Upgrading of Pyrolysis Oil with Catalytic Hydrotreatment

1
Upgrading of Pyrolysis Oilwith Catalytic
Hydrotreatment

• Agnes Ardiyanti
• Erik Heeres

2
Lignocellulosic biomass(woody biomass)

• Source wood, grass, sawmill dust
• Composition (in wt-)1
• Potential 13 EJ (minimum) in 2030

1WUR 2van Dam, 2007
3
Lignocellulosic biomass valorisation pathways
4
Fast Pyrolysis Oil
Volatiles
Fast Pyrolysis
Lignocellulosic biomass
Condensables, Fast Pyrolysis Oil
450-600 oC, 1-2 s
Char
BTG, Enschede
Bridgewater et al, Org. Geochem, 30,1999
5
Fast pyrolysis oil

• High oxygen content (up to 50)
• Immiscible with petroleum products
• Unstable upon heating and storage (coke
  formation, repolymerization)

6
ObjectiveDeoxygenation of Pyrolysis Oil
Co-feedstock for refineries (FCC, hydrocracking)
Deoxygenation
Fast pyrolysis oil
Selected process Catalytic Hydrotreatment
H2
Gas
Catalyst, P, T
Upgraded Oil
Fast pyrolysis oil
Water
-(CHxOy)- c H2 ? -(CHx)- (H2O, CO2, CH4, CO)
7
Desired product

• Low oxygen content
• Low viscosity
• Low molecular weight
• High aliphatic content
• Low coking tendency

8
Catalytic hydrotreatment
Upgraded oil properties
Process variables

• Oxygen content
• H/C, O/C ratio
• Viscosity
• Molecular weight
• Coking tendency

Catalyst Heating route Reactor design
9
Heating route
10
Why heating route?

• Polymerization is very common! ? sticky, gooey
  paste is produced, instead of a nice and liquid
  oil
• Pyrolysis oil contains 30 wt sugar ? when
  heated charring

Which condition should we apply to suppress this
reaction?
11
Pyrolysis Oil
Hypothesis1,2
HPTT
HDO
gt250oC, H2, catalyst
175-225oC

• Thermal cracking releases O mainly as H2O and CO2
  
• Repolymerisation occurrs

• O is released as H2O, H2 is consumed
• Further consumption of H2 saturates the C-C
  double bonds and cracks the large molecules
  (similar to coal liquefaction)

gt250oC, H2, catalyst
Low H/C, High Mw
High H/C, Low Mw
1 Gagnon, Ind. Eng. Chem. Res 27, 1988 2
Venderbosch, et al, J. Chem. Tech Biotech, 85,
2009
12
Experimental set-up

• 4 fixed-bed reactors in-series
• Feed forest residue pyrolysis oil (VTT, Finland)
• Catalyst Ru (5)/C
• H2 pressure 200 bar
• Variables T, WHSV
• Analysis
• Elemental composition, TGA, GPC, viscosity

BTG, The Netherlands
13
Effect of process conditions, visual observations

• High T in all 4 reactors
• Phase separation, clogging after 25 min
• Low T in all 4 reactors (Stabilization)
• Phase separation at 225 oC or higher
• Low T in first reactors, high T at the end (Mild
  Hyd)
• Phase separation, run for 3 days without clogging
• 2-stage Hyd (Hydrotreatment on Mild Hyd
  organic product)
• Top organic layer formed, no clogging observed

Py-oil
Mild Hyd
2-stage Hyd
14
Van Krevelen plot
Py-oil (dry)
Stabilization 175 oC
Stabilization 225 oC
Mild hydrotreatment
2-stage
Hydrogenation ? dehydration ? hydrogenation
15
Why H/C and O/C?
H/C 1 O/C 0
HDO
HDO
H/C 1.7 O/C 0
H/C 1 O/C 1/6
H/C 0.5 O/C 0
Coke formation
16
Physical properties during further hydrotreatment
Mw and TGA
stab
Mild
2-stage
Py-oil
Mw
residue (TGA)
Correlation between Mw and residue weight (TGA)
17
TG residue, as a function of H/C and O/C

• TGA residual weight 81.523 57.164 H/C
• 32.25 O/C

Estimation of physical properties is possible
18
Change of composition solvent fractionation

• Sugar, HMM decreases after reaction, leaving the
  apolar, low molecular weight components behind!

19
1H-NMR (organic phase)
Pyrolysis oil
Stabilization 175 oC
Mild hydrotreatment
2nd hydrotreatment

• Groups aldehydes, aromatics, carbohydrates,
  methoxy, aliphatics

20
Upgraded oil as co-feeding
In catalytic cracking

• Comparable yields are found for the petroleum
  feed (Long Residue) and mixture of Long
  residueupgraded oil

de Miguel Mercader, App. Cat. B 96, 2010
21
Summary on heating route

• Van Krevelen plot indicates the occurence of
  three subsequent processes
• hydrogenation,
• dehydration,
• hydrogenation
• During hydrotreatment, the Mw, viscosity, and TGA
  residue-weight of product oil increase during the
  stabilization step, then decrease at more severe
  conditions.
• High H/C and low O/C of the organic product is
  desired
• The change of composition can be followed by e.g.
  solvent fractionation and 1H-NMR.
• Upgraded oil can be used as co-feeding in
  refinery units

22
Catalyst
23
What type of catalyst?

• No specific reaction ? homogeneous is not an
  option
• Heterogeneous catalyst Which support, active
  metal, preparation?

24
Support

• Regenerable
• Stable in water, acid, high temperature
• ZrO2, SiO2 ? potential
• High specific surface area (less important)

Active metal

• Any metal with hydrogenation activity
• Interesting noble metals (Ru, Pd, Rh), Ni
  (usually promoted)

25
Noble metal vs cheaper transition metal

• Noble metal high activity, easy maintenance,
  very high price
• cheaper transition metal lower activity, prone
  to deactivation, cheap

www.kitco.com
26
Van Krevelen comparison of activity
Pd/C
Ru/C
27
Potential catalyst NiCu

• d-Al2O3 as support (better stability than
  ?-Al2O3)
• Various Ni/Cu ratio

28
Hydrogenation activities

• Van Krevelen plot is used to calculate the
  hydrogenation activities, blank experiment as the
  reference

16Ni2Cu and 13.8Ni6.83Cu are the most active
29
Why is Cu needed?

• Ni is a catalyst for CNT (carbon nanotube)
  formation? produces carbon whiskers, decrease
  the activity
• CNT formation is structure sensitive ? needs
  adjacent active sites
• Cu makes NixCu1-x alloy, and reduce the
  crystallite size ? the carbon formation is
  reduced
• Cu also helps the reduction

30
XRD analysis

• No Ni(0) was found at 20.8Ni after reduction at
  300 oC (reduction temperature of Ni is gt 500 oC)
• Ni(0) was formed on 13.8Ni6.83Cu after reduction

13.8Ni6.83Cu
Ni
20.8Ni
NiO
Cu does not have HDO activity, but supports the
reduction of Ni
Reduction was performed at 300 oC and 10 bar of H2
31
What about the stability?HRTEM
Fresh 16.8Ni6.83Cu
Spent 16.8Ni6.83Cu
Active metal particle size 10 nm (fresh) ? 100
nm (spent). ICP showed leaching of Ni, Cu, and Al
Dissolution and recrystallisation of NiCu seem to
occur
32
Next? Find other supports

• Carbon, ZrO2, TiO2, etc
• Ongoing research

33
Summary on catalyst selection

• A good support selection is a good start
• Noble metal vs cheaper transition metal
• Bimetallic catalyst effect of composition

Heterogeneous catalysts, S?dChemie
34
AcknowledgementRobbie Venderbosch, Vadim
Yakovlev, Sofia Khromova, Jelle Wildschut, Anja
Oasmaa, Jelmer Westra
Boreskov Institute of Catalysis SB RAS