-------
Scott Research Labs., Inc.
P ro j e c t
-------
Scott Research Labs., Inc.
Project #2608 4-4 March 6, 1970
Table 4.1 Displaced Vapor Test Results
(cont.)
C0L=
DISPL
TANK
FILL*
RATE
TAMB
RVF
TFUEL
7VAP
TVF
H/C
15-1
42.0
2
B
12.6
90
7.2
87.0
88.0
6.3
2.40
42.0
16-1
36.4
1
N
4.8
90
7.2
92.0
89.0
6.3
2.40
36.4
17-1
37.5
1
N
12.6
90
7.2
94.0
88.0
6.3
2.40
17-2
30.5
1
N
12.6
90
7.2
87.0
88.0
6.3
2.40
17-3
28.1
1
N
12.6
90
7.2
90.0
87.0
6.3
2.40
32.0
18-1 45.4 1 B 12.6 90 7.2 94.0 91.0 6.3 2.40
45.4
19-1
19-2
33.7
33.3
33.5
N
N
4.8 60 9.2 58.0 58.0 4.7 2.31
4.8 60 9.2 62.0 62.0 4.7 2.31
20-1
34.6
2
N
12.4
60
9.2
62.0
60.0
4.7
2.31
20-2
31.2
2
N
12.6
60
9.2
60.0
60.0
4.7
2.31
20-3
27.6
2
N
12,6
60
9.2
62.0
60.5
4.7
2.31
31.1
21-1
40.4
2
B
12.4
60
9.2
62.0
61.0
4.7
2.31
21-2
30.8
2
B
12.6
60
9.2
58.0
57.0
4.7
2.31
21-3
30.3
2
B
12.6
60
9.2
58.0
59.0
4.7
2.31
33.8
22-1
35.7
1
N
4.8
60
9.2
62.0
54.0
4.7
2.31
22-2
29.4
1
N
4.8
60
9.2
58.0
59.0
4.7
2.31
32.6
23-2
32,4
1
N
12.6
60
9 .2
58.5
61.0
4.7
2.31
23-3
32.4
1
N
12.6
60
9.2
59.0
59.0
4.7
2.31
23-4
27.1
1
N
12.6
60
9.2
60.0
60.0
4.7
2.31
30.7
24-1
37.0
1
B
11.8
60
9.2
60.0
60.0
4.7
2.31
24-2 •
31.1
1
B
12.6
60
9.2
59. 0
59.0
4.7
2.31
34.1
25-1
37.2
2
N
4.8
75
9.2
70.0
71.0
6.2
2.31
25-2
33.0
2
N
4.9
75
9.2 ..
76.0
76.0
6.2
2.31
35.1
* N = Nozzle Fill; B = Bottom Fill
-------
Scott Research Labs., Inc.
Project #2608
4-5
March 6, 1970
Table 4.1 Displaced Vapor Test Results
(c on t .)
CQL =
DISPL
TANK
FILL*
RATE
TAMB
RVT
TFL'EL
TVAP
TVP
H/C
26-2
25.5
2
N
12.5
75
9.2
76.5
71.0
6.2
2.31
26-3
26.9
i
N
12.8
75
9.2
76.0
72.0
6.2
2.31
26-4
34.1
->
N
12.6
75
9.2
76.0
72.0
6.2
2.31
28.8
27-1
41.3
2
B
12.6
75
9.2
72. 0
72.0
6.2
2. 31
•*>7 _ 0
40.8
2
B
12.5
75
9.2
75.0
73.0
6.2
2.31
41.0
28-1
35 .2
1
N
4.8
75
9.2
74.0
70.0
6.2
1. 31
28-2
32 .9
1
N
5.0
75
9.2
77.0
73,0
6.2
2.31
34.0
29-1
36 .8
1
N
12.6
75
9.2
73.0
70.0
6.2
2.31
29-2
31.4
1
N
12.6
75
9.2
74.0
72. 0
6.2
2.31
29-3
32.9
1
N
12.6
7 5
9.2
74.0
73.0
6.2
2.31
33.7
30-2
41 .6
1
B
12.6
7 5
9.2
73.0
72 . 0
6.2
2.31
30-3
41.9
• 1
B
12.6
75
9.2
74 .0
74.0
6.2
2.31
41.7
31-1
28.4
2
N
4.8
90
9 .2
85.0
82.0
8.2
2.31
31-2
36.5
0
N
4.8
90
9.2
88.0
88.0
8.2
2.31
32.4
32-1
23.0
1
N
12.0
90
9.2
86.0
82.0
8.2
2.31
32-2
19.4
2
N
12.6
90
9.2
88.0
82.0
8.2
2.31
32-3
32.0
2
N
12.6
90
9.2
90.0
82 .0
6.2
2. 31
24.8
33-1
54 .5
2
B
12.0
90
9.2
89.0
91. 0
8.2
2.31
33-2
53.0
2
B
12.6
90
9 .2
90.5
92. 0
8.2
2 .31
53.8
34-1
38.3
1
N
4.8
90
9.2
93. 0
83.0
8.2
2.31
34-2
34.5
1
N
4.8
90
9.2
93.0
90.0
8.2
2.31
36 .4
35-1
31.1
1
N
12.0
90
9.2
87.5
84 .0
8.2
2.31
35-2
32 .8
1
N
12.6
90
9.2
89.0
89.0
8.2
2.31
35-3
33.9
1
N
12.6
90
9.2
91.0
87.0
6.2
2.31
32.6
* N = Nozzle Fill; B = Bottom Fill
-------
Scott Research Labs., Inc.
Project #2608
4-6
March 6, 1970
Table 4.1 Displaced Vapor Test Results
(cant.)
col=
DISPL
TANK
FILL*
RATE
TAMB
RVP
TFUEL
TVAP
TVP
H/C
36-1
53.6
1
B
12.0
90
9.2
89.0
89.0
8.2
2.31
36-2
52.1
1
B
12.6
90
9.2
89.0
92.0
8.2
2.31
52.9
79-1
29 .2
2
N
5.0
45
9.8
47.5
47.0
3.8
2.40
79-2
28.8
2
N
4.9
45
9.8
48.0
48.0
3.8
2.40
29.0
-
80-1
32.4
2
N
12.6
45
9.8
44.0
44.0
3.8
2.40
80-2
31.8
2
N
12.6
45
9.8
45.0
44.0
3.8
2.40
32.1
81-1
24.8
2
B
12.6
45
9.8
45.5
44.0
3.8
2.40
81-2
27.9
2
B
12.6
45
9.8
45.0
44.0
3.8
2.40
26.4
82-1
30.2
1
N
4.8
45
9.8
46.0
44.0
3.8
2.40
82-2
26.5
1
N
5.0
45
9.8
46.0
46.0
3.8
2.40
82-3
30.1
I
N
4.8
45
9.8
47.0
44.0
3.8
2.40
28.9
83-1
26.9
1
N
12.6
45
9.8
47.0
45.0
3.8
2.40
83-2
26.1
1
N
12.6
45
9.8
47.0
44.0
3.8
2.40
26.5
84-1
26.2
1
B
12.6
45
9.8
47.0
44,0
3.8
2.40
84-2
27.0
1
B
12.6
45
9.8
47.0
44.0
3.8
2.40
26.6
85-1
36.2
2
N
4.8
60
9.8
62.0
62.0
5.1
2.40
85-2
36.4
2
N
4.8
60
9.8
59.0
58.0
5.1
2.40
36.3
86-1
32.7
2
N
12.6
60
9.8
62.0
61.0
5.1
2.40
86-2
34.6
2
N
12.6
60
9.8
58.0
58.0
5.1
2.40
33.7
87-1
32.2
2
B
12.6
60
9.8
58.0
59.0
5.1
2.40
87-2
34.1
2
B
12.6
60
9.8
58.0
59.0
5.1
2.40
33.1
88-1
30.4
1
N
5.0
60
9.8
59.0
58.0
5.1
2.40
88-2
33.3
1
N
4.8
60
9.8
58,0
59.0
5.1
2.40
31.9
* N
= Nozzle
Fill;
B = Bottom
Fill
-------
Scot t
Research
Labs
. , Inc.
Project #2608
4-7
March
6, 1970
Table 4.1
Displaced Vapor Test
Results
(cont.)
C0L=
DISPL
TANK
FILL*
RATE
TAMB
RVP
TFUEL
TVAP
TVP
H/C
89-3
31.6
1
N
12.6
60
9.8
58.0
60.0
5.1
2.40
89-4
28.5
1
N
12.6
60
9.8
58.5
58.0
5.1
2.40
30 .0
90-1
32,6
1
B
12.6
60
9.8
58.0
59.0
5.1
2.40
90^2
31.8
1
B
12 .6
60
9.8
58.0
59.0
5.1
2.40
32.2
91-1
40 .6
2
N
4.8
75
9 .8
75.0
72.0
6.7
2.40
91 -2
35.5
2
N
4.7
75
9.8
73.5
71.0
6.7
2.40
38.1
92-1
41.4
2
N
12.6
75
9.8
73.5
71.0
6.7
2.40
92-3
36.0
2
N
12.6
75
9.8
73.5
68.0-
6.7 •
2.40
92-4
34.7
2
N
12.6
75
9 .8
73.0
70.0
6.7
2.40
92-5
32.5
2
N
12.6
75
9.8
73.0
70.0
6.7
2.40
36 .2
93-1
43.8
2
B
12.6
75
9.8
73-0
73. 0
6.7
2.40
93-2
43.8
2
B
12.6
75
9.8
73.0
72.0
6.7
2.40
43.8
94-1
45 .4
1
N
4. 7
75
9.8
74.0
71.0
6.7
2.40
94-2
43.7
1
N
4.8
75
9 .8
72 . 0
72. 0
6.7
2.40
44,6
95-1
44.5
1
N
12.6
75
9.8
74. 0
73.0
6.7
2.40 ¦
95-2
41. 2
1
N
12.6
75
9.8
74.0
73.0
6.7
2.40
95-3
37.2
1
N
9.7
75
9.8
74.0
72. 0
6.7
2,40
95-4
35.0
1
N
12.6
75
9.8
74.0
72. 0
6.7
2.40
95-5
38.3 ¦
1
N
12.6
75
9.8
74.0
74.0
6.7
2.40
95-6
38.5
1
N
12.6
75
9.8
73.0
74.0
6.7
2 .40
39.1
96-1
48.2
1
B
12.6
75
9.8
74.0
72.0
6.7
2.40
96-2
46.4
1
B
12.6
75
9.8
73.0
74.0
6.7
2.40
A 7 1
NOTE; Whereas temperature of all gasoline dispensed in Tests #1 through #96
was conditioned to 60 F, temperature of gasoline dispensed in the
following tests, #101 through #20 J-5, was conditioned to equal ambient.
* N = Nozzle Fill; B = Bottom Fill
-------
Scott Research Labs,, Inc.
Project #2608
4-8
March 6, 1970
Table 4.1 Displaced Vapor Test Results
(cont,)
C0L=
DISPL
TANK
FXLL
RATE
TAMB
RVP
TFUEL
TVA?
TVP
H/C
101-1 .
42.9
2
N
12.8
76
9.2
74.0
76.0
6 .2
2.31
101 —2
38.0
2
N
12.7
74
9.2
74.0
75.5
6.2
2.31
40.5
102-1
36.6
2
B
12.6
74
9.2
76.0
74.0
6.2
2.31
102-2
35.8
2
B
12.6
76
9.2
75.0
76.0
6.2
2.31
36.2
103-1
32.4
1
N
12.6
74
9.2
76.0
74.0
6.2
2.31
32.4
104-1
33.3
1
B
12.6
73
9.2
74.0
73.0
6.2
2.31
33.3
105-1
' 58.6
2
N
12.6
89
9.2
87.0
89.0
8.2
2.31
58.6
106-1
23.4
2
B
12.6
89.5
9.2
89.0
89.5
8.2
2.31
106-2
35.4
2
B
12.6
91
9.2
91.0
91.0
8.2
2.31
29.4
107-1
51.3
1
N
12.6
90
9.2
90.0
90.5
8.2
2.31
51.3
108-1
46.4
1
B
12.6
90
9.2
90.0
90.0
8.2
2.31
46 .4
20 J-5
55.3
2
N
12.6
64
9.2
74.0
70. 0*
4.7
2.31
55.3
* Temperature of dispensed gasoline = 85.5°F and vapors displaced = 79°F in
Test if20J-5.
** N = Nozzle Fill; B = Bottom Fill
-------
Scott Research Labs,, Inc.
Project -''2608 4-9 March. 6, 1970
4,1.1 Displaced Vapor Weight, Tabular Results - continued
The relative effects on displaced vapor losses are discussed in
subsequent sections for the following parameters:
o Ambient temperature
o Fill method
o Tank configuration
o Reid vapor pressure
o Fill rate
o Dispensed fuel temperature
The graphical results shown in the following pages fall into three
major types :
o Displaced Loss (grans) versus Temperature
(Fahrenheit, Ambient)
o Displaced Loss (grams) versus Reid Vapor
Pressure (psig)
o Displaced Loss (grams) versus Fill Rate
(Gallons/Minute)
Within each type, one or more sets of graphs were prepared for
diverse parameters. Continuous functions are plotted on each graph, as well
as actual data points. These functions were obtained by means of a first or
second order polynomial regression. These curves are intended only to illus-
trate the general characteristics of the functions underlying the data pre-
sented. They do not, in any way, afford opportunity to interpolate. Cursory
inspection will show that the grouping of data collected prohibits accurate
curve fitting. Consequently, the attendant statistics to the regressions were
neither obtained from the computer, nor included in the data presented.
-------
Scott Research Labs.» Inc.
Project #2608 4-10 March 6, 1970
4.1.1.1 Ambient Temperature
Refueling operations were conducted under four ambient temperatures:
o 45°F
o 60°F
o 7 5°F
o 90°F
Vapor space and residual gasoline temperatures in the tank were equilibrated
to the ambient temperature before starting the fill.
Displaced losses computed from measurements taken after dispensing
10 gallons of gasoline, by means of an automatic nozzle at these four ambient
temperatures, are plotted in Figure 4.1. Losses are shown for 9.8 RVP gasoline
dispensed into tank configuration $2 and for 9.2 RVP gasoline dispensed into
tank #1. The same fill rate of about 12.6 GPM was held in each test shown.
The displaced loss weight did not increase with succeedingly warmer ambient
tests.
Failure of the displaced loss weight to increase with ambient
temperature rise indicates that ambient temperature alone does not affect
o o
the magnitude of displaced losses between 45 F and 75 F.
4.1.1.2 Fill Method
Gasoline was introduced to each fuel tank by two methods;
o Nozzle fill
o Bottom fill
The first method simulates a normal service station operation in which
gasoline dispensed through a nozzle splashes upon the liquid surface inside
-------
Scott Research Labs., Inc.
Project #2608 4-11
March 6, 1970
60-
50-
4 fl-
ew
w
(=»
Fill Method = Nozzle
Fill Rate - 12.6 GPM
Dispensed Temp. = 60°F
Tank Type #2
RV? = 9JL
Tank Type #1
RV? - 9.2
10-
I"" i 1 1 1 11
40 50 60 70 80 90
AMBIENT AND INITIAL TANK TEMPERATURE, (°F)
Figure 4.1 Effect of Ambient Temperature on
Displaced Hydrocarbon Loss From
10 Gallon Refueling Operations
-------
Scott Research Labs., Inc.
Project #2608 4-12 March 6, 1970
4.1.1.2 Fill Method - continued
the tank. The second method represents a proposed alternative of dispensing
gasoline through a fitting in the bottom of the tank with minimum disturbance
to the liquid surface.
Displaced losses computed from bottom and nozzle fill measure-
ments at the same set of test conditions are compared in Figure 4.2. Also
shown, is the calculated weight of a 10 gallon equivalent, volume of vapor
in equilibrium with each ambient temperature. The formula used to calculate
this weight of equilibrated vapor is given in Appendix 6.4, Bottom fill
losses rise with warmer ambients and closely agree with losses calculated
for vapor equilibrated to that ambient temperature. It can be noted that
nozzle fill losses are nearly independent of ambient temperature. This
characteristic was generally found at other test conditions.
Bottom and nozzle fill displaced losses are nearly equal at 60°F
ambient only. These two curves converge at 60°F .because that is the only
ambient where the respective temperatures of vapor displaced from both
methods are equal. The influence of 60°F dispensed gasoline on the tempera-
ture of displaced vapor, and consequently on the magnitude of the displaced
loss, is discussed in Section 4.1.1.6.
Displaced losses computed from bottom fill measurements at all
conditions tested are presented in Figures 4.3,1 and 4.3.2. The calculated
loss for vapor temperature equilibrated to the ambient is shown for each
RVP. The agreement between measured and calculated losses indicates that
the temperature of vapor displaced during a bottom fill is nearly equal to
ambient.
-------
Scott Research Labs., Inc.
Project #2608
4-13
March 6, 1970
60'
50-
40-
30"
Tank Type
RVF
Fill Rate
Dispensed Temp.
#1
9,2 psig
12.6
60°F
Calculated*
10-
0"
ottom Fill
Nozzle Fill
20- * Assumptions for Calculation (See Appendix 6.7)
Volume of Displaced Vapor
Temp, of Displaced Vapor =
Mol. Weight = 68.5 GMS
TVP
= 10 Gallons in Ft"
Ambient
= 4.7 psia <3 60°F
= 6.2 psia @ 75°F
=8.2 psia 3 90°F
—1~
40
1 1 1
50 60 70
AMBIENT AND INITIAL TANK TEMPERATURE,
T"
90
80
(°F)
Figure 4.2 Effect of
Displaced
10 Gallon
Fill Method on
Hydrocarbon Loss From
Refueling Operations
-------
Scott Research Labs•»
Project #2608
Inc.
4-14
March 6, 197 0
60-
50-
40"
o
w
CO
o
(J
a 3&i
w
04
U1
M
a
Tank Type
Fill Method
Fill Rate
Dispensed Temp.
#1
Bottom
12.6 GPM
60 °F
/
Aw -
/ 9.2 psig
RVP -
9.8 pai
EVP =
7.2 psij
20-
10-
Measured
Calculated
(See Figure 4.2)
, 1 ,
40 50 60 70
AMBIENT AND INITIAL TANK TEMPERATURE,
80
(°F)
90
Figure 4.3,1 Bottom Fill Displaced Hydrocarbon Loss
From 10 Gallon Refueling Operations
-------
Scott Research Labs.
Project #2608
Inc.
4-15
March 6, 1970
60
50
40
o
xn
CO
o
o
Ol,
w
M
a
30
20
10
Tank Type
Fill Method
Fill Rate
Dispensed Temp
#2
Bottom
12.6 GPM
60°F
.VP =
9-2 psig
RVP
9.8
RVP »
7.2 psig
Measured
Calculated
(See Figure 4.2)
-1 1 1 1
40 50 60 70
AMBIENT AND INITIAL TANK TEMPERATURE,
—I—
80
(°F)
r~
90
Figure 4,3.2 Bottom Fill Displaced Hydrocarbon Loss
From 10 Gallon Refueling Operations
-------
Scott Research Labs,, Inc.
Project #2608 4-16 March 6, 1970
4.1.1.3 Tank Configuration
Gasoline was dispensed into two tanks with distinctly different
fill pipe designs and different enclosed liquid surfaces.
Configuration #1 exhibits a nearly horizontal fill pipe attitude
and broad surface while #2 exhibits a nearly vertical pipe and narrow
liquid surface. Displaced losses computed from type #1 and #2 measurements
are compared in Figures 4.4.1 and 4.4.2.
Inspection of these figures discloses that displaced vapor losses
from both tank configurations are nearly identical at the following
conditions:
o Same RVP fuel dispensed into both tanks
o Dispensed fuel temperature (60°F) equals ambient
and initial tank temperature
Individual tank losses diverge with ambient temperatures other than equal
to the dispensed temperature.
The narrow tank with the steep fill pipe (Tank #2) is seen to
generate a lesser displaced loss from 7.2 pslg and 9.2 psig RVP fuels at
higher ambient temperatures. Although losses from nozzle fills are not
directly affected by ambient temperature alone, the Tank #2 - 7.2 and 9.2 RVP
test series displayed vapor loss reduction with warmer ambients while the
Tank #1 - 7.2, 9.2, and 9.8 RVP series showed loss increase. This divergence
could be explained by different heat transfer characteristics between dis-
pensed gasoline and displaced vapor in the two filler pipes. This mechanism
will be discussed in Section 4.1,1.6.
-------
Scott Research
Project #2608
Labs., Inc.
4-17
March 6, 1970
60-
50-
Fill Method - Nozzle
Till Rate - 12.6 GPM
Q
Dispensed Temp. - 60 F
40-
CA
o
w
OT
O
hJ
« 30-
tJ
w
Tank Type
RVP
20-
9.8 psi
RVP = 7.2 psig
10-
o 1 j 1 1 -i r
40 50 60 70 80 90
AMBIENT AND INITIAL TANK TEMPERATURE, ( F)
Figure 4.4,1 Effect of Tank Configuration on
Displaced Hydrocarbon Loss From
10 Gallon Refueling Operations
-------
Scott Research Labs., Inc.
Project #2608
4-18
March 6, 197 0
60
50
40 -
Fill Method
Fill Rate
Dispensed Temp.
=» Nozzle
- 12.6 GPM
- 60°F
%n
en
O
~J
a
0 30
Pm
to
Tank Type
#1
EVP
20
10 -
0 -® 1 1 1 1 1 r
40 50 60 70 80 90
AMBIENT AND INITIAL TANK TEMPERATURE, (°F)
Figure 4,4.2 Effect of Tank Configuration on
Displaced Hydrocarbon Loss From
10 Gallon Refueling Operations
-------
Scott Research Labs., Inc.
Project #2608 4—19 March 6, 197 0
4.1.1.4 Reid Vapor Pressure
Three different Reid vapor pressure blends of gasoline were
dispensed into each tank.
The average RVP identified for each blend is listed below:
7.2 psig
9.2 psig
9.8 psig
Displaced losses computed from measurements taken while dispensing these
three RVP fuels are plotted in Figure 4.5, Results are shown for each
ambient temperature and for each tank tested.
Displaced vapor losses in Figure 4.5 are observed to rise with
higher Reid vapor pressure gasolines. Increased losses are to be expected
because the true vapor pressure TVP rises with RVP increase at any of these
ambient temperatures. True vapor pressure, in pounds per square inch
absolute, was determined from the measured RVP and the ambient and initial
tank temperature established for each test. These data were applied to a
nomogram* from which the TVP was read-off.
Little spread is observed between the Tank it2 test results because
expected ambient temperature effects were suppressed by the same 60°F dis-
pensed gasoline temperature used throughout. Better control of fuel tem-
peratures and more precise RVP determination would have closed-up the spread
in Tank #1 test results.
* API Bulletin 2518, dated June, 1962. Evaporation Loss From Fixed-Roof
Tanks, Figure 3, page 10, "Vapor Pressures of Gasolines - 5 lb to 14 lb RVP."
-------
Scott Research Labs., Inc.
Project #2608
40
Fill Method
Fill Rate
4-20
March 6, 197 0
30
o
CO
CO
o
hJ
<
& 20
in
10
Nozzle
12.6 GPM
Dispensed Temp. ¦ 60 F
75°F
60°F
90"F Ambient &
Initial Tszik
Tank Type #2
T
50
EQ
40
o
a
CO
8
•J
o
0-
cn
30 -i
20
Initial Tank
60°F
Tank Type #1
I : r
8 9
RSID VAPOR PRESSURE, psig
10
Figure 4.5 Effect of Reid Vapor Pressure on
Displaced Hydrocarbon Loss From
10 Gallon Refueling Operations
-------
Scott Research Labs., Inc.
Project >7 2608 4-21 March 6, 1970
4.1.1.5 Fill Rate
Gasoline was dispensed into each tank with the nozzHatched
alternately in the first and then the third notch. In these positions the
approximate flow rates are respectively, 4.9 and 12.6 GPM.
Displaced losses are extrapolated between 4.9 and 12.6 GPM in
Figure 4.6. Results are shown for each tank configuration and for each
ambient and initial tank temperature tested.
The anticipated spread between losses due to different RVP fuels
is evident at each ambient temperature. Most significant in this figure
is the flat characteristic of all plots indicating that displaced loss
magnitude is not affected by fill rate.
4.1.1.6 Dispensed Fuel Temperature
Gasoline dispensed during the planned test program was condi-
tioned to 60°F throughout. After completion of the regular program,
additional tests were performed during which the dispensed gasoline was
conditioned to 75°F and 90°F.
Displaced losses computed from measurements taken while dis-
pensing 60°, 75°, and 90°F conditioned gasoline are compared in Figure 4.7.
While ambient and initial tank temperature has little apparent effect, it
can be seen that higher dispensed temperatures produced predictably higher
displaced losses. The displaced loss plots intersect with the calculated
plot for equilibrated displaced vapor and ambient temperatures only when
the dispensed temperatures equal ambient.
-------
36-
30-
Tank Type
Fill Method
Avg, of #1 & 4.
=• Nozzle
Dispensed Temp. ¦ 60°F
9.2 psig RVP
7.2 psig
Ambient & Initial Tank Temp.
24 ¦ "t i ¦ K-
90°F
42*
3 fr~
30-
u>
to
24-
Ambient & Initial Tank Temp. ¦> 73 F
9,8 psig
9.2 psig
'1.2 psig
o 36.
ft.
to
30-
24-
Ambient & Initial Tank Temp. * 60 F
9.8 psig
9.2 psig
7.2 psig
36-i
30-
24-
Ambient & Initial Tank Temp. » 45 f
T"
10
¦9.8 psig
—r
12
T
14
FILL RATE, GALLONS PER MINUTE
Figure 4.6 Effect of Fill late on Displaced
Hydrocarbon Losses From 10 Gallon
Refueling Operations
-------
Scott Research Labs., Inc.
Project #2608
60
4-23
March 6, 1970
50-
40-
O
w
w
o
J
w
-J
Pn
CO
H
«
30-
20-
10-
90 F
Tank Type
Dispensed Temp
G
/
/
Average Tank Types #1 & #2
60°F
Dispensed Temp.
Fill Method
Fill Rate
RVP
= Nozzle
= 12.9 GPM
¦ 9.2 psig
Measured
Calculated
(See Figure 4.2)
—¦T—
40
—T"
50
—J—
60
~r~
70
AMBIENT AND INITIAL TANK TEMPERATURE,
80
(°F)
90
Figure 4.7 Displaced Hydrocarbon Loss From
10 Gallon Refueling Operations at
Different Dispensed Gasoline Temperatures
-------
Scott Research Labs., Inc.
Project #2608 4-24 March 6, 1970
4.1.1.6 Dispensed Fuel Temperature - continued
Two conclusions about displaced vapor loss weight can be drawn
from Figure 4.7;
o Loss' is relatively independent of ambient temperature
o Loss increases predictably with dispensed temperature
The mechanism by which dispensed temperature determines dis-
placed losses was desired. Thermocouples were placed in the nozzle spout
and in the filler pipe mouth in order to measure the respective temperatures
of gasoline dispensed and vapor displaced. Refueling operations were per-
formed and the observed responses of both vapor weight and vapor temperature
to dispensed temperature are shown in Figure 4.8.
The lower curve follows the increase of displaced vapor tempera-
ture with dispensed temperature rise.
The upper plot represents the increase in displaced loss magni-
tude with dispensed temperature rise from Figure 4.7.
It can now be observed from these data that both the temperature
and mass of a 10 gallon volume of displaced vapor respond in a like manner
to dispensed gasoline temperature rise. Vapor loss magnitude increase
should be expected because of the greater true vapor pressure'of a given RVP
gasoline at higher temperatures. Figure 4,8 shows that the dispensed gaso-
line temperature governs displaced vapor temperature and consequently the
magnitude of displaced losses.
-------
Scott Research Labs.,
Project #2608
60
Inc.
50
ft
w
o
56
¦J '
pm 40
W u
30
Fill Method
Fill Rate
Fill Quantity
Tank Type
RVF
Ambient &
Initial Tank
Temp.
4-25
March 6, 1970
Nozzle
12.6 GFM
10 Gallons
#2
9.2 psig
60° F
#20 J-5
Test #20 0
90 -1
mxest
#20 J-5
> 70 -
xn
Test #20 ©
60 -
-r
50
T r
60 70
DISPENSED GASOLINE TEMPERATURE, (°F)
~r
80
40
Figure 4.8 Effect of Dispensed Gasoline Temperature on
Temperature and Mass of Vapor Displaced
-------
Scott Research Labs.
Project #2608
) Inc .
4-26
March 6,, 1970
4,1.1.6 Dispensed Fuel Temperature - continued
The following two conclusions can be made:
o Dispensed gasoline temperature is the major
factor determining the temperature of vapor
displaced.
o Consequently, displaced vapor is a function
of dispensed gasoline temperature.
4.1.2 Ent r ained Droplet Loss
The weight of entrained droplets was to have been determined from
the difference between nozzle and bottom fill losses. Introducing fuel
from below the tank liquid surface with minimum disturbance should have
produced no entrainnent and consequently lesser bottom fill losses than
nozzle fill losses. This differential could not be obtained in the planned
program because bottom losses, as previously discussed in Section 4.1.1.2,
were always greater at ambient temperatures higher than dispensed gasoline
temperatures.
Additional refueling operations were performed during which
the dispensed gasoline was conditioned to equal the 60°, 75°, and 90°F
ambient temperatures.
Total displaced losses computed from nozzle and bottom fill
measurements at equal dispensed and ambient temperatures are compared
in Figure 4.9 for tank type §2. The difference between these losses is
plotted at the bottom of the figure. This difference may be the entrained
droplets associated with the turbulent dispensed liquid from a refueling
operation using the nozzle. However, these data are supported by only one
test at each set of conditions. Therefore, no firm conclusions can be
drawn until replicate tests are run.
-------
Scott Research Labs., Itic
Project #2608
March 6, 1970
4-27
60 -
- 12.9 GFM
= 9.2 psig
Fill Rate
R¥P
Tank Type
Dispensed Temp. ¦ Ambient
50 -
Nozzle
Fill/
Bottom
Fill
40 -
o
*
10 _
Difference Attributed
to Entrained Droplet^.
80
50 60
AMBIENT AND INITIAL TANK TEMPERA1USE, (°F)
40
Figure 4.9 Loss Attributed to Entrained
Hydrocarbon Droplets Displaced by
10 Gallon He fueling Operations
-------
Scott Research Labs., Inc.
Project #2608
4-28
March 6, 1970
4.1,3 Spilled Liquid Loss
In these experiments complete passenger cars were placed inside
the SHED and "average technique"* refueling operations were performed.
The experimental strategy was to collect data from tests in which
both displaced losses and spillage occurred and, under the same conditions,
to collect data on tests in which only displaced losses occurred; subse-
quently to eliminate displaced losses from those tests where both losses
occurred, to obtain spillage alone. This strategy enables one to account
for the hydrocarbon contribution of spilled gasoline. i
The weight, in grams, of composite losses was computed from the
formula given in Appendix 6.3.
Seventy tests were run in an attempt to gather data on the two
eventualities (spill and vapor, vapor only) for eight vehicles. In each
test the galienage of fuel dispensed was recorded. From this data a factor
was developed for the vapor weight displaced per gallon dispensed. This
factor was computed as an average over all tests without spillage for a
given tank type;
and, F = Average displaced vapor weight per gallon dispensed.
This factor was then applied to the gallonage dispensed in each
test in which spillage occurred, and subtracted from the composite loss to
give spillage alone.
Where, n = Number of tests without spillage, in the tank type
* Described in Section 2.2.2.
-------
Scott Research Labs., Inc.
Project #2608 4-29 March 6, 1970
4.1.3 Spilled Liquid Loss - continued
Spill test results are presented in Table 4.2, for each of eight
vehicles. It will be noted that spillage was not achieved for some types.
The averaged spillage is shown as the last entry in each type. Table headings
are as follows:
TANK = Coded Tank Type
SPILL = Spillage in grans (computed: Total-Displ)
DISPL = Displaced vapor loss in grams (computed:
Factor "F" x gal.)
TOTAL •= Composite spill and displaced vapor loss
in grams (computed)
TSHED = Ambient temperature inside SHED in °R
COL // = Column (test) number
These test results are summarized in Table 4.3. The minimum',
maximum, and average spill weight observed for each tank tested are tabulated
against the individual filler pipe designs.
Liquid spill test results presented in Table 4.2 and summarized
in Table 4.3 reflect FID measurements taken in the SHED 30 minutes after
the spill had been precipitated. As mentioned in the note in Section 2.2.2,
page 2-6 of this report, only about 75% of a spilled liquid volume actually
evaporated into the SHED and was accounted for on the FID. These test
results should be treated accordingly.
-------
Scott Research Labs., Inc.
Project #2608 4-30
March 6, 1970
Table 4.2 Spilled Gasoline Test Results
from "Average Technique"
Refueling Operations
Measured (Grams)*
TSHED
TANK
SPILL
DISPL
TOTAL
°R
COL
1
62.8
15.6
78.4
518
37-01
41.0
17.9
58.9
528
37-02
27.7
15.3
43.1
529
37-03
' 72.9
15,9
88.9
537
37-04
96.9
17.2
114.2
538
37-05
29.3
15.9
45.3
540
37-06
32.7
15.6
48.3
537
37-07
1
'51.9 AVERAGE
1A
No Spills
Observed
2
14.6
20.8
35.5 •
519
38-01
11.0
21.6
32.6
520
38-02
28.8
20'. 0
48.9
523
38-03
19.6
20.4
40.0
532
38-04
14.3
22.4
36.8
537
38-05
7.6
20.8
28.4
538
38-06
14.3
20.4
34.7
538
38-07
2
15.7 AVERAGE
2P
8.0
18.3
26.4
549
40-01
11.2
17.9
29.1
550
40-02
4.3
20.0
24.3
550
40-03
7.1
22.0
29.1
550
40-04
30.7
20.0
50.7
546
40-05
1.2
20.0
21.2
534
40-07
1.1
20.8
22.0
540
40-08
2P
9.1 AVERAGE
3
No Spills Observed
4
4.2
15.6
19.9
537
42-08
15.7
17.3
33.0
537
42-09
3.5
17.6
21.1
535
42-10
4
7.8 AVERAGE
5
No Spills
Observed
6
No Spills
Observed
* FID measurement of the evaporated fraction of a spilled liquid
volume; estimated to be 75% in 30 minutes. See Section 2.2,2
-------
Scott Research Labs., Inc.
Project #2608 4-31 March 6, 1970
Table 4.3 Spilled Liquid Losses for
"Average Operator Technique"*
Fill Pipe
Tank Measured (Grams)** Angle From Length,
'vpe
Min.
Max.
Avj^
Device
horizontal
inches
Dia.
1
27.7
96.9
51.9
No
20°
13
2 1/8
1A
0
0
0
Yes
15°
8
2 1/8
t
2
7.6
28.8
15,7
No
/ r O
45
16
2 1/4
2P
1.1
30.7
9.1
No
o
40
10
2 1/4
3
0
0
0
No
85°
1
2 1/4
4
3.5
15.7
7.8
No
45°
8
2 1/4
5
0
0
0
No
90q
4
2 1/2
6
0
0
0
Yes
60°
30
2
* "Average Operator Technique" defined as:
Maximum nozzle insertion at a convenient attitude in the fill
pipe with the trigger latched in the second tooth; dispense
until three automatic cut-offs.
** FID measurement of the evaporated fraction of a spilled liquid volume;
estimated to be 75% in 30 minutes. See Section 2.2.2.
-------
Scott Research Labs•, Inc.
Project #2608 4-32 March 6, 1970
4.1.3.1 Fill Pipe Configuration
Comparison of spill losses from fuel tanks with no anti-spill
devices finds measurable effects related to:
o Fill pipe angle with horizontal
o Fill pipe length
o Fill pipe diameter
i
Measured spill losses increase as the fill pipe angle approaches
the horizontal. Greatest losses among three tanks with fill pipes at
about 45 were measured from the longest pipe; the least losses were
observed from the shortest. Largest spill losses were measured from the
smallest diameter fill pipe; the least losses were observed from the
largest diameter pipe. , •
Spill magnitude may be a function of fill rate into a given fuel
tank configuration. Fill rate, as it affects spillage, was excluded from
the scope of this investigation.
4.1.3.2 Anti-Spill Devices
Devices installed in the filler pipes of two vehicles tested were
effective in preventing spill losses during refueling operations. Automatic
cutoff of nozzle flow was obtained before any gasoline was lost in every
instance tested.
After observing no spills at about 9 GPM (second notch of nozzle
latch), the maximum flow rate of about 13 GPM was imposed on these devices.
Again no spills were observed. Relative effectiveness of different anti-spill
devices and vent tubes were not investigated in this program.
-------
Scott Research Labs., Inc.
Project "2608
4-33
March 6, 1970
4.1.4 Nozzle Drip Losses
In these experiments passenger cars were refueled and nozzle
drippings collected as the refueling nozzle was withdrawn from the filler
pipe (Post-Fill Nozz 1 e Drip Loss). After a short interval, residual fuel
in the nozzle was drained and collected (Residual Nozzle Contents). These
residual liquid contents are equal to the maximum potential pre-fill nozzle
drip loss.
The weights, in grams, of Residual and Post-Fill losses were com-
puted from the formula*
Wt. = Specific Gravity x Volume Observed
Wt. = 0.744 x cc
Average losses were computed bv assigning equal weight to the
maximum and minimum losses experienced in a group of tests.
Ninety-five tests were run, gathering data on eight vehicle types
and two nozzle withdrawal methods; normal and rotated. In the following
tables results are presented for each type and method. Each tank type is
presented on a separate page, and an average loss is computed for each
handling method within a tank type. Table headings are as follows;
TANK = Coded tank type
METHOD = Normal, Rotated
LOSSES = Collected lasses in grams (computed)
Post-Fill (Lost at Withdrawal)
Residual (Potential Pre-fill Nozzle Loss)
COL = Column (test) number
-------
Scott Research Labs., Inc.
Project #2608 4-34
March 6, 1970
Table 4.4 Nozzle'Drip Test Results
Measured (Grams)
Tank
Method
Normal
AVG.
Rotated
AVG.
Post-Fill
53.5
68.4
61.0
3.6
0.3
2.0
Residual
0.0
0.0
0.0
Col
43-0
43-0
44-0
44-0
1A
2P
Rotated
AVG,
Normal
AVG.
Rotated
AVG.
Rotated
AVG.
Normal
AVG.
Rotated
AVG.
Rotated
AVG.
Normal
AVG.
Rotated
AVG.
Rotated
AVG.
66.2
71.4
68.8
0.0
0.0
0.0
52.8
53.5
53.2
2.3
0.5
1.4
57.2
5'1.3
54.3
1.5
4.4
3.0
52.8
44.5
48.7
0.0
0.0
0.0
63.2
63.8
63.5
0.0
0.0
0.0
37.2
42 • 4
39.8
0.0
0.0
0,0
49.0
38.7
43.9
45-0
45-0
43-1
43-2
44-1
44-2
45-1
45-1
49-0
49-0
50-0
50-0
51-0
51-0
61-0
61-0
62-0
62-0
63-0
63-0
-------
Scott Research Labs,, Inc.
Project #2608 4-35 March 6, 1970
Table 4.4 Nozzle Drip Test Results - cont.
Measured (Grams)
Tank - Method Post-Fill Residual Col
3 Normal 26.0 0,0 56-0
32.7 0.0 56-0
AVG. 29.4 0.0
Rotated 2.1 - 57-0
3.6 - 57-0
AVG. -2.9
Rotated - 18.5 58-0
' - 37.9 58-0
AVG. 28.2
4 Normal 20.0 , 0.0 73-0
39.4 0.0 73-0
AVG. , 29.7 0.0
Rotated 0.7 - 74-0
1.7 - 74-0
AVG. 1.2
Rotated - 14.8 75-0
- , 8.8 75-0
AVG. 11.8
5 Normal 46,0 0.0 67-0
43.1 0.0 67-0
AVG. 44 . 6 0 .0
Rotated 2.1 - 68-0
0.3 - 68-0
AVG. 1.2
Rotated - 22.9 69-0
37.9 69-0
AVG. 30.4
6 Normal 40.1 0.0 67-1
43.1 0.0 67-2
AVG. 41.6 0.0
Rotated 0.1 - 68-1
0.1 - 68-2
AVG. 0.1
Rot ated
AVG.
20.8
23.7
22 .3
69-1
69-2
-------
c
Scott Research Labs., Inc.
Project #2608 4-36 March 6, 1970
4.1.4 Nozzle Drip Losses - continued
Equivalent weight in grams of the average Post-Fill Nozzle Drip
Loss for each technique and tank type is listed in Table 4.5. Pertinent
filler pipe design details are listed for each tank tested•
Equivalent weight in grams of the average residual contents for
each technique and tank type is listed in Table 4.6.
-------
Scott Research Labs.,
Project #2608
Inc.
4-37
March
6, lc
Table 4.5
Post-Fill Nozzle
Drip Losses
Tank
Average Grams
Fill
Pipe
Type
Normal
Rotated
Att i tude
Dia.
I
61,0
2.0
2 0°
2
1/8
Ik
68 • 8
0.0
15°
0
1/8
2
53.2
1.4
45°
2
1/4
2P
54.3
3.0
40°
2
1/4
3
29.4
2.9
85°
2
1/4
4
2 9.7
1.2
45°
2
1/4
5
44.6
1.2
90°
2
1/2
6
41.6
0.1
60°
2
-------
Scott Research Labs., Inc.
Project tV 260S 4-38
March 6, 1970
Table 4.6 Residual Nozzle Contents
Tank Average Grams
Type Normal Rotated
1 0.0 48.7
1A 0.0 63.5
2 0.0 39.8
2? 0.0 43.9
3 0.0 28.2
4 0.0 11.8
5 0.0 30.4
6 0.0 22.3
-------
Scott Research Labs., Inc.
Project #2608 4-39 March 6, 1970
4.1.4.1 Operator Technique
Intuitively, it was expected that careful operator technique
(rotated withdrawal) would significantly reduce or prevent nozzle drip
losses. This expectation was borne out in these test results.
These data also show that careful technique is most beneficial
in reduction of nozzle loss from fill pipes of large diameter and low
angle with the horizontal.
4.1.4.2 Fill Pipe Attitude
Comparison of post-fill nozzle drip losses finds measurable
effects related to the fill pipe angle with the horizontal.
The greatest post-fill nozzle losses after a careless (normal)
withdrawal were observed from the lowest angle fill pipes. Conversely,
lower post-fill losses could be derived and the greatest residual liquid
volume could be retained in the nozzle by careful (rotated) withdrawal
technique at these same low fill pipe angles.
4.1.4.3 Residual Nozzle Liquid
Residual gasoline contained by the nozzle after careless (normal)
and careful (rotated) withdrawals is listed in Table 4.6 by the tank type
from which it was obtained. Residual contents were nil after careless
withdrawals by definition. Contents after careful withdrawals varied with
fill pipe geometry. Between these two extremes lies the potential for
pre-fill nozzle drip losses during subsequent refueling operations.
-------
Scott Research Labs., Inc.
Project #2608
4—40
March 6, 1970
4.2 REFUELING LOSS FREQUENCY
4.2.1 Preliminary Results
The data obtained in the technician survey was evaluated to
determine the following:
o' The average amount of fuel acquired during a
refueling stop
o The percentage of vehicles that obtain a full
tank of fuel
o The effect of obtaining a full tank of fuel or
whether or not a post-fill nozzle loss occurs
o The expected distribution of fuel tank con-
figurations among the population
It is not purported that the technician survey is a true represen-
tative sample of refueling operations throughout the country. The results,
however, can be applied (with restrictions) for the purpose of illustrating
basic concepts and defining future survey techniques of greater sophistica-
tion.
The results of the technician survey (200 data points) indicate
the following;
o F, the frequency of obtaining a full tank at
freeway service stations is 74.3%.
o F, , the frequency of obtaining a full tank at
community service stations is 61.5%.
o F , the overall frequency of obtaining a full
tank is 68.51 (137 of 200 observations).
o G , the average number of gallons obtained at
freeway service stations is 11.4.
-------
Scott Research
Project if2608
Labs., Inc.
4-41
March 6, 1970
4.2.1 Preliminary Results - continued
o G , the average number of gallons obtained at
community service stations is 10.6.
o G, the overall average quantity of fuel obtained
is 11.0 gallons.
o The frequency of occurrence for post-fill nozzle
losses is independent of whether or not the vehicle's
tank is filled. This statement is based on calcu-
lated spill frequencies of 50.3% for tanks filled
and 49.2% for tanks not filled to capacity.
The forementioned values are applicable to the overall refueling
loss scheme. The integrity of these values is limited by the sample size,
geographical area surveyed and possibly climatic or weather influences.
4.2.2 Overall Survey Results
All refueling observations (where a full tank of fuel was obtained)
were grouped into two categories, freeway service station operations and
community service station operations. The frequency of the three types of
spill losses was determined for each of the six fuel tank configurations in
the two underlined groups.
The three types of spill losses are:
1. Spitback/overfill
2. Pre-fill nozzle loss
3. Post-fill nozzle loss
Results obtained by combining the observations from the employee
survey (554 data points) and the technician survey (136 "fill-ups" only)
yielded the following;
-------
Scott Research Labs,, Inc.
Project #2608 4-42 March 6, 1970
4.2,2 Overall Survey Results - continued
o Total observations - 690
o Overall frequency of overfill/spitbaek - 26.1%
o Overall frequency of pre-fill nozzle losses - 8.6%
o Overall frequency of post-fill nozzle losses - 34.2%
Table 4.7 shows how the forementioned values were broken down by
type of service station and fuel tank configuration. Table 4.8 presents
the results categorized by fuel tank configuration only.
From Table 4.7 it can be seen that there exists differences
between the frequencies of certain spill losses for freeway and community
service stations (of the same tank type). These differences show no
consistent trend which would indicate either type of station as being more
prone to refueling losses than the other. The total (average) frequencies
for the three types of losses agree quite well between the two types of
service stations possibly indicating that the sample size associated with
each sub-group was smaller than desired.
Table 4.8 shows the distribution of spill loss frequencies after
combining the results from both the freeway stations and the community
stations. This figure best illustrates the refueling loss picture. Therefore,
the ensuing discussion will be based on values taken from Table 4.8.
Spitback/Overfill: This phenomenon is believed to be a function
primarily of fuel system design and nozzle performance, operator technique
not being a significant factor. It is possible, however, that less than
optimum insertion of the nozzle into the filler tank could affect a spit-
back in an otherwise spill-free system,
-------
Scott Research Labs,, Inc.
Project 2608
4-43
.March 6, 1970
Table 4.7 Refueling Loss Frequencies
For All Observations I-.liere a
Full Tank of
Fuel Was Obtained
Tank
Nozzle
Losses
Samp
Type
Overfill/Spitback
Pre-Fill
Post-Fill
Siz
Community Service Stations
1
0.249
0.118
0.408
245
0
i-
0.269 -
0.07 5
0.373
67
3
0.250
0.105
0.276
76
4
0.265
0.000
0.306
49
5
0.161
0.000
0.194
31
6
0.316
0.105
0.368
19
TOTAL
0.250
0.090
0.357
487
Freeway Service Stations
1
0.296
0.065
0.259
108
1
0.345
0.06 9
0.483
29
3
0.321
0.071
0.393
28
4
0.294
0.000
0.235
17
5
0.000
0.181
0.273
11
6
0.200
0.200
0.200
10
TOTAL
0.261
0.074
0.305
203
-------
Scott Research Labs,, Inc.
Project #2608
4-44
March 6, 197 0
Table 4.8 Refueling Loss Frequencies
For All Observations Where a
Full Tank of Fuel Was Obtained
Tank Nozzle Losses
Type Overfill/Spitback Pre-Fill Post-Fill
All Observations
1
0.264
0.102
0.362
2
0.292
0.073
0.407
3
0.269
0.096
0.308
4
0.273
0.000
0.288
5
0.119
0.048
0.214
6
0.276
0.138
0.310
TOTAL
0.261
0.086
0.342
Sample
Size
353
96
104
66
42
29
690
-------
Scott Research Labs., Inc.
Project #2608 4-45 March 6, 1970
4.2.2 Overall Survey Results - continued
Referring to Table 4.8, the frequency of spitback/overfill
appears to be consistent among the different fuel tank categories with
the exception of configuration #5. This category is composed primarily
of certain foreign cars with the filler neck opening' located under the
front deck lid.
Although the frequencies of spitback/overfill all appear to be
in the .26 to .29 range (except configuration #5), laboratory tests indicate
that losses from configuration ft I vary considerably. This category includes
a large segment of the automotive population which has a generally flat
fuel tank, located beneath the trunk, and a low filler neck opening, usually
behind the license plate. One vehicle with this type fuel tank was observed
to spitback consistently during laboratory tests while another vehicle of
the same apparent configuration could not be made to spitback at all. It is
assumed that similar results (variations in performance) would have been
encountered with other fuel tank configurations if more vehicles could have
been evaluated.
The concept of classifying the vehicles by the external shape
of their fuel tank and filler neck will have to be investigated more
completely in the future. It is presumed that the existence of internal
preventative devices in the filler neck and other subtle design factors
in this area are the primary influence in preventing losses of this type,
rather than external shape.
-------
Scott Research Labs., Inc.
Project #2608 4-46 March 6, 1970
4.2.2 Overall Survey Results - continued
Pre-fill Nozzle Losses: Losses of this type were observed
infrequently as indicated by the average frequency of occurrence (.086).
Values for different fuel tank configurations ranged from .000 (type #4)
to .138 (type #6).
Pre-fill nozzle losses are thought to be primarily attributed
to operator technique. Vehicles with body panels adjacent to the filler
neck opening sometimes create accessibility problems which would result in
a higher probability of pre-fill nozzle losses occurring.
Post-Fill Nozzle Losses: Post-fill nozzle losses are related
to both operator technique and filler neck location. With the exception
of tank type number 5, the frequency of post-fill nozzle losses ranges
from .288 (type number 4) to .407 (type number 2). Type number 5 indicated
a loss frequency of only .214. Due to the underhood filler opening asso-
ciated with many vehicles of this type, it is presumed that station atten-
dants use more caution upon withdrawing the nozzle from these vehicles as
fuel would essentially be spilled inside the trunk area.
With most vehicles evaluated in the laboratory, withdrawal of the
nozzle from the filler pipe could be effected with no spillage if precautions
(rotating the nozzle) were taken. The magnitude of the values in this category
appears to be primarily related to less than optimum operator technique.
-------
Scott Research Labs,, Inc.
Project •'/2608 5-1 March 6, 1970
5. CONCLUSIONS
5.1 SIGNIFICANT FACTORS CONTRIBUTING TO REFUELING LOSS
This investigation determined that sortie of the parameters studied
significantly affect the magnitude and frequency of refueling losses while
others do not.
Those factors which were found to have no significant effect
by themselves are;
o Tank Shape (exclusive of fill pipe)
o Fill rate
o Ambient temperature
The following factors did affect refueling losses and their
relative effects are discussed in Section 4.
o Dispensed quantity
o Reid vapor pressure
o Dispensed gasoline temperature
o Operator technique
o Fill pipe design
-------
Scott• Research Labs., Inc.
Project #2608 6-1 , March 6,' 1970
6. APPENDIX
-------
Scott Research Labs. , Inc.
Project #2608
6-2
March 6, 1970
APPENDIX 6.1 TEST GASOLINE ANALYSIS
Throughout the test program, samples of all gasolines used were
taken for subsequent analysis. Ethyl Corporation performed the following
RVP, ASTM distillation, and hydrocarbon analyses, on liquid samples taken
from subject fuel tanks during the test procedure. Scott Research Labora-
tories performed the gas chromatographic analyses on vapor samples taken
from filler pipes while gasoline was being dispensed.
-------
APPENDIX 6.1 TEST GASOLINE ANALYSIS - continued
Spill &
Loss
Displaced
Nozzle Drip
Test No.
1-18
19-36 & 101-108
7 9-96
37-78
Sample
ABC
E
D
F G
H 1
I
J K
RVP
7.0 7.2 7.0
7.6
9.8
8.9 9.3
8.8
9.8
8.1 9.0
RVP Average
7.2
9.2
9.8
8.6
Hydrocarbon Analysis
% Aromatics
36 .3
31.9
32.4
28 . 0 28.5
% Olefins
11.0
7.6
8.4
5.4 5.5
% Saturates
52.7
60.5
59.2
66 .0 66 .0
ASTM Distillation
Initial Boiling
Point , °F
°F
° F
°F
101
90
91
95 92
5% Evaporated,
123
101
108
10% Evaporated,
134
114
127
136 133
15% Evaporated,
142
123
142
20% Evaporated,
°F
150
131
158
30% Evaporated,
°F
° F
°F
167
150
187
50% Evaporated,
214
201
234
221 22 3
90% Evaporated,
3 53
358
343
313 319
Final Boiling
Point, °F
424
423
41 1
398 399
Slope
1.9
2.2
3 .4
Recovery, %
98.0
98. 0
97 .5
00
O
00
o
Loss, %
0.9
0.9
1.4
1.0 1.0
GC Analysis
C
X
C4.46
C4.79
C 4.42
H
y
H10.7 1
H11.04
H10 .61
H/C Ratio
2.40
2.31
2.40
2.31
Mol. Wt.
64 .3
68.5
63.7
TJ cn
»-t n
o o
¦_i. rr
fD rt
n
rt pc
fD
^ en
r j ft)
o
o i
oo o
rr
r-
£1)
CT
cn
3
O
CTN
I
U>
K
Q>
f-t
O
Z3*
VD
O
-------
Scott Research Labs. , Inc.
Project #2608
6-4
March 6, 1970
APPENDIX 6.2 TEST PROCEDURES
REFUELING LOSSES
TEST PROCEDURE
Date: 7-10-69
Revision: 11-21-69
I. Vapor Loss Test Conditions 1-36 and 79-96
1.0 Set up test apparatus as shown in Figure 1.
1.1 Place Mini-SHED Assembly in controlled temperature
room.
1.2 Place one drum of the specified Reid Vapor Pressure
fuel and one clean empty drum in a position conducive
to rapid conditioning to the controlled ambient
temperature.
1.3 Position the Scott Model 403 Fuel Conditioning System
for convenience to the Mini-SHED and fuel drums.
Primary attention must be directed against recirculation
of explosive gases in the controlled temperature room
and unnecessary hydrocarbon additions to the SHED back-
ground. Secure static ground cable from the cart to any
tanks or drums with which it is exchanging gasoline.
Observe NO SMOKING rule.
1.4 Fill the conditioning cart from another drum of the
same Reid Vapor Pressure specified in 1.2 above.
1.5 Condition fuel in 1.4 above to 60°F. Record conditioned
fuel temperature.
2.0 Establish Conditions for Respective Test.
2.1 Raise Mini-SHED skirt. Control the ambient air and fuel
drums to the ambient temperature specified in the test matrix.
2.2 Transfer fuel from the full drum (in 1.2 above) into the
auto gas tank specified. Then return 11.0 gallons to
the same fuel drum. The fuel remaining in the auto tank
shall be called tare fuel.
-------
Scott Research Labs., Inc.
Project r/2608
6-5
March 6, 1970
APPENDIX 6.2 TEST PROCEDURES - continued
2.3 Obtain a tare fuel sample by the displaced water tecr.r.iqu
for subsequent exact RVP measurement. Take sample only
when so directed x n matrix.
2.4 Verify that the specified tare fuel temperature is
established.
3.0 Background Concentration
3.1 Plug auto tank vent, cap the fill neck and cork the
filler nozzle.
3.2 Operate the temperature recorder and flame ionization
detector recorder throughout procedures 3-3 to 4.5.
3.3 Calibrate FID.
3.4 Observe for a stable ambient hydrocarbon concentration
Record the anbient and tare fuel temperatures, and
ambient pressure.
3.5 Secure the Mini-SHED skirt in the water seal.
3.6 Observe for a stable background hydrocarbon concentration
and record on test sheet. Also record bag differential
pressure and Mini-SHED interior temperature.
4.0 Refueling Loss Concentration
4.1 Unplug tank vent, uncap fill neck and uncork filler nozzl
4.2 Observe for a stable open tank hydrocarbon concentration.
4.3 Dispense 10 gallons of 60 ? conditioned fuel through the
fill nozzle into the auto tank at the specified fill rate
The automatic nozzle passes about 6 G?M when latched in
the first tooth; 15 G?M in the third tooth. Observe for
no fuel "Spi t-back" out of fill neck.
4.4 Upon completion of the 10 gallon fill operation, perform
the following procedure without hesitation:
4.4.1 Remove the nozzle from the filler neck, taking
care to avoid any fuel spillage.
4.4.2 Plug auto tank vent, cap the fill neck and cork
the filler nozzle.
-------
Scott Research Labs., Inc.
Project #2608
6-6
March 6, 1970
APPENDIX 6.2 TEST PROCEDURES - continued
4.5 Observe for a stable refueling loss hydrocarbon concen-
tration and record on test sheet. Also record bag differ-
ential pressure and Mini-SHED' interior temperature.
5.0 Purge-
5.1 Raise the Mini-SHED skirt and permit the controlled ambient
air to purge the SHED interior of all displaced hydrocarbon
vapors.
5.2 Suitp pump all fuel from the auto gas tank into the empty
fuel drum (in 1.2 above) for temporary storage and recon-
ditioning to the controlled ambient temperature.
6.0 Second Fill Rate
6.1 Repeat procedures 2.0 through 5.2 above for the second
fill rate.
7.0 Subsurface Fill
7.1 Repeat procedures 2.0 through 5.2 with following exceptions:
7.1.1 In procedure 4.1, unplug the tank vent and uncap
fill neck ONLY. Do not uncork filler nozzle.
7.1.2 In 4.3, dispense 10 gallons of 60°F conditioned
fuel through the pipe fitting in the bottom of
the auto tank at the same fill rate as recorded
in 6.1 above.
7.1.3 Delete 4.4.1.
7.1.4 In 4.4.2, plug auto tank vent and cap the fill
neck ONLY. The filler nozzle was not uncorked
in 7.1.1 above.
8.0 Test Matrix
8.1 Adjust test conditions as described in the test matrix
for vapor losses and repeat the above procedures 2.0
through 7.1.4, where applicable, for the remaining 36
test conditions.
-------
Scott Research Labs,, Inc.
Project #2608
6-7
March 6, 1970
APPENDIX 6.2 TEST PROCEDURES - continued
8.2 In 5.2 above, continue to sump pump fuel into the
temporary storage drum until it is 'full. Then sump
pump into the other empty drum originally used to fill
the conditioning cart.
8.3 In 2.2 above, continue to transfer fuel from the originally
full drum into the auto gas tank until it is exhausted.
Then transfer from the, now full, temporary storage drum
used to recondition fuel to the controlled ambient tempera-
ture in 5.2 above.
8.4 When all the fuel has been dispensed from the conditioning
cart in 4.3 above, refill the cart from the "now full"
drum from which the cart was originally filled. Do net
dispense again until the fuel has been reconditioned to
60°F .
II. Spill Loss Test Conditions 37-42
9.0 Set up test apparatus as shown in Figure 2.
9.1 Assemble SHED in protected area capable of garaging
automob iles,
9.2 Fill the conditioning cart with fuel of the specified
Reid vapor pressure.
9.3 Condition fuel in 9.2 above to 60°F. Obtain one fuel
sample by the displaced water technique from each barrel
of 'gasoline used in the spill tests.
9.4 Identify and locate an automobile with the fuel tank
shape and location specified in the test matrix. Remove
air cleaner and seal, carburetor in plastic bag. Measure
and record length, width, and height of automobile.
9.5 Open SHED entrance and push automobile into SHED taking
care not to damage entrance zipper. Close windows, doors,
and trunk lid of automobile.
9.6
Calibrate FID before each test.
-------
Scott Research Labs., Inc.
Project #2608
6-8
March 6, 1970
APPENDIX 6.2 TEST PROCEDURES - continued
10.0 Perform Liquid Spill Test
10.1 Fill the automobile fuel tank with gasoline and then remove
five gallons. Replace the tank cap and wipe up any gasoline
spillage.
10.2 With the SHED entrance open, operate the purge fan until the
ambient hydrocarbon concentration is negligible.
10.3 Connect static cable from the nozzle to the automobile.
Observe the NO SMOKING RULE. Turn off purge fan and
secure the SHED entrance and foot door zippers. Record
barometric pressure.
10.4 Operate the FID recorder throughout procedures 10.5 to 10.9.
10.5 Observe for a stable background hydrocarbon concentration.
Record the background concentration, SHED temperature, and
bag differential pressure.
10.6 Uncap the fuel tank fill neck, uncork filler nozzle'and
insert nozzle in fill neck. Avoid pre-fill nozzle
spillage.
10.7 Squeeze trigger, latch in second tooth and dispense fuel
in a conventional manner until the automatic nozzle trips
off for the first time. Again, squeeze the trigger but
now without latching and continue to dispense for a total
of three automatic trip-offs. Make no attempt to prevent
a spill during the above procedure.
10.8 Without hesitation, remove nozzle while avoiding post-fill
nozzle drip. Cork the nozzle and cap the fill neck*
10.9 Observe for a stable hydrocarbon concentration on the FID
recorder. Record hydrocarbon concentration resulting from
the total spill, bag differential pressure, and SHED interior
temperature on the test sheet. Turn off sample pump until
a stable reading can be obtained. Complete evaporation of
the total spill must be observed.
10.10
Record total gallons of gasoline dispensed in 10.7 above.
-------
Scott Research Labs., Inc.
Project 2 6 OR
6-9
March 6, 1970
APPENDIX 6.2 TEST PROCEDURES - continued
11.0 Measure Vapor Displaced During Spill Test
11.1 Fill the automobile fuel tank with gasoline and then
remove six gallons. Replace the tank cap and wipe up
any gasoline spillage. .
11.2 Repeat procedures 10.2,through 10.10 except that in 10,7,
dispense only as many gallons of gasoline as recorded in
10.10 above and make every attempt to prevent a spill during
this fill procedure. Void and repeat this measurement if
a spill does occur.
11.3 Open SHED entrance and foot door, and operate purge fan.
Pull automobile out taking care not to damage zipper,
gasoline hoses or thermocouple leads.
11.4 Clean up any oil or fuel spillage. Inspect for and repair
any damage to SHED enclosure.
11.5 With the SHED entrance and foot door open, operate the
purge fan until the background hydrocarbon concentration
is negligible.
12.0 Repeat Spill Tests on Other Automobiles
12.1 Identify and locate additional automobiles with the other
fuel tank shapes and locations specified in the test matrix.
12.2 Repeat procedures 9,4 to 11.5 above for each automobile.
Ill. Post Nozzle Loss Test Conditions 43-78
13.0 Set up test apparatus as shown in Figure 3.
13.1 Identify and locate an automobile with the fuel tank
shape and location specified in the test matrix.
13.2 Fill the conditioning cart and condition the fuel to
60 F. Obtain specific gravity of each barrel of fuel
used.
14.0 Normal Attitude Losses ¦
14.1 Remove two gallons of 'gasoline from the previously full
automobile tank.
-------
Scott Research Labs., Inc.
Project #2608
6-10
March 6, 1970
APPENDIX 6,2 TEST PROCEDURES - continued ¦
14.2 With the automatic nozzle trigger latched in the first
tooth, dispense fuel back into the tank until three
automatic trip-offs have been performed.
14.3 Immediately after the third trip-off, place the funnel
and graduate under the tank fill neck and withdraw the
nozzle. Do NOT rotate nozzle from normal attitude during
withdrawalt but 'capture all the liquid which escapes from
the nozzle.
14.4 Record collected liquid volume on test sheet.
15.0 Rotated Nozzle Losses
15.1 Repeat procedures 14.1 through 14.4 with the following
exception:
15.1.1 During withdrawal in 14.3 above, ROTATE the
nozzle making every attempt NOT to lose liquid
from the spout. Capture any liquid that does
manage to escape, and record volume on test
sheet.
15.2 Immediately proceed to 16.0 below.
16.0 Residual Nozzle Losses
16.1 Immediately after collecting inverted losses in 15.1.1
above, point nozzle down into another graduate and collect
the liquid remaining in the spout.
16.1.1 Record residual liquid volume on test sheet.
16.2 Remove one gallon of fuel from automobile tank, dispense
fuel back into tank until third trip-off and withdraw
while rotating nozzle in an attempt NOT to lose liquid.
Hang nozzle in vertical position.
16.2.1 After five minutes in the vertical position,
point nozzle down into a graduate, and collect
and record the liquid remaining in the spout.
16.3 Repeat 16.2 above.
16.3,1 Repeat 16.2.1 above, after ten minutes instead
of five.
-------
Scott Research Labs., Inc.
Project f',' 2608
6-11
March 6, 1970
APPENDIX 6.2 TEST PROCEDURES - continued
16.4 Repeat 16.2 above.
16.4.1 Repeat 16.2.1 above, after fifteen minutes instead
of five.
17.0 Identify and locate antortioblies with the remaining fuel tank
shape and locations specified in the test matrix.
17.1 Repeat procedures 14.0 through 16.4,1 for each of the
additional auto types.
-------
-~'v>
REFUELINR LOSSES TEST SHEET HO, TECHNICIAN DATE,
I. VAPOR LOSSES ' , '
TEST
CONDITIONS 1
-35
PARA.
OBSERVATIONS
'' UNITS
1 .2
Ri'P, Nominal
pslg
1.3
Secure Stat 1c Cable
Yes/No
1.3
Observe - NO STOKING - Rule
Yes/No
1.5 ,
TEMP,, Conditioned Fuel
°F
2.2
SHAPE, Auto Tank
FtAT/RECT.
2,3
qty. , Tare Fuel Sample
qt.
3.3
Calibrate FID
Yes/No
3.4 '
CONC., BC Ambient
Deflect
Scale
3.4
TEMP. , Ambient
°F
3.4
TEMP., Tare Fuel
. °F
'
3.4
PRESS., Ambient
In. Hg
3.6
CDNC,, HC Background
Deflect
Scale ' '
3.6
PRESS., Bag Diff.
in. H.,0
3-6 '
TEMP., SHED Background
°F
4.2
CONC,, HC Open Tan':
Deflect
' Scale
4.3
METHOD, Fill
NOZZLE/BOTTOM
4.3
RATE .Fill
mm
1
4.3
Qty.. Fill
gal.
4.3
OBSERVE, "Spit-Back"
Yes/Hone
4.5
CONC,, HC Refueling
Deflect
Scale-
4,5
PRESS,, Bag 01ff.
in. H.,0-
4.5
TEMP., SHED
°F
CALCULATIONS
RVP, Analysis
psig
TVP
psia
TEMP. . SHED
°R
•
' PRESS., SHED
in. Hg
VOLUHE, SHEO Net
Ft3
RATE, Fill
¦ gpin
Cone., Refueling
Cone., Background
pSfTc
_
—
— .
—
—
Cone., Loss
pp
-------
REFUELING LOSSES TEST SHEET NO. TECHNICIAN DATE.
II.. SPILL LOSSES ,
TEST CONDITIONS
37-42
PARA.
OBSERVATIONS
'UNITS
9.2
RVP, Nominal
¦ ps 1 g
9.3
TEMP., Cond. Fuel
t} p
•9.3,
QTY,, Fuel Sample
¦' qt.
9.4 ^
Shape/Location Tank
" 9.6
Ca1
-------
REFUELING LOSSES TEST SHEET NO. TECHNICIAN DATE ^ »
III. POST NOZZLE LOSSES 8 g.
TEST CONDITIONS 43-78
PARA.
OBSERVATIONS UNITS
13.1
Shape/Location Tank
1
13.2
Temp., Cond. Fuel °F
13.2
Specific Gravity Fuel
14.3
Method of Withdrawal J; NR°™Jd
14.4
Volume, Withdrawal Loss cc
16.1.1
Volume, Residual Loss cc
16.2.1
Delay Before Collection Minutes.
CALCULATIONS
Wt., Withdrawal Loss gms
Wt. , Residual gms
-------
Scott Research Labs., Inc.
Project #2608 6-15 • March 6, 1970
APPENDIX 6.3 VAPOR WEIGHT COMPUTATION
Concentration measurements of displaced gasoline vapor arid
evaporated spills were converted to grams with the following formula:
nTrnT 20.8(12 + H/C) C V P 10~6
DISPL = ,
Where: DISPL = hydrocarbon weight, grams
20.8 x 10 ^ = Units correction factor
H/C = Hydrogen carbon ratio
C = Net concentration as carbon, ppm
3
V = Net enclosure volume, ft
P = Enclosure pressure, in. Hg,
T = Enclosure temperature, °R
C = Net concentration as carbon, ppm
J - B
C = x Q
Where: J = FID, Refueling loss
B = FID, Background
W = FID, Calibration
Q = ppm, Calibration Gas
-------
Scott Research Labs., Inc.
Project #2608 6-16 . March 6, 1970
APPENDIX 6.3 VAPOR WEIGHT COMPUTATION - continued
Example - Test No; 23-2:
Q = 29100 ppm
W = 95.9 deflections
J — 43.7 deflections
30
B = 29,5 x = .295 deflections.
41 7 - .295
C = 95 9 " 29100 = 13160 ppm
R/e = 2.31
V = 150.5 ft3
P = 28.56 - 0 = 28.56 in. Hg.
T = 60°F = 520°!
20.8 (12 + 2.31) 13160 x 150.5 x 28.56 x 10
55 1 _ — - ' • ' ' " "¦ * "
520
DISPL = 32.4 grams
*
-------
Scott Research Labs., Inc.
Project #2608 ¦ 6-17 March 6, 1970
APPENDIX 6,4 EQUILIBRATED VAPOR CALCULATION
Gram weight calculations for hydrocarbon vapor displaced at the
same temperature as ambient were made with the following formula:
= 46,2 x MW x MF x V
T
Where: MW = Mole Weight of hydrocarbon
46.2 = Units correction factor
V = Volume of vapor, gallons
T = Temperature of ambient and initial tank, lK
MF = Mole fraction of vapor
T\'P
= tl—
BAR
Where: TVP = True vapor pressure at T °F, psia
BAR = Barometric pressure, psia
Example - Test No. 23
MW = 68.5 gin/mole
V = 10 gallons
T = 60°F = 288.5°K
TVP = 4.7 psia
BAR = 14.1 ps ia
4 ,7
MF = —— = . 334
14,1
46,2 x 68,5 x .334 x 10
288.5
36.6 grams
-------