Impact of Vapor Pilot Lights on Hot Air Balloon Operations
Frank M. Bacon, PhD[1]
Hot air balloons typically have two types of pilot lights for the propane burners. One type uses vapor withdrawn from the top of the propane fuel tank; this configuration requires two hoses from the tank, one for the vapor to the pilot light and one carrying liquid propane from the bottom of the fuel tank to the burners. The other type of pilot uses liquid directly from the liquid fuel line at the burner where the liquid is vaporized at the pilot light; this configuration has only one fuel line carrying liquid propane to the burner and no separate line for vapor. This paper describes operation of a balloon with a vapor pilot light. In an earlier paper I described balloon operations for a balloon with a pilot light run off of liquid propane.[2] In the original paper, I presented the results of calculations and measurements on the impact of evaporation of the liquid propane in the fuel tank necessary to replace the volume of liquid consumed during balloon flight and how the evaporation reduced the temperature and pressure of the fuel and the flow rate of the fuel to the burners. This paper describes how the removal of the vapor for the pilot light contributes to an even greater reduction of the propane temperature, pressure and flow rate.
In the original paper on Propane Characteristics and Performance in Hot Air Balloons containing the detailed calculations dated May 21, 2014, I hypothesized that a balloon operating with a vapor pilot light might experience deterioration in balloon performance:
Vapor pilot lights will contribute to evaporation of propane in the tank and probably impact the operating conditions of the balloon in a more dramatic way at low propane levels than for balloons equipped with pilot lights operating with liquid propane. A calculation of the impact of vapor pilot lights could be made if propane flow rates were made available.
Having been unable to obtain any data on propane flow rates for vapor pilot lights from balloon manufactures, I decided that I could measure the flow rate on the pilot lights on my balloon even though the pilots are pilot converters run off of liquid propane, not vapor pilot lights. The flow rates so obtained would be a reasonable representation of flow rates on other systems.
Flow rate measurements were made by enclosing the pilot light with a deflated one-gallon plastic bag obtained from the produce rack at the local grocery store, sealing the bag around the base of the pilot, and measuring the time required to fully inflate the bag with propane from the pilot light. Measurements were made on both pilot lights, and with 90 psig[3] pressure in the propane tanks, the time required to fill the bag was one minute on each of the pilots. So in a 100-minute flight, the total vapor flow through each of the pilots would be about 100 gallons at atmospheric pressure of 12.5 psia. Translating this volume back to a pressurized tank at 100 psig would correspond to 12 gallons of vapor at 112.5 psia. These 12 gallons would have been provided by evaporation of the propane in the tank, and would be comparable to and in addition to the 14 gallons evaporated from the liquid propane to replace the 14 gallons of liquid that were used by the burner as described in the earlier paper, and the volume of liquid propane that would typically be used out of a 15 gallon tank during a 100 min. flight.
Assuming a 100-minute flight consuming 14 gal of propane with the burner, the amount of propane used by the pilot per amount used by the burner would be 0.86 gal. of vapor per gallon of liquid propane burned. So I modified the calculations presented in the earlier paper to include the additional evaporation inside the tank due to the vapor pilot and determined the impact on balloon performance. In the following graphs I present the results obtained and compare them with the earlier results for a balloon with a pilot converter, where the effect of heat transfer from the metal tank to the liquid propane was included in both calculations.
_files/image001.jpg)
Figure 1. Propane Pressure vs. % Propane Remaining for two operating conditions, one with a pilot using liquid, referred to as a pilot converter, and one with a vapor pilot light; initial pressure of 100 psig. The calculations included the effects of heat from the metal propane tank on propane pressure. Vapor flow for the vapor pilot light was assumed to be 1.0 gal/min.
_files/image002.gif)
Figure 2. Calculated Propane Equilibrium Temperature vs. % Propane Remaining for two conditions, one with a pilot converter and one with a vapor pilot light; initial pressure of 100 psig. The calculations included the effects of heat from the metal propane tank on propane pressure. (See Figs. 1 and 2 in the original paper for the pilot converter calculations.2)
In addition to these graphical results, the flow rate for the vapor pilot decreased by about30 per cent due to the reduced pressure. In burning 14 gal. of propane through the burner, 0.7 gal. of liquid propane would have been consumed due to evaporation inside the tank.
Experimental Results
Using a borrowed[4] Cameron balloon equipped with a MK IV burner with vapor pilots and two 15-gal, 42-pound stainless steel tanks, I performed a test where the tank pressure was measured during a 100 min. period simulating a 100 min. flight where the propane level was reduced to about ten percent. The pressure at the tanks was measured with a pressure gauge connected in series between the propane tank and the liquid fuel hose going to the burner; this arrangement was required because the pressure gauges on the burner were on the downstream side of the blast valves and did not indicate the static pressure in the tanks. I was unable to measure the flow rate on the pilot lights due to their mechanical configuration. Pressure on two tanks was measured, the tank connected to the burner under test and to the tank that was used as a control to monitor ambient effects on pressure during the test. At the beginning of the test on December 14, 2014, the ambient temperature was 47◦ F at 3:15 PM, and the static pressure was 101and 102 psig after having heated the tanks prior to the test. On the tank under test, the burner was operated periodically in short bursts so the propane would be used uniformly during a 100 min. test; measurements were made in five minute intervals. The level of propane could not be monitored throughout the test because the indicator for the liquid level in the tank could not register the propane level until the level was less than about 35%. Intermediate levels between 100% and 35% were estimated using interpolation; the results of the interpolation are shown in the following figure. (In the tests described in the original paper, the liquid level began registering on the tanks for that balloon system at the 65% level.)
_files/image003.jpg)
Figure 3. Percent propane vs. time. Values between 100% and 35% are interpolated based on the regression curve shown on the graph.
Experimental results are presented in the following table.
|
Time |
Elapsed Time (min) |
Measured Press Tank 2 (psig) |
Percent Propane Tank 2 |
Measured Press Tank 1 (psig) |
Percent Propane Tank 1 |
Calculated Propane Temp. ( ◦C) Tank 1 |
Calculated Propane Temp. ( ◦F) Tank 1 |
Ambient Temp. ( ◦F) |
|
1515 |
0 |
102 |
100 |
101 |
100% |
17.3 |
63.1 |
47 |
|
1516 |
1 |
96 |
99% |
15.6 |
60.0 |
|||
|
1520 |
5 |
92 |
94% |
14.2 |
57.5 |
|||
|
1525 |
10 |
89 |
88% |
13.1 |
55.6 |
|||
|
1530 |
15 |
87 |
82% |
12.4 |
54.3 |
|||
|
1535 |
20 |
86 |
76% |
12.0 |
53.6 |
|||
|
1540 |
25 |
84 |
70% |
11.3 |
52.3 |
|||
|
1545 |
30 |
83 |
65% |
10.9 |
51.7 |
|||
|
1550 |
35 |
81 |
60% |
10.2 |
50.3 |
|||
|
1555 |
40 |
80 |
55% |
9.8 |
49.7 |
|||
|
1600 |
45 |
79 |
50% |
9.4 |
49.0 |
|||
|
1605 |
50 |
78 |
46% |
9.1 |
48.3 |
|||
|
1610 |
55 |
77 |
42% |
8.7 |
47.6 |
|||
|
1615 |
60 |
76 |
38% |
8.3 |
46.9 |
|||
|
1620 |
65 |
75 |
34% |
7.9 |
46.2 |
|||
|
1625 |
70 |
73 |
27% |
7.1 |
44.8 |
|||
|
1630 |
75 |
72 |
25% |
6.7 |
44.1 |
|||
|
1635 |
80 |
71 |
22% |
6.4 |
43.4 |
|||
|
1640 |
85 |
69 |
19% |
5.6 |
42.0 |
|||
|
1645 |
90 |
68 |
16% |
5.2 |
41.3 |
|||
|
1650 |
95 |
66 |
13% |
4.4 |
39.9 |
|||
|
1655 |
100 |
96 |
100 |
64 |
9% |
3.6 |
38.4 |
45 |
Figure 4. Experimental results. The percent propane values in red font are based on interpolated results.
The pressure on the tank under test dropped five psig after the first burner operation. This sharp drop was not understood because of its inconsistency with the subsequent measurements; in a normal flight it would have been involved in the balloon inflation and would not have been a factor in the flight operations after inflation. In the following analysis, the test was assumed to start with a pressure of 96 psig. Results of the test and corresponding calculations are shown in the following figures. Two calculations were performed: one assuming a burner with a pilot converter and one with a vapor pilot light. The flow rate for the vapor pilot light was not know, so a flow rate of 1.2 gal per min was assumed for the calculations to match the experimental results. This value is not appreciably different from the value measured on a balloon with a pilot converter as discussed above. The vapor flow rate from the propane tank with the pilot converter is zero since the pilot light is fueled by liquid fuel.
_files/image004.jpg)
Figure 5. Measured pressure vs. remaining propane and calculated values for pilot converter and for vapor pilot light with 1.2 gal/min vapor flow rate. Values for percent propane between 100% and 35% are interpolated values.
_files/image005.jpg)
Figure 6. Calculated propane equilibrium temperature based on propane pressure. Pilot converter pilot with no vapor flow from tank, vapor pilot light with 1.2 gal/min vapor flow, and measurements. Ambient pressure was measured at the beginning and end of test sequence.
The ambient temperature during the test decreased from 47◦ to 45◦. The experimental data deviated from the adiabatic[5] calculations, apparently due to some thermal losses to the atmosphere when the propane temperature was greater than the ambient temperature. The pressure on the tank that was used as a control decreased about five psig, also indicating some thermal loss to the ambient atmosphere. These results are somewhat different from the earlier results2 on the Aerostar balloon with no thermal losses to the ambient, possibly due to different designs of the thermal jackets on the propane tanks.
Total propane consumed during the test was 12.0 gals. Of this value about 0.7 gal. of liquid propane was evaporated to replace the vapor through the pilot light and the vapor to replace the volume of liquid used by the burner.
Conclusions
It is clear from the above results, a vapor pilot light where vapor is removed from the propane tank to run the pilot lights will adversely impact balloon performance due to reduced propane pressure at the burner. The effects will be more pronounced than for a balloon equipped with a pilot converter where the pilot light vapor is produced at the burner without drawing vapor from inside the tank. Propane pressure will decrease about 30 psig for a vapor pilot when the propane is reduced from 100% to 10% near the end of a long flight, thus reducing the propane flow to the burner by about 30%. These conditions could be in the “marginally safe” operating range as defined by the balloon flight manual. These results are compared with a 20% reduction in pressure for a balloon equipped with a pilot converter.
Pilots of balloons with vapor pilot lights should consider operating with a pilot light as low as is practical for safe operation to minimize loss of pressure during flight.
As discussed in the earlier paper, thermal gradients in the liquid propane could contribute to sharply changing performance of the balloon near the end of a long flight which could be more pronounced with a vapor pilot light.
Given these results, a balloon equipped with a pilot converter would have an improved margin of safety over a balloon equipped with vapor pilot lights. The results presented here are consistent with my experiences over 40 years of flying hot air balloons with both types of pilot lights.
Propane tanks equipped with regulated nitrogen to maintain constant pressure would not experience pressure drops for the propane thus leading to improved safety. By design, these systems would not have vapor pilot lights.