. .

AMF - Implementing Agreement on Advanced Motor Fuels

Path: Home| > FUEL INFORMATION| > Butanol| > Compatibility and emissions|

Compatibility and emissions

How much butanol can be used in conventional cars?

Compatibility of butanol with conventional cars depends on technology and material choices. Modern cars with a closed-loop fuel control system can compensate the leaning effect of oxygen containing gasoline, but only up to a certain limit. When this limit is exceeded, control of stoichiometric air to fuel ratio fails, three-way catalyst is not operating correctly and exhaust emissions increase. In FFV cars, stoichiometric air to fuel ratio is maintained even with high oxygen content fuels based on capability to use high fuel injection rate (Read more of FFV cars). In Europe, 3.7 wt-% oxygen in gasoline is regarded to be compatible with today’s car population. This oxygen limit is equivalent to 10 vol-% ethanol or 16 vol-% butanol. Properties of butanol are closer to gasoline than properties of ethanol as concerns e.g. heating value, vapor pressure, water tolerance, corrosiveness, and polarity. Therefore butanol could be an option for spark-ignition gasoline cars, and also for non-road applications.

Material compatibility, water tolerance

Elastomer and material compatibility for butanol is better than that for ethanol. BP (2006) reported the following results from the material tests:

  • Weight and volume increase of viton and nitrile butadiene rubber with butanol is close to gasoline.
  • No significant changes in swelling or hardness of elastomers were observed for butanol when compared to gasoline.
  • Butanol passed 6-weeks corrosiveness tests with copper, brass, zinc, aluminum, steel, and lead.
  • Behavior of butanol is better than that of ethanol as concerns copper and brass.

Gevo (2011) refers to DNVs tests on corrosion. These tests showed that no stress corrosion cracking (SCC) in carbon steel was observed for 12.5 vol-% and 50 vol-% isobutanol blends. In addition, several tests with elastomers showed better compatibility with isobutanol blends than with gasoline.

Isobutanol and n-butanol are limitedly soluble in water. Phase separation risk is not high for these butanol isomers. Water solubilities of all butanol isomers are higher than those of gasoline. Thus small amount of water accumulated in the bottom of tank or infrastructure may be dissolved. Butanols can also act as anti-icing agents.


With gasoline the vapor mixture in the air space in the fuel tank is too rich, and with diesel too lean, to be ignitable in normal ambient temperature range. With butanol, the flammability range is close to that of gasoline.

Sec-butanol forms peroxides, which is typically a problem for ethers. (Wikipedia). This can be dealt with stability additives.


Volatility is a fuel parameter, which affects driveability. Volatility is specified by vapor pressure and distillation properties of gasoline. These properties are controlled in legislation and standards to ensure proper cold-starting, warm-up behavior, drivability, and to avoid unnecessarily high evaporative emissions. Modern cars with multi-point fuel injections systems are generally less sensitive than old cars towards volatility properties of gasoline.

Cold weather driveability is affected by mid-range volatility, defined in Europe by the E100 value (vol-% evaporated at 100°C). According to Stradling et al. (2009) mid-range distillation is significant for modern vehicles due to interrelated exhaust emissions under cold starting conditions. Butanol increases mid-range distillation range, which is seen as high E100 values. E100 is lower for n-butanol blends than for isobutanol blends (distillation). Butanols do not necessarily change the front-part distillation, and thus problems in hot-weather driveability are not expected (Stradling 2009).

The effect of isobutanol on driveability has been studied by Baustian et al. (2012). In this study, new driveability performance model was developed to optimize isobutanol blending without compromising cold-start and warm-up driveability performance.

One interesting factor that might play role in driveability, is high viscosity of butanols.

Combustion, power output and fuel consumption

BP (2006) studied ethanol and n-butanol blends with three cars, Audi A4 Avant 2.0 FSI and BMW CI. Ethanol and n-butanol were splash blended into baseline gasoline. The effect of n-butanol on power output was lower than that of ethanol. Significant increase in volumetric fuel consumption was observed for alcohol blends compared to gasoline due to low energy content of alcohols. The effect was higher for ethanol than for n-butanol. When 10 vol-% of n-butanol was added into gasoline, an increase in volumetric fuel consumption varied from 2 to 3.5%. (BP 2006).

Major part of studies with butanol are conducted with direct-injection engines on engine test benches. Yang et al. (2009) studied blends containing up to 35 vol-% of butanol in gasoline. They observed that engine power was maintained until butanol content increased up to 20 vol-%, but after that engine power dropped. Cooney et al. (2009) reported of the results with n-butanol and isobutanol in a direct-injection engine. Engine combustion strategy was not changed, closed-loop lambda feedback maintained stoichiometric operation. Brake thermal efficiencies were at same level with all fuels at low and medium loads, but at high load ethanol and isobutanol gained benefit from higher octane numbers when compared to gasoline and n-butanol. Niass et al. (2011) studied 2- and tert-butanol mixture in 15 and 30 vol-% ratios in standard gasoline with RON 97 by using direct-injection single-cylinder engine. Increased RON, MON, and heat of vaporization was observed. High and full load efficiency could be improved by earlier ignition timing. The gain in knock could contribute to reduction of CO2 emissions when compression ratio is increased or when downsizing an engine. At low and medium loads identical or slightly improved performance and emissions was detected. Improved efficiency for 2-butanol was observed also by Thewes et al. (2011).

Wallner et al. (2009) reported of a study using gasoline baseline fuel, 10 vol-% ethanol and 10 vol-% n-butanol blends in a direct-injection, four-cylinder, engine with varying loads. Brake specific volumetric fuel consumption was 3.4% higher for a blend containing 10 vol-% of n-butanol, and 4.2% higher for E10 compared to gasoline. Cairns et al. (2009) studied gasoline/ethanol and gasoline/butanol blends covering a range of oxygen content and octane numbers to identify key parameters. Turbocharged multi-cylinder, direct-injection TWC and EGR equipped engine was used. Under part-load conditions, brake specific fuel consumption was similar for 10% ethanol and 16 vol-% butanol containing fuels.

Black et al. (2010) reported of combustion properties and detailed chemical kinetic model for n-butanol. Higher oxygen concentration induced faster ignition (reduced ignition delay). Wallner et al. (2009) reported that burning velocity seemed to be higher for the n-butanol blend than for E10 blend or gasoline in a modern direct-injection spark-ignition engine. Relatively minor differences existed between the three fuels for the combustion characteristics (e.g. heat release rate and 50% mass fraction burned) at low and medium engine loads. At high engine loads, engine control unit retard the ignition timing substantially when using n-butanol blend due to the reduced knock resistance of the n-butanol blend. This was clear when compared to gasoline, and even more pronounced when compared to the ethanol blend. Other studies on combustion and e.g. spray formation with n-butanol are reported by Serras-Perreira et al. (2008) and Beeckmann et al. (2009).

Exhaust emissions

Heat of evaporation of butanol is lower than that of ethanol, but higher than that of gasoline. Increased heat of vaporization is desirable particularly for direct-injection engines according to Wallner et al. 2012. Cooling effect can reduce in-cylinder temperatures, and consequently, NOx emissions and knock propensity.

Tests with cars (post-catalyst)
Very limited data is available on the exhaust emissions with butanol containing gasoline. BP (2006) studied emissions with cars by using n-butanol as gasoline component. This study concluded that CO, HC or NOx emissions did not change significantly when 10 vol-% of n-butanol blend was compared to gasoline with the standard FTP cycle. In the highway cycle, n-butanol showed a slight reduction in CO emission.

Aakko-Saksa et al. (2011) observed that isobutanol containing gasoline decreased CO emission, but increased NOx emission with conventional and direct-injection gasoline cars at -7 °C, whereas an opposite result was obtained with the FFV car. Lower particulate matter emission was observed for butanol-containing fuels than for conventional gasoline with the direct-injection gasoline car. 1,3-Butadiene emission was higher in many cases for isobutanol or n-butanol containing fuels than for non-oxygenated gasoline. Butanol-containing fuels increased formaldehyde, acrolein, butyraldehyde, methacrolein, and propionaldehyde emissions. Sum of analyzed aldehydes was in most cases higher for butanol-containing fuels than for ethanol containing fuels. Emissions with n-butanol were higher than those with isobutanol. Aakko-Saksa et al. (2011) reported that butanol may reduce particulate matter associated PAHs from direct-injection car, but not necessarily Ames mutagenicity. For gasoline cars with indirect injection technology, particulate matter emission levels were low and no significant effect of butanol was seen.

IEA-AMF has an ongoing work on particle emissions with butanol (Annex XXXV-2).

Engine tests (pre-catalyst)
In a direct-injection, spark ignition engine at varying speeds and loads, no significant difference was seen in CO and HC emissions between ethanol blend, n-butanol blend and baseline gasoline. E10 showed the highest, and 10 vol-% n-butanol blend the lowest NOx emissions. The ethanol blend produced the highest peak specific NOx due to the high octane rating of ethanol and effective anti-knock characteristics. (Wallner et al. 2009). Yang et al. (2009) studied blends containing up to 35 vol-% of butanol in gasoline. Engine raw HC and CO emissions were reduced, but NOx emission increased with increasing butanol content.

Cooney et al. (2009) reported of the results with a direct-injection gasoline engine by using n-butanol, isobutanol, and ethanol in blending ratios up to 85 vol-% of the oxygenated fuel. Engine combustion strategy was not changed, closed-loop lambda feedback maintained stoichiometric operation. CO and HC emissions remained unchanged with butanol blends, but decreased when ethanol was compared to gasoline. n-Butanol showed slight increase in NOx emission at low engine loads, but NOx results at high loads were dependent on the exhaust gas recirculation (EGR) valve lift. Duration of injection is longer for alcohols than for gasoline to achieve same loads, and this changes EGR valve lift.

Wallner et al. (2010) used a direct-injected gasoline engine, from which EGR was disabled. In this study both regulated and unregulated emissions were measured. Ethanol, n-butanol and isobutanol were used as blending agents in gasoline. The following results were achieved: NOx emissions decreased with increasing alcohol content; formaldehyde and acetaldehyde emissions increased with n-butanol and isobutanol; a reduction in aromatic hydrocarbon emissions was observed with increased alcohol content; butanol increased propene, 1,3-butadiene, and acetylene emissions. Thewas et al. (2011) observed reduced NOx emissions with ethanol, 1-butanol, and 2-butanol due to lower combustion temperature in homogenous combustion with direct-injection engine. In general, HC emissions and particulate matter emissions were higher at high loads indicating a worse mixture formation. Alcohol fuels lead to higher oil dilution under cold condition than regular gasoline. In this study, tetrahydro-2-methylfuran was tested, as well.