How To Green The World's Deserts And Reverse Climate Change | TedTalk by Allan Savory
    Sunlight funnel collects light from all directions

    For years we heard about gadgets and gizmos improving fuel efficiency.  Some were snake oil, others worked to varying degrees. This story grabbed my attention because of its simplicity and credibility.

    ORNL’s Jim Szybist works with a multi-cylinder engine at the lab’s National Transportation Research Center. Credit: Jason Richards/Oak Ridge National Laboratory, U.S. Dept. of Energy

    Engines — Fueling innovation

    Gasoline-powered automobiles could achieve an 8 percent or greater fuel efficiency gain through a new combustion strategy developed at Oak Ridge National Laboratory. Scientists have demonstrated a new method for reforming fuel over a catalyst, a process that chemically converts fuel into a hydrogen-rich blend. This blend allows more work to be extracted from the engine cylinders, increasing efficiency and saving fuel. “Typically, you incur a fuel penalty when reforming fuel,” said ORNL’s Jim Szybist. “We’ve created a systematic approach that addresses that issue and can be used with conventional fuels and conventional emissions controls.” The team published the method in Energy & Fuels and is working at ORNL’s National Transportation Research Center to demonstrate similar fuel savings at a wider range of engine operation. [Contact: Kim Askey, (865) 576-2841


    Efficiency improvements in spark-ignited (SI) engines are desired as a way to meet increasingly challenging fuel economy and CO2 emission regulations.(1) External cooled exhaust gas recirculation (EGR) provides known thermodynamic benefits while maintaining compatibility with conventional three-way catalysts for emissions control.(2) Caton(3) summarized the thermodynamic benefits of EGR, which include reduced pumping work at partial-load conditions, decreased heat transfer due to lower cylinder temperature, and increased ratio of specific heats. External EGR is also a proven way to decrease the knocking propensity for a given fuel, which can be used as the basis for additional increases in efficiency through more advanced combustion phasing or higher compression ratio.(4) Lastly, EGR reduces NOx emissions over a broad range of speed and load conditions.(5)
    The amount of EGR dilution that can be used is limited due to cycle-to-cycle combustion instability, thereby limiting the potential efficiency benefit of EGR.(6-8) The root cause of the cyclic instability with EGR has been linked to a decrease in flame speed, which elongates the initial flame kernel development process, making it more susceptible to stochastic turbulence variation.(9-11) While several technologies unrelated to fuel reforming are being developed to extend the EGR dilution limit, including high-energy long-duration ignition systems(12) and incorporation of different higher turbulence combustion chamber flows,(13) the focus of this work is to extend the EGR dilution limit by using the reformed products of the fuel to increase the EGR dilution tolerance.
    The addition of the major products of fuel reforming, H2 in particular, increases the combustion rate and can increase engine efficiency because of the shorter combustion duration.(14-20) Alger et al.(21) reported that H2 can also be used to extend the EGR dilution limit, with the addition of 1 vol % H2 extending the EGR limit from 25% to more than 50% for gasoline and from 20% to 28% for compressed natural gas. Fennell et al.(22) showed that simulated H2-rich reformate could extend EGR dilution in an SI engine from 21% to 27% at the same combustion stability. Ivanič et al.(23) added reformate at levels of 15% and 30% gasoline energy equivalent in a single cylinder engine and found that, under partial-load conditions, lean dilution can improve engine efficiency by as much as 12% while EGR dilution delivers 8% improvement.
    Although there are many approaches to generating reformate on board, they can generally be classified into two broad categories. The first category is where fuel is reformed in one or more cylinders in an engine using noncatalytic processes. This category includes the Dedicated EGR (D-EGR) strategy developed by Southwest Research Institute, which is the most developed strategy in this category.(24) D-EGR uses fuel-rich combustion in one cylinder and recirculates its exhaust to the intake system, generating brake thermal efficiency as high as 42.5%(25) and demonstrating a vehicle-level fuel consumption decrease of more than 10%.(26) This category also includes injecting fuel during the negative valve overlap period for a homogeneous charge compression ignition engine to manipulate the fuel–air mixture autoignition propensity.(27-32) The second category of fuel reforming, and the category of the present work, is where a catalyst is used to reform the fuel outside of the engine cylinders(19, 22, 33-40)
    Catalytic reforming processes often use the sensible enthalpy in the exhaust to promote reforming reactions over the catalyst to improve the energetics of reforming.(18, 19, 33-36, 31-43) The range of possible combustion and reforming reactions that can occur has been presented by Ahmed and Krumpelt(44) and by Jamal and Wyszynski,(18) and is presented below for completeness using iso-octane as the starting fuel. Equation 1 shows the reaction stoichiometry for the complete combustion of iso-octane, while eq 2 defines the equivalence ratio (Φ). For the purposes of this manuscript, Φ refers to the O2/fuel ratio of complete combustion in eq 1, relative to the actual O2/fuel ratio. Equations 35 show the primary types of reactions that can occur in a reformer. When O2 is present such that the molar O/C ratio is unity, partial oxidation (POx) occurs where the parent hydrocarbon is converted to CO and H2, resulting in an exothermic process, as shown in eq 3. For an environment devoid of O2 but where H2O is present, the resultant steam reforming reaction, shown in eq 4, is highly endothermic and results in much higher concentrations of H2. Similarly, eq 5 shows a dry reforming reaction where CO2 is consumed to form H2 and CO in a process that is substantially more endothermic than steam reforming, but where the concentrations of H2 are lower. Ideally, the endothermic steam reforming reactions are driven by the sensible enthalpy in the exhaust, allowing for waste heat recovery through thermochemical recuperation (TCR).(45) In this work, we studied reforming environments that contain an insufficient amount of O2 to convert all of the fuel to CO and H2 through eq 3 and thus consist of a balance of POx, steam, and dry reforming processes.

    Complete Combustion:


    equivalence ratio:


    POx reforming:


    steam reforming:


    dry reforming:


    The work by Leung et al.(38) showed that, for ethanol, the steam reforming reaction (eq 4) is very active at engine exhaust temperatures, converting almost 100% of the fuel energy at 600 °C. In contrast, dry reforming (eq 5) produces <40% fuel conversion at the same temperature. This indicates that dry reforming is not the favored reaction pathway, likely due to the larger energy input required, especially at lower initial temperatures. Regardless of whether reforming occurs via the steam or dry reforming pathway, however, the product concentrations can equilibrate via the water–gas shift (WGS) reaction, shown in eq 6.



    It can be challenging to operate the engine and catalyst systems together, such that good performance of the catalytic reformer and engine are achieved simultaneously. Hwang et al.(36)integrated a reforming catalyst containing Rh and Pt into the exhaust manifold of a diesel engine. While they were able to produce high concentrations of H2 (>10%), the reforming system resulted in an overall engine efficiency decrease, because the reformer catalyst equivalence ratio was too low (Φcatalyst as low as 1.5), and the reforming process was too dependent on the exothermic reactions in eq 3. This is consistent with much of the diesel reforming work where high concentrations of H2 can be produced, but because of the dependency on POx reforming, there is a large fuel energy penalty to produce the H2.(46, 47) In contrast, Ashida et al.(33) used a steam reforming catalyst (4 wt % Rh/Al2O3, La additive) in the EGR loop of an SI engine without any additional O2 (Φcatalyst approaching ∞). They initially found high levels of fuel conversion and H2production, but the steam reforming catalyst began deactivating immediately, with 35 ppm of S fuel resulting in a 90% deactivation within 5 h.
    In the companion paper (Part 1, “Catalyst Performance”), we reported on the reformer catalyst performance over a range of Φcatalyst conditions in an effort to find a balance between the POx activity required for catalyst durability and the endothermic steam reforming reactions required to minimize energy losses or to achieve TCR during reforming.(48) In this paper, we characterize the multicylinder engine performance while maintaining stoichiometric exhaust emissions with the EGR-loop reforming strategy, focusing on whole-engine performance, including dilution tolerance, combustion performance, and brake thermal efficiency.


    Experimental Facility

    The engine used in this study was a 2.0 L GM Ecotec LNF SI engine equipped with the production side-mounted direct injection fueling system. Engine geometry details are presented in Table 1. The combustion chamber geometry and camshaft profiles were unchanged from the stock configuration. All engine experiments presented in this manuscript were conducted with primary reference fuel iso-octane obtained from Haltermann, as shown in Table 2. The companion paper presented a detailed characterization of the catalyst performance and equilibrium energetics,(48)while this paper focuses on performance of a multicylinder engine.
    Table 1. Engine Geometry
    bore × stroke86.0 mm × 86.0 mm
    conrod length145.5 mm
    wrist pin offset toward expansion stroke0.8 mm
    compression ratio9.2:1
    fuel injection systemdirect injection, side-mounted, production injector with opposite linear wall-directed six-hole spray pattern
    Table 2. Fuel Properties
    RON (ASTM D2699)100
    MON (ASTM D2700)100
    S (RON – MON)0
    boiling point99.3
    C84.2 wt %
    H15.8 wt %
    O0 wt %
    lower heating value, LHVa44.3 MJ/kg
    We used a laboratory fueling system for the engine in this study. A lift pump delivered fuel to the engine at a pressure of 5 bar through a Coriolis effect flow meter. A cam-driven fuel pump was used to maintain a constant rail pressure of 100 bar throughout the experiment, and a Drivven engine controller with the Combustion Analysis Toolkit package was used to control the engine and acquire crank angle (CA)-resolved data. For each experimental condition, cylinder pressure, spark discharge, and camshaft position data were recorded at 0.2°CA resolution for 1000 sequentially fired cycles. Cylinder pressure was measured using a flush-mounted piezoelectric pressure transducer from Kistler (Model 6125C), and camshaft position was obtained from the production Hall-effect sensors. Fuel injection timing started during the intake stroke and was held constant at 280°CA before firing top dead center (TDC), and the start of the cylinder 4 post-injection event (see Discussion below) was held constant at 150°CA after TDC (ATDC). The engine coolant temperature was maintained at 90 °C. Concentrations of water, CO2, CO, CH4, and iso-octane were performed using a MultiGas 2030HS Fourier transform infrared (FTIR) spectrometry instrument (MKS Instruments). The FTIR was calibrated to detect up to 20% water using an 8-point calibration and 23% CO2 using a 14-point calibration. Two different calibrations for CO were, with the low detection range being a 5000 ppm, 15-point calibration, and the high detection range being an 8%, 10-point calibration. The calibrated ranges for CH4 was 3000 ppm, using a 14-point calibration, and for iso-octane, a 1000 ppm range using a 5-point calibration was used. Hydrogen was measured using an ODYSSEY magnetic sector mass spectrometer (VTI Instruments). A gas divider was used to verify that there was a linear response up to 10% H2, thereafter a single-point calibration of 10% was used for daily calibration. Smoke emissions were monitored with a reflective filter smoke number instrument, but as all smoke measurements were negligible (filter smoke number <0.01), further discussion of smoke emissions was omitted from this manuscript.
    We used two different configurations for the engine: a conventional EGR configuration and a catalytic EGR-loop reforming strategy. All data were collected at an engine speed of 2000 rpm and a brake mean effective pressure (BMEP) load of 4 bar. Furthermore, all data were collected for CA50 combustion phasing, or the timing of 50% heat release, between 6 and 10 crank angle degrees (CAD) ATDC firing (ATDCf) in each cylinder, which typically corresponds to maximum brake torque for most engine operating conditions.
    For the conventional EGR condition, which is shown in Figure 1a, the engine was equipped with an external cooled EGR loop with the flow rate controlled by an electromechanical valve. Cooled exhaust gas was mixed with fresh air upstream of an air heater. The recirculated exhaust gas in this study was not treated with an exhaust catalyst before being recirculated to the intake. The EGR mixed with the fresh incoming air in an intake plenum that had a volume of 57 L. This arrangement allowed the intake manifold to be held at a constant temperature of 52 °C, regardless of the EGR level. The outlet of the intake plenum passed through a static mixer prior to entering the intake manifold. EGR was measured at the outlet of the static mixer using nonintrusive pressure-compensated wideband oxygen sensors in both the intake and exhaust. This method has been reported on previously(49) and is similar to that used in other investigations using conventional EGR.(11)

    Figure 1. Schematic of the engine and test cell configuration for (a) the conventional EGR configuration and (b) the catalytic exhaust gas recirculation (EGR)–loop reforming strategy.

    The catalytic reforming strategy, which is shown in Figure 1b, presents an alternative engine operating strategy that has not been reported previously. A single isolated cylinder feeds the reforming catalyst prior to passing through the EGR cooler and entering the same intake plenum and static mixer arrangement used in the conventional EGR case, thereby providing the entirety of the reformate-EGR mixture to the other three cylinders. The isolated cylinder, which will be referred to as cylinder 4, does not have the ability to breathe the EGR–reformate mixture as the other cylinders do. However, because all of the cylinder 4 exhaust is used to feed the reforming catalyst and is recirculated to the other three cylinders, it is not necessary to maintain stoichiometric conditions in cylinder 4 to have a stoichiometric exhaust. In fact, to provide the catalyst with the proper feed for good performance, cylinder 4 is intentionally operated under nonstoichiometric conditions.(50)
    To achieve high fuel conversion and robust operation of the catalyst, we operated cylinder 4 so that it provided the catalyst with a fuel-rich mixture that also contained oxygen, as our previous studies indicated was necessary.(50) This was accomplished by operating the main combustion event in cylinder 4 fuel-lean and providing a post-injection of fuel at a constant timing of 150°CA ATDC, which is near the exhaust valve opening timing. A schematic of the cylinder pressure and fuel injection schedule for this strategy is shown in Figure 2. The main injection event duration was held constant throughout this study, delivering 16 mg/injection event.

    Figure 2. Cylinder pressure and fuel injection strategy for operation with catalyst reforming. A fuel-lean mixture for the main combustion event in cylinder 4, combined with a post-injection of fuel., provides a fuel-rich mixture at the catalyst inlet, with oxygen present.

    Both the level of excess air and the post-injection fuel mass were varied. The cylinder 4 main combustion equivalence ratio (Φcombustion) was studied at four different conditions (Φcombustion = 0.91, 0.83, 0.77, and 0.71), resulting in four ideal oxygen flow rates to the catalyst ( = 6.1, 11.9, 17.5, and 23.3 g/min). It is worth noting that portions of the excess oxygen and post-injection fuel are trapped as exhaust gas residuals in-cylinder and burned in the subsequent combustion cycle, and these fuel and oxygen residuals were not quantified in this study. At each oxygen flow rate, the amount of post-injection fuel was varied such that Φcatalyst was varied from 3.0 to 10.0. The minimum amount of fuel delivered during the post-injection event was 4.8 g/min, or ∼10% of the fuel required by the three remaining cylinders to achieve BMEP = 4.0 bar at 2000 rpm, depending on efficiency. The maximum amount of fuel delivered during the post-injection event was 66.6 g/min, or ∼135% of the fuel required by the remaining three cylinders. Thus, this range of fuel flux over the catalyst, presented graphically in Figure 3, represents a wide operating range. The secondary x-axis in Figure 3 represents the molar ratio between oxygen contained in O2 and carbon in the fuel fed to the reforming catalyst. An O/C ratio of unity represents a stoichiometric mixture for POx reforming (eq 3), while O/C < 1 in the catalyst feed would require some steam reforming (eq 2) to achieve complete conversion of the fuel to CO and H2. While this full range of oxygen flow rates and Φcatalyst was investigated in the companion catalyst mapping investigation, it is noteworthy that only a portion of this operating map provided sufficient combustion stability in cylinders 1–3, as is discussed in the Results section.

    Figure 3. Fuel flow rate through the catalyst, as a function of catalyst oxygen flow rate and Φcatalyst at the catalyst inlet.

    While cylinder 4 was operated over the range of Φcombustion described above, cylinders 1–3 were operated to maintain overall stoichiometric conditions in the exhaust to maintain compatibility with conventional three-way catalyst emissions control systems. As Φcatalyst and oxygen flow rate conditions changed, the amount of fuel energy contained in the EGR–-reformate mixture also changed. Thus, the fueling rate in cylinders 1–3 was adjusted to maintain overall stoichiometric conditions and a constant brake load.

    Results and Discussion

    Effect of Exhaust Gas Recirculation Dilution on Baseline Engine

    To assess the effects of reformate on the combustion process, relative to conventional EGR, baseline experiments using conventional EGR were conducted with the same experimental engine. Figure 4 illustrates how the heat release rate changes with increasing amounts of EGR. Note that the CA50 combustion phasing was held almost constant at 8 ± 2 CAD ATDCf, necessitating that the spark timing be advanced as EGR was added. The peak heat release rate decreases by more than 35% at the highest EGR concentration, and the combustion duration increases correspondingly. The increase in the combustion duration with EGR is shown in Figure 5, where the timing of the spark, 5% mass fraction burned (CA05), 50% mass fraction burned (CA50), and 85% mass fraction burned (CA85) are indicated. While most portions of the combustion process elongate with increasing EGR, the interval between the spark and CA05 increases the most, which is consistent with previous findings.(11) Figure 6 shows the spark-to-CA05 duration and the coefficient of variation (COV) of the indicated mean effective pressure (IMEP), which is a metric of combustion stability. Combustion is stable for all EGR concentrations below 25% (COV < 3%), corresponding to spark-to-CA05 durations shorter than 40 CAD. Combustion stability decreases dramatically at higher EGR concentrations, where spark-to-CA05 durations are longer than 40 CAD. The increased instability is a direct result of the reduced laminar flame speed with EGR, causing the initial flame kernel growth to be longer and making the process more susceptible to variations in stochastic turbulence.(11, 51)

    Figure 4. Baseline heat-release rate as a function of crank angle with different exhaust gas recirculation (EGR) rates. (HRR = heat release rate; CAD = crank angle degrees.)



    Figure 5. Baseline timing of the combustion process as exhaust gas recirculation (EGR) is increased. Note that CA50 (crank angle at which 50% of the mass fraction has burned) combustion phasing is held almost constant.



    Figure 6. Baseline combustion stability as indicated by the COV of IMEP and spark-to-CA05 duration as a function of EGR.

    The brake thermal efficiency with conventional EGR is shown in Figure 7. Initially efficiency increases with EGR, but it reaches a maximum value and begins to decline due to increased combustion duration and decreased combustion stability. This is consistent with previous studies on the subject.(4, 10, 49) In the following sections, highly dilute combustion with reformate is compared to this baseline performance trend with conventionally cooled EGR.

    Figure 7. Baseline engine brake efficiency as a function of exhaust gas recirculation (EGR) rates.

    Catalytic Reforming Operating Limits and Exhaust Gas Recirculation Estimation

    In the companion paper, we mapped the reforming catalyst performance over a wide range of oxygen flow rates and Φcatalyst.(48) That range of operating conditions was reduced during multicylinder engine operation because the engine was only operated during relatively stable conditions. The engine became unstable when H2 production from the reforming catalyst decreased, likely due to reduced flame speed. Figure 8 compares the operable range of oxygen flow rates and Φcatalyst from the catalyst mapping to the full engine experiments. Note that the catalyst-out concentrations of H2 reported in the companion paper(48) translate to H2concentrations in the intake manifold of up to 5% as the reformate is mixed with incoming air.

    Figure 8. H2 concentration as a function of oxygen catalyst flow and Φcatalyst at the catalyst inlet for (a) catalyst-out concentration from the catalyst mapping study(48)and (b) intake manifold concentration for the multicylinder experiments.

    For catalytic reforming conditions, the reformate stream contains a significant portion of the fuel energy for the combustion event, and therefore it is not exhaust gas per se. However, the effective EGR rate from a pumping perspective can be estimated. Figure 9 shows a contour plot of the intake manifold pressure for the catalytic fuel reforming strategy. The intake manifold pressure for this operating condition (2000 rpm, 4 bar BMEP) varies between 65 kPa and 70 kPa, with the highest intake manifold pressures corresponding to high oxygen catalyst flow and being coincident with the highest levels of H2 formation in Figure 8. For comparison, Figure 10 shows that intake manifold pressure increases linearly with conventional EGR from 48 kPa to 60 kPa. By extrapolating this trend to the manifold pressures observed for the catalytic reforming strategy, it can be determined that the catalytic reforming strategy is operating at an equivalent of 45%–55% EGR, from a volumetric perspective.

    Figure 9. Intake manifold pressure for cylinders 1–3 for the catalytic reforming strategy, as a function of oxygen catalyst flow and Φcatalyst at the catalyst inlet.



    Figure 10. Intake manifold pressure, as a function of conventional exhaust gas recirculation (EGR) operation.

    The manifold pressure determines the thermodynamic expense of moving air into and out of the cylinder, which is quantified as the pumping mean effective pressure (PMEP). PMEP, shown in Figure 11 as a function of the manifold pressure for both the conventional EGR and reforming strategies, demonstrates an almost linear dependence on intake manifold pressure, regardless of operating strategy. Relative to the baseline operating condition without EGR (PMEP = −60 kPa), the catalytic reforming strategy can reduce the pumping work by more than 40% [PMEP = (−35 ± 3) kPa].

    Figure 11. Pumping mean effective pressure (PMEP) as a function of intake manifold pressure for conventional exhaust gas recirculation (EGR) and catalytic reforming.

    Cylinder 4, which is the engine cylinder that is feeding the reformer, is operating with an isolated intake manifold and can be at a different pressure than the manifold for cylinders 1–3. This cylinder operates fuel-lean in order to provide the required O2 to the reforming catalyst. Figure 12compares the intake manifold pressure of cylinder 4 versus the intake manifold pressure of cylinders 1–3. The intake manifold pressure of cylinders 1–3 falls in a narrow range between 66 kPa and 69 kPa, while the intake manifold pressure of cylinder 4 varies over a much larger range, from 63 kPa to 81 kPa. Thus, the intake pressure for cylinder 4 can be higher or lower than cylinders 1–3. The intake pressure of cylinder 4 is dependent on  to the catalyst, which is controlled by how lean the engine operates.

    Figure 12. Intake manifold pressure for cylinder 4 versus the intake manifold pressure for cylinders 1–3.

    Combustion Duration and Stability Analysis

    Figure 13 shows the spark-to-CA05 duration as a function of oxygen catalyst flow and Φcatalyst for the catalytic reforming strategy. The spark-to-CA05 duration is representative of the early flame kernel development time, and it was shown in Figure 5 that this combustion interval exhibits the largest increase with the presence of conventional EGR. The spark-to-CA05 duration for the catalytic reforming strategy is less than 30 CAD for a significant portion of the oxygen catalyst flow and Φcatalyst domain investigated, including a number of operating points where the spark-to-CA05 duration is less than 25 CAD. It can be seen in Figure 5 that, for the conventional EGR cases, combustion became unstable (COV of IMEP > 3%) when the spark-to-CA05 duration was greater than 40 CAD

    Figure 13. Spark-to-CA05 duration as a function of oxygen catalyst flow and Φcatalyst at the catalyst intake for reforming conditions.

    Figure 14 shows the COV of IMEP as a function of the spark-to-CA05 duration for both the catalytic reforming strategy and the conventional EGR strategy. It can be seen that, with the exception of two outliers, the conventional EGR and the catalytic reforming strategy show the same trend: combustion instability begins when the spark-to-CA05 duration is greater than 40 CAD at this speed and loading condition. This finding confirms that the mechanism by which reformate extends the dilution limit is by accelerating the initial portion of combustion through a faster flame speed.

    Figure 14. Combustion stability as quantified by the coefficient of variation (COV) of the indicated mean effective pressure (IMEP), as a function of the spark-to-CA05 duration for both conventional exhaust gas recirculation (EGR) and the catalytic reforming strategy.

    Brake Efficiency

    A plot of brake efficiency, as a function of O2 catalyst flow and Φcatalyst, is shown in Figure 15 for the catalytic reforming strategy. At this engine operating condition, 2000 rpm and 4 bar BMEP with stoichiometric exhaust, the maximum brake efficiency was as high as 32.1%. The peak efficiency occurs for an oxygen flow rate of 12 g/min and Φcatalyst = 6. This condition is a lower oxygen flow rate than where the minimum spark-to-CA05 (Figure 13) and the maximum H2 concentration (Figure 8) occur. However, in our companion study, where the equilibrium energetics of reforming were examined,(48) it was found that a lower concentration of oxygen led to a more favorable energy balance so long as there was sufficient enthalpy to drive the equilibrium products to CO and H2. If there was insufficient enthalpy, CH4 was a higher concentration equilibrium product, reducing the favorability of the energy balance.

    Figure 15. Brake thermal efficiency, as a function of oxygen catalyst flow and Φcatalyst at the catalyst inlet.

    Figure 16 shows brake thermal efficiency for both conventional EGR and the catalytic reforming strategy as a function of the intake manifold pressure. Relative to the baseline engine condition without EGR, the brake thermal efficiency increased from 29.5% to 32.1%, which is equivalent to a fuel consumption decrease of 8.3% (brake-specific fuel consumption decrease from 276.1 g/kWh to 253.1 g/kWh). It is noteworthy that all of the parametric conditions investigated with the reforming strategy have a better brake thermal efficiency than the baseline operating condition without EGR. This is due, in large part, to the reduction of pumping work at such a high dilution level at this part-load engine condition, as shown in Figure 11. This efficiency improvement demonstrates that extending the dilution limit through catalytic reforming can produce a substantial increase in the brake efficiency of a multicylinder engine.

    Figure 16. Brake thermal efficiency as a function of intake manifold pressure for conventional exhaust gas recirculation (EGR) and the catalytic reforming strategy.


    In this investigation we demonstrated significant progress toward increasing the efficiency of stoichiometric SI engines using a catalytic EGR-loop reforming strategy. The development of this combustion strategy is at a very early stage, and as a result, there are a number of additional undeveloped benefits that could produce further efficiency improvements, and there are also a number of unresolved challenges and barriers to its use in vehicle applications.
    Undeveloped Benefits

    The efficiency increase observed here was achieved without any optimization of the engine geometry or of the catalyst, and there is reason to believe that significant optimization is possible. EGR has been shown to be an effective knock suppressant(4) that could allow a substantial increase in the engine compression ratio, which could increase engine efficiency. While increasing the compression ratio benefits efficiency by extracting more work from the combustion products, it also reduces the temperature in the exhaust, thereby exacerbating a process that is already enthalpy-limited. The tradeoff between these two factors has yet to be explored.
    This engine uses a square combustion chamber geometry (stroke = bore), whereas newer engines that are highly tolerant of EGR dilution are under-square (stroke > bore). An under-square geometry increases the mean piston speed and, by extension, the in-cylinder turbulence, which scales with mean piston speed.(52) Higher turbulence shortens the duration of the early flame kernel development by transitioning to turbulent combustion sooner, thereby increasing combustion stability, and shortens combustion duration, which increases efficiency.
    Finally, no attempt was made to optimize the fuel injection timing or targeting, or more broadly, optimize the overall catalyst fueling strategy. This engine used a side-mounted fuel injector, making fuel spray impingement on the cylinder wall during the post-injection probable. A centrally mounted injector and specifically designed fuel injector targeting for these conditions would likely provide better control. Whether the post-injection fueling strategy is preferable to direct fueling to the reforming catalyst in the exhaust system has not been studied and may provide improved performance and/or improved control.
    Barriers to Implementation

    In this engine, as with any combustion strategy using high levels of EGR, the transient performance of the engine poses a difficult control challenge. This is because the reforming and EGR loop represents a substantial volume, making it difficult to rapidly change the intake manifold pressure required when varying load to follow transients in SI engines. The EGR-loop reforming system provides further transient complications because Figures 9–11 showed that achieving a steady-state temperature in the catalyst took several minutes, which is significantly longer than engine transient operation. Intake system optimization to minimize the volume combined with calibration optimization can provide improvements to the transient response, but also has risks. A smaller volume may lead to poor mixing of the reformate both spatially and temporally. Hybrid electric vehicles may not require the engine performance to be as transient-capable as conventional powertrain vehicles, thus hybridization may be synergistic with this technology.
    In addition, the current investigation focused on a single speed/load operating condition. It is not yet known whether there are limitations of the reforming process itself that prevent it from being applicable over the entire operating strategy. From idle to full power, the mass flow rate of a typical engine varies by more than a factor of 50. This will cause a significant variation in space velocity and enthalpy flux, as well as catalyst temperature. Additionally, meeting the speed/load demands of a modern vehicle likely requires operating under boosted intake manifold pressure, which further complicates the engine system.
    Finally, the long-term durability of the reforming catalyst, as well as the reforming response to a variety of fuels that a vehicle will encounter has not been investigated. This investigation used iso-octane, which is a paraffinic fuel that contains no sulfur. Marketplace gasoline will have some level of sulfur and aromatics, and the real-world H/C ratio of the fuel will change. Whether the EGR-loop reforming strategy can perform consistently has not yet been determined.

    Summary and Conclusions

    The engine experiments described here demonstrate the potential benefits of EGR-loop reforming for combustion stability and brake efficiency. Specifically, we found that the catalytic reforming strategy could produce intake manifold H2 concentrations as high as 5% at the 2000 rpm and 4 bar BMEP condition investigated. The reforming strategy also introduced a high level of dilution in the engine, equivalent to 45%–55% EGR. However, despite the high level of dilution, the strategy resulted in good combustion stability. A combined analysis of both the conventional EGR and the reformate-assisted combustion shows that combustion stability problems occur when the spark-to-CA05 duration exceeds 40 °CA. With the high flame speed components in the reformate, namely, H2, this threshold was not exceeded for large parts of the reforming conditions investigated.
    The brake efficiency for our multicylinder engine was increased substantially with EGR-loop reforming, thereby decreasing fuel consumption by more than 8%. This efficiency improvement was achieved while maintaining stoichiometric engine exhaust, meaning that this technology should be compatible with conventional three-way emissions control technology for gasoline engines. We expect that additional efficiency increases should be possible by combining EGR-loop reforming with other engine hardware and operating strategies that enhance the positive impacts of reformate on the combustion profile.

    This manuscript has been authored by UT-Battelle, LLC, under Contract No. DE-AC0500OR22725 with the U.S. Department of Energy. The U.S. government retains and the publisher, by accepting the article for publication, acknowledges that the U.S. government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for U.S. government purposes. The U.S. Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (

    The authors declare no competing financial interest.


    The authors gratefully acknowledge the support of the U.S. Department of Energy Vehicle Technologies Office, particularly program managers Gurpreet Singh and Mike Weismiller. Y.C. was also enrolled as a Ph.D. candidate at the University of Michigan at the time of this publication. She would like to express gratitude for the strong support she received from her coadvisors, Prof. Andre Boehman and Dr. Stani Bohac.

    How To Green The World's Deserts And Reverse Climate Change | TedTalk by Allan Savory
    Sunlight funnel collects light from all directions
    Twitter Auto Publish Powered By :