GDI is a strategy to improve fuel efficiency that is rapidly gaining market acceptance, as illustrated by the popularity of Ford’s GDI EcoBoost engines. (Earlier post.) However, unlike conventional gasoline engines, gasoline direct injection engines produce particulate matter, as do diesels, and emission controls will become an issue the technology must address as standards for particle mass tighten, and also as standards for particle numbers emerge.

(The EU has already set a standard for 6 × 10¹² number/km limits between 2014 and 2017, tightening to 6 × 10¹¹ number/km.)

[GDI] technology has evolved considerably over the past decade and half, and is still undergoing development. Early versions of direct injection employed a stratified injection strategy which included late fuel injection and charge air motion to produce a stoichiometric air/fuel mixture in the vicinity of the spark plug and, thereby, avoid pumping losses from the need to throttle the engine at light load. The drawback is that direct injection of fuel into the combustion cylinder risks liquid fuel impingement onto the piston and cylinder surfaces and inhomogeneous air fuel mixing; consequentially PM formation is higher than in conventional PFI gasoline engines. This is difficult to avoid during late injection as the piston approaches the top of its stroke, therefore more recent development of GDI technology has focused on early injection, or so-called homogeneous operation.

Homogeneous GDI provides fuel efficiency benefits from charge air cooling, more facile turbocharging, and the downsizing that these permit. The efficiency advantage is not as large as with diesel engines, making use of gasoline particulate filters (GPF) a less practical solution to PM emissions aftertreatment. For this reason, there are currently significant research efforts aimed at understanding how PM emissions can be reduced by engine design (e.g., fuel injection timing, number of injections, fuel pressure, injector placement, etc.) and fuel properties. These efforts are yielding reductions in PM emissions, but this is still a work in progress. As a result, surveys of current on-road GDI vehicles show a wide variety in PM emissions that is more a snapshot of technological progress than one of emissions capability.

One important real-world aspect of emissions is to understand how these change with vehicle age. Such information is needed to develop emissions inventories and understand environmental impacts. This type of data is difficult to obtain because it requires prolonged mileage accumulation and periodic vehicle testing from new to 150K miles. Over the past few years, we had the opportunity to gather such data from two current-technology GDI vehicles that were being used to study the aging of a prototype catalyzed GPF [gasoline particulate filter] and which exhibited PM levels near or below the 3 mg/mi standard.

—Maricq et al.

Schematic diagram of test vehicle engine & exhaust system and PM sampling systems. MSS = AVL micro soot sensor, APC = AVL solid particle counter, DiSC = diffusion size classifier. Credit: ACS, Maricq et al. Click to enlarge.

The two 2010 model year test vehicles represented two models, but had the same engine and similar aftertreatment. Fuel injectors were side-mounted. The standard exhaust system included close-coupled catalysts on each engine bank to assist with fast light-off and a common underbody catalyst after the two exhaust streams merge. The vehicles were fueled with an E10 ethanol−gasoline blend.

The Ford team took PM measurements just upstream of the GPF as a surrogate for tailpipe values. The vehicles retained their close coupled catalysts; the researchers monitored the GPF’s catalytic efficiency for HC removal to ascertain possible changes to PM contribution from semivolatile material. At the end of the study a laboratory-aged underbody catalyst was refitted into the exhaust system to confirm the 150K mile emissions using the standard constant volume sampling (CVS)/gravimetric filter method prescribed for PM mass emissions measurement.

Among their findings were:

• The two vehicles exhibited relatively higher, 4−14 mg/mi PM emissions during cold start, which dropped by a factor of 2−5 for warmed-up operation. This is typical behavior for GDI engines where cold piston and cylinder surfaces exacerbate liquid fuel impingement and reduce evaporation from surfaces, which produces soot when the fuel ignites, the team noted.

• Relative to a 3 mg/mi standard, the PM mass emissions remain stable over the 150K miles—except for sporadic variability in the cold start emissions of vehicle 2. Emissions began at 3−4 mg/mi at 4K miles (vehicle 1) and dropped to FTP weighted average rates of about 1.5 mg/mi at 30K and 60K, the lowest levels for both vehicles. There is a subsequent step up to 2−3 mg/mi for 90K−150K miles, but there is no trending rise in emissions. The US06 emissions follow the same pattern.

• Tailpipe solid particle number emissions for the two test vehicles are 3.5 and 3.8 × 10¹² per mile for the weighted FTP average. The present vehicles thus appear capable of achieving the first phase of EU particle number standards, but will require further development to meet the longer term standard.

• The PM is predominantly (∼80−90%) soot. This is consistent with previous optical engine studies of fuel impingement and with the observed correlation between sooting tendency and decreasing fuel volatility. The high soot content suggests a similarity to non-DPF equipped diesel PM.

• Unlike non-DPF diesel vehicles, GDI vehicle PM emissions are predominantly a cold start phenomenon.

The results of this work provide two examples to suggest that GDI vehicles, which are introduced to improve fuel economy, can meet upcoming US and California PM standards over 150K miles of operation. The instrument-recorded PM mass and number emissions are relatively stable over 150K miles; the overall variation is about a factor of 2, but the emissions remain near or below 3 mg/mi. And the final CVS tests exhibit 1−2 mg/mi gravimetric tailpipe emissions for both the FTP average and US06 drive cycles. However, further development is needed to achieve the LEV III goal of 1 mg/mi, and it is not clear if this can be accomplished solely by engine design or will require GPFs.

...The test vehicles in this study show that GDI technology has the potential to meet upcoming 3 mg/mi standards over the lifetime of the vehicle, but further work must address the question of 1 mg/mi...This engine technology requires more control parameters than conventional gasoline engines and the engine community has less collective experience in their optimization. It is likely that continued development will achieve further improvements in fuel economy as well as PM emissions over a wider range of engine designs and displacements.

—Maricq et al.

Resources

• M. Matti Maricq, Joseph J. Szente, Jack Adams, Paul Tennison, and Todd Rumpsa (2013) “Influence of Mileage Accumulation on the Particle Mass and Number Emissions of Two Gasoline Direct Injection Vehicles”, Environmental Science & Technology 47 (20), 11890-11896 doi: 10.1021/es402686z