control factors that vary by year and also defect rates that vary by year. Since these numbers are not readily accessible at this time, distribution emissions were based on the assumptions shown in Table 4-26. The control factors and defect rates were applied to M100. The baseline uncontrolled emissions were then modeled from the fuel's vapor pressure and vapor molecular weight.
Future refueling emissions might be lower due to reduced Stage 2 failures with vacuum assist nozzles and other fueling equipment; however, the extent of such reductions if any has not been quantified.
Distribution emissions are estimated from the emission control assumptions in Table 4‑26, and the fuel's properties. An underlying assumption is that the control effectiveness is equal for all fuels. There might be greater variation in control effectiveness for small volume fuel distribution (perhaps higher or lower) and underground tank breathing losses would depend upon the fueling station throughput. The number of vehicles per station will vary as more alcohol-fueled vehicles are introduced into the State. A total vehicle to fueling station ratio of about 2500 currently exists for gasoline-fueled vehicles. Many of these stations have multiple tanks. As alcohol-fueled vehicles are introduced, the ratio of vehicles to fuel tanks will be approach to the ratio for gasoline tanks. Therefore, differences in breathing losses due to differences in throughput will be small on a lb per day basis during an alcohol vehicle transition and would become negligible for a larger fleet of alcohol-fueled vehicles. Second order effects on breathing losses (per gallon) would depend on whether gasoline tanks are decommissioned if alternative-fueled cars displace gasoline-fueled cars. Given equivalent assumptions on emission control, distribution emissions depend on fuel vapor pressure and vapor molecular weight. Scenario 3 includes a downward adjustment by 50 percent of service station vapor losses to reflect reduced emissions from storage tank breathing and a potentially lower defect rate for Stage 2 control.
Vehicle Fueling Spillage
Service station fueling practices were also observed to evaluate vehicle fueling. The dispensers at numerous fuel stations were polled to determine the amount of fuel dispensed per fueling event. The amount of fuel dispensed ranged from one half to 18 gallons with an average of 8 gallons. The volume of fuel dispensed is important in determining emissions that depend on the number of fueling events rather than fuel volume. Various vapor recover nozzle types are used at service stations in California. At self-service stations, the vehicle driver dispenses the fuel. Most customers select the self-service option that results in a lower fuel price. While most vehicle operations are successful with little fuel spilled from the nozzle, occasionally a significant quantity of fuel is spilled. Fuel spills from vehicle refueling were evaluated by ARB (Morgester). This paper estimated spills from small drips as well as large refueler error type accidents. The average fuel spill size was estimated to be 0.42 lb/1000 gal. An API study estimated spill volumes at 0.31 lb/1000 gal.