Abstract
The partial replacement of diesel fuel with gaseous fuels in diesel engines allows for reducing soot, increasing the renewable fraction of the fuel and decreasing CO2 emissions. Hydrogen is a promising alternative; since it is a non-carbon compound, it can be produced from renewable sources and it has suitable combustion properties. However, the use of hydrogen in diesel engines could require some modifications on the engine calibration. Among the different phenomena involved in diesel combustion, autoignition significantly affects the engine efficiency. This work analyzes the autoignition behavior of diesel and biodiesel fuels under a H2-rich ambient. Two different liquid fuel replacements (10% and 20% by energy) have been tested in a constant-volume combustion chamber. Three different chamber temperatures (535°C, 602°C, and 650°C) and equivalence ratios (0.4, 0.6, and 0.8) have been checked. Results show that, in the case of diesel fuel, hydrogen delays autoignition and reduces the combustion rate, the latter caused by a higher fuel dilution with air. The influence of H2 in the autoignition of biodiesel is less significant. A reduction in the OH radicals pool appears as the main reason for retarding ignition. The lower pressure peaks with hydrogen suggest unburnt hydrogen to be relevant.
Introduction
Nowadays, most of the stationary engines for power generation are either diesel engines (most often questioned due to pollution) or gas engines. The former has better efficiency, whereas the latter uses a less pollutant fuel, since the particulate matter (PM) emission when the gaseous fuel is burned is very low or even nil depending on the fuel. For these reasons, the use of gaseous fuels in diesel engines is expected to reduce PM emissions, keeping a relatively high efficiency.1–4 Moreover, typical gaseous fuels, such as natural gas, also have a low specific CO2 emission, this aspect also being a priority target in the European Union (EU) directives.
In dual-fuel operation mode, and although direct injection of the gas has also been used, the gaseous fuel is normally introduced along with air in the intake manifold. 5 The diesel fuel injected at the end of the compression stroke acts as igniter. The most widely used fuels under this mode are natural gas6–9 and liquefied petroleum gas (LPG), 10 although renewable gases such as biogas,11,12 syngas,13,14 or producer gas15–17 have also been tested. However, this study focuses on hydrogen since it is a non-carbon compound, therefore, with a nil contribution to PM and CO2 emissions, which could come from renewable sources. Among the renewable techniques to produce H2, electrolysis using wind, photovoltaic or hydro power, biomass gasification, or biogas steam reforming 18 are being explored. Besides stationary engines, H2 can also be used in light applications (vehicles and small ships) since there are very well-known techniques such as hydrogen compression,6,19 liquefaction, and physical and chemical storage in metal hydrides 18 that can be used for H2 storage in relatively reduced volumes. Other less commercial techniques but with an interesting future potential are those based on hydrogen carriers such as ammonia, 20 which is already necessary for the selective catalytic reduction (SCR) of NOx in diesel vehicles, or methanol. 21
Some previous studies related to H2 addition/substitutions in a diesel engine have been carried out by Hernández and collegues22–24 and others.25–27 In these publications, a slight increase in NOx emissions and a decrease in efficiency apart from the obvious sharply reduction in PM and CO2 emissions are observed. As NOx increases when H2 is added, some techniques such as high exhaust gas recirculation (EGR) rates 28 or N2 addition29,30 are used to reduce its emission in dual-fuel engines. Hydrogen has also been proved to reduce the mechanical vibration in diesel engines due to its effect on the ignition delay and thus on the premixed combustion phase.31,32 Since autoignition, which is significantly affected by the fuel and the in-cylinder conditions, is one of the main phenomena governing diesel combustion, a deep analysis of this process is highly recommended to properly understand the effect of H2. However, and although several papers show the effect of hydrogen on the ignition delay measured in engines,33–36 there is a lack of work dealing with autoignition in the presence of H2 under well-controlled conditions such as those existing in constant-volume combustion chambers or shock tubes. Moreover, those systems also allow for isolating the role of each operating variable (temperature, pressure, equivalence ratio). Engine experiments, such as those previously referenced, are subjected to several uncertainties, which limit the reliability of the ignition delay time determination. Delay time is affected when using H2 due mainly to the lower O2 concentration (considering the same intake manifold pressure), the higher specific heat of the mixture, and/or pre-reactions leading to compounds less reactive than oxygen. For this reason, and in order to accurately evaluate the effect of H2 addition on the ignition delay, this article analyzes this effect in a constant-volume chamber equipped with a common-rail diesel injector. Conventional diesel fuel and biodiesel were used as reference fuels. In addition, two liquid fuel replacements, three different chamber temperatures, and three equivalence ratios were tested.
Experimental setup and fuels
Tests were carried out in a constant-volume combustion chamber (Cetane ID510 by PAC-Herzog). This equipment, used in previous works,37–39 consists of a 0.473 L combustion vessel and a common-rail diesel injector (operating at 1000 bar injection pressure), similar to those used in commercial engines, with six nozzles and diameter of 0.17 mm. The device, which is schematically shown in Figure 1, has three pressure sensors (a dynamic pressure sensor to correct the temperature offset of the dynamic sensor, an injection pressure sensor, and an inlet air pressure sensor) and two temperature sensors (two thermocouples type K for the chamber inner wall and the cooling jacket). A mixture of synthetic air and hydrogen, heated during approximately 1 min to reach equilibrium temperature with the walls of the combustion chamber, was introduced before the injection of the liquid fuel. The quantity of H2 in the bottles of synthetic air (two bottles with 1.4% and 2.8% by mol. each) was previously calculated considering the desired substitutions (around 10% and 20% by energy, respectively) for the diesel fuel. The injection pulse width set as reference was 2500 µs (as established in the standard for the derived cetane number (DCN) determination), pure air in the chamber, and 21 bar and 602°C in-chamber pressure and temperature, respectively. Moreover, the maximum injection pulse of the injector (3000 µs) was also considered. Once the initial temperature was reached, the coolant temperature, the chamber wall temperature, and the chamber pressure were checked to be within a close tolerance with respect to set values (2°C, 0.2°C, and 0.2 MPa, respectively). In every test, 15 fuel injections at each operating condition were carried out in order to provide the average pressure evolution considering repeatability. The equipment is normally used to measure the DCN 40 and it also allows for the determination of two different ignition delays: IDCF associated to the cool flame stage and IDM associated with the main combustion event. The procedure to determine these times is widely described elsewhere. 41

Schematic diagram of the experimental setup.
The aim of this article is to evaluate the effect of hydrogen addition on the ignition delay of diesel and biodiesel fuel. The liquid fuel replacement with hydrogen (EH2), which was achieved decreasing the injection pulse width, was calculated as the energy content of the latter with respect to that of the total fuel injected (H2 and liquid fuel), as shown in equation (1) (where the sub index “lf” refers to the liquid fuel and LHV is the lower heating value). In order to keep constant the total energy inside the combustion chamber as well as the equivalence ratio for specific initial conditions (p0 and T0), the maximum substitution with H2 was limited by the chamber pressure (around 20%, 15%, and 10% at 21, 16, and 11 bar, respectively). It is important to remark that the volume of fuel injected for a specific pulse width (mlf, Tables 3 and 4) has been calculated from a correlation shown in the standard for the determination of the DCN. 40 This correlation has been obtained for hexadecane as reference fuel, widely used as diesel fuel surrogate when physical phenomena were analyzed (atomization, jet macroscopic parameters, etc.). 42 Thus, although some mistakes when calculating the injected volume of the diesel and biodiesel fuels used are expected, which could lead to slight deviations on the total energy inside the chamber, these errors are predictable to be insignificant. Moreover, since the main target of the work is to evaluate the effect of hydrogen on the ignition delay (comparative analysis), the observed trends for a specific fuel (diesel or biodiesel) are reliable enough. Three different initial chamber temperatures (T0 = 535°C, 602.5°C, and 650°C) and pressures (p0 = 11, 16, and 21 bar) have been tested. These initial pressures correspond with approximately 0.4, 0.6, and 0.8 of equivalence ratio (Fr) considering both fuels, liquid and gas, and defined with respect to the stoichiometric one. Both the initial temperatures and pressures were selected according to the equipment limitations. Diesel fuel was supplied by the oil company Repsol S.A., without biodiesel content. Biodiesel was donated by Bio Oils and was produced from soybean and palm oils, whose main components are 47.26% C18:2, 26.22% C18:1, and 15.62% C16:0. The main properties of both fuels are shown in Table 1. Some relevant properties for hydrogen are shown in Table 2.
Properties of diesel and biodiesel.
Relevant properties of hydrogen.
Results and discussion
The results obtained when diesel and biodiesel fuels were replaced by hydrogen are detailed in Tables 3 and 4, respectively. pmax is the maximum pressure reached, while dp/dtmax represents the maximum pressure gradient. As shown in the tables, and because of the very lean mixtures tested as well as the very high heating value of hydrogen (by mass), the substitutions considered do not lead to a significant reduction on the oxygen content (mO2(g)) for an specific condition (initial p and T), thus limiting the effect of this variation on the observed results.
Results for diesel fuel.
Results for biodiesel fuel.
Figure 2 shows the pressure evolution inside the combustion chamber at 21 bar of initial pressure. As expected, a sharp decrease of the ignition delay when temperature increases was obtained. The lower maximum pressure reached when hydrogen replaces liquid fuel is mainly due to unburned hydrogen located possibly near walls. Although the diesel jet starts combustion, only hydrogen entraining the diesel flame is supposed to be burnt due to the very lean hydrogen/air mixtures tested, which difficult propagation of flames from the diesel plume (as proved in previous works). 23 Moreover, molecular expansion of combustion products with respect to reagents is lower for hydrogen than for diesel fuels, thus also leading to lower peak pressures when replacing the latter. In agreement with the previous mentioned comments, the pressure reduction is less significant for the lowest replacement.

Instantaneous pressure evolution at p0 = 21 bar.
Regarding the effect of H2 on the autoignition time, Figures 3 and 4 show this influence to be more important for diesel than for biodiesel fuel. In general, H2 addition delays autoignition. As it can be deduced from Tables 3 and 4 and the mentioned figures, conditions leading to longer delays, consisting of low temperature and/or low pressure, are more sensitive to the replacement. It is important to remark that increasing the initial chamber pressure is performed through a higher air amount, thus limiting the evaluation of the isolated effect of pressure and equivalence ratio. However, and despite of the irrelevant oxygen concentration changes (Tables 3 and 4), higher equivalence ratios could limit the access of the diesel fuel to oxygen and thus slightly delaying autoignition. The observed changes on the delay time when using H2, which are produced by pre-reactions leading to less reactive intermediate compounds (when compared to oxygen), have been reported in other works.43,44 H-abstraction reactions are the main routes governing the oxidation kinetics of long-chain hydrocarbons at low and intermediate temperature, 45 the OH radical being a very important branching specie. Thus, the increase in the ignition delay when using hydrogen could be due to the partial OH consumption by the latter through the reaction H2 + OH = H2O + H. 46 Although H2 could also produce OH radicals (H + O2 = OH + O and H2 + O = OH + H), these reactions are not likely to occur at the low temperatures preceding autoignition. 47 In case of biodiesel, the oxygen of the carbonyl group could slightly mask the mentioned reduction in the OH radicals pool, thus diminishing the effect of H2.

Instantaneous pressure evolution at 602.5°C (solid lines correspond to increasing substitutions with H2).

IDM at 602.5°C for diesel and biodiesel fuels.
Furthermore, Figure 3 also proves that the combustion velocity of the diesel fuel is lower when it is replaced by hydrogen (mainly at 535°C), especially for high H2 concentrations. This trend can be also checked in the pressure gradient values shown in Table 3, and it could be due to the more significant diesel fuel dilution (oxygen entrainment during the delay time) when hydrogen is added. The extremely high mass diffusivity of hydrogen could significantly improve the air mixing phenomena. In the case of biodiesel, the effect of hydrogen in the combustion rate appears to be less important, again due to the fuel oxygen content, which promotes combustion.
Conclusion
The main conclusions are described as follows:
A significant part of the hydrogen is not burned, as proved by the lower pressure inside the combustion chamber.
Under the conditions tested (low liquid fuel replacements), hydrogen addition promotes chemical activity prior to autoignition, which affects the ignition delay time. This effect is more significant for diesel than for biodiesel and mainly under low temperature and/or low pressure conditions.
The combustion velocity is lower when diesel fuel is replaced by hydrogen. This trend could be caused by a more significant diesel fuel dilution when hydrogen is added due to the extremely high mass diffusivity of the latter. In case of biodiesel, the structural oxygen content masks this effect.
The previous conclusions (variations on the delay time as well as on the combustion rate) suggest that slight modifications in the injection timing through engine calibration are required when diesel or biodiesel fuel is partially replaced by hydrogen in a diesel engine. On the contrary, the combustion phasing could lead to low engine efficiencies. Moreover, the expected unburnt hydrogen highly recommends the use of engine operating conditions promoting propagation of flames from the diesel jets (i.e. H2/air ratios within the flammability limits and/or limiting dual mode at high loads to achieve high in-cylinder temperatures).
Footnotes
Handling Editor: Jose Ramon Serrano
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
