Alcohol fuels for spark-ignition engines: Performance, efficiency and emission effects at mid to high blend rates for binary mixtures and pure components

James Turner, Andrew Lewis, Sam Akehurst, Christian Brace, S. Verhelst, J. Vancoillie, L. Sileghem, Felix Leach, Peter P. Edwards

Research output: Contribution to journalArticlepeer-review

37 Citations (SciVal)
157 Downloads (Pure)


The paper evaluates the results of tests performed using mid- and high-level blends of the low-carbon alcohols, methanol and ethanol, in admixture with gasoline, conducted in a variety of test engines to investigate octane response, efficiency and exhaust emissions, including those for particulate matter. In addition, pure alcohols are tested in two of the engines, to show the maximum response that can be expected in terms of knock limit and efficiency as a result of the beneficial properties of the two alcohols investigated. All of the test work has been conducted with blending of the alcohols and gasoline taking place outside the combustion system, that is, the two components are mixed homogeneously before introduction to the fuel system, and so the results represent what would happen if the alcohols were introduced into the fuel pool through a conventional single-fuel-pump (dispenser) approach. While much has been written on the effect of blending the light alcohols with gasoline in this way, the results present significant new findings with regard to the effect of the enthalpy of vaporization of the alcohols in terms of particulate exhaust emissions. Also, in one of the tests, two mid-level blends are tested in a highly downsized prototype engine – these blends being matched for stoichiometry, enthalpy of vaporization and volumetric energy content. The consequences of this are discussed and the results show that this approach to blend formulation creates fuels which behave in the same manner in a given combustion system; the reasons for this are discussed. One set of tests using pure methanol alternately with cooled exhaust gas recirculation and with excess air shows a significant increase in thermal efficiency that can be expected as the blend level is increased. The effect on nitrous oxide emissions is shown to be similarly beneficial, this being primarily a result of the enthalpy of vaporization of the alcohol cooling the charge coupled with the lower adiabatic flame temperature of the alcohol. Whereas it is normally a beneficial characteristic in spark-ignition combustion systems, one disadvantage of a high value of enthalpy of vaporization is shown in another series of tests in that, in admixture with gasoline, it is a driver on flash boiling of the hydrocarbon component in the blend in direct injection combustion systems. In turn, this causes the particulate emissions of the engine to increase quite markedly over those of ethanol–gasoline blends at the same stoichiometry. The paper shows for the first time the dichotomy of the potential efficiency improvement with the challenge of particulate control. This effect poses a challenge for the future introduction and use of such high blend rate fuels in engines without particulate filters. Although it must be stated that the overall particulate emissions of the methanol–gasoline blend are lower than for gasoline, the effect is only present at extremely low load, and that there is a likelihood that particulate filters will be adopted in production vehicle anyway, nullifying this issue.

Original languageEnglish
Pages (from-to)36-56
Number of pages21
JournalProceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering
Issue number1
Publication statusPublished - 1 Jan 2018


  • Alcohol
  • downsized
  • ethanol
  • fuel efficiency/economy
  • methanol

ASJC Scopus subject areas

  • Aerospace Engineering
  • Mechanical Engineering


Dive into the research topics of 'Alcohol fuels for spark-ignition engines: Performance, efficiency and emission effects at mid to high blend rates for binary mixtures and pure components'. Together they form a unique fingerprint.

Cite this