Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

Experimental and numerical work at Ghent University on methanol combustion in SI engines Prof. Sebastian Verhelst Ghent University

1

WHY METHANOL?

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

2

UGent ICE research focus: background

Starts from a vision on long term energy supply and energy carrier for transportation • Long term energy supply: selection of candidates based on

‣ Sustainability, of energy source and of harvesting

technology (includes e.g. recyclability)

‣ Scalability, i.e. abundance of energy source, and of resources needed for building the harvesting technology

• Conclusion: CSP (concentrated solar power) best choice

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

3

UGent ICE research focus: background

• Long term solution for transport? Options for energy carrier and powertrain?

‣ Sustainability: closed cycle for energy carrier and

powertrain materials

‣ Scalability: resources for energy carrier and powertrain

‣ Compact: need sufficient energy & power density

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

4

Conclusions for sustainable transportation

• Energy carrier: need for renewable (solar), liquid fuels ‣ Efficient, so practical and cheap distribution and storage

• Powertrain: internal combustion engine ‣ Cheap to produce:

‧ Easy to produce ‧ From abundantly available, recyclable materials ‧ Relatively little energy needed for production

‣ Fuel flexible ‣ High power density ‣ Still potential for efficiency improvement

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

5

Candidate fuels

• Simple molecules are preferred ‣ Production is more efficient ‣ Conversion (end-use) can be controlled more easily (η, emissions)

• Abundantly available building blocks: C, H, O, N, …

• Thus, most simple fuels: ‣ Hydrogen, H2 (at patm, liquid at 20K) ‣ Methane, CH4 (at patm, liquid at 91K) ‣ Ammonia, NH3 (at Tatm, liquid at 8.6 bar) ‣ Methanol, CH3OH (liquid) ‣ Dimethylether (DME), CH3OCH3 (liquid at 5.3 bar) ‣ …

LIQUIDS

6 Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

Case: methanol

• Can be produced in different ways ‣ Biomass, fossil fuels, synthesize using renewable energy

• Liquid ‣ Cheap tanks, cheap distribution ‣ Miscible with gasoline and ethanol ‣ Evolution of infrastructure possible

• High engine efficiencies possible ‣ High octane number, heat of vaporization, …

• Also building block for synthetic hydrocarbons (MTO) Has been a focus for UGent since 2009

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

7

UGent work on methanol as an engine fuel

PhD Jeroen Vancoillie 2009-2013 [1] • Problem statement: no simulation tools for engines

operating on light alcohols (MeOH/EtOH) ‣ Complexity of engines – optim. solely relying on expts. undoable ‣ Unique fuel properties – tools optimized for gasoline etc. ‣ Special working conditions – e.g. EGR%, charging pressures, …

• Contents ‣ Conversion of engine test benches, measurement database ‣ Fundamental data: laminar burning velocity, autoignition delays ‣ Evaluation of turbulent combustion velocity models ‣ Engine cycle simulations, normal + knocking operation

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

8

UGent work on methanol, cont.

MSc theses, postdoc stay • Extension of measurement database (incl. effects H2O) • Vehicle calculations PhD Louis Sileghem 2011-2015 [2] • Problem statement: introduction of methanol through use

as blend component (cfr. presentation James Turner, Univ. Bath); no simulation tools for engines on alcohol blends ‣ Non-linear behaviour burning velocity, knocking behaviour, …

• Contents: measurements; blending/mixing rules for burning velocities, auto-ignition delays; engine simulations

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

9

EXPERIMENTAL WORK Engine tests

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

10

Properties MeOH ↔ gasoline & implications

Properties: • High heat of vaporization

• High octane number

• High flame speed

Potential • Even for gasoline-optimized

engine max. achievable engine load higher when using alcohols

• Increase in volumetric eff. due to high degree of charge cooling • Spark timing less knock limited due to elevated knock resistance and higher flame speeds

• Options for dedicated engines: • High compression ratio • High EGR ratios • Lean operation

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

Methanol Gasoline

325 1100

(kJ/kg)

95 109

RON

30 45

Laminar burning velocity at NTP (cm/s)

11

Experimental work: flex-fuel

2 converted ( “flex-fuel”) PFI engines [3,4] • With same operating strategy:

‣ Relative efficiency benefits MeOH vs. gasoline order of 10% ‧ More isochoric combustion ‧ Less flow, cooling and dissociation losses

‣ Reductions of engine-out NOx levels of 5–10 g/kWh ‧ Lower combustion temperatures

• Alternative load control strategies ‣ WOT, lean; or WOT+EGR: relative eff. benefits of 5% vs. throttling ‣ Possible due to higher dilution tolerance of MeOH ‣ Also lower NOx, due to further lowering of combustion temperatures

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

12

Experimental work: dedicated methanol [3,4]

VW TDI, CR 19.5:1 • Replaced diesel injectors with spark plugs,

mounted methanol PFI system • Reproduced EPA work (Brusstar et al.)

and expanded to lower loads • Faster, more stable combustion

‣ High CR, turbocharged, high turbulence ‣ More dilution tolerance ‣ Wider range (down to 3 bar BMEP)

• Lower in-cylinder temperatures ‣ Cooling & dissociation losses ↓ ‣ Knock/pre-ignition ↓; Engine-out NOx ↓

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

13

Experimental work: dedicated methanol [3,4]

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

NOx (ppm)

BTE (%)

Unstable combustion CoV > 10% (EGR≈50 m%)

42%

Diesel-like peak BTE Part load efficiency gains up to 20% (compared to throttled operation)

Vast engine-out NOx reductions (ppm)

14

Experimental work: dedicated methanol

• Dedicated engines: diesel-like efficiencies while using cheap aftertreatment systems

• Used engine bench results in vehicle driving cycle calculations [5]

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

26.5 25.6

31.8 31.3

22.0

24.0

26.0

28.0

30.0

32.0

34.0

NEDC FTP75

Effic

ienc

y %

Vehicle 5, MeOH, Throttle Vehicle 6, MeOH, WOT_EGR

1.19

2.20

0.10 0.11

0.00

0.50

1.00

1.50

2.00

2.50

NEDC NOx g/km FTP75 NOx g/mi

NO

x em

issi

ons

Vehicle 5, MeOH, Throttle Vehicle 6, MeOH, WOT_EGR

15

Experimental work: DI engines

• Naturally asp. DI: Hyundai 2.4L at Argonne Nat. Lab [11] ‣ Tested MeOH vs. gasoline, EtOH, BuOH, E85, M56 ‣ Stock ECU – load limitations

(~injection duration; λ=1) ‣ Low-mid load: 2.7 %pt BTE increase

on MeOH vs. gasoline ‣ High load: 5.6 %pt (BTE=40%, i.e. -20% CO2)

• Turbocharged DI: Volvo 1.6L “T3” ‣ MoTeC M1 ECU ‣ In progress…

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

16

FUNDAMENTAL WORK Burning velocity measurements and computations

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

17

Building blocks for engine cycle simulations

• Laminar burning velocity ul ‣ Groups chemical effects on combustion ‣ Function of pressure, temperature, mixture composition ‣ 50% higher than for gasoline faster combustion, higher efficiency

• Lack of reliable ul data for methanol at engine-like conditions ‣ Calculated ul using carefully selected chemical kinetics scheme ‣ Built correlation for use in engine code [6] ‣ Validated correlation

‧ Flat flame burner at Lund University [7] ‧ Spherical flames at Leeds University [1]

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

18

Evaluation of ul correlation + ut models [8]

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

19

Building blocks for engine cycle simulations, cont.

• Autoignition delay time τ ‣ Determines resistance to engine knock ‣ Function of pressure, temperature, mixture comp. ‣ RON 109 higher efficiency: either through

increased CR or wider use of MBT • Again, lack of data

‣ Calculated τ using the same chemical kinetics scheme as for ul

‣ Built correlation for use in engine code [9]

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

20

NUMERICAL WORK ul & τ correlations; engine simulation framework

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

21

Multizone thermodynamic engine model

i.e. “quasi-dimensional” approach • Compromise between accuracy and computational effort • Framework for testing ul & τ correlations

‣ Power cycle: in-house GUEST code (Ghent University Engine Simulation Tool)

‣ Gas dynamics: GT-Power • Findings [10]

‣ “Puddling” model necessary: low volatility ‧ Evaporated fraction low compression slope underestimated ‧ Evaporated fraction high volumetric efficiency overestimated

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

22

Engine cycle simulations

• Evaluation of ul&τ correlations, and ut models, against experimental database obtained on 2 engines with variation of: ‣ throttle position ‣ fuel-air equivalence ratio ‣ spark timing ‣ engine speed ‣ …

• Normal & knocking combustion [10] ‣ knocking: puddling model & in-cylinder

heat transfer to be revisited

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

23

Summary

• UGent contributions on methanol-fueled ICEs ‣ Measurement database on a range of SI engines ‣ Correlations for ul, δl, τ for use in engine simulation ‣ Validation of engine cycle simulations against expt. database

• To be improved ‣ In-cylinder heat transfer models (work in progress at UGent) ‣ Proper modeling of evaporative cooling

• In progress or starting ‣ MeOH in downsized engine ‣ Dual-fuel operation: MeOH in diesel engine

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

24

Outlook

LEANShips: Low Energy And Near to zero emissions Ships • EU Horizon 2020 Mobility for Growth ‘innovation action’ • 48 partners, 8 demonstrator platforms • UGent: WP05 leader “Demonstrating the potential of

methanol as an alternative fuel” (6 partners) ‣ Conversion of high speed marine diesel engine to dual fuel

operation with methanol ‣ LCA of methanol in shipping ‣ Tools for dissemination and market uptake (pilots)

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

25

References

1. Jeroen Vancoillie, Modeling the combustion of light alcohols in spark-ignition engines, PhD Ghent University, 2013 - download link

2. Louis Sileghem, Modeling the combustion of alcohol blends in SI engines, Ghent University, PhD defense July 2015

3. Vancoillie J et al, Experimental evaluation of lean-burn and EGR as load control strategies for methanol engine, SAE 2012-01-1283

4. Vancoillie J et al, The Potential Of Methanol As A Fuel For Flex-Fuel And Dedicated Spark-Ignition Engines, Applied Energy 102:140-149, 2013

5. Naganuma K et al, Drive Cycle Analysis of Load Control Strategies for Methanol Fuelled ICE Vehicle, SAE 2012-01-1606

6. Vancoillie J et al, Laminar burning velocity correlations for methanol-air and ethanol-air mixtures valid at SI engine conditions, SAE 2011-01-0846

7. Vancoillie J et al, The temperature dependence of the laminar burning velocity of methanol flames, Energy&Fuels 26:1557-1564, 2012

8. Vancoillie J et al, The turbulent burning velocity of methanol-air mixtures, Fuel 130:76-91, 2014

9. Vancoillie J et al, Development and Validation of A Knock Prediction Model for Methanol-Fuelled SI Engines, SAE 2013-01-1312

10. Vancoillie J et al, Development and validation of a quasi-dimensional model for methanol and ethanol fueled SI engines, Applied Energy 132:412-425, 2014

11. Sileghem L et al, Experimental investigation of a DISI production engine fuelled with methanol, ethanol, butanol and iso-stoichiometric alcohol blends, SAE 2015-01-0768

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

26

Thank you for your attention!

Sebastian Verhelst sebastian.verhelst@UGent.be http://users.ugent.be/~sverhels

Methanol workshop, Lund – March 17th 2015 Faculty of Engineering and Architecture – Department of Flow, Heat and Combustion Mechanics

27