Change search
ReferencesLink to record
Permanent link

Direct link
Influence of Reaction Atmosphere (H2O, N2, H2, CO2, CO) on Fluidized-Bed Fast Pyrolysis of Biomass Using Detailed Tar Vapor Chemistry in Computational Fluid Dynamics
KTH, School of Industrial Engineering and Management (ITM), Materials Science and Engineering, Energy and Furnace Technology.ORCID iD: 0000-0002-8045-6344
Aston University, European Bioenergy Research Institute (EBRI), Birmingham B4 7ET, U.K..
KTH, School of Industrial Engineering and Management (ITM), Materials Science and Engineering, Energy and Furnace Technology.ORCID iD: 0000-0002-1837-5439
KTH, School of Industrial Engineering and Management (ITM), Materials Science and Engineering, Energy and Furnace Technology.
2015 (English)In: Industrial & Engineering Chemistry Research, ISSN 0888-5885, E-ISSN 1520-5045, Vol. 54, no 33, 8344-8355 p.Article in journal (Refereed) Published
Abstract [en]

Secondary pyrolysis in fluidized bed fast pyrolysis of biomass is the focus of this work. A novel computational fluid dynamics (CFD) model coupled with a comprehensive chemistry scheme (134 species and 4169 reactions, in CHEMKIN format) has been developed to investigate this complex phenomenon. Previous results from a transient three-dimensional model of primary pyrolysis were used for the source terms of primary products in this model. A parametric study of reaction atmospheres (H2O, N2, H2, CO2, CO) has been performed. For the N2 and H2O atmosphere, results of the model compared favorably to experimentally obtained yields after the temperature was adjusted to a value higher than that used in experiments. One notable deviation versus experiments is pyrolytic water yield and yield of higher hydrocarbons. The model suggests a not overly strong impact of the reaction atmosphere. However, both chemical and physical effects were observed. Most notably, effects could be seen on the yield of various compounds, temperature profile throughout the reactor system, residence time, radical concentration, and turbulent intensity. At the investigated temperature (873 K), turbulent intensity appeared to have the strongest influence on liquid yield. With the aid of acceleration techniques, most importantly dimension reduction, chemistry agglomeration, and in-situ tabulation, a converged solution could be obtained within a reasonable time (∼30 h). As such, a new potentially useful method has been suggested for numerical analysis of fast pyrolysis.

Place, publisher, year, edition, pages
American Chemical Society (ACS), 2015. Vol. 54, no 33, 8344-8355 p.
Keyword [en]
Secondary pyrolysis, CFD, Fluidized bed, Fast pyrolysis
National Category
Chemical Process Engineering
Research subject
Materials Science and Engineering
URN: urn:nbn:se:kth:diva-172292DOI: 10.1021/acs.iecr.5b02164ISI: 000360322400038ScopusID: 2-s2.0-84940477488OAI: diva2:846256
Swedish Energy Agency, 33284-1

QC 20150918

Available from: 2015-08-14 Created: 2015-08-14 Last updated: 2015-10-23Bibliographically approved
In thesis
1. Pyrolysis of biomass in fluidized-beds: in-situ formation of products and their applications for ironmaking
Open this publication in new window or tab >>Pyrolysis of biomass in fluidized-beds: in-situ formation of products and their applications for ironmaking
2015 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

The iron and steel industry emitted 8 % of all CO2 emissions in Sweden, 2011. Investigating alternative energy carriers is the purpose of this thesis. By pyrolyzing biomass, an energetic solid, gaseous and liquid (bio oil) fraction is obtained. If pyrolyzing biomass in a fluidized-bed reactor, the highest value may be added to the combined products. Additional understanding of pyrolysis in fluidized beds is pursued, using Computational Fluid Dynamics (CFD) and comprehensive kinetic schemes. The obtained solid product is investigated as a bio-injectant in blast furnaces for ironmaking.

A new approach of separately modeling, the primary and secondary pyrolysis, is developed in this thesis. A biomass particle devolatilizes during pyrolysis. Primary pyrolysis is the solid decomposition which results in the volatiles that can leave the particle. Secondary pyrolysis is the decompositions of these volatiles, primarily in the gas phase.

The primary pyrolysis (35 species, 15 reactions) mainly occurs in the bed-zone and as such, the model needs to take into account the complex physical interaction of biomass-particles with the fluidizing media (sand) and the fluidizing agent (gas). This is accomplished by representing the components by Eulerian phases and implementing interaction terms, as well as using a Stiff Chemistry Solver for the implemented reactions. 

The secondary pyrolysis (not considering heterogeneous reactions), mainly occurs outside the bed zone in one phase. The fluid flow is simpler but the chemistry is more complex, with a larger variety of molecules emerging. Carrying out the simulations time-effectively, for the secondary pyrolysis (134 species, 4169 reactions) is accomplished by using Dimension Reduction, Chemistry Agglomeration and In-situ Tabulation (ISAT); in a Probability Density Functional (PDF) framework.

An analysis of the numerical results suggest that they can be matched adequately with experimental measurements, considering pressure profiles, temperature profiles and the overall yield of gas, solid and liquid products. Also, with some exceptions, the yield of major and minor gaseous species can be matched to some extent. Hence, the complex physics and chemistry of the integrated process can be considered fairly well-considered but improvements are possible. A parametric study of reaction atmospheres (or fluidizing agents), is pursued as means of understanding the process better. The models revealed significant effects of the atmosphere, both physically (during the primary and secondary pyrolysis) and chemically (during secondary pyrolysis).

During primary pyrolysis, the physical influence of reaction atmospheres (N2, H2O) is investigated. When comparing steam to nitrogen, heat flux to the biomass particles, using steam, is better distributed on a bed level and on a particle level.

During secondary pyrolysis, results suggest that turbulence interaction plays an important role in accelerating unwanted decomposition of the liquid-forming volatiles. Steam, which is one of the investigated atmospheres (N2, H2O, H2, CO, CO2), resulted in a lower extent of unwanted secondary pyrolysis. Altough, steam neither resulted in the shortest vapor residence time, nor the lowest peak temperature, nor the lowest peak radical concentration; all factors known to disfavor secondary pyrolysis. A repeated case, using a high degree of turbulence at the inlet, resulted in extensive decompositions. The attractiveness of the approach is apparent but more testing and development is required; also with regards to the kinetic schemes, which have been called for by several other researchers.

The solid fraction after pyrolysis is known as charcoal. Regarding its use in blast furnaces; modelling results indicate that full substitution of fossil coal is possible. Substantial reductions in CO2 emissions are hence possible. Energy savings are furthermore possible due to the higher oxygen content of charcoal (and bio-injectants in general), which leads to larger volumes of blast furnace gas containing more latent energy (and less non-recoverable sensible energy). Energy savings are possible, even considering additional electricity consumption for oxygen enrichment and a higher injection-rate on energy basis.

A survey of biomass availability and existing technology suppliers in Sweden, suggest that all injection into Blast furnace M3 in Luleå, can be covered by biomass. Based on statistics from 2008, replacement of coal-by-charcoal from pyrolysis could reduce the on-site carbon dioxide emissions by 28.1 % (or 17.3 % of the emissions from the whole industry). For reference, torrefied material and raw biomass can reduce the on-site emissions by 6.4 % and 5.7 % respectively.

Abstract [sv]

Järn och stålindustrin stod för 8 % av alla koldioxidutsläpp i Sverige, 2011. Alternativa energibärare undersöks i denna avhandling. Genom pyrolys av biomassa, fås en energirik fast produkt, och samtidigt en gasformig och en vätskeformig produkt (bio-olja). Om en fluidbäddsreaktor används kan största möjliga mervärde tillföras de kombinerade produkterna. Djupare förståelse för pyrolys i fluidbäddar har eftersträvats med hjälp fluiddynamikberäkningar (CFD) och detaljerade kinetikscheman. Den fasta produkten har undersökts som bio-injektion i masugnar.

En ny approach för modellering av primär och sekundär pyrolys separat, har utvecklats i denna avhandling. En biomassapartikel avflyktigas under pyrolys. Primär pyrolys är nedrytningen av den fasta biomassan till intermediärer (flyktiga ämnen) som kan lämna partikeln. Sekundärpyrolys är nedbrytning av dessa flyktiga ämnen, som primärt sker i gasfas.

Primärpyrolysen (i detta arbete, 35 ämnen och 15 reaktioner) sker mestadels i bäddzonen och därmed behöver modellen ta hänsyn till den komplexa fysiska interaktionen av biomassapartiklarna med fluidbäddsmediet (sand) och fluidiseringsgasen. Detta åstadkoms med hjälp av Euleriska faser och interaktionstermer, samt en lösare för hantering av styva reaktionssystem.

Sekundärpyrolysen sker huvudsakligen utanför bäddzonen. Fluiddynamiken är enklare men kemin är mer komplex, med fler ämnen närvarande. Att tidseffektivt köra beräkningarna, för sekundärpyrolysen (134 ämnen, 4169 reaktioner) åstadkoms med hjälp Dimensionsreducering, Kemiagglomerering och In-situtabulering (ISAT); som implementerats i en sannolikhetstäthetsfunktion (PDF).

En analys av de numeriska beräkningarna antyder att de kan matchas med experimentella resultat, med avseende på tryckprofil, temperaturprofil, utbyte av gasformiga, fasta och vätskeformiga produkter. Dessutom, med några undantag, kan beräkningarna matchas ganska väl med de viktigaste gasformiga produkterna. Därmed kan de huvudsakliga fysiska och kemikaliska mekanismerna representeras av modellen men förbättringar är givetvis möjliga. En parameterstudie av reaktionsatmosfärer (dvs fluidiseringsgaser) genomfördes, för att förstå processen bättre. Modellen visade på betydande effekter av atmosfären, fysisk (både under primär och sekundärpyrolys), och kemiskt (under sekundärpyrolysen).


Under primärpyrolysen undersöktes den fysiska inverkan av reaktionsatmosfärer (N2, H2O). När ånga jämfördes med kvävgas, visade det sig att värmeflödet sker mer homogent på både bäddnivå och på partikelnivå, med ångatmosfär.

Under sekundärpyrolysen, så antyder resultaten på att turbulensinteraktion spelar en viktig roll för accelererad oönskad sekundärpyrolys av de vätskebildande ämnena. Ånga som är en av de undersökta atmosfärerna (N2, H2O, H2, CO, CO2), resulterade i den lägsta omfattningen av sekundärpyrolys. Dock så ledde en ångatmosfär varken till den lägsta residenstiden, den lägsta peaktemperaturen eller den lägsta radikalkoncentrationen; som alla normalt motverkar sekundärpyrolysen. Ett repeterat case, med hög turbulens i inloppet, gav betydande sekundärpyrolys av de vätskebildande ämnena. Attraktiviteten av approachen är given men mer testning och utveckling behövs, som också påkallats av andra forskare.

Den fasta produkten efter pyrolys kallas träkol. Angående dess applicering i masugnar, så visar modelleringsresultaten att full substitution av fossilt kol går att göra. Betydande minskningar i koldioxidutsläpp är därmed möjliga. Energibesparingar är dessutom möjligt, tack vare det höga syreinnehållet i träkol (och biobränslen generellt), vilket ger större volymer av masugnsgas med högre värmevärde (och mindre sensibel värme som inte är utvinnbar). Energibesparingar är möjliga även om hänsyn tas till högre eleffekt för syrgasanrikning i blästerluften och en högre injektionsåtgång på energibasis.

En översikt över biomassatillgången och existerande teknikleverantörer i Sverige, indikerar att all injektion i Masugn 3 (i Luleå) kan ersättas med biomassa. Baserat på statistik från 2008, så kan ersatt kol med träkol, minska de platsspecifika koldioxidutsläppen med 28.1 % (eller 17.3 % av alla utsläpp från stålindustrin). Som jämförelse kan torrifierad biomassa and obehandlad biomassa reducera utsläppen med 6.4 % respektive 5.7 %.

Place, publisher, year, edition, pages
Stockholm: KTH Royal Institute of Technology, 2015. xii, 91 p.
Biomass, Pyrolysis, Fluidized bed, CFD, Blast furnace
National Category
Chemical Process Engineering Materials Engineering
Research subject
Materials Science and Engineering
urn:nbn:se:kth:diva-172293 (URN)978-91-7595-642-8 (ISBN)
Public defence
2015-09-25, D3, Lindstedtsvägen 5, KTH, Stockholm, 10:00 (English)
Swedish Energy Agency, 35819-1, 35386-1, 33284-1, 38685-1

QC 20150827

Available from: 2015-08-27 Created: 2015-08-14 Last updated: 2015-08-27Bibliographically approved

Open Access in DiVA

fulltext(3717 kB)3 downloads
File information
File name FULLTEXT01.pdfFile size 3717 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Other links

Publisher's full textScopusPublisher's website

Search in DiVA

By author/editor
Mellin, PelleYang, WeihongBlasiak, Wlodzimierz
By organisation
Energy and Furnace Technology
In the same journal
Industrial & Engineering Chemistry Research
Chemical Process Engineering

Search outside of DiVA

GoogleGoogle Scholar
Total: 3 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Altmetric score

Total: 51 hits
ReferencesLink to record
Permanent link

Direct link