EPA-650/4-74-020
June 1974
Environmental Monitoring Series
33
V
O
UJ
-------
EPA-650/4-74-020
DEVELOPMENT OF TECHNICAL SPECIFICATIONS
FOR STANDARD GAS-DILUENT MIXTURES
FOR USE IN MEASUREMENT
OF MOBILE SOURCE EMISSIONS
by
Louis R. Reckner
Scott Research Laboratories, Inc.
Plumsteadville, Pennsylvania 18949
Contract No. 68-02-0652
Program Element No. 1HA327
ROAPNo. 26ADZ
EPA Project Officer: John H. Margeson
Quality Assurance and Environmental Monitoring Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D . C. 20460
June 1974
-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
Publication No. EPA-650/4-74-020
ii
-------
ill
SRL 1317 13 0674
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1-1
2.0 TEST PROGRAM 2-1
2.1 DESIGN OF TEST PROGRAM 2-1
2.2 DESCRIPTION OF TEST VARIABLES 2-4
2.3 ANALYTICAL PROCEDURES 2-5
2.4 ANALYTICAL SCHEDULE 2-6
3.0 RESULTS 3-1
3.1 ANALYSES OF TEST CYLINDERS 3-1
4.0 DATA ANALYSIS 4-1
4.1 DETERMINATION OF DETERIORATION RATE 4-1
4.2 REGRESSION ANALYSIS OF STABILITY DATA 4-3
4.3 SUMMARY OF EFFECTS OF TEST VARIABLES ON GAS
STABILITY 4-19
4.4 EFFECT OF EXTREME STORAGE TEMPERATURES ON
GAS STABILITY 4-20
4.5 EFFECT OF CYLINDER PRESSURE ON GAS STABILITY 4-24
4.6 EVALUATION OF THE EFFECT OF REGULATOR TYPE 4-25
5.0 DISCUSSION OF RESULTS 5-1
5.1 STABILITY OF CARBON MONOXIDE MIXTURES 5-1
5.2 STABILITY OF PROPANE MIXTURES 5-2
5.3 STORAGE OF CARBON MONOXIDE AND PROPANE MIXTURES 5-3
5.4 GENERAL COMMENTS 5-3
6.0 RECOMMENDATIONS 6-1
6.1 CARBON MONOXIDE IN NITROGEN MIXTURES 6-1
6.2 PROPANE IN AIR MIXTURES 6-2
7.0 ACKNOWLEDGEMENTS 7-1
8.0 REFERENCES 8-1
SCOTT RESEARCH LABORATORIES. INC.
-------
IV
SRL 1317 13 0674
LIST OF ILLUSTRATIONS
FIGURE 4-1 CONCENTRATION - TIME TREND FOR CARBON MONOXIDE
@ 50 PPM NOMINAL CONCENTRATION
Pace
4-2
LIST OF TABLES
TABLE 2-1 DESIGN OF VARIABLES FOR EACH OF TEN COMPONENT/
CONCENTRATION CYLINDER SETS 2-2
TABLE 2-2 DESIGNATION OF CONTROL CYLINDERS 2-3
TABLE 2-3 PROCEDURE FOR CARBON MONOXIDE ANALYSIS 2-7
TABLE 2-4 PROCEDURE FOR PROPANE ANALYSIS 2-7
TABLE 3-1 ANALYTICAL DATA FOR CARBON MONOXIDE AT 10 PPM 3-2
TABLE 3-2 ANALYTICAL DATA FOR CARBON MONOXIDE AT 50 PPM 3-3
TABLE 3-3 ANALYTICAL DATA FOR CARBON MONOXIDE AT 100 PPM 3-4
TABLE 3-4 ANALYTICAL DATA FOR CARBON MONOXIDE AT 500 PPM 3-5
TABLE 3-5 ANALYTICAL DATA FOR CARBON MONOXIDE AT 1000 PPM 3-6
TABLE 3-6 ANALYTICAL DATA FOR PROPANE AT 3 PPM 3-7
TABLE 3-7 ANALYTICAL DATA FOR PROPANE AT 10 PPM 3-8
TABLE 3-8 ANALYTICAL DATA FOR PROPANE AT 50 PPM 3-9
TABLE 3-9 ANALYTICAL DATA FOR PROPANE AT 100 PPM 3-10
TABLE 3-10 ANALYTICAL DATA FOR PROPANE AT 500 PPM 3-11
TABLE 4-1 PERCENT CHANGE IN CARBON MONOXIDE CONCENTRATION
OVER A SIX-MONTH PERIOD 4-4
TABLE 4-2 PERCENT CHANGE IN PROPANE CONCENTRATION
OVER A SIX-MONTH PERIOD 4-5
TABLE 4-3 VALUES ASSIGNED TO VARIABLES IN MULTIPLE
REGRESSION MODEL 4-6
TABLE 4-4 EXAMPLE OF REGRESSION MATRIX FOR CARBON MONOXIDE
AT 50 PPM NOMINAL CONCENTRATION 4-6
TABLE 4-5 ANALYSIS ON STABILITY OF CARBON MONOXIDE
AT 10 PPM 4-9
TABLE 4-6 ANALYSIS ON STABILITY OF CARBON MONOXIDE
AT 50.PPM 4-10
TABLE 4-7 ANALYSIS ON STABILITY OF CARBON MONOXIDE
AT 100 PPM 4-11
SCOTT RESEARCH LABORATORIES. INC.
-------
SRL 1317 13 0674
LIST OF TABLES (CONTINUED)
Page
TABLE 4-8 ANALYSIS ON STABILITY OF CARBON MONOXIDE
AT 500 PPM 4-12
TABLE 4-9 ANALYSIS ON STABILITY OF CARBON MONOXIDE
AT 1000 PPM . 4-13
TABLE 4-10 ANALYSIS ON STABILITY OF PROPANE
AT 3 PPM 4-14
TABLE 4-11 ANALYSIS ON STABILITY OF PROPANE
AT 10 PPM 4-15
TABLE 4-12 ANALYSIS ON STABILITY OF PROPANE
AT 50 PPM 4-16
TABLE 4-13 ANALYSIS ON STABILITY OF PROPANE
AT 100 PPM 4-17
TABLE 4-14 ANALYSIS ON STABILITY OF PROPANE
AT 500 PPM . 4-18
TABLE 4-15 t-TEST ON HOT-CYCLE EFFECT FOR PROPANE AT 100 PPM
NOMINAL CONCENTRATION 4-21
TABLE 4-16 SUMMARY OF HOT AND COLD STORAGE EFFECTS ON
PROPANE STABILITY 4-23
TABLE 4-17 SUMMARY OF HOT AND COLD STORAGE EFFECTS ON
CARBON MONOXIDE STABILITY 4-23
TABLE 4-18 EFFECT OF REDUCED CYLINDER PRESSURE ON STABILITY
OF TRACE GASES 4-26
TABLE 4-19 COMPARISON OF REGULATOR TYPES IN ANALYSIS OF
CARBON MONOXIDE 4-27
TABLE 4-20 COMPARISON OF REGULATOR TYPES IN ANALYSIS OF
PROPANE 4-29
SCOTT RESEARCH LABORATORIES. INC.
-------
vl
SET 1317 13 0674
ABSTRACT
The stability of gas mixtures of carbon monoxide in nitrogen
and propane in air in high-pressure cylinders was investigated in 240 test
cylinders over a six month period. The effect of several variables related
to the preparation of the mixtures by the suppliers, the storage of the
cylinders and their use by laboratories engaged in emissions measurements
was studied. The variables included:
1. Cylinder wall material
2. Cylinder valve type
3. Cylinder preconditioning procedure
4. Concentration of carbon monoxide and propane
5. Purity of diluent nitrogen and air
6. Mixing procedure after blending
7. Temperature at which cylinders are stored
8. Cylinder pressure
9. Type of pressure-reducing regulator used
The concentration data obtained by periodic analysis of the 240
cylinders over the six month period were subjected to statistical analysis
by multiple stepwise regression. The effects of the individual variables
are discussed, and recommended practices for assuring stable mixtures of
carbon monoxide in nitrogen and propane in air are presented.
SCOTT RESEARCH LABORATORIES. INC.
-------
1-1
SRL 1317 13 0674
1.0 INTRODUCTION
This report describes the work performed by Scott Research
Laboratories, Inc. under EPA Contract No. 68-02-0652, "Development of
Technical Specifications for Standard Gas-Diluent Mixtures for Use in
Measurement of Mobile Source Emissions." The primary objective of this
program was to develop technical specifications for producing stable gas
mixtures of propane in air and carbon monoxide in nitrogen in high pressure
gas cylinders. A stable mixture was defined as one in which the con-
centration of trace gas changed by no more than 3.0% from its original
concentration over a six month test period.
In carrying out this project, Scott investigated the effect of
variables related to the preparation of the gas mixtures by the suppliers
as well as those related to the storage and use by laboratories engaged
in emissions measurements. The first category included hardware type,
preparation procedures and grade of diluent gas. In the latter category
were hot and cold storage and cylinder regulator type.
This study was accomplished by utilizing 240 new cylinders
purchased from the leading manufacturer of steel cylinders. The
experimental design involved five concentration levels each of propane
and carbon monoxide, two cylinder types, two valve types, three procedures
for cylinder preparation, two mixing methods and three grades of diluent
gas. Two thirds of the cylinders were subjected to hot and cold storage
cycles with pressure reduction between each cycle. The remaining one-
third served as controls. Each cylinder was analyzed by gas chromatography
on a regular schedule.
The concentration data for each of the ten component/concentra-
tion level combinations were subjected to statistical analyses to estimate
the effect of each variable on the concentration stability. This infor-
mation was used to draw up recommendations for manufacturing and use
specifications to assure stable mixtures.
SCOTT RESEARCH LABORATORIES. IMC.
-------
2-1
SRL 1317 13 0674
2.0 TEST PROGRAM
2.1 DESIGN OF TEST PROGRAM
The test program was designed to study the effect of various
material, cylinder preparation procedures and use procedures on the
stability of five concentration levels each of carbon monoxide in nitrogen
and propane in air. It was decided that 240 test cylinders, 120 each for
carbon monoxide and propane, should be adequate to evaluate the effect of
the variables and keep the magnitude of effort within practical limits.
An examination of procedures followed by the various manufacturers of
calibration gas mixtures resulted in the selection of five parameters at
two or three levels each. These parameters and levels were believed to
represent those materials and procedures in general use which could
affect the stability of the mixtures.
For each of the ten component/concentration combinations, 24
cylinders were prepared according to the design given in Table 2-1. A
detailed description of the variables is presented in the next section.
The various levels of each variable were distributed among the 24 cylinders
so as to produce an approximately balanced design which would facilitate
subsequent data analysis.
Other variables were also studied. The effect of hot and cold
storage was determined by exposing two-thirds or 16 of each set of 24
cylinders to elevated and lowered temperatures while maintaining the
remaining one-third at constant ambient temperature. The latter group
were termed control cylinders and they are identified in Table 2-2. The
remaining 160 cylinders were each subjected to two 2-week storage periods
at 95°F and two similar periods at 0°F. The storage sequence was either
hot-cold-hot-cold or cold-hot-cold-hot. Approximately three weeks were
allowed between successive storage periods.
It was also desired to determine the effect of reducing cylinder
pressure on stability. For this purpose each of the cylinders subjected
to hot and cold storage was reduced by 500 psi after each storage. Thus,
the storage pressures were 2000 psi, 1500 psi, 1000 psi and 500 psi during
the four successive storage periods. The control cylinders were maintained
at approximately 2000 psi throughout the program.
SCOTT RESEARCH LABORATORIES. INC.
-------
TABLE 2-1 DESIGN OF VARIABLES FOR EACH OF TEN
COMPONENT/CONCENTRATION CYLINDER SETS
V)
w
o
EC
r-
O
5"
in
Cyl.
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Cyl.
Type
2
2
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
Valve
Type
1
1
1
1
1
1
2
2
2
2
2
2
1
1
1
1
1
1
2
2
2
2
2
2
Pre-
conditioning
1
2
3
1
2
3
1
2
3
1
3
2
1
2
3
1
2
3
1
2
3
1
2
3
Mixing
1
1
1
2
2
2
2
1
2
1
1
2
1
1
1
2
2
2
1
1
1
2
2
2
Diluent
Purity
1
1
1
1
1
1
2
2
2
3
3
3
2
3
2
3
2
3
1
1
1
1
1
1
KEY TO VARIABLES £
Cylinder Type ^
1. Chrome-moly steel per DOT 3AA2015 £
2. Manganese Steel per DOT 3A2015 0
Valve Type *
1. Packed brass with Teflon packing
2. Diaphragm packless brass
Preconditioning
1. Evaculation only
2. Evaculation + nitrogen flush 4- evacuation
3. Same as 2 with heat applied
Mixing
1. Mechanical
2. Thermal
to
i
Diluent Purity
Air (C-^Ha)
1. Blended HC free
2. Blended N -0
3. Water pumped
Nitrogen (CO)
99.997% pure
99.7% pure; low moisture
99.7% pure; high moisture
-------
2-3
SRL 1317 13 0674
TABLE 2-2 DESIGNATION OF CONTROL CYLINDERS
Cyl
No.*
1
2
3
4
5
6
Cone
A,B
C,D
E
A,D
C,E
A
Cyl
No.*
7
8
9
10
11
12
Cone
B,D
C,E
D
A,E
B,C
B
Cyl
No.*
13
14
15
16
17
18
Cone
B,E
C
D
A,C
B
A,B
Cyl
No.*
19
20
21
22
23
24
Cone
D,E
C,D
A
B,C
A,E
D,E
*See Table 2-1
KEY TO CONCENTRATION LEVELS
Cone.
A
B
C
D
E
CO , ppm
10
50
100
500
1000
Propane, ppm
3
10
50
100
500
SCOTT RESEARCH LABORATORIES. INC.
-------
2-4
SRL 1317 13 0674
A final study involved the effect of cylinder regulator type.
This part of the program is discussed in Section 4.6.
2.2 DESCRIPTION OF TEST VARIABLES
2.2.1 Cylinder Type
The two types of cylinders accounting for the large majority of
carbon monoxide and propane mixtures in use are chrome-moly steel and
manganese steel. The chrome-moly cylinders are manufactured to DOT
Specification 3AA2015 as detailed in the Code of Federal Regulations
(CFR) Title 49, Paragraph 178.37. The manganese steel cylinders are
manufactured to DOT Specification 3A2015 as detailed in Paragraph 178.36
of the above CFR. All 240 test cylinders were purchased from the Taylor-
Wharton Division of Harrisburg Steel Company, the leading supplier of such
type cylinders. Prior to delivery to Scott, the cylinders were hydro-
statically tested and steam cleaned by the manufacturer, as is the general
practice.
2.2.2 Cylinder Valve Type
The two types of cylinder valves used were packed brass with
Teflon packing and diaphragm packless brass. The packed brass valves
were Superior Valve Co. Model 1200-E2. The packless brass were Superior
Model 1250-E2.
2.2.3 Cylinder Preconditioning
The three types of preconditioning used were:
1. Evacuation only - evacuate cylinder as received to <0.1" Hg
absolute.
2. Evacuation + nitrogen flush + evacuation - evacuate cylinder
to <0.1" Hg, fill with pure nitrogen to 200 psi, evacuate to
<0.1" Hg.
3. Same as 2 with heat applied - evacuate cylinder heated to 150°F
for 15 minutes to <0.1" Hg, fill with pure nitrogen to 200 psi,
reheat to 150°F for 15 minutes while evacuating to <0.1" Hg.
SCOTT RESEARCH LABORATORIES, INC.
-------
2-5
SRL 1317 13 0674
2.2.4 Mixing of Contents After Blending
Mixing is accomplished commercially by mechanical or thermal
procedures or a combination of the two. It is of greater importance where
high concentrations of low volatility gases are included in a mixture
than in the current test program where the concentrations were relatively
low, and carbon monoxide and propane both have relatively high vapor
pressures.
Mechanical mixing was accomplished by placing the cylinders on
rollers for 15 minutes. Thermal mixing was performed by heating the
cylinder bottoms on a heated platen for \ hour. The cylinder contents
reached approximately 125 F.
2.2.5 Diluent Purity
Three levels of diluent purity were selected for both air and
nitrogen. They are representative of the grades of diluent in current
commercial use.
The nitrogen diluents were:
1. 99.997+ % (from liquid storage) dewpoint - -105°F.
2. 99.7% low moisture - dewpoint = -75°F.
3. 99.7% high moisture - dewpoint - -67°F.
Oxygen is the major impurity present in nitrogen of all grades. The air
diluents were:
1. Blended hydrocarbon-free synthetic air (0 = 20.4%, NZ - 79.6%)
THC - 0.1 ppm, dewpoint = -105°F.
2. Blended synthetic air (02 = 20.4%, N2 = 79.6%), THC « 5 ppm-C,
dewpoint = -85°F.
3. Water pumped air, THC = 3 ppm-C, dewpoint « -75 F.
2.3 ANALYTICAL PROCEDURES
All test cylinders were analyzed by gas chromatography. Carbon
monoxide was determined using a Varian Aerograph Model 1520 with a helium
ionization detector at the conditions shown in Table 2-3. Concentrations
were calculated by comparing the peak heights of the standard and test
cylinders.
SCOTT RESEARCH LABORATORIES, INC.
-------
2-6
SRL 1317 13 0674
Propane was determined using a Perkin-Elmer Model 900 with a
flame ionization detector. The instrument conditions are given in Table
2-4. Concentrations were calculated from peak areas measured by an
Infotronics Model CRS-100 Integrator. Total hydrocarbon analyses were
obtained with a Beckman Model 108A Flame Ionization Detector.
Except for the different instruments, the schemes for analysis
of propane and carbon monoxide were identical. Each test cylinder at a
particular level was analyzed using a single standard cylinder of CO in N
or propane in air as the reference. Each analysis consisted of three to
five injections of gas from each of the test and standard cylinders. The
two cylinders were sampled on an alternate basis to eliminate any possibility
of errors due to time trends in instrument response.
The cylinder standards were rechecked every two to four weeks
by comparison to primary standards in glass flasks. The primary standards
were prepared by a procedure used by Scott in analyzing close tolerance
gas mixtures sold commercially. This procedure has been shown to have a
high degree of accuracy and repeatability. The specific details are
proprietary. The procedure basically involves the injection of a known
volume of trace gas of known purity from a calibrated gas-tight syringe
into a calibrated 5-liter glass flask. The glass flask is then pressurized
to a measured pressure of approximately 0.5 atmosphere gauge.
The overall analysis scheme produced an accuracy of ±1% for
propane and ±2% for carbon monoxide. The error for carbon monoxide was
higher because the carbon monoxide peak fell on the tail of the nitrogen
peak. This necessitated construction of a baseline which increased
error especially at low concentration levels.
2.4 ANALYTICAL SCHEDULE
Each cylinder was analyzed within several days after preparation
and at various intervals thereafter over the six month test period. The
control cylinders were analyzed five to seven times and the others 11 to 15
times. This included analyses before and after each hot or cold storage
period. The analyses after hot and cold storage were performed within
several hours of removal from storage.
SCOTT RESEARCH LABORATORIES. INC.
-------
2-7
SRL 1317 13 0674
TABLE 2-3 PROCEDURE FOR CARBON MONOXIDE ANALYSIS
Instrument: Varian Aerograph Model 1520
Detector: Helium lonization
Column: 10' Molecular Sieve
Temperature: 75°C
Carrier Gas: Helium @ 50 psi
Sample Size: 0.25 to 1 cc
TABLE 2-4 PROCEDURE FOR PROPANE ANALYSIS
Instrument: Perkin-Elmer Model 900
Detector: Flame lonization
Column: 9* x 1/8" Porapak Q
Temperature: 150°C
Carrier Gas: Helium @ 50 psi
Flame Gas: Hydrogen @ 20 psi
Combustion Gas: Oxygen @ 50 psi
Sample Size: 2 cc
SCOTT RESEARCH LABORATORIES. INC.
-------
3-1
SRL 1317 13 0674
3.0 RESULTS
3.1 ANALYSES OF TEST CYLINDERS
The analytical data for the 120 carbon monoxide test cylinders
are shown in Tables 3-1 through 3-5. The corresponding data for the
propane test cylinders are shown in Tables 3-6 through 3-10. The cylinder
age is the number of days from the date of blending to the date of analysis.
The concentrations are in parts per million. The data shown in these
tables provided the base for the data analysis described in Section 4.0.
SCOTT RESEARCH LABORATORIES, INC.
-------
TABLE 3-1 ANALYTICAL DATA FOR CARBON MONOXIDE AT 10 PPM
Cyl.
Code
Age
13
31
45
91
113
136
153
168
183
205
223
Cyl.
Code
Age
17
31
78
99
122
139
154
169
191
209
Cyl.
Code
Ace
26
54
75
97
120
137
152
167
189
208
A11349
21111*
Cone.
iU. jy
10.81
10-33
10.55
10.72
10.41
10.39
10.35
10.56
10.40
10.41
A11359
21211
Cone.
10.08
9.66
10.56
10.18
10.00
9.86
9.98
10.13
10.10
10.18
A11369
21311
Cone .
10.16
10.27
10.69
10.81
10.16
10.45
10.59
10.69
10.53
10.40
Cyl.
Code
Age
26
54
75
97
120
137
152
167
176
189
203
Cyl.
Code
Age
22
50
71
93
116
133
148
163
172
204
Cyl.
Code
Age
22
50
72
143
157
184
A11379
21121
Cone.
9.97
10.40
10.80
10.79
10.15
10.81
11.01
10.64
10.43
10.66
10.51
All 389
21221
Cone.
10.30
10.29
10.82
10.59
10.25
10.39
10.41
10.71
10.20
10.41
All 399
21321
Cone .
9.91
10.06
10.57
10.16
10.10
10.03
Cyl.
Code
Age
7
20
36
53
142
169
203
Cyl.
'Code
Age
0
17
52
68
91
108
123
138
160
179
Cyl.
Code
Age
2
16
47
64
86
103
118
133
155
174
201
A11409
22122
Cone.
9.09
9.38
9.27
9.33
9.75
9.48
9.89
A11419
22212
Cone.
10.37
10.45
10.79
10.66
10.22
10.44
10.44
10.32
10.27
10.39
A11429
22322
Cone.
10.88
10.63
10.72
10.97
10.53
10.49
10.59
10.44
10.40
10.16
10.43
Cyl.
Code
Age
2
15
41
121
148
199
Cyl.
Code
Age
5
28
39
54
76
94
97
109
124
150
164
182
Cyl.
Code
Age
6
28
39
54
76
94
109
124
150
164
182
A11439
22113
Cone.
10.83
10.79
11.22
11.04
11.02
11.09
A11449
22313
Cone.
9.25
10.26
9.65
10.00
9.36
10.03
9.39
9.39
9.27
9.65
9.73
9.76
A11459
22223
Cone.
13.32
13.57
13.60
13.98
13.29
13.29
12.88
13.41
12.87
13.32
13.32
Cyl.
Code
Age_
12
25
41
63
147
174
208
Cyl.
Code
Age
6
39
53
69
91
109
112
124
139
165
179
197
Cyl.
Code
Age
2
35
49
65
87
105
108
120
135
141
161
175
193
A11469
11112
Cone.
9.15
9.36
9.45
9.67
9.46
9.14
9.72
A11479
11213
Cone.
10.71
11.53
11.58
11.42
10.76
11.02
10.99
10.84
10.60
10.93
11.04
11.01
A11489
11312
Cone.
10.51
10.51
10.73
10.88
10.69
10.20
10.42
9.99
9.11
9.21
9.62
9.76
9.65
Cyl.
Code
Age
1
37
48
63
85
103
106
118
133
139
159
173
191
Cyl.
Code
Age
5
28
39
54
76
94
109
124
130
150
164
182
Cyl.
Code
Age
6
28
69
125
133
165
A11499
11123
Cone.
11.52
12.44
12.24
12.11
11.54
11.82
11.63
11.75
11.14
11.65
11.52
11.64
11.54
A11509
11222
Cone.
10.09
10.47
10.43
10.41
9.62
9.65
9.55
8.81
8.93
9.06
9.23
9.37
A11519
11323
Cone.
11.27
11.57
10.94
10.88
11.23
11.32
Cyl
Code
Age
2
16
47
64
86
103
118
133
155
174
201
Cyl.
Code
Age
2
15
41
121
148
182
Cyl.
Code
Age
2
51
64
135
1 AT
175
A11529
12111
10.54
10.55
10.57
10.38
10.30
10.26
10.18
10.08
9.66
9.99
10.04
A11539
12211
Cone.
10.46
9.91
10.43
10.06
10.15
10.28
A11549
12311
Cone.
9.74
8.02
8.01
7.08
7.58
7.65
Cyl.
Code
Age
3
38
53
68
90
108
123
138
144
164
178
196
Cyl.
Code
Age
6
41
90
104
131
165
Cyl.
Code
Age
5
28
39
54
76
94
109
124
150
164
182
A11559
12121
10.09
9.40
9.11
7.97
7.05
7.26
6.57
6.41
4.59
5.15
5.30
5.00
A11569
12221
Cone.
10.42
10.72
10.15
9.95
9.91
10.09
A11579
12321
Cone.
11.20
11.82
11.38
11.95
11.34
11.30
11.37
11.22
10.93
11.21
11.30
u>
w
o
LO
K5
* See page 2-2 for explanation of code.
-------
TABLE 3-2 ANALYTICAL DATA FOR CARBON MONOXIDE AT 50 PPM
Cyl.
Code
Age
8
30
43
87
106
129
146
161
176
199
218
Cyl.
Code
Age
3
16
31
74
132
188
212
Cyl.
Code
Age
8
21
54
72
130
186
222
All 350
21111*
Cone.
50.86
51.61
'50.20
51.69
51.91
52.32
51.24
51.57
51.83
51.12
51.94
A11360
21211
Cone.
49.96
50.96
49.83
51.38
50.80
51.09
51.55
A11370
21311
Cone.
50.30
53.94
52.20
53.01
51.71
53.90
53.79
Cyl.
Code
Age
8
21
54
72
90
113
130
145
160
183
202
Cyl.
Code
Age
7
23
51
68
86
109
126
141
156
163
179
198
Cyl.
Code
Age
7
23
51
68
86
109
126
141
.156
163
179
198
All 380
21121
Cone.
51.05
52.51
51.65
52.37
52.05
52.37
51.50
51.97
51.64
51.88
52.42
A11390
21221
Cone.
50.64
49.94
51.71
52.78
52.41
52.58
51.46
51.35
52.24
52.28
51.92
52.43
A11400
21321
Cone.
51.05
50.52
52.24
52.54
52.26
52.18
52.19
51.61
52.81
51.67
51.27
52.31
Cyl.
Code
Age
2
20
37
56
83
98
113
136
154
168
184
213
A11410
22122
Cone.
60.32
59.54
59.17
59.97
59.52
58.61
58.37
58.54
57.87
59.24
58.68
58.68
Cyl A11420
Code
Age
3
17
53
73
88
103
126
144
158
174
203
Cyl.
Code
Age
1
16
48
68
83
98
121
139
153
169
201
22212
Cone.
52.49
51.77
52.65
52.76
51.88
51.53
51.88
51.13
52.14
52.51
52.92
A11430
22322
Cone.
53.54
52.83
53.83
54.05
52.80
52.52
52.80
52.21
53.05
53.06
53.13
Cyl.
Code
Age
2
16
42
64
87
104
119
135
142
157
174
197
Cyl.
Code
Age
5
41
90
124
133
165
Cyl.
Code
Age
7
31
42
59
74
89
112
129
145
159
185
189
A11440
22113
Cone.
48.42
48.08
49.05
49.43
48.32
47.89
48.88
47.84
48.55
47.11
49.91
49.94
A11450
22313
Cone.
49.37
49.89
49.31
48.51
50.71
50.19
A11460
22223
Cone.
53.10
53.76
54.00
54.03
52.55
52.85
53.30
54.18
53.78
55.17
55.13
55.56
Cyl.
Code
Age
25
42
61
147
172
208
Cyl.
Code
Age
6
41
47
62
85
102
117
133
140
158
174
195
203
Cyl.
Code
Age
2
37
43
58
81
98
113
129
136
154
170
191
199
A11470
11112
Cone.
55.40
55.04
54.97
55.44
54.08
54.19
55.49
A11480
11213
Cone.
52.40
52.44
52.97
52.65
51.71
51.44
52.46
51.37
51.60
51.13
53.50
53.01
52.65
A11490
11312
Cone.
51.94
52.65
52.00
51.39
50.80
50.55
51.22
49.17
49.36
48.70
50.63
50.08
49.89
Cyl.
Code
Age
1
22
78
113
138
174
210
Cyl.
Code
Age
5
41
90
124
133
165
Cyl.
Code
Age
7
31
42
59
74
89
112
129
145
159
185
189
A11500
11123
Cone.
47.01
48.38
47.59
47.72
47.49
49.46
49.51
A11510
11222
Cone.
50.58
50.44
49.09
47.95
49.24
49.31
A11520
11323
Cone.
53.96
54.35
54.71
53.80
53.56
53.09
53.78
54.03
53.62
53.69
53.55
53.63
Cyl.
Code
Age
1
16
48
68
83
98
121
139
146
153
169
201
Cyl.
Code
Age
2
16
42
64
87
104
119
135
142
157
174
197
Cyl.
Code
Age
2
23
52
69
84
99
122
139
155
195
199
A11530
12111
Cone.
52.30
51.13
53.07
52.68
52.27
51.72
52.25
51.38
52.06
52.67
52.77
53.72
A11540
12211
Cone.
52.04
51.46
51.48
50.33
49.03
49.50
49.73
48.22
48.88
48.18
50.84
49.78
AH550
12311
Con.
50.94
51.54
50.98
50.93 '
50. 17
49.79
50.10
50.72
50.57
50.72
50.67
Cyl.
Code
Age
3
39
56
73
88
103
108
126
143
159
173
199
Cyl.
Code
Age
6
41
90
124
133
165
175
Cyl.
Code
Age
6
41
90
124
133
157
A11560
12121
Cone.
52.65
53.00
52.43
52.36
52.10
50.96
52.18
51.7?
51.43
51.70
51.75
52.07
A11570
12221
Cone.
51.58
52.05
51.20
50.59
52.10
52.35
52.00
A11580
12321
Cone.
54.26
53.63
53.82
52.51
54.42
54.66
CO
LO
U)
o
* See page 2-2 for explanation of code.
-------
o
n
n
TABLE 3-3 ANALYTICAL DATA FOR CARBON MONOXIDE AT 100 PPM
Cyl. A11351
Code 21111*
Age Cone.
7 109.15
29 108.23
50 112.21
91 110.27
106 109.50
129 110.17
146 110.16
161 109.36
176 109.56
199 111.19
'218 108.69
Cyl. A11361
Code 21211
Age Cone.
6 107.20
17 106.65
36 109.36
77 108.53
92 107.40
115 107.00
132 107.70
147 106.71
162 107.12
185 107.44
204 106.75
Cyl. A11371
Code 21311
Age Cone.
6 111.70
54 111.95
19 111.35
75 109.90
180 111.62
204 109.72
Cyl. A11381
Code 21121
Age Cone.
6 108.17
20 109.80
54 111.95
75 110.20
90 lll.io
113 110.03
130 109.22
145 108.39
160 108.91
183 109.68
202 108.28
Cyl. A11391
Code 21221
Age
4
30
71
93
113
128
151
169
183
199
Cyl.
Code
Age
4
29
71
86
109
126
141
156
179
198
221
Cone.
111.60
113.30
111.27
109.53
112.37
109.76
109.84
110.80
112.71
109.37
A11401
21321
Cone.
108.20
111.13
108.07
109.93
107.83
108.08
106.67
106.63
1 O7 99
AU / . £ f.
106.01
107.60
Cyl. A11411
Code 22122
Age Con.
2 106.02
24 106.50
58 107.70
83 106.73
161 107.44
185 106.06
213 107.07
Cyl. A11421
Code 22212
Age Cone .
4 109.17
18 106.85
Q T 1 e\f i- n
O J
151
187
229
Cyl.
Code
Age
2
16
55
70
93
110
125
141
166
180
203
107.25
106.48
106.43
A11431
22322
Cone.
114.64
112.40
110.27
110.47
110.15
110.23
110.09
111.79
113.46
110.84
112.25
Cyl. A11441
Code 22113
Age Cone.
3 103.73
17 105.93
43 105.09
64 102.84
87 103.75
104 103.21
119 102.88
135 104.32
160 102.30
176 104.70
197 105.20
Cyl. A11451
Code 22313
Age Cone .
0 91.41
26 92.88
42 91.64
60 92.69
74 91.35
89 91.11
112 91.46
130 94.31
138 95.13
146 91.99
159 94.27
185 94.21
192 93.77
Cyl. A11461
Code 22223
Age Cone.
11 108.60
42 109.27
84 108.05
123 111.57
157 109.86
201 111.21
Cyl. A11471
Code 11112
Age Cone .
7 105.50
29 106.50
63 109.15
88 106.75
103 108.67
118 107.72
141 106.43
159 107.19
173 108.78
189 105.07
228 106.45
Cyl. A11481
Code 11213
Age Cone.
6 110.37
41 109.00
70 108.45
139 109.97
175 109.96
Cyl. A11491
Code 11312
Age Cone .
2 111.20
37 108.33
43 107.67
58 107.25
81 106.43
98 106.68
113 106.56
129 106.27
154 108.09
168 104.95
191 105.20
203 105.41
Cyl. A11501
Code 11123
Age Cone.
1 107.17
35 108.95
51 107.47
69 106.77
83 105.76
98 106.41
121 106.71
138 108.60
155 105.60
168 106.60
194 107.62
201 107.46
Cyl.
Code
Age
o
26
42
60
74
89
94
112
130
138
146
159
185
192
A11511
11222
Cone.
1 A7 }*
lu/. J3
106.60
103.55
105.10
104.10
100.34
98.91
99.40
103.04
104.01
99.97
99.12
98.48
98.05
Cyl. AH521
Code 11323
Age Cone .
11 109.33
27 108.47
46 107.10
61 105.53
84 103.72
101 104.89
118 106.37
138 103.61
157 101.49
171 101.96
185 100.61
192 100.97
Cyl.
Code
Age
0
16
78
146
182
224
Cyl.
Code
Age
3
17
43
64
87
l m
107
119
135
160
197
209
A11531
12111
Cone.
111.07
108.90
108.10
109.37
108.84
108.68
A11541
12211
Cone.
104.90
104.57
105.70
101.67
103.45
101 . ]7
101.76
101.59
101.22
102.15
102.81
101.99
Cyl. A11551 Cyl. A11581
Code 12311 Code 12321
4&i Cqnc_._ Age Cone.
J 103.57 6 TYY^o
52 IS?'?1 2? U1-03
84 98.99 160 108.71
139 100.97
148 100.06
156 99.58
169 99.85
195 99.71
202
Cyl.
Code
Age
4
56
98
137
171
Cyl.
Code
Age
6
27
46
61
84
101
118
131
157
171
185
192
99.07
A11561
12121
Cone.
109.73
109.53
106.68
110.81
107.51
A11571
12221
Cone.
116.83
114.70
112.67
111.65
108.47
109.65
109.98
109.22
108.61
110.52
109.86
110.75
u>
u>
* See page 2-2 for explanation of code.
-------
TABLE 3-4 ANALYTICAL DATA FOR CARBON MONOXIDE AT 500 PPM
Cyl. A11352
Code 21111 *
Age Cone.
9 ceo oft
J J£. . oU
29 554.45
50 542.86
93 543.00
111 550.20
126 553.30
141 556.62
164 550.90
181 545.72
196 549.73
211 556.66
Cyl. A11362
Code 21211
Age Cone.
6 537.77
21 541.35
56 526.30
79 527.35
170 529.00
192 531.86
233 530.55
Cyl. A11372
Code 213112
Age Cone.
5 555.57
19 552.25
54 551.00
77 555.87
95 551.20
110 558.67
125 554.05
148 550.92
165 546.53
180 550.81
195 553.58
* See page
Cyl. A11382
Code 21121
Age Cone .
5 522.47
19 518.15
54 518.95
77 522.73
168 525.10
190 529.12
Cyl. All 392
Code 21221
Age Cone.
3 503.80
29 510.45
63 511.50
164 521.20
184 519.82
225 520.72
Cyl. A11402
Code 21321
Age Cone .
3 544.97
18 541.60
63 543.53
88 535.01
105 546.73
120 541.07
135 535.95
158 537.25
175 532.29
190 540.86
206 538.30
234 542.88
2-2 for exp]
Cyl. A11413
Code 22122
Age Cone.
1 540.50
24 530.50
59 530.70
76 536.00
91 541.45
106 534.74
129 532.67
146 530.41
161 532.17
176 539.51
211 535.01
Cyl. A11422
Code 22212
Age Cone.
4 532.57
18 542.40
55 541.33
80 545.75
96 545.40
110 544.07
133 544.01
150 536.76
165 545.83
181 542.95
209 550.70
Cyl. A11432
Code 22322
Age Cone.
2 523.40
16 516.13
106 515.66
148 509.76
259 516.13
Lanation of
Cyl. A11442
Code 22113
Age Cone .
3 537.57
27 549.40
86 546.76
142 542.09
191 549.13
Cyl. A11452
Code 22313
Age Cone .
0 536.67
32 531.35
46 542.40
62 538.20
84 537.39
101 537.18
117 538.06
131 543.09
154 550.00
171 545.88
185 545.39
193 545.98
Cyl. A11462
Code 22223
Age Cone.
11 514.32
48 519.60
83 518.18
125 511.44
174 522.84
code.
Cyl. A11472
Code 11112
Age Cone .
6 531.60
29 535.70
64 524.90
81 528.17
96 534.20
111 526.68
134 530.35
151 522.73
154 531.17
166 519.91
181 535.20
216 529.40
Cyl. A11482
Code 11213
Age Cone.
6 550.03
27 555.93
61 549.73
76 556.80
99 542.29
116 549.41
120 549.93
132 543.55
146 552.52
169 548.18
186 559.92
201 560.78
208 556.56
Cyl. A11492
Code 11312
Age Cone.
2 513.62
O *J C 1 / £ A
iJ 514.60
80 515.06
136 503.13
185 513.34
Cyl.
Code
Age
1
20
58
75
91
105
128
145
152
160
175
195
Cyl.
Code
Age
0
32
46
62
84
101
117
131
137
154
171
185
193
A11502
11123
Cone
522.37
517.50
521.27
522.46
521.67
520.29
519.16
530.27
517.19
527.65
535.38
534.68
11512
11222
Cone.
538.67
533.60
538.40
532.30
535.97
536.39
530.51
544.14
544.66
550.88
545.04
547.26
546.52
Cyl
Cod.
Age
11
35
46
61
84
101
117
131
137
154
171
186
193
Cyl.
Code
Age
0
16
71
134
154
197
215
Cyl.
Code
Age
3
27
52
69
84
99
122
139
154
170
198
. A11522
e 11323
Cone.
546.77
540.63
544.53
546.40
541.50
538.43
534.87
545.37
553.50
525.15
554.72
554.37
549.68
A11532
12111
Cone.
537.43
543.73
542.77
548.53
548.75
545.26
545.14
A11542
12211
Cone.
527.43
530.20
528.83
539.43
530.64
533.79
534.90
525.83
535.32
535.41
541.02
Cyl.
Code
Age
2
21
59
76
92
106
129
146
153
161
176
196
Cyl.
Code
Age
4
41
49
66
82
96
103
119
136
151
167
195
AH552
12311
Cone.
501.60
499.40
503.40
499.61
499.00
496.68
494.33
504.67
495.65
509.55
518.81 '
507.88
A11562
12121
Cone.
545.23
541.57
537.10
542.23
543.88
531.17
527.56
530.12
522.17
529.06
513.51
520.64
Cyl.
Code
Age
4
33
46
62
84
101
117
131
154
171
186
195
Cyl.
Code
Age
4
33
46
62
84
101
117
131
154
171
186
195
W
P
A11572 ,_
12221 W
Cone. «j
613.20 ^
588.63 w
595.07 0
601.17 0
590.47 Ji
587.10
589.24
592.50
599.77
599.64
608.62
613.13
All 582
12321
Cone.
534.43
532.50
539.30
536.33
534.55
529.98
524.13
532.76
532.30
534.15
543.35
557.89
Ln
-------
TABLE 3-5 ANALYTICAL DATA FOR CARBON MONOXIDE AT 1000 PPM
{/>
0
£j
w
en
n
n
BB
E
o
90
i-l
O
90
MM
M
ft
Cyl.
Code
Age
10
29
43
94
184
204
240
Cyl.
Code
Age
6 '
22
55
80
105
120
134
157
176
189
206
Cyl.
Code
Age
5
20
53
78
95
110
125
148
165
180
195
*
A11353
21111 *
Cone.
1133.00
1121.00
1122.92
1128.67
1135.97
1117.83
1126.32'
A11363
21211
Cone.
1105.33
1079.00
1093.00
1091.33
1089.67
1102.33
1095.02
1083.02
1096.95
1093.08
1092.20
A11373
21311
Cone.
1064.50
1037.00
1058.50
1059.33
1052.67
1057.00
1065.00
1055.78
1049.78
1067.08
1045.83
C**fi «***n
Cyl. A11383
Code 21121
Age Cone .
5 1080.00
20 1066.50
53 1089.50
78 1082.50
95 1081.t>0
110 1085 .-GO
125 1091^7
148 1082.67
165 1076.96
180 1095.95
195 1070.11
Cyl. A11393
Code 21221
Age Cone .
3 1089.00
18 1089.00
63 1085.67
164 1088.27
184 1074.94
Cyl. A11403
Code 21321
Age Cone.
3 1126.67
18 1140.00
63 1115.00
136 1114.7%
179 1113.62
220 1104.90
A O O £s\V £*-V~r<
Cyl.
Code
Age
1
27
59
77
91
106
129
146
161
176
212
Cyl.
Code
Age
18
111
154
201
Cyl.
Code
Age
19
85
149
184
224
»1 ai-to t*
A11412
22122
Cone.
1123.33
1115.00
1112.00
1109.52
1116.50
1124.67
1118.33
1113.59
1128.34
1093.66
1107.77
All 42 3
22212
Cone.
1125.00
1121.67
1131.03
1123.19
1124.84
A1I433
22322
Cone.
1055.67
1075.00
1086.50
1084.69
1092.46
1056.54
^ f>n rtf n
Cyl.
Code
Age
6
28
52
69
84
99
122
139
142
154
171
198
Cyl.
Code
Age
3
34
49
66
81
96
119
136
153
166
187
194
st/4 A
A11443
22113
Cone.
1116.00
1112.33
1112.33
1108.50
1117.33
1110.94
1097.91
1133.86
1120.37
1103.63
1079.76
1114.59
A11453
22313
Cone.
1097.37
1113.33
1099.00
1112.10
1103.12
1103.99
1113.50
1094.95
1108.14
1134.31
1111.15
1125.39
Cyl. A11463
Code 22223
Age Cone.
10 1111.33
34 1114.67
49 1099.33
66 1111.38
81 1105.45
96 1104.15
119 1124.35
136 1103.51
151 1149.72
166 1120.25
187 1113.72
194 1130.31
Cyl. A11473
Code 11112
Age Cone .
6 1093.33
32 1095.50
64 1090.33
82 1091.33
96 1086.00
111 1098.15
134 1087.57
153 1084.00
166 1103.47
181 1065.13
217 1071.10
Cyl. A11483
Code 11213
Age Cone .
5 1095.33
27 1087.67
77 1099.00
141 1092.08
176 1117.71
211 1092.52
Cyl. A11493
Code 11312
Age Cone .
1 1100.00
23 1082.33
38 1087.00
57 1081.00
72 1094.00
86 1089.47
109 1077.37
128 1090.32
141 1088.64
158 1091.45
192 1082.08
205 1082.59
Cyl. All 503
Code 11123
Age Cone.
2 1024.50
20 1013.00
71 1033.50
135 1014.37
175 1027.53
205 1055.76
Cyl. A11513
Code 11222
Age Cone .
3 1122.00
34 1129.00
49 1116.00
66 1139.34
81 1117.78
96 1119.08
119 1131.19
136 1110.95
153 1127.16
166 1151.86
187 1128.11
194 1121.82
Cyl.
Code
Age
10
34
49
66
81
96
119
136
151
166
187
196
Cyl.
Code
Age
1
19
57
75
90
105
128
145
148
160
174
204
A11523
11323
Cone.
1115.33
1107.67
1092.33
1117.03
1097.72
1104.69
1118.98
1088.28
1140.68
119.73
1103.74
1120.46
A11533
12111
Cone.
1083.67
1098.00
1106.33
1104.67
1110.84
1093.38
1077.47
1120.73
1105.51
1081.70
1082.20
1090.56
Cyl. A11543
Code 12211
Age Cone .
6 1117.33
28 1118.67
52 1115.00
69 1116.00
84 1123.47
99 1113.01
122 1093.18
139 1134.94
142 1120.73
154 1106.42
17 1086.70
198 1115.94
Cyl. A11553
Code 12311
Age Cone .
3 1138.00
21 1123.00
72 1135.50
136 1115.28
176 1137.48
206 1152.85
Cyl. A11563
Code 12121
Age Cone .
4 1124.00
41 1118.00
48 1124.67
66 1122.00
81 1133.24
96 1107.73
103 1123.74
119 1087.52
136 1133.14
139 1124.94
151 1102.62
165 1117.45
195 1119.42
Cyl.
Code
Age
4
34
49
67
81
96
119
136
151
166
187
194
Cyl.
Code
Age
34
49
67
81
96
119
136
151
166
187
A11573
12221 |-1
Cone . i-*
989.37 -J
993.33 h-
988.73 w
992.63 o
999.28 5
996.84 *»
1001.92
989.07
972.70
1028.25
996.60
1006.35
A11583
12321
Cone.
1140.67
1146.67
1138.67
1145.36
1150.29
1144.05
1167.85
1140.11
1129.65
1169.04
1153.58
-------
TABLE 3-6 ANALYTICAL DATA FOR PROPANE AT 3 PPM
m
s
K
O
o
n
CO
i
to
U)
Cyl. A11344
Code 21111*
Age Cone.
27 3 34
43 3.32
55 3.27
79 3.27
106 3.22
111 3.27
120 3.24
127 3.24
146 3.20
160 3.26
175 3.29
197 3.29
Cyl. A11354
Code 21211
Age Cone.
13 3TT7~
29 3.14
42 3.12
65 3.09
92 3.10
106 3.08
113 3.07
132 3.02
136 3.08
146 3.12
148 3.14
162 3.11
183 3.24
188 3.16
203 3.12
Cyl
Cod<
4
26
39
54
90
95
104
111
130
144
146
160
181
201
Cyl.
Code
Age
4
26
39
54
90
104
111
130
134
144
160
166
181
201
. All 364
! 21311
. Cone.
3.20
3.19
3.23
3.14
3.16
3.16
3.10
3.10
3.19
3.20
3.20
3.20
3.16
All 374
21121
Cone.
3.17
3.19
3.12
3.13
3.13
3.15
3.17
3.10
3.16
3.17
3.20
3.21
3.19
3.11
Cyl. AU384
Code 21221
Age Cone.
23 3.18
37 3.17
51 3.20
73 3.20
100 3.17
123 3.19
140 3.22
154 3.20
170 3.19
190 3.23
191 3.20
Cyl. All 394
Code 21321
Age Cone.
3 3.24
23 3.25
37 3.23
59 3.20
80 3.21
101 3.23
123 3.25
144 3.09
148 3.26
164 3.24
185 3.24
204 3.24
Cyl. A1140
Code 22122
Age Cone.
7 3.32
13 3.25
27 3.19
49 3.19
66 3.20
114 3.24
136 3.27
167 3.27
175 3.25
189 3.25
196 3.27
Cy]
Coc
Age
17
33
53
75
98
115
129
145
165
185
200
Cyl.
Code
Age
20
34
51
68
83
98
121
139
152
168
202
L. A114H
le 22212
' Cone.
3.23
3.23
3.20
3.23
3.18
3.22
3.19
3.24
3.24
3.25
3.18
3.23
A11424
22322
Cone.
3.22
3.22
3.19
3.21
3.20
3.23
3.22
3.24
3.22
3.20
3.18
3.16
Cyl. A11434
Code 22113
Age Cone.
1
15
29
45
70
93
115
154
175
177
196
Cyl.
Code
24
41
59
68
89
111
129
143
144
160
164
186
188
3.14
3.14
3.13
3.10
3.16
3.14
3.17
3.15
3.20
3.14
3.19
A11444
22313
Cone.
3.28
3.27
3.24
3.29
3.28
3.27
3.28
3.29
3.23
3.29
3.20
3.30
3.28
3.28
*
Cyl.
Cod«
Age
4
24
41
59
68
89
111
132
143
160
188
Cyl.
Code
Age
18
32
48
68
88
110
131
134
138
152
175
186
189
194
, A11454
! 22223
Cone.
3.30
3.27
3.25
3.26
3.28
3.27
3.31
3.27
3.25
3.27
3.30
A11464
11112
Cone.
3.47
3.20
3.14
3.20
3.21
3.19
3.21
3.28
3.04
3.22
3.22
3.18
3.15
3.21
3.21
Cyl,
Cod<
Age
5
20
41
62
85
102
116
132
152
153
172
174
193
Cyl.
Code
Age
16
37
58
81
98
112
128
148
149
168
170
183
loy
. A11474
! 11213
Cone.
3.13
3.14
3.14
3.14
3.16
3.17
3.17
3.11
3.17
3.15
3.07
3.09
3.12
A11484
11312
Cone .
3.13
3.12
3.12
3.09
3.07
3.11
3.10
3.13
3.17
3.10
3.06
3.05
3.11
Cyl,
Code
Age
2
20
35
54
69
84
104
107
125
133
138
154
Cyl.
Code
Age
4
24
41
59
68
89
111
129
143
144
160
164
178
179
186
. A11494
! 11123
3.14
3.08
3.12
3.10
3.08
3.13
3.16
3.15
3.07
3.14
3.10
3.11
A11504
11222
Cone.
3.15
3.14
3.12
3.15
3.19
3.19
3.19
3.19
3.11
3.17
3.13
3.17
3.16
3.12
3.16
Cyl
Cod.
Age
19
20
28
48
95
111
132
151
188
Cyl.
Code
Age
20
34
51
68
84
98
100
139
152
168
169
189
Cyl.
Code
Age
0
15
29
45
70
93
115
120
154
175
195
196
. A11514
e 11323
3.32
3.32
3.29
3.31
3.34
3.35
3.34
3.34
3.34
A11524
12111
Cone.
3.16
3.15
3.18
3.13
3.18
3.19
3.12
3.19
3.14
3.19
3.16
3.20
3.14
3.14
A11534
12211
Cone.
3.03
2.98
2.97
2.5£
3.02
3.02
3.10
3.03
3.02
3.04
3.00
3.05
Cyl.
Code
Age
21
36
58
101
105
121
142
161
196
197
Cyl.
Code
Age
0
19
39
68
84
89
91
109
112
123
138
159
160
178
185
All 544
! 12311
Cone.
3.02
2.93
3.02
2.99
2.78
3.03
3.00
3.03
3.03 .
3.10
3.05
A11554
12121
Cone.
3.11
3.08
3.07
2.85
3.01
3.06
3.05
3.15
3.08
3.13
3.13
3.04
3.10
3.10
3.09
Cyl. A11564 ^
Code 12221 W
Age Cone.
4 37IT g
24 3.10 ^j
41 3.12 *-
63 3.08
82 3.14
95 3.17
124 3.18
146 3.15
151 3.16
167 3.14
188 i it
Cyl.
Code
Age
4
24
41
59
68
89
111
129
143
144
160
188
A11574
12321
Cone.
3.16
3.14
3.11
3.14
3.16
3.16
3.17
3.15
3.09
3.12
3.14
3.12
-------
TABLE 3-7 ANALYTICAL DATA FOR PROPANE AT 10 PPM
e
SCOTT RESEARCH LABORAT
0
n
y
M
Cyl.
Code
Age
27
43
56
79
106
120
127
146
150
160
176
182
184
197
Cyl.
Code
Age
13
29
42
65
84
104
126
147
150
154
168
191
A11345
21111*
Cone.
10.85
10.80
10.80
10.72
10.75
10.75
10.95
10.79
10.79
10.95
10.72
10 64
10.87
A11355
21211
Cone.
10.90
10.80
10.80
10.80
10.84
10.73
10.75
10.86
11.33
10.90
10.80
10.90
Cyl.
Code
Age
4
26
39
54
77
96
117
118
138
160
181
200
Cyl.
Code
Age
4
26
39
54
90
95
104
111
130
144
160
181
201
A11365
21311
Cone.
10.97
11.00
10.95
11.00
10.97
10.95
10.75
10.99
10.98
11.05
11.00
11.09
A11375
21121
Cone.
11.10
11.13
11.15
11.17
11.07
11.04
11.03
11.04
10.91
11.10
11.14
11.08
11.11
Cyl.
Code
Age
3
23
37
51
73
100
123
140
154
170
190
210
Cyl.
Code
Age
2
23
37
59
86
100
107
126
130
140
156
164
177
197
All 385
21221
Cone.
10.87
10.90
10.90
10.90
10.83
10.83
10.88
10.94
10.93
10.96
10.98
11.11
A11395
21321
Cone.
11. DO
11.00
11.10
11.13
11.03
11.03
11.06
10.96
11.17
11.13
11.21
11.01
11.12
11.11
Cyl.
Code
Age
7
13
27
49
71
85
92
111
125
141
162
182
Cyl.
Code
Age
3
17
33
53
75
98
115
129
145
165
185
Cyl.
Code
Age
5
20
34
51
68
83
98
121
139
152
168
169
202
A11405
22122
Cone.
10.47
10.30
10.40
10.40
10.38
10.39
10.38
10.35
10.44
10.36
10.46
10.42
A11415
22212
Cone.
11.07
10.97
10.95
10.93
10.95
10.95
10.98
10.99
11.15
11.05
11.29
A11425
22322
Cone.
11.03
11.00
11.00
11.02
10.97
10.89
11.06
11.06
11.02
10.99
11.11
11.02
11.14
Cyl.
Code
Age
1
15
29
45
62
77
92
115
133
146
162
196
Cyl.
Code
Age
3
24
41
63
66
82
124
146
167
186
188
Cyl.
Code
Age
4
24
41
59
68
89
111
129
143
160
188
A11435
22113
Cone.
10.53
10.60
10.55
10.53
10.51
10.42
10.59
10.57
10.57
10.65
10.66
10.62
A11445
22313
Cone.
10.80
10.74
10.71
10.58
10.79
10.78
10.74
10.80
10.72.
10.77
10.84
A11455
22223
Cone.
10.90
10.89
10.86
10.84
10.87
10.83
10.72
10.91
10.90
10.92
10.94
Cyl.
Code
Age
5
18
32
48
68
88
110
131
152
180
194
Cyl.
Code
Age
5
20
41
62
85
102
116
132
152
172
193
Cyl.
Code
Age
1
16
37
58
81
98
100
112
128
148
168
183
A11465
11112
Cone.
10.63
10.60
10.60
10.60
10.59
10.56
10.63
10.64
10.61
10.65
10.66
A11475
11213
Cone.
10.97
10.97
11.02
11.01
10.96
11.06
11.02
11.09
11.14
11.05
11.00
A11485
11312
Cone.
10.87
10.80
10.85
10.79
10.73
10.86
10.89
10.89
10.94
10.90
10.88
10.89
Cyl.
Code
Age
2
20
35
57
78
100
104
120
141
160
Cyl.
Code
Age
3
24
41
82
124
146
167
186
Cyl.
Code
Age
20
28
48
66
73
77
98
117
136
147
151
166
188
A11495
11123
Cone.
10.80
10.80
10.76
10.87
10.77
10.96
10.86
10.73
10.81
10.82
All 505
11222
Cone.
17.40
17.37
17.21
17.49
17.46
17.35
17.37
17.37
A11515
11323
Cone.
15.48
15.51
15.48
15.51
15.23
15.62
15.63
15.64
15.58
15.73
15.66
15.67
15.58
Cyl.
Code
Age
5
20
34
51
68
83
98
100
121
139
152
168
189
190
Cyl.
Code
Age
0
15
29
45
62
77
92
115
133
149
162
195
A11525
12111
Cone.
10.90
10.90
10.90
10.91
10.83
10.77
10.92
10.92
10.94
10.94
10.90
10.95
10.84
11.02
A11535
12211
Cone.
10.82
10.80
10.80
10.84
10.72
10.70
10.85
10.86
10.79
10.90
10.91
10.86
Cyl.
Code
Age
3
21
36
55
70
85
108
126
139
155
156
196
Cyl.
Code
Age
0
19
39
73
82
103
125
143
157
174
185
Cyl.
Code
Age
3
24
41
63
82
124
146
151
167
188
A11545
12311
Cone .
10.65
10.78
10.77
10.66
10.58
10.73
10.75
10.79
10.75
10.85
10.83
10%82
A11555
12121
11.07
11.10
11.12
11.02
11.09
11.15
10.98
11.14
11.21
11.23
11.06
A11565
12221
Cone .
10.70
10.77
10.62
10.53
10.66
10.75
10.86
10.79
10.83
10.73
Cyl.
Code
Age
4
24
41
63
82
124
129
146
151
167
188
A11575
12321
11.20
11.26
11.15
11.15
11.18
11.38
11.37
11.14
11.25
11.15
11.08
OT
1 »
u>
-4
h-
u>
o
ON
P-
00
See page 2-2 for explanation of code.
-------
TABLE 3-8 ANALYTICAL DATA FOR PROPANE AT 50 PPM
en
O
>
O
53
Cyl. All 346
Code 21111 *
Age Cone .
27 51.77
43 52.35
56 52.10
79 52.30
111 51.77
112 51.93
126 52.26
141 52.53
162 52.76
182 52.14
211 52.59
Cyl. A11356
Code 21211
Age Cone.
13 53.10
29 53.00
42 53. "5
65 52.77
97 52.75
98 52.80
112 53.06
127 53.30
148 53.74
168 53.30
183 53.06
197 53.33
Cyl. All 366
Code 21311
Age Cone.
4 53.27
20 52.85
39 52.55
54 52.40
76 52.80
96 52.48
138 52.61
181 52.85
200 53.15
Cyl.
Code
Age
4
. 20
39
54
75
97
118
137
152
155
167
189
Cyl.
Code
Age
3
23
37
51
72
92
107
128
148
162
177
198
A11376
21121
Cone.
54.60
54.75
54.65
54.45
54.70
54.45
55.03
54.88
55.58
55.86
54.99
55.07
A11386
21221
Cone.
53.70
53.25
52.90
52.70
53.30
52.95
53.31
53.46
53.90
53.68
53.25
53.49
Cyl.
Coda
Age
23
37
59
91
92
106
121
122
142
144
162
164
191
Cyl.
Code
Age
13
27
49
66
91
114
136
139
170
175
196
All 396
21321
Cone.
52.73
52.40
52.65
52.00
52.25
52.03
52.50
52.60
52.87
53.03
53.00
52.23
52.76
52.59
A11406
22122
Cone.
51.43
51.20
51.10
51.05
51.25
51.35
51.15
52.25
51.79
51.55
51.69
51.81
Cyl.
Code
Age
17
33
52
73
95
116
137
160
179
188
Cyl.
Age
5
20
34
51
77
99
118
121
132
147
168
187
A11416
22212
Cone.
53.30
53.45
53.30
53.51
53.37
54.11
53.93
53.87
53.80
54.10
53.95
A11426
22322
Cone.
53.70
53.30
53.40
53.60
53.42
53.41
54.39
54.32
53.58
53.74
54.08
53.93
Cyl.
Code
Age
15
29
45
71
92
112
115
126
141
162
181
202
Cyl.
Code
Age
21
41
61
62
80
103
116
132
150
171
187
Cyl.
Age
71
£>
82
146
167
187
. A11436
s 22113
Cone.
60.05
59.45
59.70
59.85
59.89
60.18
60.95
60.62
60.43
59.96
60.34
60.44
59.75
A11446
22313
Cone.
52.40
52.70
52.76
52.65
52.96
52.32
53.07
52.69
53.15
53.21
52.91
52 88
A11456
53.41
53.69
53.58
53.95
Cyl.
Code
Age
5
18
32
48
67
83
104
123
138
141
153
175
194
' Cyl.
Code
Age
20
41
63
84
88
106
126
147
166
200
A11466
11112
Cone.
52.92
52.75
S2.70
52.47
53.07
52.72
53.11
53.23
53.61
53.24
52.87
53.11
53.17
A11476
11213
Cone.
55.95
55.85
56.16
56.01
55.87
56.20
56.63
56.05
56.45
56.18
56.07
Cyl.
Code
Age
1
16
37
65
86
106
109
120
135
156
175
177
183
Cyl.
Code
Age
20
35
63
84
86
104
118
133
154
173
A11486
11312
Cone.
52.92
52.83
53.12
53.07
52.85
53.79
53.68
53.16
53.54
53.67
53.18
53.32
53.44
A11496
11123
Cone.
52.92
53.05
53.10
52.94
53.68
52.96
53.87
53.44
53.35
53.26
53.41
Cyl.
Code
Age
3
21
41
61
77
80
103
116
132
150
171
172
185
Cyl.
Code
Age
19
28
48
66
73
98
117
136
147
166
167
187
188
A11506
11222
50.90
51.20
51.40
51.42
51.64
51.12
51.30
51.15
51.50
51.62
50.82
51.45
51.78
A11516
11323
Cone.
53.30
53.25
53.53
53.49
53.47
53.97
52.94
53.44
53.90
53.11
53.60
53.17
53.27
Cyl.
Code
Age
5
20
34
51
76
99
121
124
160
181
189
Cyl.
Code
Age
0
15
29
45
71
92
112
126
141
162
163
181
202
Cyl.
Code
Age
3
21
36
64
85
105
119
134
155
174
195
A11526
! 12111
52.80
52.50
52.95
52.85
52.83
53.08
54.73
53.21
53.18
53.26
53.21
A11536
12211
Cone.
52.72
52.50
52.65
52.83
52.86
53.22
53.47
53.19
53.08
53.74
53.33
53.27
53.23
All 546
12311
Cone.
51.77
51.95
52.40
52.18
52.63
52.85
52.20
52.44
52.27
52.30
52.27
Cyl.
Code
Age
0
19
38
59
81
102
123
146
164
165
173
185
Cyl.
Code
Age
24
41
61
80
101
103
116
132
171
187
Cyl.
Code
Age
4
21
41
63
82
124
146
167
187
A11556
! 12121
54.07
53.95
54.03
54.28
54.74
54.65
54.33
54.74
53.72
54.58
49.41
54.46
A11566
12221
Cone.
52.80
53.00
52.98
53.55
53.53
53.31
53.15
53.16
53.52
53.43
53.45
A11576
12321
Cone.
54.15
54.35
54.14
54.73
54.36
54.08
54.16
54.25
54.66
i '
CO
t *
Co
O
CO
10
* See page 2-2 for explanation of code.
-------
TABLE 3-9 ANALYTICAL DATA FOR PROPANE AT 100 PPM
en
P
OJ
Cyl. A11347
Code 21111*
Age Cone.
15 105.00
27 109.00
43 108.00
56 108.50
79 108.00
111 107.75
112 107.70
126 108.51
141. 107.00
182 107.68
211 107.17
Cyl. A11357
Code 21211
Age Cone.
2 107.33
13 107.67
29 108.00
42 107.00
65 107.00
83 106.27
104 106.40
126 106.18
147 106.09
168 105.40
19 105.47
210 106.95
Cyl. All 367
Code 21311
Age Cone.
4 109.00
20 1C8.00
39 108.00
54 107.00
76 106.65
97 106.56
118 106.31
137 106.12
152 106.63
167 107.04
189 106.11
208 106.07
Cyl.
Code
Age
4
20
39
54
76
96
117
138
160
181
186
200
Cyl.
Code
Age
3
23
37
51
72
92
113
134
156
182
196
219
A11377
21121
Cone.
108.33
109.00
109.00
109.00
108.50
108.44
107.66
107.97
107.14
108.31
107.64
108.04
A11387
21221
Cone.
107.33
107.50
107.50
107.00
106.70
106.61
105.79
105.37
105.82
105.34
105.93
107.16
Cyl.
Code
Age
23
37
59
91
92
106
121
142
144
162
191
Cyl.
Code
Age
13
27
49
76
77
91
106
127
147
176
184
A11397
21321
Cone.
102.33
102.00
102.00
101.50
101.25
101.20
100.47
100.23
101.51
100.95
101.90
100.64
All 407
22122
Cone.
103.33
103.00
102.50
102.50
102.55
102.50
102.15
101.76
102.36
101.87
101.19
102.20
Cyl.
Code
Age
3
17
33
52
81
91
110
131
150
164
165
181
188
Cyl.
Code
Age
5
20
34
51
76
99
100
103
121
124
160
181
208
A11417
22212
Cone.
103.00
102.00
102.00
102.33
102.00
101.81
101.55
101.70
100.87
102.22
101.34
101.95
102.27
A11427
22322
Cone.
108.00
108.00
108.00
106.90
107.04
105.44
105.69
104.69
107.06
107.11
106.67
106.87
106.66
Cyl.
Code
Age
15
29
45
70
93
115
154
175
202
Cyl.
Code
Age
3
21
41
61
80
103
116
132
150
171
187
Cyl.
Code
Age
4
21
41
82
124
146
167
187
A11437
22113
Cone.
102.00
102.00
101.57
100.73
101.10
100.97
101.34
100.86
101.19
100.76
A11447
22313
Cone.
105.65
105.20
105.30
104.92
105.00
104.71
105.51
105.20
104.73
104.91
105.13
A11457
22223
Cone.
108.00
107.55
107.75
.106.83
106.84
106.95
107.50
107.05
Cyl.
Code
Age
5
18
32
48
68
83
104
105
123
138
153
175
194
Cyl.
Code
Age
5
20
41
81
88
113
116
132
151
162
181
200
A11467
11112
Cone.
107.00
106.50
106.00
106.00
105.70
105.90
104.85
105.12
105.22
105.68
105.82
105.78
105.93
A11477
11213
Cone.
105.00
105.00
105.05
103.87
103.63
104.92
103.83
104.74
104.73
104.81
104.89
104.49
Cyl.
Code
Age
1
16
37
59
81
102
122
143
162
183
Cyl.
Code
Agje
2
20
35
62
72
91
112
131
145
162
Cyl.
Code
Age
21
41
61
62
80
101
103
116
132
150
171
172
185
A11487
11312
Cone.
107.00
107.33
106.55
105.95
106.58
106.77
106.49
106.67
106.16
106.27
A11497
11123
Cone.
105.50
105.00
105.27
104.46
104.68
104.20
104.37
104.56
104.82
104.92
All 507
11222
Cone.
102.00
101.20
101.50
100.00
100.68
100.53
101. 3D
100.07
101.68
101.31
101.03
100.46
101.00
100.89
Cyl.
Code
Age
19
28
48
66
73
98
103
117
136
147
166
187
Cyl.
Code
Age
"o*~
5
20
34
51
76
99
121
160
181
189
A11517
11323
Cone.
104.67
104.63
102.60
103.42
103.79
105.05
104.20
104.31
104.51
105.73
105.58
105.01
A11527
12111
Cone.
107.33
106.50
106.75
106.65
106.35
106.30
106.12
106.47
106.11
106.43
Cyl.
Code
Age
0
15
29
45
70
80
99
120
139
153
154
170
202
Cyl.
**/ A.
Code
Age
3
21
36
63
73
92
113
132
146
163
195
Cyl.
Code
Age
0
19
38
67
96
117
150
151
152
167
185
A11537
12211
Cone.
105.00
105.00
105.00
103.75
104.89
103.92
103.69
103.47
104.36
104.70
103.07
104.07
104.26
A11547
12311
Cone.
102.50
103.00
102.35
102.22
102.69
102.07
102.43
102.51
102.75
102.31
102.79
A11557
12121
Cone.
108.00
108.00
107.60
107.60
107.39
107.08
108.01
105.96
106.79
106.75
107.37
Cyl.
Code
Age
3
21
41
59
68
89
111
if*
i ftt\
iOU
187
Cyl.
Code
Age
4
21
/ 1
** i
77
QQ
7O
136
1 L 7
I**/
1 &&
1 DO
107
1O /
A11567 ^
12221 <_£
C__. _
one .
102.00 2
i r\ i ne ^^
101.95 ^j
102.40 ^>
102 .00
101 . 99
101 . 88
102 .41
101 . 81
101 .95
102 . 30
i ni T\
1U1 . / j
A11577
12321
Cone.
109.60
108 . 85
I fift Q7
1UO . 7 /
i no OQ
1U3. £y
109 48
i fift An
iUO. 4U
108 78
1 OQ OQ
lUO. O7
mfifi
uo
1 Ah An
IUO HU
V
H^
O
* See page 2-2 for explanation of code.
-------
i
99
B
r*
u
O
n
z
rt
Cyl.
Code
Age
15
27
43
56
79
97
118
140
161
182
205
Cyl.
Code
Age
13
29
42
65
97
98
112
127
148
150
168
197
All 348
21111
556.33
556.00
556.50
555.50
556.33
554.40
555.20
556.97
558.81
559.40
558.75
A11358
21211
Cone.
529.67
525.67
529.50
529.00
528.67
526.70
528.07
525.10
528.79
533.54
530.84
529.80
527.89
Cyl..
Code
Age
4
20
39
54
76
97
118
137
152
167
189
208
Cyl.
Cod*
Age
26
39
54
76
97
118
137
152
155
167
189
208
AU368
21311
523.00
521.50
523.00
519.00
522.50
521.34
522.00
523.14
524.91
521.96
520.78
523.77
A11378
21121
Cone.
330.67
532.00
532.67
536.00
532.00
532.89
529.61
533.74
536.75
534.90
532.64
534.53
537.81
Cyl.
Code
Age
3
23
37
51
72
92
113
134
162
177
196
219
Cyl.
Code
Age
3
23
37
59
79
101
122
144
164
185
204
Cyl.
Code
Age
13
27
49
76
77
91
106
127
147
162
176
TABLE
All 388
21221
Cone.
547.67
552.00
555.00
553.00
553.00
552.70
552.67
549.79
556.20
552.64
548.22
560.19
A11398
21321
Cone.
600.00
605.00
605.50
602.00
604.63
602.01
607.32
607.20
610.70
607.52
605.40
A11408
22122
Cone .
538.00
538.00
537.00
534.00
536.03
538.10
535.42
540.44
540.56
540.77
538.31
538.17
3-10
Cyl.
Code
Age
3
17
33
52
73
95
116
137
160
179
188
Cyl.
Code
Age
20
34
51
99
121
160
181
202
Cyl.
Code
Age
1
15
29
45
70
80
99
120
139
153
170
196
203
ANALYTICAL DATA
A11418
22212
Cone.
517.33
517.50
515.00
518.47
518.06
516.12
519.89
522.10
517.83
516.94
519.73
A11428
22322
Cone.
557.33
555.00
555.00
556.85
553.84
556.93
556.18
555.11
551.43
A11438
22113
Cone.
528.50
525.00
525.00
527.70
524.06
525.54
526.15
529.30
526.48
525.39
526.14
521.10
526.84
Cyl.
Code
Age
3
21
41
61
80
103
116
132
150
151
171
186
187
Cyl.
Code
Age
4
21
41
66
73
117
136
147
166
187
Cyl.
Code
Age
18
32
48
67
83
104
123
138
153
175
194
A11448
22313
Cone.
564.00'
564.55
567.07
564.87
563.65
567.66
565.54
563.54
574.29
567.13
567.45
566.93
566.12
A11458
22223
Cone.
561.00
562.55
562.00
562 . 34
561.84
562.91
563.56
566.00
564.59
564.17
A11468
11112
Cone.
538.33
53B.50
541.50
536.33
538.27
537.39
539.47
540.21
541.54
539.53
544.30
S41.34
FOR PROPANE
Cyl.
Code
Age
5
20
41
63
84
106
126
147
166
194
Cyl.
Code
Age
1
16
37
72
73
91
112
127
143
161
182
183
Cyl.
Code
Age
20
35
57
78
100
141
160
A11478
11213
Cone.
548.25
546.00
547.63
544.67
549.45
549.20
550.89
549.70
545.25
545.94
A11488
11312
Cone.
526.33
521.00
524.95
519.33
528.30
518.84
525.75
522.90
523.82
524.70
524.61
524.92
A11498
11123
Cone.
513.25
512.50
514.45
512.29
515.48
517.82
517.55
513.56
w
P
AT 500 PPM £
Cyl.
Code
Age
21
41
61
80
98
116
132
150
151
171
186
Cyl.
Code
Age
19
28
48
66
73
98
117
136
147
150
166
188
189
All 508
11222
Cone.
510.67
512.55
513.03
511.46
512.29
515.44
513.92
512.43
519.38
512.53
514.10
513.56
A11518
11323
547.67
548.80
543.48
546.71
546.40
550.94
544.04
545.70
552.19
550.55
550.01
543.43
542.61
Cyl.
Code
Age
5
20
34
63
84
103
118
133
155
174
189
190
Cyl.
Code
Age
0
15
29
45
70
80
99
120
139
153
170
174
196
A11528
12111
Cone.
527.33
525.50
523.00
525.73
527.02
526.46
529.75
525.58
523.76
527.32
533.56
527.55
A11538
12211
Cone.
540.75
539.50
538.50
540.10
533.02
542.52
539.63
542.76
539.02
540.57
544.65
538.44
542.10
Cyl.
Code
Age
21
36
58
63
79
101
121
142
161
196
Cyl.
Code
Age
0
19
38
67
96
117
150
167
171
185
All 548
12311
Cone.
517.25
515.00
517.70
516.91
516.91
516.95
520.52
521.05
521.76
516.43
519.11
A11558
12121
Cone.
541.00
537.00
540.32
538.74
541.62
540.43
539.35
533.09
540.92
538.03
Cyl.
Code
3
21
41
59
68
89
111
132
143
144
160
172
179
188
189
Cyl.
Code
Age
4
21
41
66
73
98
117
136
147
166
188
A11568
12221
Cone.
526.20
528.73
526.24
529.96
529.54
534.35
531.53
521.25
527.07
534.16
526.47 .
526.52
532.56
523.51
A11578
12321
Cone.
534.75
537.97
536.90
539.78
537.93
540.22
541.18
538.85
540.18
543.43
538.33
i
O
ON
vl
.p-
* See pages 2-2 -for explanation of code.
-------
4-1
SRL 1317 13 0674
4.0 DATA ANALYSIS
The primary objective of the program was to develop specifications
for producing stable gas mixtures of propane in air and carbon monoxide in
nitrogen in high pressure gas cylinders. A stable mixture was defined as
one in which the concentration of trace gas changes by no more than 3%
from its original concentration over a six month test period.
The test cylinder concentration data shown in Tables 3-1 through
3-10 were analyzed statistically to determine the influence of each of
the five hardware type and mixture preparation variables on stability of
the cylinder mixtures. Separate analyses were performed for each of the
ten component/concentration range combinations. The test variables,
which were described in detail in Section 2.2, included cylinder type,
valve type, preconditioning procedure, mixing procedure after blending,
and purity of diluent gas.
4.1 DETERMINATION OF DETERIORATION RATE
The first step in the data analysis was to determine the average
rate of change of each of the 240 test cylinders during the six month test
period. Plots of the concentration determinations as a function of age
produced irregular lines as illustrated in Figure 4-1, which shows the
concentration-time variations for carbon monoxide at a 50 ppm nominal
concentration. The illustration is for a manganese steel cylinder using
a diaphragm packless brass valve, preconditioned by evacuation only,
mixed by thermal agitation and diluted with 99.7% pure nitrogen with a low
moisture content. In order to establish some definite quantitative measure of
stability, a general trend in the time-deterioration characteristic of the
trace gas had to be developed. The method of least squares was selected for
this purpose. The broken line in Figure 4-1 illustrates the time-concentration
regressed least squares estimate of trace gas concentration deterioration.
Once this least squares estimate was obtained, the percent change in
concentration over the six month period was computed according to Equation 1.
SCOTT RESEARCH LABORATORIES. INC.
-------
FIGURE 4-1 CONCENTRATION - TIME TREND FOR
CARBON MONOXIDE @ 50 PPM NOMINAL CONCENTRATION
o
o
M
SB
n
BE
f
w
o
30
25-
o
B
n
61.0
Least Squares Estimate
o
o
vj
ro
20
40
60 80 100 120 140 160
Time After Blending (Days)
180 200 220
-------
4-3
SRL 1317 13 0674
C±~ °f
% change = -^ x 100 (1)
Ci
where: C. = Concentration at time of preparation.
Cf = Final concentration at end of 180 days.
This percent change in concentration was used as a measure of the stability
of the trace gases. The percent changes for each of the 24 cylinders at
each of five levels of carbon monoxide and propane are shown in Table 4-1
and 4-2 respectively.
4.2 REGRESSION ANALYSIS OF STABILITY DATA
The effect of five qualitative variables on the stability of
trace gases as defined by their percent change in concentration over a
six month period was assessed. The technique of multiple regression was
selected as the method for estimating the magnitude and effect that these
variables had on the percent change in concentration of trace gases in
high pressure cylinders. Since these are qualitative variables in the
sense that they are not defined on a continuous scale of measurement but
consist instead of groups of discrete items, the method outlined in
Reference 1 was utilized.
In this method, where a variable consists of only two types of
items (e.g., cylinder type, valve type or mixing) the value '-!' is
assigned to one of the items and '+!' to the other. When appropriate,
interactions between this variable and any other variable may also be
incorporated in the selection of the regression model. Of course, in
the final regression equation that is selected, this variable can only take
the values ±1. When there are three types of items in a variable (e.g.,
preconditioning or diluent purity), a quadratic term of this variable must
t\
be included in the model, i.e., the equation must have an x. and an x
term (x.^ being the three-item variable). The three items of this variable
are given the values: -1, 0, +1. In the interpretation of the final model
2
the x and x. terms must be considered jointly.
Table 4-3 shows the values assigned to each of the items comprising
the five variables studied. Using these values, a regression matrix of
SCOTT RESEARCH LABORATORIES, INC.
-------
SRL 1317 13 0674
4-4
TABLE 4-1 PERCENT CHANGE IN CARBON MONOXIDE CONCENTRATION
OVER A SIX-MONTH PERIOD
:yi.
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
©
Cylinder
Code
21111
21211
21311
21121
21221
21321
22122
22212
22322
22113
22313
22223
11112
11213
11312
11123
11222
11323
12111
12211
12311
12121
12221
12321
SCOTT RESEARCH
10 ppm
-1.05
0.97
1.99
3.43
-0.15
-0.06
6.08
-1.87
-4.65
1.66
-0.78
-3.38
1.59
-2.83
-12.79
-4.75
-14.85
-1.93
-6.43
-0.73
-22.47
-52.24
-6.79
-3.84
LABORATORIES.
50 ppm
1.32
1.91
2.93
0.48
2.65
1.39
-2.19
0.11
-1.05
1.08
1.19
3.68
-0.91
0.12
-5.07
3.48
-4.20
-1.11
1.58
-4.80
-0.98
-2.03
0.69
0.18
INC.
100 ppm
0.07
-0.68
-0.89
-0.98
-1.45
-2.43
0.29
-1.20
-0.64
-0.40
2.80
2.31
-0.11
0.30
-3.48
-0.34
-7.35
-7.27
-0.88
-2.56
-3.25
-1.26
-4.57
-1.87
500 ppm
0.35
-0.96
-0.55
1.53
2.62
-1.15
-0.17
1.39
-0.81
0.85
2.20
0.84
-0.65
0.90
-0.94
2.26
2.48
0.98
1.02
1.47
1.78
-5.25
1.08
1.79
1000 ppm
-0.04
-0.04
0.13
0.43
-0.81
-1.86
-0.81
0.13
0.62
-0.79
1.64
1.85
-1.58
0.79
-0.51
1.93
0.58
1.03
-0.45
-1.00
0.93
-0.78
1.32
1.10
-------
SRL 1317 13 0674
4-5
TABLE 4-2 PERCENT CHANGE IN PROPANE CONCENTRATION
OVER A SIX-MONTH PERIOD
71.
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Cylinder
Code
21111
21211
21311
21121
21221
21321
22122
22212
22322
22113
22313
22223
11112
11213
11312
11123
11222
11323
12111
12211
12311
12121
12221
12321
SCOTT RESEARCH
3 ppm
-1.68
0.52
-1.55
0.37
1.02
-0.25
0.69
0.08
-0.95
1.49
-0.12
0.17
-2.67
-0.97
-0.84
0.18
-0.02
1.26
0.23
1.47
2.13
1.62
1.76
-0.80
LABORATORIES.
10 ppm
-0.45
1.30
0.53
-0.45
1.24
0.78
0.32
1.75
0.80
0.98
0.47
0.38
0.43
0.81
0.82
0.27
0.18
1.57
0.51
0.84
1.38
0.66
1.39
-0.19
INC.
50 ppm
0.83
0.76
0.40
1.53
0.71
0.51
1.25
1.35
1.17
0.86
0.93
0.80
1.01
0.50
1.06
1.02
0.57
-0.13
1.40
1.47
0.63
-1.87
1.00
0.19
100 ppm
0.04
-1.45
-1.53
-1.01
-1.21
-1.23
-1.32
-0.66
-1.14
-0.80
-0.39
-0.69
-0.81
-0.16
-0.76
-0.65
-0.45
1.63
-0.60
-0.83
0.03
-0.89
-0.12
-0.03
500 ppm
0.55
0.29
0.19
0.68
0.55
0.88
0.52
0.44
-0.41
-0.27
0.60
0.63
0.73
0.03
0.14
0.70
0.66
-0.14
0.62
0.36
0.65
-0.34
0.21
0.86
-------
W
n
ft
B
i
o
O
n
en
Carbon
Monoxide:
Propane:
TABLE 4-3 VALUES ASSIGNED TO VARIABLES IN
MULTIPLE REGRESSION MODEL
Variable
Cylinder
Material X2
Valve
Type X3
Pre-
conditioning X4
Mixing
5
Diluent
Purity X6
Diluent
Purity X6
Type
of Item
1
2
1
2
1
2
3
1
2
1
2
3
1
2
3
Description
Chrome moly steel
Manganese steel
Packed brass with Teflon
Diaphragm packless brass
Evacuation only
Evacuation + N flush + evacuation
same as 2 with heat applied
Mechanical
Thermal
99.997% pure NZ
99.7% pure N_; low moisture
99.7% pure N ; high moisture
Blended HC-free air
Blended N -0
Water pumped
Value
Assigned
-1
+1
-1
+1
-1
0
+1
-1
+1
-1
0
+1
-1
0
+1
TABLE 4-4 EXAMPLE OF REGRESSION MATRIX FOR CARBON MONOXIDE
AT 50 PPM NOMINAL CONCENTRATION
Obs.
No.
1
2
3
4
24
Dep. Variable
% Change
y
1.32
1.91
2.93
0.48
0.18
1
Independent Variables
Cyl.
Type
X2
1
1
1
1
-1
Valve
Type
X3
-1
-1
-1
-1
1
Precon-
ditioning
X4
-1
0
1
-1
1
Mixing
X5
-1
-1
-1
1
1
Oil.
Purity
X6
-1
-1
-1
-1
-1
2
X4
1
0
1
1
1
2
X6
1
1
1
1
1
p
H->
W
U)
o
^
vj
-------
4-7
SRL 1317 13 0674
the type shown in Table 4-4 was developed for each of the five nominal
concentration levels of carbon monoxide and propane.
Standard analysis of variance methods could have been used on
the data collected. Such methods often require balanced statistically
designed experiments and unfortunately, the precise requirements of these
designs cannot always be achieved in practice. In such cases, regression
analysis techniques are very suitable and provide the necessary answers
on the magnitude and effect that independent variables have on the response
of a dependent variable.
One of the most important steps in any regression analysis is the
selection of the "best" independent variables for inclusion in the final
regression equation. Once the appropriate regression model has been
ascertained, analysis of variance techniques can be applied to determine
what effect the independent variables have on the response of the dependent
variable. The dependent variable (y), in our case, is the percent change
in concentration over a six month period. The independent variables are
described in Table 4-3. Since we are considering the effect of more than
one independent variable, a multiple linear regression model was employed.
The basic form of a multiple regression model is:
y = aQ + a!(x2) + a2
-------
4-8
SRL 1317 13 0674
significnatly greater than the tabulated F (at a given P% probability level,
usually 1% to 10%), we can assert with a P% risk that the variate under
consideration contributes significantly to the regression and that there
probably exists a genuine relationship between the independent variable and
the dependent variable. The contribution of the sum of squares due to
the "non-significant" variables is incorporated in the residual term, since
the association (if any) between the "non-significant" independent
variables and the dependent variable can be attributed only to chance or
random causes.
A number of techniques are available for selection of the "best"
regression model. All of these techniques, however, use the F-test for
assessing the significance of a variable. The model which is finally
selected should be simple; incorporate those variables which contribute
significantly in explaining the variation in the data; and, reflect the
situation as observed from practical experience. For the purposes of this
study the technique of Multiple Stepwise Regression was employed in the
selection of the best set of independent variables. In this technique,
variables are added into the model one by one but, at each stage, any
variable which is already included in the model but whose extra sum of
squares contribution has declined to a non-significant level is eliminated.
However, this variable may possibly be brought in again at a latter stage.
Selection stops when all unused variables are non-significant and all the
included variables are significant.
With the percent change in concentration for each level of carbon
monoxide and propane as a dependent variable, the regression variables
shown in Table 4-4 were subject to the above technique. Besides the
variables shown in Table 4-4, interaction and second-order variable effects
were also studied, but no significant effects could be determined. The
results of the multiple stepwise regression and the corresponding analysis
or variance are presented in Tables 4-5 through 4-9 for carbon monoxide
and in Tables 4-10 through 4-14 for propane.
SCOTT RESEARCH LABORATORIES, INC.
-------
4-9
SRL 1317 13 0674
TABLE 4-5 ANALYSIS ON STABILITY OF
CARBON MONOXIDE AT 10 PPM
Multiple Stepwise Regression;
Only cylinder type (x«) is significant.*
Regression Model:
y = aQ + a. (x.) y - Z change
a - -5.2445
o
al = 5.4271
Standard Error of Estimate = 10.63%
Analysis of Variance;
Degrees Mean
Source Sum of Squares of Freedom Squares F-Value % Contribution
X2
Residual
Total
706.88
2485.50
3192.38
1
22
23
706.88 6.26
112.98
22.14
77.86
100.00
Standard Fj^ 22 - 4.30 @ 0.05 probability point.
*Note: Diluent purity most likely has an effect on the stability of carbon
monoxide in cylinders at this concentration; but the presence of extreme
values (see data points 15, 17, 21 and 22 in Table 4-1) associated with
cylinders of type 1 (chrome-moly steel) apparently tends to mask this
effect.
SCOTT RESEARCH LABORATORIES, INC.
-------
4-10
SRL 1317 13 0674
TABLE 4-6 ANALYSIS ON STABILITY OF
CARBON MONOXIDE AT 50 PPM
Multiple Stepwise Regression;
Cylinder type (x2) and diluent purity (x,) are the significant
variables.
Regression Model:
y =
y = % change
a = -2.2182
o
&l = 1.056
a2 = 0.4803
a3 = 3.1425
Standard Error of Estimate = 1.74%
Analysis of Variance;
Degrees Mean
Source Sum of Squares of Freedom Squares F-Value
% Contribution
x
2 2
VX6
Residual
Total
29.34
43.72
60.70
133.76
1
2
20
23
29.34
21.86
3.04
9.65
7.19
21.93
32.69
45.38
100.00
Standard FL 2Q = 4.35 @ 0.05 probability point.
Standard F2 2Q = 3.49 @ 0.05 probability point.
SCOTT RESEARCH LABORATORIES. INC.
-------
4-11
SRL 1317 13 0674
TABLE 4-7 ANALYSIS ON STABILITY OF
CARBON MONOXIDE AT 100 PPM
Multiple Stepwise Regression;
Only cylinder type (X) had any significant effect
Regression Model;
y -
aQ +
a = -1.4924
o
y = % change
al - 1.2260
Standard Error of Estimate
2.12%
Analysis of Variance;
Degrees Mean
Source Sum of Squares of Freedom Squares F-Value
Residual
Total
36.07
98.90
134.97
1
22.
23
36.07
4.50
8.02
% Contribution
26.72
73.28
100.00
Standard FI 22 = 4.30 @ 0.05 probability point.
SCOTT RESEARCH LABORATORIES. INC.
-------
4-12
SRL 1317 13 0674
TABLE 4-8 ANALYSIS ON STABILITY OF
CARBON MONOXIDE AT 500 PPM
Multiple Stepwise Regression:
No variables have any significant effect.
Regression Model;
y = a y - % change
a = 0.5444
o
Standard Error of Estimate = 1.74%
The percent change in concentration for all 24 cylinders is small enough
to be attributable to experimental errors alone. The % change in concentra-
tion is independent of the variables considered.
SCOTT RESEARCH LABORATORIES. INC.
-------
4-13
SRL 1317 13 0674
TABLE 4-9 ANALYSIS ON STABILITY OF
CARBON MONOXIDE AT 1000 PPM
Multiple Stepwise Regression;
No variables have any significant effect.
Regression Model;
y - a y - Z change
o
a - 0.1580
o
Standard Error of Estimate - 1.06%
The percent change in concentration for all 24 cylinders is small enough
to be attributable to experimental errors alone. The percent change in
concentration is independent of the variables considered.
SCOTT RESEARCH LABORATORIES. INC.
-------
A-14
SRL 1317 13 0674
TABLE 4-10 ANALYSIS ON STABILITY OF
PROPANE AT 3 PPM
Multiple Stepwise Regression:
Only valve type (x ) is of any significance.*
Regression Model:
y = aQ + a (x ) y = % change
a = 0.1225
o
a1 = 0.5076
Standard Error of Estimate = 1.11%
Analysis of Variance;
Degrees Mean
Source Sum of Squares of Freedom Squares F-Value % Contribution
x3 6.18 1 6.18 4.98 18.5
Residual 27.25 .22 1.24 81.5
Total 33.43 23 100.00
Standard F. ,= 4.30 @ 0.05 probability point.
i ,z/
*Note: x, is just barely significant. Moreover, the percent change in concen-
tration for all 24 cylinders is well within experimental errors and no real
significance can be attached to any cause-effect relation between valve type
and propane stability in cylinders at the 3 ppm level.
SCOTT RESEARCH LABORATORIES. INC.
-------
4-15
SRL 1317 13 0674
TABLE 4-11 ANALYSIS ON STABILITY OF
PROPANE AT 10 PPM
Multiple Stepwise Regression;
No variables have any significant effect.
Regression Model:
y = a y - % change
a = 0.6797
o
Standard Error of Estimate - 0.59%
The percent change in concentration for all 24 cylinders is small enough
to be attributable to experimental errors alona. The percent change in
concentration is independent of the variables considered.
SCOTT RESEARCH LABORATORIES. INC.
-------
4-16
SRL 1317 13 0674
TABLE 4-12 ANALYSIS ON STABILITY OF
PROPANE AT 50 PPM
Multiple Stepwise Regression:
No variables have any significant effect.
Regression Model;
y = a y = % change
a = 0.7477
o
Standard Error of Estimate = 0.69%
The percent change in concentration for all 24 cylinders is small enough
to be attributable to experimental errors alone. The percent change in
concentration is independent of the variables considered.
SCOTT RESEARCH LABORATORIES, INC.
-------
4-17
SRL 1317 13 067A
TABLE 4-13 ANALYSIS ON STABILITY OF
PROPANE AT 100 PPM
Multiple Stepwise Regression;
Only cylinder type (x») is of any significance.*
Regression Model;
y = aQ + a. (x ) y - % change
a = -0.6262
o
al = -0.3229
Standard Error of Estimate - 0.59%
Analysis of Variance;
Degrees Mean
Source Sum of Squares of Freedom Squares F-Value % Contribution
X2
Residual
Total
2.50
7.65
10.15
Standard F, ,
1
22
23
, = 4.30 <§
2.50
0.35
0.05 probab:
7.20 24
75
100
Lllty point.
.60
.40
.00
*Note; The effect of cylinder type, although statistically significant,
does not have any practical significance. The sum of squares contribution
due to x2 is minimal in this case and the F-value appears significant
because the residual term is very small in comparison.
SCOTT RESEARCH LABORATORIES. INC.
-------
4-18
SRL 1317 13 0674
TABLE 4-14 ANALYSIS ON STABILITY OF
PROPANE AT 500 PPM
Multiple Stepwise Regression;
No variables have any significant effect.
Regression Model;
y = a y = % change
a = 0.3791
o
Standard Error of Estimate = 0.38%
The percent change in concentration for all 24 cylinders is small enough
to be attributable to experimental errors alone. The percent change in
concentration is independent of the variables considered.
SCOTT RESEARCH LABORATORIES, INC.
-------
4-19
SRL 1317 13 0674
The standard error of estimate term shown in these tables gives
an estimate of the variance about regression and represents the error
with which any response in the dependent variable could be predicted from
the independent variables. Note that the square of the standard error of
estimate is the residual term. The residual term, as described earlier,
is the estimate of the variance which deviates from regression. The least
squares property minimizes this sum of squares of residuals in order to
estimate the regression coefficients. Residuals are therefore a form of
error which cannot be explained in terms of the independent variables
used in the model. The F-values shown in the tables are the ratio of
the mean squares due to a "significant" independent variable and the
mean squares due to the residual variance. Standard F-values are presented
below the analysis of variance tables.
4.3 SUMMARY OF EFFECTS OF TEST VARIABLES ON GAS STABILITY
The results of the analysis indicate that the cylinder type
has a definite influence on the stability of carbon monoxide at the
lower and middle concentration levels (10 ppm, 50 ppm and 100 ppm).
Approximately 25% of the variation in the deterioration of carbon monoxide
at these levels is explainable by the type of material used in cylinder
construction. Carbon monoxide mixtures in cylinders of manganese steel
were more stable than those stored in cylinders of chrome-moly steel.
These conclusions were reenforced by observation of rust in all the
chrome-moly cylinders, as discussed further in Section 5.0.
The effect of diluent purity could not be conclusively established,
although it appears that at the lower carbon monoxide concentration levels
(10 ppm and 50 ppm) there might be some degree of influence. No interaction
effects between cylinder type and diluent purity could be determined.
At the middle and higher concentration levels (especially 500 ppm
and 1000 ppm), the percent change in concentration is small enough to be
well within the analytical errors of concentration analysis, and none of
the variables considered had any dominating effect. Valve material, type
of mixing and preconditioning had a negligible effect on the stability of
the test carbon monoxide cylinders.
SCOTT RESEARCH LABORATORIES. INC.
-------
4-20
SRL 1317 13 0674
Analysis on the stability of propane in high pressure cylinders
indicated that at all levels of concentrations, the deterioration over a
six month period was insignificant. The maximum percent change in
concentration for a single cylinder was 2.7% at the 3 ppm level, which
is within the ±3% specification established as a criteria for stability.
4.4 EFFECT OF EXTREME STORAGE TEMPERATURES ON GAS STABILITY
Changes in the temperature of the environment in which the gas
mixtures are stored can effect the equilibrium between the gas phase and
the cylinder walls. This shift in equilibrium can be determined from
measured changes in concentration before and after the temperature
conditioning.
For the purposes of this study, 16 of the 24 cylinders at each
level were placed in hot and cold temperature storage. The remaining eight
cylinders were designated as control cylinders and not subjected to any
extreme ambient influences. Each of the 16 test cylinders were stored for
two 2-week periods at -10 to +20°F and at 90 to 100°F. Thus at each level
32 pairs of before-after concentration data were obtained for a "cold-cycle"
and a similar set of 32 paired observations determined for the "hot cycle."
A well known test for comparing "treatment" effects on paired
data is to perform a t-test on the differences of the pair observations.
By taking differences, extraneous effects which might influence both
members of a pair tend to cancel out, thus leaving only the effect (if any)
of the "treatment."
Table 4-15 is an example of the t-test performed for "hot-cycle"
effects on propane mixtures at the 100 ppm nominal concentration level.
A listing of the paired differences in ascending order of magnitude is
presented, followed by an illustration of the frequency distribution
divided arbitrarily into ten groups. The terms are defined as follows:
<$>
SCOTT RESEARCH LABORATORIES. INC.
-------
4-21
SRL 1317 13 0674
TABLE 4-15 t-TEST ON HOT-CYCLE EFFECT FOR PROPANE
AT 100 PPM NOMINAL CONCENTRATION
-1.
?n
-.MOO
-.5100
-.3700
.'V,on
.4]")°
.0 i.o
1.5100 -
-.7.200
-.1134
.17.7.36
-0.93
.'.n o.o
1.7.000
3.0
SCOTT RESEARCH LABORATORIES. INC.
-------
4-22
SRL 1317 13 0674
x. : Paired differences
R: Range = x^ & x^
Md: Median observation
n
M : Mean =
n
/ n
'^ -.2
SD = Standard Deviation = / f-
SE = Standard Error =
V
SD
t = Computed t-value
SD/- n
The test is now carried out by considering the hypothesis, H: y. = y .
That is, the "before" and "after" treatment observations came from a
universal population with equal means. In other words, there is no
effect of treatment on the two sets of observations. The assertion of
the hypothesis is stated with some degree of risk. This is termed the level
of significance (a) and for this case an a of 0.05 is chosen. From
standard t-tables, for an 0.05 level of significance and 31 degrees of
freedom, the critical t region is -2.042.04. Since the computed t = -0.93 the
hypothesis is accepted that hot temperature storage had no effect on
propane at 100 ppm (with a 5% risk of being wrong).
A summary of the effects of hot and cold storage on propane
stability is given in Table 4-16. Similar data for carbon monoxide are
given in Table 4-17. Only two t values were not between -2.04 and 2.04,
and in both cases the average change was 0.1 ppm or less. On this basis
it is concluded that hot or cold storage had no significant effect on
either propane or carbon monoxide stability at any concentration level.
SCOTT RESEARCH LABORATORIES. INC.
-------
4-23
SRL 1317 13 0674
TABLE 4-16 SUMMARY OF HOT AND COLD STORAGE EFFECTS
ON PROPANE STABILITY
Hot/Cold
Hot
Hot
Hot
Hot
Hot
Cold
Cold
Cold
Cold
Cold
Cone.
500
100
50
10
3
500
100
50
10
3
Mean
Difference
(ppm)
0.124
-0.113
0.006
0.043
-0.010
0.653
-0.005
0.058
0.005
-0.143
Standard
Dev.
3.210
0.692
0.363
0.082
0.053
2.646
0.572
0.474
0.098
0.047
t Value
0.22
-0.93
0.09
2.92
-1.03
1.40
-0.14
0.69
0.31
-1.73
TABLE 4-17 SUMMARY OF HOT AND COLD STORAGE EFFECTS
ON CARBON MONOXIDE STABILITY
Hot/Cold
Hot
Hot
Hot
Hot
Hot
Cold
Cold
Cold
Cold
Cold
Cone.
(ppm)
1000
500
100
50
10
1000
500
100
50
10
Mean
Difference
(ppm)
-3.465
1.408
-0.587
-0.239
-0.051
3.183
1.327
0.086
0.193
0.116
Standard
Dev.
20.344
7.322
1.686
0.727
0.384
19.547
8.209
1.686
1.044
0.235
t Value
-0.96
1.09
-1.97
-1.86
-0.75
0.92
0.91
0.29
1.04
2.79
SCOTT RESEARCH LABORATORIES. INC.
-------
4-24
SRL 1317 13 0674
4.5 EFFECT OF CYLINDER PRESSURE ON GAS STABILITY
At each concentration level of carbon monoxide or propane the
pressure in 16 cylinders was periodically reduced from a blending pressure
of about 2000 psi to about 500 psi at the end of six months. The eight
control cylinders were maintained at essentially the blending pressure.
To determine the effect of reduction in cylinder pressure on the
stability of the trace gases, the percent change in concentration of the
controlled and uncontrolled cylinders over a six month period was
subjected to a t-test. The t-test here is on the means of two populations
with unequal variances. The hypothesis states that the two populations
have the same mean, or in other words, there is no effect of reduction in
cylinder pressure on the stability of trace gases. A 0.05 level of
significance is chosen and the t-statistic is computed as:
and the degrees of freedom are computed as:
N
where: x = mean of concentration changes from control cylinders.
x? = mean of concentration changes from uncontrolled cylinders.
2
S1 = variance of concentration changes from control cylinders.
2
S = variance of concentration changes from uncontrolled cylinders.
SCOTT RESEARCH LABORATORIES, INC.
-------
4-25
SRL 1317 13 0674
From the computed degrees of freedom and for a 0.05 level of
significance, the critical t region is obtained at each concentration
level of carbon monoxide and propane. If the computed t value lies
outside this critical region, the hypothesis that reduction in cylinder
pressure has no effect on the stability of trace gases in cylinders is
rejected.
Table 4-18 presents the results of this t-test for carbon monoxide
and propane. All the computed t values lie well within the critical region
and hence the hypothesis that cylinder pressure has no effect on the stability
of trace gases in cylinders is accepted.
4.6 EVALUATION OF THE EFFECT OF REGULATOR TYPE
A separate test program was performed to study the effect of
cylinder regulator type on the concentration of carbon monoxide and propane
in the test mixtures. In this study three control cylinders of carbon
monoxide in nitrogen and five control cylinders of propane in air were used.
The regulator types evaluated were:
1. Scott Model 2A - single stage, forged brass body, neoprene
diaphragm.
2. Scott Model 10A - two stage, forged brass body, Teflon-faced
diaphragm.
3. Scott Model 12A - single stage, nickel plated brass body,
stainless steel diaphragm, Kel-F and nylon seal.
In addition to the three regulators, a nickel plated brass needle valve was
used without a regulator. Past experience had shown that the needle valve,
while not practical for general use, eliminated trace gas adsorption and
was particularly effective for transferring low concentrations of high
molecular weight (>C,) hydrocarbons.
o
Each test cylinder was analyzed using each regulator and the
needle valve in three separate runs. The instrument response for each
regulator was then compared to that obtained with the needle valve. The
data for carbon monoxide are shown in Table 4-19. The data show that the
regulators delivered well in excess of 99% of the concentration delivered
by the needle valve except for the 98.6% delivery found for the Model 10A
at the two higher concentrations. The slightly lower concentration found
SCOTT RESEARCH LABORATORIES, INC.
-------
©
SCOTT
go
w
MI
>
n
SC
r
S3
O
SB
o
JO
5
2
P
TABLE 4-18 EFFECT OF REDUCED CYLINDER P
Cone.
Component (ppm)
Carbon Monoxide 10
,50
100
500
1000
Propane 3
10
50
100
500
PRESSURE ON STABILITY OF TRACE GASES
- - 9 9
xl
-3.996
-0.532
-0.547
0.367
-0.120
0.658
0.445
0.609
-0.892
0.481
X2
-5.869
0.294
-1.965
0.633
0.297
0.145
0.797
0.817
-0.493
0.362
Sl
66.484
6.151
2.534
1.453
0.978
1.577
0.316
1.109
0.202
0.619
S2
180.551
5.804
7.100
3.716
1.189
1.264
0.335
0.197
0.526
0.157
df
23
15
23
23
17
14
16
9
23
9
OJ
i
1 »
Computed t Critiral R^o-inr.
0.
-0.
1.
-0.
-0.
0.
-1.
-0.
-1.
0.
42
78
63
41
94
98
43
54
66
40
-2
-2
-2
-2
-2
-2
-2
-2
-2
-2
.07
-------
©
i
99
n
o
*
>
o
|
n
UI
p
~~ , i? wnin».j.0vn ur run,UJ.AiUK TYPES IN ANALYSIS OF CARBON MONOXIDE
*}
Cone.
Cylinder No. (ppm) Run f
A- 11403 1105 1
2
3
Ave.
A-11481 110 1
2
3
4
Ave.
A-11539 10.3 1
2
3
Ave.
Ins trum
2A
19.83
19.83
19.02
19.56
15.49
15.47
15.03
15.17
15.29
6.92
6.94
7.03
6.96
ent Respon
10A
19.43
19.67
19.08
19.39
15.30
15.39
14.77
15.10
15.14
6.96
6.92
7.03
6.97
se (Cm) fo
12A
19.66
19.99
19.28
19.64
15.53
15.53
14.99
15.00
15.26
6.96
6.93
7.01
6.97
r Model Ave. Response as % of N.V. £
N-V. ZA 10A 12A ^
19.97 w
19.78 §
19.22 *"
19.66 99.5 98.6 100.0
15.56
15.56
15.05
15.27
15.36 99.5 98.6 99.3
7.00
6.96
7.05
7.00 99.4 99.6 99.6
.e-
NJ
N.V. - needle valve.
-------
4-28
SRL 1317 13 0674
with the Model 10A (two stage) was reasonably consistent from run to run,
but it is within experimental error for the analyzer.
The evaluation of regulators for use with propane in air mixtures
included measurements of total hydrocarbons as well as propane because
some regulators, especially those with rubber diaphragms, have been known
to add hydrocarbons to the gas passing through them. The propane tests
were conducted similarly to the carbon monoxide tests described above. The
data are given in Table 4-20. Two of the test cylinders (A-11526 and
A-11534) were blended with hydrocarbon-free synthetic air, and thus the
propane and total hydrocarbons are similar. The other three cylinders
were blended with synthetic air containing methane and other hydrocarbons
(see page 2-5). In these cases the total hydrocarbons exceed the propane.
The data for both propane and total hydrocarbons show variations
well within the experimental error of the instruments. This indicates
that any regulator type tested may be used. However, caution must be
exercized. The test regulators were all new and had not been exposed to
higher molecular weight hydrocarbons. These compounds can readily adsorb
on inside regulator surfaces and be desorbed in subsequent use. Thus,
when first using any regulator with uncertain previous use history, it is
desirable to test the regulator with hydrocarbon-free air to determine if
any previously adsorbed hydrocarbons will be released into the gas stream.
SCOTT RESEARCH LABORATORIES. INC.
-------
o
H TABLE 4-20 COMPARISON OF REGULATOR TYPES IN ANALYSIS OF PROPANE
VI
M
f. Cylinder No.
S A-11498
25 A-11487
5 A-11526
5 A-11505
A-11534
P
Propane Cone, (ppm) for Model
2A
514.0
106.2
53.3
17.42
3.04
10A
515.0
106.1
53.3
17.35
3.03
12A
515.3
106.5
53.4
17.40
3.01
N.V.
513.6
106.3
53.2
17.38
3.05
LO
0
Total HC ^
as Propane (ppm) for Model ^
2A
527.5
109.5
52.7
19.60
3.02
10A
528.0
109.0
53.1
19.84
3.07
12A
529.8
109.4
52.8
19.81
3.02
N>
VO
N.V. - needle valve
-------
5-1
SRL 1317 13 0674
5.0 DISCUSSION OF RESULTS
5.1 STABILITY OF CARBON MONOXIDE MIXTURES
The most significant finding of this study was the substantial
losses in carbon monoxide concentration which occurred when mixtures were
prepared in chrome-moly cylinders. This was most evident at the low
concentration levels (10 to 100 ppm). This loss of carbon monoxide was
not entirely unexpected since similar poor stability had been experienced
in actual use conditions in random cylinders. However, no study had been
made on a sufficiently large batch of cylinders to pinpoint the cause of
instability.
The interior surfaces of those cylinders in which significant
losses occurred were examined and compared to those with good stability.
The surfaces of the manganese steel cylinders which provided good stability
were reasonably smooth and coated with a gray-black film. On the other
hand, troublesome chrome-moly cylinders had surfaces with large rust
colored blotches. Those chrome-moly cylinders in which the greatest
losses of carbon monoxide occurred appeared to have more rust colored
areas than those which exhibited modest loss in carbon monoxide. It can
be theorized that the rusty surfaces lead to carbon monoxide losses by
accelerating the formation of iron carbonyls. No efforts were made to
pursue this theory because the extensive work necessary could not be
justified in terms of its practical value.
The conclusion that the use of chrome-moly cylinders for
storage of 10 to 100 ppm carbon monoxide mixtures is likely to produce
poor stability is clearly demonstrated by the data in Table 4-2. Thus,
it is recommended that manganese steel cylinders be used for these
mixtures until it can be demonstrated that other materials provide
equivalent stability. The only disadvantage of manganese steel cylinders
is their additional weight of 20 pounds which leads to higher transportation
costs. The additional costs are very modest when compared to the value of
a stable calibration gas mixture.
SCOTT RESEARCH LABORATORIES, INC.
-------
5-2
SRL 1317 13 0674
Some chrome-moly cylinders had much greater losses than others
at the same concentration level. This random effect tended to mask the
influence of the other test variables. Diluent purity produced a statis-
tically significant effect at the 50 ppm level. The regression analysis
indicates that the diluent nitrogen with lower purity (higher oxygen content)
tended to accelerate losses of carbon monoxide. Where the diluent also
contained higher moisture, this effect appears to have been reversed.
The use of high purity nitrogen (>99.997%) in manganese cylinders produced
satisfactory stability at all concentration levels. It is readily available
at low cost and thus should be the diluent of choice for carbon monoxide
mixtures, especially in the 10 to 100 ppm range.
The other test variables: Valve type, cylinder preparation
procedure and mixing procedure had no effect at any concentration level.
Any of the valves or preparation and mixing procedures used should be
satisfactory for producing stable mixtures of carbon monoxide in nitrogen.
5.2 STABILITY OF PROPANE MIXTURES
All of the 120 test cylinders of propane in air satisfied the
stability criteria of less than 3% change in six months. The only
variables shown to have a statistically significant effect at the 95%
confidence level were valve type at 3 ppm and cylinder type at 100 ppm.
In both cases the effect was less than 1%, thus within experimental error
of the analytical measurements.
While the purity of the diluent gas had no effect on the stability
of the propane, it should be pointed out that the hydrocarbons other than
propane present in the diluent can lead to serious calibration errors
when the propane mixture is used for calibration of flame ionization
detector total hydrocarbon analyzers. It is generally assumed by the user
that the propane represents the total hydrocarbons in the mixture. The
test cylinders prepared with blended air and water pumped air contained
SCOTT RESEARCH LABORATORIES. INC.
-------
5-3
SRL 1317 13 0674
approximately 1.8 and 1.0 ppm propane equivalent respectively, of other
hydrocarbons. Errors due to non-propane hydrocarbons can be reduced by
having the cylinder analyzed for total hydrocarbon by the supplier. For
best results propane mixtures should be prepared in hydrocarbon-free air
(<0.1 ppm-C) and analyzed for both propane and total hydrocarbons. This
procedure will detect cases where hydrocarbons may have entered the
cylinders from improperly flushed manifolds or other sources.
5.3 STORAGE OF CARBON MONOXIDE AND PROPANE MIXTURES
The data analysis described in Section 4.4 showed that storage
at 0° and 95° F had no measurable effect on stability of carbon monoxide
or propane mixtures in the range of concentrations tested. Low temperature
storage of higher molecular weight hydrocarbons can lead to losses through
condensation on cylinder walls. This had been a serious problem when hexane
was used as the standard hydrocarbon. High temperature storage can
accelerate reactions which cause losses of mixture components. It can also
result in desorption of material present from previous cylinder useage.
This study has shown that mixtures of carbon monoxide and
propane can be stored over a wide temperature range without loss of
stability. It is nevertheless desirable to allow the cylinders to
equilibrate to room temperature before use. In any event, the temperature
of the calibration gas and sample gas to be analyzed must be at the same
temperature if an instrument is to produce accurate data.
5.4 GENERAL COMMENTS
This study has been limited to effects where new cylinders,
valves and regulators are used. Any attempt to evaluate situations
where materials had been exposed to other mixtures would have required a
program of impractically large magnitude. In actual practice, cylinders
are refilled many times in their useful life. Most suppliers try to
re-use a cylinder for the same type of mixture as it previously contained.
Periodic cleaning of cylinders may be performed.
SCOTT RESEARCH LABORATORIES. INC.
-------
5-4
SRL 1317 13 0674
This does not mean that cylinders should not be refilled or even
that refilled cylinders are more likely to lead to more stability problems
than new cylinders. Rather, it must be realized that producing gas
mixtures following the recommendations developed in this program will
assure but not guarantee satisfactory stability. The possibility
always exists that an unknown and indeterminable factor will reduce
stability. Thus, in addition to purchasing gases prepared to the recommended
specifications, a user should plan a periodic check of his calibration
mixtures. This is best accomplished by participation in a commercial gas
cross-reference service, but comparison with primary standards or other
mixtures in possession of the user can also serve to check stability.
Inaccurate span and zero gases are most probably the largest
single contributors to errors in gas concentration measurements. Therefore,
quality assurance in the manufacture and use of gas mixtures is essential
to all programs involving air monitoring or measurement of emissions from
mobile and stationary sources.
SCOTT RESEARCH LABORATORIES, INC.
-------
6-1
SRL 1317 13 0674
6.0 RECOMMENDATIONS
6.1 CARBON MONOXIDE IN NITROGEN MIXTURES
Based on the data analysis and other relevant information acquired
in this study, the following practices are recommended for producing stable
mixtures of carbon monoxide in nitrogen in the range of 10 to 1000 ppm:
1. Mixtures of 10 to 500 ppm must be blended in manganese
steel cylinders meeting DOT Specification 3A2015. While
the use of chrome-moly cylinders is satisfactory at
concentrations above 500 ppm, substantial losses
frequently occur in the 10-100 ppm range.
2. Cylinder valves may be either brass with Teflon
packing or diaphragm packless brass.
3. For new cylinders, preconditioning by evacuation at
ambient temperature to 0.1 in. Hg absolute pressure is
adequate.
4. Mixing after blending may be accomplished by either
mechanical or thermal methods.
5. The diluent nitrogen should have a purity of 99.997%
or greater. The use of diluent gas with a lower purity
(higher oxygen content) can accelerate loss of carbon
monoxide. This effect may cause unsatisfactory
stability in the 10 to 100 ppm carbon monoxide range.
The following recommendations apply to the storage and use of
carbon monoxide in nitrogen mixtures:
1. Cylinders may be stored at temperatures from -10 to 100 F
without loss of stability. For best results, the cylinders
should be stored in the laboratory for 24 hours so that
the contents reach ambient temperature before use.
2. Cylinder regulators having bodies of brass, aluminum or
stainless steel and diaphragms of neoprene, Teflon-faced
metal or stainless steel are all satisfactory.
SCOTT RESEARCH LABORATORIES, INC.
-------
6-2
SRL 1317 13 0674
6.2 PROPANE IN AIR MIXTURES
Based on the data analysis and other relevant information
acquired in this study, the following practices are recommended for producing
stable mixtures of propane in air from 3 to 500 ppm:
1. Mixtures may be blended in steel cylinders meeting DOT
Specification 3AA2015 or 3A2015.
2. Cylinder valves may be either brass with Teflon
packing or diaphragm packless brass.
3. For new cylinders, preconditioning by evacuation
at ambient temperature to 0.1 in. Hg absolute pressure
is adequate.
4. Mixing after blending may be accomplished by either
mechanical or thermal methods.
5. Satisfactory propane stability can be achieved using
blended hydrocarbon-free air, blended nitrogen and
oxygen or water pumped air as the diluent gas.
However, the use of hydrocarbon-free air offers the
advantage of excluding methane and other hydrocarbons
which must be taken into account when using the gas
mixtures for calibration of flame ionization total
hydrocarbon analyzers.
6. All cylinders should be analyzed for total hydrocarbons
as well as propane. This will determine whether other
hydrocarbons have entered the mixture via the diluent
gas, dirty blending system, cylinder residuals, etc.
The following recommendations apply to the storage and use of
propane in air mixtures:
1. Cylinders may be stored at temperatures from -10 to 100 F
without loss of stability. For best results, the cylinders
should be stored in the laboratory for 24 hours so that
the contents reach ambient temperature before use.
SCOTT RESEARCH LABORATORIES, INC.
-------
6-3
SRL 1317 13 0674
2. Cylinder regulators having bodies of brass, aluminum
or stainless steel and diaphragms of neoprene, Teflon-
faced metal or stainless steel may all be used if
clean and new. Past experience has shown that
neoprene diaphragms can cause problems if they have been
exposed to higher hydrocarbons. Thus, extreme caution
must be exercised with neoprene diaphragms, and it is a
better practice to use metal or Teflon-faced diaphragms.
SCOTT RESEARCH LABORATORIES. INC.
-------
7-1
SRL 1317 13 0674
7.0 ACKNOWLEDGEMENTS
We wish to express our thanks to the Project Officer, Mr. John
H. Margeson and to Dr. John B. Clements, Chief of the Methods Standardization
Branch, Quality Assurance and Environmental Monitoring Laboratory for their
assistance in the planning and performance of the gas stability program.
Mr. Margeson's comments and helpful suggestions were of great benefit
throughout the entire program period.
SCOTT RESEARCH LABORATORIES. INC.
-------
8-1
SRL 1317 13 0674
8.0 REFERENCES
1. Davies, O.L., Goldsmith, P.L., "Statistical Methods in Research and
Production," Hafner Publishing Co., New York, 1972.
2. Dixon, N.J., Massey, F.J., "Introduction to Statistical Analysis,"
McGraw-Hill Book Company, Inc., 1957.
3. Stepwise Multiple Regression program obtained from UNIVAC STAT-PACK
Routine: RESTEM.
SCOTT RESEARCH LABORATORIES. INC.
-------
TECHNICAL REPORT DATA
(Please read luusiictioiis on the reverse before completing)
RtPOftT NO
EPA 650/4-74-020
2.
3. RECIPIENT'S ACCESSION-NO
j. TITLE AND SUBTITLE
Development of Technical Specifications For
Standard Gas-Diluent Mixtures for Use in
Measurement of Mobile Source Emissions
5. REPORT DATE
June 1974
6. PERFORMING ORGANIZATION CODE
7. AUTHOH(S)
Louis R. Reckner
8. PERFORMING ORGANIZATION REPORT NO
9 PERFORMING OR'ANIZATION NAME AND ADDRESS
Scott Research Laboratories, Inc.
Plumsteadville, Pa. 18949
10. PROGRAM ELEMENT NO
1HA327
11. CONTRACT/GRANT NO.
68-02-0652
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVE RED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT T~
The stability of gas mixtures of carbon monoxide in nitrogen and propane in
air in high-pressure cylinders was investigated in 240 test cylinders over a six
month period. The effect of several variables related to the preparation of the
mixtures by the suppliers, the storage of the cylinders and their use by laboratories
engaged in emission measurements was studied. The variables included:
1. Cylinder wall material; 2. Cylinder valve type; 3. Cylinder preconditioning
procedure; 4. Concentration of carbon monoxide and propane; 5. Purity of diluent
nitrogen and air,; 6. Mixing procedure after blending; 7. Temperature at which
cylinders are stored; 8. Cylinder pressure; 9. Type of pressure-reducing regulator
used.
The concentration data obtained by periodic analysis of the 240 cylinders over
the six month period were subjected to statistical analysis by multiple stepwise re-
gression. The effects of the individual variables are discussed, and recommended
practices for assuring stable mixtures of carbon monoxide in nitrogen and propane in
air are presented.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI Held/Group
Air Pollution
Carbon Monoxide
Propane
Stability
Gas Cylinders
Nitrogen
Air
Tests
Specifications
Standard Gas-Diluent
mixtures
Mobile Source Emissions
13B
7B
7C
Unlimited
19. SECURITY CLASS (This Report I
Unclassified
21. NO. OF PAGES
70
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-I (9-73)
------- |