fast pyrolysis development_venderbosch et al. 2010

Correspondence to: RH Venderbosch, BTG Biomass Technology Group B.V., Josink Esweg 34, 7545 PN Enschede, The Netherlands.

E-mail: [email protected]

178

Review

© 2010 Society of Chemical Industry and John Wiley & Sons, Ltd

Fast pyrolysis technology developmentRH Venderbosch, BTG Biomass Technology Group BV, Enschede, the Netherlands

W Prins, Faculty of Bioscience Engineering, University of Ghent, Belgium and BTG Biomass Technology Group BV,

Enschede, the Netherlands

Received October 21, 2009; revised version received November 30, 2009; accepted December 14, 2009

Published online in Wiley InterScience (www.interscience.wiley.com); DOI: 10.1002/bbb.205; Biofuels,

Bioprod. Bioref. 4:178-208 (2010)

Abstract: While the intention of slow pyrolysis is to produce mainly charcoal, fast pyrolysis is meant to convert bio-

mass to a maximum quantity of liquids (bio-oil). Both processes have in common that the biomass feedstock is

densifi ed to reduce storage space and transport costs. A comfortable, more stable and cleaner intermediate energy

carrier is obtained, which is much more uniform and well defi ned. In this review, the principles of fast pyrolysis

are discussed, and the main technologies reviewed (demo scale: fl uid bed, rotating cone and vacuum pyrolysis;

pilot plant: ablative and twin screw pyrolysis). Possible product applications are discussed in relation to the bio-oil

properties. General mass and energy balance are provided as well, together with some remarks on the economics.

Challenges for the coming years are (1) improvement of the reliability of pyrolysis reactors and processes; (2) the

demonstration of the oil’s utilization in boilers, engines and turbines; and (3) the development of technologies for the

production of chemicals and biofuels from pyrolysis oils. One important conclusion in relation to biofuel production

is that the type of oxygen functionalities (viz. as an alcohol, ketone, aldehyde, ether, or ester) in the oil should be

controlled, rather then merely focusing on a reduction of just the oxygen content itself. © 2010 Society of Chemical

Industry and John Wiley & Sons, Ltd

Keywords: pyrolysis, technology, review, bio-oil, biomass

Introduction

Environmental concerns and possible future short-

ages have boosted research into alternatives for fos-

sil-derived products. Biomass is abundantly available

worldwide and considered to be renewable. Despite its com-

plexity, the use of biomass is rapidly expanding. Agriculture,

petrochemical industries, and individual entrepreneurs,

meanwhile, have developed a signifi cant production of so-

called fi rst-generation biofuels from vegetable oils (biodiesel),

sugar and starch (bioethanol). Th e scale of production of

these fi rst-generation fuels (<100 MWth) appears to be sig-

nifi cantly lower than that of unit operations in traditional

petroleum refi neries (several GWth’s). Obviously the pro-

duction of fi rst-generation biofuels is in competition with

the food/feed industry. Th is is a non-ethical situation with

a limited CO2 reduction potential that is acceptable only

to get started while developing the techniques for second-

generation biofuels derived from lignocellulosic biomass

(residues like wood thinning, bagasse, rice husks, straw, etc.).

Second-generation biofuels can be made by (1) hydrolysis and

fermentation of cellulosic materials to ethanol; (2) gasifi cation

© 2010 Society of Chemical Industry and John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 4:178–208 (2010); DOI: 10.1002/bbb 179

Review: Fast pyrolysis technology development RH Venderbosch, W Prins

of lignocelluloses to syngas and further conversion to metha-

nol or gasoline/diesel, for example; and (3) liquefaction of

lignocelluloses with further upgrading, either by gasifi cation

(see 2), or by hydro-de-oxygenation (CO2 and H2O removal).

Large-scale implementation of biomass technologies is hin-

dered by a number of adverse biomass properties causing

diffi culties or excessive costs in its processing. Such problems

are in the highly distributed occurrence of biomass and its

low annual yields (usually less than 10 ton per hectare (dry

and ash-free), as well as in the wide variety in biomass struc-

ture and energy density (compare, for example, wood logs

with rice husks). To overcome these disadvantages of bio-

mass, one could include a pre-treatment process as a fi rst step.

In the overall conversion chain, fast pyrolysis, being such a

pre-treatment process, creates a uniform, liquid intermedi-

ate that is virtually ash-free and has a signifi cantly increased

energy density. It can be produced at a scale matching the size

of cost-effi cient biomass collection, stored until transport to

the nearest harbor, and fi nally shipped together with the pro-

duction of many other pyrolysis plants to a central site (e.g.,

a refi nery or a power station) for further conversion to heat,

electricity, chemicals, or fuels at any desirable time.

Th e handling and processing of liquids has many advan-

tages in comparison with processing solid or gaseous feed

streams. Interest in the production of pyrolysis liquids from

biomass has grown rapidly in recent years, due to the poten-

tial possibilities of:

• de-coupling liquid fuel production (scale, time, and loca-

tion) from its utilization;

• separating minerals on the site of liquid fuel production

(to be recycled to the soil as a nutrient);

• producing a renewable fuel for boilers, engines, and tur-

bines, power stations and gasifi ers;

• secondary conversion to motor-fuels, additives or special

chemicals (biomass refi nery); and

• primary separation of the sugar and lignin fractions in

biomass with subsequently further upgrading (biomass

refi nery).

An interesting aspect here is that pyrolysis could connect

(conventional) agricultural business to (petro)chemical

processes. Besides, fast pyrolysis can be integrated with bio-

logical processes in many ways, to form dedicated biorefi n-

eries (e.g., conversion of lignin residues to bio-oil (and bio-

char), fermentation of the fast pyrolysis sugar fraction, etc.).

Many reviews on pyrolysis can be found in the literature1–4

and in three handbooks edited by Bridgwater,5–7 while the

PyNe website and PyNe (now Th ermalNet) newsletters show

the state-of-the-art. Th e reviews should be read with some

care, as the data presented do not always refl ect the actual

case due to a lack of precise information on actual feedstock,

processing conditions, variations in feeds used, and a lack of

detail on the (usually proprietary) technologies.

Pyrolysis processes are carried out in the absence of oxy-

gen, at atmospheric pressure and temperatures ranging from

300 to 600°C. Charcoal is the main product of the tradi-

tional slow pyrolysis process, in which the biomass (usually

wood) is heated slowly to temperatures between 300 and

400°C. To the contrary, fast pyrolysis processes are char-

acterized by a high rate of particle heating to temperatures

around 500°C, and a rapid cooling of the produced vapors

to condense the liquids. Th is yields a maximum quantity

of dark-brown mobile liquid with a heating value roughly

equal to that of wood, which is approximately half the heat-

ing value of fossil fuel oil. Th e earliest recorded use of this

technique was in Egypt, where the product was used for

sealing boats.8 In more recent times, a number of chemicals

were derived from the liquids as well (e.g., methanol, acetic

acids, etc., and liquid smoke). While the function of slow

pyrolysis is to produce mainly charcoal and gas, fast pyroly-

sis is meant to convert biomass into a maximum quantity

of liquids. Th ey are both meant to pre-treat the biomass to

facilitate transport, storage, and utilization.

Principles

Th ermal decomposition of biomass results in the production

of char and non-condensable gas (the main slow pyrolysis

products) and condensable vapors (the liquid product aimed

at in fast pyrolysis). It is realized by rapid heat transfer to

the surface of the particle and subsequent heat penetration

into the particle by conduction. For fast pyrolysis condi-

tions, meant to maximize the liquid yield, the temperature

development inside the particle, and the corresponding

intrinsic reaction kinetics dominate the conversion rates and

product distributions. Principally, biomass is decomposed to

a mixture of defragmented lignin and (hemi)cellulose, and

180 © 2010 Society of Chemical Industry and John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 4:178–208 (2010); DOI: 10.1002/bbb

RH Venderbosch, W Prins Review: Fast pyrolysis technology development

fractions derived from extractives (if present). Th e intention

of fast pyrolysis is to prevent the primary decomposition

products (1) to be cracked thermally or catalytically (over char

formed already) to small non-condensable gas molecules on

the one hand; or (2) to be recombined/polymerized to char

(precursors) on the other. Such conditions would then lead

to a maximum yield of condensable vapors and include the

rapid heating of small biomass feed particles. Besides, it is also

essential to create a short residence time for the primary prod-

ucts, both inside the decomposing particle and in the equip-

ment before the condenser. For chemicals or fuel application,

an additional target would be to control the chemical compo-

sition of the condensables; for instance, by applying catalysts.

First reactor developers adopted the concept of fl ash

pyrolysis in which small particles (<1 mm) were used to

achieve high oil yields. Later research showed that the oil yield

is much less dependent on biomass particle size and vapor

residence times than originally assumed.9,10 Th e composition

of the oil, however, is sensitive for these parameters. High

external heat transfer to the biomass particles can be realized

by mixing the cold biomass feed stream intensively with an

excess of pre-heated, expectedly inert, heat carrier (e.g., hot

sand). A number of reactor designs have been explored that

may be capable of achieving high heat transfer rates, such

as fl uidized beds and mechanical mixing devices. For an

effi cient heat transfer through the biomass particle, though,

a relatively small heat penetration depth is required, which

limits the ‘size’ of biomass particles to, typically, 5 mm. ‘Size’

here refl ects a distance of two times the actual (heat) penetra-

tion depth of the particle. For such particles, the decomposi-

tion rate is controlled by a combination of intra-particle heat

conduction and the decomposition kinetics. Oil yield values

observed in continuously operated laboratory reactors and

pilot plants, for wood as a feedstock material, are usually in

the range of 60 to 70 wt.% (dry-feed basis). Although generally

reported in reviews, oil yields over 70% are exceptional and

only for well-defi ned feedstocks as cellulose. Energetic yields

are a bit lower, approx. 55 to 65%. It is obvious that the energy

left in the byproducts should be used as well; for example, for

drying the feedstock and/or steam-electricity production. If

the objective is to derive chemicals from the pyrolysis liquid,

it is essential to operate the process at the proper conditions

(temperature, residence time, feedstock type, and feedstock

pre-treatment) in order to maximize the yield of the specifi c

component aimed at. When fuels are required, less stringent

criteria must be met; the conversion of as much biomass

energy as possible to the liquid product is then decisive. Until

recently, most R&D work has been focused on maximizing

the overall oil yield, without paying suffi cient attention to the

product composition and quality. Next to water, the major

components of biomass are:

- cellulose (mostly glucans), with a composition roughly

according to (C6H10O5)n, and n = 500 to 4000;

- hemicellulose (mostly xylans), with an average composi-

tion according to (C5H8O4)n, and n = 50 to 200; and

- lignin, consisting of highly branched, substituted, mononu-

clear aromatic polymers, oft en bound to adjacent cellulose

and hemicellulose fi bers to form a lignocellulosic complex.

Cellulose, hemicellulose, and lignin all have a diff erent ther-

mal decomposition behavior, and each individually depends

also on heating rates and the presence of contaminants.11 A

typical temperature dependence of the decomposition through

thermo-gravimetric analysis (TGA) for reed, carried out by

the University of Groningen is given in Figure I. Th e total

mass loss rate is plotted versus the temperature in Figure

IA, while the TGA data are interpreted in terms of cellulose

(almost 30%), hemicellulose (25%), and lignin (20%) in lB. Th e

diff erential plot for these fractions is given on the left hand

side against the original biomass data. Hemicellulose is the

fi rst component to decompose, starting at about 220°C and

completed around 400°C. Cellulose appears to be stable up

to approx. 310°C, where aft er almost all cellulose is converted

to non-condensable gas and condensable organic vapors at

320–420°C. Th ough lignin may begin to decompose already

at 160°C, it appears to be a slow, steady process extending

up to 800–900°C. At fast pyrolysis temperatures of around

500°C, the conversion of lignin is probably limited to 40%. In

general, for fast pyrolysis, a solid residues remains (char, not

shown in Fig. 1), which is then mainly derived from lignin and

some hemicellulose fractions, respectively 40 and 20 wt-% of

the original sample.12 A conclusion from such TGA data can

be that the oil is derived mainly from cellulose, and only par-

tially from hemicellulose (depending on the heating rate up

to approx. 80% conversion to oil and gas) and lignin (roughly

50% conversion to oil and gas). Th e explanation is that in the



biomass structure, lignin and hemicellulose are linked through

covalent bonds (ester and ether) and cannot be released that

easily upon pyrolysis; cellulose and hemicellulose are linked

by much weaker hydrogen bonds.13 Indirect evidence for this

hypothesis is given by the composition of the pyrolysis-derived

char, which has an elemental composition close to that of

lignin. Th e pyrolysis of biomass can be either endothermic or

exothermic, depending on the temperature of the reactions

and the type of feed. For (hemi)cellulosic materials, the pyroly-

sis is endothermic at temperatures below about 450°C, and

exothermic at higher temperatures. As argued already, vapors

formed inside the pores of a decomposing biomass particle

are subject to further cracking, leading to the formation of

additional gas and/or (stabilized) tars. Th e sugar-like fractions

especially can be readily re-polymerized, increasing the overall

char yield (mostly ex-bed of the pyrolysis process). Th is may

be the purpose of slow pyrolysis but should be avoided in fast

pyrolysis. For the small particles used in fast pyrolysis, second-

ary cracking inside the particles is relatively unimportant due

to a lack of residence time. When the vapor products enter the

surrounding gas phase, however, they will further decompose

if they are not condensed quickly enough.

Although other mechanisms have been proposed as well,

Fig. 2 shows a possible reaction pathway for biomass pyroly-

sis. Schemes like these, including three lumped product

classes, were originally proposed by Shafi zadeh et al.14,15

and start with a reaction that is fi rst order in the decompos-

ing component. Unfortunately, there is great variety in the

results of reaction rate measurements, even for a ‘single’

biomass type, such as wood. Published rate and selectivity

expressions maybe useful in describing trends, but they can

hardly be used for reliable quantitative predictions.10,16–18 It

should be realized here that biomass is a natural material,

with widely varying structural and compositional proper-

ties. Despite all such uncertainties in the required input

data, many scientists still propose single particle models

based on fundamental chemical and physical phenomena

taking place inside the particles. Kinetic data are proposed

for the pyrolysis of wood, but a small variation in ash con-

tent seriously aff ects these reaction rates, and probably the

pyrolysis pathways themselves. Already stated in 1991, ‘it

should not be expected that any simple one-step kinetic

scheme can account for all the facts concerning the pyrolytic

behavior of [just] carbohydrates’.19 Although the predictive

power is limited, and insuffi cient for scaling-up purposes,

modeling is still useful to create a better understanding.18

0

0.01

0.02

0.03

0.04

0 100 200 300 400 500

Temperature (°C)

Mas

s Lo

ss R

ate

(wt.%

/°C

)

lignin

cellulose

hemicellulose

water

0

0.01

0.02

0.03

0.04

0 100 200 300 400 500

Temperature (°C)

Mas

s Lo

ss R

ate

(wt.%

/°C

)

Figure 1. Thermogravimetric analysis curve for Reed (A) and the differential plot interpreted in terms of

hemicellulose, cellulose and lignin (B). TGA curves are prepared by the University of Groningen.

Wood (s)

bio-oil (l)

Char (s)

Wood vapors (g)

Gas (g)(CO,CO2,CH4)

Char (s) + gas (CO2)+ ‘bio-oil (l)’

Primary phasedecomposition

reactions

Secondary phasecracking + condensation

Re-polymerisation

Gas (g)(CO,CO2,CH4)

450 – 550°C<1 s

400 – 500°C>1 s

atmosphericweeks / months

Figure 2: Representation of the reaction paths for wood pyrolysis.



Effect of ash

A photograph of bio-oil is given in Figure 3. Th e maximum

possible oil yield thus depends on various parameters, and

this includes feedstock properties, water content, tempera-

ture and vapor residence time. Th e maximum possible oil

yield thus depends on various parameters, and this includes

feedstock properties, water content, temperature, and vapor

residence time. In addition, ash in the biomass can have a

dominant eff ect on the oil yield and composition.20,21 Th e

trend line in Fig. 4 is based on published data points.

In general, the yields of char and gas increase signifi cantly

for higher ash contents in the biomass, viz. at the expense of

oil yields.

Sodium and potassium will have a large impact, but sul-

fur- and phosphorus-containing ammonium salts can also

dramatically aff ect oil yields and promote char formation.22

Although there is evidence that ash (including metals) can

have a catalytic eff ect on the thermal degradation of biomass

during pyrolysis, data like in Fig. 4 should be interpreted

with care as they are derived for diff erent types of biomass

as well. Figure 4 may suggest otherwise, but not all ash-like

materials are detrimental for the oil yield; silicon and metals

other than the alkalis appear quite inactive in the pyrolysis

process. It has been suggested that alkali metals have the

potential to lower the optimal pyrolysis temperature as well.

Consequently, for biomass types with high ash content,

which is generally the case for the technological interesting

low-valued residues, the oil-yield can drop to values, some-

times below 50 wt.%. Th ough extremely relevant for both the

oil yield and oil quality, limited research has been carried out

to understand the eff ects of ash on the pyrolysis reactions.

Oil properties

Representative values for wood-derived pyrolysis oil proper-

ties are collected from various resources and listed in Table 1.

It is a liquid, typically dark red-brown to almost black, a

color that depends on the chemical composition and the

presence of micro-carbon (Fig. 3). Th e density of the liquid is

Table 1. The range of elemental composition and properties for wood-derived pyrolysis oil.1–3

Physical property Pyrolysis conditionsWater content

(wt.%)15–30 Temperature (K) 750–825

pH 2.8–3.8 Gas residence time (s) 0.5–2

Density (kg/m3) 10500–1250 Particle size (µm) 200–2000

Elemental analysis Moisture (wt.%) 2–12

(wt.% moisture free) Cellulose (wt.%) 45–55

C 55–65 Ash (wt.%) 0.5–3

H 5–7

N 0.1–0.4 Yields (wt.%)

S 0.00–0.05 Organic liquid 60–75

O Balance Water 10–15

Ash 0.01–0.30 Char 10–15

Gas 10–20

HHV (MJ/kg) 16–19

Viscosity (315 K, cP) 25–1000

ASTM vacuum

Distillation (wt.%) 430 K ~10

466 K ~20Solubility

(wt.%)Hexane ~1

492 K ~40 Toluene 15–20

Distillate ~50 Acetone >95

Acetic acid >95



about 1200 kg/m3, which is signifi cantly higher than that of

fuel oil. It has a distinctive acid, smoky smell, and can irritate

the eyes. Th e viscosity of the oil varies from 25 up to 1000 cP,

depending on the water content and the amount of light

components in the oil. It is important to note that oil proper-

ties depend on feedstock and operating conditions, but may

change during storage, a process indicated as ‘ageing’, which

(by lack of defi nition) is usually noticed by an increased vis-

cosity in time and a possible phase separation of the oil in a

watery phase and a viscous organic phase. Due to the pres-

ence of large amounts of oxygenated components, the oil has

a polar nature and does not mix readily with hydrocarbons.

In general, it contains less nitrogen than petroleum prod-

ucts, and almost no metal and sulfur components. However,

some of the nitrogen is transferred to the oil product as well:

feed materials with a high nitrogen contents yield oil with

higher pH-values and larger amounts of nitrogen in the oil.

Degradation products from the biomass constituents include

organic acids (like formic and acetic acid), giving the oil its

low pH of about 2 to almost 4. Th e oil attacks carbon steel,

and storage of the oils should be in acid-proof materials like

stainless steel or poly-olefi ns. Water is an integral part of the

single-phase chemical solution. Th e (hydrophilic) bio-oils

have water contents of typically 15–35 wt.%, and water can-

not be removed by conventional methods like distillation.

Th is high water content is a serious drawback if considering

the heating values: the higher heating value (HHV) is below

19 MJ/kg (compared to 42–44 MJ/kg for conventional fuel

oils). Above a certain water-content level, viz. in the range

of 30 to 45 wt.%, phase separation may occur. Depending

on the type of feedstock and process conditions the ratio of

oil over aqueous phase varies from 50:50 to 30:70, and the

presence of these two phases can complicate the oil’s applica-

tion. In many cases, suffi cient drying of the biomass feed-

stock material prior to pyrolysis prevents phase separation.

Applying diff erent condensation temperatures will yield oils

with diff erent water contents (aff ecting the oil yield as well),

and there is quite some room for optimization.23

Usually the choice of the feedstock and process (char-

acteristics) will determine the ‘oil’ quality and possible

phases. Benefi cial eff ects of the water content have also been

reported, viz. in case of combustion. It causes a decrease

in viscosity of the oil (facilitating transport, pumping, and

atomization); it improves ‘stability’; it lowers the combustion

0

10

20

30

40

50

60

70

80

0 3 5 6 7 8

Ash content (wt.%)

Oil

an

d c

ha

r y

ield

(w

t.%

)

oil

char

gas

Trendline oil (Chariamonti et al. 2007)

1 2 4

Figure 4. Relationship between oil yield and ash content in the

biomass. The solid lines represent trend lines taken from literature,20

while the broken line is a trend line based upon more than 20 data

point for wood.21Figure 3. Fast pyrolysis oil.



temperature and, as a consequence, it may cause a reduction

of the NOx emission. Generally speaking, the (organic) oil

yield achieved in fast pyrolysis should be as high as possi-

ble. Besides, the oil should have a much higher (volumetric)

energy content than the original biomass, and must be more

stable towards biological degradation.

Composition and stability

Pyrolysis oils are produced by the rapid quenching of frag-

mented biomass, these fragments being derived from the

biomass constituents, cellulose, hemicellulose, and lignin.

Th e largest fragments that are conveyed to the condenser in

the vapor phase have a molecular mass that is far too high

for being a gas component at temperatures around 500°C.

Th ey may be present in the vapor phase as aerosols, or are

produced upon ‘freezing’ the vapors. Th is liquid product

collected in the condenser includes the complete spectrum

of oxygenated compounds, with molecular weights rang-

ing from 18 to over 10 000 g/mol, the higher values prob-

ably caused by repolymerization of the biomass fragments.

Whereas some researchers think the oil is a (micro-)emulsion

of these compounds, there is also reason to believe the oil

is a mixture of soluble components, likely with water as the

solvent and polar sugar constituents behaving as bridging

agents in the dissolution of hydrophilic lignin material.24

GC-analysis (including 2D-GC, GC-MS etc) appears, to a

certain extent, valuable in the interpretation of the oil qual-

ity. However, the usefulness of GC data is limited due to

the (unknown) destructive eff ect of the technique on the oil

composition. GC injection includes vaporization of the feed,

which is known to be diffi cult for pyrolysis oils and causing

some coking in the injection part of the system. Moreover,

chemical reactions occurring in the GC column cannot be

excluded either, and it is questionable whether the compo-

nents actually detected are really present in the feed oil. Other

techniques of which the potentials for pyrolysis-oil analysis

are being investigated are Gel Permeation Chromatography

(GPC) and High Performance Liquid Chromatography

(HPLC), but both methods need to be handled with great

care. Fortunately, development of new techniques to footprint

the oils is ongoing. As the oil can not be distilled without

severe repolymerization and thus chemical degradation, a

solvent fractionation technique, illustrated in Fig. 5, is devel-

oped to analyze the oil in an alternative way, and reveal the

presence of certain fractions present in the oil:25

• water solubles (acids, alcohols, diethylethers);

• ether solubles (aldehydes, ketones, lignin monomers, etc.);

• ether insolubles ((anhydo)sugars, hydroxyl acids);

• n-hexane solubles (fatty acids, extractives, etc.);

• DCM solubles (low molecular lignin fragment, extrac-

tives); and

• DCM insolubles (degraded lignins, high molecular lignin

fragments, including solids)

Th e ether insolubles in particular (the sugar components, a

syrup-like fraction) appear to have high oxygen contents (up

to 50%) if compared to, for example, the DCM solubles and

insolubles (25 to 30% oxygen). Ongoing research is aimed at

BIO-OIL

WATER-SOLUBLES WATER INSOLUBLES

ETHER INSOLUBLESETHER SOLUBLES

DCM SOLUBLES DCM INSOLUBLES

water extraction

ether extraction DCM extraction

Ether solubles:(Aldehydes, ketons,lignin monomers)

Sugars (anhydrosugars,anhydro-oligomers,Hydroxyacids (C<10))

LMM lignin (low molecularlignins, extractives)

HMM lignin (high molecularlignins, solids)

N-HEXANE SOLUBLES

Extractives

Water (by KF titration)Solids (by MeOH-DCM extraction)

Figure 5. Fractionation scheme for chemical characterization.25



revealing the various components in each fraction, and their

eff ects on storage, stability, upgrading, etc. Th e results from

solvent fractionation and GC-MS can be combined.25 Th e

main part of GC-eluted compounds is in the ether-soluble

fraction of the fractionation scheme, with the DCM (in-)solu-

bles not being detected in a GC. Table 2 shows the combined

results of solvent extraction, GC/MSD and CHN analyses for

reference pine liquid. Although this method may not be the

future standard for bio-oil analysis, it could be an important

technique relevant to understanding how the oxygen is actu-

ally bound to the carbon (alcohol, keton, or as an ether/ester).

An important property of pyrolysis oil is its chang-

ing characteristics over time. Such an ‘instability’ can be

observed by a viscosity increase during storage, some for-

mation of carbon dioxide, an increased water content, and

eventually by phase separation. Defi nitions to address this

unstable character are lacking. Th e detailed mechanism

of this ‘ageing’, the causes of it, and the consequences for

further use are still unclear, and will depend highly on the

various oxygen functionalities in the oil (and therefore

feedstock type, operating conditions, initial quality, stor-

age temperatures, etc.). At room temperature, the ageing of

bio-oil occurs over periods of months or years, depending

on the type of feedstock, and its ‘initial quality’. At elevated

temperatures, however, the polymerization reactions are

enhanced signifi cantly, and it is therefore recommended

to avoid (long) storage at temperatures above 50°C. Recent

work indicates that recombination/polymerization of oil

fragments, accompanied by separation and evaporation

of small molecules (including CO and CO2), could be an

important cause.25 Although the reasons for instability may

be unclear, the major chemical change in wood-derived oil

is due to an increase in the DCM insoluble fraction and a

signifi cant decrease in the ether insoluble constituents (‘sug-

ars’). Th e increase in the average molecular weight in time,

the viscosity (of the organic fraction) and pour point, and/or

changes in the molecular structure cause phase separation.

Th e instability of the oil and the varying quality of oils

produced worldwide could be hurdles to further develop-

ment of oil applications. Much depends on the eventual

Table 2. Chemical composition of reference pine oil and its fractions.25

Reference Pine Oil wet dry C H N O

Water wt-% 23,9 0

AcidsFormic acid Acetic acid Propionic acid Glycolic acid

wt-%wt-%wt-%wt-%wt-%

4,3 5,61,53,40,20,6

40,0 6,7 0 53,3

AlcoholsEthylene glycolIsopropanol

wt-% wt-%wt-%

2,23 2,90,32,6

60,0 13,3 0 26,7

Aldehydes and ketonesNonaromatic AldehydesAromatic AldehydesNonaromatic KetonesFuransPyrans

wt-%wt-%wt-%wt-%wt-%wt-%

15,41 20,39,72

0,0095,363,371,10

59,9 6,5 0,1 33,5

Sugars Anhydro-ß-D-arabino-furanose, 1,5 Anhydro-ß-D-glucopyranose(Levoglucosan)Dianhydro-a-D-glucopyranose, 1,4:3,6

wt-%wt-%wt-%wt-%

34,44 45,30,274,010,17

44,1 6,6 0,1 49,2

LMM lignin Catechols Lignin derived PhenolsGuaiacols (Methoxy phenols)

wt-%wt-%wt-%wt-%

13,44 17,70,060,093,82

68 6,7 0,1 25,2

HMM lignin wt-% 1,950 2,6 63,5 5,9 0,3 30,3

Extractives wt-% 4,35 5,7 75,4 9,0 0,2 15,4



end application: although technologies are being demon-

strated already at a signifi cant scale (up to a 100 ton biomass

throughput per day), standards and specifi cations are still

underdeveloped. Some progress has been made however in

recent years.26 Various physical and chemical methods for

the characterization and analysis of pyrolysis liquids in rela-

tion to their future applications have now been identifi ed.

Th is applies to properties such as the viscosity, water con-

tent, pH, density, elemental composition, LHV, ash content,

char content, surface tension, solubility in diff erent solvents,

ageing characteristics, and pour and fl ash points.

Catalytic pyrolysis

It has been recognized already in the early days of fast

pyrolysis R&D that application of catalysis could be of major

importance in controlling the oil quality and its chemical

composition.27,28 Without any catalyst involvement, the bio-

oil derived from fast pyrolysis is a mixture of hundreds of

diff erent, highly oxygenated chemical compounds. Catalysis

could be applied for a number of reasons, and at a number

of diff erent positions in the process. Lower pyrolysis tem-

peratures, a higher chemical and physical stability, high

yields of target components, and an improved miscibility

with refi nery streams, are all goals strived for. Bio-oils can

be upgraded by either applying catalysts in the production

process (‘catalytic pyrolysis’), or by post-treatment of the

bio-oil over a catalyst bed. Th is post-treatment may be the

thermal cracking of re-evaporated bio-oil in a hot fl uidized

bed of Fluidized Catalytic Cracking (FCC) catalyst parti-

cles (‘bio-oil FCC’), or the catalytic hydrodeoxygenation at

elevated temperatures and hydrogen pressures (‘hydrotreat-

ment’). When, in catalytic pyrolysis, the catalyst particles

are mixed into the reactor together with the inert heat car-

rier (oft en sand), there is an immediate contact between

the catalyst and pyrolysis products. Besides, in pyrolysis

processes with a separate char combustor, the catalyst can

be regenerated continuously (coke burn off ). Requirements

are then that the particle properties of the catalyst should

match those of the inert heat carrier while its activity should

be maximal at the optimal fast pyrolysis temperature. More

fl exibility regarding the design of the catalyst and the condi-

tions of the catalytic treatment is created when a separate

reactor is installed in the pyrolysis vapor stream to the

condenser. Th e latter procedure is mostly applied in research

projects. Results of laboratory investigations are, until now,

quite poor regarding the understanding of what is actually

taking place. Mostly, the approach is to carry out catalytic

pyrolysis experiments for selected types of biomass and ana-

lyze the liquids obtained for their contents of alkanes –

alkenes and aromatics. Yield numbers remain oft en unclear,

but obviously the yield of the desired liquid is far below the

theoretically maximum possible value. Signifi cant amounts

of coke, water, and carbon oxides are produced.29,30

Fast pyrolysis technologies

Th e aim of the slow pyrolysis process is to produce mainly

charcoal, whereas the fast pyrolysis process should convert

the biomass into a maximum quantity of liquid. As men-

tioned in the introduction, both processes have in common

that the energy in the biomass feedstock is concentrated in a

smaller volume by which transport costs and storage space

can be reduced. Also benefi cial is that a more uniform, stable,

and cleaner-burning product is obtained, that could serve as

an intermediate energy carrier and feedstock for subsequent

processing. In an industrial process, the byproducts char or

gas (both 10 to 20 wt.%) would be used primarily as a fuel for

the generation of the required process heat (including feed-

stock drying). But the byproducts left could also be applied

otherwise. Active carbon, carbon black, or a pelletized fuel

could be produced from the char. Th e char is also proposed

as a soil improver (‘biochar’).31 For specifi c purposes, such as

entrained fl ow gasifi cation (syngas production), recombina-

tion to a char-oil slurry is sometimes considered. Th e gaseous

byproduct, essentially a mixture of CO and CO2, could also

be used for electricity production in an engine, if cleaned

properly. Apart from possible fl ue gas emissions resulting

from the char combustion, there are no waste streams. Th e

ash in the original biomass will be largely concentrated in the

char product and is separated when the char is combusted in

the process for drying and heating the biomass feed stream.

It allows recycling of the minerals as a natural fertilizer to

the site where the biomass was grown originally.

Th e essential characteristics of a fast pyrolysis reactor for

maximal oil production are the very rapid heating of the bio-

mass, an operating temperature around 500°C, and a rapid

quenching of the produced vapors. Crucial in the pyrolysis



reactor is the ability to have high heat transfer rates to (and

preferably also inside) the solid particles. On an average

basis, it can be estimated that 1.5 MJ/kg is required, mainly

for heating the biomass (which is 2/3 of that of water evapo-

ration). Moreover, the time and temperature profi les of the

vapors produced aff ect the composition of the oil as well. In

small laboratory reactors, where very rapid transfer rates are

achieved, and vapor residence times of only a few tenths of a

second can be realized, oil yield can be maximized. For heat

transfer limited systems, and longer residence times of the

pyrolysis vapors at higher temperatures, occurring especially

in real-scale installations, the consequences of secondary

cracking can become quite signifi cant. In practice, high

external medium-to-solid heat transfer rates are required

(say >500 W/m2K); intraparticle biomass heat transfer limita-

tion should be avoided (requiring particles <5 mm to limit

the heat penetration depth); and vapor phase residence times

should be kept below a few seconds in order to maintain

the oil yield. In case a pyrolysis plant is meant to produce a

liquid fuel for combustion or gasifi cation, the process could

be designed in a way that maximizes the energy conversion

to the liquid product. When the bio-oil product is meant to

derive biofuels or chemicals from it, however, factors other

than just the vapor residence time should also be considered.

Th e composition of the oil can be steered by process condi-

tions, equipment dimensions, and the application of catalysis.

Th e latter is an ongoing research item at various laboratories.

Although laboratory studies regarding the thermal decom-

position of various organic substances have been carried

out for a much longer period, the technology development

of ‘fl ash’ and ‘fast’ pyrolysis started only some 20 years ago

when the advantages of liquefying biomass in such a simple

way were gradually recognized. During the 1980s and the

early 1990s, research was focused on the development of

special reactors, such as the vortex reactor, rotating blades

reactor, rotating cone reactor, cyclone reactor, transported

bed reactor, vacuum reactor, and the fl uid bed reactor. Since

the late 1990s, the process realization emerged, resulting in

the construction of pilot plants in Spain (Union Fenosa),

Italy (Enel), UK (Wellman), Canada (Pyrovac, Dynamotive),

Finland (Fortum) and the Netherlands (BTG). In the USA

and Canada, Ensyn’s entrained fl ow bed process is applied

at a scale of around 1 ton/hr for commercial production of

a food fl avor called ‘liquid smoke’. Dynamotive and BTG

designed and operated demonstration installations of 2 to 4

tons’ biomass throughput per hour for utilization of bio-oil

in energy production primarily. Meanwhile, many pilot-

plant projects stopped, sooner or later aft er the initial test-

ing. At the time of writing, the plants of Union Fenosa, Enel,

Wellman, Fortum, and Pyrovac’s large-scale installation in

Jonquiere, Canada, are no longer in operation. Th is may be

caused by a lack of confi dence in economic prospects and

markets at the time, or by legislative limitations.

On refl ection, the state-of-the-art in the development of

pyrolysis seems comparable with the situation in 1936 in

the petrochemical industry, when the Houdry (FCC) proc-

ess was fi rst demonstrated on a scale of 2000 barrels per

day. Today, fast pyrolysis is attracting increased interest.

Fields of science other than engineering (e.g., agriculture,

organic chemistry, catalysis, separation technology) are

getting involved, together accelerating the developments in

bio-oil applications. Oil companies and food/feed industries

are building biofuel departments and looking for existing

knowledge matching their strategies and targets regarding

renewable resources. Also, new developers of fast pyrolysis

technology are showing their intentions with the construc-

tion of pilot plants based on proprietary technology. A selec-

tion of historical developments aiming at demonstration-

scale pyrolysis technologies will be discussed later, with an

emphasis on the rotating cone technology of BTG.

Entrained down-fl ow

Early attempts of fast pyrolysis have been carried out in

entrained fl ow reactors, where biomass particles (1 to 5 mm)

were fed to a hot, down-fl ow reactor. Th e reaction was sup-

posedly complete within a residence time <1 s, if the reactor

tube was held at temperatures in between 700 and 800°C.

Unlike many other fast pyrolysis reactors later on, no extra

hot solid material was used to transport and heat the bio-

mass particles. An early process was developed at Georgia

Tech Research Institute (USA) and a fi rst unit transferred to

Egemin (Belgium) for further development and scale up in

a project funded partly by the European Commission. Th e

plant was dismantled in 1993. It appeared that the feedstock

was incompletely pyrolyzed in the reactor, particularly when

larger particles were used (6 mm). Insuffi cient heat transfer



to the solid biomass particles during their short travel in the

reactor was the likely cause.32 Th is resulted in total liquids

yields of less than 40 wt.% on dry feed basis.

Ablative reactor

Ablative pyrolysis was then considered as a possible alter-

native to entrained fl ow reactors. Th e principle of ablative

pyrolysis is given in Fig. 6. Th e surface, heated by hot fl ue

gas, is rotating, and biomass is pressed onto the hot surface

(approx. 600°C). Th e fl ue gas is produced by combustion

of pyrolysis gases and/or produced char. In the 1990s, BBC

(Canada) demonstrated an ablative fl ash pyrolysis technique

for the disposal of tires at a 10–25 kg/hr capacity.33

Th e process was licensed to Castle Capital (a holding com-

pany of several companies involved in the fabrication of

pressure vessels, military vehicle modifi cation etc.) for the

erection of a 50 t/d plant in Halifax, Nova Scotia, using solid

wastes. It is unknown to the authors what happened to these

plans. Much of the pioneering work on ablative pyrolysis was

carried out by NREL (Golden, CO, USA, formerly known as

SERI).34 In their approach, the forces to press the biomass to a

hot surface are centrifugal forces in a so-called vortex reactor.

Th e biomass, though, appeared to be insuffi ciently converted,

requiring the solids to be redirected back to the entrance. In

1989, NREL entered a consortium with Interchem Industries

Inc. (USA) to develop and exploit NREL’s ablative pyrolysis

for the production of phenol adhesives and alternative fuels.

However, the construction of a demonstration plant was never

completed. NREL is no longer contractually involved with

this fi rm, and abandoned the vortex design concept in 1997.

oil

excess

Biomass

recycle gas

char / sand loop

char combustionsand

air

gas

biomass

Oil

hot disc

air

char gas

gas

oil

vacuumvessel

char

biomass

molten salt

air

biomassbiomass

fluid bed

recyclegasrecyclegas

char excessexcess

oil

air

gas

char

sand loop

biomass

oil

air

gassand

oil

char combustion

char/sand loopbiomass

sand

air

air

gas

(a) (b)

Figure 6A. Ablative, CFB and vacuum technologies. Figure 6B. Fluid bed, screw (auger) and rotating

cone technologies.



In the 1990s, Aston University (Birmingham, UK), built

and tested a prototype rotating blade reactor for ablative

pyrolysis on a small scale of 3 kg/hr.35 Oil samples were pro-

duced in yields of up to 80 wt.%.

Work in a cyclone reactor continued at CNRS (France) in

this century, yielding up to 74% bio-oil.36 At present, the

German company Pytec is the only company developing

ablative pyrolysis technology, with a pilot plant of 250 kg/hr

in operation near Hamburg and plans for demonstration of a

2 t/hr unit in Mecklenburg-Vorpommern.37–39

In general the following limitations for this ablative tech-

nology can be expected:

• Limited heat transfer rates to the hot surface due to the

indirect heating principle. Th is is caused by both a rela-

tively small temperature diff erence between hot fl ue gas

(likely around 800oC) and the pyrolysis reactor (around

500oC), and a low heat transfer coeffi cient. Experiments

in which electricity is used as a heat source (quite usual

in R&D work) can be misleading in designing a large

scale process.

• Restrictions in feedstock morphology (particle shape,

structure and density), the particle size and its free-fl ow-

ing characteristics, because the material needs to be

pressed against a hot surface.

Finally, although strictly speaking not an ablative type

of reactor, TNO operated a 30 kg/hr pilot plant at the

University of Twente in the Netherlands. Th eir PyRos-

reactor integrates pyrolysis and high- temperature gas

cleaning in one unit, and consists of a cyclone with a rotat-

ing particle separator. Th e biomass is fed to the cyclone

by an inert transport gas, with a solid acting as a heat car-

rier.40 So far no yield data have been presented in the open

literature.

Fluid bed

A simple method for the rapid heating of biomass parti-

cles is to mix them with the moving sand particles of a

high- temperature fl uid bed. High heat transfer rates can be

achieved, as the bed usually contains small sand particles,

generally about 250 μm. Th e heat required is generated by

combustion of the pyrolysis gases, and/or char, and is even-

tually transferred to the fl uid bed by heating coils. While the

sand-to-biomass heat transfer is excellent (over 500 W/m2K),

the heat transfer from the heating coils to the fl uid bed will

be low, due to the resistance inside the coils (gas-to-coil wall

heat transfer estimated 100–200 W/m2K), and the limiting

driving force of around 300°C as a maximum (800 down

to 600°C in coils versus 500–550°C in the fl uid bed). In an

optimistic case, at least 10 to 20 m2 surface area is required

per ton/hr of biomass fed.

Th e University of Waterloo in Canada reported early work

in the beginning of the 1980s on fl uid bed pyrolysis. In 1990,

a 200 kg/hr demonstration plant was built by Union Fenosa, a

utility company in Spain for the generation, transmission, and

distribution of electricity. (Th is plant has since been disman-

tled).41 Another 200 kg/hr fl uid bed system was developed and

constructed by Enco Enterprises Inc., on a standard trailer in

the late 1980s and early 1990s, with the intention of convert-

ing peat moss into liquid fuel. Aft er extensive testing, how-

ever, it became clear that scale-up opportunities were limited.

Th e project was abandoned and a heated auger system (screw

conveyor) was adopted (more details later in this review).42

Th e Canadian company, Dynamotive Corporation, how-

ever, further commercialized the fl uid bed technology of the

University of Waterloo. Design and development of the fi rst

commercial plant at West Lorne started in 2002. A process

scheme has been presented elsewhere.43,44 Th e plant started

operations in early February 2005 with a design capacity of

100 tons per day of waste sawdust. At the beginning of 2008,

the plant was not in full production and did not reach the

designed bio-oil production capacity, presumably due to

design and construction problems. Th e company did not wait

for the West Lorne plant and started to build a second plant

in Guelph in 2006 with a design capacity of 200 tons per day.

Operational performances for both plants cannot be found

in the open literature. Th e oil of the West Lorne facility was

meant originally for combustion in Orenda’s GT 2500 gas

turbine to produce electricity. Figure 7 shows a photograph

of the West-Lorne plant. Th e Orenda turbine is an industrial

Mashproekt-designed engine, with nine axial, and one radial

stage compressor. Due to variations in the oil quality (per-

haps off -spec) and a limited supply of the oil, the turbine has

hardly been used.

Scale-up of fl uid bed pyrolysis seems thus limited in case

the heat is indirectly transferred through submersed coils



or alike, as in the case of Dynamotive, but the application

of a twin fl uid bed with solids exchange (and separation of

biomass pyrolysis and char-gas combustion) could solve

this problem (see Table 3 for mass and energy balances

of a twin fl uid bed process). An example of such a sys-

tem is a fl uid bed designed and constructed by Wellman

Process Engineering in an EU project coordinated by Aston

University in Birmingham, UK. Here the fl uid bed was sur-

rounded by the char combustor, with heat transfer through

the separating wall and by exchange of solids. Th e con-

struction of the pilot plant was indeed completed in early

1999, but due to permit problems it could never be started.

Th e installation has not been used since then. Biomass

Engineering Ltd is now erecting a similar (250 kg/hr)

installation. Improvements include the collection system

for the pyrolysis liquids, hot gas fi ltration, the handling and

combustion of the char (to provide process energy inter-

nally), and so on.

Vapo Oil and Fortum Oil together undertook another

initiative from 2001 onwards and developed a new-patented

principle for fl uid bed pyrolysis.45 Pictures of the plant are

presented in Fig. 8. Th e project was abandoned, presumably

for economic reasons.

Although fl uid bed operation seems a pretty well under-

stood technology for pyrolysis,1 experimental experiences

in fl uid bed pyrolysis still indicate a number of serious

technical problems to overcome. Referring to Fig. 6, the fol-

lowing remarks can be made:

• Cyclones are applied to separate the vapors and solids,

and thus avoid large quantities of char and bed mate-

rial ending up in the bio-oil. Char fi nes in the oil cause

increased instability, problems in pumping and, more

importantly, diffi culties in the end-use applications (tur-

bines, engines, boilers). Perhaps because of the diffi culties

in char separation, as the char is pyrophobic and easily

catches fi re, Dynamotive now deliberately leaves the char

in the oil, naming it Intermediate Bio-oil (BioOil Plus).46

Figure 7. The Dynamotive’s West Lorne plant: wood

feed hopper on left, char product hopper on right.

Figure 8. ForesteraTM pilot: reactor and product

storage area.



Table 3. Mass and energy balances for 2000 kg/hr (daf) fluid bed pyrolysis.

Input Reactor feed

Air Inert carrier gas

Pyrolysis vapours

Heat carrier

Bio-oil Surplus char

Flue gas

Stream no. 1 2 3 4 5 6 7 8 9

Organics 2,000 2,000 1200

Water 900 200 400 740

O2 600 250

N2 2,400 2,400

CO2 470

Pyrolysis gas 1,000 200

Char 155 245

Ash 20 20 12 8

Sand 66,600

Total 2,920 2,220 3,000 1,000 1,200 257 3,860

Temp. (K) 291 291 291 326 304 804 317 804 400

Pressure (bar) 1 1 1 1.2 1 1 1 1 1

Chem. heat (MW) 10.4 10.4 2.1 0.42 1.3 6.6 2.1

Heat (MW) 7.41 0.05 0.33

Water

70x70x70 mm30% moisture

<10 mm<10 %moisture

1

2

3

4

5

6

7

8

9

• Inert gas used for fl uidization of the reactor bed logically

is the non-condensable part of the pyrolysis gas. Th is gas

needs to be reheated and compressed, which requires

careful cleaning to avoid blockage of heat exchangers,

blowers, etc. For comparison, gas cleaning appears to be

one of the main hurdles in ‘conventional’ gasifi cation. A

similar problem can be noted for CFB operation.

Circulating fl uid bed (CFB)

The first CFB process was developed at the University

of Western Ontario in the late 1970s and early 1980s.

Biomass could be converted to bio-oil at yields of over

70 wt.%. The principle is shown in Fig. 6: Biomass is

screwed into a (riser or fast f luidization) reactor, where

extensive contacting between inert particles (sand) and

biomass takes place. Together with the char, sand is

entrained out of the reactor, and sent to a combustor

chamber where the char is combusted. The main

advantage of the CFB system compared to f luid bed

and ablative is the direct heat supply to the biomass by

recirculation of sand, reheated by combustion of

pyrolysis char.



In the beginning of the 1990s, Ensyn Technologies Inc., in

Ottawa, Canada developed industrial applications for their

so-called Rapid Th ermal Processing (RTP), in which woody

biomass is converted to pyrolysis liquids as a source of valu-

able chemicals and fuel.47–49 Commercialization was enabled

through the granting of an exclusive license to Red Arrow

Food Products Company Ltd of Wisconsin for certain prod-

uct applications in the food industry, mainly wood fl avors.

Ensyn and Red Arrow have been producing large quanti-

ties of bio-oil for the production of specialty products since

1990. A fairly large circulating fl uid bed pilot plant of 625

kg/hr throughput capacity has been built in Bastardo, Italy.

Th e plant is stated ‘running on demand’, which is in practice

never or hardly ever.50 A 100 kg/hr, a 40 kg/hr and a 10 kg/

hr R&D unit were available at Ensyn’s site in Ottawa, while a

20 kg/hr PDU-unit is built for research purposes at VTT in

Finland. By 2007, eight RTP™ plants are in commercial oper-

ation, ranging from 1 to 100 t/day. Details of the operation

(or operational performance) are unknown. Recently, Ensyn

went into a Joint Venture (JV) with UOP, under the name

Envergent Technologies LCC to commercialize the pyrolysis

technology for fuel substitution and electricity generation.51

Another JV was announced by Finnish companies Metso

and UPM to develop bio-oil production combined with

(preferably existing) CFB biomass combustion units. Th e

technology is based on the integration of conventional bio-

mass-based fl uidized bed boilers with a (non-disclosed)

pyrolysis reactor (in cooperation with VTT). Th e pyrolysis

unit utilizes the circulating hot sand from the boiler as a

heat source. Test production will begin at Metso’s test unit in

Tampere (Finland) in June 2009.52

Just as for fl uid bed technology, CFB technology is said to be

well understood,42 but actual operation in pyrolysis appears

problematic with substantial erosion problems, and complica-

tions due to operational parameters, such as the use of seals

between various vessels (‘dip legs’). Th is is a problem solved

in the chemical industry (see FCC Fluid Catalytic Cracking

processes). Nevertheless, and similar to fl uid bed technology,

the large amounts of gas needed for fl uidization of the reactor

should be the non-condensable part of the pyrolysis gas. Th is

gas must be reheated and compressed, which requires careful

cleaning to avoid blockage of heat exchangers, blowers, etc.

As already mentioned, gas cleaning appears to be one of the

main hurdles in ‘conventional’ gasifi cation. Finally, for real-

izing the rather low solids hold-up in riser systems at solid

fl uxes of 100 to 200 kg/m2s, the gas fl ow rate in the riser is

high, in the order of 1000 m3/hr (ton/hr biomass).

Vacuum moving bed

Th e biomass vacuum pyrolysis process was developed by Roy

at Pyrovac Institute Inc., between 1988 and 2002, aft er hav-

ing carried out R&D work at the Université de Sherbrooke

(1981–1985) and Université Laval (1985–1988), both in

Canada. Th e process includes a combination of slow and

fast pyrolysis conditions. Course solids are heated relatively

slowly to temperatures higher than those of slow pyrolysis,

while the gas is removed from the hot temperature zone rela-

tively quickly by applying a reduced pressure of less than 20

kPa in the process. An attempt to commercialize the proc-

ess was carried out by Pyrovac International by the end of

the 1990s. In this concept, biomass material was conveyed

over a long horizontal grate, which was heated indirectly by

a mixture of molten salts composed of potassium nitrate,

sodium nitrite, and sodium nitrate.53 Th e salt itself was

heated by a gas burner fed with the non-condensable gases

produced by the pyrolysis process. Limitations in external

heat transfer were avoided as much as possible by applying

a patented internal agitation raking mechanism, but obvi-

ously internal heat transfer limitations could not be avoided.

A 3.0 t/hr demonstration plant for bark residues was erected

in the City of Saguenay Quebec, Canada, and taken into

operation between 1999 and 2002 (Fig. 9). Crumb rubber

was also successfully tested in 2002. Th e system proved to be

functional except for a limitation encountered at one of the

two condensing packed towers, which somehow impaired

the organic vapor condensing performance. Th at condens-

ing tower was on its way to be modifi ed when the joint

partner of Pyrovac Group, UNA B.V. (now NUON B.V.) was

acquired by Reliant Energy Inc., in Houston. Because the

pyrolysis project did not fi t into the core business, Reliant

Energy withdrew from the JV and Pyrovac was subsequently

brought to an end in June 2002. Th e process equipment

and the building were purchased by a third party from the

bankruptcy in 2003 and the assets were kept in good shape

till today. A group of investors led by US-based NewEarth

Renewable Energy is now planning to restart the plant in



2010 for industrial production of torrefi ed wood, wood char-

coal, and bio-oils. 54–55

Auger systems

Aft er concluding that design limitations precluded the scale

up of their fl uid bed, ABRI-Tech the former Encon Enterprise

Inc., and Advanced BioRefi nery Inc. started working with

heated augers systems using a hot, high-density heat carrier

(e.g., metal beads, see Fig. 6).56 ABRI-Tech is a JV between

Advanced BioRefi nery Inc., and Forespect Inc., the latter

being a forest products company located in Namure, Quebec,

Canada. Th e system has evolved to the point where the com-

pany builds, for sale, a 1 t/d unit and a 50 t/d unit. Th e fi rst

commercial 50 t/d system is presently (October 2009) being

commissioned in Iowa and will produce bio-oil and bio-char

from agricultural residues (Fig. 10).57

Forschungszentrum Karlsruhe (FZK) developed a fast

pyrolysis reactor to convert straw to pyrolysis oil and char

to serve as a high-energy slurry feedstock for entrained

fl ow gasifi cation (the ‘Bioliq Process’).58 Th e reactor design

is adopted from the Lurgi-Ruhrgas twin-screw mixer reac-

tor that was developed decades ago for oil shale, tar sand,

and refi nery vacuum residues (LR coking). Chopped bio-

mass is mixed with hot sand in the double-screw reactor

and decomposed to vapors and char. Th e hot sand loop

is maintained pneumatically or mechanically. Just like in

BTG’s rotating cone technology, heat transfer to the bio-

mass particles takes place by intimate contact with a heat

carrier, with no need for an inert carrier gas. At the time of

Figure 9. The Pyrovac installation in Jonquière, Canada

(© Christian Roy).48

Figure 10. ABRI-Tech’s Auger pyrolysis system of 50 t/d plant. The left side of the

picture is the dryer/pulverizer. The reactor and condenser are the two modules on

the right side of the photo. (© Peter Fransham).



writing (October 2009), the construction of a 500 kg biomass

per hour pilot plant was completed but no test runs were

reported yet. Th e pilot plant uses sand as a heating medium,

whereas the R&D work was carried out while testing diff er-

ent heat carriers including metal beads.

BTG’s technology: Rotating cone reactor

Fast pyrolysis has been a continuous research item in

the Netherlands at Twente University for a few decades.

Researchers started with the principle that intense mixing

of biomass and hot inert particles is the most eff ective way

to transfer heat to the biomass, but that fl uid bed mixing

requires too much ineff ective inert carrier gas. A high-

intensity reactor for the pyrolysis of biomass was developed

where no inert gases were required, while simplifying the

reactor parts and peripheral equipment as oil condenser, gas

cleaning, etc. Th e original idea in 1989 to rely on a merely

ablative principle without inert sand was later modifi ed to

a sand-transported-bed rotating cone reactor (RCR).16 Th e

concept is depicted in Fig. 6. Instead of mixing the biomass

in a hot sand fl uidized bed driven by inert gas, the pyrolysis

reactions take place upon mechanical mixing of biomass and

sand. Similar to the CFB operation, the sand and char are

further transported to a separate fl uid bed where the com-

bustion of char takes place. Th e RCR enables a high solids

throughput and short vapor residence times.

R&D work was continued by BTG Biomass Technology

Group.59,60 Figure 11 shows the fi rst prototype built by BTG

and Royal Schelde in Vlissingen as part of a test unit for

Shenyang Agricultural University. It was shipped to China

in 1994. Biomass particles are fed near the bottom of the

rotating cone together with an excess fl ow of heat-carrier

material like sand, and then transported upwards along the

cone wall in a spiral trajectory by the centrifugal forces (up

to 600 rpm). An electrical oven in which the sand and char

are trapped surrounds the RCR. Th e produced vapors pass

through a cyclone before entering the condenser, in which

the vapors are quenched by re-circulated oil.

Subsequently, the University of Twente constructed a novel

reactor system (throughput capacity of up to 20 kg/hr).61 In

contrast with Wagenaar’s original reactor, where sand is fed

at the top, this rotating cone had a number of holes near the

bottom, through which the sand was sucked into the cone. To

compensate for heat losses and provide both the energy for

heating the biomass particles and consequently the overall

endothermic pyrolysis reactions, the char produced during

pyrolysis was burnt in a fl uid bed around the rotating cone

(‘combustion chamber’). Experiments with sand and cata-

lysts demonstrated that autothermal operation can indeed be

achieved with this system, but unfortunately, the operational

fl exibility of this advanced concept appeared to be poor,

and therefore the concept was not considered for further

scale-up.

BTG scaled up Wagenaar’s RCR technology, fi rst to about

50 kg/hr in 1997. Th e cone reactor was integrated in a circu-

lating sand system composed of a riser, a fl uid bed char com-

bustor, the RCR reactor, and a down comer. Char is burnt in

the combustor to generate the heat required for the pyrolysis

process, viz. by (re-)heating the inert sand that is re-circu-

lated to the reactor. Oil is the only product of this lab facility,

and gases were fl ared. In 2001, the system was further scaled

up to 250 kg/hr. Th rough the past ten years, about 100 tons

of bio-oil has been produced from over 50 diff erent materi-

als. Oil is partially sent out to universities, institutes, and

industrial companies for application research. In addition

Figure 11. Photo of the rotating cone reactor.



to this pilot plan installation, a smaller (5 kg/hr) test unit

has been erected in BTG’s laboratories, for quick screening

of potential feedstock materials in a continuously operated

system. Initially, researchers reported that the RCR principle

would be limited to particles < 1 mm in diameter. Over the

years, however, the method of mixing in the cone and the

pyrolysis system were improved considerably. Th e overall

effi ciency of the process has increased, the acceptable particle

diameter demonstrated to be up to 10 mm, and product qual-

ity and consistency for diff erent feedstocks have improved

considerably. In 2001, a fi rst detailed design for a 1 t/hr dia-

per sludge pyrolysis unit was prepared under a license for the

company Bio-Oil Nederland (BON). BON is now a wholly

owned subsidiary of Bio-Oil Holding N.V. (BOH), with no

relation to BTG anymore. Th e company ambitiously plans to

start up the fi rst of four, 5 t/hr installations in 2010, inten-

tionally on SRF, in Delfzijl (the Netherlands). No informa-

tion is publicly available on the technology at the moment,

but it is likely deviating from the original BTG design.

In 2004, BTG sold the world’s fi rst commercial unit of 50

t/d on so-called Empty Fruit Bunch (‘EFB’) in Malaysia.

EFB is a left over from palm oil mills. In the Malaysian

plant, EFB is taken directly from a nearby palm mill,

pressed, shredded, dried, and converted to bio-oil. From

reception to oil delivery takes about 1 h, time that is con-

sumed almost entirely by the pre-treating process. Genting

Bio-Oil Sdn Bhd (GBO) commissioned its bio-oil pilot plant

in Ayer Itam, Johor (Malaysia) in 2006. While undergoing

some engineering upgrades, full commercial operations

have been targeted for 2009.62

Th e overall (simplifi ed) scheme is given in Fig. 12, which

includes the complete chain from EFB reception, storage,

pre-treatment (pressing, shredding, and drying), storage

(approx. 4 ton) and conversion.63 Th e heat required for the

drying of the EFB (up to 70 wt.% moisture upon reception)

is taken from the pyrolysis unit, and a steam production

system is fully integrated in the production unit. Pictures

of the EFB and the pyrolysis plant are given in Fig. 13. From

mid-2005, the plant was running on a daily basis, showing

the potentials but also the shortcomings. Th e main achieve-

ments over the last years are listed below:

■ Over a thousand tons of bio-oil have been produced from

more than 5000 tons of wet (up to 70% moisture) EFB.

■ Oil was co-fi red, replacing conventional diesel in a sys-

tem located 300 km from the site.

■ Drying of de-watered EFB from around 50 wt.% mois-

ture content down to 5 wt.% is possible using the excess

heat from the pyrolysis process.

1st

shredder

press

2e shredder

drier

storage

feeding

boiler

reactor

combustor

condensor

flare

EFBbucket elevator

oil cooling

Oil

water

ash

air

airair

air

Flue gas

wetair

Flue gas

Figure 12. Process scheme of the Malaysian plant for fast pyrolysis of empty fruit bunches

(EFB), including the complete chain from EFB reception, storage, pre-treatment and

conversion.



■ Oil quality can be controlled by tuning operation

conditions.

■ Any excess energy recovered from the process aft er dry-

ing of the (very) wet EFB, is potentially available to gen-

erate electricity for on site use.

■ Th e maximum capacity of the plant achieved is about 1.7

t/hr (design capacity was 2 t/hr).

■ No problems were observed in the actual reactor system.

Technical problems related to erosion due to high-veloc-

ity sand in riser parts, cyclones, etc., could be overcome.

Some shortcomings of the system at that time should be

mentioned as well:

■ Fluid bed combustion of the char from EFB created a

high risk of blockages, but this appeared to be due spe-

cifi cally to the nature of the EFB ash (a.o. low melting

point).

■ Considerable pre-treatment of EFB is necessary and wear

and tear in pre-treatment equipment is signifi cant (for

instance in presses and shredders).

■ By average, the bio-oil yield on basis of good quality dried

EFB is 50 to 60 wt.%; the oil yield is lower than for wood

(typically 70 wt.%), while the water content is higher.

■ Th e supply rate of wet EFB from the palm mill and the

quality varies considerably, viz. from 2–2.5 t/hr during

daylight, and 1–1.5 t/hr at night with moisture contents

up to 70%. Large amounts of EFB cannot be stored easily.

■ Effi cient heat integration, together with an improved reli-

ability of the total system, will be the challenge for the

coming years.

Considering the status of the pyrolysis process at the begin-

ning of the plant design in 2004, the progress made in

Malaysia is signifi cant. From an initial set of experiments in

Figure 13. EFB Material and the pyrolysis plant in Malaysia. ((use separate

powerpoint fi le))



2003 (8 hours and maximum 100 kg/hr feeding), the system

has been scaled up to a 24 hours/day running factory, where a

direct link has been established between the palm mill and the

pyrolysis plant. In addition, the process is applied here for a

diffi cult type of biomass feedstock as, next to being fl uff y and

wet, the ash of the mineral-rich feed has a very low melting

point (below 650oC). In 2007, BTG established BTG-Bioliquids

with the objective to commercialize the technology.64 On

its website, BTG-Bioliquids announces the construction by

Empyro BV of a 5 t/hr wood based pyrolysis installation on a

site in Hengelo (the Netherlands). Th e project is supported by

the 7th framework program of the European Commission.

Bio-oil fuel applications

A key advantage of producing liquids from biomass is that

its production can be de-coupled in time, scale, and place

from the fi nal application. Th e current status of primary,

secondary, and tertiary processing of pyrolysis liquids is

further presented in Table 4. Due to its high oxygen con-

tent and the presence of a signifi cant portion of water, the

heating value of bio-oil is much lower than for fossil fuel.

Nevertheless, fl ame combustion tests showed that fast

pyrolysis oils can replace heavy and light fuel oils in indus-

trial boiler applications. In its combustion characteristics,

the oil is more similar to light fuel oil, although signifi cant

diff erences in ignition, viscosity, energy content, stability,

pH, and emission levels are observed. Problems identifi ed

in fl ame combustion of bio-oil are related to these deviating

characteristics, but can be overcome in practice. Meanwhile,

bio-oil has been used commercially to co-fi re a coal utility

boiler for power generation at Manitowoc Public Utilities

in Wisconsin (USA). It has also been approved as a fuel for

utility boilers in Swedish district heating applications. Aft er

an extensive boiler test program in Sweden in 1996 and

1997, a commercial project was said to commence in 1998,

but results were never reported. A successful co-fi ring test

with 15 tons of bio-oil was conducted in 2002 in a 350 MWe

natural-gas fi red power station in the Netherlands.65 Some

data are presented in Fig. 14 and Table 5. A four-hour co-fi r-

ing session was carried out at a bio-oil throughput of 1.6 m3

per hour, or an equivalent of 8 MWth. While co-fi ring bio-oil

in the boiler, the power output setting of the plant remained

constant at about 250 MWe, and in the test, the plant control

reduced the natural gas fl ow to the boiler to compensate

for the injection of thermal heat of the bio-oil, indeed cor-

responding to 8 MWLHV (Fig. 14). Low ash deposition rates

were also reported recently for a 100 kW combustion test rig

during the combustion of bio-oil.66

Oil from the Malaysian plant was routinely used to replace

expensive diesel for start-up of a fl uid bed combustor near

Kuala Lumpur International Airport. No results have been

reported in open literature. Since 2006, BTG has been

actively involved in research on the combustion of the oil

in a standard 250 kW hot water generation unit, to replace

Table 4. The status of primary, secondary and tertiary processing of pyrolysis products.

Primary product

Secondary processing

Secondary product

Tertiary processing

Final product

Liquid TransportCombustion2

Engine/turbine1

Stabilization2

Upgrading2

Extraction1,5

Conversion3

Conversion2

FuelHeat/steamElectricity

Stabilized oilHydrocarbons

ChemicalsChemicals

Gas

CombustionSteam turbine5

Engine/turbine1

Refi ning2

Refi ning1,5

Refi ning1,2

Fuel cell1

Heat/steam/electricityElectricity

ElectricityDiesel/gasoline

ChemicalsChemicalsElectricity

Gas CombustionEngine/turbine3

Fuel cell1

Heat/steamElectricityElectricity

Steam turbine Slectricity

(Bio)char transportcombustion5

slurrying2

active carbon

FuelHeat/powerLiquid fuel

Combustion

Combustion3

Heat/steam/electricity

Heat/power

Indices: 1 = conceptual, 2 = laboratory, 3 = pilot, 4 = demonstration and 5 = commercial (Bridgwater, 1997). Indicated in bold are the most promising options on a short time scale.



diesel and/or natural gas. For this purpose, suffi cient quanti-

ties of palm-derived oil from Malaysia were transported to

the Netherlands, and a dedicated oil lance was developed.

Results derived from a Dutch cooperation between BTG,

Stork-Th ermeq, and ECN will be reported in due course.

A subsequent project is concerned with testing in a larger,

commercial boiler set-up, owned by Stork Th ermeq.

Generally, the production of electricity is more interesting

than the production of heat because of the higher added value

of electricity, and its ease of distribution and marketing.

Diesel engines are relatively insensitive to the contaminants

present in pyrolysis oils, especially in the case of large- and

medium-scale engines, and bio-oil may be used. Tests have

been performed by diesel (re)manufacturers like Ormrod

Diesels and Wärtsilä Diesel, in collaboration with research

institutes such as Aston University, VTT, MIT, and the

University of Rostock.67–69 A review of diesel engines was pre-

pared in 2001 by one of the present authors,70 while another,

very useful review on use of such liquids in engines and gas

turbines was prepared by the University of Florence and VTT

in 2007.21 In general, diesel engine development and testing suf-

fers from insuffi cient quantities of available bio-oil and a lack

HC62 HC61

4

2

3

5

1

6

Air

NG

NG

Gasturbine

Steamcycle

NOx

Flue gas

Bio-oil

90 MWe

Boiler Stack

Stack losses

33 MWth

River cooling

274 MWth

161MWe

Figure 14. Mass and energy balances of Harculo power station.

Table 5. Power settings of Harculo station and the M&E balance as indicated in Fig. 13.

Input Output Effi ciency

Natural gas [MWth]

Heat [MWth]

Power [MWe]

Heat [MWth]

Gas turbine HC62 293.5 0 89.6 203.9 @ 520°C 30.5

Steam turbine HC61 264.8 133.8 161.5 237.2 @ 30°C 40.5

Total plant HC60 558.3 0 251.0 307.2 @ 30°C 45.0

* natural gas LHV: 35.57 MJ/Nm3; 0.808 kg/Nm3

Stream I.D. 1 2 3 4 5 6

Air Natural gas Flue gas Flue gas Natural gas Flue gas

T [°C] 16 16 520 520 16 97

P [bar] 1.00 8.45 1.06 1.02 1.2 1.00

Flow [kg/s] 358 6.67 365 123 6.02 371

LHV [MWth] 0 293.45 0 0 264.80 0

Heat [MWth] 0 0 203.88 70.05 0 32.65



of interest from engine (parts) manufacturers. Nevertheless,

the results obtained indicate that engine deterioration can be

a serious problem. Traditional diesel engines are designed to

operate on acid free fuels, however, and all engine components

are manufactured in such a way and with such (steel) materi-

als as to comply with these fossil fuels. For fast pyrolysis oils,

severe wear and erosion was observed in the injection needles,

due to the fuel’s acidity and the presence of abrasive particles.

Nozzles lasted longer when fi ltered oil was used, but it is clear

that standard nozzle materials are inadequate.

A high bio-oil viscosity and stability loss with rising tem-

peratures are other major problems, already referred to.

Damages to nozzles and injection systems, and buildup of

carbon deposits in the combustion chamber and the exhaust

valves are reported. Engines with larger cylinder bores (i.e.,

medium- and low-speed engines) are expected to be the most

suitable because of less stringent construction tolerances.

For smaller bore engines, reduction of the oil viscosity could

be needed. Injection modifi cation and/or a high turbulence

combustion chamber are required. Because the bio-oil has

poor ignition properties (cetane index below 10), it should be

enriched by the addition of cetane improvers, and the appli-

cation of a dual fuel system is most appropriate.71 Self-clean-

ing injectors are possibly required. At the end of the 1990s,

Wärtsila stopped the development work, mainly due to lack

of quality and quantity of pyrolysis oils at that time. In spite

of all these problems, it has also been reported that modifi -

cations to both the bio-oil and the engine can make pyrolysis

oils quite acceptable for diesels. One option to reduce the

need for adaptations is the use of emulsions of bio-oil in

diesel.72 Th is would not only off er prospects for stand-alone

electricity production units, but potentially for the applica-

tion of fast pyrolysis oils in the transportation sector (ships,

trucks, tractors, or busses) in the future. A substantial RTD

eff ort with involvement of manufacturers is required to

realize this application. German researchers reported a suc-

cessful 12-hour run on bio-oil in a modifi ed Mercedes diesel

engine, but exact details are not provided.73 Recently, a joint

project started between European and Russian partners to

further develop a pyrolysis oil diesel engine/turbine.74

Experience with bio-oil combustion in gas turbines is also

limited. R&D projects known were carried out by Orenda

Division of Magellan Aerospace Corporation (Canada),

ENEL Th ermal Research Center (Italy), and Rostock

University (Germany). Orenda is actively searching for

opportunities to run their Orenda GT2500 on pyrolysis oils.

Th e GT2500 uses diesel oil and /or kerosene, and unlike

aero-derived turbines, an external silo-type combustor is

adopted. Th is chamber provides a ready access to the main

components. Several modifi cations are reported necessary:75

• A complete low-pressure bio-oil supply system, including

preheating and fi ltering of the bio-oil.

• An improved bio-oil nozzle design to allow larger fuel

fl ows and dual fuel operation.

• Redesign of the hot section, including section vanes and

blades.

• Stainless steel part and modifi cation of polymeric

components.

Despite their involvement in Dynamotive’s plant in West

Lorne, the non-availability of suffi cient quality bio-oil still

remains the main reason why limited test runs are carried

out by Orenda. A 75 kWe nominal gas turbine was tested

in dual fuel mode by Rostock University in 2001, showing

deposits in the combustion chamber and turbine blades, and

higher emissions of CO and hydrocarbons.

Bio-oil may have another suitable end-application, viz.

its use as a fuel for gasifi cation. It should be noted here that

in refi neries, gasifi cation (next to combustion) is merely an

end-of-pipe technique, using (cheap) feedstocks that cannot

be used elsewhere in the process. Regarding co-gasifi cation

of biomass residues, to produce syngas for further process-

ing (e.g., methanol, Fischer-Tropsch), pyrolysis could play

an important role as a pre-treatment technique, facilitating

the cheaper transport and handling of biomass feedstocks

from origin to the site of gasifi cation, over distances that

biomass can never be shipped economically. Use of the

oil in entrained fl ow gasifi cation is the main application

Forschungszentrum Karlsruhe (FZK) is aiming at.76 Residue

gasifi ers can indeed be fed on bio-oil,77 and issues of concern

are mainly the pH (feed train) and alkaline ash content.

R&D on (small-scale) entrained fl ow gasifi cation of pyrolysis

oils have only been reported by BTG.78 Pilot experiments

were performed by BTG in UET’s (now Choren’s) entrained

fl ow gasifi er in Freiberg (Germany) at about 500 kWth with

pure oxygen (results are not yet published).



Pyrolysis oils could be used (pure or modifi ed) as a feed-

stock for the production of transport fuel, which, with

the growing demand for biofuels, is of strong interest

worldwide. Th e simplest use of bio-oil may be in the diesel

engine, but, as stated before, the oil as such is not suitable

for (even) a stationary diesel engine. Upgrading of the oil to

products more appropriate for further use is being consid-

ered by many organizations and institutes. Bio-oils can be

‘upgraded’ atmospherically using conventional FCC catalyst

in the pyrolysis process itself or, at elevated pressures, by

hydrotreating. FCC-like upgrading dates back to the 1990s,

assuming that, similar to ZSM-5 processing, oxygen will be

rejected from the oil’s structure as CO2.27,79-81 Th ey showed

a limited yield of hydrocarbon products of about 20 wt.% (or

40% energetically), mainly due to charring – coking of the

feed.

Hydrotreating of bio-oils was reviewed in 2007.82 Th e early

work originates from the late 1980s, using the slow pyrolysis

oils derived from carbonization or hydrothermal liquefac-

tion processes.83 It was shown that for deoxygenation, tem-

peratures in the order of 300 to 400oC, and residence times

> 1 hour are required. Until 2000, the real goal remained

unclear. Th e overall objective was to produce transportation

fuels (diesel and gasoline like components), and the target

seemed to be the reduction of the oxygen content in the oil.

Since the use, for instance, of MTBE, butanol and ethanol in

transportation fuels, the common believe of petrorefi neries

shift ed from absolutely no oxygen in the pool of transporta-

tion fuels, toward allowing adding small amounts of oxygen

in the appropriate functionality (viz. alcohol or/and ethers).

Since 2000, the intention of upgrading bio-oil shift ed from

using it directly as a transportation fuel (or blending compo-

nent) aft er upgrading, toward co-refi ning upgraded bio-oil

together with crude oil (derivatives). Hydrotreated bio-oils

can be well co-refi ned, with rather high effi ciencies, in FCC

processes.84 Th e University of Groningen carried out pio-

neering work on the hydrotreatment of such oils revealing

a clear resemblance with hydroprocessing of sugars.85 Aft er

an almost 10-year period of reduced activities in this hydro-

processing area, the concept has recently attracted consider-

able interest again in the USA.86

Co-refi ning in standard refi neries is the main subject

matter of a large European project BioCoup, showing that

‘upgraded’ pyrolysis oils are suitable for such co-refi ning

(Fig. 15). First papers were presented recently.87,88 Complex

chemical process engineering factors such as water evapo-

ration, dry-out phenomena, mass transfer, reaction kinet-

ics, and occurrence of parallel and consecutive reactions,

are considered and quite some progress has been made.

Successful co-refi ning of upgraded oils has been demon-

strated at lab scale,89 and papers will be published in due

course.

Primary fractionation

and liquefaction

Biomassresidues

Co-processingin conventional

petroleum refineryDe-oxygenation

Hydrocarbon-rich bio-liquid

Lignin-rich bio-liquid

ConversionDerivatives of hemicelluloses and celluloses

Conventional fuels and chemicals

Oxygenated products

OVERALL BIOREFINERY CONCEPT

incorporating fractionation with liquefaction

Energyproduction

Process residues

(blending)

Figure 15. Biocoup’s concept of co-refi ning bio-oil in existing refi neries.



Other approaches to arrive at transport fuels include

alcohol treatment (analogous to vegetable oil esterifi ca-

tion), reducing the oil’s acidity and, aft er phase separation,

water content. Th ese studies, initiated by the University of

Groningen are still exploratory in nature.90

Another interesting application may be fermentation.

Anhydrosugars that are produced by pyrolysis of biomass

can be converted by hydrolysis to fermentable solutions as

well.91,92 Feasibility studies show that this route to produce

ethanol from lignocellulose may be an attractive alternative

to acid and enzymatic hydrolysis. A further acid hydrolysis

step of the levoglucosan to free sugars is necessary, and this

should take place at elevated temperatures of around 100oC.

Fermentation of pyrolyzates and their hydrolyzates is under

investigation, and there is much uncertainty about the best

process conditions and the ‘detoxifi cation’ steps required.

Attempts to sustain yeast growth in grass-derived bio-oil

hydrolyzates were unsuccessful, but success was achieved

in fermenting a hydrolyzate from a wood-derived bio-oil

and this supports the concept of pyrolysis with ethanol as

a major product. More research is required to improve the

fermentability of wood-derived bio-oil hydrolyzate and to

establish whether its fermentability is due to the chemistry

of wood vs grass pyrolysis feedstock, the pyrolysis process

conditions existent in the commercial- vs lab-scale reactor

system, or a combination of these two variables.93

Chemicals in bio-oil

Hundreds of compounds have been recognized in (GC) anal-

ysis as fragments of the basic compounds of biomass, viz. the

lignin (amongst others: phenols, eugenols, and guaiacols), and

the cellulose or hemicellulose (sugars, acetaldehyde, and for-

mic acids). Although GC analysis may not be the most appro-

priate analysis tool for bio-oil (as discussed earlier), large

fractions of acetic acid, acetol, and hydroxyacetaldehyde are

identifi ed (Table 2). Until now, approximately 40 to 50% of the

oil’s identity (excluding the water) has been revealed, but the

large, less severely cracked or de-/re-polymerized molecules

(derived from the cellulose and the lignin) in the oils can still

not be identifi ed. Figure 5 shows that all types of (oxygen)

functionalities are present: acids, sugars, alcohols, ketones,

aldehydes, phenols and their derivatives, furans and other

mixed oxygenates. Also (poly)phenols are present, sometimes

in rather high concentrations. Th ese phenolic fractions then

include phenol, eugenol, guaiacols and their derivatives, and

the so-called pyrolytic lignin (poly-phenols) representing the

water insoluble components. It is likely that this ‘pyrolytic

lignin’ which contain the fragments of the original lignin,

also contains polymerized carbohydrate (fractions).

Components in the oil interesting to consider for future

chemicals production are the carbohydrate fragments. Th ese

are sugar derivatives such as all types of anhydrosugars and

oligosaccharides, formaldehyde, furfural alcohols, hydroxy-

acetaldehyde and so on. Due to the principle of GC analysis,

in which only the ‘distillable’ components in the oil can be

identifi ed and quantifi ed, levoglucosan is usually referred

to as an important type of sugar to be isolated (Fig. 5).

However, much more sugars, approx. 30 wt.% of the oil,

must be present.

Aspects to be considered here are that the original feed-

stock, the process conditions, and condensing parameters

are of major importance for the type of chemicals in the oil.

To complicate this further, pre-treatment of wood may result

in an increase of one particular component at the expense

of the other. As mentioned earlier, ash is known to infl uence

the reactions in the pyrolysis process, and may contribute to

higher yields of certain products.

Last but not least, the analysis of bio-oil can also be compli-

cated. As an example of the observed variation in composi-

tion of wood-derived pyrolysis oil, a study published in 1997

indicated that not a single compound of over 100 has been

identifi ed by all 10 laboratories where pyrolysis oils could be

analyzed at that time.94 It should also be repeated that GC

analysis does not reveal the identity or quantity of compo-

nents that do not evaporate in the injection system. Part of the

‘discrepancy’ in the analysis results of the various laboratories

is also caused by the continuous improvements in method-

ology during the last few years, (increased use of GC-MS,

HPLC, etc.) and the factual identifi cation of the various com-

ponents. Th e applications for bio-oil are now further detailed,

starting from application of the unfractionated oil, and appli-

cations in which fractions or isolated compounds are relevant.

Unfractionated bio-oil

Resins for MDF or OSB: Work to examine the potential

use of the pyrolysis oil as a raw material in wood panel



manufacture is ongoing. Th e use of bio-oil has been inves-

tigated over the years for the replacement of formaldehyde

– phenols in resins for particleboards. Due to the high

cross-linking capability of the lignin-derived compounds

in the bio-oil, a polymer with an improved strength can be

obtained when mixed with conventional urea-formaldehyde

resins. Research in this area published at the end of 2000,

concludes that bio-oils can be used in the manufacture of

resins in phenol substitution rates up to 50%.95 A review of

the production of renewable phenol resins based on pyrolysis

products published recently, concludes that none of the phe-

nol production and fractionation techniques available allows

a complete substitution of the resin.81 Partial substitution

seems more likely though.

Fertilizers and soil conditioners: Reaction of bio-oil

with ammonia, urea, or other amino compounds produces

stable amides, amines, etc. Th ey are non-toxic to plants and

can be used as slow release organic fertilizers. Additional

benefi ts are that the lignin degradation products and their

reaction products are good for soil conditioning, control of

soil acidity, amelioration of the eff ects of excess Al and Fe,

increasing availability of phosphate, and crop stimulation.

Furthermore, they are excellent agents for nutrient metals

such as Mo, Fe, B, Zn, Mn, and Cu. Other functional groups

in the bio-oil-derived fertilizers are nutrients such as Ca,

K and P.96,97 DynaMotive co-operated with two fertilizer

manufacturers on the commercialization of bio-oil-derived

products, but so far no specifi c commercial product outlets

have been demonstrated.

Pure bio-oil can be mixed with lime to form BioLime™,

a trademark of Dynamotive Technologies Corporation in

Canada. Injection of this mixture into fl ue-gas tunnels

should result in complete removal of sulfur oxides, but also

in a signifi cant reduction of nitrogen oxides. Th e research

on this has stopped, and no references have been found in

literature hereaft er.

Fractions derived from bio-oil

In wood-derived pyrolysis oil, specifi c oxygenated com-

pounds are present in quite substantial amounts. Th e

recovery of such pure compounds from the complex bio-oil

may be technically feasible but probably economically unat-

tractive because of the high costs for the recovery of the

chemical and its low concentration in the oil. Th e relevant

chemical components are now presented.

Wood fl avors: Th e only commercial application of wood-

derived bio-oil known to date is that of wood fl avor or liquid

smoke. A number of companies produce these liquids by

adding water to the bio-oil. A red-colored product is then

obtained, that can be sprayed over meat before further cook-

ing. Th e taste, color, and smell of the meat are thus created

‘artifi cially’. A range of food-fl avoring products, based on

pyrolysis oils, has been patented and commercialized by

Red Arrow Products Company (USA)98 and the former

Chemviron, ProFagus (Germany).

Phenolic compounds: A signifi cant part of the oil is

the phenolic fraction, consisting of small amounts of phe-

nol, eugenol, cresols, and xylenols, and larger quantities

of alkylated (poly-) phenols (the so-called water insoluble

pyrolytic lignin). Recoveries of phenolic compounds up to

50 wt.% have been reported, but only for specifi c feedstocks.

Th e amount of the smaller, more expensive, phenolic compo-

nents in bio-oil is usually limited, probably because the orig-

inal lignin in the biomass is only partly cracked. Moreover,

it is likely contaminated by re-polymerized lignin and sugar

fragments (from the hemicellulose) as well. Phenolics have

also been proposed for use as an alternative wood preserva-

tive to replace creosotes.99

Sugars: Levoglucosan, together with levoglucosenone and

HAA, are the few sugars derivatives and detectable in GC

equipment. It seems that this in itself is the main reason that

the fi rst two components especially received a lot of atten-

tion: other sugars which are present cannot be traced back

in the oil using GC. An extensive overview on levoglucosan

is given elsewhere.7 Th e existing market for levoglucosan is

very small (and the product high-priced), but it may well be

an intermediate product suitable for further fermentation,

as indicated earlier. Most pyrolysis technologies could be

adapted for levoglucosan production by pre-hydrolysis of the

feedstock and/or demineralization.100 Levoglucosenone is

said to be applicable in the synthesis of antibiotics/pherom-

ones, rare sugars, butenolide, immuno-suppresive agents,

whisky lactones, and so on, and can be present in amounts

up to 24 wt.%. Progress in the last few years to valorize

bio-oil on the basis of these two interesting products is

limited, though. On the contrary, hydroxy-acetaldehyde is



a commercial product. It can be present in relatively large

amounts in the bio-oil (up to almost 20 wt.%), and is used in

browning food (cheese, meat, sausages, and poultry or fi sh).

A possible application is the use as a precursor for glyoxal

OHC-CHO, which is an important chemical produced by

oxidation of ethylene glycol.101

Acids: Components that can also be derived from bio-oil

are mainly the carboxylic acids, from which salts such as cal-

cium acetate and calcium formate can be produced.102,103 In

the aqueous fraction of the bio-oil, these acids are present in

amounts up to 10 wt.%. Th ey have potential applications such

as road or runway de-icing, sulfur dioxide removal during fos-

sil fuel combustion, or as a catalyst during coal combustion.

Furfural-derivatives: Furfural and furfurylalcohol can be

produced from carbohydrates (glucose, maltose, cellobiose,

amylose and cellulose) in amounts up to 30 wt.%.104

Economics

A key factor in the development to commercial implemen-

tation is the economic viability of fast pyrolysis processes.

Currently the main interest in Europe is electricity generation

from biomass. CO2 mitigation, socio-economic benefi ts from

re-deployment of surplus agricultural land, and energy inde-

pendence are driving forces. Th ese have led to signifi cant fi scal

incentives. Apart from such incentives, infl ation, (in)direct

eff ects of oil prices, local costs and labor strongly aff ect the

economics of pyrolysis plants. Th erefore, it does not make

sense to discuss economics in much detail. General data, col-

lected by BTG and confi rmed in other studies, show that the

range of capital costs for the pyrolysis plant alone is in between

€200 and €500/kWth input biomass, depending on the technol-

ogy, scale, degree of heat integration, location, etc. Th e main

parts of the BTG pyrolysis plant are the reactor, riser, combus-

tor, and condenser. Th e costs of pre-treatment, feeding, build-

ings, and infrastructure are not included, but may add up to

another 50 to 100%, depending on the initial feedstock prop-

erties (size, water content, dimensions, free-fl owing behavior,

bulk density, and so on). Costs related to the heat integration

system (heat recovery, steam generation, drying, etc.) are usu-

ally not addressed in the studies of economics.

One of the main challenges of BTG’s process concept is

the effi cient generation and further use of the excess heat

generated in the system. Th e Malaysian plant showed that (1)

the heat required for drying the wet feedstock is delivered

by the process itself; and (2) electricity can be generated

from the excess heat available, even aft er use for drying.

An important aspect to consider in plant economics is the

observation that the cost for the actual pyrolysis reactor is

just a fraction of the overall plant costs. At the same time,

a proper reactor choice in particular off ers the possibility

to reduce costs upfront (i.e., in feeding and pre-treatment)

or in peripheral equipment. Th e rotating cone reactor in

the Malaysian plant, for instance, costs about 2 to 3% of

the complete plant, but reduces the overall costs of the total

plant signifi cantly because the absence of inert gases limits

the costs of secondary equipment.

Studies over the years indicated that pyrolysis oils can be

produced at costs in a range of €4 to €14/GJ (corresponding

€65 to €225/t), with feedstock costing between €0 and €100/t

(€0 to €6/GJ).105,106 Such fi gures match quite well with the

data of the Malaysian plant. But again, it should be noted

that they strongly depend on process technology, the scale of

operation, feedstock, year of construction, and so on.

Generally speaking, it can be stated that the costs of

pyrolysis processes should be rather low, because the oper-

ating conditions are less extreme than for combustion or

gasifi cation (lower temperatures and atmospheric pressure).

Biomass pre-treatment, heat integration and the required

operation reliability, however, can be factors seriously

increasing the overall investment costs.

Concluding remarks

Pyrolysis defi nitely remains an interesting pre-treatment

technique enabling (intercontinental) transport of large vol-

umes of biomass. Th e proof of principle (reactor concept),

and proof of concept (plant set-up) have been demonstrated,

amongst other by BTG (Malaysia), DynaMotive (Canada)

and Ensyn (USA and Canada). Fast pyrolysis is still an

immature technology, of which many aspects are unknown.

Although a number of installations were erected, they all

suff er from a lack of operational hours and no process has

really been ‘demonstrated’. Interested industries like heat

and electricity producers, oil companies and food/feed com-

panies, are waiting for full-scale demonstration including

continuous operation (>7000 hours per year) preferably of

multiple plants on a scale of 5 to 10 tons biomass feedstock



per hour. Here, the problem is that the opportunities for

commercial bio-oil production are limited due to lack of

economic applications. Th e oil should, for the time being, be

used to substitute fossil fuels in heat and power production,

by combustion in conventional boilers or co-combustion in

power stations. Th e focus for the next few years is to produce

oil, and to apply simple and cheap applications. Once the

process and peripherals have been proven, larger amounts of

oil will become available for the development and commer-

cial-scale demonstration of other bio-oil applications, such

as turbines, diesel engines and/or further upgrading to bio-

fuels. Obviously the construction and operation of demon-

stration installations needs to be supported by government

bodies and risk-taking investors, viz. in a sustainable way.

Past failures in biomass demonstration projects were oft en

caused by a serious lack of fi nancing aft er the fi rst period of

plant design and erection.

Challenges within the coming years are related to

improving:

• the operational reliability of demonstration scale pyroly-

sis processes;

• the feedstock fl exibility (accepting all kinds of biomass

residues, instead of only wood);

• the heat transfer to the pyrolysis reactor and from the

char combustor; and

• the process heat integration and its control.

In the meantime, R&D should be directed to improve-

ment of the quality (and stability) of the oil in relation to the

end-application envisaged. Although nice, uniform bio-oil

samples have always been available during the last 20 years,

only limited fundamental know-how has been generated on

the exact composition of the bio-oil. Much attention was paid

to (destructive!) GC analysis techniques, while nowadays it

is gradually being recognized and understood that the com-

ponents quantifi ed are not necessarily (in that concentration)

present in the original oils. In addition, the authors believe

that the compounds or fractions in the oil, causing its specifi c

characteristics (pH, ageing, viscosity, phase separation, and so

on) have not been fully identifi ed nor are the reactions taking

place understood. One particular issue is the exact role of the

various oxygen functionalities in the oil. It is important to

establish which functionalities are desired and which ones

are undesirable, and to understand how to steer the pyrolysis

process itself in this respect, for instance, by catalysis.

With respect to bio-oil upgrading, an important conclu-

sion is that reduction and control of the oxygen functionali-

ties should be the ultimate goal instead of the reduction in

oxygen content itself. Actual demonstration that pyrolysis

oils can be used as a feedstock in refi neries, as aimed at in

the European ‘BioCoup’ project, will certainly boost the fur-

ther development of fast pyrolysis.

For BTG’s system in particular, the progress made in

Malaysia from the beginning of the plant design in 2004 to

date is quite satisfactory. From an initial set of experiments in

2003 (8 hours and maximum 100 kg/hr feeding), the system

has been scaled up to a 24 hours/day running factory, where a

direct link has been established between the palm mill and the

pyrolysis plant. In addition, the process is applied to a very dif-

fi cult biomass feedstock material, which is fl uff y (not free fl ow-

ing) and wet, and has a high ash content (with a very low melt-

ing point). Problems resolved are largely related to boundary

conditions (feeding, pre-treatment, ash-related problems and

heat recovery) instead of to the actual pyrolysis process itself.

All the experiences and knowledge acquired by BTG over the

past 20 years of RTD will be condensed in the design of a new

wood based demo-pant of 125 t/d, to be erected by Empyro

BV in 2010/2011 at a site in Hengelo, the Netherlands. A BTG

daughter company called BTG-Bioliquids takes care of the

future commercialization of BTG’s fast pyrolysis technology.

Acknowledgements

Th e authors would like to thank Erik Heeres and Agnes

Ardiyanti from the University of Groningen for carrying out

the TGA experiments reported in Fig. 1, and Dietrich Meier

for allowing us to use his presentation of pyrolysis technolo-

gies as input for Fig. 6.

References 1. Bridgwater AV, Renewable fuels and chemicals by thermal processing of

biomass, Chem Eng J l 91:87–102 (2003).

2. Freel BA, Graham RG, Huffman DR and Vogiatzis AJ, Rapid thermal

processing of biomass (RTP): development, demonstration, and com-

mercialization. Energy Biomass Wastes 16:811–826 (1993).

3. Piskorz J, Radlein D, Scott DS and Czernik S, Liquid products from the

fast pyrolysis of wood and cellulose, in Research in Thermochemical

Biomass Conversion, ed by Bridgwater AV and Kuester JL. Elsevier

Applied Science, London, pp. 557–571 (1988).



4. Babu BV, Biomass pyrolysis: a state-of-the-art review. Biofuels Bioprod

Bioref 2:393-414 (2008).

5. Bridgwater AV, Fast pyrolysis of biomass: a handbook. Vol 1. CPL Press,

Newbury, Berkshire, UK (1999).





8. Mohan D, Pittman CU, Steele PH, Pyrolysis of wood/biomass for bio-oil:

A critical review. Energ Fuel 20:848–889 (2006).

9. Wang X, Biomass fast pyrolysis in a fl uidized bed. Thesis. University of

Twente, the Netherlands (2006).

10. Wang X, Kersten SRA, Prins W and van Swaaij WPM, Biomass

pyrolysis in a fl uidized bed reactor. Part 2: Experimental Validation of

Model Results. Ind Eng Chem Res 44(23):8786–8795 (2005).

11. Gupta AK and Lilley DG, Thermal destruction of wastes and plastics, in

Plastics and the Environment, ed by Andrady AL. John Wiley & Sons Ltd,

Chichester, pp. 629–696 (2003).

12. Yang H, Yan R, Chen H, Lee DH and Zheng C, Characteristics of

hemicellulose, cellulose and lignin pyrolysis. Fuel 86:1781–1788 (2007).

13. Vaca Garcia C, Biomaterials, in Introduction to chemicals from biomass,

ed by Clark J and Deswarte F. John Wiley & Sons Ltd, Chichester (2008).

14. Shafi zadeh F and Chin PPS, Wood technology, chemical aspects. Am

Chem Soc Symp Series, 43:57–81 (1977).

15. Shafi zadeh F, Pyrolytic reactions and products of biomass, in

Proceedings: Fundamentals of Thermochemical Biomass Conversion,

ed by Overend RP, Milne TA and Mudge LK. Elsevier Applied Science,

London (1985).

16. Wagenaar BM, The rotating cone reactor for rapid thermal solids

processing. Thesis. University of Twente, the Netherlands (1994).

17. Wagenaar BN, Prins W, and van Swaaij WPM, Pyrolysis of biomass in the

rotating cone reactor: modeling and experimental justifi cation. Chem Eng

Sci 49 (24B):5109–5126 (1994).

18. Di Blasi C, Modelling chemical and physical processes of wood and

biomass pyrolysis. Prog Energ Combust Sci 34:47–90 (2008).

19. Radlein D, Piskorz J and Scott DS, Fast pyrolysis of natural

polysaccharides as a potential industrial process. J Anal Appl

Pyrolysis 19:41–63 (1991).

20. Fahmi R, Bridgwater AV, Donnison I and Yates N, The effect of lignin and

inorganic species in biomass on pyrolysis oil yield, quality and stability.

Fuel 87:1230–1240 (2008).

21. Chiaramonti D, Oasmaa A and Solantausta Y, Power generation using fast

pyrolysis liquids from biomass. Renew Sust Energ Rev 11:1056–1086 (2007).

22. Di Blasi C, Branca C and Galgano A, Thermal and catalytic decompo-

sition of wood impregnated with sulfur- and phosphorus-containing

ammonium salts. Polym Degrad Stabil 93:335–346 (2008).

23. Westerhof RJM, Brilman DWF, Kersten SRM and Van Swaaij WPM, Effect

of condenser operation on the biomass fast pyrolysis oil. 16th European

Conference & Exhibition, June 2–6 2008, Valencia (Spain), 1095–1099

(2008).

24. Diebold JP and Bridgwater AV, Overview of fast pyrolysis of biomass

for the production of liquid fuels, in Developments in Thermochemical

Biomass Conversion, ed by Bridgwater AV and Boocock DGB. Blackie

Academic & Professional, London, pp. 5–23 (1997).

25. Oasmaa A and Kuoppala E, Fast pyrolysis of forestry residue. 3. Storage

stability of liquid fuel. Energ Fuel 17(4):1075–1084 (2003).

26. Oasmaa A, Elliot DC and Muller S, Quality control in fast pyrolysis

bio-oil production and use, Environmental Progress & Sustainable

Energy 28:404–409 (2009).

27. Adjaye JD and Bakhshi NN, Production of hydrocarbons by catalytic

upgrading of a fast pyrolysis bio-oil. Part I: Conversion over various

catalysts. Fuel Process Technol 45:161–183 (1995).

28. Horne P and Williams PT, Reaction of oxygenated biomass pyrolysis

compounds over a ZSM-5 catalyst. Renewable Energ 2:131–144 (1996).

29. Samolada MC, Papafotica A and Vasalos IA, Catalyst evaluation for cata-

lytic biomass pyrolysis Energ Fuel 14:1161–1167 (2000).

30. Carlson TR, Tompsett GA, Conner WC and Huber GW, Aromatic

production from catalytic fast pyrolysis of biomass-derived

feedstocks. Topics in Catalysis 52(3):241–252 (2009).

31. Lehmann J and Joseph S, Biochar for Environmental Management.

Earthscan, London (2009).

32. Maniatis K, Baeyens J, Peeters H and Roggeman G, The Egemin

fl ash pyrolysis process: commissioning and results, in Advances in

Thermochemical Biomass Conversion, ed by Bridgwater AV. Blackie

Academic & Professional, London, pp. 1257–1264 (1993).

33. Brown DB, Continuous ablative regenerator system, in Bio-oil:

Production and Utilisation. Proceedings of the 2nd EU-Canada Workshop

on Thermal Biomass Processing ed by Bridgwater AV and Hogan EN.

CPL Press, Newbury, pp. 96–100 (1996).

34. Diebold JP and Scahill JW, Improvements in the Vortex Reactor Design,

in Developments in Thermochemical Biomass Conversion, ed by

Bridgwater AV and Boocock DGB. Blackie Academic & Professional,

London, p. 242 (1997).

35. Peacocke GVC and Bridgwater AV, Ablative plate pyrolysis of biomass for

liquids, Biomass Bioenerg 7(1–6):147–154 (1995).

36. Lede J, Broust F, Ndiaye F-T and Ferrer M, Properties of bio-oils

produced by biomass fast pyrolysis in a cyclone reactor. Fuel 86:

1800–1810 (2007).

37. Schöll S, Klaubert H and Meier D, Wood liquefacation by fl ash-pyrolysis

with an innovative Pyrolysis system. DGMK proceding 2004-1 contribu-

tions to DGMK-meeting, Energetic Utzilization to Biomasses, April 19–21,

2004 Velen/Westf (2004).

38. Meier D, Aus Holz wird Oel, Forstmaschinen Profi 15(12):26–28 (2007).

39. Meier D, New ablative pyrolyzer in operation in Germany, PyNe

Newsletter 17 (2004).

40. Bramer EA, Holthuis MR and Brem G, A novel thermogravimetric vortex

reactor for the determination of the primary pyrolysis rate of biomass,

Proceedings Conference on Science in Thermal and Chemical Biomass

Conversion, Aug 30–Sept 2 2004, Vancouver, (2004).

41. Cuevas A, Reinoso C and Lamas J, Advances and development at

the Union Fenosa pyrolysis plant, in 8th Conference on Biomass for

Environment, Agriculture and Industry, ed by Chartier, Beenackers and

Grassi. Pergamom Press, Oxford, (1995).

42. Fransham P, Advances in dry distillation technology. ThermalNet

Newsletter 1: Aston University, Birmingham, (2006).

43. Sandvig E, Walling G, Brown RC, Pletla R, Radlein D and Johnson W,

Integrated pyrolysis combined cycle biomass power system concept

defi nition Final Report, Report DE-FS26-01NT41353 (2003).



44. Dynamotive, Biooil Production at Dynamotive’s Guelph plant: cellulose

based biooil aimed at industrial fuel markets and input to BTL plants.

ThermalNet Newsletter 5 (2007).

45. Nieminen JP, Gust S and Nyrönen T, ForesteraTM Liquefi ed wood fuel

pilot plant, PyNe Newsletter 16 (2003).

46. www.dynamotive.com Jan 2010.

47. Freel B, Ensyn announces commissioning of new biomass facility, PyNe

Newsletter 14 (2002).

48. Freel B and Graham RG, Bio-oil preservatives, US Patent 6485841,

NREL (2002).

49. Graham RG, Freel BA, Huffman DR and Bergougnou MA, Commercial-

scale rapid thermal processing of biomass. Biomass Bioenerg 7:

251–258 (1994).

50. Rossi C and Graham R, Fast Pyrolysis at ENEL, in Biomass

Gasifi cation and Pyrolysis: State of the Art and Future Prospects, ed

by Kaltschmitt M and Bridgwater AV. CPL Scientifi c Ltd, Newbury,

Berkshire, UK (1997).

51. Envergent, Presentation at the tcbiomass 2009 International Conference on

Thermochemical Conversion Science, Sept 16-18, Chicago, (2009).

52. Metso, Presentation at the tcbiomass 2009 International Conference on

Thermochemical Conversion Science, Sept 16-18, Chicago, (2009).

53. Roy C, Morin D and Dubé F, The biomass Pyrocycling™ process,

in Biomass Gasifi cation and Pyrolysis: State of the Art and Future

Prospects, ed by Kaltschmitt M and Bridgwater AV. CPL Scientifi c Ltd,


54. Roy C., Yang J., Blanchette D., Korving L., De Caumia B., Development

of a novel vacuum pyrolysis reactor with improved heat transfer poten-

tial, in: Developments in Thermochemical Biomass Conversion, ed by

Bridgwater A. V. and Boocock D.G.B., Blackie Academic & Professional,

London (1997), 351–367

55. http://www.newearth1.net/newearth-eco-technology.html Jan 2010

56. http://www.advbiorefi neryinc.ca/news/ Jan 2010

57. Badger P.C., Fransham P. Use of mobile fast pyrolysis plants to densify

biomass and reduce biomass handling costs. In: lEA Bioenergy Task 31:

International Workshop: Impacts on Forest Resources and Utilization of

Wood for Energy. 5-10 October 2003, Flagstaff, Arizona, 2003.

58. Henrich E, Dahmen N and Dinjus E, Cost estimate for biosynfuel

production via biosyncrude gasifi cation. Biofuels Bioprod Bioref 3:

28–41 (2009).

59. Venderbosch RH, Prins W and Wagenaar BM, Flash pyrolysis of wood for

energy and chemicals: Part 1. NPT Procestechnologie 6:18–23 (1999).

60. Venderbosch RH, Prins W and Wagenaar BM, Flash pyrolysis of wood for

energy and chemicals: Part 2. NPT Procestechnologie 1:16–20 (2000).

61. Venderbosch RH, Janse AMC, Radovanovic M, Prins W and Van Swaaij

WPM, Pyrolysis of pine wood in a small integrated pilot plant rotating

cone reactor, in Biomass Gasifi cation & Pyrolysis, State of the art and

future prospects, ed by Kaltschmitt M and Bridgwater AV. CPL Scientifi c

Ltd, Newbury, Berkshire, UK (1997).

62. Genting Berhad, Annual Report 2007, Review of Operations (2007).

63. Venderbosch RH, Gansekoele E, Florijn J, Assink D and Ng HY, Pyrolysis

of palm oil residues in Malaysia. ThermalNet Newsletter Feb (2006).

64. www.btg-btl.com Jan 2010

65. Wagenaar BM, Venderbosch RH, Prins W and Penninks FWM, Bio-oil as

a coal substitute in a 600 MWe Power Station, 12th European Conference

and technology Exhibition on Biomass for Energy, Industry and Climate

Protection. June 17-21, Amsterdam, the Netherlands (2002).

66. Khodier A, Kilgallon P, Legrave N, Simms N, Oakey J and Bridgwater T,

Pilot scale combustion of fast pyrolysis bio-oil: ash deposition and gase-

ous emissions. Environmental Progress & Sustainable Energy 28 (3):

397–403 (2009).

67. Jay DC, Sipila KH, Rantanen OA and Nylund NO, Wood pyrolysis oil for

diesel engines, ICE-Vol. 25-3, ASME Fall Technical Conference, ASME

Books, New York (1995).

68. Baglioni P, Chiaramonti D, Bonini M, Soldaini I and Tondi G, Bio-crude-

oil/diesel oil emulsifi cation: main achievements of the emulsifi cation

process and preliminary results of tests on Diesel engine, in Progress

in thermochemical biomass conversion, ed by Bridgwater AV. Blackwell

Science Ltd, Oxford, 1525–1539 (2001).

69. Kindelan JC, Comparative study of various physical and chemical

aspects of pyrolysis bio-oils versus conventional fuels, regarding their

use in engines, in Biomass Pyrolysis Oil Properties and Combustion,

ed by Milne TA. Sandia National Laboratories, Albuquerque,

p 321 (1994).

70. Venderbosch RH and van Helden M, Diesel engines on Bio-Oil: a review,

SDE Samenwerkingsverband Duurzame Energie BTG-report (2001).

71. Ikura I, Stanciulescu M and Hogan E, Emulsifi cation of pyrolysis derived

bio-oil in diesel fuel. Biomass Bioenerg 24(3):221–232 (2003).

72. Chiaramonti D, Bonini M, Fratini E, Tondi G, Gartner K, Bridgwater

AV, Grimm HP, Soldaini I, Webster A and Baglioni P. Development of

emulsions from biomass pyrolytic liquid and diesel and their use in

engines – Part 2: test in diesel engines. Biomass Bioenerg 25(1):

101–111 (2003).

73. Meier D, Scholl S, Klaubert H, Markgraf J, Practical results from Pytec’s

Biomass-to-Oil process with abltaivbe pyrolyser and diesel CHP plant,

in Success & Visions for Bioenergy: Thermal processing of biomass

for bioenergy, biofuels and bioproducts, ed Bridgwater AV. CPL Press,


74. www.bioliquids-chp.eu Jan 2010

75. Lupandin V, Thamburaj R and Nikolayev A, GT2005-68488 Test results

of the OGT2500 gas turbine engine running on alternative fuels: bio-oil,

ethanol, biodiesel and crude oil, Proceedings of GT2005 ASME Turbo

Expo 2005, June 6–9, 2005, Reno-Tahoe, Nevada, USA (2005).

76. Henrich E, Dahmen N and Dinjus E, Cost estimate for biosynfuel

production via syncrude gasifi cation. Biofuels Bioprod Bioref 3:

28–41 (2009).

77. Higman C and Van der Burgt M, Gasifi cation. Elsevier Science, London

(2003).

78. Venderbosch RH, van de Beld L and Prins W, Entrained fl ow gasifi cation

of bio-oil for synthesis gas. 12th European Conference and technology

Exhibition on Biomass for Energy, Industry and Climate Protection,

17–21 June 2002, Amsterdam, the Netherlands (2002).

79. Horne P and Williams PT, Reaction of oxygenated biomass pyrolysis

compounds over a ZSM-5 catalyst. Renew Energ 2:131–144 (1996).

80. Adjaye JD and Bakhshi NN, Catalytic conversion of a biomass-derived oil

to fuel and chemicals I: model compound studies and reaction pathways.

Biomass Bioenerg 8 (3):131–149 (1995).

81. Adjaye JD and Bakhshi NN, Production of hydrocarbons by catalytic

upgrading of a fast pyrolysis bio-oil. Part II: comparative catalyst



performance and reaction pathways. Fuel Process Technol 45: 185–202

(1995).

82. Elliott DC, Historical developments in hydroprocessing bio-oils. Energ

Fuel 21:1792–1815 (2007).

83. Elliott DC, Beckman D, Bridgwater AV, Diebold JP, Gevert SB and

Solantausta Y, Developments in direct thermal liquefaction of biomass:

1983–1990. Energ Fuel 5:299–410 (1991).

84. Samolada MC, Baldauf W and Vasalos IA, Production of bio-gasoline by

upgrading biomass fl ash pyrolysis liquids via hydrogen processing and

catalytic cracking. Fuel 14:1667–1675 (1998).

85. Wildschut J, Arentz J, Rasrendra CB, Venderbosch RH and Heeres HJ,

Catalytic hydrotreatment of fast pyrolysis oil: Model studies on reac-

tion pathways for the carbohydrate fraction. Environmental Progress &

Sustainable Energy 28 (3):450–460 (2009).

86. Elliott DC, Hart TR, Neuenschwander GG, Rotness LJ and Zacher AH,

Catalytic hydroprocessing of biomass fast pyrolysis bio-oil to produce

hydrocarbon products, Environmental Progress & Sustainable Energy

28 (3):441–449 (2009).

87. Venderbosch RH, Ardiyanti AR, Wildschut J, Oasmaa A and Heeres

HJ, Insights in the hydroprocessing of biomass derived pyrolysis oils.

Available at www.btgworld.com Jan 2010 (2009).

88. Ardiyanti AR, Venderbosch RH and Heeres HJ, Process-product stud-

ies on pyrolysis oil upgrading by hydrotreatment with Ru/C catalysts.

Proceedings of the 2009 AIChE Spring National Meeting, April 26–30,

Tampa (2009).

89. Solantausta, Presentation at the EC Bioenergy Contractors Meeting

October 15-16, Brussels (2008).

90. Mahfud FH, Melian-Cabrera I, Manurung R and Heeres HJ, Upgrading of

fl ash pyrolysis oil by reactive distillation using a high boiling alcohol and

acid catalysts. Trans IChemE, Part B, Process Safety and Environmental

Protection 85(B5):466–472 (2007).

91. Brown RC, Radlein D and Piskorz J, Chemicals and materials from

renewable resources. ACS Symposium Series 784. American Chemical

Society, Washington, DC, pp. 123–132 (2001).

92. So KS and Brown RC, Economic analysis of selected lignocellulosic –

to – ethanol conversion technologies. Appl Biochem Biotechnol 77–79:

633–640 (1999).

93. Olson ES, Sharma RK, Aulich TR, Freel B, Boulard D, von Keitz M, Final

Report: A fast-pyrolysis biorefi nery for the Northern Great Plains, 2006-

EERC-01-05, University of North Dakota, North Dakota (2006).

94. Milne T, Agblevor F, Davis M, Deutch S and Johnson D, A review

of the chemical composition of fast pyrolysis oils from biomass, in

Developments in Thermochemical biomass conversion,

ed by Bridgwater AV and Boocock DGB. Blackie Academic &

Professional, London, p. 409 (1997).

95. Effendi A, Gerhauser H and Bridgwater AV, Production of renewable

phenolic resins by thermochemical conversion of biomass: a review.

Renew Sustain Energy Rev 12:2092–2116 (2008).

96. Radlein D and Piskorz J, Production of chemicals from bio-oil, in PyNe

Final report ed by Bridgwater AV and Humphreys C. FAIR CT94-1857,

Aston University, Birmingham (1998).

97. Radlein D, Fast pyrolysis for the production of chemicals, in Bio-oil pro-

duction and upgrading, Proceedings of the 2nd EU-Canada Workshop

on Thermal Biomass Processing, ed by Bridgwater AV and Hogan EN.

CPL Press, Newbury, p. 113 (1996).

98. Underwood GL and Rozum JJ, Article and method for browning and

fl avoring foodstuffs United States Patent 7282229 (2007).

98. Suzuki T, Doi S, Yamakawa M, Yamamoto K, Watanabe T and Funaki

M, Recovery of wood preservatives from wood pyrolysis tar by solvent

extraction. Holzforschung: Mitteilungen zur Chemie, Physik, Biologie

und Technologie des Holzes. 51:214–218 (1997).

100. Brown RC, Radlein D and Piskorz J, Pretreatment processes to increase

pyrolytic yield of levoglucosan from herbaceous feedstocks, chemicals

and materials from renewable resources. ACS Symposium Series 784.

American Chemical Society, Washington, DC, pp. 123–132 (2001).

101. Oehr KH and McKinely J, Glyoxal production from biomass pyroly-

sis derived hydroxy-acetaldehyde, in Advances in Thermochemical

Biomass Conversion, ed by Bridgwater AV. Blackie Academic &

Professional, London p. 1452 (1994).

102. Oehr KH and Barrass G, Biomass derived calcium carboxylates, in

Advances in Thermochemical Biomass Conversion, ed by Bridgwater

AV. Blackie Academic & Professional, London, p. 1456 (1994).

103. Oehr KH and Barrass G, Biomass derived alkaline carboxylate road

deicers. Resour Conserv Recy 7:155–160 (1992).

104. Pavlath AE and Gregorski KS, Carbohydrate pyrolysis II. Formation of

furfural and furfurylalcohol during the pyrolysis of selected carbohy-

drates with acidic and basic catalysts, in Research in Thermochemical

Biomass Conversion, ed by Bridgwater AV and Kuester JL. Elsevier

Applied Science, London, p. 155 (1988).

105. Peacocke GVC, Bridgwater AV and Brammer JC, Techno-economic

assessment of power production from the Wellman and BTG fast pyrolysis

process, in Thermal and Chemical Biomass Conversion, ed by Bridgwater

AV and Boocock DGB. CPL Press, Newbury, Berkshire, UK (2006).

106. Ringer M, Putsche V and Scahill J, Large-scale pyrolysis oil production:

A technology assessment and economic analysis, Technical Report,

NREL/TP-510-37779, NREL, Golden Colorado (2006).



Robbie Venderbosch

After his PhD at the Twente University in 1998,

Robbie Venderbosch joined BTG Biomass

Technology Group BV. He has more than ten

years (practical) experience in thermochemi-

cal processes for biomass and its further

use. Robbie Venderbosch has expertise and

responsibilities in the field of engineering and

implementation of fluidized bed system coupled with chemical

reaction, biomass energy processes (ranging from concept to de-

tailed design), construction and operation of laboratory, pilot and

demonstration scale set-ups and process instrumentation, and

process control (both software and hardware). He was responsible

for the delivery of BTG’s commercial pyrolysis plant to a client in

Malaysia.

Wolter Prins

Wolter Prins received his Masters in Chemical

Engineering from the University of Groningen

and his doctoral degree from the University

of Twente in Enschede, the Netherlands. In

1984, he was appointed Assistant Professor

in the Department of Chemical Technology

at the University of Twente. Since 1992, he

has combined his work at the University with a position as head

of R&D in BTG Biomass Technology Group BV in Enschede. In

2008, he was appointed as Professor for Bioresources Processing

in the Bioscience Engineering faculty of the University of Ghent.

Wolter Prins published around a hundred papers in the area of

novel gas-solid reactors, heat and mass transfer in fluidized beds,

and thermochemical conversion of biomass. He participated in

the European Network for Pyrolysis (Pyne) for many years, and

was invited by NEDO and the Chinese Academy of Sciences to

present his work in Japan and China respectively.

fast pyrolysis development_venderbosch et al. 2010

Documents