Category Archives: Thermal biomass. conversion and utilization —. Biomass information. system

MATHEMATICAL MODELLING OF PYROLYSIS

2.3.1 Подпись: 2.3Introduction

Mathematical modelling may be defined as the art of obtaining a solution, given specified input data, that is representative of the response of the process to a corresponding set of inputs (54). The development of a mathematical model can be mechanistic (theoretical) using physico-chemical principles, empirical based on experimental data, statistical or judgmental as in an expert system or a combination of the above.

2.3.2 Pyrolysis Modelling Objectives

Mathematical modelling is utilised in pyrolysis to account for the effects of the interaction of the parameters on the end products. The objectives of a mathematical pyrolysis model should include:

1 the development of a diagnostic tool in order to evaluate the importance of the various process parameters such as particle size, heat of pyrolysis (reaction) and thermal properties of the feedstock and products;

2 the prediction of the effects of process parameters, i. e. heating rate, reactor temperature, particle size, moisture content, on the product yields and characteristics in order to aid optimisation of the pyrolysis process;

3 the development and establishment of better reactor design techniques in order to specify reactor type and size.

There are four types of pyrolysis model: empirical, kinetic, analytical and stagewise. All models basically derive energy and mass balances across a particle of biomass as shown in Figure 2.6.

where:

Te: environment temperature T§: surface temperature

Tq: char temperature Ту/: biomass temperature

The pyrolysis of a single particle represented above is not however applicable to conditions where particle ablation is significant or the primary method used to achieve pyrolysis. A typical example is that of Diebold’s vortex reactor where the particle is in contact with a heated surface under conditions of high applied pressure and high relative motion where conductive heat transfer is the dominant mode (48).

image021

image14

Figure 2.6 Pyrolysis Behaviour for a Single Spherical Particle

Economics and efficiency

A key factor in the continuing development of fast pyrolysis processes to eventual commercial implementation is their economic viability. The current main interest in Europe is for electricity generation from biomass with the driving forces of environmental benefits, CO2 mitigation, socio-economic benefits from re­deployment of surplus agricultural land, and energy independence. These have led to significant fiscal incentives being offered for renewable electricity in many European countries such as prices of up to 200/kWh in Italy and up to 140/kWh in the UK.

An indication of the performance and specific capital cost of various thermal conversion to electrical power technologies is summarised in Figures 3.2 and 3.3 below. This clearly shows the potential impact of a successful development of flash pyrolysis and engines on the cost of electricity production at smaller scales of operation (e. g. up to 10 MWe) since capital costs are similar for pyrolysers as gasifiers and the higher efficiencies will thus give lower production costs.

3.8 CONCLUSIONS

Fast pyrolysis is becoming more accepted as an emerging technology with commercial potential for producing high yields of liquid fuels that can be used in many applications as direct substitutes for conventional fuels or as a source of chemicals. There are still problems to be resolved but it is clear that considerable progress is being made in firstly identifying these problems and then finding and defining solutions. There are some interesting challenges to be faced in modifying the fast pyrolysis technology, in upgrading the liquids and adapting applications to accept the unusual behaviour and characteristics of the liquid product. International co-operation is one of the key routes to facilitate such developments and this needs to be encouraged as bioenergy is increasingly accepted as a

potentially significant resource. Further research and development of the fundamental science which provides the tools for the design engineer to achieve the performance improvements is also required for successful exploitation.

image21 image056

Short term opportunities are likely to be in power generation due to the higher added value of electricity and the incentives available in several European countries. Chemicals will considerably enhance the overall economics but are further away in terms of development. It is important to remember that it may be more effective to modify the application to suit the liquid bio-oil than modify the liquid to suit the application.

Efficiency vs Size for Power Generation Systems (46)

Specific capital cost, $/kWe

image22

Figure 3.3 Specific Capital Cost for Power Generation Systems (46) 3.9

FAST PYROLYSIS LIQUID PRODUCTS

6.2.1 Introduction

Many flash pyrolysis processes have been developed to pilot, demonstration and commercial scale based solely on thermal conversion without any extrinsic catalytic activity, to give a crude fuel product (2, 3). Concerns over utilisation and assimilation into a fuel market infrastructure have caused attention to be paid to in-

6-111

situ or close coupled catalytic upgrading.

The liquid product is currently attracting the most interest in Europe and North America because of its high energy density, easy transportability, ease of use and de-coupling of conversion and utilisation processes (4). The liquid approximates to biomass in elemental composition, and is composed of a very complex mixture of oxygenated hydrocarbons. The complexity arises from the degradation of lignin, cellulose and hemicellulose, resulting in a broad spectrum of oxygenated compounds from uncontrolled and interactive degradation. Its composition is determined intrinsically by the temperature, rate of reaction, vapour residence time, and temperature-time cooling and quenching process which controls the extent of secondary reactions, and extrinsically by the feed composition. The liquid is often referred to as "oil" or "bio-oil" or "bio-crude-оіГ and can be upgraded to liquid hydrocarbon fuels, as indicated in Table 2 above.

Typical properties of flash pyrolysis oil have been reported from analytical programmes sponsored, for example, by the EC JOULE and AIR programmes (5), Energy Mines and Resources Canada (6), the United States Department of Energy (7) and the International Energy Agency Bioenergy Agreement (8), as well as extensively by technology developers. The key feature from a fuel utilisation viewpoint is the high oxygen content which ranges from 30 to 55% wt. depending on the feed, the feed water content, the oxygenated product spectrum, product water content and basis of reporting.

It is important to note that there are two types of liquids produced by pyrolysis of biomass:

• Primary liquid from flash pyrolysis processes. This is produced in high yields of up to 85% weight. It has a relatively low viscosity and high water miscibility tolerance of up to 35-50% wt water and can be readily combusted in most applications such as boilers, kilns and dual fuel engines. It is relatively unstable compared to conventional fuel oils being, for example, very temperature sensitive and non-volatile in distillation due to polymerisation reactions (1, 4).

• Secondary oil or tar from conventional or slow pyrolysis processes. This is produced in low yields of up to 20% weight It is very viscous and can only tolerate up to about 20% wt water before phase separation occurs. While it can also be combusted in many applications, the high viscosity and potential for water separation require careful handling.

The characteristics of both slow and flash pyrolysis oils are summarised in Table 6.2. Both products may be upgraded by any of the processes described in this chapter, but the low yields of slow pyrolysis liquid products will give very poor overall yields of upgraded products, thus adversely affecting the economics.

Table 6.2 Detailed Characteristics

of Wood

Derived Flash

Pyrolysis Oils

Physical DroDertv

Flash pyrolysis

Slow Dvrolvsis

Moisture content

20%

14.6%

pH

2.5

2.0

Specific gravity

1.21

1.195

Elemental analysis (moisture free)

C

56.4%

61.9%

H

6.2%

6.0%

N

0.2%

1.05%

S

<0.01%

0.03%

Ash

0.1%

1.5%

0 (by difference)

37.1%

29.5%

C/H ratio

9.1

10.3

HHV (moisture free basis)

23 MJ/kg

26.3 MJ/kg

HHV as produced

19.3 MJ/kg

Viscosity (@ 40°C)

51 cp

300 cp

Kinematic viscosity @ 25®C

233 cSt

@ 40°C

134 cSt

ASTM vacuum distillation 160 °С

10%

193 °С

20%

219 °С

40%

Distillate

50%

Pour point

-23 °С

27°C

Solubility hexane insoluble

99%

toluene insoluble

84%

acetone/acetic acid insoluble

0.14%

Feed preparation

The heat transfer rate requirements described above impose particle size limitations on the feed for some reactors. The cost of size reduction in financial and energy terms is clear qualitatively but data is not available to define such a penalty associated with the small particle sizes demanded of fluid bed and circulating fluid bed systems. Reactor performance as for example liquid yields is, therefore, not an adequate criterion by itself.

Drying is usually required to less than 10% wt water unless a naturally dry material such as straw is available. As moisture is generated in flash pyrolysis, bio-oil always contains at least about 15% water at an assumed product yield of around 60 wt% organics and 11 wt% reaction water. This water cannot be removed by conventional methods such as distillation. The effect of water is complex in that it affects stability, viscosity, pH, corrosiveness, and other liquid properties. Selective condensation may reduce the water content of one or more fractions but at the

expense of operating problems and a possible loss of low molecular weight volatile components,

Description

image25

Figure 5.1 is a flowsheet of the process (1, 2), Feedstocks tested include wood chips, olive husks, straw and vine trimmings. The feed was screened, rechipped and dried in a rotary drier before entering the fluid bed reactor. The drying fuel was propane, but the product fuel gas would be used in a commercial venture. Air was added to the reactor to give a partial gasification reaction to provide reaction heat. This results in a poor quality, low heating value fuel gas. The reaction temperature and pressure are about 500°C and 1 atm respectively. The char formed is separated from the vapour stream in a hot gas cyclone. The char is cooled in a sequence of water cooled screw conveyors and stored in a silo. The reactor was not well defined — it was specified as a fluid bed, but seemed to behave more like a stirred bed. As the primary objective was the production of oil for testing, little instrumentation was available.

The vapour and gas streams passed through a quench vessel where they were cooled and condensed by direct contact with recycle product water. The mixture of oil and water was separated in a settlement tank before sending the oil to storage and recycling the water through an air cooler. Excess water was removed to
maintain a constant height interface in the settlement tank. The resulting water condensate had a very high COD at around 150000 and would require treatment. Further liquid was removed from the product gas in a cyclone and filter, before burning the gas in a flare. The gas from the drier was also sent to the flare to eliminate pollutants. The gas could be used internally as a fuel source for drying the feed or other process heat applications. A typical mass balance and energy output is shown in Table 5.2.

TECHNO-ECONOMICS

Techno-economic assessments are important for identifying more promising process routes, and for identifying major areas of uncertainty where further R&D would have a significant impact on technical feasibility and commercial viability. The technical and economic viability of liquid fuels production that include catalytic processing has been investigated by a number of researchers and research groups as summarised in Tables 6.10 and 6.11.

Table 6.10 Techno-economic Assessment Groups

Colorado School Mines, USA

chemicals and liquid fuels

166

IEA Bioenergy Agreement

liquefaction test facility

167

liquefaction & pyrolysis

117, 168, 169, 170

gasoline via zeolites

159

Science Applications Inc., USA

liquid fuels

171,172, 173, 174

SRI, USA

Albany liquefaction plant

175

Statens, Sweden

wood and peat liquefaction

167

University of Aston, UK

liquid fuels: gasification mainly

176, 177, 178

flash pyrolysis, upgrading

165, 179, 180, 181

VTT, Finland

liquefaction

159, 167- 170, 182

Zeton, Canada

upgrading

183

flash pyrolysis

184

Table 6.11 Techno-economic Assessment of Catalytic Biomass Conversion Processes

TODiC

Oraanisation

References

Primary conversion

Pyrolysis

IEA

117, 168-170

SAI

171-174 285-288

University of Aston

165, 179 — 181

Zeton

184

Liquefaction

IEA

117, 167, 168- 170

SRI

175

Statens

167

VTT

159, 167- 170, 182

Gasification

University of Aston

4, 176- 178

Upgrading

Zeolites

IEA

159

University of Aston

165, 179- 181

Hydrotreating

IEA

117, 168-170

SAi

171 — 174

University of Aston

165, 179- 181

Zeton

183

Synthesis

University of Aston

176-178

Applications

Chemicals

BC Research

*15

Red Arrow

*25

NOTES * Market assessment

Many individual cost estimates and economic evaluations have been carried out which are difficult to compare as the bases are rarely published in sufficient detail for valid comparisons to be made. The IEA Bioenergy Agreement has also been sponsoring process, technical and economic assessments for over 10 years, beginning with the conceptual design of a biomass test liquefaction plant (167) and progressing through direct liquefaction, flash pyrolysis and upgrading (117, 159, 168 — 170) to the current examination of electricity generation systems. Some dedicated computer packages have been developed by Aston University which include process simulations and a common approach to cost estimation to provide a consistent comparison of alternative routes (176 — 181). This enables the user to examine alternative technologies, process conditions, routes, feedstocks and products.

Typical estimates of transport fuel costs from direct liquefaction processes are summarised in Figure 4 (296). In both cases the current pre-tax cost of fossil fuel derived products is shown. The conclusions agree well with those carried out by the IEA Bioenergy Agreement over the last few years (117).

For gasification derived synfuels, Table 13 summarises the ratio of costs of biomass derived liquid fuels to conventionally produced fuels as an example of analysis of results from cost estimates (177).

Basis: Capacity 1000 t/d d. a.f. feed,

Feedstock costed at 50 ECU/dry t (US$40/dry t)

Table 6.12 Comparison of Fuel Costs to Prices (292)

Product

Feed

Route

Uncertainty

Cost/Price

1

Methanol

straw

gasification

low-moderate

1.17

2

Gasoline

wood

pyrolysis + zeolites

high

1.48

3

Methanol

wood

gasification

low-moderate

1.46

4

Fuel alcohol

wood

gasification

moderate

1.56

5

Methanol

RDF

gasification

moderate

1.58

6

Gasoline

wood

pyrolysis + hydrotreating *

high

1.69

7

Gasoline

wood

liquefaction + hydrotreating

* high

1.81

8

Gasoline

wood

gasification + MTG

low

2.14

9

Diesel

wood

gasification + SMDS

low-moderate

2.15

10

Gasoline

wood

gasification + MOGD

moderate

2.52

11

Diesel

wood

gasification + MOGD

moderate

3.56

* includes consideration of lower quality product

The overall conclusion from all these assessments is that transport fuels cannot currently be produced economically from purpose grown biomass or energy crops without some economic support. For site specific wastes, there are opportunities for fuel gas and liquid fuels for heat and electricity production. The short term prognosis is, however, very hopeful as the more promising products only require modest improvements in process performance or cost to be competitive with fossil fuels (159). Catalyst development is likely to play a key role in meeting these objectives.

2.5 CONCLUSIONS

The use of catalysts to improve either the yield or quality of gas and liquid fuels from thermochemical biomass conversion processes is still in its infancy. While there is extensive fundamental work underway, considerably more research is necessary to explore the wide range of conventional and unconventional catalysts. Of particular potential significance is the integration of catalytic processes into the thermal conversion process to improve efficiency and reduce costs.

For fast pyrolysis processes, there are considerable opportunities for production of conventional and unconventional fuels for both electricity generation and fuel and chemical synthesis. The R&D requirements here are for more fundamental research into catalyst selection and evaluation for higher product specificity and/or higher yields of marketable products, since the products from these processes are much more complex than from gasification. Chemicals are always of greater potential interest due to their higher value compared to fuels. Here also catalysis has a significant role to play, and is more likely to justify more intensive R&D. Integrated fuel and chemicals production is the most likely scenario.

Union Electrica Fenosa fSpainl

Union Fenosa have successfully scaled up the Waterloo fluid bed process to 160 kg/h and this has been operating successfully since September 1993 producing around 400 kg/month (26). Total liquid yields of 55% wt on dry feed at 15% water are being achieved which continues to increase as the liquids recovery section is improved (3). Concepts have been announced for units up to 2 t/h (26)

3.5.2.6 University of Aston (ЦЮ

An ablative plate pyrolyser has been successful designed, built and operated at capacities up to 3 kg/h giving liquid yields up to 80% wt on dry feed (27). This will be further developed to establish design and performance prediction models as well as producing liquids for testing and chemicals recovery. A small conventional fluid bed has also been commissioned. Work on chemicals extraction and production has recently started including studies on catalytic pyrolysis.

3.5.2.7 University of Leeds (ЦЮ

Fixed and fluid bed pyrolysis has been carried out on biomass and wastes with catalytic upgrading of the products in the fluid bed work including co-processing with methanol. The fluid bed pyrolysis reactor is externally heated and nitrogen is used as a carrier gas (28). The maximum yield of liquid was 58.8 wt%, obtained at 720*C, Detailed analysis of the liquids by SEC, FTIR have been performed. A dual fluid bed has been used to pyrolyse biomass in an externally heated 75 mm diameter bed, 1 m high with nitrogen as the fluidising gas. Part of the reactor freeboard was packed with ZSM-5 catalyst and a secondary fluid bed has also
been used for zeolite upgrading. Liquids were collected before and after pyrolysis for comparison.

NATIONAL RENEWABLE ENERGY LABORATORY, USA

5.13.1 Summary

The original aim at the beginning of the 1980s was to convert biomass into non­equilibrium gases in a low pressure process without the addition of a catalyst in an analogous concept as the work at the University of Western Ontario which lead to the formation of Ensyn. This aim was modified in 1984 to produce liquid fuels in high yield, and more recently it has been modified again to consider production of hydrocarbon fuels such as gasoline, aromatics such as benzene, xylene and toluene, and chemical fractions such as polyphenols.

Interchem attempted to adapt the concept of the vortex reactor in their first pilot/demonstration plant at Springfield, MO, USA, (q. v.) but this was not successful. The second attempt was based on an NREL designed ablative vortex reactor and is shown below. While the reactor was built, the completed plant was never operated and the project is understood to have been abandoned.

NREL are continuing to operate their original pilot plant in Golden and has commissioned a second unit. There is more recent work on fluid bed fast pyrolysis (56), on MBMS fast pyrolysis (57) and on close coupled zeolite upgrading (see Chapter 6).

The description below focuses on the ablative vortex reaction system as the major R&D activity in fast pyrolysis.

5.13.2 Description

Initially a smooth walled vortex reactor with a gas recycle loop, made from Inconel 800H in order to withstand temperatures of 1000°C, was used. However, early experimentation demonstrated that severe coke deposits were formed at wall temperatures much above 625°C (58). Since a lower reactor temperature of 625°C was needed, a vortex reactor made from stainless steel was designed, constructed and tested. Very high organic vapour yields resulted from this lower wall temperature operation. The design capacity of the vortex reactor is 50 kg/h biomass but the maximum throughput achieved to date is 36 kg/h (59, 60, 61).

Figure 5.11 shows the current configuration of the reactor system. Biomass, with a particle size of about 5 mm, is metered into the system, where it is entrained and mixed with the recycle stream. The biomass particles, entrained in the carrier gas, enter the vortex reactor tangentially at speeds of over 400 m/s so that the particles are forced to the reactor wall by high centrifugal forces.

The reactor is made from 316 stainless steel with a diameter of 13.4 cm and a length 70 cm. The reactor is heated externally by three electric cylindrical furnaces. To force the particles into tighter helical paths than would naturally occur, a helical rib having a pitch of 25 mm and width and height of 3 mm was machined from the wall of the reactor. An insulated recycle loop is also added tangentially at the exit of the reactor to recycle partially pyrolysed feedstock and any large char particles.

GAS at 5°C

Подпись: HeatПодпись: LIQUIDПодпись: productПодпись: Steam ejector Подпись: wall heatersПодпись:image35The fine char, gases and vapours in the reactor leave through the axial exit which extends part way into the reactor. The reactor has a very high specific capacity and can in principle be easily scaled up. Very high heat transfer rates are achieved between the hot wall and biomass particles centrifuging against the hot reactor wall.

The wall temperature has to be limited to a maximum of about 625°C to ensure production of a liquid film between the wall of the reactor and the particle which then vaporises and leaves the reactor. The product stream then passes through a char cyclone where the char is removed. The diameter is 4 in. (10 cm) and operates at 475-500°C. The vapours pass to the first heat exchanger which is a 38 cm diameter cyclone. The condensed liquids and water are retained in the receiver. The cooled gas stream at about 80°C is then passed to a series of heat exchanger before passing through an orifice meter, and then to flare.

Empirical Modelling

This is the simplest modelling approach in that the model is based on the overall mass balance as shown in equation M.1.

biomass — > a char + b liquid + g gas M.1

Such model provides a simple stoichiometry specific to the reactor conditions used and are therefore highly specific and of limited use. functional or empirical relationships between yield and the process parameters may be derived but requires a substantial amount of experimental data.

2.3.3 Kinetic Modelling

Kinetic models are usually derived from particles with the hope that the Arrhenius type equations derived will represent the intrinsic reaction rates of pyrolysis. With small particles mass transport effects will be minimal and may therefore be neglected.

The complexity of the pyrolysis process implies that there are numerous homogeneous and heterogeneous reactions occurring either simultaneously and/or consecutively depending on the reactor conditions. Kinetic modelling is

therefore an attempt to represent the overall kinetics as individual reaction pathways are very difficult to determine.

PYROLYSIS LIQUIDS PRODUCTION TECHNOLOGY

4.1 INTRODUCTION

In the descriptions of pyrolysis pathways and models in Chapter 3, it has been shown how a high yield of fast pyrolysis liquids can be obtained from an optimal combination of very high heating rates; vapour temperatures below about 600°C; very low vapour residence times; and rapid quenching of the resultant vapours. Although the three main products of gas, liquid and solid are formed, both the gas and char are minimised through careful control of temperature and residence time at typically 15-20% weight of the dry feedstock. There is an trade-off between reaction temperature and residence time that has not yet been explored. Higher temperatures with low residence times gives an olefin rich gas which caused much interest in the early developments of flash pyrolysis until it was found that the upper limit on olefin yield was too low to be of commercial interest.

The high temperature required for pyrolysis can be obtained in several ways:

• heating the gas-solid reacting mix through the wall of the reactor;

• heating with a heat transfer medium that may be gas such as preheated recycle gas, or liquid such as molten metal or molten salt;

• heating through exothermic chemical reactions inside the reactor such as partial oxidation;

• heating the particle directly through the wall of the reactor into an ablative pyrolysis.

In all cases the heat transfer mechanism and controlling step is significant in the design of the reactor, the performance of the reactor and its ability to be scaled up 0).

Since high liquid yields result from very high heating rates, low vapour residence times and moderate temperatures, the method of energy transfer has to be very efficient and well controlled. This imposes constraints on the particle size and method of heat transfer. The thermal conductivity of most biomass is relatively low and thus rapid heating of a sufficiently high proportion of the particle to achieve flash pyrolysis conditions and give high liquid yields imposes an upper particle size of 3 to 5 mm. Above this characteristic dimension, the rate of heat penetration into the particle becomes too slow to achieve high heating rates and also results in severe degradation of primary products as they diffuse back through the hotter outer shell of the pyrolysing particle.

This chapter is devoted to the major pyrolysis processes that have been developed from first principles into working processes, and that in some cases have achieved demonstration scale and commercial operation. Direct thermal liquefaction has been reviewed by Elliott et al. (2), and a review of pyrolysis processes for liquid production carried out by Bridgwater and Bridge (1). A survey of commercial and advanced technologies has also recently been published which includes pyrolysis as well as gasification (3).

4.2.1 introduction

Before providing a detailed review of individual processes, the current situation is Europe is described and compared to that in North America.