plastic and rubber components of the fuel system.
Evaporative emissions depend mainly on ambient temperature, fuel volatility and fuel system design . As far as the fuel is concerned, Vapour Pressure (Dry Vapour Pressure Equivalent - DVPE) is a key parameter: higher values of DVPE, which is measured at 37.8 °C, mean that the fuel is more volatile or, in other words, that more fuel will evaporate at that temperature. When blended with gasoline, ethanol is known to cause an increase of the ethanol/gasoline blend DVPE compared to the base gasoline.
As a consequence of the European policy aiming to promote the use of ethanol and other biofuels  , the question has arisen whether the vapour pressure limit for ethanol/gasoline blend should be relaxed with respect to the current fuel specifications laid down in the Directive 98/70/EC.
The programme described in this report was conceived to assess specifically the effect of ethanol/gasoline blends on evaporative emissions with the objective of providing a technical basis for discussion on this issue.
A range of seven gasoline passenger cars representative of current EURO 3-4 technology were tested for evaporative emissions with ten different test fuels. The test fuel matrix comprised 60 and 70 kPa hydrocarbon base fuels with 5 and 10% ethanol splash blends and 5 and 10% ethanol matched volatility blends. The evaporative emission tests were carried out according to a test protocol agreed by the partners of the programme and based on the European homologation test procedure.
This test procedure turned out to have a considerable influence on the results; in particular repeating the test procedure, without any additional conditioning between tests above that prescribed by the legislation, resulted in an increase of the carbon canister weight with successive tests. In other words, the test protocol was not able to return the vehicle to a
consistent condition at the start of each test. The increase of the canister weight, which is indicative of the accumulation of gasoline/ethanol vapours probably due to insufficient canister purging, may not represent real-world operating conditions as the canister load depends on the typical driving patterns. Running at these increased canister weights is a very severe test of the evaporative control system. The canister weight problem made it more difficult both to obtain representative emission measurements from the various vehicles and to clearly determine fuel effects. However the programme has provided valuable information and several clear conclusions can still be drawn from the results.
The vehicles tested differed in their level of evaporative emissions and in the extent of their response to fuel changes. All cars met the 2 g/test emission limit on the first test on fuel A, the reference fuel with DVPE of 60 kPa. Some vehicles slightly exceeded the limit on subsequent tests on fuel A, probably related to increased canister loading in later tests.
The test results confirmed that vapour pressure (DVPE) is a key fuel variable for evaporative emissions; in general, increasing fuel vapour pressure above that of the 60 kPa DVPE reference fuel used for system development increased evaporative emissions. However the effect of vapour pressure is strongly non-linear, as expected for a process in which a vapour breakthrough effect may occur. The ethanol blends with final DVPE around 75 kPa gave considerably higher evaporative emissions than the other lower volatility fuels in most of the vehicles. Differences between fuels with DVPE in the range 60-70 kPa were small.
Furthermore, due to the combination of DVPE variations, the presence or absence of ethanol, and to significant changes of canister weight it is difficult to draw any reliable conclusions on the influence of individual parameters. The results obtained in a few tests where extra purging of the canister was carried out suggest that differences in evaporative emission measurements on fuels in this volatility range could be reduced if a more extensive canister conditioning procedure was adopted. The engineering margin built into the system may also explain the reduced fuel effect. The evaporative emission control system is designed for the DVPE of the reference fuel (60 kPa) used in the homologation test but, as for other emission control devices, the manufacturer introduces a certain margin to take into account the production variability.
Ethanol might influence evaporative emissions also via different mechanisms than the increased vapour pressure of ethanol/gasoline blends . For example ethanol is known to be more difficult to purge from carbon canisters (as are heavy hydrocarbons), so could reduce their working capacity. To explore this possibility ethanol/gasoline blends matching the vapour pressure of the pure hydrocarbon base fuel were included in the fuel matrix. The increase in canister loading noticed during the programme could be due to increased adsorption of hydrocarbons or ethanol in the canister, or both. Unfortunately the poor repeatability of the main data set does not allow us to quantify the relative size of these effects. However multiple additional tests on one vehicle showed that ethanol containing fuels with matched volatility gave higher emissions than the hydrocarbon fuels. Ethanol was also found in the VT SHED vapour of tests on pure hydrocarbon fuels following use of ethanolcontaining fuels. Further research would be required to clarify these effects.
Extra diurnal emission tests were carried out on two vehicles with the canister vented outside of the VT SHED. The results of these tests suggest that fuel permeation through plastics and rubbers could be a significant contributor to evaporative emissions. As demonstrated by other studies, ethanol does increase the fuel permeation rate   . Measurements of regulated exhaust emissions showed few statistically significant differences between fuels. However, the test programme was not designed to look at exhaust emissions and this could explain the noticeable variability of the exhaust emission measurements. Specific conditioning between tests and multiple testing is essential for such studies.
Nevertheless the data show clearly that volumetric fuel consumption (litres/100 km) increased with increasing ethanol content. This increase was roughly proportional to the oxygen content of the fuel. However there was no effect of ethanol on energy consumption.
The test programme was designed to explore only the effects of ethanol and fuel vapour pressure on evaporative emissions from a range of latest generation canister-equipped gasoline cars using the EU Evaporative Emissions test procedure. Other parameters like test temperature profile, presence of ethers in the fuel, fuel permeation and the long term effect of ethanol and water on carbon canister working capacity have not been addressed by this programme.