Mixture Formation in Hydrogen Engines

This is a preview of the paper, limited to some initial content. Full access requires DieselNet subscription.
Please log in to view the complete version of this paper.

Abstract: Hydrogen has unique thermochemical properties—such as low density, high diffusivity, and wide flammability limits—that offer both opportunities and challenges for engine development. Most hydrogen engine development effort has been focused on high-pressure direct injection (HPDI) systems that offer a high degree of flexibility over combustion control. HPDI can be further divided into different injection strategies, including pre-mixed HPDI, stratified HPDI, split-injection, and diffusive HPDI. Meeting the mixture preparation needs for the various hydrogen injection strategies requires designing specialized, hydrogen-specific fuel injectors.

Introduction

Hydrogen as a Fuel

As energy systems de-fossilize and low carbon electricity production increases, hydrogen is promoted as an energy storage medium that could be used during periods of excess electricity production. In this context, the prospects of hydrogen fueled internal combustion engines (hydrogen ICE) as a transport decarbonization pathway are increasingly being explored, and they are thus the subject of active research and development. There is particular interest from the heavy-duty and nonroad sectors where powertrain electrification is challenging because of high power density and “range” requirements [5982][5800]. From a technical feasibility standpoint, the hydrogen ICE has been demonstrated as a viable technology (e.g., Ford [5984], BMW [5985], JCB [5800], Mazda [5986]) but for it to be competitive against other decarbonization options, most notably electrification and hydrogen fuel cells, its efficiency and power density need to be improved and low pollutant emissions (mainly NOx) need to be ensured. There are limitations to such improvements for port fuel-injected (PFI) hydrogen engines, which can potentially be addressed by direct-injected (DI) engine designs. This article presents an overview of the direct-injected hydrogen engine concept and discusses the potential of different injection strategies and their current state of development.

Hydrogen has unique thermochemical properties that offer both opportunities and challenges for engine development [5982][5988][5987]. A summary of these properties and their impact on engine performance relative to gasoline engines is provided in Table 1. For the topic at hand (injection and mixture formation), relevant properties of hydrogen are its:

Moreover, hydrogen has a higher ratio of specific heats, thermal conductivity, and speed of sound than comparable natural gas and gasoline mixtures [5982].

**Table 1**
Comparison of hydrogen to gasoline as an ICE fuel
Property	Comparison with Gasoline	Positive Effects	Negative Effects
Density	Lower (0.08 vs 692 kg/m³ at 300 K and 1 bar [5982]	High diffusivity and mixing rates	Low energy density No vaporization-induced charge cooling Limited on-board storage
Volumetric energy density	Lower (10.7 vs 33×10³ MJ/m³ at 1 bar and 273 K [5987]		Large volume of fuel needed (low power density)
Heating value	Higher (LHV: 120 vs 44.3 MJ/kg [5982])	Compensation for low density^a
Laminar Flame Speed^b	Higher (185 vs 40 cm/s at φ = 1, 1 bar, 298 K)	Thinner flames and lower minimum ignition energy Faster / complete combustion Lower turbulence requirements	Increased peak pressure, pressure rise rate, and wall heat transfer
Minimum ignition energy	Lower (0.02 vs 0.28 mJ at 300 K, 1 bar [5982])	Wide flammability limits Hot-surface assisted ignition possible	Surface-ignition (pre-ignition, backfire)
Quench distance	Shorter (0.64 vs 2.84 mm at φ = 1 [5988])	More complete combustion	Increased wall heat transfer (thinning of the thermal boundary layer) Crevice combustion and backfire [5988]
Flammability Limits	Wider (φ = 0.1-7 vs 0.66-3.85 at 300 K, 1 bar^c [5982])	Ultra-lean combustion In-cylinder NOx abatement Qualitative load control Reduced turbulence requirements	Backfire Crevice combustion Crank-case combustion
Mass diffusivity (in air)	Higher (0.61 vs 0.07 cm²/s at 1 bar and 300 K [5982])	Faster mixing and combustion	Thermodiffusive instability Preferential diffusion instability
Auto-ignitability	Unclear (higher RON, lower MON, higher AI temperature, low methane number)^d	Higher compression ratios possible	Long ignition delay Difficult to compression ignite (CI)
^a The Power deficit of a stoichiometric hydrogen PFI engine with 29% lower volumetric efficiency than a comparable gasoline engine decreased to 15% because of the higher LHV [5984]. ^b The suitability of using LFS as a flame speed index for hydrogen is unclear because of the rapid onset of cellularity, which increases the actual flame speed and makes the determination of LFS difficult. Limited LFS data is available at engine-relevant conditions [5987]. ^c Flammability limits widen with temperature and pressure. For engine-relevant conditions, practical lean limits are around φ = 0.25. ^d Ambiguity exists about the relative auto-ignitability of hydrogen and the suitability of using existing auto-ignition indices. High RON and low MON values of 130 and 60-65, respectively, have been reported for hydrogen. The low MON values are a consequence of hydrogen’s low minimum ignition energy, which at the high-temperature MON conditions leads to knocking. On the methane number scale, hydrogen has a value of 0 which suggests that it has a very high propensity to knock compared to other gaseous fuels. The auto-ignition temperature of hydrogen, 585°C, is higher than diesel (> 225°C) and gasoline (230-450°C) [5991].

Mixture Formation Needs

Modern hydrogen engines need to simultaneously meet multiple operational requirements. These include:

Figure 1 qualitatively illustrates the trade-offs between these operational requirements. The combustion strategy employed to meet these requirements determines an engine’s mixture preparation needs. It defines the operating equivalence ratio (φ) and in turn the amount of hydrogen and air that need to be supplied, the injection system type, timing of hydrogen supply, and the nature of the combustion mixture formed (i.e., homogeneous or stratified).

As illustrated in Figure 1, lean combustion (φ < 0.5) can abate NOx emissions and thus minimize the need for NOx aftertreatment. However, engine power is reduced as less chemical energy (from the fuel) is available. Supercharging can help mitigate this. NOx emissions increase rapidly at φ > 0.5 and peak around φ = 0.75. Ultra lean operation (φ < 0.3) is undesirable from an engine efficiency perspective as flame speed decreases and combustion slows down, adversely affecting combustion phasing and combustion efficiency.

Lean operation has the additional benefit of reducing the frequency of abnormal combustion events because the minimum ignition energy of hydrogen-air mixtures increases rapidly when φ is decreased from stoichiometric [5992].

Stoichiometric operation is desirable from a power output perspective even though it increases NOx emissions and the probability of abnormal combustion. Stoichiometric combustion does, however, permit the use of NOx reduction catalysts (similar to three-way catalysts) whose conversion efficiency drops significantly and rapidly at lean conditions. For moderately lean mixtures (0.5 < φ < 1) selective catalytic reduction (SCR) can be used.

Figure 1 presents the general behavior for well-mixed hydrogen-air mixtures. For stratified mixtures that have a range of φ, the cumulative effects of the combustion of different φ strata would be witnessed. For example, in a universally lean but stratified mixture, regions with φ > 0.5 will produce high NOx emissions.