Assessment of the safety of onboard-refueling vapor-recovery systems. Final report, Jun-Jul 91

The test program was performed to investigate potential fire hazards from the use of larger charcoal canisters for on-board fuel vapor recovery, especially for fuel vapor recovery system malfunctions or during vehicle crashes. Crushed canisters from four crashed test vehicles were inspected. Testing was then conducted using new and used charcoal canisters with different vapor saturation levels (0% to 100%), different canister case integrities (intact, two holes, crushed by impact), and different ignition sources (spark, flame, hot surface). Sudden bursting of saturated canisters sometimes resulted in rather explosive fires. Less saturated canisters resulted in varying degrees of fire. Some tests were so low in vapor content that no fire resulted. To estimate typical levels of gasoline vapors accumulated in the canisters, three vehicles were refueled and parked outside for five to seven days. These canisters achieved approximately 40% to 65% of saturation capacity. Tests were also conducted to simulate broken or separated fuel vapor recovery hoses using a modified fuel tank. The tank and gasoline (3 gal) was allowed to heat-soak and water was injected into the tank to simulate a refueling rate of approximately 8 gallons per minute. Three ignition sources were used to ignite the fuel vapors atmore » the open end of the vent hose and the end of the hose became a flame thrower during the filling operation. It was noted that the evaporative emission from the tank was sufficient to sustain a small flame even without the filling process to force vapors out.« less