Propane overfilling fires are very common. It is reported that at least several hundred occur each year in the United States alone. Overfilling fires are generally not understood by those who cause them or by those who witness them. Every hot air balloonist should understand how they happen, how they can be prevented, and other means available to provide adequate vapor pressure for flying safely.
The classic propane overfilling fire occurs in cold weather where a propane container is filled and then subjected to a warmer environment. To understand how overfilling fires occur, it is necessary to understand the physical characteristics of commercial liquefied petroleum gas more commonly referred to as propane.
Commercial propane is a mixture of propane with its near chemical relatives ethane and butane. Propane is normally odorized with small amounts of strong-smelling sulfur compounds (ethyl mercaptan) as an aid to leak detection. Balloonists should recognize that commercial propane has a range of physical properties, depending on its ethane and butane content.
The diagram in Figure 1 shows many things with regard to the physical properties of propane. First consider the curve marked "vapor pressure". As long as both liquid and vapor propane are present in the tank, the pressure in the tank will be given by this curve. The curve shows the vapor pressure of propane in PSIA. This normal boiling point of propane is -43.7 F (pressure = 14.7 PSIA or 0.0 PSI Gauge). This means that liquid propane can only exist in an open container or as a cloud of liquid droplets in air if its temperature is -43.7 F or lower. At room temperature (70 F) its vapor pressure is 375 PSI Gauge (the normal setting for the pressure relief valve) at a temperature of 162 F. Thus a properly working propane container, properly filled, can be heated to about 162 F before its pressure relief valve can be expected to open.
Figure 2 shows the fractional expansion curves for liquid propane, and liquid water (3,4), both under their own vapor pressure. One sees that liquid propane expands much more on heating than does liquid water. From 60 to 160 F liquid propane expands 25%, while liquid water expands only 2.2%. One may also compute that over this temperature range ordinary steels expand 0.19%, so liquid propane expands roughly ten times as much on heating as does water and roughly 100 times as much as does steel. This is the cause of the problem with overfilling propane containers.
From Figure 2 we can also see that if the tank is filled to 85% liquid with liquid propane at a temperature lower than 60 F then heating will fill the container full of liquid at a temperature lower than 144 F. If it were filled 85% full at 35 F then the total mass of liquid in the container is (1/0.965) = 1.036 times the mass of propane that would fill it 85% full of liquid at 60 F. (If you buy propane, or any other liquid by the gallon, you get more for your money in cold weather). This amount of liquid would expand to completely fill the container when the volume was (1.18/1.036) = 1.14 times to 60 F volume, which we see from Figure 2 corresponds to a temperature of about 138 F. This is less than the 144 F shown above, but still substantially higher than any temperature a propane tank is likely to encounter in normal usage. Unlike gasoline, which is normally stored in underground tanks, and dispensed at practically the same temperature all year, propane is normally stored in above-ground, uninsulated tanks. For that reason the temperature of the liquid propane placed in a customer's container is practically the same as the temperature of the liquid propane in the supply tank, which is generally about the average outdoor air temperature over the past few days and nights.
Overfilling
Now, what happens if, for example, the tank is filled completely (100%) with
liquid, instead of only being 85% filled with liquid? In this case the pressure-
temperature relation will not be the vapor pressure relation for gas and liquid, but the
pressure-temperature relation for liquid at a constant volume. At constant volume the
pressure of liquid propane rises (at 60 F) by about 45 PSI/ F. If we correct this value to
take into account the thermal expansion and elastic stretching of the tank, we find that
the best estimate of the rate or pressure rise for a typical container full of liquid
propane is about 37.6 PSI/ F. So if the tank were 100% full at 60 F, we would expect
that raising the temperature by about 8 F would cause the pressure to reach 375 PSI
Gauge and cause the pressure relief valve to open.
It is difficult to fill such a tank 100% full of liquid. But if we filled one 99% full of liquid at 60 F, we could then trace its behavior on Figure 1 as follows. As the temperature was raised, as long as there was gaseous propane present the pressure- temperature behavior would follow the vapor pressure curve. From Figure 2 we see that the liquid expansion at 60 F is roughly 0.14%/ F, so that it would take an increase of about (1%/0.14%0 = 7.2 F for the tank to become full of liquid. Further heating would then cause the liquid to increase in pressure at constant volume. By similar calculation one can see that if the tank is 95% full of liquid at 60 F it would take about (5%/0.14%) = 35 F temperature rise for the system to become liquid filled. This suggests that a 99% full tank is almost as dangerous as a 100% full tank, but a 95% full tank has considerable margin of safety for temperature increase before it becomes totally full of liquid.
Similar curves could be constructed for various other starting temperatures, and various degrees of liquid filling but they all illustrate the same point, which is that once the system becomes liquid-filled, increasing the temperature very little causes the pressure to rise very much. In contrast, if both liquid and vapor are present, the rate of pressure rise is small; at 60 F is 1.7 PSI/ F.
It is plausible to expect that a temperature rise of 1.5 or more degrees F can occur naturally. For any filling greater than about 99% that temperature rise will cause the pressure relief valve to open. Once the high pressure relief valve opens, there will be a considerable flow of liquid propane out through the open valve. One would hope that the valve would close and reseat, shutting off the flow when the pressure inside the tank had fallen back to the 375 PSIG setting of the valve, but most such valves do not reseat quickly under such circumstances, and as a result a significant burst of liquid propane is emitted. Once this vaporizes, spreads and finds an ignition source, a fire results which then generally further heats the container, so its temperature rises even more rapidly and it continues to vent liquid propane, sustaining the fire.
Relief Valves
One might ask why we have a pressure relief valve in this situation? The
pressure relief valve prevents a worse accident. The standard small DOT portable
propane container is designed for 240 PSIG (255 PSIA) service pressure. It is
hydrostatically tested to twice its service pressure (480 PSIG) and the pressure relief
valve is set at 75% of the test pressure which would be 360 PSIG, but is conventionally
set at 375 PSIG. These containers have a design safety factor of 4, indicating that they
would be expected to burst at 960 PSIG. That pressure would be expected to occur
about 22 degrees F above the temperature at which the tank became totally filled with
liquid. If the tank burst, then all its contents would be released at once. That would
most likely lead to a much larger and more destructive fire than the one caused by the
intermittent release of small slugs of liquid propane from the high pressure relief valve.
Many people have escaped from propane overfilling fire, because the accidents often
produce a warning in the form of a screeching or sputtering sound from the relief valve
and a visible white cloud before ignition, and because the fires start at a moderate rate,
increasing in size as the heat from the fire speeds the process of propane release. A
sudden rupture of an entire tank would give no such warning and would result in a very
large, quickly-spreading fireball, which no nearby person would survive.
In addition the pressure relief valve protects the lives of firefighters who are fighting fires. In a fire, if there is a properly filled propane tank involved, and the temperature of the tank and contents rises past 160 F the pressure relief valve will open, and allow gas to gradually escape. (If the tank is not vertical, then liquid may escape instead of gas). This gradual escape limits the internal pressure in the tank and thus prevents an explosion of the tank, which would release the entire contents at once, causing a fireball likely to kill any nearby firefighters. The protection provided by the safety relief valve is not total, under some circumstances the tank can rupture, even if protected by such a safety valve, often killing firefighters.
The probability of a propane overfilling fire will be increased if abnormal means are utilized to increase the temperature of the propane liquid for the purpose of increasing vapor pressure. As previously discussed, a propane tank refilled at an outside air temperature of 35 F will only produce a vapor pressure of approximately 59 PSI. Vapor pressures of between 90 to 160 PSI are recommended for both performance and safety. When temperatures are below 50 F and tanks have been exposed to colder temperatures for a period of time some means of increasing vapor pressure are usually required to achieve desired performance.
One of these methods employed widely in hot air ballooning is through the use of electrical heat tape, with or without a thermal tank cover. While this method most certainly will increase the temperature of the liquid propane and result in an improved vapor pressure, few if any balloonists may be aware of the potential danger of employing this method. Let's look again at what happens during winter flying conditions.
As previously noted a 10 F rise in temperature will result in approximately a 2% increase in propane volume. If a tank is filled to 85% capacity, then a temperature increase of approximately 75 F will increase the liquid volume to 100% of tank capacity. If the tank was fueled at an outside temperature of 35 F this point of total volume capacity will be reached at approximately 110 F (35 F + 75 F). The method of utilizing electrical heat tape will raise the temperature of the liquid propane to 120 F or more depending on time, insulating characteristics of any tank covers being used, etc. Once the expansion of the liquid propane reached full capacity each additional degree of heating will increase the internal tank pressure by approximately 45 PSI. Thus we can expect that the pressure relief valve would operate at approximately 123 F or within 30 F of being at full volume. Once activated the relief valve will release both vapor and liquid to the atmosphere. The vapor will expand to 270 times its initial volume, is heavier than air and will move along the ground until it finds an ignition source. When ignited the flame will flash back to the source increasing the heating efforts on the tank further continuing the cycle. Even if the relief valve were to close following the initial opening, further heating in the area will continue the cycle.
Any method used to increase the vapor pressure through heating the liquid propane will result in similar hazardous conditions. What, therefore, are the alternatives for the balloonist who desires to fly in colder temperatures?
One promising solution for consideration is the recent development of a nitrogen pressurization system from Aerostar. Properly employed, the pressurization of the liquid propane immediately prior to flying in colder temperatures will provide artificially adequate vapor pressures for acceptable performance and yet eliminate the exposure to risk of fire from other more popular methods currently being employed.
Ultimately the decision on what means or method will be used to ensure adequate pressures for acceptable performance is solely that of the pilot. That decision should be an informed one carefully made and recognizing the risks to be encountered and the benefits to be derived. When it comes to safe winter flying, hopefully this information will help make that decision a safe one.
William Peterson is Fire Chief, City of Plano (Texas) Fire Department and active hot air balloon pilot. He has lectured on this subject at balloon safety seminars.
