Potential of advanced combustion for fuel consumption reduction in the light-duty fleet

Transcription

1 Potential of advanced combustion for fuel consumption reduction in the light-duty fleet Paul Miles Combustion Research Facility Sandia National Laboratories International Summit on Breakout Technologies of Engine and Fuel (ISEF2018) Tianjin, China August 20-23,

2 Overview Is there potential for a significant reduction in fuel consumption from IC engine powered vehicles? Can we project historical trends into the future? What will the estimated impact on petroleum displacement or GHG emissions be?? Source: ORNL What are the knowledge gaps/barriers that are preventing us from reaching this potential?

3 Our emphasis will be on the light-duty fleet Projected 46% of total transportation energy use in 2050 (EIA AEO 2018)

4 Baseline definition 2015 fleet average fuel consumption (EPA Combined) 24.6 mpg ( Unadjusted ~31.4 mpg) Car 57.4%: 28.2 mpg (~36.5 mpg = 6.44 L/100 km) 1620 kg, 150 kw, ~7.7 s 0-60 mph (Malibu, Fusion, Camry) cf adjusted ~ 25.7 mpg (32.6 mpg) Truck 42.6%: 21.1 mpg (~26.5 mpg = 8.88 L/100 km) 2130 kg, 200 kw, ~7.7 s 0-60 mph (F150, Odyssey, Colorado) cf adjusted ~ 18.8 mpg (23.4 mpg)

5 Best-in-class Best-in-class fuel consumption Car Mazda 6: 32 mpg (43.4 mpg) (1470 kg, 140 kw, ~7.5 s 0-60) 16% fuel consumption reduction from baseline fleet Truck Ford F-150 : 22 mpg (28.5 mpg) (2060 kg, 242 kw, <7.5 s 0-60) 7% fuel consumption reduction from baseline fleet

6 Projected FC reductions Argonne National Laboratory (2016) 2045 FC reduced ~ 17% relative to 2015 baseline (Combustion alone; based on part-load η and adj. via 2010/2015 fleet avgs) USDRIVE (2018) 2025 FC reduced ~ 11% relative to 2015 baseline (Downsized-boosted tech.; 2025 stretch goals ref. to 2010 baseline; adj. via 2010/2015 fleet avgs) 6

7 Projected FC reductions EPA/CARB/NHTSA (2016) 2025 FC reduced 19% relative to PFI + VVT baseline NRC (2015) 2025 FC reduced 18% relative to 2016 vehicle, +5% thereafter (2030)

8 What is the potential of advanced technologies? 2-Cycle Fuel Economy Improvement (relative to 2015 baseline) Best-in-class car 16% Best-in-class truck 7% Adv. comb. car 27% Adv. comb. truck 21% Lab Car 31% Lab Truck 25% Lab Car w/fuel 34% Lab Truck w/fuel 29% BSFC [g/kwh] Co-optimized with fuel Engine displacement scaled to 120 kw Torque [N-m] 1500 rpm Fuel economy improvements of 34% for cars and 29% for trucks over the 2015 baseline are conceivable with combustion improvement alone

9 Several additional technologies can be brought to bear Friction reduction (lubricants and mechanical design) Cylinder deactivation Accessory electrification Transmissions Low friction brakes Hybridization ~ 30% additional FC reduction Gasoline conventional midsize vehicles consume from 20% to 44% less fuel by 2045 compared with the 2015 reference (adj. via 2010/2015 fleet avgs) 2018 Camry 32 mpg 52 mpg A 39% reduction 9

10 researchimpactnetwork.wordpress.com Impact

11 Combustion improvement will play an important role in reducing 2050 fleet petroleum usage Petroleum consumption reduction targets due to combustion enhancement alone: 2015 fleet average baseline Car ~ 34% Truck ~ 29% Estimated per vehicle 2050 Fleet average energy consumption reduction, using very optimistic BEV market penetration (35%): (0.32)* 65% = 21% With additional ICE technology: (0.50)* 65% = 33% Does not include impacts of renewable fuels 11

12 How great is the CO 2 emission reduction advantage for BEVs? Like-utility vehicle comparison 300 mile range Full vehicle lifetime Full LCA accounting for manufacture, maintenance/repair and disposal Proper accounting for carbon intensity of the grid Account for taxes Allow for technology improvement Allow for expected CO 2 reductions resulting from the use of mandated biofuels

13 A closer look at CO 2 emissions from the electric grid Source: Energy Policy 44: (2012) In this example (UK ) marginal CO 2 emissions are 54% higher than the average EPA egrid 2016 puts US non-baseload emissions 50% higher than the national grid average

14 Comparing similar-utility vehicles and using marginal grid CO 2 emissions indicates current ICEs are superior Best-in-class vs. BEV300 (Marginal CO 2 in-use): Best-in-class hybrid vs. BEV300 (Marginal CO 2 in-use): Until we see significant changes to the grid, introduction of PHEVs/BEVs is likely to increase GHG emissions over even current fleet average vehicle emissions

15 ec.europa.eu Barriers (Light-duty)

16 Where should we invest our research effort? Several manufacturers and government agencies have identified multi-mode combustion as a likely future trend (Mazda, GM, US EPA ) Hybridization and cylinder deactivation will de-emphasize low-load region Research should emphasize strategies to increase mid-to-high load efficiency HCCI rpm Lean SI Stoichiometric SI

17 Both fuels and engine architectures are constrained for multi-mode combustion Fuels need to allow: o Stoichiometric operation at full load (Requires autoignition resistant fuel) o Low-temperature, lean combustion at low-to-moderate loads (Benefits from more reactive fuel) Engines need to support: o High-tumble, fast-burn SI combustion o Residual gas retention, low-load boost o Controlled, precise mixture formation o High load ignition & low load combustion timing control Cold-start/re-start emissions exacerbated by hybridization Source: GM, UW Symposium June

18 Stability and combustion phasing control are significant barriers to lean combustion Stabilization of lean, stratified SI combustion o Key flow features causing cyclic variability o Understanding of stabilizing flow/spray interactions (effects of multiple injections) Spark-assisted CI phasing control techniques for HCCI-like combustion o Fuel effects o Fundamental understanding of impact of boost/tin/egr etc., interactions with φ CI combustion phasing control using mixture stratification synergies with SACI

19 Enhanced ignition systems will be an enabler for lean combustion technologies Plasma-assisted ignition systems have potential to support multiple combustion strategies using the same hardware o More robust ignition for ultra-lean or high-egr SI operation o Improved tolerance to high flow velocities o Control and/or reduced intake temperature requirements for inherently low NO x /PM HCCIlike strategies Turbulent jet ignition systems have a similar broad application range, improving both lean and stoichiometric combustion Combustion phasing control and intake temperature requirements can be significantly impacted by O 3

20 Improved emissions will continue to be key Particulate/PAH emission control will be key to future ICE acceptance o GPFs increase residuals and back-pressure, impacting FE through knock & pumping losses o PM dominated by cold-start o Current modeling tools and understanding of fuel effects inadequate EU limit for benzopyrene is 1 ng/m3 Source: J. Ind. Eco. 17: (2013) IC engines are worse than BEVs for photo-chemical oxidation potential Reduced engine-out NOx, HC, and HCO emissions will be essential o Improved, low-temperature lean NO x aftertreatment technologies

21 Sprays and mixture formation are foundational to all high-efficiency, low-emissions combustion strategies SI boosted engine knock & emissions sensitive to mixture formation Sprays are key to cold-start soot/hc emissions DI stratified fuel consumption and COV highly sensitive to spray structure Source: Scomak AMR 2017 Advanced strategies(mazda s SPCCI) rely on precise control of mixture formation Fuel 183: (2016) Improved modeling capabilities will be essential especially multi-phase (enabled by ECN) Sub-cooled Collapse

22 In-cylinder flow control also crosscuts across multiple technologies Key to fast-burn (high tumble) combustion systems need to understand impact of VVT/VVL o Atkinson/Miller cycle implementations o Exhaust gas retention / re-breathing Reduced cyclic variability o More robust ignition and flame kernel growth in lean/dilute systems o Mixture formation variability in stratified systems Optimal scavenging and mixing of residuals for improved: o Knock control in boosted SI engines o Auto-ignition control in HCCI-like system Source: Daimler, THIESEL 2016

23 ec.europa.eu Barriers (Heavy- and Medium-duty)

24 Mixing is key for diesel efficiency improvement Shortening/advancing the heat release rate is the mainstream strategy being pursued by OEMs. Implementation barriers include: o Cavitation erosion issues o Maintaining high late-cycle mixing rates (bowl shape effects) o Optimized multiple injection approaches [Understanding both physical & chemical interactions between injection events] o Maintaining or reducing NO x emissions [Complex aftertreatment systems are expected to incur a ~2% FE penalty] Volvo wave piston Source: Dieselnet.com/news/2016/09sae.php

25 Improved models are also key Turbulence-chemistry interactions Predictions of ignition delay change by an order of magnitude SI flame kernel can either extinguish or double growth rate depending on conditions Two-phase flow and atomization models, including cavitation-turbulence interactions Fuel dependent impacts on spray structure, impacting wall-wetting, emissions, and efficiency Improved soot formation models (particle inception) Plasma-assisted ignition

26 Science-based fuel surrogate formulation for Mixing-Controlled Compression Ignition (CRC) Particulate Matter Index works well for conventional fuels Realistic surrogate fuels enable coupling of real fuel properties to numerical engine optimization. How much complexity is required? Approach matches target fuel chemical structure as well as key physical properties (cetane, density, distillation )

27 Hybridization of medium-duty diesels presents additional cold-start challenges US regulations consider the combination of NO x /NMOG Lean (i.e., diesel) combustion prolongs catalyst light off by approximately a factor or 5 compared with stoichiometric SI Fuel effects (higher cetane) have been shown to allow more aggressive catalyst heating strategies Key objective: Understanding fuel effects on cold-start performance

28 A fuel metric that captures sooting propensity in diesel engines is needed The Particulate Matter Index (PMI) has been shown to predict soot Particulate Matter Index works emissions well for GDI engines well for HT: conventional Bob McCormick, NREL fuels nn PPPPPP = ii=1 DDDDDD = 2CC+2 HH 2 Tendency to form soot DDDDDD ii + 1 VVVV(443KK) ii WWWW ii Tendency to evaporate & mix An equivalent metric for mixing-controlled combustion will also need to incorporate a factor that captures ignition delay variation Key objective: Understanding fuel effects on engine-out emissions

29 Large engines with slower transients are good candidates for full-time kinetically controlled combustion Fuels with identical RON and MON can exhibit very different auto-ignition behavior John Dec, SNL CA50 [ CA] Co-Op - E30, Premixed Co-Op - E30, Early-DI Co-Op - Aromatic, Premixed Co-Op - Aromatic, Early-DI P in = 1.0 bar abs. φ = Intake Temperature (T in ) [ C] John Dec, SNL CI combustion modes often use stratification for CA50 control High sensitivity fuels sacrifice LTHR/ITHR needed for control 29

30 Ducted fuel injection concept holds promise for a large improvement in the soot/nox trade-off A simple new technology with potential for improved efficiency and reduced after-treatment costs compared with conventional diesel o Compatible with existing and new, renewable fuels o Soot formation reduced by 2 orders of magnitude Key research questions: What are the optimal fuel properties for DFI? Can oxygenated fuels enable a similar reduction in NOx?

31 Summary Advanced combustion concepts can reduce ICEV fuel consumption by over 30% compared with the 2015 fleet average Combined with additional measures, we see potential to reduce fuel consumption by 50% or more Renewable fuels can further reduce GHGs and petroleum consumption Improved ICEVs will be an essential element of a transportation GHG reduction technology portfolio through at least mid-century Engines are far from a mature technology. Research and development challenges are numerous, for both light- and heavy-duty engines and can provide rewarding careers

32 Questions?