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)

-------