EPA-450/3-74-042
July 1974
FIELD SURVEILLANCE
AND ENFORCEMENT GUIDE
FOR PETROLEUM REFINERIES
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Triangle Park, North Carolina 27711
-------
JEEA-450/3-74-Q42_
FIELD SURVEILLANCE
AND ENFORCEMENT GUIDE
FOR PETROLEUM REFINERIES
by
Anker V. Sims
The Ben Holt Co.
201 South Lake Avenue
Pasadena, California 91101
Contract No. 68-02-0645
EPA Project Officer: Rayburn M. Morrison
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Ojoality Planning and Standards
Research Triangle Park, N.C. 27711
July 1974
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - as supplies permit - from the Air
Pollution Technical Information Center, Environmental Protection Agency,
Research Triangle Park, North Carolina 27711; or, for a fee, from the
National Technical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22151.
This report was furnished to the Environmental Protection Agency by The
Ben Holt Co. , Pasadena, California, in fulfillment of Contract No. 68-02-
0645. The contents of this report are reproduced herein as received from
The Ben Holt Co. The opinions findings, and conclusions expressed are
those of the author and not necessarily those of the Environmental Protec-
tion Agency. Mention of company or product names is not to be considered
as an endorsement by the Environmental Protection Agency.
Publication No. EPA-450/3-74-042
11
-------
ABSTRACT
A field surveillance and enforcement guide was prepared for
use by air pollution control officers in petroleum refineries and
natural gas processing plants. The processing facilities used by the
petroleum and natural gas industries are described and air pollution
sources are identified. The guide includes methods for estimating
emission rates and describes instruments for monitoring air
pollution sources. Procedures for inspection and surveillance of
refineries are presented. Qualification requirements for the field
enforcement officers are discussed.
111
-------
CONTENTS
LIST OF FIGURES xiii
LIST OF TABLES xvi
SUMMARY 1
INTRODUCTION 5
Purpose 5
Sources of Information 5
Role of Refineries in the Petroleum Industry 6
REFERENCES 8
I. PETROLEUM REFINING AND NATURAL GAS PROCESSING 9
A. Introduction 9
B. Crude Distillation and Desalting 15
1. Crude Distillation Unit 15
2. Naphtha Stabilizer 19
3. Crude Desalter 19
4. Vacuum Unit 21
5. Pollution Sources 23
C. Thermal Cracking 24
1. Process Objective 24
2. Cracking Furnaces 25
3. Process Description 25
a. Visbreaking 25
b. Thermal Cracking 27
4. Pollution Sources 29
D. Coking 30
E. . Deasphalting 36
-------
F. Catalytic Cracking 40
1. The Fluid Catalytic Cracking Process 41
2. Catalytic Cracking with Bead Type Catalysts 44
3. Control Strategies 46
G. Alkylation 46
1. The Sulfuric Acid Process 47
2. The Hydrofluoric Acid Process 49
3. Pollution Sources 49
H. Isomerization 51
I. Hydrotreating 55
J. Reforming 57
K. Hydroc racking 60
L. Hydrogen Production 66
M. Sweetening 69
N. Asphalt Air Blowing 73
O. Acid Gas Treating 75
1. Acid Gas Absorption Processes 76
a. Absorption in Monoethanolamine (MEA) 76
b. Absorption in Diethanolamine (DEA) 78
c. Absorption in Hot Carbonate 78
P. Sulfur Recovery 78
1. Process Objective 78
2. Process Description 79
3. Instrumentation 81
4. Pollution Sources 82
5. Tail-Gas Treatment Processes 83
a. Modified Stretford Process 83
b. Solution Claus Process 84
c. Sulfreen Process 84
d. Stack-Gas Treating Processes 84
O. Sour Water Stripping 85
1. Refluxed Sour Water Steam Strippers 85
2. Nonrefluxed Sour Water Steam Strippers 87
R. Natural Gas Processing 87
1. Gas Dehydration Using Liquid Absorbent 88
2. Gas Dehydration Using Solid Adsorbent 89
3. Acid Gas Removal 92
VT.
-------
4. LPG and Natural Gasoline Recovery 93
by Compression
5. LPG and Natural Gasoline Recovery 95
by Refrigeration
6. LPG and Natural Gasoline Recovery 97
by Oil Absorption
S. Light Ends Recovery 99
T. Wastewater Systems and Solids Disposal 102
U. Flare and Slowdown System 105
V. Storage 106
1. Floating Roof Tanks 108
2. Variable Vapor Space Tanks 109
3. Flares and Incinerators 110
4. Vapor Recovery Systems 110
5. Estimated Hydrocarbon Losses 111
6. Potential Point Sources of Pollutants 119
W. Loading and Transfer 119
1. Loading Equipment 120
2. Vapor Recovery Systems 120
a. Vapor Recovery to Fuel Gas 120
b. Vapor Recovery by Absorption in Gasoline 120
c. Vapor Disposal to Flare 122
3. Loading Losses 122
X. Fuel Gas Systems 123
Y. Steam Generation 127
Z. Cooling Tower 129
AA. Electric Power Generation 132
BB. Catalyst Regeneration 133
II. REFINERY EQUIPMENT 137
A. Introduction 137
B. Pumps 137
1. Centrifugal Devices 138
a. Centrifugal Pumps 138
b. Axial Pumps 143
c. Turbine Pumps 147
2. Positive Displacement Devices 147
a. Reciprocating Piston Pumps 149
b. Plunger Pumps 151
vii
-------
c. Diaphragm Pumps 151
d. Rotary Vane Pumps 153
e. Rotary Gear Pumps 153
C. Compressors 156
1. Positive-Displacement Compressors 156
a. Reciprocating Piston Compressors 156
b. Rotary Lobe Blowers 157
c. Rotary Sliding Vane Compressors 159
2. Centrifugal and Axial Compressors 159
D. Heat Exchangers 163
1. Direct Heat Exchangers 163
2. Indirect Heat Exchangers 163
E. Furnaces 169
F. Jet Ejectors 177
G. Pipe Valves and Fittings 180
1. Pipe 180
2. Valves 181
3. Flanges 181
4. Vents and Drains 183
H. Pressure-Reli.ef Devices 183
1. Spring-Loaded Relief Valve 184
2. Rupture Disc 186
3. Relief Hatch 188
I. Flares 188
J. Knockout Drums 191
K. Scrubbers 192
L. Fractionators 195
in. PROCESS INSTRUMENTATION 199
A. Introduction 199
B. Identification of Process Instruments 199
1. Indicators 201
2. Recorders 201
3. Transmitters 202
4. Controllers 202
5. Control Valves 203
Vlll
-------
C. Flow Measurement 203
1. Positive Displacement Meters 203
2. Variable-Area Meters 203
3. Variable-Head Meters 204
D. Temperature Measurement 204
1. Thermocouples 205
2. Thermometers 205
3. Radiation Pyrometers 205
4. Resistance Thermometers 206
E. Pressure Measurement 206
1. Elastic Elements 206
2. Gravity-Balance Manometer 206
3. Electrical Pressure Instruments 207
F. Level Measurement 207
1. Float Devices 207
2. Displacer Devices 208
3. Hydrostatic Methods 208
G. Analytical Instruments 208
H. Computers 209
IV. MONITORING INSTRUMENTATION 211
A. Source Monitoring 211
1. Monitoring Systems 211
a. Approaches 211
b. System Components 212
c. Monitoring Strategy 215
2. Source Monitoring Interfaces 216
a. Probe and Materials o f Construction 217
b. Sample Conditioning 218
c. Sample Transport and Flow Measurement 220
3. Calibration 223
4. Source Monitoring Instruments 224
a. Gaseovis Contaminants 224
b. Particulates 231
B. Perimeter Monitoring 2^6
1. Continuous Monitoring 237
2. Integrated or Static Monitoring 239
C. Portable Sampling Equipment 240
1. Use of Portable Sampling Equipment 241
2. Types of Portable Sampling Equipment 241
ix
-------
References 245
V. MAINTENANCE OF REFINERY RECORDS FOR USE BY 247
AIR POLLUTION CONTROL FIELD ENFORCEMENT
OFFICERS
A. Introduction 247
1. Necessity of Keeping Records 247
2. Availability of Records to the Inspector 248
3. How Records will be Used 249
a. Permit and Source Registration Identification 249
b. Emissions Inventories 249
c. Emergency Action 251
• d. Legal Action 251
B. Format of Records 251
1. Period of Time Covered 253
2. Person Responsible for Keeping the Records 253
3. Brief Description of Process or Equipment 253
for Which the Record is Maintained 253
4. Data 254
C. Types of Records 255
1. Permit or License Files 255
2. Maintenance Records 255
3. Shutdown and Startup 256
4. Ground-Level Perimeter Monitoring and 257
Continuous Source Monitoring Records
a. Particulates 257
b. Gases (SO2, H2S, NH3, NOX> Amines, RSH) 258
References 260
VI. ESTIMATING AND ASSESSING EMISSIONS 261
A. Inspection and Surveillance Procedures 261
1. Initial Refinery Survey 262
a. Environmental Observations 263
b. External Observations of Facility 266
(1) Visible emissions (plumes) 266
(2) Evaluation of visible emissions 271
!3) Investigation of odor potentials of 272
emission sources
(4) Relating source strength to control 274
requirements
x
-------
c. Management Interview 276
d. Process and Equipment Inventory 277
2. Physical Inspection and On-Site Testing 278
a. Preparing for the Plant Visit 278
(1) Review of records and regulations 278
(2) Review of safety precautions and 297
procedures
b. In-Plant Inspection and Testing 300
(1) Sensory observations 301
(2) Observing process instrumentation 304
(3) On-site testing 304
(4) Grab sampling 308
(5) Source testing 309
3. Resurveys and General Surveillance 311
a. Updating the Process Inventory 312
b. Assessing the Quality of Maintenance 313
B. Estimation of Losses 314
1. Direct Estimation Techniques 315
a. Source Testing and Monitoring 315
b. Direct Observation 317
2. Indirect Estimation Techniques 318
a. Data from Process Instruments 318
b. Equipment Inspection and Operational Data 320
c. Data from Air Sampling Equipment 323
References 325
VII. MAINTENANCE 327
A. Description of Refinery Maintenance Operations 327
B. General Plant Maintenance 328
C. Maintenance of Air Pollution Control Equipment 329
1. Flare and Blowdown System 329
2. Particulate Matter Control Equipment 330
3. Sulfur Recovery Plants 331
D. Maintenance of Monitoring Equipment 332
VIII. PERSONNEL 333
A. Manpower Reqvu'rements 333
B. FEO Functions 336
C. FEO Personnel Qualifications 337
XI
-------
D. Training 340
References 341
GLOSSARY 343
XII
-------
LIST OF FIGURES
Figure Page
1 Flow Diagram of a Small Oil Refinery 13
2 Flow Diagram of an Intermediate Oil Refinery 14
3 Crude Distillation Unit 16
4 Naphtha Stabilizer 18
5 Crude Desalter 20
6 Vacuum Unit 22
7 Visbreaker 26
8 Thermal Cracking Unit 28
9 Coking Unit 33
10 Propane Deasphalting 38
11 Fluid Catalytic Cracking Unit 42
12 Moving Bed Catalytic Cracking Unit 45
13 Sulfuric Acid Alkylation Process 48
14 HF Alkylation Process 50
15 Butane Isomerization 52
16 Hydrotreating 56
17 Reforming 58
18 Hydrocracking Reaction Section 62
19 Hydrocracking Distillation Section 63
20 Hydrogen Production by Steam Reforming 67
21 Gasoline Sweetening 72
22 Asphalt Air Blowing 74
23 Acid Gas Treating 77
24 Sulfur Plant 80
25 Sour Water Stripping Process 86
26 Glycol Dehydration 90
27 Adsorbent Dehydration 91
28 Amine-Water & Glycol Amine Process Using 94
Breakdown Turbine
29 Two-Stage Gas Compressor 96
30 Oil Absorption Plant 98
31 Light Ends Recovery 101
32 Wastewater System & Solids Disposal 103
33 Flare and Blowdown System 107
34 Vapor Recovery for Storage 112
35 Vapor Pressures of Gasolines and Finished 114
Petroleum Products
36 Working Loss of Gasoline from Fixed-Roof Tanks 1 15
Xlll
-------
Figure Page
37 Breathing Loss of Gasoline from Fixed-Roof Tanks 116
38 Loading Rack Vapor Recovery System 121
39 Loading Losses for Tank Trucks 125
40 Loss from Loading Tankers and Barges 126
41 Refinery Fuel Gas System 128
42 Steam Regeneration 130
43 Cooling Tower 131
44 Centrifugal Pump 139
45 Multiple Stage Centrifugal Pump 141
46 Canned Motor Pump 142
47 Close-Coupled Pump 144
48 Balanced Internal Mechanical Seal 145
49 Axial Flow Propeller Pump (Elbow Type) 146
50 Vertical Turbine Pump 148
51 Reciprocating Piston Pump (Double Acting) 150
52 Diaphragm Pump 152
53 Rotary Vane Pump 154
54 Rotary Gear Pump (Two-Impeller) 155
55 Reciprocating Compressor (2-Stage) 158
56 Rotary Lobe Blower (Two Impeller) 160
57 Rotary Compressor (Sliding Vane) 161
58 Labyrinth Seal 162
59 Barometric Condenser 164
60 Shell & Tube Heat Exchanger 166
61 Double-Pipe Heat Exchanger with Longitudinal Fins 167
62 Air-Cooled Heat Exchanger 168
63 Vertical Cylindrical Furnace 170
64 Horizontal Box Type Fired Heater 171
65 Gas Burner 173
66 Combination Gas and Oil Burner 175
67 Steam Jet Ejector 178
68 Flanges 182
69 Relief Valve 185
70 Typical Rupture Disc Installation 187
71 Pressure Relief Hatch 189
72 Elevated Flare 190
73 Knockout Drum 193
74 Scrubber 194
75 Fractionator 196
76 Typical Stack Monitoring System 221
77 Fuel Use/Sulfur Balance Report 250
78 Odor Survey Form 252
79 Activity Status Report 284
xiv
-------
Figure Page
80 Process Flow Diagram of a Sour Water 288
81 Symbols Used in Petroleum Flow Diagrams 289
82 Bulk Plant Data Sheet 291
83 Truck Loading Inspection Data Sheet 292
84 Oil-Water Separator Inspection Sheet 293
85 Tank Inspection Report 295
86 Natural Gasoline, Gas, and Cycle Plant 296
Survey Summary
xv
-------
LIST OF TABLES
Page
1. Products From Three Typical Refineries 11
2. Sulfur Content in Coker Products 31
3. Reactor Products 41
4. Alkylation Processes Operating Temperatures 47
and Pres sures
5. Sulfur and Nitrogen Distribution in the High 61
Pressure Separator
6. H2S Distribution in the Low-Pressure Separator 65
7. Loading Losses: Motor Gasoline, Turbine Fuels, 124
Aviation Gasoline
8. Instrument Identification Codes 200
9. Approaches to Source Monitoring 212
10. System Components and Operations for Stationary 214
Source Monitoring
11. Contaminants that can be Tested in the Field 306
with Portable Devices
xvi
-------
SUMMARY
This Field Surveillance and Enforcement Guide was prepared to
familiarize state and local air pollution control officials with the
operation of petroleum refineries and natural gas processing plants
and to aid agency personnel developing surveillance, inspection,
monitoring, reporting, and enforcement procedures.
The Guide is divided into eight chapters. The first three
describe present-day refinery equipment and operation with emphasis
on air pollution sources. The remaining chapters describe monitoring
equipment, record-keeping requirements and enforcement programs
and procedures. A glossary of petroleum industry terms is given at
the end of the guide.
CHAPTER I PETROLEUM REFINING AND NATURAL GAS
PROCESSING
A typical refinery consists of a crude distillation unit and a
variety of units designed to separate, react, and blend petroleum
fractions. Several air pollution sources are common to many refinery
process units. Sour water streams contain dissolved sulfur compounds
-------
and ammonia. Fuel gases containing hydrogen sulfide are produced
at various locations. Heaters are widely used and produce combustion
products such as sulfur oxides, nitrogen oxides, and particulates.
Special air pollution problems associated with specific
operations are discussed. These include sulfur oxide emissions from
sulfur plants, particulates, carbon monoxide and sulfur oxides from
catalytic crackers and catalyst regeneration operations, hydrocarbon
emissions1 from storage and loading facilities, and emissions from
flares and blowdown systems.
CHAPTER II - REFINERY EQUIPMENT
Individual equipment items located throughout a refinery have
characteristic air pollution problems. Each type of process equipment
is described with emphasis on air pollution sources. In general,
pumps, compressors, heat exchangers, valves and fittings emit air
contaminants by leakage. Furnaces and heaters are a source of
combustion pollutants. Relief valves are designed to release gases
at certain pressures and represent an important potential source of
air pollution.
CHAPTER III - INSTRUMENTATION
Process instrumentation and control equipment, including
devices for measuring pressures, temperature, liquid levels, and
flow rates, are described. The application of commonly used
process instruments to the detection and measurement of air
-------
pollution is discussed. A table showing the derivation of the codes for
instruments and controls is included.
CHAPTER IV MONITORING INSTRUMENTATION
Equipment and strategy for monitoring air pollution sources are
discussed. Sampling and source monitoring on a local and continuous
basis is described. Instruments for determination of sulfur dioxide,
nitrogen oxides, carbon monoxide, hydrocarbons, and particulates are
commercially available. Instruments may be required for monitoring
ambient air quality at locations where widespread leakage occurs, such
as hydrocarbon emissions from valves, pumps, compressors, etc. Use
of portable instruments is discussed.
CHAPTER V MAINTENANCE OF REFINERY RECORDS FOR USE BY
AIR POLLUTION CONTROL FIELD ENFORCEMENT OFFICERS
The importance of record keeping in refinery surveillance and
control activities is emphasized. Records available from the refiner
are described. Records should be kept on emission rates, source
registration, permits to operate, complaints, episode histories, and
compliance p]ans. Programs, strategics, and formats for record
keeping are described.
CHAPTER VI ESTIMATING AND ASSESSING EMISSIONS
The procedures for inspection and surveillance of refineries
include the initial survey, physical inspection, and follow-up surveys.
-------
The initial survey consists of environmental and external observations
including visible emissions, odor detection, and management inter-
views. The physical inspection includes on-site review of records,
sensory tests, instrument readings, equipment inventories and effluent
sampling. A resurvey is required to update inspection data and review
maintenance records. Techniques for estimating loss rates are dis-
cussed.
CHAPTER VII MAINTENANCE
Refinery maintenance is an important element of refinery oper-
ations. Maintenance operations are a source of air pollution, and
maintenance records are a valuable source of information to the control
officer. Chapter VII describes typical maintenance operations. Main-
tenance of air pollution equipment and monitoring instruments is dis-
cussed.
CHAPTER VIII PERSONNEL
Manpower requirements for state and local air pollution control
agencies depend upon the number and complexity of refineries and
other chemical process related plants in the jurisdiction. The function
of the field enforcement officer to survey, inspect, and investigate is
defined. Personnel qualifications and a list of programs for training
and recruiting are given.
-------
INTRODUCTION
PURPOSE
This Guide was prepared to aid state and local air pollution con-
trol officials in carrying out effective surveillance and control of air
pollution sources in petroleum refineries and natural gas processing
plants. The Guide is designed to familiarize control officials and field
enforcement officers with the operation of petroleum refineries and
natural gas processing plants. In the description of refinery equipment
operations, particular emphasis is given to the air pollution aspects.
Another important purpose is to provide guidance to agency personnel
in developing surveillance, inspection, monitoring, reporting and en-
forcement procedures.
SOURCES OF INFORMATION
Private and public sources were used to prepare the Guide.
Public sources include textbooks and journals which are widely distri-
buted in the petroleum industry. Some general references are listed at
the end of this Introduction. Manufacturers and suppliers were relied
upon for information on equipment and monitoring instrumentation.
Catalogs of air pollution control equipment suppliers are regularly
-------
published by Chemical Engineering and Environmental Science and
Technology. The files of The Ben Holt Co. and Pacific Environmental
Services, Inc. were also sources of information.
ROLE OF REFINERIES IN THE PETROLEUM INDUSTRY
The petroleum industry is divided into the following functional
operations: exploration, production, transportation, marketing and
distribution, and refining. An oil company may perform only one or a
few or all'of these functions. Exploration and production companies
search for oil and gas reserves, drill wells, and develop and produce
the resource. Many companies provide only transportation services -
using pipeline, barge, rail, tanker, and truck facilities. Marketing
and distribution generally refers to petroleum products only. This
function includes sales, advertising, and distribution to retail and com-
mercial outlets.
The refining function includes a broad range of processing and
manufacturing operations designed to convert crude oil and natural gas
to saleable products. In the early days of the oil industry, the refining
function consisted only of batch distilling of crude oil. As the number
and quality of petroleum products increased, a myriad of refining pro-
cesses were developed to effect both chemical and physical changes in
petroleum. Today, there are literally hundreds of these processes
available. Each refinery consists of a combination of them. The con-
figuration of each refinery and natural gas processing plant depends on
-------
the feed stock and product specifications. Field enforcement officers
should not expect to find all or even a majority of the process equip-
ment and abatement techniques presented (in Chapters I and II). Many
factors, such as throughput, geography, and slate and product split,
affect not only the number and types of processes utilized, but also the
abatement controls necessary. By understanding the function and
operation of each refinery process, the field enforcement officer will
gain the knowledge of the refining industry necessary to the effective
performance of his tasks.
-------
REFERENCES
TEXTBOOKS AND HANDBOOKS
Danielson, J. A. Air Pollution Engineering Manual, 2nd Ed. ,
Washington, D. C. , Air Pollution Control District, County of
Los Angeles, and Environmental Protection Agency, 1973. p. 987.
Guthrie, U. B. Petroleum Products Handbook. New York, McGraw-
Hill Book Co. , 1960. p. 837.
Lund, H. F. Industrial Pollution Control Handbook. New York,
McGraw-Hill Book Co., 1971. p. 886.
Nelson, W. L. Petroleum Refinery Engineering. New York,
McGraw-Hill Book Co. , 1958. p. 960.
Perry, J. H. (ed.). Chemical Engineers' Handbook, 4th Ed.
New York, McGraw-Hill Book Co. , 1962. p. 1884.
Stern, A. C. (ed.). Air Pollution, 2nd Ed. 3 vols. New York,
Academic Press, 1968.
JOURNALS
Chemical Engineering
Environmental Science and Technology
Hydrocarbon Processing
Journal of the Air Pollution Control Association
Oil and Gas Journal
-------
I. PETROLEUM REFINING AND
NATURAL GAS PROCESSING
A. INTRODUCTION
Crude petroleum is a mixture of hydrocarbons and small
amounts of sulfur, nitrogen, oxygen, and various metals. Crude oils
are classified as being derived from paraffinic, naphthenic, or inter-
mediate base stocks. The characteristics of a given crude can vary
from a clear liquid in the gasoline range to a pitch that is so viscous it
must be heated to be pumped. Crudes from geographically related oil
fields tend to have similar compositions and properties.
Refineries are designed to convert one or more types of crude
oil into saleable petroleum products. The types of oil processed and
the products desired determine the facilities of a refinery. Theoreti-
cally, it is possible to produce any type of product from any crude oil,
but in practice economic considerations usually decide which products
will be produced.
The trace elements in petroleum also affect the processing
steps and, to some extent, the pollutant emissions from the refinery.
A high sulfur content in the crude oil increases the corrosive charac-
teristics of the oil and its products. After processing in the refinery,
-------
sulfur tends to concentrate in the light ends and gases in the form of
hydrogen sulfide and in the heavy ends as complex sulfur compounds.
Generally, the hydrogen sulfide is selectively removed from the process
gas by treating with ethanolamines. The hydrogen sulfide is regenerated
from the ethanolamines and is usually sent to a sulfur recovery unit.
New air pollution control regulations will require removal of hydrogen
sulfide and its conversion to sulfur, or the use of some other mechanism
for controlling sulfur emissions. The residual fuels which contain high
percentages of sulfur can be treated by hydrotreating or hydrocracking
processes to reduce the sulfur content of the fuel oil to a level that is
acceptable to local air pollution control agencies. The sulfur is con-
verted to hydrogen sulfide and enters the process gas stream.
Trace metals such as iron, nickel and vanadium are present in
crude petroleum and act as a poison to some catalysts. As the metal
accumulates on the catalyst, the activity of the catalyst decreases.
Crudes with high concentrations of metal deposits require more frequent
catalyst regeneration and replacement. These operations generally
release pollutants to the atmosphere and should be carefully controlled.
The products from a refinery vary widely with location,
climate and time of year. In winter, there is a high demand for heating
fuel oils, and the gasoline products must contain a high percentage of
volatile components for cold weather starts. In summer, the demand
for fuel oil declines, the demand for gasoline increases, and the vola-
tility of the gasoline must be reduced to minimize vapori?.ation losses
10
-------
and carburetor vapor lock. Refineries must be sufficiently flexible to
meet these varying demands..
Table 1 shows the type of products that might be prepared at
three typical refineries that differ in size, location and climate.
Table 1. PRODUCTS FROM THREE TYPICAL REFINERIES
Products
Feed, bbl/day
LPG
Olefins
Gasoline
Jet Fuel
Kerosene
Diesel
Fuel Oil
Heavy Fuel Oil
Naphthenates
Solvents
Asphalt
Lubricating Oils
Greases
Petroleum Coke
a
Refinery 1
10,000
X
X
X
X
X
Refinery 2
80,000
X
X
X
X
X
X
X
X
X
Refinery 3
200,000
X
X
X
X
X
X
X
X
X
X
X
X
X
X
a. Refinery 1 is assumed to have a crude capacity of 10,000 bbl/day
b. Refinery 2 is assumed to have a capacity of 80,000 bbl/day.
c. Refinery 3 is assumed to have a capacity of 200, 000 bbl/day.
As the size of the refinery increases, the number of products
increases and the processing operation becomes more flexible. The
production rate of each of the products shown can be varied signifi-
cantly by making relatively minor changes in refinery processing con-
ditions. Hydrocarbon fractions can be shifted from one product to
another to' meet product demands.
11
-------
Figure 1 is a flow diagram of a low-capacity, basic refinery
which contains only the minimum number of elements to supply a local
area with fuels. This type of refinery is relatively inflexible and pro-
duces only a limited number of products.
Figure 2 is a flow diagram of an intermediate capacity refinery.
This facility has more processing units and a much wider range of
products. These diagrams show only the most fundamental processing
operations. A large, complex refinery would be about the same with
the addition of specialty processes.
Operations such as sweetening, hydrotreating, hydrocracking,
sour-water treatment, and sulfur plants are now needed for pollution
control in most refineries. They are not shown because they do not
affect the principal product streams, but they are essential in modern
refinery practice and will be found in every modern refinery.
Many refinery products, for which a special but limited market
exists, are manufactured by only a few refineries. In this category
are:
Asphalt Petroleum Coke
Benzene Petroleum Solvents
Cresols Phthalic Anhydride
Detergents Tar
Greases Toluene
Lubricating Oil White Gasoline
Naphthenic Acid Xylene
Aliphatic Solvents
The F£O should expect to encounter at least a few of them.
12
-------
FUEL GAS
LPG
FELED
ictun.oSfc.Mt.
01
OIL.
Figure 1. Flow diagram of a small oil refinery.
-------
FiLLO
STA&ILIZ
P'•>"[• I Ll.A"p(OsJ
T
PL.UIC?
CATALYTIC
1
PU£L GAS
t*
LPG
T r
1
1
•C-ATA t_YT' C-
\~L £- ^-& {T- Nrt Jr. [i.
J
l
- |
T
| ; H
SLPAKATc?f
N
:
1
i
^
/
fc
aA;>OL
-------
Refineries may be divided into on-site facilities and off-site
facilities. On-site facilities consist of the petroleum processing units.
Off-site units are made up of the support facilities which include waste-
water treatment, power generation, steam plant, water treatment, and
feed and product storage. Both on-site and off-site processing units
are discussed in detail in this chapter. Flow diagrams illustrate the
processing units. On the diagrams, the point sources of pollution dis-
cussed in the text are indicated by numbered diamonds -- <£>, <£>, etc.
The identification codes used in the figures are explained in Table 8
within Chapter III, Process Instrumentation.
B. CRUDE DISTILLATION AND DESALTING
1. Crude Distillation Unit
The crude distillation unit (Figure 3) is the first refinery unit to
process the crude oil. The unit separates the incoming crude oil into
light naphtha, heavy naphtha, middle distillates, and bottoms residue.
The bottoms residue is usually further processed in a vacuum unit to
produce heavy gas oils and a vacuum residuum.
The crude oil is pumped from tankage, preheated by exchange
with the products to about 250-300° F, and then desalted. Booster
pumps, downstream of the desalter, pump the crude oil through addi-
tional heat exchange and through a crude furnace. The temperature of
the oil coming out of the furnace is usually 650 - 700° F. From the
furnace, the crude oil passes to the crude distillation column. Light
15
-------
TO 6TAe>ii_iZ.e-ri.
N&ATfc-fU
Figure 3. Crude distillation unit.
-------
naphtha and reflux are condensed in the overhead condenser. Water
remaining in the crude and entering the tower from the stripping steam
is also condensed overhead. The foregoing products plus any noncon-
densed gases pass to the overhead accumulator where the phases are
separated. Gas may be compressed to the fuel gas system or for light
hydrocarbon recovery. In some instances, it may be vented to the
flare system. The light naphtha is pumped to the stabilizer. The
reflux is pumped back to the crude tower for control of the overhead
temperature. The water is pumped off plot for treating.
The crude tower usually produces three or more sidestreams
which are stripped of their light ends by the use of steam in a side-
stream stripper tower. Each sidestream (usually heavy naphtha, kero-
sene, and gas oil) is heat exchanged with crude oil, cooled in a water or
air-cooled exchanger, and pumped to storage.
The hot residual is steam stripped in the base of the crude
tower and then, in most refineries, it is pumped directly to a vacuum
unit heater. Alternatively, the hot residual oil may be cooled by ex-
change with the incoming crude oil and with cooling water before being
pumped to storage.
The point sources of pollution in the crude distillation unit are
the crude heater (point 1 on the figure), the overhead accumulator vent
(point 2), water from the overhead accumulator, and the stream
sample connections.
17
-------
I—&
(TCV-*-
, '•
ACCUMULATOR
-------
2. Naphtha Stabilizer
Light naphtha from the crude unit is preheated by exchange with
the stabilized naphtha and sent to the stabilizer (Figure 4). This is
usually a 20 to 30-tray tower in which only mild fractionation (essen-
tially stripping) takes place. The light gases are removed overhead in
the tower in order to reduce naphtha vapor pressure. The naphtha may
then be used as a JP-4 jet fuel or gasoline component. Reflux is con-
densed overhead and the gas is usually sent directly to fuel gas through
a back pressure controller. Alternatively some liquid may be produced
overhead in a high-pressure light ends recovery unit. The stabilizer
bottoms are reboiled by means of a steam heated reboiler. The stabi-
lized light naphtha is sent to storage after suitable cooling in an ex-
changer and water cooler.
3. Crude Desalter
Desalters are not always required, but when used are the first
unit operation applied to the crude. The crude oil passes through the
desaltcr to remove salt, silt, sand, water, and other crude oil con-
taminants (Figure 5). Water and sometimes demulsifying chemical
are added to the crude oil stream usually before preheating. The
mixture is passed through an electrostatic field inside the drum after
preheating by exchange to 250- 300° F. The desalter can usually be
blocked in, bypassed, and drained during normal operation. A point
source of pollution is the desalter water effluent (point 1 in the figure)
which will contain a mixture of oil, water and chemicals.
19
-------
Pt- fcM
CWt.MlC.AL
VALAJfc.
,
r "*"
I
E-f-FLUtUll
ZOKlfc.
2 50-500*
IOOP6IQ
Figure 5. Crude desalter.
-------
The sour water effluent from the desalter often must be pro-
cessed in a sour water stripper or oxidizer for purification. Some-
times this water is degassed and then recycled back to the crude oil.
The gas must be treated for H2S removal and is a possible point source
of pollution. The desalter is usually equipped with these sampling
connections: oil-water emulsion (feed), desalted oil, desalter internal
sampler, and effluent water sampler.
Sometimes two-stage desalting is employed to further remove
salts, etc. , from the crude oil. This is merely two desalter drums in
series and the point sources of pollution are the same for both vessels.
4. Vacuum Unit
In the vacuum unit, hot residual bottoms from the crude unit
enter the vacuum tower by way of a vacuum heater which heats the oil
to 700 - 750° F (Figure 6). The process here again is distillation, but
this time at a reduced pressure (vacuum). The pressure will usually
be 25 - 100 mm Hg absolute (equivalent to 735 - 660 mm Hg vacuum).
The combination of higher temperature and lower pressure allows addi-
tional distillation to take place.
Any vacuum gas oils are taken off from the side of the vacuum
tower, exchanged with feed, cooled in a water-cooled heat exchanger
and pumped to storage. A small amount of gas oil plus any noncondens
ables are taken overhead to the jet ejector system. Normally three to
five jet ejectors are employed with intercooling, and aftercooling is pro-
vided. The overhead streams are sent to an accumulator where the
21
-------
STt-AM
LT VACUUM QA5 Pit.
VACUUM
HEATER.
RESIDUALS
Figure 6. Vacuum unit.
-------
phases are separated. Gas is vented to a heater for burning or to a
gas recovery unit for purification. The hydrocarbon liquid is usually
sent to a slops tank. The water phase is sent to a sour water stripper
for purification. The hot vacuum residual is sent either to storage via
crude exchange and water cooling or directly to a coking unit.
5. Pollution Sources
The major point sources of air pollution for the crude distilla-
tion unit are shown in Figure 3. The crude heater is a fired heater and
the stack gas (point 1) contains pollutants (see ChapterII). The over-
head accumulator vent (point 2) may contain H2S and hydrocarbons.
Depending on the amount and composition of the vent gas, the gas may
be flared or sent to the fuel gas system.
The naphtha stabilizer is a closed system and should have no
major point sources of air pollution.
The air pollution point source for the desalter is shown in
Figure 5. The water effluent (point 1) may be sour and, if so, it
should be sent to a sour water stripper.
Air pollution point sources for the vacuum unit are shown in
Figure 6. The vacuum heater is a fired heater and the stack gases
(point 1) will contain pollutants (see Chapter II). The overhead accumu-
lator vent (point 2) may contain H2S and hydrocarbons and should be
vented to a heater for burning or should be sent to a gas recovery unit.
23
-------
C. THERMAL CRACKING
1. Process Objective
Thermal cracking of petroleum is one of the oldest refinery
operations and continues to play an important role in many refineries
today. Thermal cracking is the process whereby large hydrocarbon
molecules are converted into smaller molecules through the process
of thermal decomposition. The objectives of the process vary widely
depending oh the feed stock and the severity of cracking. Cracking
feed stocks vary from very light materials such as butane to very
heavy vacuum residual tar. Cracking of light hydrocarbon gases and
distillates to produce ethylene, propylene and butylene will not be
considered here. Olei'in plants are commonly considered petrochemi-
cal processes even though olefin generation is important in the refinery
scheme. Severe cracking of vacuum tars to produce petroleum coke
•
and cracked distillates will be covered in the "Coking" section.
Present day cracking operations can be divided into once-
through and recycle plants. In recycle plants, the partially thermally
cracked feed is distilled to separate the uncracked portion. Part or
all of this uncracked material is then recycled to the cracking section.
Visbreaking is a once-through cracking process where a heavy residual
fuel oil is mildly cracked to reduce fuel oil viscosity. Recycle thermal
cracking plants often involve a variety of cracking and distillation steps
to produce a range of petroleum products. The most common recycle
cracking plants today feed gas oil and produce a high yield of cracked
24
-------
gasoline. Thermal reformers have been used to convert straight run
naphthas to high octane gasoline. Today, however, catalytic reforming
is used for this purpose. See Section J, Reforming.
2. Cracking Furnaces
The cracking operation takes place in a tubestill furnace similar
to other furnaces in the refinery. Cracking furnaces are designed to
control residence time and temperature. Unlike other refinery heaters,
cracking furnaces are designed to heat the process stream to tempera-
tures in the 800 to 1100°F range. The furnace consists of radiant and
convection sections. Burners located near the floor provide radiant
heat to the tubes on furnace walls. Heat is recovered from the combus-
tion gases in the convection section at the top of the furnace. The con-
vection section may be used to preheat air, generate steam, or heat
other process streams. Coking of tube surfaces is an important design
consideration. The hot tube surface causes coke to deposit on the
tubes especially in liquid phase cracking operations. As coke accumu-
lates, heat transfer is impaired and the tube wall temperature rises.
Regular cleaning of tube surfaces is required to maintain operating
efficiency.
3. Process Description
a. Visbreaking - The process flow diagram for a typical
visbreaker is shown in Figure 7. Vacuum residual tar is pre-
heated by heat exchange with visbroken fuel oil and fed to the
visbreaker furnace. Mild cracking in furnace tubes produces a
25
-------
VACUUM
RESIDUALS
Figure 7. Visbreaker.
-------
mixture of residual oil, naphtha and gas. The reaction products
are quenched with a recycle stream and fractionated in a distil-
lation tower. In most visbreakers, no cracked material is re-
cycled to the cracking furnace. The visbroken fuel oil is
blended with other components to meet viscosity, sulfur and
pour point specifications. Visbreaker naphtha is taken over-
head and stripped of light ends in the stabilizer. The naphtha
may be desulfurized and reformed before blending into gasoline.
Light ends containing hydrogen sulfide are taken from the over-
head accumulators on the fractionator and stabilizer. These
gases may be processed further in a gas recovery plant or fed
directly to the refinery fuel gas system. A side cut is taken off
the fractionator to provide internal cooling and heat for the
stabilizer reboiler.
b. Thermal Cracking - The process flow diagram for a
typical thermal cracking process with a single recycle stream
is shown in Figure 8. The feed stock in most units is a reduced
crude. The feed is preheated by exchanging with other process
streams and charged to the quench tower. The feed serves to
quench the furnace discharge to about 800 to 850 °F. The
quench tower separates residual fuel oil from distillate petro-
leum. In this scheme, none of the residual material (black oil)
is' cracked thus reducing coke buildup in the furnace. The
quench tower contains several bubble trays to ensure an
27
-------
00
Figure 8. Thermal cracking unit.
-------
efficient separation of residual oil. The overhead vapors, con-
taining straight run and cracked distillate are separated in the
fractionator. The heavy gas oil from the bottom of the frac-
tionator is fed to the cracking furnace which is operated at
elevated pressure. The pressure at the furnace outlet is
dropped sharply to the quench tower pressure. One or more
side cuts are taken off the fractionator. The overhead vapors
containing naphtha are separated in a stabilizer similar to the
one used in a visbreaker.
Most thermal cracking plants are combination plants
containing two or more cracking furnaces. Side cuts from the
main fractionator are recycled to separate furnaces to provide
selective cracking of residual fuel, gas oil, and naphtha. In
some plants, the residual fuel oil from the quench tower is fed
to a low-pressure separator followed by vacuum distillation.
Regardless of complexity, the major operations of cracking,
quenching, and fractionation are present.
4. Pollution Soxirces
The major point sources of air pollution are shown in Figures 7
and 8. The cracking furnaces (point 1) are fired with fuel oil, natural
gas, or refinery fuel gas. Stack gas emissions such as NOX, SO2, and
particulates will be present. Cracking furnaces may also emit pollu-
tion during the maintenance shutdown period (point 2). Furnace tubes
are cleaned with steam-air mixtures. Coke deposits contain sulfur and
29
-------
nitrogen. Removal of coke by burning will result in SO2, NOX, CO,
and particulate emissions. Steam condensate will require stripping to
remove sour constituents. Flue gases from cleaning operations may
be routed to a furnace or otherwise incinerated to complete the com-
bustion process. Mechanical cleaning of furnace tubes will greatly
reduce the air pollution problems, but the refiner is then faced with a
significant solids waste disposal problem.
Thermal cracking also serves as a desulfurization process,
with the degree of desulfurization varying from 3% for mild visbreaking
to 80% for severe thermal cracking. The sulfur is removed in the form
of H2S in the cracked gases (point 3), and these gases should be sent to
an acid gas treating plant before being used for refinery fuel.
D. COKING
Coking is a thermal cracking process in which crude oil residue
and other decanted oils and tar-pitch products are cracked at high
temperature (900 - 1, 080° F) and low pressure (atmospheric) into lighter
products and petroleum coke. The objective is to produce gas oil and
lighter petroleum stocks from the crude residue. These materials are
further processed and blended with other stocks to produce premium
products such as gasoline, jet fuels, and diesel fuels.
The crude oil residue feed to the coking unit contains most of
the impurities of the crude oil. It contains most of the heavy metals
(nickel, vanadium), essentially all the asphaltene, resin, and ash, 40%
to 60% of the sulfur, and 80% to 90% of the nitrogen.
30
-------
.A rough estimate of the distribution of sulfur in coker products
follows in Table 2.
Table 2. SULFUR CONTENT IN COKER PRODUCTS
_ , , Percent of Feed Sulfur
Product „
on Products
Fuel Gas
Coker Gasoline
Heavy Oils
Coke
12
3
25
60
100
In the coking process, the residue is heated to a cracking temperature
in which the longer hydrocarbon chains in the residue are severed into
smaller more volatile components. These components evaporate at
this high temperature and are collected and separated into the desired
products by distillation. The asphaltene, resin, ash, metals, and
residual carbon precipitate out of the liquid to form the highly cross-
linked polymerized structure of the product coke. This process con-
tinues until all the volatile components are removed by vaporization
and all the nonvolatile components have formed coke.
There are two principal coking processes: the fluid coking pro
cess and the delayed coking process. The most widely used is the
delayed coking process and very few fluid coking units are now in
service.
31
-------
In the fluid coking process, the crude residue is fed to the
reactor where it is mixed with recycled hot coke particles. The
hydrocarbon portion of the liquid feed cracks and evaporates while
the nonvolatile material is deposited on the suspended (fluidized) coke
particles. The coke particles thus grow in size, sink to the bottom of
the reactor and flow to the burner. In the burner, the particles are
fluidized with air, partially burned and are recycled back into the
reactor. A portion of the coke produced in the reactor is withdrawn
as product.
A point source of pollutant emission is the burner. Emission
control equipment could be similar to that used with fluid catalytic
cracking units. Meeting new air pollution regulations may require the
use of electrostatic precipitators and CO boilers.
In the delayed coking process (Figure 9), the charge stock is
fed to the bottom section of the fractionator where material lighter
than the desired end point of the heavy gas oil is flashed off. The
remaining material combines with recycle and is pumped from the
bottom of the fractionator to the coking heater where it is rapidly
heated to above 900 ° F. The liquid-vapor mixture leaving the coking
heater passes to a coke drum.
A unit usually has two drums with one being filled while the
other is being decoked. Large units may have four or even six drums.
The coke drums are most often sized so that each one operates on a
48-hour cycle, thus permitting decoking of a drum to be scheduled at
the same time each day on a 24-hour cycle.
32
-------
rue.u
rr~~^~i
^ ' } VT£AM ? T
^JL OuT J-.
Figure 9. Coking unit.
-------
Under the time-temperature conditions in the drum, coke is
formed and accximulates in the vessel and the more volatile compo-
nents former! leave in I he overhead vitpors. The coke drum overhead
vapors enter the lower section of the fractionating tower for separa-
tion into gas, gasoline, gas oil., and recycle stock.
After the first coke drum is filled with coke, the operation is
shifted to the second drum and the first drum is ready for coke
removal.
The. initial step in coke removal is the cooling of the hot drum
with steam. Live steam is blown into the drum where it absorbs heat,
evaporates .some hydrocarbon material and ^ntrains some coke par-
ticles. After leaving the drum, the vapors are cooled to condense the
steam and hydrocarbons. The cooled stream separates into three
parts :
1. Water, with coke particles, that should be added to the
coke removal system.
2. Hydrocarbon liquid that should be added to the slops
system.
3. Noncon'iens abl e.s dial normally go to a fired heater or a
flare for disposal. This stream is primarily fuel gas and usually con-
tains small amounts of sulfur compounds.
The .second step is cutting the coke. A high-pressure water jet,
2,000 psi or more is used to cut the coke free from the drxim. The coke
particles are washed out with the water and are separated from the
34
-------
water on vibrating screens. Fines remaining in the water are removed
in a thickener, and the water is recycled for cutting and transporting
the coke.
Pollution Sources
Point sources of pollutant emission from this process include:
1. The steam from the steam-out operation, if not
properly condensed and separated (point 1 in the figure).
2. Accumulator fuel gas is normally rich in H2S. This
stream has to be treated before entering the fuel gas system of the
refinery (point 2).
3. Coker gasoline is normally treated for the removal of
H2S (point 3). (See Section M, Sweetening. )
4. Water drawn from the overhead accumulator contains
H2S and should be routed to a sour water stripper (point 4).
5. The fired heater is a source of pollution; see Chapter II
(point 5).
6. Coking units are in general covered with coke dust.
Unless the units are cleaned and washed thoroughly, these fine
particles will blow with the wind and may create pollution.
7. Most delayed coking units use water for cutting the coke.
The water is recycled in this operation and stored in open containers.
Since this water contains some sulfur compounds, it may be the source
of objectionable odors.
35
-------
E. DEASP HALTING
Deasphalting is used to separate oil and asphalt. The feed
stock is reduced crude from the crude distillation process. Reduced
crude is the heavy fraction of the crude from which as much distillate
has been removed as is practical with the existing crude unit. Deas-
phalting produces products comparable to those produced by high-
vacuum distillation, but deasphalting is generally capable of more
complete oil-removal. A variation of the process is sometimes used
to remove trace amounts of asphalt from lube-oil stocks.
Asphalt-oil separation is accomplished by selective extraction
(liquid-liquid extraction) of the oil by a light hydrocarbon, usually
propane. A heated mixture of reduced crude and liquid propane settles
into two liquid phases, an upper phase containing oil and most of the
propane and a lower phase containing the asphalt and some propane.
The phases are separated, and propane is recovered for recycling by
flashing and steam stripping.
Liquid-liquid extraction is a unit operation commonly used in
refineries and chemical plants. In this operation, a mixture is sepa-
rated into two components by means of a selective solvent. The addi-
tion of the solvent to the feed mixture must result in a two-phase
mixture with an appreciable density difference between the two phases.
The two phases will contain different ratios of the two feed components.
Although a single step of mixing and settling will produce some
separation, multiple-step operations are frequently used to achieve
36
-------
more complete separation. Usually counter cur rent extraction is em-
ployed, with solvent entering near one end of the column and feed at
the other. The two liquid phases leave the column at the top and bottom.
Mixing and settling within the column can be accomplished in several
ways. Perforated plates or packing are often used, as well as mechan-
ical agitators driven by a common shaft running through the column.
A flow diagram for a typical propane deasphalting process is
shown in Figure 10. Reduced crude is metered to the process under
flow control and passes through the feed heater where it is brought to
operating temperature by indirect steam heat. The heated reduced
crude is fed to the center of the extraction column where it is brought
into contact with a rising stream of liquid propane. The propane, a
recycle stream, is metered through a flow controller and through the
propane heater into the bottom of the extraction column. Asphalt,
containing some propane, leaves the bottom of the extractor under
level control. The remaining propane and the extracted oil leave the
top of the extractor. A steam coil in the upper section of the column
is used to heat the rising propane-oil stream, reducing the solubility
of asphalt in that stream. In this way, final traces of asphalt are re-
moved from the oil, forming a separate phase that flows downward
through the column.
The asphalt phase from the column passes through a furnace
and then to a flash drum, where most of, the propane is removed as a
37
-------
00
/I TV bUB-Gt.
( 1 pnjj M
WAJEJL
^\ TO SE-WCA.
S. "y
r/iopAMt.
WtATf-
t)TM
STM
id
COMPfLE.'ibOR-
nu
f-LASH
OIL.
6TM
X
HJJ... J
TO
•6TM.
•6AS OIL
•ASPHALT
Jfc.T
Figure 10. Propane deasphalting.
-------
vapor. Since the remaining asphalt still contains a small amount of
propane, this stream is fed to the asphalt stripper. The remaining
propane is steam stripped from the asphalt, and the asphalt product
leaves the bottom of the stripper under level control. The stripping
stream and propane leave the top of the stripper and pass through the
jet condenser to condense the steam.
The oil-propane phase from the extractor is fed to a propane
vaporizer, where the bulk of the propane is vaporized by indirect
steam heating. The remaining propane is removed from the oil in the
oil stripper by steam stripping. Product oil leaves the stripper under
level control. The stripping steam and propane join the asphalt
stripper overhead and are sent to the jet condenser.
Vaporized propane from the flash drum and the propane vapor-
izer are combined and sent directly to the propane condenser, where
the propane is condensed to a liquid by cooling with water. The propane
flows to a surge drum and is pumped back to the process as recycle.
Steam and propane vapor entering the jet condenser are contacted with
water to condense the steam. The water and condensed steam leave
the bottom of the jet condenser and are sewered. The propane vapor
leaves the top of the jet condenser at a pressure that is lower than that
in the extraction section, so a compressor is used to send it to the
propane condenser.
Pollution Sources
There are two potential sources of air pollution in this process
39
-------
that should be given special attention. One, point 1 on the flow dia-
gram, is the -water to sewer from the jet condenser. This stream will
contain small amounts of propane and, depending on the character of
the reduced crude feed, may also have an odor. The other potential
source, point 2, is the water to sewer from the propane surge drum.
This stream is small and may therefore be drained intermittently under
manual control. The water will contain a small amount of dissolved
propane, but more importantly, relatively large amounts of propane
can be lost if the operator is careless about closing the valve after
manually draining the water.
F. CATALYTIC CRACKING
There are two catalytic cracking processes in use today: the
fluid catalytic cracking process which uses a powdered catalyst, and the
Houdriflow or TCC process, no longer in general use, which uses a bead
catalyst. Catalytic cracking is a high-temperature, low-pressure
process which is used to convert gas oil feed stock into fuel gas, liquified
petroleum gases (LPG), high octane gasoline, and distillate fuel.
Feed stocks to the catalytic cracking unit may be gas oils from
the crude unit, thermally cracked gas oils, and/or deasphalted oils.
The products from the reactor are given in Table 3. Catalytic cracking
plants are normally operated to produce a maximum amount of gasoline,
but the units are very flexible and operating conditions can be varied to
produce other products.
40
-------
Table 3. REACTOR PRODUCTS
Product
Wt. %
Water
Gas, C2 and Lighter
Coke
Liquid Hydrocarbons
1 to 2
3 to 7
3 to 8
Remainder
1. The Fluid Catalytic Cracking Process
Figure 11 shows a schematic diagram of a typical fluid catalytic
cracking unit. Gas oil feed is mixed with hot catalyst and introduced .
into the reactor. Steam is added at the base of the reactor to fluidize
the catalyst bed and purge the spent catalyst. The volatile hydrocarbon
products are withdrawn from the top of the reactor and sent to a frac-
tionator where the product streams are separated.
The coke and catalyst are withdrawn from the base of the reactor
and sent to a regenerator. A controlled amount of air is introduced into
the regenerator with the catalyst to burn the coke and reheat the cata-
lyst. The exhaust gas flows through a series of cyclone separators
located inside the regenerator to remove the catalyst dust. An electro-
static precipitator or a third stage cyclone separator, located outside
the regenerator, can be used to remove catalyst fines from the effluent
gas.
The catalyst is continuously circulated from the reactor to the
regenerator. The hot regenerated catalyst is returned to the reactor
through a separate line into which the feed to the reactor is introduced.
41
-------
Jk
pft.OPUC.TS
FLUE GAS
^h
/N A,
/N
/\ A
FUEL GAS
METER
TCP BOILER t_-6FW ppEClPlTATOR
'
STEAM STEAM
LlUfe.
Figure 11. Fluid catalytic cracking unit.
42
-------
Fresh catalyst is added to the system to keep up the activity of
the reaction. Some catalyst is continuously lost to the atmosphere and
constitutes a form of particulate pollution which should be controlled.
The flue gas from the regenerator contains five to ten percent
carbon monoxide and can be burned to yield a considerable amount
of heat energy. This gas can be burned in a boiler, with an auxiliary
fuel, to generate steam. This procedure removes the toxic carbon
monoxide gas from the flue gas and makes it suitable for discharge
to the atmosphere. The boiler is usually monitored with suitable
instrumentation to assure that complete combustion of the carbon mon-
oxide occurs.
Pollution Sources
The catalytic cracking process converts about half of the sulfur
in the feed stock to hydrogen sulfide. Part of this material appears in
the condensate from the fractionator and from the gas plant. The
wastewater is saturated with the gas and must be treated in a sour
water treating process as discussed in Section Q, Sour Water Stripping.
Some of the sulfur stays with the coke and is ultimately burned
to SO? in the regenerator. The effluent gas from the regenerator
(point 1 on the figure) may contain the following contaminants:
Aldehydes Hydrocarbons
Ammonia Oxides of nitrogen
Carbon dioxide Sulfur dioxide
Carbon monoxide Sulfur trioxide
Catalyst fines
The amount of each contaminant will vary with the type of plant and the
43
-------
effectiveness of pollution controls in the plant. Aldehydes, ammonia,
carbon monoxide and hydrocarbons are controlled by combustion in a
CO boiler. Catalyst fines are controlled by an electrostatic precipi-
tator. The remaining pollutants are not controlled at present.
2. Catalytic Cracking with Bead Type Catalysts
Figure 12 is a schematic diagram of a typical moving bed cata-
lytic cracking process. This type of plant can operate with a wide
variety of feed stocks to maximize production of either gasoline or
burning fuels. The process employs a moving bed of hot catalystbeads
•which flow downward from a surge hopper into a reactor where the
beads contact fresh gas oil feed. The gas oil is cracked yielding a
mixture of hydrocarbons known as synthetic crude. The descending
bed of catalyst is purged with steam at the base of the reactor. The
mixture of steam and cracked gases flows from the reactor to a
system of fractionating columns which separate it into fuel gas, LPG,
gasoline, gas oil, and water. The catalyst drops from the reactor
into the regenerator where air is used to burn off the coke which has
deposited on the beads.
Pollution Sources
The flue gas from the regenerator (point 1 in the figure) is
similar to that from a fluid catalytic cracking unit. A precipitator or
cyclone separator can be used to remove catalys-t fines, although the
problem is not as severe as with fluid units. The flue gas can be
burned in a waste heat boiler to eliminate the carbon monoxide.
44
-------
FINES
SEPARATION
PR.&.5M
CATALV5T
Figure 12. Moving bed catalytic cracking unit.
45
-------
Another possible source of particulate pollution is the catalyst
transfer system. The catalyst is lifted from the regenerator outlet to
the catalyst feed hopper with a pneumatic lift. A disengaging drum at
the top of the lift recovers the catalyst, but some catalyst fines may be
discharged to the atmosphere from this source (point 2) if it is not
controlled.
3. Control Strategies
a. CO Boiler:
Excess CO in Flue Gas - Increase air to CO boiler.
Excess SO2 in Flue Gas Reduce sulfur in feed stock.
b. Electrostatic Precipitator
Excess Fines in Flue Gas - Inspect and clean fines
from Fluid Unit removal equipment.
Increase fines withdrawal
from the plant at the pre-
cipitator.
G. ALKYLATION
Alkylation is a process in which an olefin hydrocarbon reacts
with an aromatic or a paraffinic hydrocarbon. An acid catalyst is used
to reduce the temperature and pressure required for the reaction to
proceed. Alkylation units are used to produce high octane gasoline
components and synthetic chemicals such as cumene and ethyl benzene.
The two principal materials currently being used as alkylation
catalysts are sulfuric acid and hydrofluoric (HF) acid. There are
about 140 alkylation units in operation in the United States, with the
number of units of each type being about equal. Both processes operate
46
-------
at moderate temperatures and pressures as shown in Table 4.
Table 4. ALKYLATION PROCESSES OPERATING TEMPERATURES
AND PRESSURES
Process Catalyst
Sulfuric Acid
Hydrofluoric Acid
Temperature
°F
30- 50
75- 105
Pressure
psig
5
125
Both acids are highly corrosive, and the containment systems are
prone to develop leaks.
1. The Sulfuric Acid Process
Figure 13 is a schematic diagram of a typical sulfuric acid
alkylation process. The feed stream could be a mixture of an olefin
such as butene and a paraffin hydrocarbon such as isobutane. The two
streams are mixed and introduced into a horizontal staged reactor A
circulating stream of sulfuric acid and isobutane flows through the
reactor. The reaction of the hydrocarbons is exothermic, and vapor
is withdrawn and condensed to cool the reactor. The hydrocarbon
liquid from the reactor is washed with caustic and water and then
fractionated to separate the alky late product from the isobutane which
is returned to the reactor. The process shown uses autorefrigeration.
It is also possible to cool the reaction by indirect cooling.
The spent acid (point 1 in the figure), which may be shipped off-
site for regeneration, is saturated with volatile hydrocarbons which
47
-------
00
r
I i
I I i i
~l
( ) )
Figure 13. Sulfuric acid alkylation process.
-------
should be contained. The spent caustic (point 2) and water (point 3)
streams may also entrain some hydrocarbon which will vaporize if the
stream is discharged to the atmosphere. These sources should be
confined to prevent atmospheric pollution.
2. The Hydrofluoric Acid Process
Figure 14 is a schematic diagram of a typical HF acid alkylation
plant. The hydrocarbon feed to the plant is mixed with hydrofluoric
acid in a reactor. The mixture is settled and the hydrocarbon product
fractionated to produce recycle isobutane and product alkylate. A
small acid stripping column regenerates the catalyst and gives a by-
product tar which may be burned (point 1 on the figure). A stripping
column is provided to separate an acid rich fraction, which is recycled,
from the LPG product which is washed with caustic and water. The
caustic waste (point 2) and the sour water (point 3) should not be vented
to the atmosphere.
Hydrofluoric acid is a volatile liquid which is toxic and corro-
sive. HF lines may be jacketed to contain any possible leakage, and a
water spray or alkali dump system is usually provided in the event that
a mechanical failure should release any of the acid.
3. Pollution Sources
The principal source of pollution from either type of alkylation
plant during normal operation would result from atmospheric disposal
of spent caustic and water streams (points 2 and 3 on the figures) and
these streams should be confined with vapors vented to flare.
49
-------
f^&e.^
X
u*
1
HP
stn-ippfca
t
Figure 14. HF alkylation process.
U/A5H
f
-------
Hydrocarbon and, more particularly, HF leaks are a potential source of
serious air pollution. Tar from HF alkylation (point 1, Figure 14) is
often disposed of by incineration. This may be a source of pollution if
the burning is not properly controlled.
H. ISOMKRI7.ATION
Isomerization is used to upgrade normal paraffins (straight-
chain hydrocarbons) to isoparaffins (branched chain). The process is
usually applied to butane or to mixtures of pentane and hexane. Where
butane is the feed stock, the isobutane product is normally used as
feed to an alkylation-unit. Pentane-hexane feeds, from crude distilla-
tion or catalytic reforming, are processed to improve their octane
ratings, and the product is blended to gasoline.
Two methods have been used to bring about the isomerization
reaction in petroleum refining. The old method used aluminum
chloride in either vapor or liquid-phase reactions, but this practice is
now obsolete. The new method employs noble metal catalysts on a
solid catalyst support, and the feed is mixed with hydrogen to suppress
unwanted reactions.
Figure 15 is a schematic diagram for a butane isomerization
process. Since the butane feed usually contains a mixture of normal
and isobutane, the mixture is first separated into its normal and iso
components by distillation in the deisobutanizer. The overhead product,
isobutane, is sent to storage or directly to an alkylation unit. The
51
-------
r\j
7E.ISOIbuTAkll2.fcJL.
PR.OPUCT
Jrt ACTOR,
Figure 15. Butane isomerization.
-------
bottom product, normal butane, is mixed with hydrogen and heated by
indirect heat exchange with reactor effluent.
The feed stream is further heated in the fired heater and then
passed over the catalyst in the reactor. Reactor effluent is cooled,
first by exchange with the feed and then in a water-cooled effluent
condenser. The condenser is followed by a separator in which the
liquid and gas phases are separated.
The gas phase from the separator, primarily hydrogen, is
recycled to the reactor by means of the compressor. The liquid phase
contains dissolved hydrogen and other gases, which are removed in the
stabilizer. The stabilizer is a distillation column in which all liquid
reflux is returned to the column. Gas from the stabilizer accumulator
is sent to the fuel gas system. Stabilizer bottoms is a mixture of iso
and normal butane and is, therefore, fed to the deisobutanizer for sepa-
ration into isobutane product and normal butane. Normal butane is
recycled until it is converted to isobutane.
The following procedure is used for regeneration in place. The
unit is shut down, and the reactor is depressurized to the refinery flare
system. Inert gas, supplied by an inert gas generator, is used to free
the reactor of combustible gas. The catalyst is brought to temperature
by recycling inert gas through the heater and reactor, and carbon is
burned off the catalyst by adding a controlled amount of air to the cir-
culating gas stream. Since the gas stream contains products of incom-
plete combustion, it is good practice to incinerate the off gas that is
53
-------
bled from the system to balance the air added. This off gas can be
incinerated in the fired heater.
Pollution Sources
Since the feed to this process must be nearly sulfur free in
order to protect the catalyst, there is no problem with hydrogen sulfide
contaminated gas streams. There is, however, a possible potential
source of air pollution in the gas-to-fuel stream (point 1 in the figure).
In some versions of the isomerization process, particularly with pen-
tane-hexane feed, an organic chloride is added to the feed to increase
catalyst activity. This chloride eventually shows up in the vapor
streams as hydrogen chloride. Most of the hydrogen chloride is
recycled to the process, but some of it is eliminated with the gas-to-
fuel stream. In such cases, this stream should be treated to remove
the hydrogen chloride before the gas is burned as fuel. Sometimes a
caustic scrubber is used for this purpose.
The combustion gases from the reactor feed heater (point 2)
are a source of air pollution (see Chapter II).
Another potential source of air pollution is the off gases from
catalyst regeneration, not shown in Figure 15. The catalyst used in
isomerization is very stable and can be expected to last for two years
or more before regeneration is required. In fact, some refineries do
not regenerate the catalyst, preferring to replace it and return the
spent catalyst to the manufacturer.
54
-------
I. HYDROTREATING
Hydrotreating or hydrodesulfurization processes are used to
remove sulfur from liquid petroleum fractions. Some nitrogen removal
and saturation of olefin bonds may also occur. Sulfur removal is
accomplished by reacting the sulfur containing compounds with hydrogen
in the presence of a catalyst to form hydrogen sulfide. The hydrogen
sulfide is separated by simple vapor-liquid separation.
The processes have been applied to a full range of feeds, from
gasoline to fuel oil. Reaction temperature is normally kept within the
range of 600 to 750°F, while pressures are in the range of 300 to 500
psig for easily treated fractions and from 700 to 1, 000 psig for fractions
requiring more severe treatment.
Figure 16 is a diagram of a typical hydrotreating process. The
hydrocarbon feed is heated in an exchanger and mixed with a hydrogen -
rich gas stream. The mixed feed is heated in a fired heater and passed
through a catalyst bed, where the hydrogen reacts with sulfur and
nitrogen compounds. Reactor effluent is cooled and a small quantity of
water is added to absorb ammonia compounds. The liquids are sepa-
rated from the vapors, and the water phase is withdrawn. The hydro-
carbon liquids are fractionated into separate product streams.
Pollutioii Sources
Possible point sources of pollutant emission are:
The w-ater phase (points 1, 2, and 3 in the figure) contains
55
-------
VTAMi.iz.e-tu
M V PH. & T fZ-fcATl NJ&
01
Figure 16. Hydrotreating.
-------
ammonia and hydrogen sulfide and should be sent to a sour water treat-
ment plant.
The gas phase (point 4) contains hydrogen, methane, and hydro-
gen sulfide and should be treated to remove H2S before being used for
fuel gas.
The catalyst loses its activity due to the accumulation of car-
bonaceous deposits and to the deposition of trace metals. As the unit
continues to operate the pressure drop across the bed builds up and
eventually the process must be shut down. The catalyst may be re-
generated by burning off the carbonaceous material (point 5), or it may
be replaced by new catalyst. With a mild treatment, regeneration may
be required at yearly intervals, but where treatment conditions are
severe and with the older type catalysts, regeneration will be required
at more frequent intervals. (See Section B B, Catalyst Regeneration. )
J. REFORMING
Catalytic reforming is used by refineries to economically up-
grade low octane naphthas to produce premium quality motor fuels,
high yields of aromatic hydrocarbons, high quality aviation gasoline
components, and liquified petroleum gases. The reactions involved in
reforming normally result in the production of hydrogen which is used
either in other refinery processes or in the plant fuel gas system.
A typical reforming process design is shown in Figure 17.
Prior to reforming, the naphtha is hydrotreated for the removal of
57
-------
oo
Figure 17. Reforming.
-------
essentially all the sulfur in the feed. (See Section I, Hydrotreating
Processes. ) Sulfur free naphtha charge and hydrogen rich recycle gas
are heated in a furnace to reactor inlet temperature and then passed
through catalyst beds in a series of reactors. As the major reactions
are endothermic, the gas temperature drops across each reactor and
furnaces are required to reheat the gas between the reactors. The
effluent gas is condensed and separated into a liquid stream and hydro-
gen rich gas. The liquid is processed through the stabilizer and with-
•
drawn as finished reformate. A portion of the gas is recycled and the
remainder used as either a source of hydrogen for other processes or
fuel.
A number of reforming processes are currently in use, but the
basic process is that shown in the figure. The major differences be-
tween the processes are in the composition of the catalyst used in the
reactors and in the methods for regenerating the catalyst. Until
recently, catalyst beds were fixed and were regenerated in place.
However, continuously regenerated beds have now been introduced.
Catalyst regeneration is required because coke is deposited on the
catalyst surface during normal operation. This results in reduced
catalyst activity. During regeneration, the coke is burned off the
catalyst under carefully controlled conditions.
Pollution ^Sources
The gases evolved during regeneration may contain air pollu-
tants and should be incinerated (point 1). (See Section BB, Catalyst
Regeneration. )
59
-------
The fuel gas stream contains hydrogen sulfide and is a potential
source of air pollution (point 2). The fuel gas stream is sent to an
amine unit where hydrogen sulfide is removed.
Combustion gases from the heater (point 3) are another source
of air pollutant emissions.
K. HYDROCRACKING
Hydrocracking is used to convert heavy feed stocks into lighter,
more valuable products. In this cracking process, the feed stock is
converted into shorter chain hydrocarbon molecules in the presence of
hydrogen and a catalyst. The feed stock to the hydrocracking unit is
normally gas oil or middle distillate, and the usual products are fuel
gas, gasoline, and jet fuel. The process is flexible in that the pro-
duction of either gasoline or jet fuel can be maximized as needed.
Hydrocracking employs high pressure (1, 500 to 3, 000 psi),
high temperatures (500 to 750T), and a special catalyst. The reac-
tion section is usually divided into two stages. The first stage is
used to remove all the nitrogen and sulfur (see Hydrotreating), while
the second stage is used for cracking. The first stage is needed in
most cases because sulfur and nitrogen are catalyst poisons for the
cracking catalyst used in the second stage. However, a single-stage
reaction section is sometimes used where the feed stock is low in
sulfur and nitrogen. Single-stage catalysts are more sulfur-resistant
than those used in the second stage of a two-stage unit.
60
-------
Figures 18 and 19 are flow diagrams of a reaction section and
a distillation section of a typical hydrocracking unit. The feed is pre-
treated and cracked in the reaction section, and products are recovered
in the distillation section.
Gas oil and/or distillate feed containing some sulfur and nitrogen
impurities is pumped from the feed drum and heated by exchange with
the first reactor product. Hydrogen is added to the feed and additional
heat is supplied by a fired heater. The feed enters the reactor at about
3, 000 psi and 700 °F. The reaction is exothermic (heat producing) and
the temperature is maintained by cooling the reactor catalyst beds by
adding cold hydrogen. The reaction products are cooled, water is in-
jected into the product stream and the reaction products and water
stream enter the high pressure separator.
At this point all of the sulfur in the feed has been converted to
H?S and all the nitrogen in the feed to NH3. The approximate distribu-
tion of sulfur (H2S) and nitrogen among the streams leaving the high-
pressure separator is shown in table 5.
Table 5. SULFUR AND NITROGEN DISTRIBUTION IN THE HIGH
PRESSURE SEPARATOR
Product
Hydrogen rich gas
Hydrocarbon liquid
Water
Total
Wt. % Feed
Sulfur
50
40
10
100
Nitrogen
Nil
5
95
100
61
-------
I ST 6TA6 fe-
ll &.A£.TOn_
TO A
HEAT fl££0V£A.Y
I—&
Figure 18. Hydrocracking reaction section.
-------
FR-OM IVT 4-
-S
b 6-f*A 11. A TO u. »>
U)
M;
Figure 19. Hydrocracking distillation section.
-------
The hydrocarbon liquid from the first-stage high-pressure
separator is combined with the liquid from the second-stage high-
pressure separator and is fed to the stabilizer in the distillation section
to separate products from unreacted feed. The hydrogen rich gas from
the first-stage high pressure separator is recycled back to the reactor
with the addition of sufficient makeup hydrogen to replace that con-
sumed by the reaction. Hydrogen purity is maintained by sending a
bleed stream to a H2S absorber.
Fractionator bottom liquid from the distillation section is un-
reacted feed. This stream is fed to the second reaction stage. Pro-
cess flows in the second reaction stage are the same as in the first
reaction stage except that no sulfur or nitrogen is present in the feed.
Cracking occurs at 1, 500 psi and 600° F.
The distillation section is similar to those used in other crack-
ing processes such as fluid catalytic cracking and thermal cracking.
The total flow from both high pressure separators is fed to the low
pressure separator in the distillation section. The pressure is
reduced and fuel gas is flashed out of the hydrocarbon fluid.
The low-pressure separator liquid is fed to the stabilizer. In
the stabilizer, all the light ends and the H?S are stripped out and are
taken overhead. Both the gas and the light ends (overhead) liquid have
to be treated for the removal of H2S. The stabilizer bottoms are
separated into products and fractionator bottoms in the fractionator and
the fractionator bottoms are recycled to the second reaction stage.
64
-------
The approximate distribution of H^S in the streams leaving the
low-pressure separator is as follows:
Table 6. H?S DISTRIBUTION IN THE LOW-PRESSURE SEPARATOR
Phase
Liquid
Vapor
Total
Wt. % Feed Sulfur
85
15
100
Pollution Sources
Point sources for possible pollutant emission in a hydrocracking
unit are:
1. Sour water containing H2S and NH3 leaves the first stage
high pressure separator, the low pressure separator, and the stabilizer
accumulator. All these streams have to be treated in a sour water
stripper before being discharged or reused (point 1 on Figure 18 and
points 1 and 2 on Figure 19).
2. The bleed gas from the hydrogen rich gas in the first
reaction stage may contain as much as 0. 5 vol. % to 1. 5 vol. % H2S on
Figure 1 8.
3. The catalyst is periodically regenerated in situ by
burning off the accumulated coke (points 3 and 4 in Figure 18). (See
Section BB, Catalyst Regeneration. )
4. Fuel gas produced in the distillation section contains H2S
65
-------
and needs to be treated. Points 3 and 4 on Figure 19.
5. Light ends produced in the distillation section contain
some H2S and need to be treated (point 5 on Figure 19).
L. HYDROGEN PRODUCTION
Hydrogen is an intermediate material in refining operations. It
is used as a reactant in operations such as hydrogenation, hydrodesul-
furization, and hydrocracking. The principal method for producing
hydrogen is steam reforming of some available hydrocarbon like
natural gas, refinery gas, propane, butane, or naphtha. Hydrogen is
also produced in the refinery as a by-product in the reforming process.
The sequence of processing steps for hydrogen production by
steam reforming is sulfur removal, reforming, shift conversion, car-
bon dioxide absorption in monoethanolamine (MEA) solution and meth-
anation. The plant is in steam balance, that is, high-pressure steam
for reforming and low-pressure steam for MEA regeneration and de-
aeration is produced as a by-product. The flow diagram is shown on
Figure 20. The gas feed to the plant normally contains traces of sulfur
which are removed by adsorption on activated carbon. Two carbon
beds are used, normally operating in series flow. During regeneration,
one bed is valved out of the normal flow and returned after regeneration
to the downstream position in the flow sequence.
The sulfur-free gas is mixed with high-pressure steam and
pre-heated in the convection section of the reformer furnace. The
66
-------
NA7V1AI-
OH.
ri
FJ
t>T<^
HtAPtfl.
>> t
I
X
Me.i»A>-jATg>r>_
X
X
JHIPT
X
' — (
*i-V ' „
?v;
i ' '
i ; '
' : '
1 , 1
1
1
~xy
I
HI&H Ptt£SvVT?\
iTM . PdOr<-| /\by
-N
^3-^
R.£pO/VM6fl.
; fUfX>4ACfc_
@-^-
\
i.
1 Fufci.'
— — , •(
I
auc^cooM-j
j
— L_ C-TFA
H/AVrtMCAT \ —
&o't-tn. y , L
Figure 20. Hydrogen production by steam reforming.
-------
mixed gas flows downward through catalyst-filled tubes where steam
reacts with methane and other hydrocarbons to produce hydrogen, car-
bon monoxide, and carbon dioxide. The high-temperature effluent gas
from the reformer furnace flows through the tubeside of a steam gener-
ator producing high-pressure steam. Additional high-pressure steam
is generated in the convection section of the furnace. This steam is
consumed in the reforming reaction. More than three-quarters of the
total hydrogen is produced in the reforming reaction. The remaining
hydrogen is produced by the shift conversion of carbon monoxide to
carbon dioxide. The catalytic conversion occurs in two stages, one at
high temperature and the second at low temperature. The combination
of converters shifts over 98 percent of the total carbon oxides to car-
bon dioxide and hydrogen. Reaction heat is removed between the two
stages of shift conversion by generating low pressure steam. The gas
temperature between the catalyst beds is also lowered to give more
.favorable conditions for the shift reaction.
The crude hydrogen gas from the shift converter is further
cooled before mixing with MEA solution in an exchanger. The heat of
reaction between carbon dioxide and MEA is largely removed in this
precontact stage ahead of the amine absorber. The remaining carbon
dioxide is removed from the gas stream by reaction in the absorber.
The heat for MEA regeneration is supplied partly from the process
gas and partly from low pressure steam. An MEA reclaimer is pro-
vided to maintain a clean active solution. Other means of absorbing
68
-------
carbon dioxide may be used instead of MEA; these include Sulfinol and
activated hot carbonate.
The last traces of carbon monoxide and carbon dioxide are con-
verted to methane by reaction with hydrogen in the methanator. This
catalytic reaction requires a preheat temperature of about 600° F. The
residual carbon oxide content of the product hydrogen is less than
10 ppm.
Pollution Sources
There are two potential sources of pollution in the steam re-
forming process. During carbon tower regeneration (point 1 in the
figure), H2S and organic sulfur compounds are removed using steam.
This steam should be condensed and transferred to the sour water
system for processing.
The periodic replacement of catalyst beds could result in pollu-
tion if the removed catalyst is allowed to weather in an open area (point
Z). The spent catalyst should be put in containers when removed from
process equipment and disposed of in an acceptable manner.
M. SWEETENING
A distillate is said to be sour if it contains noticeable amounts
of sulfur compounds, particularly the odoriferous mercaptans. A pro-
cess that removes these compounds or converts them to less objection-
able forms is called "sweetening". Hydrotreating, which could be
called a sweetening process is discussed in Section I.
69
-------
Sweetening can be accomplished by removing the mercaptan,
usually by extraction or by converting it to a disulfide. Frequently, the
solutions used for extraction are regenerated by converting themercap-
tans in them to disulfides and then removing the disulfides. In treating
light distillates or light naphtha, lighter mercaptans can be satisfac-
torily removed by extraction with caustic solutions containing solubility
promoters. The high molecular weight mercaptans associated with the
heavier distillates, such as'full-range gasoline or kerosene, are more
difficult to extract and are normally converted to disulfides and left in
the distillate.
Doctor sweetening, copper sweetening, and hypochlorite sweet-
ening are old processes and are not in general use at this time. In
doctor sweetening, the distillate is treated with alkaline sodium plum-
bite solution to oxidize the mercaptan to disulfide. The lead is reduced
to lead sulfide and is discarded or regenerated by air blowing. In
copper sweetening, cupric chloride is reduced to cuprous chloride,
followed by regeneration with air to recover the cupric chloride. In
hypochlorite sweetening, sodium or calcium hypochlorite is used as
the oxidizing agent without regeneration.
There is a variety of sweetening processes in use today, for
instance, treatment with sulfuric acid or absorbing mercaptan with
molecular sieves. However, the more widely used processes usually
employ sodium hydroxide with added catalysts or promoters. Most
70
-------
frequently a caustic solution containing the dissolved catalyst or pro-
moter is employed, but a fixed bed of catalyst can also be used.
Figure 21 is a flow diagram for a gasoline-sweetening process
that employs a sodium hydroxide solution (caustic) containing a dis-
solved catalyst. Sour gasoline feed is metered into the extractor, where
it is brought into contact with recycled regenerated caustic solution.
Partially treated gasoline, with part of the mercaptan removed, flows
from the top of the extractor and is mixed with recycled caustic solution
and air before entering the bottom of the sweetener. The remaining
mercaptan is oxidized to disulfide in the sweetener and remains in the
treated gasoline stream. Caustic solution is separated from the treated
gasoline in the solution settler and is recycled to the sweetener. Caustic
solution from the extractor, containing dissolved mercaptan, is mixed
with air and sent to the oxidizer. The mercaptan is oxidized to disulfide
in the oxidizer and then flows to the air separator. Excess air is vented
from the air separator, and the caustic solution and disulfide flow to the
disulfide separator. The insoluble disulfide layer separates and is
withdrawn from the system, and the regenerated caustic is recycled to
the extractor.
Pollution Sources
There are two points of possible pollutant emission of special
interest in the process. One is the disulfide product stream (point 1 in
the figure). If the disulfide cannot be sold, it will normally be disposed
of by burning as fuel or by incineration. Since the stream is small,
71
-------
(W) •
SOUR,
Aili.
CAU5TIC SOLUTION
-D- - -,
AIR.
r -.
7
o
t-
o
o
£
3
O
7I2.E1AT&-D
&XTRACTQE.
Figure 21. Gasoline sweetening.
-------
this may be permissible, but in some cases it may be necessary to use
an incinerator equipped for sulfur dioxide recovery. Incinerator emis-
sions may be the subject of future control regulations. The other pol-
lution emission is the excess air from the air separator (point 2).
Since this air will contain disulfide, it may be necessary to incinerate
it.
N. ASPHALT AIR BLOWING
Asphalt used for composition roofing and shingles is usually
blown with air to oxidize the material. The oxidation reaction increases
the hardness and raises the melting point of the asphaltic material and
improves its resistance to weathering. In some processes a catalyst
such as ferric chloride or phosphorous pentoxide is added to the asphalt
to produce a product with a very high melting point and hardness.
Figure 22 is a typical schematic diagram of an air-blowing
asphalt-treating facility. The feed to the asphalt plant will usually be a
residuum from a vacuum distillation of a topped crude. The reaction
is usually carried out in a batch operation. The feed is preheated to
400 to 600" F in a fired heater and then-pumped into the reactor. The
reactor is usually a vertical vessel which is partially filled with feed
stock. Air is compressed to a pressure high enough to permit dis-
charging it into the residuum through a sparger at the base of the
reactor. As the gases rise through the liquid, an exothermal reaction
takes place between the oxygen and the hydrogen in the oil. Simulta-
neous chemical reactions cause polymerization and formation of oxygen
73
-------
&.
FUM
_pUfcL.
AAS
A1/2.
Figure 22. Asphalt air blowing.
-------
linkages. The blowing reaction may be continued for periods of from
one to 24 hours.
Pollution Sources
The gases which are vented from the reactor contain hydrocar-
bon vapors and aerosol particles of oil. These gases used to constitute
one of the major forms of air pollution from a refinery. Now the vent
gases are normally incinerated which eliminates the objectionable
constituents in the vent gas. However, the incinerator remains a
potential source of air pollution (point 1 on the figure).
O. ACID GAS TREATING
Hydrogen sulfide (H2S) and carbon dioxide (CO2) are called acid
gases, and a gas stream containing H2S is called sour gas. Since sour
gas is produced in a number of processes during normal refinery opera-
tion, untreated refinery fuel gas can be expected to be sour. It is gen-
erally necessary to treat the gas for H2S removal before it is used as
refinery fuel. This is done to avoid the air pollution that results from
the sulfur dioxide (SOE) formed in burning the H2S.
Acid gases are removed by absorbing them in an alkaline solu-
tion. The alkaline material, monoethanolamine for instance, is chosen
so that the chemical bond formed during absorption can be broken by
heating. The acid gas is stripped from the heated solution and the
solution, after cooling, is ready for reuse. Solution containing acid
gas is called "rich" and the regenerated solution is called "lean".
75
-------
Several acid gas treating processes are available but the differ-
ences are primarily in the choice of alkaline absorbent. The processes
are similar in that the acid gas is absorbed in the alkaline solution
under pressure, and the solution is regenerated by heating at a low
pressure.
1. Acid Gas^ Absorption Processes
a. Absorption in Monoethanolamine (MEA) The absorbing
medium is a 10 to 20 weight percent solution of MEA in water.
Figure 23 shows the process flow for a typical MEA absorption
system. Acid gas is removed from the sour gas by contacting
with the MEA solution in the absorber. Sour gas enters the bot-
tom of the column and cool, lean amine enters at the top.
Treated gas leaves the top of the absorber and passes to the re-
finery fuel gas system. Rich amine, containing the absorbed
acid gas, is used to cool the lean amine and is fed to the top of
the stripper. Steam used for stripping the rich amine is gener-
ated by boiling the stripper bottoms in the reboiler. Acid gas
and steam leave the top of the column and steam is condensed.
Condensate and acid gas are separated in the acid gas separator
and the condensate is pumped to the stripper as reflux. The
acid gas flows to the sulfur plant for conversion to elemental
sulfur. This stream is a potential source of air pollution (point 1
in the figure). Hot lean amine from the stripper reboiler is
cooled and filtered before returning to the absorber.
76
-------
-OM*S
v—/ MONITOR
SUMP
Figure 23. Acid gas treating.
-------
b. Absorption in Diethanolamine (PEA) - The absorbing
medium is a 20 to 30 weight percent solution of DEA in water.
The process flow scheme is identical to.that shown for MEA in
the figure.
c. Absorption in Hot Carbonate - The absorbing medium is
a 15 to 30 weight percent solution of potassium carbonate in
water. There are a number of variations to this process em-
ploying additives to improve solution performance. The major
equipment and process flow are again similar to the MEA
process.
The treated gas (point 2 in the figure) may be a source of air
pollution if H2S removal is incomplete. The H2S content of this stream
should be checked periodically, if not monitored continuously.
P. SULFUR RECOVERY
1. Process Objective
The sulfur recovery process, also known as the Glaus process,
is used to convert hydrogen sulfide to elemental sulfur. The feed
stream contains acid gases (CO2 and H?S) obtained from the acid-gas
recovery plant, but hydrocarbon impurities may also be present. The
sulfur plant is normally designed to convert 90 to 95% of the H2S to
elemental sulfur, which requires two or three reaction stages, with the
unrecovered sulfur being burned to SO2 and vented to the atmosphere.
However, recently implemented pollution regulations have, in some
areas, required higher degrees of recovery. Tail gas cleanup processes
78
-------
have been developed for this purpose that reduce the SO2 emissions
from the sulfur plants.
2. Process Description
Hydrogen sulfide is converted to elemental sulfur in two steps.
In the first step, H2S is partially burned to SO2 with air. The H2S/SO2
mixture is then reacted over a catalyst to produce sulfur and water.
This reaction is known as the shift conversion and is carried out in
two or three stages with sulfur removal after each stage. The design
of a sulfur recovery plant depends upon the acid gas composition. If
the concentration of H2S in the feed is high, a "straight-through" pro-
cess is used. In the straight-through configuration, all of the acid gas
and air are fed to the burner. If the H2S concentration in the feed is
low, a "split-flow" or "sulfur recycle" process is used. In the "split-
flow" process, a portion of the feed is burned completely to SO2 and
combined with the remainder of the feed to provide the proper H2S/SO2
ratio for the shift conversion. In the "sulfur recycle" process, the
product sulfur is recycled to the burner to support combustion. A
fourth type of sulfur recovery process uses the "direct oxidation"
approach. This configuration, which is for very lean feeds, eliminates
the burner by feeding the acid gas/air mixture directly to a catalytic
burner.
Most sulfur plants in refineries are the "straight-through" or
"split-flow'" type. A flow diagram for a typical refinery sulfur recovery
process is shown in Figure 24. The acid-gas stream containing H2S,
79
-------
CD
o
Figure 24. Sulfur Plant.
-------
CO2, water, and minor amounts of hydrocarbons is fed to an inlet
separator where any entrained liquid is removed. The acid gas and
air are fed to the sulfur boiler. Fuel gas lines are provided to assist
in plant startup. Boiler feed water is fed to the sulfur boiler to gener-
ate low-pressure steam. The sulfur boiler usually contains three tube
passes. A portion of the gases is diverted from the boiler after two
passes to provide preheat for the shift converter feeds. In someplants
an auxiliary burner is provided to furnish preheat to the reactors.
Liquid sulfur is separated from the boiler effluent gases and is sent to
the sulfur pit for storage. The gases are mixed with the reheat stream
and fed to the first-stage shift converter. The effluent from the reac-
tor is passed through the condenser. Sulfur is separated and sent to
the sulfur pit and the gases are fed to the second-stage converter.
Approximately 5% of the sulfur fed remains in the tail gas. This sulfur
is converted to SO2 in the tail-gas incinerator.
In some sulfur plants, the sulfur boiler and condenser are com-
bined into a single unit. The catalytic converters are often combined
into a single horizontal or vertical vessel. Combining the units in this
manner permits many smaller plants to be shop fabricated and skid-
mounted.
3. Instrumentation
The acid-gas feed to the sulfur plant is on pressure control.
The plant-is designed to accept acid-gas flows as they develop in the
refinery and a plant may be designed to operate at 25% of capacity to
81
-------
allow for future demand. The ratio of acid gas to air is controlled by
a special ratio flow controller. To set the proper ratio, the acid gas
must be analyzed regularly to determine the H2S concentration. The
proper H2S/air ratio is critical to the operation of the plant since the
concentration of SO2 in the incinerator stack gases will increase if the
improper ratio is set. Automatic stream analyzers can be used to set
the proper ratio on a nearly continuous basis. The volume of reheat
streams is controlled to set the inlet temperatures to the shift con-
verters. If these temperatures are allowed to fall, the reaction will be
incomplete and sulfur recovery will drop. Incinerator stack tempera-
ture is controlled by setting fuel consumption and air bypass. A high
temperature is required to ensure complete combustion of sulfur com-
pounds to SO2 and good dispersal of gas to the atmosphere.
4. Pollution Sources
The stack gas emissions (point 1 in the figure) are the major
source of pollution from sulfur plants. The concentration of SO2 in the
stack gas will depend on the number of stages, feed gas composition,
operating efficiency, amount of dilution air and plant upsets. In a typical
plant with two reactor stages, the stack gas will contain up to 10% of
the feed sulfur. The presence of hydrocarbons in the feed will be
detrimental to sulfur recovery and pollutant emissions. Some of the
hydrocarbons are converted to carbon disulfide and carbonyl sulfide in
the burner These compounds are essentially inert and pass through to
the incinerator where they are emitted as SO,. Hydrocarbons can also
82
-------
foul the shift conversion catalyst and reduce sulfur recovery.
5. Tail-Gas Treatment Processes
Tighter air pollution regulations have forced many sulfur plant
operations to consider "hang-on" plants to reduce SO2 emissions.
Numerous processes have been developed to treat sulfur plant tail gas.
The more important tail-gas treating processes are discussed below.
a. Modified Stretford Process - The Stretford Process has
been used in Great Britain for many years to recover hydrogen
sulfide from natural gas and convert it to sulfur. The feed gas
is passed through an absorption tower which removes the H2S.
The absorbent is an organic liquid which also serves to oxidize
the dissolved H2S to sulfur. The sulfur is removed from the
liquid by filtration, and the solvent is regenerated by air oxida-
tion. Very high conversions of H2S to sulfur are possible with
this process.
Since the Stretford Process is not suitable for use with
feed gases containing SO2, two modified versions have been de-
veloped for use with sulfur plant tail gas. The Beavon Process
uses a small catalytic reactor to hydrogenate the SO2, COS, and
CS? to H^S before feeding the gas to a Stretford tower. The
Cleanair Process hydrolyzes the impurities to H2S and SO2.
The gas is then cooled to permit the Claus reaction to take place.
At low temperatures the conversion of SO2 to sulfur is complete
and the remaining H2S is fed to a Stretford Tower. Adding a
83
-------
Modified Stretford unit can reduce the incinerator stack concen-
tration of SO 2 to 250 pprn.
b. Solution Glaus Process - The Claus reaction between
H?S and SO2 has a higher sulfur yield at low temperatures. In
the IFF Process (developed by the Institut Francais de Petrole)
H2S and SO2 react in stoichiometric amounts in a liquid-phase
solvent. The temperature is kept above the melting point of
sulfur so that molten sulfur can be drained from the bottom of
the tower. The solvent is cooled and recycled to the tower.
Treated gas is incinerated as in other processes. The IFP
Process can reduce SO2 in the stack gas to 2, 000 ppm.
c. Sulfreen Process - This process uses a low-tempera-
ture Claus reaction on an activated carbon catalyst. The sulfur
produced is adsorbed on the carbon and is removed by stripping
with inert gas in a second step. Reduction of SO2 in the stack
gas is comparable to that of the IFP Process.
d. Stack-Gas Treating Processes Another approach to
reducing stack-gas emissions is to treat the stack gas rather
than the tail gas. Several schemes have been proposed to re-
move or recover SO2 from stack gases. One such process that
has received some attention is the Wellman-SO2 Recovery Pro-
cess. The stack gases are cooled and fed to a tower where SO2
is absorbed in sodium sulfite solution to form sodium bisulfite.
The sodium bisulfite is heated to drive off the SO2 and the sulfite
84
-------
is crystallized, redissolved, and recycled to the tower. The
SO2 is recycled to the sulfur plant. Effluent gas concentrations
may be reduced to 100 ppm SO2.
Q. SOUR WATER STRIPPING
Water containing sulfides is called sour water. Refinery opera-
tions produce sour water from processes such as steam stripping and
whenever steam is condensed in the presence of gases containing hydro-
gen sulfide. Sour water may also contain ammonia and phenols. Steam
stripping has been used by refineries to reduce the level of contaminants
in sour condensate to allow either further use of this condensate or its
return to public waters. There are two types of sour water steam
strippers: refluxed and nonrefluxed.
1. Refluxed Sour Water Steam Strippers
Figure 25 shows the process flow for a typical refluxed stripper.
Sour water feed may be flashed to release some vapor and then stored
in a sxirge tank. The sour water is then pumped through a preheat ex-
changer and into the top of the stripper column. Steam is fed into the
bottom of the column. Sour gas, containing steam and contaminants,
leaves the top of the stripper and is partially condensed. Condensate
and sour gas are separated in the surge tank and the condensate is re-
cycled to the stripper. The sour gas is disposed of either by conver-
sion to sulfur in a sulfur plant or by incineration, where air pollution
control regulations allow the emission of the sulfur dioxide resulting
85
-------
00
Figure 25. Sour water stripping process.
-------
from incineration. The stripped water is removed from the bottom of
the column and either is used as a process water stream or flows to the
waste water treatment system.
The pollution potential of the sour gas stream (point 1 in the
figure) is apparent because of its H2S content. Further processing of
this stream is necessary before emission to the atmosphere. The
stripped water (point 2) may be a source of odor problems if traces of
dissolved H2S remain and the stream flows to an open waste water
treatment system.
2. Nonrefluxed Sour Water Steam Strippers
In these systems there is no overhead condensing system and
the column overhead goes directly to either flare or incineration. This
stream has proved hard to dispose of because of the large volume of
steam and is a potential source of pollution.
R. NATURAL GAS PROCESSING
Natural gas processing refers to processes used to remove
impurities and to recover heavier hydrocarbons from natural gas.
Natural gas is a mixture of methane and ethane that occurs in
nature as deposits trapped in geologic structxires. The gas may occur
in combination with crude oil or alone and at pressures varying from
hundreds of pounds per square inch to substantial vacuums. Usually
water or brine is produced along with the gas, and the gas can be ex-
pected to contain nonhydrocarbon impurities and heavier hydrocarbons.
87
-------
Water, brine, and crude oil where present are usually sepa-
rated from the gas at the production location. The field separators or
gas traps, used for this purpose, are simple physical separation de-
vices. The separated gas can be expected to be saturated with water
vapor and will frequently contain heavier hydrocarbon vapors, carbon
dioxide and hydrogen sulfide.
Mixtures of water vapor and natural gas, at elevated pressures
and low temperatures, can form gas hydrates. These icelike solids
plug lines and otherwise interfere with transport and storage equip-
ment. Water vapor, carbon dioxide, and hydrogen sulfide are corro-
sive and hydrogen sulfide is a potential air pollutant. For these
reasons it is frequently desirable to remove such impurities from
natural gas.
The heavier hydrocarbon vapors are often separated to recover
LPG and natural gasoline which are valuable products. LPG is pro-
pane, butane, or mixtures of both. Natural gasoline is a component of
motor gasoline consisting of mixtures of butanes, pentanes, hexanes,
and lesser amounts of still heavier hydrocarbons.
1. Gas Dehydration Using Liquid Absorbent
Natural gas is dehydrated by using either absorption in a hygro-
scopic liquid or adsorption on a solid desiccant. In processes employ-
ing liquid absorbents, the liquid is continuously regenerated and
recycled. Solid adsorbents are usually regenerated periodically, with
two or more vessels used to provide continuous operation.
88
-------
Diethylene glycol, triethyiene glycol, and calcium chloride
brine are commonly used liquid absorbents. The glycol dehydration
process which is typical of the processes using absorbents is shown in
Figure 26. Gas is brought into the system through an inlet scrubber
to remove any entrained liquid water or hydrocarbon. The gas is then
dried by countercurrent contact with the absorbent in the absorber.
Dehydrated gas leaves the system from the top of the absorber and the
absorbent containing water leaves from the bottom. Since the absorber
is normally operated at pressures of several hundred pounds per
square inch, some gas will be dissolved in the absorbent. This gas is
separated in a flash vessel at reduced pressure and delivered to the
fuel gas system. The absorption liquid is then fed to a distillation
column, or still, for regeneration. Water is distilled overhead, along
with a minor amount of gas which is sent to the flare. The regenerated
absorbent is recycled to the absorber after cooling by exchange with
the feed stream and cooling water.
Water from the inlet scrubber (point 1 in the figure) and the
still overhead (point 2) may contain sulfides. If so, these streams
should be routed to a sour water stripping system. Flash tank gas
(point 3) may contain hydrogen sulfide and may require treatment
before being used for fuel.
2. Gas Dehydration Using Solid Adsorbent
Alumina, silica gel, and molecular sieves are three commonly
used solid adsorbents. Figure27is a flow sheet for atypical adsorbent
89
-------
OVERHEAD
ACCUMULATOR
Figure 26. Glycol dehydration.
-------
iHLtT
^
Vk/AT1^!-
_
I
i,
r-7
n
n
L Al
r*
i
Figure 27. Adsorbent dehydration.
•«•
-------
dehydration process. Since the adsorbent is regenerated in place, two
dehydration vessels are provided to permit continuous operation. One
vessel is in service while the other is being regenerated. In the follow-
ing description, it is assumed that one desiccant tower is in service
while the other tower is being regenerated. Gas is brought into the
system through an inlet scrubber to remove any entrained liquid water.
The main flow, to the # 1 desiccant tower, is controlled by a flow-
control valve taking its signal from, a flow sensor in the bypass used
for regeneration. Gas flows downward through the tower and dehydra-
ted gas leaves the process from the bottom of the tower.
The rr 2 desiccant tower is regenerated while the # 1 tower is on
stream. A bypass stream from the main gas flow is heated and then
passed through the # 2 tower. Gas and water vapor from the tower are
cooled to condense the water. The water is separated from the gas in
the condensate separator and the gas is returned to the main gas
stream. After regeneration, the desiccant bed is cooled by bypassing
the heater and passing cool gas through the tower.
Depending on the pressure of operation and on the amount of
hydrogen sulfide in the gas, it may be desirable to treat the water re-
moved from the process to control dissolved hydrocarbon and sulfide
emissions (points 1 and 2).
3. Acid Gas Removal
The processes applied to natural gas are essentially the same
as those covered under Section O, Acid Gas Treating. There are,
92
-------
however, some differences attributable to the usually much higher
pressures of the gas to be treated.
Glycol is often added to the amine in an amine process to pro-
vide simultaneous dehydration and acid gas removal.
In amine plants, a flash tank and reabsorber are normally added
to the rich amine circuit between the absorber and the stripper. This
arrangement is shown in Figure 28. Natural gas absorbed at the high
absorber pressures is released at a lower pressure in the flash tank.
Since the flashed gas is sour, a small reabsorber column is mounted
on the flash tank. A slip stream of lean amine is fed to the reabsorber
to remove the acid gas from the flashed gas before the gas is sent to
the fuel gas system.
Since the absorber pressure is much higher than the stripper
pressure, a large amount of power is required to pump lean treating
solution from the stripper to the absorber. Some of this power can be
supplied by utilizing the pressure difference available in the rich solu-
tion circuit. A pressure breakdown turbine is inserted in the rich
solution circuit and the turbine is used to drive the lean solution pump.
A motor or steam turbine is used to supply the balance of the power
required.
4. LPG and Natural Gasoline Recovery by Compression
Natural gas is often transported through high pressure pipelines
as a matter of economy. Where the gas is produced at low pressure,
the gas must first be compressed. Although natural gas is seldom
93
-------
on.
Figure 28. Amine-water and glycol amine process using breakdown turbine.
-------
compressed solely for the purpose of LPG or natural gasoline recovery,
significant amounts of these products are recovered from compressor
stations. Under pressure, the heavy hydrocarbons are condensed and
separated from the natural gas. Since the increase in pressure per
stage, expressed as the ratio of outlet to inlet pressure, is limited by
practical considerations, several stages of compression may be needed
to reach the required pressure.
Figure 29 is a flow sheet for a typical two-stage compressor
station. Gas enters through an inlet scrubber or knock out drum to
remove entrained liquid. The gas is compressed in the first stage
cylinder, cooled by a cooling water exchanger and sent to the first stage
accumulator. Water and hydrocarbon are separated from the gas under
liquid level and interface level control. The hydrocarbon is sent to a
distillation unit for recovery of LPG and natural gasoline (see Section S,
Light Ends Recovery). The gas is then compressed in the second
stage in a similar manner. The first and second stage cylinders are
usually driven by a single engine or motor.
The water streams (points 1, 2, and 3 in the figure) may contain
sulfides and hydrocarbons. If so, they should be routed to separators
and a sour water stripper.
5. LPG and Natural Gasoline Recovery by Refrigeration
The amount of heavy hydrocarbon vapor that can be held at satu-
ration by natural gas decreases with decreasing temperature or increas-
ing pressure. Increased recovery of LPG and natural gasoline can be
95
-------
Figure 29. Two-stage gas compressor.
-------
achieved in a compressor plant if refrigeration is used in place of
cooling water in the compressed gas coolers.
6. LPG and Natural Gasoline Recovery by Oil Absorption
The absorption process for recovery of LPG and natural gaso-
line is more complex and generally more expensive than the compres-
sion and refrigeration process, but it is also more efficient. An
absorption oil, usually in the kerosene range, is used to absorb the
heavy hydrocarbon vapor under pressure. The hydrocarbon product is
recovered from the absorption oil by distillation.
Figure 30 is a flow diagram for a typical oil absorption process.
Gas is fed to the bottom of the main absorber where it is brought into
countercurrent contact at relatively high pressure with cold lean
absorption oil. Treated gas leaves the system from the top of the
main absorber and rich absorption oil flows from the bottom to a flash
tank. Enough product vapor is flashed along with the dissolved natural
gas to warrant the use of a reabsorber column. Flashed gas and gas
from the still are combined and fed to the reabsorber where they are
brought into contact with a slip stream of lean absorption oil at low
pressure. Overhead gas from the reabsorber goes to fuel. Rich oil
from the reabsorber and the flash tank are combined and heated by
exchange with still bottoms. The combined stream is heated by ex-
change with steam or hot gas oil and is sent to the still for regeneration
by distillation. A small amount of steam is injected directly into the
still to help strip the lean absorption oil. Most of the resulting
97
-------
oo
TREATED
OUT
Figure 30. Oil absorption plant.
-------
condensate is removed from the top trays of the still by the use of a
draw off tray. Product hydrocarbon from, the reflux accumulator is
sent to a fractionater for distillation into LPG and natural gasoline.
Hot lean absorption oil from the bottom of the still is cooled by feed-
bottoms exchange and by cooling water exchange and is recycled to the
absorbers.
Water entering the process in the feed gas and stripping steam
is condensed in the system and leaves in streams (points 1-5 in the
figure). These streams may contain sulfides. If so, they should be
routed to a sour water stripper. Gas from the reabsorber (point 6)
may contain hydrogen sulfide and may therefore require treatment
before using as fuel.
Some absorption plants are operated at high pressure, on the
order of 1, 000 psig. In such cases, it may be desirable to employ two
stills for absorption oil stripping. One still is operated at high pressure
and one at low pressure. The advantage of this procedure is that better
recoveries of product and cleaner separations are possible.
S. LIGHT ENDS RECOVERY
Refinery light ends are usually hydrocarbon compounds having
four or less carbon atoms, including methane, ethane, propane, butane,
and isobutane. The objective of a light ends recoveryunit is to separate
these components into saleable products and fuel gas. In small refin-
eries, or in refineries where little or no cracking processes take place,
99
-------
light end recovery units may not be economical and most of the light
ends are used as fuel. Larger refineries, on the other hand, may have
more than one light ends recovery system, perhaps located in different
sections of the refinery.
Normally, the ethane and methane go into the fuel gas system.
Propane is separated as a product because of its value as a feed stock
in the petrochemical industry. The butanes are either added to the
gasoline pool, or separated to isobutane and normal butane andused as
feed for other units in the refinery (see Section G, Alkylation Process).
The process configuration and recovery of light ends will vary with the
particular needs of the refinery. A typical unit is presented in Figure 31.
The feed to the recovery unit is unstabilized naphtha (a gasoline
cut containing dissolved light ends) and a gas stream rich with propane
and butanes. The naphtha feed is stabilized in the stabilizer by removing
a portion of the light ends. Some butanes are left in the stabilizer bot-
toms product for maintaining the proper naphtha vapor pressure. The
stabilizer overhead product is combined with other refinery gases and
fed to a deethanizer. Methane and ethane are removed as fuel gas from
the deethanizer accumulator and the deethanizer bottom product, essen-
tially propane and butanes, is fed to the depropanizer for separation
into the two products.
Pollution Sources
Possible point sources of pollutant emission are sour water
from the deethanizer accumulator (points 1 and 2 in the figure). This
100
-------
L^-2^1
Figure 31. Light ends recovery.
-------
stream should be sent to a sour water stripper. Fuel gas (point 3)
from the unit may contain H2S, in which case it should be treated for
H2S removal before being used as fuel.
T. WASTEWATER SYSTEMS AND SOLIDS DISPOSAL
Refineries generate a significant amount of process water, which
has been in contact with oil and chemicals. This wastewater stream
requires extensive treatment before discharge into a body of water. In
addition, large quantities of runoff from refinery process and storage
areas during rainy periods require treatment before discharge. Fig-
ure 32 shows a relatively complex system for handling these water
streams. The unit operations involved vary from refinery to refinery
depending on local requirements. All refineries can be expected to
have an API separator. Dissolved air flocculators are common, many
units having been installed in the past few years. Biological oxidation
units are less common, and there are only one or two carbon units in
service. For more detail refer to Air Pollution Engineering Manual.
During dry weather operation, the process water streams are
collected and flow to an API s eparator. The oil skimmings are pumped
to a storage tank to await treatment. The sludge that settles in the
separator may be either pumped to an incinerator or removed by tank
truck for disposal. The pH of the water leaving the separator is con-
trolled at 7. 0 to 7. 5 by the addition of either acid or caustic.
This water is frequently pumped to a biological oxidation system
102
-------
ro LX t LtcTn.0 tyrt.
o
UJ
A
\*4CiMg.rLATog.
iua&*.
TAWK.
EppUUC-WT
5UMP
r
•
L
r
t
CAIUOU
t,to
\
h
1
r
i
CAfcBnJ
bU7
t
h
i
r
i
CAAA
6*1
*
L^
I i>e.*Je.B,
/
(Lft-^c-fuvoin.
32. Waste water system and solids disposal.
-------
to reduce the biological oxygen demand (BOD). Further removal of
suspended solids and oil from the water is effected by chemical floc-
culation, using a combination of alum and a polyelectrolyte flocculent,
followed by air flotation. Sludge from biological oxidation and the
skimmings from air flotation may be either pumped to anincinerator
or removed by tank truck for disposal. The water has now been
treated sufficiently to be pximped to a surge tank and then to a muni-
cipal sewer if one is available.
Further treatment may be required before the water can be
discharged to a natural water body. The wastewater is collected in a
reservoir, and the final or tertiary treatment for the removal of dis-
solved organics is effected by passage through activated granulated
carbon beds. The water is then chlorinated and flows to a retention
sump before being discharged from the plant.
The carbon beds are reactivated by backwashing with treated
water. Removed contaminants flow back to the wastewater-reservoir.
When the carbon is spent, it is removed from the beds in slurry form
and placed in storage tanks. The carbon is regenerated by educting
from the tank to a dewatering screw and feeding to a multiple hearth
furnace. Contaminants are burned off as the carbon drops through the
furnace. The carbon falls into a quench tank and is educted to a stor-
age tank ready for reuse. Flue gases from the furnace are quenched
and then scruhbed before venting.
104
-------
During periods of rain, the process waters, mixed with rain
water, continue to be processed through the normal treating facilities.
The maximum rate through this system is limited by the capacity of
the biological oxidation and air flotation units. Water runoff in excess
of the capacity of the system is pumped to the reservoir. Oil skim-
ming facilities should be provided in the reservoir for surface oil
removal. This water can then be pumped and processed through the
activated carbon adsorption plant before being discharged from the
plant.
Pollution Sources
Tanks that are open to the atmosphere are potential sources of
pollutants, thus the API separator (point 1) should be designed with a
cover where required to comply with pollution control standards for
oil-effluent water separation equipment.
Another source (point 2) is the stack gas from the incinerator.
This equipment should also be designed to comply with local pollution
control standards.
A third potential source of emission (point 3) is the vent gas
from the carbon regeneration system.
U. FLARE AND SLOWDOWN SYSTEM
During refinery processing plant upsets and plant emergency
conditions, such as power failures, higher than normal pressure may
be generated in certain equipment. To protect this equipment from
105
-------
damage, pressure relief devices are installed and set to open at a
pressure below the design pressure of the equipment. Process mate-
rials released when these valves open are collected and burned by a
flare (see Figure 33). Process waste gas is sometimes also flared.
Under normal operating conditions, when no systems are
relieving to the flare header, a small purge of fuel gas is used to keep
a positive pressure in the line and the flare flame alight. During
emergency conditions the relieved process fluids flow through the flare
header to the knockout drum, where any entrained liquid is separated.
The vapors from this drum flow through a liquid seal to the flare and
are burned. The liquid is pumped to a slop tank.
Pollution Sources
The only potential source of air pollution from th'is system is
the stream containing the products of combustion of the flare (point 1
on the figure). Any of the refinery processing plants could relieve to
the flare, including those containing H2S and other pollutants; but air
pollution regulations do not usually cover emergency situations. A
continuing emission problem could occur if relief valves do not reseat
properly, leaking process fluids to the flare system. Preventing such
leakage is a matter <>t proper maintenance. Additional information on
flares is presented in Chapter II.
V. STORAGE
Normal refinery operation requires the storage of large vol-
umes of crude oil, intermediates, and finished products. Many of
106
-------
T TL
1—*r ****•
PUC.L
FLANI
AiR-
FUAO.ES.
LIOUlP
SBAL,
/A
T
Figure 33. Flare and blowdown system.
-------
these materials are volatile and, unless properly controlled, storage
can be a major source of hydrocarbon pollutants.
Hydrocarbon emissions from volatile fuel storage can be con-
trolled by not allowing space for vapor formation, confining the vapors
within the storage system, collecting and burning the vapors, and by
collecting the vapors and recovering the hydrocarbon portion. Although
adequate control can be achieved for any storage facility by any of
these methods, the best method to use depends on the characteristics
of the storage facility and fuel to be stored.
1. Floating Roof Tanks
The principle used in floating roof tanks is the elimination of
vapor spaces. This is accomplished by floating a rigid deck or roof
on the surface of the stored liquid. The roof then rises and falls
according to the depth of stored liquid. The roof is equipped with a
sliding seal at the tank wall so that the liquid is completely covered.
No additional roof is required; however, many tanks are equipped with
a standard fixed roof that covers the floating roof. Floating roof tanks
are suitable for volatile fuel storage where the fuel is stored below its
boiling point. That is, the vapor pressure of the fuel must be below
atmospheric pressure at storage conditions.
Sliding seals are an important feature of all floating roofs. The
ideal seal will be vapor tight, long lasting, and require little mainte-
nance. Seals are required at the rim of the roof, at support columns,
108
-------
and at all points where tank appurtenances pass through the roof. Two
basic types of seals are in common use today, the metallic and the
nonmetallic seal. The metallic seal consists of a sheet metal shoe held
against the tank wall by springs or counter weights. The space between
the roof and the shoe is covered by a flexible fabric membrane. The
fabric is often protected by a metal cover.
The nonmetallic seal is usually made of a hollow flexible plastic
tube filled with plastic foam, liquid, or compressed air. Column and
guide cable seals are usually close-fitting, flexible plastic sheets. The
sheets cover holes cut in the floating roof and are sealed at the edges of
the holes by resting on plastic or metal rims fitted around the holes.
The sheet is sometimes allowed to slide horizontally on the rim to
provide for vertical misalignment of the column.
Floating roofs are taken as the standard of effective emission
control for storage tanks. Their effectiveness depends on the material
stored and other factors, but their use results in about a 90% reduction
of emissions.
2. Variable Vapor Space Tanks
There are two general types of variable vapor space, or conser-
vation, tanks that have been used in volatile fuel storage systems.
Lifter roof tanks have a movable roof that rises or falls as the volume
of vapor changes. Flexible diaphragm tanks have a diaphragm installed
within the tank and attached to the wall so as to provide a variable
109
-------
vapor space in the'lower half of the tank, with the diaphragm protected
by the upper half of the tank.
It would be possible to provide enough vapor space within a
system to prevent the escape of vapor during both loading and normal
breathing, but the volume required would be almost equal to the liquid
storage volume. Such a system would be too expensive to be practical.
In the past, lifter roof tanks have been used to prevent tank breathing
losses. No attempt to prevent loading losses was made, the excess
volume being vented to the atmosphere. Since the degree of control
provided by such systems will not meet current pollution control re-
quirements, the use of variable vapor space tanks as primary control
devices has been discontinued. They are, however, still used to pro-
vide surge volume in systems employing absorber units and the like.
(See Section W, Loading and Transfer. )
3 . Flares and Incinerators
Hydrocarbon emissions from storage systems can be controlled
by piping the vapors to a flare or incinerator for burning. Since the
hydrocarbons are destroyed in the process, this procedure results in
an economic loss. The procedure is sometimes justified for isolated
tankage, but most refineries employ floating roof tanks and/or vapor
recovery systems.
4. Vapor Recovery Systems
Occasionally vapor recovery systems of the type described in
Section W, Loading and Transfer, are used for control of storage
110
-------
systems (or tank farms). A more generally used system, however, is
one in which the storage tanks are blanketed with natural gas. (That is,
the tanks are manifolded together and a slight positive pressure of
natural gas is maintained in the manifold. ) When vapor is generated in
this system, the excess is compressed and sent to the refinery fviel
system. In larger systems, where the cost of the additional equipment
is justified, light ends recovery may be employed (see Section S).
Figure 34 is a schematic diagram for a typical refinery storage
vapor recovery system. Standard cone roof tanks are interconnected
with a piping manifold. Since the tanks can withstand a pressure or
vacuum of only a few inches of water in the vapor space, the tanks are
equipped with combination pressure-vacuum relief valves. The system
should be designed so that these valves remain closed. The proper
pressure is maintained by admitting natural gas to the manifold when
the pressure falls, and by removing vapor by means of the compressor
when pressure rises. Normally, the compressor is run continuously so
as to maintain a vacuum in the surge tank at the compressor inlet. The
fuel gas stream may contain sulfur compounds (point 1 in figure).
5. Estimated Hydrocarbon Losses
Annual fuel losses can be estimated by a method given in a
manual prepared by the American Petroleum Institute, API Publication
No. 4080, Recommended Procedures for Estimating Evaporation and
Handling Losses of Volatile Petroleum Products in Marketing Operations,
July 1971. The appropriate nomographs from this publication are
111
-------
ScMXC.6.
fAKlK.
Figure 34. Vapor recovery for storage.
-------
shown in Figures 35, 36 and 37. The factors affecting hydrocarbon
loss include:
Daily temperature change
Fuel volatility
Loading frequency
Paint
Storage temperature
Tank diameter
Tank outage
Fuel volatility is normally expressed in terms of vapor pres-
sure. Reid Vapor Pressure is commonly used. It is the vapor pres-
sure at 100°F as determined by a test method that employs a specified
apparatus. Because of the apparatus used, Reid Vapor Pressure is
not a true vapor pressure. If the true vapor pressure is required, it
is usually found by using an experimentally developed correlation.
Vapor pressure is a measure of the pressure developed by a
liquid, at a given temperature, in the vapor space over the liquid. It
applies to a closed container containing only the liquid and vapor from
the liquid. The vapor pressure depends only on the composition of the
liquid and on the temperature. If the vapor pressure is equal to atmo-
spheric pressure, the liquid is at its boiling point. Materials such as
gasoline are normally stored at temperatures below their boiling
points. The vapor space of the tanks, therefore, contains mixtures of
hydrocarbon and air. For practical purposes, the volume percent of
hydrocarbon in the mixture at equilibrium is equal to the ratio ofliquid
vapor pressure to total system pressure, expressed as percent. The
113
-------
0.20
— O.3O
Ul
t-
0
m
CD
<
x
o
z
w
<
^J
e»
<0
c
Ul
0.
O
z
D
O
a.
z
" *
Ul
ff
3
«0
40
ec.
a.
K
0
a.
>
Ul
D
t-
— 0.40
— O.50
— O.60
— 0.7O
— o.eo
— O.9O
— I.OO
— 1.50
-
— 2.0O
-
_^
— 2.50
-
±- 3.00
^- 3.50
—
— 4.00
—
_
M
— 5.0O
~
-
~ 6.00
—
^
^- 7.00
—
^
— 8.00
-
— 9.00
— IO.O
— 1 1.0
— 12.0
— 13.0
— 14. 0
— I5.O
— 1 6.O
— 17.0
— 18.0
— 1 9.O
— 20.0
— 21.0
— 22.0
— 23.0
<=- 24. 0
S
,1 0
43
nil 1
iim •
IT
1 ,
\
1
\
n
1
L
V
1-4" 2
M «
3
\\ 3 «
Vc 4 8
nt
1 Jf S
ixi p
6^
<
ffln 8 i
iM/t 10 *
/
AH/f
' T/V
<1V> 14
. y% 16
• MA**
'w/x20
'W
v
120-:
^
I 1 0 ^~
'.
~
1 00 "™"
1
~
90-^
;
^
«
I
80-^
;
;
70-^
"
-
"3
™
60-^
•^
™
50-^
™
.1
~
I
40* i
f
•™
-
30-:
i
^
;
20 -f
S a SLOPE OF THE ASTM DISTILLATION CURVE AT I
10 PER CENT EVAPORATED » ~
DEG F AT 15 PER CENT MINUS
1 0
1 W
DEG F AT 5 PER CENT 10 —
;
^
IN THE ABSENCE OF DISTILLATION DATA THE FOLLOW- o :
^>
u
X
z
Ul
(C
X.
"•
Ul
Ul
oe
o
S
Z
Ul
-------
EXAMPlEi
MOM Barral Tank
Huovghpvl=5oOOOO tarrcb p«f To«
Turnover* =10
Tr«« Vapor >rouvro= 5.B p»ia
Working UH = 975 e«rr.li P«r r«or
JO-
IS
10 -
20
- IS
1000 n
1500 -
2000-
3000-
4000 -
5000-
6000
7000
8000
9000
10000
I
I
10.0-
9.0T-;
1.0-
7.0-
6.0-.
5.0-
4.0-
J.O-
1.3-
Pivol
e
* « —
o- .5 •
•= ? 5
1*1
E • j
•• o **
Note: The throughput is divided by a number (1, 10, 100, 1,000) to bring it
Into the range of the scale. The working loss, read from the scale, must
then be multiplied by the same number.
(API PUBL. NO. 4060)
- U
- 1.0
- 0.9
- 0.8
- 0.7
- 0.6
•
-04
- 0.4
i
J- OJ
0.2-^
:;
0.15
0.10
0.09
•- 0.2
• '
'
. -0.15
:-
0.0« 3"
^- 0.10
0.07-J. Q0o
0.06 "T" 0.08
V 0.07
u.
&
c
^
J
e>
c
^
o
*
•
c
1
£
Figure 36. Working loss of gasoline from fixed-roof tanks.
115
-------
(API PUBL. NO. 4060)
T - U*r
I, - 1.11
H - 31 rm
D - 100 rY.1
I, - I150*orr*li »«' y.«»
10
r
X
15
S
s
\
\
s
s
\
\
\
s
s
\
20
X
X
\
\
\
^ i
\
X
25
N,
\
\
X
s
10 >
\
\
\
\
Tonh Color
tool
•nil*
wMta
•|«*IM« (ipocvlor)
•Iwin.x Mill.-)
fighl gray
•wditia gray
ih.ll
whita
wfcito
•IwHllNtim (lp*Cvl«f)
OfwM[MW«(tp«i>>.» IdWuu)
Kokt gray
M«diwM gray
Po«» factor, F,
Point in Good Condition
1.00
1.04
1.1*
1JO
1.10
1-1?
1JO
143
1*1
Point In Poor Condition
1.15
1.11
1.24
1.2*
l.M
1.44
1.4*
u
1.4
\
1Jv
IJ,
'•'v
LO
^
30*
b
N
sx
23'
X
s
1
20
Average AtmotpKork
Temperature Chang*
in DvgiMi Fohranh
15 .
\
\
N
\
s
\
j
> « r 1 1 ! i i ii — i
40
x
s
\
\
s
s
x
\
x
s
\
\
\
s
50
\
x
\
\
\
\
\
\
s
X
40
\
^
s
x
\
\
\
\
\
x
X,
.70 (0 M 100 125
Xl X X X X J _X
\
\
s
X
s
\
\
\
\
\
s
X
x
S
\
s
\
\
\
\
\
N
\
x
S
\
s
\
s
s
\
X
\
. X
\
X
s
X
ss
\
s
s
s
\
s.
\
v
\
N
X
s
X
S
x
s
\
\
s
\
s
V
\
\
N
\
s^X
S
\
v
S
V
\
\
\
S.
X
\
\
N
150 175 200
x. rv r\
X\
X
S
\
V
S
\
\_
sv
S
\
X
\
\
k
s\
\
\
X
v
S
s
\
\
\
s
\
\
S
\
\
\
s^\
x,s
N
V
\
\
s
S
v,
S
s
k
S
\
\
\
N\
N
\
N
S
S
\
\
S
v
V
^
S
\
10 15 20
f £— « 1 1 »• fill 1 1— f-1 1— « 1 1
k
\
s
\
s
\
\
^v
\v
\
\
N
V
K
S
\
^
S
\
\
\
\\
v
s'
\
\
IS
s
N
\
\%
X,
s
s
s.
S
\
y
s^
x^
\
s
s
\
s
s
s
s^
s,
s
\
v
S
\
s
sj"
\jv
Njv
^
s!
**^
xS
s
\
\
s
\
>s
s
s
\
K>
X
s
V
S
\
X
N
N
s
s
'0
S,
v
s
s
~N
\
y
\\
\s
s
s
v
s
S
\
s
\s
V
x,
SI
s
s
x,
S
\
^
v\
x
s
s
\
S
\
y
s\
v
s
x
X
i
•it.
s
V
s
V
s
N
\
\
x^
X
X
V
^
ss
X
;''N
s
S
X
s
S
\
\
V
s
V
\
—
x
x
X
X
s
10 '
N_
—
/
/
^
i
>
i
4
7
'*
10 ,
200
175
--
/
/
r
/
'
\
s
\
\
s
\v
X
s\
\
\
/
/
/
/ 1
/
y
'
\
\
s
V
\
k \
\
s\
s
N
/
/
/
//
/
' k
I
f i
,.
/ ' " '
'/
(I
f
/
/
//
A
/ /
^
X
\
\
\
V
s\
x
S
-t
*r
X
\
s,
\v
X
\
\
//
/
/
H
/
'/ f
f j
I
*
\
\
\
\
\
v
\
/
//
//
{/
i
$
'•/
7
V
",
1
1
1
1
N
N
^
V
iv
\
\
s
S
//
//
//
/t
§
/
^
\
\
\v
\
s\
s
r/
//
/*
V
• f
v
* *<
\
\
V
^
s
9
r
»
s
^
x
^
\
v
v
25 101540 50 M 70 10 »0 100 125 150
•"I It >-l tl 1 > 1 II t 1 1 1 M 1 »— OooooVw
Figure 37. Breathing loss of gasoline from fixed-roof tanks.
-------
amount of hydrocarbon lost in a given volume of vapor displaced from a
tank is thus a direct function of the vapor pressure of the stored liquid.
Loading frequency is a measure of the number of times a tank
is filled and emptied in a given period of time. It is usually expressed
as "turnovers per year". When liquid is loaded into a lank, an essen-
tially equal volume of vapor is displaced. The amount of hydrocarbon
lost is a function of other factors as well bul is strongly dependent on
the total volume of vapor displaced and, therefore, on the turnovers
per year.
The temperature at which a given fuel is stored determines the
vapor pressure and hence the equilibrium vapor composition. The
higher the temperatxire, the higher the hydrocarbon content of the vapor
and the greater the loss per volume of vapor. In vapor loss correla-
tions, the Reid Vapor Pressure and the storage temperature arc used to
specify vapor composition, and the composition itself may not ever be
determined explicitly.
Daily temperature change causes a cyclical change in the tem-
perature of the stored liquid and of the vapor space above the liquid.
Increasing liquid temperature results in increasing liquid vapor pres-
sure, and some of the liquid is vaporized. At constant pressure, not
only is the percent of hydrocarbon in the vapor increased, but vapor is
displaced from the tank. Another effect of increasing temperature is
that the vapor in the vapor space expand's. I'nless the tank pressure is
increased, vapor is displaced from the tank. When temperature drops,
1 17
-------
the reverse occurs and air is drawn into the tank vapor space.
Breathing losses are a function of vapor space volume (or tank
outage); the greater the volume, the greater the loss. Losses from
tanks that are nearly empty will be greater than losses from tanks that
are full. For a tank of a given diameter, the vapor volume can be
specified by stating the height from the top of the tank to the liquid
surface. This height is called outage. When calculating breathing
losses, it is common practice to assume that the tanks are half full;
that is, outage is taken as equal to one-half of the total tank height.
If the liquid and vapor in a tank were always in equilibrium,
tank diameter would only be a means of stating volume. In actual field
testing, however, it has been found that the tank vapors are not always
saturated and that vapor losses are somewhat less than would be other-
wise expected. Smaller tanks were found to have lower losses, but the
difference is not significant for tank diameters of 30 feet or more.
The color and condition of the paint on a tank have been found to
have an important effect on breathing losses. Using white paint in good
condition as a standard, white paint in poor condition may lead to a 15%
increase in loss. A medium grey paint may lead to a 46% increase.
All of these factors are included in the nomographs. The nomo-
graph in Figure 35 can be used to obtain the true vapor pressure from
the more commonly available Reid vapor pressure. That in Figure 36
can be used to estimate working loss, that is, the loss due to loading
118
-------
operations and the nomograph in Figure 37 can be used to estimate
breathing loss, that is the loss due to daily temperature changes.
6. Potential Point Sources of Pollutants
Vents on uncontrolled cone roof tanks storing volatile hydrocar-
bons are a source of pollutants and should be controlled.
Sliding seals and gage-opening covers on floating roof tanks
must be properly maintained or they will leak.
Conservation valves on tanks connected to a vapor recovery
system may leak. The leakage may be due to a system pressure thai
is higher than the set pressure of the valve or to an improperly main-
tained valve seal.
W. LOADING AND TRANSFER
Gasoline and other petroleum products require distribution
from the refinery to the consumer. This is achieved by pumping from
refinery storage tanks to a loading terminal where the products are
loaded into tank cars, barges, and tank trucks by means of loading
racks. Products are also loaded into ocean-going tankers at bulk
marine terminals by pumping from storage. Marine terminals may
have facilities for unloading crude oil from tankers into storage tanks.
During all these loading and transfer operations large amounts of air
containing hydrocarbon vapors are generated. If not controlled, these
vapors can be an important source of air pollution.
119
-------
1. Loading Equipment
Loading racks are structures containing the platforms, piping,
vapor collection devices, control devices, and loading arm assemblies
required for transferring the product from storage to the transport
vehicle. Bottom loading normally requires simpler equipment than
overhead loading. The method used for loading marine tankers is
similar to the bottom-loading operation. Liquid is delivered to the
bottom of a compartment while vapors are vented through a manifold.
To avoid atmospheric pollution the produced vapors are col-
lected at the tank vehicle hatch using specially designed closure de-
vices. For overhead loading these are plug-shaped devices that have
a central channel for the liquid to flow into the tank and an annular
space for the vapor to flow out of the tank into a pipe connected to a
vapor disposal system. For bottom loading a vapor take-off line is
connected between the vapor space of the tank and a vapor disposal
system.
2. Vapor Recovery Systems
a. Vapor Recovery to Fuel Gas When a suitable fuel gas
system is available, the vapors can be used for fuel. The
loading system is gas blanketed to avoid explosive mixtures,
and the vapors are collected in a vapor holder. The vapors are
fed to a compressor and discharged to the fuel gas system.
b. Vapor Recovery by Absorption in Gasoline Figure 38
presents a typical system for the absorption of hydrocarbons in
120
-------
gasoline. Explosive mixtures cannot be permitted in this unit
so the amount of hydrocarbon in the air is raised substantially
above combustible limits by saturating with gasoline. This is
accomplished by countercurrent contact of the air with gasoline
in the saturator prior to storage in the gas holder. The vapors
are compressed, cooled and introduced into an absorption column
where absorption of the hydrocarbon in gasoline takes place.
The air is vented to the atmosphere through a back pressure
control valve. The gasoline is returned to storage after the
dissolved air is removed in the two-stage flash separator.
Although the design recovery of hydrocarbon vapor by this
system can be in excess of 90 percent, the vented air (point 1
in the figure) may be a potential source of pollution and may
require checking. If the hatch closure (point 2) is not operating
properly, air pollution may occur.
c. Vapor Disposal to Flare The vapors can be satisfactorily
disposed of by burning in a smokeless flare.
3. Loading Losses
Loading losses are influenced by many variables. The volume
of vapors produced during loading is influenced by the mode of loading
employed. The modes in general use in refinery operations are
overhead loading and bottom loading. Overhead loading is sub-
divided into two types, splash filling and submerged filling. In
splash filling, the outlet of the delivery tube is normally above the
122
-------
Figure 38. Loading rack vapor recovery system.
-------
liquid level, while in submerged filling the outlet is below the liquid
level. The former generates more hydrocarbon vapors.
To calculate loading losses, the total throughput information is
recorded on a form such as that shown in Table 7. The true vapor
pressure (TVP) is determined from the nomograph in Figure 35 for
each product. The curves in Figure 39 can be used to calculate the
evaporation as volume percent of load for splash loading and sub-
merged loading in tank cars and tank trucks. The correlation curve in
Figure 40 can be used to determine marine evaporation losses. A
detailed discussion of these calculations can be found in API Publication
No. 4080 (July 1971).
X. FUEL GAS SYSTEMS
Operators would like to maintain a fuel gas balance in their
refineries to produce enough fuel gas to supply the heat required in the
refining processes. However, production and use of fuel gas depend on
the refinery processes, the crude processed, and economics. In
general, additional fuel gas must be purchased. The crude unit and all
of the cracking process units produce fuel gas. In most cases these
gases contain sulfur compounds and have to be treated before entering
the fuel gas system.
The fuel gas system is the storage and piping network by which
the refinery stores, blends and distributes the gas internally in the
refinery. The input to the fuel gas system is in general from two
123
-------
Table 11. LOADING LOSSES: MOTOR GASOLINE, TU RBINE FUELS, AVIATION GASOLINE
Plant Company
State Year
Location: City
Product
Total
Thruput
M Gallons
Total Deliveries
M Gallons
Tank Car or Tank Truck
Marine
Splash (1)
.
.
Sub-
Surface (2)
Product
Temp.
°F
Vapor Pressure
RVP
PSI
•
TVP
PSIA (3)
Losses
Bbls/Yr.
(3)
(1) Splash Loading - free fall of product during loading. (2) Subsurface Loading - product delivered to bottom
of compartment without splashing. Direct bottom delivery or top loading thru long spout achieves this effect.
(3) To be calculated.
-------
DURING SPLASH LOADING CAM
EVAPORATION LOSS BY
OP rHR£E FOLD
l/APOR.
6 789
(TVP)iPS/A
Figure 39. Loading losses for tank trucks.
125
-------
0. 10
Tl
ri
0.08
"o 0.06
W
CO
O
c
o
.^H
-1-1
ri
O
0.04
S* 0.02
W
0 2 4 6 8 10
True Vapor Pressure (TVP) psia
Figure 40. Loss from loading tankers and barges.
\26
-------
sources; from the treating unit where the sour gases produced in the
refinery are treated for the removal of sulfur and from the gas
purchased outside.
A diagram of such a system is presented in Figure 41. A knock-
out drum is provided to collect any liquid that may condense from the
gas stream and to provide minimal storage for averaging high and low
demands by the refinery. The fuel gas produced by the refinery flows
into the system on a back-pressure control from the higher pressure
level in the treating unit. When the demand exceeds the supply of gas
produced within the refinery, the pressure in the knockout drum de-
«
clines and purchased gas is allowed into the system. When the supply
of refinery gas exceeds the demand of the refinery, the excess gas is
flared. The system in general is well contained, with minimum emis-
sion sources. However, fuel gas produced in the refinery should be
checked to see that it is free of sulfur compounds (point 1 in the figure)
and any excess gas should be flared (point 2).
Y. STEAM GENERATION
Steam is used as a heating medium in various refinery opera-
tions and as a process fluid in others suchas hydrogen production and
steam stripping.
Plant steam systems are normally closed cycles in which the
steam generated yields its heat to process streams in heat exchangers
by condensation. The condensate is returned to the steam generators
127
-------
00
Figure 41. Refinery fuel gas system.
-------
to be vaporized again. A relatively small make-up water stream is
required to replace losses and blowdown from the system. The pro-
cess flow sheet is shown on Figure 42.
Utility water is normally used as make-up water to the system.
This water is treated by either softening or deionization. Air dissolved
in the treated water is removed by steam stripping in the deaerator
after which it is mixed with the recycle condensate. The water is
pumped from the deaerator to the steam drum for conversion into
steam.
The boiler may be fired by either fuel gas, fuel oil, or coal.
Whichever fuel is used, the stack gases may be a source of atmo-
spheric pollutants (point 1 in the figure). The deaerator vent could
emit pollutants to the atmosphere if there is leakage of hydrocarbons
from the process side into the steam side of a heat exchanger. This
stream (point 2) should be checked regularly for hydrocarbon contami-
nation.
Z. COOLING TOWER
Water is used as the medium for removing heat from various
refinery streams. Plant cooling water systems are normally closed
cycle (see Figure 43). Water from the cooling tower picks up heat
from process heat exchangers and is returned to the cooling tower.
The heated water flows to the top of a tower, which is open to the atmo-
sphere, and is allowed to flow down the tower over packing. Atmospheric
129
-------
pH^SRH Alt-
is),
Figure 42. Steam generation.
-------
n
a
w»—
8
^
r~*T»
UTILITY
ACID
» f
•— (PI
Figure 43. Cooling tower.
-------
air is either forced or flows by natural convection up the tower and the
water is cooled by partial vaporization into the air flow. The cooled
water is collected in the cooling tower basin prior to being pumped
back to the process users. Utility water is added as make-up to the
basin to replace water lost by evaporation in the tower, entrainment,
and to blowdown. Chemicals, such as sulfuric acid, are added to
maintain water quality.
The cooling tower (point 1 in the figure) could be the source of
atmospheric pollutants if there is leakage of either hydrocarbons or
other pollutants from the process side into the water side of a heat
exchanger. Evaporation and entrainment losses can result in the
release of significant quantities of pollutants into the atmosphere. A
periodic check of the cooling tower basin for odor and visual surface
contamination would indicate the presence of this type of leak.
AA. ELECTRIC POWER GENERATION
Each operation and process in an integrated refinery adds to the
overall power requirement of the refinery. Requirements of the indi-
vidual processes can vary from 0. 3 to 5. 0 kw of electrical power per
daily barrel of throughput. Maintenance of adequate and reliable sup-
plies of electrical power is a major concern.
Most refineries now use purchased electric power for normal
operation. Commercial sources of power are reliable and power can
be obtained at a reasonable cost. Refineries, however, frequently have
132
-------
emergency shutdown procedures that must be followed to prevent a
major loss of product or damage to equipment in the event of a local
power failure. Some pumps or blowers must continue to operate, and
some instruments and alarms must be available to warn if a hazard
exists. Normally, a refinery will have an emergency power generation
system to meet this need. Emergency power generation equipment
may be driven by steam turbine or by gasoline or diesel engine.
Remotely located refineries may not have access to a commer-
cial source of power and may, therefore, generate their own power.
The motive source to drive the main refinery generators may be a
steam or gas turbine or a gas engine. The air pollution resulting from
operation of gas turbines or gas engines is usually not excessive if low
sulfur fuel gas is used. The air pollution characteristics of steam
boilers are the same as those of other fired heaters (see Chapter II,
Refinery Equipment).
BB. CATALYST REGENERATION
The catalyst in most of the refinery catalytic processes needs
to be regenerated periodically. Whenever hydrocarbon feed stock is
cracked, reformed, isomerized, or hydrotreated, some coke deposit
will be formed on the catalyst and thus deactivate the catalyst. In most
cases as soon as the catalyst gets partially deactivated, the yields
decline and higher cracking or reforming severity is needed to achieve
the desirable product characteristics. Higher severity in most cases
133
-------
means higher operating pressure and temperature which in turn de-
posits more coke on the catalyst. So at some point, the process has to
stop and the catalyst has to be rejuvenated. The frequency of catalyst
regeneration will depend upon the type of process, the feed stock and
the severity of the operation. Regeneration frequency can vary be-
tween once every two or three years to once a week. In some pro-
cesses, the catalyst bed may be removed, replaced with freshcatalyst
and the deactivated bed shipped to be regenerated elsewhere. Most
catalysts, however, are regenerated in place either continuously or
periodically. For continuous catalyst regeneration, see the process
description of catalytic cracking, especially fluid catalytic cracking.
Periodic in situ catalyst regeneration procedures are similar
for most processing units. The only way coke can be removed without
disturbing the catalyst bed is by burning it with air. Some processes
may use steam and air to burn the coke while others can use inert gas
(mainly nitrogen) and air when the catalyst is sensitive to water. The
decoking operation will normally proceed as follows:
1. Purge and depressurize the reactor.
2. Inert gas or steam circulation.
3. Coke burning.
4, Inert gas and combustion products purge.
5. Gas purge and repressure.
The initial step is to cut off the feed to the reactor and purge
the unit into the relief header and to the flare. At this stage, steam or
134
-------
inert gas (less than 0. 1% O2) may be added to the system and the purge
will continue until all the volatile combustible materials are out of the
system. Pockets of combustible material in the system could cause an
explosion during the coke burning step. When the gas purge is clear of
hydrocarbons or other combustible gases, the inert gas or steam is
recirculated through a compressor and a heater to bring up the temper-
ature of the catalyst bed to the coke ignition level. Some of the coke
will break loose at this stage and flow out of the system with some
catalyst dust. Both coke and dust have to be removed, for instance, by
wet scrubbing. Also, if steam is used, some CO2, CO and H2 will be
formed and some of the circulating gas needs to be continuously purged.
At the preset temperature (500-700 °F) air is allowed into the circu-
lating inert stream, and the coke burn begins. Excess gases are con-
tinuously purged, and coke particles and dust are scrubbed out of the
recirculatmg stream. The circulating gas temperature is allowed to
go to about l,000°Fby regulating the amount of air allowed into the
system. The "hot spot" where the coke burns slowly proceeds down
the catalyst bed until all the coke deposit is consumed. At this point,
the circulating gas is purged to the flare with fresh inert gas. This
purge continues until the oxygen level in the system is reduced and the
ash and dust are removed. The inert gas is then circulated to reheat
the system to operating temperature and feed is introduced while the
inert gas is purged to the flare.
135
-------
Gaseous air pollutants are formed throughout the regeneration
operation by the reaction of steam air and coke, and mechanical action
forms dust. If sulfur is present in the coke, SO2 will be formed as a
part of the coke combustion products. Coke dust, ash, and catalyst
dust also are present throughout this operation and have to be removed
before the combustion gases are discharged to the atmosphere. Wet
scrubbing of the purged gases with a caustic wash may be used in some
cases to reduce the SO2 and the dust in the combustion gases. How-
ever, this scheme is not in general use and cannot be used where
water is a catalyst poison.
136
-------
II. REFINERY EQUIPMENT
A. INTRODUCTION
In the same way that individual chemical processes have charac-
teristic kinds of pollutant emission, the pieces of equipment used in
these processes have characteristic points of pollutant emission.
This chapter contains a summary of the major types of process
equipment and their performance characteristics. The internal mech-
anisms of each are illustrated. Emphasis is directed toward the mech-
anisms by which each item of equipment might release material .to the
atmosphere under normal and abnormal operating conditions.
The objective of this chapter is to provide sufficient information
about process equipment so that the FEO will recognize the equipment,
understand its function, and be sufficiently knowledgeable regarding its
operation to decide whether the unit is being operated properly. If im-
proper operation is evident, he should be able to recommend changes to
the system that would reduce pollutant emissions to the point where the
facility can comply with environmental pollution standards.
B. PUMPS
A refinery uses many different types of pumps to move fluids.
137
-------
Pumps vary in capacity up to 100, 000 gallons per minute and in pres-
sure differential up to 30, 000 pounds per square inch. A pump is de-
signed to perform a specific function, and it is limited to a rather
narrow range of operation above and below the design condition. For
this reason, hundreds of pumps of different styles and modes of opera-
tion are used. The types most often used in refineries fall into two
categories: centrifugal and positive displacement devices. Centrifugal
devices are centrifugal, axial and turbine pumps. Positive displace-
ment devices are reciprocating piston, plunger, diaphragm, rotary
vane, and gear pumps. Other specialty pumps are available for unusual
applications, but this list covers the types that will be encountered most
often..
1. Centrifugal Devices
a. Centrifugal Pumps A centrifugal pump consists of a
rotating element, called an impeller, and a casing which sur-
rounds the impeller. Liquid enters the pump and flows to the
eye of the impeller. As the impeller rotates, it throws the
liquid outward by centrifugal force. The casing collects the
liquid that is discharged from the impeller converting a part of
the kinetic energy in the liquid into fluid pressure. The centri-
fugal pump is a flow device which continuously imparts energy
to a flowing fluid. Figure 44 is a cutaway view of a centrifugal
pump.
The amount of energy imparted to the fluid is a function
of the top speed of the pump impeller. The discharge pressure
138
-------
PUMP
u>
rfcE.6.
Figure 44. Centrifugal pump.
-------
of the pump increases with the diameter of the impeller and
the speed of the pump. Low-pressure electrical pumps nor-
mally operate at 1, 750 rpm, and higher head pumps will rotate
at ?, 500 rpm. Even higher pressures are available if a tur-
bine drive is used to increase the impeller speed. Many differ-
ent types of centrifugal pumps are used in the refinery. Figure
44 is typical of a single-stage, horizontal centrifugal pump.
Where a high head is required, multiple impellers can be
mounted on the same shaft as shown in Figure 45. The same
type of pump is sometimes mounted in a vertical position to
conserve space.
Figure 46 shows a sealed rotor, or canned motor,
single-stage pump in which the rotor is exposed to the process
liquid. The pump motor case is fully enclosed and therefore
no seals are required for the pump. This pump is frequently
used where the fluid pumped is particularly corrosive or toxic.
The centrifugal pump is a variable capacity pump. The
liquid capacity of the pump can be regulated by adjustment of a
valve on the discharge of the pump liquid. Closing the valve
increases the head that the pump must provide, and the liquid
capacity of the pump is reduced to meet the new requirement.
As the valve is closed, the horsepower requirement of the
pump is minimized. However, if the flow of fluid is stopped,
the fluid in the pump will overheat and the pump could be
damaged.
140
-------
Figure 45. Multiple stage centrifugal pump.
-------
Figure 46. Canned motor pump.
142
-------
The centrifugal pump shown in Figure 47 has three
points where leakage might be expected to occur: at the inlet
flange (point 1 in the figure), at the outlet flange (point 2), and
at the shaft seal (point 3). If the system is properly constructed,
the flanges will not leak and the principal spot where leakage
will occur is at the shaft seal. The shaft seal may be a simple
packed stuffing box which must be replaced periodically when
leakage becomes excessive, or it may be a mechanical seal
(see Figure 48). Mechanical seals work very satisfactorily and
allow only minimum leakage with clean fluids, but deteriorate
rapidly if the fluid contains abrasive particles.
Pump leakage is readily visible if the products do not
vaporize. Where the liquid pumped has a high vapor pressure,
vaporization will occur as soon as the product is released.
Evidence of leakage may appear as an accumulation of conden-
sate or frost around the point of leakage.
b. Axial Pumps - The axial pump uses both mechanical
impulse and centrifugal force to pump large quantities of fluid
where the head requirements are low. Figure 49 shows a sec-
tion of a typical axial pump. The rotating element is a pro-
pellor, which is sometimes followed by stator blades to assist
in recovery of the fluid energy.
Axial pumps are normally used for recirculation of
fluids where the head requirement is low. One application is
143
-------
(OUTLET)
( INJLET)
Figure 47. Close-coupled pump.
1 A A
-------
Stationary
Seal Ring
Rotating Seal Ring
Static Seals
Shaft
Shaft
Internal Mechanical Seal
Packing
Packed Seal
Figure 48. Seals
145
-------
Pi 5CMA£.<££.
/NJJ-B.T
Figure 49. Axial flow propeller pump.
(elbow type)
146
-------
in the circulation of large tanks of stocks such as gasoline.
Side-entering mixers which are a type of axial pump are fre-
quently used to obtain uniform blends of mixed stocks. If the
discharge of an axial pump is throttled, the horsepower in-
creases and the pump efficiency decreases. Normally these
pumps are operated for maximum capacity with minimum
regulation.
Hydrocarbon emission from axial pumps can occur at
the. points where the pump is connected into the system (point 1
in the figure) and at the shaft seal (point 2). Leakage at the
shaft seal is most common. The discussion of the problem
under the heading Centrifugal Pumps is also applicable here.
c. Turbine Pumps - Turbine pumps combine the character-
istics of centrifugal and axial pumps. The impeller of a. turbine
pump, as shown in Figure 50, causes the liquid to move axially
and radially as it passes the impeller. These pumps may have
several stages in series and are frequently mounted vertically.
They are often immersed in fluid to be pumped - for example,
when they are xised in wells.
2. Positive Displacement Devices
There are many pumps that move liquids by mechanical dis-
placement of the liquid from a fixed chamber. These pumps are
usually low-volume, high-differential pressure pumps. The capacity
of the pump may be variable but is limited by the mechanical volumetric
147
-------
Figure 50. Vertical turbine pump.
148
-------
displacement of the pump.
A positive displacement pump produces a discharge pressure
that meets the head requirements of the system. The maximum head
obtainable with the pump is fixed by the power limitations of the drive
mechanism and it is not possible to throttle the discharge of the pump
to adjust the fluid capacity. Instead, the rate of displacement of the
pump must be adjusted.
a. Reciprocating Pistoji Pumps - The reciprocating piston
pump is one of the oldest types of pumps in refinery service.
Figure 51 shows a motor-driven piston pump which could be
used to pump fluids in a high-pressure process system.
Each displacement stroke of a piston pump produces a
flow pulsation. Some pumps are double acting, i. e., each
movement of the piston causes fluid displacement to produce a
more continuous flow. To smooth the flow even more, two or
more pistons may be coordinated so that the flow pulsations
are alternated. By these measures, the flow pulsations can be
minimized until the flow is essentially continuous. Piston
pumps are most susceptible to leakage through the packing on
the pump shaft. This packing must slide along the shaft for the
full stroke of the pump and is continuously subject to wear. To
prevent excessive leakage, this packing should be regularly
adjusted and maintained.
149
-------
Figure 51. Reciprocating piston pump.
(double acting)
-------
Piston pumps have check valves on the suction and dis-
charge sides of each cylinder. These valves are subject to
wear and must be serviced periodically. The valves are im-
mersed in the fluid that is pumped and, to service the valves,
this fluid must be removed. Removing it can release a signifi-
cant amount of hydrocarbon to the atmosphere.
b. Plunger Pumps - Plunger pumps are similar to piston
pumps in function, but instead of a piston they use a plunger
which moves in and out of a fluid-filled chamber. A packed
seal is used to prevent leakage around the plunger. The plunger
punp is prone to leak through the packed seal in the same way
that a piston pump leaks around the pump shaft. The valves of
this type of pump must also be maintained regularly, and this
operation could result in air pollution if not carefully performed.
c. Diaphragm Pumps - A diaphragm pump uses the move-
ment of a flexible diaphragm to displace the fluid that is being
pumped. Check valves on the inlet and outlet of the pump pre-
vent the liquid from flowing backward. A typical diaphragm
pump is illustrated in Figure 52.
The use of a diaphragm eliminates the need for a shaft
or plunger packing to contain the fluid being pumped. However,
the diaphragm itself is subject to failure and must be watched
carefully. A leak in the pump diaphragm would allow the pro-
cess fluid to escape to the atmosphere through the mechanism
151
-------
Figure 52. Diaphragm purnp.
153
-------
that is used to move the diaphragm.
d. Rotary Vane Pumps - Rotary vane pumps are frequently
used at moderate pressures and temperatures to pump fluids.
Figure 53 shows a cutaway drawing of a typical rotary vane
pump. As the rotor turns counterclockwise, liquid is carried
from left to right by the vanes which slide in and out of the
.slots in the rotor to maintain constant contact with the wall.
The cylinder space is circular but it is offset from the axis of
the rotor. Each revolution of the rotor moves a positive volume
of liquid through the pump. The flow through the pump is non-
pulsating, and the pump can handle fluids and gas without devel-
oping a vapor lock.
The shaft of this type of pump is the principal point of
external leakage. However, effective shaft seals are available
so that, with a well maintained pump, leakage should not be a
problem.
e. Rotary Gear Pumps - Rotary gear pumps are used for
low-head, low-capacity services with clean fluids. Figure 54 is
a cutaway section of a dual-shaft, two-gear pump. As this
pump rotates, the gear teeth carry fluid from one side of the
pump to the other. The meshing teeth prevent the liquid from
flowing backward.
The principal point of leakage with a gear occurs where
the shafts penetrate the pump case. With a regular maintenance
program, leakage at this point should be negligible.
153
-------
Figure 53. Rotary vane pump.
154
-------
Figure 54. Rotary gear pump.
(two-impeller)
155
-------
G. COMPRESSORS
Mechanical devices used to compress gases also fall into posi-
tive-displacement and centrifugal categories. Positive-displacement
category includes piston, rotary-vane, and rotation-lobe compressors.
The centrifugal category includes centrifugal and axial compressors.
The mechanism, used to drive the compressors, can sometimes
become a secondary source bf air pollution. Therefore the type of
drive used with each compressor is also discussed briefly.
In addition to these devices, there are a number of hydraulic
systems for compressing gas, but they are seldom found in refinery
service and, therefore, are not discussed.
A unique type of gas-compression device which is described is
the jet ejector, a simple device that uses the dynamic energy of one
fluid to compress another.
1. Positive-Pis placement Compressors
a. Reciprocating Piston Comprejsors The reciprocating
piston compressor is the most common type of gas compression
device in use. It is used for pressure differentials from five to
several thousand psi and for capacities from a fraction of a
cubic feet per minute (cfm) up to the maximum capacity of a
system. In refineries, reciprocating piston compressors are
likely to be found in service compressing natural gas, hydrogen,
or liquified petroleum gas (propane and butane).
156
-------
Where pressure differentials are high, several stages of
compression may be required to achieve a desired processing
pressure. For high capacity gas compression requirements,
several cylinders can be designed to operate in parallel. In
many instances, a single multicylinder compressor may be used
to compress two or more stages of several different gases.
For high-pressure systems most pistons are single-
acting, two-stage devices as shown in Figure 55. At low pres-
sure, the usual practice is to use double-acting pistons with a
packed seal on each end of the piston rod. The cylinder is often
supported by a sleeve or distance-piece and the crankcase is a
second low-pressure packing designed to prevent loss of crank-
case oil. Any gas which escapes from the cylinder is vented
into the distance-piece, and in some cases, where the gas being
compressed is toxic or flammable, the distance-piece is en-
closed and a pressure vent to the flare is provided.
Large reciprocating gas compressors may be driven by
electric motor or by gas or diesel-powered engine. Where a
reciprocating combustion device is used to drive the compressor,
the device could become a major source of pollution. The
exhaust gas from such engines should be monitored as a part of
the complete refinery inspection.
b. Rotary Lobe Blowers - Rotary lobe blowers are used for
high capacity, low-differential pressure systems. A section of
157
-------
« /VAU/es
Figure 55. Reciprocating compressor.
(2-stage)
158
-------
such a blower is shown in Figure 56. Leakage occurs primarily
at points where the shafts come through the case.
c. Rotary Sliding Vane Compressors - The rotary sliding
vane compressor operates at relatively low pressures and capa-
cities. As shown in Figure 57, a single rotor with sliding vanes
turns in an eccentric cylinder, compressing the gas as it moves
through the unit. As with the rotary lob£ blower, leakage occurs
primarily where the shaft comes through the case of the blower.
2. Centrifugal and^ Axial Compressors
Centrifugal and axial compressors may be used in various pro-
cesses in the refinery. The centrifugal compressor is effective in
compressing the relatively high molecular weight hydrocarbon gases.
It is frequently used to compress the mixed gases from a catalytic
cracking unit.
Axial compressors are suitable for high-capacity gas require-
ments, but they are not as versatile as the centrifugal compressor and
have not been used extensively in refineries.
Figure 58 shows a centrifugal compressor shaft with a labyrinth
seal used to prevent loss of gas. The seal consists of a number of re-
strictions and openings through which the escaping gas must flow. If
this seal is not properly maintained, it can be a major source of loss of
the process fluid. The labyrinth seal is normally vented at some mid-
point and bled back to a lower pressure stage or to the compressor
159
-------
Figure 56. Rotary lobe blower.
(two impeller)
160
-------
Figure 57.
Rotary compressor.
(sliding vane)
161
-------
Figure 58. Labyrinth seal.
162
-------
suction. When loss of gas must be restricted, an inert gas or liquid
can be injected into the seal at the midpoint of the labyrinth.
Centrifugal and axial compressors may be driven by an electric
motor, a steam turbine, or a gas turbine. If a gas turbine is used, it
could be a secondary source of pollution and, therefore, the exhaust
gas from the turbine should be monitored.
D. HEAT EXCHANGERS
The exchanging of heat between refinery fluids is one of the
methods by which heat is added or removed from refinery streams.
There are two types of mechanisms for effecting this exchange:
the direct and indirect heat exchangers.
1. Direct_He_at Exchangers
The single most important direct heat exchanger is the baro-
metric condenser (Figure 59) used mainly for the condensation of steam
in vacuum systems. Heat is exchanged by contacting the incoming
steam directly with cooling water. The steam condenses and is pumped
out of the system with the coolant. Noncondensables in this system are
removed with a steam jet or a compressor.
Any pollutants in the steam will be distributed between the con-
densed steam and the noncondensables based on the degree of their solu-
bility in water (point 1 in the figure). In most cases, however, the
steam is relatively free of pollutants.
2. Indirect Heat Exchangers
Indirect heat exchangers are more common. In these units, the
163
-------
WATER. STEAM
STEAM
SUCTION
HOT WELL
NONCONDEN^/ABLES
FUME INCINERATOR
TO
CONDENSER TAIL PIPE
figure 59. Barometric condenser.
164
-------
two fluids exchange heat through a surface plate or a tube that separates
them. This group of exchangers includes: the shell and tube, the
double pipe, and the air-cooled.
Shell and rube heat exchangers (Figure 60) consist of a tubular
shell 12 to 40 inches in diameter and 10 to 24 feet long with 3/4 or
1-inch in diameter tubes inside. One fluid thus flows in the shell and
the other through the tubes.
A double-pipe exchanger (Figure 61) consists of two concentric
pipes, one inside the other. The outer pipe may be 2 to 3 inches in
diameter. The inner pipe is usually about 1 inch in diameter. The two
fluids flow countercurrently through the two pipes, exchanging heat
through the walls of the pipes.
Air-cooled heat exchanger (Figure 62) uses ambient air as the
cooling medium. The fluid to be cooled passes through the inside of
the tubes and a fan induces or forces a flow of air to the outside sur-
faces of the tubes. The fluid is cooled by transferring some of its heat
through the walls of the tubes to the air. These units are relatively
large, varying between 10 and 20 feet wide and 10 to 40 feet long. Often
these units are stacked on top of and along the pipeway in the refinery.
Miscellaneous exchangers include plate coils, steam coils, tank
heaters, and box coolers. Most of these exchangers are in steam heat-
ing, water cooling, or emergency services and are of little importance
from the pollulant emission standpoint.
165
-------
OUfLfc-T
Figure 60. Shell and tube heat exchanger.
-------
cr-
-vl
x^*=
t
t
Figure 61. Double-pipe heat exchanger with longitudinal fins.
-------
AXIAL
(2.1 U<£
Figure 62. Air-cooled heat exchanger.
168
-------
Heat exchangers normally do not emit pollutants. However,
because of the corrosion of the tubes or leakage between the shell and
the tubes, it is possible to contaminate the lower pressure fluid with
the higher pressure fluid. When the lower pressure fluid is cooling
water, the hydrocarbon contamination will be released in the cooling
tower (see Section Z, Cooling Water System in Chapter I). If steam
condensate gets contaminated with hydrocarbons, these contaminants
will be released to the atmosphere with the noncondensables in the
deaerator. By checking the composition of the effluent gas leaving the
deaerator and the vapor leaving the cooling tower, leakage into the
steam and cooling exchange systems can be detected. In some cases,
leakage to the cooling water system can be detected by observing an oil
film on the water in the cooling-tower basin.
E. FURNACES
Furnaces are used for heating refinery fluids. These units are
also known as "fired heaters", "tube stills", or "pipe stills". In
general, refinery proces s streams are heated by exchanging heat with
other hot streams up to 400°F. When a process requires a higher
temperature level or when other hot streams are not available, the
direct-fired furnace will be used.
Refinery furnaces have many shapes and forms and varying
firing and tube arrang ements. Two of the more common units, the
vertical cylinder and thr horizontal box are illustrated in Figures 63
and 64.
169
-------
x\
Figure 63. Vertical cylindrical furnace.
170
-------
Figure 64. Horizontal box type fired heater.
171
-------
Most furnaces have two main sections: the convection section
and the radiant section.
In the convection section, heat is exchanged between the hot
combustion gases on the outside of the convection tubes and the process
stream inside the tubes. Often additional tubes will be added in this
section of the heater for superheating steam. The recovery of heat in
this section greatly influences the total heater efficiency and reduces
the stack flue-gas temperature. Flue gas temperatures in the stack
vary between 500-700° F in an efficient heater and 900-1300°F in an
inefficient furnace with reasonable amounts of excess air.
In the radiant section, heat is transferred to the tubes mainly by
radiation. The refractory walls of the heater get red hot and emit heat.
The tubes absorb this heat and transfer it to the process streams inside.
The temperature of the walls in this section will vary from 1000 to 2000°F.
Fuel is delivered to the furnace through the burners. The
•function of the burners is to mix the fuel and the air, maintain a flame
of proper shape, size, and stability, and ensure complete combustion.
There are some 20 basic burner designs and many variations of each
basic design.
A typical gas burner is shown in Figure 65. Natural gas or
refinery gas is delivered to the burner at 3 to 20 psig. The fuel gas is
mixed with the primary air, and the mixture is injected into the fire
box of the furnace; the secondary air is drafted or forced through the
air registers into the fire box where the combustion takes place. The
172
-------
AlR.
Figure 65. Gas burner.
173
-------
fire box is maintained at a. low vacuum (up to 0. 5 inches of water) and
thus any flow of air is inward with little danger of the flame burning
outside the firebox.
When heavy fuel oil is used for firing the furnace, the oil must
be atomized, either by mechanical means or by steam. The combina-
tion oil-gas burner shown in Figure 66 is a steam-atomized fuel oil
burner. The steam and fuel oil are fed simultaneously to the oil gun at
about 100 psig. They form an emulsion like mixture in the oil gun, mix
with the primary air, and are injected into the fire box much like the
mixture in the natural gas burner.
The combustion product leaves the furnace through the stack.
Stacks are designed to induce draft (low-level vacuum) in the fire box,
dispose of the hot gases high enough above ground so that the fire dan-
ger is minimized, and reduce the ground-level concentrations of the
combustion products. Both the burners and the stack create noise and
may require mufflers.
Coke will deposit on the inside of the tubes when the furnace
processes hydrocarbons. These deposits may be removed from the
tubes either by mechanical cleaning (turbining) or by steam air de-
coking. The steam air decoking process is divided into two portions.
First the tubes are heated by partial firing of the furnace to about 300°F
and steam is allowed to flow through the tubes. The external heating of
the tubes produces shrinking and cracking of the coke (spalling) inside
the tubes and the steam blows the loose coke out of the tube.
174
-------
TILE
Figure 66. Combination gas and oil burner.
175
-------
Some of the steam will react with the coke to form CO2, CO,
and H2. Coke particles, steam, and gases come out of the tube and
into a concrete pit or drum; the decoking effluent is then cooled with
water. All of the gases and most of the steam with some coke particles
are discharged from the decoking drum into the atmosphere. When the
spalling stops, compressed air is slowly added to the steam, and the
coke deposits inside the tube start burning. The reaction of coke and
oxygen produce CO2 and CO. At this point, if any sulfur is present in
the coke, SO2 is formed. Current practice is to discharge the decoking
effluents into the atmosphere. This may not be a severe problem if the
furnace has to be decoked infrequently. Otherwise, flaring the effluents,
or incineration of the gases in another furnace, may be required.
The emission of pollutants from furnaces is dependent on the
fuel used for firing, the proper operation of the furnace, and - to a
lesser degree -the design of the fire box and burners. Natural gas or
refinery gas is the least polluting fuel if it is properly treated for the
removal of all the sulfur components. All of the sulfur in the fuel oil
or the fuel gas will be converted to SO2 or SO3 in the process of com-
bustion.
Noncombustible residue in fuel oil, such as ash and metals, will
be discharged into the atmosphere as soot and ash. Improper atomiza-
tion of the fuel oil in the burner will produce unburned carbon in the
flue gas. Refinery furnaces, in general, operate with high excess air
176
-------
(15-25% excess air when gas fired and 30-50% excess air when oil fired).
High excess air produces complete burning, reduces flame temperature,
and minimizes the amount of carbon monoxide in the flue gas. However,
high excess air reduces furnace efficiency. NOX compounds are pro-
duced in the furnace from both the nitrogen in the air and the nitrogen
compounds in the fuel. Only a very small portion of "the nitrogen in air
is reacted, while the generally accepted opinion la that most of the fuel
nitrogen is converted to NO . Recent research on the formation of NOX
compounds, in furnaces indicates that a reduced flame temperature can
reduce the rate of formation of NO . Redesign of burners and fire
boxes may be required in the future to reduce the flame temperature
and with it the formation of NOX.
F. JET EJECTORS
A jet ejector is a gas compressor that vises the dynamic energy
of one fluid to compress another. The unit shown in Figure 67 consists
of a nozzle, a vacuum chamber, and a diffuser tube. A high-pressxire
gas (normally steam) is expanded in the nozzle. It passes through the
vacuum chamber at high velocity, entraining the surrounding vapor.
The mixed pases then enter the diffuser where the kinetic energy in the
gas is recovered as the gas decelerates, leaving the ejector at a pres-
sure significantly higher than the pressure in the vacuum chamber.
A jet ejector can be used to move large amounts of gas from a
low-pressure system to a higher pressure with a very low investment
177
-------
Figure 67. Steam jet ejector.
178
-------
in equipment. Consequently, ejectors are widely used in refineries in
such services as tank evacuation, batch distillation, drying, and re-
moval of inert gases from condensing operations.
The effluent gas from a steam-jet ejector consists of a mixture
of process gas and steam. Frequently, these gases are discharged
directly to the atmosphere (point 1 in the figure). If these gases con-
tain significant amounts of hydrocarbon or sulfurous gases, this dis-
charge could amount to significant pollutant emissions.
In many systems, the discharge from the jet ejector is cooled
to condense the steam. The condensate removes some of the hydrocar-
bons and some of the acid gases such as H2S. The condensate should be
checked for sulfides and hydrocarbons and routed to the appropriate
treating facility if they arc present. The efflxient gas from such a con-
denser should consist primarily of inert gases, but the vent gas should
be analyzed to assure that the gases are being effectively scrubbed.
To obtain a high vacuum, two (or more) electors are sometimes
connected in series. In some cases, an interstage condenser may be
used to reduce the volume of gas to the second-stage ejector. The inter'
stage condenser will operate at a reduced pressure and, therefore, will
not be effective in absorbing acid gases and light hydrocarbons. Where
these materials are present in the process gas, a condenser on the
effluent of the second-stage ejector would be beneficial in reducing
pollutant emission.
179
-------
G. PIPE VALVES AND FITTINGS
1. Pipe
The pipe used in refineries in hydrocarbon service is primar-
ily carbon and alloy steel. Carbon steel is used in ambient to moder-
ate temperature service, and alloy steel is used in high-temperature
service. In general practice, threaded connections are used for the
smaller sizes (1-1/2 inches and under) and flanged connections are
used for piping that is 2 inches and larger. Wherever possible,
welded joints are used to connect sections of pipe to minimize leakage.
Specifications for pipe and fittings for refinery service are nor-
mally very conservative. It is reasonable to assume that if the pipe is
in good condition, it will perform satisfactorily and will not contribute
o
to atmospheric pollution. However, if the piping is allowed to corrode
on either the interior or e.xterior surface, a significant reduction of
wall thickness can occur with resultant leakage of the process fluid.
Threaded connections, when first assembled, can be made leak
tight. However, if a threaded connection is assembled and disassem-
bled many times, the threads become deformed and leakage becomes
a probability.
Piping that is designed to hold very hot or very cold fluids must
be capable of expansion or contraction. In some cases an expansion
joint is installed in the line, but in most cases the line is designed with
sufficient flexibility to absorb the change in length. Such lines are in-
clined to leak either at the expansion joint or at a terminal point in the
line that is highly stressed by the pipe expansion.
180
-------
2. Valves
A refinery uses many types of valves. Gate or ball valves are
used where complete shutoff is required. Globe or plug valves are
used where the valve is used for throttling or process control. Check
valves are used in a flow line to prevent backflow. Control valves are
used to regulate process flow.
All of these valves tend to leak through the valve seat when
handling low viscosity fluids and/or when the pressure differential
across the valve is large. Where a single valve is used to separate a
material that is considered a pollutant from the environment, the outlet
of the valve should be closed with a blind flange, a blind, or a plug, to
assure that the material does not escape.
The second major source of valve leakage is the packing that is
used to .seal I ho valve stem. This packed joint must slide along the
valve stem as the valve is opened and closed. A regular program of
inspection should be scheduled to ensure that valve stem packings are
properly maintained to prevent leakage to the atmosphere.
3. Flanges
Flanges provide a removable connection between pipe and ves-
sels or other items of equipment. Flanges are specified by pres sure
rating and by facing. Pressure ratings used are: 150, 300, 400, 600,
900, 1500, and 2500 psi. The most common flange facings, shown in
Figure 68, are flat face, raised face, tongue and groove, and ring
joint. It is important that two opposing flanges have the same rating
181
-------
00
Kg
m
/)
m
MALE.
PLAT
Figure 68. Flanges.
-------
and facing. If mismatched flanges are connected together, there is a
good possibility that the joint will warp and leak.
4. Vents and Drains
Most process piping is installed with vents on the high points of
the lines and with drains at the low points of the line. These connec-
tions facilitate startup and shutdown procedures. They also constitute
a prime candidate for leakage of the process fluid. Vent and drain con-
nections on lines containing volatile hydrocarbons should be closed and
sealed with a pipe plug or blind flange to assure that no leakage will
occur at those points.
Many vessels containing hydrocarbons are equipped with manual
drain valves which can be used to separate water from the hydrocarbon
that is in the vessel. In this operation, an operator is supposed to
visually observe when the water flow stops and the hydrocarbon flow
starts. At this time the valve should be shut to minimize loss of hydro-
carbon. Unfortunately, such manual separation operations are fre-
quently left unattended, with the result that large quantities of hydro-
carbon are dispersed into the environment.
H. PRESSURE-RELIEF DEVICES
All vessels, tanks, heat exchangers, and other equipment
capable of being pressurized should be equipped with pressure-relief
devices that will protect them from too much pressure. Many different
types of pressure-relief devices are used for this purpose. The
183
-------
principal types of relief devices that may be encountered in a refinery
are the spring-loaded relief valve, the rupture disc, and the relief
hatch. Each device has different operating characteristics and failure
modes. Knowledge of these characteristics will permit the FEO to
determine if a given device is performing effectively.
Relief devices may be installed for relief of gaseous or liquid
pressure. The liquid relief devices are normally installed to relieve
thermal expansion and are less inclined to leak than vapor relief
devices. Vapor relief devices are emphasized in this discussion
because they are the primary source of pollutant emissions.
1. Spring-Loaded Relief Valve^
A typical spring-loaded relief valve is shown in Figure 69.
Fluid pressure is maintained in the throat of the valve. If the operating
pressure exceeds the valve set pressure, the valve will open and relieve
the system pressure.
Relief valves may discharge into a closed flare system or di-
rectly to the atmosphere depending on local air pollution regulations.
If the discharge from the valve goes to a flare system, then leakage
through the valve does not constitute pollutant emission. However, if
the relief valve discharges to the atmosphere, significant amounts of
material could be dispersed.
Relief valves, as delivered from the manufacturer, arenormally
gas-tight. However, after the valve has been actuated one or more
times, it is possible that the valve may leak. In an effective refinery
184
-------
Figure 69. Relief valve.
185
-------
maintenance program, all relief valves would be removed and checked
for performance and leakage at least once a year.
Most relief valves achieve a seal by forcing a metal plate
against a metal seat. These seats are durable but tend to leak. A
more effective seal can be achieved by use of an "0" ring or plastic
seat where process conditions permit their use.
Relief valves are normally set to activate at a pressure that is
10% or 25 psi above the normal operating pressure of a vessel. If the
operating pressure is too close to the set pressure, the valve will tend
to lift off the seat during normal operating cycles with resultant dis-
charge of material through the valve and possible damage to the valve
seat. The set pressure of each valve should be reviewed in relation to
the operating pressure of the system to assure that the two pressures
are separated sufficiently to prevent loss of material.
2. Rupture Disc
The rupture disc or burst diaphragm is a metal diaphragm
which is designed to rupture at a predetermined pressure. Figure 70
shows a typical rupture disc assembly. The rupture disc is used to
protect systems which might experience a sudden pressure rise or
which must vent large quantities of gas in a short time.
In continuous processes, use of the rupture disc is limited be-
cause the entire system must be depressurized to replace the disc.
Rxipture discs are usually used on batch systems which can be readily
shut down.
186
-------
Figure 70. Typical rupture disc installation.
187
-------
Rupture discs can provide a vapor-tight seal when initially in-
stalled. However, it is possible for the disc to develop a pinhole leak
due to atmospheric or internal corrosion. If such a leak should develop,
significant amounts of material could be lost before the leak is dis-
covered.
The effluent from a rupture disc may be vented directly to the
atmosphere or it may be discharged into a flare system. Rupture
discs which discharge directly into the atmosphere should be checked
frequently for leakage.
3. Relief Hatch
The relief hatch is designed to provide a large emergency relief
opening for handling large volumes of vapor at low pressure or in case
of internal explosion. Figure 71 is a typical section of such a device.
These relief devices normally discharge directly to the atmos-
phere. If the device becomes warped or if the gasket surface deterio-
rates, significant quantities of material may escape to the atmosphere.
I. FLARES
The primary purpose of a flare is the safe disposal of waste
gases by combustion. Except during rare emergencies, it should be
possible to accomplish the complete combustion of the waste gases
without producing smoke or noise.
The flare shown in Figure 72 is representative of one of a num-
ber of acceptable designs. It is an elevated flare with a smokeless tip.
188
-------
VACUUM
Figure 71. Pressure relief hatch.
189
-------
PILOT BURNER
FLARE TIP
STE/4M LINE
P/L07
IGNITER
SUPPLY
L/N£
^3 f* ^J ^\ i M
^^ ^^ /\ ^x^^y
ELECTRIC
IC,MITtOM
SOX
T— G^IS
>« / a
X DPX\IN
Figure 72. Elevated flare.
190
-------
The flare consists of a vertical pipe extending to an elevation suffi-
ciently higher than surrounding equipment and the ground so that the
heat given off by the flame will not harm personnel or equipment. A
pilot burner and ignition system is provided to insure that a flame is
maintained during venting. The flare burner tip is equipped with a
series of steam jets so placed that turbulence is produced in the flame,
which leads to smokeless combustion. Smaller flares sometimes use
air provided by a blower in place of the steam, but larger flares gen-
erally employ steam. To conserve steam, a control valve in the steam
line is activated by a flow sensor in the flare line.
Another type of flare frequently encountered in refineries is the
burning pit. They are generally reserved for very large gas flows
during major emergencies and are connected to the blowdown system
through a deep liquid seal that only opens when the elevated flare is
overwhelmed. A simple burning pit may consist of a circular area
enclosed by a wall, with inward-facing burners piercing the wall at
intervals.
For additional information refer to "Air Pollution Engineering
Manual" (Second Edition), John A. Daniels en, available from the
Government Printing Office as EP4. 9:40-2.
J. KNOCKOUT DRUMS
A knockout drum is a device for separating entrained liquid
from a vapor stream. The separation is accomplished by increasing
191
-------
the area of the cross section available to the flowing stream, thus re-
ducing the flow velocity and allowing the entrained liquid droplets to
settle by gravity. In some knockout drums, the separation is improved
by adding a mist eliminator in the upper section of the drum just ahead
of the vapor outlet. The mist eliminator can take the form of a knitted
wire pad that separates fine liquid droplets from the vapor by impinge-
ment. Figure 73 is a sketch of a typical vertical knockout drum.
Horizontal knockout drums are also very commonly used. The level
control shown on the figure is not a required feature and may be omitted
where the volume of liquid to be separated is small.
K. SCRUBBERS
Scrubbers are devices for contacting a vapor with a liquid for
the purpose of removing a contaminant from the vapor. For instance,
a hydrocarbon gas stream containing H7S can be scrubbed with a sodiurn
hydroxide solution (caustic solution) to remove the H2S. Another
example would be the scrubbing of a gas stream with water to remove
dust carried by the gas.
Figure 74 is a sketch of one of the many possible scrubber de-
signs. In the sketch, ceramic saddle packing is used to promote inti-
mate mixing of vapor and liquid. Other forms of packing and various
types of trays can be used, or the vessel can be left empty with vapor-
liquid contact being achieved by spraying the liquid into the vapor. The
scrubber shown uses closed liquid circulation, implying that the liquid
192
-------
VAPOR
OUTLET
INLET
VAPOR
LIQUID
MIST 5L/M/NM70R
P/P£
LEVEL
VALVE
LIQUID
CONTROL OUTLET
Figure 73. Knockout drum.
193
-------
OUTLET VAPOR
LIQUID
DISTRIBUTOR
CERAMIC
SADDLE
PACKING,
/=>n/P U—
VAPOR
^ /'N X'N
SCRU&&IN&
LIQUID
fl-
-tx-
CIRC UL ATION
PUMP
Figure 74. Scrubber.
194
-------
will be replaced periodically. Other systems might employ once-
through liquid circulation.
L. FRACTIONATORS
Fractionators are sometimes called fractionating towers, or
distillation columns, or stills. They are liquid-vapor contacting de-
vices that achieve a separation of feed components based on the differ-
ence in boiling points of the components. When a mixture of hydrocar-
bons is boiled the vapors produced are richer in the lighter, or lower -
boiling-point components. If the vapors are condensed and the process-
repeated further enrichment of the lighter components will be obtained
in the vapor. Similarly, if the liquid from the first step is boiled, the
remaining liquid will contain a still higher concentration of heavy com-
ponents. The fractionator shown in Figure 75 is a device to accomplish
these steps in a continuous fashion.
Feed is introduced near the middle of the column onto the feed
tray. Vapor rising from the tray below condenses in the liquid on the
tray, vaporizing some of the liquid which travels up to the next tray.
In this way, vapor originally formed by the reboiler at the bottom of
the column supplies a vapor stream that passes up through the column,
leaving as overhead vapor. The overhead vapor is condensed to liquid
in the condenser. Part of this liquid is overhead, or light, product
and the remainder is returned to the column as reflux to provide a
source of liquid to the trays.
195
-------
OVERFLOW
WEIR
t\
J
Q
COND&NSBR
REFLUX
ACCUMULATOR
•*—S
REFLUX
PUMP
TRAY
35£ C K
T/?/W
REROILER
RE&OILtK
VALVE
X^>
BnnB^
BOTTOMS
KISEIZ
SLOT
BOTTOMS PUMP
Figure 75. Fractionator.
196
-------
The figure shows a fractionator that uses trays as vapor-liquid
contacting devices. Details are shown for both valve trays and bubble
cap trays, which are two of many types of trays. The combination of
downcomer and overflow weir shown is used to maintain a liquid level
on the tray and a vapor seal so that the vapor is forced to flow through
the valves or caps. It is also possible to use shaped solids such as
Raschig rings, Intalox saddles or Pall rings as contacting material.
The column is filled with the rings or saddles and is called a packed
column.
197
-------
III. PROCESS INSTRUMENTATION
A. INTRODUCTION
Process instruments are widely used in refineries to indicate,
record, and control the operating conditions and plant performance.
Early processing plants contained only indicating instruments to assist
the operator to control the plant operation manually. Modern refin-
eries are highly automated with instruments which control process
flows, temperatures, and pressures. Some refineries have computers
which process operating data and recommend or execute changes in
operating conditions.
B. IDENTIFICATION OF PROCESS INSTRUMENTS
Process instruments can be classified according to the variable
being processed and its function. The variety of instrument types is
best illustrated by the Instrument Society of America (ISA) code system.
This code system is used on all flow sheets in this manual. A two or
three-letter symbol is used for each instrument. A summary of the
commonly used code letters is presented in Table 8. The symbol
"FRC", for example, indicates an instrument which records and con-
trols the flow rate. Other symbols are used to indicate whether the
199
-------
Table 8. INSTRUMENT IDENTIFICATION CODES
Letter
A
C
D
E
F
G
H
I
L
M
0
pH
P
R
S
T
U
V
W
Process Function (s)
(First Position) (Second and Third Position)
Alarm
Conductivity Controller
Density
Electric variable Element
Flow
Glass or gage
Hand or manual
activation
Indicator
Level Logging, scanning
Moisture
Orifice, restriction
Acidity
Pressure
Recorder
Speed Safety
Temperature
Unit
Viscosity Valve
Weight Well
Example: PSV = Pressure Safety Valve
200
-------
instrument is board mounted in the control room or locally mounted
within the plant.
A process instrumentation and control system consists of sen-
sors that measure the process variable, transducers that convert the
signal to electrical or mechanical form, gages, recording devices,
transmitters that relay the signal to receiving instruments, control
valves, and other instruments that affect the process stream behavior.
Each of the major instrument categories is discussed below.
1. Indicators
An indicator shows the immediate condition of the process
stream. It might be a thermometer gage, level glass, or manometer.
In most refineries, operators take readings of various indicators in the
plant at regular intervals. These data are generally recorded on oper-
ating sheets and filed for a limited time.
2. Recorders
A recorder takes an electric or pneumatic signal from a remote
sensor and records it on a moving chart. Some recorders use circular
charts that rotate as a pen traces the process signal. The normal time
coverage of these charts is 24 hours. This type of recorder is widely
used, but the current trend is to use strip charts which trace a process
signal on a roll of paper as it moves from one roll to another. A strip-
chart instrument will accumulate data from several days' operation on
a single roll, which makes it easier to review and analyze the process
operation over an extended period.
201
-------
Care must be exercised in reading instrument charts. The
charts are frequently graduated in units from one to ten or from one to
one-hundred. The proper instrument value is obtained by multiplying
the chart reading by a factor which is unique to that instrument system.
These factors may be written on the instrument face or recorded in the
instrument log book. The field engineer should check with the oper-
ating foreman to confirm that the correct factors are being used.
Care is also required in reading pulsating or cycling signals.
An average of the high and low signals can be used to estimate the
average value. The maximum reading can be used to estimate the
instantaneous peak value.
3. Transmitters
Transmitters are devices used within the plant to relay signals
from the measuring point to the receiving device. Transmitters are
generally the electric or pneumatic type. The pneumatic transmitter
in conjunction with force-balance measuring devices is in wide use
today.
This system operates on a low-pressure instrument air system
with signals transmitted through small-bore tubing.
4. Controllers
A controller is an instrument that receives a signal from the
sensing element or transmitter, compares the signal to a predeter-
mined value or set point, computes the action required to align the pro-
cess variable, and executes the corrective action. Controllers operate
Z02
-------
on electrical, mechanical, hydraulic, and pneumatic signals. The
pneumatic system is the most common type found in refineries.
5. Control Valves
The control valve is the final device in a control system. The
valve opening is varied according to the pneumatic signal received.
The valve position is adjusted through a diaphragm or bellows which
responds to pressure changes. Control valves are discussed in more
detail in Chapter II, Refinery Equipment.
C. FLOW MEASUREMENT
Flow metering devices fall into three categories: positive dis-
placement, variable-area, and variable-head meters .
1. Positive Displacement Meters
The nutating piston or nutating disk meter is the most commonly
used positive displacement meter. This meter is used primarily for
liquids. The lobed rotor meter contains two lobed impellers and is
used chiefly for high volume gas flow metering. Other types include the
rotary vane meter (gases and liquids) and the liquid-sealed gas meters.
Velocity meters are based on the turbine principle. The velocity
of the fluid actuates an impeller whose speed varies with flow rate. The
turbine meter is the inost frequently used type of velocity meter.
2. Variable-Area Meters
The piston meter and rotometer are based on the variable-flow
area principle. A rotometer is a vertical tapered tube containing a
203
-------
float. The level of the float is determined by the balance between the
weight of float and the force of the fluid passing between the float and
the tube wall. The float level is read directly on a scale on the trans-
parent tube.
The piston meter contains a piston which rises as the flow in-
creases. As the piston rises additional orifice area is exposed. The
level of the piston provides a direct indication of flow rate.
3. Variable-Head Meters
The most frequently encountered variable-head meters measure
the pressure drop across a constriction in the line. An orifice plate or
flow nozzle is often used to create the pressure loss. Streamlined
restriction tubes, known as venturi tubes, are also used. A pitot tube
measures the local fluid velocity in the line. A tube is placed in the
flowing fluid facing upstream. The velocity pressure is converted to
static-head reading. Pitot tubes can be used as portable instruments
to measure gas flow rates from vents and ducts.
Weirs and flumes are used to measure flow rates in open
channels. A weir is a dam with a notched opening. The height of the
level in the notch indicates the flow rate. A flume is a narrow throat
in the open channel. The level of the quiet fluid upstream of the flume
indicates the flow rate.
D. TEMPERATURE MEASUREMENT
Temperature measurements in refineries are made with thermo-
couples, thermometers, and radiation pyrometers.
204
-------
1. Thermocouples
Thermocouples are widely used in refineries to measure the
temperature of process streams. When two dissimilar metals are con-
nected in a circuit, an electric current will flow if a temperature differ-
ence exists between the two junctions. The thermocouple is placed in
a protective tube or thermowell. Voltage signals from thermocouples
are used to operate controllers, recorders, and multiple readout sta-
tions in control rooms.
2. Thermometers
The common liquid-in-glass thermometer is used as a tempera-
ture indicator in refineries. More common, however, is the bimetallic
thermometer or temperature gauge. When two dissimilar metal strips
are attached, a temperature change will cause unequal changes in
length in the strips resulting in a deflection. Temperature gauges are
placed in the plant to obtain local readings. Often only a thermowell is
provided requiring the operator to insert a thermometer to obtain a
reading.
3. Radiation Pyrometers
Radiation pyrometers are generally used where temperatures
are extremely high or where other devices cannot be used in direct
contact with the heat source. Radiation pyrometers measure the in-
tensity of radiation from a hot source. A photocell or other detector
converts the radiation to electrical current. Optical pyrometers com-
pare the radiation intensity from the object in question to another hot
205
-------
source of known temperature such as a tungsten filament.
4. Resistance Thermometers
The resistance thermometer and filled-system thermometers
are also used in refinery applications. The resistance thermometer is
based on the change in electrical resistivity of metals with temperature.
The element is placed in a protective tube similar to a thermowell.
The filled-system thermometer measures the pressure change of an
enclosed volume of gas or liquid.
E. PRESSURE MEASUREMENT
Pressure measuring instruments can be classified as elastic
elements, gravity-balance manometers, and electrical pressure in-
struments.
1. Elastic Elements
Pressure indicating gauges commonly found throughout refin-
eries operate on elastic elements, such as bellows, diaphragms, or
bourdon tubes. When the elastic element is strained, a deflection indi-
cates the pressure level. Sealing fluids are often used to protect the
element from the process fluid.
2. Gravity-Balance Manometer
A U-tube or well filled with a liquid column will register a level
differential under pressure. One side of the manometer is connected to
the unknownpressure source and theother side is connected to atmo-
sphere, vacuum, or other known pressure. A manometer can also be
used to directly measure differential pressures on flow elements such as
an orifice plate.
206
-------
3. Electrical Pressure Instruments
The strain gauge operates on the principle of change in electric
resistivity of a wire under strain or deformation. The strain gauge is
used as a transducer to convert pressure signals to electric signals.
High-vacuum pressures are also measured using electric or electronic
principles. Resistance elements in a vacuum can be used to measure
pressure indirectly by measuring gas thermal conductivity. The
Knudsen gauge is a complex device for measuring vacuum pressures by
measuring, the deflection of molecules off heated vanes. Other types
are based on the principle of gas ionization.
F. LEVEL MEASUREMENT
The measurement of liquid levels and liquid-liquid interface
levels is widely practiced in refineries to control process operations.
The simplest method of level measurement is direct observation. This
can be done with a manually operated gauge tape or an externally
mounted sight glass. Indirect methods of level determination arebased
on the location of an internal float, differences in physicalproperties of
the two phases, or static head. Each of these is discussed brieflybelow:
1. Float Devices
The simple ball-float mechanism is widely used in refinery ser-
vice to control liquid levels. A rod, attached to the ball float, is used
to operate a level indicator, inlet valve, or pilot relay which, in turn,
operates a pneumatic control system. In some cases, the float is
207
-------
located in a float cage connected to the vessel. Magnetic floats are also
used to indicate levels.
2. Displacer Devices
The displacer type of liquid-level indicator has largely replaced
the ball float. The displacer is a tube lighter than the liquid whose
level is being measured. As the liquid level rises the displacer under-
goes a buoyant force. This force is translated to torque in a rod or to
pressure in a bellows located externally. The rod rotation or the bel-
lows pressure is used to indicate level or transmit signals to a control
system.
3. Hydrostatic Methods
Numerous hydrostatic methods are used to measure the static
head of the liquid in a vessel. One method connects the liquid under
static pressure to a diaphragm box. The diaphragm pressure is then
metered to determine level. Mercury manometers and other differen-
•tial pressure instruments are also used to measure static head.
G. ANALYTICAL INSTRUMENTS
A variety of special instruments is used to monitor product
quality and control refinery operations. Some of these instruments are
unique to the petroleum industry. Each refinery has a laboratory for
quality control testing. In large refineries, the laboratory may contain
sophisticated analytical equipment and perform limited research func-
tions. In most refineries, the laboratory performs only routine tests.
Z08
-------
These tests include determination of gravity, boiling point, flashpoint,
viscosity, and sulfur and metals content. Gas chromatographs are
occasionally used to analyze the composition of light hydrocarbon
streams.
In most cases, refinery laboratory data can provide only clues
to possible emission sources. Routine data on chemical and physical
properties of process streams are useful in determining the normal
plant operating charactistics. Deviation from normal operation indi-
cates a possible plant upset which could result in temporary emission
of pollutants.
H. COMPUTERS
Digital computers find widespread use in modern refineries and
are often applied in process control. Shortly after computers became
generally available, many people felt that within a few years entire
refineries would be under the direct control of on-line computers. This
did not come to pass, and at this time direct computer control is usu-
ally limited to sections of a processing unit or to a single piece of
equipment. Computers are, however, very widely used to collect and
display data and to provide information used in making process control
decisions.
Where computers are used for on-line control, electric rather
than pneumatic sensors and controllers are usually used. Signals from
primary sensing elements are converted to low voltage electric signals
209
-------
that are in turn converted to binary codes acceptable to the computer.
The computer is programmed to produce process control decisions
which are converted from binary code to low voltage signals for trans-
mission. The low voltage signal is used to control electric motors
that position valves, etc.
210
-------
IV. MONITORING INSTRUMENTATION
A. SOURCE MONITORING
1. Monitoring Systems
a. Approaches - Today source emissions are still mea-
sured manually. A sample is extracted from a point or series
of points in a stack or duct and are analyzed in the laboratory
or at some other remote site. The result of the analysis repre-
sents an integrated average of the emission parameter measured
over the length of time the sample was collected.
While the field enforcement officer may request that
such tests be made and may observe or even participate in
the conduct of these tests, his principal concern will be in data
that are immediately available and that can be correlated with
physical observations made during an inspection or investiga-
tion. Greater use is now being made of newer measurement
methods that permit immediate determination of some measure-
ment variables related to emissions and that also can be used
in the continuous monitoring mode.
Nader has categorized these newer methods according
211
-------
to three criteria: (l)Is measurement made in situ (at the point
of extraction)or on extracted sample; (2) Is a point measure-
ment made or is the measurement integrated across a path;
and (3) Is sensing on-site or remote? Since all extractive
monitoring inherently involves point sampling, only in situ or
remote sensing can involve the integrated sample mode.
Table 9 lists those approaches that are at present in use or that
seem feasible, considering the above criteria. Specific exam-
ples of these techniques will be given later.
Table 9. APPROACHES TO SOURCE MONITORING
Approach
Example
1. 'Extractive monitoring
2. In situ point monitoring
3. Integrated in situ
monitoring
4. Integrated off-site
monitoring
Continuous analyzer on stack
gas sample extracted from
point or series of points in
stack
Sensor placed directly in
stack at single point or
series of points
Spectrometric or optical
measurement made across
stack
Remote sensing Spectrometric
or optical technique. May
have active source or use
solar energy
b.
System Components - A source monitoring system con-
sists of far more than the sensor or analysis device used. In
the same reference, six component or component/operation
212
-------
combinations are given as parts of a source monitoring system.
These are reproduced in Table 10.
A major problem may be encountered in satisfying the
requirements of the site selection criteria as applied to source
monitoring. In conventional extractive source testing where
samples are removed for subsequent off-site analysis the New
Source Performance Standards of the Environmental Protection
Agency call for a minimum of twelve sample traverse points in
a stack if the sample location is eight stack or duct diameters
downstream and two duct diameters upstream from any flow
disturbance such as a bend, expansion, contraction or visible
flame. If these location criteria cannot be met, additional
traverse points are called for.
In the case of a continuous point source extractive moni-
tor or an in situ point source monitoring device which is used in
a permanent fixed installation, there will be only one sample
point. This point should be relatable in terms of contaminant
concentration to the true mean concentration established by
traverse prior to the fixed installation. This relation cannot be
expected to remain constant unless process and stack conditions
(e. g. , temperature and flow rate) continue to be the same as
when the relation was originally established. Therefore, the
field enforcement officer should determine, when observing such
monitoring equipment, the precautions which have been taken to
213
-------
Table 10. SYSTEM COMPONENTS AND OPERATIONS
FOR STATIONARY SOURCE MONITORING
Component /Ope ration
Requirement
1. Sample site selection
and execution
Sample transport
(when applicable)
Sample treatment
(when applicable)
4. Sample analysis
5. Data reduction and
display
6. Data interpretation
Representative sampling, consistent
with intended interpretation of
measurement
Transporting sample extract with
minimum and/or known effects on
sample integrity
Physical and/or chemical conditioning
of sample consistent with analytical
operation with controlled and/or
known effects on sample integrity
Generation of qualitative and quanti
tative data on pollutant or parameter
of interest
Calibrating and processing analog
data and display of final data in
format consistent with measurement
objectives
Relating the measurement data to the
source environment within the
limitations of the sampling and
analytical operations
Z14
-------
assure that the relation between the contaminant concentration
as determined by the fixed monitor and the true concentration
can be reasonably established. This problem is much less
severe with an integrating or across the stack monitor, but still
cannot be neglected.
A variety of other problems may be related to sample
point or sample point environment. For example, in the case of
an extractive monitor for sulfur dioxide in the tail gases from a
sulfur recovery unit, any residual hydrogen sulfide may con-
tinue to react forming sulfur which may cause erroneous ana-
zer readings or plug the sample line. The American Petroleum
Institute has suggested that the following environmental in-
fluences be considered when locating process or source moni-
toring analyzers: radian heat, mechanical shock and vibration,
vulnerability to damage, electrical hazards, and weather.
Further, analyzer installation points should have safe access
for calibration, servicing and maintenance.
c. Monitoring Strategy - In addition to the sometimes
severe environmental problems associated with source moni-
toring, e.g., high temperatures, excessive moisture, and
presence of high contaminant concentrations, a major problem
exists in obtaining representative test data. This is a result of
the time variation in the process* and distributional variation
across stack or duct and fluctuations in contaminant loading.
215
-------
The' strategy approach in source monitoring should be to
(1) examine the process and equipment for likelihood of these
variations, (2) to pick a sampling point at which the variations
will be at a minimum, and (3) to select a monitoring approach
which, within available options, is best able to compensate for
the variations.
One of the primary arguments in favor of source moni-
toring, of course, is the potential for obtaining a continuous
record of contaminant concentrations, thus enabling a more
accurate determination of the true emission rate. This is
particularly true for gaseous contaminants. In the case of
particulate extractive monitors where time variations are
accompanied by flow rate changes, there should be an ability to
vary the sampling rate to maintain isokinetic sampling condi-
tions.
In the case of distributional variation, nonextractive
across-the-stack integrating monitoring devices are much
better able to minimize the effect of such variation.
2. Source Monitoring Interfaces
To present a source sample to a monitoring device in a form
that will ensure a correct analysis and to ensure that the monitoring
device will be able to function continuously and reliably, some acces-
sory items are usually required. These are commonly classified as
216
-------
interfacing equipment. This equipment is designed (at least in the case
of extractive monitors) to remove the sample from the stack; to trans-
port it; to clean it; to alter temperature, water content, and possibly
pressure; and to measure, and if necessary, to control its flow rate.
The selection and design of this interfacing equipment in many cases is
as important as the selection and design of the actual analytical device.
It often costs as much or more to acquire and install the interfacing
equipment as it does for the monitor itself. For example, in the case
of an electrochemical type extractive monitor for sulfur dioxide, the
purchase price of the analyzer is approximately $1, 100 and the total
installation cost typically ranges from $6,000 to $8, 000. In the case of
an extractive photometric analyzer costing approximately $12,000, with
installation charges ranging from $6, 000 to $8, 000 the total installed
cost can be $18, 000 to $20, 000. The important categories of inter-
facing equipment are discussed briefly below:
a. Probe and Materials of Construction - A probe is a de-
vice that is placed in the stack of an extractive monitor. Under
clean, dry conditions, it may be a simple tube curved at the end
to face the exhaust flow, or it may have an integral particulate
filter as a preliminary clean-up device. These filters are -
usually of ceramic or sintered metal design to withstand tem-
perature and to minimize adsorption effects for reactive gas
sampling. Further, the probe may require heating to prevent
217
-------
moisture condensation or may incorporate internal static pres-
sure taps to determine isokinetic sampling conditions in the
4
case of particulate monitors.
Materials of construction should be such as to reduce to
a minimum wall losses through reaction, adsorption, or other
depositions. Smooth, clean, nonreactive tubing such as boro-
silicate glass, Vycor, or stainless steel are preferred mate-
rials. If copper is used, it should be refrigeration grade. All
probes and tubing should be cleaned thoroughly prior to instal-
lation. All valves, fittings, and seals should be compatible
with the system and must be leak-proof,
b. Sample Conditioning - Sample conditioning is one of the
major problems in source monitoring. This operation can
include temperature reduction, moisture removal, particulate
» *
removal, and pressure reduction - all to be accomplished with
a minimal or known effect on the component of interest.
Temperature reduction and moisture removal may both
be done by cooling. In turn, the cooling may be accomplished
directly by refrigeration or indirectly by dilution, although the
latter step requires additional flow measurement and control.
Unfortunately, the removal of moisture by cooling leaves a
condensate which must be purged from the system. This is
done by a trap - a device permitting discharge of water without
loss of the gas sample. In the case of soluble gases such as
218
-------
sulfur dioxide, the moisture removal process by condensation
will also result in the removal of some SO2. Considerable
effort has been expended by manufacturers" of commercial equip-
ment to alleviate this problem. One approach is to pass the gas
through condensing systems as rapidly as possible with the trap
located away from the main gas flow path. This reduces expo-
sure of the gas stream to the condensate. Nevertheless, the
probable loss of SO2 should be evaluated, preferably by in-place
sample spiking or by calibration.
Unwanted particulate matter which may clog lines or
interfere with the measurement process is usually removed by
filtration. Depending upon the loading and the operating sched-
ule, a built-in inert gas back-flush system may be required to
periodically remove accumulated particulate matter which in-
creases back pressure and may cause increased sample losses.
The filter media itself should be nonreactive, strong, resistant
to moisture and corrosion, and nonadsorptive.
Recently, reverse permeation devices have been studied
which will permit a gaseous contaminant of interest to pass
from the sample stream through a semipermeable membrane
barrier into a nonreactive carrier gas which then flows to the
detector. The unwanted contaminating moisture and particu-
lates are retained on the sample*-gas side of the barrier.
Because the permeation of the gas of interest is not quantitative
219
-------
but is a function of partial pressure (concentration) in the orig-
inal stream, a careful calibration is required to translate the
apparent analyzer concentration to the true concentration.
These devices have not as yet been used extensively, but if
successful could become an attractive solution to sample
clean-up problems.
Pressure reduction may sometimes be necessary. If
so, those making the installation should be cautious about
effects of .possible freezing, ice deposit, or hydrate formation.
Figure 76 illustrates a typical sample system for an extractive
monitoring device showing major interface components.
c. Sample Transport and Flow Measurement - In the case
of extractive soxirce monitors, a sample gas stream, must be
made to flow through the interfacing equipment to the actual
sensor. This is accomplished by a vacuum-pro due ing device,
usually a pump, but sometimes an air or steam jet ejector.
The latter require auxiliary services which are usually avail-
able in a refinery, but they are quite reliable and less subject
to wear than are mechanical pumps.
In the case of concentration measuring sensors, flow
rate does not directly affect results unless pressure limits are
exceeded or unless pressure conditions inside the detector are
changed from those at calibration. Where an indirect measure
of mass is made, the sample flow rate enters directly into the
220
-------
ts>
SPAM GAS
1
Z.ERO GAS
_L
STACK WALL
PROBE/
PRIMARY
FILTER
* INCLUDES
PARTI C.LE FILTER,
PUMP; 4 SLOWDOWN
CONTROLS
SAMPLER/BLOWDOWN
UNIT*
SAFE DISCHARGE
ANALYZER RECORDER
FLOWMS.TER.
NEEDLE VALVE
DRYER
TRAP
COMPRESSED
A\ R
AUTOMATIC
DRAIN
Figure 76. Typical stack monitoring system.
-------
final calculation of emission rate. In any case, flow should be
at a rate to minimize wall losses and to avoid unusually long
response times which are not consistent with possible rate-of-
concentration variation.
Sample flow rates may be measured by a variety of
measuring devices including calibrated nozzles and orifices and
rotameters. The latter are probably the most commonly used.
They are indirectly sensitive to temperature by its effect on
carrier gas viscosity and are subject to wear of float and tube
which also affects calibration. They should be used with care.
As earlier mentioned, isokinetic flow conditions must
be maintained when sampling particulates where particle diam-
eter exceeds about 3 /jm. If sampling velocity in the probe is
higher than that in the stack, a disproportionate number of
smaller particles will be present in the sample. If probe ve-
locity is lower than stack velocity, a higher proportion of
larger particles will be present in the sample stream. It may
be desirable to have a sample flow rate controller which is
actuated by a stack flow rate sensing device.
If mass emission rates are to be obtained, total exhaust
flow must be measured regardless of whether extractive or
nonextractive monitoring equipment is used. Although pitot
4
tubes may be used as in conventional batch source testing, an
instrumental approach using a transducer which produces an
222
-------
electrical signal may be useful for permanent installations or
those used to control sample flow rate. Recently, a heated
thermopile velocity sensor has been marketed in which fluidic
control of an inert gas flow over the thermopile is utilized
instead of the stack gas itself. This permits operation in con-
taminated conditions without the necessity of cleaning the gas
stream to be measured.
3. Calibration
Calibration requires relating the readout of the source monitor
to the true concentration of the contaminant in the source being mea-
sured. A full-scale calibration involving the in-place introduction of a
known concentration of the contaminant to the analyzer under stack
conditions would be desirable, but to date, this has been difficult to do
"e 1"
-------
devices, the zero transmittance point can be obtained by blocking the
light path, but a true 100% transmittance can only be obtained during
shutdown conditions.
4. Source Monitoring Instruments
In recent years, many firms have been working on the develop-
ment of continuous source monitoring instruments. Relatively few of
the devices have reached full commercial status. The field enforce-
ment officer should be generally familiar with the types of continuous
monitoring systems in use or soon to be available. The principal types
for the major air pollutants and their principal points of application are
described here.
a. Gaseous contaminants
(1) Sulfur dioxide - The potential for sulfur dioxide emission
exists in any combustion operation in which sulfur-containing
fuel is burned - sulfur recovery plants, catalyst regeneration
and other decoking operations, and some treating units.
Most source monitors available for sulfur dioxide de-
termination are based upon the extractive monitoring concept.
The two having the widest application involve the nondispersive
infrared (NDIR) principle and an electrochemical approach. The
nondispersive infrared technique involves the absorption of
infrared energy by sample gas in a cell. Instead of using dis-
persive optical elements to obtain specificity, the infrared
absorbing properties of the gas of interest are used. This is
224
-------
done by using the gas of interest either in a sensitizing cell or
in a detector cell. In the first case, a differential thermopile
is used for detection and, in the second, a capacitor micro-
phone is used. The thermopile type is less sensitive but is also
less sensitive to vibration and shock. In any case, the NDIR
method is quite specific. It requires moisture and particulate
removal to keep optical surfaces clean. The measurement is
not highly sensitive to flow rate, but cell pressure must be
maintained at the same level as during calibration.
The electrochemical sensor operates by producing an
electric signal through electrooxidation of sulfur dioxide. In
this particular class of electrometric instruments, the bulk gas
flow does not pass through an electrolyte, but rather the gas of
interest passes through a semipermeable membrane to the cell.
As in the case of the nondispersive infrared analyzer, the in-
strument is not directly sensitive to flow rate but must be
operated at constant pressure. Moisture and particulates must
be removed and temperature is normally limited to 1 10° F.
Water removal systems for all sulfur dioxide monitors
should be designed so as to minimize contact time between the
gas stream and condensate to reduce the possibility of sulfur
dioxide adsorption. The dilution approach to cooling and hu-
midity reduction might be particularly attractive in sulfur
dioxide analysis systems.
225
-------
At least one commercial in situ sulfur dioxide analyzer
is available. It is based upon the principle of ultraviolet corre-
lation spectrometry. Although it is a dispersive technique,
specificity is obtained by comparing the spectrum obtained
with a correlation mask instead of selecting a particular wave-
length interval as representative of the compound of interest.
Physically, a tubular slotted probe is inserted in the stack so
that flue gas continuously passes through the slot. Ultraviolet
light shines through the flowing sample and is reflected by a
mirror at the end of the slot back to the spectrometer head.
An electrical signal proportional to the sulfur dioxide concen-
tration results which is processed and displayed. Compensa-
tion for particulate accumulation on the optics is provided by
automatic adjustment of the gain on the photomultiplier in the
detector to reflect overall light intensity. Obviously, this
accumulation cannot be permitted to continue indefinitely. Very
little practical experience is available with this analyzer.
(2) Carbon monoxide - The principal source of carbonmon-
oxide emissions from refinery operations is the regenerator of
a fluid catalytic cracking unit. Carbon on the catalyst is burned
off with less than the stoichiometric quantity of air. This
results in appreciable carbon monoxide formation instead of
total conversion of the carbon to carbon dioxide. In most
cases there will be an afterburner, commonly called a "CO
226
-------
boiler" in which the carbon monoxide is burned. While there
are other possible sources such as other types of decokers,
fluid cokers, engines, and incinerators, the most likely point
of application for carbon monoxide monitoring is in connection
with the fluid catalytic cracking regenerator.
The most frequently used technique for carbon monoxide
monitoring is the extractive nondispersive infrared method.
Fundamentally, the same type of sample conditioning equipment
and analyzer is used as previously described for sulfur dioxide.
The sensitizing cell would, of course, be filled with a carbon
*
monoxide mixture instead of sulfur dioxide.
Electrochemical sensors used in conjunction with extrac-
tive monitoring techniques are also available, but there is much
less experience with this method available to date.
It is entirely possible that in situ open path infrared
spectrophotometric procedures will come into use, but they are
still in the research and prototype stages.
(3) Nitrogen oxides - In refineries, almost all emissions of
nitrogen oxides are from combustion processes and are pre-
dominantly in the form of nitric oxide (NO). Therefore, heaters,
boilers, catalyst regenerators, furnaces, engines, flares, and
other miscellaneous combustion processes may be considered
as potential candidates for monitoring. It is likely, though, that
only the larger sources, even in the future, will have monitoring
227
-------
equipment installed.
There are many techniques for monitoring nitrogen
oxides. These include both extractive, in situ, and remote
sensing. These techniques are for the most part specific for
either nitric oxide or nitrogen dioxide. Thus in the case of the
extractive monitoring approach, where this is the case, con-
version of NO to NO2 to NO may have to be employed. Obviously
this is not practicable in the case of in situ open-path or remote
sensing techniques.
The extractive monitoring techniques available include:
• Electrochemical sensors - Fundamentally, these
use the same general principle as similar devices for
sulfur dioxide and carbon monoxide. The semipermeable
membrane and electrolyte are optimized to give the
necessary specificity. The commercially available sen-
sors respond to both nitric oxide and nitrogen dioxide,
although the relative response to nitric oxide is usually
greater. Where the ratio of nitric oxide to nitrogen di-
oxide is known and constant, this is not a serious
deficiency.
• Chemiluminescence - These analyzers operate
on the principle of light emission resulting from the
reaction between nitric oxide and ozone. The analyzer
incorporates a built-in ozonizer which provides a
228
-------
stream of ozonized air to react with the sample stream.
If nitrogen.dioxide is to be determined, the sample
stream must first pass through a catalytic decomposi-
tion section. Both particulate contamination and water
must be removed. Special care must be taken or opera-
tion must be at reduced pressure if SO2 or CO2 are pre-
sent in significant quantities, as both compounds act to
quench the chemiluminescent reaction. Several manu-
facturers offer commercial instruments. All must be
operated, calibrated, and serviced with considerable
care.
Ultraviolet-visible spectrometry - At least one
manufacturer offers a split-beam filter photometer of
this type for analysis of NO2. If NO is to be determined,
it must first be oxidized to NO2. One approach used is
to oxidize the NO with O2 at approximately five atmo-
spheres pressure. As is the case with most extractive
gas monitors, particulates and water must be removed.
Calibration is usually accomplished with a standard
optical filter, but the zeroing operation uses a zero gas
supply.
In situ open path or remote sensing is possible
for nitrogen dioxide with the use of ultraviolet correla-
tion spectrometry. The same general approach is used
229
-------
as when the technique is applied for sulfur dioxide.
Because nitric oxide is not detected, the application is
limited.
(4) Hydrocarbons - The major potential hydrocarbon losses
in a refinery are from storage and product transfer. In most
cases, source monitoring is not applicable. There are many
other smaller sources such as blow down systems, vacuum jets,
barometric condensers, air blowers, emergency vents, and
fume incinerators where monitoring conceivably could be
applied. So far, little effort has been made in this direction.
Three extractive monitoring techniques are applicable
to hydrocarbons. One is the nondispersive infrared method
previously discussed. The second is hydrogen flame ionization
detection in which the presence of hydrocarbons in a sample
stream flowing through a hydrogen flame increases the ion flow
between two electrodes thus producing a signal proportional to
the hydrocarbon concentration. The selective combustion ana-
lyzer technique is a third. Particulates and most water should
be removed from, the sample although the flame ionization
method is not so sensitive to the presence of water as the other
techniques.
(5) Miscellaneous gaseous contaminants - Hydrogen sulfide,
mercaptans, ammonia, phenols, and other organic compounds
may be discharged from catalyst regenerators, treating units,
230
-------
air blowers, hydrogen sulfide recovery systems, and fume
incinerators. One of the techniques previously mentioned for
other contaminants can also be applied to most of these com-
pounds. Monitoring for hydrogen sulfide is a more common
practice than is monitoring for the contaminants of lesser im-
portance. Again, most of the techniques applicable would be
based upon extractive monitoring.
b. Particulates - The term particulates, here, covers
smoke, dusts and fumes, and liquid phase aerosols. It is
obvious that a great variety of sizes, particles, densities,
shapes, optical properties, rigidity, and surface characteristics
may be covered by this definition. In a refinery, we shall
mainly be concerned with smoke, catalyst dust, coke breeze,
and liquid aerosols. Because of the high reliability of combus-
tion control now available, the major particulate problem from
refineries is catalyst fines lost from fluid catalytic cracking
unit regenerators.
(1) Optical - A variety of devices may be included in this
category such as single-particle, light-scattering instruments,
nephelometers, and opacity meters. The latter is the one in
current use for source monitoring and includes the long avail-
able smoke meters. Using the earlier mentioned classification
system, opacity instruments would be listed under the in situ
open path class.
231
-------
In recent times, considerable effort has been devoted to
the adaptation of opacity instruments so as to enable the deter-
mination of mass concentration from light transmittance read-
ings. Ensor and Pilat have recently described in some detail
the theory and assumptions necessary to make such calcula-
tions. In their expression, mass concentration is a function of
particle density, optical path length, optical density (In —V and
\ Io/
a "K" factor. In turn, this "K" factor is a constant which is
dependent upon particle diameter, wavelength of source light,
refractive index of the particle, size frequency distribution,
and scattering efficiency based upon particle shape. It is quite
obvious that the "K" developed for any given particulate emis-
sion is quite unique. Any change in the several parameters
listed above can affect the validity of the relation between light
transmittance and mass concentration.
A number of installation and operational problems affect
the reliability of opacity measurements in addition to those
related to fundamental design. In large part, these are related
to source and receiver alignment, cleaning of optical surfaces
exposed to the stack gases, and to problems of calibration and
zeroing. A variety of techniques has been used to facilitate
maintaining cleanliness of optical surfaces. They include using
recessed mounts with partial isolation of surfaces by purge air
to cleaning technique based upon air blast and water jets. Even
232
-------
with the best of these, some manual cleaning is necessary at
varying intervals.
Both infrared and visible light sources are used for
instruments now available commercially. While most systems
have the transmitted light sensing unit mounted opposite the
light source, at least one now offers a dual beam unit with a
corner cube reflector making possible the mounting of the
source and receiver in the same housing.
It is important that the field enforcement officer under-
stand the exact nature of any opacity device used for mass-load-
ing determination, its limitations, and its operating require-
ments .
(2) Particle mass sensors - At least two devices are now
available which are stated to be direct mass sensing devices .
Both are in fairly early stages of development as far as source
monitoring applications are concerned. The first described
uses the principle of beta radiation attenuation. In this approach,
an extractive sampling system dilutes and moves a sample
stream of stack gas to a sequential filter tape mechanism where
an integrated filter sample is obtained over some pre-set time
period. Following this, the tape is advanced so as to be exposed
to a beta radiation source. On the other side of the tape a
sensing device measures the attenuation of the beta radiation.
This attenuation has been found to be almost directly proportional
233
-------
to the mass between the source and detector. The procedure
does not give a truly continuous measurement but rather a
series of integrated values over finite time periods. As is the
case with an extractive sampling procedure for particulates,
provision should be made for isokinetic sampling. Also care
must be taken to maintain the integrity of the sample. Systems
presently available have three major components as follows:
(1) sampling probe and diluter, (2) collector unit containing
pump, filter tape system, and beta radiation gage, and
(3) control and readout units in which computation of mass
concentration is made from mass, and flow data. The collector
unit should be near the source while the control and readout
unit may be located remotely.
Another type of mass sensor involves the use of the
piezoelectric principle. Here particles in a sample stream are
deposited by electrical precipitation on a piezoelectric quartz
crystal oscillating at its resonant frequency. Since the reson-
ant frequency is mass dependent, the rate of change of this
frequency can be measured by an appropriate detector circuit
and related to the increase of mass on the crystal caused by
the particulate deposit. At higher concentrations, dilution
may be required. Also the effective particle size range of
one instrument is stated to be limited to . 01 to 20;um. Among
the operational problems that have to be considered are the
234
-------
requirement for periodic cleaning of the crystal and ensuring
that liquid water is not deposited and cause spurious readings.
(3) Electrical - One type of particulate monitoring device
is an extractive monitor which utilizes the charge transfer
principle. In this instrument, particles in the sampled gas
stream pass by a sensor without being collected but result in an
electrical disturbance changing the current flow in the detector
with a logarithmic dependence upon particle mass flow. Rela-
tively little experience is available with this type of analyzer.
Calibration is obviously a problem as is sample handling.
(4) Costs - A brief word is in order about present day costs
of source monitoring instruments and installations keeping in
mind that such figures may become dated rapidly. One of the
lower cost sensors is the electrochemical sensor. For the
sensor only, costs for a single parameter instrument are in the
range of $1, 000 to $1, 500. Total installed cost including site
work, sample handling system, and recorder may easily reach
$6, 000 to $8, 000. At the higher end of the range, a completely
installed spectrophotometric system may reach $18,000 to
$20, 000, based upon an analyzer cost of $14, 000. The installed
cost of one of the more expensive opacity instruments, including
purge air systems, isolation shutters, and other accessories
can reach $6, 000 to $7, 000. Other types of particle monitoring
devices including the mass sensors may have installed costs
235
-------
ranging up to $20, 000. It seems obvious that even for a large
company the installation of continuous monitoring devices can-
not be considered a trivial expense, particularly when the con-
tinuing costs for servicing, calibration, repair and data
processing are considered.
B. PERIMETER MONITORING
Contaminant monitoring as applied to the field surveillance of
petroleum refineries can be more broadly interpreted than direct
source monitoring alone. For one thing, not all sources of air con-
taminants are easily subject to direct source testing or monitoring.
Such sources would include pump seals, valves, fittings, storage tanks,
barometric condensers, oil water separators, pressure relief systems,
most emergency flares, and other relatively minor sources. Further,
it will be some time in the future before even most potentially major
sources will be equipped with fully effective direct source monitoring
instruments. To fill these gaps, the concept of perimeter or source
oriented ambient air monitoring may be applied.
Source, meteorological, and receptor oriented criteria are all
important in designing and establishing a perimeter air monitoring
effort. The source oriented information would include an emission
inventory for normal operation including points of emission and de-
tailed information on potential major point sources of emission includ-
ing stack location, stack height and maximum expected emission rate
in the event of control equipment failure.
236
-------
The meteorological data required involves wind speed and
direction and atmospheric stability. If available, a wind rose and
stability wind rose are most useful. A wind rose gives the frequency
of occurrence for each wind direction (usually 16 points) and wind
speed class (nine classes in standard Weather Bureau use) for the
period under consideration. A stability wind rose gives the same type
of information for each stability class. For maximum utility both
seasonal and annual wind roses should be obtained.
The source and climatological data thus developed may be
applied through the use of a diffusion model to predict points of maxi-
mum expected concentration and concentration isopleths for various
averaging times and seasons. Turner has prepared a workbook on
the practical application of these models while Stern has reported on
the current status of a variety of modelling techniques.
Receptor and effects oriented information are also useful in
planning a perimeter monitoring procedure. Should significant informa-
tion be available relating to locations of odor complaints and materials
or vegetation damage that clearly is related to past refinery operations,
it should be used to modify or support monitoring planning based upon
emission and meteorological data.
1. Continuous Monitoring
This approach involves the use of automatic, continuous, re-
cording air sampling instruments. Such instruments are normally
located in specially prepared fixed sampling sites or in mobile
237
-------
equipment. A significant investment is usually required and special
precautions are necessary to ensure proper calibration, operation, and
maintenance. Descriptions of such monitoring networks have been pre-
Q
pared by Bryan and the Office of Air Programs of the Environmental
9
Protection Agency .
Both fixed and mobile sites have advantages. In some cases the
criterion will be availability of land or shelter where the optimum site
locations have been determined through diffusion modelling process.
Because the modelling process can only approximate expected condi-
tions, the additional flexibility of a mobile station may be advantageous.
On the other hand, a mobile station is generally more costly than
equipping a permanent site and can require more manpower. It is
particularly important in the case of perimeter or source oriented
ambient monitoring systems to obtain wind speed and direction data
concurrently with the contaminant data.
As a general principle, perimeter monitoring is most effective
at points reasonably close to the probable source of expected contami-
nants. As distances increase, the contribution from a single source to
the total ambient air loading of a given contaminant is reduced and a
point will be reached where this contribution is nearly indistinguishable
by ordinary monitoring means. Obviously, the contribution from a
large point source in a remote area having no other point or area
sources could be distinguished at greater distances than could the same
source in the midst of similar sources.
238
-------
At the other extreme, contaminants discharged from elevated
point sources will not reach the surface of the ground closer than some
finite distance from the source. Even under unstable plume looping
conditions this distance may be several stack heights. Even though
these latter conditions give rise to high concentrations at point of im-
pact, the values can be expected to fluctuate rapidly due to vertical and
lateral plume eddying movements. Thus some compromise must be
reached in station siting so as to satisfy two conditions: (1) the sites
must be close enough so that the contribution from the source involved
can be clearly distinguished, and (2) the sampling or monitoring net-
work must be dense enough at the given distance from the source that
the emissions will be detected.
2. Integrated or Static Monitoring
Integrated monitoring systems may be either active or passive.
That is they may involve the active sampling of air by means of a
powered device moving air through or over a collection or sampling
device, such as a bubbler train or a filter, or they may involve use of
effects packages which are merely exposed to the normal free passage
of ambient air. This latter group would include sticky paper, greased
glass slides, sulfation plates, fabrics, metal plates, etc. In general,
a much denser network of intermittent or static sampling sites is prac-
tical as compared to continuous monitoring. This results from the
lower cost, lesser demands for space and power, and lower level of
manpower skills required.
239
-------
The resolution of data obtained from intermittent or static
monitoring equipment takes longer than that from continuous monitor-
ing so it is less suitable for episodic sampling. On the other hand, in
preplanned situations such as might be the case with the application of
an intermittent control strategy sufficient notice should be available to
make the use of intermittent air sampling devices practicable and
us eful.
C. PORTABLE SAMPLING EQUIPMENT
The field enforcement officer may find it advantageous to use
special portable sampling and analysis equipment during some phases
of refinery inspection or in the investigation of complaints. For one
thing not all potential sources of air pollution in .refineries are well-
defined stacks serving a major process or operation. Such miscel-
laneous sources would include leaks from pumps, fittings, and relief
valves either into air or into cooling water later passed through cooling
towers; oil-water separators; sumps; wastewater handling andstorage;
low-level minor process vents; storage tanks; low-level flares and
spillage.
In many of these sources, general housekeeping and mainten-
ance is closely related to the degree of contaminant loss. Further,
the character of contaminants likely to be lost from these miscellaneous
sources includes mahy of those of significant nuisance potential because
of odor. These include hydrogen sulfide, mercaptans, aromatic hydro-
carbons, amines, and other nitrogen bases to mention but a few.
240
-------
1. Use of Portable Sampling Equipment
The field enforcement officer should be very careful in the use
and interpretation of data collected by portable equipment. First of all,
by very nature of the sources involved, many of these losses may be
concentrated but very small in volumetric or mass flow. The actual
ambient concentration, particularly near the source, will be highly
variable and greatly affected by atmospheric or meteorological condi-
tions. Thus the inspector should record over a period of time the
values of a particular measurement made in any given area of the re-
finery or close proximity to establish a range of normal or expected
* •
values. In some cases, it will be more satisfactory to collect an inte-
grated sample using nonreactive flexible bags or miniaturized integrated
sampling devices, such as impingers, to obtain a more representative
value. Of course, if one is leak tracing, a fast response system is
desirable.
The field enforcement officer may also use portable sampling
equipment during investigation of complaints. This equipment is oper-
ated at the point of the complaints or in a systematic perimeter check
of the suspected source. Some knowledge of wind speed and direction
and atmospheric stability will be needed to interpret such data.
2. Types of Portable Sampling Equipment
A variety of sampling and analysis equipment is available
ranging from manually operated devices to portable, continuous,
direct-reading analyzers.
241
-------
One of the commonest portable devices is that based upon the
use of indicator tubes. These tubes are sealed ampules containing a
reagent on a solid substrate which are inserted into an air sampling
device immediately following the breaking of the two sealed ends. A
known amount of air is drawn through the tube by a squeeze bulb, hand
pump, or a battery-driven air pump. Following exposure, the tube is
compared against a standard by use of a length of stain measurement
or degree of coloration to determine the concentration of the contami-
nant in the air sampled. A great variety of indicator tubes is avail-
able, some of which have been tested by the Bureau of Mines or other
groups concerned with occupational health. The accuracy and sensi-
tivity are, of course, generally poorer than instrumental methods of
analysis. In some situations, it may be the only practicable approach.
As earlier mentioned, the field operation may be limited to
sampling only with analyses to be performed in the laboratory. Com-
monly this approach involves the use of sampling bags, liquid im-
pingers, filters, and perhaps precipitators or impactor devices. This
type of collection should be performed as directed by a competent
chemist or other air sampling specialist.
Finally, there are a substantial number of portable continuous
analyzers. Generally, they are equipped with a direct-reading meter
and only occasionally used with a miniature recorder. Perhaps the
most common such instrument is the combustible gas analyzer. In
this instrument an air sample is drawn over a heated element connected
242
-------
in an electric bridge circuit. Any combustibles present are oxidized
over the surface of the element causing a temperature rise and conse-
quent change in resistance of the element. By appropriate circuitry
this resistance change causes a meter displacement proportional to
concentration. However, the concentration shown is only relative and
the instrument must be calibrated in terms of the material of interest.
Other combustible gases or vapors will cause a response but the con-
centration indicated will not be correct.
Combustible gas analyzers have also been long available for
carbon monoxide. In this analyzer, the resistance element is in a
special catalyst and held at a temperature so that only carbon monoxide
will be oxidized and thus cause a response. There can be some inter-
ference from large concentrations of the more easily oxidized hydro-
carbons, such as ethylene.
Portable sampling instruments are available for hydrogen sul-
fide and sulfur dioxide. One for sulfur dioxide is based upon the con-
ductivity principle and uses a plunger pump to draw the sample. Other
instruments for hydrogen sulfide use the bromocoulometric or titri-
metric approach.
In recent years there has been considerable development of the
electrochemical cell type of instrument which is essentially the same
in principle as those used for fixed source or ambient air monitoring.
Analyzers based upon this principle are'available for H2S, NO2, NO,
C12, HCN, COC12, SO2, and CO. Specificity is obtained by the use of
Z43
-------
selective prefilters, semipermeable membranes, electrolyte material,
electronic circuitry» or a combination of one or more of these
approaches.
244
-------
REFERENCES
1. Nader, J. S. Developments in Sampling and Analysis Instru-
mentation for Stationary Sources. (Presented at the 65th Annual
Meeting of the Air Pollution Control Association. Miami Beach.
June 18-22, 1972.) Paper 72-39.
2. Standards of Performance for New Stationary Sources. Federal
Register. 3^:24876-24895, December 23, 1971.
3. Manual on Installation of Refinery Instruments and Control
Systems, Part II-Process Stream Analyzers, 2nd Ed. American
Petroleum Institute, Division of Refining, Washington, D.C.
Publication Number API RP 550. May 1965.
4. Morrow, N. L. , R. S. Brief, and R. R. Bertrand. Sampling and
Analyzing Air Pollution Sources. Chemical Engineering. 79:84-98,
January 24, 1972.
5. Ensor, D. S., and M. J. Pilat. Calculation of Smoke Plume
Opacity from Particulate Air Pollutant Properties. Journal of the
Air Pollution Control Association. 21:496-550, August 1971.
6. Txirner, D. B. Workbook of Atmospheric Dispersion Estimates.
Department of Health, Education and Welfare, Public Health
Service, National Air Pollution Control Agency, Washington, D.C.,
Publication Number 999-AP-26. Rev. 1969.
7. Stern, A. C. (ed. ). Proceedings of Symposium on Multiple
Source Urban Diffusion Models. Environmental Protection Agency,
Air Pollution Control Office, Research Triangle Park, N. C.
Publication Number AP-86. 1970.
8. Bryan, R. J. Air Quality Monitoring. In: Air Pollution, Vol. II,
Stern, A. C. (ed. ). New York, Academic Press, 1968.
9. Field Operation Guide for Automatic Air Monitoring Equipment.
Environmental Protection Agency, Research Triangle Park, N.C.,
Publication Number APTD-0736. November 1971.
245
-------
V. MAINTENANCE OF REFINERY RECORDS
FOR USE BY AIR POLLUTION CONTROL
FIELD ENFORCEMENT OFFICERS
A. INTRODUCTION
1. Necessity of Keeping Records
The recording and filing of operational data is a requisite for
efficient and economical plant operation in refineries and most other
industries. Much of this information is related to throughput, turn-
arounds, upsets, emergency venting, maintenance, source
monitoring, and ambient air monitoring and is, therefore, valuable
in the effective enforcement of air pollution control rules and
regulations. Field enforcement personnel, in addition to making
observations for violations of visible emission standards, perform
other duties which, in part, depend upon information from refinery
records. These other duties include:
• Emissions Inventories cataloging point sources according
to type and quantity of air contaminants emitted;
• Source Registration Monitoring - determining that all
sources covered by the ag'ency's regulations have been
duly registered with the agency;
247
-------
• Permit Compliance Investigations - checking to ensure that
permits have been granted for all applicable processes and
equipment and their modifications.
• Complaint Investigation - determining cause of complaint,
recording pertinent data, issuing violation notices if
appropriate, and ascertaining adequacy of plans for pre-
vention of future incidents;
• Episode Management - periodically reviewing emergency
procedure plans; checking that all shutdown procedures are
being implemented during periods of process curtailment;
coordinating with other agencies participating in pollution
reduction effort;
• Compliance Plan Status Inspection - checking to see that
engineering, procurement, installation, and testing of
equipment is proceeding according to the approved plan.
• Source Compliance Monitoring - determining that all
sources are in compliance with applicable emission
standards (particularly important where the agency does
not have a permit system).
2. Availability of Records to the Inspector
To conserve time and effort, data files should be kept in a
specified location preferably where the enforcement officer meets the
individual designated to accompany him on inspections. Refinery man-
agement personnel must also be aware of the location and content of
248
-------
these files so that in the event the primary contact is not available, an
alternate can be assigned. Files kept at the home office, if not on the
refinery premises, are of little immediate value. The unavailability of
data may result in time delays that reduce the field enforcement offi-
cer's effectiveness and waste the time of plant personnel.
3. How Records will be Used
Data contained in these records will have four primary uses.
a. Permit and Source Registration Identification - Flow
charts, engineering drawings, and equipment description will be
used to designate the exact location, capacity and configuration
of a system or permit unit. This descriptive information is
necessary to preclude modifications without agency concurrence,
to estimate increases of emissions due to process or feed
changes, and to determine the location of monitoring systems
(see Chapter IV).
b. Emissions Inventories - Data usable for estimating
emissions are contained in special reports such as sxilfur
balances in fuel gases (Figure 77) and estimates of losses from
pumps, valves, emergency flaring, tanks, and transfer pro-
cesses. Additional direct data can be found in logged values
from source monitors, air monitoring devices, plant operations
reports, analyzers, sulfur and hydrocarbon balance forms, and
odor surveys.
249
-------
REFINERY DAILY FUEL, USE REPORT
Company Name
Add res a
Zip
Tel. No.
Day
of
Month
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Ga^
Cur-
tailed
Total
Fuel Use Data
Fuel
Oil
Barrels
Weight
%
Sulfur
Gravity
API
Natural
Gaa
MCF
Refine ry
G*.
MCF
•
•
SOj
Refinery
Gas, Tons
Remarks
*
Signature of Company Representative_
Title
Figure 77. Fuel use/sulfur balance report
Z50
-------
c. Emergency Action - Plans and procedures for emer-
gency action are among the documents which will be duplicated
in the enforcement agency's files. These records will serve as
a checklist for the field enforcement officer during periods of
curtailment of operations and during dry runs to"simulate
emergency operations. These plans must be updated periodi-
cally to account for process changes, procedural modifications,
and errors detected during practice runs.
d. Legal Action - Incident and complaint investigations,
court actions, and variance board activity will require data
from all of the records described above. For example, the
point of emission of excessive odors may be traced from an
incident described in an operator's log or from an inplant odor
survey record (Figure 78).
Some of the information retained for agency use will be
sensitive if not classified in the military sense. The field en-
forcement officer must respect the confidentiality of any infor-
mation so designated by the refinery management. Any process,
control, or cost data that the company feels is proprietarymust
be treated as such by the air pollution control agency represen-
tative.
B. FORMAT OF RECORDS
Data management and records control procedures may be dif-
ferent at each refinery. It is necessary, therefore, to arrange with
251
-------
Date
12-8 8-4 4-12
Time
Wind
Direction
4
/
L
N
M St.
Remarks
12-8
8-4
4-12
Odor Complaints
F rom:
Time:
Investigation^
Figure 78 - Odor survey form
252
-------
refinery management for the maintenance of data in the format most
useful for air pollution control field personnel. In general, written
files should contain the following information:
1. Period of Time Covered
Continuous records, such as daily logs, are usually prepared by
shift, i. e. , 12 to 8, 8 to 4, and 4 to 12 for specified dates. Incident
records must accurately pinpoint the time of day, duration of theoccur-
rence, and date. Weekly, monthly, quarterly, and annual records or
data summaries must note the time span from start date to end date,
2. Person Responsible for Keeping the Records
Most refineries have designated a member of management to be
contacted by the field enforcement officer and to serve as the coordina-
tor for all environmental control functions. Generally, this individual
will also be responsible for keeping the records. Since there is the
potential for inspections at off-hours and over weekends, supernumer-
aries who are familiar with the environmental files should be assigned
to be contacted during these times.
3. Brief Description of Process or Equipment for Which the
Record is Maintained
There are two basic methods for describing the processes or
equipment for which data are recorded. The first is to use commonly
accepted nomenclature accompanied by a location indicator such as
"Fluid Catalytic Cracking Unit No. 1. " The second appropriate
approach to process identification is a numbering system. Some of the
253
-------
options available include air pollution control agency permit numbers,
source registration number or an "Emission Point Number". These
designators may be assigned to each point within a refinery capable of
emitting air contaminants. A numbering system starting with 001 and
continuing until all points are numbered is acceptable. When a num-
bering system is employed, the descriptive details of the equipment
and location must be kept in a master file.
4. Data
Recorded information will generally be presented as a narrative
or in tables. Typed data are always preferred but carefully written or
printed details are acceptable. Each chart or table should have an
appropriate heading to describe the data, i. e. , "Hydrocarbon Losses
from Tanks", "Semiannual Sulfur Emission Balance", or "S<52 Losses
from Emergency Flaring". Column headings for tabular information
should be precise, clearly defined units (such as Bbl/day, Tons/hr,
etc. ), allow space for remarks and identify totals and subtotals where
applicable.
A significant quantity of information is stored in computer data
banks. Printouts of these data may use abbreviations to conserve file
space. These abbreviations may not be easily understood by the field
enforcement officer so it is advisable that descriptions of the abbre-
viations and symbols be kept with the hard copies.
254
-------
C. TYPES OF RECORDS
1. Permit or License Files
The field enforcement officer will probably use the permit
records more than any other single file. During routine inspections,
he observes refinery operations for compliance with permit conditions
and looks for equipment or process modifications. Permit certificates
may only partially describe a given unit, but additional details are
necessary for an efficient inspection. "As built" drawings and flow
charts will provide the necessary details to show the field enforcement
officer if unregistered modifications have been performed. Each per-
mit unit description should provide operating data which includes
throughput, sulfur content of processed crude, fuel usage, or rated
capacity.
2. Maintenance Records
Maintenance procedures and schedules including preventive
maintenance are essential for a sincere air pollution control effort.
Maintenance is not only the repair or replacement of valves, flanges,
compressors, etc., after failure; it is a planned, co-ordinated effort
to prevent breakdowns which will affect pollutant emission rates. Since
maintenance in most refineries is a planning function based upon inputs
from the operating departments, replacement, repair and modification
schedules will be kept in the planning department files. The field
enforcement officer should determine where these records are kept,
and who is responsible for the data. Periodically, he should discuss
255
-------
the maintenance and replacement schedules with designated personnel.
Often the maintenance plans also include modifications or system
alterations for which new permits will be required. A thorough review
of all such proposed modifications may preclude legal action at some
future date for noncompliance.
Another group of maintenance records which is of concern to
field personnel is monitoring instrument repair and replacement.
Chapter IV contains a detailed discussion of in-stack and peripheral
monitoring systems. Maintenance responsibility for these instruments
may reside with the planning department of the refinery, a special
instrument repair shop, or the environmental control office. The field
enforcement officer must determine where the records are maintained
and institute a regular evaluation of the plans for repair, installation,
or modification of the instruments.
3 . Shutdown and Startup
Reports of equipment malfunctions, upsets, and overload con-
ditions reported by operating personnel should be compiled, recorded
and maintained in the environmental control office. As an example, if
sulfur recovery facilities are unable to receive the sulfur bearing gases,
the gases must be flared causing the release of significant quantities of
SO2. The record of this event will contain the quantity of sulfur burned
(usually in tons), the duration of the procedure, time of day and date.
Emergency plans and procedures also come under this category
of data. Shutdown procedures for refinery operations call for very
256
-------
stringent safety precautions for both personnel and equipment which
are the overriding consideration in the suspension of operations. This
coupled with the possibility of secondary effects of greatly increasing
emissions demands that curtailment plans be very specific regarding
operations and time. The plan is a narrative with accompanying sched-
ules, designated responsible personnel and special safety precautions.
The plan must be periodically reviewed by management and agency
field officers to assure the workability of the procedure.
4. Ground-Level Perimeter Monitoring and Continuous Source
Monitoring Records
The design and placement of ground-level perimeter and con-
tinuous monitoring systems are discussed in Chapter IV. Measure-
ments are recorded in analog form on strip or circular charts, in
digital form on punched or magnetic tape, or are hand logged from dials
and gages. Contaminants monitored are aerosols and gases that are
reported in the following manner.
a. Particulates - The ambient measurement reporting units
depend upon the sampling or monitoring methods. In the case of
a high volume sampler, the units are usually mass concentra-
tion units such as mg/m . In the case of sequential filter tape
samplers, the reporting parameters normally are in reflection
or transmittance units such as RUDS (Reflective Units of Dirt
Shade) or COH (Coefficient of Haze). These units are arbi-
trary and cannot be translated directly into mass units unless a
257
-------
consistent conversion factor has been determined. It is not
likely that LJDAR or nephelometers will be encountered, but
such instruments essentially look at numbers of particles, and
reporting units are usually in terms of visibility.
In the case of a source monitoring instrument, the re-
cording is actually a fraction or percentage of the full-scale
range of the instrument. Only if the instrument has been cali-
brated and a scale attached to the instrument can a meaningful
reporting parameter be observed. Further, most monitoring
devices measure concentration so that stack exhaust flow must
be known before mass per unit time such as Ibs/hr can be
derived. The most common particulate monitoring device that
will be encountered will be an opacity meter. These instru-
ments are usually calibrated to give a direct readout in terms
of opacity. In some cases, they may be calibrated in terms of
mass concentration units such as gr/ft , mg/m , etc. The new
beta gage-filter tape samplers are calibrated to read out
directly in mass concentration units.
b. Gases (SO,. H,S. NH^. NOX. Amines, RSH) -Monitoring
for gases almost uniformly results in the reporting units being
produced in volumetric concentration units. Depending upon
the concentration these units may be percent by volume, parts
per million (ppm), or even parts per billion (ppb). Most instru-
ments would be calibrated to permit the use of a direct-reading
258
-------
scale. In some cases where a multirange instrument is used,
the scale would be only in percent of full scale.
259
-------
REFERENCES
1. Administration of the Permit System. County of Los Angeles,
Air Pollution Control District, Los Angeles, Calif. Internal
Document. January 1968. pp. 1-12.
2. Enforcement Management System Users Guide. Environmental
Protection Agency, Research Triangle Park, N. C. Publication
Number APTD-1237. September 1972. pp. 3.4-3.5.
3. Stein, A. Guide to Engineering Permit Processing, Environ-
mental Protection Agency, Research Triangle Park, N. C.
Publication Number APTD-1164. July 1972. pp. 5.8 - 5.9.
260
-------
VI. ESTIMATING AND ASSESSING EMISSIONS
A. INSPECTION AND SURVEILLANCE PROCEDURES
An air pollution inspection consists of entering a refinery to
determine if the equipment or processes meet the standard and
comply with the rules and regulations of the air pollution control
agency.
Some inspections, especially initial ones, are comprehensive
and designed to gather information on all equipment and processes of the
refinery. Others are conducted for specific purposes, such asr
• obtaining information for source registration;
• gathering evidence relating to violations;
• checking permit or compliance plan status of equipment;
• investigating complaints;
• following up on a previous inspection;
• obtaining emissions information by source testing.
Field surveillance is a field operations activity that provides
for the systematic detection and observation of emission sources.
Observations are made of surrounding areas and the exteriors of
facilities for visible emissions, odors, contaminant damage, new
261
-------
construction or expansion, and other visual or sensory manifestations
of air pollution or changes in emission potentials.
Surveillance is conducted mostly by vehicle patrol. However,
aircraft, television, ambient air-sampling devices, and pollutant -
detection instruments can be used.
An initial refinery inspection lays the groundwork for evalu-
ating potential emissions of pollutants from a facility and for assessing
the relative magnitude of pollution control problems requiring correc-
tion, reinspection, or further attention.
The initial inspection has two phases: a refinery survey and
a physical inspection of the equipment and processes. After this
inspection is complete, routine surveillance continues. Periodic
reinspections are scheduled and occasional special-purpose inspec-
tions (unscheduled) may be required.
1. Initial Refinery Survey
The elements of a refinery survey include:
• Environmental Observations - the examination of possible
effects of emissions on property, persons, and vegetation
adjacent to the source; include the collection of samples
or specimens that exhibit possible pollution-related
damage.
• External Observations of Facility - observation of all
possible points of emission, all visible emissions, odors,
and pollution-related activities.
262
-------
• Management Interview - interview with refinery managers
and equipment operators establishes corporate identity,
ownership of the organization, operations, and air pollu-
tion control performance.
• Process and Equipment .Inventory - the inventory consists of
obtaining complete descriptions and records of all processes
equipment capable of emitting air contaminants. The inventory
inspection can also serve to initiate a source registration
system as a method for rigorously accounting for equipment
and processes capable of air pollution or confirm the infor-
mation provided in permit applications.
a. Environmental Observations - The surroundings of any
refinery or petroleum operation should be surveyed for odors
and for damage to vegetation and materials. Findings should be
confirmed by questioning residents in the neighboring communi-
ties. Soiling of surfaces of automobiles, residences, and other
structures should be noted and, where found to be severe, should
be investigated. Any pattern of increasing intensity of soiling or
staining of materials in the immediate vicinity may be a clue to
previously unrecognized emissions and should be studied.
Hydrocarbon vapors and gases are likely to be particu-
larly prevalent near petroleum production and refining facilities.
When the odors are intense, they indicate the occurrence of high
263
-------
emission le-vels since odor thresholds for these compounds are
relatively high and their quality not especially offensive.
Sulfur compounds associated with crude oil production
and with some refinery operations are readily detectable by
odor. Sulfur dioxide may be produced by combustion of fuels
or waste gases (for example in boilers, catalyst regenerators,
incinerators, and flares). This gas has an acrid, suffocating
odor with a threshold at one part per million (1 ppm). Reduced
sulfur compo-unds, including hydrogen sulfide, various mer-
captans, and other sulfur-bearing organics, have characteris-
tically very offensive odors and very low threshold levels,
which have been estimated at less than one part per billion
(1 ppb).2
In reporting on occurrences of odors allegedly causedby
petroleum facilities, the inspector should note wind conditions
(speed, as well as direction) at the time of his observations,
and he should develop systematic procedures for patrolling and
characterizing odors.
Odors may be described on an intensity scale in terms of
subjective evaluation, or a procedure using a scentometer or
other comparison devices may be adopted. A community odor
panel can help to establish the significance of day-to-day
variations in odor intensity and quality of the ambient atmo-
sphere, as well as to alert the field enforcement officer to the
264
-------
occurrence of unusual conditions.
Sulfur compounds, particularly hydrogen sulfide, can
discolor and damage lead-based paints often used in residential
areas. They also accelerate tarnishing of silver and copper
surfaces.
Sulfur dioxide and trioxide may be responsible for
causing damage to vegetation, such as evidenced by yellow to
brown blotchy spots on many varieties; sulfuric acid mist
formed from the trioxide occasionally causes a sort of pock-
marking injury to plants. Diagnosis of plant damage due to air
pollutants is difficult, however, and should be confirmed by
consultation with someone experienced in the field.
Data relevant to determining the speed and direction of
dispersion of contaminants in the atmosphere include tempera-
ture, humidity, wind speed, and wind direction measurements.
These data are continuously provided at weather observing
stations, but they may be required at other locations during
special investigations.
Equipment suitable for such purposes includes thermo-
meters, psychrometers, especially the "sling" variety and
anemometers or wind gages. Such equipment is discussed in
4
detail by Hewson. For measuring surface wind speeds, rota-
tion anemometers are most satisfactory; a particularly con-
venient type is the totalizing cup anemometer, which counts and
265
-------
records each tenth, sixtieth, or whole mile of wind past the
instrument.
b. External Observations of Facility - A fundamental in-
spection technique consists of identifying, describing, and
evaluating air pollution emissions and the factors contributing
to their formation.
The field enforcement officer must demonstrate that his
observations prove the presence of an air contaminant (that is,
an effluent that is not water vapor or a natural constituent of
the atmosphere) and that the emission violates a standard. He
must also determine what factors caused the emission to violate
the standard. He should be prepared to describe the events
occurring in each stage or element of the air pollution problem.
(1) Visible emissions (plumes) - The air contaminant of
primary interest is the plume. It is the "discharge" or "emis-
sion" regulated or prohibited in most statutes or rules.
Since all substances become liquid, solids, and gases at
certain temperatures, the plume may consist of a variety of
contaminants in various states of matter. Smoke, for instance,
contains visible aerosols - carbon particles and solid or liquid
particles of partially burned fuels - gases such as sulfur di-
oxide, oxides of nitrogen, and unburned vapors.
The identity ascribed to the plume is usually made in
terms of its outstanding visible characteristic. For example,
266
-------
even though sulfur dioxide may be the most significant of the
pollutants emitted from a given stack, the effluent in which it is
contained is frequently described as smoke due to the visible
soot, carbon particles, and fly ash contained in the plume.
The mere observation of a plume, however, does not
result in its conclusive identification. Knowledge of the specific
conditions which caused the contaminants is required. The dis-
tinction between smoke and fumes cannot be made unless the
processes by which they are generated are known. Effluent
plumes may be smoke, dusts, mists, gases or vapors.
• Smoke is the visible effluent resulting from in-
complete combustion. It consists mostly of soot, fly ash, and
other particles less than one micrometer in diameter. Depend-
ing upon the composition of the fuel or materials being burned
and the efficiency of combustion, various volatilized gases and
organics such as aldehydes, various acids, sulfur oxides, nitro-
gen oxides and ammonia may also be emitted. Because of the
low vapor pressures and slow settling properties of the parti-
cles, the smoke may be carried considerable distances from
the source and many submicron particles will be permanently
dispersed in the atmosphere.
Smoke will vary in color but will be generally
observed as grey, blue, black, brown, white, or yellow,
depending upon types of fuel or material and the conditions under
267
-------
which they are burned. ' The color of smoke is generally a good
indication of the type of combustion problem encountered.
Grey or black smoke may indicate that material is
being burned with insufficient air or inadequate mixing of fuel
and air. This, for example, will occur during the flaring of
waste gases when steam injection is not functioning properly.
• Dxtsts are minute solid particles released in the
air by natural forces or by mechanical processes such as crush-
ing, grinding, melting, drilling, demolishing, shoveling,
sweeping, and sanding. Dust particles are larger and less con-
centrated than those in colloidal systems, such as smoke and
fumes, and will settle fairly quickly on surfaces. A dust
effluent may also contain many submicroscopic particles.
Catalyst regenerators are a major source of dust
emissions in refineries. High efficiency collection systems
such as electrostatic precipitators in series with cyclones or
multiple cyclones in series are generally used to control this
pollutant. Bag filters are not generally used in petroleum re-
fineries though baghouses may be required to meet stringent
air pollution control regulations.
• Mists consist of liquid particulates or droplets
smaller than raindrops, such as fog, and are formed by con-
densation of a vapor or atomization of a liquid by mechanical
spraying. Mist droplets may contain contaminant material in
solution or suspension.
268
-------
In large oil-burning installations, sulfur trioxide
is formed as a gas and, after contact with sufficient moisture in
the air, forms as a white-to-blue plume several feet above the
stack (detached plume). After further contact with moistxire in
the air, the sulfur trioxide is transformed to a sulfuric acid
mist.
• Gases A gas is a fluid of freely moving mole-
cules tending to expand infinitely and to diffuse and mix readily
with other gases. Gas pollutants include a large variety of in-
organic and organic compounds which may have noxious, mal-
odorous, toxic or corrosive effects or may have an effective
smog-producing potential. These include carbon monoxide (CO),
ozone (O3), oxides of nitrogen (NO, NO2), sulfur dioxide (SO2),
hydrogen sulfide (H2S), hydrocarbons and their oxidation products,
halogens (chlorine, bromine, fluorine, iodine) and their deriva-
tives such as hydrogen fluoride (HF), and vario\is chlorinated
solvents such as those used in industrial degreasing and dry
cleaning. Other important toxicants include ammonia (XH-J,
arsine (AsH3), fluorine (F,)( hydrogen chloride (HC1), phosgene
COC12) and hydrogen cyanide (HCN).
Gases that commonly occur in refinery air pollu-
tion problems as a. result of direct emission are sulfur dioxide,
hydrogen sulfide, and mercaptans.
• Sulfur dioxide is a common stack gas produced
269
-------
from the combustion of sulfur-containing fuels such as coal and
fuel oil, the burning off of residue on catalyst in oil refining
operations, the burning of tail gases from the recovery of sulfur
from refinery waste gases, and various other chemical and
metallurgical processes. A major source of sulfur dioxide is
the burning of fuel oil by refineries. Fuel oil may have a sulfur
content from less than 1 percent sulfur to 5 percent in some of
the heavier fuels. SO2 has a noticeable odor at comparatively
low concentration and will damage certain species of vegetation
at 0. 25 ppm. Sulfur dioxide gases and sulfuric acid mists can
hasten the corrosion of wires, metals, and other materials.
Cracking processes in oil refining operations
convert the sulfur contained in the crude oil into hydrogen sulfide
and mercaptans. When hydrogen sulfide is released to the atmo-
sphere as a gas, it manifests a characteristic rotten egg odor.
From relatively small gas concentrations, mercaptans also
exhibit'strong unpleasant odors such as garlic, decayed garbage,
skunk, or onions. Hydrogen sulfide is detectable at 0. 12 ppm
and mercaptans at 0. 001 to 0, 041 ppm. Under humid conditions,
H2S will also discolor some surfaces painted with lead pigments.
• Vapors - A vapor is the gaseous phase of a sub-
stance which at normal temperature and pressure is a liquid or
solid.
The most important vapor problem results from
270
-------
the evaporation of petroleum products, such as the unburned
gasoline vapors in automobile exhaust. Gasoline vapors also
originate from processes in which volatile products are main-
tained in storage tanks and from the operation of pumps, com-
pressors, and blowers required for moving liquid and gas
streams.
(2) Evaluation of visible emissions Once a plume is identi-
fied as an air contaminant, it must be measured against some
standard to determine whether a violation of the law has
occurred or it must be evaluated to determine the size or
severity of the pollution.
Compliance with applicable emission standards is
determined by visual evaluation of visible emissions and by
source testing of emissions which are invisible or near the
threshold of vision.
Visual observation of plumes by field personnel can be
an effective and economical method of determining compliance
with air pollution regulations, provided the regulations are
based on the visible aspect of plumes or on other properties
that can be shown to be directly related to the visible aspect.
The benefits of basing smoke statutes on opacity or
density are quite evident, even though equipment and fuel regu-
lations have increasingly assumed precedence in control legis-
lation. When the visual standard is specific with reference to a
271
-------
cut-off point and time interval, it is simply and directly en-
forced. To cite a violator for excessive smoke, enforcement
officers need only observe an emission of an opacity or density
beyond that allowed for a specific period of time. Although the
visual standard is limited to estimations of particulate pollution
which obscures vision, its application simultaneously tends to
reduce grain loading and gaseous contaminants. (As the grain
loading in the plume increases, the light transmission decreases
exponentially. ) To comply with the opacity standard, more
efficient combustion or equipment operation is necessary. The
Ringelmarm standard can be applied not only to smoke, but to
fumes, dusts, and mists arising from a variety of problems and,
therefore, is most versatile in identifying and controlling atmo-
spheric pollutant emissions in a community.
However, while large reductions can be anticipated, they
cannot always be precisely predicted or evaluated. Determina-
tion of opacity and shade of any emission alone gives no specific
measurement of the qxiantities of contaminants being emitted.
A complete description of the theory and use of the
Ringelmann chart is discussed in a Bureau of Mines Information
Circular and in EPA Field Operations Manual .
(3) Investigation of odor potentials of emission sources
• Plant Inspection - On suspicion of odor nuisance
emissions, an inspection may be undertaken supplemented by
272
-------
source testing for evaluation of the odor potential. As soon as
practical after identifying the suspected source of odor emis-
sions, the field officer should enter the refinery: (1) to gather
the evidence needed to prove that the violation has occurred, for
example, that someone discharged into the atmosphere, a con-
taminant in greater amount or density than allowed and for mo re
than the specified time; (2) to determine the cause: and (3) to
ascertain the necessary corrective measures.
By proper interrogation, the field officer should
establish the circumstances leading to the emission violation.
He should be alert for observations he can make to verify the
accuracy of the statements made to him.
Next, if not already a part of the plant record,
the equipment data are obtained. These should include the make,
type, size, and capacity of all equipmenfor processes involved.
General conditions bearing on the air pollution potential of the
equipment should be noted. Observations should be made of
gapes and monitoring instruments, particularly temperature
charts on incinerators, and load charts on boiler instrument
panels. Information on operating failures which lead to exces-
sive emissions is being published in the technical literature.
Much can be learned from process studies which identify
operating conditions that cause high pollution discharge.
273
-------
• Evaluating Odor Concentrations - If the odor
being investigated has been identified as caused by a known
odorant, its concentration should be measured by chemical or
physical means in the laboratory. This is especially true when
the known odorant also has toxic or irritant potential, as in the
case of hydrogen sulfide, sulfur dioxide, ammonia, chlorine,
and various aldehydes. In many such cases, the criteria for
acceptable concentrations in ambient air are already estaUished
in terms of mass concentrations which are lower than odor
thresholds, so that evaluation in terms of odor units is super-
fluous .
However, when the odor nuisance is the only
suspected effect, or the identity of the odorant is in doubt, or
more specific methods of measurement are unavailable, the
samples collected at the source should be evaluated by an odor
8
panel using dilution techniques.
(4) Relating source strength to control requirements - The
contaminants responsible for an odor should be controlled so that
threshold levels are never reached in the outdoors. Some indus-
tries assume that they have no odor problems, because they con-
sider discharges from their processes to be unobjectionable or
even pleasant. However, the presence of any odor which per-
sists and is not normally associated with the daily routine of
living will be a source of annoyance to the neighborhood.
274
-------
The odor evaluations of source samples provide esti-
mates of odor concentrations in terms of odor units per unit
volume. These estimates can serve as guidelines in the devel-
ment of control methods. Thus, if a stack effluent is normally
diluted by a factor of 1, 000 before it arrives at a breathing level
in the surrounding neighborhood, an odor concentration of 1, 000
odor units per standard cubic foot could be considered to be on
the verge of acceptability, while an odor concentration of 10, 000
would require at least 90% control.
This sort of guideline can be refined by calculating an
odor emission rate in odor units per minute. This is equal to
the product of the odor concentration and the volume rate of the
stack exhaust, in standard cubic feet per minute.
Dilution factors required for positive control can be
estimated either by surveying ambient air in the vicinity to de-
termine the maximum odor concentrations observable or by-
standard engineering design procedures based on plume dilution
equations or community experience. It should be remembered
that dilution of odorous gas to the median odor threshold level
can be expected to render it undetectable by only about half of
the people in the community; therefore the use of an additional
safety factor in design for positive control is advisable. Also,
dilution factors work better near the source and tend to break
down with distance.
275
-------
An application of odor measurement in improving neigh-
borhood odors would be to survey all the operations in a plant
and determine the odor emission rate from each. Listing such
emissions together with estimates of costs for control can help
management pick out the largest odor sources (rather than the
largest stacks or largest volume discharges) and concentrate
effort initially on those which are likely to provide the greatest
Q
improvement per dollar of expenditure.
c. Management Interview - The purpose of the management
interview is to obtain from management and supervisory per-
sonnel the identification of processes and equipment as possible
sources of pollutant emissions, preparatory to making the de-
tailed equipment inventory.
In particular, the field enforcement officer should obtain:
• business and ownership data, including former
owners, .and responsible management personnel
• plot plans showing disposition of all major units of
the facility
• flow charts for each major processing unit, indicating
purpose, operating conditions, and normal processing
capacity.
The FEO should determine what procedures are employed to
control or eliminate the discharge to atmosphere of noxious or
malodorous emissions through the purging or depressurizing of
276
-------
tanks or vessels. These procedures may include the instal-
lation of special instrumentation, high-level or high-pressure
alarms, liquid knock-out drums on fuel gas systems, pressure
relief or manually controlled discharge from process equipment
to blowdown vessels of either variable or fixed capacity which
are served by vapor recovery compressors, flare systems, or
both, and fixed roof tankage tie-in to properly sized vapor
recovery or fume disposal systems.
d. Process and Equipment Inventory - The initial
process and equipment inventory may be developed from
information acquired through administration of a permit
system or a source registration system or from direct inter-
views and conferences with management personnel. It consists
of complete records of all equipment and processes capable
of emitting air contaminants that are located at all facilities
within the jur i sdictional area.
Inventory of equipment is called an equipment list.
This list enumerates al] items capable of emitting air
contaminants that are located at a given refinery and thf
status of that equipment with respect to compliance with the
permit system and the rules and regulations.
In preparing equipment lists, the agency should
consider using the format of procedures devised by the
277
-------
Environmental Protection Agency for the National Emissions
Data System (NEDS). Equipment and process classifications
readily usable for an automated data processing system have
been enumerated which may also serve as an aid in preparing
equipment lists. Complete details of this system are con-
tained in the EPA Guide for Compiling a Comprehensive
Emission Inventory.
2. Physical Inspection and On-Site Testing
a. Preparing for the Plant Visit - The objectives of
the field enforcement officer are not only to determine which
elements of the operation are governed by rules and regu-
lations, but to determine as well the degree of compliance to
them. His inspection procedures are adapted to the specific
air pollution regulations that apply to the type of unit being
inspected. All refineries have standard safety procedures for
employees and visitors. These procedures also concern the
• "%,
field enforcement officers.
(1) Review of records and regulations - Prior to the
physical inspection, the field enforcement officer should review
and organize the records contained in the field file. He should
determine what specific rules and regulations apply to the
operations of the facility to be inspected and he should be
278
-------
prepared to furnish appropriate copies or extracts of these
regulations to the management of the facility.
On the occasion of the initial physical inspection, the
field file for the facility should contain the equipment list, plot
plans, flow charts and auxiliary information furnished by
management in the interview phase of the inspection. At this
stage, the file will probably be incomplete, and a major objec-
tive of the physical inspection is to fill in the gaps and pin-point
problem areas. The visit will also be the subject of an Activity
Status Report on findings and recommendations.
Because of the operational complexity of petroleum re-
fineries, and allied activities, a degree of specialization, and
training and experience is required to make an effective FEO.
It is necessary to write technical reports of processes and to
prepare graphic presentations to describe the air pollution po-
tentials of the process units being inspected.
Separate inventory forms for each type of equipment,
operation, process, and plant are helpful at inspections to
ensure coverage of specific points. Coverage is complicated
becaus e:
• Similar process vessels arc used in various source
activities and are grouped interdependently. Attempts
to itemize individual pieces of equipment often lead to
confusion and disorientation.
279
-------
• Air pollution potential can be better determined from
an inventory of functions of process vessels than
from itemization of equipment units. Process inven-
tories may also require field surveys of product
flows, throughput capacities, and emission factors.
• Refinery and chemical plant inventories thus cate-
gorize, itemize, and present such data as will
directly determine compliance not only with permit
regulations but also with equipment regulations.
To make these data readily available, a special inven-
tory system should be adapted for each refinery. To cover each
of the multiple operations of a refinery adequately, the plant
area is subdivided into process units-. (Units with the gfeat'S-st
air pollution potential are subsequently assigned more frequent
inspections. )
Plant ownership data is recorded separately on a Plant
Card. It is most important to know who the responsible officials
are and how they can be quickly contacted. Where accurate
field data exist in the inventory files, it is possible to make a
preliminary investigation of refinery problems by telephone.
The inventory records for each refinery consist of a
group of file folders, each folder dealing with a single process
unit. The folders are numbered and filed sequentially. As one
or more process units may constitute a source activity, an index
280
-------
of source activities cross referenced by folder number is main-
tained at the head of the file.
A source activity folder contains:
• A general description of the process and an
analysis of its purpose and function in the processing
sequence. Generally, the analysis traces the flo\v of
materials from introduction through various sidestrearns
to final effluents.
• A list of pieces of equipment contained in the
process unit and their function.
• A discussion of the air pollution potential of
the process or equipment including an analysis of any
important problems and, if possible, estimates of the
contaminants emitted and their chemical designations,
odor quality and intensity, opacities, or physiological
effects, as well as the potential hazards of the stocks
or products released should equipment failure occur.
• Estimates of throughput of volatile materials
and of emissions from known sources. This may be
determined from results of "material balances", e.g.,
estimates of sulfur derivatives lost as calculated from
the differences between input and final output.
• Results of any tests or analyses of effluents,
fuels or other materials.
281
-------
• A process flow chart and plot plan, which can be
used for reference and verification on follow-up inspec-
tions. These give the flow rates, pressures, tempera-
tures, etc., in process vessels and lines, where neces-
sary to estimate air pollution potential and to locate
points of emission.
Flow diagrams and plot plans are of particular impor-
tance in accounting for all equipment in a production sequence
which otherwise might be overlooked. They are of value in
showing the potential of an existing production system for
growth or change. They also show the capacity for such systems
to accomodate increased production. Comparison of the flow
chart and plot plans with conditions existing at a later inventory
reinspection will show exactly how the process may have
changed.
Flow charts and plot plans are drawn according to con-
ventional engineering rules. Pertinent liquid or gas feed lines
are shown and gas or liquid effluents indicated. Overhead dis-
charge and drainage from columns or vessels are generally
shown by vectors indicating method of disposal. Features not
essential to the understanding of the air pollution problem are
omitted.
The flow lines are clearly labeled and vectored as to
direction and content, for example: "Refinery Gas In", "To
282
-------
Fuel Gas System", and "To Oil-Water Separator". The process
lines indicate whether the flow originated in, or entered at the
top, side, or bottom of columns or equipment. Functions of
equipment and columns should be clearly labeled, unless they
can be depicted by symbols (such as heat exchangers , and con-
densers, and by plant number). It is of utmost importance in a
flow diagram to illustrate clearly all sources of air pollution,
including stacks, flares, and pressure relief valves, and to
identify the problem areas and the contaminants which may be
emitted. Process vessel and line operating conditions recorded
on pressure and temperature gages, manometers, continuous
recorders, and relief valve pressure settings should be
indicated wherever pertinent.
A sample of an Activity Status Report covering a sour
water treatment and disposal plant and accompanying flowcharts
are shown in Figures 79 and 80. The symbols used in chart
preparation are shown in Figure 81.
Some sources are so routine that standard inventory
forms are used to .report them. These in chide bulk plants,
truck loading facilities, oil-effluent water separators, tanks,
and natural gasoline plants. Examples of these forms and the
Activity Status Report arc those used by the Los Angeles County Air
Pollution Control District.
283
-------
ACTIVITY STATUS REPORT M.R.No.
Firm Name: Sunrise Oil Company, Inc. , Unit II Sector: 1 3
Address of Premises: 1325 Court Street City:
Responsible Person Contacted: J. R. Hicks Title: Plant Engineer
Nature of Business: Petroleum refining
Assigned Inspection New Activity Change of Status
Description General Usage Name of Equipment - System or Process
Inspected: Sour water oxidizing unit - Unit II
Field Enforcement Officer; J. R. Hardy Date: 10-15-59
FEO's Conclusions and Recommendations: The odors detected at this
time were not great enough to result in a public nuisance. This unit
remains, however, one of the greatest potential sources of odor prob-
lems in this refinery since it comprises the processing area for sour
waste water containing malodorous components formed during the
cracking operation.
Modifications made for compliance with Rule 62 have reduced the
possibility of excessive SO2 emissions from the vacuum heater. Since
the materials processed are both highly malodorous and corrosive,
the present inspection frequency of three times per year should be
continued to insure adequate maintenance.
Figure 79 . Activity Status Report from an Inspection Made of a Sour
Water Oxidizing Unit at an Oil Refinery (from Los Angeles
County Air Pollution Control District).
284
-------
FEO's Findings: The purpose of this unit is to deodorize the sour
water pumped from the accumulators at the crude, thermal, and
catalytic cracking units. This consists of the following equipment:
(1) a 10, 000 barrel cone-roof tank, (2) a neutralizing column
tower, (3) a waste-water stripper, (4) an aeration column, (5) a waste-
water cooler, (6) a sour water degasifier drum, and (7) necessary
pumps, piping, and instrumentation. These are shown in the attached
flow diagram.
The sour water is pumped from the accumulators to the degasi-
fier drum. The gas removed from this drum flows through a back
pressure regulator valve to a low pressure H2S removal plant.
The sour water and waste caustic are collected in the 10, 000 bbl.
capacity tank venting to the vapor recovery system.
The sour water is pumped from the tank to the neutralizing
column where it contacts 98% sulfuric acid. The mercaptans released
from the water by the sulfuric acid, along with other waste gases in
the overhead line from the caustic regeneration unit, are condensed
and fed into the cracking unit for conversion to H2S and recovery.
This accounts for the disposal of most of the mercaptans in the system.
The neutralized water is then pumped to the waste-water
stripper, and live steam is introduced in the column to strip out H2S
and mercaptans. Sweet gas with 7 or less grains of H2S per 100 CF
Figure 79 (continued) Page 2 of 4.
285
-------
from the secondary scrubber at the H2S removal plant is introduced
into the bottom of the stripper at the rate of 350, 000 to 500, 000 CF/D
to sweep the released gases from the water. This sour gas from the*
stripper goes to the H2S absorption plant. A pressure relief valve on
the stripper vents to the flare.
The stripped water then flows to an aeration column where it is
contacted counter-currently with an air stream and caustic and is
oxidized to non-odorous thiosulfate. The resulting foul air from this
vessel is sent to the firebox of the vacuum unit heater for deodorizing.
The water flows from the bottom of the serator through a cooler to
the covered waste-water separator.
Prior to the introduction of the air pollution control program,
the principle sources of air pollution at this unit resulted from the
introduction of (1) mercaptans from the neutralizing tank to the burners
of the vacuum unit heater, and (2) the H2S from the waste-water
stripper column to the refinery fuel gas system. The APCD test
team determined that previously 0. 5 ton/day of H2S was contributed by
this unit to the refinery fuel gas system. This is equivalent to a loss
of 2, 000 Ibs/day of SO2 to the atmosphere. After studying data
disclosed by extensive physical inspection, testing, and industrial
cooperation on waste gas streams throughout the refinery, Rule 62 was
introduced to control those waste gas streams to be incinerated and
Figure 79 (continued) Page 3 of 4.
286
-------
containing significant quantities of sulfur derivatives. To comply
with this rule, this refinery adopted the following solutions to meet
the problems resulting from its particular operating methods:
a. Enlarged its H2S absorption facility.
b. Made provision for introduction of condensable mercaptans
into the cracking plant for conversion of H2S and its
eventual recovery.
The waste gas stream now burned in the vacuum unit heater was
found to comply with Rule 62 during a test conducted by the APCD test
team on 10-13-59.
Negligible mercaptan odors were noted in the vicinity of equip-
ment at this time. Equipment was in good condition and operating
under permit conditions and requirements. No visible emissions
were observed at this time from the vacuum heater. A sample of
treated water taken from the cooler (after oxidation) was free of
noxious odors.
Figure 79 (continued) Page 4 of 4.
287
-------
PMBUM
VACl'W VAlVf
, VAfO« UlCOVtin
lUTt
H.O 1MK
ROM
ew. i
,»
1
1
10UH NjO
IIUI OF KOI
xmoultccuu
TO
lit "CAM All
oisrosAi
UNIT.
:• PVVPH.
iTf
MTJOI ---J
r
yy
«v
OiTOVr«>M
II
1
ir nines
TO* Ait
J_
«
T>
H..UA.
>luin
C-IV
f(«»f
Lf Hj!
TDCRACXMO
» WIT
» ru«
» i.r.n^
VAC HIATM
TOW WO
Wt
U . rUkHTAJH
TOCOVCMD
• Hit MICH
HPMATM
tun IATK
SIHIPPH
r.nx
COOK*
roiuwc-ut
*-* "S
FLOW DIAGRAM
..urnrtrri-nt Sunrl" ou ^3^^, Inc. tm,in IV5 C«,rt .t«.l
r.i, ,«pirnn 10, l>/\9 KEO J-*- "'"'J'
-
REVISIONS
VJA0 ojd?7
Figure 80. Process flow diagram of a sour water oxidizing unit (page 4 of the
Activity Status Report).
288
-------
SYMBOLS USED IN PETROLEUM FLOW DIAGRAM*
XJ
I (S»
©
vV
©
©
TYPES OF FLOATING ROOFS
Chart Number Two
10
1L
ftr
ft>
cjl
DOVBK 0(CK
POMOV. 1ITHOU1
CfMHPCmOOH
i 0!C<
MOCi «iih
Mt< 4(lC«T
|l>'l .ili .UT
CI'.HH PU11COK
vf.iiutio
PO »
MllUttO
PI'. >MH
Figure 81. Symbols used in petroleum flow diagrams.
289
-------
• Bulk Plant Data This inventory form (Figure 82)
is used to record data obtained from inspection of loading racks,
storage tanks, pumps, vapor controls, and associated equipment
located at bulk plants. Bulk plants are used to store and distri-
bute various petroleum products and may be found at airport
facilities, distributing centers, marine terminals, etc. The
form is used to emphasize degree of compliance in equipment
and operation with the applicable rules. This Bulk Plant Data
Sheet presents a master inventory for each bulk plant.
• Truck Loading Inspection Data Sheet - An
inspection sheet (Figure 83) is made for each trxick loading
facility. It lists the number of racks and spouts, the permit
status of each rack, and throughputs of tank truck loading racks.
From this form, the total losses of hydrocarbons
are determined using the following emission factors:
» Emission of hydrocarbons in gallons from
uncontrolled equipment - Approximately 1/10
of 1 percent of the average gallon throughput
per day.
• Emission of hydrocarbons in gallons from
controlled equipment 8 percent of the
above.
• Oil-Effluent Water Separator Inspection The
example shown in Figure 84 is of a single refinery oil-effluent
water separator which derives its influent from treating vessels,
290
-------
BULK PLANT DATA
Name Sunrise Oil Company Date 12-1-56
Address 1325 Court Street Conversation with H. Smith
No. of Storage Tanks 6 No. Storing Gasoline 3
Reid V. P. of Gasoline 11 IDS. Av. Gal/Day 49, OOP
Max. Gal/Day 58, 500 Plant Throughput 980, OOP Gal/Mo.
Units Loaded/Mo. 150 No. of Loading Racks 2
Length of Racks 20 ft.
Loading Schedule 12 Hrs/Day 5 Days/Wk.
Peak Operation Hours: 6-11 A.M.
No. of Filler Spouts: 2
Method of Filling (% splash and/or bottom fill) Through vapor closures
Rate of Filling 250 Gal/Min.
Vapor Recovery on Loading: Yes (x) No( ): If Yes, Type Vapor
absorption system
Possible Sources of Vapor Loss; Filling and breathing of storage tanks
No. of Tanks Under Rule 56 3 No. of A. P. Controls Floating roof.
Fill out tank data summary forms on these tanks.
Remarks: Used for blending stocks.
No. of Sumps 1 No. of Slop Tanks 1
No. of Separators None No. Under Rule 59 None
Fill out Separator Inspection Forms on All of Above.
Remarks:
Actual Loading Rack Total Hydrocarbon Losses 20 Lbs/Day
FEO <2- A? /
J
Figure 82. Bulk plant data sheet.
291
-------
TRUCK LOADING INSPECTION DATA SHEET
CO* ANY
Sunrise Oi 1
Conp any
•'
'•
••
-^^V_
LOCATION
1325 Court St.
Onyx, Calif.
"
'•
••
^^J
RACK
NO. OR
NAME
1
2
1
2
--'
GAL/DAY GASOLINE
TO TRUCKS
AVERAGE
23,000
26.000
24,500
26.000
^V.
MAXIMjM
28 , SOO
30,000
47.000
30,000
—
SPOUTS
TOTAL
s
6
S
7
^•\
OVER
4«
RVP
2
2
2
2
J
UNDER
40
RVP
3
4
3
S
^- — »».
RULE
61
S
2
2
2
\
PERMIT
STATUS
A-7432
A- 473 3
A- 473 2
P.R.
^^
RECHECK
7-8-59
7-8-S9
^-
DATE
CONTROLS
FIRST
USED
12- 1- 56
t t
^^
INSPECTED
BY
J.R.H.
J.R.H.
J.R.H.
J.R.H.
_. ^
Figure 83. Truck loading inspection data sheet.
292
-------
OIL-EFFLUENT WATER SEPARATION INSPECTION
Company Sunrise Oil Company Date 12-1-59
Address I 325 Court St. f Onyxp Ca. Conversation with H. Smith
Designation of separator Oil-effluent water separator
Description A reinforced concrete basin consisting of one primary
and three subsequent compartments separated by wooden curtain walls
with primary compartment covered by floating roof.
Method of skimming 4 - 2" swing pipes manifolded to 4" suction line
Size of inlet & outlet 12" dia. inlet and outlet
Location of separator At #3 treater area in south area
Size: Length 47'11" Width 2Q'-10" Depth 10'
Description of A. P. controls Floating roof on primary compartments
Primary compartments, No. & Size 11' x 20' x 0'
Subsequent compartments. No. & Size Three 11' x 20' - 10" _
Oil on surface 3" _ Temperature 90" _ Odor Moderate
Quantity & disposition of oil 300 bbls/day to oil recovery facility
Disposition of water & estimate of quantity 15,000 bbls/day to _
_ drainage system _
Sources of oil & water From treating shells at FCCTCCU south
tank farm -t-2243 and alkylalion plants _
Quantity of oil & water from each source Major source: _
cracking area _
Oil sample
.E.O. JjL Vp.
Figure 84. Oil-water separator inspection sheet.
293
-------
south tank farm, thermal and catalytic cracking units, and
alkylation plants in the general cracking area of the refinery.
The influent wastewater is traced on the flow diagram by the
inspector. The dcsc ription of the controls, (i.e., "floating roof
on primary compartments") and other information indicates
whether the equipment is in compliance with pertinent rules.
• Tank Inspection Report - This inventory form
(Figure 85) records the type of tank, vapor control, function,
dimensions, products stores, Reid vapor pressure, storage
temperature, etc. , of each tank to determine compliance with
rules.
• Data Sheet for Natural Gasoline, Gas, and
Cycle Plants - This form (Figure 86) is used to
inventory plants located away from refinery facilities but near
crude oil production facilities. Information to be recorded in-
cludes the status of tanks and separators subject to emission
control regulations.
Another aid to the field enforcement officer is the infor-
mation incorporated in applications to operate the equipment.
The permit status of equipment should be routinely checked to
detect any changes in equipment or process that might invalidate
an existing permit or conflict with variance conditions.
The factors affecting the permit status are (1) change of
ownership, (2) change of address, (3) new construction of
294
-------
NO
AREA GRID NO. 8 TRJ
M.R.NO. _fii
u
FIRM NAME l/ii Aniylrs tYlrolfim, Gsrpany (
AonRFSs OF PRFMISFS 12W, s. Oil iioj.i
NATIJRF OF flllSINF•
30'
12'
30'
sr
IS' 1
Si'
Dia.
30'
30'
CO1
40'
HO1
11'
401
Type
c
c
c
c
F
11. P
F
Gen.
Cond.
G
G
G
G
G
G
G
Product
Stored
G-isoline
Sieve Oil
..
\vialion Gnso.
(in^ol inc
n-'litane
Dl.icl
RVP
(lb.)
9
t^ftt.
6
9
50
nrrl.
Prod.
Stg.
Temp.
Mbient
1 1
airhirnt
..
..
•'
Service
*tor*R«
i«
..
%i nrm*?
..
..
'•
Vapor
Control
vn
N
N
V»
F (SS)
PVV-W
F (SS)
Permit
Status
IVIr 13
U.Ic ll-K-5
n.,ic )3
n-,u 13
«10002
• 104AJ
Uil- ll-h-3
/tl^l-K-S
Rules
Affected
RuU 10. S6
None
Non-
nil <• 10 V.
It.l* 10. V,
None
IV fe 10
Remarks
No. off
Vnoor llrrnvrrv
Tc«^wr»rily
nil nf vr
-------
NATURAL GASOLINE, GAS, AND CYCLE PLANT
SURVEY SUMMARY
Company Sunrise Oil Co. of Calif.. Inc. + 3 Date 3r28-56
Address 1400 Bliss St. City Onyx, Calif.
Phone RA 6-3251
Information by L. M. Black Title Production Engineer
Type of Plant: natural gasoline (x): Absorption ( ), Booster ( ),
Other ( ) Specify: Gasoline plant
Source of material process Wet gas from oil wells
Total No. of wells
Throughput: Wet gas 20 mm SCFD: Dry 2 mm SCFD
Natural gasoline (bbl)475/day Propane (bbl) 0
N-butane(bbl) 0 Butane-propane mixture (bbl) 0
Isobutane(bbl)
Boilers: Number 5 , Type hrt Heaters: Number 2 Type steam
Type fuel Plant gas Source Plant residue
Quantity of fuel used 400 MCF/day
Is flow diagram available^ Can it be obtained?
Storage & handling:
Tanks Oil-Effluent Water Separators
No. of Tanks 8 No. of separators 1
Vapor recovery 3 No. questionable None
No. under rule 56 None Type of control None
Pressure • 4 Floating roof None
Other (specify) 1 cone No. under rule 59 None
No. controlled 1 No. controlled under rule 59 None
Totally enclosed None
Vapor recovery system None
Other (specify) None
FEO's remarks re equipment & plant conditions: No leakages.
_ No losses noted.
FEO
>Li K-
u
Figure 86. Natural gasoline, gas, and cycle plant survey summary.
296
-------
equipment, and (4) alteration of equipment. The equipment list
is constructed to provide reference data which will enable en-
forcement officers on inventory reinspections to determine
whether the permit status has changed. Any untested equipment
found in the plant, capable of air pollution will require a permit.
Similarly, alteration of equipment is frequently detected
by discrepancies in the equipment description or on the flow-
chart or by changes noted on engineering applications in the
permit file.
(2) Review of safety precautions and procedures - The FEO
is accompanied to the unit or units to be inspected by the air
pollution representative within the plant or by such other in-
formed refinery personnel as he might indicate.
Personal protection is necessary in many of the indus-
trial locations that an enforcement officer may be required to
visit. Safety equipment such as hat, goggles, steel-toed shoes,
ear plugs, heavy gloves, gas masks or respirators, and safety
flashlights should be available. The field enforcement officer
must never enter a plant without the proper safety equipment.
The following is a list of do's and don'ts for field enforce-
ment officers to adhere to during refinery inspections.
1. When entering a plant to make an investigation, do
ENTER and leave by the MAIN GATE or GUARD POINT set up
by company regulations. Comply with company SECURIT Y RULES
297
-------
by registering and wearing a BADGE OF IDENTIFICATION if
requested to do so.
2. Do wear a head covering while in a plant, preferably
a HARD SAFETY HAT.
3. Do wear RUBBER GLOVES, when necessary, such
as when sampling strong acids and alkalies, a RESPIRATOR in
an atmosphere of heavy dust or gases, GOGGLES to protect the
eyes from caustic solutions, flying particles of steel, sand, and
hot oil.
4. In areas containing noxious gases, such as hydrogen
sulfide or hydrogen fluoride, a respirator should be worn unless
the gas previously has been tested by an MSA DETECTOR and
has been found to be below the toxic limit.
5. In the event of FIRE in the area of your inspection,
do immediately LEAVE, and remain outside the area until the
"ALL OUT" signal is sounded.
6. When inspecting an area where inflammable liquids
are being processed, do make certain that there is an avenue of
escape before starting the assignment. Look around. If the
area appears to be unsafe, leave it until you are assured of its
safety by someone in control of the area.
7. When working in an area where strong acid or caustic
solutions are handled, do note the location of the nearest water
shower for quickly washing off a person accidentally sprayed
298
-------
with these chemicals.
8. If your clothing becomes sprayed with light oil such
as kerosene or gasoline, do change clothing as soon as possible
to prevent damage to skin from contact with the oil and to elimi-
nate the hazard of the clothing catching on fire.
9. Do use the buddy system when taking a sample or
gauging a tank of volatile or gaseous hydrocarbons. The Field
Enforcement Officer should be accompanied by another person
and the two persons should remain together until the job is
completed.
10. If about to sample a tank, do make certain that all
equipment is in workable order before leaving the ground. The
uncertain perch atop of a high tank is no place to untangle a line
or to force a stubborn bottle in or out of a sampler.
11. Do not smoke or carry "strike anywhere matches "or
cigarette lighters which ignite when dropped within an oil re-
finery. Most plants allow smoking in the main office and certain
approved areas with permission of management.
12. Do not open or close VALVES, or start or tamper
with any equipment.
13. Do not enter a tank or other confined area where
gases may be present unless equipped with a GAS MASK and an
ASSISTANT is in attendance outside. Make use of the EXPLOSI-
METER to determine the quantity of hydrocarbons present.
299
-------
14. Do not ascend to the ROOF OF A TANK or other
large vessel in the course of an inspection unless accompanied
by a company representative, except when you have his permis-
sion to proceed alone. Do not step or walk on the roof of a tank
unless planks have been laid to distribute the weight of the body,
except advised otherwise by the company engineer.
15. Do not poke a flashlight into an open tank hatch or
confined area and snap it on. The light may be defective and an
explosion might result. It is better to turn the light on away
from the hatch and bring it no closer than necessary. Only
approved safety flashlights will be used in oil refineries.
16. Do not place the face close to a tank hatch and peer
inside to get a better view unless someone is close by to offer
assistance in case the gas happens to be toxic.
17. Do not watch a welding torch in action. Failure to
observe this precaution could result in painful eye burns.
.18. Do not use petroleum distillate or benzol to remove
grease from any part of the body unless the action is followed
by a thorough rinse with plenty of soap and water. This pre-
caution may prevent a serious case of dermatitis.
19. Do not enter into the immediate area of a flare which
might go off at any moment and scorch everything within reach.
b. In-Plant Inspection and Testing - After the preparatory
review and paper work have been completed, the FEO undertakes
the m-plant inspections.
300
-------
For gathering evidence in the field, it is useful to include
in an inspection kit: binoculars, camera, stop watch, flashlight,
maps, compass, smoke tube, and the required forms and regu-
lations.
Cameras are used primarily to photograph excessive
emissions from stacks and vehicles and to photograph equipment
and operating personnel for identification purposes. Cameras
which permit rapid development of photographs at the site of an
investigation are especially useful. Moving picture cameras
may be desirable for special investigations.
Stopwatches should be of the accumulative type for use in
recording total time of excessive emissions within a given period
of observations.
The actual techniques which may be used in any given
inspection depend xipon the specific information to be gathered.
For a complex facility such as a refinery, completing the initial
inspection may require a number of visits. To optimize the
efficiency of the total inspection, individual trip objectives
should be planned in advance. Among the inspection techniques
which may be useful are:
(1) Sensory observations - Sensory evidence, such as visible
discharge and odors disclosed by thorough physical inspection,
may be sufficient to determine noncompliance or equipment
breakdown in the cases of Ringelmann number readings or excess
301
-------
hydrocarbon vapor discharge from vapor controlled tankage,
for example. However, in other situations, sensory evidence
may only be a preliminary or corroborating step to a broader
investigation, either because the regulations affected call for
evidence not obtainable in this manner (e. g. , percentage sulfur
in the fuel oil, t^ S grain loading in gaseous products burned
and its Btxi value, weight of particulate discharge, concentration
of SO2 in discharge of flue gas) or because the control equipment
does not discharge a waste effluent directly to the atmosphere.
In such cases, the inspector must either rely on data indicating
temperature, density, pressure, vacuum or throughput recorded
on gages, continuous recorders, high level or density alarms,
voltmeters and ammeters, or he must provide for special
testing.
Emissions caused by leakage are primarily recognized
by sensory evidence. Their detection depends on the ability of
the field enforcement officer to recognize odors and trace them
to their points of origin. Probable points of origin within refin-
eries include pres sure-relief valves, storage vessels, bulk-
loading facilities, pump glands, pipeline valves, flanges and
blinds, and cooling towers.
The t'EO must also note any visible air turbulence
caused by light hydrocarbon leakage, frosting of valves or
pump glands caused by light hydrocarbon evaporation, liquid
30Z
-------
leakage, local area discolorations caused by vapor condensate,
visible emissions or changes in flow (surging) of an emission,
extinguished flare pilot lights or detection of audible gas leaks.
These may disclose a violation or a potentially critical situation
that should be corrected quickly.
Visible plumes must be noted and their apparent causes
recorded for further investigation. Plumes may be caused by
incomplete combustion of waste gases in flares, by dust from
catalyst regeneration operations, or by breakdowns in various
process unit operations.
The field enforcement officer inspecting refineries should
thoroughly survey the vapor recovery systems and the facilities
for gathering and processing (sour water, wastewater, sour gas,
spent caustic, and acid sludge). Even though in a modern
refinery most of these streams are treated, their extremely
noxious and malodorous characteristics make even the most
isolated uncontrolled stream a potential air pollution problem.
Wherever possible, the inspector should point out condi-
tions having a high pollution potential so that the refinery's tech-
nical staff may have the opportunity to assess the problem and
solve it. More effective use of existing control equipment may
be achieved by extending its service to as many uncontrolled
sources as is possible without overloading its capacity.
303
-------
The inspector may find
• isolated streams of sour gas fuel untreated for H2S
removal and recovery
• sour water discharged to open drains with live steam
which has not been first deodorised by processing in
sour water oxidation or H2S stripping facilities
• odors and hydrocarbons emitted to the atmosphere
from oil-water separators
• malodorous or noxious acid slxidge stored in
uncontrolled tanks, which release fumes and odors
due to breathing and filling losses, or loaded into
tank cars and trxicks with similar results.
(2) Observing process instrumentation The process flow
charts obtained from plant management in the preparatory phase
of the survey should be annotated with operational information
relevant to emission evaluation. Operating conditions may be
indicated on pressure and temperature gages, manometers,
continuous recorders, and other devices. Recording the pres-
sure settings for relief valves can be useful.
(>) On-site testing At times, the enforcement officer will
be called upon to make quick and sometimes crucial, estimates
of air pollution problems in any environment. While he cannot
make accurate determinations of concentration on the basis of
sense perceptions only, he may be able to identify pollutants,
304
-------
allow for hazardous concentrations, and trace them to a logical
source. To eliminate guesswork and to establish identity and
concentration within a reasonable degree of accuracy, some
field sampling equipment is required. Such equipment, to be
of use in enforcement, must be portable, require a minimum
amount of bench and field preparation, be of a direct-reading
type, yet be sxibstantially accurate. Table 11 lists a variety of
contaminants that can be detected or measured by means of
simple portable equipment.
Noxious gases, odors, vapors and phenomena for which
tests can be made in the field and which require no collection of
samples for laboratory analysis are aldehydes, ammonia, aro-
matic hydrocarbons (benzene, toluene, slyrene, xylene) atomic
radiation, carbon dioxide, carbon monoxide, chlorine, com-
bustible gases and vapors, organic halides, humidity, hydrogen
cyanide, hydrogen sulfide, mercaptans, oxygen (deficiency), and
sulfxir dioxide.
Certain types of sensitized papers will change color in
the presence of physiologically significant concentrations of
certain noxious gases, fumes, and dusts. These can be vised to
test for or to verify the existence of certain suspected contami-
nants such as ammonia, hydrogen sulfide and phosgene. For
example, ammonia reacts with litmus to produce a red to blue-
color change. Concentrations of ammonia from 0 to 1,000 ppn>
305
-------
T.ble II. CONTAMINANTS THAT CAN BE TESTED IN THE FIELD WITH PORTABLE DEVICES
Contaminant
Aldehyde*
Ammonia
A romalic
Hydrocarbons
1. Benzene
2. Toluene
S. Xvlene
4. St\rene
A r sine
Ca rbon
Dioxide
Carbon
Monox' -if
Chlorine
Combustible
Gates
Hydrocyanic
Acid Gas
H> drogen
Fluoride
HvJropen
Sull.de
Dioxide
0:one
O r ! ; c i r n C V
rh^s^ne
Phosphmr
Sulfur
Dioxide
Reason for
Source Te»t
Eye
Irritation
Complaints
Odor com-
plaint!
Plating
Operations
E xhaust
Compla mt s
E xhaust
Compla mts
Cylinder
Loading fa
Bleach Mfg.
Venting
Storage
Tanks Odor
Complaints
Plating
Phosphate
Rock
Odor Com-
plaints
R efmene s
^. Chemical
r rocesaes
; r o m Air
Room
T herR-.al D.--
t. ompof ition
.-f Organic
MfS. of
Aretylene
Equipment Used
Absorption in
Sodium Bi*ulfite
M.S. A. Midget
Impinger
Red Litmus L Stop
Watch
M. S. A. Aromatic
Hydrocarbon
Detector
M. S. A. Arsine
Detector
Fynle CO.
Analyter
M.S. A. CO
Detector
O. Tolidine in the
M.S. A. Midget
Impinger
M.S. A. Model 40
Indicator
M.S. A. Hydro-
Detector
M.S. A. Hydro-
cyanic Fluoride-in-
air Detector
M.S. A. H.S.
Detector
F\ r . :e Ox v &;<• n
I rented Filter
Pipen
Treated Granules
M.S. \ SO detector
Treated Granules
Reic h' • Ir *t
T 'J r*' e : I f T
». Prrparation in Minutes
Time
Treatment or Required
Jodometnc Titration
10 5 10
Color Change to Blue 1 1 0
Colors. Treated
Granule*. Stain Length
Measured 10 2 0
10 2 0
10 2 0
10 2 0
Change Color 10 2 1
Absorption in Caustic
b Measure Vol. 1 3 0
Change
Colors. Treated
Granules • Color 122
Change
Color Intensity Com-
pared to Standard* 10 S 2
Direct Reading
Treated Granule*
Change Color 10 2 0
Treated Filter Papers
Change Color 550
Treated Granule*
Change Color 1 2 0
Color Change
Time Interval of
Cracking Measured ' 10 20 2
Absorption Measure
Volun.e Change 550
Color Change Com-
pared to Standards 5 S )
Color Stain Length 552
Measured
Length of Bleaching
Action Measured 1020
Gas Titration 10 10 5
Gas Titration 10 10 S
o. Test or Sampling
Concentration
0-1000 ppm
10- 100 ppm
0- 100 ppm
0- 400 ppm
0- 400 ppm
Qualitative
0- 100 ppm
0-20%
0-1000 ppm
0- 70 ppm
0-20 x LEL
0- SO ppm
.5- 5 ppm
0- 50 ppm
0- 10 ppm
u- 100 ppm
„.„«
1 - 100 ppm
1- 10 ppm
1- ISO ppm
0- 1000 ppm
50 Grams/ft3
c. Calculation
Eight-Hour
Threshold
. 5 to 5. 0 ppm
100 ppm
25 ppm
200 ppm
200 ppm
100 ppm
0. OS ppm
5000 ppm
100 ppm
1 ppm
...
10 ppm
3 ppm
20 ppm
?• ppm
10 pphm
18-21'.
1 ppm
5 pphm
5 ppm
and Interpret*!
Sufficient
Warning Without
Te.ling
Ye*. Eye
Irritation
Ye*. Odor
Ye*. Odor
Yes, Odor
Yes, Odor
Ye*. Odor
No
No
No
Ye*. Odor for
Immediately
Dangerous level*.
No. for Low Cone.
Some Yes, Odor
Ye*. Odor.
by Trained
Personnel
Ye*. Odor for
Immediately
No. for Low Cone.
No, Odor is
not reliable
Not reliable
No
No
No
No
Yes. Odor
ion
306
-------
can be detected by this method. Similarly, hydrogen sulfide
may be detected with lead acetate, phosgene with diphenylamine,
etc.
Another detection system, useful for various contami-
nants including aromatic hydrocarbons, carbon monoxide,
hydrogen cyanide, hydrogen sulfide, and sxilfur dioxide uses
glass tube ampules containing treated granules which change
color when exposed to contaminated air. Air is aspirated
through the tubes by rubber squeeze-bulbs, and the length of
the discoloration produced is proportional to the concentration
of the contaminant.
Combustible gas meters, also called explosimeters,
may be used in testing for high concentrations of hydrocarbon
gases, vapors, and other combustible gases including carbon
monoxide.
More accurate measurements may be made, usually less
conveniently, with portable kits for specific chemical tests.
Among these are gas-liquid reaction systems such as Tutweilers
apparatus (for H,S, SO2, NH3, and CO2), Reich's test ;for SO,
by reaction with iodine), and Fyrite analyzers (for CO, and O,).
Midget impingers incorporate a hand-cranked pump to
bubble air through impingcr tubes for collection of particulate
matter or for absorption of soluble gases. For some contami-
nants, they offer a convenient means of gas-liquid titration, but
307
-------
for other contaminants it is necessary to return the samples to
a laboratory for analysis. Fritted glass bubblers may be used
in a similar manner, except that insoluble particulates are not
conveniently manageable with these devices.
(4) Grab sampling Where on-site testing is inconvenient or
the necessary apparatus for filtering and absorption is unavail-
able, a simple expedient is to obtain samples for later analysis
in the laboratory. "Grab samples" are often obtained by filling
gas sampling tubes or inert plastic bags with air by means of
motor or hand-powered pumps. The collected air sample is
analyzed in the laboratory. For sampling of liquid fuels, as
may be required to enforce fuel composition regulations,
special containers for flammable liquids are required. Fuel
oils of low volatility can be collected in quart or half-gallon
tins.
Systems used for the collection of particulate contami-
nants include sedimentation and settling devices, such as fall-
out jars and gummed paper stands, miniature cyclone collectors,
blower and filter systems, impingers and impactors, electro-
static samplers and thermal precipitalo rs . It is becoming
more important to obtain size -discriminating samplers in place
of total mass samplers.
The sampling of gaseous contaminants involves sepa-
rating them from the air in which they are entrained. Such
308
-------
techniques are adapted either to the sampling of specific Caseous
compounds or to the determination of gross total concentrations
of gaseous contaminants. Specific methods are available- for
sampling many inorganic gases and some reactive organic com-
pounds. Inmost cases, these methods involve absorption by
bubbling the air through a reactive liquid agent. Mixed gases
are usually trapped by adsorption or freeze-out techniques. In
any event, an appropriate sampling train must be devised,
usually augmented by a suction pump and a wet or dry gas meter.
(5) Source testing Source tests may be re-quired for a
variety of reasons. Among the most critical is the establish-
ment of compliance or noncompliance of the equipment in ques-
tion with emission standards. Where qualitative estimates are
nonconclusive, the field enforcement officer will be the initiator
of the test request. The efficiency of the testing operation may
depend largely upon the field enforcement officer's adcptness at
preparing the test request and his presence during the test to
observe the operation of the system under scrutiny. This is
especially important in regard to operating parameters that
must be maintained as specified during the test.
The request for analysis must state in detail the- operat-
ing condition of the suspect equipment during the source test.
This is necessary so that the appropriate test procedures may
be prepared. Basic instructions for the test should include
309
-------
the points to be tested, anticipated contaminants for which the
test is run, accessibility of test points (scaffolding), and avail-
ability of electric power near the test point. Provisions should
be made for portable hoods, or other specialty items where the
equipment to be tested does not have an exhaust system. In
summary, sufficient data should be made available to the team
to avoid surprises during the test. Operating conditions must
be defined for the basic system and for the air pollution control
system.
• Basic equipment
Description of the type, quantity, and rate of
material to be processed by the equipment
during the test
Type, quantity, and rate of usage of fuel
Phase of operation during which the source
test is to be conducted if the process is not
continuous
• Air pollution control system
Pressure drop across the control device
For scrubbers water rate
For electrostatic precipitators current and
voltage reading, rapping frequency, operating
temperature, gas velocity
For baghouses shaking frequency
310
-------
Duration and frequency of control device
down-time, if any, during the test
The air pollution control system must be in operation
during the test. If it is desirable, samples may be taken at the
inlet to the air pollution control device as well as the outlet to
confirm collection efficiencies.
3. Resurvcys and General Surveillance
Because of the complexity of the petroleum industry, unit pro-
cesses must be inspected systematically and regularly. The frequency
of reinspection of any process is based upon the findings during the
initial inspection and the recommendations of the FEO and his super-
visor. The schedules are printed monthly for each area or special
assignment and forwarded to the FEO and his supervisor. The reinspec^
tions are scheduled so that they can be completed within a month. The
number of reinspections assigned per area is based on the estimate
that all required inspections can be completed within one year.
The enforcement officer may have occasion to inspect plants
out of schedule because of complaints or violations. In these cases,
he does not make a formal inventory reinspection, but uses the copy of
the previous inventory record (equipment list), from his files as a
check on status of the permit, compliance, or other situation. When a
specific air pollution problem is involved, it is best to concentrate on
that problem rather than on the inventory of the entire plant. The
equipment list can thus be updated during unscheduled inspections.
31 1
-------
Vehicle patrol is the principal surveillance method. Field
enforcement officers drive their vehicles throughout a defined area -
such as a zone, sector, or district - and major traffic arteries to
observe evidence of emissions and to detect possible violations of the
rules and regulations. The patrol route is laid out to bring the greatest
area under view over the shortest distance.
As the enforcement officer becomes familiar with his area, he
concentrates on sources requiring the greatest attention and on areas
of high source density. He may employ a check list of facilities that
are currently involved in permit cases, hearing board actions,
recurrent violations, and complaints.
a. Updating the Process Inventory - The process inventory
can be used as a tool in gathering evidence. This is especially
true in public nuisance cases where it is desirable to eliminate
from suspicion all processes or equipment that do not contribute
to the nuisance. In such instances, the equipment list serves as
a check list.
On an assigned inspection, the enforcement officer must
check all equipment units in the plant against those on the equip-
ment list. He is careful to note not only that all equipment
listed is identical in important respects, but that they have not
been replaced, since a replacement can affect permit regulations.
This is usually determined by comparing manufacturer's
serial numbers. He also'checks for new equipment, alteration
312
-------
of equipment, posting of permits, operation contrary to condi-
tions of permit, etc. Any discrepancies noted are recorded in
detail on an inspection report.
Information from the process inventory is input to the
permit system data base whenever an enforcement officer pre-
pares or updates an equipment list. The data management
system then outputs data from the permit system to schedule
inventory inspections and, while doing so, prints out the exact
permit descriptions of the equipment.
b. Assessing the Quality of Maintenance - The effective
operation of various control systems in a refinery is of basic
importance to efficient air pollution control.
Reinspcctions and surveillance afford the opportunity to
observe changes in the condition and performance of control
equipment and to detect any development of, or increase in,
sources of leakage within a refinery.
Cyclones, electrostatic precipitators and vapor recovery-
plants are subject to corrosion and other destructive forces
which reduce efficiency. In addition, a change in process feed
or feed rates due to altered product requirements may also
result in overloading or otherwise upset operating conditions in
the air pollution control system.
In a vapor recovery system serving tankage, peak loads
develop in the morning hours when the heat of the sun produces
313
-------
maximum volumes of hydrocarbon vapors in the space above
the liquid. Uneven loading schedules at tank truck loading
facilities tend to create a similar situation.
Since a gradual reduction in control efficiency does not
always alter the effective operation of the process unit, it may
not be noted by operating personnel until a major breakdown
occurs accompanied by a serious air pollution situation consti-
tuting a public nuisance. It is therefore essential for adequate
air pollution control that such areas be inspected regularly
and frequently.
Proper design and maintenance of this equipment, to-
gether with adequate housekeeping, are required for the effi-
cient control of emissions from refineries and other petroleum
industry operations. It is field enforcement officers' responsi-
bility to observe and report, from the physical evidence, the
managements' degree of success or failure in this endeavor.
B. ESTIMATION OF LOSSES
Most air pollution control enforcement activity involves investi-
gating compliance with statutory regulations and abatement orders.
Regulations include emission prohibitions and performance standards.
These standards require that the losses or discharge to the atmosphere
of specific contaminants be determined. Inspection of equipment for
general emission inventory purposes or for support of permit system
314
-------
activities may also require that losses be estimated.
A variety of direct and indirect techniques, having different
levels of accuracy and precision, may be used to estimate losses in
petroleum refinery operations. Some direct techniques are source
testing and monitoring and physical observation. Indirect techniques
employ material balance, related process variables, perimeter moni-
toring, and emission factors.
1. Direct Estimation Techniques
a. Source Testing and Monitoring - When performance or
emission standards are given in allowable mass rates of dis-
charge or concentrations, source testing or monitoring is
normally used to achieve the accuracy required for enforcement.
This does not mean that a reasonable estimate of losses cannot
be made for a specific source when a previous correlation has
been developed between an observation and actual test data.
Source testing differs from source monitoring in that the
former is limited in scope to a specified amount of time while
the latter is continuous. Also, samples extracted for source
testing are examined or analyzed elsewhere while samples ex-
tracted for source monitoring are examined in situ. The field
enforcement officer in a large diversified control agency nor-
mally would not participate in source testing except as an ob-
server of the process or of visual emissions. In a smaller
organization, and depending upon his qualifications, the FEO
315
-------
may also participate in the actual test operations. In both
situations, the field enforcement officer will be expected to
initiate requests for source testing when he has reason to
believe that a violation of a regulation exists.
Readings from source monitoring instruments which the
field enforcement officer may take during on-site inspections
can be used to estimate losses although nomograms or further
calculations may be required to convert the instrument readings
to useful emission terms. Most of the source monitoring instru-
ments normally used in a refinery (see Chapter IV) measure
concentrations of contaminants. If performance standards are
given in terms of mass flow rates, the FEO must be able to
estimate the loss rate in those terms. Suppose, for example,
that the emission standard was given in terms of pounds of
particulate matter lost per pound of material processed, and
the analyzer was calibrated in terms of Ibs. /ft. (at stack condi-
tions), then
TT- ^^
= ^
where ^
_ . Ib. contaminant
E. = emissions,.
'Ib. process weight
Ib. contaminant*
C = concentration, —7^1 ;
ft. stack gas
Q = stack gas flow rate, —-
Ib.
Wp= process weight, ~~~
Both C and Q at the same stack conditions or corrected to standard
temperature and pressure.
316
-------
If an instrument does not measure in the units required, then a
calibration chart will be needed to convert to the proper units.
b. Direct Observation - The field enforcement officer uses
his powers of direct observation of an effluent stream to esti-
mate losses at many points. The most common example is in
determining the Ringelmann Number of black smoke or the
equivalent opacity of any visible plume. Also included would be
observation of liquid hydrocarbon leaks from pump seals,
flanges, and relief valves and from blind changing. It could
include estimation of losses from flare operation, from condi-
tions in oil-water separators, from odor intensity inside plant
boundaries, and from fugitive dust (that is, dust from open
storage of bulk materials, from uncovered materials-handling
equipment, and from unpaved roads and exposed soil).
Training is the key to achieving an acceptable level of
proficiency in using observational estimation techniques. This
requires the repetitive making of estimates which are verified
by an objective measurement procedure at the time of observa-
tion. Opacity observations, for example, are accepted in court.
When carried out by a trained observer under correct conditions,
procedures for making visual opacity determinations are to be
found in the Field Operations and Enforcement Manual for Air
14
Pollution Control and in the EPA Standards of Performance
o • o 15
for new Stationary Sources.
317
-------
2. Indirect Estimation Techniques
a. Data from Process Instruments - Process instruments
are used to secure efficient and safe operation of the various
process units in a refinery. Where the parameter being mea-
sured is closely related to the potential for effluent losses or to
the efficiency of control equipment, the process instrument is a
useful tool in the estimation of losses.
Most process instruments are used to measure flow,
pressure, and temperature. Normally, a measurement per se
may not permit the estimation of losses. However, any evi-
dence of unstable or off-limits operation may give an indication
that excessive demands are being placed upon control equipment
and that losses may increase as a result.
There are many situations in which pressure, flow, or
pressure/vacuum measurements may aid in the estimation of
losses or potential for losses. First, the temperature and flow
rate of gas entering a carbon monoxide afterburner (CO boiler)
serving a fluid catalytic cracker catalyst regenerator unit will
indicate whether there is sufficient residence time and a high
enough temperature for the complete conversion of carbon
monoxide to carbon dioxide to take place. Minimum tempera-
ture and residence times may be specified by permit conditions
and are provided for in the new source performance regulations
of the EPA for fluid catalytic cracking units in petroleum
refineries .
318
-------
Though not routine, flow rate monitoring for waste gas
or emergency flares will yield data that may be used to estimate
losses if the composition of flare gases is known.
Vacuum recording at vapor recovery systems can give
an indication of the adequacy of compressor and recovery sys-
tem capacity. Where insufficient capacity results in other types
of venting or flaring, estimates of the losses can be made.
Voltage measurements on electrical precipitators can be
used to estimate some of the performance characteristics of
such control equipment although specific knowledge of the design
specifications of the equipment would be necessary to interpret
this data.
Pressure drop measurements across cloth filter units
are valuable in estimating the degree of bleed-through, whether
capacity is being exceeded, whether the cleaning cycle is ade-
quate, and in some cases, whether bags are ruptured. Again,
specific knowledge of the design pressure drop is necessary to
interpret these findings.
Absorber solution circulation rates and indications of
strength such as pH are process measurements that may be use-
ful in estimating effluent losses in acid gas scrubbers or ab-
sorbers. Because of the high dependence of scrubber perfor-
mance on these parameters, very good estimates may be
possible.
319
-------
The collection efficiency of almost all air pollution
control equipment decreases with an increase in the effluent
gas flow rate. A possible exception is the centrifugal collector
class, but increased flow in these devices results in such high
pressure drops that capacity is somewhat self-limiting . There-
fore, as a general case, effluent gas flow rates in excess of
design capacity can be expected to increase losses both by
virtue of the increased flow and the loss in collection efficiency
as well.
A very special case of process monitoring involves the
analysis of fuel oil and fuel gases for sulfur content. This is
particularly applicable to the sulfur content of refinery "make
gases". These ar-e gases-that res-ult from refinery operations.
They have a reasonably high heating value and are used as sup-
plementary fuel for heaters rather than being further processed.
Sulfur content information so obtained, together with usage
rates, -can be used directly to calculate sulfur losses to atmo-
spheres.
b. Equipment Inspection and Operational Data Estimates
of contaminant losses may sometimes be made by observation of
operating techniques, by inspection of equipment condition, by
determining level and adequacy of maintenance, and by evaluation
of process data not determined analytically.
Hydrocarbon losses from storage tanks vary with color
320
-------
and condition of paint, with product throughput, with vapor pres-
sure of product and, in the case of floating roof tanks, with the
condition of seals and interior surface. Specific calculation
procedures for determining losses from storage tanks have been
o
reported elsewhere and are not normally applied in the field.
However, the field enforcement officer should be aware of the
influence of these factors so that estimates can be made of the
potential influence of changes in them. The conditions andproper
functioning of floating roof seals is, for example, very important
in maintaining losses at the expected levels.
The field enforcement officer must learn the kind and
frequency of maintenance necessary to keep control equipment
in effective operating condition. For example, fluid catalytic
cracker dust is sufficiently abrasive to wear through the 'high
efficiency cyclones generally used as collection equipment in
front of an electrostatic precipitator. If excessive wear results
in high dust loadings to the precipitator, losses to atmosphere
will increase.
Losses would also be expected to increase as a result of
inadequate maintenance of relief valves and pump seals although in
these situations visible emissions may also increase. Electro-
static precipitators, cloth filters, and mist eliminators also
require regular maintenance. The principal maintenance prob-
lems with electrostatic precipitators are related to the power
321
-------
supply, particularly voltage level and regulation. Broken
electrodes must be replaced, collection plates and tubes must
be kept in the proper condition, alignment must be maintained,
and rapping mechanisms must operate properly.
Fabric filters require regular inspection and replace-
ment of bags. Based upon the fabric used, the severity of
service, and the cleaning procedure, an estimation can be made
of expected bag life. Bags must, of course, be mounted prop-
erly to avoid overstressing the fabric and to preclude leaks at
points of fastening. The bag cleaning mechanism whether in-
volving shaking, rapping, or reverse flow must be maintained.
Timers, solenoid valves, and mechanisms, all require main-
tenance and are subject to failure.
Mist eliminators, such as those used on sulfuric acid
plants and on the discharge of some absorption columns, are
passive devices and somewhat less subject to failure than the
other equipment. In some cases, however, depending upon the
materials used and the severity of service, corrosion or clog-
ging may occur which will disrupt design flow patterns or reduce
collection efficiency. Sagging or compression of the eliminator
media (usually metal or glass fiber mesh) may also result in
loss of efficiency.
Operational data that does not appear on process instru-
ments may also give an indication of possible emission losses.
322
-------
Examples would include catalyst make-up rates, change in
nature of materials being processed, and operator comments in
shift logs.
c. Data from Air Sampling Equipment - Information avail-
able to the field enforcement officer from ambient sampling
which might be of use in assessing losses will for the most part
be obtained by the use of portable or hand-held sampling equip-
ment or by perimeter monitoring equipment. The portable or
hand-held devices are most likely to be useful inside or close to
the plant boundary, although under unusual conditions they may
be used to verify outside complaints. Sampling would be con-
ducted regularly at a series of points within and near the
refinery boundary. These points would be selected on the basis
of past experience, proximity to known sources, and subjective
criteria such as location where certain odors have been or are
detected.
Sampling devices or indicator tubes are available to mea-
sure hydrogen sulfide, sulfur dioxide, carbon monoxide, ammo-
nia, mercaptans, phenols, and hydrocarbons. These measure-
ments are useful mainly in a qualitative sense because of
limitations in accuracy, specificity, and knowledge of the amount
of dilution of the contaminants after discharge from the source.
That is to say they are most useful in identifying unusual loss
or emission conditions which may be related to leaks or
323
-------
operational malfunctions.
Perimeter monitoring as use,d here means formalized
monitoring at sites established outside the refinery boundaries.
Contaminants monitored at these sites would most likely be
those identified as being of major importance, either because of
actual or potential volume of release, toxicity, or impact on
meeting air quality-standards. Expected values for the contami-
nants monitored, based upon known release rates and meteoro-
logical conditions, through use of diffusion modeling should be
determined. Then a normal pattern should be established,
which the field enforcement officer should be able to use to
estimate changes in source strength through observation of the
recorded contaminant concentrations.
324
-------
REFERENCES
1. Weisburd, M. I. Air Pollution Control Field Operations Manual,
A Guide for Inspection and Enforcement. Department of Health,
Education and Welfare, Public Health Service, Division of Air
Pollution, Washington, D.C. Publication Number 93 7. 1962.
p. 285.
2. Douglass, I. B. The Chemistry of Pollutant Formation in Kraft
Pulping. Department of Health, Education and Welfare, Public
Health Service, National Council for Stream Improvement.
In: Proceedings of the International Conference on Atmospheric
Emissions from Sulfate Pulping, Hendrickson, E. R. (ed. ).
Gainesville, Fla. , April 28, 1966.
3. Brandt, C. S., and W. W. Heck, Effects of Air Pollutants on
Vegetation. In: Air Pollution, Vol. I, Stern, A. C. (ed. ).
New York, Academic Press, 1968. pp. 401-443.
4. Hewson, E. W. Meteorological Measurements. In: Air Pollution,
Vol. II, Stern A. C. (ed. ), New York, Academic Press, 1968.
pp. 329-391.
5. Kudlich, R. Ringelmann Smoke Chart. Department of the Interior,
Bureau of Mines, Washington, D. C. Information Circulars
Numbers 77 1 8 and 8333 (revised). May 1967.
6. Weisburd, op. cit. p. 65.
7. Gruber, C. W. Source Inspection, Registration and Approval.
In: Air Pollution, Vol. II, Stern, A. C. (ed. ). New York,
Academic Press, 1968. pp. 561-595.
8. Danielson, John A. Air Pollution Engineering Manual, Second
Edition, A. P. - 40.
325
-------
9. Byrd, J. F., and A. H. Phelps, Jr. Odor and Its Measurement.
In: Air Pollution, Vol. II, Stern, A. C. (ed. ). New York,
Academic Press, 1968. pp. 305-327.
10. Guide for Compiling a Comprehensive Emission Control Inventory
(revised). Environmental Protection Agency, Research Triangle
Park, N. C. Publication Number APTD-1135. March 1973.
pp. 1-1 - 1-2.
11. Weisburd, op. cit. pp. 260-265.
12. Weisburd, M. I., and P. Roberts. Inspector's Manual. County
of Los Angeles, Air Pollution Control District, Enforcement
Division, Los Angeles, Calif. Internal Document. May 1957.
pp. 95-97.
13. Stein, A. Guide to Engineering Permit Processing. Environ-
mental Protection Agency, Research Traingle Park, N. C.
Publication Number APTD-1164. July 1972. pp. 7.12-7.13.
14. Weisburd, M. I. Field Operations and Enforcement Manual for
Air Pollution Control, Vol. I. Environmental Protection Agency.
Research Triangle Park, N. C. Publication Number APTD-1100.
August 1972. pp. 4.16-4.36.
15. Standards of Performance for New Stationary Sources. Federal
Register. 36^24876-24895, December 23, 1971.
326
-------
VII. MAINTENANCE
A. DESCRIPTION OF REFINERY'MAINTENANCE OPERATIONS
Maintenance is a major activity at all refineries. A large
portion of the work force is assigned to maintenance operations.
These operations can be categorized as routine or emergency.
Emergency maintenance is required to clean and repair facilities
damaged by accident. Emergencies such as ruptured lines, fires,
and explosions must be considered from the safety standpoint first
and from the pollution standpoint second. Other emergencies, such
as a broken compressor shaft, can be handled with more attention to
pollution control.
Routine maintenance can be categorized as minor or major.
Minor maintenance includes the day-to-day upkeep of individual
equipment items. Major maintenance involves the shutdown and
repair of entire processing units or even the entire refinery. These
"major turnarounds" are performed on a regular basis, usually at
intervals of two or three years. Routine maintenance is carefully
scheduled and planned whether it is minor servicing of equipment or
a major refinery turnaround.
327
-------
Record keeping is an essential part of a refinery maintenance
program. Maintenance records are kept primarily to facilitate
planning and work schedules. In some large refineries, electronic
data processing equipment is used to improve the efficiency of main-
tenance operations. Maintenance records are useful to determine the
state of repair of individual pieces of equipment. These records can,
therefore, be used to review the maintenance history of equipment
items known to be pollution sources. Emission control equipment
with histories of frequent upsets and malfunctions can often be detec-
ted easily by reviewing the maintenance records before beginning the
field inspection.
B. GENERAL PLANT MAINTENANCE
Routine maintenance operations within the refinery may con-
tribute to air pollution. Major turnarounds are often handled by
contractors who specialize in refinery maintenance. The procedure
in a major shutdown will vary widely depending on the individual
refiner's practices. In most cases, all liquid material is pumped to
product and intermediate storage or slops tanks. As the plant is
depressurized, gases are delivered to the blowdown system.
If acid gases are present, these can be delivered to the fuel-
gas system or the acid-gas recovery unit. Tanks and vessels are
steamed out to remove toxic and flammable.gases. Generally, the
steam vapors are vented to the atmosphere and pollutants escape.
328
-------
The vapors can, however, be vented through a condenser to a flare
where such an arrangement is justified. The shutdown procedure
followed in each case will greatly affect the quantity of air pollutants
emitted. At present, complete recovery of air pollutants during
shutdown is not practiced. Ventilation of vessels and tanks is
required for safety reasons during repair. Noticeable odors are
generally unavoidable at these locations.
The repair and cleanup operations represent a source of air
pollution during both major turnarounds and day-to-day operations.
Tanks, vessels, and towers require sandblasting and painting. Sand-
blasting can be the source of particulate air pollution resulting in
property damage or health hazard. In some cases, protective
tarpaulins or wet blasting can be used to reduce dust emissions.
Smaller items can be removed to remote sandblast facilities
equipped with dust collectors. The use of steel shot or grit can also
be used in place of sand. Paint spraying results in particulate and
hydrocarbon emissions. Nonreactive paints and solvents should be
used. Outdoor paint spraying is illegal in some areas. Solvent
cleaning may also be a source of air pollution. Nonreactive solvents
should be used.
C. MAINTENANCE OF AIR POLLUTION CONTROL EQUIPMENT
1. Flare and Slowdown System
A blowdown system consists of relief valves, blowdown
329
-------
piping, liquid knock-out drum, and flare stack. The operation of a
blowdown system is discussed in Chapter I, Section U. Relief valves
are discussed in Chapter II, Section H. The blowdown system is
designed primarily as an emergency shutdown system. The proper
operation of the system, however, is essential to the reduction of air
pollution from relief valves and flare stacks.
Relief valves are normally removed from service and are
shop repaired on a regular basis. Relief valves should be tested for
correct operation at regular intervals. If this is not possible, spare
valves should be installed to permit removal and servicing. Service
records can be reviewed to determine the inspection frequency. It is
essential that the valves be clearly marked and replaced in the proper
location. Tank vents and flame arresters also require regular
inspection to assure proper operation.
The flare stack is designed to operate under a wide range of
conditions. Since materials of wide ranging properties are processed
by the blowdown system, plugging of lines is a possibility. Blowdown
lines and knock-out vessels should be regularly inspected to maintain
them free and clear. Control systems on the liquid knock-out drum
must operate properly to prevent flooding of the flare stack. The gas
pilot equipment and purge gas lines must be in good operating condi-
tion if the blowdown material is to be combusted completely.
2. Participate Matter Control Equipment
Particulate matter is controlled in refineries through the use
330
-------
of cyclone separators, electrostatic precipitators, and to a lesser
extent - baghouses. Cyclone separators can plug at the dust outlet
if viscous material is introduced. Pressure buildup should be moni-
tored to determine if the cyclone is operating as designed.
Electrostatic precipitators are used to recover catalyst fines
in some refineries. These units can malfunction if tars or other
materials foul the collection plates. Regular inspection and cleaning
is required. Ionizer wires will corrode and require replacement.
Electrical control systems require checking to assure the proper
voltage and rapping frequency and intensity. Operating conditions
may have to be adjusted if particulate properties vary with time.
Baghouse filters require regular maintenance to ensure
efficient operation. This includes regular removal of the collected
dust, inspection of the bags, and lubrication of the moving parts. If
required, the bags should be replaced. Precoating of bags after each
cleaning cycle with dust is recommended to improve collection
efficiency, but may not be required in all cases.
3. Sulfur Recovery Plants
Sulfur recovery plants are an important source of sulfur
dioxide pollution in refineries. An upset can result in sharply
increased emission rates. The operation of sulfur plants is
discussed in detail in Chapter I, Section P. A sulfur plant generally
requires little attention. Proper operation of the inlet mixture con-
331
-------
trol valves is essential to complete conversion of sulfur compounds
to elemental sulfur. Sudden surges in feed rate or composition can
also upset the performance of the plant. Frequent inspection is
required to determine catalyst activity. The tail-gas incinerator
must be inspected for corrosion. Air/fuel mixture control instru-
ments should be inspected. Since a sulfur recovery plant is pro-
cessing a poisonous gas, alarm systems are provided to announce
plant upsets.
D. MAINTENANCE OF MONITORING EQUIPMENT
A wide variety of monitoring equipment is available. Sam-
pling devices and analytical instruments are commercially available.
Manufacturers' instructions should be carefully followed for use and
maintenance of monitoring equipment. Portable instruments can be
easily damaged in transit. Instruments located outdoors can be
damaged by the elements. Attempts to use equipment under condi-
tions other than those intended by the manufacturer may also cause
damage. Titrating solutions should be sealed tightly when not in use.
Fresh solutions should be used and frequently checked for concen-
tration. Regular calibration and inspection of instruments is essen-
tial. A maintenance record should be maintained on each monitoring
device.
332
-------
VIII. PERSONNEL
A. MANPOWER REQUIREMENTS
Of the field and enforcement functions performed by air pollution
control agencies, those relating to petroleum refineries and to the
petrochemical and chemical industries are probably the most special-
ized. Since these industries are likely to constitute large point sources
of pollution, as well as the cause of public complaints, an organizational
component within the agency (or at the very least, special assignments)
should be dedicated to the surveillance of these industries. The man-
power and the degree of specialization required will depend on the
number and types of such installations located within the jurisdiction of
the air pollution control agency.
Few air quality control regions contain a sufficient number of
petroleum refineries to justify a special operating unit. In most cases,
petroleum refineries are grouped with other plants requiring chemical
engineering expertise and with activities related to the petroleum
economy of the area. ' These include:
Asphalt manufacturing plants
333
-------
Chemical plants, including manufacturers
of sulfuric acid, vinyl chloride, paint
and varnish, and fertilizer
Gasoline absorption plants
Natural gas processing plants
Oil reclaiming plants
Petrochemical plants
Petroleum marketing and consumption
Petroleum marketing stations, service
stations, bulk plants, marine terminals
Soap and detergent manufacturing plants
Sulfur recovery plants
Tank farms
With a few exceptions (such as power plants and ships), these
industries are organized from unit processes - those involving a
chemical change in one or more reactants (e.g., nitration, polymer-
ization, hydrogenation) and unit operations - those involving physical
changes only (e.g., distillation, absorption). Both unit processes
and operations may be defined in terms of the process unit, that is,
the equipment or process vessels interrelated by flow systems in
which materials are progressively transformed towards a desired end.
Surveillance work levels tend to be related to the number of
process units with significant air pollution potentials rather than to
334
-------
the number of industrial plant address-locations. On the average, a
process unit in this industry may require three hours of surveillance
and inspection time. This includes surveillance, equipment inspection,
plant personnel interviews, and report write-up, exclusive of detailed
engineering evaluation, permit processing, and coxirt and hearing
board activities. At an inspection frequency of four times per year,
1Z man-hoxirs of field time per year per process unit would be re-
quired. Thus about 160 process units is equivalent to one man-year
of surveillance for the industry. This would be equivalent to eight
petroleum refineries each averaging 20 process units that are signi-
ficant from an air pollution standpoint.
Of course, these averages will vary among agencies, depending
on the personnel available and delegation of responsibilities. Field
personnel who are also responsible for permit evaluation (as opposed
to a separate permit processing unit) or for source testing will require
significantly more man-hours per process unit. Ideally, for maxi
mum efficiency, permit and source testing operations should be sepa-
rated from surveillance operations. A grouping of installations in
terms of process/work units and in terms of the types of surveillance
and enforcement functions required will help to determine the
manpower, organizational structure, and support services that will
be needed.
335
-------
B. FEO FUNCTIONS
The functions of the field enforcement officer to be performed
in connection with petroleum refineries are discussed in Chapter V,
"Maintenance of Refinery Records" and Chapter VI, "Estimating and
Assessing Emissions". To summarize, these functions are:
1. Maintain surveillances of all process, equipment, and activi-
ties associated with assigned petroleum, petrochemical, or chemical
installations, by means of vehicle patrol or other exterior proce-
dures. The purpose of this surveillance is to detect visible emis-
sions, odors, new construction, or other obvious changes in plant
conditions which may affect emission rates or permit status.
2. Conduct surveys and inspections of all processes, equipment,
and activities that have air pollution potential to establish, at each
and every definable source of pollution, compliance or noncompliance
with all rules and regulations.
3. Investigate all citizen complaints made in connection with
these installations.
To perform these duties, certain capabilities are required
of the air pollution control agency, the enforcement operation, and
the individual officers.
The aii- pollution control agency and, desirably, the field
enforcement operation should have personnel with baccalaureate
degrees in chemical engineering and previous work experience in the
industrv. Even with such training and experience, personnel should
336
-------
receive as much as 100 hours of formal on-the-job training.
Training should include courses in petroleum technology, and air
pollution engineering and enforcement. Courses may be available
from a nearby university or from the EPA or they may be created
and implemented by the air pollution control agency itself.
The specialized personnel should be available to perform such
functions as permit evaluation, plan review, engineering super-
vision, supervision of specialized enforcement activities, technical
services, and general agency planning and evaluation. In any event,
the field enforcement operation should either be supervised by, or
have access to, these personnel. In larger agencies, particularly
agencies responsible for large petroleum, petrochemical, and
chemical plant complexes, highly trained and experienced personnel
should occupy senior field enforcement positions.
C. FEO PERSONNEL QUALIFICATIONS
If the expertise described above is directly available to the
enforcement operation, then it is not necessary for FEOs to possess
degrees and previous experience in chemical engineering. Using
such personnel in the field may be inefficient, since not all of the
engineering capabilities of such persons would be needed in the field
operations program.
The qualifications of field enforcement personnel should be
looked upon, primarily, in terms of aptitude, including ability to
337
-------
receive special-purpose training and, secondarily, in terms of
general education and previous work experience. With respect to
aptitudes, FEOs should be able to:
1. Assimilate and comprehend (1) engineering information
_\
related to chemical and physical processes and (2) legal and adminis-
trative information related to enforcement activities;
2. Recognize equipment, process configurations, and operating
conditions or parameters (particularly pressure and temperature)
that affect air pollution emissions;
3. Relate equipment, processes and conditions to specific pro-
cessing functions (unit processes or operations) and to the specific
substances (input and output) involved in such operations;
4. Identify abnormal operating conditions, e.g., process
upsets, overloads, breakdowns, equipment failures;
5. Become familiar with the chemicals and materials employed,
particularly hazardous substances, and understand the technology
and terminology of the industry;
6. Evaluate visible and nonvisible emissions, including odors,
and materials-damaging substances;
7. Prepare and interpret process flow data and make preliminary
estimates of material losses; and
8. Prepare concise, accurate, and complete reports which
effectively communicate the technical information necessary to
338
-------
establish compliance or noncompliance of specific sources with the
rules and regulations.
These personnel aptitudes or capabilities can be determined
from:
1. General intelligence tests. Candidates should have high
verbal and arithmetical scores, including reading comprehension and
vocabulary skills.
2. Scholastic records and courses pursued in high school, junior
college, college, and technical institutes with respect to exposure
and performance in technical and scientific courses.
3. Previous work and military experience, especially in tech-
nical areas and responsibility and level of contact with public
(e.g. engineering sales activities).
4. Special aptitude tests. These can be developed and used to
identify and select potential FEO candidates. Skills and aptitudes
that can be evaluated include interpretation of flow charts, compre-
hension of technical information with respect to chemistry, heat
transfer, principles of conservation of energy and mass and fluid
flow. Personality tests can be used to help assess judgment,
emotional stability, ethics, and responsibility.
5. The degree of motivation. Motivation is a very important
factor, since FEOs generally operate in the field under minimal
supervision and are responsible for planning and scheduling much of
their own work. In effect, they are responsible for the installations
339
-------
they are assigned to, and it is up to them to conduct all necessary
research and collect all information they may need to understand the
operations of these installations.
6. Approach and appearance. Neatness, maturity, objectivity,
and responsibility are essential qualities in relating to operating
personnel at all levels.
D. TRAINING
Personnel meeting these qualifications should receive at least
100 hours of classroom and on-the-job training. They also must work
vinder supervision of experienced field personnel prior to entering and
inspecting petroleum refineries on their own. Courses should include:
Air Pollution Control Technology
Care and Use of Inspection Equipment
Drivers Training
Field Orientation
Legal Authority
Monitoring Instrumentation
On-the-Spot Field Testing
Petroleum and Petrochemical Technology
Report and Notice Writing
Smoke School
Supervised Field Training
340
-------
REFERENCES
1. Weisburcl, M. I. Air Pollution Control Field Operations
Manual, A Guide for Inspection and Enforcement. Depart-
ment of Health, Education and Welfare, Public Health
Service. Division of Air Pollution, Washington, D. C.
Publication Number PUS 937. 1962. pp. 98-99.
2. Weisburd, M. I., and P. Roberts. Inspector's Manual.
County of .Los Angeles, Air Pollution Control District, Enforce
ment Division, Los Angeles, Calif. Internal Document.
May 1957. pp. 84-97.
3. Weisburd, M. I. Field Operations and Enforcement Manual
for Air Pollution Control, Vol. I. Environmental Protection
Agency, Office of Air Programs, Research Triangle Park,
N. C. Publication Number APTD-1100. August 1972.
pp. 1.49-1.52.
341
-------
GLOSSARY
ABSORPTION: A process whereby a liquid dissolves a gas, such as
oil which absorbs light hydrocarbons from natural gas.
ABSORPTION GASOLINE: Gasoline which is obtained from natural
gas by absorption in oil.
ACCUMULATOR: A vessel which serves as a surge tank and holds
intermediate product, usually overhead distillate.
ACID GAS: A gas consisting mostly of hydrogen sulfidc and carbon
dioxide.
ACID TREATING: A process in which petroleum products are
contacted with sulfuric acid.
ACTIVATED CARBON: A form of carbon or charcoal which has a
high adsorptive capacity for gases, vapors or solids.
ADDITIVES: Chemicals added to petroleum products to improve
performance or obtain needed product characteristics.
ADSORPTION: A process in which a gas or vapor physically adheres
to the surface of a solid such as activated carbon.
AEROSOL: A continuous dispersion of solids and liquids in a gas,
usually air, such as a haze or fog.
AIR BLOWING: A process in which hot asphalt is oxidized by passing
air through it.
ALIPHATIC HYDROCARBONS: Open chain hydrocarbons such as
paraffins and olefins.
343
-------
ALKYLATION: A process for combining isoparaffins and olefins such
as isobutane and butylene, to form alkylate, a gasoline
component.
AMINE UNIT: A process in which acid gases are removed from
hydrocarbon gases by absorption in amine solution.
ANILINE POINT: An index for measuring the solvent capacity or
aromatic content of hydrocarbons.
ANTI-KNOCK COMPOUNDS: An additive in gasoline, such as tetra-
ethyl lead, for improving combustion characteristics in internal
combustion engines.
API: American Petroleum Institute.
API GRAVITY: An index for measuring the density of petroleum.
API SEPARATOR: A device for separating oil from water by gravity.
AROMATIC HYDROCARBONS: Hydrocarbons with an unsaturated
closed ring structure, such as benzene, toluene and xylene.
ASH: A nonvolative, incombustible component of fuels which remains
after combustion.
ASPHALT: A high, boiling, semi-solid hydrocarbon refined from
crude oil. A component of asphaltic-base crude oils.
ASTM: American Society for Testing Materials.
AVIATION GASOLINE: A grade of gasoline for reciprocating aircraft
engines.
AZEOTROPIC DISTILLATION: A process for separating hydro-
carbons of the same boiling point.
B
BARREL: A volume unit used in the petroleum industry consisting of
42 U. S. standard gallons.
BENZENE: An aromatic hydrocarbon present in some crude oils.
BFW: Boiler feed water.
344
-------
BITUMEN: See Asphalt.
BLACK OIL: Petroleum containing residual crude oil.
SLOWDOWN: Material purged from the refining process during
startups, shutdowns, and pressure relieving. The blowdown
system collects and disposes of the waste material.
BS&cW: Bottom settlings and water material found in tank bottoms.
BTU: British thermal unit - used to define heating value of fuels.
The heat required to raise one pound of water one degree
Fahrenheit.
BUNKER FUEL OIL: A heavy residual fuel oil.
BURNING OIL: See Kerosene.
CALCINED COKE: See Coke.
CASINGHEAD GASOLINE: See Natural Gasoline.
CATALYST: A substance used to accelerate chemical reactions.
CAUSTIC: A term used for solutions of sodium hydroxide used in
treating processes.
CETANE: A paraffin used as a standard for diesel fuel quality.
CLAUS PROCESS: A process in which hydrogen sulfide is converted
to elemental sulfur.
•
COKE: Solid carbonaceous residue obtained from coking residual
crude oil. Calcined coke is coke that has been heat treated
to remove volatile materials.
COKING: A process in which crude oil is destructively distilled to
produce petroleum coke.
CONDENSATE: A liquified hydrocarbon gas such as obtained from
natural gas wells.
CONVERTER: See Shift Converter.
345
-------
CRACKED GASOLINE: Gasoline obtained by cracking heavier
petroleum fractions.
CRACKING: A process in which large hydrocarbon molecules are
divided into smaller molecules. Process may be catalytic
or thermal cracking.
CUTBACK: Petroleum fractions used to reduce viscosity of fuel oils
and asphalt.
CYCLE STOCK: Intermediate refinery product which is recycled to
other process units.
D
DEA: Diethanolamine, an absorbent used to remove acid gases for
sour gas streams.
DEAERATOR: A device used to remove dissolved oxygens from
boiler feed water.
DEASPHALTING: A process in which asphalt is removed from
reduced crude.
DEBUTANIZER: See De-ethanizer.
DE-ETHANIZER: A distillation column which removes ethane and
lighter hydrocarbons from propane and heavier hydrocarbons
The terms depropanizer and debutanizer are also used for
similar operations.
DEHYDRATING: A process in which water is removed from hydro-
carbon gases and liquids.
DEPROPANIZER: See De-ethanizer.
DESALTING: A process in which salts are removed from crude oil.
• DEWAXING: A process in which wax is removed from lubricating
oils.
DIESEL FUEL: A petroleum product used as fuel in diesel engines
consisting of gas oils.
DISTILLATE: The light material taken overhead in a distillation
column and condensed.
346
-------
DISTILLATION: A process in which a hydrocarbon feed is separated
in two or more components of different boiling points.
DOCTOR TEST: A test used to determine the mercaptan content or
odors of petroleum products. Odor-free products are termed
"Doctor Sweet".
DRUM: A container which holds 55 gallons. Also, a vessel used in
the refinery process for storage or separation.
DRY GAS: A hydrocarbon gas, usually natural gas, which does not
condense easily. Usually contains mostly light hydrocarbons,
such as methane and ethane.
ENTRAINMENT: Liquid droplets or mist contained in vapors
leaving a boiling liquid.
EXTRACTIVE DISTILLATION: A distillation process in which hydro-
carbons with similar boiling points are separated by selective
absorption in a solvent.
FLARE: A device used for burning waste gases. See Blowdown.
FLASH DRUM: A vessel used to separate vapors and liquids after a
pressure reduction.
FLASH POINT: The minimum temperature at which vapors above a
petroleum fraction or product will ignite in the presence of a
flame.
FLOATING ROOF: A roof which floats on surface of liquid in a
storage tank and reduces evaporation losses.
FRACTIONATOR: See Distillation.
FUEL GAS: Light hydrocarbon gases generated in the refinery
process used for firing process heaters and furnace.
FURFURAL: An organic compound used as a solvent in refining
lube oils.
347
-------
FURNACE OIL: Distillate fuel oils used for residential and
commercial heating.
GAS OIL: A fraction obtained in the distillation of petroleum
generally used in distillate fuel oil.
GASOLINE: Refined petroleum naphtha used in internal combustion
reciprocating engines.
GRAVITY: See API Gravity.
H
HEATING OILS: See Furnace Oils.
HOT CARBONATE: A potassium carbonate solution used to absorb
acid gases from light hydrocarbon streams.
HYDRODESULFURIZING: A process in which sulfur is removed
from petroleum in the presence of a catalyst by combining
the sulfur with hydrogen.
HYDROTREATING: A process in which petroleum is reacted with
hydrogen in the presence of a catalyst to remove sulfur or to
hydrogenate unsaturated compounds.
ILLUMINATING OIL: See Kerosene.
ISOMERIZATION: A process in which normal hydrocarbons are
converted to isomers by rearranging the molecular
structure. A typical isomerization is the conversion of
butane to isobutane.
JET FUEL: A kerosene based fuel for use in gas turbine powered
aircraft. JP-4 and JP-5 are common grades of jet fuel.
348
-------
K
KEROSENE: A petroleum distillate boiling between naphtha and gas
oil. Used in jet fuels and heating oils.
KNOCK: A property of gasoline related to octane rating and engine
knocking.
KNOCKOUT DRUM: A process vessel used to remove entrained
liquid from gases.
LPG: Liquified petroleum gas. A petroleum product containing
propane and butane.
LAMP OIL: See Kerosene.
LEAD: Refers normally to lead additives in gasoline, such as
tetraethyl lead.
LEAN OIL: An absorption oil which contains no dissolved light
hydrocarbons.
LIGHT: A relative term applied to petroleum fractions to denote the
lower boiling material, such as light naphtha, light gas oil.
LIGHT ENDS: Light liquid hydrocarbons, typically ethane, propane
and butane.
LOADING RACK: A structure used to load petroleum products into
tank trucks, rail tank cars or barges.
LONG RESID: Residual oil obtained from crude distillation
containing neutral oils.
LUBE OILS: A petroleum fraction, generally heavy gas oils, vacuum
gas oils and neutral oils used for lubricating purposes.
M
MEA: Monoethanolamine, an absorbent used to remove acid gases
from sour gas streams.
349
-------
MERCAPTAN: An organic compound present in "sour" crude oil.
Mercaptan compounds contain sulfur and have a strong odor.
METHANATION: A process in which carbon monoxide is converted
to methane by reaction with hydrogen.
MOTOR OILS: See Lube Oils.
MOTOR SPIRIT: See Gasoline.
N
NAPHTHA: A petroleum fraction boiling in the gasoline range.
NAPHTHENES: A group of hydrocarbons having a saturated ring •
structure such as cyclohexane found in naphthenic crude oils.
NAPHTHENIC ACID: A corrosive organic acid found in some
naphthenic crude oils.
NATURAL GAS: Light hydrocarbon gases naturally formed in the
earth. May also refer to the finished product or pipeline gas.
NATURAL GASOLINE: A mixture of light hydrocarbons boiling in
the gasoline range recovered from natural gas.
NEUTRAL OILS: Distillate petroleum fractions, generally heavy gas
oils, having specific viscosity properties and used in lube oils.
O
OCTANE NUMBER: An index used to measure the anti-knock
properties of gasoline. Research, Motor and Road Octane
Numbers are three different octane ratings.
ODORANT: A material added to fuel gas to impart a distinctive
odor and permit human detection.
OLEFINS: A class of paraffin hydrocarbons which are "unsaturated1
or deficient in hydrogen, such as ethylene, butylene.
ON STREAM: A term to denote that a refinery or process unit is in
normal operation.
350
-------
OVERHEAD: The vapors which are boiled off the top of a distillation
tower or the lightest product obtained in the distillation
*
process.
PALE OIL: A distillate lube oil, yellow in color.
PARAFFIN: A series of linear and branched hydrocarbons fully
saturated in hydrogen, such as methane, propane. Also
known as alkanes. High molecular weight paraffin in solid
form is known as paraffin wax.
PETROCHEMICAL: A chemical compound, intermediate or product
derived from natural gas or crude oil.
PETROLATUM: A semisolid product obtained by filtration containing
residual oils and wax.
PETROLEUM COKE: Coke derived from crude oil. See Coke.
PETROLEUM SPIRITS: A distillate product used in solvents,
varnishes and paint thinners.
PHENOL: An organic chemical used in solvent extraction
processes.
PHOTOCHEMICAL REACTION: The process of chemical change in
the presence of radiation, such as the reaction of hydro-
carbons in sunlight to form smog.
PIPELINE GAS: Refined natural gas sold to residential, commercial
and industrial customers.
PLUME: The path taken by visible discharges from a stack or
chimney.
POLYMER GASOLINE: A gasoline component obtained by combining
two olefins.
POLYMERIZATION: A process in which two or more molecules are
combined. Typically, refers to the combination of two
olefins, such as propylene and butylene, to form polymer
gasoline.
351
-------
POUR POINT: The minimum temperature at which a petroleum
fraction will flow.
PRESSED DISTILLATE: The oil obtained when wax is separated from
paraffin distillates.
PRESSURE DISTILLATE: Distillate obtained from cracking stills.
QUENCH: The process of cooling hot gases or liquids by direct
contact with cold liquid. Usually refers to quench tower or
quench oil.
R
RAFFINATE: That portion of the oil which is not absorbed by the
solvent in the solvent refining process.
RANGE OIL: See Kerosene.
RAW GASOLINE: See Wild Gas.
REBOILER: A heat exchanger used to boil liquid to provide vapors
to the bottom of a distillation column.
RED OIL: A lube oil which is red in color.
REDUCED CRUDE: The crude oil remaining after distillate
products have been removed in the crude distillation process.
REDWOOD VISCOSITY: A measure of viscosity used in the
petroleum industry.
REFLUX: That portion of the overhead vapors that is condensed
and returned to the distillation column.
REFORMING: A process in which the octane rating of naphtha is
increased by catalytic reaction or mild thermal cracking.
The reformed product is termed reformate.
REID VAPOR PRESSURE TEST: A standard test used to measure
the vapor pressure of gasoline and other petroleum products.
35Z
-------
RESIDUAL FUEL OIL: Fuel oils containing reduced crude.
RICH OIL: An absorption oil containing dissolved hydrocarbons.
SAYBOLT-FUROL, SAYBOLT-UNIVERSAL: Measures of viscosity
used in petroleum industry.
SCALE WAX: Wax obtained by sweating the oil obtained from slack
wax, that is separating excess oil from oily wax.
SEDIMENT: See BS&W.
SHIFT CONVERTER: A reactor used to convert two compounds to
two different compounds, such as are used in sulfur plants
and hydrogen plants.
SHORT RESIDUAL: The residual oil obtained after neutral oils have
been removed by distillation.
SKIMMING: Distillation of crude oil to remove light fractions only.
SLACK WAX: Crude wax obtained by pressing paraffin distillates.
SLOPS OIL: Mixture of oils lost and recovered in the refining
process.
SLUDGE: Degradation residue obtained when treating petroleum.
SMOKE POINT: An index of diesel and kerosene fuels which measures
smoking tendency when burned.
SOLVENT NAPHTHA: See Stoddard Solvent.
SOUR: Containing sulfur compounds such as hydrogen sulfide,
mercaptans, as in sour gas or sour crude.
SPINDLE OIL: A grade of lube oil.
SPRAY OIL: A grade of oil used as a pesticide.
STABILITY: Resistance to change, generally refers to oxidation
resistance of gasoline, other products in storage.
353
-------
STABILIZER: A distillation process which removes light ends,
generally butanes, from naphthas.
STILL: A distillation tower.
STODDARD SOLVENT: A naphtha used in dry cleaning or as a
general solvent.
STRAIGHT RUN: Products directly obtained from distillation of
crude oil before undergoing chemical change, such as
c racking.
STRIPPING: The removal of volative products by heating.
SWEATING: A process in which oil is removed from wax by heating.
SWEET: Containing little sulfur or sulfur compounds, such as
hydrogen sulfides and mercaptans.
TEL: Tetraethyl lead.
TAIL GAS: Sulfurous gases unreacted in sulfur recovery process.
TAR: Highly viscous polymerized residue produced in vacuum
distillation, cracking coils. By-product of the cracking
process.
TEMPERING OIL: Neutral oils.
THERMAL CRACKING: See Cracking.
THERMAL REFORMING: See Reforming.
THIEF: A device for taking samples of petroleum from specific
location in the tank.
TOLUENE: An aromatic hydrocarbon derived from crude oil.
TOPPED CRUDE: Residual crude oil obtained in topping plant.
TOPPING: See Skimming.
TOWER: A vertical vessel in which petroleum is distilled, or gases
are absorbed, etc.
354
-------
TREATING: A process in which petroleum is contacted with chemicals
to improve product quality.
TURNAROUND: A maintenance operation in which a refinery or
process unit is shut down and repaired.
U, V
VACUUM DISTILLATION: Separation of crude oil by distillation
below atmospheric pressure.
VAPOR PRESSURE: Pressure exerted by a liquid at a given
temperature in a closed vessel in the absence of air or
other compounds.
VAPOR RECOVERY: A system used to collect hydrocarbon vapors
from vents and relief devices for reuse in the refinery.
VIRGIN STOCK: See Straight Run.
VISBREAKING: A process of mild thermal cracking in which oil
viscosity is reduced.
VISCOSITY: A measure of resistance to flow, often determined by
the time for liquid to pass through standard orifice.
W
WATER WHITE: A grade of oil color.
WAX DISTILLATE: A neutral oil containing recoverable paraffin wax.
WAX TAILING: Heavy tarlike distillate recovered in coking process.
WET GAS: Light hydrocarbon gas containing heavy hydrocarbons
which are easily condensed.
WHITE OIL: A grade of colorless, light lube oil.
WILD GAS: Natural gasoline containing dissolved light ends.
X, Y, Z
XYLENE: An aromatic hydrocarbon derived from crude oil.
YELLOW SCALE: Low-grade paraffin wax.
355
-------
' TECHNICAL REPORT DATA
(Please read iHUntiituns on the n-\\-rK bi'Jorc completing)
EPA-450/3-74-Gffi
4,'TITLE AND SUBTITLE
Field Surveillance & Enforcement-Guide for Petroleum
Refineries
6. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Anker V. Sims
8. PERFORMING ORGANIZATION REPORT NO,
9. PERFORMING OR'ANIZATION NAME AND ADDRESS
The Ben Holt Company
201 South Lake Avenue
Pasadena, California 91101
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRAN1 NO.
68-02-0645
12. SHONSCFUNG AGENCY NAME AND ADDRESS
Control Program Development Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final Report
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The report entitled "Field Surveillance and Enforcement Guide for Petroleum
Refineries" describes petroleum refining and natural gas processing, refinery
equipment, process instrumentation, air pollution monitoring instrumentation,
maintenance of refinery records for use by air pollution control personnel,
estimating and assessing emissions, plant and equipment maintenance and the
qualifications and training requirements of field enforcement personnel. The
guide was prepared to familiarize state and local air pollution control officials
with the operation of petroleum refineries and natural gas processing plants and
to aid agency personnel in developing surveillance, inspection, monitoring,
reporting and enforcement procedures.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution
Petroleum Refining
Natural Gas
Refineries
Standards
Instrumentation
Inspection
Personnel
"Reporting
Particles
Sulfur Oxides
Hydrogen Sulfide
Organic Sulfur Com-
pounds
13 DiSI Rl BUTION ST ATtMt NT
Release Unlimited
b.IDENTIFIERS/OPEN ENDED TERMS
Air Pollution ConTroT
Petroleum Refining
Natural Gas Processing
Inspection Procedures
19. SECURITY CLASS (I'lm Kcport>
Unclassified
20. SECURITY CLASS lT>ns pu
Unclassified
COSATi F icId/Group
21. NO. OF PAGES
377
22. PRICE
EPA Form 2220-1 (9-73)
------- |