Tailor-Made Fuels from Biomass - RWTH Aachen University

Tailor-Made Fuels from Biomass Status Update Core Interaction Field „Fuel Design“

TMFB Status Update Core Interaction Field „Fuel Design“

25.06.2015, 3rd TMFB International Conference

S. Pischinger

TMFB Approach: The Vision of Tailor-Made Fuels from Biomass

2 von 52

From Biomass

to Biofuels

From Biofuels

to Propulsion

Model-Based Description of Synthesis and Production Routes

Model-Based Specification of Target Properties

Fuel Design:

Overall optimum of

biofuel production and

biofuel combustion



S. Pischinger

In the CIF, the Fuel Design Process is carried out…

3 von 52



S. Pischinger

…but what does this really mean?

4 von 52

From Biomass

to Biofuels

From Biofuels

to Propulsion

Fuel Design:

Overall optimum of


biofuel combustion

Proposal of fuel

molecules

Screening of Fuel

Properties

Screening of

production feasibility

Proposal of fuel

molecules

Which molecules are

proposed and why?

Which descriptors do we

use to measure the

properties and why?

? ?


use to describe the

production and why?

?

TMFB

Which molecules are

proposed and why?



S. Pischinger

Fuel Design Step by Step! Molecules proposed by

5 von 52

Proposal of fuel

molecules

Screening of

Fuel Properties

Diesel (CI) candidate

Gasoline (SI) candidate

Why furans?:

State-of-the art biofuel candidates

(comparably) easily accessible from lignocellulosic

biomass

2-MF

Good anti-knock behavior (gasoline engine)

Novel pathways were identified (last year)

3-MTHF

Novel molecule (not available anywhere!)

Promising self-ignition behavior Diesel fuel

candidate?

Variation of chain-length

What is the effect of the side chain length?

What is the effect of the position of the side chain?



S. Pischinger

The properties are numerous, it is our goal to bridge the gap with selected property descriptors

6 von 52

Aim: bridge the gap!

Molecule’s features

Thermo-physical

properties

Engine performances

C content

H/C-ratio

O/C-ratio

C saturation

Branching

Density

Viscosity

Surface tension

Vapor pressure

Vaporization enthalpy

Combustion enthalpy

Boiling behavior

Mixture formation

Air entrainment

Evaporation rate

Ignition delay

Combustion

Emissions

Engine

performance

indicators

Feedback to the fuel design process

QSPR*

GCM*

*Quantitative Structure Properties Relationship, Group Contribution Method

[1] B. Graziano et al. SAE Int. J. Fuels Lubr. 8(1):2015, doi:10.4271/2015-01-0890.

Screening of

Fuel Properties



S. Pischinger

Screening of fuel properties: Which phenomena can be characterized by which descriptor?

Screening of

Fuel Properties

Mixture

Formation Emissions Ignition

Physico-chemical

Ignition delay

Chemical

Ignition delay

Spray behavior

and evaporation Smoke points (SP)

Derivate

Cetane

Number

Lubricity

Material

[5]

Wear Scar

Diameter (WSA)

7 von 52

Threshold

Sooting

Index

𝟏

𝑾𝑺𝑫

Ignition

Delay

Time

3D -Computational

Fluid Dynamics (CFD)

Ignition Quality

Tester (IQT)

Advanced

Fuel Ignition Delay

Analyzer (AFIDA)

Shock Tube (ST)

Rapid Compression

Machine (RCM)

Laminar Flow

Reactor (LFR)

ASTM D1322

SP Lamp

High frequency

reciprocating rig

(HFRR)

Oxidation

Potential

Number



S. Pischinger

Results: RCM and Shock tube experiments

9 von 52

CI engines SI engines

[2] R. Tripathi, et al. Gas kinetic symposium 2014, Poster presentation.

[3] Unpublished data from PCFC

[4] U. Burke, et al. in preparation for Combustion and Flame 2015.

[5] L. Cai, et al. Proceedings of the Combustion Institute , 01/2015; 35(1):419-427.

0.01

0.1

1

10

100

0.6 0.8 1 1.2 1.4 1.6

Ign

itio

n D

ela

y t

ime

/ m

s

1000/T / K-1

2Methylfuran RCM 20 bar [2]2MTHF RCM 20 bar [4]2 butanone RCM 20 bar [3]2MTHF ST 20 bar [4]2 butanone ST 20 bar [3]n-Octanol ST 20 bar [1]DnBE ST 20 bar [4]3MTHF RCM 20 bar[4]

[2]

[3]

[4]

[3]

[4]

[5]

[3]

[3]

2-MF is less reactive than

2MTHF due to the influence of

double bond in the ring

structure.

Di-n-butylether is highly reactive

compared to n-octanol due to

the position of the oxygen atom

in the long chain.

2-Butanone is less reactive than

2MF, & 2MTHF and its kinetic

behavior is under study.

TMFB fuels results:

Screening of

Fuel Properties

Ignition



S. Pischinger

Excursus: Ignition delay is not ignition delay

10 von 52

RCM IQT / AFIDA CFR

Charateristic

Number Ignition Delay / ms

Derived Cetane Number

DCN

Cetane Number

CN

Pressure and

Temperature variable

T = 823 K

p = 22 bar variable

Ignition Delay variable variable 2.3 ms

13 °CA

Boundary

Conditions /

Phenomena

Chemical ignition delay Chemical and physical

ignition delay

Chemical and physical

ignition delay

+ in-cylinder flow



S. Pischinger

Results: RCM and Shock tube experiments

9 von 52

CI engines SI engines

[2] R. Tripathi, et al. Gas kinetic symposium 2014, Poster presentation.

[3] Unpublished data from PCFC

[4] U. Burke, et al. in preparation for Combustion and Flame 2015.

[5] L. Cai, et al. Proceedings of the Combustion Institute , 01/2015; 35(1):419-427.

0.01

0.1

1

10

100

0.6 0.8 1 1.2 1.4 1.6

Ign

itio

n D

ela

y t

ime

/ m

s

1000/T / K-1

2Methylfuran RCM 20 bar [2]2MTHF RCM 20 bar [4]2 butanone RCM 20 bar [3]2MTHF ST 20 bar [4]2 butanone ST 20 bar [3]n-Octanol ST 20 bar [1]DnBE ST 20 bar [4]3MTHF RCM 20 bar[4]

[2]

[3]

[4]

[3]

[4]

[5]

[3]

[3]

2-MF is less reactive than

2MTHF due to the influence of

double bond in the ring

structure.

Di-n-butylether is highly reactive

compared to n-octanol due to

the position of the oxygen atom

in the long chain.

2-Butanone is less reactive than

2MF, & 2MTHF and its kinetic

behavior is under study.

TMFB fuels results:

Screening of

Fuel Properties

Ignition

Ignition Delay in the

CFR engine (CN

determination): 2.4 ms

Temperature in the IQT

(823 K)



S. Pischinger

From Biofuels to propulsion: DCN as descriptor for global ignition characteristics

11 von 52

fuel

candidates

for

SI engines

fuel

candidates

for

CI engines

Ignition Quality Tester (IQTTM)

Standardized constant

volume combustion

chamber

T = 823 K

p = 22 bar

Applicable to diesel-

and gasoline-like fuels

Rapid screening,

small fuel amount

(<200 ml fuel, 32

repetitions, <20 mins.)

derived cetane number

(DCN)

5.2 7.0 15.0 100.0

7.6 18.7 30.9 40.3

IQT ignition delay [ms]

MTBE, toluol, furan,

ethylbenzene, …

isooctane, isocetane,

butanol, …

heptane, hexadecane,

dibutyl ether, …

more prone to

auto-ignition

less prone to

auto-ignition

The axes do not scale linearly.

[6] M. Dahmen & W. Marquardt, submitted to Energy & Fuels.

Screening of

Fuel Properties

Ignition



S. Pischinger

From Biofuels to propulsion: DCN as descriptor for global ignition characteristics

12 von 52

Side chains have only minor influence on furans

Similar DCNs within group of furans

2-EF and 2-FFOH have very similar DCN negligible effect of alcohol group in side chain

Side chain length determines ignition behavior of tetrahydrofurans

[7] A. Sudholt et al., PROCI, 35 2957-2965, 2015, doi: 10.1016/j.proci.2014.06.147.

Screening of

Fuel Properties

Ignition

fuel

candidates for

CI engines

fuel

candidates for

SI engines



S. Pischinger

From Biofuels to propulsion: Group Contribution Modeling of IQTTM Ignition Delay Data (1/2)

13 von 52

Fuel Screening & Modeling

Ignition delay data for 161

(oxygenated) hydrocarbons, e.g.,

Model is applicable to CxHyOz

structures except for alkynes,

acids and structures with

consecutive double-bonds

OH

OH

OO

O

O

O

OO

O

O

O

OO

O

OH


Screening of

Fuel Properties

Ignition

-CH3

-CH2-

-CH2-

(ring)

>CH-

(ring)

-O-

(ring)

-CH3

=CH-

(ring)

=CH-

(ring)

>CH=

(ring)

-CH2- -OH

-O-

(ring)

=CH-

(ring)

>CH=

(ring)

-CH3

butyltetrahydrofuran

n-butanol 2-methylfuran

toluene

(Illustrative examples)

Group decomposition based on the set of structural groups

proposed by Joback and Reid, Chem. Eng. Commun., 57, 233–243, 1987.



S. Pischinger

0 20 40 60 80 100 1200

20

40

60

80

100

120

model DCN

measure

d D

CN

training

external validation

From Biofuels to propulsion: Group Contribution Modeling of IQTTM Ignition Delay Data (2/2)

14 von 52

Simple, yet predictive estimator for DCN

of pure oxygenated hydrocarbons

Group Contribution Model

Two-dimensional molecular graph

can be decomposed into

structural groups.

Number and type of groups have

been chosen carefully in the view

of a limited amount of

experimental ignition delay

data.

Low parametric uncertainty and

correlation in DCN model.

Model-Based

Fuel Design

TMFB


Screening of

Fuel Properties

Ignition

Training Set

Mean DCN Error 5.8

Leave-Many-Out Cross-Validation

Mean DCN Error 6.6

External Validation Set

Mean DCN Error 5.8



S. Pischinger

So what do we learn from this?

1st preliminary conclusion

15 von 52

In a first screening step, the DCN (measured or/and calculated) as well as

the pure chemical ignition delay serve to differentiate between a fuel’s

potential for SI- or CI-type combustion

The longer the side chain, the higher the DCN

suitability for CI engines, not suitable for SI engines

But: how high should the DCN (or CN) be for optimum low emission (soot,

HC, CO, noise) combustion?

Screening of

Fuel Properties

Proposal of fuel

molecules



S. Pischinger

What should the fuel look like to enable alternative, soot- and NOx-free combustion?

16 von 52

Ind

icate

d S

pecif

icP

M-e

mis

sio

ns /

mg

/kW

h

0.0

15.0

30.0

45.0

Ind

icate

d S

pecif

icH

C-e

mis

sio

ns /

g/k

Wh

0.0

1.5

3.0

4.5

Ind

icate

d S

pecif

icC

O-e

mis

sio

ns /

g/k

Wh

0

10

20

30

Cetane number / -

20 30 40 50 60 70 80

n-decane iso-octane/

n-heptane blends

Vh = 0.39l, n = 2280 1/min, IMEP = 9.4 bar, ISNOx = 0.4 g/kWh

Lower cetane number

Reduced PM-emissions

Significantly elevated

HC- and CO-emissions

Worsened combustion

controllability

Cold engine start

• How to further reduce soot emissions without the drawbacks of HCCI?



S. Pischinger


17 von 52

Ind

icate

d S

pecif

icP

M-e

mis

sio

ns /

mg

/kW

h

0.0

15.0

30.0

45.0

Ind

icate

d S

pecif

icH

C-e

mis

sio

ns /

g/k

Wh

0.0

1.5

3.0

4.5

Ind

icate

d S

pecif

icC

O-e

mis

sio

ns /

g/k

Wh

0

10

20

30

Cetane number / -

20 30 40 50 60 70 80

iso-octane/ n-heptane blends /

n-decane

EN590 Diesel


Lower cetane number

Lower aromatic content

Less soot precursors

Reduced PM-emissions

HC-, CO- and noise

emissions not affected

• How to further reduce soot emissions by adjusting the fuel characteristics?



S. Pischinger

18 von 52

Ind

icate

d S

pecif

icP

M-e

mis

sio

ns /

mg

/kW

h

0.0

15.0

30.0

45.0

Ind

icate

d S

pecif

icH

C-e

mis

sio

ns /

g/k

Wh

0.0

1.5

3.0

4.5

Ind

icate

d S

pecif

icC

O-e

mis

sio

ns /

g/k

Wh

0

10

20

30

Cetane number / -

20 30 40 50 60 70 80



n-decane

EN590 Diesel

1-decanol n-decane

Lower cetane number

Lower aromatic content

Increased oxygen

content

Inhibitory effect of alcohol

functional group on ignition

At same cetane number,

PM-emissions are reduced

due to locally increased

AFR

But: OH-group lowers the

fuel’s volatility and thus and

worsens air/fuel mixing


• How to increase the molecular oxygen content while maintaining high volatility?

EN590 Diesel n-decane GtL iso-octane/n-heptane blends 1-alcohols di-n-butyl ether

1-alcohols



S. Pischinger


19 von 52

Ind

icate

d S

pecif

icP

M-e

mis

sio

ns /

mg

/kW

h

0.0

15.0

30.0

45.0

Ind

icate

d S

pecif

icH

C-e

mis

sio

ns /

g/k

Wh

0.0

1.5

3.0

4.5

Ind

icate

d S

pecif

icC

O-e

mis

sio

ns /

g/k

Wh

0

10

20

30

Cetane number / -

20 30 40 50 60 70 80



n-decane

EN590 Diesel


1-alcohols

Ind

icate

d S

pecif

icP

M-e

mis

sio

ns /

mg

/kW

h

0.0

15.0

30.0

45.0

Ind

icate

d S

pecif

icH

C-e

mis

sio

ns /

g/k

Wh

0.0

1.5

3.0

4.5

Ind

icate

d S

pecif

icC

O-e

mis

sio

ns /

g/k

Wh

0

10

20

30

Cetane number / -

20 30 40 50 60 70 80 90 100 110 120

Ind

icate

d S

pecif

icP

M-e

mis

sio

ns /

mg

/kW

h

0.0

15.0

30.0

45.0

Ind

icate

d S

pecif

icH

C-e

mis

sio

ns /

g/k

Wh

0.0

1.5

3.0

4.5

Ind

icate

d S

pecif

icC

O-e

mis

sio

ns /

g/k

Wh

0

10

20

30

Cetane number / -

20 30 40 50 60 70 80

EN590 Diesel n-decane GtL iso-octane/n-heptane blends 1-alcohols di-n-butyl etherdi-n-butyl ether

Lower cetane

number

Lower aromatic

content

Increased oxygen

content

Increased volatility

E.g. ether functional

group with strongly

enhanced volatility

(high vapor pressure,

low viscosity,

low density)

Very high ignition

propensity



S. Pischinger


20 von 52

Ind

icate

d S

pecif

icP

M-e

mis

sio

ns /

mg

/kW

h

0.0

15.0

30.0

45.0

Ind

icate

d S

pecif

icH

C-e

mis

sio

ns /

g/k

Wh

0.0

1.5

3.0

4.5

Ind

icate

d S

pecif

icC

O-e

mis

sio

ns /

g/k

Wh

0

10

20

30

Cetane number / -

20 30 40 50 60 70 80



n-decane

EN590 Diesel


1-alcohols

Ind

icate

d S

pecif

icP

M-e

mis

sio

ns /

mg

/kW

h

0.0

15.0

30.0

45.0

Ind

icate

d S

pecif

icH

C-e

mis

sio

ns /

g/k

Wh

0.0

1.5

3.0

4.5

Ind

icate

d S

pecif

icC

O-e

mis

sio

ns /

g/k

Wh

0

10

20

30

Cetane number / -

20 30 40 50 60 70 80 90 100 110 120

Ind

icate

d S

pecif

icP

M-e

mis

sio

ns /

mg

/kW

h

0.0

15.0

30.0

45.0

Ind

icate

d S

pecif

icH

C-e

mis

sio

ns /

g/k

Wh

0.0

1.5

3.0

4.5

Ind

icate

d S

pecif

icC

O-e

mis

sio

ns /

g/k

Wh

0

10

20

30

Cetane number / -

20 30 40 50 60 70 80

EN590 Diesel n-decane GtL iso-octane/n-heptane blends 1-alcohols di-n-butyl etherdi-n-butyl ether di-methyl ether (DME)*

Lower cetane

number

Lower aromatic

content

Increased oxygen

content

Increased volatility

Also DME is a

promising candidate

(but not liquid!)

How to charac-

terizesuch

advanced

fuels properly?

• Split up mixture formation, ignition and emission formation for fuel characterization!



S. Pischinger

From Biofuels to propulsion: approach to derive the OPN

21 von 52

[7] Pischinger S., "Internal Combustion Engines", 6th Edit. VKA, 2013.

l /

Te

mp

era

ture

/ K

Air utilization (AU) to monitor the mixture formation

Def.:

Soot formation

Mixture

2500

2000

1500

1000

500

00 0.5 1.0 1.5 2.0

Burned

mixture

Soot oxidation

[8]

Screening of

Fuel Properties

Mixture Form.

Environmental demand:

Nozzle outlet flow conditions

Atomization and air entrainment

Evaporation rate

Fuel path aspects

lean unused rich

Oxidation Potential Number

160 180 200 220 240 260 280 300

0.0

0.2

0.4

0.6

0.8

1.0

Rich mixture (l < 1.0)Lean mixture (1.0 < l < 2.0)Unused air EVO-SOI (l > 2.0)

Soot formation

Soot oxidation

Air

uti

liza

tio

n /

-

Crank angle / °CA ABDC

SOI EVO

… by integrating the air utilization curves …

Total area|SOI

EVO

.

Lean mixture

Rich mixture Unused air

.

OPN :

OPN ↑ Soot oxidation ↑

CFD calculation w/o combustion

Environmental demand:

Engine load point

Injection strategy

Combustion system layout

Engine-related aspects



S. Pischinger

From Biofuels to propulsion: definition of OPN

22 von 52

[8] B. Graziano et al. SAE Int. J. Fuels Lubr. 8(1):2015, doi:10.4271/2015-01-0890.

Better degree of mixing enhances convective and diffusive transport of oxygen

At fixed engine load point, fuel thermo-physical properties drive the air/fuel mixing process

Def.: Oxidation Potential Number


160 180 200 220 240 260 280 300

0.0

0.2

0.4

0.6

0.8

1.0


Soot formation

Soot oxidation

Air

uti

liza

tio

n /

-


SOI EVO


Total area|SOI

EVO

.

Lean mixture


.

OPN :



160 180 200 220 240 260 280 300

0.0

0.2

0.4

0.6

0.8

1.0


Soot formation

Soot oxidation

Air

uti

liza

tio

n /

-


SOI EVO


Total area|SOI

EVO

.

Lean mixture


.

OPN :


Screening of

Fuel Properties

Mixture Form.

160 180 200 220 240 260 280 300

0.0

0.2

0.4

0.6

0.8

1.0

Rich mixture (l < 1.0)Lean mixture (1.0 < l < 2.0)Unused air (l > 2.0)

Soot formation

Soot oxidation

Air

uti

lizati

on

/ -

Crank angle () / °CA ABDC

SOI EVO

𝐴𝑈1<λ<2𝑑𝛼 ∙ 𝐸𝑉𝑂 − 𝑆𝑂𝐼𝐸𝑉𝑂

𝑆𝑂𝐼

𝐴𝑈λ<1𝑑𝛼 ∙𝐸𝑉𝑂

𝑆𝑂𝐼 𝐴𝑈λ>2𝑑𝛼𝐸𝑉𝑂

𝑆𝑂𝐼

…by integrating the air utilization curves ….



S. Pischinger

1E-3 0.01 0.120

40

60

80

100

120

140

LP4

LP4

LP4LP3

LP3

LP3

LP2

n-Octane

1-Octanol

Di-n-butylether

OP

N / -

ISPM / g/kWh

LP2

Engine load

increase

From Biofuels to propulsion: OPN vs engine out soot emissions, Test at constant EURO6 NOx

23 von 52

Example: C8 straight chained fuels

[9] B. Graziano et al. SAE Paper No: 2015-01-1934, JSAE Paper No.: 20159305 , 2015.

Screening of

Fuel Properties

Mixture Form.

OPN shows promising correlation with

engine out soot emissions for each fuel

C8 oxygenated allow reducing drastically

soot emissions (independent of OPN?)

Di-n-butylether (DNBE) allows maximum

OPNs:

best mixture formation process.

OPN to be put in context with DCN

(i.e. ignition delay)

Soot formation chemistry to be

included

Summary:

LP 02: n = 1500 rpm, IMEP = 6.8 bar

LP 03: n = 2280 rpm, IMEP = 9.4 bar

LP 04: n = 2400 rpm, IMEP = 14.8 bar



S. Pischinger

From Biofuels to Propulsion: Smoke Point Measurements to describe sooting tendency of a fuel

27 von 52

Smoke Point Determination Sooting tendency of C8 fuels

𝑂𝐸𝑆𝐼 = 𝑎𝑥 +

𝑦4−𝑧2

ℎ𝑆𝑃+ 𝑏

TSI / OESI ↑ Soot tendency ↑

Where: a, b: apparatus-dependent constants; MW: fuel molecular weight in g/kmol; x, y, z: carbon, hydrogen and oxygen molar

concentrations; hSP: smoke point height in mm.

𝑇𝑆𝐼 = 𝑎𝑀𝑊

ℎ𝑆𝑃+ 𝑏

[1] [2]

[14] H.F. Calcote, D.M. Manos. Combustion and flame: 49(1–3), 1983.

[15] E.J. Barrientos et al., Combust. Flame. 160 (8): 1484–1498, 2013.

C8 fuels tested at constant oxygen mass fraction (YO)

in base blend (35 vol.% toluene / 65 vol.% heptane).

2

4

6

8

10

12

14

16

0.2 0.7 1.2

OE

SI /

mm

-1

Base blend molar fraction / -

n-Octane

Di-n-butylether

Octanal

1-Octanol

Base blend

YO = 7%

YO = 5% YO = 9% YO = 4%

YO = 6%

YO = 3%

YO = 1%

[16] B. Graziano et al. “C8 Oxygenated Fuels In-Engine Characterization“-

3rd TMFB International Conference –Aachen- 2015.



S. Pischinger

So what do we learn from this?

2nd preliminary conclusion

28 von 52

In a first screening step, the DCN (measured or/and calculated) as well as the

pure chemical ignition delay serve to differentiate between a fuel’s potential for

SI- or CI-type combustion

The longer the side chain, the higher the DCN

suitability for CI engines, not suitable for SI engines

But: how high should the DCN (or CN) be for optimum low emission (soot, HC,

CO, noise) combustion?

With more descriptors we can define the desired properties more specifically!

Screening of

Fuel Properties

Proposal of fuel

molecules

Gasoline Diesel

DCN < 10 DCN > 30 (40)

Enthalpy of vaporization < 60kJ/kgair Lubricity: HFRR < 460 µm

Tboil < 120°C High DCN only with high OPN, Low TSI

These desired values can be used for Fuel Design!



S. Pischinger

From Biofuels to propulsion: to derive characteristic numbers, physical properties are needed

25 von 52

Interdisciplinary CIF interactions!

Quantum mechanics model [10, 11]

OPN

Engine type & load

Pv,

Hv

ρ, σ, η

PC-SAFT equation of state

Fuel thermo-physical properties

Parameters prediction

3 parameters needed

to predict properties!

0 or 1 experiment needed!

Screening of

Fuel Properties

Mixture Form.

[10] Umer M. & Leonhard K, J. Phys. Chem. A, 2013, 117 (7), pp 1569–1582, doi:10.1021/jp308908j.

[11] K. Leonhard et al. Fluid Phase Equilibria. (362) 25 January 2014, Pages 41–50, doi:10.1016/j.fluid.2013.08.037.



S. Pischinger

PC-SAFT parameter prediction and improvement

26 von 52

[12] Moser & Kistenbacher 1987, adjusted for inflation

Screening of

Fuel Properties

Mixture Form.

saturation pressure

density

enthalpy of vaporization

etc.

PC-SAFT equation of state

3 parameters per component!

≥3 experiments per component!

fitting

#exp. costs psat accuracy

3 10,000 €1 0.9%

1 3,500 €1 2.0%

0 50 € 37% [12]

[10] Umer M. & Leonhard K, J. Phys. Chem. A, 2013, 117 (7), pp 1569–1582, doi:10.1021/jp308908j.

[11] K. Leonhard et al. Fluid Phase Equilibria. (362) 25 January 2014, pp 41–50, doi:10.1016/j.fluid.2013.08.037

[13] manuscript in preparation.

≙ 18 K in normal

boiling temperature

1 experiment

per

component!!!

Parameter

Improvement method [13]

rapid

screening! Quantum mechanics model

Parameter prediction method [10,11]



S. Pischinger

From Biofuels to Propulsion: CI Candidates Screening

29 von 52

# s

tru

ctu

res

TMFB

C2-C14 only

Structural pattern

constraints

Pro

du

cti

on

syste

m

dep

en

din

g r

eq

uir

em

en

ts

DCN > 30

Oxygen content > 10 wt-%

Heating value > 30 MJ/kg

Boiling point 60..300 ºC

Co

mb

usti

on

syste

m

dep

en

din

g r

eq

uir

em

en

ts

CI engine

structure

generation

Bio-derived

intermediates

TMFB Proposal of fuel

molecules [17] M. Dahmen et al., SAE Int. J. Fuels & Lubr. 5(3),990-1003,2012.

[18] M. Dahmen & W. Marquardt, manuscript in preparation.



S. Pischinger

From Biofuels to Propulsion: CI Candidates Screening

30 von 52

typical (sub-)patterns[1,3]

A full range of novel candidates

for the compression-ignition engine!

DCN* 38

* DCN prediction [6]

DCN* 75

DCN* 72

DCN* 98 DCN* 34

DCN* 71

DCN* 77

DCN 34

TMFB

CI engine

Proposal of fuel

molecules [6] M. Dahmen & W. Marquardt, submitted to Energy & Fuels.

[18] M. Dahmen & W. Marquardt, manuscript in preparation.

[19] J. Klankermayer, ITMC, manuscript in preparation.

TMFB fuel candidate

1-octanol



S. Pischinger

Data-driven fuel design approach: filtering all possible molecules according to the desired fuel properties

31 von 52

# s

tru

ctu

res

mo

del

co

mp

lexit

y

mo

del

co

mp

lexit

y

TMFB

Oxygen content

Boiling point

Ignition Delay

Accessibility

score

Reaction yield

Life-cycle

analysis

… …

… …

Pro

du

cti

on

syste

m

dep

en

din

g r

eq

uir

em

en

ts

Co

mb

usti

on

syste

m

dep

en

din

g r

eq

uir

em

en

ts O

O

OH

…

Solving the inverse problem!

Proposal of fuel

molecules



S. Pischinger

Fuel design for SI fuel candidates

32 von 52

[20] M. Hechinger, Dissertation, AVT-PT, RWTH, Aachen,2014.

[21] M. Dahmen et al., SAE Int. J. Fuels & Lubr. 5(3),990-1003,2012.

279 molecular structures

#str

uctu

res

TMFB

C1-C8 only

Structural pattern

constraints

Pro

du

cti

on

sys

tem

dep

en

din

g r

eq

uir

em

en

ts

Enthalpy of vaporization

<60 kJ/kg(air)

Oxygenate

Heating value > 30 MJ/kg

Boiling point 50..100 ºC

Co

mb

us

tio

n s

ys

tem

de

pe

nd

ing

re

qu

ire

me

nts

All mathematically feasible

CxHyOz structures

based on the valence rules

SI engine Proposal of fuel

molecules



S. Pischinger

From Biofuels to propulsion: SI candidates screening

33 von 52

[20] M. Hechinger, Dissertation, AVT-PT, RWTH, Aachen,2014.

[21] M. Dahmen et al., SAE Int. J. Fuels & Lubr. 5(3),990-1003,2012.

RON 117

279 molecular

structures

furans, dihydrofurans,

tetrahydrofurans,

pyranes

typical (sub-)patterns

acylic ethers and acetals

aldehydes and ketones

alcohols

2-butanone

DCN <5

methyl-

isopropyl-

ketone

DCN <5

di-isopropyl-

ketone

DCN 16.6

methyl-

isobutyl-

ketone

DCN 12.6

Ketones selected for investigation in an

Ignition Quality Tester (IQT) TM

(ASTM D6890)

Ketones: attractive for SI engines! MON 107

TMFB

fuel candidate

2-butanone

Proposal of fuel

molecules



S. Pischinger

And it works: butanone shows excellent full load behavior…

34 von 52

Improved injector layout and high knock

resistance of both fuels also lead to an

optimal MFB50 % at very high loads

High efficiency up to 42 % at high loads for

both fuels due to an optimal MFB50 % in

combination with a high compression ratio

Same spark timing for both fuels

Better combustion stability for 2-butanone

at higher loads

2-Butanone compensates lower

evaporative cooling compared to

ethanol by lower ignition delay times

0

20

40

60

80Indicated efficiency / %

nMot = 2000 min-1, IMEP = 24 bar, CR = 13.5, EVC = 4° CA BTDC, IVO = 10° CA ATDC,SOI = 300° CA BTDC, pRail = 200 bar, TCoolant = TOil = 90 °C, lSpindt = 1.0

Ethanol 2-Butanone

0

5

10

15

20Mass fract. burned 50 %/ ° CA ATDC

opt.

0

100

200

300

400

Indicated mean effective pressure / bar

9 18 24

Peak cylinder pressure/ bar

0.0

0.2

0.4

0.6

0.8

9 18 24

IMEP standard deviation / bar

6

8

10

12

14Mass fraction burned 50 %/ ° CA ATDC

opt.

0

20

40

60



Ethanol 2-Butanone

0.1

0.2

0.3

0.4

0.5

Intake air temperature / °C

40 55


150

160

170

180

190


40 55


Limit

6

8

10

12

14Mass fraction burned 50 %/ ° CA ATDC

opt.

0

20

40

60



Ethanol 2-Butanone

0.1

0.2

0.3

0.4

0.5


40 55


150

160

170

180

190


40 55


Limit

(CR = 13.5) (CR = 13.5)

Engine settings:

SOI = 300° CA BTDC

Spark timing: MFB50 % = 8° CA ATDC

if not knock restricted

pRail = 200 bar

TCoolant = TOil = 90 °C

TIntake = 25 °C

Proposal of fuel

molecules



S. Pischinger

…and even allows for improvement under crucial operating conditions (catalyst heating)!

35 von 52

RON95, ethanol and 2-butanone

similar concerning their combustion

stability

Improved mixture formation with both

2-butanone and 2-methylfuran

compared to ethanol due to lower

heat of vaporization and higher vapor

pressure resulting in:

Lower HC emissions

Higher efficiency

Lower oil dilution

nMot = 1200 min-1, IMEP = 3 bar, CR = 13.5, EVC = 4° CA BTDC, IVO = 10° CA ATDC,SOI1 = 260° CA BTDC, SOI2 = 20° CA ATDC, ti2 = 0.2 ms, SP = 25° CA ATDC, lSpindt = 1.0pRail = 200 bar, TCoolant = 30 °C, TOil = 40 °C

0.1

0.2

0.3

0.4

0.5

RON95

E10

Ethan

ol

2-M

ethy

lfura

n

2-But

anon

e

IMEP standard deviation/ bar

0

5

10

15


0

2

4

6

8

RON95

E10

Ethan

ol

2-M

ethy

lfura

n

2-But

anon

e

Oil dilution (calc.)/ mg/cycle

0

10

20

30

40Indicated specific HCemissions / g/kWh

Engine settings

nEng = 1200 min-1, IMEP = 3 bar

SOI1 = 260° CA BTDC

SOI2 = 20° CA ATDC, ti2 = 2 ms

Spark timing = 25° CA ATDC

pRail = 200 bar

Proposal of fuel

molecules



S. Pischinger

The process efficiency needs to be assessed just like the combustion properties!

Example: development of biorefinery concepts for biofuel production.

36 von 52

platform

chemicals fuel candidates

O

OH

CH2 O

OH

O

O H

O

OH

…

sugars

Early stage process pathways ranking

…

?

biomass

lignin

cellulose

hemicellulose

? TMFB

?

Typical bottleneck in biofuel processes

- High-boiling components in H2O

- H2O removal energy-intensive

Screening of

production

feasibility

Proposal of fuel

molecules



S. Pischinger

From Biomass to Biofuels: process performance screening

Example: screening of fermentative platform chemicals.

37 von 52

[22] Voll & Marquardt, AIChE Journal (2012) 58:1788-1801.

[23] Ulonska, et al.12th Int. Symp. Process Systems Engineering - 25th European Symposium on Computer Aided Process

Engineering ,1331-1336, 2015. .

Fermentation Purification

Yield

Productivity

Concentration

Aeration

pH

Sugars

TM

TB

pKa…

Product

Accessible from

lab-experiments Accessible

property data

New extended methodology → process performance [19]

Fast evaluation of a high number of fermentation setups and products possible

Reaction Network Flux Analysis[22] Evaluation of pathways according to reaction yields

No assessment of process energy requirement

Screening of

production

feasibility



S. Pischinger

Performance Indicators for Biofuel Production

38 von 52

Example Products → Relative Potential (Basis Ethanol)

Hydrogen demand

Auxiliaries (i.e. solvents) 3. Global

Warming Potential

Raw material costs

Yield Δhcomb,efficiency

Carbon efficiency

Heating demand

Electricity demand

2. Primary Energy Demand 1. Production cost

…

1. Production Cost 2. Primary Energy Demand 3. Global Warming Potential

&

Global

Warming Potential

Future improvement steps

1. Itaconic acid

→ productivity increase (ferm.)

2. 2,3-Butanediol

→ downstream processing



S. Pischinger

From Biomass to Biofuels: performance indicators for biofuels production

39 von 52

3-Methyltetrahydrofuran (3-MTHF) via Itaconic Acid Fermentation

2-Methylfuran (2-MF) via biphasic 2-Methylfurfural synthesis

2-Butanone via 2,3-Butanediol Fermentation

RO &

Crystallization Hydrogenation Fermentation 3-MTHF Distillation Sugars

Phase

separation Decarbonylation Dehydroxylation 2-MF Distillation Cellulose

RO &

Extraction Dehydration Fermentation 2-Butanone Distillation Sugars



S. Pischinger

From Biomass to Biofuels: performance indicators for biofuels production

3-MTHF 2-MF 2-Butanone

Pathways

Indicators

Sugars → IA → 3-

MTHF

Cellulose → 2-MF Sugars → 2,3 BDO → 2-Butanone

Hydrogen demand 5 mol H2/mol MTHF No H2 required No H2 required

Carbon balance 1 CO2 lost per IA 1 CO2 lost per 2-

MF

2 CO2 lost per 2,3-BD

Yield 0.3 kgMTHF/kgsugars 0.3 kgMF/kgcellulose 0.3 kgMEK/kgsugars

Primary Energy

Demand

>0.7-1.3 MJ/MJfuel >0.1-0.3 MJ/MJfuel

>0.2-0.4 MJ/MJfuel

Measures continuous

fermentation

with in-situ product

removal

reaction

system for higher

yield and viable

processing

integrated separation and

dehydrogenation of 2,3-BDO from

fermentation broth

Integration of all process concepts with biomass pretreatment!

Improvement of conversion and separation in case of fermentation products!

Establish process development in particular for chemical synthesis of 2-MF!

only 60% yield in 1st reaction

Ref.: Ethanol:

0.2 MJ/MJfuel

low productivity in fermentation

large H2 demand tricky separation

40 von 52



S. Pischinger

From Biofuels to propulsion: Assessing the eco-toxicity impact of TMFB fuels

Diesel 2-MF 2-MTHF0

10

20

30

40

50

600

800

1000

1200

EC

50

[mg

HC

/L]

0.16

27

932

Acute daphnid assays can be used for a rapid toxicity

screening of biofuels and comparison to reference

fuels

Concept of Green Toxicology: „Test early, produce safe“ &

Fail early, fail cheaply“

Biofuel candidates and diesel fuel were tested in

different exposure

Direct dosing (Biofuels) vs passive dosing (Diesel)

Only indirect comparison possible!

Present results strongly indicate a lower aquatic

hazard potential for biofuels!

Toxic potential: Diesel > 2-MF >> 2-MTHF

Next steps: Screening of further fuel candidates

EC50-values in [mg hydrocarbon / L] of diesel, 2-MF and 2-MTHF found in the D. magna immobilization assay

(OECD 202) after 48 h. EC50-value for diesel was determined by GC-FID measurement. Hydrocarbon

concentrations for the biofuels were calculated from EC50-values measured by HPLC. Numbers denote the EC50-

value. Red arrow indicates increasing aquatic toxicity. n=3

EC50-value: Concentration with 50% effect (Immobility);

lower EC50-values indicate higher toxicity!

Increasing aquatic toxicity

Screening of

production

feasibility

41 von 52

[22] S. Heger, H. Hollert: manuscript in preparation.



S. Pischinger

Core Interaction Field Fuel Design: Characteristic Numbers

42 von 52

CIF-1

(Klankermayer)

Performance Indicators

for Biofuel Production

CIF-3

(Leonhard)

Model-Based

Fuel Design

CIF-2

(Pitsch)

Performance Indicators

for Biofuel Combustion

IRF-A IRF-B

Primary Energy

Demand

Production

Costs

OPN

DCN

TSI

1/WSD

Liquid-liquid

Equilibria

Global Warming

Potential Ecotoxicity



S. Pischinger

Fuel Design: the answers to the questions raised before

43 von 52

From Biomass

to Biofuels

From Biofuels

to Propulsion

Fuel Design:

Overall optimum of


biofuel combustion

Proposal of fuel

molecules

Screening of Fuel

Properties

Screening of

production feasibility

Proposal of fuel

molecules

Which molecules are

proposed and why?


use to measure the

properties and why?

DCN TSI


use to describe the

production and why?

?

TMFB

Which molecules are

proposed and why?



S. Pischinger

Conclusions

44 von 52

Descriptors for both fuel synthesis as well

as fuel usage have been developed and

are further improved

Already at an early design stage, the

following attributes can be described

Energy efficiency of production

Production cost

GWP

Ignition behavior (DCN, Ignition Delay)

Mixture formation behavior (OPN)

Sooting behavior (TSI)

Lubricity*

Liquid-Liquid Equilibria*

Ecotoxicity

Which descriptors do we use to describe the

fuel properties and production and why?

*not shown today

Which molecules are proposed and why?

Based on the mentioned descriptors, the

fuel properties can be chosen freely:

High production efficiency and low carbon

footprint

Adjusted combustion properties according to

its application

SI engines CI engines

Finally, the Fuel Design is carried out

New fuel molecules are proposed

Here, promising candidates for SI and CI

combustion were identified



S. Pischinger

Summary

45 von 52

Desired properties

Energy efficiency of production HIGH

Production cost LOW

GWP LOW

Ignition behavior (DCN, Ignition Delay) < 10 > 40 (bet. > 60)

Mixture formation behavior (OPN) HIGH > 50

Sooting behavior (TSI) - LOW (< 10)

Which fuels do we want?

SI engines CI engines



S. Pischinger

Tailor-Made Fuels from Biomass - RWTH Aachen University

Documents

Transcript of Tailor-Made Fuels from Biomass - RWTH Aachen University