Although HCCI engines promise low NOx emissions with high efficiency, they suffer from a narrow operating range between knock and misfire because they lack a direct means of controlling combustion timing. A series of previous studies showed that reformer gas, (RG, defined as a mixture of light gases dominated by hydrogen and carbon monoxide), can be used to control combustion timing without changing mixture dilution, (λ or EGR) which control engine load. The effect of RG blending on combustion timing was found to be mainly related to the difference in auto-ignition characteristics between the RG and base fuel.
The practical effectiveness of RG depends on local production using a fuel processor that consumes the same base fuel as the engine and efficiently produces high-hydrogen RG as a blending additive. Depending on the base fuel, the reforming technique and the reformer operating condition, the molar ratio of hydrogen to carbon monoxide in the RG may vary from more than 3/1 down to around 1/1. One possible barrier to using RG for combustion control is the variation of H2/CO ratio in the RG composition for practical small-scale reformers.
This paper reports on a series of experimental studies using RG blending to control ignition timing with three base fuels: n-Heptane (representing diesel fuel), iso-octane (representing gasoline), and natural gas (commonly used for industrial SI engines). These fuels were tested in a CFR engine operating in the HCCI mode with blends of two different simulated RG compositions characterized by H2/CO ratios of 3/1 and 1/1 to cover a range of actual fuel processor output.
It was found that, for all three fuels, RG blending could provide combustion timing control despite the wide range of reformer gas composition. Taking the n-Heptane case as an example, RG blending retards combustion timing for both RG compositions but the retardation is less for the low H2 fraction RG (1/1) than for the high H2 fraction RG (3/1). It can be concluded, (and modeling supports), that the lower H2 fraction RG has less capability to suppress the radical pool between 1st and 2nd stage ignition. The study results support the possibility of RG blending for HCCI timing control. They also illustrate a trade-off between RG quality, (measured by H2 concentration), and RG quantity required for effective timing control.