of benzene.
o The same possible variants as discussed for the cracking refinery apply to the
petrochemical refinery. :
Thus, Figure 5 depicts only one of the many processing configurations that could be
classified as a petrochemical refinery. '
! ' i
Lubricating Oil Refinery (Category D)
' Figure 6 is a flow diagram of a lubricating oil (lube oil) refinery. It is
' identical with the cracking refinery in Figure 4 with the exception that it includes pro-
cessing units for producing a range of lubricating oils. *' !
i . i
The feedstocks for the lubricating oil processes are gas oil and residual oil
obtained from the vacuum distillation unit. The various processes that may be included
• <-> the lubricating oil unit are:
o SDA of the vacuum residual oil feedstock.
i i
o Solvent extraction of aromatics and resins from the feedstocks for improving
the viscosity index of the endproduct lubricating oils.
o Solvent extraction of waxes from the feedstocks for lowering the pour point of
the endproduct lubricating oils.
! i i
! Figure 6 depicts only one of the many processing configurations that could be classified as
i lubricating oil refineries. All of the possible variants discussed for a topping refinery and
for a cracking refinery apply to a lubricating oil refinery as well,.
! , ' . • j !
: Integrated Refinery (Category. E) , !
Figure 7 is o flow diagram of an integrated or "combination" refinery, which is
'one that combines all of the elements of cracking, petrochemical and lubricating oil
j refineries. , '
i • • ' !
i .
As with the other refinery categories, Figure 7 depicts only one of the very
many processing configurations that could be classified as integrated refineries.
-------
Sulfur
(Ua
UNSAIDRATED
GAS RECOVERY-
AND SHEETEN1MG
Lt FCC t Co Her Naphtha
Hvy FCC (. CoKer Naphtha
Keroftine
Jet Fuel
Diesel Oil
Fuel Oil
Refinery Fuel Oil
J
Lubt icating
Oils
FIGURE.6. An Example of a Category D ReHnery (Lube Oil Refinery)
-------
MJLFUM
-
Pipeline
Crude Oil
Tanker
Crude Oil
CRUDE OIL
. STORAGE
. CRUDE OIL
DESALTING
Tanker
Ballast Mater
Gas
ATMOSPHERIC |
DISTILLATION P
.
VACUUM
DISTILLATION
Naphtha
•
Gas.
r"2 -cf4"
J DESULFURIZIN22C_!!2£.hth<,
f-"2 r^*
»4 HYDRDTHEATING ." "" slne!
| Diesel
Gas
Oil
f
Gas
Oil ["
Hvv -
Gas Oil
BALLAST
MATER
TREATMENT Residuum
Treated
Ballast Mater
Residuu*
Gas!
. I r,,,Tn * 1 Naphtha
CATALYTIC
. CRACKtMG *•
1 ' r u,
Coker Gas Oil *
Lube Stock 1
r- r-J'I I
DISTILLATE '
i- HYPPO Hvy Naphtha f
CRACKING
Gas,
1 Naphtha f I 7
b COKING i
1 — : ^ Coke J
7
ASPHALT . ... 1
* BLOWING *" AaPn*lt
1 +. fuel g«»a
1 C^C4
Gas
H2— ,
• CATALYTIC
^_ REFORMING
H2S
f— oa_sl _
C3/C4
T
CAS RECOVERY 1
, AND SWEETENING
|
, * ^ FCC Gaa Oil . -IK
Kerusine* Diesel ^
HydrocrdcKed Diesel ^
1 Residuum E
i r3
i i
—JIT T-
Matei
1 1
I . 1 :„ 1 Lt Virgin Naphtha .
1 L \ A*OMAT1CS 7 rJiScnj:
It EXTKACTION
( T • Benzene
J Benzene ^ 1
L». PETRO- «-«nyioenrene
/Cj CHEHICALS T-^stVren«
1 | »• Propane. Butane
* .1 ALI'YLATIOM Alkylate
- f".'j^Tw%»"l Lt FCC i. Coker Naphtha^
. 1 ..^r^.^ 1 HVY FCC L CoKer Naphtha
». Keroaine
S ». Jet Fuel
a ^ Diesel Oil
u » Fuel Oil
« •• Refinery Fuel Oil
Lt Hydrocracked Naphtha
H2S
1
5OUR
• STKIPPER ' .Mater
LUBE OIL | Lubricating
PLANT | * Oil*
FIGURE 7. An Example of a Category E Refiner/ (Integrated Refinery)
-------
REFINERY WASTEWATERS
Thus for, fhis section has dealt primarily wifh Hie unit processes used in
petroleum refineries and with the wastewaters generated by the individual processes (as
summarized in Table 4. However, there are a number of non-process wastewater sources
in most refineries, and the purpose of this section is to discuss all of the wostewaten that
may be encountered in refineries including non^process as well as process sources.
Wastewater Sources
Figure 8 depicts all of the wastewater sources'typicajly encountered in petroleum.
refining:
o Wastewatsrs from process units, which includes:
(a) Sour waters (process steam condensates and wash waters) as previously
discussed and summarized in Table 4.
(b) Miscellaneous solvents and chemical wastes (also summarized in Table 4)
as well as drips and drains from valves, pumps, sample taps, etc.
(c) Rainwater runoff from process unit areas.
o Drainage from tank farms (storage tank diked areas).
o Ballast water withdrawn from the cargo holds of incoming empty tankers at
those refineries which ship some or all of their endproducts by tankers.
o Cooling tower blowdown ai those refineries which use circulating cooling
water systems with evaporative cooling towers.
o The wastewater brines from the treatment (softening and/or demineralizing) of
the refinery's intake water supply.
o Blowdown from the steam-generating boilers in a refinery, which includes
utility steam boilers as well as process heat recovery steam generators (waste
heat boilers).
o Sanitary water effluents (from wash .oom and toilet facilities).
o Clean rainwater runoff from non-process areas such as roads, parking lots,
utility plant areas, etc.
o Wostewaters from the refinery laboratory (not shown in Figure 8 for the sake
of simplicity).
It must be emphasized that some refineries may include all of the wastewater sources
shown in Figure 8, but other refineries may not. For example, many refineries do not
ship products by tanker and hence would not have ballast water to handle. As another
example, some refineries utilize once-through cooling water systems rather than circu-
lating cooling water systems and therefore would not have any coolina tower blowdown.
55
-------
TAUK FARM
CONTROLLED
DRAINAGE
OIL
TRAP
COOLING
TOWER
SLOWDOWN.
13API or CPIS
Sj OIL-WATER p
SEPARATOR
"IP^AiR" '1
:i aOTATION* S
;| EQUAL! ZING
;j BASIN
1 BIOLOGICAL
j TREATING '
; MIXED
?! MEDIA
i? FILTRATION
_HIGH TDS HATERS (inorganic salt J
INTAKE WATER
TREAT;CNT
WASTES
BOILER
SLOWDOWN
TREATED
SANITARY
EFFLUENTS
CLEAN
RAINWATER
RUNOFFS
* Or chemical coagulation and settling
FIGURE 8. Wostewoter sources and wastewater management in-a refinery
J FINAL ,
3 POND
DlSCHARGt
-------
Figure 8 also depicts a wastewater management or treatment sequence which
includes:
o Primary removal of oil and suspended solids from the oily water sewer flow,
using a gravimetric oil-water separator of either the API design or the
corrugated plate interceptor (CPI) type.
o Secondary removal of oil and suspended solids, using either air flotation or
chemical coagulation and settling.1
o An equalizing basin which provides intermediate storage and mixing of the
incoming wastewater to "smooth out" any fluctuations in wastewater flow and
composition so as to obtain a uniform fsed to the biological treatment step.
o Biological treatment which may utilize:
(a) An aerated lagoon system,
(b) A trickle filter process
(c) An activated sludge process
(d) Or various combinations of aerated lagoons, trickle filters and activated
sludge units.
o Mixed media filtration for the final removal of suspended biological solids.
o A final storage pond which provides for testing and quality control before
final discharge of the tre<:*ed effluent.
o A separate sewer for non-oily wastewater, which contains a relatively high
concentration of total dissolved solids (TDS) primarily in the form of
inorganic salts. Such non-oily, high TDS wastewaters may be routed
directly to the final pond as shown in Figure 8.
The total v.o$fewater flow generated by petroleum refining varies very widely
from one refinery to another. There are many factors that influence the wasiewaicr flow
from any given refinery, some of which are:
o The refinery's processing configuration as defined by the refinery categoric
discussed herein.
o The annual rainfall at the refinery's geographical location.
o The type of cooling water system used in the nsfinery (that is, whether once*
through or circulating). :
o Whether or not the refinery must handle tanker ballast water.
o The age of the refinery and the degree of good "housekeeping" practiced
within the refinery. .
i
o The degree of air-cooling and of wostewater reuse so as to minimize the
overall water demand of the refinery.
57
-------
!\Based upon data obtained in 1972 from 94 U.S. refineries (each with less than 3 percent
• of their cooling needs provided by once-through cooling water), the EPA derived these
median, dry-weather wastewater flows":
TABLE 6. MEDIAN DRY-WEATHER WASTEWATER FLOWS'
Median wastewater flow
Refinery Category (gallons of wostewoter/barrel of crude oil)
Topping (Category A) , 20
Cracking (Category B) 25
Petrochemical (Category C) '30
Lubricating. Oil (Category D) , 45
Integrated (Category E) . 48
Translated to a refinery processing 100,000 barrels per day of crude oil, the above
• median wastewater flows range from 1,390 gpm (Category A) to 3,330 gpm (Category E).
; It must be'emphasized that the range of extremes (lowest to highest) was very much
. broader than the above range of median flows. In any event, these data may provide the
reader with a perspective as to the wastewater flow rates encountered in petroleum
refineries.
(
Wastewater Pollutants
Refineries generate a spectrum of wastewater pollutants which includes
inorganic salts, hydrocarbon oils, suspended solids, ammonia, su If ides, phenols,
! mercaptans, cyanides, solvents, heavy metals, etc. To discuss each and every specific
' pollutant or to classify those which might be considered "hazardous" or "toxic" is beyond
; the scope of this refinery characterization section. Accordingly, this section is limited
1 to a discussion of the major pollutants or pollutant parameters which have generally been !
used in the petroleum refining industry to characterize the level of pollutants in refinery .
' wastewater streams: : \
i. ! ' ;
BODj. A parameter which measures the oxygen consumed by the bio-
logical oxidation of wastewater pollutants over a 5-day period,
commonly called the 5-day biological oxygen demand.
! • ,
COD A parameter which measures the oxygen required to chemically
1 ; .oxidize wastewater pollutants using a chemical oxidant and a
specific reaction time, commonly called the chemical oxygen
demand. There are many different COD measurement methods,
and the most common one in the United States uses a dichromate
oxidant with a 2-hour reaction time. There is no generalized
correlation between the BODc and the COD parameters.
58
-------
Suspended
Solids
Oil
Ammonia
Phenol?
Sulfide-.
Total
Chromium
A parameter which .measures the amount of organic carbon
present in a wastewater, commonly called the total organic
carbon. There is no reliable, generalized correlation between .
the BOD5, COD and TOC parameters. The TOC measurement
can be made by an automated, on-line analyzer which provides
a rapid determination of organic carbon levels in wastewaters.
For specific wostewotert in specific refineries, TOC may be
correlated with BOD5 and COD to provide a continuous indi-
cation of the biological and chemical oxygen demand of the
wastewater. .-.as long as the operating conditions in the
refineries remain within the range for which the TOC was corre-
lated with BOD5 and COD.
The suspended solids content of a wastewater includes organic
solids (tar, grease, fibers, hair, sawdust, e:c.) and inorganic
solids such as sand, silt and clay.
The oil.content of a wastewater includes any oil or grease which
is extractable by a specific solvent, usually Freon.
The ammonia content of a wasrewater includes free or dissolved
ammonia as well as combined ammonia such as in ammonium
hydrosulfide. It is usually reported as ammonia nitrogen (NH3~
N), which is the nitrogen fraction of the ammonia amounting to
82 weight percent of the amn.onia.
The phenols content of a wastewater includes phenol, cresols
and xylenols as well as perhaps other phenolic species.
The suifide content of a wastewater includes free or dissolved
hydrogen suifide as well as combined suifide such as in ammonium
hydrosulfide or in sodium hydrosulfide.
The total chromium content of a wastewater includes trivalent
and hexavalent chromium.
• M&ny if not all of the laboratory test methods for measuring the above pollutants and
| pollutant parameters have problems from interfering substances and have varying degrees
of repeatability (in the same laboratory by fhs same analyst on the same sample) as well
: as varying degree* of reproducibility (by different ioboratories or different analysts on the
i same sample). This makes it quite difficult to obtain meaningful pollutant data com par- '
i isons from one refinery to another even when using specified or standardized test methods.
In fact, it is often difficult to obtain meaningful comparisons of pollutant dare from one_
K,-wastewoter stream to another within the same refinery.
.59
-------
5*-^. Each of the above pollutants and pollutant parameters are discussed below in .
terms of the refinery wastewater sources ([depicted in Figure 8) in which those pollutants
I can normally be expected to be present. ' , i
i ' i , '
! 5-doy BOD- :
.,_.. Most of the sour wastewaters, from the normal operation of refinery process units
*' -(»ee Figure 8) contain biodegradable pollutants ond have a significant 5-day BOD. - As
: shown in Table 4, almost all refinery unit processes generate wastes containing oils,
solvents and other substances, all of which are biodegradable. Since the process unit
i oreas are usually pavecl'and graded to drain into the oily water sewer, the rainwater run-
; off and miscellaneous drains and drips from those process areas contain biodegradable
• pollutants and also have a significant 5-day BOD (although perhaps somewhat lower than
. the 5-day BOD of the operational sour wastewaters).
;
' After stripping the sour waters for hydrogen sulfide and ammonia removal, the
stripped waters are normally routed through a crude oil desalter. The crude oil absorbs
. and removes a gooa part of the phenols that may be present in the stripped waters .
'However, the oity and salt-laden effluent water from the desalter to the oily water sewer
, has a 5-day BOD which may range from 70 to 600 ppm'.
The oily water drainage from the storcge tank farm has a significant 5-day BOD,
as does the oily ballast water removed from tankers arriving to be loaded with outgoing
refinery products.
t
( The blowdown from cooling towers in cooling water systems which are in good
; physical condition do not usually contain any oils or other pollutants that create a 5~day
' BOD. However, in older systems which may have leaks in their water-cooled process
i heat exchangers, the cooling tower blowdown may contain a significant amount of oil !
; which creates a significant 5-day BOD. i ,
i '•
In summary, the probable wastewater sources of 5-day BOD are these:
1 - i
o The operational sour or phenolic or oily wastewaters from the refinery unit
processes. ! j
i
o The rainwater runoff and miscellaneous drips and drains from the refinery unit
processes. ; ,
o The oily ballast water >'n those refineries shipping products by tanker.
• i i
o The oily brine effluent from crude oil desalters.
o The oily drainage water from storage tanks and from their diked enclosures.
o The cooling tower blowdown from refineries in which there are significant
leaks in the water-cooled process heat exchangers.
-------
COD-7 '*, .
In generol,, the wastewater sources of COD in a refinery are identical with the
5-day BOD sources discussed just above. Although there is no generalized correlation,
between COD and 5-day BOD, the level of COD in refinery wastewaters is typically
higher than the level of 5"day BOD by a factor of 1 .5-2.0 or more.
TOC— -•' • -- ' — '-....... :
• In general, the wastewater sources of TOC in a refinery are also identical with
the 5-day BOD sources. There is no generalized correlation between TOC, COD and
i 5-day BOD and the level of TOC in refinery wastewaters may be higher or lower than the
level of 5-day BOD (but TCC is generally lower than COD).
Oil-- .
The sources of oily wastewaters in o. refinery are identical with those for 5-day
BOD, and that is all of the sources shown as routed into the oily water sewer in Figure 8.
i ,
Suspended Solids—
;' Atl of the wastewaters in a refinery, whether oily or non-oily, are sources of
suspended solids. That includes the wastewaters which are not routed ir.to the oily water
sewers such as cooling tower blowdown, boiler blowdown, intake water treatment wastes,
treated sanitary wastes and clean rainwater runoff. However, the sources of oily
suspended solids are the same as for oil or for 5-day BOD.
i
TDS- . i
; TDS in refinery wastewoters consist mostly of inorganic salts and their major
sources are; . , :
o Cooling tower blowdown t
' o Boiler blowdown i
! o Intake water treatment wastes j
• I ' !
; Most refinery cooling towers operate within the range of two to seven cycles of
concentration due to the evaporation required to remove the heat absorbed by the ci-cu-
lating cooling water. Thus, the cooling to\ver acts like an evaporative concentrator and
the concentration of inorganic saKs in the tower blowdown is two to seven times the con-
centration of the salts entering the tower in the treated makeup water (which replaces
th,e evaporated water and the blowdown water). :
! : i
Boiler blowdowns also contain high levels of TDS due to the evaporation of ';
• water to generate steam, which leaves behind a highly concentrated inorganic salt j
solution. Th« level of TDS in a boiler blowdown depends upon the degree to which the ,
(boiler feedwoter has been softened or demineralized, and upon the pressure level of the
steam generated which determines the amount of blowdowr needed to maintain a tolerable
> level of dissolved salts. In general, the TDS in refinery boiler blowdowns ranges from '
, IjJOOO ppm in high pressure steam boilers to 4,000 ppm in low pressure boilers. _
' ' . • 1
-------
Jf _. The degiree of intake water treatment required and whether the treatment is
• softening or demineralizirig depends largely upon the composition of the intake water
,• supply and upon the steam generation pressure levels in a refinery. In any event, the
! minerals or salts removed by the intake water treatment generate a high IDS wastewater
effluent or brine. The level of TDS in those wastes depends upon the specific treatment
unit design and the specific treatment requirements of each refinery,.
i Ammonia and Su If ides— : ' , '
i As discussed earlier, sour wastewaters are those which contain either dissolved |
i hydrogen sulfide or ammonia and hydrogen sulfide combined in the form of ammonium
; hydrosulfio'e. All of the sour wastewaters are derived from the unit processes within a
refinery and are usually routed through a sour water stripped (see Figure 8). The stripper
removes as much as 99 percent of the hydrogen sulfide and perhaps 80-95 percent of the
: ammonia. Thus, the stripped sour water routed to the oily water sewer (or reused in the :
' crude oil desalter) normally has only a few ppm of hydrogen julfide and perhaps 50-200,
.' ppm of ammonia. >
,-=a- • - Spent caustic 'solutions from Merox and other treating units (used to scrub
j refinery liquids for removal of hydrogen sulfide and mercaptans) are also sources of
{ sulfidic wastewaters. Stripping of spent caustics without prior neutralization is largely
: ineffective in terms of removing hydrogen sulfide, The options for spent caustic disposal
. are: -<•,••. •
o Neutralization followed by stripping prior to routing into the oily water
sewer. , | ,•'.,;
o Incineration in a fluidized bed incinerator.
o Deep well disposal in a suitable underground geological strata.
i
o Offseit disposal by contmct with commerciol disposal firms.
Phenols--
As discussed in the introduction, most of the phenols appearing in refinery
wastewaters are generated by the thermal cracking and fluid catalytic cracking processes
used in refineries: Thus, the sour Wastewaters from the refinery process units contain ,
significant levels of phenols (especially if the refinery includes cracking processes). '
Stripping of the sour water may remove about 20-30 percent r'. the phenols. If the •
stripped water is then reused in a crude oil desal'cr, a gcjd part of the phenols remaining
in the stripped water will bo removed by absorption.into the crude oil. '
Phenolic sprit caustics are those which have been used to scrub cracked
naphthas and heavier liquids within the refinery. If the phenolic spent caustics are
neutralized, the phenols may be removed and recovered as saleable byproduct . The
posol options for the phenolic spent caustics:are the same as for the sulfidir spent
caustics discussed just above.
: NUMt'EF!
-------
'Total Chromium—*t,
Organic and inorganic chromates are used as corrosion inhibitors in the large
majority of refinery cooling water systems, especially in circulating systems with evapor-
' ative cooling towers. Thus, the major source of chromium in refinery wastewaters is the
blowdown from cooling towers.
SUMMARY AND PERSPECTIVE
All of the wastewater sources (process and non-process) typically encountered
in petroleum refineries have been discussed and explained in this section. Those sources
and their routing through a typical wastewater management or treatment sequence hove
been depicted in Figure 8. To provide a perspective, Table 6 presents EPA data on the .
median, overall dry-weather wastewater flows from each of the EPA's five refinery cate-
gories. Those flows ranged from 20 to 48 gallons of wastewater per barrel of crude oil
processed. .
The pollutants and pollutant parameters commonly used to characterize refinery
Wastewaters have been defined and discussed in this section. Then each of those
poiiutants and pollutant parameters have been discussed in terms of identifying the
specific wastev/oter sources in which they can normally be expected to appear.
i
As a final order-of-magnitude perspective, Table 7 presents data derived by the
EPA as to the median pollutant loads (and their equivalent wastewater pollutant concen-
trations) from each of the EPA's five refinery; categories. 'Excluding'the' category of
topping refineries, the median, overall refin«ry wastewater pollutant concentrations ir.
Table 7 may be summarized as:
' ' i
i
5-day BOD 120-240 ppm •
, COD 290-650 ppm
TOC, '70-210 ppm
i ' I
Suspended Solids 20- 70 ppm
Oil 150-110 ppm
Ammonia '20- 50, ppm
Phenols ' 3- 11 ppm
Sulffdes < 1- 2 ppm
total Chromium ^O.l-OM ppm
•I i ,
It -nu$t be emphasized that the above concentrations refer to the wastewaters prior to any J
treatment or removal other than the primary removal of oi- and suspended solids in an API
or CPI oil-water separation unit. I
-------
TABLE 7. REFINERY POLLUTANT LOADS AND THE EQUIVALENT WASTEWATER
POLLUTANT CONCENTRATIONS
POLLUTANT LOAD as pounds
of pollutant per thousand
barrels of crude oil pro-
cessed in refinery
' BOD$
COD
TOC
Suspended solids
Oil
Ammonia
Phenols
Sulfides
Total chromium
REFINERY CATEGORY
TOPPING CRACKING PETROCHEMICAL LUBE OIL INTEGRATED
1.2
13.0
2.8
4.1
2.9
0.42
0.012
0.019
0.0025
25.5
76.0
14.5
6.4
10.9
9.9
1.4
0.33
0.088
60.0
162.0
52.0
17.0
18.5
12.0
2.7
0.30
0.085
76. P
190.0
38.0
25.0
42.0
8.5
2. :9
0.005
0.016
69.0
115.0
48.6
20.3
26.2
7.2
1.3
0.70
0.17
EQUIVALENT WASTEWATER
POLLUTANT CONCENTRATION as
pps by weight in the waste-
water flow
BOD5 7 122 240 203 173
COD 78 365 648 507 288
TOC 17 70 208 101 122
Suspended solids 35 31 68 67 51
Oil 17 52 74 112 66
Anmonia 3 48 48 23 18
Phenols <1 7 11 83
Sulfldes <1 2 1 <1 2
; Totai cnxaaJiun < 0.1 0.4 0.3 <0.1 0.4
-------
.-FUTURE TRENDS
General
The future trends in petroleum refinery wostewater generation and the pollutants
' in ".-cstewaters is a function of many inter-related variables. Perhaps the key variables
affecting' those future trends ore those associated with;
o Future environmental regulations
o New refinery technology
o, New refinery feedstocks
o Water reuse ond conservation
Each of the above key variables are discussed in qualitative terms in this section. It
would be very difficult if not impossible to quantify the future trends, and any attempt to
do so is beyond the scope of this section.
FUTURE ENVIRONMENTAL REGULATIONS
i
Th<; generation of wastewaters in refineries will be strongly influenced by the
need to CO: ply with the ever-increasing stringency of environmental regulations. This
.applies r.or only to effluent wastewater regulations. It applies just as well to regulations
which O-H directed at air emissions and at solid wastes. The additional technology and
facilities needed to comply with air emission and solid waste regulations very often
results in additional water usage and additional wastewater generation. For example, if
flue gas scrubbers are required on refinery boilers, process heaters and fluid catalytic
cracking unit regenerators in order to meet future limitations on sulfur dioxide emissions,
wastewoter slurries may be generated which will require handling, treatment and disposal
of a type not currently practiced in refineries.
As another example, future environmental regulations on the quality of refinery
, products such as on the permissible constituents of gasoline (for example, constraints on
the aromatic content or ret roe thy I lead content of gasolines) will have o direct effect on
forcing changes in refinery processing or the development of new processes. Those,
changes or new technology will undoubtedly hove an effect upon the generation and
composition of refinery wastewaters.
Any future environmental regulations directed toward "toxic" or "hazardous"
pollutant discharges will have a direct effect upon the wastevnter treatment technology
required at the "end of the pipe". However, such regulations may also require process
changes or new processes to control the generation of such pollutants. Again, those
changes or new processes may well have a secondary effect upon the generation of
wastewaters.
-------
Obviously, any regulations directed toward an ultimate "zero discharge"
would have a profound effect upon the generation of refinery wastewaters by forcing new
technology for the reuse of treated waste wo ten.
NEW REFINERY TECHNOLOGY
- -The development of new technology for petroleum refining is on on-going,
evolutionary process. As the world's supply of petroleum crude oil diminishes and the
"energy shortage" grows more and more severe, the search for new, technology to obtain
more gasoline and other products from each barrel of crude oil will be greatly intensified.
The new technology that wil,! emergy from that search will bring with it either additional
problems or new problems related to wastewater generation.
While it is difficult to foresee just what, new refinery technology is over the
horizon, one can predict with some certainty that refineries will become increasingly
complex. And it can bs predicted that the increasing complexity will mean additional
problems of wastewater generation and treatment. A brief comparison of the topping
refinery and the integrated refTnery categories in Tables 6 and 7 makes it obvious that
increasing the complexity of refineries not only results in an increased wustewater gener-
ation but it also results in a higher level of pollutants within the wastewaters.
NEW REFINERY FEEDSTOCKS
A* the world's supply of low-sulfur crude oil becomes scarcer, refineries will
have to process increasingly higher-sulfur crude oils. As discussed earlier herein, pro-
' cessing higher-sulfur crude oils will require more and more hydrogen synthesis units in
refineries. Since hydrogen synthesis has a very large steam demand, there will be on
increase in the wostewaters generated by boiler feedwater treatment (softening or
demineralization) and there will be on increase in the generation of high TDS boiler blow-
down. Hydrogen synthesis also produces a process steam condensote which, if not reused
as boiler feedwater, constitutes a major woitewoter source. ,
The search for new or alternative energy supplies will inevitably lead to coal '
liquefaction and to extraction of oil from U.S. western shale deposits?. The crude oils
from those sources will still require refining to produce the conventional range of
petroleum refinery products. Those feedstocks will have very much higher nitrogen
, contents and sulfur contents than the low-sulfur crude oils currently processed in
refineries. The processing of those new feedstocks will compound the need for additional
hydrogen synthesis as well o» possibly create new wastewater generatior problems other
than those associated with additional hydrogen synthesis. ;
WATER REUSE AND CONSERVATION
i x
, Thete is a vast body of technical literature regarding industrial wastewater reuse
and conservation of fresh water intake. It it not the intent of this section to discuss the
w-mony reuse technologies that hove been proposed. Nor is it the intent to review the
66
-------
histories of those specific cases in which water reuse has been successfully demonstrated
or commercially practiced. However, it is quite safe to predict that there will be on
increasing trend toward water reuse in industrial plants of all kinds in order tc comply
with environmental regulations. , .
'As fcr petroleum refineries, their oily wastewaters can be treated so as to
produce an effluent amenable to selected reuse within the refineries. However, most
refineries have four wastewoters which contain a high leve! of inorgunic salts (see Figure
8) and it will be most difficult to treat,those four wastewaters for reuse:
o Desolter effluent brine
o Cooling tower blowdown
o Boiler blowdown
o Intake water treatment wastes (brines).
The technologies do in fact exist for desalinotion of such wastewaters, but they are very
expensive and require a very large consumption of energy. Removal of the inorganic,
weter-soluble salts from those wastewoters creates the problem of how to dispose of the
salts in o manner which prevents their disriolution in the ^ext rainfall. In other words,
desalination of wostewcter brines and transport of the removed salt to a disposal dump
merely transfers the problem without providing an ultimo*" resolution of the problem.
The trend towards water reuse and fresh water conservation in petroleum
refineries will probably include:
o Replacement of once-through cooling water systems with circulating systems
using evaporative cooling rowers. '
o Raising rhe level of concentration cycles within existing circulating cooling
water systems by reducing the amount of blowdown.
I o More usage of air-cooling rather than water-cooling.
' o More intensive efforts to reduce water-cooling and steam-heating needs by
1 utilizing more process heat recovery.
o Replacement of injected steam into refinery side-cut strippers by external,
1 steam-hooted reboilers in order to reduce the generation of sour condensates.
i o. Reuse of biotrested wastewaters (containing low levels of inorganic salts) in
i selected services such as low pressure boiler feedwater.
i •
! o More research into the problems of desalination and disposal of the resultant'
i , inorganic salt wastes.
-------
SECTION 4, ' ,
A REVIEW OF POLLUTANTS IN PETROLEUM REFINERY WASTEWATERS
AND EFFECTS ON AQUATIC ORGANISMS
Sterling L. Burks
Oklahoma .State Univarsity
Stillwater, Oklahoma
ORGANIC CHEMICALS
. ' •
Petroleum refineries use water to wash t!ie crude oil stock or fractions derived
from the crude oil stock. The process of washing crude oil with water for removal of
inorganic salts is referred to as desalting. During the desalting process, the contact
water is contaminated with dissolved or emulsified organic chemicals from the crude oil
stock. Water is also used for stripping of undesirable chemicals such as hydrogen sulfide
and ammonia from overhead flue gases in thermal and thermo-catlytic process units. Con-
densed water from stripper units will also be contaminated with volatile organic chemicals
produced by cracking of heavier hydrocarbons. Phenolic type compounds may occur at
high concentration levels in stripper condensate waters. Several other compounds, such
as chemical compounds used within the refinery to control scale and corrosion on cooling
towers, to reduce frothing in desalters, and other miscellaneous uses, in addition to those
originally present in the crude oil may be picked up as contaminants in the wastewater.
Therefore, the potential number of organic chemical compounds which can occur in
wastewater from a petroleum refinery is essentially equal to not only the number of com-
pounds present in the original crude oil stock but also to that resulting from fracturing
processes and chemical additives used within the refining operations.
It is beyond the scope of this section to discuss all of the organic chemical com-
pounds which could potentially occur in petroleum refinery wastewaters. Essentially, any
or all of the compounds identified in crude oils could occur atfrace levels in either
suspension, emulsion, or solution form in the wastewater. Several analyses of crude oils
from various geographic locations and geological strata have indicated that while most
crude oils contain many compounds in common, the percentage composition of these
common compounds varies tremendously. Also, there may be uniquely different compounds
in oils derived from different geological formations. The reader is referred to other
reviews for'a more detailed discussion of the organic chemical composition of crude
68
-------
The complexity of organic contaminants in petroleum refinery wastewaters has
stimulated investigation and development of many different methods of extraction for
isolation and identification of some or all of the components. For the sake of brevity,
n-.ost of the isolation techniques can be grouped into major categories. A brief discussion
of the techniques is presented. '
Liquid/Liquid Extraction
i .
This extraction procedure is based upon the principal of differential solubility of
organic chemicals in polar and'nonpolar solvents . Water is a very polar solvent and,
in relative terms, will not dissolve large quantities of nonpolor solutes such as hydro-
carbon molecules. In contrast, nonpolar solvents such as hexane, benzene, and
methylene chloride will, in relative terms, dissolve large quantities of nonppiar solutes.
Mixing of two immiscible solvent phases such as water and hexane will result in a
selective partitioning of the nonpolar solutes from the aqueous polar phase into the non-
polar hexane phase. If this process is repeated several times, most of the nonpolar
sol'jtes can be extracted from the aqueous polar phase into the nonpolar solvent phase.
This procedure is widely used for extraction of nonpolar organic contaminants from water
and wastewaters. The most extensive use of this procedure has been for extraction of
. pesticides and herbicides from water. Percentage recoveries of spiked samples.have been
90 percent or better for most pesticides.
Adjustment of pH of thr aqueous phase prior to extraction results in a change of
the polarity of solutes in the aqueous phase. Organic acids exist ai ionized polar mole-
cules at neutral or alkaline pH's. A decrease in pH towards the acidic range will shift
equilibria of the organic acids towards the un-ionized nonpolar form, which is more
soluble in a nonpolar organic solvent. Similarly, adjustment of ,pH of an aqueous solution
towards the alkaline range will shift ionization equilibria of basic organic molecules
toward the un-ionized nonpolar form, which can then be extracted by a nonpolar organic
solvent. By extraction of an aqueous solution with a nonpolar organic solvent at neutral,
acidic, and alkaline pH's, three different groups of organic solutes can be isolated. This
procedure simplifies the resultant extracted fractions and benefits subsequent identifi-
cation. This procedure has been adopted by EPA for analysis of the "consent decree"
priority pollutants. The protocol procedure calls for extraction of the water sample with
pH adjusted to 12 to obtain a base-neutrai fraction end subsequent extraction of the
water with pH adjusted to'2 to obtain an acidic fraction.
Adsorption/Solvent Extraction
This procedure utilizes the unique property of some solid substances to adsorb
nonpolar solutes from aqueous solutions. The most commonly used substance for this
purpose is activated carbon. Activated carbon will adsorb approximately 20 to 30 per-
cent of its weight of nonpolar organic solutes from an aqueous solution. This unique
property has stimulated its use for concentrating trace organic chemicals from aqueous
solutions* . The adsorbed compounds were sequentially extracted with a nonpolar
organic solvent (chloroform) and a semi-polar organic solvent (ethanol) to obtain
' . 69
-------
nonpolar and polar fraction, respectively. The carbon adsorption method (CAM) was
successfully used to isolate and identify several organic compounds responsible for taste
and odor problems in drinking.water^ . There have been numerous studies of the
advantages and disadvantages of this procedure which have b>een reviewed ' .
180
Weber and Morris reported that the rate-limiting factor in the carbon
adsorption process was introparticulate diffusion of the solute molecules within the m?cro~
pore structure of the granular activated carbon. Aqueous solubility of the solute com-
pound affects the initial surface adsorption. A slightly soluble organic compound with a
structurally large molecule adsorbs rapidly from aqueous solution but diffuses slowly with-
in the micropore structure of the carbon. Thus, the compound rapidly saturates the
carbon surface and the adsorption rate declines rapidly. Thus the efficiency of activated
carbon for extracting trace organic solutes from water appears to be related vo two
factors; i.e., surface adsorption and intraparticulate diffusion within the microporous
structure of the activated carbon.
One of the major advantages of the CAM was the capability to adsorb trace
levels (ng/£ or Ug/-4) of organic contaminants from large volurves of water. The resultant
chloroform extracts would yield several grams of organic contaminants which could then
be further fractionated by classical solubility separation procedures. Prior to ihe advent
of sensitive analytical instrumentation it was necessary to obtain gram quantities of
extract to provide enough sample for identification of individual compounds.
Some of the disadvantages of the CAM were: non-quantitative recovery of some
organic solutes from the activated carbon, catalytic or microbial mediated changes in
chemical structure on the carbon surface, preferential adsorption of some solutes, and
selective displacement of solutes1 ^'83'^" .
Another solid adsorbent which has been used for concentrating trace organics
from water is XAD-2 resin. This resin which is a polymer of styrene-divinyl benzene is
classified as having a low polarity adsorption surface. Its efficiency has been evaluated
by Junk**^ and varied from 89 percent for phenols extracted from acidified water to 101
percent for organic acids extracted from acidified water. The range of compounds tested
were alcohols, aldehydes, esters, acids, phenols, halogenated compounds, polynuclear
aromatic?, alkyl benzenes, nitrogen, and sulfur containing compounds, pesticides and
herbicides. The general procedure involves passage of from 1 to 100 liters of water
through 2 grams of XAD-2 resin contained in a glass column. The adsorbed solutes are
subsequently eluted with 25 m-fc of diethyl ether, which is then dried and concentrated to
1 nrvi. The concentrated eluates are subsequently analyzed by gas chromatography (GC)
or a combination GC-mass spectrometry (MS).
Physical Extraction Techniques
The mixture of organic substances extracted by liquid/liquid extraction tech-
niques or adsorption/solvent elution techniques are often too complex for adequate
resolution by conventional GC. As a result, several physical separation techniques
4 .
•' 70 !
-------
have bean developed to isolate simpler fractions. One of the earliest methods utilized
was steam distillation to extract volatile compounds which were subsequently extracted
and concentrated in organic solvents. The resultant extract contained a less complicated
mixture of organic compounds which could be resolved by conventional GC. Sttam
distillations have been used by Ogata and Miyake and Ogata and Ogura to identify
petroleum derived compounds causing odors in fish flesh, by Acknaaxind Noble to
identify petroleum hydrocarbons from fair-ted fish, by Dorris, et at. to identify
aliphatic hydrocarbons in oil refinery wastewaters.
Another procedure which has been developed recently is the use of an inert gas
such as nitrogen or helium to strip volatile organic' compounds from water which are
adsorbed on a Tenax-GC trap. The compounds are subsequently thermally eluted directly
onto a GC column for analysis . This procedure has proven to be effective for analysis
of volatile organic compounds in drinking water and industrial wastewaters. Severs!
trihalomethanes, halogenated ethanes, and volatile hydrocarbons such as naphthalene,
toluene, and benzene have been identified in surface waters using this technique .
ANALYTICAL INSTRUMENTATION . • • - -
Gas Chromorography
A major obstacle to identification of trace organic contaminants has been the
complexity of the mixture of organic compounds dissolved in wastewaters. Classical
' solubility separations are time consuming and usually not adequate for resolving complex
mixtures into individual components for identification. Also, fractional distillations can-
not resolve individual compounds from complex mixtures. In gas liquid chromatography
(GLC), the separation of a mixture of organic compounds is dependent upon partitioning
between a moving gaseous phase and a stationary liquid phase coated on an inert support.
Differences in affinity of the individual compounds for liquid phase results in a varying
partition rote between the moving gas phase and the stationary phase and therefore
separation ensues''''**. In GLC, separation or resolution of compounds depends upon the
equilibrium distribution of substances between the gas phase and the column stationary
liquid phase. For fixed conditions of gas flow, column material, size and length, tern- •
perature, etc., there is, a specific ebiion time for most chemical constituents. Therefore
the chromatographic column is the heart of any GC 'nstrument, and the degree of
resolution that can be obtained is dependent upon the characteristics of the column.
i
Perhaps the best example of the value of GC for separation and identification of
complex mixtures of organic compounds was the progress made on the API Project ^6. In ,
1927, API established research project #6 at the Notional Bureau of Standards. In 1950,
the project was moved to the Carnegie Mellon Institute of Technology in Pittsburgh,
Pennsylvania. In the early years of this investigation, only systematic distillation was
j used until around 1935 when fractibnation by selective adsorption on silica or alumina
' was also introduced. It was not until 1956, when GLC finally came into use for
separations. Rossini reported isolation of 175 hydrocarbon compounds from the crude oil.
.This was the result of 33 man-years of work. In 1961. Destv reoorted on the GC analvsis
\ f
71
-------
of the €4 through €9 fraction of Ponca crude, representing about 30 weight percent of
the total' crude oil. By using a 900 feet x 0.006 inch ID glass open hjbular column
having over 1 million effective plates and operated at room temperature, 122 peaks were
resolved in 24 hours. In contrast, in the API project *6 only 89 hydrocarbons had been .
.separated in this region.
. . Capillary columns have long been recognized as probably the types of columns
which had the greatest potential for resolution of the complex mixtures of organic com-
pounds extracted from water. However, capillary columns can accommodate only very
small injection volumes without overloading and subsequent loss of resolution. The
maximum quantity of sample which could be injected ranged from 0.01 to 0.1 M^ including
the carrier solvent. This limitation lead to the development of special attachments which
split the injection 100:1 or 1000:1, venting most of the' sample away from the cojumn.
Most environmenrcl contaminants ran occur at concentrations of a few ng/4'or Mg/^ rind
GC detectors could not detect rhe small quantities of compounds after split injection.
This limitation prevented successful application of capillary column GC for analysis of
environmental contaminants. Therefore most of the analyses have been performed with
packed columns which could b« injected with from 1 to 10 U/- of sample, without affecting
resolution of the mixture, but had limited resolution for very complex mixtures. However
several researchers successfully used GLC retention times on packed column* to identify
trace organic compounds extracted from surface waters 14,08,/y^
The GC was combined with other confirmatory techniaueS such as infrared to
identify compounds associated with petroleum refinery wastes^ and MS to identify
petroleum products''°. These investigations were perhaps influential in the development
of combined GC with MS, the most powerful analytical development at the piesent time.
', Combination Gas Chromotogroph~Moss Spectrometer'System • I
( *,
' ' Two developments hove occurred within the last few years which have signifi-
; cant.ly improved capabilities to identify trace organic contaminants in water and waste" '
waters. Perhaps the'most significant was development of a combined GC-MS system. ,
Early attempts to combine the GC with the MS encountered many problems. One of the
most troublesome was the interface between the two instruments. The GC uses high
: pressure carrier gas as the mobile phase to affect separation of the compounds, whereas
the analyzer section of an MS must maintain a high vacuum of approximately 10~4 Torr.
' Therefore, special interfaces hod to be developed which could remove most of the carrier
gas without removing all of the organic compounds of interest. A second major problem
was the time of elution of a chromatographic peak. Many peaks from GC columns elute '
.within seconds or at least fractions of.a minute. Existing MS usually require several i,
i minutes to scan a compound. Therefore an MS which could scan o GC peak in a few
i seconds or less had to be developed. \
: , I
; Another major development was the splitless injection technique of Grob and
•! Grob' . This technique permitted injection of samples dissolved in a solvent without i
{splitting, which prevents loss of sensitivity without overloodina the caoillarv column with
72
-------
solvent. With the aid of the Grob-t-. ->e splitless injection technique, glass capillary GC
with up to 600,000 theoretical plates can be used to resolve up to 125-150 specific com-
pounds. Combining the resolution capability of glass capillary GC ro separate complex
mixtures with the capability of the MS to provide molecular weight and election impact
fragmentation data has resulted in the most powerful instrument ever developed for '.
analysis of environmental contaminants. A simultaneous development of computers
capable of rapidly accumulating the data from the MS was also necessa-y to permit full
utilization of the capability of the combination GC-MS.
The impact of these developments is clearly illustrated by Hie fact that in 1970
'only 66 specific organic chemical compounds had been positively identified in surface
freshwaters of the world. In 1975, Junk reported identification of over 300 specific
orCjKJnic compounds from drinking water alone. In a recent symposium on "Identification
and Analysis of Organic.Pollutants in Water," 31 of 36 papers us-ad GC-MS as the major
instrumental technique'^.
Another factor which has had a major impact upon the current stare of knowledge
concerning the composition of trace organic compounds in wastewaters was the consent
decree established between the Natural Resources Defense Council and the Director of
EPA/ Russell Train. Thfs decree mandated EPA to perform a survey of 128 priority
pollutants in industrial wastewaters to determine if these toxic or deleterious compounds
were being discharged in final effluents. This judicial decree initiated the most intensive
and brood-scale investigation of trace organic contaminants in the wprld. The API .
collaborated with EPA in a survey of 17 oil refineries to determine if any of the 128
piiority pollutants were present in either the influent feedwarer, the influent wastewcter,
the influent wastewater to the treatment system, or persisting after treatment in tho final
affluent10. :
; i
COMPOUNDS IDENTIFIED !N OIL REFINERY WASTEWAfERS
Prior' to development of combination GC-MS, most investigators were unsuccess*
ful at identification of specific organic contaminants in oil refinery wastewaters.
Researchers in the 1950 to 1970 period used GC retention times as one of the major
analytical parameters. Solvent extracts from oil refinery wastewaters were so complex,
that the chromatograms contained many unresolved peaks. A large number of unresolved
peaks were often encountered in the region between where n-decane and dodecane '•
'standard hydrocarbons would elute. The unresolved peaks were so numerous that the '
chrcmatograph contained a hump in this region. This was encountered so frequently, that
: it was used as a distinguishing characteristic for petroleum derived chemical contaminants.
; Rosen and Middleton used the ccrbon adsorption method to concentrate trace
organic compounds from lake water in order to obtain sufficient quantities of organic
'.compounds to identify. The compounds were solvent extracted from the activated carbon
j and chroma tog raphed on silica gel columns to obtain aliphatic, aromatic, and oxygenated
groups of, compounds. The resultant fractions were analyzed by infrared (IR) spectro-
.»photometry. The IR "fingerprint" spectra of extract* from the lake water were similar to ,.
73
-------
extracts from on oil refinery wcstewater. Thus, although specific compounds could not
be identified, similarities of the IR spectra were judged sufficient to indicate contami-
nation of the lake water by aliphatic and aromatic compounds discharged from the
refinery .
A similar procedure was used by MJddleton, et al. to isolate odoriferous
compounds from the drinking wo'fot iopply of Nitro, West Virginia. The steam volatile
aromatic frcction was the most odorific. The IR spectra of this fraction was very similar
to that of kerosine (APCO blend).
• 1 1 R
Melpolder, et al . demonstrated the utility of MS to identify and quantify
known mixtures of petroleum type hydrocarbons in aqueous solutions.' The hydrocarbons
were stripped from the aqueous solutions with hydrogen gas and collected on liquid
nitrogen traps. The collected compounds were thermally eluted into an MS through a
sintered glass inlet. Quantitative recoveries were obtained for levels of kerosine and
furnace oil at 0.1 mg/A, gasoline at O.C1 mg/2, and a standard mixture of hydrocarbons
from pentane (£5^12) to decone (ClQ,H22) a* 0.014 mg/£.
In 1967, Brady piesentad a summary of an API sponsored project conducted by
A. D. Little, Inc., to identify the taste and odor causing compounds in oil refinery
wastewaters. Using pentjne extraction, column chromarography, GC, and nigh
resolution MS; the investigators were able to identify several groups of compounds in
odor-rich fractions. A group, of compounds responsible for the "burnt-rubber" odor was
identified as diary I sulfides. One particularly odorific compound identified in this group
'we: diphenyl cJisoifide. Other compounds identified in this fraction were dimethyl
naphthalene and dimethyl anthracene.
Steam distillation fractions of oil refinery wcstewaters appeared to concentrate
• acutely lethal substances from oil refinery wastewaters as indicated by acute static
. Daphnio bioassays^* Solvent extract's of the steam volatile fraction were subjected to
analysis by GO-MS. A homologous series of normal aliphatic hydrocarbons from ,
; n-undecane (C]]H24) to n-octodecane (ClgHsg) was identified in extracts from four
•different refineries. The compounds m-cresol and dioctyl phthalate were identified in
, extracts from two different refineries, however no compounds were identified which could
t fully account for the acutely lethal effects observed.
' ' ' *5Q
i Burlingame, et al. analyzed solvent extracts from three different locations at
pn oil refinery wastewoter treatment plant. The extracts were designated as phenolic,
acidic, basic, and neutral fractions based upon pH of wastewater during extraction. The
extracts were analyzed on a capillary column GC-high resolution MS with computer
, analysis of the data. Based upon the mass spectral data, four specific, compounds in the
. phenolic fraction and one in the acidic fraction were identified (Table 8). In addition,
1 11 different classes of compounds were detected in the acidic fraction and 15 different
! classes of compounds in the neutral fraction (Table 9). There were several isomers present
1 within, most classes of compounds. Most of the compounds or classes of compounds identi-
>.*fied by Burlingame, et al. ^ hove been identified as components of crude oils by
74
-------
TABLE'S. PARTIAL LIST OF COMPOUNDS IDENTIFIED IN EXTRACTS FROM OIL
REFINERY WASTEWATERS (MODIFIED FROM BURLINGAME. ETAL., 1976)
Coc.pound Type
i
Anisole
Methyl Anisole
C Anisole
Methyl Benzoate
Saturated Methyl esters
Saturated Ethyl esters
Olef inic Ethyl esters
Saturated Propyl esters '
Cyclic Alkyl Methyl esters
Alkyl-sub. Methyl Benzoatef
Phenylalkyl Methyl esters
Alkyl-sub. Naphthenic Metnyi
esters
Indenic Methyl esters
Sulfur-sub. Aromatic Methyl'
_ _. A. - _
Formula Present in
No. of Scans
C7H8°
C8H10°
SH12°'
?8HB°2
CnH2n°2
CnH2n°2 '
CnH2n-2°2,
CnH2n°2
CnH2n-2°2
CnH2n-8°2
CnH2n-8°2
CnH2n-«°2
CnH2n-12°2
CnH2n-6°2S
•
1
1
1
1
28
, 6
1
1
4 '
16
5
1
3
Fract ion
Phenol ic Acidic
X X
X
X
X ' X
X
X
X
X
X
X
X
X
X
X
Alkyl-sub. Methyl su 1 fides
75
-------
TABLE 9. PARTIAL LIST OF COMPOUNDS IDENTIFIED IN NEUTRAL EXTRACTS FROM
OIL REFINERY WASTEWATERS BY BURLINGAME, ET AL., 1976
Compound Type
n-alkanes
branched alkanes'
mono-saturated or
mono-cyclic alkanes'
alkyl benzenes
naphthalenes
phenenthrenes or
anthracenes
pyrene or f luoroanthrene
alkyl biphenyls
•methyl indan
i
elkylated phenols
'Thlacyclanes
benzothiophenes
Formula
CnH2n+2
CnH2n+2
,CnH2n
CnH2r-6
CnH2n-12
CnH2n-18
CnH2n-22
CnH2n-14
C10H12
CnH2n-6°
CnH2nS
CnH2n-10S
Samp
K-l-N
n = 11-33
series
n = 11-28
n.= 9 ,13)
n = 10 (6)
*n = 11
*n = 10-13
(several )
n = 14-19
n = 16
minor
n = 13
sign! f icant
n = 14
trace
n = 7-12
*n = 8 4 9
n = 6
n = 8
n » 8-12
le location
K-2-N
n = 12-33
n. d.
n = 9
n = 10 (yesl
n = 10-14
n = 14-19
**n = 17
n = 16
m i no:-'
i
n. d.
n. d.
n - 8-11
n » 8-11
K-3-N
n = 15
n. d.
n. d.
n = 8 (2)
n = 9 13)
n = 10 (3)
n = 12-14
(yes)
n = 15-17
trace
n. d..
n = 12
n = 13
trace •
n. d.
n = 7-13
n = 6-11
(several )
n = 9
n - 10
(few),
* rrwjor constituent of extract.
** relatively rbundant
Numbers In parentheses indicate number of isome.-s detected.
4
76
-------
previous researchers. Therefore, Burlingame's analyses confirm what many researchers .
have suspected; any compound originally present in crude oil could potentially occur in
oil refinery wastewaters. Perhaps the most significant result of their analyses was the
detection of some isomers of n-alkanes, olkyl benzenes, naphthalenes, phenanthrenes or
' anthracene;, alkyl biphenyls, etc. in the influent and effluent from the biological treat"
ment system. These resulrs confirm that some classes of compounds are resistant to bio-
'logical treatment ond may persist in the final effluents.
Raphaelian ond Harrison'^ analyzed 24-hour composite samples of wastewaters
from dissolved air flotation (DAF) and final clarifier (FC) units of Class B refinery acti-
vated sludge treatment system. Solvent extracted neutral, acidic, and basic fractions
from the wastewater were analyzed by Grob-type'splitless injection .glass capillary GC~ ,
MS. Over 300 compounds were identified in the neutral fraction from the DAF unit
wastewaters based upon GC retention time and presence of major ions.in mass spectral
scans. The concentration of the compounds was semi-quantified by either the area or
peak height of major ions in the mass spectra (Table 10). In addition, Raphaelian and
Harrison also analyzed the trace organics after passage of tho oil refinery wostewater
through a mixed-media filter/activated carbon (MMF/AC) pilot-scale treatment system.
Identification of over 300 compounds in the neutral fraction from t!ie DAF unit waste-
waters illustrates the complexity of organic contamination in oil refinery effluents. The
majority of the compounds identified were aromatic and alkylqromatic types.
The relative efficiency of biological and MMF/AC treatment systems for removal
of specific compounds can be determined from Raphael ian and Harrison's data. The
activated sludge unit was mere effective in removal of aromatic than long chain aliphatic
hydrocarbons. The long chain aliphatic hydrocarbons, greater than C]Q, were
especially difficult to remove. The effectiveness of the activated sludge treatment
system for removal of polynuclear aromotics could not be determined since the concen-
trations of these compounds were too low for accurate quantification.
; I
Several organic compounds were detected in final effluents during the cooper-
ative survey of 17 petroleum refineries by the EPA and API. The most frequently
detected compounds were methylene chloride, benzene, toluene, chloroform, trichloro-
ethylene> ethyl benzene, pyrene, di-n_-butyl phthalote, and bis(2~ethylhexyl) phthaiate.
The concentrations of these contaminants ranged from less than detectable to a maximum
of60ug/X. , , !
; A significant result of the cooperative EPA-API survey was a measure of the
! accuracy and reproduability of analyses for trace organics. The range in percent
recovery of spiked samples for polynuclear aromatics was from'56 to 72 percent for the 5
| to 20 Ug/4 range ond from 78 to 104 percent for the 33 to 220 Vg/t range. The overall
' percent recovery of phenol spike* was 65 percent and the quantity of the spike did hot
appear to offect extraction efficiency. The efficiency of liquid/liquid extraction for
, recovery of" phthaiate spikes was very erratic with a range from 4 to 300 percent. Over-
1 all, the accuracy and reproducibility of the liquid/liquid extraction procedure appears
to be inadequate at this time. Considerable effort should be expended to improve the
\ . ...""•.',"" : " • • " " '*•
.._ 1 :_ .. :? '.:..: 77 ..:... _ ...
-------
TABLE 10. LIST OF ORGANIC COMPOUNDS IDENTIFIED OR DETECTED IN NEUTRAL
EXTRACTS FROM THE EFFLUENTS FROM A DISSOLVED AIR FLOTATION
(DAF), AN ACTIVATED SLUDGE FINAL CLARIFIER (FC), OR A PILOT-
SCALE MIXED-MEDIA FILTER/ACTIVATED CARBON (MMF/AC) TREAT-
MENT SYSTEM. MODIFIED FROM RAPHAELIAN AND HARRISON, 1978
'
Compound Name
Ch lorof orm
1 ,1 >l-tr ich lor oe thane
benzene
• carbcin tetrach lor ide
eye lohexone .
toluene
ethyl benzene
p-xy lene
n>-xylene
o-xylene
n-nonane
I-oropyl benzene
n-propyl benzene
m-ethyl toluene
p-«thyl toluene
1 ,3,5-tr imethy 1 benzene
o-«thyl toluene
1,2,4-trimethy 1 benzene
cycloalkane
i-butyl .benzene
6-butyl benzene
n-decane
J, 2, 3-tr imethy 1 benzene
m-isopropyl toluene
o-!sopropyl toluene
p-isoprr.pyl toluene
-Indan
Indene
m-diethyl benzene
m-n-propyl toluene
p-n-propyi toluene
n-.'iutyl benzene •
1 ,3-dimethy l-5-ethy 1 benzene
o-n-propyl toluene
1,4-dimethy l-2-ethy 1 benzene •
ethyl styrene
1,3-dimethy l-4-ethy 1 benzene
tthyl styrene
1 ,2-dimeth/l-4-ethy 1 benzene
Relat i ve
Cone.
In DAF Neu-
tral Traction
•
h igh
high
med i urn
h igh
high
h igh
low ,
high
high
med i urn
low
trace
low
' medium
medi urn
low
low
high
trace/medi urn
trace
trace
med i urn
med i um
trace
trace
trace
medium
trace'
low
tow
low
trace
low
low
low
low
low,
med ! um
low
Presence)*) ,
Absence(-) '
(FC effluent)
4
4
4
4
1 4
4
4
4
4
+
•f
•f
+
•f
•f
•»•
•f
+
T/*
+
•»•
•»•
+
T
-
-
4
4
, T
* • ,
4
T
4
4
NM
4
4
4
Presence! 4) ,
, Absence (-)
(MMF/AC
effluent)
4
4
4
4
4 ' ,
4'
4
4
4
4
-
-,
•V
-
.
«•*
T
4
-
-
-
4
-
4
_
-
-
, -
-
-
-
NM
• -
-
, -
-
(continued)
•t -:
78
-------
TABLE 10 continued
Compound Name
1,3-dimethy l-2-ethy 1 'benzen'e
1 ,2-dimethy l-3-ethy 1 benzene
C.-benzene
5
1 ,2,4,5-tetramethy 1 benzene
1 ,2,3,5-tetramethy 1 benzene
n-undecane
2-methyl indan
1-methyl indan
1 ,2.3,4-tetramettvy 1 benzene
tetra'in • . ,
nephtha'ene
C,-benzene (16)
o
n-dodecane
ethyl indan
C.,-elkane
dTftethyl indan ,13)
ffiwthy 1 tetral ! n
C,- indan
methyl benzothiophene (4)
methyl ethyl indan
2-methyl1 naphthalene •
trimethyl indan (3)
C -indan/Cj-tetral in (7)
dimethyl tefral in
n-tr idecane
blphenyl
dimethyl benzothiophene (5)
et;\yl benzothiophene (2)
tthyl naphthalena
dimethyl naphthalene 16)
C -elkane (2)
n-f etradecanw
acenephthene' ,
methyl biphenyl 12;
C.-naphthalene (14)
C^.-alkane
n-pentadecane
f luorene
C.-biphenyl i4!
mlthyl acenaphthene ,(3)
n-hexedecane
Relat i ve
Cone.
in DAF Neu-
tral Fraction
low
low
trace
low
medium
h igh
medium
medium
medium
low
h igh
trace
h igh
low
h igh
medi urn'
med i urn
trace
low
trace
h igh
trace
trace
low
high
low
tr.ace
trace
med i urn
high
high
high
trace
low
low to, high
high
high ,
low
trace
low
high
Presencel+l ,
Absence(-)
(FC effluent)
:
T
NM
•f
+
, -f
T
•f
•f
-
+
•»•
+
-
+
T
T
_
+
- .
•*•
-
+ /-
T '
4
•»•
•f
+
•f
4
•f
4
+ ' ,
•f
+
4
•f
NM
NM
NM/+
4
Presence!-*-) ,
Absence(-)
(MMF/AC
ef f luent)
^
NM
^m
+
'_
_
^B
'_
4
—
4
_
«»
T
-
^m
*•
^*
+
-
«•
•^
*
T
••.
-
'
4
4
4
-
•»
4
4
4
NM
-
-
4
(continued)
79
-------
TABLE 10 continued
Compound Name
C.-blphenyl (5)
mifhyl fluorehe (3)
C_-ecenephthene (5)
n-heptadecane
d i benzot h i ophene
pr Istane
anthrecene/phenanthrene
C_-fluorene (7)
n-octadecane
methyl di bensot hi ophene (2)
phytane
methyl ph»?nanthrene (3)
2 -methyl anthracene
1-methyl anthracene'
C, fluoreof (2)
n-nonedecr.ne
C '-di ben: ot hi ophene
C.-phenonthrene/anthracene (8)
ffuoranthrene
C--phenanthracene/ anthracene
n-eicosane
C,-phenanthrene/enthracene (<5)
pyrene
n-henei cosane
C -H PNA .(61
n-docosane
C H PNA (3)
cnrysene
1,2-benzanthraccne
n-tttracosane
n-pcntacosene
phthalbte 12)
Relative
Cone.
in DAF Neu-
tral Fraction
1 trace
low
i
low
high
low
high
,high
low
high '
low
med i urn
medium
low
low
trace
' high
trace
trace/ low
trace
trace
high
trace
low
rr.ed i um
trace
med i urn
trace
trace
low
low
low
medium/ high
Presence!-*-) ,
Absence!- I
(FC effluent)
NM
*
NM
•*•
T
•*•
•f
NM
NM/f
+
NM '
•i
•f
+
•f
^»
-/+
NM
NM
NM/T
•f
+
NM/T
•*
4/T
•»•
•f
+
•»•
Presence!-*-)
Absence(-)
(MMF/AC
effluent
-
-/T
— .
4
-
-
•f
_
•f ' '
1
+
T
T
T
T/-
+
-
-
-
•t
-
-
*
^
^
^
-
-
•f
NM
•»
!
80
-------
accuracy and precision of this procedure or to develop new and better p.ocedures.
The results cited in this review illustrate the rapid advances in analyses of trace
organics which have occurred since the development of commercial computerized GC-
MS systems. Additional investigations in the next few years will undoubtedly identify
many more new organic compounds. Detailed knowledge of chemical compounds in the
Influent and effluent streams from wastewater treatment systems will permit more realistic
evaluations of their effectiveness. In addition, delerterious effects of oil refinery waste-
waters upon aquatic organisms can be correlated with concentration of specific toxicants
rather than attempting to correlate with collective measures of organic chemicals such as
COD or BOD.
i i '
EFFECTS,OF PETROLEUM REFINERY WASTEWATERS
Acute Lethal Effects
Short term lethal effects of wastewaters are customarily determined with a bio-
Ossay test. Basically, this test consists of exposing a number of individuals (6-10) of tie
same species of organism to selected percentage concentrations of a wastewater for a
predetermined interval of time (1-4 days) and recording response ol the organisms. Death
of test organisms, sometimes referred to as the quanta! or "all or none" response, is an
easily recognized and the most frequently used measure of effect. The most reproducible
measure of response of a group of organisms is the median or 50 percent response. There-
fore, the results of acute tests are usually expressed as median lethal concentration or
median tolerance level for a specified time of exposure, i.e., 96 hour LC50 or 96 hour
TLm. The specific procedures for performance of acute toxicity tests with aquatic
organisms are presented in detail elsewhere (Doudoroff, et ai., APHA, and EPA). There-
fore, u detailed discussion will not be presented in this section. However, some of the
subjective decisions which may affect the results of the test and possibly the interpre-
tation of the data will be discussed.
The choice of dilution water is very important because of the probable occur-
rence of other chemicals in the dilution water which could affect the toxicity of some
contaminants in the effluent If possible, dilution water should be obtained from the
receiving stream above the point of wastewater discharge. Change in pH can in crease cv
decrease the lethal effects of ufniT.onia (Wurhmann and Woker ), hydrogen sulfide
(Smith, et ol.)/ o*1^1 cyanide (Doudoroff0'). In general, shifts in pH which result in ,
greater concentration of the un-ionized equilibrium form result in an increase in lethal
effects of the contaminant. b:creq»e$ in hardness of water decrease the lethal effects of
heavy metals (Mount; Pickering, and Hunderson'^8) and also the presence of organic
chelators decrease the lethal effects of heavy metals (Sprogue, et at. ^'). Lee pre-
sented a good summary of many chemical factors which should be considered in toxicity
testing. Obviously, the dilution water should duplicate conditions in the receiving
stream, after allowance for mixing, as near as possible. If water from the receiving ,
strsam cannot be used for dilution, then the following choices should bo considered in
Border of preference; untreated well wofe/, reconstituted water prepared from distilled
" ! • ; ' V,
. .81
-------
wafer, and dechlorinated tap-wafer.
It is preferable to use continuous-flow for exposure of test organisms to waste-
waters due to the potential for a decrease in dissolved oxygen concentration and an
increase in metcbollc waste products in static methods of exposure. A comparison of 96
hour LCSO's of fathead minnows exposed to oil refinery wastewaters in continuous-flow
versus static conditions indicated that me continuous—flow exposure was not os lel'iai as
the static exposure (Kleinholz 3). The differences in 96 hour LC50 wens not statistically
significant between the two methods of exposure, however the fish in the continuous-flow
tests appeared healthier.
Species of aquatic organisms differ in susceptibility to chemical toxicants and
therefore care must be used in selection of a species for testing. Within the fishes,
salmon id species such as rainbow trout and salmon are considered to be more susceptible
than most warm water fishes such as largemouth bass, sunfish, and channel catfish.
Irwin" compared the susceptibility of 57 different species of fish by performing 96 hour
static acute toxicity bioosscys of on oil refinery wastewater. The response was used to
rank the species of fish relative to the most resistant, the common guppy (Lebesfes
reticulotus), which was assigned a value of 100.00. The relative ranking of some fish
commonly, used in bioassay tests were: mosquito fish (Gombusio affirms), 74.02; channel
catfish (Ictolurus punctotus), 60.15; goldfish (Corossius ouratus), 51 ?5; fathead minnow
(Pimepholes promelos), 49.19; bluegill sunfish (Lepomis mocrochiais), 54.10; and rainbow
trout (Solmo goirdneri), 34.66. The aquatic invertebrate species have not been used
extensively as test organisms, therefore it is difficult to make generalizations concerning
their relative susceptibility to chemical toxicants. The water flea (Dophnio sp.) is the
most commonly used and is much more sensitive than most species of fish" .
Acute toxicity bioassay tests of oil refinery wastowcters with fish indicate that
' effluents from properly operated biological treatment systems do not cruse greater than
50 percent mortality of fathead minnows or rainbow trout' . Properly operated seconr
dory biological wastewater treatment systems can produce final effluents which cause no
mortality in 96 hours at 100 percent (undiluted) concentrations. Matthews' '^ found that
activated sludge treatment systems reduced the toxicity of the influent feed water from
18-56 percent (volume/volume percent dilution) to 100 percent in the final effluent as
measured by 24 hour TLjg in four different refineries.
, The Oil Refiners Waste Control Council (ORWCC) in Oklahoma has been
cooperating with OkLhoma State University (OSU) in a monthly acute toxicity testing
program since 1959. These acute toxicity tests are performed in accordance with APHA
standard methods for static 96 hour fish biocssay tests^. Graham and Dorris^ reported
some of the data generated by OSU-ORWCC for 1960-1962. The wastewaters from
refineries which used either activated sludge or waste stabilization lagoons consistently
had a mean annual TL^ of greater than 100 percent, i.e., more than 50 percent of the
test fish could survive in 100 percent effluent. Wastewaters from refineries which had
not been treated adequately with some type of biological secondary treatment system
extremely tqxic,_wjth a mean annual TLm of 19 and 21, respectively.
I
; 62. „ ..
-------
Since Graham and Dorris<^ reported Miei<- data, all of the refineries within the
ORWCC have installed the equivalent of activated sludg& secondary treatment or better.
All final effluents tested in the last five years have consistently had Tl-m's &reoter rn°n
100 percent. In fact, rhe method of reporting data has been changed to the convention
of reporting TLjQQ; i.e., the concentration (percentage of effluent) in which 100 percent'
of the test organisms will survive.
Graham and Dorris'^ also performed 32-day continuous-flow fa'heod minnow
bioouays. Wastewaters which were highly toxic in 96 hour acute tests hod to bo diluted
by 90 percent to obtain greater than 80 percent survival. Wostewaters from the biolog-
ical treatment units were relatively non-toxic in short term tests; i.e., 23/25 or 92 per- '
cent of the dilutions tested had greater than 90 percent survival after 96-hour exposure.
However, the wastewaters contcmed chemical toxicants which had either delayed or
cumulative effects; i.e., 17/25 or 68 percent of the dilutions tested after 16 days of
exposure had greater than 90 percent survival whereas only 9/25 or 36 percent had
greater than 90 percent survival after 32 days of exposure. Graham and Doiris observed
that many fish appeared to stop feeding and appeared emaciated prior to death. The
Observation was supported by the condition index, K = weight In grams times 100,000 ,
divided by total length in mm"* (Coriander), which was uniformly 0.52 for all fish that
died, indicating that death occurred whan the fish reached a certain degree of emaciation.
reviewed the literature on effects of oil refinery wastewaters upon
aquatic organisms and was uncbie to locate many articles. Most of the articles were
concerned with the effects of compounds identified from crude oil or crude oil spills.
Cote also reported on the result* of some acute toxicity rainbow trout bioassays performed
on Canadian oil refinery effluents. In general, effluents which had been adequately
treated in biological systems caused less than 50 percent mortality of rainbow trout at
100 percent concentration for 96 hour exposure. Pessah, et al. '* reported similar
results on Canadian petroleum refinery wastewaters. :
In the most comprehensive study located, Sprogue, et al. performed several
types of sublethal tests in order to determine: (1) the sub lethal effects of biologically
treated si! rsfinery effluents, and (2) the type of test which appeared to yield the best
measure of sublethal effects in the shortest interval of time and was simple to perform.
They used rainbow trout, small tropical flagfish, and Dophnio moqno as test organisms in
1 long term, continuous-flow exposures to the wastewaters. Wostewaters were collected and
; transported to the laboratory for performing the tests. Most of the effluent samples tested
over a two-year period were considered to be non-lerh al based upon 96 hour rainblow
trout bioassays. Growth of rainbow trout indicated a .breshold level in the vicinity of
,9.4 percent effluent. The threshold concentration in a 3.5-month chronic test of flog-
. fish was about 9.1 percent for an appreciable effect on final size of males. However,
major effects were only observed in 28 percent or higher concentration o' Affluent. A
significant increase in cough-reflex response by rainbow trout occurred b< tween 25 ond
50 percent effluent. Dophnio experienced a 50 percent failure in reproduction when
jBxposed to 3.1 percent of the effluent for 14 days. The "safe" level at which only 5 _
->*perrent impairment of Dophnio reproduction would occur was predicted to be 0.52
-------
. -vpercent effluent. '
The fish cough-reflex has been suggested to be very promising for predicting
potential long-term effects based upon responses obtained during short-term tests.
Carlson and Drummond^' used the fish cough-reflex of bluegill sunfish lo measure the
effects of an activated sludge treated oil refinery wostewater. Analyses of three different
'-effluent sample* collected over on 8-month period of time, showed the effluent caused
significant ir.cre:,jes in coughs at concentrations as low as 0.8 percent following an upset
in the biological treatment system. After the system had returned to normal, 12 percent
effluent caused a significant increase in cough frequency. Previous investigators had
defected a 200 percent increase i>t rate of coughs at concentrations of zinc just above the
LCcn value of 6.0 mg/4. A large and rapid rise in frequency of cough-reflex of fish
. exposed to wostewaters or chemical toxicants indicates potential toxic effects. Smaller
increases in frequency of cough responses may be indicative of potentially sublethal long-
tnrm effects, Sprogue, et al.'°^ detected the threshold of significant increase in
frequency of coughs by rainbow trout at between 25-50 percent oil refinery wastewater.
In contrast, the "safe" level of effluent for Dophnio was predicted to be 0.52 percent
•effluent. - , '
The detection of an increase in tumors in fish collected from a polluted river
may have serious implications for Seng-term contamination with trace levels of chemicc!
sA. '
toxicants. Brown, et al. found 4.48 percent of the fish collected from the polluted
Fox River had tumorous growths compared to only 1.03 percent in fish from a non-polluted
river. Several classes of contaminants were detected in the polluted river; crude oil,
JQCsoiine, aromatic hydrocarbons, polynuclear aroma tics, ethers, organic acids, organo-
phosphates, triazines, and chlorinated hydrocarbons. The concentrations ranged from
1 0.002 to 0.1 mg/£. In subsequent articles, Brown, et al. also showed an increase in
j incidence of disease in fish from the polluted river compared to the fish from the non-
polluted river. The increase in tumors in fish fron the polluted river may have been
I portly due to an increase in viral infections".
I • ,
I The acute toxicity of oil refinery wastewaters has been demonstrated with many
different types of tests. Dophnio sp. have frequently been used as very sensitive bioassay
! organisms for comparison of other tests'"^. Although Dophnio hove long been recognized
,0$ a very sensitive bioassay organism, they have not been widely used because of.
;unexplainable mortality in laboratory cultures and in controls. Buikema, et al. performed
j some inter-laboratory Dophnia sp. bioassay tes'.'s to determine reproducibility of the test.
j Their results indicated that refinery personnel could be quickly trained to obtain satis- j
i factory reproducibility with Dophnio bioassays. Richardson, et ol. ' showed cell j
j culture assays of oil refinery wastewaters to br more sensitive than 24-hour Dophnio bio- '
I assays. The eel! culture assay would permit exposure of two generations of cells during a
i40-hour test and thus could be valuable for detecting mutagenic effects in a short period
I of time. Another short-term test using enzyme inhibition studies of glucose-6-phosphote
j dehydrogenase was more sensitive than fish bioassay* but not Dophnio or grass shrimp
i^f' bioassays (Rutherford, et al.). The grass shrimp (Poloemonetes pugio) has been found to
;r **4>e as sensitive as Dophnio for detecting a cure lethal effects of finery wastewaters ,
-------
__ ... ._. .. _.. J.. .. _,.,__ ..... . -. - . _ . , -.
-' The chemical toxicants In oil refinery wastewaters which cause short-term acute
r - •*• i f i
toxiciry can be removed by good biological treatment systems. These substances have not
been identified and probably will not be since most refineries within the U.S. have the
equivalent of biological treatment or better i However, as some long-term toxiciry tests
have shown, biological treatment generally does not remove a!i chemical toxicants.
Some substances persist which have been shown to be lethal after 16 to 32 days of
-«nposore70» 30. These substance* 'Con be removed by activate*1 carbon adsorption, which '
. indicates that the substances may be organic in nature. Many of the acutely lethal com-
ponents of oil refinery wastewaters can be removed by simple steam stripping, which
, would indicate volatile substances . Since the long-term toxiciry can be reduced by
'•• activated carbon treatment, it would indicate that these components are organic in
1 nature. In the EPA-API survey of priority pollutants, most of the compounds identified
in tfte base-neutral fraction would have to be judged deleterious to organisms in the
receiving environment.
'•. Long-Term Effects j ,
,""* The long-term effects of any chemical toxicant are difficult to measure and even
; more difficult to predict. The effects may be lethal to selected stages in the life cycle
, of the organi?m aid might not be detected in short-term tests. Larval stages of fish and
oouatic invertebrates seem to be especially susceptable. Chemical toxicants con also
affect the ability or an organism to escape predators or pursue food and thus indirectly
impair the organisms chance of survival. Other sub lethal long-term effects may result in
, a change in the norm.il behavior pattern of organisms and thus impair spawning
' r»^5\ities'** or affect response time of avoidance reflexes'* . Since the long-term
effects ot eve.; ::r*»le rhemical toxicants are difficult to measure, there have been few
investigations of the long-Term effects of complex mixtures such as petroleum refinery
! wostewo ten. ,
.INF
• One method that has been used to obtair. an integrated assessment of the effects
:of all environmental factors is an evaluation of the structure'of the biological communities
j in receiving waters. The balance of aquatic communities can bs used as integrated
'. response to mo.ny different stresses such as chemical pollution, artificial enrichment, or
i severe changes *n physical factors (Hawkes) J Wilhm and Dorris utilized this technique
to assess the comWnad effects of municipal and oil refinery wastewater discharges into a
: creek. A numerical index, diversity of the distribution of individual organisms among ;
• species, was used to compare effects at various distances downstream from the discharges.
j The diversity index of the benthic macroinveftebrate organisms directly below the di$- i
; charges was less than 1.0, whereas it was greater than 3.0 downstream a distance of 60
km. Wilhm and Dorris proposed that a species diversity index of less than 1.0 indicated !
a severe effect upon the benthic community, whereas diversity indices of greater than |
3.0 indicated a non-disturbed community. ' 't
Although numerical indices have been criticized for over-simplification and for
not being sensitive to rare species, several indices have proven to be very useful for
^summcrizing large, quuntities of biojogicoldcto. Severgj indiceat should be used to
85
-------
s- provide the «naximum amount of information possible. In some cases, the disappearance
' of species of organisms known to be sensitive to chemical toxicants should be used to '
supplement the numerical indices. : j
t,l t
I The long-term effects of a bunker oil spill in a river was studied over a two-year
.period following the spill . Many species of phytoplonkton and benthic macroinverte-
'; 'brot«» w«c« «limtnot«doft«r-tt>e spill • Using the indicator species Approach/ McCouley
i showed that Gamma rus, Agrion nymphs, and Dugesio disappeared from the benthic
1 community after the spill. Tne tolerant species of Tubifex, Tenc'ines larvae, Nemota and
i and Hjpjdineo remained in1 the area after the spill . Thus the presence or absence of ' ;
; sensitive species can sometimes be used to interpret water quality. However, this '
! approach must be used with cautior since many factors other than poll ition can affect th^
' desirability of the environment for a particular species. i
< ' - ' ! !
Periphyton, a collective term referring to organi.rms which attach to a substrate
• in a flowing stream, composed mainly of algae and diatoms have been used as effective j
• monitors of water quality. Cooper and Wilhm used plexiglass plates suspended in a
Trtreain receiving domestic and oil refinery effluents to measure productivity, species KCH
I diversity, and pigment diversity of colonizing periphyton. Comparison of the periphyton
' responses directly below the discharges and 60 km downstream indicated that species
' diversity was I'owest near the outfalls and increased downstream. These data were inter- ;
' preted as a measure of an improvement in water quality downstream from the outfalls. i
i
t ! '
Burks and Wilhrrr ' uswd artificial substrates (Hester Dendy samplers) colonized
:wirh benthic mocroinvsrtebrotfiS from a ronpolluted stream as bioassay colonies. The I
i authors demonstrated that species diversity, total number of species, and mean density of!
I individuals could be used to evaluate the effects of an oil refinery wastewoter during a i
1 Sunday exposure in artificial streams. In a comparison of an activated sludge treated, I
t dual-media filtered, and activated carbon treatment methods, activated carbon treated j
; wastewater caused no significant effects upon the benthic macro in vertebrate colonies. In'
1 contrast, (he activated sludge treated wastewater caused a significant decline in species !
; diversity, number of individuals, and mean density of individuals of the benthic macro- I
.invertebrate colonies. i
used *he benthic macroinvertebrafe bioassay* and continuous-flow fat-
'• - 1
! head minnow bioassays to evaluate the effectiveness of sequential pilot-scale dual-media;
filtration and activated cr.-bon adsorption treatment technologies as "add on" systems to |
improve oil refinery wastrwaters from biological treatment systems. The combined results'
of the fathead minnow and benthic macroinvertebrate bioassays clearly showed that
activated carbon would remove toxic substances not removed by biological treatment
j systems. The range in median lethal time (LT^g) of fathead minnows to the biologically
I treated oil refinery effluents was from 11 .5 hours to 28 days. The acutely toxic waste-
water, i.e. LT«jo = 11.5 hours, was from on'octivated sludge system which was
i hydrw!:ca)ly overloaded from rainfall runoff within the refinery. All of the ether bio- I '
f;!:hogfcol treatment systems were operating at near optimum efficiency. In seven, tests per-j |
reformed at four different refineries, there w*?e on'v twn f*»»K in u/h:<-ti <4i»ra um. )»«« fU«»« i
I
-------
percent mortality of the test fish at the end of 32 doys of exposure. Therefore, even ;
' though the effluents were not twvic in short-term exposures, the fish could not survive >
long-term exposures. .Species diversity and total number of faxo of benthic macro- '
1 invertebrates were not as sensitive >o. the effluents as the fathead minnows. In three of ,
; the seven tests, there was no significant effect upon species diversify or number of taxa '
• of benthic macroinvertebrotes exposed to the) biologiccliy treated effluents for 32 doy$i -
j Another long-term effect of oil refinery wastewaters is an impairment of the
! taste/flavor of fish in the receiving waters. Studies inflated in 1952 and conducted ;
i intermittently until 1968 showed a direct relationship between the threshold v^lc.* number
i (TON) of oil refinery wastewaters and tlie presence cf an "oily" taste/flavor of rainbow
I trout"^. Dilutions of oil refinery wastewaters of greater than 1/100 were required to ;
; prevent occurrence of the "oily*1 taste/flavor in trout exposed to the effluent. A panel '
. reported the odor was nor like the odors of naphtha, turbo fuel or diesel fuel. Ogata and
j Ogura exposed greenfish (Girella punctate) and eels (Angulla rostra to) to an aqueous
'solution of u mineral oil composed of 81 percent paraffinic and naphthenic hydrocarbons ,
sand 19 percent aromatic hydrocarbons. The taste and odor in the exposed organisms '•
"appeared to be caused by severe! unsaturated oliphotic~hydr6carbons and possibly some ^"^
aromatic compounds. |
! • ' • '
! Steam volatile extracts of salmon collected from the Great Lakes contained over
185 different organic compounds identified by. GC-MS. Many of the compounds were
suspected to be derived from natural compounds synthesized by foodfish (alewives; 1 - '
octen-3-one) or microbial activity (geosmin). However several of the compounds were j
suspected to have curie from industrial waste discharges; p-methoxy thiophenol,
I naphthalene, 2-isopropyl phenol, 3-iiopropyl phenol, 4-isopropyl phenol, t-butyl
phenol, and chlorobenzaldehyde.
l
Several aquatic toxicoiogists have observed tfat behavior of aquatic organisms
should be a very sensitive indicator of deleterious effects of chemical toxicants'°*.
However, there hove been very few studies of the effects of chemical pollutants upon
behavior. A recently completed study at OSU indicates the utility of using changes in
normal agonistic behavior of orange-spotted;*unfish (Lepomis humilis) for detecting sub-
lethal effects of oil refinery wastewoters. Agonistic behavior, .aggressive acts such as
bites and chases, is important in establishing territories and spawning areos of many fish.
Exposure to non-lethal oil refinery wastewaters for periods of 10 days significantly
decreased the agonistic behavior of the exposed fish compared to controls. The eco-
logical significance of changes in agonistic behavior have not been determined yet,
therefore the practical significance of these effects have yet to be demonstrated. How-
ever, such tests may be important for monitojring other more significant deleterious sub-
lethal effects upon populations ol organisms Within receiving waters of oil refinery waste
waters. , '
-------
SECTIONS
AN EVALUATION Or EXISTING AND EMERGING CONTROL TECHNOLOGY FOR
THE TREATMENT OF PETROLEUM REFINERY WASTEWATERS AND SLUDGES
Davi: L. Fore1
I • Engineering-Science, Inc.
i ' Austin, Texas
INTRODUCTION !
) i *' i
| Waste water, spent acids, spent caustic, and other liquid materials are generated
;by petroleum refining operations and present disposal problems. The wastewaters :
.emanating from refinery and petrochemical operations can be divided into five general
categories.- ',,,'.
' 1. wastes containing a principal raw material or product resulting from the
I ' stripping of the material from solution;
! I
2. reaction by-products; . |
! i
3. spills, slab washdowns, vessel cleanouts, sample point overflows, etc.;
4. cooling tower and boiler blowdown, steam condensete, water treatment
wastes, and general washing water; and
5. storm waters from contaminated drainage drea'° .
L . •' ' !
jTh*. first consideration in the evaluation of treatment and control technology for o j
(particular industry requires an estimation of the characteristics of the various source flows.
iA logical approacn in formulating these .estimates is to categorize the pollurionaLinputs j
{into select source components. These components, as defined by ford'"" are: (1) process'
[operation wastes; (2) utility operation wastes; (3) sanitary sewage; (4) ballast water blow-!
I down; and (S) contaminated storm runoff. The utility wastewaters include the blowdown
'from boiler and cooling systems. In addition, the miscellaneous flows resulting from
spills, turnarounds, and other inordinate discharges also must be considered.
The principal contaminants in the wastewaters includa organics from residual
products and by-products, oils, suspended solids, acidity, heavy metals and other toxic
materials, color, and taste and odor producing compounds. The principal parameters I
to characterize organic wastewater constituents are BOD, COD, YOC, total oxygen]
88
-------
(TOD), T&S, grease and oil, and phenolic compounds. The principal parameters j
i used to characterize inorganic wastewater constituents are acidity, alkalinity, TDS, j
i ammonia-nitrogen, sulfidcs, and specific inorganic ions. In many cases, it is important !
• to measure the inorganic dissolved ions in solution such as chlorides, sulfates, nitrates,
land phosphates. This is important in the quality control of cooling tower and.boiler blow-
-down waters. The pH of refinery wastewateh is normally alkaline, but may vary con- •
r~stderab'y * depending* on* utsposaf or spent OCIOT, *causttc3/*and actd washes. ' ~" '
I
PROCESS OPERATION WASTEWATERS ' ' '
I i
Wastewaters from plants manufacturing similar or even the same compounds
I usually display dissimilar characteristics. This con be ascribed to the use of different
j manufacturing processes coupled with the fact that the by-product disposal pattern may !
• occur in a number of different ways. The many combinations of process operation, type
i of crude charge, or age of plant make classification of a "typical" refinery petrochemical
i wastewater difficult. The differences in in-plarit effluent segregation systems, in-plant ,
! treatment systems, and process design, operation, and maintenance also contribute to :
pftiese vdnoti6nsr1The large variety of compounds produced~wrthfn~the refinery-petro- ""']
| chemical industry make the task of treating process wastewaters difficult and complex. •
| Hence, a wastewater treatability study is often required before treatment options can be
considered. | ' !
The most common approach to predating the quality of petroleum refinery
wastewaters is to relate the quantity and quality of pollutant produced by a unit process
I to production'™<"!^. This information, when combined with general wastewater
j characterization data, offers insight into the nature of production and processing wastes.
The principal processes which characterize the refining industry are: (1) crude and
product storage; (2) crude desalting; (3) crude oil fractionation; (4) catalytic cracking;
(5) thermal cracking; (6) hydrocrocking; (7)jreformmg; (8) polymerization; (9) alkylaticn;
(10) isomerizotion; (11) solvent refining; (12) dewaxing; (13) hydrotreating; (14) de-
asphalting; (15) drying and sweetening; (16)! wax manufacture; (17) grease manufacture; -
i (18) lubricating oil finishing; (19) blending Ond packaging; and (20) hydrogen monu-
! facture. A brief (discussion of the wa*tewat^rs and solid wastes associated with each of
the aforementioned processes is presented herein.
The wastes associated with storage of crude oil and products are mainly in rht,
form of free and emulsified oil and suspended solids. The storage of crude oM produces a
waste liquor which i* both high in COD and! BOD and also high in an oily bottom sludge
which must be removed intermittently. Crude oil storage is frequently the source of poly
sulfide-bearing wastewaters and suspended solids. Product storage can produce high BOD,
alkaline wastewaters, as well as tetraethyl lead. Tank cleaning can contribute large
amounts of oil, COD, and suspended solids, and a minor amount of BOD.
The wastes associated with crude o
^corrosion. This waste stream is relatively hi
desalting contain emulsified and fre
-------
^.significant levels ,pf chlorides and other dissolved materials which contribute to the over_-j
(all dissolved solids concentration in the combined wastewater. The temperature of the
:desalter waste often exceeds 200°F. '
i i i <
The wastes from crude oil fractionation generally result from the wastewater
associated with overhead accumulators, the wastewater discharged from oil sample lines,
. and the vt<.stewoter associated with barometric condensers. The wostcTvaters from the- - ••
-!accumulators are a major source of sulfides, especially when sour crudes are processed. \
j and'it also contains significant amounts of oil, chlorides, mercaptans, and phenols. The :
jwastewoters characteristic of the barometric condensers which are used to create the
'reduced pressure in vacuum distillation units ore o source of stable oil emulsions. Like-
\wise, the major source of wastes in thermal cracking is the overhead accumulator on the •
j fractionotor. These wastewaters usually contain various oil fractions and may be high in
BOD, COD, ammonia, phenol, and sulfides. |
| i ;
| Catalytic cracking units are one of the largest sources of sour waters in o
j petroleum refinery. Pollution from catalytic cracking generally derives from the steam
isrrippers and overhead accumulators on fractionators used to recover and separate various
(hydrocarbon fractions. The major pollutants resulting from catalytic cracking operations
'arc oil, sulfides, phenols, and ammonia. These waitewaters are alkaline with high BOD
'and COD concentrations. . ' •
I 'A •'.''' :
j Information concerning waste production associated with hydrocracking has not
(been published; nevertheless, a waste stream from the process could be high in sulfides
since hydrocracking reduces the sulfur content of the material being cracked. This waste
stream would probably be generated in the product separation and fractionation units j
following toe hydrocracking reactor.
* i
Reforming, which is o relatively clean process, will generate a small volume of
wostewater. This waste stream is alkaline, and the major pollutant is sulfide which is [
derived from the Overhead accumulator on the stripping tower used to remove light hydro-
carbon fractions from the reactor effluent. In addition to sulfides, the waste contains
small amounts of ammonia, mercaptans, and oil. < . <
< • i ' !
Polymerization is characterized by a major pollution loading per barrel of
charged material, but because of the small polymerization capacity in most refineries,, the
total waste production from the process is small. Even though the process makes use of '
acid catalysts, the waste stream is alkaline, because the acid catalyst in most of the sub-1
processes is recycled, and any remaining acid is removed by caustic washing. Most of 1
the waste material, which is high in ;ulfidesr mercaptans, and ammonia, comes from the '
pretreatment of the feedstock to the reactor.' These materials are' removed from the feed- <
stock in ccustic scrubbers and woshwoier towers. The spent caustic must be removed j
periodically and contributes to the solids disposal problem.
!>.)£ The two principal alkylation processes are cascade sulfuric acid alkylation and
r ^hydrofluoric acid olkylotipn. The general sourcesi ofi watte in q_ sulfuric acid alkylation
I 1 I j .
90
-------
Tjn.it are the overtyped accumulators in the fracHonation section; the alkylation reactor;
and the caustic wash. The wastewater from the overhead accumulators contains varying
amounts of oil, sulfides, and other contaminants, but is not considered the major source i
of waste in this process. The waste from the' reactor consists of spent acids and has a pH j
less than 3. This stream seldom enters the sewer system because most refineries process it!
«4o recover clean acids. Consequently, the ^nojor contaminants entering the, sewer froavo-|
•-sulfurfe-actd-atkyJotiofTvmt are generally spent caustics from the neutralization of the—\
hydrocarbon stream leaving the alkylation reactor. Hydrofluoric acid alkylation units do
not have a spent ccid or spent caustic.waste stream; consequently, the major source of
waste material is the overhead accumulator on the froctionator.
Dora are; not available concerning waste discharge: from isomerization processes,
but it is the general contention jf the industry that these wastewaters are low in phenolic'
compounds and oxygen demand; and sulfides and ammonia are not likely to be present.
The potential.pollutants from the various solvent refining processes are the
j solvents themselves. Many of the solvents used in the process, such as phenol, glycols,
pttid~ amines", con'produce a high BOD." Under Tdeaf conditions, fhs solvents are con- ,
j tinually recirculated; but in reality, some solvent is lost to the sewer through pump seals,;
flange leaks, etc. The principal source of wastewater is from the bottom of the fraction-i
at ion towers. Oil and solvent are the major waste constituents. Likewise, leaks, and
[spills aie the major source of wastes in solvent dewaxing processes which use MEK. A j
spill of MEK can result in a waste stream which is high in BOD. Propane dewaxing is a j
cleaner process that poses few water pollution problems. !
The'hydrotreating process is used tb saturate olefins, it also will remove sulfur,
nitrogen, and oxygen compounds and other contaminants from either straight-run or
cracked petroleum fractions. The strength and quantity of waste generated by a hydro-
treating process depends on ihe type of process used and the material being hydrotreated.
The major waste streams derive from overhead accumulators on fractionators and steam
strippers and sour water stripper bottoms. The major pollutants are sulfides and ammonia.
Phenols also may be present if the boiling range of the feed is high enough.
Data ore not available concerning waste discharges from deaspholting processes.
Indications are that wastewater does not result from the actual deasphalting step, but is
generated from overhead condensers on the steam strippers that are used to separate the
asphalt, deasphalted oil, and propane. Thii "sour water" generated from the condensers
probably contains small amounts of sulfides, oil, and ammonia.
The principal waste stream from drying and sweetening operations is spent
caustic. The sperjt caustic is characterized as phenolic or sulfiHic depending on which is
present in the largest concentration. Phenolic spent caustics certain phenol, cresols,
xylenols, sulfur compounds, and some neutral oils. Sulfidic spunt caustics are rich in
sulfides, but do not contain any phenols. Both spent caustics are usually high in BOD am
;COD. Other waste streams from the process' are generated from water washing of the
"Hregted product and regeneration of the treating solution.• These waste streams will
, - r^-. r -- - "•;,:•.. f;.:,7 " ~
-------
}-J.55.nra'n smo'' omounts of oil and the treating material; namely, sodium plumbite or j
'• copper chloride. j '
. i i • i
! Data are not available concerning waste discharges from wax manufacturing pro-
| cesses, but there is little reason to believe that such wastes pose a significant source of
• • '.pollution. Likewise, only small volumes,of wastewater are discharged from grease martu-
•vj focruring processes, .In the casa of both manufacturing processes, a wnall amount of oil
• may be lost to the wastewater system through leaks; but the largest waste loading occurs
i when batch units are washed. This results in soap and oil discharges t.o the sewer.
1 Acid treatment of lubricating oils produces acid-bearing wastes occurring as
rinse waters and sludges. In addition, there art discharges from sampling points, leaks,
j and shutdowns. These waste streams are usually high in dissolved and suspended solids, ;
'sulfate's, sulfonates, and stable oil emulsions. The acid sludges produced in this process
; create a significant solids handling problem. I
' i ! • .
Blending and packaging processes pose few pollution problems because care is
."token to avoid loss of product through spillage. The primary source of waste material is "
! from the washing of railroad tank cars or tankers prior to loading finished products. These
wash waters are high in emulsified oil. Sludges from gasoline storage tanks can contain
large amounts of tetraethyl lead, a highly toxic compound, which could be flushed into
! the wastewater system during washing operations. i
1 ' I i :
j Information concerning wastes generated during hydrogen manufacture is not ;
;available; however, the process appears to be relatively' clean. In the special case of the
steam reforming process, a potential waste stream produced during desulfurization would j
contain oil, sulfur compounds, and phenol. '
UTILITY OPERATION WASTEWATERS i
I I
, t
Utility operations which are an integral part of a petroleum refinery contribute
i to the wastewater flow in the form of blowdown from boilers, process steam generators,
iond cooling towers. In order to control excessive scaling and fouling of heat transfer
surfaces in boilers and process steam generators, a limitation is placed on the level of
dissolved solids and c!/;alinity present in the water within steam drums. Blowdown occurs'
'each time treated makeup water is fed to the system so that allowable quality limits can '
'be maintained. Although the volume of process steam generation blowdown will vary from
, system to system, .it normally represents less than 5 percent of the total treated feed water
i used to produce steam. This is a smo I! portion of the total utility operation wastewater i
jvolume; although, the concentration of key water-quality constituents may be significant^
j Blowdown from cooling towers represents a sizable portion of the combined wastewater
[flow, and constitutes the majority of the utility wastewater volume.
i
In general, cooling tower effluent quality will be similar to the feed water
supply for a once-through cool
-------
jsojt concentrations since pure water is continuously evaporated to the surroundings.
i Coding tower additives, TDS, and contaminants entering the water through heat
'exchanger leaks constitute the pollution cha'roctonstics of cooling rower blowdown. the
fwilt riot or?ly~vory with time during the,course of a "storm, but also will change
j with each individual area within a refinery-petrochemical complex since each area has
i its own geometric characteristics which influence patterns of surface runoff. The investi-
jgation of storm water flows is usually oriented toward obtaining data regarding the prob-
i able wastewater volume, quality characteristics, and runoff hydrograph shape. The
i anticipated rainfall volume estimate is developed statistically from long-term precipi-
tation records with sufficient accuracy to allow computation of meaningful probability
statements concerning future events. ^
The character of contaminated storm runoff is similar in water quality to the
wastewaters generated during operation of the petroleum refinery inasmuch as there will
exist a significant, oxygen demand, TSS, and oil and grease level in the runoff which mus
be removed before1 these waters can be discharged to a receiving body. The analysis of
storm flow COD concentration relationships from numerous refinery-petrochemical com-
plexes has indicated that between 60 and 80 percent of the organic moss is washed from
the surface of the facility during the first hour of the rainfall-runoff event'"'. Similar
relationships exist, for other pollutant constituents.
BALLAST WATER BLOWDOWN
! '
Tankers
-------
* water to the refinery wastewater collection system has been found to be the most
effective und economical means of minimizing pollution problems. Among the principal
wastes that could be treated separately ore oil-in-woter emulsions, sulfur-bearing waters,
acid sludges, and spent cnustic wastes. A complete evaluation of the effectiveness of in-
plant processing practices in reducing wastewater pollution requires detailed information
on the wastewater flows and pollutant concentrations from all types of refinery process
unit* end storage facilities. Unfortunately, this kind of information is not readily avail-
able. . '
Despite the lock of specific process wastewoter data, there exists information of
a more general nature which indicates that substantial wustewatcr pollution reduction
could be achieved through changes in processing facilities and practices. For instance',
hydrocracking and hydrotreating are two processes that generate much lower waste '
lo.dings thon the processes they replace. However, the greatest potential for in-plant
waste load reduction appears to be in improvement of genera! operating and housekeeping
practices rather than in changing processes or subprocesses.
"*-* ~ ~ Although the central thrust of this treatise is to evaluate existing and emerging
end-of-pipe wcstewate.' treatment technology as it applies to the petroleum refining
' industry, one should be aware of the existence of numerous in-plant source reductions,
procesr modifications, and conrrol practices which contribute to the overall goal of
pollution control. Such in'plant treatment control methods include stripping and
• recovery operations, neutralization and oxidation of spent caustics, ballast water treot-
• meht, slop oil recovery, and temperature control. These assorted in-plant treatment
. practices not only reduce the waste loadings tc the wastewater treatment facility, but
[also enhance its performance. In some cases, in-plant control will show a cost credit in <
• the form of product recovery.
I l
| Oil-Water Separation
. *
| 1,1
1 Oily wastewater treatment systems are found in the three phases of the petroleum
: industry; namely, production, refining, and marketing. These systems will vary in size
1 and complexity, although their basic function is to collect and recover valuable oils and
to remove undesirable contaminants before discharge to a receiving body. The wostewotei
Treatment systems found in refineries are larger and more elaborate than those found in the
production phase. The wastewater collection system of a modern refinery usually includes
gathering lines,, drain seals, junction boxes, and channels of vitrified clay or con.crete ;
which transport wastewater from processing units to oil-water separators. These oil-water
separators are designed to receive wastewater from all process sources, sometimes even
including storm runoff and ballast water blowdown. The discharge of liquid wastes to '.
these systems originates from c variety of sources such as pump gland and accumulator '
leaks, process spills, cleanouts, sample ports, relief valve«, etc. I
' i :
The removal of oil contamination from wastewaters can be accomplished by the '
f U.\C' UIe °f *everal well-known and widely.accepted techniques. The performance of any . j
;E>T >E-g?ven separation technique will depend^entirely on the condition of the oil-water mixture:
_ _
-------
thera'ore, the nottire of a particular oily waste stream most be determined before the
proper treatment device can be selected. The types of oil-water mixture which carl be
encountered may be classified as wostswaters with oi! present as free oil, Hispersed oil,
emulsified oil, and dissolved o- soluble oil. Free oil is usually characterized by an oil-
water mixture with droplets larger than 150 microns in size, wh;'e o dispersed oil mixture
will have a droplet size range between 20 and 150 microns. .An emulsified oil mixture
^wJH have droplet sizes smaller than 20 mjcr<5ns. A wostewoter which is characterized by
an oi|-wot«r mixture where the oil is said to be soluble, is o liquid waste where oil is not
present in .'he form of droplets. Soluble oils are, in actuality, nonoily materials such as
phenolic type aromatic compounds and colloidal sulfur compounds which are selectively
extracted to varying degrees by the solvents used in the analysis.
Several professional societies and industrial institutes have offered definitions •
for oil in the form of methods for measuring oil concentration in water and waste water.
The commonly used solvents in oil and grease analysis are hexane, petroleum ether,
benzene, chloroform, carbon tetrachloride, and freon. Tlr»ce solvents exert selective
extraction of specific greases and oily constituents. Since oily matter and grease may be
"Of mineral, .animal, or vegetable origin, the solvent action, exerted on material of such
.different chemical structure, will vary to a marked degree. The application of a test
method for oil and grease analysis to such materia! will produce a variety of results each
depending on the solvent used. Therefore, the definition of grease and oily matter is, by
necessity, based on the procedure used for analysis.
Theory and Practice of Gravity Oil Separation '
i The three main forces acting on a discrete oil droplet are buoyancy, drag, and
•gravity. The buoyancy of an oil droplet is proportional to its volume and the drag is
' proportional to the projected area of the droplet. As th« diameter of an oil droplet .
decreases, the ratio of its volume to surface area also decreases. Because of this droplet
. size relationship, larger droplets tend to rise while smaller droplets remain suspended.
, This relationship is defined by the well-known "Stokes Law." The concept assumes that
' the terminal velocity of a rising droplet is: (1) proportional to the specific gravity
i difference between the oil and .water; (2) proportional to the square of the oil droplet
diameter; and (3) inversely proportional to the viscosity of the water.
! • '
The oil removal efficiency exhibited by a particular separator is a function of
the droplet size distribution, the amount of oil present in the contaminated water, and
the presence of surfactcnts or chemical emulsifiers which may or may not be indigenous to
the crude oil. '
• ! j •
If a free or dispersed oily water mixture is brought to a relatively quiescent state
and given sufficient time, the oil droplets will coalesce and eventually separate from the
wastewater. The droplets which collect on the surface of the water will coalesce, pri-
marily because of, their proximity to othor droplets, and form a continuous floating oil
Jgyer which may be decanted from the water. This process is called "gravity separation,.1'
95
-------
r Generally, oil droplets 20 microns and smaller cannot be separrted practically
with a gravity separator because the net buoyancy force is o- ercome by the drag force of
the continuous water phos«. These droplets which were pre iously classified as emulsified
oils are so small that the random movement of the water keeps them suspended. In
addition, impurities present in both the oil and water phases also may affect the separ-
ability.
Nearly all oil;- wastewaters have some amount of free, dispersed, and emulsified
oil present. Rainfall runoff which has not been pumped may typically be contaminated
with oils which hcve between 8 and 10 percent of the oil droplet sizes below 20 microns.
Pumped mixtures may have up to 2C percent of the oil droplets below microns. Contami-
nated bilge and ballast water will typically have from 10 to 15 percent or more of the oil
. in a mechanically emulsified state.
The principal objective of gravity differentia I-type oil-water separators such as
, the API separator and the parallel or tiltcd-plate separator is to establish an environment
in which suspended solids are settled coincident wirh the separation of oil in the influent.
The current design principles of the API separator are based co extensive studies into the
effect of inlet and cy*!»f arrangements on oil separation efficiency as well as the impact
of appurtenances on the hydraulic ciK.rcct»nsties in the separation chamber. The prin-
cipal factors which affect the design of oil-water separators *;*: ,(1) 'the specific gravity
of the oil; (2) the specific gravity of the wastewater; (3) wastewater temperature; (*} nw
percentage of emulsified oils; and (5) the concentration of suspended solids. The specific
gravities of both oil and wafer which are a function of temperature'determine the rate of
• oil separation. The cforementioned factors set the allowable hydraulic overflow rate for
the particular oil-water separator design.
The parallel-plate interceptor (PPI) separator essentially reduces the distance
that a particle of oil must travel before reaching a collection surface. In this case, such
a surface consists of a number of parallel plates, evenly spaced and set at an angle from
the horizontal. The oil particles coalesce on the underside of these plates and creep up
to the water surface. Therefore, the distance traveled by an oil droplet is a few inches
instead of several feet. An alternate design uses a variant of th« normal PPI separator
known as the tilted-plate separator (TPS) or the CF'I which have only recently been
' applied to the removal of free oil from refinery effluents.
! A coogulation-flocculotion step is commonly employed in refinery wastewater
treatment to increase the overall removal of oil and suspended solids. The flocculated
; suspension may be removed by simple gravity separation but an air flotation scheme is
, more common. The addition of a chemical coagulant to an oil con ram inc ted wostewater
: serves to promote, the aggregation of dispersed and emulsified droplets into flocculated
' clusters. It also may negate the effects of indigenous crude oil surfactants and permit
1 smaller droplets to coalesce into larger ones. This piocess is irreversible, leading to a
decrease in the number of oil droplets and finally to complete demulsificption. The net
. effect of coagulation-flocculation is to render the suspension more amenable to phase
^separation. '
96
-------
*£ When a •quantity of an aluminum (Aj ') or iron (Fe' ' ) salt sufficient to exceed
' the solubility limit of ths metal hydroxide is added to water, a series of complex
hydrolysis reactions occurs. These proceed from the production of simple hydro-complexes
through the formation of a metal hydroxide precipitate. The reactive form of these com-
plexes is a function of system kinetics, expressed as metal ion concentration, pH, and
alkalinity. Chemical theories attribute coagulation to precipitation of metal hydroxides
' which o«t to*om«th colloidal particles, specific chemical reactions, surface adsorption,
and specific chemical interactions involving coagulant hydrolysis products.'
* . i
The impact of coagulation-flocculotion on removal of emulsified oils is repre- ,
sented by the data presented in Figure 9. These data indicate the results of a series of
batch tests conducted to examine the means by which aluminum added as alum and iron
, added as ferric sulfate interact with oil droplets'?'. The curves presented in Figure 9
! reveal that performance response is a strong function of pH. The factors believed to be
responsible for the oil removal at pH=5.5 and 8.5 are the charge on the oil droplet and
the nature of the hydroxy-metal species. The combined effects of charge reduction,
sorption to the droplet surface, and interdroplet bridging are required to effect significant
*b!l removal. ' ~ _....-.
i
Synthetic organic polymeric electrolytes or polyelectrolytes are used with
increasing frequency in water and wastewater treatment. They function both as primary
coagulants and as coagulant aids when used in conjunction with metal coagulants. As a
coagulant aid, their principal function is to strengthen the floes formed by metal coagu-
lants such as alum. A series of batch tests was conducted in order to examine the effects
of polyelectrolyte addition on oil removal efficiency and these data are presented in
Table il . Specific factors evaluated were polymer type (anionic, nonionic, or
' cotionic), polymer dose, and order of addition. The data presented in Table 11 indicate
: that the anionic polymer is least effective for oil removal and that the non ionic and
! cationic varieties are approximately equal. However, none of the polyelectrolytes
' examined resulted in sufficient oil removal to justify its use as a primary coagulant. ;
1 i i
Since both the polyelectrolytes and the oil droplets are highly charged, there
' exists a very strong electrostatic attraction between the two substances. Consequently,
, this may result in a nonspecific interaction in which the polymer coats the oil droplet
. surface instead of attaching at a point and extending into solution to act as a bridging '
'< agent for other droplets. Moreover, the relatively high ionic strength of petroleum
I refinery wastewaters may inhibit the polymers from fully uncoiling in solution,,thereby .
i raducing their effectiveness. i
i I
; i
i These same polyelectrolytes were evaluated as coagulant aids in conjunction
with alum as the primary coagulant. The results of these tests are presented in Figure 10
! and indicate that polymer addition does not significantly improve oil removal efficiency
i but may even hinder ft'' . |
97
M'Mofri
-------
a s
o o
w
S10
7
tt o
a •—
' oe
O X Ul
90 t-
— 141
p e si
853!
90
£ 80
Ul
a.
UJ
Ul
VI
Ul
te.
o
70
eo
50
40
30
20
10
0
e.s
'pH - 5.5
20 40 60
ALUM DOSE (mg/1)
60
100
K) 20 30 40
FERRIC SULFATE DOSE (mg/1)
FIGURE 9. Experimental observotions concerning oil ond greose reduction
with coagulant addition
98
-------
TABLE 11. POLYELECTROLYTES AS PRIMARY COAGULANTS FOR CRUDE OIL DISPERSIONS
Polyelectrolyte
Dose
(mg/D
0.1
°-2
0.5
1.0
2.0
5.0
10.0
Anlonlc
Polymer*
16
15
16 .
15
12
12
10
Percent Oil and Grease Removal
Non1nn1c
Polymer**
23
20
26
29
29
27
27
Cat1on1c
Polymer***
18
27
27
30
29
31
31
*An1on1c polymer • polystyrene sulfonate, approximate molecular weight 2 x 10 .
**Non1on1c polymer « polyacrylamlne, approximate molecular weight 1 x 10 .
***Cat1on1c polymer « polyacrylamlde. approximate molecular weight 1 x 10 .
-------
VOTE:
INITIAL ALUM DOSE - 10 m/1
0 0.2 0.4 0.6 O.i 1.0 L2 1.4 1.6 1.8
POLYELECTROIYTE DOSE (mg/1)
.0 2.2 2.4
FIGURE 10. Pol/electrolytes evaluated as coagulant aids in conjunction with
alum as the primary coagulant
(,
> 100
-------
-Flotation Oil-Water Separation
l ' '
Flotation is one of the most common oil refinery wastewater pretreatment
| techniques and generally includes a chemical coagulatiori-flocculotion step. It is a unit
operation specifically desianec! for the separation of liquids and solids from wastewate/.
Natural flotation occurs to some extent in -rovity oil-water separation techniques, but
. is generally limited to the removal of oil globules greater than 40 microns in size. •
i '
j The flotation procen relies on uniting gas bubbles with the dispersed oil phase
j resulting in a reduction in the specific gravity of the oil and a subsequent increase in .
' rise velocity. Flotation is strongly influenced by the surface characteristics of the dis-
1 persed oil phase and to a lesser degree by the oil droplet size.
l Flotation methods are divided into dispersed-gas and dissolved-gas processes.
! Dispersed or entrained gas flotation utilizes bubbles generated by one of the following
techniques: (1) mechanical shear; (2) gas diffu:>on through a porous media; or (3) homo-
' genization of the gas into the wastewater. Dissolved gas flotation generates gas bubbles
fcy the precipitation of gas from a super-saturated solution. The processes differ in the
i size of the gas bubble produced. In dissolved gas flotation, bubbles average 80 microns
: in diameter, while'they are generally an order of magnitude larger in dispersed gas
: flotation. The gases commonly used for flotation of dispersed oils are air amJ nitrogen.
Air will oxidize any ferrous iron in the wastewater which may precipitate as ferric
: hydroxide. The ferric hydroxide, however, may act as a coagulating agent and improve
| overall oil removal efficiency. ;
DAF is the process commonly used in refinery and petrochemical installations to
enhance oil and suspended solids removal. It is generally preceded by a gravity oil
i separator which will remove gross quantities of free oil and oily suspended solids. The
j process involves, the pressurization of wastewater in the presence of air, thereby creating
: a super-saturated solution which when passed into a flotation chamber at atmospheric
1 pressure will precipitate the air from solution in the form of small bubbles. These bubbles
, unite with the dispersed oil phase to form a collection of distinct oil droplets (coagule)
• and carry it to the surface. The float is removed to disposal or recovery by mechanical ,
' flight scrapers, whil« the underflow represents the clarified effluent. j
i The .mechanisms by which the.air bubbles and dispersed oil droplets interact are '
I generally classified as follows: (1) the adhesion of oil droplets to air bubbles; (2) <
entrapment of air bubbles in a coagule or a flocrulant structure; and (3) absorption of air'
bubbles into thy coagule or floe structure. J '
r the DAF process are: (1) full wastewater
stream pressurization; (2) split stream pressurizarion; and (3) recirculation stream pressuri-
zation. Most applications of the DAF process for the treatment of refinery wastewaters j
use recycle stream pressurization as the principal mode of operation. This system hr-s the
advantage of requiring a smaller pressurizing pump than in full-stream pressurization so 4
sUhot capital am.' opsratinj cosh are reduced'._ln addition ,_pymp life is increased because
PAG!: Ml
-------
solids or? not pumped since they ore separated from the recycle stream in the __ j
'flotation chamber prior to pressorization. Flocculant structures formed in the wastewater ,
•. "stream are not subjected to the shearing forces of the pressuriration pump. Moreover,
'emulsion formation due to shear is minimized. A potential disadvdntage of recycle oper-
•ation as compared' :o full-stream or split-stream operation is the necessity of an enlarged
-flotation chamber; , —
' i ' '
i ' The use of chemical agents has historically been an integral part of the flotation
{process. 'These chemicals function by modifying the surface properties of one or more
phases. For instance, rhere are chemical agents (frothers) which serve to lower the inter-;
ifacial tension between the air bubble and ?he wastewater. Other chemical agents
'(collectors) either reduce the interfaciol tension between the dispersed oil phew and the
'wastewater or increase the interfocial tension between the air bubbles end the oil phase.
iBoth conditions tend to increase bubble-droplet adhesion. Coagulating chcrr.iccU SyCn as
aluminum and iron salts are often used to enhance the bubble capture mechanisms of ;
{flotation. As disrussed previously, these coagulants function to improve the flocculant
{nature of the dispersed oil phase and enhance the capture of small oil d.oplets. ;
--
| The cragulants may be' introduced into the process stream in many ways depending
'on the mode of operation and the type of chemicals used,., In some installations, the
chemical is injected downstream of the pressure release; while at others, it is injected
directly into the suction side of the pressurizing pump. When chemicals are injected
directly into the pressurizing pump suet ion j mixing is enhanced; but the relatively short '.
jtime available for floe formation along with the shear forces encountered may be detri-
imental to the overall system performance. i • ,
I ' . - , i
{ In recycle or split flow operation, the coagulant may be added directly to the ,
iwastewater in a separate reucror for better control of the coogulation-flocculation process.
jThe flocculated wastewater stream is then transferred by gravity to the flotation chamber
jwhere it is combined with the pressurized stream. This mode of operation provides for •
more effective use of coagulants by increasing reaction time as well as improving floe ;
'formation and increasing separation rates. I ' j
Oil-Water Separator Performance j
! , i '
Oil-water separator treatment efficiency is a function of many design and oper~
attng variables, the two factors which are most significant, in impacting upon separator ,
performance are flow rate and influent oil concentration. Separators must be protected
in order to prevent flushing during periods of high flow. Bypass and overflow lines \
generally are used to provide this hydrt . 'ic protection. The influence of initial pit and i
;gr*QJe ccncsr.trstSor on Joe rreuhnent performance of conventional oil-water separators is
[illustrated in Figure 1 1 . Assuming the systems are operated properly, these data indicate !
'that the influent oil concentration is a significant factor with respect to process capacity :
and efficiency. This effect is underscored by data for both the conventional API separator
•M u1"^. *nc chemical flocculation OAF unit inasmuch as higher influent oil concentrations are!
'
.
£=*emoved with greater .efficiency than that of low infjyent_pil concentrations. _
— ~ ..... ~~~ ~~ ....... ~~~ ......... "" ~~
-------
8
no
•00
900
400
• 900
M too
MO
•O
•o
TO
40
90
Otff
9 8 3 £ 288
M mrmmKc cw*t
mrouiMtt tunvt
(UPCKTIO VM.UII IIKIUM CHtHICM. AMITlm)
IHFlUtKT OIL AND CREASE fo/l)
III!
LS
FIGURE 11. Impact of influent oil and grease concentration on treatment performance of conventional
oii-water separators.
-------
-r _ Although gravity taparation will not provide a satisfactory effluent oil quality ___]
1 for direct discharge to biological treatment, it is the most economical ana efficient
iopproach to the removal of high concentrations of incoming free oil. lh«s effluent oil <
! concentrations from API separators generally range from 70 to 150 nvg/t, although devi- :
j ations on either side of this range ore common due to the great vrsriety of wostewaters j
Ureoted witl. this type of oil renoval equipment, Nt»vertr.sless, under excellent operating
^HCondirtomy-an apparent-lower limit of oil »n o gravity seporotor effluent n usually around
150 mg/t. It should' be «rnphas?zed that the removal efficiency of all gravity separation
! techniques is a function of temperature and the density difference between oil and water.
i ; ' ' •
; As observed from the data presented in Figure 11, DAF unit effluent oil levels of
i 10 to 30 mg/t can be achieved if the influent oil concentration remains below 200 mg//.
, Although it is possible to obtain an effluent oil quality of 10 mg/Jt under certain con-
:ditions with a chemical flocculation OAF unit, consistent performance at a 10 mg/Jt
'effluent oil concentration would appear to be difficult if not impossible to sustain .
'This is particularly trve when considering the fluctuations in influent.oil concentration
; inherent with most petroleum processing operations, and the corresponding effect on
fgravTty'and DAF oil removal systems.
i j
( The effect of peak hydraulic loading on API separator effluent oil concentrations
i is illustrated by the data presented in Figure 12. In general, the ideal API separator
'design seeks to limit the extent to which turbulence and short-circuiting will affect the
\ operation of the separator. The effects of turbulence increase with the magnitude of the
I ratio of the horizontal velocity to the rate of oil droplet rise. Theoretically, turbulence
ionly can be compensated for by decreasing the overflow rate. It is apparent from the
I data presented in Figure 12 that a deterioration of effluent oil quality will occur when
j hydraulic surges exceed the design maximum. Unfortunately, a theory does not exist
'which defines the magnitude of a maximum absolute value of the horizontal velocity. '•
! ' • '
i Data collected during the operatic* of a pilot-scale tilted-plat separator are .
I presented graphically in Figure 13. The pilot unit was operated as part of a comprehensive
j freatobilit;' study on wostewaters from an integrated petroleum refinery. The apparent oil
i removal efficiencies tend to vary between a low of 50 percent to a high of 80 percent at
'the design average hydraulic load. Like the API separators, .ne treatment performance of
j the TPS unit is affected by hydraulic surges,^ond the overall suspended solids and oil ,
removal efficiency of this pilot unit did decrease when the system was operated at 1.5
| times the design average hydraulic load. Nevertheless, there is broad application for
j tilted-plots separators in the treatment of refinery-petrochemical wostewaters since so
{little space is required for their installation: The TPS unit can be installed in series with
Jan existing API separator which is either overloaded or improperly designed, K ereby
| increasing the overall oil removal efficiency! normally obtained by this gravity separator!
jThe TPS unit also can be installed to operate1 in parallel with existing API separators, '
j reducing the hydraulic load and enhancing the oil removal capacity of the system. ; '
i' 'I ' • . i
Several cose histories where DAF units are used to treat refinery oily waste- '
g'waters ore presented in Figures 1_4 and 15. These curves_representjtotistica| analyses of ' •
. ,. , r- - -.--p. ----- _- ^
104
-------
KM n.0* WUTIO*
Utl T«M 15 HIMTM
HA* WftFACe UMDINC (t»«/ft2)
FIGURE 12. Impact of p«ak hydraulic loading en API separator effluent quality.
' !±
1 v 105
-------
a.
v>
3
Wl
OwU
ton
rt
,
,
y
/*•
X
^ a
cP« *
|
•
k
•
tOO 200 300 400 ,900 600
INFLUENT OIL AND GREASE (mg/l)
FIGURE 13. Treatment performance of pilot-scale Hired-plate separata
T 106
C'SL'L OJ
-------
too
700
600
500
400
800
too
too
•0
.•0
TO
£ »
1 w
u 40
t/>
2 so
<3
a
3 >o
_<
o
10
t
7
e
4
t
•
•
4
J
_o
i ;.
1 />
•
•
/
^
r [
/
;
^^
^
\^
r*
jf*
^
(
j» _*.
r
r
^
•
.
r*
^
>
1
luJMC
-9-^
— -
— •
••^H
••I^H
mrit
o im
.
:*T
ItKT
'
.1 1 10 l» M 90 40 SO «Q TO *0 §3 *0 IJ »8
r
-------
KV»0
ipoo
•poo
TjOOO
•jooa
»poo
4,000
1,000
IJDOO
1,000
800
•00
TOO
I800
400
SCO
u
e
too
100
10
•0
TO
(0
so
40
BO
to
10
X
mriunrr
UtXT
t 6 » 15 tO 90 40 SO «P TO *0 §4 90 *» Utf
PERCENT PROM6ILITY THAT CONCENTRATION WILL IE LESS THAN STATED VALUE
FIGURE 15. Variation in DAF treatment performance for oil and grease removal1,
I , ' ' 7 108
-------
*• DAF process efficiency data obtained from wostewater surveys performed within petroleum
refineries which employ flotation oil-water separation equipment for secondary oil
recovery. The first case history which is graphically presented in Figure 14 is charocter-
. istic of the expected treatment performance associated with a DAF unit operated in series
with an API seor.ator. These data indicate that the DAF unit reduced the oil level from
• ,68 mg/1 to 15 mg// 50 percent of the time, and from 105 mg/£ to 25 mg/4 at a 90 per-
""centile probability level. The second COM, history (Condition B) w!i!ch is presented in
Figure 15 depicts the variation in DAF treatment performance for oil and grease removal
at a much higher level of influent concentration. These data indicate that this DAF unit
reduced the median oil level from 580 mg/4 to 68 mg/t and the 90 percent! le probability
value from 1 , 930 mg/i to 1 28 mg// .
DAF oil removal efficiency is a function of many factors; namely, design over-
flow rate, retention time, recycle volume, pressurization level, air-to-solids ratio, type
and volume of chemical odditi^n, and the concentration and form of the influent oil.
The data tabulated in Table 12 account for these variables inasmuch as the operational
overflow rates fall within the accepted design spectrum of 1 .5 to 3.0 gallons per minute
(gpm) per square foot and all use the pressurized recycle mode of operation. Ail of the
cases cited use coogulation-flocculation with the principal chemical being alum at a
' dosage of between 100 and 150 mg/4 on the average. A few operate with a small poly-
electrolyte addition.
In general, the DAF unit experiences better oil and grease removal at the higher
influent concentration levels. This also is the case regarding the separation of influent
suspended solids and COD. The overall COD removal efficiencies presented in Table 12
are not surprising when 'one considers that a large fraction of the COD is associated with
suspended solids. I
! ' •
! ' . ' '
| The information presented in Table 13 summarizes the design parameters which
characterize full-scale DAF units treating petroleum refinery wastewaters both with and
i without coagulant addition. It should be emphasized that air flotation without chemical
I addition is not widely used in refinery wastewater treatment. Moreover, the addition of
• pah/electrolytes has not substantially increased oil-water separation by air flotation
': systems to warrant their use as anything other than a coagulant aid. This is indicated b)
the data presented in Figure 16 which depicts air flotation process efficiency as a
function of influent oil concentration. The data specify treatment performance of both
DAF units and induced air flotation (IAF) units with, various ranges of polyelectrolyte
! addition. It is apparent from these data that polyelectrolyte addition did not improve
j flotation treatment efficiency, and in some cases, it may have even hindered effective
| oil removal. ; ' :
,i |
The variation in the treatment performance of a pilot-scale IAF unit used pri-
! marily for oil-voter separation is presented In Figure 17. Pollutant removal efficiencies
! are essentially equal to those achieved by the DAF units, but .operation is more difficult
! and chemical flocculant requirements are significantly higher. Moreover, the skimming
n-are much less concentrated and represent a 'significant waste solids disposal problem.
Pi .jr. N-. MbfR
-------
TABLE 12. EVALUATION OF DISSOLVED AIR FLOTATION TREATMENT PERFORMANCE
Oil and Grease Re»val
Sample Influent
Observation («g/l)
1
2
3*
4
5
«•
7
8
9*
10
11
12
13
1.781
2.029
1,034
3,500
700
970
1.530
230
420
720
310
324
170
Effluent
("•9/O
20
35
50
54
60
160
73
60
_190
130
90
72
70
Percent
Removal
99
98
95
98
91
84
55
74
55
82
71
78
59
Influent
("9/D
1.335
4.115
2.813
4.525
1.660
1.680
3,663
793
1.200
4.432
4,702
?86
966
COO Rem»9l
TSS Rewoval
Effluent Percent
(mg/t) Removal
492
552
-
1.125
656
615 .
555 :
453
675
1.968
2.551
513
523
63
87
-
75
60
63
85
43
44
56
46
48
46
Influent
(mg/1)
740
1,304
650
805
244
229
466
75
152
222
115
142
95
Effluent
("9/1 )
86
48
36 -
170
41
27
50
32
121
106
34
77
70
Percent
Removal
88
96
94
79
83
88
89
57
20
52
70
46
26
•Addition of five «g/1 polyelectrolyte.
-------
TABLE 13. DISSOLVED AIR FLOTATION DESIGN PARAMETERS
Example Refinery
Design Parameter
Average Inlet 011 and
Grease, mg/1
Hydraulic Loading Rate,
gpm/ft'
Detention Period, minutes
1n flotation chamber
Recycle Pressure,. pslg
Recycle, Percent of Feed
Air-Water Ratio,
SCF/IOO gal
Chemlcul Addition prior
to Flotation
Inlet pH
Design Flow, mgd
011 Removal Efficiency,
Percent
Suspended Solids Removal
Efficiency, Percent.
A
-
3.5
19.5
50
25
0.8
Yes
-
30
70-85
30-50
B
270
2.3
23
40
33
1.0
No
- •
4.3
60
*•
C
70
2.9
2>
40
33
1.0
Yes
-
5.7
92
72
D
112
2.3
20
45
see note
0.36
Yes
6.5-10.0
2.5
66
"•
E
- - •
4.0
in
40
50
-
Yes
8.5-10.0
2.5
75
"
F G H
100 125
2.5 2.9 2.3
?5
40 -
35 25 see note
1.0 - -
Yes Yes No
-
. - ." . -
92 72
" • • •
NOTE: Full flow pressurlzatlon method.
-------
MM
0*00
m '
u
5
. OlOO
t oosoi
2,0063
= 0070
- 0060
O Mwn
O
I OOOT
= 0006
£0003
S
g 0002
MOi
<
O
f
1
'
^••^^^
^^•M^»
<
•
<
t
,
• \U WITH > 10 «f/l W
O IV WITH 10 « I « »
A I«T WITH « S •»/' Ml
• B*f WITH i $ »s/l Wl
2
^
< 1
a
•
A
1
•
trull UOITIOII
•1/1 mmi MOITION
ma AOOITIM
TKtl JUOtTIOM
1
I
,
1
t
I
, . »
'
1
] )
i , '
'
;
1
i !
•
I
1 i j
li,-
1 1
1 1
& 8 ? S S S888 g § § § 1 8J88 8 8 8 || |||g
IHTLUtHT OIL AND CRtASC («g/t)
FIGURE 16. Air flotation efficiency as a function of influent oil and grease concen-
.-*- . trotion.
112
-------
sort:
mOT-SCAlt OPERATION
2 9 10 19 20 30 40 SO 60 70 *0 85 60 99 66
PERCENT PROBABILITY THAT CONCENTRATION WILL BE LESS THAN STATED VALUE
FIGURE 17. Variation in IAF treatment performance for oil and grease removal
u
• ? 113
-------
data fqf an IAF unit treating petroleum refinery ballast wastewater are pre-
•.sented in Table 14. this,pilot-scale unit was operated under closely contrclled hydraulii
, conditions. In general, suspended solids removal was unimpressive, exhibiting a mean
,| removal of 33 percent with greatly decreased efficiency at the higher influent solids con-
'centratfons. Conversely, effluent oil concentrations were consistently low with increase
'•»o?l removal efficiency of the higher influent oil concentrations. In addition, the IAF --
^unit produced o targe volume of float material,which was between four ond nine percent
of the throughput volume.
! , ' '
I ' ' '
I The variations in the treatment performance of both a TPS unit and an IAF unit
'used for suspended solids removal from petroleum refinery ballast waste waters are pre-
sented in Figure 18. These data are depicted as a function of anticipated probability of
[occurrence. The curves presented in Figure 18 indicate that the overall removal of
suspended solids by the aforementioned processes will increase with increased influent
solids concentrations which also was observed with the removal of oil and grease.
I , ' ; ' , ' '
j Curves which depict the relative treatment performance of oil-voter separators
•"for the removal of dispersed and some emulsified oil fractions from refinery wostewaters ''
; are presented in Figure, 19. As discussed in the preceding paragraphs, gravity differentia
'oil-water separators of either the API or TPS type are needed to absorb shock oil,loadings
I since excessive influent oil concentrations will reduce the performance of DAF and IAF
' oil' removal systems. Of the two gravity separators, both perform reasonably well and arc
| low energy and low maintenance systems. Although API separators are more widely used
: for primary oil removal, the current trend is to install TPS units because of their more
J economical price and reduced space requirements. The data presented in Figure 19
j indicate that a TPS unit also will experience better oil removal efficiency than the con-
! ventional API separator. Since gravity oil separation by itself cannot be depended upon
to meet the influent oil criteria of biological treatment systems, additional oil removal b;
a DAF or an IAF unit will be necessary. The data presented in Figure 19 also indicate
that an IAF unit is less stable than a DAF unit and it will obtain slightly lower oil remove
!efficiencies, as illustrated. j
Recently, pilot-scale fibrous-bed coalescers have been applied to wastewoter
treatment for the removal of residual secondary oil-water emulsions that are not separable
under prolonger detention in gravity separators. A fibrous-bed coalescer is a'fixed-filter
element constructed of fiberglass or other materials which act to coalesce oil droplets one
I break emulsions. I In a continuous water phase, oil droplets hove a greet affinity for the
j special fibrous materials (some fibers are lesi than 0.5 microns in diameter) contained in
J the coalescing element. When oil-water emulsions are forced through the element, the
micronic oil droplets preferentially adhere to the fibers where coalescence takes place.
As the oil droplets grow, they migrate through the elements. The droplets emerge from
the coalescer elements large enough to rapidly rise to the surface of the water. The
velocity of flow is controlled by the design of the separator to prevent oil droplets from
being entrained with the flow of wostewoterl Trie oil is removed from the water surface
[by skimmers or other withdrawal devices one
the oil free water leaves the bottom of the
14
-------
TABLE 14. EVALUATION OF INDUCED AIR FLOTATION TREATMENT PERFORMANCE*
o>
Influent
Parameter
Oil and Grease, ng/1
Suspended Solids, ng/1
BOOj. ng/1
COO. ng/1
TOC, ng/1
Phenol, rag/1
Turbidity, standard units
Color, standard units
Mean
29
70
92
239
115
9
122
2
Standard
Deviation
13
40
13
74
18
6
J7
1
Effluent
Mean
6
45
57
126
77
6
63
1
Standard
Deviation
4
38
15
49
13
5
26
1
Percent ftesnoval Efficiency
Mean
76
33
38
47
33
32
47
54
Standard
Deviation
17
37
15
16
10
29
24
41
•Evaluation performed with a r"lot-scale IAF unit on petroleum refinery ballast wastewater.
-------
•00
700
600
500
400
500
too
J 100
*0
2 «°
- TO
8 60
0 SO
I 40
W
i *
i»
to
t
7
t
.
'
=3
v^
•/
•
^
y^f*~
•
i
/
X'
/
X"
(
•
/•
S*
^
K-
s
if
•T)
^
•*"
1
>^
f
•if
^
,
>
^^
^\
iutt>-fuu t
irruitur
1
-J
.INO
l/r
<
r
veto
UMT
,
'M». ...
X
X
trOMTOI
L.
r T
AI« not
,
AT 10*
t • 10 M tO M 40 60 (0. 70 W SS SO t» »•
HRCENT mOBABIUTY T.iAT CONCENTKATION Will IE LESS THAN STATED VALUE
FIGURE 18. Variation in TPS and IAF treatment performance for suspended solids removal
•s~
*t*
116
-------
o
900
230
200
=50
5 too
50
NOTES:
OILY WATER MIXTURE CHARACTERIZED
IV OESPERSED AND EMULSIFIED OIL
OIL DROPLET
IAF AND DAF
ADDITION
'•,
,
.
/
. ^
S LESS THAN 1
OPERATED Wll
/
/
/
50 MICRONS
'H CHEMICAL
/
/
S
.. — • —
API
^
^
S**^
• -,
"" ""
• '
•
-TfS
^..- — — "
••»••"*"**
.
'
'
1
IAF
-~~WF
100 . 200 300 400 500
INFLUENT OIL AND GREASE (mg/1)
600
FIGURE 19. Relative performance of oil-wo tor separators for treatment of
dispersed and emulsified oils
117
-------
*£ _ Cooiescffice In a fibrous b«d Involves three steps; namely: (1) interception of__i
;flne oil droplets by fibers; (2) attachment of droplets to the fiber* or retained droplets; j
and (3) release of enlarged droplets from the fibers. The actual process of demulsification
I is due to the rupture of the interfacial film which surrounds each droplet when the j
', emulsion flows through the small passages in the fibrous media. The subsequent coo- j
!-)«
-------
IW
•0
TO
to
to
40
~ JO
^,
aw
1
i
.1 »
0 •
E:
i *
M ft
4
I
f
OW." VITN CMWItAL ABOITIOM
ar.M&ut «toiA coAitscm
,
/
•
^/
^
I
r
/
O
o
V
x
a/
z
mn>
wi «
.
X
n IAS n
£
<*
OOUCTI
•
0
a^
r
<
a
M M
/
£>
JTfi
/
MIC
9
/s
II
7
s,
tO M 40 90 W TOK 1100
•0 I
OIL AND CRCASt (•9/1)
FIGURE 20. Relative oil-water separator performance when treating low levels of oil in
wastewater.
i ' . **•
? 7 119
-------
TABLE 15. PROCESS COMPARISON FOR OIL AND GREASE REMOVAL
Oescrlptlo*
Advantages
Disadvantage*
*lr Flotation
OMerical Motcwlatlon
nitration
CMtttcenc*
ProetM
•1oloo,1ca1 Processes '
Carton Adsorption
i. CM. m. •*>!
OAT. ur
IKed I* COftJMCtta* «ltk
gravity separation end «lr
flotation.
Sand, anthracite, ewttl-
ecdla, crushed gr«p>lt«.
Hbroot M4U
ttVtrt*
Activated tlwlgt. «$
MC vlll act both «*
•f'tctlwt.fntratlM
and ceaV*scal of 'rr* and
dlspci-sfd oil; staple and economical
treatment operation.
Potential 'or treatment of suspended
solids; effective rernwl of dispersed
and e«w1slfled oil with cnexlcal ad-
dition; reliable process whtch effec-
tively treats shock loads.
Potential for treatment of high sus-
pended solids.
Cffectlm) suspended solids renovalt
has application to the separation of
free, dispersed and ewlslfled oil.
tffect<«* r«wvt! of %11 oil ccwpo-
ncnts eicept soluble oils.
. Soluble ell removal dewmStraUd
laboratory tests.
tffectlv* soluble oil removal.
tffectlv* rewval of all oil
•ents »i«cludln9 soluble *IU.
Halted efficiency for re«»val »f tmi'f
lifted oil; «111 not remve soluble •U|
restricted to the rtnoval of ell
droplets greater than ?0 etlcron*.
~ •
Cheolcal tludoe handllnj vhen coafitMU
are used.
Chnlcal sludo« produced.
Requires bacln>:-.«fhv snd_ the-
creates t subsequent treateient probleev.
levels of suspended Solids «1II I*.
duce fouling; potential for biological
foutlnq; needs eitentive pretreatnUt
not demnstrated as * practical process
In full-scale operation.
low flui rates; oeotirane fouling tad
Halted «e«br
-------
[jand carbon adsorption treatment processes will be presented in subsequent sections of ___
| ffiis report. , |
Intermediate Treatment Processes
The need to equalise wastewater surges from refinery and petrochemical pro-.-
iti-os-orv intermediate, step in a treatment system-is, well established. Biological
('processes as well as physical-chemical systems operate more efficiently if the pollutant
I concentration and flow of the wastewater are relatively constant. The objec'.'ve of
j equalization is to minimize or coiitrol fluctuations in wastewater characteristics, thereby
! providing optimum conditions for subsequent treatment processes. The size and type of
i the required equalization bo?in will very with'the quantity of wastewater and the vori-
I ability of the pollutant concentration. Many of the wastewater discharges within a ;
I refinery complex are from washdowns, tank cleanings, batch operations, and inadvertent
i spills. Adequate equalization would provide damping of the organic fluctuations associ-
I ated with these intermittent wastewater sources in order to prevent shock loading of the
j biological treatment sysrem. • 'j i
l^—TJJ. ——. —. —. . —f — *• — — •- —• -—•• -f—- -— — • —• --* -~— - -• — -• •— - - f._ .-(
I t ,
| Mixing Is usually orovided to ensure adequate pollutant concentration equali-
zation and to prevent settloable solids from depositing in the basin. In general, the
design problem consists of determining a functional relationship between the size of the I
I equalization facility and the probcHe reduction in the fluctuations associated with the
j pollutant concentration of the wastewater. When flow varies considerably, and not in
the same manner as the pollutant concentration; the equalisation facilities should be
physically arranged to provide additional storage beyond the normal required volume in
order to dampen both the flow rate and concentration. Such off-specification impound-
ment of hydraulic surges and high organic loadings would require additional pumping
capacity. The required size and mixing features of the equalization basin depend on the
wastewater characteristics, process plant operating schedule, and any unusually severe
operational fluctuations. \
The presence of an equalization basin will often minimize the chemical require**
merits necessary for neutralization of wastewqters by mixing acidic and alkaline streams '
which would normally be discharged from the refinery at different times. Although !
refineries and petrochemical installations hove many individual streams which are highly ,
acid or alkaline, the combined wastewater discharge is generally slightly basic within a J
pH range between1 7 and 10. Consequently,! the combined wastewater from many install- •
ations does not reqjire neutralization; however, the separate discharge of ce-»s:n streams!
must be neutralized. 'Such streams include dilute acid or alkaline was waters, spent !
caustics from caustic treating operations, acid sludges from alkylatibn, acid treating
processes, and spent acid catalysts.
Typical replications where neutralization is required include: '(1) emulsion
breaking through acidification and neutralization following gravity separation; (2)
neutralization prior to biological treatment; (3) neutraliiation of specified streams to
prevent corrosion;^ (4) pH adjustment to prevont unwanted precipitation of certain
«4
-------
^'constituents; and (5) pH adjustment to enhance coagulation and sedimentation. j
i" : . • !
I Design of a neutralization system is predicated on titration curves which deter- j
jmine the to to I alkalinity or acidity of a solution. These curves graphically depict the pH
[change per milliequivolent of reagent. The design and control of a neutralization pro- j
,c«ss includes consideration for mixing, reagent feed, and pH control systems. The -*- >
"'economics of the neutralization system depend on th« rharoctertstics end quantity of - - - J
j wastewater neutralized and the source and supply of neutralizing agents. ;
, I i !
' Biological Treatment Processes
i
j The treaiiftcnt of wastewaters discharged from refinery-petrochemical install- ;
'ations encompasses the remove! of many pollutants, by! most of the attention in rwfinery '
pollution control is focused on organic removal. Biological treatment is generally con-
sidered the most effective techniquw available for removing organic materials from !
/petroleum refinery wastewaters. Essentially, the basic biological treatment processes
include suspended growth activated solids systems, fixed microbial slime systems, and
iNvaste stabilization lagooning. Usually wastewater pretreotment is required to remove "*"'
!oi!s, suspended solids, and toxic substances. Additional unit operations required to
{maximize process stability are provided by such processes as neutralization, equalization,'
land surge or holding capacity (off-specification impoundment) as was previously discussed.
The principal suspended growth activated solids system which is most effective in
• the treatment of refinery-petrochemical wastewaters is the activated sludge process. This
i is a continuous treatment system in which a microbial population is contacted with waste-}
[water in a completely-mixed reactor and then separated by gravity from the treated !
i liquor. The basis for the design of an activated sludge process is predicated on main- !
jtaining an1 environment in the bioreactor which is optimal for the growth and activity of
select populations, of acclimated microorganisms.
Treatability studies using bench- or pilot-scale process simulation techniques ore
Jused to formulate the basic design criteria for suspended growth systems and predict
j treated effluent quality. These treatability studies are conducted: (1) to vertify the
{applicability of biologically stabilizing the organic constituents in the waste water; (2) to:
develop the basic'kinetics for process design) (3) to establish the impact of various
organic and inorganic wastewcter constituents upon biological process efficiency; and
(4) to predict the treated effluent quality and evaluate this quality in terms of effluent
requirements. This process simulation will pfovide predictor relationships in the form of
i mathematical expressions which describe the activated sludge process arid its applicability
to the treatment of the given wastewater.
The actual degree of treatability characteristic of a refinery wastewater is a.
function of the type of refinery, the type of'crude charge, the age of the facility, the ,
type of collection system, the quantity of utility water blowdown, and the degree of in- j
pjont control. For these reasons, the actualjdesign of the basic activated sludge processj
vary from one installation to another.
T -I 1
-------
; _ Petroleum refinery wastewaters, while highly variable in nature, rend to • '
exhibit similar biological treotobility characteristics. Consequently, if the wostewater
characteristics for a given refinery are known and the capacity of the various refinery
processes are considered, it is possible to use historical data collected during bench-scale
treatability studies on other petroleum refinery wastewaters to obtain the required design
coefficients. The pertinent design coefficients fit mathematical models which describe
^he biological oxidation process. These models
-------
"'oxidized by endogenous respiration; Xo is the influent suspended solids; Xe is the ',
I effluent suspended solids; and AX is the net sludge accumulation'™. The basic theory of
biological woitewatb.' treatment and the kinetic development behind each of Hie afore-
mentioned models are adequately described elsewhere . . ;
Engineering-Science, Inc. (ES) has conducted numerous investigations into the
^reotraent ^>f petroleum refinery wostewaters tn which wesfe «harocteftr,ot?on und bench- ,
scale cctivated sludge treatability data were collected. Table 16 summarizes the results
of nine different investigations into the biological treatment of refinery wastewsters and
lists the biological coefficients which characterize these wastes. The majority of these
data are presented in terms of COD. In general, the organic removal rote coefficient in
terms of COO varied from a low of 0.007 liters per milligram of VSS per day ro a high of
0.017 liters per milligram of VSS per day at 24°C for the refinery wastewaters investi-
gated. The highest value for the organic removal rate coefficient presented in Table 16
indicates the likelihood of extensive stripping of organics to the atmosphere during
aeration of the wastewater. A typical magnitude of the organic removal rate coefficient
: characteristic of petroleum refinery wastewater treatment would probably be situated
'between 0.008 and 0.01 liters per milligram of VSS per day at 24°C. *""
With several exceptions, the majority of the sludge accumulation coefficients
characteristic of refinery wastewaters varied between 0.4 and 0.5 for the magnitude of
"a" and from 0.06 ro 0.1 for "bn. The range associated with the oxygen requirement
coefficients a' and b' is 0.35 to 0.6 and 0.01 to 0.14, respectively.
i The activated sludge process, by virtue of its high MLVSS concentration, is
classified as a high-rate biological process which can tolerate more concentrated con-
. taminant levels in its influent stream. In general, biotoxicity in an activated sludge
: process con be minimized or completely circumvented by in-plant controls, effective
• equalization, and complete mixing of aeration basins. Most plant upsets result from
inadvertent dumps or spills into the process sewer which shock the biological system. The
.discharge of excessive levels of chromates, sulfides, ammonia, free oil, or other known
biotoxicants common to refinery-petrochemical wastewaters also will severely upset bio-
logical treatment.
I ; ' . • ' ;
i The major design considerations for the successful activated sludge treatment of
, refinery,wastewaters include: (1) the organic loading; (2) the organic removal kinetics;
! (3) the organic and inorganic fluctuations of the wastewater; (4) the free oil loading;
i (5) the temperature effects; (6) the potential biotoxic or biostutic effects; (7) the oxygen
utilisation and sludge production rates; and (8) the settleability and thickening character-
i istics of the biological sludges. The design organic loading for most activated sludge
, systems will range from a low of 0.10 pounds of BODs pier pound of MLVSS per day for
• extended aeration to a high of 1.0 pounds of BOD5 per pound of MLVSS per day. Higher.
j loadings generally result in poorer treatment efficiency and higher treated effluent ,
| organic levels. ! ' • ' '
124
-------
TABLE 16. BIOLOGICAL TREATMENT COEFFICIENTS FOR PETROLEUM REFINERY WASTEWATERS
IM
Cast Category
1 C
f e
1 0
4 C
~ J** C
t 1
7 1
• e
t i
n^tat. ti
"Wat-mi* 1
Orfmtc ••*<
s in '«|isr
f (HB.PD) M)D{ Basts
406 0.017
-60
M
63
427
427 0.0035
20S O.OI/
n
. tt 0.006
«*tr*t*T«-24*C.
>U1o«lca1 rtacter.
w»«1 IUU* Sl*4ot GroMth Oxy«t* ^t«i i«tu coo IMU wo, to«
-------
A lowTtjte suspended growth octivoted solids system in which oxygenation and_ '
, mixing are normally provided by mechanical surface aerators Fs the aerated lagoon bio- ,
I logical treatment process. The mechanical aeration system is designed such that the j
power levels are normally too low to maintain dissolved oxygen in all portic.n of the •
1 basin or to completely mix the system. The operating MLVSS level is normally between !
2 and 10 percent of that maintained in an activated sludge process; consequently, aerated
te-4fonsi«o»-orgon5e 4oodings, toxic substances/ or»d temper- *
, oture effects. ,
i ' i i
i
Aerated lagoons are operated at high organic loading levels because of the low
concentration of microorganisms suspended in the wastewater. These loadings usually vary
I between 1.0 and 2.0 pounds of BOD5 per pound of MLSS per day. The aerated lagoon is
effective for the treatment of most hydrocarbons such as those present in the discharge
; from topping plants, but exhibits a reduced efficiency for the treatment of more complex
i wastewaten such as those discharged from integrated refineries' . j
{ Because of the low mixing levels in aerated lagoons, large volumes of oily
"SJlrds tend to settle within the basin where they decompose anaerobically releasing " s'""
additional BOD to the upper layers of the lagoon. This feedback can be os high as 20 ',
percent of the influent organic load during the summer montfis. Moreover, winds tend to
i stir up bottom sediments, causing a deterioration of effluent quality in terms of suspended
solids and partially stabilized oily material. ,
A tabulation of effluent-quality data from a modified aerated lagoon treating '
j petroleum refinery wastewaten is presented In Table 17. This aerated lagoon was oper-
• oted with intermittent solids recycle and post-clarification of the effluent, HHJS providing
, a greater degree of operational control. These data indicate a better effluent quality j
j than normally associated with aerated lagoon .treatment. The conventional approach to j
' aerated lagoon design specifies that the treatment unit will function either as a process :
! preceding waste stabilization lagoon ing or as an interim treatment process which can be |
i converted to an activated sludge system as was the case illustrated by the data presented !
! in Table 17. j , j i
Temperature is a particularly important variable in the biological treatment of '
refinery wastewaten, and the influence of temperature on biochemical reactions is well-.
documented'"'. Excessive temperature losses through a'biological treatment system
during winter months in a northern climate might lower microbial activity to the point of ;
failing, to meet effluent BOD quality standards. In general, temperature effects are more
pronounced with the increasing solubility and complexity of the wastewater and activated
sludge treatment would be significantly affected by aeration basin temperature.
I ' ' ,
The level of oil and grease discharged to a biological treatment system can have
c significant impact on treatment performance. In general, hexane extractable oils will |
adversely affect a biological system as the concentration of free and emulsified oils in <
j the mixed liquor approaches 50 to 75 mg/t.i The most significant problem associated with
presence of oil'In biological systems is attributed to the lowering of floe density to a [
-------
TABLE 17". EFFLUENT QUALJTY FROM~A MODIFIED AERATED LAGOON TREATING TETROLEUM RHFINERY WASTE
WATERS* (MONTHLY AVERAGE DATA)
Parameter
Month
January
February
March
April
May
June
July
August
September
October
November
December
Suspended
Sol Ids
(mg/1 )
28
28
- 30
25
21
26
24
33
22
17
27
23
COO
(mg/1)
221
196
122
99
88
81
90
90
72
108
131
110
BODs NH3-N
(mg/T) (mg/1)
52
22
23
15
9
10
14
12
11
15
27
18 ,
20
41
38
45
36
28
22
37
30
29
48
32
Phenols
(mg/1)
0.024
0.114
0.044
0.028
0.024
0.036
0.031
0.024
0.017
0.024
0.011
0.015
Total
Cyanide
(mg/1)
0.119
0.167
0.40
0-052
0,076
0.041
0.0)
0.0)
0.015
0.045
0.22
0.17
011 and
Grease
(mg/1 )
8.3
3.3
2.8
2.S
2.0
1.7
1.5
1.9
1.0 .
1.8
2.7
3.5
*EPA Category D,
-------
, IE
whore the sludge settling properties ore destroyed. Free oils will coat the bio- . ._
; logical floe and prohibit the efficient transfer of oxygen and .substrate within the biomass,
The apparent quantities of oil associated with the loss of suspended solids from a secon-
dary clarifier are Indicated graphically in Figure 21. The accumulation of oil on sludge
|particles will .contribute to effluent quality deterioration. Therefore, .pretreatment
Jiocilitiis for oil-water separation should be capable of removing oil to an acceptable —
,*4£Oft€£o i rot wxy i£V6i fOf "^t rt~ct WVF opcrottcw^'O* Titc bfOiOQtc^i ~»rcGFfflCJi* process * ~" *
Biological treatment processes are very limited in their potential for oil removal
i While microorganisms are capable of oxidizing most soluble oil fractions present in
i refinery wastewaters, they are limited in their ability to degrade, free and emulsified oils
within the contact times characteristic of most biological treatment processes. The'best
documentation of biological process effectiveness for the removal of oil is the analysis of
: full-scale plant operation. ES has compiled effluent oil quality data from selected bio-
i logical treatment plants handling refinery wastewaters. These data are presented in
i Figure ?.2 as a function of anticipated probability of occurrence. These case histories
1 present operational data for an activated sludge process operated both with and without
"fJblymer" addition to the secondary clarifler, an activated sludge process with tertiary ""
'filtration, an aerated lagoon process, and an aerated lagoon process with tertiary
'filtration. Although these data were obtained for treatment of wastewaters from refinerie:
which represent various production subcategories (as defined by EPA), refine cruHes of
varying characteristics, and have other process idiosyncrasies which impact on the raw
|wastewater characteristics; they do indicate a probable range of obtainable effluent oil
•and grease levels. The median values range from a low of 3.0 mg/4 to a high of 12 mg/4
teffluent oil and grease, while two percent of the observed values (98 percentile prob-
ability) exhibited a range from 20 to 60 mg// effluent oil and grease concentrations.
! With the addition of tertiary filtration, biological treatment processes can meet allowable
;Best Practicable Control Technology Currently Available (PPfTCA) discharge levels for
{the presence of oil and grease in .treated refinery effluents ' .
i I ' .
j The second type of biological treatment process to be considered herein is the
tfixed microbial slime system. In this system; attached microorganisms remove suspended
land dissolved organic material from the wastewater as it flows over fixed slime surfaces.
I At the same time,; oxygen is absorbed into the wastewater from the air which is induced
through the system by natural draft. At the surface of the slime, oxygen and organics are
made readily available to the microorganisms by mixing and diffusion in the wastewater.
The effective film depth is that which is in immediate contact with the wastewater. The
area between the filter media and the effective film depth is considered anaerobic be-
cause all oxygen is consumed .before it can diffuse to the inner layer. The relative sizes
of these two regions depend on the concentration of oxygen in the wastewater and the
turbulence between the liquid film and the attached slime surface.
I
A typical fixed microbial slime system is the conventional trickling filter in
which granular media are generally used to support the biological mass. The introduction
qfjynthetic filter media, with low bulk density, has resulted in the use of deeper filters_
••operated at gieotly increased organic and hydraulic loading rates. Trickling filters ore
I "I ~ -I " . ... | . . ; ' "^
128
-------
•OV
170
160
150
140
130
^ 120
J! no
i/»
o
3 100
0
o 90
o
£ eo
3
j 70
<
° 60
50
40
30
20
10
o
•
<
*
/
/
9 /
/
*
1
•
& j
/
i
'
1
A
/
.
.
/
r
•
•
•
/
/
'
'A
/
'
•
/
/
f
'
EQUATION OF LINE:
OIL AMI/ GREASE • 0.6436 TSS - 29.13
•
OK) 20 30 40 SO 60 70 BO 90 CO IK)
OIL AND GREASE (mg/1)
FIGURE 21. Relationship between suspended solids and oil and grease
concentrations in biological treatment plant effluent
? f 129
-------
2 9 10 IS 20 30 40 60 60 70 8O 65 BO , C5 98
PERCENT PROBABILITY THAT CONCENTRATION WILL BE LESS THAN STATED VALUE
FIGURE 22. Effluent oil and grease from biological unit processes for the
treatment of refinery wastewaters
130
-------
•not used extensively for the treatment of refjnery-petrochemical wastewaters since the j
trend is toward hi^h-rate biological processes. Nevertheless, they do have application
as roughing devices preceding other biological or physical-chemical units.
I i
i,l'
The rotating biological surface (RBS) unit is a fixed microbial slime system which
.has application as a secondary wastewater treatment system for handling refinery waste---,
e-diom«ref oloiHc media mounted on o -rotating 4>orizonto!
shaft such that approximately 40 percent of the surface area comes in contact with the
wastewater at any one vime. Trm large microbial population which is fixed to the
rotating surface permits a high degree of treatment iii a short retention period. During '
rotation, the contactor carrieS'O film of wastewater into the air which absorbs oxygen
while trickling dpwn the biological surface/ Microorganisms in the biomass then remove ;
both the dissolved oxygen and organic material from the wastewater. Further treatment i
occurs as the biological surfaces continue rotation through the bulk of the wastewater. i
The shearing forces exerted on the biomass as it passes through the wostewator cause ;
excess slime to slough from the surface of tht media. This prevents clogging of the media
surfaces and maintains a constant microorganism population on the media. The turbulence
traced by the rotation of the biological surfaces through "The wastewater wTtl maintain1"*'
ploughed solids in. suspension until the flow of treated wastewater is passed to secondary '
clarifiers for solids separation and disposal . i
The blomoss associated with the RBS unit is shaggy in appearance with many
elongoted macroscopic filaments which project outward into the adjacent .liquid film of ,>
the wostewater. This provides a much larger active biological surface area than that I
defined by the surface area of the support medium. It also permits substrate and dissolved
oxygen to reach a greater portion of the biomass. The nature of the attached biological '
growth is due primarily to the continual drag induced by contactor rotation through the
wastewater and the draining of entrained liquor from the contactor surface.
The results of a pilot-scale study into the application of the RBS unit for the
treatment of petroleum refinery wastewaters are presented in Table 18. In general, the
effluent quality from this pilot-scale RBS unit compares favorably with other biological
treatment processes. These data indicate that the effluent quality associated with an RBS:
unit is a function of hydraulic loading. Forjinstonce, better removals of soluble organicsj
oil and grease, and sulfides were obtained at the lower hydraulic loadings. On the other
hand, a degree of stable nitrification was achieved at each of the hydraulic loadings
examined during the study. Air stripping of ammonia was not considered responsible for
the ODoarsnt reductions observed during the ktudy because of the operating pH and temper -
ature levels. Tests conducted in microorganism-free reactors indicated that the loss of
lest than 10 percent of tha ammonia may be attributed to air stripping**"'. Since the
apparent ammoniaj reductions experienced wi
a degree of nitrification did occur.
The variation ?n pilot-scale RBS treatment performance for both COD and oil
removal is presented in Figure 23. In geneml, the RBS unit did not perform consistently
well for the treatment of oil and grease indicating the need for tertiary treatment
th the RBS unit were greater than 25 percent,
1
-------
TABLE 18. PILOT-SCALE RBS EFFLUENT QUALITY FOR TREATMENT OF PETROLEUM REFINERY WASTEWATERS'
0.9 gpd/ft*
Hean (mg/1)
Standard Deviation («g/1)
95 Percentlle («g/1)
Retnval Efficiency Percent
1.9 gpd/ft2
Item (mg/1)
Standard Deviation (mg/1)
95 Percent Me (•9/1)
Removal Efficiency Percent
*.« qpd/ft2
Mean (wg/1)
Standard Deviation (»g/1)
95 Percent lie (wg/1)
Retwval Efficiency Percent
COO
UO
142
448
58
156
70
293
44
•
212
71
351
35
BOO,
19
7
33
90
25
: 6
36
87
33
21
74
69
TSS
20
18
55
-
28
11
SO
-
65
48
159
-
OM and Great*
18
10
38
65
-
12
11
34
. 72
-
37
18
73
66
NH3-»
"
48
25
".
26
48
18
83
»
4
6
15
32
Sulftdt
0.2
0.2
0.6
99
0.3
0.1
0.5
99
0.7 .
0.7
2-1
88
Phenols
0.04
0.02
0.08
79
0.03
0.01 .
3.05
84
-
0.03
0.01
0.05
77
•EPA Category 0 Refinery.
-------
mOT-SCAU MTATINC IIOlOCICAl
IWr«Ct AT MtBMUlIC lOMIM Of
l.t
ft 10 IS 10 90 40 SO «4, 70 SO •» K> M fid
FRoeAeitm THAT cwctrrnATiw WILL it uss THAN STATIC VALUE
^FIGURE 23. Variation in RBS treatment performance for COD and oil and grease removal
A .. . . • • • ;.'
133
-------
, -' "' - ______ _ _ L_ _ £*
^fprpcesses tc enhor^ce effluent qunlity. The variation In pilot-scale RBS treatment per- __
! formance for phenol removal is presented in Figure 24. The median effluent phenol con-
centration is approximately 0.025 m
The third alternative biological treatment process for refinery wastewater treat-
r*»«nt is that of waste stabilization Icgooningj. In general, waste stabilization ponds — i
jepand-on the noh>fo4-equo»k: processes of bocteriol-.o*^ «4
-------
s
I.OOO
0*00
ocoo
Q700
0600
OSOO
0400
0.900
6JOO
OIOC
009 C
0063
0070
00(3
0050
0040
0030
0020
0010
0038
oooe
0007
oooe
0005
0004
0003
ooor
'
•
V^VH^H^V^I
turiu
Trip
'
^H^^
—
•
,
,
;« j
XT J
1
X
x
i
T
/
/
/
/
/
/
[/
X
/
•
,
*
/
X
•
(
•— •
•
1
•'
1
— • »-^
^^^^
^•^tf*
,
r*
«
,
J
i
.
MTIt:
fllOI-KMf «OT»TI«C IIOlOCICAL
tUUU.1 *T HYDMULIC LOA9INI Or
t ^ J ^
PA UT1&MT » Wf IH|«T
2 » 10 is >o so 40 w> to rc oo e> BO »> ' on
rtKCENT ^KOOABILITY THAT COKCtmAI IOM WILL (C LESS THAN S1A11D VA1 Ht
FIGURE 24. Variotion in RBS treatment performance for phenol removal.
! 135
-------
-jqmount of porticulpte matter escaping the filter exceeds the effluent guidelines or when J
j the headless through ?he filter caused by solids capture exceeds the limiting value. j
1 ' • ! . i ' ' J
' A comparison of both the upflow aiid down flow mode of filter operation as applies
to the tertiary treatment of petroleum refinery wastewaters is presented in Table 19. Both
of these pilot-scale filtration units were fed biologically treated secondary effluents, —,
chemically -conditioned before application to the filter. Run times were considerably —[
longer with the upflow sand filtration unit and the overall operational performance
. characteristics were more favorable.. : • i
i ' ' ' I
I Irrespective of ,heir relative efficiency, all filters require the periodic remove! |
of deposited material. Cleaning of a filte; M the end of a run cycle is generally accom-,
plished by high-velocity baclcwashing and the overall efficiency of a filtration unit is
greatly affected by the amount of backwash water required. In general, the contaminated
.backwash water cr.n be treated separately or rechanneled back through the biological j
:process. i • '• t
• •' ! ' !
r1* '" ~~ A measurement of filter efficiency is specific capture which represents the masV.
'of par.ticulate matter retained in the filter'per square foot of surface area and foot of ',
head loss. Specific capture i» affected by approach velocity and loading rate since filters
i operating at higher loading rates will experience higher intergranular velocities, thereby
! driving solid particles deeper into the filter bed and increasing specific deposit. A
• number of values for specific capture determined from the operation of both shallow and
i deep-bed filtration units are presented in Table 20. In general, an increase in the '
j magnitude of the specific capture or efficiency of the filter is offset by increased oper- !
'ational difficulties and maintenance problems. !
I j '.
General'characteristics common ro most filtration processes used for the treat- I
iment of biologically treated secondary effluents are: (1) that the variation in wastewaterj
flow and effluent quality from biological treatment will not adversely affect filter
! efficiency; (2) that BPCTCA discharge limitations for suspended solids can be met with
either deep-bed or shallow-bed filtration units; and (3) that filter performance is
extremely sensitive to changes in media size and configuration.
i
A second tertiary treatment process applied to the removal of refractory trace
organics from petroleum refinery wastewaters is carbon adsorption. Activated carbon
adsorption involves the use of high surface area activated carbon as a physical media for
surface adsorption, of organic contaminants from wastewaters. The controlling design
parameters are: (1) contcminant concentration in the waste stream; (2) temperature; (3) !
adsorbent area per unit volume; (4) adsorption characteristics of the contaminants; and 1
(5) diffusion considerations. The principal advantages of activated carbon treatment j
include: (1) the ability to remove select organic molecules; (2) the ability to withstand
shock loadings; (3) the ability to remove extremely low concentrations of sorbable con- , =-.
taminants; and (4)'the relative ease of operation. Carbon adsorption is a valuable treat* : "
I ment process for the removal of potentially toxic organic materials. Its disadvantages
3»include; (1) a high capital and operation cdst; (2) fouling; (3) the inability to remove
-------
TABLE 19. COMPARISON OF UPFLOW SAND AND DEEP-BED TERTIARY FILTRATION FACILITIES FOR THE TREAT-
MENT OF REFINERY WASTEWATERS*
Parameter
Upflow Sand
Deep Bed PVC Media
.Hydraulic Loading (gpm/ft*)
Cumulative Loading at Breakthrough
(gal/ft2)
Solids Loading at Breakthrough
(Ib dry solids/ft2)
Pressure Drop at Breakthrough
(psl/ft of bed)
Average Effluent TSS (mg/1)
Average Effluent £OD& (mg/1)
Average Effluent COO (tm,/l)
Shape of Breakthrough Curve
Effect of Chemical AddH'cn
4-10
1.800 - 4,000
0.10 - 0.28
0.35 - 0.90
5
3
73.
Sharp Breakthrough
Alum and polyelectrolytc. addition
did not significantly effect sol Ids
loading, but did Increase the
pressure drop.
6-15
200 - 850
0.05 - 0.20
2.5 - 8.0
16
R
74
Continuous Deterioration of
Effluent Quality
Alum and polyelectrolyte addition
Increased the solids loading.'
•Pilot-scale study for EPA Category B Refinery.
-------
TABLE 20. DESIGN CHARACTERISTICS FOR TERTIARY FILTRATION FACILITIES
Filter*
Deep Bed
Shallow ted
Shallow Bed
Shallow Bed
Deep Bed
Media Type
and
Configuration
24" sand and 36" anthrafllt**
16" sand***
16" sand*
16' sand+t
24" sand and 36" i«thraf11t+ft
Hydraulic
Loading
(gpm/ft2)
7
2
2
2
3.5 .
Specific Capture.
(Ib dry solids/ ft*
per ft H20)
0.015
0.010
0.046
0.022
0.07
Effluent TSS
("9/1 )
4
10
2
4
2
*P1lot-scale study for EPA Category C Refinery.
r> **0.5 to 1.0 m sand below 1.0 n* anthrafllt.
***0.5 to 1.0 m sand above 1.2 on sand.
f0.6 to 0.65 m sand above 1.2 nm sand.
f+0.6 to 0.65 IM sand above 1.0 ni sand.
ttf0.5 to 1.0 m sand below 13 anthrafllt.
-------
-'many inorganic compounds and some organic compounds; ond (4) the necessity for
extensive carbon regeneration equipment.
; The two types of carbon adsorption technology applied to the treatment of
refinery wostewaters are: (1) GAC in a continuous flow-through system and (2) the
addition of PAC to an activated sludge process. Various approaches to the application of
the GAC adsorption process ore: {1} carbon-btclogicel series treatment; (2) biological-
carbon series treatment; and (3) carbon adsorption as a total process. Each of these
explications requires primary treatment for the removal of oil and suspended solids.
Of the three GAC process applications indicated, only the series biological-
carbon treatment scheme.will probably be prevalent in the immediate future since many'
refineries already have made the capital investment in secondary biological treatment.
The carbon-biological series treatment option has some apparent advantages in that it
allows for a more effective use of carbon, dampens organic surges to the bioreactor, pre-
vents biological upset by removal of biotoxic substances, and allows for a reduction in
, excess sludgs production. Nevertheless, many of these advantages also are characteristic
"of the PAC adsorption process. Some obvious disadvantages are: (1) a potential effluent'
suspended solids and color problem associated with biological treatment; (2) a less
efficient utilization of the biological process; and (3) a strong dependence on a sensitive
biological process to consistently produce an effluent which will meet stringent quality
requirements.
The results of a pilot-scale GAC process treating the effluent from a conventional
API separator are presented in Figure 25 in terms of a probability analysis. As illustrated
in Figure 25, the effluent TOC concentration from the pilot carbon columns was con-
; sistently less than 30 mg/£, although the influent TOC levels also were fairly low due to
dilution with cooling water. The probability distribution of long-term mean effluent COD
i values from pilot-scale and full-scale GAC treatment systems treating refinery or rotated
; wostewaters is presented in Figure 26'". These data indicate that,the effluent concen- (
1 trot ions and removal capabilities of activated carbon vary significantly and must be
, assessed on a case-by-case basis. In general, a total carbon treatment system is not '
! satisfactory because of excessive organic leakage to the effluent, and it should not be '
' considered a process panacea for producing high-quality effluents2^'. . I
• ; i i
• I The application of full-scole GAC treatment to a relatively low-strength
j refinery waste water consisting primarily of contaminated »torm water indiccted mixed I
• results in ierms of both removal of organic materials and cost effectiveness. The operation
'of full-scale carbon facilities has indicated that carbon capacity, in rerms of pounds of
jorganics removed per pound of cPlfc°n applied, has generally been less than that deter- '';
i mined during pilot-scale studies . This can be attributed to the difference in
adsorptive capacity between virgin carbon which is often used in pilot studies and regen-
• crated carbon which is recycled in full-scale systems. '
l • ! .
The PAC process may have wider application for the treatment of refinery waste1* ,,
since it has been shown to improve the performance of biological treatment systems •
,._.._.^_.. . __ .^
-------
100
90
60
70
60
50
40
SO
20
? ,10
*- 9
o 0
I
X
)00
3
-INFLUENT
IFFLUENT
SHANULAIl ACTIVATED CARBON PILOT
PLANT 'OPERATED AT HYDRAULIC
LOADING OF 6 gpm/ft2
CFA CATEGORY
EFFLUENT
B REFINERY UASTEWATER
LJJI J
.2 & M> 15 20 SO 40 SO 60 70 60 65 BO 95 96
PERCENT PROBABILITY THAT CONCENTRATION WILL BE LESS THAN STATED VALUE
FIGURE 25. Variation in GAC treatment performance for TOC removal
140
-------
o
o
<
LU
I
O
000
coo
70C
600
too
<00
500
200
100
90
60
70
60
50
40
50
20
Ift
^^•«to
I j
• PILOT-SCAU SYSTEM TREAT
MFINERY UASTCVATER
• FULL-SCALE SYSTEM TREATI
BIOLOGICAL TREATMENT PL A
4 FULL-SCALE SYSTEM TREATI
REFINERY WASTtVATER
•
,
•
«—
i
J
^
,
^
ING "
MC -
NT
•
NO
^
s
f
V
•
y
/
f
/
V
t
/
/
I
•
1
•
29 10 IS 20 ,50 40 00 6Q 70 60 65 60 95 96
fERCENT PROBABILITY THAT CONCENTRATION WILL BE LESS THAN STATED VALUE
FIGURE 26. Effluent COD attainable from activated carbon system
¥ , Ml
-------
with a relatively minor addition of equipment to an existing plant. The primary opor-
"oting parameter that defines the performance of a PAC process is the equilibrium curbon
! concentration in tine aeration basin. This equilibrium carbon concentration is a function
i of the carbon dosage to the wostewater, the amount of carbon leaving in the final
i effluent, the quantity of carbon/biological sludge wasted, and the hydiaulic retention
'-time. It is estimated that a PAC variation of the activated sludge process can achieve o
MxkJJtiorKjf 54- pef c«r»*~TOC removol->>ver conver»tion«lra<:tivoted sludge treatment when -
! • *5^%O
treating refinery wastewaters . In general, the performance of n PAC process is high!
{dependent on the type cf carbon used. |
l i
i ,GAC treatment has application both as a tertiary process for the removol of
1 soluble oils from secondary treatment effluents and as a coalescing media for the sepa-
; ration of dispersed and emulsified oils. The oil and grease removal performance of a
• pilot-scale GAC system treating the effluents from a conventional API separator is pre-
, sented in Figure 27. These data are illustrated in terms of the probability of occurrence
.and indicate that the median effluent oil and grease concentration was approximately 2..
jmg//. The variation in pilot-scale GAC treatment performa:..e for oil and grease
'Tnemovdf "from"tfercfWd lagoon effluents is Tllustrcted'by trie probability distributions pre~~
| sented in Figure 28. These data indicate the characteristic GAC treatment potential for
j the removal of soluble oils. The data presented in Figure 28 also indicate that the medic
> effluent oil and grease concentration was less than 0.5 mg//. Both systems were precede
jby *and filtration.. In generul, the aforementioned pilot-scale performance character-
| ijtics agree with other experimental observations . Nevertheless, current operating
(experience with full-seals GAC treatment sytffim* has indicated median oil and grease
concentrations in the neighborhood of 8 mrj/£ . . Two documented full-scale case
histories indicating effluent oil and grease quality from GAC treatment of API *eparator
effluents'are presented in Figure 29. Although these data fall within the same range as
oil and grease effluent qualities from the activated sludge and aerated lectori processes,
they do indicate that application of carbon adsorption treatment to refinery wastewaters
is most effective as an organic removal step following biological treatr.ient and sand
filtration rather than as a replacement process for biological treatment.
SOLID WASTE DISPOSAL
Any attempt to improve the quality'of a wastewater is generally accompanied by
a solid waste disposal problem. Solid waste handling can be extremely costly, particular
in the case of wastewater sludges that must be dewctered prior to economical ultimate
disposal. Because; of the associated cosh and increasingly stringent regulatory controls,
solids handling and disposal operations must be considered an integral part of any treat-
ment program in order for engineering or economic evaluations to b« relevant.
The types of sludges generated by tie petroleum refining industry are: (1) once-
through cooling sludge; (2) cooling tower sludge; (3) alkylation.sludge; (4) waste catalyfi'
cracking catalyst;| (5) spent treating clays; (6) tank bottoms; (7) storm water silt; (8) oil-
_woter separator bottoms; (9) air flotation f!o$t; and (10) waste biological sludge. The
^general characteristics of these waste solids were discussed in previous sections of this
-------
IJBCO
*00
•00
TOO
600
600
400
too
100
»0
00
90
5 id
•nUfXT
tminm
X
VOTIti
tMmn.** WTi»«Tto CMHN mot rum
OHMTI0 AT NTOMUIIC iOMIRC OF
tM t»n6o«T i tiriinr
» » U »0 >0 4« SO W TO M •» 90 t5 »8
PKOMBIIITT THAT COMCINTMTION WILL 1C LCSS THAN STATED VALUE
. FIGURE 27. Variation in GAC treatment performance for oil and grease removal from
>•*- API separator effluent. ;
' ' .
f
143
-------
•cm i
(MNUtM ACTIVATIO CAMON HIOT PLANT'
CMMTU AT NYMAUUC LOAOIN6 Of
I t W I* 10 10 40 50 «0 TO 90 19 W »S ••
fCRCCNT PROEABILITT THAT CONCIHTMTION WILL 1C LCSS THAN STATCO VALUE
FIGURE 28. VarJrtlwn Jn GAC treatment performance for oil and .grease
removal from aerated lagoon effluent.
i
144
-------
til
tc
o
9V
60
70
60
SO
40
SO
20
to
9
e
7
6
5
4
9
2
.
'
•
X
'
X
X
•
x
,/
.a
^
•
X
X
,x'
1
FULL-SCALE
TREATMENT
SYSTEM
CASE 1 >
X
X
x
X
x^
f
f
X
,
X
^
jf
&
s
/
/
X
X*"
^
FULL-S
TREATM
SYSTEM
CASE 2
•
,
X
x
:ALE
ENT
/
ll
•
X
jf
y
•
'
,
2 5 10 15 20 30 40 50 CO 70 6O 85 90 95 96
.PERCENT PROBABILITY THAT CONCENTRATION WILL BE LESS THAN STATED VALUE
FIGURE 29. Effluent oil ond grease from GAC unit processes for the treatment
of refinery wastewaters
145
-------
H- report. The specific constituent nature of the sludges is adequately deicribed else-
: where188'204.
1 >
' i , '
Sludges from refinery operations and waste treatment processes can be nandled in
• numerous way:.. Historically, the method most'often used was lagoon storage of sludges
with ultimate disposal of the combustible material. As land resources diminished, the
" necessity for concentration of sludges prior to disposal become apparent. In general, the
• methods of sludge,concentration vary with the type of sludge. Oily sludges from storage
tanks and oil-water separator bottoms can be concentrated by precoat vacuum filtration or
centrifugation. These methods permit a certain amount of oil recovery from the sludges
and such recovery is warranted considering the quantity of oil likely to be present in
1 gravity separator bottoms. A chcrocterizotion of oil contaminated sludges from both an
API separator and a TPS unit is presented in Tables 21 and 22, respectively. ' These data
1 indicate that a greater amount of oil can be found in API separator bottoms than in TPS
sludges; nevertheless, a significant quantity of oil is present in the sludges from both of
these units.
' Sludges from boiler treatment blowdown and chemical or biological treatment of
( refinery wastewaters also can be thickened and subsequently dewotered by vacuum
filtration or centrifugation. The ultimate disposal of dewatered sludge is normally either
; by incineration, landfill/landfarming, or ocean disposal. Acid and caustic sludges from
refinery processes generally require neutralization before dewatering and ultimate dis-
posal.
: The only available method of disposal for spent treating cloys and crocking/dt>-
' suffurization catalysts is landfill ing due to the potentially hazardous nature of these
1 materials. Some refineries have reported landfarming of spent clays, although this
i method is unacceptable in areas where the soils have a relatively high permeability and
, would run the risk of groundwater contamination. The spent clays and catalysts should be
• considered hazardous solid wastes and should bo disposed of in accordance with proper
hazardous material landfill procedures. Cooling tower sludges can be landfarmed or land-
j filled, providing that neither chroma re nor zinc is present. If these constituents are
• present in the sludge in significant concentrations, this would preclude the use of land-
i farming as a disposal method. Consequently, some cooling tower stodges must be disposed
I of in a hoacrdous material landfill . ' ;
j With the exception of incineration and digestion where waste solids are actually
oestroyed, the majority of existing solids handling processes are oriented toward the con-
centration of sludges. Since the economics of ultimate disposal are predicated on the
volume of sludge handled, the concentration of waste solids can result in definite cost
savings. The mosr c nv.:On sludge dewatering processes are: (1) rotary drum vacuum
filtration; (2) pressure filtration; and (3) centrifugation. ;
Vacuum filters ore the most widely u.«ed type of mechanical sludge dewatering
device for the concentration of wastewater sludges. It is a filtration process in which
^f-sol ids are separated from the liquid phase by means of a porous media which retains the
i i . , i
146
-------
TABLE 21. CHARACTERIZATION OF OIL CONTAMINATED API SEPARATOR BOTTOMS*
API Separator
Swales
(grab)
1
2
3
4
' 5
6
7
8**
Oil fr«t
Dry Solids
Percent Weight
6.6
9.2
2.S
14.5
9.9
15.1
13.1
26.5
Oil Fr«e
Volatile
Dry Solids Oil
Percent Weight Percent Weight
2.1
2.7
1.1
4.1
3.1
4.0
4.9
3.7
16.5
5.7
34.0
4.5
5.8
1.9
3.0
31.7
Ratio of 011
To Oil Free
Dry Solids
2.5
0.6
13.5
0.3
0.6
0.1-
0.6
1.2
Total
Sulfur
Percent
-
0.2
:
0.4
4.6
0.7
0.7
1.2
*£PA Category C Refinery.
**Co»pcs1t« taken over 24 hours.
-------
TABLE 22. CHARACTERIZATION OF OIL AND SLUDGE PHASES FROM TILTED-PLATE SEPARATOR*
Case
1
2
3
4
5
'
Percent
Sediment
Volume
5.0
6.0
11.5
3.0
0.5
011 Pnese
Percent
Oil
Volume ,
' 17
. 82
86 .
85
99
Specific
Gravity
of Oil
-
0.921
0.897
0.905
-
Percent
Sediment
Volume**
11
14
25
9
-
Sludge Phase .,
Percent
Oil TSS VSS
Volume " (nig/1 ) - (i"?/!)
2 25,000 21.000
2 57.000 45.000
5
6 •
-
*Thesc data arc characteristic of a pilot scale tllted-plate separator treating petroleum refinery Mastawatar
**The sludge phase exhibited an average dry solids content of four percent by might.
-------
solids but allows the liquid to pass through. Media employed fr>- this purpose include
nylon, dacron, polyethylene cloth, steel mesh, or tightly-wound steel coils. The
filtration process is accomplished by means of a horizontal drum covered with filter media
which rotates in a tank with approximately one quarter or .more of the drum submerged in
wet sludge. As the drum rotates, a vacuum is applied on the inner side of the filter media
which draws water from the sludge and produces a moist cake of solids on the outer surface
which is removed prior to re-entering the tank. Precoats are often used to speed filtration
rates or collect more of the fine particles in the slurry. The principal sludge properties
which affect vacuum filtration dewatering are: (1) solids concentration; (2) viscosity; (3)
compressibility; and (4) chemical characteristic*.
The filter press, which is the simplest of all pressure filters, is an assembly of
alternate solid plates, the faces of which are grooved to permit drainage. Between these
plates rest hollow.spaces in which the cuke collects during filtration. A filter medium,
usually some sort of fabric, covers both faces of each plate and coke formation occurs on
each face. Fiitar presses can operate at pressures up to 1,000 psio. Water one! oHs are
able to pass through the filter medium while solids are retained on the fabric. Data from
a pilot-scale filter press unit used to dewater oily solid wastes are presented in Table 23.
Filter cake solids concentrations rangtr from a low of 40 percent by weight to a high of 71
percent, depending on the nature of the filtration aid applied during operation. The
sludge properties which affect filter press dewatering are the same as those which impact
upon vacuum filtration process efficiency.
Centrifugation permits the mechanical dewatering of sludges through centrifugal
force. Within a centrifuge, centrifugal force acts on a sludge particle causing it to
settle through the liquid. The variables which affect gravity sedimentation also affect
sedimentation within centrifuges; namely, particle size, density differential, and liquid
viscosity. Sludge solids in suspension are a combination of particles both granular and
fibrous in nature. Extremely fine particles which will not settle under normal gravity con<*
ditions will separate at higher gravitational forces. In general, those sludges which
separate most readily and concentrate to a fair degree during plain sedimentation will de-
water most efficiently by centrifugation. The introduction of flocculation aids, such,as
polymers, has increased the range of materials that can be satisfactorily dewatered by
centrifuges. The degree of solids capture can be regulated over a wide range by adjusting
the amount of chemical coagulant applied during dewatering. Data from a pilot-scale
vertical solid bowl centrifuge unit used to dewater oily solid wastes are presented in Table
24. Bowl coke solids concentrations range from a low of 43 percent by weight to a high
of 62 percent, depending on operating conditions. The sludge dewatering performance
tevell for The centrifugation process were comparable to that observed with the filter press
unit.
i ' i '
.' Londforming of organic, biodegradable petroleum refinery solid wastes is an
environmentally sound solid waste management practice when careful consideration is
glv*n *c -.',',0 selection, solid*, topography, and surrounding lard us«s in order to minimize
adverse impacts that could arise from the operation,. The types of sol's usually'considered
best suited for waste disposal by landfarming are those that cpntain high proportions of
149
-------
TABLE 23. PILOT-SCALE FILTER PRESS UNIT FOR DEWATERING OILY SOLID WASTES*
r«< • c.u " "
•?.£»
•
t
»
4
*
«
»
•
<-
10*
II
AM (ik tiir ••»
o.t
O.t
6.t
O.t
O.I
0.54
O.M
I.M
l.-f
0.17
l.t
MtUlM^ 1
0.11
O.It
-
0.01*
.V
0.07
0.07
0.07
0.10 '
O.Oti
-
t£nr>i.
i»
M
I7t
140
IN
1*0
IM
IM
to
171
IM
CrcU
t
f •
t
t
t
t
1
t
t
t
t
m
12
It
11
H
' tl
It
17
17
11
IS
n
nil
I.S
a.i
U.I
It.S
l.t
•
-
.
-
11.1
W.7
ISS
M
40.
M
41
11
17
M
II
71
U
11
on
I7.»
M.O
11.0
H.I
1.4
».7
10.1
«.t
-
- 7.1
11. t
tllo»1»c «i1»
(ITU/It)
4.700
- '
-
.
l.OM
4.SW
t.na
•
-
I4tO
l.TtO
*lf* CatefOry C K«fti*«ry.
toflntrjr »l1y tludqt *1it*jr« conttttt4 •* et«U"lMtH ATI tep«r«tor ilvdft, iMiM MI4 nonlttdt^ 9»t«1fiit tUrtf* t*r> betteat i*t *I1 fr»t «1«4*jt
IV
-------
TABLE 24. PILOT-SCALE VERTICAL SOLID BOWL CENTRIFUGE UNIT FOR DEWATERING OILY SOLID WASTES
EPA CATEGORY E REFINERY
o>
.Sludge
Feed Operating Percent
Mixture* Conditions** TSS
1
2
3
.4
5
6
7f
8"
- 8.3
lire addition 9.5
emulsion break- 3.0
Ing chemical
10.3
9.4
11.0
unheated feed 8.5
1.0
Feed
Centra te
Percent
Percent TSS
Oil Recovery
5.4
11.9
26.5
15.2
22. D
10.5
-
0.8
98
82
86
81
78
88
96
95
Percent
OH Percent
Recovery TSS
99
97
99
99
99
94
-
98
0.2
1.8
1.3
2.0
2.1
1.4
0.34
0.50
Cake
Percent
TSS
53.8
54.9
56.0
62.4
59.0
60.3
54.4
42.7
.
Percent
Oil
0.3
3.4
2.5
0.9
1.2
4.5
-
0.7
*Ref1nery oily sludge mixture consisted of contaminated API separator sludge, leaded and unleaded gasoline
storage tank bottoms and oil free sludge material.
**Feed heated to 175'F.
+Uncontam1.-.dted API bottoms.
tt
Thickened IAF float solids.
-------
cloy and organic matter to adsorb and filter the applied waste material. Wastes to be
landfa'ned should contain organic constituents that are susceptible to biodegrodation and
not subject to significant leaching while the degradation process proceeds. In general,
refinery oily solid wastes and waste treatment biosolids meet these requirements.
The rote and extent of biodegradation of waste in the soil is strong!/ influenced
by many chemical and physical factors. Some of the principal factors controlling biode-
grodation are: (1) the composition of the waste; (2) contact between waste and soil micro-
organisms; (3) presence of adequate oxygen; (4) soil temperature; (5) soil pH; (6) presence
of available inorganic nutrients; and (7) moisture content of the soil. The presence of
adequate oxygen in the soil is essential to effective biodegradation. Aerobic biodegra-
dation of organic matter is .much more rapid and complete than is anaerobic waste stabili-
zation . Generally, adequate drainage and proper.waste loading will prevent anaerobic
conditions. • , • • '
The prevsnce of available nitrogen and phosphate in the soil is essential to
achieve the maximum biodegradation rote. These nutrients are normally suppled by
common agriculture'! fertilizers. The amounts required ere dependent on available nitrogen
and phosphate in the waste, fertilizer persistence in the soil, rate of waste application,
and waste biodegradation rate.
Data from pilot-scale londfarm plots treating oily solid wastes are presented
graphically in Figure 30. These data, which depict both fertilized and unfertilized con-
ditions, indicate that the removal of oil and grease by landfarming is represented by first-
order kinetics. Namely, that the fraction of oil and grease remaining in rhe soil of the
landfarm site (F) is directly proportional to the time elapsed since the initial application
(t) as follows:
/ F = e"1" . (4) , ,
where k is the first-order rate constant. The rote constant may be obtained from the slope
of the line of best fit as is indicated in Figure 30. The impact of fertilization upon oil
and grease biodegradation rate is apparent from these data.
The application rate of 800-960 barrels of sludge per acre per year is considered
acceptable for landfarming oily sludges from refinery operations and ballast waste treat-
ment. It is recommended that groundwater quality in the landfarm area be monitored
periodically. If deterioration is observed,, the application rate is generally reduced to a
level which precludes further contamination. Due to the potentially high salt content of
ballast oily solid;, however, it is necessary to elutriate these sludges before landfarming.
It is acceptable to mix together the elutriated ballast solids and the refinery sludges in a
single landfarming operation.
Landfill ing also is an acceptable treatment process for final disposal of petroleum
refinery oily process and ballast water sludges. Landfill ing consists of the basic operations
of dumping, spreading, compacting, and covering w.aste solids which are classified either
5 ' 7 152
-------
o
z
IU
at:
o
1.0
0.9
o.e
0.7
0.6
0.5
0.4
0.3
0.2
O.I
\ V
5
1
0
o
k (FERTILIZED) • -0.0137
(UNFERTILIZED) • -0.0096
\
V
OFERTILIZED SOIL
• UNFERTILIZED SOIL
0 20 40 60 80 K>0 120 140 160 180 200
ELAPSED TIME (DAYS)
FIGURE 30. Oil and great* removal as o function of londfarm residence time
t ,
V
• 153
-------
hazardous or nonhazardous. The sludge classification will ultimately determine the pro-
cedures and requirements for safe disposal of the various solid wastes. Usually, landfill
'sites are located in thick, relatively impermeable formations such as massive clay beds.
Soils with a high clay and silt content also are acceptable as landfill sites. Artificial
impermeable liners will be required for most hazardous solid wastes, as well as other pre-
cautionary measures such as landfill monitoring and leachate collection systems.
A two-level leachate collection system is usually installed beneath a pit, pond,
or earthen facility receiving hazardous wastes. The primary level of protection consists of
an impermeable synthetic liner which is, in turn, protected by a layer of permeable
washed sand. A network of perforated pipes is usually placed immediately above the
primary lir,->r to facilitate leachate collection. A secondary barrier of compacted clay
will usually underlie the primary barrier and serve as further protection for underlying
groundwater. The disposal site also will include grouncKvoter—quality monitoring wells,
and the performance of the leachate management system is usually monitored on a regular
basis.
Oily sludges from gravity oil-water separators and tank bottoms (other than,
leaded) may be either londfarmed o* landfilled. Any leaded tank bottom sludges will hov«
to be disposed of in a hazardous material landfill in order to prevent potential groundwatei
contamination from the leaching of lead. Landfarming rather than landfill ing is the
recommended disposal method for all oily sludges including air flotation float material pro1
duced in a refinery wastewater treatment plant. In general, oily sludges other than leadei
tank bottoms have been successfully degraded by the landfarming process ^ . The analysis
of soil from experimental landfarm sites used for the disposal of oily solid wastes is pre-
sented in Table 25. Data from these experimental plots which are operated both with and
without fertilizer addition to the soils indicate that slight increases in the quantity of lead
nitrate, and chromium have occurred over the test period. Major increases in the quantity
of ommonio, potassium, and sulfaie present in these soils also are evident. These
increases are due, in port, to the fertilization process. The storm water runoff and leach-
,ote characteristics from the experimental landfarm sites are presented in Table 26. These
data illustrate the need for proper runoff and leachate collection measures at a landfarm
site in order to minimize any harmful effects to the environment.
, , l '
i Incineration is used to reduce the volume of a solid waste by combustion to an
•ash, which is subsequently landfilled. At present, it is costly in comparison to other land
disposal methods and normally only is feasible where available land is scarce. A primary
consideration in the cost-effectiveness of sludge incineration is the effect of sludge feed
composition on auxiliary fuel requirements. The heat yield of a given sludge is a function
of the relative amounts and elemental composition of the contained combustible elements.
Pretreatment methods such as chemical conditioning and dewatering reiiult in a substantial
'reduction in incineration fuel requirements, but at the expense of creating increased
energy demands on other unit processes. The combustibility of a sludge is dependent on
the volatile content of the sludge. The inert content o'f the feed sludge will reduce its
.heat value and effect the requirements for complete combustion.
154
-------
TABLE 25. ANALYSIS OF SOIL FROM LANDFARM SITES FOR THE DISPOSAL OF OILY SOLID WASTES
Ul
tn
Parameter
PH
Organic Matter, Ib/acre***
Cation Exchange Capacity,
•eq/100 9
Chlorides, ineq/100 g
Sulfate, Ib/acre
Nitrate, Ib/acre
Amonla. "b/acre
Potassium. Ib/acre"
Cadmium, mg/1
Chromium, mg/1
Lead, mg/1
Mercury, mg/1
Nickel, ng/1
Z1nc, »g/l
Control*
7.7
0.25
1.66
2.1
H11
2
5
25
<0.0015
<0.003
<0.015
Nil
<0.01
0.1
Experimental Plot Experimental Plot
With Fertilizer** Without Fert'.11zer*«
6.2
3.75
6.13
2.7
>250
18
25
285
<0.02
3.90
0.9
-
0.14
5.5
7.J
J.6
4.15
2.1
>250
22
7
25
<0.02
3.54
1.28
-
<0.1
4.8
*L1me applied.
•*l1rae applied; total oil and gre»se load - 3.35 1b/ft2.
***The Ib/acre represents seven-Inches of soil over one acre.
-------
--M.J
TA5LE 26. STORM WATER RUNOFF AND LEACHATE CHARACTERISTICS FROM A LANDFARM SITE FOR THE DISPOSAL
OF OILY SOLID WASTES
Parana ter
PH
B005, mg/1
COO. i»9/l
Oil and Grease, ng/1
TOC. mg/1
Chlorides, ng/1
Sulfate. ng/1
Anmonla. ng/1
70S. mg/1
Cadmium, mg/1
Chromium, mg/1
Lead, ng/1
Hercuvy, ng/1
Nickel, mg/1
Zinc," ng/1
Site Runoff
From
Experimental Plot*
5.8
406
1,041
Nil
268
201
Nil
27
1 ,950
Nil
<0.003
Nil
<0.002
Nil
0.38
Forced Leachate
From
Experimental Plot**
6.5
174
1,304
Nil
295
380
4
140
4,950 .
<0.0015
<0.003
<0.0'5
-
<0.1
0.21
Forced Leachate
From Control
7.2
35
40
N11
13
5
4
Nil _
122
-------
'* Incineration is a two-step process involving both drying and combustion. The
drying step essentially involves some form of mechanical dewatering which reduces the
moisture content of the sludge to within a range where the heat required to evaporate the
remaining water within the solid mass nearly balances the available heat from combustion
of dry solids. The practical limit for mechanical dewatering is normally between 50 and
60 percent moisture whitk is not sufficient to sustain the combustion reaction; consequently,
the remaining portion of the drying sequence which occurs in the incinerator, will require
auxiliary fuel addition. Self "sustained combustion is often possible with de watered row
sludges once the burning of auxiliary fuel rcises incinerator temperatures to the ignition
point.
PROCESS PERFORMANCE VARIABILITY
' . • ' ' .'
i • Candidate wastewater treatment, solids handling, and disposal processes are pre-
sented schematically in Figures 31, 32, and 33. Considering these schematic drawings, it
,is apparent that a multitude of sequential process permutations are possible, ranging from
•direct contract disposal to a complex system involving primary, secondary, and tertiary
wastewater treatment, conditioning, dewatering, incineration, and ultimate disposal of
isludges. Selection of the optimum process or system is usually based primarily on economic
.considerations. Moreover, since the economics of disposal are contingent upon such
; dynamic factor* as landfill availability and regulatory criteria, the system must be flexible
'enough for modification. Therefore, the optimum disposal system should provide for a
jleast-cost solution which ultimately allows for the efficient implementation of future
(expansions as dictated by treatment objectives. , '
1 , i •
' Regardless of the type of wastewater or waste solids processing system finally
jimplemenred, a residue I of material will require ultimate disposal to either a landfarm,
• landfill, or some other treatment option. Moreover, Hie implemsntation of a solids :
.handling process requires a decision as to whnther it is necessary to continue further solids
processing or take the option of contract disposal . For instance, aerobic or anaerobic j
digestion presents an alternative process for solid waste handling. Such processes not only!
reduce the volume of degradable solids, but the remaining solids also are less susceptible |
to the production of noxious odors and are rendered more amenable to several of the '.
potential dewatering processes. Since waste activated sludge is the only waste solid <
.category presented in Figure 30 which is subject to biodegrodation by digestion, it would ;
constitute the only feed to a digester. However, landfarming can be used to dispose of |
waste biological solids and oily process sludges which makes it a more favorable sludge i
^disposal alternative. ! I
i !
it is axiomatic that biological, chemical, or physical wastewater treatment pro- ;
cesses produce effluents of varying quality and this inherent variability cannot be signifi-
icontly altered by in-plant control, operational changes, or changes in concept and design.
'Although effluent variability is common to all w* srewater treatment processes, it is most i
'noticeable with biological systems. The variability in effluent quality from a properly
; designed and operated biological treatment system is attributable to the basic nature of
"the treatment process, the raw waste load characteristics, and geographical and
'
?
157
-------
FIGURE 31. Petroleum refinery wastewater treatment plant schematic.
-------
SOURCE
or
SSH. IDS
CtNTMI.
COLLECTION
AND COKOITIOHINC
CONCENTRATION
•tCTCU*
OCVATCRIHO
INCINERATION
OIL
RECOVERY
uiremn
OiSfOSAL
FIGURE 32. Schematic of solids handling and disposal alternatives.
-------
CARBON BACKWASH
s
SUPERNATANT
DAF
SLUDGE
THICKENER
THICKENED
—•SLUDGE TO
INCINERATOR
CARBON TRANSPORT WATER
INCINERATOR SCRUBBER
RETURN
REGENERATION SCRUBBER
RETURN
FIGURE 33, Tertiary carbon adsorption addition to petroleum refinery was*ewater treatment plant.
-------
"•climotological conditions. This is the minimum variability which can be practically
obtained assuming irbper system design, management, and operational control.
A change
-------
TABLE 27. DAILY VARIABILITY FACTORS FOR TREATMENT OF PETROLEUM REFINERY WASTEWATERS
i>
to
•
Nr*wt*r
no
en
IK
TU
til ««*«
MrMlt
•I,-.
bin*
Cr»
Cr*
•• ft.ll. .
AttiwM $l«*tt A*r*t*tf IWMII «ntf A«r **4 VCTCA
r fletttlM
AMItlw
Awigt
S.8
M
M
r.i
i.t
T.t
J.»
HI
M
M
.1
.1
.1
.»
.0
.1
.1
.«
.1
-
W/M«i M4U1CM
lUngt A**r«gt ktngt
4.
4.
M M
-4.
1.
t.
Mf
M
M
M
M
M
7. M
M M • M
M MM
11.4 M .
-
Amrm
M .
M
M
M
M
M . .."-
M
M
M
M
* M . l«f*m*tlw Mt n*M*»1t.
*» tnk la«lutt* •>• w« •» Ml«n.
-------
TABLE 28. MONTHLY VARIABILITY FACTORS FOR TREATMENT OF PETROLEUM REFiNERY WASTEWATERS
Acttxtrt Stirtit A*r*t»4 l*«oa> *ntf *•-••.*< Ltfowi M* VCTU Activated $l«t«t A*r*t«tf Up**
•*UtMn« *«•* Olttolvn: Atr FUUt'w !»«d fUtntiM . M»ttf M4U riltrtttM
W/Ckmlctl Mtltko* - . U/Alw AMItlw
rtiaatu *§»»• A«*r*f* fcinge /)*tng» lUn^e Ave,-»g« Hug* Atrtrigt ktngt A«*r*|*
•0 t.t-4.
COO 1.S-I.
nc
ns :.s->.
•11 > CTMM 1.»-t.
PkMel* l.«-J.
-
Mj-« 1. 1-4.1
telfldt
trT 1 .«->.•
1.1-t.
1.4-J.
M
l.«-t.
t.O-S.
t.?-4.
1.7-f.
.0 -* 3.9
.8 M** HA
U M M
.t - ».»
.1 • I.I
.1 - 1.;
.1 - ?.».
M M W M
M HA HA M
Cr** Nft M M M M M
.
.
.
,
,
.
.
.1
- .
.
M
.
*
.
•
.1 • l.»
.t M M
MM M
.1 - 1.0
.« - s.t
.1 • - m
.1 M
M .M M - M
r M M M M
,4 1.* HA KA
ca
luft 1*tf1«U* OM Mt «f Mlm.
M - tafimitlen Mt •*•«•*)•.
-------
that the effluent from an activated sludge system Is more variable than that from an
oerated lagoon and polishing pond. The lower effluent variability of the aerated lagoon/
polishing pond system is attributable to the long hydraulic contact times required of such
biological treatment units to obtain long-term effluent qualities comparable to those
attainable with the activated sludge process. In general, en aerated kaocn/pollshing
pond system provides low effluent variability, although the control of suspended solids in
these effluents is difficult. .Moreover, aerated lagoons do not offer the same operational
flexibility as activated sludge systems and may not meet long-term effluent quality speci-
fications for BPCTCA conditions.
The calculated variability fccfors for the various pollution abatement systems
indicated in Tobies 27 and 28 ars greater than the EPA variability factors specified in the
Development Document for BPCTCA conditions ' . The addition of a filtration unit
after biological treatment does reduce the monthly variability. However, the magnitudes
of ths variability factors are still greater than those specified in the Development Docu-
ment.
It should be recognized that the distribution of effluent concentrations or mass
loadings is bounded at the lower end by the nonremovable portion of a particular waste-
water constituent. Such a boundary does not exist at the upper end of effluent concen-
trations Of loadings. Consuquently, the effluent variability factor will increase as the
effluent concentration or loading from a treatment system decreases. Moreover, those
variability analyses.w.hich are based solely on concentration information ignore that vari-
ability present in mass discharge which :s due primarily to variations in flow.
An examination of the variability factors for each pollution abatement system
indicates that ths variability may be high for one waste constituent and low for another
in the same treatment process. This is undoubtedly related to the differences in con-
stituent concentration present in different refinery wastewaters which can, in turn, be
related to product mix and refinery complexity. These differences are often manifested in
the form of different effluent qualifies from two similar treatment systems receiving the
same type of wasrewarer. These data illustrate the diffuulty in keeping the daily and
monthly maximum allowable effluent qualities from refinery wastewoter treatment systems
within the allowable venation limits set by EPA, even when long-term effluent quality
constraints are attained.
164
-------
SECTION 6
EVALUATION OF POLLUTANT LEVELS OF PETROLEUM
REFINERY WASTEWATtRS AND RESIDUALS
E. H. Snider
F. S. Manning
' Department of Chemical Engineering
The University o^ Tulsa
Tulsa, Oklahoma
INTRODUCTION
In 1978 the EPA awarded the University of Tulsa contract No. R-805 099 010 to
develop an environmental assessment for petroleum refining: wastewaters and residuals.
One portion of this assessment is the preparation of a state-of-the-art report on pollutant
emission levels. Accordingly this section summarizes a careful search of the recent liter-
ature on emission loads; avenues of emission; and includes estimates of accuracy,
precision, variance and causes of variance whenever possible.
WASTEWATER EMISSIONS
Wastewater emissions are discussed under two classifications: raw waste loads
and treated effluent loads. .
Raw Waste Loads
Raw waste loads are reported for most of the so-called "traditional" pollutants,
and for several priority pollutants, primarily those (such as cyanide, phenol, mercury)
which have been monitored .extensively. Median row waste loadings for the five sub-
cateaories of refineries .we're collected in a 1972 study, conducted jointly by API and
EPA . In this study, the net raw waste loading was defined as the amount of a c,'ven
contaminant added to intake woter per unit volume of crude oil charge to a refinery .
The median concentration values of these pollutants based on the refineries studied are
presented in Table 29 in units of js.t.b. of crude oil processed. Since an average value
of gallons of wastewater produced per barrel of oil processed was also given, a simple
conversion gives the results in mg/JJ, as also reported in Table 29. In addition the
folloving priority pollutants were included and the results are included in Table 29:
cyanides, phenolics, copper, lead, and zinc.
165
-------
TABLE 29. MEDIAN RAW WASTE LOADINGS FOR REFINERY SUBCATEGORIES A
THROUGH E AS REPORTED IN API-ERA SURVEY206
Porometer0
A
Ib
1000 bbl mgA
gol/bbl 18.00
BOO.
COO
TOC
Oil & Greo*e
Suspended Solids
Dissolved Solids
Sulfides
Chromium •
Ammonio
Nitrogen
Orgonic
Nitrogen
Nitrate
Nitrogen
Acidity
Alkalinity
Total .Phosphate
Cyanide
Chloride
Fe
Cu
Pb
Zn
2.92
1X33
2
3
4
103
0
0
0
0
-0
-0
0
0
0
15
0
0
0
0
.49
.13
A1
.01
.43
.49
.03
,00
.34
.04
.01
.96
.53
.02
.00
.75
.06
.00
.00
.00
19.4
88.7
16.6
20.8
Dm
.07
29.5
689
0120
<0.07
2.3
0.27
0.00
0.00
3.5
0.13
<0.07
105
0.40
<0.07
<0.07
«0.07
I
Ib
1000 bbl mgA
, 40. i4
38.29
105.77
17.76
13.74
Jo
.52
11.84
210.66
0.34
0.03
7.79
2.39
0.00
C.CO
12.35
0.08
0.00
65.34
0.22
0.00
0.00
0.07
•
113
313
52.6
40.7
4c
.5
35.1
624
1.00
0.09
23.1
7.10
<0.03
<0.03
36 A
0.24
<0.03
194
0.65
«0.03
<0.03
0.21
Class'''6
C
Ib
1000 bbl mgA
42.65
57.99
168.38
54.26
18.09
4AM
.0
18.62
186.04
0.71
0.19,
16.95
2.29
0.00
0.00
33.81
0.20
0.05
76..T7
0.25
9.00
0.01
0.06
,
163
473
152
50.8
52.3
523
2.00
0.53
47.6
6.40
<0.03
<0.03
95.0
0.56
0.14
214
0.70
<0.03
0.03
0.17
D
Ib
1000 bl
47.34
59.79
184.00
50.91
56.82
39.23
200.. 12
0.02
0.06
8.39
2.85
moA
151
466
129
144
99.3
506
0.05
0.15
21.2
7.20
0.00 <0. 02
-0.26
32.94
0.19
0.03
67.43
0.06
0.00
0.00
0.03
0.00
83.4
0.48
0.08
171
0.15
<0.02
<0.02
0.08
E
Ib
1000 bbl
86.96
129.40
296.59
58.65
50.16
21.73
299.28
1.40
0.31
16.55
4.38
-0.02
-0.56
53.04
0.19
0.00
50.69
0.27
0.01
0.01
0.03
moA
1
178
409
80.8
69.1
3t
.3
29.9
412
,1.93
0.43
22.8
6.0
0.00
0.00
73.1
0.26
<0.01
69.8
0.37
0.01
0.01
0.04
Color
Turbidity
Volatile Suspended
Solids
Cend. etonce
Nitrite Nitrogen
Total KjeloWil
Nitrogen
49
178
15
0.06
6.6
8.2
38.3
83.8
19.8
1712
0.15
99.5
8.0
104
1064
23.8
1312
0.80
448
8.4
534
122
0.10
543
8.1
21.7
41.2
0.17
23.3
7.8
"Units for all uoremeters are mgA wl* Hie following exceptions: color (color units), turbidity (J.T.U.),
conductance (^mho/em), and pH (pH units).
Values reported for color, turbidity, VSS, conductance, nitrite nitrogen, TKN, and pH were obtained
from limited row date reported In the appendices to this reference and do not represent average* based
I on all the refineries which rrode up the A PI-EPA survey.
°A negative value for a parameter in the lb/1000 bbl column represents a case In which the plant
Intake water concentration 'exceeded the row waste load by the Indicated amount, and was reported
a* a *ere concentration In mgA. A zero value In the lb/1000 bbl column was Interpreted m
meaning a loading of lets than 0.01 lb/1000 bbl ond so this limit was used to arrive at the mgA value.
166 '
-------
In 1978 o Department of Energy (DOE)-EPA study determined the variation in
trace organics concentrations across the wastewater treatment system of a typical Class B
refinery . In addition to the trace organics several traditional pollutants1 were moni-
tored in the raw waste stream; these are included as part of Table 30, which olso reports
concentrations at severe! stages of treatment for this waste stream as well as for the intake
water. • ,
The organic analyses in, the DOE-EPA study were performed by GC-MS on
methylene chloride extracts of the wastewater sample*. Three extracts were collected -
a neutral extract, an acid extract, and o base extract. Water samples extracted and
analyzed included tha raw wastewater after DAF treatment, effluent from activated
sludge (AS) treatment, and effluent from activated carbon (AC) columns. The details of
the extraction and analysis procedures are contained in the report of the study ^ . The
results, which indicated that 304 trace organic compounds were detected, are presented
in Tables 31 through 37 for the neutral extracts, Table 38 for the acid extracts, and
Tables 39 through 41 for the base extracts.
Evaluation of the £&$*$}&£ levels of the priority pollutants was the purpose of a
1978 study by EPA and API208'209. Intake waters, wastewater feeds to biotreatment,
and final effluent streams from 17 refineries were sampled and analyzed for the 129 sub-
stances on the EPA Priority Pollutant list. A summary of the data from this stud/ is pre-
sented in Tables 42 through 45. , ,
Treated Effluent Loads ;
Several sources of data on the effluent concentrations of various pollutants after
differing modes of treatment are available.
; In the DOE-EPA itudy which has been described previously, data were presented
for effluents from an ACFC, MMF, and AC columns . These data are presented in ;
Tables 30 through 41. : j
1 ' I
In the 1972 API-EPA survey2**0 data were collected to determine the median '
performance of AS treatment plants for five refineries - three Class R and two Class C.
These data are summarized in Table 46. , '
1 Data for treated effluent loadings of priority pollutants from 17 refineries were
reported as part of the 1978 API-EPA study . Concentration ranges for the priority
pollutants in treatment plant effluents are presented in Tables 42 through 44 described
previously.
In addition, at six of the refineries considered in the above study^"", pilot-
scale GAC results were obtained for several traditional pollutants as well as a number of
priority pollutants., These results were reported in a separate document*™., The inlet
wastewaters to the carbon columns were treated usinj multimedia filtration, and results
were obtained for both virgin and regenerated carbons. The data summarizing the ranges
167
-------
TABLE 30. AVERAGE INTAKE AND WASTEWATER EFFLUENT CONCENTRATIONS
AFTER VARIOUS TREATMENT PROCESSES IN A CLASS B REFINERY207
Porometer Intake
Oil & Grease <10
Cyanide < 0.02
Phenol 0.02
COD <15
BOD <10
TOC 18
TSS 21
DAF°
24
0.25
3.9
150
103
57
38
Concentration,
FCb
<10
0.14
0.02
43
17
24
<10
mpA
MMFC ACd
10 <10
0.15 . < 0.02
0.02 < 0.01
44 <15
17
-------
TABLE .31. CONCENTRATION OF n-ALKANES IN THE NEUTRAL FRACTION OF THE
DAF EFFLUENT FROM A CLASS B REFINERY AND PERCENT REMOVAL BY
THE ACTIVATED SLUDGE AND ACTIVATED CARBON UNITS207
Compound
n-nonone
n-decone
n-vndecone
n-dodecane
n-tridecone
n-tetradecane
n-ponradecane
n-hexadecane
n-heptodecane
n-octodecone
n-nonodecane
n-«icosane
n-heneicosane
n-dorosane
n-tricosane
n-tetracosane
n-pentacosane
n-hexacosone
n-heptocosane
n-octocosane
n-nonocosane
n-frfocotane
n-heneitriocotane
Concentration
In DAF (cob)
32
128
349
544
675
683
651
493
'355
261
205
160
107
64
61
43
32
27
19
13
11
NM
NM
Percent Removal
by Activated
Sludqe0
NM
99 +
99 +
99 +
99 +
99 +
99 +
99 +
99 +
, 99 +
99 +
99 +
99 +
99 +
99 +
99 +
99 +
99 +
99 +
99 +
NM
T
T
i
Percent Removal
by MMF + AC
ND
ND
T
98
96
94
91
91
88
87
88
85
79
80
83 ;
79
70
78
T
*»
T
T ;
1
1
1
"Neutral fraction of finol-cl.orifier effluent.
T Trace concentration present
NM Not measurable due to interferences
ND Concentration not detectable
169
-------
;1 TABLE 32. CONCENTRATION OF CYCLOALKANES AND ALKANES OTHER THAN
: n-ALKANES IN THE NEUTRAL FRACTION OF THE DAF EFFLUENT FROM A
CLASS B REFINERY AND PERCENT REMOVAL BY THE ACTIVATED SLUDGE
AND ACTIVATED CARBON UNITS207
Concentration
Compound in DAF (ppb)
Alkones
C,y,-Aikane 159
C.«-Alkane 77
1 O
C14-Alkane . 196
C.,-Alkane 246
14
Pri stone 157
Phytane 67
i
Cycloolkones
30
; 21
' ' 57
31
i • , 41
I • 29
i .27
71
42
43
Percent Removal
by Activated
Sludge0
99 +
99 +
99 +
99 +
99 +
99+ ,
99 +
99 +
NM
99 +
99 +
99 +
99 +
99 +
99 +
99 +
Percent Removal
ly MMF + AC
T
ND
T
79
78
59
(
ND
ND
ND
ND
ND
NO
ND
ND
ND
ND
Neutral fraction of final-clorifler effluent.
Also includes alkenes; exact qualitative analysis was not made; the 10 entries represent
10 cycloalkane~alkene compounds.
T trace concentration present
ND Concentration not detectable
NM Nor mensurable due to interference*
170
-------
TAELE 33. CONCENTRATION OF ALKYLATED BENZENES !N THE NEUTRAL
FRACTION OF THE DAF EFFLUENT FROM A CLASS 3 REFINERY AND PER-
CENT REMOVAL BY THE ACTIVATED SLUDGE AND ACTIVATED CARBON
UNITS207
Compound
, Toluene*
Ethyl benzene*
p end m-Xylenei
o-Xylene
l-fropyl benzene
n-Propyl benzene
, «- Ethyl toluene
o-Ethyl toluene
1,3,5-Trlmethyl
1,2,4-TrimetTiyl
1,2,3-Trimetfiyl
benzene
n-6utyl benzene
•v-n-Propyl toluene
o-n-Propyl toluene
•r-Diethyl benzene
1 ,3-Dimerhyl-5-
eftiyl benzene
l,3-D!methyl-4-
•thyl benzene
1,2-DlmethyM-
•thyl benzene
1,3-D?methy|-2-
ethyl benzene
1,2-Dimethy|-3-
othyl benzene
1,2,4,5-Tefromethyl
benzene
1,2,3,5-Tetromethyl
1,2,3,4-Tetromethyl
benzene
Concentration
In OAF (ppt)
101
35
187
101
5
13
93
32
43
176
96
8
19
13
13
29
37
43
16
13
27
48
64
Percent Removal
by Activated
99 +
99 +
99 +
99 +
ND
99 +
99 +
99 +
99 +
99 +
99 +
T
99 +
9? +
T
99 +
99 +
99 +
ND
T
99 +
99 +
99 +
rVrcnnt Removal
by MMF + AC
84
67
77
77
NO
t ,
71
T
ND
45
60
ND
,ND
ND
NO
ND
ND
ND
ND
ND
ND
T
31
Neutral fraction of final clerlfler effluent.
T Trace concentration preient
NO Concentration not deferable
•An !.•>. Mortty folluiant
171
rm .'.>>' t
-------
34. CONCENTRATION OF INDAN AND TETRALIN AND RELATED COM-
POUNDS AND THEIR ALKYLATED DERIVATIVES IN THE NEUTRAL
FRACTION OF THE EFFLUENT FROM1 A CLASS B REFINERY AND PERCENT
REMOAL BY THE ACTIVATED SLUDGE AND ACTIVATED CARBON
UNITS''
Compound
Indon •
i
1 -methyl indan ,
2-methyl indon .
Ethyl indon
Dimethyl indon
. Dimethyl indon
Dimethyl indon
Trimethyl Indan
Tetralin
i
;' Methyl tetrolin
Ethyl tetralin
Dimethyl tetralin
Ethyl styrene
Ethyl styrene
C_-Styrene
•3
C,-Styrene
C«-Styrene
j <*
i Cg-Sryrene
Concentration
in DAF (ppb)
93
104
61
27
61
11
35
35
11
64
27
21
19
48
19
72,
21
53
Percent Removal
by Activated
Sludge0 '
99 +
99 +
99 +
T
99 +
ND
99 +
T
ND ,
T,
99 +
T
ND
99 +
T
99 +
i
99 +
99 +
Percent Removal
by MMF + AC
50
. T
ND
ND
T
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND !
ND
ND
ND
°Neutral fraction of final clarifier effluent.
Several dimethyl indon and C,-Styrene compounds were observed - see succeeding
compounds. Reference 2 provides structures.
T Trace concentration present
ND Concentration not detectable
172
-------
-TABLE 35. CONCENTRATION OF NAPHTHALENE AND ALKYLATED NAPHTHALENES
IN THE NEUTRAL FRACTION OF THE DAF EFFLUENT FROM A CLASS B
REFINERY AND PERCENT REMOVAL 3Y THE ACTIVATED SLUDGE AND
ACTIVATED CARBON UNITS207
Concentration
Compound In DAF (ppb)
Naphthalene*
1 -Methyl ' naphthalene
2-Methy! naphthalene
Ethyl naphthalene
Dimethyl naphthalene
Dimethyl naphthalene
Dimethyl naphthalene
Dimethyl naphthalene
Dimethyl naphthalene
Dimethyl naphthalene
C_-Nophthalene
Cg-Naphtholene
C -Naphthalene
C.-Naphthalene
C.-Naphthalene
C.-Nophtholene
Cj-Nophtholene
Cg-Naphthalene
C.-Nophtholene
C,-Nophthalcne
C.-Nophtha!ene
197
448
259
77
192
NM
267
203
96
45
24
21
160
45
37
51
99
125
85
80
93
Percent Removal
by Activated
Slodge0
99 +
99+ ,
99 +
99 +
99 +
—
99 +
99 +
99 +
99 +
99 +
ft
99 +
99 +
99 +
99 +
99*
99 +
99 +
99 +
99 +
99 +
Percent Removal
by MMF+AC
38
45
33
56
38
--
27
1
12
37
T
88
92
T
T, ;
t
70 i
61
45
T
T
Neutral fraction of final clarifier effluent
NM Not measurable due to interferences
T Trace concentration present
I Increase In concentration
*An E. P.A. f rioriJy Pollutant
y 173
-------
TABLE 36. CONCENTRATION OF ALKYLATED BENZOtHIOPHENES AND DiBENZO-
THIOPHENES IN THE NEUTRAL FRACTION OF THE DAF EFFLUENT FROM
A CLASS B REFINERY AND PERCENT, REMOVAL BY THE ACTIVATED
SLUDGE AND ACTIVATED CARBON UNITS^
Percent Removal
Concentration by Activoted Percent Ramovo!
Compound in DAF (pob) Sfudqe° ' by MMF + AC
Methyl benzothiophene
Methyl benzothiophene
Methyl benzothiophene
Methyl benzothiophene
Ethyl benzothiophene
Dimethyl benzo-
• thiophene
D'methyl benzo-
thiophene
Dimethyl benzo-
thiophene
Dimethyl benzo-
thiophene
D? benzothiophene
21
16
,13
32
11
11
16
8
8
13
T
99 +
99 +
99 +
99 +
99 +
99 +
99 +
99 +
T(N)
ND
71
ND
83
ND
ND
ND
ND
ND
' ND
Neutral fraction of final clarifier effluent.
T Trace concentration present
N Noisy, possibly due to column bl«ed
ND Concentration not detectable
174
-------
TABLE 37. PNAs AND ALKYLATED PNAs OTHER THAN NAPHTHALENES IN THE
NEUTRAL FRACTION OF THE OAF EFFLUENT FROM A CLASS B REFINERY
AND PERCENT REMOVAL BY THE ACTIVATED SLUDGE AND ACTIVATED
CARBON UNITS207
Con
Compound ' In t
Anthracene*
Methyl phenonthrene
Methyl phenonthrene
l-Merhyl anthracene
. 2-Methyl onthrocene'
C_-Fh« nonthrene/
Anthracene
Cj-Phenonthrene/
Anthracene
C_-K>enonthrene/
Anthracene
C»-Phenonthrene/
Anthracene
Cj-Phenonthrene/
Anthracene ,
' Cj-Phenonthrene/
Anthracene
Fluor»ne*
I f-i*thyl fluorerto
Methyl fluorene
Methyl fluorene
AC6flOpn*n9AQ
i
; Methyl ocencphthene
1 Methyl ocenophthene
Methyl ocenophthene
•ionenyl
Methyl blphenyl
' Methyl btphenyl
' Fyrene*
C,_ H.. PNA («ueh at
17 '12 •
methyl pyrene)
^^ 4
1,2-leraontt.rocefte
fercent Removal
«entrat!on by Activated Percent &wnowal
JAf (cott) Sludge9 by MMF + AC
168
72
80
27
27
5
5
16
37
40
11
27
29
35
16
3
35
24
16
24
19
11
29
11
5
13
99 +
N
N
99 +
99 +
N
N
N
N
N
N
N
99 +
99 +
99 +
1
N
N
N
T(N)
, . '(N) .
T(N)
?9+
99*
99 +
53
T
T
T
T
NO
ND
ND
i
ND
i
NO
NO
ND
ND
T
NO
ND
ND
ND
NO
T
ND
7o
NO
ND
NO
*Nov*re! fraction of the flnel clo/IRar effluent.
N Nolly, pcmlbly due to unrotolved Interfering orpordo
T Trace concentration prewnt
ND Concentntlan not detectable
I incfeaie1 In concentration
•An I.P.A. fHority
175
-------
TABLE 38. PHENOLS !N THE ACID FRACTION OF THE DAF EFFLUENT FROM A CLASS
B REFINERY AND PERCENT REMOVAL BY THE ACTIVATED SLUDGE AND
ACTIVATED CARBON UNITS207
Ct
Compound In
. ' •«-!--«*
rnenof
Cmol
p-Cretol
Elhyl phenol
Ethyl phenol
Dimethyl phenol*
2,3-Dtmethyl phenol
Dimethyl phenol*
n-Propyl phenol
I-Propyl phenol
l-Propyl phenol
t-Prwl phenol
l-f ropyl phenol
'-Propyl phenol
m.-fmjLju_,t rJt^ortl JL
n^TTwpri pnvnov QI
Methyl ethyl phenol
n-Propyl phenol &
Methyl ethyl phenol
Methyl ethyl phenol
Methyl ethyl phenol
2,4,3-Trlmethyl phenol
Methyl •ffiyl pn*>nol ft
C--Ph«nol
KUtfiyl «fV phenol &
C--Ph»nol
S-- -» ^ iiM_j__ii_nJ
VMI V. . | IWIHM
^* ^^^^ ^^ ^filk^^M^U
\M vm \> * t iwtiui
OJethyl phenol
Methyl pH^ol
ncenrraNon
DAF(aob)
22
33
50
4
7
29
16
8
1
2
4
10
1
1
4
<,
2
3
3
1
<1
1
1
1
*
Percent Removal
by Activated
Siu4ae0
NO
99 +
NO
ND
NO
ND
99 +
NO
NO
ND
NO
NO
ND
ND
ND
ND
NO
ND
99 +
ND
ND
ND
NO
ND
ND
Percent Removal
byMMF+AC
NO
NM
.NO
ND
ND
ND
ND
ND
NO
ND
NO
NO
ND
ND ,
ND
MO
NO
ND
NO
NO
NO
NO
ND
ND
ND
*Acld fraction of nSe final clartfler effluent.
ND Concentration not detectable
NM Not measurable
•An E.P.A. Priority Pollutant
176
-------
TABLE 39. 'ALKYLATED PYRIDINES IN THE BASE FRACTION OF THE DAF EFFLUENT
FROM A
SLUDGE
Compound
Picoline
Ethyl pyridine
4-Ethyl pyridine
Lutidine
Lot! dims
Ethyl picolinn
2-Efhyl picoline
Ethyl picoline
2,4,6-Cbllidine
2,3,6-Collidine
2,3,5-Collidine
Collidine
Collidine
C.-Pyridine
C3-Pyridine
Ethyl lutidine
Ethyl lutidine
Ethyl lutidine
Ethyl lutidine
Ethyl lutidine ,
CLASS B REFINERY AND PERCENT REMOVAL
AND ACTIVATED CARBON UNITS207
Concentration
in DAF (ppb)
<1
<1
<1
<1
2
NM
I
6
2
<1
<1
<1
2
<1
<1
1
<1
1
<1
<1
Percent Removal
by Activated
Sludqe0
- •
ND
93
ND
98
NM
ND
ND
ND
ND
ND
67
ND
ND
ND
ND
ND
ND
ND
NO
BY ACTIVATED
Percent Removal
by MMF + AC
-
ND
ND
T
ND
ND
ND
ND
ND
ND
ND
99
ND
ND
ND
ND
ND
ND
, ND
. ND
BOM fraction of the final clorifier effluent.
NM Not measurable due to interferences
NO Concentration not detectable
T Trace concentration present
177
-------
TABLE 40. ALKYLATED QUINOLINES iN THE BASE FRACTION OF THE DAF
EFFLUENT FROM A CLASS B REFINEUY AND PERCENT REMOVAL BY
ACTIVATED SLUDGE AND ACTIVATED CARBON UNITSZO/
Concentration
Compound in DAF (i>nb)
Quinoline
Methyl quinoline
Methyl quinoline
Methyl quinoline
Methyl quinoline
Methyl quinoline
Methyl quinoline
Ethyl quinoline
Dimethyl quinoline
Dimethyl quinoline
Dimethyl quinoline
Dimethyl quinoline
Dimethyl quinoline
Dimethyl quinoline
C~-Quino!ine
C.-Quinoline
J '
6
4
1 ,
<1
<1
2
1
<1
2
1
2
<1
2
1
2
2
Percent Removal
by Activated Percent Remove!
Sludge0 by MMF + AC
ND
ND
ND
ND
85
ND
ND
ND
94
ND
ND
ND
92
ND
97
97
ND
ND
ND
ND
ND
ND
ND
ND
89
ND
ND
ND
ND
ND
ND
ND
Base fraction of the final clorifier Affluent.
ND Concentration not detectebl*
178
-------
TABLE 41. ALKYLATED ANILINES IN BASE FRACTION OF THE OAF EFFLUENT FROM
A CLASS B REFINERY AND PERCENT REMOVAL BY ACTIVATED SLUDGE
AND ACTIVATED CARSON UNITS207
• >
Percent Removal
Concentration by Activated Percent Removal
Compound in DAF (ogb) Sludge0 by MMF + AC
Aniline 27
o-Toluidine 29
Toluidine 10
N,N-D!methyl
aniline <1
99 +
NM
NM
89
T
ND
ND
ND
Base fraction of the final clarifier effluent.
T Trace concentration present
ND Concentration not detectable
NM Not measurable due to interferences
179
-------
TABLE 42. CONCENTRATION RANGES OF PRIORITY POLLUTANTS - VOLATILE
ORGANICS CATEGORY REPORTED IN API-EPA PRIORITY POLLUTANTS
SURVEY208
JET
M.*rl«~OI«ri4>.
C»W» T.».cMorM.
«*~.
til.l-TridtlOT-
•*•«•
1,1-CicMw
•ArUn.
I,1,J.J-T.»«-
Twn-I.I-OicMvcr
A
i:. run
> JO >»
< J < 5
'
< to 100 <10
>IOO «10
>100 <10
A . *. m
a.*' .w.*1
< U.S-HOO < 0.2- M
< O.V 10 < 0.)- 14
< 0.1-< O.J
< 1 t 10 < 1 -<10
< 0.9-< 10 < 0.3-<1Ci
« 0.7- 1 « 0.7
< 1 - 3 < 1 - 1
J - 4 < 0.4
< 0.7- J < 0.7-< 1
<10 -4700 « 8.5-<10
«10 -4200 < 1 -*10
<10 •»!« < 1 -<40
10 10
« 10 <10
< 10 <10
< to 109 «10
MOD IOO <10
1
»'•' w"-J
O* • w •
<10D-»IOO <10
>IOC <40
10- U <10
< 10 <10
< 10- 107 <10
< 10 <10
< 10 <10
< 10 <10
< 10 <10
< 10 <10
« 10 <10
t»*100 4
*too u
« 10- » <10
B»>' B!.'7
4.f 4
< 0.3 ^.J
< 1 . «1
< 0.4 <0.4
J1CU) «1
44000 <1
< 0.5 ^.S '
• V T
hr *• (!••.
lin* l*«ir l« Ift III ik»
M
180
-------
TABLE 43. CONCENTRATION RANGES OF PRIORITY POLLUTANTS - SEMI-
VOLATiLE EXTRACrABULS CATEGORY REPORTED IN APi-EPA PRIORITY,
POLLUTANTS SURVEY'
*•*>
a..
fWI
-
o.* -
« 1 -
0.7 -
« 1 -
««
«o
07
oj
ia»
7 J
M
1 J
«i
< i
«i
< i
«i
«i
«i
*o.»
ID
O.I - I.I
«t
«o.e>-
O.I-
« I
« IJ
* J O.I - 1.*
« 1
« 1.1
OJ
u
0.1
«»
«o.ot
* -1,
• e •» o» i -»; iMt<(M>w a— • *.
181
-------
TABLE 44. CONCENTRATION RANGES OF PRIORITY POLLUTANTS - TOTAL METALS CATEGORY REPORTED IN
API-EPA PRIORITY POLLUTANTS SURVEY208
Category*
CO
K>
taA)
Afl
At
8.
Cd
Cr
Co
HO
. Nl
rt>
s« -•
Sb
Tl
Zn
A
Bio. Final
Eff.*' Eff.*2
<5 < 5
12 <10
< 3 < 3
< 1 < 1
32 5
17 < 5
< 0.5 < 0.5
23 <15
64 <15
-------
TABLE 45. CONCENTRATIONS OF PESTICIDES AND ASBESTOS CATEGORIES
OF PRIORITY POLLUTANTS OBSERVED IN DATA OBTAINED
DURING API-EPA PRIORITY POLLUTANTS SURVEY209
Pofwn
-------
TABLE 46. MEDIAN PERFORMANCE OF ACTIVATED
PLANTS FOR THREE CLASS B REFINERIES
REFINERIES206
SLUDGE TREATMENT
AND TWO CLASS C
Porofnctar
B005
COO
TOC
Oil end Grvow
PUooli
Suspended Solidt
Dit»oly»d Solids
Sutfidei
Chromium
Ammonia
Organic Nitrogen
Nitrate Nitrogen
Acidity
Alkalinity
Phtopnofe
Cyanides
Chloride
Iron
Copper
leed
Zinc
NJtrlte
T.X.N.
pH
Conductivity
Color
Twfbidlty
Volatile SuspemWd
Solids
Mec'ion Influent
Concentration'
85
213
55.5
29.2
3.4
36
968
2.88
1.43
12.5
3.64
0.13
0
139
. 0.31
0 22
188..'
1.02
0.14
0.1
0.22
O.I
16.56
'1.6
1475
30
27
21
Median Effluent
Concentration
8.5
70
25
U
0.01
25
1170
0.275
0.26
11.25
2.9
0.055
8
100
0.68
0.1
175.5
0.78
0.05
0.11
0.16
0.02
14.58
7.3
1450
' 30
18.5
1S.5
MHion Treatment
Efficiency (% Removal)
89.2
54.6
41.9
60.0
99.6
23.8
-8.8
90.0
58.2
16.2
22.5
43.6
0
55.1
-1.4
30.9
6.6
37.1
37.7
0.0
32.0
27.*
17.2
16.5
•5.9
0
39.2
t.6
*AII vr)fi y* «o/l etcept pH, wfi',ch li r«port«
-------
of pollutants cbterv^J in this study ore presented in Tables 47 through 49.
Furthermore, four of the refineries studied above participated in a pilot-scale
PAC study. The study was conducted using
-------
TABLE 47. CONCENTRATION RANGES OF CLASSICAL PARAMETERS AND
THE PRIORITY POLLUTANT TOTAL METALS OBSERVED IN PILOT-
SCALE GRANULAR ACTIVATED CARBON TREATMENT OF THREE
CLASS B REFINERIES209
Poromrf.rO'D ,
BOO,
sob3
COO
TOC
Total Vii^endvd
Solidi
Ammonia
Nitrogen
0*
Sulfldx
Oil & Greow
pH
T. Cyenlde
Phenol*
Ho
Aff
At
fe
Cd
O (total)
Cv
Nl
Pb
Sb
Se
Tl
Zn
Filter
Influent
< 6 - 27
4 - 23
24 -120
n - 54
,
4 - J6
3.9 - 22
< 0.02 - 0.02
< 0.1 - 0.5
4 - 40'
7.2 - 7.7
0.01 - 0.12
0.012- 0.090
< O.CW
< 0.001-< 0.005
< 0.020
< 0.001-5 0.003
< 0.001- 0.010
< 0.005- 0.060
< 0.005- 0.032
< 0.005- 0.016
< 0.015- 0.020
< 0.025- 0.470
< 0.020- 0.062
< 0.015
0.015-* 0.060
Rlter ,
<6 - 30
4 - 23
20 -160
8 - 70
1 r 34
2.2 - 22
<0.02 - 0.02
<0.1 - 0.4
4 - 24
6.8 - 7.9
0.01 - 0.12
0.013- 0.092
<0.0005
<0.001-< 0.005
<0.020
<0.001-< 0.003
0.050
<0.015
0.020-<0.060
Carbon
Hfluent
2-14
5 , -6
10 -81
3 -70
1 -11
3.4 -22
O.02
<0.1 - 0.2
2 -20
7.2 • 7.6
0.01 - 0.07
<0.005-O.01
<0.0005
<0.001-<0.005
<0.020
<0.001-«0.003
<0.001-<0.002
<0.005- 0.040
<0.005- 0.020
<0.005-<0.015
•^.015-^O.ffW
-l.ion of I.P.A, All onolytM for porom»»«f» from
T. Cyenld* througrt Zn w«r« p«rform*d by rh« R. S. K«rr Envlronmontal R*March Laboratory of
186
-------
TABLE 48. CONCENTRATION RANGES OF CLASSICAL PARAMETERS AND THE
PRIORITY POLLUTANT TOTAL METALS OBSERVED IN PiLOT-SCALE
GRANULAR ACTIVATED CARBON TREATMENT OF TWO CLASS C
REFINERIES209
•orom*t*r '
BO01
SOD3
COO
TOC
Totol Suspended
Solid.
Nitrogen
Cr**
Sulfidet
Oil & Great*
pH
T. Cyonlde
Phenoli
HO
Ag
At
l«
CA
Cr (total)
Cu
Nt
ft
Sb
s*
Ti
Zn
Filter
Influent
10 -<12
6 - 13
75 -150
17 - 55
•
7 - 42
<1 - 3.4
<0.02 - 0.02
0.3 - 0.8
5 - 30
7.5 - 8.4
<0.02 - 0.13
<0.01 - 0.038
-------
TA3LE 49. CONCENTRATION RANGES OF CLASSICAL PARAMETERS AND THE
PRIORITY POLLUTANT TOTAL, METALS OBSERVED IN PILOT-SCALE
GRANULAR ACTIVATED CARBON TREATMENT Of A REFINERY OF
UNKNOWN CLASSIFICATION209
*sror«eter0'b ]
BOO,
*°°3
COD
TOC
Totol A'tpended
Solidi
Aonnorio
Nitrogen
G*
Sulfldei
Oil & Greow
pH
T. Cyonide
Phenols
Hg
: A*
Be
Cd
G (total)
Gi
Nl
Pb
* ,
S*
Yl
Zn
Filter
Infl-jent
<6 -<30
-
96 -170
32 - 58
33 - 66
6.2 - 11
e» for parameter! In tnli lilting from BOD) throu^ pH were perfprmed by loborotorlet
under contract to the Effluent GuloVllnei tXvIilon of E.P.A. All onalytet for poremetert from
T. Cyonlde rhrough Zn were performed by I he R. S. Kerr (tnvlronmentol Reieorch Loboretory of
I.P.A.
188
-------
TABLE 50. CONCENTRATION RANGES OF CLASSICAL PARAMETERS AND THE PRIORITY POLLUTANT TOTAL
METALS OBSERVED IN PILOT-SCALE POWDERED ACTIVATED CARBON TREATMENT OF REFINERY
WASTEWATERS20*
oo
>o
0.0
BOO!
BOOa
COO
TOC
Total Suspended
.Solid*
Amonio
Nitrogen
G/1*
SulRdes
OH & Cream
pH
T. Cyanlda
Phenol*
HO
Ag
As
Be
Cd
Cr (total)
Co
Pb
Sb
T1
Zn
B
- Pilot Pkinl Pilot Plant
Influent Effluent
110 -200
110 -220
340 -480
58 -160
1
22 - 78
3.9 - 10
•< 0.02 - 0.09
0.7 - 11
16 - 81
8.1 - 9.3
0.01 - 0.07
18 -80
< 0.0005
< 0.001-< 0.005
< o.ow
< 0.001-< 0.003
< 0.001- 0.007
0.081- 0.700
< 0.005- 0.014
< 0.005- O.C16
< 0.01 5-* O.f'20
< 0.025- 0.4-.0
< 0.020
< 0.015
0.030- 0.110
5 - 14
<6 - 16
63 -190
20-56 -
30 -140
3.6 -22
•*0.02 - 0.02
0.7 - 2.7
5 - 16
7.i» - 8.2
*tM -< 0.03
<0.01 - 0.99
.«
0.0! - 0.28
5.2 - 6.4
< C.OOU5
0.002-* O.OCJ
< 0.020
0.002-< 0.003
•: 0.001-< 0.002
0.393- 0.500
0.017- 0.020
0.010-< 0.013
< 0.015-* C.020
< 0.025
0.023
< 0.015
0.250- 0.300
L- sir — ,i — — — ' .»•..,•.
8 - 13
-
84 -140
17 - 32
35 -62
10 - 15
<0.020
0.5 - 0.7
9 - 2S
5.6 - 5.8
ire performed by labor^He* under conrroct to the Affluent Guldeltn
OJvltlon of E.P.A. All analyse* for porametf* from T. Cyanide rrrough Zn were pertained by the «. S. Kerr Environmental Research
Laboratory of E.P.A.
'Four refineries comprised thl* part of Ac study - 2 CUm 8. 1 QOM C. and I Undetignotexl Cl«n§.
-------
TABLE 51. CASE HISTORY LONG-TERM EFFLUENT CONCENTRATIONS FOR PUNTS
CONFORMING TO ABATEMtNT LEVELS I AND II210
CM
kA^-i^ » i
v^ff"Wv •
ft •^^•t! 1 Cu*4ab*^d
L^^Vf 1 JrWW
l
_
10
II
12
. 14
IS
13
' 14
Sr - M
-------
TABLE 52. CASE HISTORY ACTIVATED CARBON APPLICATIONS IN PETROLEUM REFINERY
WASTEWATER TREATMENT210
I A*W~M U>*t .BUM* 0.05 to 0.07 V$=^ 4-S5 0 SS II-* 17 JO
^_-J fl|l^
J- « Pilot PW
I ATI!
0.05 H 0.«
SvWflltor
5* • HI* Mmt
t AH!
S- « HI* PlOTt
4 MtSMblyiPM
I AM
0.4* »• o.n
«-. «*'*«•-<
a.!****
W At^ v^pVPMUV c^nVVMV I •• *BT^^^^
5* » Rial FW
Sl.nMUlHM.
ION4
O)h
M MI-WU TiMiMii o*
..»
b. ?•••
W *J
43 0.»-t.4
<.OOt-.OV O.OU
t-n t
t-IS U
ii
-------
TABLE 52 continued
r»i,uinf
C006MAI AM. TCC fc«A> AM. OH* Ofta» t«aA| AM. *«*ehfe»aAi AM
ii nw»n»«i e.i6 ^ff? 4»-i70 io» M u-4t 2* a
DM CnUMHl '
Act. Stufr* DHu»««
OiMMlonN^
U«w Chitlflw
4-1* • CwSrn CoV
II nbtHvt 7t 41 J7 0.0*5 0.01 M
SC^wO
It AH Ti • O*MI*
l«i-*-» CoJ. J-17
r«o«. . t-u - «»
i - r«4*»Ofc. >•!} •
-------
31E53. ANALYTICAL DATA FOR PRIORITY POLLUTANTS FROM ONE REFINERY
211
•ffc~»
CLASSICAL
TOTAL 'CYAN! DES(moA)*
TOTAL PHENOL
TOTAL METALS
Anenlc
Selenium
Codmium
Hen/Ilium
Copper
Antimony
Chromium
Nicfcel
Zinc
Silver
Thallium
Lead
Mercury
OtGANICS (GAS CHROMATOGRAPHY)
PURGEABLES
Methylene chloride
Chloroform
lenxene
1 , 1 , 2, 2-Tefroehloroefhylene
Toluene
ctnyiDensene •
POtYNUCLEAR AROMATICS
fc
ryrenv
linio-o-pyrene ,
Otrytene
FlvorannVene
Phenonfhrene/AntHrocene
Naphthalene
A i inniitiliiin*
Fluorene
PHENOLS
A j fu » IrJiannl
PHTMALATE ESTERS
DtM^tnvl pnfnOiO*V
Dltttiyt ^rttote*
Df*n*wWrjfl pttfrioMfQ
KH2-^r«n.«yl)phtr-Wre
*Mo0fc: Total CyonJdw •xpt^tMd In n^X«
NA - Nor Applicable
Influent
< .05
417
<10
*10
« 1
< 3
45
<10
280
17 ,
390
^10
^10
40
<0.6
<10
<10
320
<10
695
56
<10
<44
^^0
^iO
<10/<10
282
24
<16
14}
>^»
<10
^C|fl
4J(lA
<16
"BT
< .05
3C6
260
400
31
20
5,800
< 10
r" jOO
j,200
i/,000
22
< 10
12.000
< 10
.
NA
NA
NA
NA
NA
NA
< 12
< 106
< 14
< 12
i <14/<16
< 14
« 10
< 39
tn
IV
< 10
< 10
< 10
< 16
Return PAC Flnol
Sludge Effluent Effluent
0*8^) (*&A) (fB/l)
< .05 < .05 < M
67 42 29
150
-------
TABLE 54. AIR-STRIPPER SAMPLES ANALYSIS RESULTS FROM ONt REFINERY211
i
Priority Pollutant
SporgrjdAir, XAD-2
R«in
PAC
Control
POLYNUCLEAR AROMATJCS
Nophfholene
2-Chloronoph thane
Acenaphthalene
Acenophfhene
Fluorene
Phenanthrene/Anthrocene
Fluoronthene
Pyrene
ND
30
ND
220
1 , 2-Benzonthrocene
Chrysene
3,4-Benzopyrene
1,, 2:5, 6-Dibenzonthrocene
PHENOLICS
2-Chlorophenol
2'Nitrophenol
Phenol
2,4-Dimethylphenol
2,4-Dichlorophenol
2,4, 6-Tri chlorophenol
4~Chloro-m"cre$ol
2,4-Dinitrophenol
4, 6-Di ni tro-o-cresol
Pcntochlorophenol
4-Nitrophenol
20
20
<35
ND
<25
<25
<10
<25
<50
<10
PO
<100
<25
<25
<25
<10
<10
<35
ND
<25
<25
<10
<25
<50
200
30
<100
<25
<25
<25
ND - Not Detectable, or leu than detectable limits.
194
-------
TABLE 55. PROPERTIES OF SIMULATED OILY SLUDGES USED IN STUDY OF DISPOSAL
BY SOIL CULTIVATION2'2
Tcnk
Parameter Bottoms
Spec. Grov., 60/60°F
Lbs/Gol
Pour Point, °F
Viscosity, SU, 60°F
,SSF, 122°F
Hydrocarbon Type, %w
Saturate
Resins
Aromatic
Total Sulfur, %w
Total Nitrogen, %w
Total Phosphorus, PPM
Total Ash, PPM 31,000
Calcium 1,880
Magnesium 375
Sodium 3,135
Iron 1,255
Copper 94
L*od Nil
Crude
Oil
0.86
7.12
-5
60
36
8
56
0.47
0.09
6
•
SLUDGE
Bunker C
1.03
8.57
40
19,000
120
18
26
56
1.96
0.41
11
320
10
1
15
20
1
Nil
Waxy
Product
0.85
7.08
95
59
'•
90
-
10
0.04
<0.0005
5
8
Nil
Nil
Nil
Nil
Nil
-
195
-------
samples, chronologies! dcto, penetration depths, hydrocarbon types, and microbial pro~
Tiles are also presented in the EPA report^'^.
1 01 1
In Hie EPA indicatory fate study described previously , samples of return
sludg.es from both the conventional and PAC systems, as well as influenrs and effluents
were analyzed. These data are contained in Table 53.
ANALYTICAL CONSIDERATIONS
Wastewater and wastewater residual samples generated by the petroleum refining
industry are routinely analyzed by standard techniques. The analytical methods are sum-
' *5AA *JA7 vflfl 01 1
marized in several of th« reports considered in this study*"0'*u/'*uo'*'° and will not be
reiterated here.
However, the topics of accuracy, precision, and variance in refinery waste
analysis deserve consideration. Accuracy is a measure of how close analytical results are
to the "true" value, while precision is a measure of the repeatability of the analytical
results. Precision data are usually easy to obtain, while accuracy dote are often difficult
or impossible to obtain. Most accuracy measurements ore based on a "spiked" sample, in
which a known concentration of a parameter, to be determined is added to the wastewater
sample under consideration, and the "recovery" of the spike is determined*'^. A
summary of precision and accuracy data for refinery wastewater analysis i. presented in
Table 56, and further results on metals are shown in Table 57.
There are a number of reasons that analytical variability and errors exist.
Several contributing factors cr* as follows:
1. Repeatability of the t«st method itself208.
2. Interferences from contamination of samples208, either in Held or laboratory.
3. Presence of interfering compounds as part of the wastewotor208.
4. Variability which inevitably exists among laboratories using the same test
methods2"8.
5. Use of different analyticc'procedures208.
6. Interlaboratory analysis of parallel, not true duplicate, samples .
7. Inability to perfectly preserve samples for transport from fiold to laboratory
and time delays due to transport^*.
Most of the traditional parameters, such as COD, TQC, oil and grease, and
TSS, are anolyzoble to a reasonably high degree of accuracy, since they routinely exist
at levels significantly above the lower detection limit in wastowoters. Biological
196
-------
TABLE 56. ANALYTICAL VARIABILI1Y OF WASTEWATER PARAMETERS
Pester
BOO
COO
TOC
Oil & Create - Hemne
- Frecm
-Freon with Infrared
onolysis
-Freon by seporotory
runnel
-Freon by Soxhlet
, extraction
Phenol
*fcenolics
-By extraction
-By direct photometric
onolysis
Total Suspended Solids
Svlfldes
Chromium
,
•
Ammonia Nitrogen
-By dittil lotion
'
.
-By specific Ion electrode
,
-By automated enolytlt
T.K.N.
^V|rFO*v ^flTPOQvO k
Nitrite Nitrogen
Test Concentration
or Bonge (mp/l.)
175
12. J
137.3
270
11.2
11.1
13.86
13.0
12.3
0.120
0.330
0.330
O.OC85
0.029
0.004 -O.C09
0.014 -0.029
0.084 -O.:330
5.5637
0.0090
0.0483
0.0935
4.7
43.2
97.0
18.6
*
0.353 (0.370)°
0.380 (0.407)
O.OT-2 (0.074)
0.064 (0.093)
0.0102 (0.0074)
0.016(0.015)
8.6
0.21
0.26
1.71
1.92
1.00
0.77
0.19
0.13
0.43 -1.41
Standard
Deviation
26
4.15
'3.3
17.8
.
3.5
1.7
1.4
0.9
1.1
1
•
0.6.VM
0.00099
0.0031
0.0042
O.'B
0.43
1.58
4.3
0.105
0.128
0.029
0.035
0.0078
0.009
0.7
0.122
0.07
, 0.24
0.28
0.04
0.17
0.007
0.003
0.005
Spike
Recovery
95.'
101
98
42
45
42
68
55
68-92
55-93-
25-89
85
• .
95.6
,
.
100
101
101
Reference
Eng-Sci,lne. (1975)
•
Myers, «tal. (1976)
Eng-Scl, Inc. (1975)
Myers, etol. (1972)
Myers, etol. (1976)
. "
,
Eng-Sci>e. 0975)
•
•
"
Myers, et el. (1972)
A.P.I. (1978)
1
,
i
Myers, et al. (1976)
Eng-Scl, Inc. (1975)
•
*
•
"
"
Myers, et al. (1976)
Myers, et ol. (1972)
Eng-Sci, Inc. (1975)
Myers, etol. (1976)
Er-j-Sel, Inc. (1975)
1
,
,
Myers, et el. (1972)
• '
•
*TnM concentration In
tynnSetle Mmple
(conftnu-sd)
197
-------
TABLE 56 continued
Test Concentration Standard
flaronxUf or Rong* (rr^/L) D«viot!on
(moA)
CjroniiM
0.06 0.005
0.13 0.007
0.28 0.031
0.62 0.094
Cadmium 0.070 (0.071) 0.02)
0.074 (0.078) 0.018
0.0168(0.014) 0.011
• . 0.0183(0.018) 0.010
0.0033(0.0014) 0.005
0.0029 (0.0028) 0.003
PMV. • 0.005 -0.007
, 0.018 -0.033
0.054 -0.220
Phifeolote btort 0.006 -0.025
0.140 -0.610 ,
rW-*ol«n« 0.074
0.210
0.210
0.0046
0.018
Fluo»«n« 0.054
0.160
0.160
O.OC51
0.019
Ph«nonthr«n*/onthroc«n« 0.078
0.220
0.220
0.005
0.020
Fxrafw 0.033
0.095
0.095
0.0073
0.023
B-n-butyl phtholot* .0.140
0.380 .
0.380
•II P-Erirylhcxy!) pfctholotc 0.220
0.610
0.610
2-Otl
1
'
46-70
67-84
19-112
4-7
41-420
26
81
100
46
67
94
106
110
49
84
113
104
91
70
75
73
79
116
48
70
300
25
55
88
89
77
90
67
41 .
92
n
6.5
4
IUhr*nc«
*y«n, «t ol. (1972)
•ng-Sci, Inc. (1975)
.P.I. (1978)
1 ,
(
,
'
'
Ttw ipllut MRipovnd wot prKoonftif»n«.
198
-------
TABLE 57. ANALYTICAL VARIABILITY OF METAL CONCENTRATIONS IN REFINERY WASTEWATERS
208
Refinery *1
Concentration, ng/g
Element
Zinc
Chromium
Copper
leod
oerylllum
Antimony
TViliium
Nickel
Artenlc
Selenium
Silver
Cadmium
Mercury
Initial
55
110
130
6.6
0.2
5.3
4.8
3.4
22
11
1.3
1 I
0.5
Spike
too
40
33
5.0
1.7
10
50
10
50
5
1.7
2.0.
0.05
Found
250
150
iTO
21
1.9
20
32
13
M
15
5.8
3.3
0.5
Percent
Recovery
161
100
104
181
100
131
58
V
117
94
193
106
91
Refinery
»2
Concentration, "fl/g
Initial
82
46
16
5.3
0.7
68
9.5
3.6
U
15
0.8
0.8
0.4
Spike
500
SO
100
5.0
1.0
100
50
10
50
SO
2.0
2.0
0.05
found
720
140
130
18
3.5
220
75
14
62
30
3.0
3.0
0.5
Refinery '3
Percent
Recovery
124
146
112
175
130
131
126
103
94
46
107
> 107
111,
Concentration
Initial
27
94
52
3
< .1
36
< 2
2.1
14
74
4.2
1.1
O.C
Spike
250
.100
100
100
1 -
SO
100
25
100
100
5
10
5
."9/8
Found
280
190
130
63
.7
36
48
21
120
130
9
12.6
2
Percent
Recovery
101
98
86
61
70
42
48
77
105
75
98
114
34
-------
analyses such as BOD are limited in accuracy by the variability of microorganism be-
havior.
i '
Among the priority pollutants analytical accuracy and precision problems arise
from several causes. Many priority pollutant metals are present at quite low concen-
trations, often very near or below the lower quantifiable level, and often below the limit
of detection. As a general rule, the nearer to the limi* of detection on analytical result
is, the larger the percent error in its determination. Hence many of the metals analyses
presented may be reported accurately only by indicating a concentration less than a
lower quantifiable level.
Organic priority pollutant analyses are subject to the some errors described
•above, and other's due to the method of analysis. The preferred analytical technique for
these tests is analysis by a GC coupled to o MS. in addition to problems of contami-
nation, low-level analysis inaccuracies, and differing procedures of analysis^~, several
problems due to the nature of the analytical method arise as summarized below :
1. Obtaining standard samples of the priority pollutant compounds is often a
problem.
i
2. GC-MS do to are only semiquantitative at best.
3. MS and data system quadrupole peak errors exist.
4. Low concentration samples can be "lost" in the chroma tog rophy column.
5. Efficiency of extraction of the organics from the aqueous phase to an organic
so'vent phase is variable.
6. Variation* in injection volume and technique occur.
7. On many systems there is a loss during the purge step.
8. Inaccuracies arise in dilution and concentration steps.
Despite these limitations, GC-MS analysis provides the most sensitive and most accurate
method widely available for determining organic priority pollutants'. '
Another type of variability arises in the analysis of similar but not identical or
duplicate wastewaters. This is the variability which exist) in wastewater samples taken
from one location of different times, defined by some authors os Inherent variability .
Several factors influence inherent variability, including the following:
1. Row waste loading variations; >
2. Temperofure variations; •
200
-------
3. Dissolved solids variations;
4. pH variations; .
5. Nutrient level variations;
6. Dissolved oxygen level variations; and .
7. Toxic and/or inhibitory species variations.
Furthermore/ obtaining a "representative" sample rroy involve variabilities. For
wastewaten a representative sample may be obtained by continuously withdrawing a small
fraction of a wast* stream and compositing. A less accurate representative sampling
technique, but one often used, is the collection of "grab" scmples over specified time
Intervals and compositing. For sampling solid wastes such as sludges, obtaining a repre-
sentative sample is quite difficult, and special sampling procedures are often called for
in these cases*'*. • , •
A detailed discussion of analytical variability of five wastewater parameters in
the refining industry is provided in on EPA repot? published in 1976^3.
ADDITIONAL INFORMATION AVAILABILITY,
Since the purpose of this section ts to summarize the available data on refinery
wastewater pollutant loadings, not all the available data are appropriate for inclusion
herein, in tfi.'r jection a brief description of other available information will be provided.
First, tn this section little has been said about the volumes of wastewaters emitted
by refinery operations. Details of the volumes produced by various operations ore pre-
sented in reference 209.
Second, data ore available concerning the waste loading to a tree, went facility
cjus«d by once-through cooling water, and a comparison is mode to these Derations <
without once-through cooling water. These data ore tabulated in reference ^06. Also,
data on row waste loadings by desalter brines are presented in reference 210.
Long term expected effluent loadings for a number of treatment technologies ore
discussed (n reference 210. Multimedia environmental goals " for olr, water, and solids -
ore considered from the standpoint of health effects and ecology effects in referenco 215.
These goals are outlined for over 650 pollutants.
Information on the frequency with which the priority pollutants are us«d as raw
materials, intermediate materials, or ore the products of manufacturing processes is pre-
sented In reference 209.
201
-------
The amount* of solid wastes generated by refineries are reported in reference
210. The expected levels of a number of metals in sludges from refineries and other
manufacturing and processing operations are given in reference 214.
Further information on analytical variability is also available. Data compar-
isons of intralaboratory and interloboratory results are shown in reference 213 for five
wastewater param-sturj, onr1 variability in PNA analysis is discussed in reference 208.
Compounds which caused contamination of blanks are listed also in reference 208.
202
-------
SECTION 9
RECOMMENDED RESEARCH NEEDS
INTRODUCTION AND PURPOSE
The amount and composition of wdstewaters and sludges generated by petroleum
refineries are highly variable in nature and are dependent upon a complex matrix of
factors which include216:
o The types of refineries and their product slates.
o Composition of the feedstock <..rude oils which, in rum, depends upon the geo-
graphic sources of the crude oils.
o The processes and technologies used in the refineries.
o The geographic location of the refineries ana' the prevailing climate patterns
at those locations.
i
o The type cf cooling water systems used in ths refineries, and whether or not
the refineries must handle tanker ballast waters.
o The degree of air-cooling and waste water reuse utilized within the refineries.
o The physical age of the refineries end the degree of good housekeeping
practiced within the refineries.
There is a large body of literature and data concerning the characteristics and variability
of refinery wastewaters and sludges*1®'* . There is also a large body of literature, data
case histories concerning the existing control technologies for the treatment of refinery
wastewaters and iludges* °. The purpose of this section is to delineate and to discuss key
areas in which the existing body of literature end data needs to be improved by more
research and development so as 13 provide:
o More comprehensive, more valid and better correlated data on the amount,
composition and variability,of refinary wastewaters and sludges.
o More information related to the long-term toxicity effects of refinery waste*
waters,upon fish. , .
o Better demonstrated and documented data on the design parameters,
economics and performance capabilities of existing and emerging control
technologies for the treatment of refinery wastewaters and sludges.
o More and better guidance as to the technological and economic feasibility of
methods proposed for the ultimate disposal and/or reuse of treated wastewoters,
concentrated pollutant brines and residual sludges.
203
-------
Rather than dealing v/ith generalized areas or categories of research, this section focuses
upon a number of specific research needs and discusses each specific need on an
.individual basis.
IDENTIFYING THE SPECIFIC REFINERY WASTEWATER POLLUTANTS
WHICH ARE TOXIC TO FISH ON LONG-TERM EXPOSURE
, As discussed by Burks*'", short-term (96-hour) fish bipossay tests of the wast>s-
waters from a number,of refineries clearly indicate that the effluents from properly oper-
ated biotrearers consistently had a mean annual TLm of greater than 100 percent, which
means that more Iron half of, the test fish could survive in the undiluted effluents. On the
other hand, wastewaters from refineries without any biotreating had a mean annual TL,p
of 19-21 percent, which means that more than half of the test fish could not survive unless
the effluents were diluted about 4-to~l. In other words, biotreating of refinery waste~
waters consistently removed or degraded those pollutants which are lethal to fish on a
short-term (96-hour) exposure.
, ' 01 o ' '
However, other tests have shown that biolreated refinery wosrewoters retain
some pollutants which are lethal to f*r.h on a long-term (16-32 day) exposure. There may
also be some long-term, sub-lethal effects such as the impairment of spawning activities,
or the impairment of the ability to escape predators or to pursue food.
I* is not known if the long-term toxiciry of biotreated refinery wastewaters is due
to low residual levels of the short-term lethal pollutants (which are largely removed by
biotreatment) or is due to a completely different group of pollutants. It should be of con-
siderable usefulness to both industry and to regulatory agencies to undertake research
studies directed toward: '
o ldent:fying the specific pollutants retained in biotreated refinery wastewaters
which are toxic to fish on a long-term exposure basis.
o Determining the concentration levels at which the specific effluents exhibit
long-term toxicity.
o Determining the interaction effects (if any) in mixtures of the specific
pollutants, which either increase or decrease the long-term toxiciry of the
individual pollutants.
At. a lower level of priprity, it is recommended that research also be undertaken
to develop techniques for the more rapid determination of long-term toxiciry effects. In
other words, fhe development of a technique which could correlate the short-term
responses of aquatic organisms with long-term toxicity effects (either lethal Or sub-lethal)
would be of great value as a research tool as well as a practical effluent monitoring tool.
One such biological monitoring technique, based on fish venlilotory responses (breathing
behavior), i:. already under development for use as an on-line, continuous and automated
22a ,
204
-------
VARIABILITY OF TREATED EFFLUENT QUALITY
presents o comprehensive discussion of rhe problems associated with the
variability of treated effluent quality resulting from process performance variability. The
variability in process performance of wastewater treating is dependent upon o multitude of
factors, including: . '
o The inherent variability of biological treatment processes attributable to the
basic nature of the process, the influent waste load characteristics, the geo~
graphic and climatic conditions prevailing at the plant site, and the sensi-
tivity of biological treatment to seasonal changes ir. ambient temperatures.
Q Changes in raw wastewater flow and composition due to changes in the
refinery's feedstock crude oil.
o Changes in raw wastewater flow and composition resulting from events such as
maintenance turnarounds, short-term production of specialty products, '
seasonal changes in product slates, seasonal changes in rainfall, etc.
To some extent, the impacts of the above variability factors may be minimized by
installing appropriate surge and equalization ponds, but their impacts cannot be com-
pletely mitigated.
Snider and Manning*'' summarize the many factors involved in rhe variability of
analytical test methods for determining pollutant levels in wastewaters. Bombarger^'
discusses the specific problems and inadequacies encountered ir> the analytical determin-
ation of pollutants in refinery sour wastewaters. Burks*" also reviews some inadequacies
in the methods for analyzing organic pollutants in refinery waste waters.
Both process performance variability and analytical test method variability con-
tribute to the overall problem of treated effluent quality variability. It is recommended
that research be undertaken to: (1) determine how much analytical test method vari-
ability contributes to overall effluent quality variability, (2) determine which test methods
are the least reliable, and (3) develop new, more reliable test methods if possible.
ACTIVATED SLUDGE BIOTREATMENT
There is a most extensive literature base on the theory and design of AC bio-
treavment of wastewaters . The majority of that literature concerns the theoretical aspects
of the biochemical reaction process, such es: reaction kinetics, cellular energy and
material balances, sludge yields and sludge characteristics, volatile and non-volatile
suspended solids levels in the mixed liquor, food-to-microorgonisms ratios, methods of
supplying oxygen to the reaction, etc. Such studies have contributed a great deal to our
understanding of the theoretical mechanisms of AS biotreorment, but have done little to
improve the basic process performance. In fact, Schoumberg and Marsh have pointed out
in a JWPCF editorial222:
205
-------
o The number of technical studies concerning activated sludge biotreatment
published in the JWPCF over the past 65 years has risen from about 0-25 per
year (during the-period 1910-1965} to.about 145 per year in 1975.
o Yet, the BOD removal eFficiency of the activated sludge process has remained
constant over the past 65 years (at a level of about 88-96 percent).
It would appear that all of the years of research directed at understanding the process
have not improved the process performance beyond that which had been achieved in the '
early 1900's. It would also appear thot there may be'merit in re-directing AS research
towards correlating key design pyrometers with process performance instead of further
studying of theoretical mechanisms. For example, instead of studying reaction models
based upon assuming completely mixed reaction systems, research should be directed
toward defining how much mixer horsepower is needed to achieve complete mixing in real-
world'biotreatment systems. As another example, instead of continuing to study reaction
kinetics in glassware systems, there should be more benefit to be derived from correlating
reaction basin retention time with process performance in real-world, full-scale bio-
treat ers operating at.various parametric levels of mixing horsepower input.
There can be little doubt that improving the performance capabilities of the
AS biotreotment process would be most desirable. It is recommended that studies be under-
taken to correlate key design parameters, such as mixer horsepower and recction basin
retention time, with actual process performance in the full-scale AS units currently
operating in refineries. ' .
GRANULAR ACTIVATED CARBON AS END-OF-PIPE TECHNOLOGY
GAC systems follcwing a biotrecter at an end~of-pipe technology is one of the
options considered by the EPA as the regulatory basis for their latest BAT effluent limi-
tations for petroleum refineries^^. The EPA's latest Development Document**-* con-
cerning refinery effluent limitations presents pilot plant test data on the performance of
GAC in treating refinery wastewoters. However, the fact remains that only two t/iajor
refineries hove ever installed full-scale GAC systems and operated them on a long-term
basis. One of those two systems was trended only for use in treating impounded storm
water for a small part of the year. The other system was used as the main tree-mien t pro-
cess, in that there was no biotreater ahead of the AC system. Tho latter system was
eventually shut down because of poor performance.
' i '
If GAC systems are to be considered as viable option's in arriving at BAT or any
other regulatory effluent limitations for refineries, then it would appear that there is a
real need for the long-term testing of such systems on a demonstration -scale basis. It is
recommended that research grants be used to fur.d the demonjt'ration of GAC units in
petroleum refineries, in sizes capable of treating at least 200-400 gpm of biotreated >
wastewnters and including the facilities for the necessary carbon regeneration.
206
-------
ENHANCEMENT OF ACTIVATED SLUDGE BIOTREATMENT BY PAC
The direct addition of PAC has been shown to be a very cost-effective method of
enhancing the performance of AS biotrearment of refinery wostewaters*"' * , if it is
necessary fo achieve an effluent quality better than that which can be achieved by con-
ventional AS processes. Oil industry studies by Sun Oil Company, Amoco and Exxon have
provided an extensive amount of data on the PAC enhancement of AS biotreaters. How-
ever, the development of 011 economic method of recovering and regenerating the spent
PAC would be a major improvement to the process in terms of cost-effectiveness. Research
directed toward that end should be worthy of consideration.
EFFLUENT TREATING VIA CHEMICAL OXIDATION
i .
A number of small refineries utilize chemical oxidation of their wastewater
effluents by the addition of hydrogen peroxide or chlorine. Ozonation may also be
effective for oxidation. Very little data is available in the literature regarding such
applications. A research stud)* cimed at collecting, correlating and publishing actual
case history data on the chenical oxidation of refinery wastewafars should be most useful. .
LANDTARMING OF WASTEWATER SLUDGES
Lanctfarming of organic, biodugradable sludges derived from refinery wastewater
treatment is an environmentally sound practice when careful consideration is given to site
location, soils, hydrology and surrounding land uses in order to minimize the adverse
impacts tfmt might arise*™. Londf arming application rotes eppeor to range from 100-120
barrels of sludge per acre per application with about eight applications per year
.(allowing four winter months per year when applications are not advisable). Thus, it is
possible to landfarm about 800-900 barrels of sludge per acre per year. A good many
refineries have tested and practiced landfarming for disposal of oily
There are two areas of research concerning the landfarming of oily sludges which
merit consideration:
) . The relationship between oily sludge vapor pressure and problems of odor
control and air emissions, if any, and the methods for mitigating such
problems (by sub-surface injection, for example).
2. The problems associated with landfarming vis-a-vis RCRA regulations and the
methods of resolving any such problems.
ULTIMATE DISPOSAL OF CONCENTRATED POLLUTANT BRINES
AND FEASIBILITY OF METHODS PROPOSED FOR "ZERO DISCHARGE"
Technologies such as reverse osmosis and vapor compression evaporation do exist
which may be technically feasible for the concentration and desalination of treated
wosfewaters from refineries, although such technologies have not yet been practiced on a
207 .
-------
large scale in tfie refining indu.jtry. Such technologies g.re very costly and energy-
intensive, and their en-^product is essentially a concentrated pollutant brine. The ulti-
mate disposal of such brines in an environmentally sound and secure method which pre-
vents their dissolution by subsequent rainfall poses a very difficult problem. Transport of
pollutant brines to some disposal site merely transfers the problem without providing an
ultimate resolution of the problem.
Injection of concentrated brines, or residual wastewaters, into underground dis-
posal wells may be environmentally acceptable in those locations where the sub-surface
geology and hydrology are deemed suitable for such disposal.
Evaporation-percolation ponds are listed in the EPA's Development Document1-"
in support of their NSPS requirement of 'no discharge' from refineries into Hie waters of
the United States.' Such ponds ma/ provide a short-term method for the ultimate disposal
of residual wastewaters in those locations where the soil'i permeability and the sub-
surface r.ydrology are suitable', or where the ponds are provided with liners or barriers to
prevent wostewater pollutants from permeating to sub-surface aquifers. However, in the
long-term, those ponds will eventually fill with accumulated, concentrated poMutants
which must be removed and disposed of in a manner that secures them from dissolution in
subsequent rainfall.
Research studies are needed to define the long-term problems of 'no discharge1
or 'zero discharge'. The ultimate disposal of wastewater pollutants in evaporation-
percolation ponds and in disposal wells does, in fact, const!rute 'no discharge* into '
waters of the United States over the short-term. However, there remain the problems of
protecting underground aquifers from long-term contamination and the problems of long-
term disposal of the concentrated pollutants accumulated in the evaporation-percolation
ponds. Studies of these problems sSould include a realistic assessment of the magnitude
of improvements in the quality of the nation's waters to be expected from the refinery
NSPS 'no discharge* requirement, and whether that requirement can be justified on a
cost-benefit basis relative to control of other sources such as non-point
206
-------
REFERENCES
SECTIONS
i
1. Beychok, M. R., "Aqueous Wastes From Petroleum and Petrochemical Plants",
John Wiley & Sons, Ltd., London, 1967.
2. Engineering-Science, Inc., "Petroleum Refining Industry, Technology and
Costs of Wostewater Control", The National Commission on Water Quality,
June 1975. ,
3. Cantrell, A., "Annual Refining Survey", Oil c.->d Gas,Journal, March 26,
1979.
4. Hatch, L. F. and Motor, $., "From Hydrocarbons to Petrochemicals, Ports 1
through 15", Hydrocarbon Processing, May 1977 through March 1979.
5. Beychok, M. R., "State-of-the-Art in Sour Water Stripping", Open Forum on
Management of Petroleum Refinery Wastewater, E PA-A PI-NPRA-University of
Tulso, Tulsa, June 1977.
6. "Development Document for Effluent Limitations Guidelines and New Source
Performance Standards for the Petroleum Refining Point Source Category", U.S.
EPA, EPA-440/1-74-014-c, April 1974.
7. Beychok, M. R., "Process and Environmental Technology for Producting SNG
and liquid Fuels", EPA-660/2-75-011, May 1975.
SECTION 4 ,
8. Ackman, R. G. and Noble, D. 1973. Steam distillation: A simple technique
for recovery of petroleum hydrocarbons from tainted fish. J. Fish. Board.
Con., 30(5)711-714.
9. Alabaster, J. S. 1970. Testing the Toxicity of Effluents to Fish. Chemistry
and Industry, pp. 759-764.
10. American Petroleum Institute. 1978, Analysis of refinery wastewaters for the
EPA priority pollutants. API Publ. No. 4296. Washington, D. C.
209
-------
11. Anderson, BenH. 1943. The text city thresholds of various substances found
in industrial wastes as determined by the use of Daphnia magna. Sewage
Works Journal, 16:1156-1165.
12. Atlas, R. M. and Bartha, R. 1973. Fate and effects of polluting petroleum .
in rSe norine environment. In: Residue Reviews, Vol. 49:49-85. Gunther/
F. A. aid Gunther, I.D., W«. Springer-Verlog, New York.
13. Baird, R. B., etal. 1974. Fate of phenolics in wastewoter. Determination
by direct-injection GLC (Gas-Liquid Chromatography) and Warburg respiro-
metry. Arch. Environ. Contam. Toxicol., 2(2)165-178.
14. Baker, R. A. 1964. Chromatogrcphic evaluation of activated carbon. J.
, Amer. Water Wks. Assoc., 56:92-98.
15. Batt»sl'e Memorial Institute. 1975. Ea-'iy warning systems for toxic substance!
Proceedings of a seminar. NTIS Accession No. PB-244-412.
16. Baumgardner, R. K. 1966. Oxygen balance in u stream receiving domestic
and oil refinery effluents. Ph.D. Thesis, Okla. State Univ., Stillwater, Ok
la. 70 p.
17. Beck, M. B. 1975. Dynamic modelling and control applications in water
quality maintenance. Water Research, 10:575-595.
18. Bellar, T. A. and Lichtenberg, J. J. 1974. Amei. Water Wks. Assoc., 66:
739.
19. Bestougeff, M. A. 1967. Petroleum hydrocarbons. Chap. 3, pp. 77-108 In:
Fundamental Aspects of Petroleum Geochemistry. Nagy, B. and Colombo, U,
(eds.). Elsevier Publ. Co., New York. 388 p.
20. B ire hard, E. C. 1978. Development of a rapid fish toxicity test utilizing a
freeze concentration technique for routine petroleum refinery wastewater
monitoring. Prog. Water Technol., 9:749-759.
21. Booth, R. L., English, J. N. and McDermott, G. M. 1965. Evaluation of
sampling conditions in,the carbon adsorption method. J. Amer. Water Wks.
Assoc., 57:215.
22. Brady, S. O. 1968. Tastes and odor components in refinery effluents. Re-
print of Paper Presented at API Div. of Refining Meeting, Philadelphia, Pa.
23. Braus, H., Middleton, F. M. and Walton, G. 1951. Organic chemical
compounds in raw and filtered surface waters. Anai Chern., 23:1160-1164.
210
-------
24. Brown, E. R., et ol. 197:5,. Tumon in fish caught in polluted waters: possible
explanations. Comparative Leukemia Research, Univ. Tokyo Press, Tokyo,
Basel, 1975.
25. Brown, Richc.-d. 1971. The use of biological analysis as indicators or water
quality. Annual Ed. Conference, Portland, Ore. June 1971.
26. Brown, V. M. 1968. The calculation of the acute toxicity of mixtures of
poisons to Rainbow Trout. Water Res., 2:723-733.
27. Browny,V. M., Jordan, D.H.M. and Tiller> B. H. 1967. The effect of
temperature on the acute toxicity of phenol to Rainbow Trout in hard water.
Water Res., 1:587-594..
.28. Brown, V. M., Jordan, D.H.M. and Tiller, B. H.;<1969. The acute toxicity
to Rainbow Trout of fluctuating concentrations and mixtures of ammonia,
phenol, and zinc. J. Fish. BioL, 1:1-9.
29. Brown, V. M. and Dalton, R. A. 1970. The acute lethal toxicity to Rainbow
Trout of mixtures of copper, phenol, zinc, and nickel. Jour, of Fish. BioL,
2:211-216. ,
30. > Burks, S. L. 1977. Biological evaluation of dual media filtration and acti-
vated carbon treatment of oil refinery waste water. Technical completion
report >o USDI, Office of Water Resources Technology. NTIS Accession No.
PB-272-281.
31. Burks, S. L. and Wilhm, J. L. 1.977. Bioassays with a natural assemblage of
bantnic mocr bin vertebrates., In: Aquatic Toxicology and Hazard Evaluation,
ASTM STP 634, pp 127-136. Mayor, F. L. and Homelink, J. L. (eds.).
American Society for Testhj and Materials, Philadelphia, Pa.
32. Burlfngome, A. L., Cox, R. E. and Derrick, P. J. 1974. Mass spectrometry,
Anal. Chem., 46:248R-287R.
1 !-•
33- 3ur!inaoro«, A. L., •t>oj. 1976. The characterization of trace organic con-
stituents in petroleum refinery woitewater by capillary gas chromatography/
real-time high-resolution mass spectrometry: A preliminary report. Identif. '
Anal. Org. Pollut. Water, pp 587-624.
34., Burlingame, A. L. 1977. Assessment of the trace organic molecular com-
position of industrial and municipal wastewoter effluents by capillary gas
chromatography/real-time high-resolution mass spectrometry: A preliminary
report. Ecotoxicology and Environmental Safety, 1:111-150.
211
-------
JO. , Burlingome, A. L., ct al. 1V/6. -Ihe motecular nature and extreme complex-
ify of trace organic constituents in Southern California municipal wcitewater
effluents. Identif. Anal. Org. Pollut. Water, pp 557-585. .
36. Coriander, K. D.. 1953. Handbook of Freshwater Biology with th« First Sup-
plement. Brown, Dubuque, Iowa.
37. Carlson, R. M. and Long,! D. W. 1978. Chlorinated organics as o factor in
reduced biological oxygen demand (BOD). J. Environ. Sci. Health, Fart A;
A13(2)177-186.
1 i
38. Cochran, L. G. and Bess, F. S. 1966. Waste monitoring by gas chromato*
grcphy. J. Wat. Pollut. Control .Fed., 38:2002-2008.
39. Cole, H. A. 1971. Objectives of biological pollution studies. Proc. Roy.
Soc., Series B, London, 177:277-278.
40. , Conger, K. A., et, aK 1978. Shell growth of unfed oysters in the laboratory:
A *ubletbal bioassay~7/stem for pollutants. Uf3 Sci., 22:245-254.
41. Cooper, J. M. and Wilhm, J. L. 1975. Spatial and temporal variation in
productivity, species d'varsity, and pigment diversity of periphyton in an
stream receiving domestic and oil refinery effluents. The Southwestern Nat-
uralist, 19:413-428.
42. Cooper, R. L. and Wheotstone, K. C. 1973. The determination of phenols
in aqueous effluents. Water Res., 7:1375-1384.
43. Copeland, B. J., Ward, J. T. and Fruh, E. G. 1969,, Pert 3: Phytoplank-
ton toxic!ty. In: Final report to Texas Water Quality Board, Feb. 1970.
Ecological Studies of Galveston Bay, pp 179-237.
44. Copeland, B. J. and Dorris, T. C. 1964. Community metabolism in ecosys-
tems receiving oil refinery effluents. Limnol. and Oceanogr. 9:431-447.
45. Cote1, R. P. 1976. Trie effects of petroleum refinery liquid wastes on aqua-
tic life, with special emphasis on the Canadian environment. Pubt. by
National Research Council of Canada. NRCC No. 15021. ,
ii
46, Cowell, E. B. (editor) 1971. The Ecological Effects of Oil Pollution on
Littoral Communities. Applied Science Publishers Ltd, Institute of Petrolf urn,
London.
47. Cox, G. V. 1974. Industrial waste effluent monitoring. American Lab,
July: 36-40. ,
212
-------
48. Grossman, J. S. and Cairns, Jr. 1974. A comparative study between two diff-
erent artificial substrate samplers and regular sampling techniques. Hydrobio-
logio, Vol. 44, No. 4, 1974.
49. Cummins, Kenneth W. (editor) 1969. The stream ecosystem. An American
Association for the Advancement of Science Symposium. Technicol report No.
- 7, Institute of Wa»er Research, Michigan State University.
50. Curtis, E. J. C. and Harrington, C. W. 1970. Effects of organic wastes on
riven. Process Biochemistry, April: 1-4.
51. Dawson, G. W., Jennings, A. L., Drozdowslci, D., and Rider, E. 1977. The
acute toxicity of 47 industrial chemicals to fresh and salt water fishes. J.
Hazard. Mater., 1:303-318.
** '
52. Dicks, B. 1976. The effects of refinery effluents: the case history r»f a salt-
march, pp 227-245 in,Marine Ecology and Oil Pollution, edited by J. M.
Baker. John Wiley & Sons, New York, NY.
53. Dorris, T. C., Burks, S.I., Waller, G. R. 1974. Effects of Residual Toxins
in Oil Refinery Effluents o/s Aquatic Organisms. Technical completion report
to USDI, Office of Water Resources Research. NTIS OK-B023.
54. Dorris, T. C., Patterson, Don, and Cope land, B. J. 1962. Oil refinery
effluent treatment in ponds. Water Pollut. Control Fed., Toronto, Can., Oct.
11, pp 1-14.
55. Dorrfs, T. C., Gould, W. and Jenkins, C. R. 1959. Toxicity bioassay of
oil refinery effluents in Oklahoma.. Trons. 2nd Seminar on Biological Problems
in Water Pollution. R. A. Taft San. Engr. Center., Cincinnati, Ohio, Tech.
Report W60-3:276-285.
56. Doudoroff, P., et al. 1951. Bioassay methods for the evaluation of acute tox-
icity of industrial wastes to fish. Sew. & IrH. Wastes, 23:1380-1397.
1 .• ,
57. Doudoroff, P., Luloc, G., and Schneider, C. R. 1966. Acute toxicity to
fish of solutions containing complex meta! cyanides, in relation to concentra-
' . tions of molecular hydrocynic acid. Trans. Amer. Fish. Soc., 95:6-22.
58. Docdoroff, P. 1976. Toxicity to fish of cyanides and related compounds.' Env-
iromentol Protection Agency, EPA-600/3-76-038.
i • ' '
59. Durkin, P. R. 1978. Biological impact of various chlorinated phenolics and
related compounds on Dophnio mogno. Proc. of Tappi -Environ. Conf., 165-169.
213
-------
60. Duce, R. A., and Duursma, E. K. 1977. Inputs of organic matter to the
ocean. Marine Chemistry, 5:319-339.
61. E'oerhardt, L. L. 1975. Methodology for appraising contaminants in aquatic
systems. J. Fish. Res. Bd. Can., 32:1852-1859.
62. Ellis, P. S. 1975. The medical significance ot marine'pollution by organic
• chemi'cols. Proc. Roy. Soc. Land., (B), 189:443-458.
63. Ettinger, M. B. 1965. Developments in detection of trace organic contamir
nants. J. A.W.W.A., 57:453-457.
64. Faust, S. D. 1977. Chemical mechanisms affecting the fate of organic pollu-
tants in natural aquatic environments. Adv. Environ. Sci. Technol., 8:317-
.365.
65. Farrington, J. W. and Quinn, J. G. 1973. Petroleum hydrocarbons and fatty
acids in wastewater effluents. J. Wat. Pollut. Contr. Fed., 45:704-712.
66. Gardner, G. R. 1975. Chemically induced lesions in estuarine or marine
teleosrs. op 657-693 in The Pathology of Fishes. The Univ. of Wiic. Press,
Modison, Wise., edited by W. E. Ribelin and G. Migaki.
' i i '
67. Gorrett, J. T. 1957. Toxicity considerations In pollution control. Ind.
Wastes, 2:17-19.
68. Gtoyno, E. F. and Ford, D. L. 1970. The characteristics and pollufional
problems associated with petrochemical waste*, summary report. Prepared by
1 Engineering-Science, Inc. for the Federal Water Pollution Control Administra-
tion, USDI. Program No. 12020.
69. Gould, W. R. and Dorris, T. C. 1961. Toxicity changes of stored oil refinery
effluent. J.Wat. Pollut. Control Fed., 1107-1111.
70. Graham, R. J. and Dorrii, T. C. 196(3. Long-term toxicity bioassay of oil '
refinery effluents. Water Res., 2:643-663. . , .
, i
71. Grob, K. and Grob, K.., Jr. 1974. Isothermal analysis on capillary columns
without stream splitting. The role of tKs solvent. ' Jour, of Chrom., 94:53-64.
72. Hall, L. Ir., Buikema, A. L. and Cairns, J. Jr. 1978. The effects of a
Simulated refinery effluent on the gross shrimp, Palbemonetes pugio. Arch.
Environ. Contam, & Toxico'., 7:23-25.
73. Home I ink, J, L. and Space, A. 1977. r-ish and Chemicals: the process of
accumulation. Ann. Rev. Pharmacol. Toxicol., 17:167-177.
214
-------
^"7^i Howkes, H. A. 1968. Some effects of industrial effluents on the biology of
rivers. Water Pollution Control. H. M. Stationery Office, London, Engl.
75. Heftman, E. 1975. Chrornotography. 3rd edition. Van Nostrorsi-Reinhold.
New York, N.Y.
76. Herbes, S. E. 1976. Organic contaminants in aqueous coal conversion efflu-
ents: environmental consequences and research priorities. In Trace Substances
' in Environmental Health, edited by D. D. Hemp. Univ. of Missouri, Columbia
Mo., X:295-3C3.
77. Herbert, D.W.M. 1962. The toxicity to rainbow trout of spent still liquors fro
the distillation of coal. Ann. Appl. Biol., 50:755-777.
78. Hildebrond, S. G. 1976. The potential toxicity and bioaccumulation in
aquatic systems of trace elements present in aqueous coal conversion effluents.
pp 305-313 in Symposium on Trace Substances in Environmental Health-X.
"** edited by D. D. Hemphill. Univ. of Missouri, Columbia, Mo.
79. Hfndin, E., May, D. S. and Dunstan, G. H. 1965. Determining volatile
neutral compounds in sewage by gas chromatography. Wat. & Sew. Wks., 112-
346.
60. Hites, R. A. 1973.. Analysis of trace organic compounds in New England
Rivers. Jour, of Chroma>ogr. Sci., 11:570-574.
81. Hites, R. A. 1973. Phthalates in the Charles and the Merrimock Rivers. Env.
Health Perspectives, January.
82. Hites, R. A. and Biemann, K. 197,2. Water pollution: organic compounds In
1 the,Charles River, Boston. Paper presented before Division of Water, Air and
Wast* Chemistry, American Chemicnl Society, New York, August.
83. Hook, R. D. 1962. Recovery and identification of organ!cs in water. Inter-
; not. Jour. Air and Water Pollut., 6:521.
» i ' , ,
84. , Howgate, P., et al. 1977. Petroleum tainting in fish, Rapp. R.-V. Reun.,
i Cons. Int. ExploVTMer., 171:143-146.
f
85. Imhoff, K. R. 1974. Water quantity and quality management in the Ruhr
Valley. Jour. Water Pollut. Control Fed., 46:1663-1673.
86. Ingols, R. S. and Gaffney, P. E. 1965., Biological studies of holophenols.
i Proe. S. Water Resour. Pollut. Control Conf., 14:175-181.
215
-------
87. Irwin, VV. H. 1964. Fifty-seven species cf fish in oil-refinery waste bioassay.
Thirtieth N. Amer. Wildlife Conf., 89-99.
88. Junk, G. A. and Stanley, S. E. 1975. Organ' .n drinking water. Parti.
Listing of identified chemicals. IS-3671, NTIS, P. E. Box 1553, Springfield,
VA.
89. Junk, G. A. et al. 1974. Use of a macroreticulcr resin in the analysis of
water for organic contaminants. Jour. Crtromatogr., 99:745.
i
90. Junk. G. A., etal. 1976. Appiications of resins for extracting organic
components from water. Fresenius1 Z. Anal. Ch«m., 282:331-337.
91. Kaiser, Rudolph 1963. Gas Phase Chrcmatography, Vol. 1. Butterworth,
Wash., D. C.
92. Keith, L. H. 1976. Identification and Analysis of Organic Pollutants in
Water, Ann Arbor Science Publ., Inc. Ann Arbor, Ml
93. Kleinholz, C. W. 1978. Evaluation of the Pollution Potential of Petroleum
Refinery Wastewaters using Fathead Minnow Bioassays. M. S. Thesis, Okla.
State Univ., Stillwoter, OK.
94. Kloth, T. C. and Wohlschlog, D. E. 1972. Size-related metabolic responses
of the Pinfish, Logodon rhomboides, to salinity variations and sublethal petro-
chemical pollution. Contributions in Marine Science 16:125-137. Univ.
Texas, Austin, TX
95. Kopfler, F. C., etal. 1977. Human exposure to water pollutants. Adv.
Environ. Sei. TecTTnol. 8(Fate Pollut. Air/Water Environ., pt. 2):419-433.
96. K.'ishnoiwami, S. K. and Kupchanko, E. E. 19o9. Relationship between
ocfor of petroleum refinery wastewoter and occurrence of "oily" taste-flavor in
rainbow trout, Solmo goirdneri. Jour. Water Pollut. Control Fed.,'41 :R189-
R196.
97. Lammerhy, M. W. and Burbonk, N. C. 1961. The toxicity of phenol, o-
chltxophenol, and o-nitrophenol to bluegill sunfish. Proc. 15th Ind. Waste
Conf., Purdue Univ., 45 (2) 541-555.
98. Learner, M. A., Williams, R., Harcup, M., and Hughes, B. D. 1971. A
survey of the macro-fauna of the River Cynoi, a polluted tributary of the,
River Toff (South Wales). Freshwater Biology, 1:339-367.
* T 216
in.I
-------
''99j Lee, G. Fred 1973. Review Paper. Chemical aspects of bioassay techniques
• for establishing water quality criteria. Water Res., 7:1525-1546.
.100. Lichtenberg, J. J. 1975. Methods for the determination of specific organic
; pollutants in water and wastewater. IEEE Transactions on Nuclear Science.
Vol. NS-22(2):874-891.
•101. Little, A. D., Inc. 1967. Isolation and identification of persistent and proK-
' lem odors from refinery waste water. Annual progress reports to CDRW Sub-
committee on Chemical Wastes, American Petroleum Institute, Wash., D.C.
,102. Lively, L., Rosen, A. A. and Mashni, C. I. 1965. Identification of petro-
leum products in water. Purdue Univ. Engr. Bull. Ext. Ser., 118:657-663.
103. Ludzack, F. L., Ingram, W. M. and Ettinger, M. B. 1957. Characteristics ,
; of a stream composed of oil refincy and activated sludge effluents. Sew. &
Indust. Wastes, 29:1177-1189.
104. Ludzack, F. J. and Kinkead, Diana 1956. Persistence of oily wastes in
polluted water under a
-------
'112. Matsumcto, G., et al. 1977. Gas chroma tographic-mass spectometric iden-
• tification of phenols and aromatic acids in river waters. Water Res., 11:693-
698.
113. Matthews, J. E. 1976. Acute toxic.effects of petroleum refinery wastewahsrs
on redear sunfish. Environmental Protection Agency. Report No. EPA-600/2—
•-... 76-241.
i <
114. Matthews. J. E. and Myers. L. H. 1973. Static bioassay tests of petroleum
refinery wastewaters using redear sunfish (Lepomis microlophus). Presented at
Workshop on Pathological Effects of ChemicaU on Aquatic Organisms, Pensa-
cola, Fla., May 8-9, 1973.
i
115. Mottsor, V. R., Arthur, J. W. and Walbridge, C. T. 1976. Acute toxicity
: of selected organic compounds to fathead minnors. Environmental Protection
' Agency, Report No. EPA-600/3-76-097.
116. McCauley, Rita 1966. The biological effects of oil pollution in a river.
Limnol. &Oceanogr., 11:475-486.
117. McLafferty, F. W. 1973. Interpretation of Mass Spectra. 2nd edition. The
B«njcmin/Cummings Publ. Co., Inc. Reading, Muss.
118. Melpolder, F. W., Warfield, C. W.; and Headington, C. E. 1953. Mass
spectrometer determination of volatile contaminants in water. Anal. Chem.,
: 25:1453-1456.
119. Meyerson, Seymour 1965. Organic ions in the gas phase. XV. Decomposi-
tion of n-alkanes under electron impact. Jour, of Chem. Physics, 42:2181-
! 2186.
. i
120. Meyerson, Seymour 1955. Correlations1 of alkylbenzene structures with mass
, spectra. Presented before the Division of Petroleum Chemistry, 127th Meeting
of the American Chemical Society, April, 1955.
' i '
121. Middleton,,F. M. and Lichtenberg, J. J. I960; Measurements of organic
contaminants in the nation's livers. Irtdust. & Engr. Chem. 52:99A-102A.
122. Middleton, F. M. and Rosen, A. A. }956. Organic contaminants affecting
the quality of water. Public Health Reports 71(11)1125-1133.
'i i • , i
123. Minor, P. S. 1974. Organic chemical industry's wastewaters. Environ. Sci.
Technol., 8:620-625. . • ,
..218
-------
>24. Minter, K. W. 1964. Standing crop and community structure of plankton in
oil refinery effluent holding ponds. Ph.D. thesis, Oklahoma State Univ.,
Stillwoter, OK.
125. Mohler, E. K. and Jacobs, I. N. 1957. Determination of phenolic-type
compounds in water and industrial waste waters - a comparison of analytical
methods. Anal. Chorr,., 29:1369.
126. Moore, S. F. and Dwyer, R. L. 1974. Effects of oil on marine organisms: a
critical assessment of published data. Water Res., 8:819-827.
127. Mozley, S. C. 1975. Bent hie community responses to energy related effluents
in the Great Lakes. Proc. Fed. Conf. Great Lakes, 2nd:438-447. Publ. by
Great Lakes Basin Comm., Ann Arbor, Ml.
128. Myers, L: H. 1976. Analytical variability of five wastewater parameters.
Petroleum refining industry. Environmental Protection Agency, Report No.
EPA600/2-76-234.
129. National Academy of Sciences, National Research Council 1978. The Tro-
pospheric Transport of Pollutants and Other Substances to the Oceans. Publ.
by; Notional Academy of Sciences, Washington, D. C.
130. , Neff, W. A. 1974. Petroleum Refinery Effluent Regulations and Guidelines.
Environment Cancda, Environmental Protection Service, Report EPA NWP-74-1.
131. Ogata, Masana 1977. Transfer to fish of petroleum paraffins and organic sul-
fur compounds. Water Res., 11:333-338.
132. Ogata, M. end Ogura, T. 1976. Petroleum components and objectionable
malodorous substances in fish flesh polluted by boiler fuel oil. Water Res.,
10:407412.
133. Parke, D. V. 1968. The Biochemistry of Foreign Compounds. Vol.5, in
International Series of Monographs in Pure and Applied Biology. Pergamon
Press, New York, NY
134. Porker, B. L., Brammer, J. D., Whalon, M. E. and Berry, W; O. 1976.
Chronic oil contamination and aquatic organisms with emphasis.on Dipfera:
status and bibliography. Water Resour. Bull., 12:291-305.
135. Parker, P. L., Winters, K., Van Boalen, C., Batterton, J. C. and Scalan,
R. 5. 1977. Petroleum pollution: chemical characteristics and biological
effects, pp 256-269 in Sources, Effects and Sinks of Hydrocarbons in Aquatic
Environments. Publ. by American Institute biological Sciences, Arlington, Va.
219
-------
t36. Pessah, E., Loch, J. L. and MacLeod, J.C, 1973, Preliminary report on
the acute toxicity of Canadian petroleum refinery effluents to fish. Fisheries
Res. Bd. of Canada, Technical Report No. 408.
137. Pettyjohn, W. A. 1971. Water pollution by oil-field brines and related in-
dustrial wastes ih Ohio. Ohio J. Sci., 71:257-269.
138. Pickering, Q. and Henderson, C. 1966. Acute toxicity of seme important
petrochemicals to fish. J. Water Pollut. Control Fed.. 38:1419-1428.
139. Pipes, W, O. i975. Biodegrodation of taste and odor compounds, pp 651-
657 in Proc. Int. Biodegrodation Symp. edited by Sharpley, J. M. and Kaplan,
A. M. Applied Science Pub!., Barking, Eng.
140. Po-Yung, Lu and Metcalf, R. L. 1975.' Environmental fa»e and biodegrada-'
bility of benzene derivatives as studied in a model1 aquatic ecosystem. Environ,
Health Perspectives, 10:269-284. ,
141 „ Polisois, G., Tessier, A., Campbell, P.G-C. and Villeneuve, J. P. 1975.
Dcgradat.'on of pheriolic compounds downstream from a petroleum refinery
complex. J. Fish. Res. Bd. of Canada, 32:2125-2131.
142. Prata, L., etal. 1971. Assessment of surface water quality by a single index
of pollution. Water Res., 5:741-751.
143. Price, K. S., Waggy, G. T. and Conway, R. A. 1974. Brine shrimp bioassay
and seawater BOD of petrochemicals. J. Water Pollut. Control Fed., 46:63-
77.
i
144. Raphaelian, L. A. and Harrison, W. 1973. Trace organics variation across
th« wasrewater treatment system of a class-B refinery and estimate of removal
of refractory organics by add-on mixed-media filtration and grcnular activated
carbon at pilot scale. Interagency Energy/Environment R & D Program Report.
Dept. of Energy and Environmental Protection Agency. DOE Report No. ANL
/WR-78-2. EPA Report No. EPA-600/7-78-125.
145. Ratcliffe, D. A. 1974. Ecological effects of mineral exploitation in the
United Kingdom and their significance to nature conservation. Proc. Roy.
Sbc. London, Series A, 339(1618)355-372.
146. Reynolds, J. H., MSddlebrboks, E. J., PorceUa, D. B. and Grenney, W. J.
1975. Effects of temperature on oil refinery waste toxicity. J. Water Pollut.
Control Fed., 47:2674-2693.
220
-------
J47. Richardson, D., Dorris, T. C., Burks, S. L., Browne, R. H., Higgins, M. L.,
and Leach, F. R. SV77. Evaluation of a cell culture assay for determination
of oil-refinery effluents. Bull. Environ. Contam. Toxicol., 18:683-690.
i ' > .
148. Rosen, A. A. and Middleton, F. M. 1955. Identification of petroleum refin-
ery wastes in surface waters. And. Chem., 27:790-794.
149. Rosen, A. A., Skeel, R. T. and Ettinger, M. B. 1963. Relationship of river
water odor to specific organic contaminants. J. Water Pollut. Control Fed.,
35:777-782.
i , i
150. Rossini, F. D., Mair, B. J. and Streiff, A. J. 1952. Hydrocarbons from
Petroleum. Rheinhold, New York, NY
151. Ruchhoff, C. C., Middleton, F. M., Braus, H. and Rosen, A. A. 1954.
Taste and odor producing components in refinery gravity oil separator effluents.
Indust. & Engr. Chem., 46:284.
152. Shackerford, W. M. and Keith, L. Hi 1976. Frequency of organic compounds
identified in water. Environmental Protection Agency, EPA Report No. 600/4-
76-062
153. Shriner, R. L. and Fuson, R. C. 1948. The Systematic Identification of Or-
ganic Compounds. 3rd edition. John Wiley & Sons, New York, NY
154. Sichels, T. W. and Stafford, D. J. 1977. Environmental sampling method:
rapid identification by GC/MS/DS and spectral search. Amer. Lab.,, May:17-
24.
155. Smith, J. W. and Grigoropoulos, S. G. 1968. Toxic effects of odorous trace
organics. J. Amer. Water Wks. Assoc., Aug.:969-979.
156. Southworth, G. R., Beauchamp, J. J. and Schmieder, P. K. 1978. Bioacc-
umulotion potential and acute toxicity of synthetic fuels effluents in freshwater
biota: azaarenes. Env. Sci. Techno!., 12:1062-1066.
157. Sprague, J. B., El son, P. H. and Sounders, R. L. 1965. Sub lethal copper-
zinc pollution in a salmon river - a field and laboratory study. Int. J. Air
& Water Pollut., 9:531-543.
158. Sprogue, J. B. 1969. Avoidance reactions of Solmonid fish to representative
pollutants. Advances in Water Pollution Research. Proc. of 4'rh Intern. Conf.,
Vol 1:169-179. Publ. by Pergamon Press, Oxford, Eng.
159. Sprogue, J. B. 1969. Review paper: Measurement of pollutant tox'city to fish.
I. Bioassay methods for acute toxicity. Water Res., 3:793-821.
* ? 221
-------
160. Sprague, J.B. 1969. Review paper: Measurement of pollutant toxicity to fish.
II. Utilizing and applying bioassay results. Water Res., 4:3-32.
161. Sprague, J. B. 1971. Review paper: Measurement of f.ollutant toxicity to fish.
III. Sublethal effect? and ''safe" concentrations. Witer Res., 5:245-266.
162. Sprague, J. B. 1976. Current $tavi/s of sublethal tests of pollutants on aquatic
organisms. J. Fish. Res. B,d. Canada, 33:1988-1992.
163. Sprague, J. B., Rowe, D. W., Westlake, G. F., Heming, T. A. and Brown, K
T. 1978. Subletha! effects of treated liquid effluent from a petroleurr. refinery
on freshwater organisms. 297 pp. Publ. by Petroleum Assoc. for the Conserv
of the Canadian Environment, Ottawa, Ont., Canada.
164. Sproul, O. J. and Ryckman, D. W. 1963. Significant physiological characteris-
tics of organic pollutants. J. Water Pollut. Control Fed., 35:1136-1145.
T65. Statrkim, C. N. 1976. Bioconcentration of xenobiotics in trout bile: a proposed
monitoring aid for some waterborne chemicals. Science, 193: 680-681.
166. Sugar, J. W. and Conway, R. A. 1966. Development of GLC techniques for
wastewater analysis. Int. J. Air & Water Pollut., 19:591.
167. Symons, J. M., Bellar, T, A., CdrsweH, J. K., DeMarco, J,, Kropp, K. L.,
Robeck, G. G., Seegar, L?. R., Slocum, C,. J., Smith, B. L, and Stevens, A.
A. 1975. National organic* reconnaissance survey for halogenated organics.
J. Arrie;-. Water Wks. Assoc., 67:634-647.
168. Taken.ura,, N., Akiyama T., and Nakajuma, C. 1965, A survey of the pollution
of the Sumida River, especially on the aromatic aminei in the water. Int. J. Air
& Water Pollut., 9:665.
169. Tanacredi, J. T. 1977. Petroleum hydrocarbons from effluents: detection in
marine environment. J. Water Pollut. Control Fed., 49:216-226.
170. Tertipis, G. G. and Pyburn, C. M. 1974. Water quality studies in petroleum
refineries. Water Pollut. Res., Canada, 9:320-327.
171. Thibodeaux, L. J. and Millican, J. D. 1977. Quantity and relative desorption
, rates of air-itrippable organics in industrial waste water. Environ. Set. Technol.,
11:879-883.
172. Tsonopoulos, C./ Co>;ison, D. M. and Inmon, L. B. 1976. lonization constants
of water pollutant-,. J. Chem. Eng. Data, 21:190-193.
222
-------
T73. Tubb, R. A. and Dorrii, T. C. 1965. Herbivorous Ir.j«:t populations in oil re-
finary effluent holding pond series. Limnol. AOceonogr.,. 10:121-134.
174., Vandermeulen, J. H. and Ahern, T. P. 1976. Effect of petroleum hydrocarbons
on algal physiology: review and progress report, pp 107-125 in Effects of Pollu-
tants on Aquatic Organisms, edited by A.P.M. Lockwood. Cambridge Univ.
Press, London, Eng.
175. Veit'n, G. D. and Kuvas, I. M. 1977. An exhaustive steam-distillation and
solvent extraction unit for pesricide and industrial chemicals. Bull. Environ.
Contom. AToxicol., 17:631-636.
176. V-gars, G. A. and Maynard, A. W. 1977. The residual oxygen bioassay: a
rapid procedure to predict effluent toxicity to Rainbow Trout. Water Res.,
11:343-346.
177. Werner, R. E., Peterson, K. K. and Borgman, L. 1966. Behavioural pathology
in fish: a quantitative study of sublethal pesticide toxicction. Jour. Appl.
Ecol., 3{Supplement)223-247.
178. Webb, R. G. 1978. Solvent extraction of organic water pollutants. Int. J:
Environ. Anal. Chem., 5:239-252.
179. Webb, R. G., Garrison, A..W.r Keith, L. H. and McGuire, J. M. 1973.
Current practice in GC-MS (gas chrotnatographyTfnass spectromerry) analysis of
organics in water. U.S. Net .Tech. Inform. Serv., PB Rep. No. 224947/2GA.
94pp.
180. Weber, W. J. and Morris, J. C. 1963.. Kinetics of adsorption on carbon from
solution. J. San. Engr. Div., A.S.C.E., 89:SA2, Proc. paper No. 3483, pp
31-59.
181. , Weir, P.A. and Mine, C. H. 1970. Effects of various metals on behavior of
conditioned goldfish. Arch. Environ. Health, 20:45-51.
182. Wilhm, J. L, and Dorris, T. C. 1969. Species diversity of benthic macroin-
vertebrates in a stream receiving domestic and oil refinery effluents. The Amer.
Midland Naturalist, 76(2)427-449.
153. Williams, .B.R.H. 1974. Biological moniio,Ing of Affluents. Fish bioassays.
Chem. Ind., 3:89-91.
184. Wright, A. • 1976. The use of recovery as a criterion for toxiclty. Bull. Environ.
Contom. AToxicol., 15:747-749.
5 * 223
-------
185. Wright, R. M. and Bowen, C. E. 1971. Water pollution. Environmental hea-
lth hazard. J. Scii Res. Counc. Jam., 2(2)66-89.
186. Wuhrmann, K. and Woker, H. 1948. Beitrage zur toxikologie der fische. II.
Experimentelle untersuchungen uber die ammoniakund bluasaure-vergiftung.
Schweiz. Z. Hydrol., 11:210-244.
187. Zitko, V. 1976. Comments on the status and directions of environmental re- " :
search. Presented at the NSF/IDOE Biological Effects Program Workshop, Texas
A & M Univ., College Station, TX. , May 17-19, 1976.
188. Jorws, H. R., Pollution Control in the Petroleum Industry, Noyes Dcta Corpora-
tion, New Jersey, 1973.
189. Azod, H. S., (editor), Industrial Wostewater Management Handbook, "Water
Pollution Control in the Petroleum Industry," McGraw-Hill Book Company, New
York, 1976.
190. . Environmental Protection Agency, "Development Docximent'for Effluent Limita-
tions Guidelines and New Source Performance Standards for the Petroleum Refin-
int Point Source Category," Washington, D. C., April 1974, EPA-440/1-74-
014-c.
191. 'Churchill, R. J. and Kaufman, W. J., Waste Processing Related Surface Chemi-
stry of Oil Refinery Wostewoters, University of California, August 1973, SERL'
Report No. 73-3.
192. Engineering-Science, Inc., "Treatment of Comb I.ted Sewer Overflows by Dis-
solved Air Flotation," Report to the City and County of San Francisco, 1974.
193. Chieu, J. N., Schechter, R. S., Humenick, M. J. and Gloyna, E. F., Co-
alescence of Emulsified Wastes by Fibrous Bed, CRWR, University of Texas,
August 1975, Technical Report EHE-75-05.
194. ' Experimental Evaluation of Fibrous Bed Coolescers for Separating Oil-Water Emu-
lsions, U.S. EPA Water Pollution Control Research Series, 12050-DRC-l 1/71.
H95... Brown and Root, Inc., Determination of Best Practicable Control Technology
Currently Available to Remove Oil From Water Produced with Oil and Gas,
Sheen Technical Subcommittee, Offshore Operators Committee, March 1974.
196. Peoples, R,. F., Krishman, P. and Simonsen, R. N., "Nonbiological Treatment
of Refinery Wastewater", Journal Water Pollution Control Federation, Volume
44, Mo. 11, November 1972.
224
-------
J97. Davis, B., "Filtration of Solid Stabilized Oil Emulsion", Thesis, University of
Texus, 1973. ,
198. Eckenfelder, W. W. and Ford, D. L., Water Pollution Control, Pemberton
Press, New York, 1970.
199. Engineering-Science, Ins., Petroleum Refining Industry Technology and Costs or
Wastewoter Control, The Notional Commission on Water Quality, January 1975.
200. Churchwell, R. L., Ford, D. L. and Kachtick, J. W., "A Comprehensive
Analysis of Nitrifying and Chemical Processing Wastewaters", unpublished.
201. Environmental Protection Agency, "Treatment of Petroleum Refinery, Petrochem-
ical and Combined Industrial-Municipal Wastewaters with Activated Carbon,"
Robert S. Ker'r Environmental Research Laboratory, September 1978, EPA-600/
2-78-200.
202. Engineering-Science, Inc., An Assessment of Granular Activated Carbon And
Powdered Activated Carbon for Removal cf Trace Organics in Petroleum Refinery
Wcstewater, Prepared for Argonne National Laboratory, October 1978
203. McCrodden/ B. A., "Water Pollution Abatement at BPOil Corporation's Marcus
Hook Refinery", 46th Annual Conference, Water Pollution Control Federation,
October 1973.
204. Engineering-Science, Inc., The 1976 API Refinery Solid Waste Survey, Port 2; '
Solids Management Practices, Prepared for the7»merican Petroleum Institute,
April 1978.
205. Huddieston, R. L., "Solid-Waste Disposal* lordfarming", Chemical Engineer-
, mo,, Volume 86, No. 5, February 26, 1979.
206. Myers, L., M. Wood, T. Short, Jr., F. Pfeffer, B. DePrater, F. Mayhue and
J. Kingrey: National Petroleum Refining Wastewater Characterization Studies,-
U.S. E.P.A., R. S. Kerr Environmental Research Laboratory, Ada, OK, 1972.
207. Raphael ion, L. A. and W. Harrison: Trace Organics Variation Across the
Wastewater Treatment System of a Class-B Refinery, EPA-600/7-78-125, 1978.
208. Analysis of Refinery Wastewaters for the EPA Priority Pollutants, Interim Report,
American Petroleum Institute, Washington, D. C. 1978.
209. Draft Development Document Including the Data Boj,e for the Review of Effluent '
Limitations Guidelines (BATJIA), New Source Performance Standards, and Pre-
treatment Standards for the Petroleum Refining Point Source Category, U.S. E.
P..'A., 1978. .
? . • '< . • :.225 .... ..'.'.
-------
*~,210. Petroleum Refining Industry Technology and Costs of Wastewater Control, Eng-
ineering-Science, Inc., PB 243-830, 1975.
211. U. S. Environmental Protection Agency Indicatory Fate Study, R. S. Kerr En-
vironmental Research Laboratory, Ada, OK, in preparation.
' 2*2. - Kineonnon, C. B.t Oily Waste Disposal by Soil Cultivation Process, EPA-R2-
72 11D, U. S. Environmental Protection Agency, 1972.
213. Myeri, L. H., T. E. Short, B. L. DeiVatsr, F. M. Pfeffer: Analytical Vari-
ability of Five Wastewater Parameters - Petroleum Refining, EPA-600/2-76-234,
1976.
214. Jeffus, H. M.t Manual of Practice: The Disposal of Combined Municipal/
Industrail Wastewater Residues (Metals), EPA-600/2-79-052, U. S, E. P. A.,
1979.
215. Cleland, J. G. and G. I . Kingsbury: Multimedia Environmental Goals for
Environmental Assessment, E. P. A. Contract No. 68-02-2612, 1977.
i '
2V-. Beychok, M. R., "Characterization of Petroleum Refining Industry and Petro* ,
leum Refinery Wastewaters", Section 3 of rhis report.
217. Snider, E. H. and Manning, F. S., "Evaluation of Pollutant Levels in the Pet-
roleum Refinery Wajtewaters and Residuals", Section 6 of this report.
218. Ford, D. F., "An Evaluation of Existing and Emerging Control Technology for
the Treatment of Petroleum Refinery Wastewatrrs and Sludges", Section 5 of
; thii report. ,
i i '
'219. Buries, S, L., "A Review of Pollutants In Petroleum Refinery Wastewcters and
Effects on Aquatic Organisms", Section 4 of this report.
220. Gruber, D. et al, "Recent Concepts and Development of an Automated Biolog-
ical Monitoring System", JWPCF, March 1980.
221. Bomberger, D. C., "Evaluation of Ammonia Fixation Components in Actual
Refinery Sour Waters", API Publication 954, January 1978 (study performed by
Stanford Research Institute).
222. Schaumb«rg, F. D. and Marsh, B. E., "65 Years of Efficiency Progress in Acti-
vated Sludge", JWPCF, January 1980 (editorial).
223. "Development Document for Effluent Limitation Guidelines, New Source Perfor-
mance Standards and Pretreatment Standards for t!ie Petroleum Refining Point
"- r. Source Ccteaory", EPA 440/1 -79/014-b, December 1979.
226
-------
r224. Grieves, C. G. eral, "Powdered Versus Gianular Carbon for Oil Refinery
Wastewater Treatment", JWPCF, March 1980.
t
225. Anon., "Landfarming Fills an HP! Need", Hydrocarbon Processing, June 1980.
227
-------
-------
Reproduced by NTIS
National Technical Information Service
U.S. Department of Commerce
Springfield, VA 22161
This report was printed specifically for your
order from our collection of more than 2 million
technical reports.
O .5
For economy and efficiency, NTIS does not maintain stock of its vast
collection of technical reports. Rather, most documents are printed for
each order. Your copy is the best possible reproduction available from
our master archive. If you have any questions concerning this document
or any order you placed with NTIS, please call our Customer Services
Department at (703)487-4660.
Always think of NTIS when you want:
• Access to the technical, scientific, and engineering results generated
by the ongoing multibillion dollar R&D program of the U.S. Government.
• R&D results from Japan, West Germany, Great Britain, and some 20 >
other countries, most of it reported in English.
NTIS also operates two centers that can provide you with valuable
information:
• The Federal Computer Products Center - offers software and
datafiles produced by Federal agencies.
• The Center for the Utilization of Federal Technology - gives you
access to the best of Federal technologies and laboratory resources.
For more information about NTIS, send for our FREE NTIS Products
and Services Catalog which describes how you can access this U.S. and
foreign Government technology. Call (703)487-4650 or send this
sheet to NTIS, U.S. Department of Commerce, Springfield, VA 22161.
Ask for catalog, PR-827.
Name
Address.
.3
- Your Source to U.S. and Foreign Government
Research and Technology.
------- |