United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
January 1981
Air
Guideline Series
Control of Volatile
Organic Fugitive
Emissions from Synthetic
Organic Chemical,
Polymer, and Resin
Manufacturing Equipment
Preliminary Draft
-------
NOTICE
This document has not been formally released by EPA and should not now be construed to represent
Agency policy. It is being circulated for comment on its technical accuracy and policy implications.
Control of Volatile Organic
Fugitive Emissions from Synthetic
Organic Chemical, Polymer, and
Resin Manufacturing Equipment
Emission Standards and Engineering Division
Contract No. 68-02-3168
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
. Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
January 1981
-------
GUIDELINE SERIES
The guideline series of reports is issued by the Office of Air Quality
Planning and Standards (OAQPS) to provide information to state and local
air pollution control agencies; for example, to provide guidance on the
acquisition and processing of air quality data and on the planning and
analysis requisite for the maintenance of air quality. Reports published in
this series will be available - as supplies permit - from the Library Services
Office (MD-35), U. S. Environmental Protection Agency, Research Triangle
Park, North Carolina 27711, or for a nominal fee, from the National
Technical Information Service, 5285 Port Royal Road, Springfield, Virginia
22161.
11
-------
TABLE OF CONTENTS
Page
List of Tables v
List of Figures viii
Chapter 1.0 Introduction and Summary 1-1
1.1 Introduction 1-1
1.2 Summary of Model Regulation 1-2
1.3 Hazardous Materials 1-3
Chapter 2.0 Processes and Pollutant Emissions 2-1
2.1 General 2-1
2.2 Facilities and Their Emissions 2-1
2.3 Model Units 2-17
Chapter 3.0 Emission Control Techniques 3-1
3.1 Leak Detection and Repair Methods 3-1
3.2 Equipment Specifications for Open-Ended Valves . . 3-12
Chapter 4.0 Environmental Analysis of RACT . . . 4-1
4.1 Air Pollution 4-1
4.2 Water Pollution 4-4
4.3 Solid Waste Disposal 4-4
4.4 Energy 4-4
Chapter 5.0 Control Cost Analysis of RACT 5-1
5.1 Basis for Capital Costs 5-1
5.2 Basis for Annualized Costs 5-3
5.3 Emission Control Costs 5-8
5.4 Cost Effectiveness 5-12
-------
TABLE OF CONTENTS (continued)
Page
Chapter 6.0 Model Regulation and Discussion 6-1
6.1 Model Regulation 6-1
6.2 Discussion 6-5
Appendix A. Emission Source Test Data A-l
Appendix B. List of Chemicals Defining Synthetic Organic
Chemical, Polymer, and Resin Manufacturing
Industries B-l
Appendix C. Method 21. Determination of Volatile Organic
Compound Leaks C-l
Appendix D. Example Calculations for Determining Reduction
in Emissions from Implementation of RACT D-l
iv
-------
LIST OF TABLES
Table 2-1
Table 2-2
Table 2-3
Table 2-4
Table 3-1
Table 3-2
Table 3-3
Table 3-4
Table 3-5
Table 4-1
Table 4-2
Table 4-3
Table 4-4
Page
Fugitive Emission Sources for Three Model Units . . 2-18
Uncontrolled Fugitive Emission Factors in Process
Unit Equipment 2-21
Estimated Total Fugitive Emissions from Model
Unit
2-22
Average Percent of Total Fugitive Emissions Attributed
to Specific Component Types 2-23
Percentage of Emissions as a Function of Action
Level 3-5
Estimate Occurrence and Recurrence Rate of Leaks
for a Quarterly Monitoring Interval 3-6
Average Emission Rates from Sources Above
10,000 ppmv and at 1000 ppmv 3-8
Impact of Monitoring Interval on Correction Factor
Accounting for Leak Occurrence/Recurrence (For Example
Calculation) 3-11
Example of Control Efficiency Calculation 3-13
Estimated Hourly Emissions and Emissions Reduction
on a Model Unit Basis 4-2
Estimated Annual Emissions and Emissions Reduction
on a Model Unit Basis 4-2
Emission Factors for Sources Controlled Under RACT . 4-3
Example Calculation of VOC Fugitive Emissions from
Model Unit A Under RACT 4-5
-------
Table 5-1
Table 5-2
Table 5-3
Table 5-4
Table 5-5
Table 5-6
Table 5-7
Table A-l
Table A-2
Table A- 3
Table A-4
Table A-5
Table A-6
Table A- 7
Table A-8
Table A-9
Table A-10
Table A-ll
Capital Cost Data
Capital Cost Estimates for Implement!' ng RACT. . . .
Labor-Hour Requirements for Initial Leak Repair
Under RACT
Basis for Annual i zed Cost Estimates
Annual Monitoring and Leak Repair Labor Requirements
for RACT
Recovery Credits
Annual i zed Control Estimates for Model Units Under
RACT
Frequency of Leaks from Fugitive Emission Sources in
Synthetic Organic Chemical Units
Twenty- four Chemical Process Units Screened for
Fugitive Emissions
Summary of SOCMI Process Units Fugitive Emissions . .
Average Fugitive Emission Source Screening Rates .
Sampled Process Units from Nine Refineries During
Refinery Study
Leak Frequencies and Emission Factors from Fugitive
Emission Sources in Petroleum Refineries
Comparison of Leak Frequencies for Fugitive Emission
Sources in SOCMI Units and Petroleum Refineries. . .
Frequency of Leaks from Fugitive Emission Sources
in Two DuPont Plants
Frequency of Leaks from Fugitive Emission Sources
in Exxon's Cyclohexane Unit
Summary of Maintenance Study Results from the Union
Oil Co. Refinery in Rodeo, California
Summary of Maintenance Study Results from the Shell
Oil Company Refinery in Martinez, California. . . .
Page
5-2
5-2
5-4
5-5
5-7
5-9
5-11
A-4
A-5
A-8
A-9
A-10
A- 12
A-13
A-14
A-16
A-18
A-20
-------
Page
Table A-12 Summary of EPA Refinery Maintenance Study Results . . A-21
Table A-13 Maintenance Effectiveness Unit D Ethylene Unit
Block Valves A-22
Table D-l Uncontrolled Fugitive Emission Factors in Process
Unit Equipment D-2
Table D-2 Controlled Emission Factors for Equipment Affected
by RACT D-2
Table D-3 Example Calculation of Uncontrolled Emissions from
an Illustrative Process Unit D-3
Table D-4 Uncontrolled Emissions from Components Affected
by RACT D-4
Table D-5 Controlled Emissions from Components Affected by RACT. D-4
Table D-6 Emission Reduction Expected from RACT D-5
-------
LIST OF FIGURES
Figure 2-1
Figure 2-2
Figure 2-3
Figure 2-4
Figure 2-5
Figure 2-6
Figure 2-7
Figure 2-8
Figure 2-9
Figure 2-10
Figure 2-11
Figure 2-12
Figure 2-13
Figure 2-14
Figure 2-15
Figure 2-16
Figure 3-1
Figure 3-2
Figure C-l
Figure C-2
Page
General schematic of process levels that make up
the organic chemical industry 2-2
Diagram of a simple packed seal 2-3
Diagram of a basic single mechanical seal 2-4
Diagram of a double mechanical seal 2-5
Diagram of a double mechanical seal 2-5
Diaphragm pump 2-7
Labyrinth shaft seal 2-8
Restrictive-ring shaft seal 2-8
Mechanical contact shaft seal 2-10
Liquid film shaft seal with cylindrical bushing . . . 2-10
Diagram of a gate valve 2-11
Example of bellows seals 2-12
Diagrams of valves with diaphragm seals 2-12
Diagram of a spring loaded relief valve 2-13
Diagram of hydraulic seal for agitators 2-15
Diagram of agitator lip seal 2-16
Cumulative distribution of total emissions by screening
values - valves on gas streams 3-10
Cumulative distribution of sources by screening
values - valves on gas streams 3-10
Calibration Precision Determination C-8
Response Time Determination C-9
vm
-------
1.0 INTRODUCTION AND SUMMARY
1.1 INTRODUCTION
The Clean Air Act Amendments of 1977 require each State in which there
are areas in which the national ambient air quality standards (NAAQS) are
exceeded to adopt and submit revised state implementation plans (SIP's) to
EPA. Revised SIP's were required to be submitted to EPA by January 1, 1979.
States which were unable to demonstrate attainment with the NAAQS for ozone
by the statutory deadline of December 31, 1982, could request extensions for
attainment with the standard. Such extensions could not go beyond
December 31, 1987. States granted such an extension are required to submit
a further revised SIP by July 1, 1982.
Section 172(a)(2) and (b)(3) of the Clean Air Act require that
nonattainment area SIP's include reasonably available control technology
(RACT) requirements for stationary sources. As explained in the "General
Preamble for Proposed Rulemaking on Approval of State Implementation Plan
Revisions for Nonattainment Areas," (44 FR 20372, April 4, 1979) for ozone
SIP's, EPA permitted States to defer the adoption of RACT regulations on a
category of stationary sources of volatile organic compounds (VOC) until
after EPA published a control techniques guideline (CTG) for that VOC source
category. See also 44 FR 53761 (September 17, 1979). This delay allowed
the states to make more technically sound decisions regarding the application
of RACT.
The CTG documents provide State and local air pollution control agencies
with an information base for proceeding with development and adoption of
regulations which reflect RACT for specific stationary sources. Consequently,
CTG documents review existing information and data concerning the technology
and cost of various control techniques to reduce emissions. The CTG documents
also identify control techniques and suggest emission limitations which EPA
considers the "presumptive norm" broadly representative of RACT for the
entire stationary source category covered by a CTG document.
1-1
-------
The CTG documents are, of necessity, general in nature and do not fully
account for variations within a stationary source category. RACT, however,
is defined as the lowest emission limitation that a particular source is
capable of meeting by the application of emission control technology that is
reasonably available considering technical and economic feasibility. Thus,
reasons may exist for regulations developed by States to deviate from the
"presumptive norm" included in a CTG document. The CTG document, however,
is a part of the rulemaking record which EPA considers in reviewing revised
SIP's, and the information and data contained in the document is highly
relevant to EPA's decision to approve or disapprove a SIP revision. Where a
State adopts emission limitations that are consistent with the information
in the CTG, it may be able to rely solely on the information in the CTG to
support its determination of RACT. Where this is not the case, the State
must include documentation with its SIP revision to support and justify its
RACT determination.
This draft CTG document includes a model regulation based upon the
"presumptive norm" considered broadly representative of RACT for the
stationary source category covered by this document. The sole purpose of
this model regulation is to assist State and local agencies in development
and adoption of regulations for specific stationary sources. This model
regulation is not to be construed as rulemaking by EPA.
This CTG document is being released in working draft form to achieve
two objectives. First, to provide an opportunity for public review and
comment on the information and regulatory guidance contained in the document;
and second, to provide as much assistance and lead time as possible to State
and local agencies preparing RACT regulations for specific stationary sources
covered by this document.
1.2 SUMMARY OF MODEL REGULATION
The model regulation applies to the synthetic organic chemical
manufacturing industry and plants manufacturing specific polymers and resins.
A list of the chemicals, polymers, and resins produced by plants to which
the "model regulation" applies is presented in Appendix B.
1-2
-------
The RACT selected for control of fugitive emissions in process units is
a leak detection and repair program. The "model regulation" requires that
specific components in contact with process fluid containing at least
ten percent volatile organic compounds (VOC) by weight must be monitored
with a hydrocarbon detection instrument once every three months. These
components are: pumps in light liquid service, compressors, valves in light
liquid and gas service, and pressure relief valves in gas service. Pumps in
light liquid service must be visually checked weekly for indications of
leaks. Components having concentrations of VOC at or above 10,000 ppmv are
considered leaking components. Leaking components must be repaired within
fifteen days of the date of detection. A tag must be affixed to a component
when a leak is detected and must remain in place until the leak is repaired.
A plant must keep a record of leaks detected and a record of leaks
repaired. A plant must keep a copy of the inspection log at the plant for
two years and make it available to the Director of the State air pollution
agency or authorized representative upon request.
Once each quarter, the plant must submit a report to the State including
the number of leaking components detected but not repaired and the total
number of leaking components found. The plant must also report the total
number of components inspected during the quarter.
Provisions are made in the model regulation for the Director to approve
a plant's alternative program for control of fugitive emissions. The
alternative program must be: equivalent to the leak detection and repair
program in the State reguletion. A recommended procedure for determining
equivalency of alternative programs is outlined in Chapter 6.
1.3 HAZARDOUS MATERIALS
The EPA has proposed e carcinogen policy under section 112 of the Clean
Air Act (44FR58643, 44FR58C-62, October 10, 1979) outlining procedures for
regulation of substances de'termined to be carcinogenic which pose a signifi-
cant risk to public health. Where appropriate, following listing of a sub-
stance as hazardous under s.ection 112, this policy would result in immediate
proposal of regulations limiting emissions of this substance. These regulations
1-3
-------
would consist primarily of requirements to institute a leak detection and
repair program similar to the program outlined above. The major difference
would most likely be the time interval between inspections of equipment
leaks. The inspection interval in the leak detection and repair program
associated with the proposed carcinogen policy is monthly, whereas the inspection
interval included in the leak detection and repair program recommended in
this document is quarterly.
A few of the organic chemicals listed in Appendix B are currently under
investigation and may eventually be found to be carcinogenic. If so, more
stringent leak detection and repair programs may be proposed by EPA under
section 112 of the Clean Air Act for these chemicals in the future.
1-4
-------
2.0 PROCESSES AND POLLUTANT EMISSIONS
2.1 GENERAL
The synthetic organic chemical manufacturing industry (SOCMI) is a
segment of the chemical industry consisting of some of the higher volume
intermediate and finished products. A list of these chemicals is presented
in Appendix B, Table I.
Most of the SOCMI chemicals produced in the United States are derived
from crude petroleum or natural gas. The ten principal feedstocks used in
the manufacture of organic chemicals are produced primarily in petroleum
refineries. After chemical feedstocks are manufactured from petroleum,
natural gas, and other raw materials, they are processed into chemical
intermediates and end-use chemicals (see Figure 2-1). Approximately
12 percent of the plants in the United States produce less than 5,000 megagrams
(Mg) annually. Another 12 percent have production capacities in excess of
500,000 Mg.
The equipment in SOCMI process units is similar to equipment in polymer
and resin manufacturing process units. Therefore, the information and
discussion in this chapter and subsequent chapters applies equally to SOCMI
plants and polymer and resin manufacturing plants where process equipment isV
similar. A list of polymers and resins is presented in Appendix B, Table II.
2.2 FACILITIES AND THEIR EMISSIONS
2.2.1 Potential Source Characterization and Description
In this document, fugitive emissions from process units are considered
to be those volatile organic compound (VOC) emissions that result when
process fluid (either gaseous or liquid) leaks from plant equipment. There
are many potential sources of fugitive emissions in a typical process unit.
The following sources will be considered in this chapter: pumps, compressors,
in-line process valves, pressure relief devices, open-ended valves, sampling
connections, flanges, agitators and cooling towers. These potential sources
are described below.
2-1
-------
RAW MATERIALS
(CRUDE OIL, CRUDE NATURAL GAS, ETC.)
REFINERIES
CHEMICAL
FEEDSTOCK
PLANTS
CHEMICAL
FEEDSTOCKS
CHEMICAL
PLANTS
CHEMICAL
PRODUCTS
Figure 2-1. General schematic of process levels that make up
the organic chemical industry.
2-2
-------
2.2.1.1 Pumps. Pumps are used extensively in process units for the
movement of organic liquids. The centrifugal pump is the most widely used
pump. However, other types, such as the positive-displacement, reciprocating
and rotary action, and special canned and diaphragm pumps, are also used.
Chemicals transferred by pumps can leak at the point of contact between the
moving shaft and stationary casing. Consequently, all pumps except the
shaftless type (canned-motor and diaphragm) require a seal at the point
where the shaft penetrates the housing in order to isolate the pump's
interior from the atmosphere.
Two generic types of seals, packed and mechanical, are currently in
use on pumps. Packed seals can be used on both reciprocating and rotary
action types of pumps. As Figure 2-2 shows, a packed seal consists of a
cavity ("stuffing box") in the pump casing filled with special packing
material that is compressed with a packing gland to form a seal around the
shaft. Lubrication is required to prevent the buildup of frictional heat
between the seal and shaft. The necessary lubrication is provided by a
2
lubricant that flows between the packing and the shaft. Deterioration of
the packing will result in process liquid leaks.
Fluid ^
End C
1
Stuffing
Box
I \
nn
(xixixixixixixr-
Packing
/ Gland
— — •) Atniospnere
I) FnH
v\
[XIXIXIXIXCKIXl
f
1
Packing
.
n_4^
\
/ Possible
Leak
Area
Figure 2-2. Diagram of a simple packed seal.3
2-3
-------
Mechanical seals are limited in application to pumps with rotating
shafts and can be further categorized as single and double mechanical seals.
There are many variations to the basic design of mechanical seals, but all
have a lapped seal face between a stationary element and a rotating seal
ring. In a single mechanical seal application (Figure 2-3), the rotating-seal
ring and stationary element faces are lapped to a very high degree of flatness
to maintain contact throughout their entire mutual surface area. As with a
packed seal, the seal faces must be lubricated to remove frictional heat,
however, because of its construction, much less lubricant is needed.
A mechanical seal is not a leak-proof device. Depending on the
condition and flatness of the seal faces, the leakage rate can be quite low
(as small as a drop per minute) and the flow is often not visually
detectable. In order to minimize fugitive emissions due to seal leakage,
4
an auxiliary sealing device such as packing can be employed.
PUMP
STUFFING
BOX
GLAND
'RING
FLUID
END
STATIONARY
ELEMENT
POSSIBLE
LEAK AREA
SHAFT
ROTATING
SEAL RING
Figure 2-3. Diagram of a basic single mechanical seal.
In a dual mechanical seal application, two seals can be arranged
back-to-back or in tandem. In the back-to-back arrangement (Figure 2-4), the
two seals provide a closed cavity between them. A seal liquid, such as water
or seal oil, is circulated through the cavity. Because the seal liquid
surrounds the double seal and lubricates both sets of seal faces in this
arrangement, the heat transfer and seal life characteristics are much better
2-4
-------
SEAL LIQUID
POSSIBLE LEAK
INTO SEALING
FLUID
FLUID END
PRIMARY
SEAL
v
SECONDARY
SEAL
GLAND
PLATE
Figure 2-4. Diagram of a double mechanical seal
(back-to-back arrangement)
PRIMARY
SEAL
BUFFER LIQUID
OUT IN
(TOP) (BOTTOM)
SECONDARY
SEAL
GLAND
PLATE
70-1767-1
Figure 2-5. Diagram of a double mechanical seal
(tandem arrangement)8
2-5
-------
than those of the single seal. In order for the seal to function, the seal
liquid must be at a pressure greater than the operating pressure of the
stuffing box. As a result some seal liquid will leak across the seal faces.
Liquid leaking across the inboard face will enter the stuffing box and mix
with the process liquid. Seal liquid going across the outboard face will
exit to the atmosphere. '
In a tandem dual mechanical seal arrangement (Figure 2-5), the seals
face the same direction. The secondary seal provides a backup for the
primary seal. A seal flush is used in the stuffing box to remove the heat
generated by friction. The cavity between the two seals is filled with a
buffer or barrier liquid. However, the barrier liquid is at a pressure lower
than that in the stuffing box. Therefore, any leakage will be from the
stuffing box into the seal cavity containing the barrier liquid. Since this
liquid is routed to a closed reservoir, process liquid that has leaked into
the seal cavity will also be transferred to the reservoir. At the reservoir,
the process liquid could vaporize and be emitted to the atmosphere. To
ensure that VOC's do not leak from the reservoir, the reservoir can be
g
vented to a control device.
Another type of pump that has been used is the shaft!ess pump which
includes canned-motor and diaphragm pumps. In canned-motor pumps the cavity
housing the motor rotor and the pump casing are interconnected. As a result,
the motor bearings run in the process liquid and all seals are eliminated.
Because the process liquid is the bearing lubricant, abrasive solids cannot
be tolerated. Canned-motor pumps are being widely used for handling organic
solvents, organic heat transfer liquids, light oils, as well as many toxic or
hazardous liquids, or where leakage is an economic problem.
Diaphragm pumps (see Figure 2-6) perform similarly to piston and plunger
pumps. However, the driving member is a flexible diaphragm fabricated of
metal, rubber, or plastic. The primary advantage of this arrangement is the
elimination of all packing and seals exposed to the process liquid. This is
an important asset when hazardous or toxic liquids are handled.
2-6
-------
DISCHARGE
CHECK VALVE
INLET
CHECK VALVE
DIAPHRAGM
PISTON
Figure 2-6. Diaphragm pump.
12
2.2.1.2 Compressors. Gas compressors used in process units are
similar to pumps in that they can be driven by rotary or reciprocating
shafts. They are also similar to pumps in their need for shaft seals to
isolate the process gas from the atmosphere. As with pumps, these seals
are likely to be the source of fugitive emissions from compressors.
Shaft seals for compressors may be chosen from several different
types: labyrinth, restrictive carbon rings, mechanical contact, and liquid
film. All of these seal types are leak restriction devices; none of them
completely eliminate leakage. Many compressors may be equipped with ports
in the seal area to evacuate gases collecting there.
The labyrinth type of compressor seal is composed of a series of close
tolerance, interlocking "teeth" which restrict the flow of gas along the
shaft. A straight pass labyrinth compressor seal is shown in Figure 2-7.
Many variations in "tooth" design and materials of construction are available.
Although labyrinth type seals have the largest leak potential of the different
types, properly applied variations in "tooth" configuration and shape can
reduce leakage by up to 40 percent over a straight pass type labyrinth.
2-7
-------
PORT MAY BE ADDED
FOR SCAVENGING OH
INERT-GAS SEALING
ATMOSPHERE
Figure 2-7. Labyrinth shaft seal.
14
SCAVENGING
PORT MAY BE
ADDED FOR
VACUUM
APPLICATION
ATMOSPHERE
Figure 2-8. Restrictive-ring shaft seal.
15
2-8
-------
Restrictive carbon ring seals consist of multiple stationary carbon
rings with close shaft clearances. This type of seal may be operated dry
or with a sealing fluid. Restrictive ring seals can achieve lower leak
rates than the labyrinth type. A restrictive ring seal is shown in
Figure 2-8.
Mechanical contact seals (shown in Figure 2-9) are similar to the
mechanical seals described for pumps. In this type of seal clearance
between the rotating and stationary elements is reduced to zero. Oil or
another suitable lubricant is supplied to the seal faces. Mechanical seals
can achieve the lowest leak rates of the types described here, but they are
not suitable for all processing conditions.
Centrifugal compressors also can be equipped with liquid film seals.
A diagram of a liquid film seal is shown in Figure 2-10. The seal is
formed by a film of oil between the rotating shaft and stationary gland.
When the circulating oil is returned to the oil reservoir, process gas can
I O
be released to the atmosphere. To eliminate release of VOC emissions
from the seal oil system, the reservoir can be vented to a control device.
2.2.1.3 Process Valves. One of the most common pieces of equipment
in organic chemical plants is the valve. The types of valves commonly used
are control, globe, gate, plug, ball, relief, and check valves. All except
the relief valve (to be discussed further below) and check valve are activated
by a valve stem, which may have either a rotational or linear motion,
depending on the specific design. This stem requires a seal to isolate the
process fluid inside the valve from the atmosphere as illustrated by the
diagram of a gate valve in Figure 2-11. The possibility of a leak through
this seal makes it a potential source of fugitive emissions. Since a check
valve has no stem or subsequent packing gland, it is not considered to be a
potential source of fugitive emissions.
Sealing of the stem to prevent leakage can be achieved by packing
inside a packing gland or 0-ring seals. Valves that require the stem to
move in and out with or without rotation must utilize a packing gland.
Conventional packing glands are suited for a wide variety of packing
materials. The most common are various types of braided asbestos that
2-9
-------
INTERNAL
GAS PRESSURE
CLEAN OIL IN
PRESSURE
/- BREAKDOWN
I SLEEVE
CONTAMINATED
OIL OUT
Figure 2-9. Mechanical (contact) shaft seal.
CLEAN OIL IN
ATMOSPHERE
CONTAMINATED
OIL OUT
OIL OUT
Figure 2-10. Liquid film shaft seal with cylindrical bushing.
15
2-10
-------
contain lubricants. Other packing materials include graphite,
graphite-impregnated fibers, and tetrafluoroethylene. The packing material
19
used depends on the valve application and configuration. These conventional
packing glands can be used over a wide range of operating temperatures. At
high pressures these glands must be quite tight to attain a good seal.20
PACKING
GLAND
PACKING
VALVE
STEM
POSSIBLE
LEAK AREAS
21
Figure 2-11. Diagram of a gate valve.
Elastomeric 0-rings are also used for sealing process valves. These
0-rings provide good sealing but are not suitable where there is sliding
motion through the packing gland. Those seals are rarely used in high pressure
22.
service and operating temperatures are limited by the seal material.
Bellows seals are more effective for preventing process fluid leaks
than the conventional packing gland or any other gland-seal arrangement.^
This type of seal incorporates a formed metal bellows that makes a barrier
between the disc and body bonnet joint. An example of this seal is presented
in Figure 2-12. The bellows is the weak point of the system and service
life can be quite variable. Consequently, this type of seal is normally
backed up with a conventional packing gland and is often fitted with a leak
detector in case of failure.2^
2-11
-------
BELLOWS
BODY
BONNET
Figure 2-12. Example of bellows seals.
25
A diaphragm may be used to isolate the working parts of the valve and
the environment from the process liquid. Two types of valves which utilize
diaphragms are illustrated in Figures 2-13(a) and (b). As Figure 2-13(b)
shows, the diaphragm may also be used to control the flow of the process
fluid. In this design, a compressor component pushes the diaphragm toward
the valve bottom, throttling the flow. The diaphragm and compressor are
connected in a manner so that it is impossible for them to be separated under
normal working conditions. Mhen the diaphragm reaches the valve bottom, it
seats firmly against the bottom, forming a leak-proof seal. This configuration
is recommended for fluids containing solid particles and for medium-pressure
service. Depending on the diaphragm material, this type of valve can be used
at temperatures up to 205°C and in severe acid solutions. If failure of the
seal occurs, a valve employing a diaphragm seal can become a source of
?fi
fugitive emissions.
STEM
DIAPHRAGM
DIAPHRAGM
DISK
70-177M
?7
Figure 2-13. Diagrams of valves with diaphragm seals.
2-12
-------
.2.2.1.4 Pressure Relief Devices. Engineering codes require that
pressure-relieving devices or systems be used in applications where the
process pressure may exceed the maximum allowable working pressure of the
vessel. The most common type of pressure-relieving device used in process
units is the pressure relief valve (Figure 2-14). Typically, relief valves
are spring-loaded and designed to open when the process pressure exceeds a
set pressure, allowing the release of vapors or liquids until the system
pressure is reduced to its normal operating level. When the normal pressure
28
is reattained, the valve reseats, and a seal is again formed. The seal
is a disk on a seat, and the possibility of a leak through this seal makes
the pressure relief valve a potential source of VOC fugitive emissions.
Two potential causes of leakage from relief valves are: "simmering or
popping", a condition due to the system pressure being close to the set
pressure of the valve, and improper reseating of the valve after a relieving
29
operation.
Possible
Leak Area
Process Side
Figure 2-14. Diagram of a spring-loaded relief valve.
2-13
-------
Rupture disks are also common in process units. These disks are made
of a material that ruptures when a set pressure is exceeded, thus allowing
the system to depressurize. The advantage of a rupture disk is that the
disk seals tightly and does not allow any VOC's to escape from the system
under normal operation. However, when the disk does rupture, the system
depressurizes until atmospheric conditions are obtained. This could result
in an excessive loss of product or a corresponding excessive release of
fugitive emissions.
2.2.1.5 Cooling Towers. The purpose of cooling towers is to cool the
plant's process cooling waters which have been heated while removing heat
from various process equipment (reactors, condensers, heat exchangers).
This cooling process is achieved by evaporation when the process cooling
water and air are contacted. Under normal operating conditions, a cooling
tower would not be considered a fugitive emission source. However, if a
leak occurs in the process equipment and if thi.s equipment is operating at
a pressure greater than that of the cooling water, organic chemicals can
leak into the water. When the process water is recirculated to the cooling
30
tower, these chemicals can be released to the atmosphere.
2.2.1.6 Agitators. Agitators are commonly used to stir or blend
chemicals. Like pumps and compressors, agitators may leak organic chemicals
at the point where the shaft penetrates the casing. Consequently, seals
are required to minimize fugitive emissions from agitators. Four seal
arrangements are commonly used with agitators. These are compression
31
packing (packed seal), mechanical seals, hydraulic seals, and lip seals.
Packed seals for agitators are very similar in design and application to
the packed seals for pumps (Section 2.2.1.1).
Although mechanical seals are more costly than the other three seal
arrangements, they offer a greatly reduced leakage rate to offset their
higher cost. The maintenance frequency of mechanical seals is, also,
32
one-half to one-fourth that of packed seals. In fact, at pressures
greater than 1135.8 kPa (150 psig), the leakage rate and maintenance frequency
33
are so superior that the use of packed seals on agitators is rare. As
with packed seals, the mechanical seals for agitators are similar to the
design and application of mechanical seals for pumps (Section 2.2.1.1.)
2-14
-------
The hydraulic seal (Figure 2-15) is the simplest and least used agitator
shaft seal. In this type of seal, an annular cup attached to the process
vessel contains a liauid that is in contact with an inverted cup attached
to the rotating agitator shaft. The primary advantage of this seal is that
it is a non-contact seal. However, this seal is limited to low temperatures
and pressures and can only handle very small pressure fluctuations. Organic
chemicals may contaminate the seal liquid and then be released into the
atmosphere as fugitive emissions.
INVERTED CUP-
ANNULAR CUP-
Figure 2-15. Diagram of a hydraulic seal for agitators.
35
A lip seal (Figure 2-16) can be used on a too-enterina agitator as a
or vapor seal. The sealing element is a spring-loaded elastomer. Lip
seals are relatively inexpensive and easy to install. Once the seal has
been installed the agitator shaft rotates in continuous contact with the
lio seal. Pressure limits of the seal are 2 to 3 osi because it operates
without lubrication. Operating temperatures are limited by characteristics
of the elastomer. Fugitive VOC emissions could be released through this
seal when this seal wears excessively or the operating pressure surpasses
the pressure limits of the seal. °
2-15
-------
37
Figure 2-16. Diagram of agitator lip seal.
2.2.1.7 Open-Ended Valves or Lines. Some valves are installed in a
system so that they function with the downstream line open to the atmosphere.
Examples are purge valves, drain valves, and vent valves. A faulty valve
seat or incompletely closed valve would result in leakage through the valve
and fugitive VOC emissions to the atmosphere.
2.2.1.8 Sampling Connections. The operation of a process unit is
checked periodically by routine analyses of feedstocks and products. To
obtain representative samples for these analyses, sampling lines must first
be purged prior to sampling. The purged liquid or vapor is sometimes drained
onto the ground or into a sewer drain, where it can evaporate and release VOC
emissions to the atmosphre.
2.2.1.9 Flanges. Flanges are bolted, gasket-sealed junctions used
wherever pipe or other equipment such as vessels, pumps, valves, and heat
exchangers may require isolation or removal. Normally, flanges are employed
for pipe diameters for 50 mm or greater and are classified by pressure and
face type.
Flanges may become fugitive emission sources when leakage occurs due to
improperly chosen gaskets or a poorly assembled flange. The primary cause of
flange leakage is due to thermal stress that piping or flanges in some services
undergo; this results in the deformation of the seal between the flange
faces. 38
2-16
-------
2.3 MODEL UNITS
This section presents model process unit parameters. The model units
were selected to represent the range of processing complexity in the industry.
They provide a basis for determining environmental and cost impacts of
reasonably available control technology (RACT).
2.3.1 Model Units
Available data show that fugitive emissions are proportional to the
number of potential sources, but are not related to capacity, throughput,
39
age, temperature, or pressure. Therefore, model units defined for this
analysis represent different levels of process complexity (number of sources)
rather than different unit sizes.
2.3.1.1 Sources of Fugitive Emissions. Data from petroleum refineries
40
indicate that cooling towers are very small sources of VOC emission.'
Differences in operating procedures, such as recirculation of process water,
might result in cooling tower VOC emissions, but no data are available to
verify this. The number of agitator seals in the industry is not known.
Furthermore, the emission rate from agitator seals has not been measured.
Since there are no data from similiar sources in other industries, no estimates
of emission rate can be made. Because of these uncertainties, cooling towers
and agitator seals are not included in the Model Units.
2.3.1.2 Model Units Components. In order to estimate emissions, control
costs, and environmental impacts for process units on a unit specific basis,
three model units were developed. The equipment components comprising the
model units are shown in Table 2-1. These three model units represent the
range of emission source peculations that may exist in SOCMI process units.
The number of equipment comoonents for each model unit was developed from a
data base compiled by Hydroscience, Inc. The data base included equipment
source counts from 62 SOCMI plants which produce 35 different chemicals.
These plant sites represent approximately 5 percent of the total existing
SOCMI plants and include large and small capacities, batch and continuous
production methods, and varying levels of process complexity. The source
counts for the 35 chemicals include pumps, valves, and compressors. These
2-17
-------
TABLE 2-1. FUGITIVE EMISSION SOURCES FOR THREE MODEL UNITS
Equipment component
Pump seals
Light liquid service
Single mechanical
Double mechanical
Sealless
Heavy liquid service
Single mechanical
Packed
In-line valves
Vapor service
Light liquid service
Heavy liquid service
Safety/relief valves
Vapor service
Light liquid service
Heavy liquid service
Open-ended valves and lines
Vapor service
Light liquid service
Heavy liquid service
Compressor seals
Sampling connections0
Flanges
Cooling towers
Number of
Model unit
A
5
3
0
5
2
90
84
84
11
1
1
9
47
48
1
26
600
~e
components in
Model unit
B
19
10
1
24
6
365
335
335
42
4
4
37
189
189
2
104
2400
-e
model unit
Model unit
C
60
31
1
73
20
1117
1037
1037
130
13
14
115
581
581
8
320
7400
~e
Equipment components in VOC service only.
Sample, drain, purge valves.
cBased on 25 percent of open-ended valves. From Ref. 1, pg. IV-3.
52 percent of existing SOCMI units are similar to Model Unit A.
33 percent of existing SOCMI units are similar to Model Unit B.
15 percent of existing SOCMI units are similar to Model Unit C.
Ref. 1, pg. IV-1.
eData not available.
2-18
-------
counts were used in combination with the number of sites which produce each
d?
chemical in order to determine,the average number of sources per site.
Hydroscience estimates that 52 percent of existing SOCMI plants are similar
to the Model Unit A, 33 percent are similar to B, and 15 percent are similar
to C.
Data from petroleum refineries indicate that emission rates of sources
decrease as the vapor pressure (volatility) of the process fluid decreases.
Three classes of volatility have been established based on the petroleum
refinery data. These include gas/vapor service, light liquid service, and
heavy liquid service. 43 The split between light and heavy liquids for the
refinery data is between naphtha and kerosene. Since similar stream names
may have different vapor pressures, depending on site specific factors, it is
difficult to quantify the light-heavy split. The break point is approximately
at a vapor pressure of 0.3 kPa at 20°C. The data collected by Hydroscience
were used to estimate the split between gas/vapor and liquid service for each
source. 44 in order to apply emission factors for light and heavy liquid
service, it is assumed that one-half of SOCMI liquid service sources are in
light liquid service. There are no data available on the actual distribution
of sources in volatility ranges. It is assumed that all packed seal pumps
are in heavy liquid service. This assumption is reasonable, since more
volatile liquid are more suitable for mechanical seal applications, and newer
process units tend to use fewer packed seals. Sampling connections are a
subset of the open-ended valve category. Approximately 25 percent of
open-ended valves are used for sampling connections.45 Emissions which occur
through the valve stem, gland, and open-end are included in the open-ended
valve category. The emission factor for sampling connection applies only to
emissions which result from sample purging.
2.3.1.3 Uncontrolled Fugitive Emission Estimates. Data characterizing
the uncontrolled levels of fugitive emissions in the SOCMI are presently
unavailable. However, data on this type have been obtained for the refining
industry. These data are presented in Table 2-2, and represent the average
uncontrolled emission rate from each of the components of a specific type in
the process unit. Because the operation of the various process equipment in the
2-19
-------
SOCMI is not expected to differ greatly from the operation of the same
equipment in the refining industry, the refinery fugitive emission data can
be used to approximate the levels of fugitive emissions in SOCMI.
The total amount of VOC emitted from fugitive sources can be estimated
for each Model Unit. Total hourly emissions can be calculated by multiplying
the number of pieces of each type of equipment (Table 2-1) by the corre-
sponding hourly emission factor (Table 2-2). The total annual emissions
have been calculated by multiplying the total hourly emissions for each
Model Unit by the number of hours in a year (8,760 hours/year). These
estimated annual emission rates appear in Table 2-3.
The average percent of total VOC emissions attributed to each component
type is presented in Table 2-4. The percent attributed to each component type
is the same for each model unit.
2-20
-------
TABLE 2-2. UNCONTROLLED FUGITIVE EMISSION FACTORS
IN PROCESS UNIT EQUIPMENT
Uncontrolled emission
Fugitive emission source factor,3 kg/hr
Pumps • ,
Light liquids
With packed seals 0.12
Wtih single mechanical seals ^.12
With double mechanical seals 0.12
With no seals 0.0
Heavy liquids
With packed seals 0.020
With single mechanical seals 0.020
With double mechanical seals 0.020
With no seals 0.0
Valves (in-line)
Gas . 0.021
Light liquid^ 0.010
Heavy liquid0 0.0003
Safety/relief valves
Gas . 0.16
Light liquid: 0.006
Heavy liquid 0.009
Open-ended valves
Gas . 0.025
Light liquid^ 0.014
Heavy liquid0 0.003
Flanges 0.0003
Sampling connections 0.015
Compressors 0.44
Cooling towers 13.6-1107
Agitators NAf
aThese uncontrolled emission levels are based upon the refinery data presented
in reference 39-
bLight liquid is defined as a fluid with vapor pressure greater than 0.3 kPa
at 20°C. This vapor pressure represents the split between kerosene and naphtha
and is based on data presented in reference 39.
°Assumes the inner seal leaks at the same rate as single seal and that the VOC
is emitted from the seal oil degassing vent.
Heavy liquid is defined as a fluid with vapor pressure less than 0.3 kPa at 20°C.
This vapor pressure represents the split between kerosene and naphtha and is
based on data presented in reference 40.
eThese levels are based on cooling tower circulation rates that range from
0.05-3.66 m3/sec (714-58,000 GPM). Reference 46-
NA = no data available.
2-21
-------
TABLE 2-3. ESTIMATED TOTAL FUGITIVE EMISSIONS FROM MODEL UNITS3
Model unit Model unit Model unit
ABC
Estimated total
emissions (Mg/yr) 67 260 800
aBased upon equipment component counts in Table 2-1, the uncontrolled
emission factors in Table 2-2, and 8,760 hours/yr.
2-22
-------
TABLE 2-4. AVERAGE PERCENT OF TOTAL FUGITIVE EMISSIONS ATTRIBUTED
TO SPECIFIC COMPONENT TYPES
Percent of
total uncontrolled
emissions attributed to
to component type
Component for model units A,B,C
Pump seals
Light liquid service
Heavy liquid service
In-line valves
Gas service
Light liquid service
Heavy liquid service
Safety/relief valves
Gas service
Light liquid service
Heavy liquid service
Open-ended valves
Gas service
Light liquid service
Heavy liquid service
Compressor seals
Sampling connections
Flanges
12
2
26
11.
23
3
9
2
4
5
2
Less than one percent.
2-23
-------
2.4 REFERENCES
1. Erikson, D.G., and V. Kalcevic. Emissions Control Options for the
Synthetic Organic Chemicals Manufacturing Industry, Fugitive Emissions
Report, Draft Final. Hydroscience, Inc., 1979. p. II-2.
2. Ref. 1.
3. Ref. 1, p. II-3.
4. Ramsden, J.H. How to Choose and Install Mechanical Seals. Chem. E.,
85(22):97-102. 1978.
5. Ref. 1, p. II-3.
6. Ref. 4, p. 99.
7. Ref. 4, p. 99.
8. Ref. 4, p. 99.
9. Ref. 4, p. 99.
10. Perry, R.H., and C.H. Chilton, Chemical Engineers' Handbook, 5th Ed.
New York, McGraw-Hill Book Company, 1973. p. 6-8.
11. Ref. 10, p. 6-13.
12. Nurken, R.F. Pump Selection for the Chemical Process Industries, Chem.
E., Feb. 18, 1974. p. 120.
13. Nelson, W.E. Compressor Seal Fundamentals. Hydrocarbon Processing,
56(12):91-95. 1977.
14. American Petroleum Institute, "Centrifugal Compressors for General Refinery
Service". API Standard 617, Fourth Edition, November, 1979, p. 8.
Reprinted by courtesy of the American Petroleum Institute.
15. Reference 14, p. 9.
16. Ref. 13.
17. Ref. 13.
18. Ref. 1, p. 11-7.
19. Lyons, J.L., and C.L. Ashland, Jr. Lyons' Encyclopedia of Valves. New
York, Van Nostrand Reinhold Co., 1975. 290p.
20. Templeton, H.C. Valve Installation, Operation and Maintenance. Chem.
E., 78(23)141-149, 1971.
2-24
-------
References (continued)
21. Ref. 1, p. II-5.
22. Ref. 18, p. 147-148.
23. Ref. 18, p. 148.
24. Ref. 18, p. 148.
25. Ref. 18, p. 148.
26. Pikulik, A. Manually Operated Valves. Chem. E., April 3, 1978.
p. 121.
27. Ref. 24, p. 121.
28. Steigerwald, B.J. Emissions of Hydrocarbons to the Atmosphere from Seals
on Pumps and Compressors. Report No. 6, PB 216 582, Joint District,
Federal and State Project for the Evaluation of Refinery Emissions. Air
Pollution Control District, County of Los Angeles, California. April 1958.
37 p.
29. Ref. 1, p. II-7.
30. Cooling Tower Fundamentals and Application Principles. Kansas City,
Missouri, The Marley Company, 1969. p. 4.
31. Ramsey, W.D. and G.C. Zoller. How the Design of Shafts, Seals and
Impeller Affects Agitator Performance. Chem. E., 83(18): 101-108.
1976.
32. Ref. 29, p. 105.
33. Ref. 29, p. 105.
34. Ref. 29, p. 105.
35. Ref. 29, p. 106.
36. Ref. 29, p. 106.
37. Ref. 29, p. 106.
38. McFarland, I. Preventing Flange Fires. Chem. E. Prog., 65_(8): 59-61.
1969.
39. Wetherold, R.G., et al. Emission Factors and Frequency of Leak Occurrence
for Fittings in Refinery Process Units, interim report. EPA Contract
No. 68-02-2665. Austin, Texas, Radian Corporation, February 1979.
pp. 11-49.
2-25
-------
References (continued)
40. Radian Corporation. The Assessment of Environmental Emissions From Oil
Refining. Draft Report, Appendix B. EPA Contract No. 68-02-2147,
Exhibit B. Austin, Texas. August, 1979. pp. 3-5 through 3-16.
41. Ref. 1, pp. IV-1, 2.
42. Ref. 1, p. II-9-13.
43. Ref. 37,p p. 11-23.
44. Ref. 1, p. 11-10.
45. Ref. 1, p. IV-8.
46. Letter with Attachments from J.M. Johnson, Exxon Company, U.S.A., to
Robert T. Walsh, U.S. EPA. July 28, 1977.
2-26
-------
3.0 EMISSION CONTROL TECHNIQUES
Sources of fugitive VOC emissions from process unit equipment were
identified in Chapter 2. This chapter discusses the emission control
technique which is reasonably available control technology (RACT) for these
sources. The estimated control effectiveness of the technique is also
presented. Qualitative discussions of effectiveness and references to
technology transfer from similar industries are presented wherever applicable.
3.1 LEAK DETECTION AND REPAIR METHODS
Leak detection and repair methods can be applied in order to reduce
fugitive emissions from process unit sources. Leak detection methods are
used to identify equipment components that are emitting significant amounts
of VOC. Emissions from leaking sources may be reduced by three general
methods: repair, modification, or replacement of the source.
3.1.1. Individual Component Survey.
Each fugitive emission source (pump, valve, compressor, etc.) is
checked for VOC leakage in an individual component survey. The source may
be checked for leakage by visual, audible, olfactory, or instrument techniques.
Visual methods are good for locating liquid leaks, especially pump seal
failures. High pressure leaks may be detected by hearing the escaping
vapors, and leaks of odorous materials may be detected by smell. Predominant
industry practices are leak detection by visual, audible, and olfactory
methods. However, in many instances, even very large VOC leaks are not
detected by these methods.
Portable hydrocarbon detection instruments are the best method for
identifying leaks of VOC from equipment components. The instrument is used
to sample and analyze the air in close proximity to the potential leak
surface by traversing the sampling probe tip over the entire area where
leaks may occur. This sampling traverse is called "monitoring" in subsequent
3-1
-------
descriptions. The VOC concentration of the sampled air is displayed on the
instrument meter. The performance criteria for monitoring instruments and a
description of instrument survey methods are included in Appendix C.
The VOC concentration at which maintenance is required is called the
"action level". The RACT action level is 10,000 ppmv. Components which
have indicated concentrations higher than this "action level" are marked for
repair. Emission data indicate that large variations in mass emission rate
may occur over short time periods for an individual equipment component.
3.1.2 Repair Methods
The following descriptions of repair methods include only those features
of each fugitive emission source (pump, valve, etc.) which need to be
considered in assessing the applicability and effectiveness of each method.
They are not intended to be complete repair procedures.
3.1.2.1 Pumps. Many process units have spare pumps which can be
operated while the leaking pump is being repaired. Leaks from packed seals
may be reduced by tightening the packing gland. At some point, the packing
may deteriorate to the point where further tightening would have no effect
or possibly even increase fugitive emissions from the seal. The packing can
be replaced with the pump out of service. When mechanical seals are utilized,
the pump must be dismantled so the leaking seal can be repaired or replaced.
Dismantling pumps may result in spillage of some process fluid causing
emissions of VOC. These temporary emissions have the potential of being
greater than the continued leak from the seal. Therefore the pump should be
flushed of VOC as much as possible before opening for seal replacement.
3.1.2.2 Compressors. Leaks from packed seals may be reduced by the
same repair procedure that was described for pumps. Other types of seals
require that the compressor be out of service for repair. Since most
compressors do not have spares, repair or replacement of the seal would
require a shutdown of the process. If the leak is small, temporary emissions
resulting from a shutdown may be greater than the emissions from the leaking
seal.
3-2
-------
3.1.2.3 Relief Valves. In general, relief valves which leak must be
removed in order to repair the leak. In some cases of improper reseating,
manual release of the valve may improve the seat seal. In order to remove
the relief valve without shutting down the process, a block valve may be
required upstream of the relief valve. A spare relief valve should be attached
i
while the faulty valve is repaired and tested. After a relief valve has been
repaired and replaced, it is possible that the next over-pressure relief will
result in another leak.
3.1.2.4 Valves. Most valves have a packing gland which can be tightened
while in service. Although this procedure should decrease the emissions from
the valve, in some cases it may actually increase the emission rate if the
packing is old and brittle or has been overtightened. Plug-type valves can
be lubricated with grease to reduce emissions around the plug. Some types of
valves have no means of in-service repair and must be isolated from the
process and removed for repair or replacement. Other valves, such as control
valves, may be excluded from in-service repair by operating procedures or
safety procedures. In many cases, valves cannot be isolated from the process
for removal. Most control valves have a manual bypass loop which allows them
to be isolated easily although temporary changes in process operation may
allow isolation in some cases. If a process unit must be shut down in order
to isolate a leaking valve, the emissions resulting from the shutdown might be
greater than the emissions from the valve if allowed to leak until the next
process change which permits isolation for repair.
Depending on site specific factors, it may be possible to repair process
valves by injection of a sealing fluid into the source. Injection of sealing
fluid has been successfully used to repair leaks from valves in petroleum
3
refineries in California.
3.1.2.5 Flanges. In some cases, leaks from flanges can be reduced by
replacing the flange gaskets. Most flanges cannot be isolated to permit
replacement of the gasket. Data from petroleum refineries show that flanges
4
emit very small amounts of VOC.
3-3
-------
3.1.3 Control Effectiveness of Leak Detection and Repair Methods
There are several factors which determine the control effectiveness of
a leak detection and repair program; these include:
• Action level (leak definition),
, • Inspection interval (monitoring frequency),
'< • Achievable emission reduction of maintenance, and
• Interval between detection and repair of the leak.
Some of these factors can be estimated by using data collected from petroleum
5
refineries.
3.1.3.1 Action Level. The action level is the VOC concentration
observed during monitoring which defines a leaking component which requires
repair. The choice of the action level for defining a leak is influenced by
a number of important considerations. First, the percent of total mass
emissions which can potentially be controlled by the monitoring and repair
program can be affected by varying the leak definition, or action level.
Table 3-1 gives the percent of total mass emissions affected by the 10,000 ppmv
action level for a number of equipment types. The choice of an appropriate
leak definition is most importantly limited by the ability to repair leaking
components. Test data indicate that about 50 percent of valve leaks with
initial screening values equal to or greater than 10,000 ppmv can be successfully
repaired. Similar data indicate that attempted repair of valve leaks with
initial screening values of less than 10,000 ppmv can increase instead of
decrease emissions from these valves. From these data it is concluded that
repairing leaks with screening values in the 1,000-10,000 ppmv range may not
result in a net reduction in mass emissions.
3.1.3.2 Inspection Interval. The length of time between inspections
should depend on the expected occurrence and recurrence of leaks after a
piece of equipment has been checked or repaired. The choice of the interval
affects the emission reduction achievable since more frequent inspection
will result in leaking sources being found and fixed sooner. In order to
evaluate the effectiveness of the quarterly monitoring interval which is
RACT, it is necessary to estimate the rate at which new leaks will occur and
repaired leaks will recur. The estimates which have been used to evaluate
quarterly monitoring are shown in Table 3-2.
3-4
-------
TABLE 3-1. PERCENTAGE OF EMISSIONS AS A FUNCTION OF ACTION LEVEL3
Fraction of mass
emissions for 10,000 ppmv
action level (as %)
Source type
Pump seals
Light liquid service 87
Heavy liquid service 21
Valves
Vapor service 98
Light liquid service 84
Heavy liquid service 0
Safety/relief valves 69
Compressor seals 84
Flanges 0
aThese data show the fraction of the total emissions from a given source
type that is attributable to sources with leaks above the various action
levels. Reference 4.
Level of emission at which repair of the source is required.
3-5
-------
TABLE 3-2. ESTIMATED OCCURRENCE AND RECURRENCE RATE OF LEAKS FOR A
QUARTERLY MONITORING INTERVAL
Component
type
Estimated Percent of sources
percent of leaking at quarterly
components inspection from leak
leaking occurrence, recurrence,
initially and leaks not repairedb>c
Pump seals
Light liquid service
Heavy liquid service
23
2
2.3
0.2
Valves
Vapor service
Light liquid service
Heavy liquid service
10
12
0
1.0
1.2
0.0
Safety/relief valves
Compressor seals
8
33
0.8
3.3
Flanges
0
0.0
Approximate fraction of components with a concentration greater than or
equal to 10,000 ppmv prior to repair.'
Estimated that 10 percent of initial leaks are found leaking at subsequent
quarterly inspections. This estimate is based on engineering judgement.
r+
"Estimated percent of components found leaking at subsequent inspections.
3-6
-------
3.1.3.3 Allowable Interval Before Repair. If a leak is detected, the
equipment should be repaired within a certain time period. The allowable
repair time should reflect an interest in eliminating a source of VOC
emissions but should also allow the plant operator sufficient time to obtain
necessary repair parts and maintain some degree of flexibility in overall
plant maintenance scheduling. The determination of this allowable repair
time will affect emission reductions by influencing the length of time that
leaking sources are allowed to continue to emit pollutants. Some of the
components with concentrations in excess of leak definition action level may
not be able to be repaired until the next scheduled unit shutdown.
The allowable interval before repair for RACT is fifteen days. The
percent of emissions from a component which would be affected by the repair
interval if all other contributing factors were 100 percent efficient is
97.9 percent. The emissions which occur between the time the leak is detected
and repair is attempted are increased with longer allowable repair intervals.
3.1.3.4 Achievable Emission Reduction. Repair of leaking components
will not always result in complete emission reduction. To estimate the
emission reduction from repair of equipment it was assumed that leaks are
reduced by maintenance to a level equivalent to a concentration reading of
1,000 ppmv. The average emission rates of components above 10,000 ppmv and at
1,000 ppmv are shown in Table 3-3.
3.1.3.5 Development of Controlled Emission Factors. The uncontrolled
emission levels for the emission sources that are typically found in the model
plants were previously presented in Chapter 2 (Table 2-2). Controlled VOC
emission levels can be calculated by a "controlled emission" factor. This
factor can be developed for each type of emission source by using the general
expression:
Controlled emission factor = Uncontrolled factor - uncontrolled
factor x emission reduction efficiency
3-7
-------
TABLE 3-3. AVERAGE EMISSION RATES FROM SOURCES ABOVE
10,000 PPMV and at 1000 PPMV9
Source type
Pump seals
Light liquid service
Heavy liquid service
In-line valves
Vapor service
Light liquid service
Heavy liquid service
Safety/relief valves
Compressor seals
Flanges
(Y)
Emission rate
from sources above
10,000 ppmv
(kg/hr)
0.45
0.21
0.21
0.07
0.005
1.4
1.1
0.003
(X)
Emission rate
from sources at
1000 ppmvb
(kg/hr)
0.035
0.035
0.001
0.004
0.004
0.035
0.035
0.002
/Y-X\ t-inn\
( Y M 1UU/
Percentage
reduction
92.0
83.0
99.5
94.0
20.0
97.5
97.0
33.0
Average emission rate of all sources, within a source type, having screening
values above 10,000 ppmv.
Emission rate of all sources, within a source type, having screening values.
of 1000 ppmv.
3-8
-------
The reduction efficiency can be developed by the following expressions and
correction factors:
g
Reduction efficiency = AxBxCxD
Where:
A = Theoretical Maximum Control Efficiency = fraction of total mass
emissions for each source type with VOC concentrations greater
than the action level (Table 3-1, Figure 3-1).
B = Leak Occurrence and Recurrence Correction Factor = correction
factor to account for sources which start to leak between inspections
(occurrence) and for sources which are found to be leaking, are
repaired and start to leak again before the next inspection (recurrence)
(Table 3-2, 3-4).
C = Non-Instantaneous Repair Correction Factor = correction factor to
account for emissions which occur between detection of a leak and
subsequent repair; that is, repair is not instantaneous.
D = Imperfect Repair Correction Factor = correction factor to account
for the fact that some sources which are repaired are not reduced
to zero emission levels. For computational purposes, all sources
which are repaired are assumed to be reduced to a 1000 ppmv
emission level equivalent to a concentration of 1000 ppmv (Table 3-3).
These correction factors can, in turn, be determined from the following
expressions:
(l)B-l- 2N
„ 365 - t
(3) D = 1 -
Where:
n = Total number of leaks occurring and recurring over the monitoring
interval.
N = Total number of sources at or above the action level (Figure 3-2),
3-9
-------
Upper limit of 90 percent
confidence interval
Estimated percent of
total mass emissions
Lower limit of 90 percent
confidence interval
percent of total mass
emissions -
indicates the percent of
total emissions attribu-
table to sources with
screening values greater
than the selected value
10
SCREENING VALUE ( ppmv ) ( LOG 1Q SCALE )
Figure 3-1. Cumulative distribution of total emissions by screening
values - valves on gas streams.
100
90
80
M
UJ
O 70
§ 60
O 50
K
u 40
O
£ 30
20
10
0
'Upper limit of 95 percent
confidence interval
Estimated
percent of sources
percent of sources -
indicates the percent of
sources with screening
values greater than the
selected source
Lower limit of the 95 percent
confidence interval
10°
SCREENING VALUE ( ppmv ) ( LOG 1Q SCALE )
Figure 3-2. Cumulative distribution of sources by screening
values - valves on gas streams1.1
3-10
-------
TABLE 3-4. IMPACT OF MONITORING INTERVAL ON CORRECTION FACTOR ACCOUNTING
FOR LEAK OCCURRENCE/RECURRENCE (FOR EXAMPLE CALCULATION)
Monitoring a — b
interval m m
3 months 0.2Nd 0.1N 0.90
a n = Total number of leaks which occur, recur, and remain between
monitoring intervals.
n = Half the number of leaks which occur, recur, and remain between
monitoring intervals. This represents the emissions expected
from the total number over the monitoring period.
c B = Correction factor accounting for leak occurrence and recurrence.
N = Total number of components at or above the action level.
3-11
-------
t = Average time before repairs are made (with a 15-day repair limit,
7.5 is the average used).
f = Average emission factor for sources at the average screening value
achieved by repair.
F = Average emission factor for all sources at or above the action level.
An example of a control effectiveness calculation is presented in Table 3-5.
Support data for this calculation are presented in Tables 3-1, 3-2, 3-3,
and 3-4, as well as in Figures 3-1 and.3-2.
3.2 EQUIPMENT SPECIFICATION FOR OPEN-ENDED VALVES
Fugitive VOC emissions can be reduced by installing equipment which
will reduce leakage. The RACT equipment specification for open-ended valves
is presented in this section.
3.2.1 Open-Ended Valves
Fugitive emissions from open-ended valves are caused by leakage through
the seat of the valve. Emissions may also occur through the stem and gland
of the valve, and these emissions may be controlled by the methods described
for valves in Section 3.1.2. Approximately 28 percent of valves (excluding
safety/relief and check valves) in VOC service are open-ended. They include
drain, purge, sample, and vent valves. Fugitive emissions from open-ended
valves can be controlled by installing a cap, plug, flange, or second valve.
to the open end of the valve. In the case of a second valve, the upstream
valve should always be closed first after use of the valves. Each time the
cap, plug, flange, or second valve is opened, any VOC which has leaked
through the first valve seat will be released. These emissions have not
been quantified. The control efficiency will be dependent on the frequency
of removal of the cap or plug. Caps, plugs, etc. for open-ended valves do
not affect emissions which may occur during use of the valve. These emissions
may be caused by line purging for sampling, draining or venting through the
open-ended valve.
3-12
-------
TABLE 3-5. EXAMPLE OF CONTROL EFFICIENCY CALCULATION
Assume:
1) A leak detection and repair program to reduce emissions from
valves in gas/vapor source.
2) Action level = 10,000 ppmv.
3) Average screening value after directed repair = 1,000 ppmv.
4) Leak detection monitoring interval = 3 months.
5) Allowable repair interval = 15 days.
6) Number of valves having new or recurring leaks between repair
intervals, n = 0.2N (see Table 3-4).
Calculations:
A = 0.98 (from Figure 3-1 for a screening value of 10,000 ppmv)
B = 0.9 (from Table 3-4)
C = 0.979 (for 15-day interval)
where:
F = A(Avg. uncontrolled emission factor)9 .
Fraction of sources screening 1 10,000 ppmv
= (0.98)(0.021 kg/hr)/0.10 = 0.206 kg/hr
f = Emission factor at 1,000 ppmv3
= 0.001 kg/hr
and D = (1 - ^°jl) = 0.995
Overall percentage reduction =AxBxCxD
= (0.98) x (0.9) x (0.979) x (0.995)
= 86 percent
Therefore:
Control effectiveness factor = 0.021 kg/hr - (0.86)(0.021 kg/hr)
= 0.003 kg/hr
Reference 4.
From Figure 3-2.
3-13
-------
3.3 REFERENCES
1. Hustvedt, K.C., and R.C. Weber, Detection of Volatile Organic Compound
Emissions from Equipment Leaks.. Presented at 71st Annual Air Pollution
Control Association Meeting, Houston, Texas, June 25-30, 1978.
2. Ref. 1.
3. Teller, James H. Advantages Found in On-Line Leak Sealing.
Journal, 77 (29):54-59, 1979.
Oil and Gas
4. Wetherold, R.G., and L.P. Provost. Emission Factors and Frequency of
Leak Occurance for Fittings in Refinery Process Units. Interim Report.
EPA/600/2-79-044. Radian Corporation. February 1979. p. 2.
5. Ref. 4.
6. Valve Repair Summary and Memo from F.R. Bottomley, Union Oil Company.
Rodeo, California. To Milton Feldstein, Bay Area Quality Management
District, April 10, 1979.
7. Ref. 4.
8. Ref. 4.
9. Tichenor, B.A., K.C. Hustvedt, and R.C. Weber. Controlling Petroleum
Refinery Fugitive Emissions Via Leak Detection and Repairs in Proceedings:
Symposium on Atmospheric Emissions from Petroleum Refineries (November 1979,
Austin, Texas). EPA-600/9-80-013. Radian Corporation. March 1980.
pp. 421-440.
10. Ref. 4
11. Ref. 4.
3-14
-------
4.0 ENVIRONMENTAL ANALYSIS OF RACT
The environmental impacts that would result from implementing reasonably
available control technology (RACT) are examined in this chapter. Reasonably
available control technology is weekly visual inspection of pumps in light
liquid service, capping of open-ended valves and quarterly monitoring of the
following components: pumps in light liquid service, valves in gas and
light liquid service, compressors, and safety/relief valves in gas service.
Included in this chapter are estimates of VOC fugitive emissions before and
after implementation of RACT. The percent reduction of VOC emissions achievable
with RACT is estimated. The beneficial and adverse environmental impacts of
RACT on air pollution, water pollution, solid waste generation, and energy
consumption are also discussed.
4.1 AIR POLLUTION
Implementation of RACT would reduce VOC fugitive emissions from process
units. A significant beneficial impact on air pollution emissions would
result. The hourly and annual emissions from each model unit before and
after control by RACT are presented (Tables 4-1 and 4-2). There are no
adverse air pollution impacts associated with RACT.
4.1.1 Development of VOC Emission Levels
The: uncontrolled emission factors for process unit equipment were
previously presented in Chapter 2 (Table 2-2). Emission factors were developed
for those sources that would be controlled by the implementation of RACT.
These controlled fugitive emission levels were calculated by multiplying the
uncontrolled emissions from this equipment by a control efficiency. The
control efficiency is determined by several factors which are described and
presented in Chapter 3. The controlled VOC emission factors for each source
are presented in Table 4-3.
4-1
-------
TABLE 4-1. ESTIMATED HOURLY EMISSIONS AND EMISSIONS REDUCTION
ON A MODEL UNIT BASIS.
Estimated emissions
(kg/hr) Average percent
Level of Model unit reduction from
control ABC uncontrolled level
Uncontrolled 7.7 29.3 91.2
RACT 2.59 10.2 31.6 66
TABLE 4-2. ESTIMATED ANNUAL EMISSIONS AND EMISSIONS REDUCTION
ON A MODEL UNIT BASIS.
Estimated emissions
(Mg/yr) Average percent
)del uni'1
Level of Model unit reduction from
control ABC uncontrolled level
Uncontrolled 67 260 800
RACT 23 89 277 66
4-2
-------
TABLE 4-3. EMISSION FACTORS FOR SOURCES CONTROLLED UNDER RACT
-p*
1
CO
Uncontrolled
emission source
Pumps
Light liquid service
Valves
Gas service
Light liquid service
Safety/relief valves
Gas service
Compressors
Uncontrolled'
emission
Inspection factor,
interval kg/hr
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
0
0
0
0
0
.120
.021
.010
.160
.440
i
Ab
0.87
0.98
0,84
0.69
0.84
Correction
factors
BC
0.90
0.90
0.90
0.90
0.90
cd
0.98
0.98
0.98
0.98
0.98
De
0.92
0.99
0.94
0.97
0.97
Control
efficiency
(AxBxCxD)
0.70
0.86
0.70
0.59
0.72
Controlled
emission
factor,
kg/hr
0
0
0
0
0
.036
.003
.003
.066
.123
aFrom Table 2-2.
Theoretical maximum control efficiency. Reference 1.
°Leak occurrence and recurrence correction factor. Reference 2.
Non-instantaneous repair correction factor - for 15-day maximum allowable repair time, the correction
factor is [365 - (15/2)] v 365. Reference 2.
elmperfect repair correction factor, calculated as 1 - (f * F), where f = average emission rate for sources
at 1000 ppmv and F = average emission rate for sources greater than 10,000 ppmv. References 1, 2.
Controlled emission factor = uncontrolled emission factor x [1 - (AxBxCxD)].
-------
In calculating the total fugitive emissions from model units controlled
under RACT, the uncontrolled and controlled emission factors were used.
These emission factors were multiplied by the equipment source inventories
for each model unit. An example calculation for estimating emissions from
model unit A under RACT is shown in Table 4-4.
4.1.2 VOC Emission Reduction
The emission reduction expected from the implementation of RACT can be
determined for each model unit. The emission reduction is the difference
between the amount of fugitive emissions before RACT is implemented and the
amount of fugitive emissions after RACT is implemented. These amounts are
presented in Tables 4-1 and 4-2. The reduction in emissions for the model
units after RACT is implemented is 66 percent.
4.2 WATER POLLUTION
Implementation of RACT would result in no adverse water pollution
impacts because no wastewater is involved in monitoring and leak repair.
Some liquid chemicals may already be leaking and entering the wastewater
system as runoff. A beneficial impact on wastewater would result from
implementation of RACT since liquid leaks are found and repaired. This
impact, however, cannot be quantified because no applicable data on liquid
leaks are available.
4.3 SOLID WASTE DISPOSAL
The quantity of solid waste generated by the implementation of RACT
would be insignificant. The solid waste generated would consist of used
valve packings and components which are replaced.
4.4 ENERGY
The implementation of RACT calls for an emission control technique that
requires no additional energy consumption for any of the model unit sizes.
A beneficial impact would be experienced by saving VOC which has been heated,
compressed, or pumped.
4-4
-------
TABLE 4-4. EXAMPLE CALCULATION OF VOC FUGITIVE EMISSIONS FROM
MODEL UNIT A UNDER RACT
Number of
sources in.
model unit0
(N)
Emission^ Emissions
from sources,
kg/hr
(N x E)
factor,
kg/hr-source
(E)
Emission Source:
Pumps .
Light liquid single
mechanicaljseal
Light liquid double
mechanical seal
Heavy liquid single
mechanical seal
Heavy liquid packed seal
In-line valves
Gas service .
Light liquid service
Heavy liquid service
Safety/relief valves
Gas service .
Light liquid service
Heavy liquid service
Open-ended valves
Gas service ,
Light liquid service
Heavy liquid service
Compressors
Sampling connections
Flanges
5
3
5
2
90
84
84
11
1
1
9
47
48
1
26
600
Total
0.036
0.036
0.020
0.020
0.003
0.003
0.0003
0.066
0.006
0.009
0.003
0.003
0.0003
0.123
0.015
0.0003
emissions
0.180
0.108
0.100
0.040
0.270
0.252
0.025
0.726 .
0.006
0.009
0.027
0.141
0.014
0.123
0.390
0.180
2.591
aFrom Table 2-1.
•RACT emission factors include uncontrolled factors from Table 2-2 and controlled
factors from Table 4-3.
°Sources in VOC service.
Light liquid is defined as having a vapor pressure equal to or greater than
0.3 kPa at 20°C.
eHeavy liquid is defined as having a vapor pressure less than 0.3 kPa at 20°C.
Open-ended valve factor is equivalent to the in-line valve factor because of
capping the open end.
4-5
-------
4.5 REFERENCES
1. Wetherold, R. and L. Provost, Emission Factors and Frequency of Leak
Occurrence for Fittings in Refinery Process Units. EPA-600/2-79-044,
February 1979.
2. Tichenor, B.A., K.C. Hustvedt, and R.C. Weber. Controlling Petroleum
Refinery Fugitive Emissions Via Leak Detection and Repair, Draft.
Symposium on Atmospheric Emissions from Petroleum Refineries, Austin,
Texas.
4-6
-------
5.0 CONTROL COST ANALYSIS OF RACT
The costs of implementing reasonably available control technology (RACT)
for controlling fugitive emissions of volatile organic compounds (VOC) from
process units are presented in this chapter. Capital costs, annualized costs,
and the cost effectiveness of RACT are presented. These costs have been
developed for the model units presented in Chapter 2. All costs presented
in this chapter have been updated to second quarter 1980 dollars.
5.1 BASIS FOR CAPITAL COSTS
Capital costs represent the total cost of starting a leak detection and
repair program in existing process units. The capital costs for the implemen-
tation of RACT include the purchase of VOC monitoring instruments, the
purchase and installation of caps for all open-ended lines, and initial leak
repair. The cost for initial leak repair is included as a capital cost
because it is expected to be greater than leak repair costs in subsequent
quarters and is a one-time cost.
The basis for these costs is discussed below and presented in Table 5-1.
Capital cost estimates for model units under RACT are presented in Table 5-2.
Labor costs were computed using a charge of $18 per labor-hour. This rate
includes wages plus 40 percent for related administrative and overhead
costs .
5.1.1 Cost of Monitoring Instrument
The cost of a VOC monitoring instrument includes the cost of two
instruments. One instrument is intended to be used as a standby spare. The
cost of $4600 for a portable organic vapor analyzer was obtained from a
2
manufacturer.
5-1
-------
TABLE 5-1. CAPITAL COST DATA
Cost value
used in analysis
Item (June 1980 dollars) Cost basis Reference
Monitoring instrument 2 x 4600 = 9200/model unit One instrument used as a spare
Caps for open-ended lines 53/1ine Based on cost for 1" screw-on .
type valve. Cost June 1980 = $35.
Installation = 1 hour at $18/hour.
TABLE 5-2. CAPITAL COST ESTIMATES FOR IMPLEMENTING RACT
(thousands of June 1980 dollars)
Capital cost item
Model Unit A
1. Monitoring instruments .
2. Caps for open-ended lines (104 caps)
3. Initial leak detection and repair cost0
Total
Model Unit B
1. Monitoring instruments .
2. Caps for open-ended lines (415 caps)
3. Initial leak detection and repair cost
Total
Model Unit C
1. Monitoring instruments .
2. Caps for open-ended lines (1277 caps)
3. Initial leak detection and repair costc
Total
Level
Uncontrolled
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
of control
RACT
9.2
5.5
' 4.48
19.18
9.2
22.0
14.35
45.55
9.2
67.7
40.34
117.24
aBased on capital cost data presented in Table 5-1.
bFrom Table 2-1.
Initial leak detection and repair are treated as capital costs because they are one-time cost.
5-2
-------
5.1.2 Caps for Open-Ended Lines
Fugitive emissions from open-ended lines and valves can be controlled by
installing a cap, plug, flange, or second valve to the open end. These
pieces of equipment are all included in the definition of a cap for an
open-ended line. The cost of a cap for an open-ended line is based on a
cost of $35 for a one-inch screw-on type globe valve. This cost was supplied
by a large distributor. A charge of $18 for one hour of labor is added to
$35 as the cost for installing one cap. Therefore, the total capital cost
for installing a cap on an open-ended line is $53.
5.1.3 Initial Leak Repair
The implementation of RACT will begin with an initial inspection which
will result in the discovery of leaking components. The number of initial
leaks is expected to be greater than the number found in subsequent inspec-
tions. Because initial leak repair is a one-time cost, it is treated as a
capital cost. The number of initial leaks was estimated by multiplying the
percentage of initial leaks per component type by the number of components
in the model unit under consideration. Fractions were rounded up to the
next highest integer. The repair time for fixing leaks is estimated to be
80 hours for a pump seal, 40 hours for a compressor seal, and 1.13 hours for
a valve. The repair time for fixing pump seals and compressor seals includes
the cost of a new seal. These requirements are presented in Table 5-3.
The initial repair cost was determined by taking the product of the number
of initial leaks, the repair time, and the hourly labor cost of $18.
5.2 BASIS FOR ANNUALIZED COSTS
Annualized costs represent the yearly cost of operating a leak detection
and repair program and the cost of recovering the initial capital investment.
This includes credits for product saved as the result of the control program.
The basis for the annualized costs is presented in Table 5-4.
5-3
-------
TABLE 5-3. LABOR-HOUR REQUIREMENTS FOR INITIAL LEAK
REPAIR UNDER RACT
Number of
components
per model
Source type
Pumps (light liquid)
Single mechanical seal
Double mechanical seal
Valves (in-line)
Gas
Light liquid
Safety/relief valvesd
(gas service)
Valves on open-ended lines
Gas
Light liquid
Compressor seals
TOTAL
A
5
3
90
84
11
9
47
1
unit
B
19
10
365
335
42
37
189
2
Estinated
number of
initial
leaksd
C
60
31
1117
1037
130
115
581
8
A
2
1
9
11
0
le
6e
1
B
5
3
37
41
C
4e
23e
1
C
14
8
112
125
0
12e
70e
3
Repair Labor-hours
time, required
hrs A
80b 160
80b 80
1.13° 10
1.13C 12
0 0
1.13C 1
1.13C 7
40b 40
310
B
400
240
42
46
0
5
26
40
799
C
1120
640
127
141
0
14
79
120
2133
Based on the percent of sources leaking at _> 10,000 ppm. From Table 3-2.
blncludes labor-hour equivalent cost of new seal. Reference 6.
Weighted average based on 75 percent of the leaks repaired on-line, requiring 0.17 hours per repair,
and on 25 percent of the leaks repaired off-line, requiring 4 hours per repair. Ref. 5, p. B-12.
It is assumed that these leaks are corrected by routine maintenance at no additional labor
requirements. Ref. 6.
p
The estimated number of initial leaks for open-ended valves is based on the same percentage of
sources used for in-line valves. This represents leaks occurring through the stem and gland of the
open-ended valve. Leaks through the valve seat are eliminated by adding caps.
5-4
-------
TABLE 5-4. BASIS FOR ANNUALIZED COST ESTIMATES
1. Capital recovery factor for
capital charges
• Caps on open-ended lines
• Monitoring instruments
2. Annual maintenance charges
• Caps on open-ended lines
• Monitoring instruments
3. Annual miscellaneous charges
(taxes, insurance, administration)
• Caps on open-ended lines
• Monitoring instruments
4. Labor charges
5. Administrative and support costs
for implementing RACT
6. Annualized charge for initial
leak repairs
0.163 x capital0
0.23 x capitalb
0.05 x.capital0
$3,000°
0.04 x capital;;
0.04 x capital
$18/hourf
0.04 x [monitoring + repair
labor)9
i (estimated number of leaking
components per model unit x
repair time) x SlS/hr"" x 1.49
x 0.163n
7. Recovery credits
$410/Mg
Ten year life, ten percent interest. From Ref. 5, pp. IV-3,4.
Six year life, ten percent interest. From Ref. 5, pp. IV-9,10.
cFrom Ref. 5, pp. IV-3,4.
Includes materials and labor for maintenance and calibration. Reference 6.
Cost index = 242.7 * 209.1 (Reference 7 and 8).
eFrom Ref. 5, pp. IV-3,4,9,10.
Includes wages plus 40 percent for labor-related administrative and overhead
costs. Cost (June 1980) from Ref. 1.
gFrom Ref. 5, pp. IV-9,10.
Initial leak repair amortized for ten years at ten percent interest.
•References 9,10,11.
5-5
-------
5.2.1 Monitoring Labor
The implementation of RACT requires visual and instrument monitoring of
potential sources of fugitive VOC emissions. The monitoring labor-hour
requirements for RACT are presented in Table 5-5. The labor-hour require-
ments were calculated by taking the product of the number of workers needed
to monitor a component (1 for visual, 2 for instrument), the time required
to monitor, the number of components in a model unit, and the number of
times the component is monitored each year. The monitoring times for the
various components are presented in Table 5-5. They are 0.5 minute for
visual inspection, 1 minute for in-line valves and open-ended valves,
5 minutes for pump seals, 8 minutes for safety valves, and 10 minutes for
4
compressor seals. Monitoring labor costs were calculated based on a
charge of $18 per hour.
5.2.2 Leak Repair Labor
Labor is needed to repair leaks which develop after initial repair.
The estimated number of leaks and the labor-hours required for repair are
given in Table 5-5. The repair time for each component is the same as
presented for initial leak repair. Leak repair costs were calculated based
on a charge of $18 per hour.
5.2.3 Maintenance Charges and Miscellaneous Costs
The annual maintenance charge for caps is estimated to be five percent
5
of their capital cost. The annual cost of materials and labor for maintenance
fi 7 ft
and calibration of monitoring instruments is estimated to be $3000. An
additional miscellaneous charge of four percent of capital cost for taxes,
insurance, and associated administrative costs is added for the monitoring
instruments and caps.
5.2.4 Administrative Costs
Administrative and support costs associated with the implementation of
RACT are estimated to be 40 percent of the sum of monitoring and leak
repair labor costs. The administrative and support costs include record-
keeping and reporting requirement costs.
5-6
-------
TABLE 5-5. ANNUAL MONITORING AND LEAK REPAIR LABOR REQUIREMENTS FOR RACT
I
—I
Source type
Pumps (light liquid)
Single mechanical
seals
Double mechanical
seals
Valves (in-line)
Gas
Light liquid
Safety/relief valves
(gas service)
Valves on open-ended
lines"
Gas
Light liquid
Compressors seals
TOTAL
Number of
components per
model unit
ABC
5 19 60
3 10 31
90 365 1117
84 335 1037
11 42 130
9 37 115
47 189 581
1 2 8
Type ofa
monitoring
Instrument
Visual
Instrument
Visual
Instrument
Instrument
Instrument
Instrument
Instrument
Instrument
Monitoring
Monitoring
time,b
mm
5
0.5
5
0.5
1
1
8
1
1
10
Times
monitored
per year
4
52
4
52
4
4
4
4
4
4 .
Monitoring
Estimated
labor- number of j
hours required0 leaks per year"
'A
3.3
2.2
2.0
1.3
12.0
11.2
11.7
1.2
6.4
1.3
52.6
B
12.8
8.2
6.8
4.3
49.0
44.8
44.8
4.9
25.2
2.7
203.5
C
40.
26.
20.
13.
149.
138.
139.
15.
77.
10.
630.
ABC
0125
0
8113
4
0 4 15 45
4 4 16 50
0
3125
6 2 9 28
7 .1 1 2
2
Leak repair
Repair Leak repair labor-
time, hours required6
hrs A
80b 80
80b 80
1.13f 4.5
1.13f 4.5
O9 0
1.136 1.1
1.13e 2.3
40b 40
212.4
B
160
80
17.0
18.0
0
2.3
10.2
40
327.5
C
400
240
50.9
56.5
0
5.7
31.6
80
864.7
Two workers for instrument monitoring, one for visual.
Includes labor-hour equivalent cost of new seal. Reference 6.
Monitoring labor-hours = number of workers x number of components x time to monitor (total is minimum of 1 hour) x number of times monitored per year.
dFrom Table 3-2.
eLeak repair labor-hours = number of leaks x repair time.
Weighted average based on 75 percent of the leaks repaired on-line, requiring 0.17 hour per repair, and on 25 percent of the leaks, repaired off-line,
requiring 4 hours per repair. Ref. 5, p. B-12.
g
It is assumed that these leaks are corrected by routine maintenance at no additional labor reouirements. Ref. 4
estimated number of leaks per year for open-ended valves is based on the same percent of sources used for in-line valves. This represents leaks
occurring through the stem and gland of the open-ended valve. Leaks through the seat of the valve are eliminated by adding caps.
-------
5.2.5 Capital Charges
The life of caps for open-ended lines is assumed to be ten years and
the life of monitoring instruments is assumed to be six years. The cost of
repairing initial leaks was amortized over a ten-year period since it is a
one-time cost.
The capital recovery is obtained from annualizing the installed capital
cost for control equipment. The installed capital cost is annualized by
using a capital recovery factor (CRF). The CRF is a function of the
interest rate and useful equipment lifetime. The capital recovery can be
estimated by multiplying the CRF by the total installed capital cost for the
control equipment. This equation for the capital recovery factor is:
where i = interest rate, expressed as a decimal
n = economic life of the equipment, years.
The interest rate used was ten percent (June 1980). The capital recovery
factors and other factors used to derive annualized charges are presented in
Table 5-4.
5.2.6 Recovery Credits
The reduction of VOC fugitive emissions results in saving a certain
amount of VOC which would otherwise be lost. The value of this VOC is a
recovery credit which can be counted against the cost of a leak detection and
repair program. The recovery credits for each model unit are presented in
Table 5-6. The VOC saved is valued in June 1980 dollars at $410/Mg.9'10j11
5.3 EMISSION CONTROL COSTS
This section will present and discuss the emission control costs
of implementing RACT for each of the three model units. Both the initial
costs and the annualized costs are included.
5-8
-------
TABLE 5-6. RECOVERY CREDITS
Model
unit
A
B
C
Uncontrolled
emissions,
Mg/yr
67
260
800
Emissions
under RACT,
Mg/yr
23
89
277
Emission
reduction,
Mg/yr
44
171
523
Recovered9
product
value,
$/yr
18,040
70,100
214,400
Based on an average price of $410/Mg. References 9,10,11.
5-9
-------
5.3.1 Initial Costs
The cost of initially implementing RACT consists of capital costs and
initial leak repair. The capital cost of $9200 for two monitoring instruments
is the same for all model unit sizes. Caps for open-ended lines will cost
$5500 for model unit A, $22,000 for model unit B, and $67,700 for model
unit C. The one-time initial leak repair cost is $4480 for model unit A,
$14,350 for model unit B, and $40,340 for model unit C. The total initial
capital costs for implementing RACT are $19,180 for model unit A, $45,550
for model unit B, and $117,240 for model unit C.
5.3.2. Recovery Credits
The value of VOC saved each year as a result of implementing RACT is
included as an annual credit against the net annualized costs. The implemen-
tation of RACT will result in saving $18,040 worth of VOC annually in model
unit A, $70,100 worth of VOC in model unit B, and $214,400 worth of VOC in
model unit C.
5.3.3 Net Annualized Cost
The net annual cost for controlling emissions is the difference between
the total annualized cost and the annual recovery credit for each model
unit. Net annualized control cost estimates for model units under RACT are
presented in Table 5-7. Capital cost data were previously presented in
Table 5-1.
For model unit A, the annualized capital charges are $3920 and the
total annual operating costs are $10,550. Produce recovery credits total
$18,040. The net annual ized cost for model unit A is a negative $3570,
which means that $3570 is actually gained every year by preventing loss of
VOC.
The annual ized capital charges for model unit B are $8050 and the total
annual operating costs are $18,730. The recovery credit is $70,100 per
year. The net annualized cost for model unit B is a negative $43,320, which
means that $43,320 is saved every year by controlling VOC emissions.
5-10
-------
TABLE 5-7. ANNUALIZED CONTROL COST ESTIMATES FOR MODEL UNITS
UNDER RACT (thousands of June 1980 dollars)
Cost item
Model unit
B
Annualized capital charges
1. Control equipment
a. Instrument
b. Caps
2. Initial leak repair
Subtotal
2.12
0.89
0.91
3.92
2.12
3.59
2.34
8.05
2.12
11.04
6.58
19.74
Operating costs
1. Maintenance charges
a. Instrument
b. Caps
2. Miscellaneous (taxes, insurance,
administration)
a. Instrument
b. Caps
3. Labor
a. Monitoring labor ,
b. Leak repair labor
c. Plant and payroll overhead
Subtotal
Total before credit
Recovery credits
Met annualized cost
3.0
0.275
0.37
0.22
0.95
3.82
1.91
10.545
14.47
18.0
(-3.57)
3.0
1.1
0.37
0.88
3.66
5.9
3.82
18.73
26.78
70.1
(-43.32)
3.0
3.39
0.37
2.71
9.9
15.56
10.2
45.13
64.87
214.4
(-149.53)
Sum of labor hours for monitoring in Table 5-5 multiplied by $18/hour.
Sum of labor hours for leak repairs in Table 5-5 multiplied by $18/hour.
p
Based on 40 percent of monitoring labor plus leak repair labor costs.
These costs are credits.
5-11
-------
Model unit C has annualized capital charges of $19,740 and total
operating expenses of $45,130. The recovery credit is $214,400 per year.
The net annualized cost for model unit C is a negative $149,530, which is an
annual savings as a consequence of controlling fugitive VOC emissions.
5..3.4 Differences in Net Annualized Costs
The cost for RACT is different for each model unit. The cost for caps
for open-ended lines varies because the number of open-ended lines is
different for each model unit. Because the larger model units have more
components, more labor-hours are needed for monitoring and leak repair. For
this reason, labor costs will increase as model unit size increases.
5.4. COST EFFECTIVENESS
Cost effectiveness is the annualized cost per megagram of VOC controlled
annually. The cost effectiveness of RACT for each model unit is the net
annualized cost for implementing RACT divided by the emission reduction
gained under RACT. The cost effectiveness of RACT is summarized in Table 5-8.
The implementation of RACT on model unit A results in a net annualized
cost which is a credit of $3570. The emission reduction associated with
RACT is 44 Mg/yr. The cost effectiveness is -$81/Mg.
The implementation of RACT in the case of model unit B results in a net
annualized cost which is a credit of $43,320. The emission reduction associated
with RACT is 171 Mg/yr. The cost effectivess is -$253/Mg.
The implementation of RACT in the case of model unit C results in a net
annualized cost which is a credit of $149,530. The emission reduction
associated with RACT is 523 Mg/yr. The cost effectiveness is -$286/Mg.
A comparison of the cost effectiveness of RACT for each model unit
reveals that cost effectiveness increases as model unit size increases. The
strong influence of recovery credits is responsible for the increase in cost
effectiveness.
5-12
-------
TABLE 5-8. COST EFFECTIVENESS FOR MODEL UNITS UNDER RACT
Annual i zed cost before credit ($1000)
Annual recovery credit ($1000)
Net annualized cost ($1000)a
Total VOC reduction (Mg/yr)
Cost effectiveness ($/Mg VOC)a
A
14.47
18.0
(3.53)
44.0
(81.0)
Model unit
B
26.78
70.1
(43.32)
171.0
(253.0)
C
64.87
214.4
(149.53)
523.0
(286.0)
(xxx) = net credit
5-13
-------
5.5. REFERENCES
1. Letter with attachments from Texas Chemical Council to Walt Barber,
U.S. EPA. June 30, 1980.
2. Purchase order from GCA/Technology Division to Analabs/Foxboro, North
Haven, Connecticut. July 3, 1980.
3. Telecon. Samuel Duletsky, GCA Corporation with Dave Myer, Piedmont
Hub, Greensboro, N.C. September 25, 1980. Price of 1" screw-on
type valve.
4. Letter with attachments from J.M. Johnson, Exxon Company, U.S.A., to
Robert T. Walsh, U.S. EPA. July 28, 1977.
5. Erikson, D.G., and V. Kalcevic. Emission Control Options for the
Synthetic Organic Chemicals Manufacturing Industry, Fugitive Emissions
Report, Draft Final. Hydroscience, Inc. 1979. p. IV-9.
6. Environmental Protection Agency. Control of Volatile Organic Compounds
Leaks from Petroleum Refinery Equipment. EPA-450/2-78-036, OAQPS
No. 1.2-111. June 1978.
7. Economic Indicators. Chem. Eng. Vol. 86 #2. January 15, 1979.
8. Economic Indicators. Chem. Eng. Vol. 87 #19. September 22, 1980.
9. Letter from Vincent Smith, Research Triangle Institute, to Russell
Honerkamp, Radian Corporation. November 30, 1979.
10. Reference 8.
11. Economic Indicators. Chem. Eng. Vol. 87 #1. January 14, 1980.
5-14
-------
6.0 MODEL REGULATION AND DISCUSSION
This chapter includes a model regulation based on the "presumptive norm"
which is considered broadly representative of RACT for the synthetic organic
chemical manufacturing industry and the polymer and resin manufacturing
industry. The model regulation is included solely as guidance to assist
state and local agencies in drafting their own specific RACT regulations.
Consequently, the model regulation is illustrative in nature and is not to be
construed as rulemaking by EPA.
6.1 MODEL REGULATION
iXX.010 Applicability
(A) This regulation applies to facilities that are in the following
areas:
(B) This regulation applies to components which contact a process
fluid that contains at least 10 percent volatile organic compounds
by weight in synthetic organic chemical manufacturing plants and
polymer and resin manufacturing plants. Synthetic organic chemical
manufacturing plants are facilities that produce, as intermediates
or final products, one or more of the chemicals listed in [Table I
of Appendix B]. Polymer and resin manufacturing plants are facilities
that produce one or more of the polymers and resins listed in
[Table II of Appendix B].
(C) Components specifically exempted from the requirements of this
[Regulation] are components in vacuum service, valves not externally
actuated, and pressure relief devices which are connected to an
operating flare header.
6-1
-------
(D) This regulation is not applicable to petroleum refinery units as
they are defined in "Control of Volatile Organic Compound Leaks
from Petroleum Refinery Equipment", EPA-450/2-78-036.
§XX.020 Definitions
(A) Except as otherwise required by the context, terms used in this
Regulation are defined in the [General Provisions, General Statutes],
or in this section as follows:
"Component" means a piece of equipment, including but not
limited to pumps, valves, compressors, and pressure relief valves,
which has the potential to leak volatile organic compounds.
"In Gas Service" means that the component contacts process
fluid that is in the gaseous state under operating conditions.
"In Light Liquid Service" means that the component contacts a
liquid with a concentration greater than 20 percent by weight of
volatile organic compounds having a vapor pressure greater than
0.3 kiloPascals at 20°C.
"Leak" means a volatile organic compound concentration greater
than or equal to 10,000 parts per million by volume (ppmv) as
shown by monitoring or dripping of process fluid.
"Leaking component" means any component which has a leak.
"Monitor" means to measure volatile organic compound concentration
by EPA Reference Method 21.
"Repair" means to reduce the volatile organic compound concentration
of a leaking component to below 10,000 ppmv as shown by monitoring.
"Unit turnaround" means unit shutdown and purge for internal
inspection and repair.
"Volatile Organic Compound" means any organic compound which
participates in atmospheric photochemical reactions or is measured
by the applicable test method or equivalent State method.
6-2
-------
§XX.030 Standards
(A) Each owner or operator shall monitor quarterly the following
components: each pump in light liquid service; each compressor;
each valve in gas and light liquid service; and each pressure
relief valve in gas service.
(B) Each owner or operator shall monitor:
(1) Each pressure relief valve within 24 hours after it has
vented to the atmosphere;
(2) Within 24 hours of discovery a component which sight, smell,
or sound indicates might be leaking.
(C) From the date a leaking component is detected, each owner or
operator shall:
(1) Affix within 1 hour a weatherproof and readily visible tag to
the component, bearing an identification number and the date.
This tag shall remain in place until the component is repaired;
(2) Repair the leaking component within 15 days; or
(3) Repair the leaking component at or before the next scheduled
unit turnaround if not able to do so within 15 days.
(D) Each owner or operator shall visually inspect all pumps in light
liquid service weekly for indications of leaks, and repair each
pump within 24 hours after visual inspection indicates it is
leaking.
(E) The Director may require early unit shutdown or turnaround based
on the number of leaking components which cannot be repaired while
the unit is operational.
(F) Except for pressure relief valves, an owner or operator shall seal
all open-ended valves which are in contact on one side of the seat
with process fluid and open to the atmosphere on the other side of
the seat either directly or through open piping. These open-ended
•valves shall be sealed with a second valve, a blind flange, a cap,
or a plug. The sealing device may be removed only when a sample
is being taken or during maintenance operations.
6-3
-------
(G) An alternative volatile organic compound emission control program
may be submitted after the requirements of §XX.030 (A) and (B) are
performed for four quarters. The owner or operator shall provide
any calculations, data, or other evidence which is necessary to
demonstrate equivalency.
§XX.040 Component Monitoring and Inspection
(A) Each owner or operator shall conduct monitoring as follows:
(1) The volatile organic compound detection instrument shall meet
the performance criteria of proposed EPA Reference Method 21.
(2) The instrument shall be calibrated before use on the day of
use by the methods specified in proposed EPA Reference Method
21.
(3) Calibration gases shall be a mixture of methane and air at a
concentration of approximately 10,000 ppmv methane.
§XX.050 Recordkeeping
(A) Each owner or operator shall record in an inspection log the
following information for each leaking component found:
(1) The tag number.
(2) The type of component.
(3) The date on which the leak was detected for the component.
(4) The date on which the leaking component was repaired.
(5) Identification of those leaking components which cannot be
repaired until unit turnaround and reason why repair must be
delayed.
(B) Each owner or operator shall record in a log the date of the
calibration of the monitoring instrument and the actual calibration
gas concentration.
6-4
-------
(C) Each owner or operator shall retain a copy of the inspection log
at the plant for a minimum of two years after the date on which
the report for the inspection period was prepared, and shall make
the inspection log available to an authorized representative of
the Director upon request.
§XX.060 Reporting
(A) The owner or operator shall submit to the Director quarterly, on
dates to be specified by the Director, a report that includes the
following information:
(1) The number and types of leaking components that were located
during the previous quarter but not repaired.
(2) The number and types of components inspected, the number and
types of leaking components found, and the number and types
of components repaired within the fifteen day period.
§XX.070 Compliance Schedule
(A) The owner or operator shall adhere to the increments of progress
contained in the following schedule:
(1) Submit a leak detection and repair program to the Director by
(date) . This program shall contain, as a minimum,
a list of the process units, a copy of the log book format,
and the make and model of the volatile organic compound
detection instrument to be used.
(2) Submit the first quarterly report to the Director by [the
date specified by the Director],
6.2 DISCUSSION
6.2.1 Introduction
The purpose of the regulation is to have owners and operators of plants
implement a leak detection and repair program. A plant is in violation of
the regulation if they are not trying to find leaks and repair them.
6-5
-------
Adequate enforcement of the regulation consists of plant inspections and
review of the quarterly reports submitted by the plants. The quarterly
reports provide the number of components that have not been repaired and the
number of components that should have been repaired. This information can
alert the Director to problems with the leak detection and repair program at
a plant. Inspections provide a more thorough check on plant compliance with
the regulation.
6.2.2 Applicability
The recommendations of this CTG document apply to components in VOC
service in the synthetic organic chemical, polymer, and resin manufacturing
industries (as listed in the tables in Appendix B). The CTG document,
"Control of Volatile Organic Compound Leaks from Petroleum Refinery Equip-
ment", EPA-450/2-78-036, is applicable to control of fugitive VOC emissions
from petroleum refinery units. Although these two documents apply to
different process units, the recommendations of both these documents are
similar.
6.2.3 State Inspections
States should make inspections as frequently as required for adequate
enforcement of the regulation, but at least annually. A review of the
quarterly reports will be one factor in determining the frequency of
inspections. An annual inspection should include a review of the inspection
log, a walkthrough inspection using a VOC detection instrument, and a check
of the VOC detection instrument used by the plant.
The inspection log should be checked for signs of non-compliance.
Information such as the number and type of unrepaired leaking components, the
number of components awaiting unit turnaround, and the time intervals between
leak discovery and leak repair should be checked. The log should be up-to-date
and contain the information required by the regulation.
6-6
-------
A walkthrough inspection of the plant should be made using a VOC
detection instrument. The inspector should monitor several components that
were leaking but have been repaired, and several components not identified
as leaking. Leaks may occur and recur randomly and spontaneously, so
detecting VOC concentrations above 10,000 ppmv does not indicate a violation
of the regulation. However, the discovery of an abnormally high percentage
of leaking components may indicate that the plant is not conducting an
effective leak detection and repair program.
In order to check a plant's leak detection and repair procedure, an
inspector should arrange to accompany the plant VOC leak detection personnel
on a day when routine leak detection is performed. The inspector can then
observe leak detection procedures first-hand.
6.2.4 Inspection Equipment
Inspection personnel should have access to a VOC detection instrument
for use during walkthrough inspections. The State may be able to arrange
for inspectors to use plant-owned VOC detection instruments or for a plant
employee trained in the use of the instrument to accompany the inspector and
perform monitoring. The State may need to purchase a VOC detection instrument.
In any event, inspectors should have a calibration gas (10,000 ppmv methane
in air) to calibrate their own VOC detection instrument or for checking the
accuracy of instruments owned by the plant.
6.2.5 Equivalency
The purpose cf the ecuivalency provision, §XX.030(G), is to allow a
plant to develop an equally effective leak detection and repair program
which is specific to the plant. Plants may be able to design an alternative
program which will result in less cost to the plant. For equivalency, a
plant must demonstrate that fugitive emissions expected from process equipment
leaks under an alternative leak detection and repair program are less than
or equal to fugitive emissions from leaks under the State regulation.
6-7
-------
A plant wishing to implement an alternative emission control program
could apply for a variance from the State regulation. Where the plant has
implemented the State regulation for four quarters and where the plant can
provide calculations or other data which indicate that the alternative
program will reduce emissions at least as much as the State regulation, the
State may allow a variance for a one-year trial period of the alternative
program.
To establish emissions under the State regulation, a plant should
implement the State regulation for at least one year in order to determine
the actual leak occurrence and recurrence rate for each of the various
component groupings (pumps, valves, etc.). Using this data, a plant can
estimate the percent of emissions affected by leak occurrence and recurrence
and new controlled emission rates under the State regulation for each component
grouping. The new controlled emission rates can be used to determine fugitive
emissions under the State regulation (See Appendix D).
Leak occurrence and recurrence data from the last two quarters should be
given the most weight in determining leak occurrence and recurrence rates
under the State regulation for the component groupings. The leaks detected
during the last two quarters are leaks which occur after leak detection and
repair efforts have been implemented for at least two quarters. This allows
for the high number of initial leaks accumulated prior to the initiation of
leak detection and repair program to be disregarded. The leaks detected
during the last two quarters, therefore, are representive of leak occurrence
and recurrence under the State regulation.
In applying for a variance from the State regulation to implement an
alternative leak detection and repair program, the plant should provide the
logic which indicates that fugitive emissions under the alternative program
should be equal to or less than the fugitive emissions under the State regu-
lation. In addition, the plant should outline what data or information would
be gathered during a trial period of the alternative program to estimate
fugitive emissions under the alternative program. This would have to be
based on various assumptions concerning the ability of the alternative leak
detection and repair program to reduce fugitive emissions. Little data,
6-8
-------
aside from that gathered from similar plants wiich might be implementing the
alternative program, are likely to be available concerning the actual ability
of the alternative leak detection and repair program to reduce fugitive
emissions. Consequently, if the assumptions seem reasonable and the alternative
program appears to have the potential to result in fugitive emissions that
are less than or equal to emissions under the State regulation, the State
should grant a one-year variance for a trial pariod of the alternative leak
'detection and repair program.
During the trial period, the plant should gather the data necessary to
substantiate or refute the validity of the assumptions made concerning the
ability of the alternative leak detection and repair program to reduce
fugitive emissions. For example, this might consist merely of comparing the
actual leak occurrence and recurrence rate for each component grouping under
the alternative program to the program required by the State regulation.
At the end of the trial period, the plant should estimate fugitive
emissions under the alternative leak detection and repair program. Each of
the assumptions originally made by the plant concerning the ability of the
alternative program to reduce fugitive emissions should be examined in light
of the data gathered during the trial period. The plant should explain how
this data substantiates or refutes each of these assumptions. Finally, the
plant should compare estimated fugitive emissions from the plant under the
alternative leak detection and repair program with estimated fugitive
emissions from the plant under the State regulation. This information should
then be submitted to the Director along with requests for a continuation of
the variance from the State regulation and a SIP revision permitting use of
the alternative program at the plant.
The Director should declare the plant's alternative leak detection and
repair program equivalent to the State regulation if estimated fugitive
emissions under the alternative program are no greater than those under the
State regulation and the Director feels the alternative program is equally
enforceable. If the alternative program does not result in fugitive emissions
equal to or less than emissions under the State regulation, then the Director
could require modifications to the alternative program and declare it to be
6-9
-------
equivalent, or the Director might choose instead to require the plant to
gather additional data during another trial period to demonstrate equivalence
of the modified alternative program. Finally, the Director could require the
plant to implement the State regulation for another year and then allow the
plant to submit another alternative program for consideration for a one-year
trial period as before.
Where an alternative is determined to be equivalent to the State
regulation, the plant's request for continuation of the variance from the
State regulation should be granted and activity initiated to submit a SIP
revision to EPA.
Plants may develop several types of alternative programs. Some possible
approaches are:
• Inspect components which have higher emission rates, more
frequently and inspect components w'th lower emission rates less
frequently. This approach might be practiced, for example, by
inspecting pressure relief valves and pump seals more frequently
than quarterly and inspecting other valves less frequently than
quarterly.
• Identify those components for which leaks recur more frequently and
inspect them more frequently; those components for which leaks
recur less frequently are inspected less frequently. For example, a
plant might identify ten percent of the valves in the plant as
being responsible for most leaks from valves for any inspection
interval. By inspecting those valves more frequently than quarterly
and inspecting all other valves less frequently than quarterly
emissions might be reduced more than if all valves were inspected
quarterly.
t Replace components with new components that will not leak or will
tend to leak less frequently. This might be practiced by installing
leakless valves and double mechanical seals in place of existing
components.
Up to this point, the discussion of equivalency has focused on alternative
work practices to those outlined in the model regulation. Some plants,
however, might prefer to comply with performance standards rather than work
6-10
-------
practices. Performance standards would not require a plant to implement
specific procedures to reduce fugitive emissions from leaking process components,
but would require a plant to reduce fugitive emissions from leaking process
components below some numerical emission limit.
The most straight forward and simple form of performance standards would
be specific limits on the number or percentage of leaking process components.
A regulation might state, for example, that no more than "x" percent of
valves or "y" percent of pumps could be leaking at any time. Thus, where
enforcement of a work practice regulation focuses on whether or not a plant
has implemented specific procedures to detect and repair process component
leaks, enforcement of a performance standard regulation focuses only on
whether or not a plant has more than a specific number of leaking process
components.
As with alternative work practices, to be equivalent to the model regulation
a plant must demonstrate that fugitive emissions from process component leaks
under a performance standard would be less than or equal to fugitive emissions
from leaks under the State regulation. Unlike alternative work practices,
however, this demonstration would be immediate and obvious.
A plant should implement the State regulation for at least one year to
determine the leak occurrence and recurrence rate for each of the various
process component groupings (pumps, valves, etc.). Analyzing the data from
the last two quarters, the plant should make an estimate of the percentage
or number of leaking process components under the State regulation. This
estimate should exclude those components that are awaiting repair at the
next unit turnaround. This analysis of the data and the estimate of the
number of leaking process components under the State regulation should then
be submitted to the Director, along with a request for a SIP revision,
permitting the plant to comply with a performance standard.
If the Director determines that the data are sufficient to support a
performance standard, he should declare this performance standard equivalent
to the State regulation and initiate activity to submit a SIP revision to
EPA. Alternatively, the Director could require the plant to implement the
6-11
-------
State regulation for an additional period of time to gather more leak
occurence and recurrence rate data.
The effectiveness of alternative work practices or performance
standards must be evaluated on a plant-by-plant basis. A plant might
request to implement a combination of these as well. These and other
alternatives might be more effective in reducing emissions and be less
costly than implementing the State regulation for a specific plant.
6-12
-------
APPENDIX A. EMISSION SOURCE TEST DATA
-------
APPENDIX A
EMISSION SOURCE TEST DATA
The purpose of Appendix A is to describe testing results used in the
development of the Control Techniques Guideline (CTG) document for VOC
fugitive emissions from the Synthetic Organic Chemicals Manufacturing Industry
(SOCMI) and the polymer and resin manufacturing industry. The information
in this appendix consists of a description of the tested facilities, and the
sampling procedures and test results of VOC fugitive emissions studies in
SOCMI and the petroleum refining industry.
Fugitive emission sources of VOC in SOCMI and in the petroleum refining
industry are similar. Considerable data exist concerning both the incidence
and magnitude of fugitive emissions from petroleum refineries. Studies of
fugitive emissions in SOCMI have been undertaken by EPA to support the use
of emission factors generated during studies of emissions in petroleum
refineries for similar sources in the Synthetic Organic Chemicals Manufacturing
Industry. The results of the EPA SOCMI studies, EPA data from a study of
fugitive emissions from petroleum refineries, and some industry studies of
fugitive emissions are discussed in Section A.I.
Section A.2 consists of the results of three studies on the effects of
maintenance on reducing fugitive VOC emissions from valves in petroleum
refineries and one study on maintenance of valves in a SOCMI process unit.
These results are included as an indication of the reduction in emissions
which could be expected as a function of the designated action level, and by
applying routine on-line maintenance procedures.
A.I FUGITIVE EMISSIONS TEST PROGRAMS
Three SOCMI test programs have been conducted by EPA. One was a study
performed by Monsanto Research Corporation of a small number of fugitive
emission sources in four SOCMI units. More intensive screening was performed
at six SOCMI units in another study. The third EPA study of SOCMI fugitive
A-l
-------
emissions was a screening and sampling program conducted at twenty-four
SOCMI units. The results of these studies are presented in this section.
Similar types of studies have been performed by industry. This section also
contains the results of an Exxon study of fugitive emissions in cyclohexane
unit and a DuPont study of fugitive emissions in unidentified process units.
The results of a study on fugitive emissions from petroleum refineries
are also presented in this section. Data on fugitive emissions were obtained
from 64 units in thirteen refineries located in major refining areas throughout
the country. Data on the effects of maintenance were obtained at the last
four of these refineries. These results are presented later in Section A.2
of this Appendix.
A.1.1 Study of Fugitive Emissions at Four SOCMI Units
Monsanto Research Corporation (MRC) conducted EPA Industrial Environmental
Research Laboratory (EPA-IERL) sponsored study of fugitive emissions at four
SOCMI units. The process units were monochlorobenzene, butadiene, ethylene
oxide/glycol, and dimethyl terephthalate. Due to the small number of
plants/processes sampled and the experimental design of this study, the
results were not considered to be comparable with the results of other
studies. Since the data generated by the MRC study could not be considered
representative of the SOCMI and valid conclusions could not be drawn
concerning the relative magnitude of fugitive emissions in the SOCMI, the
results of the study were not used in the development of standards for
fugitive emissions control. This study demonstrated the need for more
intensive sampling and screening which was undertaken by EPA.
A.1.2 Description and Results of EPA Study of Six SOCMI Units2'3'4'5
The objective of this test program was to gather data on the percentage
of sources which leak (as defined by a VOC concentration at the leak interface
of >_ 10,000 ppmv calibrated with methane). To achieve this objective, an
attempt was made to screen all potential leak sources (generally excluding
flanges) on an individual component basis with a portable organic vapor
analyzer. The test crews relied on plant personnel to identify equipment
handling organics. Normally, all pumps and compressor seals were examined,
and the percentage of valves carrying VOC which were screened ranged from 33
A-2
-------
to 85 percent. All tests were performed with a Century Systems Corporation
Organic Vapor Analyzer (OVA), Model 108, with the probe placed as close to
the source as possible. The results of this study are shown in Table A-l.
Six chemical process units were screened. Unit A is a chlorinated
methanes production facility in the Gulf Coast area which uses methanol as
feedstock material. The individual component testing was conducted during
September 1978. Unit B is a relatively small ethylene production facility
on the West Coast which uses an ethane/propare feedstock. Testing was
conducted during October 1978. Unit C is a chlorinated methanes production
facility in the Midwest. This plant also uses methanol as the basic organic
feedstock. Over the last few years, several pieces of equipment have been
replaced with equipment the company feels is more reliable. In particular,
the company has installed certain types of valves which they have found do
not leak "as much" as other valves. The individual component testing was
conducted during January 1979. Unit D is an ethylene production facility on
the Gulf Coast, using an ethane/propane feed. The facility is associated
with a major refinery, and testing was conducted during March 1979. Units E
and F are part of an intermediate size integrated petroleum refinery located
in the North Central United States. Testing was conducted during November
1978. Unit E is an aromatics extraction unit that produces benzene, toluene,
and xylene by extraction from refined petroleum feedstocks. Unit E is a new
unit and special attention was paid during the design and startup to minimize
equipment leaks. All valves were repacked before startup (adding 2 to 3
times the original packing) and all pumps in benzene service had double
mechanical seals with a barrier fluid. Unit F produces benzene by hydro-
deal kylation of toluene. Unit F was originally designed to produce a different
chemical and was redesigned to produce benzene.
A.1.3 Description and Results of an EPA Study of 24 SOCMI Units10
The U.S. EPA Industrial Environmental Research Laboratory coordinated
a study to develop information about fugitive emissions in the SOCMI. A
total of 24 chemical process units were selected for this purpose. The
process units were selected to represent a cross section of the population
of the SOCMI. Factors considered during process unit selections included
annual production volume, number of producers, volatility, toxicity, and
value of the final products. Table A-2 shows the process unit types selected
for screening.
A-3
-------
TABLE A-l. FREQUENCY OF LEAKS FROM FUGITIVE EMISSION SOURCES IN
SYNTHETIC ORGANIC CHEMICAL UNITS (Six Unit Study)
Unit Aa
Chloromethanes
Number
of
Equipment sources
type tested
Valves 600
Open-ended lines 52
Pump seals 47
Compressor seals -e
Control valves 52
Pressure relief
valves 7
Flanges 30
Drains -e
aSource: Reference 6.
Source: Reference 7.
°Source: Reference 8.
Snnrrp: Rpfprpnrp Q.
Percent with
screening
values
j>10,000 ppmv
1
2
15
6
0
3
Unit Ba
Ethyl ene
Number
of
sources
tested
2,301
386
51
42
128
e
_e
_e
Percent with
screening
values
MO, 000 ppmv
19
11
21
59
20
Unit Cb
Chloromethanes
Number
of
sources
tested
658
e
39
3
25
_e
e
e
Percent with
screening
values
> 10, 000 ppmv
0.1
3
33
0
Number
of
Unit Dc .
Ethyl ene
Percent with
screening
sources values
tested
862
90
63
17
25
e
e
39
>10,000 ppmv
14
13
33
6
44
10
Unit
Ed
BTX Recovery
Number Percent with
of
sources
tested _>
715
33
33f
e
53
e
_e
e
screening
values
10,000 ppmv
1.1
0.0
3.0
4.0
Unit Fd
Toluene HDA
Number
of
sources
tested
427
28
30
e
44
e
e
e
Percent with
screening
values
>10,000 ppmv
7.0
11.0 .
10.0
11.0
No data.
Pump seals in benzene service have double mechanical seals.
-------
TABLE A-2. TWENTY-FOUR CHEMICAL PROCESS UNITS SCREENED FOR FUGITIVE
EMISSIONS
Unit Type
1. Vinyl Acetate
2. Ethylene
3. Vinyl Acetate
4. Ethylene
5. Cumene
6. Cumene
7. Ethylene
8. Acetone/Phenol
9. Ethylene Dichloride
10. Vinyl Chloride Monomer
11. Formaldehyde
12. Ethylene Dichloride
13. Vinyl Chloride Monomer
14. Methyl Ethyl Ketone
15. Methyl Ethyl Ketone
16. Acetaldehyde
17. Methyl Methacrylate
18. Adipic Acid
19. Trichloroethylene/Perch!oroethylene
20. 1,1,1-Trichloroethane
21. Ethylene Dichloride
22. Adipic Acid
23. Acrylonitrile
24. Acrylonitrile
Source: Reference 11
A-5
-------
The screening work began with the definition of the process unit
boundaries. All feed streams, reaction/separation facilities, and product
and by-product delivery lines were identified on process flow diagrams and
in the process unit. Process data, including stream composition, line
temperature, and line pressure, were obtained for all flow streams. Each
process stream to be screened was identified and process data was obtained
with the assistance of plant personnel, in most cases. Sources were screened
by a two-person team (one person handling the hydrocarbon detector and one
person recording data).
The Century Systems Models OVA-108 and OVA-128 hydrocarbon detectors
were used for screening. The HNU Systems, Inc., Model PI 101 Photoionization
Analyzer was also used to screen sources at the formaldehyde process unit.
The detector probe of the instrument was placed directly on those areas of
the sources where leakage would typically occur. For example, gate valves
were screened along the circumference of the annular area around the valve
stem where the stem exits the packing gland and at the packing gland/valve
bonnet interface. All process valves, pump seals, compressor seals,
agitator seals, relief valves, process drains, and open-ended lines were
screened. From five to twenty percent of all flanges were randomly selected
and screened. For the purpose of this program "flange" referred to any
pipe-to-pipe or tubing-to-tubing connection, excluding welded joints.
.Each screening instrument was calibrated at least daily. The model
OVA-108 instruments, with a logarithmic scale reading from 1 ppmv to 10,000
ppmv, were calibrated with high (8,000 ppmv) and low (500 ppmv) concentration
methane-in-air standards to ensure accurate operation at both ends of the
instrument's range. The model OVA-128 instruments, with a linear readout
ranging from 0 ppmv to 1,000 ppmv, were also calibrated with high and low
concentration standards. A pre-calibrated dilution probe was required with
the OVA-128 when calibrating with the 8,000 ppmv standard.
The HNU Photoionization instrument, used to screen the formaldehyde
process unit, was calibrated with isobutylene which has an ionization potential
close to that of formaldehyde.
A-6
-------
Results of the screening program at the 24 process units are summarized
in Table A-3.
The fugitive emission sources in the study were screened at an average
rate of 1.7 minutes per source for a two-person team (or 3.4 person-minutes
per source). This average screening rate includes time spent for instrument
calibration and repair. Table A-4 presents screening time data on a unit-
by-unit basis. These time requirements are somewhat higher than would be
expected for routine monitoring because of the extensive record keeping
associated with the screening project.
12
A.1.4 Description and Results of Refinery Fugitive Emissions Study
Data concerning the leak frequencies and emission factors for various
fugitive sources were obtained primarily at nine refineries. More complete
information for compressors' and relief valves' emissions was obtained by
sampling at four additional refineries. Refineries were selected to provide
a range of sizes and ages and all of the major petroleum refinery processing
units were studied. The type of process units and the number of each studied
in the first nine refineries are listed in Table A-5.
In each refinery, sources in six to nine process units were selected
for study. The approximate number of sources selected for study and testing
in each refinery is listed below:
Valves 250-300
Flanges 100-750
Pump seals 100-125
Compressor seals 10-20
Drains 20-40
Relief Valves 20-40
There were normally 500-600 sources selected in each refinery.
The distribution of sources among the process units was determined
before the selection and testing of individual sources was begun. Individual
sources were selected from piping and instrumentation diagrams or process
flow diagrams before a refinery processing area was entered. Only those
preselected sources were screened. In this way, bias based on observation
of individual sources was theoretically eliminated.
A-7
-------
TABLE A-3. SUMMARY OF SOCMI PROCESS UNITS FUGITIVE EMISSIONS
(Twenty-four unit study)
00
Source Type
Flanges
Process Drains
Open Ended Lines
Agitator Seals
Relief Valves
Valves
Pumps
Compressors
Other3
Service
Gas
Light
Heavy
Gas
Light
Heavy
Gas
Light
Heavy
Gas
Light
Heavy
Gas
Light
Heavy
Gas
Light
Heavy
Light
Heavy
Gas
Gas
Light
Heavy
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
(1)
Number
Screened
1,443
2,897
607
83
527
28
923
3,603
477
7
8
1
85
69
3
9,668
18,294
3,632
647
97
29
19
33
2
(2)
% Not
Screened
4.
2.
2.
23.
1.
0.
17.
10.
21.
46.
11.
66.
72.
40.
66.
17.
12.
9.
4.
40.
9.
9.
19.
33.
6
6
4
1
9
0
5
4
5
1
1
7
7
5
7
5
2
9
3
5
4
5
5
3
(3)
% of Screened Sources
with Screening Values
a 10,000 ppmv
4
1
0
2
3
7
5
3
1
14
0
0
3
2
0
11
6
0
8
2
6
21
6
0
.6
.2
.0
.4
.8
.1
.8
.9
.3
.3
.0
.0
.5
.9
.0
.4
.4
.4
.8
.1
.9
.0
.1
.0
(4)
95% Confidence Interval
for Percentage of Sources
a 10,000 ppmv
(3.
(0.
(0.
(0.
(2.
(0.
(4.
(3.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(10.
(6.
(0.
(6.
(0.
(0.
(6.
(0.
(0.
6,
9,
0,
3,
3,
9,
4,
3,
5,
4,
0,
0,
7,
3,
0,
8,
1,
2,
6,
3,
9,
0,
7,
0,
5.
1.
0.
8.
5.
23
7.
4.
2.
57
36
97
10
10
70
12
6.
0.
11
7.
22
45
20
84
8)
8)
6)
4)
8)
.5)
5)
6)
8)
.9)
.9)
.5)
.0)
.1)
.8)
.1)
8)
7)
.1)
3)
.8)
.6)
.2)
.2)
Includes filters, vacuum breakers, expansion joints, rupture disks, slight glass seals, etc. Source: Ref. 13
-------
TABLE A-4. AVERAGE FUGITIVE EMISSION SOURCE SCREENING RATES
(Twenty-four Unit Study)
Average Screening
Number of Time Per
Process Unit Type Screened Sources Source, Minutes
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
Vinyl Acetate
Ethyl ene
Vinyl Acetate
Ethyl ene
Cumene
Cumene
Ethyl ene
Acetone/Phenol
Ethyl ene Di chloride
Vinyl Chloride Monomer
Formaldehyde
Ethylene Bichloride
Vinyl Chloride Monomer
Methyl Ethyl Ketone
Methyl Ethyl Ketone
Acetaldehyde
Methyl Methacrylate
Adi pic Acid
Tri chol oroethyl ene/Perchl oroethyl ene
1,1,1-Trichloroethane
Ethylene Di chloride
Adipic Acid
Acrylonitrile
Acrylonitrile
1,391
5,078
2,780
5,278
1,025
1,573
3,685
3,207
1,430
868
230
744
2,619
585
679
1,148
2,019
1,577
Z,720
570
42
664
1,406
1,864
2.0
1.3
0.9
1.5
0.9
1.0
1.9
3.2
2.6
1.8
1.6
1.6
2.2
1.2
0.9
0.7
1.6
1.9
2.5
1.9
Total 43,182 1.7
aAverage source screening time was determined for a two-person team,
one person screening with a portable hydrocarbon detector and one
person recording data. Average screening time includes time spent
for instrument calibration, maintenance, and repair.
Source: Ref. 14
A-9
-------
TABLE A-5. SAMPLED PROCESS UNITS FROM NINE REFINERIES
DURING REFINERY STUDY
Refinery Number of
process unit sampled units
Atmospheric distillation 7
Vacuum distillation 4
Thermal operations (coking) 2
Catalytic cracking 5
Catalytic reforming 6
Catalytic hydrocracking 2
Catalytic hydrorefining 2
Catalytic hydrotreating 7
Alkylation 6
Aromatics/isomerization 3
Lube oil manufacture 2
Asphalt manufacture 1
Fuel gas/light-ends processing 11
LPG 2
Sulfur recovery 1
Other 3
Source: Ref. 15
.A-10
-------
The screening of sources was accomplished with portable organic vapor
detectors. The principal device used in this study was the J. W. Bacharach
Instrument Co. "TLV Sniffer" calibrated with hexane. The components were
tested on an individual basis and only those components with VOC concentrations
in excess of 200 ppmv were considered for further study.
A substantial portion of these leaking sources was enclosed and sampled
to determine both the methane and nonmethane emission rates. An important
result of this program was the development of a correlation between the
maximum observed screening value (VOC concentration) and the measured nonmethane
leak rate.
Emission factors and leak frequency information generated during this
study are given in Table A-6.
A.1.5. Comparison of Fugitive Emissions Test Data from Refineries and
SOCMI Units
The results of the SOCMI studies and those of the refinery emissions
study are compared in Table A-7.
A.1.6 Description and Results of the DuPont Study
DuPont conducted a program of VOC fugitive emission measurement from
pumps and valves at two of their plants. The processes of the 5 and 10 year
old plants were not revealed. The OVA-108 was used for screening (leak
identification) and for leak rate determination (analysis of collected leak
vapors). The leak rate was determined by taking Tedlar bags partially
filled with air and enclosing the leaking valve. The hydrocarbon concentration
in the bags was recorded as a function of time. Visual estimates of the
initial bag volume were assumed to be ±5 percent. Dupont did not have a
dilution probe and, therefore, measurements above 10,000 ppmv were not
made. Analysis of the data collected indicates that no significant difference
in leak rates exists between manual and automatic control valves. Significant
trends were observed with changes in product vapor pressure. It also seemed
that full open or closed valve seat positions resulted in lower leak rates
than intermediate positions. The results of the DuPont study are shown in
Table A-8.
on oc
A.1.7 Description and Results of the Exxon Study '
A fugitive emissions study was conducted by Exxon Chemical Company at
the cyclohexane unit at their Baytown plant. The total number of valves,
pumps and compressor seals, and safety valves was determined. For all
A-ll
-------
TABLE A-6. LEAK FREQUENCIES AND EMISSION FACTORS FROM FUGITIVE
EMISSION SOURCES.IN PETROLEUM REFINERIES
Percent of
sources having
screening values
Equipment >. 10,000 ppmv
type TLV-Hexane
Valves
Gas service
Light liquid service
Heavy liquid service
Pump seals
Light liquid service
Heavy liquid service
Compressor seals (hydrocarbon
service)
Pressure relief valves
Gas service
Light liquid service
Heavy liquid service
Flanges
Open-ended lines
Gas service
Light liquid service
Heavy liquid service
NA
10
12
0
NA
23
2
33
8
0
NA
Estimated emission
factor for
refinery sources,
kg/hr-source
NA
0.021
0.010
0.0003
NA
0.12
0.02
0.44
0.086
0.16
0.006
0.009
0.0003
NA
0.025
0.014
0.003
Source: Ref. 17
A-12
-------
TABLE A-7. COMPARISON OF LEAK FREQUENCIES FOR FUGITIVE EMISSION
SOURCES IN SOCMI UNITS AND PETROLEUM REFINERIES
Equipment Type
Valves (all)
Gas
Light Liquid
Heavy Liquid
Percent of SOCMI Sources
Having Screening Values
10,000 ppmv, OVA-108
(six unit study)
11
Percent oi SOCMI Sources
Having Screening Values
10,000 |ipmv, OVA-108
(24 unit study)
11.4
6.4
0.4
Percent of Petroleum
Refinery Sources Having
Screening Values
4,121 ppmv, TLVC
15.09
16.98
0.38
Open-ended lines (all)
Gas
Light Liquid
Heavy Liquid
10
5.8
3.9
1.3
N/A
Pumps (all)
Light Liquid
Heavy Liquid
17
8.8
2.1
31.70
5.48
Compressors (Gas)
Pressure Relief Valves (all)
43
0
6.9
54.23
Gas
Light Liquid
Heavy Liquid
Flanges (all) 3
Gas
Light Liquid
Heavy Liquid
Process Drains (all) N/A
Gas
Light Liquid
Heavy Liquid
Agitator Seals (all) N/A
Gas
Light Liquid
Heavy Liquid
Otherd N/A
3.5
2.9
0.0
4.6
1.2
0.0
2.4
3.8
7.1
14.3
0.0
0.0
5.92
1.89
6.67
1.36
1.14
0.00
0.00
7.00
3.74
N/A
aSource: Ref. 18, 19, 20, 21
bSource: Ref. 22
cSource: Ref. 23. Screening with OVA-108 (methane) at 10,000 ppmv is equivalent to screening with TLV (hexane)
at 4,121 ppmv.
Includes filters, vacuum breakers, expansion joints, rupture disks, sight glass seals, etc.
A-13
-------
TABLE A-8. FREQUENCY OF LEAKS0 FROM FUGITIVE EMISSION
SOURCES IN THO DuPONT PLANTS.
Equipment
type
Valves
Gas
Light liquid
Heavy liquid
No. of
leakers
48
35
11
1
No. of
non-leakers
741
120
143
478
Percent
leakers
6.1
23.1
7.1
0.2
Pumps
Light liquid
Heavy liquid
1
0
36
6
29
2.7
14.3
0
aLeak defined as 10,000 ppmv or greater.
Source: Ref. 26
A-14
-------
sources, except valves, all of the fugitive emission sources were sampled.
For valves, a soap solution was used to determine leaking components. All
leaking valves were counted and identified as either small, medium or large
leaks. From the set of valves found to be leaking, specific valves were
selected for sampling so that each class of leaking valves was in approximately
the same proportion as it occured in the cyclohexane unit.
Heat resistant mylar bags or sheets were taped around the equipment to
be sampled to provide an enclosed volume. Clean metered air from the
filter apparatus was blown into the enclosed volume. The sampling train was
allowed to run until a steady state flow was obtained (usually about 15
minutes). A bomb sample was then taker for laboratory analysis (mass
spectrometry). Table A-9 presents the results of the Exxon study.
A.2 MAINTENANCE TEST PROGRAMS
The results of four studies on the effects of maintenance on fugitive
emissions from valves are discussed in this section. The first two studies
were conducted by refinery personnel at the Union Oil Co. refinery in Rodeo,
California, and the Shell Oil Co. refinery in Martinez, California. These
programs consisted of maintenance on leaking valves containing fluids with
Reid vapor pressures greater than 1.5. The third study was conducted by
EPA. Valves were selected and maintained at four refineries. The fourth
study was conducted by EPA at Unit D (ethylene unit). The study results and
a description of each test program are given in the following sections.
29
A.2.1 Description and Results of the Union Maintenance Study
The Union valve maintenance study consisted of performing undirected
maintenance on valves selected from 12 different process units. Maintenance
procedures consisted of adjusting the packing gland while the valve was in
service. Undirected maintenance consists of performing valve repairs
without simultaneous measurement of the effect of repair on the VOC concentration
detected. This is in contrast to directed maintenance where emissions are
monitored during the repair procedure. With directed maintenance, repair
procedures are continued until the VOC concentration detected drops to a
specified level or further reduction in the emission level is not possible.
Also, maintenance may be curtailed if increasing VOC concentrations result.
A-15
-------
TABLE A-9. FREQUENCY OF LEAKS3 FROM FUGITIVE EMISSION
SOURCES IN EXXON'S CYCLOHEXANE UNIT
Equipment
Source
Valves
Gas
Light
liquid
Safety
valves
Pump .
seals
Compressor
seals
Total
in Unit
136
201
15
8
N/A
Screened and
Sampled
136
100
15
8
N/A
Percent
Leaking
32
15
87
83
100
Emission
f actor (kg/hr)
0.017
0.008
0.064
0.255
0.264
99.8% Confidence
Interval (kg/hr)
0.008 -
0.003 -
0.013 -
0.082 -
0.068 -
0.035
0.007
0.5
0.818
1.045
N/A - Not available
aLeak defined as 10,000 ppmv or greater.
Double mechanical seal pumps and compressors were found to have negligible
leaks.
Source: Reference 27,28
A-16
-------
The Union data were obtained with a Century Systems Corporation Organic
Vapor Analyzer, OVA-108. All measurements were taken at a distance of 1 cm
from the seal. Correlations developed by EPA have been used to convert the
data from OVA readings taken at one centimeter to equivalent TLV readings at
the leak interface (TLV-0). This facilitates comparison of data from
different studies and allows the estimation of emission rates based on
screening values-leak rate correlations.
The results of the Union study are given in Table A-10. Two sets of
results are provided; the first includes all ^epaired valves with before
maintenance screening values greater than or of 5,300 ppmv, obtained with
OVA at 1 cm from the leak interface, is equivalent to a screening value of
10,000 ppmv measured by a Bacharach Instrumen; Co. "TLV Sniffer" directly at
the leak interface. The OVA-1 cm readings have been converted to equivalent
TLV-0 cm readings because:
1) EPA correlations which estimate leak ^ates from screening values
were developed from TLV-0 cm data.
2) Additional maintenance study data exists in the TLV-0 cm format.
3) EPA Reference Method 21 specifies 0 cm screening procedures.
The results of this study indicate that Maintenance on valves with
initial screening values above 10,000 ppmv (O/A-108) is much more effective
than maintenance on valves leaking at lower rites. In fact, this study
indicates that emissions from valves are reduced by an average of 51.8
percent for valves initially over 5,300 ppmv while valves with lower initial
screening values experienced an increase of 30.5 percent.
A.2.2 Description and Results of the Shell Maintenance Study
The Shell maintenance program consisted of two parts. First, valve
repairs were performed on 171 leaking valves. In the second part of the
program, 162 of these valves were rechecked and additional maintenance was
performed. Maintenance consisted of adjusting the packing gland while the
valve was in service. The second part of the program was conducted approximately
one month after the initial maintenance period. It was not determined
whether the maintenance procedures were directed or undirected, based on the
information reported by Shell.
A-17
-------
TABLE A-10. SUMMARY OF MAINTENANCE STUDY RESULTS FROM THE UNION OIL CO.
REFINERY IN RODEO, CALIFORNIA3
i—"
00
All valves
with initial
screening values
>_5300 ppmvb
Number of repairs attempted
Estimated emissions before maintenance, kg/hr
Estimated emissions after maintenance, kg/hr
Number of successful repairs (<5300 ppmv after maintenance)
Number of valves with decreased emissions
Number of valves with increased emissions
Percent reduction in emissions
Percent successful repairs
Percent of valves with decreased emissions
Percent of valves with increased emissions
133
9.72
4.69
67
124
9
51.8
50.4
93.2
6.8
All valves
with initial
screening values
<5300 ppmv
21
0.323
0.422
--
13
8
-30.5
--
61.9
38.1
Source: Ref. 33.
value of 5300 ppmv, taken with the OVA-108 at 1 cm, generally corresponds to a value of
10,000 ppmv taken with a "TLV Sniffer" at 0 cm.
-------
Fugitive VOC emissions were measured one centimeter from the source
using the OVA-108. These data have been transformed to TLV-0 cm values as
were the Union data. The same methods of data analysis described in
Section A.2.1 have been applied to the Shell data. The results of the
Shell maintenance study are given in Table A-ll.
32
A.2.3 Description and Results of the EPA Maintenance Study
Repair data were collected on valves located in four refineries. The
effects of both directed and undirected maintenance were evaluated. Maintenance
consisted of routine operations, such as tightening the packing gland or
adding grease. Other data, including valve size and type and the processes'
fluid characteristics, were obtained. Screening data were obtained with the
Bacharach Instrument Company "TLV Sniffer" and readings were taken as close
to the source as possible.
Unlike the Shell and Union studies, emission rates were not based on
the screening value correlations. Rather, each valve was sampled to determine
emission rates before and after maintenance using techniques developed by
EPA during the refinery emission factor study. These values were used to
evaluate emissions reduction.
The results of this study are given in Table A-12. Of interest here is
a comparison of the emissions reduction for directed and undirected maintenance.
The results indicate that directed maintenance is more effective in reducing
emissions than is undirected maintenance, particularly for valves with lower
initial leak rates. The results showed an increase in total emissions of
32.6% for valves with initial screening value:; less than 10,000 ppmv which
were subjected to undirected maintenance. However, this increase is due to
a large increase in the emission rate of only one valve.
QC
A.2.4 Description and Results of Unit D (Ethylene Unit) Maintenance Study
Maintenance was performed by Unit D personnel. Concentration measurements
of VOC were made using the OVA-108, and readings were obtained at the
closest distance possible to the source. The results of this study are
shown in Table A-13. Directed and undirected maintenance procedures were
used. The results show that directed maintenance results in more repairs
being successfully completed than when undirected maintenance is used.
A-19
-------
TABLE A-ll.
SUMMARY OF MAINTENANCE STUDY RESULTS FROM THE SHELL OIL.COMPANY
REFINERY IN MARTINEZ, CALIFORNIA3
ro
O
Number of repairs attempted
Estimated emissions before
maintenance, kg/hr
Estimated emissions after
maintenance, kg/hr
Number of successful repairs
(<5300 ppmv after maintenance)
Number of valves with decreased
emissions
Number of valves with increased
emissions
Percent reduction in emissions
Percent successful repairs
Percent of valves with decreased
emissions
Percent of valvei witn increased
emissions
March
All repaired valves
with initial screening
values >5300 ppmvb ,
161
11.08
2.66
105
161
0
76.0
65.2
100.00
0.0
maintenance
All repaired valves
with initial screening
values <5300 ppmv
11
0.159
0.0
--
11
0
100.0
--
100.0
0.0
April
All repaired valves
with initial
(March) screening
values >5300 ppmv
152d
2.95
0.421
45
151
1
85.7
83.3
99.3
0.7
maintenance
All repaired valves
with initial
(March) screening
values <5300 ppmv
lle
0.060
0.0
--
11
0
100.0
--
100.0
0.0
Source: Reference 34.
The value of 5300 ppmv, taken with the OVA-108 at 1 cm, generally corresponds to a value of 10,000 ppmv taken with a "TLV Sniffer" at 0 cm.
cShell reported the screening value of all valves which measured <3000 ppmv (<1500 ppmv-TLV at 0 cm) as non-leakers. Emissions estimates obtained from
emission factors. Reference 14.
Initial value of 90 of these valves was <1500 ppm-TLV at 0 cm, 54 valves screened >5300 (note nine valves from initial data set not rechecked in April).
elr,itial value of 10 of these valves was <1500 ppm-TLV at 0 cm.
-------
TABLE A-12. SUMMARY OF EPA REFINERY MAINTENANCE STUDY RESULTS
ro
Number of valves repaired
Measured emissions before
maintenance, kg/hr
Measured emissions after
maintenance, kg/hr
Number of successful repairs
(<10,000 ppmv after maintenance)
Number of valves with decreased
emissions
Number of valves with increased
emissions
Percent reduction in emissions
Percent successful repairs
Percent of valves with decreased
emissions
Percent of valves with increased
emissions
Repaired values
screening values
Di rected
maintenance
9
0.107
0.0139
8
9
0
87.0
88.9
100.0
0.0
with initial
>10,000 ppmv
Undirected
maintenance
23
1.809
0.318
13
21
2
82.4
56.5
91.3
8.7
Repaired values
screening values
Directed
maintenance
10
0.0332
0.0049
--
6
4
85.2
--
60.0
40.0
with initial
<10,000 ppmv
Undirected
maintenance
16
0.120
0.159
--
15
1
-32.6
--
93.8
6.3
Source: Ref. 36
-------
TABLE A-13. MAINTENANCE EFFECTIVENESS
UNIT D ETHYLENE UNIT BLOCK VALVES
1. Total number of valves with VOC > 10,000 ppm
from unit survey 121
2. Total number of valves tested for
maintenance effectiveness 46
% Tested 38%
UNDIRECTED MAINTENANCE
3. Total number subjected to repair attempts 37
4. Successful repairs (VOC<10,000 ppm) 22
% Repaired 59%
Followup
DIRECTED MAINTENANCE
5.
6.
Number of valves unrepaired by undirected
maintenance subjected to directed maintenance
Number repaired by followup directed maintenance
14
5
% of unsuccessful repaired by
directed maintenance 36%
7. Total number repaired based on undirected 27
maintenance subset (3) above
% Repaired 73%
8. Total number of repairs including leaks not 29
found before initial maintenance
Total % repaired 63%
Total % not repaired 37%
Source: Reference 37
A-22
-------
A.2.5 Comparison of Maintenance Study Results
Generally speaking, the results of these maintenance programs would
tend to support the following conclusions:
1) A reduction in emissions may be obtained by performing
maintenance on valves with screening values above
10,000 ppmv (measured at the source).
2) The reduction in emissions due to maintenance of valves
with screening values below 10,000 ppmv is not as
dramatic and may result in increased emissions.
3) Directed maintenance is preferable to undirected maintenance
for valve repair.
The information presented in Tables A-10, A-ll, A-12, and A-13 has been
compiled with the objective of placing the data on as consistent a basis as
possible. However, some differences were unavoidable and others may have
gone unrecognized, due to the limited emount of information concerning the
details of methods used in each study. Therefore, care should be exercised
before attempting to draw specific quantitative conclusions based on direct
comparison of the results of these studies.
A-2.3
-------
A.3 REFERENCES
1. Tichenor, Bruce A., memo to K.C. Hustvedt, U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. Research Triangle
Park, N.C. October 27, 1980.
2. Muller, Christopher, memo to files. U.S. Environmental Protection
Agency, Emission Standards and Engineering Division. Research Triangle
Park, N.C. January 18, 1979 (Plants A & B).
3. Muller, Christopher, memo to file. U.S. Environmental Protection
Agency, Emission Standards and Engineering Division, Research Triangle
Park, N.C. March 19, 1979 (Plant C).
4. Muller, Christopher, memo to file. U.S. Environmental Protection
Agency, Emission Standards and Engineering Division. Research Triangle
Park, N.C. March 19, 1979. (Plant D).
5. Hustvedt, K.C., trip report to James F. Durham* Chief, Petroleum
Section. U.S. Environmental Protection Agency. January 5, 1979
(Plants E & F).
6. Reference 2.
7. Reference 3.
8. Reference 4.
9. Reference 5.
10. Blacksmith, J.R., G.E. Harris, and G.J. Langley, Frequency of Leak
Occurence for Fittings in Synthetic Organic Chemical Plant Process
Units. EPA Contract Numbers 68-02-3176-01,06/68-02-3173-02,11/68-02-
3171-01/68-02-3174-04. Radian Corporation, Austin, Texas. September
1980.
11. Reference 10.
12. Wetherold, R.G., and L.P. Provost, Emission Factors and Frequency of
Leak Occurence for Fittings in Refinery Process Units. EPA-600/2-79-
044. Radian Corporation, Austin, Texas. February 1979.
13. Reference 10.
14. Reference 10.
15. Reference 12.
16. Meeting Report. Meeting held between U.S. Environmental Protection
Agency and DuPont at Durham, N.C. June 12, 1979.
A-24
-------
17. Reference 12.
18. Reference 2.
19. Reference 3.
20. Reference 4.
21. Reference 5.
22. Reference 10.
23. Reference 12.
24. Fugitive loss study summary and memo from James B. Cox, Exxon Chemical
Company, Baytown, Texas, to J.W. Blackburn, Hydroscience Incorporated.
February 21, 1978.
25. Letter from James B. Cox, Exxon Chemical Company, Baytown, Texas, to
R.T. Walsh, U.S. Environmental Protection Agency, Chemical and Petroleum
Branch. March 21, 1979.
26. Reference 16.
27. Reference 24.
28. Reference 25.
29. Valve Repair Summary and Memo from F.R. Bottomley, Union Oil Company,
Rodeo, Calfornia, to Milton Feldstein, Bay Area Quality Management
District. April 10, 1979.
30. Reference 12.
31. Valve Repair Summary and Memo from R.M. Thompson, Shell Oil Company,
Martinez Manufacturing Complex, Martinez, California. To Milton
Feldstein, Bay Area Quality Management District. April 26, 1979.
32. Radian Corporation. Assessment of Atmospheric Emissions from Petroleum
Refining, Final Report, Appendix B, Detailed Results. EPA Report No.
600/2-80-075C, Exhibits. Austin, Texas. April, 1980.
33. Reference 29.
34. Reference 31.
A-25
-------
35. Air Pollution Emission test at Phillips Petroleum Company, Sweeney,
Texas. EMB Report No. 78-DCM-12E, December 1979.
36. Reference 32.
37. Reference 35.
A-26
-------
APPENDIX B. LIST DEFINING SYNTHETIC ORGANIC CHEMICAL, POLYMER,
AND RESIN MANUFACTURING INDUSTRIES
-------
APPENDIX B
LIST OF CHEMICALS DEFINING SYNTHETIC
ORGANIC CHEMICAL, POLYMER, AND RESIN
MANUFACTURING INDUSTRIES
TABLE I: Synthetic Organic Chemicals Manufacturing Industry
OCPDB No.*
20
30
40
50
65
70
80
90
100
110
120
125
130
140
150
160
170
180
185
190
200
210
220
Chemical
Acetal
Acetaldehyde
Acetal dol
Acetamide
Acetanilide
Acetic acid
Acetic anhydride
Acetone
Acetone cyanohydrin
Acetom'trile
Acetophenone
Acetyl chloride
Acetyl ene
Acrolein
Aery 1 amide
Acrylic acid and esters
Acrylonltrile
Adipic acid
Adiponitrile
Alkyl naphthalenes
Allyl alchohol
Ally! chloride
Aminobenzolc acid
OCPDB No. Chemical
230 Aminoethylethanolamine
235 p-aminophenol
240 Amyl acetates
250 Amyl alcohols
260 Amyl amine
270 Amyl chloride
280 Amyl mercaptans
290 Amyl phenol
300 Aniline
310 Aniline hydrochloride
320 Anisidine
330 Anisole
340 Anthranillc acid
350 Anthraquinone
360 Benzaldehyde
370 Benzamide
380 Benzene
390 Benzenedisulfonic add
400 Benzenesulfonic acid
4:0 Benzil
420 Benzilic acid
430 Benzole acid
440 Benzoin
*The OCPDB Numbers are reference indices assigned to the various chemicals
in the Organic Chemical Producers Data Base developed by EPA.
B-l
-------
OCPDB No.
Chemical
OCPDB No.
Chemical
450 Benzonitrile
460 Benzophenone
480 Benotrichloride
490 Benzoyl chloride
500 Benzyl alcohol
510 Benzyl amine
520 Benzyl benzoate
530 Benzyl chloride
540 Benzyl dichloride
550 Biphenyl
560 Bisphenol A
570 Bromobenzene
580 Bromonaphthalene
590 Butadiene
592 1-butene
600 n-butyl acetate
630 n-butyl acrylate
640 n-butyl alcohol
650 s-butyl alcohol
660 t-butyl alcohol
670 n-butylamine
680 s-butylamine
690 t-butylamine
700 p-tert-butyl benzoic acid
710 1,3-butylene glycol
750 n-butyraldehyde
760 Butyric acid
770 Butyric anhydride
780 Butyronitrile
785 Caprolactam
790 Carbon disulfide
800 Carbon tetrabronide
810 Carbon tetrachloride
820 Cellulose acetate
840 Chloroacetic acid
850 m-chloroaniline
860 o-chloroaniline
870 p-chloroaniline
880 Chlorobenzaldehyde
890 Chlorobenzene
900 Chlorobenzoic acid
905 Chlorobenzotrichloride
910 Chlorobenzoyl chloride
920 Chlorodifluoroethane
921 Chlorodifluoromethane
930 Chloroform
940 Chloronaphthalene
950 o-chloronitrobenzene
951 p-chloronitrobenzene
960 Chlorophenols
964 Chloroprene
965 Chlorosulfonic acid
970 m-chlorotoluene
980 o-chlorotoluene
990 p-chlorotoluene
992 Chlorotrlfluoromethane
1000 m-cresol
1010 o-cresol
1020 p-cresol
1021 Mixed cresols
1030 Cresylic acid
1040 Crotonaldehyde
1050 Crotonic acid
1060 Cumene
1070 Cumene hydroperoxide
1080 Cyanoacetic acid
1090 Cyanogen chloride
1100 Cyanuric acid
1110 Cyanuric chloride
1120 Cyclohexane
1130 Cyclohexanol
1140 Cyclohexanone
B-2
-------
OCPDB No. Chemical
1150 Cyclohexene
1160 Cyclohexylamine
1170 Cyclooctadiene
1180 Decanol
1190 Dlacetone alcohol
1200 Diaminobenzoic acid
1210 Dichloroaniline
1215 m-dichlorobenzene
1216 o-dichlorobenzene
1220 p-dichlorobenzene
1221 Dichlorodifluoromethane
1244 l,2-d1chloroethane (EDC)
1240 Dichloroethyl ether
1250 Dichlorohydrin
1270 Dichloropropene
1280 Dlcyclohexylamine
1290 01 ethyl amine
1300 Diethylene glycol
1304 Diethylene glycol diethyl ether
1305 Diethylene glycol dimethyl ether
1310 Diethylene glycol monobutyl ether
1320 Diethylene glycol monbutyl ether acetate
1330 Diethylene glycol monoethyl ether
1340 Diethylene glycol monoethyl ether acetate
1360 Diethylene glycol monomethyl ether
1420 Diethyl sulfate
1430 Difluoroethane
1440 DUsobutylene
1442 D11sodecyl phthalate
1444 Dilsooctyl phthalate
1450 Diketene
1460 Dimethylamine
1470 N,N-d1nethylaniline
1480 N,N-d1methyl ether
1490 N.N-dimethylfomiamide
OCPDB No.
1495
1500
1510
1520
1530
1540
1545
1550
1560
1570
1580
1590
1600
1610
1620
1630
1640
1650
1660
1661
1670
1680
1690
1700
1710
1720
1730
1740
1750
1760
1770
1780
1790
1800
1810
1830
Chemical
Dimethylhydrazlne
Dimethyl sulfate
Dimethyl sulfide
Dimethyl sulfoxide
Dimethyl terephthalate
3,5-dinitrobenzo1c acid
Dinltrophenol
Dinitrotoluene
Dioxane
Dioxolane
Diphenylamine
Dlphenyl oxide
Diphenyl thlourea
Dipropylene glycol
Dodecene
Dodecylanillne
Dodecyl phenol
Epichlorohydrin
Ethanol
Ethanol amines
Ethyl acetate
Ethyl acetoacetate
Ethyl acrylate
Ethylamlne
Ethyl benzene
Ethyl bromide
Ethyl cellulose
Ethyl chloride
Ethyl chloroacetate
Ethyl cyanoacetate
Ethyl ene
Ethyl ene carbonate
Ethyl ene chlorohydrin
Ethylenediamine
Ethylene dlbromide
Ethyl ene glycol
B-3
-------
OCPDB No. Chemical
1840 Ethylene glycol diacetate
1870 Ethylene glycol dimethyl ether
1890 Ethylene glycol monobutyl ether
1900 Ethylene glycol monobutyl ether acetate
1910 Ethylene glycol monoethyl ether
1920 Ethylene glycol monoethyl ether acetate
1930 Ethylene glycol monomethyl ether
1940 Ethylene glycol monomethyl ether acetate
1960 Ethylene glycol monophenyl ether
1970 Ethylene glycol monoprppyl ether
1980 Ethylene oxide
1990 Ethyl ether
2000 2-ethylhexanol
2010 Ethyl orthoforroate
2020 Ethyl oxalate
2030 Ethyl sodium oxalacetate
2040 Formaldehyde
2050 Formamide
2060 Formic acid
2070 Fumaric acid
2073 Furfural
2090 Glycerol (Synthetic)
2°91 Glycerol dichlorohydrin
2100 Glycerol triether
2110 Glycine
2120 Glyoxal
2145 Hexachlorobenzene
2150 Hexachloroethane
2160 Hexadecyl alcohol
2165 Hexamethylenediamine
2170 Hexamethylene glycol
2180 Hexamethylenetetramine
2190 Hydrogen cyanide
2200 Hydroquinane
2210 p-hydroxybenzoic acid
2240 Isoamylene
2250 Isobutanol
2260 Isobutyl acetate
2261 Isobutylene
2270 Isobutyraldehyde
2280 Isobutyric acid n n
OCPDB No. Chemical
2300 Isodecanol
2320 Isooctyl alcohol
2321 Isopentane
2330 Isophorone
2340 Isophthalic acid
2350 Isoprene
2360 Isopropanol
2370 Isopropyl acetate
2380 Isopropylamine
2390 Isopropyl chloride
2400 Isopropylphenol
2410 Ketene
2414 Linear alkyl sulfonate
2417 Linear alkylbenzene
2420 Maleic acid
2430 Maleic anhydride
2440 Malic acid
2450 Mesityl oxide
2455 Metanilic acid
2460 Methacrylic acid
2490 Methallyl chloride
2500 Methanol
2510 Methyl acetate
2520 Methyl acetoacetate
2530 Methyl amine
2540 n-methylaniline
2545 Methyl bromide
2550 Methyl butynol
2560 Methyl chloride
2570 Methyl cyclohexane
2590 Methyl cyclohexanone
2620 Methylene chloride
2530 Methylene dianiline
2635 Methylene diphenyl diisocyanate
2640 Methyl ethyl ketone
2644 Methyl formate
2650 Methyl isobutyl carblnol
2660 Methyl isobutyl ketone
2665 Methyl methacrylate
2670 Methyl pentynol
2690 c. -methyl styrene
-------
OCPDB No. Chemical
2700 Horpholine
2710 a -naphthalene sulfonlc add
2720 B-naphthalene sulfonic add
2730 a-naphthol
2740 B-naphthol
2750 Neopentanoic add
2756 o-nitroaniline
2757 p-nltroaniline
2760 o-n1troanisole
2762 p-nitroanisole
2770 Nitrobenzene
2780 Nitrobenzoic add (o, m, and p)
2790 Nitroethane
2791 NUrotnethane
2792 Nitrophenol
2795 Nltropropane
2800 Nltrotoleune
2810 Nonene
2820 Nonyl phenol
2830 Octyl phenol
2840 Paraldehyde
2850 Pentaerythritol
2851 n-pentane
2855 1-pentene
2860 Perchloroethylene
2882 Perchloromethyl mercaptan
2890 o-phenetidine
2900 p-phenetldlne
2910 Phenol
2920 Phenolsulfonic adds
2930 Phenyl anthranilic acid
2940 Phenylenediamine
2950 Phosgene
2960 Phthalic anhydride
2970 Phthalimide
2973 s-picoline
2976 Piperazine
OCPDB No. Chemical
3000 Polybutenes
3010 Polyethylene glycol
3025 Polypropylene glycol
3063 Propionaldehyde
3066 Propionic add
3070 n-propyl alcohol
3075 Propylamine
3080 Propyl chloride
3090 Propylene
3100 Propylene chlorohydrln
3110 Propylene dichlorlde
3111 Propylene glycol
3120 Propylene oxide
3130 Pyridlne
3140 Quinone
3150 Resordnol
3160 Resorcylic add
3170 Salicylic acid
3180 Sodium acetate
3181 Sodium benzoate
3190 Sodium carboxymethyl cellulose
3191 Sodium chloroacetate
3200 Sodium formate
3210 Sodium phenate
3220 Sorbic acid
3230 Styrene
3240 Sucdnic acid
3250 Sucdnitrile
3251 Sulfanilic add
3260 Sulfolane
3270 Tannic add
3280 Terephthallc acid
3290 & 3291 Tetrachloroethanes
3300 Tetrachlorophthalic anhydride
3310 Tetraethyllead
3320 Tetrahydronaphthalene
3330 Tetrahydrophthalic anhydride
3335 Tetramethyllead
B-5
-------
OCPDB No. Chemical
3340 Tetramethylenediamine
3341 Tetramethylethylenediamine
3349 Toluene
3350 Toluene-2,4-diamine
3354 Toluene-2,4-diisocyanate
3355 Toluene diisocyanates (mixture)
3360 Toluene sulfonamide
3370 Toluene sulfonic acids
3380 Toluene sulfonyl chloride
3381 Toluidines
3390, 3391 Trichlorobenzenes
& 3393
3395 1,1,1-trichloroethane
3400 1,1,2-trichloroethane
3410 Trichloroethylene
3411 Trichlorofluoromethane
3420 1,2,3-trichloropropane
3430 l,l,2-trichloro-l,2,2-trifluoroethane
3450 Triethylamine
3460 Triethylene glycol
3470 Triethylene gylcol dimethyl ether
3480 Triisobutylene
3490 Trimethylamine
3500 Urea
3510 Vinyl acetate
3520 Vinyl chloride
3530 Vinylidene chloride
3540 Vinyl toluene
3541 Xylene (mixed)
3560 o-xylene
3570 p-xylene
3580 Xylenol
3590 Xylidine
TABLE II: Polymer and Resin Manufacturing Industries
High-density polyethylene
Low-density polyethylene
Polypropylene
Polystyrene
Styrene-butadiene copolymers
Acrylics
Alkyds
Mel amine Formaldehyde
Nylon 6
Nylon 66
Phenol Formaldehyde
Polyester Fibers
Polyvinyl Acetate
Polyvinyl Alcohol
Unsaturated Polyester Resins
Urea Formaldehyde
B-6
-------
APPENDIX C. METHOD 21. DETERMINATION OF VOLATILE ORGANIC
COMPOUND LEAKS
-------
. APPENDIX C
METHOD 21. DETERMINATION OF VOLATILE
ORGANIC COMPOUND LEAKS
1. Applicability and Principle
1.1 Applicability. This method applies to the determination of volatile
organic compound (VOC) leaks from organic process equipment. These sources
include, but are not limited to, valves, flanges and other connections,
pumps and compressors, pressure relief devices, process drains, open-ended
valves, pump and compressor seal system degassing vents, accumulator vessel
vents, and access door seals.
1.2 Principle. A portable instrument is used to detect VOC leaks from
individual sources. The instrument detector is not specified, but it must
meet the specifications and performance criteria contained in paragraph 2.1.
2. Apparatus
2.1 Monitoring Instrument. The monitoring instrument shall be as follows:
2.1.1 Specifications.
"a. The VOC instrument detector shall respond to the organic compounds
being processed. Detectors which may meet this requirement include, but
are not limited to, catalytic oxidation, flame ionization, infrared absorption,
and photoionization.
b. The instrument shall be intrinsically safe for operation in explosive
atmospheres as defined by the applicable U.S.A. Standards (e.g., National
Electrical Code by the National Fire Prevention Association).
C-l
-------
c. The instrument shall be able to measure the leak definition
concentration specified in the regulation.
d. The instrument shall be equipped with a pump so that a continuous
sample is provided to the detector. The nominal sample flow rate shall be
1-3 liters per minute.
e. The scale of the instrument meter shall be readable to ±5 percent
of the specified leak definition concentration.
2.1.2 Performance Criteria. The instrument must meet the following
perfrrmance criteria. The definitions and evaluation procedures for each
parameter are given in Section 4.
2.1.2.1 Response Time. The instrument response time must be 30 seconds
or less. The response time must be determined for the instrument system
configuration to be used during testing, including dilution equipment. The
use of a system with a shorter response time than that specified will reduce
the time required for field component surveys.
2.1.2.2 Calibration Precision. The calibration precision must be less
than or equal to 10 percent of the calibration gas value.
2.1.2.3 Quality Assurance. The instrument shall be subjected to the
response time and calibration precision tests prior to being placed in
service. The calibration precision test shall be repeated every 6 months
thereafter. If any modification or replacement of the instrument detector
is required, the instrument shall be retested and a new 6-month quality
assurance test schedule will apply. The response time test shall be repeated
if any modifications to the sample pumping system or flow configuration is
made that would change the response time.
C-2
-------
2.3 Calibration Gases. The monitoring instrument is calibrated in terms of
parts per million by volume (ppmv) of the compound specified in the applicable
regulation. The calibration gases required for monitoring and instrument
performance evaluation are a zero gas (air, <3 ppmv VOC) and a calibration
gas in air mixture approximately equal to the leak definition specified in
the regulation. If cylinder calibration gas mixtures are used, they must be
analyzed and certified by the manufacturer to be within ±2 percent accuracy.
Calibration gases may be prepared by the user according to any accepted gaseous
standards preparation procedure that will yield a mixture accurate to within
±2 percent. ^Alternative calibration gas species may be used in place of the
calibration compound if a relative response factor for each instrument is
determined so that calibrations with the alternative species may be expressed
as calibration compound equivalents on the meter readout.
3. Procedures
3.1 Calibration. Assemble and start up the VOC analyzer and recorder
according to the manufacturer's instructions. After the appropriate warm-up
period and zero or internal calibration procedure, introduce the calibration
gas into the instrument sample probe. Adjust the instrument meter readout
to correspond to the calibration gas value. If a dilution apparatus is
used, calibration must include the instrument and dilution apparatus assembly.
The nominal dilution factor may be used to establish a scale factor for
converting to an undiluted basis. For example if a nominal 10:1 dilution
apparatuses used, the meter reading for a 10,000 ppm calibration compound would
be set at 1000. During field surveys, the scale factor of 10 would be used
to convert measurements to an undiluted basis.
C-3
-------
: 3.2 Individual Source Surveys.
3.2.1 Case I - Leak Definition Based on Concentration Value. Place the
probe inlet at the surface of the component interface where leakage could
occur. Move the probe along the interface periphery while observing the
instrument readout. If an increased meter reading is observed, slowly probe
the interface where leakage is indicated until the maximum meter reading is
obtained. Leave the probe inlet at this maximum reading location for
approximately two times the instrument response time. If the maximum observed
meter reading is greater than the leak definition in the applicable regulation,
record and report the results as specified in the regulation reporting
requirements. Examples of the application of this general technique to
specific equipment types are:
a. Valves—The most common source of leaks from valves, is at the seal
between the stem and housing. Place the probe at the interface where the
stem exits the packing gland and sample the stem circumference. Also, place
the probe at the interface of the packing gland take-up flange seat and sample
the periphery. In addition, survey valve housings of multipart assembly at
the surface of all interfaces where leaks can occur.
b. Flanges and Other Connections—For welded flanges, place the probe
at the outer edge of the flange-gasket interface and sample around the
circumference of the flange. Sample other types of nonpermanent joints (such
as threaded connections) with a similar traverse.
c. Pumps and Compressors—Conduct a circumferential traverse at the
outer surface of the pump or compressor shaft and seal interface. If the
source is a rotating shaft, position the probe inlet within one centimeter
of the shaft seal interface for the survey. If the housing configuration
C-4
-------
prevents a complete traverse of the shaft periphery, sample all accessible
portions. Sample all other joints on the pump or compressor housing where
leakage can occur.
d. Pressure Relief Devices—The configuration of most pressure relief
devices prevents sampling at the sealing seat interface. For those devices
equipped with an enclosed extension, or horn, place the probe inlet at
approximately the center of the exhaust area to the atmosphere for sampling.
e. Process Drains—For open drains, place the probe inlet at
approximately the center of the area open to the atmosphere for sampling. For
covered drains, place the probe at the surface of the cover interface and
conduct a peripheral traverse.
f. Open-Ended Lines or Valves—Place the probe inlet at approximately
the center of the opening to the atmosphere for sampling.
g. Seal System Degassing Vents and Accumulator Vents—Place the probe
inlet at approximately the center of the opening to the atmosphere for
sampling.
h. Access Door Seals—Place the probe inlet at the surface of the
door seal interface and conduct a peripheral traverse.
3.2.2 Case II-Leak Definition Based on "No Detectable Emission".
a. Determine the local ambient concentration around the source by
moving the probe inlet randomly upwind and downwind at distance of one to
two meters from the source. If an interference exists with this determination
due to a nearby emission or leak, the local ambient concentration may be
determined at distances closer to the source, but in no case shall the distance
be less than 25 centimeters. Note the ambient concentration and then move
the probe inlet to the surface of the source and conduct a survey as described
C-5
-------
in 3.2.1. If a concentration increase greater than 2 percent of the
concentration-based leak definition is obtained, record and report the results
as specified: by the regulation.
b. For those cases where the regulation requires a specific device
installation, or that specified vents be ducted or piped to a control device,
the existence of these conditions shall be visually confirmed. When the
regulation also requires that no detectable emissions exist, visual observations
and sampling surveys are required. Examples of this technique are:
i. Pump or Compressor Seals—If applicable, determine the type of
shaft seal. Perform a survey of the local area ambient VOC concentration
and determine if detectable emissions exist as described in 3.2.2.a.
ii. Seal System Degassing Vents, Accumulator Vessel Vents, Pressure
Relief Devices—If applicable, observe whether or not the applicable ducting
or piping exists. Also, determine if any sources exist in the ducting or
piping where emissions could occur prior to the control device. If the
required ducting or piping exists and there are no sources where the emissions
could be vented to the atmosphere prior to the control device, then it is
presumed that no detectable emissions are present.
4. Instrument Performance Evaluation Procedures
4.1 Definitions.
4.1.1 Calibration Precision. The difference between the average VOC
concentration indicated by the meter readout for consecutive calibration
repetitions and the known concentration of a test gas mixture.
4.1.2 Response Time. The time interval from a step change in VOC
concentration at the input of the sampling system to the time at which
90 percent of the corresponding final value is reached as displayed on the
instrument readout meter.
C-6
-------
4.2 Evaluation Procedures. At the beginning of the instrument performance
evaluation test, assemble and start up the instrument according to the
manufacturer's instructions for recommended warmup period and preliminary
adjustments. If a dilution apparatus is used during field surveys, the
evaluation procedure must be performed on the instrument-dilution system
combination.
4.2.1 Calibration Precision Test. Make a total of nine measurements
by alternately using zero gas and the specified calibration gas. Record the
meter readings (example data sheet shown in Figure 21-1).
4.2.2 Response Time Test Procedure. Introduce zero gas into the
instrument sample probe. When the meter reading has estabilized, switch
quickly to the specified calibration gas. Measure the time from concentration
switching to 90 percent of final stable reading. Perform this test sequence
three times and record the results (example data sheet given in Figure 21-2).
4.3 Calculations. All results are expressed as mean values, calculated
by:
where:
x. = Value of the individual measurements.
1C = Sum of the individual values.
x = Mean value.
n = Number of measurements.
C-7
-------
Instrument ID
Calibration Gas Data
Calibration = ppmv
Run Instrument Meter Difference
No. Reading, ppm ppm
1.
2.
3.
4.
5.
6.
7.
8.
9.
(1)
Mean Difference
Mean Difference
caltlonaconcentratlon
Calibration Precision =
IT)Calibration Gas Concentration - Instrument Reading
Figure C-l. Calibration Precision Determination
C-8
-------
Instrument ID
Calibration Gas Concentration
90% Response Time:
1. Seconds
2. Seconds
3. Seconds
Mean Response Time Seconds
Figure C-2. Response Time Determination
C-9
-------
APPENDIX D. EXAMPLE CALCULATIONS FOR DETERMINING REDUCTION IN
EMISSIONS FROM IMPLEMENTATION OF RACT
-------
APPENDIX D
Example Calculations for Determining Reduction in Emissions from
Implementation of RACT
The purpose of this appendix is to provide emission factors and procedures
for computing the emission reduction associated with reasonably available
control technology (RACT) for control of fugitive emissions in the synthetic
organic chemical, polymer, and resin manufacturing industries. The emission
factors and procedures allow the reader to compute total process unit uncontrolled
emissions, uncontrolled emissions from equipment affected by RACT, controlled
emissions from equipment affected by RACT, and the emission reduction from
implementing RACT. The emission factors and computation procedures in this
appendix should be used only where specific plant fugitive emission data are
not available.
Uncontrolled emission factors for each type of component are presented
in Table D-l. The uncontrolled emissions from a process unit can be computed
by multiplying the uncontrolled emission factor for each component type by
the number of components of the given type, then summing the emissions from
all component types.
The average emissions from components controlled under RACT (pumps in
light liquid service, compressors, valves in gas service and light liquid
service, and pressure relief valves in gas service) following implementation
of RACT are presented in Table D-2. The basis for these factors has already
been discussed in Chapters 3 and 4.
The number of components in a process unit may be estimated (at the
Director's discretion) or derived from actual count. A sample calculation
of uncontrolled emissions from an illustrative process unit is presented
below in Table D-3.
D-l
-------
TABLE D-l. UNCONTROLLED FUGITIVE EMISSION FACTORS
IN PROCESS UNIT EQUIPMENT3
Uncontrolled
Fugitive emission factor,
emission source kg/hr
Pumps
Light liquids 0.12
Heavy liquids 0.020
Valves (in-line)
Gas service 0.021
Light liquid service 0.010
Heavy liquid service 0.0003
Pressure relief valves
Gas service 0.16
Light liquid service 0.006
Heavy liquid service 0.009
Open-ended valves
Gas service 0.025
Light liquid service 0.014
Heavy liquid service 0.003
Flanges 0.0003
Sampling connections 0.015
Compressors 0.44
aFrom Table 2-2.
TABLE D-2. CONTROLLED EMISSION FACTORS FOR EQUIPMENT
AFFECTED BY RACT
Controlled
Fugitive emission factor,
emission source kg/hr
Pumps in light liquid service 0.036
Valve in gas service 0.003
Valve in light liquid service 0.003
Gas pressure relief valves 0.066
Compressors 0.123
aFrom Table 4-3.
D-2
-------
TABLE D-3. EXAMPLE CALCULATION OF UNCONTROLLED EMISSIONS FROM
AN ILLUSTRATIVE PROCESS UNIT
Fugitive emission source
Pumps
Light liquid service
Heavy liquid service
In-line valves
Gas service
Light liquid service
Heavy liquid service
Pressure relief valves
Gas service
Light liquid service
Heavy liquid service
Open-ended valves
Gas service
Light liquid service
Heavy liquid service
Compressors0
Sampling connections
Flanges
Number
of sources
(N)
16
10
j
150*
120dd
100d
17
2
3
12
19
17
2
34
750d
Emission
factor .
kg/hr-source
(E)
0.12
0.020
0.021
0.010
0.0003
0.16
0.006
0.009
0.025
0.014
0.003
0.44
0.015
0.0003
Total emissions
Emissions from
source, kg/hr
(N X E)
1.92
0.2
3.15
1.2
0.03
2.72
0.012
0.027
0.3
0.266
0.051
0.88
0.51
0.225
= 11.49 kg/hre
Determined by actual count or by estimate.
bFrom Table D-l.
clnspected under RACT.
Because of the large number of components, an estimate may be appropriate.
eThe expected annual emissions are 100 65 Mg/yr (8,760 hr/yr).
[-3
-------
The overall emission reduction associated with RACT can be computed by
finding the emission reduction for those process units maintained under
RACT.. The emission reduction is the difference between the emissions expected
from components in the unrepaired (uncontrolled) condition and in the repaired
(controlled) condition. The uncontrolled emissions are presented in
Table D-4:
TABLE D-4. UNCONTROLLED EMISSIONS FROM COMPONENTS AFFECTED BY RACT.
Emission source
Light liquid pump
Gas service valve
Light liquid service valve
Gas pressure; relief valve
Compressors
Number of
sources
(N)
16
150
120
17
2
Emission
factor
kg/hr-source
(E)
0.12
0.021
0.010
0.16
0.44
Emissions from
sources, kg/hr
(N X E)
1.92
3.15
1.2
2.72
0.88
9.87
The emissions expected after implementing RACT are also presented in
Table D-5:
TABLE D-5. CONTROLLED EMISSIONS FROM COMPONENTS AFFECTED BY RACT.
Emission source
Light liquid pump
Gas service valve
Light liquid service valve
Gas pressure relief valve
Compressors
Number
of sources
(N)
16
150
120
17
2
Emission
factor
kg/hr-source
(E)
0.036
0.003
0.003
0.066
0.123
Emissions, from
sources, kg/hr
(N X E)
0.576
0.45 .
0.36
1.12
0.246
2.75
From Table D-2.
D-4
-------
The difference between emissions before and after the implementation of
RACT can be expressed as an annual emission reduction. This emission
reduction is presented in Table D-6:
. TABLE D-6. EMISSION REDUCTION.EXPECTED FROM RACT
1. Uncontrolled emissions from process unit 100.65 Mg/yr
(11.49 kg/hr)
2. Uncontrolled emissions from components affected 86.46 Mg/yr
by RACT (9.87 kg/hr)
3. Controlled emissions from components 24.09 Mg/yr
affected by RACT (2.75 kg/hr)
4. Total emission reduction:
86.46 Mg/yr - 24.09 Mg/yr = 62.35 Mg/yr (7.12 kg/hr)
5. Percent emission reduction:
62.35 Mg/yr
100.65 Mg/yr
D-5
------- |