United States Off ice of Air Quality EM B Report No. 80-OSP-2
Environmental Protection Planning and Standards June 198'
Agency Research Triangle Park NC 27711
v>EPA
On-Shore Production of
Crude Oil and Natural Gas
Fugitive Volatile Organic
Compound Emission Sources
Emission Test Report
AMOCO Production Company
Hastings Plant
Brazoria County, Texas
-------
DCN 81-222-018-04-28 EMB Report No. 80-OSP-2
EMISSION TEST REPORT
FUGITIVE VOC TESTING
AT THE
AMOCO HASTINGS GAS PLANT
Prepared by:
G. E. Harris
Radian Corporation
8501 Mo-Pac Boulevard
Austin, Texas 78759
Prepared for:
Winton Kelly
U. S. Environmental Protection Agency
ESED/EMB (MD-13)
Research Triangle Park, North Carolina 27711
EPA Contract No. 68-02-3542
Work Assignment No. 4
July 1981
-------
CONTENTS
Section Page
1 INTRODUCTION 1
2 SUMMARY OF RESULTS 2
3 PROCESS DESCRIPTION 5
4 TESTING METHODOLOGY 8
11
-------
SECTION 1
INTRODUCTION
This report presents the results of testing for fugitive VOC (Volatile
Organic Compounds) emissions at the Amoco Hastings gas plant. The testing
was performed by Radian Corporation on October 14 through October 17, 1980.
This work was funded and administered by the Emission Measurement Branch
of the U.S. Environmental Protection Agency. The purpose of this testing
was to develop data to be used in support of New Source Performance Standards
for onshore production facilities.
The testing described in this report consisted of a screening survey
using two fugitive emission detection methods, EPA Method 21 using portable
analyzers and soap scoring. The objectives of this testing were to:
1) Determine leak frequency by each method, and
2) Collect comparative data on each method so that emission
data from other sources could be used to support New
Source Performance Standards.
The following sections present a summary of results, a description of
the process configuration, and the testing methodology. A full listing of
the data and other supplemental information are included in the appendices.
-------
SECTION 2
SUMMARY OF RESULTS
This section presents a summary of the screening data. A full data
listing is included as Appendix A.
The gas plant screening results are summarized in Tables 2-1 and 2-2.
Table 2-1 presents the distribution of VOC concentration readings for each
source type, while Table 2-2 presents similar data on soap scores.
It should be noted that the source type called flanges actually includes
a variety of pipe-to-pipe connections including threaded fittings, unions,
and compression-type tubing fittings. Welded joints were not included in
this survey. The "other" category represents a group of sources that were
too few in number to warrant separate listing. Included in the "other"
category were sight glasses, vacuum breakers, meters, pig traps, control
valve diaphragm vents, and thermowells.
-------
TABLE 2-1. SUMMARY OF RESULTS. AMOCO HASTINGS PLANT
VOC CONCENTRATION OCCURRENCE DISTRIBUTION
Source Type
Screening Process Open Relief Pump
Value Flanges Drains Ended Lines Valves Valves Seals Compressors Other
(ppmv) # % # % // %//%//%//% //%//%
0 to 199 264 94.0 2 40.0 44 67.7 6 85.7 323 63.8 5 55.6 0 0.0 13 86.7
200 to 11 3.9 2 40.0 9 13.8 0 0.0 98 19.4 0 0.0 0 0.0 0 0.0
9,999
>= 10,000 6 2.1 1 20.0 12 18.5 1 14.3 85 16.8 4 44.4 4 100.0 2 13.3
Total Sources
Screened 281 18.2 5 100.0 65 97.0 7 53.8 506 89.6 9 64.3 4 50.0 15 100.0
Sources Not
Screened 1264* 81.8 0 0.0 2 3.0 6 46.2 59 10.4 5 35.7 4 50.0 0 0.0
Total
Sources 1545* 5 67 13 565 14 8 15
// - Number of sources
% - Percent of total sources screened
* - Estimated value - e-very fifth flange was surveyed
-------
TABLE 2-2. SUMMARY OF RESULTS: AMOCO HASTINGS PLANT
SOAP SCORING OCCURRENCE DISTRIBUTION
Source
Soap Score
0
1
2
3
4
Flanges
# %
254 94.8
2 0.7
3 1.1
5 1.9
4 1.5
Process
Drains
0
0
0
0
0
Open
Ended Lines
19
1
1
0
4
76.0
4.0
4.0
0.0
16.0
Type
Relief Pump
Valves Valves Seals Compressors
#%#%//% # %
3
0
0
0
0
100.0 317
0.0 18
0.0 44
0.0 44
0.0 66
64.8 1 25.0
3.7 0 0.0
9.0 1 25.0
9.0 0 0.0
13.5 2 50.0
0
0
0
0
0
Other
// %
14
0
0
0
1
93.3
0.0
0.0
0.0
6.7
Total Sources
Screened
Sources Not
Soaped
Total
Sources
268 17.3
1277* 82.7
1545*
0 0.0
5 100.0
5
25
42
67
37.3
62.7
3
10
13
23.1 489
76.9 76
565
86.5 4 28.6
13.5 10 71.4
14
0 0.0
8 100.0
8
15
0
15
100.0
0.0
# - Number of sources
% - Percent of total sources screened
* - Estimated value - every fifth flange was surveyed
-------
SECTION 3
PROCESS DESCRIPTION
The Amoco Hastings gas plant removed natural gas liquids by the cryogenic
separation principle. The feed gas to the unit was a mixture of lift gas* and
newly produced gas from area oil wells. Natural gas liquids were removed at the
plant and separated into an ethane/propane stream, which was transported via
pipeline to a chemical plant, and a butane-plus gasoline stream which was routed
to a refinery. The treated gas stream (primarily methane) was split between
lift gas and pipeline sales gas. The cryogenic unit was operating at its rated
capacity of about 30 MMSCFD (million standard cubic feet per day) during the
testing. The total plant capacity, including the gas by-passed directly to
lift-gas, was about 76 MMSCFD.
A simplified schematic of the process is shown in Figure 3-1. The raw
feed gas was compressed and then chilled by heat exchange, propane refrigeration,
and expanded at the Turbo expander. Condensed liquids were separated and
routed to the depropanizer and split into the ethane/propane stream overhead
and the butane-plus gasoline stream as the bottom product.
Fugitive emission testing was performed on all facilities which were
considered to be an integral part of the cryogenic separation unit, in-
cluding heat exchangers, chillers, expansion/compression turbines, and the
distillation columns. In addition, the ethane/propane and propane refrigera-
tion compressor area, the ethane/propane metering area, and the product and
*"Lift Gas" is defined as dehydrated and recompressed natural gas, which is then
returned to the area oil wells to artifically lift the oil to the surface by
decreasing the density of the column of fluid. (This is an alternate method to
pumping produced fluids that will not flow to the surface by natural reservoir
pressures).
-------
RAW FEED
GAS
1
NOT
TESTED'
• TESTED
LI FT GAS
COMPRESSION
6
PROPANE
COMPRESSION
RECOMPRESSION
| TURBINE
ETHANE/PROPANE
PRODUCT
i
ETHANE/PROPANE
COMPRESSOR
DEMETHANIZER
EXPANSION
TURBINE
LIFT SALES
GAS GAS
REFRIGERANT
AND PRODUCT
STORAGE
T
DEPROPANIZER
NOT
TESTED'
TESTED
BUTANE-PLUS
GASOLINE
70-2097-1
Figure 3-1. Simplified flow diagram.
-------
refrigerant storage area were tested. The data were recorded so as to dis-
tinguish between these facilities by assigning a unit number to each area:
Unit 1 - Cryogenic unit,
Unit 2 - Ethane/propane and refrigerant compressor area,
Unit 3 - Ethane/propane metering facilities, and
Unit 4 - Product and refrigerant storage.
The lift gas compressors and the methane recompressors were not tested.
-------
SECTION 4
TESTING METHODOLOGY
The fugitive emissions testing at this site was limited to "screening."
Screening is a generic term covering any quick, portable method of detecting
fugitive emissions. Two screening methods were used in parallel on this task,
instrumental screening (using the Century Systems OVA-108) and soap scoring.
The instrumental screening was done according to the procedures specified
in EPA Proposed Method 21,l which is included as Appendix B. The instrument
performance evaluations are included as Tables 4-1 and 4-2. Method 21 only
requires the exact concentration to be recorded if it is over the leak definition
specified in the applicable standard, but since this effort was more oriented
to standards support than to regulatory monitoring, the maximum screening value
was recorded for all sources.
The soap scoring method was modeled after a method used in screening fugitive
2
emissions from petroleum production facilities. The soap solution was prepared
from 100 ml. of rug shampoo (HR Professional Formula) mixed with a gallon of
either distilled water or a mixture of distilled water and ethylene glycol.
The solution was applied using a common garden sprayer.
Each source was sprayed with soap solution, being sure to coat all areas
of potential leakage. A careful inspection was then conducted to detect any
bubble formation. A soap score was then assigned based on the estimated bubble
volume generated in a six-second observation:
federal Register, v46 n2 Monday Jan. 5, 1981, pp. 1160.
zEaton, W.S., et al. "Fugitive Hydrocarbon Emissions from Petroleum Production
Operations." API Publication No. 4322, American Petroleum Institute (1980).
-------
TABLE 4-1
CALIBRATION ERROR DETERMINATION
Instrument ID
Centurv Svsf-pms OVA-108
Serial Number: 2158
Run
No.
1.
2.
3.
4.
5.
6.
7.
a.
9.
Mean Difference
Calibration Gas Data
Calibration = "990 ppmv
Instrument Meter Di
Reading, ppm
8000
8200
3000
8000
8000
8400
8100
8500
8200
(2}
Calibration Error - Mean Differencev ' .. 1rtr>
^Calibration
Calibration Gas Concentration " >ww
Gas Concentration - Instrument Reading
fference^ '
ppm
-10
-210
-10
-10
-10
-410
-110
-510
-210
-166
-2.1
-------
TABLE 4-2
RESPONSE TIME DETERMINATION
Instrument ID
Century Systems OVA-108
Serial Number: 2158
Calibration Gas Concentration
7990 ppmv
1-9-81
90% Response Time:
Without Dilution Probe
1. 5.8
2. 7.0
3. 5.5
Response Time
Seconds
Seconds
Seconds
6.1
With Dilution Probe
7.1 Seconds
9.5 Seconds
7.0 Seconds
Seconds
7.8 Seconds
10
-------
Soap Score Estimated Bubble Volume
0 No detectable bubbles
1 0 to 1 cc/6 sec
2 1 to 10 cc/6 sec.
3 10 to 100 cc/6 sec.
4 >100 cc/6 sec.
The screening methods outlined above were used on every accessible
source except for flanges. Approximately 20 percent of the flanges were
screened because of their large population. Sources screened included valves,
flanges, pumps, compressors, open-ended lines, drains, relief valves, and
other miscellaneous sources. The survey was conducted on a line-by-line basis
to minimize the time required to obtain process data, such as the composition
and phase of the material in the line. For those sources that were not screened
due to either physical inaccessibility or safety problems which prevented close
approach, entries were recorded on the data sheets to insure that a complete
source inventory was obtained.
11
------- |