Tesis Doctoral VALORIZACIÓN DE BIOMASA DE ORIGEN …

UNIVERSIDAD DE CASTILLA-LA MANCHA

FACULTAD DE CIENCIAS Y TECNOLOGÍAS QUÍMICAS

DEPARTAMENTO DE INGENIERÍA QUÍMICA

Tesis Doctoral

VALORIZACIÓN DE BIOMASA DE ORIGEN VEGETAL

MEDIANTE PROCESOS TÉRMICOS Y TERMOQUÍMICOS

DIEGO LÓPEZ GONZÁLEZ

Ciudad Real, 2013

UNIVERSIDAD DE CASTILLA-LA MANCHA

FACULTAD DE CIENCIAS Y TECNOLOGÍAS QUÍMICAS

DEPARTAMENTO DE INGENIERÍA QUÍMICA

VALORIZACIÓN DE BIOMASA DE ORIGEN VEGETAL

MEDIANTE PROCESOS TÉRMICOS Y TERMOQUÍMICOS

Memoria que para optar al grado de Doctor en Ingeniería Química

presenta

DIEGO LÓPEZ GONZÁLEZ

Directores: Dr. José Luis Valverde Palomino Dra. María Luz Sánchez Silva Composición del tribunal: Dra. Paula Sánchez Paredes Dra. Mª Pilar Coca Llanos Dr. Javier Dufour Andía Profesores que han emitido informes favorable de la tesis: Dr. Fernando Dorado Fernández Dra. Antonio Monzón Bescós

Ciudad Real, Julio de 2013

D. José Luis Valverde Palomino, Catedrático de Ingeniería Química de la

Universidad de Castilla-La Mancha, y Dª. María Luz Sánchez Silva, Profesor Titular

de Ingeniería Química de la Universidad de Castilla- La Mancha,

CERTIFICAN: Que el presente trabajo de investigación titulado: “Valorización e

biomasa de origen vegetal mediante procesos térmicos y termoquímicos”, constituye

la memoria que presenta D. Diego López González para aspirar al grado de Doctor en

Ingeniería Química y que ha sido realizada en los laboratorios del Departamento de

Ingeniería Química de la Universidad de Castilla-La Mancha bajo su supervisión.

Y para que conste a efectos oportunos, firman el presente certificado

En Ciudad Real a 2 de Julio de 2013

José Luis Valverde Palomino María Luz Sánchez Silva

vii

TABLE OF CONTENTS

Descripción del trabajo realizado 1

A. INTRODUCCIÓN 2

A.1. Cambio Climático y Sostenibilidad 2

A.2. Biomasa 3

A.3. Tipos de Biomasa 6

A.4. Aprovechamiento energético de la biomasa......................... 15

A.5. Energía Termosolar............................................................... 27

A.6. Objetivo del trabajo............................................................. 32

B. MATERIALES Y MÉTODOS.......................................................... 34

B.1. Reactivos empleados........................................................ 34

B.2. Instalación experimentales.................................................... 35

B.2.1. Instalación para el estudio termoquímico de biomasa 35

B.2.2. Instalación para el estudio de degradación de fluidos

de intercambio de calor.................................................... 35

B.3. Técnicas de caracterización............................................. 36

B.3.1. Específicas de biomasa......................................... 36

B.3.2. Caracterización de fluidos de intercambio de calor 38

C. DISCUSIÓN DE RESULTADOS...................................................... 41

D. CONCLUSIONES Y RECOMENDACIONES............................... 48

E. BIBLIOGRAFÍA......................................................................... 52

Abstract..................................................................................................... 58

CHAPTER 1: PYROLYSIS, COMBUSTION AND GASIFICATION CHARACTERISTICS OF NANNOCHLOROPSISGADITANAMICROALGAE

60

Table of contents

viii

1.1. Introduction 63

1.2. Experimental 66

1.3. Results and discussion 73

1.3.1. Pyrolysis of the NG microalgae 73

1.3.2. Combustion of the NG microalgae 78

1.3.3. Gasification of the NG microalgae 83

1.4. Conclusions 96

1.5. References 96

CHAPTER 2: THERMOGRAVIMETRIC-MASS SPECTOMETRIC ANALYSIS OF LIGNOCELLULOSIC AND MARINE BIOMASS PYROLYSIS

101




2.3.1. Thermogravimetric study of pyrolysis of lignocellulosic and marine biomass

112

2.3.2. Effect of heating rate 117

2.3.3. Gas products analysis 122

2.4. Conclusions 134

2.5. References 134

CHAPTER 3: THERMOGRAVIMETRIC-MASS SPECTROMETRIC STUDY ON COMBISTION OF LIGNOCELLULOSIC AND MARINE BIOMAS

135




3.3.1.Combustion of lignocellulosic biomass 144

3.3.2. Combustion of marine biomass 167

3.3.3. Combustion of Canadian biomass 184


Table of contents

ix

3.5. References 200

CHAPTER 4: 207




4.3.1. Thermogravimetric analysis 215

4.3.2. Gasification kinetic analyses 223

4.3.3. Gas evolution analyses 229


4.5. References 234

CHAPTER 5: CHARACTERIZATION OF DIFFERENT HEAT TRANSFER FLUIDS AND DEGRADATION STUDY BY USING A PILOT PLANT DEVICE OPERATING AT REAL CONDITIONS

238




5.3.1. Heat transfer fluids characterization for their use as thermal fluids in parabolic trough plants

252

5.3.2. Pilot plant assembly and tuning 261

5.3.3. Model validation 268


5.5. References 271

CHAPTER 6: General Conclusions and Recommendations 273

6.1. CONCLUSIONS 273

6.2. RECOMMENDATIONS 276

LIST OF PUBLICATIONS AND CONFERENCES 278

Descripción del trabajo realizado

1

DESCRIPCIÓN DEL TRABAJO

REALIZADO

Este trabajo da comienzo a una línea de investigación centrada en el desarrollo de

nuevas tecnologías para la valoración integral de biomasa en el Departamento de

Ingeniería Química de la Universidad de Castilla-La Mancha.

En particular, esta Tesis Doctoral tiene como objetivo la evaluación de los

principales procesos de conversión termoquímica de biomasa, principalmente pirólisis,

combustión y gasificación, mediante el sistema experimental de termobalanza

acoplada a un espectrómetro de masas. Adicionalmente, se estudió la degradación de

fluidos de intercambio de calor en su aplicación en plantas termosolares de

concentración. Este trabajo se encuadra dentro del proyecto CENIT VIDA basado en

un consorcio de colaboración de instituciones públicas (ministerio de economía y

competitividad) y privadas para el desarrollo de un nuevo concepto de BIO ciudad

basada en el aprovechamiento de biomasa. Concretamente, este proyecto ha recibido

la financiación de la empresa CT Ingenieros. Por otro lado, una parte de esta tesis ha

sido realizada en colaboración con el instituto de investigación canadiense


2

IRDA(Institut de recherche et de developpement en agroenvironnement) y el centro de

investigación francés CNRS (Centre national de la recherchescientifique).

A. INTRODUCCIÓN

A.1. Cambio Climático y Sostenibilidad

La demanda energética se ha incrementado exponencialmente en los últimos años

debido al crecimiento de la población mundial. Este hecho, junto con el agotamiento

de recursos fósiles y el auge de la conciencia global sobre la degradación del medio

ambiente son las principales razones que se proponen para realizar un cambio hacia

una sostenibilidad global.

El desarrollo sostenible se define como el “desarrollo que satisface las necesidades

del presente sin comprometer la capacidad de las generaciones futuras para atender

sus propias necesidades”, este cambio debe producirseen base a tres pilares

fundamentales: eficiencia energética, dependencia energética y razones

medioambientales.

La eficiencia energética supone la mejora de los procesos energéticos actuales en

términos de ahorro y desarrollo de nuevas tecnologías. Respecto la dependencia

energética, la fuerte dependencia de nuestra sociedad de las fuentes de energía de

origen fósil no renovable (petóleo, carbón y gas natural, principalmente) derivan en un

continuo agotamiento de las mismas. Además, los yacimientos de origen fósil se

encuentran concentrados en pocas regiosnes, lo que facilita las presiones políticas por

parte de los países productores. En la Figura A.1 se representa el consumo de energía

primaria en España (a) y mundial (b), para el año 2011 observándose que para ambos

casos alrededor del 85% de los recursos energéticos que se utilizan son de origen fósil:

petróleo, carbón y gas natural [1]. Por último, el deterioro del medioambiente debido

el aumento rápido e importante de las concentraciones de gases de efecto invernadero

(GEIs) siendo consecuencia del uso masivo e incontrolado de combustibles fósiles

desde la época industrial hasta la actualidad.


3

Las energías renovables, por su carácter autóctono, contribuyen a disminuir la

dependencia de los suministros externos, aminoran el riesgo de un abastecimiento

poco diversificado, favorecen el desarrollo tecnológico y la creación de empleo y

tienen un menor impacto medioambiental [1]. La utilización de tecnologías de

energías renovables como la eólica, la geotérmica, la hidráulica, la solar y la obtenida

a partir de la biomasa se presentan como alternativas para el reemplazo de los

combustibles fósiles. El presente trabajo está dedicado a dos de ellas: los procesos de

conversión de biomasa en energía y la mejora en la eficiencia de las plantas

termosolares.

Figura 2.1.Consumo de energía primaria (a) mundial y en (b) en España expresado en millones de toneladas equivalentes de petróleo [Mtep], año 2011 [1].

A.2. Biomasa

La biomasa ha sido desde siempre la mayor fuente de energía para el ser humano y

se ha estimado que actualmente contribuye un 14% al abastecimiento de la energía

mundial [2]. Una de las razones de que la biomasa haya tomado tanta importancia en

los últimos años es la elevada disponibilidad de la misma, estimándose en

aproximadamente 220 billones de toneladas secadas al año [3].

La biomasa se puede definir como “toda sustancia orgánica de origen vegetal o

animal que puede ser convertida en energía”[4]. Sin embargo, esta definición resulta

incompleta ya que estamos hablando de un vector energético que, a corto plazo, puede

ser básico en nuestra sociedad. Este término hace referencia a toda materia orgánica

Petróleo, 69,5

Gas natural, 28,9

Carbón, 14,9

Nuclear, 13

Hidráulica, 6,9Renovables,

12,7

CONSUMO DE ENERGÍA PRIMARIA EN ESPAÑA AÑO 2011, [Mtep]

Petróleo, 4059,1

Gas natural, 2905,6

Carbón, 3724,3

Nuclear, 599,3

Hidráulica, 791,5

Renovables, 194,8

CONSUMO DE ENERGÍA PRIMARIA MUNDIAL AÑO 2011, [Mtep]

a) b)


4

originada de forma inmediata en un proceso biológico, espontáneo o provocado,

utilizable como fuente de energía[1]. La biomasa abarca un amplio rango de materias

orgánicas que se caracterizan por su heterogeneidad.

A.2.1. Aplicaciones de la biomasa.

La existencia de diferentes tipos de biomasa y métodos de transformación de la

misma, permite utilizarla como combustible para producción de energía térmica y

eléctrica, o como materia prima para la producción de biocombustibles líquidos y

gaseosos.

• Producción de Energía Térmica. Este tipo de energía se obtiene

principalmente de la combustión directa de residuos forestales, agrícolas, de

industrias transformadoras de la madera y algunos agroalimentarios (orujillo de

aceituna, orujo lavado de uva, cáscara de almendra, etc.). En el proceso se genera

calor, tanto para su uso doméstico como industrial.

• Producción de Energía Eléctrica. Este tipo de energía también se obtiene

por la combustión principalmente de diferentes tipos de residuos como los

utilizados para la producción de energía térmica, pero también de los cultivos

energéticos y del biogás procedente de la digestión anaerobia de algunos residuos.

El rendimiento de las plantas que emplean biomasa para producción de energía

eléctrica no es muy elevado debido al elevado porcentaje de humedad que presenta

la biomasa.

• Producción de Biocombustibles Líquidos. La producción de

biocombustibles líquidos que suplan a los derivados del petróleo (gasolina y diesel)

es una opción muy ventajosa en cuanto al empleo de energías renovables y

reducción de problemas medioambientales. Existen dos tipos de biocombustibles

líquidos: los bioalcoholes (bioetanol) que se obtienen a partir de la fermentación

mediante levaduras de materiales azucarados como caña de azúcar, remolacha,

maíz, etc., y los biogasóleos (biodiesel) que se obtienen del proceso de

transesterificación de materiales oleaginosos como girasol, colza, etc., o bien de

grasas animales.


5

• Producción de Biocombustibles Gaseosos. La producción de

biocombustibles gaseosos a partir de procesos biológicos anaerobios es una opción

que presenta muchos beneficios. Este gas obtenido está formado principalmente

por metano, y aunque tiene bajo poder calorífico puede utilizarse en las propias

instalaciones donde es generado para producir tanto electricidad como calor. La

gasificación también conlleva la producción de un gas combustible rico en

hidrógeno y sobre todo en carbono.

A.2.2. Características de la biomasa para su aprovechamiento energético.

Son las propiedades inherentes de la biomasa las que van a determinar el proceso

de conversión y las consecuentes dificultades de proceso que puedan surgir[4]. Para

procesos de conversión de biomasa seca las propiedades más importantes son:

• Humedad:

Para la conversión térmica de biomasa son de interés aquellas biomasas que posean

baja humedad.Se pueden considerar dos formas de humedad en biomasa: humedad

intínseca, referida al contenido de humedad de la biomasa sin la influencia de los

efectos de la climatología, y humedad extrínseca, contenido de humedad de la biomasa

debido a los efectos de la climatología.Los procesos de conversión termoquímica

requieren materias primas con un contenido bajo de humedad (< 50%). Se podrían

usar materiales con mayor humedad pero el balance energético global para el proceso

de conversión se ve perjudicado por los procesos de secado.

• Poder calorífico (Energía/Masa) (MJ/kg):

El poder calorífico de un material es una expresión del contenido energético liberado

cuando el mismo se quema en aire.Se puede expresar de dos formas:

- HHV ( Higher heating value): Energía total liberada cuando el combustible es

quemado, incluyendo la del calor latente contenido en el vapor de agua y, por

tanto, representa la cantidad máxima de energía potencialmente recuperable dada

una fuente de biomasa determinada.


6

- LHV ( Lower Heating Value): Contenido energético sin contar el calor latente

contenido en el vapor de agua.Se puede decir que el calor latente contenido en el

vapor de agua no puede ser usado efectivamente y, por lo tanto, el LHV será el

valor apropiado para considerar la potencialidad de uso de la biomasa como

combustible.

• Proporción de carbón fijo y volátiles:

Este parámetro da una medida de cómo de fácil una biomasa determinada puede

ser inflamada y, consecuentemente, gasificada u oxidada.

- Contenido en volátiles: Porción liberada de gas mediante calentamiento (950

ºC durante 7 min).

- Carbón fijo: Es la masa que queda después de la liberación de los volátiles,

excluyendo la ceniza y humedad.

• Contenido de Ceniza/Residuo:

La rotura de los enlaces de la biomasa por procesos termoquímicos o bioquímicos

produce un residuo sólido.El contenido de ceniza puede afectar al manejo y a los

costes de proceso. En los procesos termoquímicos, la magnitud del contenido en

ceniza afecta a la cantidad de energía disponible en el combustible.

• Contenido en metales alcalinos:

Los principales metales alcalinos contenidos en la biomasa son Na, K, Mg, P y Ca.

El contenido de estos metales en la biomasa es un parámetro importante ya que

estos metales pueden catalizar/inhibir los procesos dde conversión de biomasa en

energía. Además. pueden reaccionar con los componentes de la ceniza produciendo

compuestos que pueden producir bloqueos en los equipos.

A.3.Tipos de biomasa

A.3.1. Clasificación de biomasa

La clasificación de la biomasa más ampliamente aceptada responde a su origen:


7

• Biomasa natural. Es la biomasa que se produce espontáneamente en la

naturaleza sin ningún tipo de intervención humana (recursos generados en las

podas naturales de un bosque).

• Biomasa residual. Es la biomasa que genera cualquier actividad humana. Se

distingue entre biomasa residual seca (aquella que procede de recursos generados

en las actividades agrícolas y forestales, en las industrias agroalimentarias y en las

industrias de transformación de la madera) y biomasa residual húmeda, como son

los vertidos biodegradables formados por aguas residuales urbanas e industriales,

los residuos ganaderos (generalmente purines) y también se incluyen los residuos

sólidos urbanos (materiales biodegradables sobrantes del ciclo de consumo

humano).

• Cultivos específicos (energéticos). Son cultivos realizados en terrenos

agrícolas y forestales dedicados exclusivamente a la producción de biomasa con

fines no alimentarios, sino únicamente energéticos (cardo, girasol, caña de azúcar,

maíz, remolacha, etc.). Éstos se dividen en leñosos y herbáceos.

• Biomasa marina. Como pueden ser algas, hierbajos marinos, juncos, etc.

El mayor punto de controversia encontrado en el uso de biomasa como fuente de

energía primaria reside principalmente en la competitividad con el abastecimiento de

humano comida. En este trabajo, se discutirá la conversión de biomasa lignocelulósica

y marina principalmente

A.3.2. Biomasa lignocelulósica

Una parte importante de la biomasa es lignocelulósica, siendo la celulosa, la

hemicelulosa y la lignina sus tres componentes principales. A diferencia de los

hidratos de carbono o el almidón, la lignocelulosa no es fácilmente digerible por los

seres humanos. Por ejemplo, se puede comer el arroz, que es un hidrato de carbono,

pero no podemos digerir la paja, que es lignocelulosa. La biomasa lignocelulósica no

forma parte de la cadena alimentaria humana y, por lo tanto, su uso para la obtención


8

de biogás y de energía, no suponen una amenaza para el suministro mundial de

alimentos[5].

La celulosa no es un material fácilmente accesible como es el almidón o el azúcar,

al encontrarse íntimamente unida a otros materiales como son la lignina o las

sustancias pécticas.Las paredes lignocelulósicas son estructuras complejas y de difícil

accesibilidad para algunos componentes (Figura A.2). El material lignocelulósico está

constituido por tejidos vegetales que presentan una pared celular constituida por un

entramado de microfribillas de celulosa sobre las que se forman capas de

hemicelulosas y, posteriormente, se deposita la lignina.

En este sentido, el aprovechamiento global del material requiere métodos de

pretratamiento o fraccionamiento. Estos procesos son complejos y están alejados de

rendimientos elevados. Además, no son capaces de aislar completamente cada

componente sin modificarlo o degradarlo.

Para comprender qué es un material lignocelulósico y poder aprovecharlo

completamente, se deben conocer cuáles son los componentes principales de las

paredes celulares, cómo se distribuyen en la propia pared y qué tejidos las contienen.

Figura A.2..Matriz lignocelulósica[6].


9

• Composición de la biomasa lignocelulósica.

Los componentes de los materiales lignocelulósicos se clasifican en estructurales y

secundarios.

Los componentes estructurales los forman tres polímeros, la celulosa, la lignina y

la hemicelulosa. Del total de compuestos que forman los materiales lignocelulósicos,

casi la mitad son celulosa y un 20% lignina.La unión entre celulosa y lignina puede

producirse directamente o generalmente a través de las hemicelulosas, como se

observa en la Figura A.2. En las paredes no lignocelulósicas aparece otro componente

formado por sustancias pécticas (pectina). En general, se puede establecer que entre un

60 y un 80% de los vegetales están constituidos por polisacáridos de elevado peso

molecular como son las holocelulosas. Entre las holocelulosas podemos distinguir

entre unos polímeros lineales de alto grado de polimerización, la celulosa y otros que

resultan fácilmente extraíbles en álcalis, las hemicelulosas.

- Celulosa: es un homopolímero lineal de elevado peso molecular y grado de

polimerización; entre 200 y hasta 10.000 unidades en estado nativo de β-D-

glucopiranosa unidas por enlace glicosídico o de tipo éter entre el carbono 1 y 4

(β,1�4).

En la Figura A.3 se muestra la estructura polimérica de la celulosa de forma

detallada.

Figura A.3. Estructura primaria de la celulosa[5].

- Hemicelulosas: forman cadenas ramificadas de menor grado de

polimerización que la celulosa y no tienen, por tanto, zonas cristalinas. Además, los


10

puentes de hidrógeno son menos eficaces, haciendo de las hemicelulosas

polisacáridos más accesibles al ataque de reactivos químicos.

El xilano se usa como compuesto representativo de la hemicelulosa por ser uno de

los compuestos principales de la hemicelulosa y se ha demostrado que tiene un

comportamiento térmico parecido(Wang y col., 2008; Yang y col., 2006) En la

Figura 2.5 se muestra la estructura del xilano.

Figura A.4. Estructura molecular del xilano .

- Lignina : es un polímero aromático de estructura tridimensional bastante

compleja, muy remificada y amorfa, formada por la condensación de precursores

fenólicos unidos por diferentes enlaces. En la Figura A.5 se muestran las unidades

estructurales de la lignina.La variedad de enlaces y estructuras del polímero lignina

son debidas a la diversidad de reacciones de acoplamiento (al azar) entre las

distintas formas resonantes de los radicales fenóxido.

Figura A.5. Unidades estructurales de la lignina .

Los componentes secundarios se clasifican en solubles en agua, disolventes

orgánicos e insolubles.

- Solubles en agua y disolvente orgánicos: conocidos como terpenos, son

considerados polímeros del isopreno. Por otro lado, las resinas que contienen una


11

alta variedad de compuestos no volátiles como son grasas, ácidos grasos, alcoholes,

resinas ácidas, fitoesteroides y otros compuestos neutros. Los fenoles, como los

taninos y también son solubles algunos hidratos de carbono de bajo peso

molecular, alcaloides y lignina soluble.

- Insolubles: dentro de este grupo se encuentran las cenizas, que son

principalmente carbonatos y oxalatos. Otros más raros y de poca proporción, pero

que también pueden ser insolubles, son pequeñas cantidades de almidón, pectinas o

proteínas.

A.3.3. Biomasa marina

Es la biomasa que producen los ecosistemas silvestres que se encuentran en los

océanos y corresponde al 40% de la biomasa que se produce en la Tierra (algas,

hierbajos marinos, juncos, etc.).

Las algas han atraído la atención desde hace mucho tiempo como posible materia

prima para la obtención de bioenergía[7-9], pero también la existencia de algunas

especies ricas en lípidos pueden ser explotadas como una alternativa interesante para

la producción de biodiesel[10; 11]. Son una biomasa muy prometedora por las

siguientes razones: alta velocidad de crecimiento, alto rendimiento por área, alta

eficiencia en la captura de CO2 y en la conversión de energía solar y no compiten con

la agricultura de alimentos. Además, pueden crecer en aguas abiertas (océanos o

estanques) y en bio-fotoreactores de tierra no cultivables [12]. La fijación de CO2 y las

principales etapas de transformación de biomasa marina se ilustran en la Figura A.6.

Las algas, que pertenecen al reino Protoctista y constituyen un grupo de

organismos muy variado y complejo, se encuentran en una amplia variedad de

ecosistemas acuáticos y terrestres gracias a su alta plasticidad y diversidad metabólica

y se pueden clasificar de acuerdo a su tamaño en los siguientes grupos:

• Microalgas: Incluyen todo tipo de microorganismos fotosintéticos,

procariotas o eucariotas, unicelulares o filamentosos, de tamaño inferior a 0,02

cm.


12

• Mesoalgas: Se trata de microorganismos fotosintéticos, procariotas o

eucariotas, unicelulares o filamentosos, unialgal o plurialgal, cuyo tamaño se

encuentra entre 0,02 y 3 cm.

• Macroalgas: Engloba a algas pluricelulares de diversas formas y tamaños que

van desde los pocos centímetros a varios metros de largo.

Las microalgas han recibido más atención que las macroalgas para la producción

de biofuel, las cuales pueden ser cultivadas en estanques o fotobiorreactores con

suministro de nutrientes o aguas residuales [13; 14].

Figura A.6. Fijación de CO2 y principales etapas de transformación de biomasa marina [15].

Las microalgas contienen en diferentes proporciones proteínas (6-52%),

carbohidratos (5-23%) y lípidos (7-23%) [16]. De acuerdo con Ross y col. (2010)[17],

las microalgas con un alto contenido en lípidos pueden ser una futura fuente de

biocombustibles de tercera generación y productos químicos.

CO2 en atmósfera CO2 en atmósfera H2O

Conversión bioquímica

Sustancias iniciales biofijación

Crecimiento de microalgas

Procesamiento microalgas

Conversión termoquímica

Conversión directa Bioalcoholes

Biodiesel Biogás

Biohidrógeno

Aceite de algas Combustión

Alimentos origen animal Fertilizante

Biofuel Biocrudo Biodiesel

Gas

Luz solar

Organismos fotosintéticos


13

A.3.4. Selección de la biomasa sometida a estudio

Como se ha comentado anteriormente, este trabajo se centra en el estudio de

biomasa lignocelulósica y marina, especialmente algas. La selección de los diferentes

tipos de biomasa sometidas a estudio se realizó en función de su composición.

• Selección de biomasa lignocelulósica

La elección de la biomasa depende, principalmente, de sus propiedades inherentes,

del proceso de conversión y de las dificultades de procesamiento posteriores que

puedan surgir. Las principales propiedades de interés para el tratamiento de biomasa

como fuente de energía son las siguientes como se comentó anteriormente:contenido

de humedad (MC), porcentaje de carbono fijo (FC) y proporción en volátiles (VM); la

relación cenizas / residuos (AC / AR), valor calorífico, contenido de metal alcalino y

la relación de celulosa / lignina[4].

En este sentido, se clasificaron diferentes especies de biomasa en un diagrama

ternario basándose en sus análisis inmediatos, realizados a partir de los datos

publicados por Yaman (2004) [18]. Se consideraron los siguientes parámetros:

cenizas, materia volátil y el contenido de carbono fijo (Figura A.7).

La biomasa se seleccionó de acuerdo con los siguientes criterios:

-Biomasa con alto contenido de VM y AC bajos.

-Biomasa de alto contenido FC y bajo AC.

De acuerdo con estos criterios, se identificaron dos zonas en el diagrama

claramente diferenciadas (señaladas con un círculo). La biomasa en estas dos zonas

correspondió a: madera de abeto y madera de eucalipto (ambos con elevada

proporción en volátiles) y corteza de pino (con el mayor contenido en carbono fijo).


14

Figura A.7. Diagrama ternario con diferentes tipos de biomasa terrestre según

su análisis inmediato [19].

• Seleción de biomasa marina (microalgas)

Para la selección de la microalga a utilizar en esta investigación se ha llevado a

cabo un intenso estudio bibliográfico teniendo en cuenta su composición y las

cantidades recomendadas de sus componentes para lograr los productos deseados.

Para ello, se realizó un diagrama ternario con sus tres componentes principales:

proteínas, carbohidratos y lípidos (Figura A.8) en base a los datos publicados por

Brrown y col. (1991) [20] y Renaud y col. (1999) [21].

El criterio que se empleó, se basó en la selección de la biomasa con mayor

contenido en lípidos [10]. Atendiendo a esto, se determinó que las microalgas que

reunían mejores propiedades fueron la Nannochloropsis Gaditana, la Scenedesmus

Almeriensis y la Isochrysis sp. De éstas, se seleccionaron las

microalgasNannochloropsis Gaditana (microalga NG) y Scenedesmus Almeriensis por

0,0 0,2 0,4 0,6 0,8 1,0

0,00,00,00,0

0,20,20,20,2

0,40,40,40,4

0,60,60,60,6

0,80,80,80,8

1,01,01,01,00,0

0,2

0,4

0,6

0,8

1,0

Carbón fij o

Caña de azúcarUvaMaízOlivaColzaCáscara de arrozSerrínGirasolHierbajo marinoJacinto de aguaAbetoTabacoPinoDesechos de algodónEucaliptoPaja

Cen

iza

Volátiles


15

su fácil disponibilidad y su comercialización en forma de polvo. Adicionalmente, se

seleccionó una especie de microalga con elevado contenido en proteínas a modo

comparativo. La especie de mciroalga con elevado contenido en proteínas elegida fue

la Chlorella Vulgaris.

Figura A.8. Diagrama ternario que representa la composición en proteínas, carbohidratos y lípidos de diferentes especies de microalgas.

A.4.- Aprovechamiento energético de la biomasa

Existen multitud de procesos para el aprovechamiento energético de la biomasa.

En la Figura A.9 se esquematizan los procesos más destacados. Este trabajo está

centrado en el aprovechamiento energético de biomasa mediante procesos de

conversión termoquímica, sin embargo se darán unas breves reseñas de otros procesos

de converión de biomasa.

0,0 0,2 0,4 0,6 0,8 1,0

0,0

0,2

0,4

0,6

0,8

1,00,0

0,2

0,4

0,6

0,8

1,0

Carbohidratos

Scenedesmus quadricauda Scenedesmus dinorphus Chlamydomonas rheinhardii Chlorella vulgaris Chlorella pyrenoidosa Spyroga sp. Dunaliella salina Tetraselmis maculata Porphyridium cruentum Spirulina maxima Synechoccus sp. A. coffeaformes Nitzschia sp. Cryptomonas sp. Rhodomonas sp. Nephroselmis sp. Tetraselmis sp. NT Isochrysis sp. Rhodosorus sp. Tetraselmis sp. TEQL Nannochloropsis gaditana

Prot

eína

s

Lípidos


16

Figura A.9. Procesos para la conversión energética de biomasa[5].

A.4.1. Procesos de conversión bioquímica.

Consisten en la aplicación de diversos tipos de microorganismos que degradan las

moléculas de biomasa. Se utilizan para la transformación de biomasa húmeda en

compuestos simples de gran contenido energético. Dos de las técnicas más

importantes son:

• Digestión anaerobia:

Es un proceso de fermentación bacteriana en ausencia de oxígeno donde se genera

una mezcla de gases, principalmente metano y dióxido de carbono, conocida como

biogás, y también una suspensión acuosa o lodo que contiene los compuestos no

degradados y los minerales. Se utiliza principalmente para la fermentación de biomasa

húmeda del tipo de residuos ganaderos, aguas residuales urbanas o biomasa marina

húmeda. En este caso se deben controlar una serie de variables como temperatura

(aprox. 35 ºC), acidez (pH 6.6-7.6), contenido en sólidos (< 10%), nutrientes (carbono,

nitrógeno, fósforo, azufre y sales minerales para el crecimiento y la actividad

bacteriana) y compuestos tóxicos (bajas concentraciones de amoníaco, sales

Biomasa

�Bioquímicos

�Termoquímicos

oCombustión

oGasificación

o Pirólisis

oLicuefacción

oTratamientoHidrotérmico

Digestión anaerobiao

Fermentaciónalcohólicao


17

minerales, detergentes y pesticidas que inhiben la actividad bacteriana). El biogás

puede utilizarse como combustible, mientras que el efluente (lodo) se puede utilizar

para la fertilización de suelos.

Este proceso ocurre en tres etapas consecutivas: hidrólisis, fermentación y

metanogénesis. En la hidrólisis, los compuestos complejos se dividen en azúcares

solubles. En ese momento, las bacterias fermentativas los convierten en alcoholes,

ácido acético, ácidos grasos volátiles y un gas que contiene H2 y CO2, el cual es

metabolizado principalmente en CH4 (60-70%) y CO2 (30-40%) por metanógenos

(Brennan y col., 2010).

• Fermentación alcohólica:

En el proceso de fotosíntesis las plantas almacenan la energía solar aportada en

forma de hidratos de carbono simples (azúcares) o complejos (almidón y celulosa). A

partir de estos hidratos de carbono se puede obtener por fermentación un bioalcohol,

denominado bioetanol, empleando diferentes etapas según el tipo de biomasa a

transformar. Estas etapas son las siguientes:

a) Pretratamiento: transformación de la materia prima para favorecer la

fermentación por medio de la trituración, molienda o pulverización.

b) Hidrólisis: transformación, en medio acuoso, de las moléculas complejas en

hidratos de carbono simples (azúcares) por medio de enzimas (hidrólisis

enzimática) o mediante reactivos químicos (hidrólisis química).

c) Fermentación alcohólica: conversión de los azúcares en bioetanol por la

acción de microorganismos (levaduras) durante dos o tres días bajo condiciones

controladas de temperatura (27-32 ºC), acidez (pH 4-5) y concentración de

azúcares (< 22%)

d) Separación y purificación del bioetanol: destilación de la masa fermentada

para obtener bioetanol comercial del 96% o destilación adicional con un disolvente

(benceno) para obtener un bioetanol absoluto del 99,5%.


18

El bioetanol es utilizado como combustible alternativo a las gasolinas, o bien

mezclado con ellas, en el campo de la automoción.

A.4.2. Procesos de conversión termoquímica.

Se utilizan para la transformación de biomasa seca, es decir, residuos cuyo

contenido en humedad no es muy elevado (principalmente paja, madera, orujillo,

huesos, cáscaras). Son métodos basados en la utilización del calor como fuente de

transformación de la biomasa donde se distinguen tres tipos de procesos según la

cantidad de oxígeno aportada:

• Pirólisis:

Se puede definir como la degradación de la biomasa mediante calor en ausencia de

oxígeno, resultando la producción de un sólido carbonoso (carbonilla o char),

biocombustible (líquido) y fuel gas [22]. A través de la variación de los parámetros del

proceso de pirólisis es posible influir en la distribución y características de sus

productos.El proceso de pirólisis se puede representar como la siguiente reacción:

�� → � �� +�� +��í�� !

�"� + ��#ℎ %�

Desde un punto de vista térmico el proceso se puede dividir en cuatro etapas

principalmente:

- Secado (100ºC): Ocurre en la etapa inicial de calentamiento a baja

temperatura, perdiéndose la humedad y, por tanto, el agua que está débilmente

enlazada.

- Etapa inicial (100-300ºC): En esta etapa, se produce la deshidratación

exotérmica de la biomasa, liberándose agua retenida y gases de bajo peso

molecular como el CO y el CO2.

- Etapa Intermedia (>200ºC): Se produce una pirólisis inicial, entre 200 y

600ºC, produciéndose la mayor parte del vapor o precursor de bio-combustibles. Se


19

comienza a romper las moléculas más grandes, descomponiéndose en el producto

sólido (Primary char), gases condensables (vapor y precursores del producto

líquido) y gases no condensables.

- Etapa Final (≈300-900ºC): La etapa final de pirólisis conlleva el craqueo

secundario de volátiles en producto sólido y gases no condensables. Si el tiempo de

residencia de la biomasa es suficientemente elevado, se puede producir el craqueo

de cadenas de elevado peso molecular en los gases condensables, incrementando el

rendimiento hacia el producto sólido y gases.Esta etapa ocurre principalmente por

encima de 300ºC. Si los gases condensables se retiran rápidamente del lugar de

reacción se produce la condensación hacia bio-combustibles o alquitrán.

• Combustión u oxidación:

La combustión es el proceso más directo para la conversión de biomasa en energía

útil, siendo usado en numerosas aplicaciones. Se basa en la oxidación completa de la

materia orgánica de la biomasa con exceso de oxígeno (cantidad de oxígeno superior a

la estequiométrica) convirtiendo la energía almacenada en calor, energía mecánica o

electricidad[23]. Además de calor, en el proceso se genera dióxido de carbono, agua y

cenizas.

El proceso de combustión se puede representar como la siguiente reacción:

�� +�" → ��" +�"� + ��#ℎ %� + � &�% La ignición de la biomasa requiere elevadas temperaturas (≥ 550 ºC),

constituyendo la etapa más costosa del proceso el comienzo del mismo.

A pesar de su aparente simplicidad, la combustión es un proceso complejo desde

un punto de vista tecnológico, donde tienen lugar elevadas velocidades de reacción y

grandes cantidades de calor liberado. Además de obtenerse muchos productos y

caminos de reacción.

De forma general el proceso de combustión se divide en las siguientes etapas:

- Secado: Evaporación del agua contenida en el combustible.


20

- Pirólisis y reducción: Descomposición térmica del combustible en volátiles y

un producto sólido (char).

- Combustión de los volátiles: Los productos obtenidos en la etapa anterior

son quemados en presencia de oxígeno.

- Combustión del char: Se produce la combustión del producto sólido.

• Gasificación:

La gasificación es un proceso termoquímico complejo que consiste en un número

de reacciones químicas elementales en presencia de un agente gasificante,

generalmente en atmósfera de aire, pobre de oxígeno (cantidad de oxígeno inferior a la

estequiométrica) o vapor de agua[24].

La importancia de este proceso se puede resumir en los siguientes puntos [5]:

- Incremento del valor calorífico de un combustible a través de la eliminación

de componentes como el nitrógeno y el agua.

- Eliminación de compuestos nocivos para el medioambiente, como pueden ser

los óxidos de nitrógeno y azufre.

- Reducción de la relación C/H del combustible.

- Obtención de productos químicos de gran interés comercial.

- Eliminación del oxígeno que constituye el combustible y, por lo tanto, se

produce un incremento de su densidad energética.

En general, cuanto mayor sea el contenido de hidrógeno de un combustible, menor

será la temperatura de vaporización y mayor la probabilidad de que el combustible

esté en estado gaseoso. La gasificación aumenta el contenido de hidrógeno en el

producto mediante una de las siguientes formas:

- Directa: Exposición directa al hidrógeno a alta presión.


21

- Indirecta: Exposición al vapor de agua en unas condiciones de temperatura y

presión elevadas, donde el hidrógeno (producto intermedio) se añade al

producto. Este proceso también incluye el reformado con vapor.

Las principales reacciones que ocurren en el proceso de gasificación se describen a

continuación [4] :

C + O" → CO" (Combustión Completa)

C + 1 2+ O" → CO (Combustión Incompleta)

La presencia de agua como agente gasificante, permite aumentar la proporción de

hidrógeno generado de la siguiente forma:

C + H"O → CO + H" (Reacción Water gas)

En presencia de dióxido de carbono, el carbono de la materia orgánica reacciona

para producir monóxido de carbono, según la reacción de Boudouard:

C + CO" → 2CO (Reacción de Boudouard)

También son importantes en el proceso de gasificación las reacciones de

metanización:

� + 2�" → ��- (Reacción de metanización)

�� + 3�"↔��- +�"� (Reacción de metanización)

Al igual que la reacción de water gas, la reacción de water gas shift tiene lugar

cuando existe vapor de agua en el medio de gasificación.

�� + �"� ↔ ��" +�" (Water gas shift reaction)

Las flechas indican que las reacciones están en equilibrio y se pueden producir en

cualquier dirección, dependiendo de la temperatura, presión y concentración de las


22

especies reaccionantes. De esto se deduce que el gas producto procedente de la

gasificación consiste en una mezcla de monóxido de carbono, dióxido de carbono,

metano, hidrógeno y vapor de agua.

• Otros Procesos:

Otros procesos para el aprovechamiento energético de la biomasa son el

tratamiento hidrotérmico y la licuefacción. El primero convierte la biomasa en una

atmósfera húmeda bajo presiones elevadas en hidrocarburos oxigenados parcialmente,

no obstante, este proceso está todavía en fase de planta piloto.

La licuefacción es la conversión de la biomasa en hidrocarburos líquidos estables

aplicando bajas temperaturas y elevadas presiones de hidrógeno. Este proceso está

atrayendo menos interés que la pirólisis ya que los reactores y los sistemas de

alimentación son más complejos y costosos [4].

A.4.3. Evaluación de procesos de conversión termoquímica mediante

tecnologías de análisis térmico (TA).

• Análisis termogravimétricos.

Durante los procesos de conversión termoquímica se presentan reacciones para las

cuales el estudio cinético resulta muy interesante. En esta tarea, el análisis

termogravimétrico (TGA) (realizado en un equipo denominado termobalanza) supone

una herramienta muy potente y es una de las técnicas más utilizadas a escala

laboratorio [25]. Consiste en medir la masa o el cambio de masa que experimenta una

sustancia en función de la temperatura mientras la muestra se calienta (o se enfría) con

un programa de temperaturas y bajo una atmósfera controlada (Montero, 2011).La

variación de masa puede ser una pérdida o una ganancia de la misma. El registro de

estos cambios nos dará información sobre si la muestra se descompone o reacciona

con otros componentes. La principal ventaja del análisis termogravimétrico es que

necesita un peso muy pequeño de muestra (escala de miligramos) para caracterizar un

proceso.


23

La termogravimetría se está usando muy ampliamente acoplada a otras técnicas,

como por ejemplo el análisis térmico diferencial (DTA) o calorimetría diferencial de

barrido (DSC), y también técnicas de gases producidos (EGA) ya que permiten

obtener información complementaria sobre el comportamiento de la muestra.

• Calorimetría diferencial de barrido (DSC).

La calorimetría diferencial de barrido (DSC; diferential scaning calorimetry)

permite el estudio de aquellos procesos en los que se produce una variación entálpica

como puede ser la determinación de calores específicos, puntos de ebullición y

cristalización, pureza de compuestos cristalinos, entalpías de reacción y determinación

de otras transiciones de primer y segundo orden.

En general, el DSC puede trabajar en un intervalo de temperaturas que va desde la

temperatura del nitrógeno líquido hasta unos 600 ºC. Por esta razón, esta técnica de

análisis se emplea para caracterizar aquellos materiales que sufren transiciones

térmicas en dicho intervalo de temperaturas.

La finalidad de la calorimetría diferencial de barrido DSC es registrar la diferencia

en el cambio de entalpía que tiene lugar entre la muestra y un material inerte de

referencia en función de la temperatura o del tiempo, cuando ambos están sometidos a

un programa controlado de temperaturas. La muestra y la referencia se alojan en dos

crisoles idénticos que se calientan mediante resistencias independientes. Cuando en la

muestra se produce una transición térmica, se adiciona energía térmica bien sea a la

muestra o a la referencia, con objeto de mantener ambas a la misma temperatura. Por

tanto, la DSC permite medir la energía que es necesaria suministrar a la muestra para

mantenerla a idéntica temperatura que la referencia. La energía térmica es

exactamente equivalente en magnitud a la energía absorbida o liberada en la

transición. Por tanto, mediante el uso de la técnica DSC se puede evaluar la cantidad

de calor liberado durante los procesos de conversión termoquímica.

• Técnicas de análisis térmico acopladas al análisis de gases residuales.


24

La principal limitación del análisis térmico en el estudio de procesos es que no te

proporciona informacción sobre los productos generados en los mismos. En este

sentido, se suelen utilizar técnicas complementarias acopladas al análisis térmico y

denominadas EGA, de sus siglas en inglés Evolved Gas Analysis. Existen diferentes

técnicas que se pueden utilizar con este fin, como la cromatografía de gases (GC),

espectroscopia de infrarrojo por transformada de Fourier (FTIR) ó la espetrometría de

masas (MS). Entre todas ellas destaca el acoplamiento de termogravimetría con la

especctrometría de masas (TGA-MS), siendo la única técnica experimental capaz de

monitorizar en tiempo real la distribución de productos generados en el proceso

partiendo de una muestra de bajo peso, enriqueciendo significativamente la

información del mecanismo de descomposición correspondiente [26].

A.4.4. Revisión de trabajos bibligraficos de los procesos de conversión

termoquímica de biomasa mediante TGA. DSC y TGA-MS.

• Referentes a biomasa lignocelulósica.

El proceso de pirólisis de biomasa lignocelulósica ha sido ampliamente estudiado

en biliografía mediante TGA y DSC [26]. Desde el estudio de descomposición de los

principales componentes de la biomasa lignocelulósica (celulosa, hemicelulosa y

lignina) [27-31] hasta diferentes ipos de madera [32; 33] u otros tipos de residuo [34].

Por otro lado, los estudios sobre el proceso de pirólisis de microalgas es

comparativamente mucho menor. Babich y col. (2011) [12] estudiaron la conversión

pirolítica de la microalga Chlorella mediante la técnica de TGA acoplada con MS. Las

muestras líquidas de biocombustible se recogen a partir de experimentos llevados a

cabo en un reactor de lecho fijo. Demirbas y col. (2011) [15]estudió la producción de

biocombustibles a partir de dos muestras de algas (Cladophora fracta y Chlorella

protothecoid). Para ello, investigó el efecto de la temperatura sobre la cantidad de

hidrógeno producido en los procesos de pirólisis y gasificación con vapor, estudiando

los gases producidos en dichos procesos. El proceso de pirólisis es muy importante ya

que es considerado como el primer paso de los procesos de combustión y gasificación.


25

El estudio de combustión de biomasa lignocelulósica mediante TGA es

significativamente menor comparado con el de pirólisis. Zhang y col. (2011)[35]

investigaron sobre las características de combustión de biomasa como la “paja” de

arroz y la celulosa contenida en ella. Estudiaron las diferencias entres las curvas TG-

DTG-DSC y estimaron los parámetros de TG y los índices de ignición de las muestras

de biomasa, obteniendo información sobre sus características básicas de combustión.

Por otro lado, Zhang y col. (2012)[36], mediante sus estudios de los procesos de

combustión obtuvieron resultados que mostraban que el proceso de combustión se

puede describir como una reacción de primer orden. Joaquín Collazo y col. (2012)[37]

investigaron sobre un método para la determinación del máximo error de muestreo y

los intervalos de confianza de las propiedades térmicas medidas mediante TGA-DSC.

Mustafa Versan Kok y Emre Özgür (2013)[38] estudiaron las características de

combustión de muestras de biomasa como miscanto, madera de álamo, y cascarilla de

arroz.Amutio y col. (2012) [39]analizaron la pirólisis oxidativa de biomasa

lignocelulósica con diferentes concentraciones de oxígeno para establecer un modelo

cinético para dicho proceso. El estudio de combustión de microalgas en cambio, ha

sido poco estudiado. Chen y col. [40] evaluaron el fecto de la concentración de O2 en

la combustion de la microalga Chlorella Vulgaris.

El proceso de gasificación de biomasa es probablemente el menos estudiado. La

mayoría de los estudios han sido dirigidos a la evaluación del comportamiento de

diferentes tipos de carbón gasificándolos con vapor de agua o dióxido de carbono. En

este sentido, Shabbar et al. [41] analizó la termodinámica de carbones bituminosos.

Además, Tay et al. [42] evalúo el effecto de diferentes agentes gasificantes en

diferentes tipos de carbones. Sin embargo, el estudio del proceso de gasificación de

biomasa ha sido mucho menos estudiado. Mohammed et al. [43] evalúo las cinéticas y

las características térmicas de residuo de frutas. Otros estudios han ido dirigidos a la

evaluación del proceso de gasificación del char obtenido a partir de la pirólisis de la de

diferentes tipos de biomasa[44-46]. En cambio, estudios de gasificación de biomasa

marina mediante TGA no han sido encontrados hasta la fecha.


26

Finalmente, el estudio EGA de los procesos de conversión termoquímica son

escasos en literatura. Huang y col. (2011)[47] investigaron la composición y las

propiedades térmicas de hemicelulosa, celulosa y lignina. Barneto y col. (2009)[48]

con el fin de optimizar el proceso térmico de pirólisis y tener un mayor conocimiento

de la evolución de los gases volátiles en el mismo para analizar dos muestras de

biomasa lignocelulósica.Li y col. (2003)[49]analizaron el comportamiento térmico y

caracterizaron los gases obtenidos en el proceso de combustión de trece especies

procedentes de China.Chul Yoon ycol. (2012) [50] estudiaron la pirólisis y la

gasificación de biomasa lignocelulósica y de sus principales componentes mediante

una combinación de termogravimetría y cromatografía de gases empleando aire o

vapor como agentes gasificantes en diferentes proporciones. Aghamohammadi y col.

(2011)[51] investigaron la emisión de los gases durante la combustión de madera

tropical, bambú, tronco de aceite de palma, acacia y madera de caucho utilizando la

técnica de análisis termogravimétrico acoplado a un espectrómetro de masas (TGA-

MS). Fang y col. (2006)[52] analizaron la pirólisis y la combustión de la madera bajo

diferentes concentraciones de oxígeno mediante la técnica TGA-FTIR, así como la

cinética de ambos procesos. Haykiri-Açma (2003)[53] estudió las características de

combustión de algunas muestras de biomasa terrestre tales como la cáscara de girasol,

las semillas de colza, el algodón y la piña mediante termogravimetría.

Babich y col. (2011) [12] estudiaron la conversión pirolítica de la microalga

Chlorella mediante la técnica de TGA acoplada con MS. Las muestras líquidas de

biocombustible se recogen a partir de experimentos llevados a cabo en un reactor de

lecho fijo. Demirbas y col. (2011) [15]estudió la producción de biocombustibles a

partir de dos muestras de algas (Cladophora fracta y Chlorella protothecoid). Para

ello, investigó el efecto de la temperatura sobre la cantidad de hidrógeno producido en

los procesos de pirólisis y gasificación con vapor, estudiando los gases producidos en

dichos procesos. Phukan y col. (2011)[14]caracterizaron el alga Chlorella sp mediante

espectroscopía FTIR y realizaron un estudio termogravimétrico de la misma a

diferentes velocidades de calentamiento para evaluar su viabilidad para la conversión

termoquímica. Miao y col. (2004)[54] utilizaron dos especies en sus experimentos,


27

Chlorella protothecoides y Microcystis aeruginosa, para investigar la pirólisis rápida

de ambas especies en un reactor de lecho fluido en una atmósfera inerte de N2 para

proceder, posteriormente, a la comparación con resultados obtenidos de pirólisis lenta

en un autoclave. Minowa y col. [55] realizaron el proceso termoquímico de

licuefacción a la especie de microalga Botryococcus braunii para la obtención de

combustibles líquidos y la recuperación de hidrocarburos.

A.5. ENERGÍA SOLAR DE CONCENTRACIÓN: COLECTOR CILIN DRO

PARABÓLICO.

A.5.1. Generalidades.

El empleo de colectores cilindro-parabólicos se remonta a 1880, John Ericsson

construyó un sistema de espejos cilindro-parabólicos para alimentar un motor de aire

caliente. Frank Shuman y C.V. Boys, fueron los primeros en utilizar este tipo de

espejos para la generación de energía de forma significativa, construyendo en 1912

una planta para el bombeo de agua con vapor en Meadi (Egipto) utilizando espejos

con una superficie total de captación de 1200 m2. A pesar del éxito alcanzado, la

planta se cerró en 1915 debido al inicio de la Primera Guerra Mundial y a los bajos

precios del petróleo.

Debido a la crisis del petróleo renació el interés en este tipo de tecnología siendo

principalmente el Departamento de Energía de los Estados Unidos y el Ministerio de

Investigación y Tecnología alemán los que impulsaron diversos prototipos solares

cilindro-parabólicos para la producción de vapor y para el bombeo de agua.

Posteriormente y basándose en la tecnología de espejos cilindros-parabólicos se

consiguió producir electricidad solar para cubrir las necesidad de miles de habitantes

en California (900 GWh/año). Estas centrales podían funcionar en modo solar o en

combinación con gas natural, asegurando de esta forma su disponibilidad

independientemente de las condiciones climatológicas o del ciclo día-noche. Las

centrales se encuentran en el desierto de Mojave y hoy continúan su funcionamiento


28

con 354 MW de potencia instalada, planificándose la construcción de más centrales en

sus alrededores.

En 1981, la Agencia Internacional de la Energía construyó y probó un sistema para

la producción de electricidad a base de captación solar mediante espejos cilindro

parabólicos de 500 Kw de potencia en la Plataforma Solar de Almería (Tabernas).En

esta plataforma se está investigando con todas las tecnologías termosolares. En 2008,

entró en funcionamiento Andasol I, en Granada, con 50 MW instalados, y también

cabe destacar la central de Iberdrola de 50 MW situada en Puertollano (Ciudad Real)

Aún así, este tipo de energía se considera que está en una fase de demostración de

viabilidad a gran escala, surgiendo cada día nuevos proyectos, con importantes retos

tecnológicos como el almacenamiento de calor o la hibridación con biomasa o gas

natural.

• Funcionamiento de una planta termosolar de colector cilindro-parabólico.

El esquema de funcionamiento de estas plantas es bastante simple. Se basan en un

campo de espejos con forma parabólica, que concentran la luz solar sobre un eje,

donde se encuentra una tubería por la que circula un fluido de intercambio de calor

(generalmente aceite). Este fluido caliente se introduce a la zona de generación, un

ciclo termodinámico convencional, donde se calienta agua para la producción de vapor

para el accionamiento de una turbina. Además, este tipo de centrales son combinadas

con otros tipos de combustibles, para los períodos de baja insolación.

En la Figura A.10 se muestra una imagen de un colector cilindro-parabólico en la

central termosolar de Almería y el diagrama de flujo de una planta termosolar (Flaberg

Solar International).Este tipo de plantas son capaces de calentar el fluido de

intercambio de calor hasta unas temperaturas entre 300 y 400 ºC (Razón de

concentración: 15-50), obteniendo rendimientos de hasta el 60 % y con capacidad de

320 MW.


29

Figura A.10.- Planta termosolar de colector cilíndro-parabólico en España (Plataforma

Solar de Almería) y diagrama de flujo de una planta termosolar (Flaberg Solar International)

(Forristal, 2003).

A.5.2. Fluido de Intercambio de Calor (HTF).

Los fluidos utilizados comercialmente son principalmente aceites compuestos por

mezclas eutécticas de óxido de difenilo y óxido de bifenilo. Estos HTF presentan una

serie de inconvenientes que se describen a continuación:

• Riesgos para la salud de operarios de planta. La degradación del aceite

térmico puede tener como consecuencia la aparición de aromáticos, que son

nocivos.

• Son compuesto tóxicos e inflamables.

• Producen una disminución en su función de transmisor de energía y daños

provocados en los equipos y tuberías por los que circula el fluido.

• Poseen una presión de vapor elevada, generando elevadas sobrepresiones.

Esto incrementa el coste de los recipientes para el almacenamiento de energía.

• Tienen una temperatura de degradación baja, alrededor de los 300 ºC,

disminuyendo la eficiencia del ciclo termodinámico para la producción de energía.

Por tanto uno de los principales retos que presenta este tipo de tecnología es el

cambio del fluido de intercambio de calor (HTF). Diversos autores, se han

encaminado en la búsqueda de fluidos capaces de reemplazar a los utilizados


30

comercialmente. Los principales esfuerzos, se han dirigido hacia las llamadas sales

fundidas. Estos estudios están encabezados por el Departamento de Energía de los

Estados Unidos[56; 57]. Otro tipo de fluidos, los líquidos iónicos, han abierto un

camino interesante para su sustitución [58; 59].

En la presente investigación, se pretenden estudiar diferentes HTF que puedan

mejorar los que actualmente se están empleando en la industria.

Con este fin, es importante un buen conocimiento de las propiedades específicas

requeridas para un buen intercambio de calor. Para el estudio preliminar de las

mismas, se consideró una lista de especificaciones propuesta por el Laboratorio

Nacional de Energías Renovables (NREL, 2000). En esta se especifica que la

capacidad de almacenamiento tiene que ser mayor de 1,9 MJ/m3, con un punto

decongelación inferior a 0 ºC y una estabilidad térmica por encima de los 430 ºC. La

presión de vapor debe ser inferior a la atmosférica para reducir el coste de recipientes

y debe tener una viscosidad adecuada para disminuir los costes de bombeo. Además,

como fluido de referencia se utilizarán las propiedades suministradas por el proveedor

del aceite térmico Therminol® VP-1 (Tabla A.1).

Tabla A.1.- Propiedades del HTF comercial Therminol®-VP1.

Propiedades Therminol-VP1

Punto de Cristalización 12 ºC

Humedad 300 ppm

Viscosidad Cinemática (40ºC) 2,48 cSt

Densidad 1060 kg/m3

Calor de fusión 97,3 Kj/kg

Temperatura de ebullición 257 ºC

Calor de vaporización 206 Kj/kg

Rango óptimo de uso, líquido 12-400 ºC

Rango óptimo de uso, vapor 260-400 ºC

Capacidad calorífica 100ºC 1,78 J/g ºC

Conductividad Térmica 100 ºC 0,1276 W/m K


31

• Propiedades de un Fluido de Intercambio de Calor (HTF).

- Punto de congelación.

El punto de congelación de un líquido es la temperatura a la que dicho líquido se

solidifica debido a la reducción de temperatura.Este parámetro es muy importante, ya

que un punto de congelación elevado (>0ºC) limita el uso de la planta en climas fríos,

derivando en un elevado coste asociado a la protección a la congelación que

requerirían las tuberías.

- Estabilidad Térmica.

La estabilidad térmica de los fluidos proporciona el límite de temperatura en el cual

se puede operar.La necesidad de establecer estos parámetros debidamente, se traduce

en dos aspectos, cuanto mayor sea la temperatura de descomposición el fluido va a ser

capaz de almacenar más energía térmica, por lo que hace más eficiente el ciclo

termodinámico para la producción de energía.

- Viscosidad.

Al tratarse de sistemas de fluidos en movimiento la viscosidad aparece como una

propiedad importante a la hora de operar en la planta solar.

- Capacidad calorífica:

Mide la cantidad de energía térmica que un cuerpo puede almacenar. La

importancia de su cálculo, se debe a que es necesario su determinación para el cálculo

de la capacidad de almacenamiento energético.

- Densidad.

No es una propiedad térmica, pero su cálculo es importante, puesto que es

necesaria para el cálculo de la capacidad de almacenamiento energético como se

describirá posteriormente.


32

- Capacidad de almacenamiento térmico: Calor sensible y Calor latente.

Esta variable define la capacidad de los mismos para almacenar calor.

La capacidad de almacenamiento térmico sensible, se puede calcular fácilmente

mediante la ecuación siguiente:

�0 = 2 ∙ �4 ∙ ∆6ª (A.1)

donde

HS = Capacidad de almacenamiento sensible (MJ/m3)

ρ = Densidad del fluido (kg/m3).

Cp= Capacidad calorífica del fluido (J/(kg K)).

∆T= Diferencia entre la temperatura de entrada y de salida del campo solar.

Para el cálculo de la capacidad de almacenamiento térmico latente se utilizará la

ecuación:

�8 = 2 ∙ ∆� (A.2)

donde

HL = Capacidad de almacenamiento latente (MJ/m3)

ρ = Densidad del fluido (kg/m3).

∆H= Entalpía de fusión/vaporización (J/kg).

A.6.- Objetivo del presente trabajo

En los apartados anteriores se ha puesto de manifiesto la importancia de las

energías renovables en el futuro desarrollo de nuestra sociedad. Entre estas tecnologías

cabe destacar el uso de la biomasa y la energía solar térmica como fuentes de energía

renovable. Sin embargo, el grado de desarrollo de las mismas no ha alcanzado una


33

madurez tecnológica que permita un cambio en el modelo energético actual basado

principalmente en el consumo de combustibles fósiles.

Los procesos de conversión termoquímica de biomasa son los procesos más

interesantes para el aprovechamiento energético de biomasa puesto que permiten

transformar la energía química de la biomasa en diferentes formas, como la

trasformación directa en energía (combustión) ó en combustibles líquidos, sólidos y

gaseosos (pirólisis y gasificación) para su posterior procesamiento.

Por otro lado, el cambio de los fluidos de intercambio de calor (HTF) utilizados

comercialmente en plantas termosolares de concentración basados en hidrocarburos

(mezas de difenilo y bifenilo) por otros obtenidos desde fuentes de energía renovable

con la capacidad de incrementar el ciclo térmico para la obtención de energía es

necesario para la optimización de estos procesos.

Por todo lo anterior, se consideró de interés realizar una investigación enfocada al

estudio de los principales procesos de conversión termoquímica (pirólisis, combustión

y gasificación) de diferentes tipos de biomasa (lignocelulósica y marina).

Adicionalmente, se evaluaron las propiedades físico-químicas de diferentes HTF para

su uso en plantas termosolares de concentración de colector cilindro-parabólico y se

puso en marcha una planta piloto para la evaluación de los mismos a escala semi-

industrial.

A tal fin, se planteó el siguiente programa de investigación:

- Revisión bibliográfica y puesta a punto de las distintas instalaciones

experimentales (equipos de análisis, calibración de gases, equipos de reacción, etc.).

- Diseño y construcción de una planta piloto para el estudio de degradación de

HTF para su aplicación en plantas termosolares de concentración.

- Definición de las principales características de HTF.

- Selección de biomasalignocelulósica y marina en base a su composición

química.


34

- Evaluación de las condiciones de operación óptimas en el sistema

experimental TGA-MS para el estudio de los principales procesos de conversión

termoquímicas (pirólisis, combustión y gasificación).

- Estudio de los procesos de pirólisis, combustión y gasificación de los

diferentes tipos de biomasa seleccionada.

- Modelización cinética de los procesos de pirólisis, combustión y gasificación.

- Caracterización de los HTF a estudio y selección del más apropiado para su

uso en plantas termosolares de concentración de colector cilindro-parabólico.

- Puesta a punto de la planta piloto para el estudio de degradación de HTF.

- Modelización de la degradación térmica del fluido comercial

MOBILTHERM® 605.

B. MATERIALES Y MÉTODOS

A continuación, se detallan tanto los reactivos como los gases utilizados, indicando

su concentración o pureza y la empresa suministradora.

B.1. Materiales.

Reactivos.

• Celulosa microcristalina con un tamaño de partícula medio de 50 µm. Fue

suministrada por la empresa Sigma-Aldrich.

• Lignina alcalina en forma de polvo marrón con un tamaño de partícula medio

de 50 µm. Fue suministrada por la empresa Sigma-Aldrich.

• Xilano elaborado a partir de madera de haya con un tamaño de partícula

medio de 100 µm. Se usó como referencia de la hemicelulosa y fue

suministrado por la empresa Sigma-Aldrich.

• Abeto, eucalipto y pino recogidos en la región de Castilla-La Mancha

(España). Estas muestras se secaron en un horno durante 5 horas y se

tamizaron para conseguir un tamaño de partícula medio entre 100 y 150 µm.


35

Gases.

• Argón, envasado en botellas de acero a 200 bares con pureza superior al

99,996% y suministrado por la empresa PRAXAIR.

• Nitrógeno, envasado en botellas de acero a 200 bares con pureza superior al


• Oxígeno, envasado en botellas de acero a 200 bares con pureza superior al


B.2. INSTALACIÓN EXPERIMENTAL

A continuación se detallan los diferentes equipos que se utilizaron para realizar los

diferentes desarrollos durante la presente investigación.

B.2.1. Calorimetría diferencial de barrido (DSC)

La diferente materia lignocelulósica fue analizada por calorimetría diferencial de

barrido (DSC) en un equipo TGA/DSC modelo 1 STAReSystem de METTLER

TOLEDO

B.2.2. Análisis termogravimétrico (TGA)

La pérdida de peso de los diferentes compuestos con la temperatura se analizó

usando un equipo TGA/DSC modelo 1 STAReSystem de METTLER TOLEDO. Este

equipo permite registrar con gran precisión la pérdida de masa de la muestra en

función de la temperatura/tiempo. Para ello, se debe establecer una secuencia de

calentamiento y configurar los gases circulantes por la cámara de reacción. La muestra

se coloca en unos crisoles de alúmina preparados para soportar las altas temperaturas

del ensayo.

B.2.3. Análisis termogravimétrico – Espectrometría de masas (TGA-MS).

Los productos liberados en el proceso de combustión se analizaron mediante el

acoplamiento de un espectrómetro de masas, ThermosStar-GSD320 con un analizador

de masa cuadrupolar y un potencial de ionización de 70 eV de PFEIFFER VACUUM

a un equipo TGA/DSC modelo 1 STAReSystem de METTLER TOLEDO.El principio

de funcionamiento de esta técnica se basa en la ionización de los componentes


36

producidos por la degradación térmica de la muestra, separándolos por su relación

masa carga (m/z).

Los espectrogramas obtenidos en cada experimento son almacenados y

cuantificados por el propio software informático suministrado con el equipo.

B.2.4. Análisis elemental

El análisis elemental permite obtener el contenido de la muestra en los

principales elementos químicos, como son carbono (C), hidrógeno (H), nitrógeno (N),

oxígeno (O) y azufre (S). Para llevar a cabo este tipo de análisis se utiliza un

analizador elemental, que es un equipo capaz de detectar todos los elementos citados

mediante diversos mecanismos y dar el resultado en porcentaje en masa de cada uno

de ellos en base seca. En el analizador elemental la separación de elementos de la

muestra se produce por combustión a alta temperatura (950 ºC) mediante la inyección

de una dosis elevada de oxígeno puro. Antes de ser introducidas en el mismo, las

muestras deben ser secadas para eliminar el hidrógeno y el oxígeno procedente de su

humedad, y así poder obtener los resultados en base seca.

El porcentaje de C, H, N y S será una media de los valores obtenidos en los

diez ensayos realizados a la muestra. El analizador elemental calcula automáticamente

estos datos. El porcentaje de oxígeno (O) de la muestra se calcula según la ecuación

[4.1]:

� = 100 − �� + � + ; + < + #=>? �� [4.1]

siendo O, C, H, N, S y cenizas los porcentajes en masa de oxígeno, carbono,

hidrógeno, nitrógeno, azufre y cenizas en base seca, respectivamente. El porcentaje en

cenizas de la muestra se determina mediante el análisis inmediato.

B.2.4. Análisis inmediato

El análisis inmediato permite determinar cuatro de las características químicas más

importantes de cualquier tipo de combustible:

• Humedad. Es la proporción de masa de agua libre que contiene el

combustible. El agua en el combustible puede encontrarse de dos formas

diferentes: libre o combinada. El agua libre se denomina humedad y es la que

se puede separar del combustible por simple calentamiento a 105 ºC. El agua


37

combinada forma parte de la estructura interna del combustible que, durante el

calentamiento, se combina con otros elementos para dar lugar principalmente

a hidrocarburos y para eliminarla es necesario calentar el combustible a

temperaturas comprendidas entre 150-185 ºC.

• Volátiles. Son las combinaciones de carbono, hidrógeno, oxígeno y otros

gases que contiene el combustible. El desprendimiento de volátiles es un

proceso exotérmico (desprende calor en el proceso de descomposición) que

ayuda al proceso de combustión de la biomasa.

• Cenizas. Son el residuo no orgánico de la combustión compuesto,

principalmente, por las materias minerales que acompañan al combustible. Se

trata de un residuo sólido no combustible, generalmente polvoriento, que

queda después de la combustión completa de la biomasa.

• Carbono fijo. El carbono fijo es la fracción residual de combustible,

descontadas las cenizas, que queda tras la desvolatilización del mismo. El

contenido en carbono fijo es un parámetro indicativo de la calidad del

combustible.

El equipo utilizado para realizar el análisis inmediato es el

analizadortermogravimétrico y permite medir la pérdida de peso de la muestra en

función de la temperatura en una atmósfera controlada. Para llevar a cabo los ensayos

se ha empleado el analizador termogravimétrico TGA/DSC modelo 1 STAReSystem

de METTLER TOLEDO.

El método utilizado para llevar a cabo este estudio consistió en calentar la muestra

de 25 a 950ºC a una velocidad de calentamiento de 10ºC/min en presencia de N2 con

un caudal de 70 ml/min. A continuación, se mantuvo la temperatura 950ºC durante 60

minutos en presencia de O2 con un caudal de 20 ml/min.

Las gráficas proporcionadas por el analizador termogravimétrico son pérdida de

peso vs temperatura y derivada peso vs temperatura, denominadas TGA y DTGA,

respectivamente. La curva TGA proporciona el contenido en volátiles, carbono fijo y

cenizas, mientras que la curva DTGA proporciona la velocidad de pérdida de masa en


38

cada punto de calentamiento dando una idea de la estabilidad térmica de la

descomposición de la muestra. Mediante el análisis de las gráficas TGA y DTGA en

atmósfera inerte y oxidante se puede determinar los contenidos en volátiles, carbono

fijo y cenizas de una muestra.

Finalmente, el contenido en carbono fijo de la muestra en base seca se calcula

según la ecuación [4.2]:

%� %A�>�BC� = 100 − �%D�&áF&=� +%�=>? �� [4.2]

donde los contenidos en volátiles y cenizas están expresados también en base seca.

B.2.5.Espectroscopía de emisión atómica de plasma acoplado por inducción (ICP-

AES)

Mediante esta técnica espectroscópica se determinó la composición química de la

biomasa objeto de estudio. En concreto, se utilizó para calcular el porcentaje en peso

de los distintos elementos metálicos de la muestra. El equipo utilizado para realizar los

análisis es el modelo VARIAN LIBERTY RL sequential ICP-AES de análisis

multielemental. La espectroscopia de emisión atómica se fundamenta en la excitación

de los átomos metálicos mediante un plasma de Argón, capaz de alcanzar 10000 K,

asegurando la completa atomización de la muestra en estado líquido. Al cesar la

excitación, tiene lugar la emisión de radiación por parte del metal para volver al estado

enérgico fundamental. La intensidad de dicha emisión permite cuantificar la

concentración del elemento ya que depende de la cantidad de átomos del mismo.

B.2.6. Microscopía electrónica de barrido (SEM)

Para evaluar la morfología y el tamaño de la microalga NG se utilizó un microscopio

electrónico de barrido Quanta 250 SEM con filamento de wolframio.

El microscopio electrónico de barrido es un instrumento que permite la

observación y caracterización superficial de materiales orgánicos e inorgánicos,

proporcionando información morfológica del material analizado. La formación de la

imagen se produce por la dispersión de los electrones. Esta capacidad de dispersión va

a depender de las distintas estructuras atómicas de la muestra. El microscopio

electrónico funciona como un microscopio convencional cuando las muestras son


39

conductoras. En cambio, cuando las muestras no son conductoras se pueden observar

utilizando el régimen de bajo vacío, y cuando las muestras son orgánicas se emplea el

régimen ambiental (ESEM). Para caracterizar la microalga NG se utilizó un detector

modelo GSED (Gaseous SED Detector).

B.2.7.Espectroscopía dispersiva de Rayos-X (EDAX)

Es una técnica analítica utilizada para el análisis elemental o caracterización

química de una muestra. Es una de las variantes de espectroscopia de fluorescencia de

Rayos X que se basa en la investigación de una muestra a través de interacciones entre

la radiación electromagnética y la materia, analizando los Rayos X emitidos por la

materia en respuesta al choque con partículas cargadas. Sus capacidades de

caracterización se deben en gran parte al principio fundamental de que cada elemento

tiene una única estructura atómica permitiendo que los Rayos X característicos de la

estructura atómica de un elemento sean identificados unos de otros.Para realizar este

análisis se utilizó el modelo APOLLO X acoplado a un microscopio electrónico de

barrido Quanta 250 SEM con filamento de wolframio.

B.2.8. Determinación de la cantidad de celulosa, hemicelulosa y lignina

El contenido de celulosa, lignina y xilano en las muestras de biomasa

lignocelulósica se calculó de acuerdo con el método descrito por Yang y col.

(2006)[27].

La determinación de la cantidad de extractos se llevó a cabo por extracción con

disolvente (100 ml de acetona para 1 gramo de muestra de biomasa seca) a 60 º C.

Después, la muestra de la biomasa se secó en un horno (110 º C) hasta que se obtuvo

un peso constante. Posteriormente, el residuo sólido se enfrió a temperatura ambiente

en un desecador y, finalmente, se pesó. La diferencia de peso antes y después de la

extracción es la cantidad de extractivos.

Para la determinación de la cantidad de hemicelulosa se añadieron 150 ml de

solución de NaOH (20 g / l) a 1 gramo de muestra de biomasa seca libre de extractos,

y la mezcla hirvió durante 3,5 h con agua destilada. El residuo se filtró y se lavó hasta

pH neutro y se secó en un horno. El residuo se enfrió posteriormente a temperatura


40

ambiente en un desecador y posteriormente se pesó. La diferencia de peso antes y

después de este tratamiento es la cantidad de hemicelulosa (Li y col., 2004).

La determinación de la lignina se llevó a cabo por el método de Klason. Se

añadieron 30 ml de H2SO4 (72%) a una muestra de 1 gramo de biomasa seca libre de

extractos. La mezcla se calentó y se agitó durante 2 horas. Posteriormente, la mezcla

se diluyó al 4% de concentración de H2SO4. La mezcla resultante hirvió durante 4

horas con agua destilada. El residuo se filtró y se lavó. Por último, se secó y se enfrió

a temperatura ambiente en un desecador. La diferencia de peso antes y después del

tratamiento es la cantidad de lignina.

Finalmente, se calculó la cantidad de celulosa por diferencia de peso asumiendo

que las muestras de biomasa están compuestas principalmente por extractivos,

celulosa, hemicelulosa y lignina.

B.3. PROCEDIMIENTO EXPERIMENTAL

B.3.1. Análisis termogravimétrico del proceso de combustión

La combustión de la biomasa lignocelulósica así como sus principales

componentes se llevó a cabo en el equipo TGA (TGA-DSC 1, METTLER TOLEDO).

Las muestras se precalentaron a 105 ºC durante 10 minutos para eliminar la humedad.

Después, la biomasa se calentó desde 105 ºC hasta 1000 ºC empleando diferentes

velocidades de calentamiento (10, 20, 40 y 80 ºC/min) en una atmósfera compuesta

por un 21% de oxígeno y un 79% de argón. Estos estudios se realizaron de acuerdo

con los trabajos realizador por Sánchez-Silva y col. (2013) para evitar las limitaciones

de transferencia de materia y calor. En este sentido, la cantidad de muestra inicial fue

de 6 mg, el tamaño de partícula se mantuvo en un rango entre 100-150 µm y se utilizó

un caudal constante de 100 Nml/ min.

B.3.2. Análisis de los productos gaseosos desprendidos en el sistema TGA-MS.

El análisis de los productos gaseosos desprendidos durante el proceso de

combustión se llevó a cabo en una termobalanza (TGA-DSC 1, METTLER TOLEDO)

acoplada a un espectrómetro de masas (Thermostar-GSD320 con analizador de masa

cuadrupolar; PFEIFFER VACUUM) con un potencial de ionización de 70 eV que


41

proporciona espectros hasta 300 a.m.u. La línea de conexión entre los equipos estaba

envuelta con hilo calefactor para evitar la condensación de los gases en esta zona.

Se realizó un análisis semicuantitativo usando un procedimiento de normalización.

Para ello, las intensidades de los iones se normalizaron con la intensidad del isótopo 38Ar para eliminar errores de instrumentación causados por la fluctuación del gas

portador, el peso de la muestra y cambios en la sensibilidad del espectrómetro de

masas [60]. Se calculó el área bajo la curva obtenida para cada uno de los gases

desprendidos, tomándose como criterio comparativo entre las diferentes muestras [61].

C. RESULTADOS Y DISCUSIÓN

El criterio empleado para la selección de la biomasa marina, se basó en la elección

de la microalga con mayor contenido en lípidos y una menor cantidad de proteínas y

carbohidratos. Por tanto, se realizó un diagrama ternario donde se determinó que la

microalga Nannochloropsis Gaditana (microalga NG) reunía mejores propiedades

para llevar a cabo este estudio.

En el Capítulo 1, el estudio de la pirólisis, combustión y gasificación de la

microalga NG se llevó a cabo mediante análisis termogravimétricos (TGA) y la

novedosa técnica de termobalanza acoplada a un espectrómetro de masa (TGA-MS),

siendo esta última la única herramienta capaz de detectar los compuestos que se

desprenden de una muestra de bajo peso a tiempo real.

Para la selección de las condiciones óptimas de operación en los procesos de

pirólisis y combustión se evaluaron las siguientes variables: masa inicial de muestra,

tamaño de partícula y caudal de gas reactivo.

En el estudio del proceso de pirólisis se observó que la microalga NG posee 3

etapas de degradación. Una primera etapa asociada a la eliminación de agua y

componentes más volátiles a temperaturas < 160ºC. La segunda etapa, donde se

produce la mayor pérdida de peso, asociada a la degradación de proteínas,


42

polisacáridos y lípidos. Y una tercera etapa (> 450ºC) donde se produce la

degradación térmica de la carbonilla.

En el proceso de combustión de la microalga NG se dividió también en 3 etapas.

Una primera etapa de secado asociada a la pérdida de agua a temperaturas < 125ºC. La

segunda etapa, donde se produce la mayor pérdida de peso, asociada a la

descomposición de proteínas, hidratos de carbono y lípidos. Y una tercera etapa (>

450ºC) donde se produce la oxidación de la carbonilla resultante.

Del estudio de las diferentes variables se observó, de forma general, que al

aumentar la masa inicial de muestra se produce una aceleración en el proceso de

combustión así como un aumento en la velocidad de descomposición, a diferencia de

lo que ocurría en el proceso de pirólisis. En el proceso de pirólisis, el efecto del

tamaño de partícula tiene poca influencia en el mismo, mientras que en el proceso de

combustión la muestra de menor tamaño es la más reactiva y se volatiliza antes. El

caudal de gas no afecta a ninguno de los dos procesos.

Posteriormente, se llevó a cabo un análisis de las condiciones óptimas de operación

en el proceso de gasificación mediante TGA. Al aumentar la temperatura de

gasificación se produce un aumento en los valores de reactividad y conversión. La

reactividad aumenta y la conversión disminuye cuando la masa inicial de muestra

disminuye y la porosidad de la muestra aumenta. Al aumentar el caudal de Ar,

disminuye la conversión y aumenta la reactividad, al igual que ocurre al aumentar el

% de vapor de agua.

En la segunda parte de este trabajo se estudió la distribución de los productos

gaseosos generados en los procesos termoquímicos de pirólisis, combustión y

gasificación utilizando las condiciones óptimas para cada uno de ellos y empleando la

técnica TGA-MS.


43

Mediante este análisis se concluyó que los productos generados durante el proceso

de pirólisis se liberan en tres etapas. En la primera etapa, se desprendió agua a

temperaturas < 160ºC. Posteriormente, en un rango de temperaturas entre 160 y 450ºC

se identificaron la mayor parte de los componentes, siendo el principal compuesto

detectado el CH3+ debido a la descomposición de los grupos carboxilos en las

proteínas y los polisacáridos, junto con HCN, CH4N, CO2, C3H8N, CO, C6H6 y otros

hidrocarburos volátiles como C2H5, C2H2 y CH4. Los compuestos nitrogenados son

liberados debido a la degradación térmica de las proteínas. Finalmente, a temperaturas

> 450ºC se produce la liberación de H2.

En cuanto a la segunda parte de la investigación, el estudio de la pirólisis de los

diferentes tipos de biomasa fue llevado a cabo mediante análisis termogravimétricos

(TGA) y mediante la novedosa técnica de termobalanza acoplada a un espectrómetro

de masas (TGA-MS). El uso de esta técnica se ha demostrado como la única capaz de

detectar los compuestos que se desprenden de una muestra de bajo peso a tiempo real.

Los tipos de biomasa sometidos a estudio fueron: la celulosa, la hemicelulosa y la

lignina (componentes mayoritarios de la biomasa terrestre), la madera de abeto y una

variedad de microalga (Nannochloropsis gaditana).

En el Capítulo 2, primero se llevó a cabo un análisis de las condiciones óptimas de

operación en el TGA-MS. Para el uso óptimo de esta técnica se tuvo en cuenta que a

elevadas cantidades de muestra inicial conllevan efectos de transferencia de materia y

de calor en el análisis termogravimétrico. Sin embargo, un bajo peso de muestra

inicial disminuye la detectabilidad en el espectrograma de masas. Por lo tanto, hay que

llegar a un compromiso en el que se utilice la mayor cantidad de masa inicial sin que

se produzcan limitaciones en el proceso por efectos de transferencia de materia y de

calor. Otras variables a estudio fueron la influencia del caudal de gas portador (He) y

de la velocidad de calentamiento. Las condiciones óptimas resultantes de este estudio

fue el empleo de un peso de muestra inicial de 10 mg, un flujo de gas portador He de

200 Nml/min y una velocidad de calentamiento de 40ºC/min


44

Posteriormente se llevó a cabo la evaluación del proceso de pirólisis para los

diferentes tipos de biomasa seleccionados mediante un análisis termogravimétrico. En

primer lugar se observó que la madera de abeto en el proceso de pirólisis se

descomponía en cuatro etapas, una primera etapa asociada a la eliminación de agua a

temperaturas ≤ 120ºC y tres etapas posteriores atribuidas a la descomposición de la

hemicelulosa (≈ 220ºC), de la celulosa y la lignina (300-400 ºC) y la lignina (>400ºC)

respectivamente. También se comprobó que el proceso de pirólisis tuvo lugar en un

rango de temperaturas de 200 a 500 ºC, intervalo de temperaturas donde ocurre la

mayor parte de la descomposición de la biomasa. Por último, se observó que la

biomasa marina estudiada se degrada a mayores temperaturas (≈ 1000ºC) que la

biomasa terrestre (≈ 700ºC).

Una vez estudiado el proceso de pirólisis, se evaluó el efecto de la velocidad de

calentamiento en la descomposición térmica de la biomasa mediante

termogravimetría. Para llevar a cabo este estudio se utilizó 5, 15 y 40ºC/min. La

velocidad de calentamiento influyó significativamente sobre la temperatura a la que

comienza el proceso de pirólisis y en la que se produce la mayor pérdida de peso. En

cambio, la velocidad no mostró un efecto tan claro sobre la cantidad de residuo

generado.

Empleando la técnica TGA-MS se estudió la distribución de los productos

generados durante la pirólisis de la biomasa. Mediante este análisis se concluyó que

los productos en la pirólisis se liberan en tres etapas. En la primera etapa, se

desprendió principalmente agua a temperaturas ≤ 120ºC. Posteriormente, en un rango

de temperaturas comprendido entre 200 y 450 ºC se identificaron la mayor parte de los

componentes volátiles, siendo el principal compuesto detectado el CO2 junto con CH4,

C2H6 y pequeñas cantidades de CO. Finalmente, a temperaturas ≥500 ºC se produce la

liberación de H2.

Finalmente, se desarrolló un modelo cinético que permitió estudiar, mediante

termogravimetría, el comportamiento de la biomasa durante su pirólisis a diferentes


45

velocidades de calentamiento. Con este fin, se empleó un modelo teórico de múltiples

saltos de descomposición basado en una expresión de velocidad tipo Arrhenius,

obteniéndose los parámetros cinéticos (energía de activación, factor pre-exponencial y

orden de descomposición) a cada velocidad considerada.

En el Capítulo 3, el estudio del proceso de combustión de la biomasa

lignocelulósica y de sus principales componentes se llevó a cabo mediante análisis

termogravimétricos (TGA), análisis de calorimetría diferencial de barrido (TGA/DSC)

y la novedosa técnica de termobalanza acoplada a un espectrómetro de masa (TGA-

MS), siendo esta última la única herramienta capaz de detectar los compuestos que se

desprenden de una muestra de bajo peso a tiempo real.

El proceso de combustión de la biomasa lignocelulóscia se divide en dos etapas

prinicipalmente. La primera etapa, donde se produce la mayor pérdida de peso,

llamada etapa de desvolatilización, en la cual se descomponen los principales

componentes de la biomasa (celulosa, hemicelulosa y lignina). Y una segunda etapa (>

441ºC) donde se produce la oxidación del residuo carbonoso (char) resultante.

La información obtenida mediante el análisis termogravimétrico se completó

mediante el estudio del proceso de combustión de las muestras de biomasa

lignocelulósica y de sus principales componentes por calorimetría diferencia de

barrido (DSC). En general, se observaron dos regiones exotérmicas, la primera se

atribuye a la etapa de desvolatilización y la segunda a la oxidación del char.

Una vez estudiado el proceso de combustión, se evaluó el efecto de la velocidad de

calentamiento en la descomposición térmica de la biomasa mediante

termogravimetría. Para llevar a cabo este estudio se utilizaron 10, 20, 40 y 80 ºC/min.

La velocidad de calentamiento influyó significativamente sobre la temperatura a la

que comienza el proceso de combustión y en la que se produce la mayor pérdida de

peso. En cambio, la velocidad de calentamiento no mostró un efecto tan claro sobre la

cantidad de residuo generado.


46

Posteriormente, se estudió la distribución de los productos gaseosos generados en

el proceso termoquímico de combustión empleando la técnica TGA-MS. Mediante

este análisis se concluyó que los principales productos gaseosos generados fueron:

CO, CO2 y H2O. También se produjeron hidrocarburos ligeros, atribuidos a reacciones

secundarias, como son CH4 y C2H5. La mayoría de los productos detectados fueron

generados durante la etapa de desvolatilización, mientras que sólo el NO2, C2H5O+,

CO y CO2 se detectaron en la segunda etapa (etapa de oxidación). Se detectaron

compuestos de nitrógeno, en mayor proporción que los compuestos de azufre,

liberados en forma de aminas primarias y NOx.

Finalmente, se desarrolló un modelo cinético que permitió estudiar, mediante

termogravimetría, el comportamiento de la biomasa durante su combustión a

diferentes velocidades de calentamiento. Con este fin, se empleó un modelo teórico de

múltiples saltos de descomposición basado en una expresión de velocidad tipo

Arrhenius, obteniéndose los parámetros cinéticos (energía de activación y factor pre-

exponencial) a cada velocidad considerada. Se encontró el mejor ajuste de los datos

experimentales con los modelos basados en el orden de reacción (Oi), la nucleación

(Ni) y la difusión (Di). Mediante una aplicación Excel-VBA se evaluó el conjunto de

ecuaciones diferenciales ordinarias que definen el modelo cinético. Se obtuvo un

modelo teórico y se comparó con los datos obtenidos experimentalmente.

En el proceso de combustión también se pueden observar tres etapas en la

liberación de las emisiones gaseosas. En la primera etapa, se desprendió agua a

temperaturas < 125ºC. Posteriormente, se produce la liberación de los compuestos

CO2, SO2, NO2, C3H8N y NH3 en la segunda etapa del proceso. La liberación de SO2

se atribuye a los radicales sulfatos existentes en los polisacáridos y a la degradación de

los sulfuros en los residuos orgánicos. Finalmente, a temperaturas > 450ºC se produce

la liberación de H2, NH3, NO2, CO y CH3+. Los hidrocarburos volátiles como CH4,

C2H2 y C2H5 no se generaron durante el proceso.


47

Los principales productos que se detectaron en el proceso de gasificación fueron

CO2, CO y H2, junto con CH4, C2H6, C2H5, C2H4 y C2H2 que también fueron

generados. Además, se estudió la influencia de la concentración de vapor en la

distribución de productos obtenida, utilizándose diferentes concentraciones de vapor

de agua. A medida que se incrementa la cantidad de vapor en el medio, también lo

hace la concentración de H2 y se produce la disminución de la producción de CH4.

Estos hechos indican que las reacciones water gas y water gas shift se ven favorecidas

al incrementar la concentración de vapor en el agente gasificante y la reacción de

reformado de metano podría estar teniendo lugar:

En el Capítulo 5 se concluyó que las propiedades más importantes que deben

reunir los fluidos térmicos para su aplicación en una planta termosolar son: un amplio

rango de temperaturas en estado líquido, una elevada temperatura de degradación, una

viscosidad baja, una temperatura de fusión baja, una densidad adecuada y una

capacidad calorífica elevada.

Posteriormente se realizó un estudio comparativo de cuatro tipos de HTF, dos sales

fundidas: Sal Solar y Hitec XL y dos líquidos iónicos: ([BMIM][BF4]) y

([EMIM][BF 4]). Del análisis comparativo se descartaron las sales fundidas ya que,

aunque poseían la mayor resistencia térmica (temperaturas de degradación ≥ 500ºC) su

temperatura de fusión es elevada, 230ºC para la Sal Solar y 120ºC para la Hitec XL,

estando limitada su aplicación a temperatura ambiente. Entre los dos líquidos iónicos

se determinó que el [BMIM][BF4] poseía las mejores propiedades para ser utilizado en

plantas solares, debido a que poseía una temperatura de degradación mayor y una

temperatura de fusión menor, y además contaba con propiedades térmicas similares al

[EMIM][BF 4].


48

D. CONCLUSIONES Y RECOMENDACIONES

De los resultados obtenidos en esta investigación se pueden obtener las siguientes

conclusiones finales:

De los resultados obtenidos en esta investigación se pueden extraer las siguientes

conclusiones.

1. En el estudio termogravimétrico del proceso de pirólisis de la microalga NG

se identifican tres etapas. La primera pérdida de peso atribuida a la pérdida de

agua y componentes más volátiles de la microalga (< 160ºC). En la segunda

etapa se distinguen tres hombros, el primero a 180ºC asociado a la

degradación de proteínas y polisacáridos solubles y dos picos a altas

temperaturas (271 y 411ºC) atribuidos a la degradación de la celulosa de la

pared celular de la microalga y otros polisacáridos insolubles y lípidos,

respectivamente. En la última etapa (> 450ºC) se produce la degradación

térmica de la carbonilla.

2. En las curvas TGA-DTGA asociadas al proceso de pirólisis se puede observar

como, en el estudio del efecto de la masa inicial de muestra un aumento de la

misma desplaza el proceso térmico a temperaturas más elevadas y disminuye

la velocidad de descomposición. En el caso del estudio del efecto del tamaño

de partícula y del caudal de gas portador (Ar), se observa que no tienen

influencia en el proceso ya que todas las curvas se solapaban indicando que no

existen limitaciones por transferencia de materia y de calor.

3. Las condiciones de operación óptimas para llevar a cabo el proceso de

pirólisis de la microalga NG fueron: masa inicial de muestra de 9 mg con un

tamaño de partícula de 100-250 µm, caudal de Ar de 200 ml/min y velocidad

de calentamiento de 40ºC/min.


49

4. Del estudio termogravimétrico del proceso de combustión se puede concluir

que el proceso está dividido en tres etapas. La primera etapa, que corresponde

a la etapa de secado, se atribuye a la pérdida de agua libre y débilmente ligada

a las biomoléculas. La segunda etapa (180-450ºC) se caracteriza por una

pérdida de peso importante debida a la descomposición de las proteínas, los

hidratos de carbono y los lípidos que constituyen la microalga. La última

etapa (> 450ºC) se corresponde con la oxidación de la carbonilla resultante.

5. En las curvas TGA-DTGA asociadas al proceso de combustión se puede

observar como la masa de muestra inicial tiene un efecto significativo en el

proceso. Al aumentar, desplaza el proceso de combustión a temperaturas más

bajas y aumenta la velocidad de descomposición. En el caso del efecto del

tamaño de partícula, la muestra con el menor tamaño es la más reactiva y se

volatiliza antes debido a que posee una mayor superficie que ofrece una

resistencia menor a la difusión. En cuanto al efecto del caudal de O2 no tiene

influencia sobre el proceso debido a que la concentración de O2 es constante.


combustión de la microalga NG fueron: masa inicial de muestra de 10 mg con

un tamaño de partícula de 100-250 µm, caudal de O2 de 100 ml/min y

velocidad de calentamiento de 40ºC/min.

7. En las curvas TGA-DTGA del proceso de gasificación se pueden observar las

distintas variables estudiadas. Un aumento en la temperatura de gasificación

produce un aumento en los valores de reactividad y conversión. La reactividad

aumenta y la conversión disminuye cuando la masa inicial de muestra

disminuye y la porosidad de la muestra aumenta. Al aumentar el caudal de Ar,

disminuye la conversión y aumenta la reactividad, al igual que ocurre al

aumentar el % de vapor de agua.


50


gasificación de la microalga NG fueron: temperatura de gasificación de 850ºC,

20 mg de muestra inicial con un tamaño de partícula de 100-250 µm en

presencia de 200 ml/min de Ar y un 5% de vapor de agua.

9. El estudio TGA-MS del proceso de pirólisis permitió obtener la distribución

de los productos generados en el mismo. Se identificaron principalmente tres

etapas. En la primera, se desprendió agua a temperaturas < 160ºC.

Posteriormente, en un rango de temperaturas entre 160 y 450ºC se

identificaron la mayor parte de los componentes, siendo el principal

compuesto detectado el CH3+ debido a la descomposición de los grupos

carboxilos en las proteínas y los polisacáridos, junto con HCN, CH4N, CO2,

C3H8N, CO, C6H6 y otros hidrocarburos volátiles como C2H5, C2H2 y CH4.

Los compuestos nitrogenados son liberados debido a la degradación térmica

de las proteínas. Finalmente, a temperaturas > 450ºC se produce la liberación

de H2.

10. En el proceso de combustión también se pueden observar tres etapas en la

liberación de las emisiones gaseosas. En la primera etapa, se desprendió agua

a temperaturas < 125ºC. Posteriormente, se produce la liberación de los

compuestos CO2, SO2, NO2, C3H8N y NH3 en la segunda etapa del proceso.

Finalmente, a temperaturas > 450ºC se produce la liberación de H2, NH3, NO2,

CO y CH3+. Cabe destacar que en este proceso el compuesto con el pico de

mayor intensidad es el CO2 y la liberación de SO2 puede atribuirse a los

radicales sulfatos existentes en los polisacáridos y a la degradación de los

sulfuros en los residuos orgánicos.

11. Mediante la comparación de las curvas del MS de los procesos de pirólisis y

combustión es posible evaluar el efecto de la presencia de oxígeno en la

degradación térmica o liberación de productos. En la combustión de la

microalga se produjo una mayor cantidad de CO2, observándose además, a


51

diferencia del proceso de pirólisis, emisiones de NO2 y SO2. Sin embargo, los

hidrocarburos volátiles como CH4, C2H2 y C2H5 no se generaron durante el

proceso.

12. Los principales productos que se detectaron en el proceso de gasificación

fueron CO2, CO y H2, junto con trazas de hidrocarburos ligeros como CH4,

C2H6, C2H5, C2H4 y C2H2. Además, se estudió la influencia de la

concentración de vapor en la distribución de productos emitida. A medida que

se incrementa la cantidad de vapor en el medio, también lo hace la

concentración de H2 y se produce la disminución de la producción de CH4.

Con objeto de ampliar y completar los resultados obtenidos en esta investigación se

recomienda:

- Investigar otros procesos menos comunes pero que pueden tener un gran interés

medioambiental como puede ser la oxy-combustión con CO2 y atmósfera deficiente de

O2.

- Escalar los experimentos de pirólisis, combustiión y gasificación en sistemas

experimentales para validar los resultados obtenidos mediante TGA-MS

- Uso de catalizadores para incrementar la generación de productos de interés,

como puede ser el H2 en el proceso de gasificación.

- Testeo de diferentes aceites de origen vegetal en la planta piloto diseñada y

optimizada para su uso en plantas termosolares de concentración.

- Evaluación de los aceites usados en el proceso de combustión, gasificación y

pirólisis en el sistema experimental TGA-DSC-MS.


52

E. BIBLIOGRAFÍA

[1] I.D.A.E. Instituto para la Diversificación y el Ahorro de Energía. Biomasa.

(http://www.idae.es/index.php/idpag.233/relcategoria.1037/relmenu.48/mod.pags/me

m.detalle).

[2] Harun, N.Y., Afzal, M.T. 2010. Thermal decomposition kinetics of forest residue.

Journal of Applied Sciences, 10(12), 1122-7.

[3] WEC. World Energy Council 2010. Survey of Energy Resources.

[4] McKendry, P. 2002. Energy production from biomass (part 1): Overview of

biomass. Bioresource Technology, 83(1), 37-46.

[5] Basu, P. 2010. Biomass Gasification and Pyrolysis: Practical Design and Theory.

Elsevier Science.

[6] Bidlack, J., Malone, M., Benson, R. 1992. Molecular Structure and Component

Integration of Secondary Cell Walls in Plants.

[7] Catallo, W.J., Shupe, T.F., Eberhardt, T.L. 2008. Hydrothermal processing of

biomass from invasive aquatic plants. Biomass and Bioenergy, 32(2), 140-5.

[8] Patil, V., Tran, K.Q., Giselrød, H.R. 2008. Towards sustainable production of

biofuels from microalgae. International Journal of Molecular Sciences, 9(7), 1188-95.

[9] Miao, X., Wu, Q. 2006. Biodiesel production from heterotrophic microalgal oil.

Bioresource Technology, 97(6), 841-6.

[10] Chisti, Y. 2007. Biodiesel from microalgae. Biotechnology Advances, 25(3),

294-306.

[11] Xu, H., Miao, X., Wu, Q. 2006. High quality biodiesel production from a

microalga Chlorella protothecoides by heterotrophic growth in fermenters. Journal of

Biotechnology, 126(4), 499-507.

[12] Babich, I.V., van der Hulst, M., Lefferts, L., Moulijn, J.A., O'Connor, P., Seshan,

K. 2011. Catalytic pyrolysis of microalgae to high-quality liquid bio-fuels. Biomass

and Bioenergy, 35(7), 3199-207.


53

[13] Bae, Y.J., Ryu, C., Jeon, J.K., Park, J., Suh, D.J., Suh, Y.W., Chang, D., Park,

Y.K. 2011. The characteristics of bio-oil produced from the pyrolysis of three marine

macroalgae. Bioresource Technology, 102(3), 3512-20.

[14] Phukan, M.M., Chutia, R.S., Konwar, B.K., Kataki, R. 2011. Microalgae

Chlorella as a potential bio-energy feedstock. Applied Energy, 88(10), 3307-12.

[15] Demirbas, A. 2011. Biodiesel from oilgae, biofixation of carbon dioxide by

microalgae: A solution to pollution problems. Applied Energy, 88(10), 3541-7.

[16] Brown, M.R., Jeffrey, S.W., Volkman, J.K., Dunstan, G.A. 1997. Nutritional

properties of microalgae for mariculture. Aquaculture, 151(1-4), 315-31.

[17] Ross, A.B., Biller, P., Kubacki, M.L., Li, H., Lea-Langton, A., Jones, J.M. 2010.

Hydrothermal processing of microalgae using alkali and organic acids. Fuel, 89(9),

2234-43.

[18] Yaman, S. 2004. Pyrolysis of biomass to produce fuels and chemical feedstocks.

Energy Conversion and Management, 45(5), 651-71.

[19] Sanchez-Silva, L., López-González, D., Villaseñor, J., Sánchez, P., Valverde,

J.L. 2012. Thermogravimetric-mass spectrometric analysis of lignocellulosic and

marine biomass pyrolysis. Bioresource Technology, 109, 163-72.

[20] Brown, M.R. 1991. The amino-acid and sugar composition of 16 species of

microalgae used in mariculture. Journal of Experimental Marine Biology and Ecology,

145(1), 79-99.

[21] Renaud, S.M., Thinh, L.V., Parry, D.L. 1999. The gross chemical composition

and fatty acid composition of 18 species of tropical Australian microalgae for possible

use in mariculture. Aquaculture, 170(2), 147-59.

[22] Demirbaş, A., Arin, G. 2002. An overview of biomass pyrolysis. Energy

Sources, 24(5), 471-82.

[23] Quaark, P., Knoef, H., Stassen, H.E. 1999. Energy from Biomass: A Review of

Combustion and Gasification Technologies. WORLD BANK PUBN.

[24] 2012. Review of the Literature on Catalytic Biomass Tar Destruction: Milestone

Completion Report. BiblioBazaar.


54

[25] Brown, M.E., Gallagher, P.K. 2011. Handbook of Thermal Analysis and

Calorimetry: Recent Advances, Techniques and Applications. Elsevier Science.

[26] White, J.E., Catallo, W.J., Legendre, B.L. 2011. Biomass pyrolysis kinetics: A

comparative critical review with relevant agricultural residue case studies. Journal of

Analytical and Applied Pyrolysis, 91(1), 1-33.

[27] Yang, H., Yan, R., Chen, H., Zheng, C., Lee, D.H., Liang, D.T. 2006. In-depth

investigation of biomass pyrolysis based on three major components: Hemicellulose,

cellulose and lignin. Energy and Fuels, 20(1), 388-93.

[28] Grønli, M., Antal Jr, M.J., Várhegyi, G. 1999. A round-robin study of cellulose

pyrolysis kinetics by thermogravimetry. Industrial and Engineering Chemistry

Research, 38(6), 2238-44.

[29] Lin, Y.C., Cho, J., Tompsett, G.A., Westmoreland, P.R., Huber, G.W. 2009.

Kinetics and mechanism of cellulose pyrolysis. Journal of Physical Chemistry C,

113(46), 20097-107.

[30] Nakamura, T., Kawamoto, H., Saka, S. 2007. Condensation reactions of some

lignin related compounds at relatively low pyrolysis temperature. Journal of Wood

Chemistry and Technology, 27(2), 121-33.

[31] Liu, Q., Wang, S., Zheng, Y., Luo, Z., Cen, K. 2008. Mechanism study of wood

lignin pyrolysis by using TG-FTIR analysis. Journal of Analytical and Applied

Pyrolysis, 82(1), 170-7.

[32] Rao, T.R., Sharma, A. 1998. Pyrolysis rates of biomass materials. Energy,

23(11), 973-8.

[33] Stenseng, M., Jensen, A., Dam-Johansen, K. 2001. Investigation of biomass

pyrolysis by thermogravimetric analysis and differential scanning calorimetry. Journal

of Analytical and Applied Pyrolysis, 58-59, 765-80.

[34] Caballero, J.A., Conesa, J.A., Font, R., Marcilla, A. 1997. Pyrolysis kinetics of

almond shells and olive stones considering their organic fractions. Journal of


[35] Zhang, C.C., Lu, H.J., He, Z. 2011. Combustion characteristics of biomass and

its cellulose, Vol. 306-307. Jinan, pp. 143-6.


55

[36] Zhang, C., Lu, H., He, X. 2012. Combustion process's kinetics study of biomass

and its cellulose based on biomaterials in material engineering, Vol. 568. Wuhan, pp.

328-31.

[37] Collazo, J., Pazó, J.A., Granada, E., Saavedra, Á., Eguía, P. 2012. Determination

of the specific heat of biomass materials and the combustion energy of coke by DSC

analysis. Energy, 45(1), 746-52.

[38] Kok, M.V., Özgür, E. 2013. Thermal analysis and kinetics of biomass samples.

Fuel Processing Technology, 106(0), 739-43.

[39] Amutio, M., Lopez, G., Aguado, R., Artetxe, M., Bilbao, J., Olazar, M. 2012.

Kinetic study of lignocellulosic biomass oxidative pyrolysis. Fuel, 95, 305-11.

[40] Chen, C., Ma, X., Liu, K. 2011. Thermogravimetric analysis of microalgae

combustion under different oxygen supply concentrations. Applied Energy, 88(9),

3189-96.

[41] Shabbar, S., Janajreh, I. 2013. Thermodynamic equilibrium analysis of coal

gasification using Gibbs energy minimization method. Energy Conversion and

Management, 65, 755-63.

[42] Tay, H.L., Kajitani, S., Zhang, S., Li, C.Z. 2013. Effects of gasifying agent on

the evolution of char structure during the gasification of Victorian brown coal. Fuel,

103, 22-8.

[43] Mohammed, M.A.A., Salmiaton, A., Wan Azlina, W.A.K.G., Mohamad Amran,

M.S. 2012. Gasification of oil palm empty fruit bunches: A characterization and

kinetic study. Bioresource Technology, 110, 628-36.

[44] Khalil, R., Várhegyi, G., Jäschke, S., Grønli, M.G., Hustad, J. 2009. CO2

gasification of biomass chars: A kinetic study. Energy and Fuels, 23(1), 94-100.

[45] Mani, T., Mahinpey, N., Murugan, P. 2011. Reaction kinetics and mass transfer

studies of biomass char gasification with CO2. Chemical Engineering Science, 66(1),

36-41.

[46] Gómez-Barea, A., Ollero, P., Fernández-Baco, C. 2006. Diffusional effects in

CO2 gasification experiments with single biomass char particles. 1. Experimental

investigation. Energy and Fuels, 20(5), 2202-10.


56

[47] Huang, Y.F., Kuan, W.H., Chiueh, P.T., Lo, S.L. 2011. Pyrolysis of biomass by

thermal analysis-mass spectrometry (TA-MS). Bioresource Technology, 102(3), 3527-

34.

[48] Barneto, A.G., Carmona, J.A., Alfonso, J.E.M., Blanco, J.D. 2009. Kinetic

models based in biomass components for the combustion and pyrolysis of sewage

sludge and its compost. Journal of Analytical and Applied Pyrolysis, 86(1), 108-14.

[49] Li, X., Matuschek, G., Herrera, M., Wang, H., Kettrup, A. 2003. A study on

combustion of Chinese coals by TA/MS. Journal of Analytical and Applied Pyrolysis,

67(2), 393-406.

[50] Yoon, H.C., Pozivil, P., Steinfeld, A. 2012. Thermogravimetric pyrolysis and

gasification of lignocellulosic biomass and kinetic summative law for parallel

reactions with cellulose, xylan, and lignin. Energy and Fuels, 26(1), 357-64.

[51] Aghamohammadi, N., Nik Sulaiman, N.M., Aroua, M.K. 2011. Combustion

characteristics of biomass in SouthEast Asia. Biomass and Bioenergy, 35(9), 3884-90.

[52] Fang, M.X., Shen, D.K., Li, Y.X., Yu, C.J., Luo, Z.Y., Cen, K.F. 2006. Kinetic

study on pyrolysis and combustion of wood under different oxygen concentrations by

using TG-FTIR analysis. Journal of Analytical and Applied Pyrolysis, 77(1), 22-7.

[53] Haykiri-Açma, H. 2003. Combustion characteristics of different biomass

materials. Energy Conversion and Management, 44(1), 155-62.

[54] Miao, X., Wu, Q., Yang, C. 2004. Fast pyrolysis of microalgae to produce

renewable fuels. Journal of Analytical and Applied Pyrolysis, 71(2), 855-63.

[55] Dote, Y., Sawayama, S., Inoue, S., Minowa, T., Yokoyama, S.y. 1994. Recovery

of liquid fuel from hydrocarbon-rich microalgae by thermochemical liquefaction.

Fuel, 73(12), 1855-7.

[56] Bradshaw, R.W., Cordaro, J.G., Siegel, N.P. 2009. Molten nitrate salt

development for thermal energy storage in parabolic trough solar power systems.

Conference Molten nitrate salt development for thermal energy storage in parabolic

trough solar power systems, 2, 615-24.

[57] Kearney, D., Herrmann, U., Nava, P., Kelly, B., Mahoney, R., Pacheco, J.,

Cable, R., Potrovitza, N., Blake, D., Price, H. 2003. Assessment of a molten salt heat


57

transfer fluid in a parabolic trough solar field. Journal of Solar Energy Engineering,

Transactions of the ASME, 125(2), 170-6.

[58] Valkenburg, M.E.V., Vaughn, R.L., Williams, M., Wilkes, J.S. 2005.

Thermochemistry of ionic liquid heat-transfer fluids. Thermochimica Acta, 425(1-2),

181-8.

[59] Wu, B., Reddy, R.G., Rogers, R.D. 2001. Novel ionic liquid thermal storage for

solar thermal electric power systems. Proceedings of the Solar Forum 2001 Solar

Energy, 445-51.

[60] Jakab, E., Till, F., Várhegyi, G. 1991. Thermogravimetric-mass spectrometric

study on the low temperature oxidation of coals. Fuel Processing Technology, 28(3),

221-38.

[61] Jones, J.M., Harding, A.W., Brown, S.D., Thomas, K.M. 1995. Detection of

reactive intermediate nitrogen and sulfur species in the combustion of carbons that are

models for coal chars. Carbon, 33(6), 833-43.

Chapter 1:

PYROLYSIS,COMBUSTION AND

GASIFICATION CHARACTERISTICS OF

NANNOCHLOROPSISGADITANA

MICROALGAE

Pyrolysis, combustion and gasification characteristics

ofNannochloropsisGaditanamicroalgae (NG microalgae)were

investigatedby thermogravimetric analysis (TGA).NG microalgae

pyrolysis and combustion could be divided into three main stages:

dehydration, proteins and polysaccharides degradation and char

decomposition. The effects of the initial sample mass, particle size

and gas flow on the pyrolysis and combustion processes were

studied. In addition, gasification operation conditions such as

temperature, initial sample mass, particle size, sweep gas flow and

steam concentration, were experimentally evaluated.

Chapter 1

63

Theevolved gases were analyzed online using mass

spectroscopy (MS).In pyrolysis and combustion processes, most

of the gas products were generated at the second degradation

step. N-compounds evolution was associated with the

degradation of proteins. Furthermore, SO2releasefrom

combustion could be related tosulphated

polysaccharidesdecomposition.The main products detected

during gasification wereCO2, CO, H2, indicating that oxidation

reactions, water gas and water gas shift reactions,were

predominant.

1.1. INTRODUCTION.

Recently,the utilization of biomass for transport fuels, chemical commodities,

power generation and reduction of CO2 emissions is growing interest[1]. Thus,

biomass has the potential of being an important renewable energy source.

Algae are a very promising biomass for the following reasons: a high growth rate,

high yield per area, high efficiency in CO2 capture and solar energy conversion and no

competition with food agriculture. In addition, they can be grown in open water (sea

water or ponds) and in bio-photo reactors on non-arable lands[2].

The generic term microalgae refer to a large group of very diverse photosynthetic

micro-organisms of microscopic dimensions. They have received more attention than

Chapter 1

64

macroalgae for biofuels production, which can be cultured in ponds or

photobioreactors with supply of nutrients or wastewater . Moreover, the production of

microalgae does not require of high quality arable land and therefore it does not

compete with food crops.

Generally, microalgae varied in their proportions of protein (6-52 wt.%), carbohydrate

(5-23 wt.%) and lipid (7-23 wt.%). Eustigmatophytes are rich in one or both of the

20:5(n-3) and 22:6(n-3) polyunsaturated fatty acids. According to Ross et al. (2009),

microalgae with high lipid content could be a future source of third generation

biofuels and chemicals.The oil content itself can be estimated to be 64.4 % of the total

lipid component. Thus, Nannochloropsis Gaditana (NG) microalgae, belongs to

Eustigmatophytes microalgae specie, have been proposed as a candidate to carry out

this study.

Interest towards the quality and characteristics of bio-oil from microalgae is

revived nowadays, due to growing concerns over emissions, energy supply, oil prices

and availability [2]. The conversion technologies for utilizing microalgae biomass can

be divided into two basic categories of conversion: thermochemical and biochemical .

Thermal technologies to process algae include direct combustion, pyrolysis and

gasification. Combustion is the conversion of biomass fuels to several forms of useful

energy in the presence of air or oxygen. Pyrolysis is a process that can be employed to

convert algal biomass material into biofuel and gas in the absence of air or oxygen

Chapter 1

65

(350-700ºC). Gasification involves the partial oxidation of biomass into a combustible

gas at high temperatures (800-900ºC).

On the other hand, biological processes can allow the conversion of biomass into

other fuels by means of anaerobic digestion, alcoholic fermentation and

photobiological hydrogen production. Despite its inherent potential as a biofuel

resource, the commercial viability of algal biofuel technology has not been achieved

yet.

A large number of researches on microalgae pyrolysis have been carried out in

recent years[2-4].Very few studies have been focused on the combustion

ofmicroalgae. Chen et al., (2011) reported the combustion behavior of Chlorella

vulgaris microalgae under different oxygen concentrations by thermogravimetric

analysis (TGA). Furthermore, Tang et al. [5] investigated the combustion of Chlorella

protothecoides microalgae in N2/O2 and CO2/O2 atmospheres by means of same

technique. [6]studied the effects of temperature in the combustion of marine algae.

However, at the best of our knowledge, the gasification process of marine biomass has

not been explored yet.

The pyrolysis behavior of brown algae has been investigated using

thermogravimetry and pyrolysis-GC-MS (PY-GC/MS) [1; 7] and thermogravimetry-

differential scanning calorimetry (TG/DSC) [3].

During the process of thermochemical conversion of biomass, the composition of

the gas emissions should be determined before industrial application. In this sense, the

Chapter 1

66

evolution with time on stream of the volatile products evolved in the marine biomass

pyrolysis or combustion processes has been carried out using the on-line combination

of TGA and Fourier Transform Infrared Spectrometry (FTIR) [8]and thermal analysis-

mass spectrometry (TA-MS) [9; 10].

As aforementioned, thermogravimetric analysis coupled with mass spectrometry

(TGA-MS) could be a useful technique to obtain information at real-time of mass loss

and evolved gases for pyrolysis, oxidation and gasification processes.

The aim of the present study was to investigate the pyrolysis, combustion and

gasification characteristics of theNannochloropsisGaditanamicroalgae by means of

TGA.In addition, the effects of different operation conditions were studied. Moreover,

evolved gases for the thermochemical conversion of NGmicroalgae were also

evaluated using the MS technique.

1.2. EXPERIMENTAL

1.2.1. Materials

NannochloropsisGaditana(NG microalgae) from Alga EnergyCompany was used

in this work. It was collected in Cadiz Bay (Spain) and delivered in green powder with

100 µm average particle size.Its composition in dry basis is about 17.6 wt. % of lipids,

12.6 wt. % of fatty acids and 24.1 wt.% of proteins.

The proximate and ultimate analyses of the NG microalgae are shown in Table

1.The proximate analyses were carried outaccording to the technical specifications

UNE-EN UNE-EN 14775:2010, UNE-EN 15148:2010 and UNE-EN 1474-2 for ash,

Chapter 1

67

volatile matter and moisture determination, respectively. Metal salts contained in

biomass have a significant impact on the thermal conversion processes[4]. In this

research, the content of metals in the sample was determined by Inductively Coupled

Plasma Spectrometry (ICP) (Table 1).

Table 1.Proximate, ultimate analysis and mineral content determined by ICP of the

NannochloropsisGaditanamicroalgae.

Ultimate analysis (wt.%)

Biomass C H N S Oa

NannochloropsisGaditana

47.26 7.03 6.72 0.49 38.5

Proximate analysis (wt. %)

Moisture Ash Volatile matter

Fixed Carbon

5.12 10.68 75.91 8.29

Mineral content (ppm)

Ca Fe Na K P Mg Zn

8652

5

170

7 23817 1385 9042

189

2 127

a % of oxygen calculated from difference of C, H, N and S.

Chapter 1

68

a) b) c)

Figure 1 shows the particle size distribution of theNG microalgae sample.The

morphology and the overall appearance of sample are shown in Figure SS2..

Figure 1.NG microalgae particle size distribution

Chapter 1

69

Figure SS2. (a) SEM micrograph of NG microalgae sample. (b) SEM micrographs of the

resulting char after the devolatilization step for 25-50 µm and (c) SEM micrographs of the

resulting char after the devolatilization step for >250 µm.

1. 2.2. Equipment and Procedures

Pyrolysis, combustion and gasification experiments were carried out in a TGA

apparatus (TGA-DSC 1, METTLER TOLEDO). Each sample was analyzed at least

three times, being the average value recorded. The experimental error in theweight

loss and temperature measurements was ± 0.5% and ± 2 ºC, respectively.

1.2.2.1. Thermal Analysis for the Pyrolysis process

The sample was heated from 40 to 1200ºC at a heating rate of 40 ºC/min under

Argon (99.996 %) atmosphere. Initial sample weight, Argon flow rate and particle size

of the sample were varied in order to obtain the most suitable operating conditions to

avoid the effects of heat and mass transfer limitations.

1.2.2.2. Thermal Analysis for the Combustion process

The sample was preheated at 125 ºC for 10 min in order to remove the moisture

content.Subsequently, the sample was heated from 125 to 1000 ºC under a reactive

atmosphere of pure oxygen (99.99 %). Initial sample weight, oxygen flow rate and

particle size of the sample were evaluated in order to obtain the most suitable

Chapter 1

70

operating conditions to avoid the effects of heat and mass transfer limitations. Finally,

the oxygen concentration was evaluated. On this account, experiments were performed

under atmospheres containing 20 %, 40 %, 60% and 80 % of Oxygen in Argon.

1.2.2.3. Thermal Analysis for the Gasification process

Figure SS1 shows the experimental set-up used for the gasification process.

Gasification experiments were conducted in the presence of water vapor generated by

a bubbler system. Ar was bubbled through degassed water heated to different

temperatures. The effect of the gasification temperature, initial sample weight, Argon

flow rate, particle size of the sample and water vapour concentration were evaluated.

The gasification of the sample was performed in three steps:

• Drying: the sample was heated in an inert atmosphere of pure Ar from 30 to

125 ºC at a heating rate of 15 ºC/min.

• Pyrolysis: the sample was heated from 125 ºC to the operating temperature at

a heating rate of 40 ºC/min. Ar was used as the carrier gas (200 ml/min (25 ºC,

0.9 atm)).

• Gasification: the char obtained in the pyrolisis process was later gasified with

the reactive gas mixture (Ar + H2O) at the test temperature for one hour.

In this paper, X was the char conversion, which is defined as:

� =�� −�

�� (eq. 1)

Chapter 1

71

FIC

FIC

Bubble

Flow meter

Thermobalance (TGA)

Mass

spectrometer

(MS)

N2

CO

O2

He

TGA Flow meter

Bubbling system

01

O 1

Ar

PC

PC

where w and w0 are the weight of char at any instant and at initial conditions,

respectively.

The reactivity R(s-1) was defined as:

� =1

� − ��∙�

=

1

1 − �

�

(eq. 2)

The reactivity at 50% char conversion was taken as a representative value[11].

Chapter 1

72

Figure SS1. Experimental set-up for the gasification process.

1.2.2.4. TGA-MS Analysis of the Gaseous Products

The analysis of the gas products distribution coming from the thermal analysis was

carried out in a thermogravimetricanalyzer (TGA-DSC 1; METTLER TOLEDO)

coupled to a mass spectrometer (Thermostar-GSD 320/quadrupole mass analyzer;

PFEIFFER VACUUM) with an electron ionization voltage at 70 eV and provided

mass spectra up to 300 a.m.u. The interface was wrapped with heating wire to

circumvent condensation of exhausting gases. Pyrolysis, combustion and gasification

experiments were carried out under the selected operating conditions. In order to

identify ions with m/z in the range 0-300, a preliminary broad scan was performed at a

heating rate of 40 ºC/min.

Although a quantitative analysis was not performed in this work, a comparison of the

intensity peak areas between different samples (semiqualitative analysis) was carried

out by using a normalization procedure. The method used in this work was based on

the relative integrated peak linear intensity normalized to total integrated peak linear

intensities and to sample weight (eq. 3), that are reported elsewhere [12; 13].

�(� )� =� �

((∑ � �) ∙ �)

(eq. 3)

where R(Int)i is the relative integrated peak linear intensity, Inti is the integrated

intensity of a gas species, and m is the mass of the sample. Ion fragments with R(Int)

Chapter 1

73

0

20

40

60

80

100

0

20

40

60

80

100

100 200 300 400 500 600 700 8000

20

40

60

80

100

(c)

(b)

4 mg 7 mg 9 mg 15 mg 24 mg

(a)

0.0

0.1

0.2

0.3

0.4

0.5

Wei

ght (

%)

25-50 µm 50-100 µm 100-250 µm > 250 µm

0.0

0.1

0.2

0.3

0.4

0.5

Wei

ght l

oss

rate

(%

wt./

ºC)

Temperature (ºC)

50 ml/min 100 ml/min 150 ml/min 200 ml/min

0.0

0.1

0.2

0.3

0.4

0.5

< 0.5 nA min/mg were not considered as their intensity is considered to be too close to

the noise level [13].

1.2.2.5. Scanning electron microscopy (SEM) observation.

The surface features and porosity of samples were evaluated usingQuanta 250 (LFD)

SEMequipped with an energy dispersive X-ray spectroscopy (EDS).

1.3. RESULTS AND DISCUSSION

1.3.1. Pyrolysis of the NG microalgae

Figure 2 shows the weight loss curvesof the pyrolysis of NG microalgae for different

initial sample weights, particle sizes and sweep gas flows at a heating rate of

40ºC/min. Table 2 summarizes the most relevant pyrolytic characteristics of NG

microalgae.

Chapter 1

74

Figure 2.Thermogravimetric (TGA) and differential thermogravimetric (DTG) curves of the

NG microalgae pyrolysis process as a function of: (a) initial weight, (b) particle size and (c) gas

flow rate at a heating rate of 40 ºC/min.

Chapter 1

75

Table 2.Pyrolysis characteristics of the Nannochloropsisgaditana microalgae at different conditions.

Pyrolysis

Initial sample weight (mg) Particle size (µm) Gas flow (ml/min)

4 7 9 15 24 25-50 50-100 100-250 > 250 50 100 150 200

T pyr (ºC)* 156 163 167 172 195 165 165 166 166 169 164 169 166

Tm (ºC)** 1st peak 67 69 68 70 74 65 71 75 74 69 69 67 66 2nd peak 295 305 307 315 317 304 307 308 309 309 310 309 307

(dw/dT)max (wt. %/ ºC) ***

1st peak 0.1 0.07 0.07 0.07 0.06 0.08 0.06 0.05 0.05 0.08 0.07 0.02 0.07 2nd peak 0.49 0.47 0.46 0.44 0.44 0.45 0.46 0.46 0.45 0.46 0.45 0.45 0.46

Residue yield (wt. %) 15.75 19.35 20.24 21.90 22.80 20.69 20.57 21.37 20.90 20.19 20.22 20.29 20.32

* Temperature at which pyrolysis started. ** Temperature at which a peak in the DTG curve was observed. *** Maximum weight loss rate.

Chapter 1

76

In good agreement with literature [3; 7-9], the thermogravimetric (TGA) and

differential thermogravimetric (DTG) curves revealed three degradation steps

common to all studied work conditions.The first stage (40-160ºC) was associated with

a small weight loss due to dehydration (cellular water and external water). The second

stage represented the main devolatilization reactions, where most of the sample weight

was lost as volatile matter (160-450 ºC). Three shoulders canbe distinguished in this

stage, being the low-temperature peak (180 ºC)mainly associated to the degradation of

protein and soluble polysaccharide whereas the higher temperature peaks (271 and

411 ºC) would correspond to the degradation of crude cellulose in the cell wall, other

insoluble polysaccharides and crude lipid[14]. Finally, the laststagetook place at

temperatures above 450 ºC leading to char formation[8].

The effect of the initial mass of the sample on the NG microalgae pyrolysis was

also examined (Figure 2a). Experiments were performed using different initial sample

weights (4-24 mg) with particle sizeof 100-250 µm. Argon flow rate was fixed at200

ml/min (25 ºC, 0.9 atm). In agreement with Antal (1998) and Stenseng et al. (2001),

increasing sample mass shifts the pyrolysis process to higher temperatures turning into

higher residue yields (from 16 to 23 wt.%). The height of the DTG peaks

decreasedwhereas the width increased withincreasing sample weights. However,

TGA/DTG curves for weights of 7 and 9 mg overlapped, indicating negligible

internal-thermal and external-mass transfer limitations. As described by Antal (1998),

the lower peak height would correspond with the increase in the peak width and

ahigher char yield.

Chapter 1

77

Therefore, high mass loadings caused heat-transfer and mass-transfer problems

delaying the pyrolytic process[15]. In addition, the shape of the peak was slightly

distorted at the high mass sample[15]. On the basis of the results described above,

initial mass sample of 9 mg was selected for the following experiments.

Figure 2b shows TGA/DTG plots versus temperature obtained from the pyrolysis

of the NG microalgae at different particle sizes (25-50, 50-100, 100-250 and >250 µm)

with initial mass of 9 mg and Ar flow rate of 200 ml/min (25 ºC, 0.9 atm). In all

cases,the second and third stagesof thepyrolyticTGA/DTG profiles were similar for

the second and third stages. However, the first stage was delayed for sample

sizesbigger than 50µm. According to Mani et al. (2010), this fact could be attributed to

the fact thatsmaller particles have larger surface area leading to less diffusion

resistance for the pyrolysis reaction.On the other hand, the residueproduced (≈21

wt.%)remained constant regardless of the particle sizeused(Table 2). Therefore, a

particle size of 100-250 µmwas selected to avoid the grinding or milling of the sample

due to the NG microalgae sample had a narrow particle size distribution centered on

this particle size range (Figure 1a).

TGA/DTG curves for different sweep gas flows (50, 100, 150 and 200 ml/min) (25

ºC, 0.9 atm) using initial mass of 9 mg and a particle size range of 100-250 µm are

shown in Figure 2c. As it canbe seen, the gas flow did not affect the pyrolysis

outcomes [15]. In all cases, the amount of residue obtainedwas practically constant (20

wt.%). However, higher sweep gas flows were required in order to avoid secondary

reactions due to long residence times inside the TGA[16].

Chapter 1

78

0

20

40

60

80

100

0

20

40

60

80

100

200 300 400 500 600 700 8000

20

40

60

80

100

200 300 400 500 600 700 8000

20

40

60

80

100

4 mg 8 mg 10 mg 17 mg 24 mg

0.0

0.5

1.0

1.5

2.0

2.5

25-50 µm 50-100 µm 100-250 µm > 250 µm

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Temperature (ºC)

Wei

ght (

%)


0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

W

eigh

t los

s ra

te (

% w

t./ºC

)

1.371.321.160.880.65

Wei

ght (

%)

Temperature (ºC)

20 % O2

40 % O2

60 % O2

80 % O2

100 % O2

S* 109

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Wei

ght l

oss

rate

(%

wt./

ºC)

1.3.2. Combustion of the NG microalgae

The effect of the initial sample weight, the particle size and the oxygen gas flow at

a heating rate of 40ºC/min on the NG microalgae combustion is shown in Figure 3.

Figure 3.TGA/DTG profiles for the NG microalgae combustion process as a function of: (a)

initial weight, (b) particle size and (c) gas flow rate at a heating rate of 15 ºC/min.

As reported by other authors [17-19], the combustion of NG microalgae took place

inthree stages. The first stage occurred in the 30-125 ºC range,which corresponded to

the loss of free water and water loosely bound to biomolecules. In this process, the cell

structure was progressively destroyed, and phenomena such as alteration of lipid

structures and protein thermal unfolding occurred. The second one, ranging from 180

to 450 ºC, was characterized by a major weight loss, which involved the

decomposition of proteins and carbohydrates[10; 20]leading to the char formation.

Finally, the last stage(450-600 ºC) corresponded to the oxidation of the formed

Chapter 1

79

remaining char.At the end of this stage, it was observed that between 25 and 8 wt.% of

the char was not completely oxidized, depending on the conditions used.The main

thermogravimetric features in the combustionof NG microalgae aresummarizedin

Table 3.The influence of the initial weightwas investigated under oxygen atmosphere

for different initial masses (4, 8, 10, 17 and 24 mg)of the NG microalgae samplewith

particle sizeranging from100 to 250 µm and an oxygen flow rate of 100 ml/min (25

ºC, 0.9 atm)(Figure 3a). The initial weight had a significant effect on the thermal

degradation behavior. In agreement with some studies reported in the literature [21;

22], the higher the sample weight, the lower both the temperature needed for the

combustion process and the residue yields were (from 25 to 14 wt.%). Likewise, the

DTG peaks heightincreased and the width decreased with increasing sample weights.

As observed for the pyrolysis process, TGA/DTG curves for sample weights of 8 and

10 mg overlapped, indicating negligible internal-thermal and external-mass transfer

limitations[15]. On the basis of the results discussedabove, a sample weight of10 mg

was chosen for the experiments of the next section in order to ensure the detection of

the gas products distribution coming from the combustion of the NG microalgae by

means of a mass spectrometer.

Chapter 1

80

Table 3.Combustion characteristics of the Nannochloropsisgaditana microalgae at different conditions.

Combustion

Initial sample weight (mg) Particle size (µm) Gas flow (ml/min) OxygenConcentration (%)

4 8 10 18 24 25-50

50-100

100-250

> 250 50 100 150 200 20 40 60 80 100

Td (ºC)* 199 202 203 206 207 202 203 206 204 204 199 202 201 206 203 203 201 200

To (ºC)** 481 472 472 470 466 475 477 477 481 474 476 480 482 502 496 491 486 478

Tf(ºC)*** 616 614 610 607 603 618 630 635 646 632 625 630 622 685 645 638 630 623

Tm (ºC)****

1ststage 253 247 247 240 242 289 260 259 263 252 252 250 250 270 263 258 256 253

2ndstage

541 519 520 495 485 503 524 523 537 524 525 525 525 560 553 538 530 525

3rdstage 847 896 900 931 954 891 888 910 914 911 905 912 916 902 905 905 903 903

(dw/dT)max 1ststage 0.34 0.38 0.38 0.49 1.03 0.39 0.37 0.37 0.33 0.41

0.41 0.42 0.42 0.36

0.35 0.35 0.36

0.37

(wt. %/ ºC)*****

2ndstage

0.41 0.65 0.63 2.38 2.36 3.47 0.67 0.76 0.50 0.59

0.66 0.67 0.66 0.33

0.42 0.53 0.58

0.63

3rdstage 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.05 0.06 0.04

0.05 0.05 0.05 0.06

0.06 0.06 0.06

0.05

Residue yield

25.34

18.91

17.14

14.43

13.94

19.50 12.53 12.75 13.64 8.32

10.01

10.36

10.10

8.28

10.18

10.26

9.48

9.59 (wt. %)db

* Temperature at which initial decomposition took place, ** Temperature at which combustion started, *** Temperature at which combustion ended.,**** Temperature at which a peak in the DTG curve was observed, ***** Maximum weight loss rate, dbDry basis

Chapter 1

81

The effect of the particle size on the combustion of the NG microalgae was studied

for four different particle sizes (25-50, 50-100, 100-250 and >250 µm) usingan initial

weightof 10 mg and an oxygen flow rate of 100 ml/min (Figure 3b). According to the

TGA/DTG curves, the most reactive sample was the NG microalgae with the smallest

particle size(25-50 µm). In addition, the residue yield was slightly reduced at higher

particles sizes (from 19 to 13 wt.%).Chouchene et al. (2010) noted that the char

oxidation of the finest particles samplestook place at lower temperaturesthan those

corresponding to higher sizes. These results are in agreement with those reported in

previous works focused on the study of the influence of the particle size on biomass

combustion[18; 23].Finally, TGA/DTG profiles for particle sizesranging from 50 to

100 andfrom 100 to 250 µm overlapped.

Figure 3c shows TGA/DTG curves of the NG microalgae for different pure oxygen

gas flows (50, 100, 150 and 200 ml/min) (25 ºC, 0.9 atm), an initial sample weight of

10 mg and particle sizesranging from 100 to 250 µm. Regardless the oxygen flow

rates used, the combustion TGA/DTG profiles remained practically the same. Thus,

the combustion behavior of the NG microalgae was not significantly affected by the

oxygen flow rate due to the oxygen concentration was kept constant anyway.

The effect of oxygen concentration on the combustion of the NG microalgae was

studied for five different oxygen concentration (20/80, 40/60, 60/40, 80/20 and 100/0

oxygen/Argon ratios) using an initial weight of 10 mg, a total flow rate of 100 ml/min

and a particle size range of 100-250 µm (Figure 3.d). TGA-DTG profiles show that

the first decomposition step (180-400 ºC) was not influenced by the oxygen

Chapter 1

82

concentration as the curves almost overlap. This phenomenon is attributed to the

thermal decomposition of the sample is in the kinetic control zone, being mainly

affected by the temperature, and the effect of oxygen concentration is almost

negligible [17]. However, in the temperature range between 380 and 475 ºC a small

peak appeared in the DTG curves for oxygen concentrations of 20, 40 and 60 % being

unappreciated for higher values of oxygen concentration. The effect of the oxygen

concentration is highly remarked between 480 and 600 ºC, where the oxidation of the

char was taking place. As the oxygen concentration is increased both, the initial

oxidation temperature and the peak temperature in the DTG curve shifted to lower

temperatures, whereas the final temperature of oxidation was decreased. On the other

hand, the maximum weight loss was higher for increasing values of oxygen (Table 3).

Thus, the oxygen concentration enhanced the combustion of the remaining char. These

results agreed well with literature [18; 24], being attributed to the fact that the

combustion reaction of the NG microalgae is in the diffusion control zone and the

oxygen concentration becomes the major influencing factor [24]. In these types of

tests, combustion characteristic index S is usually used to evaluate the combustion

behavior of biomass [24]. S is defined as follows:

� =(��

��)�� ∙ (

��

��)��

�� ∙ ��

(eq. 4)

where (dw/dt)max and (dw/dt)mean are maximum and average mass loss rates,

respectively. Ti and Tb are the ignition and burnout temperatures. It is established, than

the bigger the value S is, the higher the combustion activity [24]. Figure 3.d shows the

Chapter 1

83

value of S for the different oxygen concentrations studied. It can be observed that S

increased as the concentration of oxygen was increased. However, for oxygen

concentrations above 60 %, the increasing trend of S was stabilized. Thus, optimum

values of oxygen concentrations were found to be between 60 and 100 %.

1.3.3. Gasification of the NG microalgae

Biomass gasification depends mainly on the biomass type and the operating

conditions, such as particle size, char porosity, temperature and partial pressure of the

gasifying agents [11; 25; 26]. The gasification process generally includes a

devolatilization step (pyrolysis) and a char gasification step.The char obtained after

the devolatilization step was later gasified.In this study, a devolatilization step was

carried out by heating the samples from 125 to 850 ºC at a heating rate of 40ºC/min

under Aratmosphere.Figure 4 displays the char conversion (X) vs timeon stream

obtainedat different temperatures, initial sampleweights, particle sizes,Argongas flows

and steam concentrations. Furthermore, Table 4 lists the most relevant gasification

characteristics of the NG microalgae.

Chapter 1

84

Figure 4.Char conversion vs time plots for the NG microalgae gasification process as a

function of (a) temperature (b) initial weight, (c) particle size, (d) gas flow rate and (e) water

vapour concentrations at a heating rate of 40 ºC/min.

0 10 20 30 40 50 600

20

40

60

80

100

0 10 20 30 40 50 600

20

40

60

80

100

0 10 20 30 40 50 600

20

40

60

80

100

0 10 20 30 40 50 600

20

40

60

80

100

0 10 20 30 40 50 600

20

40

60

80

100

(d)

Con

vers

ion

(%)

Time (min)


(b)

7 mg 9 mg 15 mg 20 mg

Con

vers

ion

(%)

Time (min)

Con

vers

ion

(%)

Time (min)

550ºC 650ºC 750ºC 850ºC

(a)

(c)

Con

vers

ion

(%)

Time (min)

25-50 µm 50-100 µm 100-250 µm > 250 µm

(e)

Con

vers

ion

(%)

Time (min)

3.7% 5.5% 7.3%

Chapter 1

85

Table 4.Gasification characteristics of the Nannochloropsisgaditana microalgae at different

conditions.

Nannocholoropsisgaditana (NG) Temperature(ºC)

550 650 750 850

Time X50 (min) >60 >60 >60 19.9

Reactivities ((s-1)*104) - - - 9.1

Initial sample weight (mg)

7 9 15 20

Time X50 (min) 7.1 9.8 25.8 30.2

Reactivities ((s-1)*104) 18.3 16.4 9.2 8.5

Particle size (µm)

25-50 50-100 100-250 > 250

Time X50 (min) 32 30.2 24.2 23.6

Reactivities ((s-1)*104) 6.3 7.4 7.6 8.2

Gas flow (ml/min)

50 100 150 200

Time X50 (min) 33.0 30.1 27.9 24.4

Reactivities ((s-1)*104) 5.3 7.5 9.9 48.2

Water vapour concentration (%)

3.7 5.5 7.3

Time X50 (min) 33 30.1 27.9

Reactivities ((s-1)*104) 6.2 7.5 8.3

Chapter 1

86

The effect of the temperature(550, 650, 750 and 850 ºC) on the NG microalgae

chargasification was studied considering an initial weight of 9 mg, a particle size

ranging from 100 to 250 µm,an Argon flow rate of 200 ml/min, a steam concentration

of 5 vol.% in Argon and a heating rate of 40ºC/min. As expected, the higher the

gasification temperature, the higher both, the char conversion and the reactivitywere.

The same behavior was reported elsewhere[11; 25; 26].50% of char conversion at

850ºC was achieved after 20 min, whereas at lower temperatures this level of

conversion was achieved after 60 min. In fact, for the period of 60 min, only 48% of

the char was converted at 750 ºC. According to Mani et al. (2011), the char

gasification at temperatures lower than 1000 ºC, the chemical reaction was the rate-

determining step.This way, 850 ºC was chosenas the gasification temperature for the

following experiments due to the fact that a high amount of char is converted in less

time.

Figure 4b shows the conversionof the NG microalgae char in the gasification

process versus time on stream for different initial weights (7, 10, 15 and 20 mg) at

850ºC for particle sizes ranging from 100 to 250 µm, an Ar flow rate of 200 ml/min

(25 ºC, 0.9 atm), a steam concentration of 5 vol.%in Arand a heating rate of 40ºC/min.

The reactivity increasedwith decreasing initial sample weights.It is clear that the

sample weight had a significant influence on the time to reach a plateau in all

conversion curves. In fact, a constant conversion (50%) for 7 mg was obtained after 18

Chapter 1

87

min, whereas for 20 mg was achieved after 30 min. As expected, the reactivity

decreased with the sample weight. On the basis of the results described above, an

initial weightof the sample of 20 mg was selected. This value allowed to achieve

reproducible experimental TGA data and to improve the sensitivity of the mass

spectrometer.

The effect of the particle size on the gasification of the NG microalgae char was

studied for four different particle sizes (25-50, 50-100, 100-250 and >250 µm)at

850ºC foran initial sample weight of 20 mg, an Ar flow rate of 200 ml/min (25 ºC, 0.9

atm), a steam concentration of 5 vol.% in Argon and a heating rate of 40ºC/min

(Figure 4c).Among the particle size ranges investigated, the higher the particle size,

the shorter the time to reach a constant conversion was.However, according to the

reported literature [26; 27], the reactivity should decrease with increasing particle

sizes as a consequence of an increase in the diffusion resistance in the gasification

process.In order to explain this finding, the surface feature and the porosity of the

resulting char after the devolatilization step for each particle size was evaluated

bySEM analyses (Figure SS2). SEM micrographs clearly showed that the porosity

increasedwith the particle size justifying this unexpected behavior.In order to confirm

SEM results, Nitrogen adsorption-desorption isotherms were carried out to the

samples according to J.A. Díaz procedure [28]. This way, bigger samples of the NG

microalgae were pyrolyzed in a flux bed reactor keeping the same operating

conditions (Ar flow rate of 200 Nml/min, heating rate of 40 ºC/min and a final

temperature of 850 ºC) for obtaining a bigger amount of char. The results obtained

Chapter 1

88

backed up the SEM analyses showing than the char produced from the pyrolysis of the

NG microalgae had a non-very porous structure. Isotherms obtained can be assigned

to a type II according to the IUPAC classification with a small hysteresis loop due to

capillary condensation [28]. Type II corresponds to non-porous materials. The char

left from the minimum particle range sample was the least porous (0.006 g/cm3), and

the sample within the maximum particle range shown the highest porosity (0.01

g/cm3).The increase in the sample porosity affected the profile of total conversion-

time on stream relationship (Figure 4c). The reactivity increased with porosity due to

the reduction in gasifying agent diffusion resistance, which consequently decreased

the time to reach a plateau in conversion[29].According to these results, particle sizes

ranging from100 to 250 µm were selected for the following experiments.

The sweep gas flow determines the gas residence time during the biomass gasification.

Figure 4d shows the conversion versus time on stream curves of the NG microalgae

charat 850ºC for different argon gas flows (50, 100, 150 and 200 ml/min) (25 ºC, 0.9

atm), an initial weight of 20 mg, a particle size ranging from 100 to 250 µm, a steam

concentration of 5 vol. % in Argon and a heating rate of 40ºC/min.As the argon gas

flow increased, the conversion vs time curves shifted to a higher conversion

rate.According to Zhang et al. (2010), low sweep gas flow (long residence time)results

in the formation of carbon deposits, consequently decreasing the gas yield. Thus, a

value of argon flow of 200 ml/min (25 ºC, 0.9 atm)was chosen in order to minimize

secondary reactions such as thermal cracking, repolymerization and

recondensation[30] and increase the char conversion [29; 31].

Chapter 1

89

The effect of the steam concentration on the gasification of the NG microalgae char

was studied (3.7, 5.5 and 7.3 % in Argon) at 850ºC, for an initial weight of 20 mg, a

particle size ranging from 100 to 250 µm, an argon flow rate of 200 ml/min (25 ºC, 0.9

atm)and a heating rate of 40ºC/min (Figure 4e). As expected, the reactivity increased

with increasing steam concentrations. According to Florin and Harris (2008), water

vapour has been identified as a catalyst for the char formation mechanism, enhancing

the concentration of H2during the char gasification due to the occurrence of the water-

gas shift reaction. This is in agreement with previous reported studies [26]. However,

a steam concentration of 5.5 % was selected for the following experiments since at

higher concentrationsoverpressure problems in the bubbler system were observed.

1.3.4. Gas product analysis

The main productsderived from the pyrolysis, oxidation and gasification of theNG

microalgae were evaluatedby TGA-MS analysis. On the basis of a preliminary scan, a

list of key molecular ions was compiled by rejecting signals when the maximum

intensity was close to the noise level[32].Thedatabase of National Institute of

Standards and Technology (NIST) were usedfor the atomic mass units (a.m.u.)

selection.Furthermore, elemental analyses by Energy Dispersive X-ray Spectroscopy

(EDS) were performed on the NG microalgae and the resulting solid residue obtained

after the pyrolysis and oxidation process (Table 5).Characteristic peaks of C, N, O,

Na, Mg, P, S, Cl, K and Ca were presented in the analysis of the NG microalgae.

However, peaks correspondingto the N element were not detected in the EDS analysis

of the solid residues obtained after the pyrolysis and combustion of the microalgae.

Chapter 1

90

Table 5.EDS analysis of both, the NannochloropsisGaditana microalgae and the solid residue

obtained after pyrolysis and combustion.

Samples Elements (wt. %)

C* N O Na Mg P S Cl K Ca

NannochloropsisGadita

na

47.3 ± 0.5

10.6 ± 0.5

16.7 ± 0.5

8.7 ± 0.5

0.6 ±

0.5

2.6 ± 0.5

1 ± 0.5

8.8 ± 0.5

1.7 ±

0.5

1.9 ±

0.5

Pyrolysis residue

37.7 ± 0.5

0 ± 0.5 17.0 ± 0.5

13.2 ± 0.5

1.5 ±

0.5

8.5 ± 0.5

0 ± 0.5

10.7 ± 0.5

5 ± 0.5

6.3 ±

0.5

Combustion residue

0 ± 0.5 0 ± 0.5 48.4 ± 0.5

24.6 ± 0.5

3.1 ±

0.5

22.3 ± 0.5

1.6 ±

0.5

0 ± 0.5

0 ± 0.5

7.9 ±

0.5

Table 6 shows the molecular ions/ion fragments that were detected during the

pyrolysis, oxidation and gasification of the NG microalgae using the selected

operating condition fixed in former sections.

Chapter 1

91

Table 6. Molecular ions and probable parent molecules detected in the pyrolysis, combustion

and gasification processes for the NannochloropsisGaditanamicroalgae.

m/z

NannochloropsisGaditanamicroalgae

Key molecular ions/Ion fragment

Probable parent molecule

Pyrolysis Combustion Gasification

2 H2+ H2 X X X

15 CH3+ CH4 X X -

16 O+; CH4+ CH4 X - X

17 NH3+ NH3 - X -

18 H2O+ H2O X X -

26 CN+; C2H2+ C2H2(acetylene) X - X

27 HCN+; C2H3+ HCN (nitriles) X - X

28 C2H4+; CO+ CO X X X

29 C2H5+

C2H5 (ethyl derivates)

X - X

30 C2H6+;

CH2NH2+

CH4N (primary amines)

X - X

44 CO2+ CO2 X X X

46 NO2+;

C2H4O+

NO2 - X -

56 C3H6N+;

C4H8+

C4H8 (alquenes) - X -

58 C3H8N+ C3H8N (amines) X X -

64 SO2+ SO2 (sulfones) - X -

78 C6H6+ C6H6 (benzene) X - -

Mass spectra of the pyrolysis and oxidation processes for NG microalgae are

shown in Figure 5.MS curves could be divided into three stages that could be related

to the three degradation steps described in the TGA/DTG curves.

Chapter 1

92

200 300 400 500 600 700 800

H2OCO

H2

CH+

3

CH4N

C2H

2

CH4

CO2

C6H

6

C3H

8N

C2H

5

Inte

nsity

(a.

u.)

Temperature (ºC)

HCN

200 300 400 500 600 700

CO2

H2

NO2

SO2

CH+

3

H2O

C4H

8

NH3

C3H

8N

Inte

nsity

(a.

u.)

Temperature (ºC)

CO

(a) (b)

Figure 5.Mass spectra of the (a) pyrolysis and (b) combustion of the

NannochloropsisGaditana microalgae.

Figure 5a shows the MS curves of the gaseous products released during the

pyrolysis process. H2O, CO and CO2 were detected in the MS spectrum at <160 ºC

due to the moisture content in the NG microalgae. In this case, most of the gas

products (H2O,CO, C6H6, C3H8N, CO2, CH4N, HCN)and volatile hydrocarbons such

as CH3+, CH4, C2H2, C2H5were generated at the second degradation step (160-450 ºC).

As reported elsewhere [2; 8; 33], the pyrolysis of the carbohydrates and proteins of

algaewould mainly take place in this stage.The N-containing compounds in the NG

microalgae could be released in form of amines (C3H8N andCH4N) andnitrile

(HCN)due to the thermal degradation of proteins[33]. According to other authors [8],

CO2 was mainly produced by the cracking and reforming of carboxyl groups in protein

and saccharides.HCN, C2H5, C2H2, CH4Nand CH3+were also producedin the last

decomposition step (>450ºC), corresponding to the slow decomposition of the solid

residue [8].A slow H2 release was observed at around 450-650 ºC, which could be

caused by the further dehydrogenation of remaining carbonaceous species[2].Among

Chapter 1

93

various gaseous hydrocarbons released, the content of CH3+ was the highestone [4].

Similar decomposition profiles have been reported for different algae species by many

authors [8; 14]. These results agree with the low nitrogen and carbon contentsin the

pyrolysis solid residuemeasured by EDSanalysis (Table 5).

The evolution profiles of the gaseous species, CO2, NO2, CO, SO2, C3H8N, H2O,

NH3, CH3+, H2and C4H8, from the combustion of the NG microalgae are shown in

Figure 5b. Two peaks at 265 and 515 ºCfor the H2O intensity associated with moisture

and combustion of volatiles and char were obtained [10].The release of CO2and CO

were observed at around 250-350 and 475-600 ºC due to the combustion of fixed

carbon [10].Furthermore, the SO2release took place at225-300 and 350-425 ºC, which

could be relatedto the decomposition of sulphated polysaccharideexisting in the NG

microalgae [10; 14].The N-containing compounds evolution (NO2, C3H8N and NH3)at

225-350 ºCwas associated with the degradation of protein in the NG

microalgae.Finally, H2, NH3,NO2and CH3+were detected in the last step (>500 ºC)

corresponding to decomposition of the solid residue.

By comparing pyrolysis and combustion MS curves, it is possible to evaluate the

effect of the oxygen presence on the gas emission.As expected, the content of CO2in

the combustion processwas the highestone whereas no generation of volatile

hydrocarbons such as CH4, C2H2, and C2H5was observed. In addition, NO2 and SO2

were also released.

Finally, as it can be seen in Table 5, some of the inorganic materials that were

present in the EDS analysis of the NG microalgae were not found in the combustion or

Chapter 1

94

pyrolysis results, as chloride and potassium for combustion and sulfur in the pyrolysis

residue, pointing out that the evolution of these compound took place during the

combustion and pyrolysis of the sample. This fact is mainly due to the signals of these

ion fragments were rejected as they were too close to the noise level. Further studies

are required for the evaluation of the release of inorganics during combustion and

pyrolysis as they are a potential source of contaminants.

TGA/DTG-MS curves of the gasification process of the NG microalgaeare shown

in Figure 6a.As can be seen in TGA/DTG curves, the time to reach the total

conversion was 55 min. The main products detected during the gasification at 850ºC

were CO2, CO, H2, indicating that oxidation reactions, water gas and water gas shift

reactions were the predominant ones.In addition, traces of CH4, C2H6, C2H5, C2H4and

C2H2were also generated. Similar evolution profiles have been reported elsewhere[31;

34; 35].

Figure 6b shows the yield of the main gases during the gasification process (CO2,

CO, CH4 and H2) at different steam concentrations. It can be observed that as the

proportion of water in the reactive gas was increased, the product distribution varied,

enhancing the production of H2 and decreasing the CH4yield. This fact would indicate

that water gas (C + H2O � CO + H2), water gas shift (CO + H2O � CO2 + H2)and

methane reforming (CH4+ H2O � CO + 3H2)reactions werebeing promoted[32;

35].Anyway, CO and CO2 emissions were kept constant.

Chapter 1

95

0 15 30 45 60

0

20

40

60

80

100

In

tens

ity (

a.u.

)

Time (min)

CH4

C2H

2

C2H

5

C2H

4

C2H

6

CO2

H2

CO

Wei

ght (

%)

TGA

0.0

0.5

1.0

1.5

2.0

2.5

DTG

Wei

ght l

oss

rate

(%

wt./

ºC)

(a)

H2 CH4 CO CO2

0

10

20

30

40

50

Pro

duct

Yie

ld (

%)

Main Products

3.7 % 5.5 % 7.3 %

H2 CH

4 CO CO

2

(b)

Figure 6.(a) TGA-DTG-MS curves for the gasification process of the

NannochloropsisGaditana microalgae and (b)Product yield for the gasification process of the

NannochloropsisGaditana microalgae at different steam concentrations.

Chapter 1

96

1.4. CONCLUSIONS

Pyrolysis, combustion and gasification of NG microalgae were analyzed by means

of TGA-MS. Pyrolysis and combustion processes were divided into three stages.

During pyrolysis, the main devolatilization step took place between 160 and 450 ºC,

associated to the degradation of protein and soluble polysaccharide. In combustion the

oxidation of the sample took place between 450 and 600 ºC. As oxygen concentration

increased, the oxidation of the char shifted to lower temperatures.

N-compounds evolution was associated with the microalgae proteins degradation.

SO2release during combustioncould be related to sulphated polysaccharides

decomposition.H2production was enhanced by steam concentration.

1.5. REFERENCES

[1] Ross, A.B., Anastasakis, K., Kubacki, M., Jones, J.M. 2009. Investigation of the

pyrolysis behaviour of brown algae before and after pre-treatment using PY-GC/MS

and TGA. Journal of Analytical and Applied Pyrolysis, 85(1-2), 3-10.



and Bioenergy, 35(7), 3199-207.

[3] Li, D., Chen, L., Zhang, X., Ye, N., Xing, F. 2011. Pyrolytic characteristics and

kinetic studies of three kinds of red algae. Biomass and Bioenergy, 35(5), 1765-72.

[4] Pan, P., Hu, C., Yang, W., Li, Y., Dong, L., Zhu, L., Tong, D., Qing, R., Fan, Y.

2010. The direct pyrolysis and catalytic pyrolysis of Nannochloropsis sp. residue for

renewable bio-oils. Bioresource Technology, 101(12), 4593-9.

[5] Tang, Y., Ma, X., Lai, Z. 2011. Thermogravimetric analysis of the combustion of

microalgae and microalgae blended with waste in N2/O2 and CO2/O2 atmospheres.


Chapter 1

97

[6] Pane, L., Franceschi, E., De Nuccio, L., Carli, A. 2001. Applications of thermal

analysis on the marine phytoplankton, Tetraselmis suecica. Journal of Thermal

Analysis and Calorimetry, 66(1), 145-54.

[7] Anastasakis, K., Ross, A.B. 2011. Hydrothermal liquefaction of the brown macro-

alga Laminaria Saccharina: Effect of reaction conditions on product distribution and

composition. Bioresource Technology, 102(7), 4876-83.

[8] Marcilla, A., Gómez-Siurana, A., Gomis, C., Chápuli, E., Catalá, M.C., Valdés,

F.J. 2009. Characterization of microalgal species through TGA/FTIR analysis:

Application to nannochloropsis sp. Thermochimica Acta, 484(1–2), 41-7.

[9] Sanchez-Silva, L., López-González, D., Villaseñor, J., Sánchez, P., Valverde, J.L.

2012. Thermogravimetric-mass spectrometric analysis of lignocellulosic and marine

biomass pyrolysis. Bioresource Technology, 109, 163-72.

[10] Wang, S., Jiang, X.M., Han, X.X., Liu, J.G. 2009. Combustion Characteristics of

Seaweed Biomass. 1. Combustion Characteristics of Enteromorpha clathrata and

Sargassum natans. Energy & Fuels, 23(10), 5173-8.

[11] Ollero, P., Serrera, A., Arjona, R., Alcantarilla, S. 2002. Diffusional effects in

TGA gasification experiments for kinetic determination. Fuel, 81(15), 1989-2000.

[12] Arenillas, A., Rubiera, F., Pis, J.J. 1999. Simultaneous thermogravimetric-mass

spectrometric study on the pyrolysis behaviour of different rank coals. Journal of


[13] Widyawati, M., Church, T.L., Florin, N.H., Harris, A.T. 2011. Hydrogen

synthesis from biomass pyrolysis with in situ carbon dioxide capture using calcium

oxide. International Journal of Hydrogen Energy, 36(8), 4800-13.

[14] Wang, S., Jiang, X.M., Wang, N., Yu, L.J., Li, Z., He, P.M. 2007. Research on

pyrolysis characteristics of seaweed. Energy and Fuels, 21(6), 3723-9.

[15] Lin, Y.C., Cho, J., Tompsett, G.A., Westmoreland, P.R., Huber, G.W. 2009.

Kinetics and mechanism of cellulose pyrolysis. Journal of Physical Chemistry C,

113(46), 20097-107.

Chapter 1

98

[16] Shuping, Z., Yulong, W., Mingde, Y., Chun, L., Junmao, T. 2010. Pyrolysis

characteristics and kinetics of the marine microalgae Dunaliella tertiolecta using

thermogravimetric analyzer. Bioresource Technology, 101(1), 359-65.



3189-96.

[18] Chouchene, A., Jeguirim, M., Khiari, B., Zagrouba, F., Trouvé, G. 2010.

Thermal degradation of olive solid waste: Influence of particle size and oxygen

concentration. Resources, Conservation and Recycling, 54(5), 271-7.

[19] Ross, A.B., Jones, J.M., Kubacki, M.L., Bridgeman, T. 2008. Classification of

macroalgae as fuel and its thermochemical behaviour. Bioresource Technology,

99(14), 6494-504.

[20] Yu, L.J., Wang, S., Jiang, X.M., Wang, N., Zhang, C.Q. 2008. Thermal analysis

studies on combustion characteristics of seaweed. Journal of Thermal Analysis and

Calorimetry, 93(2), 611-7.

[21] Stenseng, M., Jensen, A., Dam-Johansen, K. 2001. Investigation of biomass

pyrolysis by thermogravimetric analysis and differential scanning calorimetry. Journal

of Analytical and Applied Pyrolysis, 58-59, 765-80.

[22] Antal Jr, M.J., Várhegyi, G., Jakab, E. 1998. Cellulose Pyrolysis Kinetics:

Revisited. Industrial and Engineering Chemistry Research, 37(4), 1267-75.

[23] Bridgeman, T.G., Darvell, L.I., Jones, J.M., Williams, P.T., Fahmi, R.,

Bridgwater, A.V., Barraclough, T., Shield, I., Yates, N., Thain, S.C., Donnison, I.S.

2007. Influence of particle size on the analytical and chemical properties of two

energy crops. Fuel, 86(1-2), 60-72.

[24] Luo, S.Y., Xiao, B., Hu, Z.Q., Liu, S.M., Guan, Y.W. 2009. Experimental study

on oxygen-enriched combustion of biomass micro fuel. Energy, 34(11), 1880-4.

[25] Mani, T., Mahinpey, N., Murugan, P. 2011. Reaction kinetics and mass transfer

studies of biomass char gasification with CO2. Chemical Engineering Science, 66(1),

36-41.

Chapter 1

99

[26] Gómez-Barea, A., Ollero, P., Fernández-Baco, C. 2006. Diffusional effects in


investigation. Energy and Fuels, 20(5), 2202-10.

[27] Mani, T., Murugan, P., Abedi, J., Mahinpey, N. 2010. Pyrolysis of wheat straw

in a thermogravimetric analyzer: Effect of particle size and heating rate on

devolatilization and estimation of global kinetics. Chemical Engineering Research and

Design, 88(8), 952-8.

[28] Díaz, J.A., Díaz-Moreno, R., Silva, L.S., Dorado, F., Romero, A., Valverde, J.L.

2012. Nickel supported carbon nanofibers as an active and selective catalyst for the

gas-phase hydrogenation of 2-tert-butylphenol. Journal of Colloid and Interface

Science, 380(1), 173-81.

[29] Kirubakaran, V., Sivaramakrishnan, V., Nalini, R., Sekar, T., Premalatha, M.,

Subramanian, P. 2009. A review on gasification of biomass. Renewable and

Sustainable Energy Reviews, 13(1), 179-86.

[30] Pütün, A.E., Özean, A., Pütün, E. 1999. Pyrolysis of hazelnut shells in a fixed-

bed tubular reactor: Yields and structural analysis of bio-oil. Journal of Analytical and

Applied Pyrolysis, 52(1), 33-49.

[31] Zhang, M., Chen, H.P., Gao, Y., He, R.X., Yang, H.P., Wang, X.H., Zhang, S.H.

2010. Experimental study on bio-oil pyrolysis/gasification. BioResources, 5(1), 135-

46.

[32] Florin, N.H., Harris, A.T. 2008. Enhanced hydrogen production from biomass

with in situ carbon dioxide capture using calcium oxide sorbents. Chemical

Engineering Science, 63(2), 287-316.

[33] Maddi, B., Viamajala, S., Varanasi, S. 2011. Comparative study of pyrolysis of

algal biomass from natural lake blooms with lignocellulosic biomass. Bioresource

Technology, 102(23), 11018-26.

[34] De Lasa, H., Salaices, E., Mazumder, J., Lucky, R. 2011. Catalytic steam

gasification of biomass: Catalysts, thermodynamics and kinetics. Chemical Reviews,

111(9), 5404-33.

Chapter 1

100

[35] Kojima, T., Assavadakorn, P., Furusawa, T. 1993. Measurement and evaluation

of gasification kinetics of sawdust char with steam in an experimental fluidized bed.

Fuel Processing Technology, 36(1-3), 201-7.

(1)

Chapter 2:

THERMOGRAVIMETRIC -MASS SPECTROMETRIC ANALYSIS OF

LIGNOCELLULOSIC AND MARINE BIOMASS

PYROLYSIS

The pyrolysis characteristics of three lignocellulosic biomasses

(Fir Wood, Eucalyptus and Pine Bark) and a marine biomass

(NannochloropsisGaditana microalgae) were investigated by

thermogravimetric analysis coupled with mass spectrometry (TGA-

MS). Thermal degradation of lignocellulosic biomass was divided

into four zones, corresponding to the decomposition of their main

components (cellulose, hemicellulose and lignin) and a first step

associated to water removal. Differences in volatile matter and

cellulose content of lignocellulosic species resulted in different

degradation rates. Microalgae pyrolysis occurred in three stages

due to the main components of them (proteins), which are greatly

different from lignocellulosic biomass. Heating rate effect was also

studied. The main gaseous products formed were CO2, light

Chapter 2

102

hydrocarbons and H2O. H2 was detected at high temperatures,

being associated to secondary reactions (char self-gasification).

Pyrolysis kinetics were studied using a multiple-step model. The

proposed model successfully predicted the pyrolitic behaviour of

these samples resulting to be statistically meaningful.

2.1. INTRODUCTION

Depletion of world fossil fuel reserves and the external dependence that these types

of fuels produce together with the environmental risks derived from its use are the

main reasons of the increasing attention that renewable energies sources are receiving.

In this context, biomass conversion for transportation fuels, chemical commodities and

power generation is getting growing interest.

Biomass is a term for all organic material that stems from plants including algae,

trees and crops that are susceptible to be converted into energy (McKendry, 2002).

One of the most controversial points in the use of biomass as an energy source is its

possible competition with human food supply. Nevertheless, there are different types

of biomass that fit into this definition without compromising world food supply. The

main ones refer to lignocellulosic biomass (Basu, 2010) and marine biomass

(especially algae). Algae are a good candidate because they present the following

advantages: large amount available, fast growth, low priced, environment protection

and suitable for pyrolysis (Wang, 2006).

There are many conversion technologies for utilizing biomass, such as direct

combustion, thermochemical, biochemical and agrochemical processes.

Chapter 2

103

Thermochemical conversion of biomass is considered as one of the most promising

processes for biomass utilization (Shen et al., 2010). There are four thermochemical

technologies: pyrolysis, gasification, combustion and liquefaction. Pyrolysis is of

special interest since it is a prior step in combustion and gasification processes.

Therefore, it seems essential to obtain a deep knowledge of biomass pyrolysis in order

to gain further understanding of the combustion and gasification processes.

Pyrolysis can be described as the biomass conversion by heat in the absence of

oxygen in a relatively low range of temperatures (300-600 ºC), which results in the

production of charcoal (solid), bio-oil (liquid) and fuel gas products. Thermal analysis

has shown to be a powerful tool for investigating the pyrolysis of biomass. Numerous

studies based on thermogravimetic analysis (TGA) and derivative thermogravimetry

(DTG) have been carried out. Many of them focused on the main components of

lignocellulosic biomass, mainly constituted by cellulose, hemicellulose and lignin

(Wang et al., 2008; Yang et al., 2006), different types of lignocellulosic biomass

(Barneto et al., 2011; Stenseng et al., 2001), and algae (Li et al., 2011; Peng et al.,

2001). The effects of heating rate and amount of sample have been also reported in the

literature (Lin et al. 2009).

Pyrolysis kinetics is other of the aspects that has been widely studied by thermal

analysis. Most studies have been focused on cellulose pyrolysis (Grønli et al., 1999,

Lin et al., 2009), lignin and xylan (Rao and Sharma, 1997), lignocellulosic biomass

(Órfão et al., 1999) and marine biomass (Wang et al., 2006). The determination of the

kinetics corresponding to biomass thermal decomposition involves the knowledge of

the reaction mechanisms. However, pyrolysis is an extremely complex process, where

Chapter 2

104

numerous reactions take place, practically making impossible to develop a kinetic

model that takes into account all these reactions. Thus, the pyrolysis is usually studied

in terms of pseudo-mechanistic models (Caballero et al., 1997). White et al. (2011)

reported that kinetics of biomass decomposition can be divided into three principal

types of models: single-step global reaction models, multiple-step models and semi-

global models.

Thermal analysis itself might not seem sufficient for a thorough study based on

kinetics. Therefore, other techniques must be used to obtain valuable results (White et

al., 2011). The combination of thermogravimetric analysis coupled with mass

spectrometry (TGA-MS) appears to give a deeper insight of the process. Some studies

concerning TGA-MS of the biomass pyrolysis have been carried out (Grønli et al.,

1999; Widyawati et al., 2011). One of the most attractive advantages of TGA-MS is

that it is able to afford real-time and sensitive detection of evolved gases, which is an

important and often a difficult task in many thermal applications (Huang et al., 2011).

The aim of this study was to investigate the pyrolysis characteristics and gas

products distribution of lignocellulosic (Fir Wood, Eucalyptus Wood and Pine Bark)

and marine (Nannochloropsisgaditana microalgae) biomass by means of the TGA-MS

technique. This work pretends to establish and gain further understanding of the

possible relationships among these components, from those present in lignocellulosic

materials to those that constitute terrestrial and marine biomass. Moreover, the effect

of heating rate on the pyrolysis behaviour of these samples was also studied. Finally,

experimental data obtained using thermogravimetric analysis were interpreted using a

multi-step kinetic model.

Chapter 2

105

2.2. EXPERIMENTAL

2.1.1 Materials

Cellulose, Xylan and Lignin were purchased from Sigma Aldrich. Xylan was used

as a representative of the hemicellulose component in the pyrolysis (Wang et al.,

2008; Yang et al., 2006). These chemicals are as follow: Cellulose (microcrystalline

cellulose with 50 µm average particle size), Lignin (alkali lignin in brown powder

form with 50 µm average particle size) and Xylan (xylan processed from beechwood

with 100 µm average particle size, was used as hemicellulose). The selected terrestrial

biomass (FirWood, EucalyptusWood and PineBark) were taken from the region of

Castilla-La Mancha (Spain). These samples were dried in an oven for 5 hours, milled

and sieved to less than 240 µm. The microalgae NannochloropsisGaditana(NG

microalgae) were purchased from AlgaeEnergy Company. This compound is delivered

in green powder with 100 µm average particle size.

2.2.2. Biomass selection

The choice of biomass mainly depended on its inherent properties, determining the

conversion process and any subsequent processing difficulties that may arise. The

main properties of interest for the biomass processing as an energy source are the

following (McKendry, 2002): moisture content (MC); proportion of fixed carbon (FC)

and volatiles (VM); ash/residue content (AC/AR); calorific value; alkali metal content;

cellulose/lignin ratio.

Chapter 2

106

The first three characteristics are determined by means of the named proximate

analysis. This analysis gives an idea of how good is the biomass to be converted into

energy. Proportion of fixed carbon (FC) and volatile matter (VM) are two ways to

represent the chemical energy stored in the biomass. The higher the VM/FC ratio is,

the larger the available energy that biomass is able to be released. On the other hand,

the moisture (MC) and ash content (AC) are two parameters that have adverse effects

on the quality of the fuel. High values of HM decreases the calorific value of the fuel

driving to an uneven overall energy balance whereas high values of AC leads to an

increase of operational costs. In order to select the most proper kind of terrestrial

biomass, a preliminary study of biomass species using a proximate analysis was done

(Table 1).

Table 1.Proximate analysis of lignocellulosic and marine biomass.

Biomass

Volatile Matter

(wt. %)

Fixed Carbon

(wt. %)

Ash

(wt. %)

Cellulose 92.8 6.1 1.1

Lignin 55.4 41.5 3.1

Xylan 76.1 21.6 2.3

Fir wood 78.1 17.1 4.8

Eucalyptus wood 76.4 16.3 7.3

Pine bark 69.4 27.5 3.1

Nannochloropsis

Gaditana

microalgae

83.1

10.1

6.8

Chapter 2

107

Taking into account the analysis listed in this Table and those reported by Yaman

(2004) for different biomass, a ternary diagram was plotting by considering the

following parameters: ash, volatile matter and fixed carbon contents (Figure 1).

Biomass to be used in this study was selected according to the following criteria:

- Biomass with high VM content and low AC.

- Biomass with high FC content and a low AC.

According to these criteria, two areas in the diagram were clearly identified

(they have been outlined with a circle). Biomass within these two zones corresponded

to: Fir Wood and Eucalyptus Wood (both with pretty higher VM) and Pine Bark (with

the highest FC).

Figure 1.Ternary diagram with different kinds of terrestrial biomass according to their

proximate analysis (biomass data was obtained from Yaman et al., 2004).

0.0 0.2 0.4 0.6 0.8 1.0

0.00.00.00.0

0.20.20.20.2

0.40.40.40.4

0.60.60.60.6

0.80.80.80.8

1.01.01.01.00.0

0.2

0.4

0.6

0.8

1.0

Fixed Carbon

Sugarcane bagasseGrapeMaizeOliveRapeseedRice huskSawdustSunflowerBrown Kelp GiantWater hyacinthFir woodTobaccoPine barkCotton wastesEucalyptus woodStraw

Ash

Volatile Matter

Chapter 2

108

2.2.3. Equipment and Procedures

2.2.3.1. Pyrolysis

The pyrolysis of biomass components was firstly carried out in a TGA apparatus

(TGA-DSC 1, METTLER TOLEDO). The sample was heated from 40 ºC to 900ºC at

different heating rates (5, 15 and 40 ºC/min). From previous studies in TGA analysis

the following operating conditions were chosen in order to avoid the effects of heat

and mass transfer limitations: sample weight was kept at 10 mg, helium (99.99 %) at a

flow rate of 200 ml/min was used as the carrier gas to provide an inert atmosphere and

the particle size was kept lower than 300 µm. Each sample was analyzed at least three

times, and the average value was recorded. The experimental error of these

measurements was calculated, obtaining an error for all studied samples of ± 0.5% in

weight loss measurement and ± 2 ºC in temperature measurement.

2.2.3.2. TGA-MS

The analysis of the gas products distribution coming from the pyrolysis was carried

out in a thermogravimetricanalyzer (TGA-DSC 1; METTLER TOLEDO) coupled to a

mass spectrometer (Thermostar-GSD 320/quadrupole mass analyzer; PFEIFFER

VACUUM) with an electron ionization voltage at 70 eV and provided mass spectra up

to 300 a.m.u. The interface was wrapped with heating wire to circumvent

condensation of exhausting gases. Approximately 10 mg of sample was loaded into an

alumina crucible pan and heated from room temperature to 900 ºC at a heating rate of

40 ºC/min. In all experiments, He was used as the purge gas (99.99 %) with a

constant flow rate of 200 ml/min. In order to identify ions with m/z in the range 0-300,

Chapter 2

109

a preliminary broad scan was performed at a heating rate of 40 ºC/min. The signals

identified corresponded to the mass spectra of 2, 15, 18, 27, 28, 30, 32, 44 a.m.u,

which corresponds to the main components of the pyrolysis gas (H2, CH4, H2O, C2H4,

CO, C2H6 and CO2, respectively).

2.2.4. Kinetics

Kinetic data from solid state pyrolysis was obtained using thermogravimetric

analysis. The model proposed in this work is similar as that reported by Sun et al.

(2011) for the determination of the kinetic parameters of decomposition of bagasse

fibers, and is based on works previously reported (Órfão et al., 1999). Considering

npyrolizable compounds, the kinetic rates of thermal decomposition of a material

assuming independent parallel nth-order reactions and an Arrhenius dependence of the

rate constants are:

( )inn

ii

s

iaioi RT

Ekc

dt

d∑=

−

−=

11exp αα

(1)

( ) ini

s

iaio

iRT

Ek

dt

d αα −

−= 1exp

(2)

where α is the degree of conversion of the material, kioand Eiaare the pre-exponential

factor and the activation energy for the individual components; R is the gas constant;

ni is the reaction order; and αi is the degree of conversion for the individual

component defined by

Chapter 2

110

io

ipioi m

mm −=α

(3)

wheremio and mip represent the mass at t=0 and t=t for each component, respectively.

The constant ci is related to the initial composition of the different components.

Finally, Ts is the actual sample temperature that may differ from the external

temperature by a thermal lag. Lin et al. (2009) proposed the following equation as a

function of a fitting experimental factor C to correct the thermal delay of the

apparatus:

( )CtbTT os −+=

(4)

In this equation, To and b represent the initial temperature and the heating rate,

respectively.

2.2.4.1. Parameter estimation

A VBA-Excel application was developed to solve this model (de la Osa et al.,

2011; Valverde et al., 2004). The Bader-Deufhard method was used in the evaluation

of the set of ordinary differential equations (Press, 1992), whereas the Marquardt-

Levenberg algorithm was used in the nonlinear regression procedure (de Lucas et al.,

2006; Froment et al., 1990; Marquardt, 1963). The ordinary differential equation, Eq.

(1) was solved by considering the following initial conditions:

α(t=0) = 0 and αi(t=0) = 0 (5)

The weighted sum of the squared differences between the observed (Exp) and the

calculated (Pr) degree of conversion was minimized according to the following

equation described elsewhere (de Lucas et al., 2006):

Chapter 2

111

( )∑ ∑= =

−=n

i

m

jjiExpiSSQ

1 1

2Pr αα

(6)

wherei represents the number of equations to be fitted, j the specific experimental data

and n and m the total number of equations and experiments (more than 4500 if all the

data generated in a TGA analysis are considered), respectively.

AF-test is a statistical test in which the test statistic has a F-distribution under the

null hypothesis. The procedure was based on the comparison between the tabulated F-

value (F-test) and Fc, which was defined elsewhere (Froment, et al., 1990):

( )

( ) ( )∑ ∑

∑ ∑

= =

= =

−∗−=

n

i

m

jjiExpi

n

i

m

ji

c

pmn

k

F

1 1

2Pr

1 1

2Pr

/

/

αα

α

(7)

wherep represents the total number of parameters.

If Fc is larger than F(p, n-p, 1-α), assuming a value of α = 0.05, 95% confidence

level, the regression was considered to be meaningful, although there is no guarantee

that the model is statistically suitable since the meaningfulness of each parameter in

the model must be also evaluated.

Hence, a complementary test, named t-test, was used. The t-test considers that the

statistical hypothesis test follows a Student’s t distribution and allows to verify if the

estimate of the parameter bfi differs from a reference value (generally zero). Thus, a

parameter is meaningful (at α = 0.05) each time that the following inequality occurs:

Chapter 2

112

( )

−−>=2

1,α

pntbV

bt

iif

fici

(8)

where[V(bf)]ii represents the diagonal ith term of the covariance matrix used in the last

step of n-linear regression procedure.


2.3.1. Thermogravimetric study of pyrolysis of lignocellulosic and marine

biomass

The TGA/DTG profiles of the main components of the biomass here considered

(hemicellulose (Xylan), lignin and cellulose) as a function of temperature at a heating

rate of 15ºC/min during the pyrolysis process are shown in Figure 2a. There are

substantial differences between the pyrolysis behaviour of these components. The

decomposition of Xylan showed two peaks, starting at about 200 ºC and reaching its

maximum weight loss rate at 250ºC, being the residue yield equal to 28 wt.%. These

two peaks can be associated to the decomposition of the xylan side units and the

cracking of the main xylan chains (Severian, 2008). The pyrolysis of Cellulose

occurred between 290 ºC and 390 ºC reaching its maximum value at 340 ºC. It can be

noticed the pronounced DTG profile of this sample. The residue yield of cellulose was

the lowest one (9 wt.%). Lignin showed the highest thermal stability, decomposing in

the whole range of temperatures studied (200-700 ºC). Furthermore, the DTG profile

of Lignin was the flattest being the residue yield obtained the highest one (>40 wt.%).

This fact could be due to the slow carbonization of lignin, being the main responsible

Chapter 2

113

for the biomass char formation (Yang et al., 2006). According to Wang et al. (2008),

the differences in the thermal behaviour of these components can be attributed to their

different chemical structures. Hemicellulose has a random and amorphous structure

with little strength whereas cellulose has a crystalline and strong structure and it is

resistant to hydrolysis. On the other hand, Lignin is the most different one due to the

fact that it is a complex, heavily cross-linked and highly branched polymer (Basu,

2010).

Figure 2b shows the TGA/DTG plots versus temperature for Fir Wood,

Eucalyptus Wood and Pine Bark at a heating rate of 15 ºC/min. The criterion for

biomass selection was explained in Section 2.2. Generally, the thermal degradation

profiles of lignocellulosic biomass are interpreted as the addition of the independent

degradations of their main components (Caballero et al., 1997). According to this fact,

the pyrolysis process can be divided into four stages: moisture evolution (<120ºC);

hemicellulose decomposition (150-310 ºC); lignin and cellulose degradation (310-400

ºC) and lignin decomposition (> 450ºC).

In spite of these four well identified stages, there are some differences in the

behaviour of these materials. The second zone, corresponding to the hemicellulose

decomposition, took place at different temperatures as a function of the raw materials.

This area is represented by a shoulder in the DTG curve (Grønli et al., 2002). In the

case of the Fir Wood sample, hemicellulose decomposition took place at lower

temperatures than for Eucalyptus Wood and Pine Bark. Nevertheless, regardless the

sample, the third zone occurred in a similar range of temperatures (340-350 ºC). This

Chapter 2

114

temperature range agrees with that of the volatilization of cellulose shown in Figure

2a. In the last zone ascribed to lignin decomposition (zone IV), Pine Bark and Fir

Wood were almost overlapped whereas that of Eucalyptus occurred with different

weight loss rates, showing two peaks in the DTG curve at 505 ºC and 650ºC,

respectively. Finally, the residue yield was 34, 29 and 26 wt.% for Pine Bark,

Eucalyptus Wood and Fir Wood, respectively.

The differences in the thermal behaviour of all lignocellulosic samples can be

attributed to content variations of hemicellulose, lignin and cellulose. The higher

content of hemicellulose in Fir Wood could explain the peak in the DTG curve

obtained at low temperatures. In addition, the different contents of lignin could justify

the differences among the residue yield observed, which is related to the fixed carbon

content present in the sample. Pine Bark had the highest fixed carbon content, being

the one leading to the largest residue yield. Finally, the differences in the maximum

weight loss rate (in the order: Eucalyptus Wood> Fir Wood > Pine Bark) can be

explained attending to the volatile matter and cellulose content in these samples

(Damartzis et al., 2011).

Figure 2c shows the TGA/DTG curves for the thermal decomposition of the

variety of microalgae NannochloropsisGaditana(NG microalgae). Its thermal

degradation behaviour can be divided into three zones. The first zone was attributed to

a dehydration process at temperatures below 130ºC. The second zone, between 140 ºC

and 540 ºC, corresponded to a devolatilization process. Three shoulders can be

distinguished in this zone, being mainly associated to the decomposition of different

Chapter 2

115

kind of triglycerides and other hexane-soluble compounds (Marcilla et al., 2009).

Finally, the last zone took place at temperatures above 540 ºC, being mainly related to

residue decomposition.

Chapter 2

116

Figure 2.Thermogravimetric (TGA) and differential thermogravimetric (DTG) curves of the

pyrolysis process of the biomass samples studied: (a) Cellulose, Xylan and Lignin; b) Fir

Wood, Eucalyptus Wood and Pine Bark; (c) a variety of NannochloropsisGaditana microalgae.

Sample holder: alumina; gas flow rate: 200 ml/min; sample size: 10 mg; heating rate: 15

ºC/min.

0

20

40

60

80

100

150 300 450 600 750 9000,00

0,15

0,30

0,45

0,60

Wei

ght (

%)

Fir Wood Eucalyptus Wood Pine Bark

ZO

NE

IV

ZO

NE

III

ZO

NE

II

ZO

NE

I

ZO

NE

IV

ZO

NE

IIIZ

ON

E I

I

Wei

ght l

oss

rate

(w

t%/º

C)

Temperature (ºC)

ZO

NE

I

0

20

40

60

80

100

150 300 450 600 750 9000,0

0,4

0,8

1,2

1,6

2,0

2,4

W

eigh

t (%

)

Cellulose Lignin Xylan

Wei

ght l

oss

rate

(w

t%/º

C)

Temperature (ºC)

140 280 420 560 700 840 980 11200

20

40

60

80

100

ZO

NE

III

ZO

NE

II

ZO

NE

I

Wei

ght l

oss

rate

(w

t.%/º

C)

Wei

ght (

%)

Temperature (ºC)

0,0

0,1

0,2

0,3

0,4

0,5 TGA DTG

a) b)

c)

Chapter 2

117

To sum up, NG microalgae showed higher thermal stability than

lignocellulosicbiomass, decomposing in a broader temperature range (the residue does

not remain constant until temperatures above 900 ºC). Furthermore, the residue yield

obtained for the microalgae was lower (about 15 wt. %) than those for the other types

of biomass considered in this work. Nevertheless, the main loss weight, corresponding

to the pyrolytic process, occurred at the same temperature range as that of the

terrestrial biomass (200-500 ºC). On the other hand, the shape of the DTG curves for

the analyzed samples showed evident differences. For the lignocellulosic biomass, a

well-defined shoulder was observed in the DTG curves whereas three little humps

were detected in the marine one. These differences were attributed to the different

compositions of these materials. NG microalgae are mainly composed of proteins (>60

%) whereas lignocellulose biomass is constituted by cellulose, hemicellulose, lignin

(>90%) and a little amount of extractives (Shuping et al., 2010).

2.3.2. Effect of heating rate

Figure 3 shows TGA/DTG plots versus temperature obtained from the

pyrolysis of Cellulose at different heating rates (5, 15 and 40 ºC/min). This figure

represents the general trend of biomass samples studied during the pyrolysis process.

Chapter 2

118

Figure 3.Effect of the heating rate in the pyrolysis process of Cellulose at 5, 15 and 40 ºC/min.

Sample holder: alumina; gas flow rate: 200 ml/min; sample size: 10 mg.

Table 2 shows the most relevant experimental results for all raw materials. As

it can be seen in Table 2, the behaviour of all of them is quite similar. Generally, as

the heating rate increased, the pyrolysis temperature (Tpyr) and all characteristic

temperatures shifted to higher values (Table 2). It can also be observed that the

maximum weight loss rate decreased as the heating rate was increased. These results

are similar as those reported by other authors (Li et al., 2010; Peng et al, 2001).

0

20

40

60

80

100

250 300 350 400 450 5000,0

0,5

1,0

1,5

2,0

2,5

Wei

ght (

%)

C/min؛ 5 C/min؛ 15 C/min؛ 40

Wei

ght l

oss

rate

(w

t%/º

C)

Temperature (ºC)

Chapter 2

119

Chapter 2

120

Table 2. Pyrolysis temperatures for Cellulose, Lignin, Xylan, Fir Wood, Eucalyptus Wood, Pine Bark and Nannochloropsisgaditanamicroalgae at

different heating rates.

* Temperature at which the pyrolysis started, ** Temperature where a peak in the DTG curve is formed, *** Maximun weight loss rate, Sh= Shoulder

Heating Rate

(ºC/min)

Primary components of biomass Lignocellulosicbiomasss Marine Biomass

Cellulose Xylan Lignin Fir Wood Eucalyptus Wood Pine Bark NannochloropsisGaditana(NG)

1st

peak

2nd

peak

Sh 1st

peak

Sh 1st

peak

2nd

peak

3rd

peak

Sh 1st

peak

Sh 1st

peak

2nd

peak

Tpyr*

(ºC)

5 275 191 202 167 170 180 140

15 290 200 210 184 183 200 145

40 300 209 220 199 201 208 162

Tm **

(ºC)

5 322 235 271 340 218 328 279 305 486 625 297 342 186 305 804

15 340 249 295 355 236 346 292 330 503 667 312 353 205 330 854

40 356 265 310 368 257 368 305 342 514 690 319 365 213 342 911

(dw/dT)

Max***

(wt.%/ºC)

5 2.62 0.5 0.54 0.24 0.18 0.63 0.36 0.47 0.1 0.06 0.29 0.43 0.13 0.47 0.06

15 2.29 0.56 0.52 0.26 0.19 0.59 0.35 0.46 0.09 0.05 0.28 0.4 0.11 0.46 0.05

40 1.94 0.57 0.51 0.28 0.2 0.55 0.33 0.46 0.06 0.04 0.27 0.36 0.10 0.46 0.05

Residue

yield

(wt.%)

5 8.83 26.9 46.2 26 28.46 35.4 5.96

15 9.05 28.2 45.0 25.2 26.63 35.3 9.46

40 9.13 28.6 43.8 24.4 25.58 33.9 10.917

Chapter 2

121

These changes could be mainly attributed to changes in the decomposition

kinetics (Peng et al., 2001) and the fact that an increase of the heating rates provided

higher thermal energy, ensuring a better heat transfer between the surrounding

environment and the sample inside (Li et al., 2010). On the other hand, the residue

yield is one of the parameters that did not remain constant. Several differences can be

observed between the primary components of biomass and lignocellulosic biomass.

First of all, Cellulose and Lignin followed opposite trends than Xylan. The residue

yield for Cellulose and Lignin increased at increasing heating rates, whereas Xylan

residue decreased. Shen et al. (2010) ascribed this fact to structure differences. The

structure of cellulose is chemically and physically rearranged after the “preheating

process”, enhancing the final production of char residue (Maschio et al. 1992). Thus,

the char residue from Cellulose pyrolysis would increase with the longer pre-heating

process at the low heating rate. In Xylan, the structure of the char residue formed at

the low heating rate is less stable than that formed at the high heating rate, leading to

secondary cracking reactions. The char formation in Lignin is enhanced at low heating

rates. According to Nakamura et al. (2007), the formation mechanism of char in

Lignin is attributed to condensation reactions.

The effect of the heating rate in the residue yield for lignocellulosic biomass

followed the same trend in all cases. The higher the heating rate, the lower the residue

yield was. According to White et al. (2011), this fact can be attributed to the

completion of thermal degradation reactions at high heating rates. Finally, NG

microalgae showed a decrease in the residue yield with increasing heating rates. This

is due to the fact that lower heating rates resulted in longer residence times inside the

Chapter 2

122

reactor favouring secondary reactions such as cracking, re-polymerization and re-

condensation, thus leading to char formation (Shuping et al., 2010).

2.3.3. Gas products Analysis

The pyrolysis behavior of biomass by means of TGA-MS has been studied by

different authors (Lin et al., 2009; Widyawati et al., 2011). TGA-MS measurements

reproduce the evolution of the main gas products during the pyrolysis of biomass. This

technique is the only one to simultaneously measure in real time the thermal

decomposition and the gas product distribution of a very small sample. The present

study was focused on the main volatile products of biomass pyrolysis on the basis of

both their relative intensities across the temperature range 40-900ºC and on their

relevancy. H2, -CH3, CH4, C2H4, C2H6, CO, CO2 assigned to the ion/mass intensities

(m/z) 2, 15, 16, 27, 28, 30 and 44, respectively (according to the database of National

Institute of Standards and Technology (NIST)). Mass spectra curves for all samples

are shown in Figure 4.

Mass spectrometry analysis for Xylan, Cellulose and Lignin are shown in

Figure 4a. As aforementioned, the pyrolysis process for Cellulose and Xylan occurred

in a relatively narrow range of temperature (200-500 ºC) coincidental with most of the

gas product detection whereas thermal decomposition of Lignin took place in a wide

temperature range. The main gas detected was in all cases CO2. Compared to Xylan

and Lignin, Cellulose pyrolysis released most of gaseous products in a narrow

temperature range (300-400 ºC). On the other hand, Xylan and Lignin released CH4

and -CH3 groups at 500 ºC.

Chapter 2

123

Firstly, CH4 was generated at 450ºC in Xylan and at 500 ºC in Lignin.

Secondly, H2 was produced as CH4 and –CH3 groups are consumed. This fact can be

attributed to CH4 steam reforming reactions (Eq. (9)) (Widyawati et al., 2011). Finally,

Lignin showed the highest reactivity in the whole range of temperatures. These results

are in good agreement with previous works using TGA-MS techniques (Widyawati et

al., 2011). On the other hand, most of the H2 formation was observed at high

temperatures (>500ºC). H2 production is attributed to secondary reactions as steam

reforming of methane and/or tar cracking (Widyawati et al., 2011; Huang et al. 2011):

CH4 + H2O ↔ CO + 3H2 CH4 steam reforming (9)

CnHmOp + (2n-p)H2O ↔ nCO2 + (1/2m + 2n-p)H2 Tar steam reforming (10)

CnHm ↔Cn-xHm-y+H2+CH4+C Thermal craking (11)

Figure 4b shows the MS spectra for lignocellulosic biomass as a function of

temperature. The process could be divided into 4 stages. Firstly, peaks at low

temperatures (<150 ºC) represented the drying process of the samples. Furthermore,

methyl groups were also detected in a similar way than in Lignin pyrolysis mass

spectra. In the second stage (150-250 ºC), the main pyrolysis products detected were

CO2, CO, H2O and light hydrocarbons (CH4 and C2H6). CH4 and CO2 productions

were also detected at temperatures ranging from 400 to 500ºC. Additionally, at

temperatures above 500 ºC two peaks were detected. The first peak occurred in all

samples at a similar temperature (about 530 ºC); the second one shifted depending on

the sample (650, 675 and 698 ºC for Fir Wood, Pine Bark, and Eucalyptus Wood,

respectively). In spite of these differences, the CO2 and CH4 evolution was similar. On

Chapter 2

124

the other hand, two peaks related to H2 evolution was observed when the rate of CO2

and CH4 formation was decreasing, reaching its maximum values at about 750 ºC.

The product distribution observed in the last stage (470-800 ºC) suggested that

secondary reactions took place. These reactions could be attributed to tar cracking,

(Eqs. 10 and 11), being CO2, CH4 and H2 mainly formed, Eqs. (10-11), self

gasification of samples (Eq. 12) (Huang et al., 2011), and CH4 consumption by steam

reforming (Eq.(9)) (Widyawati et al., 2011).

C+H2O ↔ CO2 + H2 Self Gasification (12)

This way, it can be concluded that most of the H2 produced from

lignocellulosic biomass pyrolysis came from secondary reactions (Widyawati et al.,

2011).

Chapter 2

125

150 300 450 600 750 900 150 300 450 600 750 900

150 300 450 600 750 900

Inte

nsity

(a.

u.)

Temperature (ºC)

O2

CO

C2H

4

CH4

H2

H2O

C2H

6

-CH3

CO2

Xylan Cellulose

O2CO

C2H

4

CH4

H2

H2OC

2H

6

-CH3

Temperature (ºC)

CO2

Lignin

Inte

nsity

(a.

u.)

Temperature (ºC)

O2

CO

C2H

4

CH4 H

2

H2O

C2H

6

-CH3

CO2

150 300 450 600 750 900 150 300 450 600 750 900

150 300 450 600 750 900

Eucalyptus wood

Inte

nsity

(a.

u.)

Temperature (ºC)

O2

CO

C2H

4

CH4

H2

H2O

C2H

6

-CH3

CO2

Fir wood

Temperature (ºC)

O2COC

2H

4

CH4

H2

H2O

C2H

6

-CH3

CO2

Pine bark

O2COC

2H

4CH

4

H2

H2O

C2H

6

-CH3

Inte

nsity

(a.

u.)

Temperature (ºC)

CO2

a)

b)

Figure 4.Mass spectra corresponding to the pyrolysis of different biomass feedstocks: a)

Xylan, Cellulose, Lignin. b) Eucalyptus Wood, Fir Wood and Pine Bark.

Chapter 2

126

Figure 5 shows the mass spectra of NG microalgae. No previous studies have

been found in the literature about the pyrolysis process of microalgae by means of

TGA-MS technique. Nevertheless, the pyrolytic characteristics of microalgae have

been studied by using TG-FTIR (Marcilla et al., 2009) and GC-MS (Ross et al., 2008).

As mentioned above, the pyrolysis process was divided into three stages (Figure 2c).

The first zone (<140 ºC) corresponded to the loss of moisture and very light volatiles

compounds. H2O and CH4 were released in a similar process as that described for

lignocellulosic biomass. In the second zone, associated to the major weight loss, three

well-identified products were detected. A first peak for CO2 and CH4 were detected at

190 ºC. Then, the main pyrolysis products (CO2, CO, CH4 and H2O) were detected

between 240 and 440 ºC. The third stage corresponded to a similar process than that

described for lignocellulosic biomass, where CH4, CO, CO2 and H2 were evolved. This

product distribution agrees well with that reported by Ross et al. (2008). Four stages

were identified: decomposition of carbohydrates (180-270 ºC) and proteins (320-450

ºC), loss of volatile metal and carbonate decomposition (<500 ºC), and char

decomposition (>750ºC) present in NG microalgae leading to H2 and CO2 evolution,

together with a significant proportion of inorganic material decomposed, probably

metal carbonates (Ross et al., 2008).

From the viewpoint of the pyrolysis of lignocellulosic biomass and marine

biomass, two main differences can be observed. Firstly, pyrolysis products from NG

microalgaewere detected at temperatures below 200 ºC. This behaviour is mainly due

to the fact that microalgae are composed by different kind of extractives, triglycerides

and hexane soluble components. These components are less thermal resistant than

Chapter 2

127

150 300 450 600 750 900 1050 1200

Inte

nsity

(a.

u.)

Temperature (ºC)

Algae

O2

CO

C2H

4

CH4

H2

H2O

C2H

6

-CH3

CO2

hemicellulose, cellulose and lignin present in about 90 % of the lignocellulosic

biomass composition (McKendry, 2002). Secondly, H2 production from microalgae at

high temperatures was lower than that from lignocellulosic biomass samples. This fact

could be due to the fact that the char, formed during the pyrolysis of NG microalgae

was less reactive than that occurred for Pine Bark, Fir Wood and Eucalyptus Wood.

Figure 5.Mass spectra of the pyrolysis of NannochloropsisGaditana microalgae.

2.3.4. Kinetic model

Figure 6 shows the experimental (solid line) compared to the predicted curve

(dotted line) obtained by non-linear regression of the kinetic model described in

Section 2.1 for Cellulose, Eucalyptus Wood and NG microalgae pyrolysis at a heating

rate of 40 ºC/min. It can be observed that the proposed model adequately reproduces

the experimental values.

Chapter 2

128

200 400 600 8000

20

40

60

80

100

200 400 600 800 10000

20

40

60

80

100200 400 600 800

0

20

40

60

80

100

Wei

ght (

%)

Experimental Theoretical


Wei

ght (

%)

Temperature (ºC)

Wei

ght (

%)


Figure 6.Comparison between experimental and theoretical results for the pyrolysis of a)

Xylan; b) Eucalyptus Wood; c) NannochloropsisGaditana microalgae.

The Marquardt-Levenberg algorithm was used to obtain kinetic parameters

(Marquardt, 1963). Table 3 shows the weight loss steps, the activation energy, the pre-

Chapter 2

129

exponential factor and the reaction order (n) for each weight loss step during the

pyrolysis of biomass samples.

Table 3.Estimated kinetic parameters for the pyrolysis of different types of lignocellulosic and

marine biomass.

Sample

Step

EA (KJ/mol)

log Ko

(log mol1-nln-1/s)

n

Cellulose 1 191.25 14.54 1

Xylan

1 94.09 14.22 2

2 125.26 12.72 3

3 181.35 13.29 6

Lignin

1 88.93 13.84 1

2 94.13 12.07 2

3 99.14 8.61 4

Fir Wood

1 95.58 14.83 1

2 128.28 13.91 2

3 154.22 14.01 5

Eucalyptus

Wood

1 57.14 8.43 1

2 129.68 12.98 2

3 159.97 13.94 2

4 176.75 12.30 2

5 202.82 11.29 3

Pine Bark

1 91.35 14.05 1

2 142.56 12.90 2

3 166.44 13.20 8

Nannochloropsis

Gaditana microalgae

1 93.64 14.17 2

2 83.46 8.60 2

3 122.71 12.55 4

4 132.38 6.03 3

Chapter 2

130

Cellulose pyrolysis kinetics has been broadly studied (Grønli et al., 1999; Lin et al.,

2009) due to two facts: cellulose is the main component in biomass structure and, its

structure is more homogenous than hemicellulose and lignin (Basu, 2010). Cellulose

thermal decomposition kinetics can be well fit by a single step first order reaction with

an activation energy in the range of 180-240 KJ/mol and a pre-exponencial factor

(log(ko)) of 14-19 log (1/s). (Grønli et al., 1999; Lin et al., 2009; Órfão et al., 1999).

These values agreed well with the experimental results obtained (Table 3).

Xylan and Lignin have also been studied in literature (Rao and Sharma, 1998). The

decomposition of Xylan and Lignin was divided into three steps. Both of them showed

lower activation energies than cellulose in the pyrolysis temperature range (step 2).

The values of activation energies for the three steps in the pyrolytic decomposition of

Lignin were the lowest ones, showing that is the most active material in the whole

range of temperatures.

Lignocellulosic biomass pyrolysis was divided into three steps: moisture evolution

(0-150 ºC), main devolatilization process (150-400 ºC) and char decomposition (400-

900 ºC). In the case of Eucalyptus Wood, the last step was divided into two more

substeps since the char showed higher reactivity. The kinetics parameters for each

substep were in the same range of values, although no similarities were found in

literature (Caballero et al., 1997; Órfão et al., 1999; Shen et al., 2010).

For the kinetic evaluation of NG microalgae pyrolysis and that of for

lignocellulosic biomass four steps were considered (Table 3). The value of the kinetics

parameters obtained in the pyrolysis of lignocellulosic biomass samples and

Chapter 2

131

microalgewere in a similar range, though a lower activation energy for the microalgae

pyrolysis were observed. These results agreed well with those reported by Li et al.

(2010) under similar operating conditions. However, the values of activation energies,

pre-exponential factors and reaction orders did not show a general trend. This fact

suggested than the pyrolysis kinetics was greatly influenced by the type and

composition of the biomass feedstock (Li et al., 2010; Shuping et al., 2010). Table 4

shows the comparison between the pyrolysis kinetic parameters of different types of

biomass sources and the results obtained in this work.

As aforementioned, the discrimination of kinetic parameters was done applying the

F-test and the t-test at the 95% confidence level.

The resulting parameters obtained from the computational non-linear regression are

summarized in Table 5. In terms of statistical results, F-test considered the regression

to be suitable in all cases since the corresponding values to the Fc/Ftest ratio was

larger than one. The t-test was also used for evaluating each parameter in the model.

As shown in Table 5, the values of tc/t-test ratio were also larger than one, showing the

statistical significance of the proposed models and their corresponding parameters.

Chapter 2

132

Table 4.Comparison of pyrolysis kinetics parameters for different types of biomass (Cellulose,

Hemicellulose, Lignin, Fir Wood, Eucalyptus Wood, Pine Bark and different types of

microalgae).

Biomass

Temperature range (ºC)

β (ºC/min)

k (log mol1-n ln-

1/s) Ea

(KJ/mol)

Hemicellulose (xylan)

Huang et al., 2011 200-400 5 4.2 67.6

Rao and Sharma, 1998 270-320 20 9.3 105.0

Present study 260-350 40 12.72 125.3

Cellulose

Grønli et al., 1999 270-360 40 17 222.0

Huang et al., 2011 340-360 5 15.4 210.8

Órfão et al., 1999 200-400 5 16.8 201

Rao and Sharma, 1998 280-350 20 5.7 82.7

Present study 240-390 40 14.5 191.2

Ligning

Huang et al., 2011 250-305 5 0.12 35.4

Rao and Sharma, 1998 300-390 20 4.7 67.0

Present study 225-375 40 13.8 88.9

Pine Bark

Órfão et al., 1999 120-375 5 3.4 48.5

375-650 -1.33 20.4

Present study 150-320 40 15.3 210.8

320-400 5.7 82.7

Fir wood Present study 150-250 40 15.3 210.8

250-500 5.7 82.7

Eucalyptus wood

Órfão et al., 1999 175-400 5 6.72 79.4

400-650 0.16 20.2

Present study 150-320 40 15.3 210.8

320-400 5.7 82.7

Algae

Peng et al. 2001

(S. Platensis) 220-540 40 6.8 82.7

(C. protothecoides) 190-540 3.6 42.2

Present study (NannochloropsisGaditana) 180-530 40 8.6 83.5

Chapter 2

133

Table 5.Estimated statistical parameters for the pyrolysis of different types of lignocellulosic

and marine biomass

Sample Step tc (EA) tc (ko) tc (n) t-test Fc (* 10-3

) F-test

Cellulose 1 3351.79 10096.78 60.00 1.96 954 2.37

1 36.89 2.11 34.53

Xylan 2 62.46 2.05 62.15 1.96 1325 1.83

3 210350 71.87 339

1 48.08 2.52 39.27

Lignin 2 84.93 2.78 68.44 1.96 3659 1.83

3 107.85 4.34 123.14

1 2.66 82.38 75.43

Fir Wood 2 87.99 55.08 89.98 1.96 14544 1.67

3 44.14 3221907 326.4

Eucalyptus Wood

1 77.63 3.98

2 44.22 2.42

3 54.26 3.62 1.96 1.96 871 1.67

4 60.07 60.07

5 53.39 53.39

1 77.58 2.32 89.98

Pine Bark 2 88.42 2.69 78.77 1.96 4218 1.88

3 13976008 107 662

Nannochloropsis

Gaditana

microalgae

1 48.32 2.39 89.98

2 171.35 5.89 9.74 1.96

14544

1.67

3 159.07 4.57 219.95

4 219.95 9.74 219.33

Chapter 2

134

2. 4. CONCLUSIONS

Thermal characteristics and gas formation during pyrolysis of Fir Wood,

Eucalyptus Wood, Pine Bark, NG microalgae and three individual components of

lignocellulosic biomass (hemicellulose, lignin and cellulose) were analyzed by TGA-

MS. Pyrolysis of lignocellulosic biomass was divided into four zones: moisture

evolution, hemicellulose decomposition, lignin and cellulose degradation and lignin

decomposition. NG microalgae showed the highest thermal stability. The main

products (CO2, light hydrocarbons and H2O) were generated between 200 and 450 ºC.

H2 was produced at high temperatures (>700 ºC). Kinetic model satisfactorily

predicted the pyrolysis of biomass. Furthermore, the statistical significance of the

model was proved.

2.5. REFERENCES

(1) Barneto, A.G., Vila, C., Ariza, J., 2011. Eucalyptus kraft pulp production:

Thermogravimetry monitoring. Thermochim. Acta. 520, 110-120.

(2) Basu, P., 2010. Biomass Gasification and Pyrolysis: Practical design and

theory, first ed. Elsevier, Oxford.

(3) Caballero, J.A., Conesa, J.A., Front, R., Marcilla, A., 1997. Pyrolysis kinetics

of almond shells and olive stones considering their organic fractions. J. Anal. Appl.

Pyrol. 42, 159-175.

Chapter 2

135

(4) Damartzis, Th., Vamvuka, D., Sfakiotakis, S., Zabaniotou, A. 2011. Thermal

degradation studies and kinetic modelling of cardoon (Cynaracardunculus) pyrolysis

using thermogravimetric analysis (TGA). Bioresour. Technol. 102, 6230-6238.

(5) De la Osa, A.R., De Lucas, A. Romero, A., Valverde, J.L., Sánchez, P., 2011.

Kinetic models discrimination for the high pressure WGS reaction over a commercial

CoMo catalyst, Int. J. Hydrogen. Energ. 36, 9673-9684.

(6) De Lucas, A., Sánchez, P., Dorado, F., Ramos, M.J., Valverde, J.L., 2006.

Kinetic model of the n-Octane hydroisomerization on PtBeta agglomerated catalyst:

Influence of the reaction conditions, Ind. Eng. Chem. Res. 45, 978-985.

(7) Froment, G.F., Bischoff, K.B., 1990 Chemical reactor analysis and design,

Wiley, New York.

(8) Grønli, M., Antal, M.J., Varhegyi, G., 1999. A round robin study of cellulose

pyrolysis kinetics by thermogravimetry. Ind. Eng. Chem. Res. 38, 2238-2244.

(9) Grønli, M.G., Várhegyi, G., Di Blasi, C., 2002. Thermogravimetric analysis

and devolatilization kinetics of wood. Ind. Eng. Chem. Res. 41, 4201-4208.

(10) Huang, Y.F., Kuan, W.H., Chiueh, P.T., Lo, S.L., 2011. Pyrolysis of biomass

by termal analysis-mass spectrometry (TA-MS). Bioresour. Technol. 101, 1968-1973.

(11) Li, D., Chen, L., Yi, X., Zhang, X., Ye, N., 2010. Pyrolytic characteristics and

kinetics of two brown algae and sodium alginate. Bioresour. Technol. 101, 7131-7136.

(12) Li, D., Chen, L., Zhang, X., Ye, N., Xing, F., 2011. Pyrolytic characteristics

and kinetic studies of three kinds of red algae. Biomass bioenergy. 35, 1765-1772.

(13) Lin, Y-C., Cho, J., Tompsett, G.A., Westmoreland, P.R., Huber, G.W., 2009.

Kinetics and Mechanism of Cellulose Pyrolysis. J. Phys. Chem. C. 113, 20097-20107.

Chapter 2

136

(14) Marcilla, A., Gómez-Siurana, A., Gomis, C., Chápuli, E., Catalá, M.C.,

Valdés, F.J., 2009. Characterization of microalgal species through TGTA/FTIR

analysis: Application to nannochoropsis sp. Thermochim. Acta. 484, 41-47.

(15) Marquardt, D.W., 1963. An algorithm for least-squares estimation of

nonlinear parameters, J. Soc. Ind. Appl. Math. 11, 431-441.

(16) Maschio, G., Koufopanos, G., Lucchesi, A., 1992. Pyrolysis, a promising

route for biomass utilization. Bioresour. Technol. 42, 219-231.

(17) McKendry, P. 2002. Energy production from biomass (part 1): overview of

biomass. Bioresour. Technol. 83, 37-46.

(18) Nakamura, T., Kawamoto, H., Saka, S., 2007. Condensation reactions of some

lignin related compounds at relatively low pyrolysis temperature. Journal of wood

chemistry and technology. 27, 121-133.

(19) Órfão, J.J.M., Antunes, F.J.A., Figueiredo, J.L., 1999. Pyrolysis kinetics of

lignocellulosic materials-three independent reactions model. Fuel. 78, 349-358.

(20) Peng, W., Wu, Q., Tu, P., Zhao, N., 2001. Pyrolytic characteristics of

microalgae as renewable energy source determined by thermogravimetric analysis.

Bioresour. Technol. 80, 1-7.

(21) Press, W.H., Teukolsky, S.A., Vetterling, W.T., Flannery, B.O., 1992.

Numerical recipes in Fortran, second ed. Cambridge University Press, Cambridge.

(22) Rao, T.R., Sharma, A., 1998. Pyrolysis rates of biomass materials. Energy. 23,

973-978.

Chapter 2

137

(23) Ross, A.B., Jones J. M., Kubacki, M.L., Bridgeman, T. 2008. Classification of

macroalgae as fuel and its thermochemical behavior. Bioresour. Technol. 99, 6494-

6504.

(24) Severian, D., 2008. Polysacharides: Structural diversity and functional

versatility, Second ed. Marcel Dekker, New York.

(25) Shen, D.K., Gu, S., Bridgwater, A.V., 2010. The thermal performance of the

polysaccharides extracted from hardwood: Cellulose and hemicellulose. Carbohydr.

Polym. 82, 39-45.

(26) Shuping, Z., Yulong, W., Mingde, Y., Chun, L., Junmao, T., 2010. Pyrolysis

characteristics and kinetics of the marine microalgae Dunaliellatertiolecta using

thermogravimetric analyzer. Bioresour. Technol. 101, 359-365.

(27) Stenseng, M., Jensen, A., Dam-Johansen, D., 2001. Investigation of biomass

pyrolysis by thermogravimetric analysis and differential scanning calorimetry. J. Anal.

Appl. Pyrolysis. 58-59, 765-780.

(28) Sun, L., Chen, J.Y., Negulescu, I.I., Moore M.A, Collier B.J., 2011. Kinetics

modeling of dynamic pyrolysis of bagasse fibers, Bioresour. Technol. 102, 1951-1958.

(29) Valverde, J.L., De Lucas, A., Carmona, M.S., González, M., Rodríguez, J.F.,

2004. A generalized model for the measurement of effective diffusion coefficients of

heterovalent ions in ion exchangers by the zero-length column method, Chem. Eng.

Sci. 59, 71-79.

(30) Wang, G., Li, W., Li, B., Chen, H., 2008. TG study on pyrolysis of biomass

and its three components under syngas. Fuel. 87, 552-558.

Chapter 2

138

(31) Wang, J., Wang, G., Zhang, M., Chen, M., Li, D., Min, F., Chen, M., Zhang,

S., Ren, Z., Yan, Y., 2006. A comparative study of thermolysis characteristics and

kinetics of seaweeds and fir wood. Process Biochem. 41, 1883-1886.

(32) White, J.E., Catallo, W.J., Legendre, B.L., 2011. Biomass pyrolysis kinetics:

A comparative critical review with relevant agricultural residue case studies. J. Anal.

Appl. Pyrolysis. 91, 1-33.

(33) Widyawati, M., Church, T.L., Florin, N.H., Harris, A.T., 2011. Hydrogen


oxide. Hydrogen Energy. 30, 1-14.

(34) Yaman, S. 2004. Pyrolysis of biomass to produce fuels and chemical

feedstocks. Energy Convers. Manage. 45, 651-671.

(35) Yang, H., Yan, R., Chen, H., Zheng, C., Lee, D.H., Liang, D.T., 2006. In-

Depth Investigation of Biomass Pyrolysis Based on Three Major Components:

Hemicellulose, Cellulose and Lignin. Energy Fuels. 20, 388-393.

Chapter 3:

THERMOGRAVIMETRIC -MASS SPECTROMETRIC ANALYSIS ON COMBUSTION

OF LIGNOCELLULOSIC AND MARINE BIOMASS

The combustion characteristics of lignocellulosic biomass and

marine biomass were investigated by means of TGA-DSC-MS.

Additionally, Canadian crops were investigated. The combustion

process can be mainly divided into two main steps: devolatilization

(Dev. stage) and char oxidation stage (Oxid. stage). The thermal

behavior of both types of biomass could be related to the

decomposition of their main components. During the Dev. stage of

lignocellulosic biomass the hemicellulose and cellulose decompose

whereas the oxidation stage was mainly ascribed to lignin. On the

other hand, in the Dev. stage for microalgae all of their main

components decompose in the following order: carbohydrates,

lipids and proteins. Combustion kinetics were studied by Pseudo

multi-components separate-stage models (PMSM). Models based

on reaction order (Oi), nucleation (Ni) and diffusion (Di) achieved

the best fitting to the experimental data. Additionally, the process

was successfully modeled obtaining errors below ± 3.35 %. CO,

CO2 and H2O were the main components evolved from

Chapter 3

136

combustion. Additionally, light hydrocarbons (CH4 and C2H5) were

also present. Finally, nitrogen compounds were in a higher

proportion than sulfur compounds being released as primary

amines and NOx. The NOx release was higher for the combustion

of microalgae than for lignocellulosic biomass due to their high

initial nitrogen content. The high ash content of microalgae and

Canadian biomass samples catalyzed the volatile compounds

release and shifted the process to lower temperatures. Furthermore,

this fact implied that sample pre-treatment is required before being

used in thermal applications Finally, the main pollutants released

during the combustion process of Canadian biomass were

analyzed. Nitrogen (NO, NO2 and HCN) and sulfur (SO and SO2)

compounds were found in high proportions. Nitrogen compounds

were released in both combustion stages, whereas sulfur

compounds evolved mainly in the a lower temperature range. Other

pollutants were found in lower concentrations (CH3Cl and C6H6).

3.1. INTRODUCTION

As already mentioned in Chapter 2, thermochemical conversion of biomass is

considered as one of the most promising processes for biomass utilization. These

processes are employed for power generation, production of liquid biofuels, chemicals

and charcoal.Thus, a good understanding of the decomposition of biomass during

thermochemical conversion is important for developing efficient processing

technology [1].

Combustion can be defined as the conversion of biomass fuels to several forms of

useful energy in the presence of air or oxygen. Thermogravimetric techniques have

commonly been used to investigate the thermochemical conversion of solid raw

materials as coal and woods [2-4].Unlike pyrolysis, the combustion of biomass using

Chapter 3

137

thermogravimetric analysis has not been studied intensively yet [5].Recently, the

combustion behavior of different types of wood has been performed[1; 5;6].

Furthermore, Kay et al., (2011) studied the effect of biomass components on the co-

combustion of biomass and coal.

Biomass characteristics and kinetics of biomass combustion are essential for

modelingthe combustion in industrial processes [7]. Furthermore, a knowledge of the

kinetics of the process has great importance for a correct design and product yield

control [8]. Although there are significant differences in operating conditions,

thermogravimetric analysis provides a powerful tool to accomplish preliminary kinetic

studies on the thermal decomposition of solids.Algae are a very promising biomass for

the following reasons: rapid growth rate, high yield per area, high efficiency in CO2

capture and solar energy conversion and no competition with food agriculture. Among

the different types of algae, microalgae have received more attention than others

because they can be cultured in ponds or photobioreactors with supply of nutrients or

wastewater [9].

In Canada, the average annual wood cut has been estimated at 167.5 million m3

creating over 60 million tons of residues. The annual harvests are taken from

approximately 1 million ha, constituting of only about 0.25 % of the total forestland in

Canada[10]. Thus, due to the great potential of Canada soil and variety, it seems

mandatory to invest on crops for energy production.

Dedicated energy crops can come from multiple sources. However, it is recognized

the high potential of woody crops (hardwoods and pines) and non-woody, high-

yielding annual and perennial crops (Miscanthus, switchgrass and sweet sorghum) due

to multiples advantages. Some of the benefits of energy crops include: less capital-

intensive conversion technologies, attractive opportunity for local and regional self-

sufficiency, reduction in greenhouse gas emissions and viable alternative to fossil fuel

use[11]. In spite of the environmental advantages, some aspects concerning the release

of contaminants during biomass combustion must be taken into account. In this regard,

NOx and SOx emissions depend on raw biomass composition, which usually is

Chapter 3

138

variable[12]. The control of NOx and SOxfrom biomass combustion can contribute to

decrease the emissions of these pollutants in Canada which are around 0.7 and 1.2

million tons per year for NOx (as NO2) and SO2, respectively[13]. Furthermore, the

chloride amount in biomass might turn into operational problems such as corrosion.

Other organic compounds such as benzene and toluene are considered to be part of the

most dangerous emissions from biomass combustion causing diseases as lung

infection and leukemia[14]. Therefore, the knowledge of pollutant release during

biomass conversion is truly important in order to reassure the use of biomass from the

environmental point of view.

As reported by Naik et al. (2010), suitable mathematical models can be derived for

a better comprehension of the thermal behavior of these complex feedstock that allow

to perform economic analysis and develop technology for a more efficient biomass

conversion[10].

The aim of this work was to perform a comprehensive study of the combustion

behavior of lignocellulosicand marine biomassby means of the TGA-DSC. Firstly,

three different types of lignocellulosic biomass(fir wood, eucalyptus wood and pine

bark) and their main components (cellulose, hemicellulose and lignin)

technique.Secondly, three different microalgae species were selected:Chlorella

vulgaris, NannochloropsisGaditana and ScenedesmusAlmeriensis. Finally, two

different types of Canadian biomass were considered: woody crops (black spruce and

Pinusbanksiana mixtures and willow) and different non-woody and perennial crops

(common reed, reed phalaris and switchgrass). In addition, the effect of the heating

rate on the combustion behavior of lignocellulosicsamples was also studied.

Furthermore, the kinetics of the oxidation process were evaluated. Finally,the gases

released during the combustion process were analyzed by MS. The main pollutant

gases released during the combustion process of Canadian biomass were analyzed.

3.2. EXPERIMENTAL

3.2.1. Materials

Chapter 3

139

The lignocellulosic and marine biomass samples used in this study are those

mentioned in Chapter 1. Furthermore, the combustion behavior of Canadian biomass

samples was also evaluated. In this regard, two woody crops (black spruce and

Pinusbanksiana mixtures and willow) and three different non-woody and perennial

crops (common reed, reed phalaris and switchgrass) were harvested and collected

from the province of Quebec (Canada). These samples were dried in an oven for 5 h,

milled and sieved to an average particle size between 100-150 µm.The proximate

analysis of Canadian biomass samples are shown in Table 3.1. The metal content in

samples was determined by Inductively Coupled Plasma Spectrometry (ICP) is shown

in Table 3.2.

Table 3.1.-Proximate and ultimate analysis of biomass samples (black spruce and

Pinusbanksiana mixtures (BP), willow (W), common reed (CR), reed phalaris(RP) and

switchgrass (S))

Ultimate Analysis (wt. %)*daf

Biomass C H N S O*diff

BP 47.1 6.1 0.108 0.06 51.6

W 45.1 5.9 0.584 0.54 30.9

S 43.5 6.2 0.624 0.11 55.3

CR 39.2 4.9 0.627 0.03 53.1

RP 43.7 5.5 0.653 0.08 41.7

Proximate Analysis (wt. %)*daf

Moisture Ash VM* FC* diff

BP 3.7 1.1 79.8 15.4

W 5.2 6.3 68.9 19.5

S 4.4 7.5 74.4 13.7

CR 3.9 16.1 66.9 13.1

RP 3.7 6.7 72.4 17.2

BP 3.7 1.1 79.8 15.4

Chapter 3

140

*daf dry and ash free basis; VM: Volatile matter; FCdiff : wt. % of Fixed carbonwas calculated from difference from

moisture, ash and volatile matter; Odiff: wt.% of oxygen calculated from difference of C, H, N and S.

Table 3.2.-Characterization of biomass samples (black spruce and Pinusbanksiana

mixtures (BP), willow (W), common reed (CR), reed phalaris(RP) and switchgrass (S))


Biomass Cl P K Ca Mg Al Cu Fe

BP 14 64.8 471 1532 276 137 1 55

W 34 2154 5884 13675 1587 1529 10 618

S 129 1570 2286 6242 1529 1683 6 528

CR 1219 625 2114 4402 1468 1499 11 2648

RP 473 11371 6354 2127 960 178 5 333

Cr Na Ni Ba Sr Si Mn Zn

BP 2 105 2 13 5 407 151 15

W 9 340 1 41 37 7159 132 269

S 10 335 2 21 18 24048 97 36

CR 18 671 3 22 - 20314 99 71

RP 1 19 2 11 1 12358 61 27


3.2.2.1. Thermogravimetric analysis for the combustion process

The combustion of biomass components was firstly carried out in a TGA apparatus

(TGA-DSC 1, METTLER TOLEDO). The sample was preheated at 105 ºC for 10 min

to remove the moisture content. Subsequently, the sample was heated from 105 to

1000 ºC at 40 ºC/min under a reactive atmosphere of 21% of oxygen and 79 % of

Argon. Additionally, the effect of the heating rate on the combustion process of

lignocellulosic biomass was evaluated. In this regard, different heating programmes

were used (10, 20, 40 and 80 ºC/min) Previous studies were carried out according to

the procedure described in Chapter 1 in order to avoid heat and mass transfer

Chapter 3

141

limitations. In this sense, sample weight was kept at 6 mg, the particle size was kept in

the 100-150 µm range and a constant flow rate of 100 Nml/min was used.

3.2.2.2. TGA-MS analysis of the Gaseous Products


carried out in a thermogravimetricanalyzer (TGA-DSC 1; METTLER TOLEDO)




circumvent condensation of exhausting gases. A semiquantitative analysis was

performed by using a normalization procedure. The ion intensities were normalized to

the intensity of 38Ar isotope to eliminate systematic instrumental errors [15].

3.2.2.3. Kinetic analysis

Kinetic data from solid-state combustion was obtained using thermogravimetric

analysis. The devolatilization curve is usually obtained as a sum of the corresponding

individual components contributions [2; 16]. However, solid-state reactions are a more

complex process, involving processes such as nucleation, adsorption, desorption and

surface/bulk diffusion [17].

The model proposed in this work is similar to that reported by Gil et al.[2] for the

determination of the combustion kinetic parameters of coal/biomass blends.The

kinetic rates were based on the Arrhenius equation[2; 6].

�� = � ∙ �(�) (1)

� = � ∙ � �/�∙� (2)

wheref(α) represents the hypothetical model of the reaction mechanism;k is the

reaction rate; Arepresents the pre-exponential factor (min-1); Eis the activation energy

(kJ mol-1);Tis the absolute temperature (K); trepresents the time (min), and αis the

degree of conversion defined by:

Chapter 3

142

� = (�� −��) (�� −��)� (3)

wheremoand mtrepresent the mass at t=0 and t=t, respectively, and mfis the final mass

of the sample.

For a constant heating rate β(K min-1), β= dT/dt, Eq. (1) can be transformed into:

�� (�)� = � �� (4)

By integrating Eq. (4) gives:

�(�) = � �� (�)�� = � ��

� �� (5)

whereg(α) is the integral function of conversion.

Eq. (5) is integrated by using the Coats-Redfern method [18]:

ln[�(�) �$� ] = ln[� ∙ & � ∙ ' ∙ (1 − 2&� '� +� ] − ' & ∙ �� (6)

Generally, the term 2RT/Ecan be neglected as it is much less than unity[19]. It has

been demonstrated that for both, the temperaturesof combustion range and most values

of E, the expression ln[AR/βE] in Eq. (6) is essentially constant [20]. Thus, if the

correct expression of g(α) is used, the plot of ln[g(α)/T2] against 1/T should give a

straight line with a high correlation coefficient of linear regression analysis, from

which the values of E and A can be respectively calculated from the slope of the line;

and by the intercept term in Eq. (6).The functions f(α) and g(α)referredto the different

models for reaction are presented in Table 3.3(White et al., 2011).The Marquardt

algorithm was used in the linear regression procedure to obtain the kinetic parameters

(E and A). Furthermore, the statistical significance of the estimated parameters based

on the F-test and t-test was performed according to the procedure described

elsewhere[21].

Chapter 3

143

Table 3.3.- Expressions for the most common reaction mechanisms in solid state reactions [17]

Reaction model f(α) g(α)

Reaction order

O0 (1-α)n α

O1 -ln(1-α)

O2 -(1-α)-1

O3 1/2 (1-α)-2

Phase boundary controlled

reaction

R2 (1-α)(1-1/n) 1-(1-α)(1/2)

R3 1-(1-α)(1/3)

Power Law

P1 n(α)(1-1/n) α1/4

P2 α1/3

P3 α1/2

P4 α3/2

Nucleation and growth (Avrami-Erofeev equation)

N1 n(1-α)[-ln(1-α)](1-1/n) [-ln(1-α)](1/1.5)

N2 [-ln(1-α)](1/2)

N3 [-ln(1-α)](1/3)

N4 [-ln(1-α)](1/4)

Diffusion

D1 1/2α α2

D2 [-ln(1-α)]-1 (1-α)ln(1-α)+α

D3 3/2(1-α)2/3[1-(1-α)1/3]-1 [1-(1-α)1/3]2

D4 3/2[(1-α)1/3-1]-1 1-2/3α-(1-α)2/3

Chapter 3

144

3.3. RESULTS AND DISCUSSION.

3.3.1. Combustion of lignocellulosic biomass

Thermogravimetricstudy on combustion of lignocellulosic biomass

Figure 3.1 shows the TGA-DTG profiles ofcombustion between 105 and 1000 ºC

for the main components of lignocellulosic biomass (cellulose, xylan and lignin) and

different types of lignocellulosic biomass (eucalyptus wood, fir wood and pine bark)

at a heating rate of 40 ºC/min. Table 3.4 summarizes the most relevant combustion

characteristics of lignocellulosic biomass at heating rates of 10, 20, 40 and 80 ºC/min.

Figure 1.-Thermogravimetric curves for the combustion process of: a) main components of

lignocellulosic biomass (cellulose, lignin and xylan) and b) lignocellulosic biomass (eucalyptus

wood, fir wood and pine bark).

0

20

40

60

80

100

125 250 375 500 625 750 875 10000.0

0.5

1.0

1.5

2.0

2.5

3.0

0

20

40

60

80

100

125 250 375 500 625 750 875 10000.0

0.2

0.4

0.6

0.8

Wei

ght (

%)

Lignin Xylan Cellulose

Wei

ght l

oss

rate

(%

wt./

ºC)

Temperature (ºC)

Wei

ght (

%)

Eucalyptus Wood Fir Wood Pine Bark

Wei

ght l

oss

rate

(%

wt./

ºC)

Temperature (ºC)

Chapter 3

145

Table 3.4.-Combustion characteristics for cellulose, lignin, xylan, fir wood, eucalyptus wood and pine bark at different heating rates

Primarycomponents of biomass Lignocellulosicbiomass

Cellulose Xylan Lignin Firwood Eucalyptuswood Pine bark

1st

peak 2ndpeak 1st

peak 2ndpeak 3rd

peak 4rd

peak 1st

peak 2ndpeak 3rd

peak 4rd

peak 1st

peak 2ndpeak 3rd

peak 1st

peak 2ndpeak 3rd

peak 1st

peak 2ndpeak 3rd

peak 4th

peak

Tdo

*(ºC) 244 196 138 184 185 186

262 205 142 203 196 196

270 217 147 212 206 199

283 226 166 228 214 212

Tpo* (ºC) 244 438 196 262 385 578 138 348 455 - 184 372 575 185 397 566 186 357 443 -

262 455 205 277 396 561 142 302 434 - 203 395 587 196 407 588 196 370 446 607

270 479 217 288 429 - 146 309 446 - 212 413 607 206 413 608 199 380 452 616

283 518 226 304 443 - 166 313 459 832 228 434 640 214 436 624 212 388 481 653

Tpf* (ºC) 413 556 262 368 530 693 348 455 555 - 360 505 652 355 484 653 342 443 531 -

429 584 277 380 556 703 302 434 619 - 383 529 658 378 505 675 356 446 561 645

443 616 288 383 618 - 309 442 757 - 393 553 679 390 525 699 368 452 589 667

459 649 304 396 680 - 313 459 832 911 405 622 711 418 606 735 388 481 653 715

Tp* (ºC) 325 493 243 285 485 617 285 444 481 - 318 424 608 290 431 620 314 422 469 -

339 517 253 289 491 605 297 395 499 - 330 438 630 303 438 636 322 427 473 627

354 532 264 298 550 - 309 397 518 - 340 455 652 314 424 667 333 431 491 645

373 544 274 304 557 - 312 404 537 880 350 449 682 328 451 694 344 465 528 658

(dw/dT)max*

(dwt.%/ºC) 3.247 0.142 0.675 0.616 0.387 0.256 0.144 0.590 0.905 - 0.719 0.532 0.042 0.701 0.458 0.057 0.542 0.440 0.450 -

3.129 0.106 0.778 0.619 0.335 0.302 0.141 0.277 0.603 - 0.692 0.506 0.024 0.697 0.435 0.029 0.521 0.431 0.435 0.016

2.739 0.102 0.841 0.631 0.254 - 0.132 0.271 0.336 - 0.678 0.329 0.021 0.657 0.405 0.027 0.509 0.363 0.37 0.012

2.264 0.081 0.978 0.699 0.152 - 0.123 0.213 0.301 0.02 0.652 0.222 0.020 0.634 0.222 0.027 0.451 0.248 0.080 0.08

Residue yield (%)

0.13 5.81 5.48 4.82 5.43 5.98

0 3.46 5.51 2.74 3.48 2.69

0 4.44 2.61 3.49 2.68 1.76

2.24 4.05 2.93 3.21 3.47 2.89

Chapter 3

146

There are substantial differences in the thermal behavior of the main components

of lignocellulosic biomass (Figure 3.1a), which are mainly attributed to their different

chemical structures [3]. Lignin was the first component to decompose (146 ºC)

whereas xylan and cellulose started decomposing at 187 and 266 ºC, respectively. This

behavior is attributed to the fact that lignin and xylan have methoxy functional groups,

which tend to break easily[3].

Lignin was the most thermal stable component decomposing in two steps, starting

at 146 and 550 ºC, respectively. DTG curve for lignin oxidation presented the smallest

and widest peaks. This fact is due to lignin is polymeric in nature with a three-

dimensional structure consisting of phenylpropane coupled with C-C or C-O-C bonds

whose activity covers a wide range of temperatures [3]. The DTG curve for lignin

combustion shows two main peaks at 397 and 518 ºC (decomposition rate of 0.295

and 0.326 %/ºC, respectively). The burnout temperature was established at 757 ºC.

Xylanwas the least thermally stable component of biomass. Two strong

decomposition peaks overlapped in the temperature range of 147-371 ºC, having the

maximum DTG peak at 264 ºC (0.841 %/ºC). This degradation step is mainly

attributed to C-O-C and some pyranose C-C bonds breakdown[5]. A second step was

observed between 429 and 615 ºC.

The primary decomposition of cellulose, in which 87 % by weight was lost, took

place between 266 and 423 ºC. The DTG peak for cellulose oxidation was found at

354 ºC and presented the highest weight loss rate of all samples (2.74 %/ºC). During

this stage, a complex set of reactions as denitration and deacetylation, scission of O-N,

C-O, CC and C-H bonds might take place [22]. A smaller peak was found between

441 ºC and 637 ºC (0.10 %/ºC) that can be attributed to char oxidation.

Figure 1b shows the TGA-DTG profiles for the combustion process of

lignocellulosic biomass (eucalyptus wood, fir wood and pine bark). The thermal

degradation of lignocellulosic biomass is often reported as the addition of the primary

decomposition of their main components [3; 5;23]. As can be seen from Figure

Chapter 3

147

3.1.b,the lignocellulosic combustion biomass presented similar TGA-DTG profiles,

exhibiting three decomposition peaks. In the temperature range between 180 and 388

ºC, a marked peak with a shoulder corresponding to the decomposition of cellulose

and hemicellulose, was detected. According to different authors [3; 24;25], this stage

represents the release of volatiles and their ignition leading to char formation. Then, a

broad peak between 368 and 600 ºC,related to char oxidation,was observed. Lignin is

the main contributor in this stage as it is the main responsible for biomass char

formation [23]. Finally, a smaller peak was observed for all samples at temperatures

above 625 ºC. This step is mainly related to inorganic matter decomposition as

carbonates [26].

In spite of their similarity, several differences can be observed in the combustion

behavior of the lignocellulosic samples here considered. Eucalyptus wood is the

biomass sample with high cellulose content and low hemicellulosecontent. According

to Kai et al.[3], a high content in cellulose shifts the devolatilization stage to lower

temperatures, increasing the decomposition rate. Furthermore, the shoulder in the

DTG curve, indicating hemicellulose decomposition, was less marked. On the other

hand, pine bark had the minor cellulose content, presenting the lowest weight loss rate,

whereas the shoulder related to the hemicellulose decomposition shoulder was the

most pronounced. The second stage, corresponding to the oxidation of the char formed

during the devolatilization stage, was characterized by the presence of lignin in the

corresponding tested sample [3]. In this case, the DTG peak for the pine bark

combustion was the widest one, from 368 to 589 ºC compared to the fir and

eucalyptus woods. Finally, the last stage was similar in all samples studied, pointing

out that cellulose, hemicellulose and lignin content had no influence on the residue

generated at the end of the thermal process[3].

Effect of the heating rate.

Figure 3.2 shows the DTG profiles for the combustion of biomass main

components (cellulose, xylan and lignin) and lignocellulosic biomass(eucalyptus

Chapter 3

148

wood, fir wood and pine bark) at heating rates of 10, 20, 40 and 80 ºC/min. Table

3.4shows the most relevant characteristics of the combustion process.

Figure 3.2.- Effect of the heating rate in the combustion process of: a) main components of


wood, fir wood and pine bark) at 10, 20, 40 and 80 ºC/min.

Figure 3.2a shows the DTG plots for the combustion of the main components of

biomass at different heating rates. Generally, the higher the heating rate, the higher the

temperature at which the peak appeared. This fact is attributed to the poor thermal

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

125 250 375 500 625 750 875 10000.0

0.2

0.4

0.6

0.8

0.0

0.2

0.4

0.6

0.8

1.0

0.0

0.2

0.4

0.6

0.8

0.0

0.2

0.4

0.6

0.8

125 250 375 500 625 750 875 10000.0

0.1

0.2

0.3

0.4

0.5

Wei

ght l

oss

rate

(%

wt./

min

)

10 ºC/min 20 ºC/min 40 ºC/min 80 ºC/min

Cellulose

Lignin

Temperature (ºC)

10 ºC/min 20 ºC/min 40 ºC/min 80 ºC/min

Xylan

Fir wood

Wei

ght l

oss

rate

(%

wt./

min

)

Eucalyptus wood

Pine bark

Temperature (ºC)

Chapter 3

149

conductivity of lignocellulosic biomass, resulting in a particle gradient temperature

[17]. Furthermore, the weight loss rate decreased with increasing values of theheating

rate. Cellulose main peak shifted from 325 ºC at 10 ºC/min to 373 ºC at 80 ºC/min,

whereas char oxidation peak turned from 493 ºC to 544 ºC. The maximum weight loss

rate shiftedfrom 3.247 %/ºC to 2.264 %/ºC whereas the peak for the char oxidation

varied from 0.142 to 0.081. Xylan and lignin combustion behavior followed a similar

trend, although several differences can be observed. Maximum weight loss rate

increased in the devolatilization stage with increasing heating rates for Xylan (from

0.675 to 0.978 %wt./ºC).These findings agree well with those obtained by William

and Besler[27]. Furthermore, at low heating rates,10 and 20 ºC/min, an additional

peak was found at temperatures above 500 ºC. This peak is mainly attributed to the

presence of impurities that could not be extracted from the raw material [5]. These

small impurity peaks vanished at 40 and 80 ºC/min by the overlapping effect caused

by the use of high heating rates.Finally, lignin experimented an irregular behavior. In

this case, three peaks were identified at 10, 20, and 40 ºC/min, whereas an additional

peak was found at temperatures above 832 ºC. At 10 ºC/min the combustion of lignin

was mainly performed between 300 and 525 ºC, occurring two peaks at 444 and 481

ºC in a similar way as that reported by Cheng et al. [5]. However, when the heating

rate was increased, the first peak shifted to lower temperatures whereas the second one

did to higher temperatures. As aforementioned, both peaks became broader by the

overlapping effect caused by using high heating rates.

Figure 3.2b shows the DTG combustion profiles for lignocellulosic biomass.

Concerning the heating rate effect on lignocellulosic biomass combustion behavior, all

the samples followed the general trend. The higher the heating rate, the higher the

peak temperature was whereasthe maximum weight loss rate, for all DTG peaks,

decreased. In all cases, the oxidation peak was broader when heating rate was

increased.

Chapter 3

150

Kinetic analysis

The kinetic model used in this work was derived fromthe PMSM (Pseudo multi-

component separate-stage models) approach. In this type of models, the biomass

sample is composed of multiple pseudo components [19]. In this regard, the kinetic

parameters can be determined assuming single separate reactions for the different

stages of thermal conversion. As aforementioned, the DTG plots represented different

decomposition peaks dividing the combustion process of biomass samples in different

stages. Each stage represents a separate reaction. Biomass combustion was clearly

defined by three main stages: devolatilization stage (Dev. stage), char oxidation stage

(Oxid. stage) and remaining char burning stage (Rem. stage). However, some samples

experimented additional decompositions. For example, the Dev. Stage for xylan was

represented by two peaks. In order to differentiatethem, the Dev. stagefor xylan was

divided into two stages:Dev. stage A and Dev. stage B. Furthermore, xylancombustion

showed a peak athigh temperatures named as Imp. stage.Additionally, the Oxid.stage

for lignin and pine bark combustion was characterized by two peaks. In a similar way,

these stages were named as Oxid. stage A and Oxid stage B, respectively.This way,

eq. (6) was used separately to each of the stages commented above.In order to obtain

reliable kinetic data, operating parameterswere established according to the procedure

described in Chapter 1.

The model representing the form of g(α)(Table 3.3), which delivered the highest

correlation coefficient,was considered to be the function representing the mass loss

kinetics for the samples under study. The function g(α) depends on the mechanism

controlling the reaction and the size and shape of the reacting particles [28]. Figure 3.3

and 3.4 shows the plots of ln[g(α)/T2] versus 1/T that provided the best linearity at 10,

20, 40 and 80 ºC/min for biomass main components and lignocellulosic biomass

samples, respectively. Table 3.5 and 3.6 summarizes themain kinetic parameters of

biomass samples.It can be seen from Figure 3.3,Table 3.5 and Table 3.6 that all the

stages fitted well into a straight line. All cases showed an acceptable linear fit

regression (r2> 0.9).However, only models based on reaction order (Oi), nucleation

Chapter 3

151

Figure 3.3.- Plot of ln(g(α)/T) vs 1/T for the combustion process of: a) main components of


wood, fir wood and pine bark)at 10, 20, 40 and 80 ºC/min.

0.0012 0.0016 0.0020 0.0024

-24

-22

-20

-18

-16

-14

-12

-10

0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

0.0008 0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

0.0008 0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

ln (g

( αα αα)/T

2 )

Dev. Stage - D3

Oxid. Stage A - O3

Oxid. Stage B - O1

Dev. Stage - D3

Oxid. Stage A - O1

Oxid. Stage B - O1

10 ºC/min 80 ºC/min40 ºC/min

20 ºC/min Dev. Stage - D3

Oxid. Stage A - O1

Oxid. Stage B - O1

1/T (K -1)

Dev. Stage - D3

Dev. Stage - O1

Dev. Stage - O1

Dev. Stage - O1

0.0012 0.0016 0.0020 0.0024

-22

-20

-18

-16

-14

-12

-10

0.0012 0.0016 0.0020 0.0024

-22

-20

-18

-16

-14

-12

-10

0.0012 0.0016 0.0020 0.0024

-22

-20

-18

-16

-14

-12

-10

0.0012 0.0016 0.0020 0.0024

-22

-20

-18

-16

-14

-12

-10

ln (g

( αα αα)/T

2 )

Dev. Stage - O1

Oxid. Stage - O1


Dev. Stage - O1

Oxid. Stage - O1

20 ºC/min

1/T (K -1)

Dev. Stage - O1

Oxid. Stage - O1

Dev. Stage - O1

Oxid. Stage - O1

Cellulose

Lignin

0.0008 0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

-8

-6

0.0008 0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

-8

-6

-4

0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

-8

-6

0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

-8

-6 Dev. Stage A-N1

Dev. Stage B-O2

Oxid. Stage -O1

Imp. Stage -O2

ln

(g( αα αα

)/T2 )

Dev. Stage A-N1

Dev. Stage B-O2

Oxid. Stage -O1

Imp. Stage -O1


20 ºC/min Dev. Stage A-N1

Dev. Stage B-O2

Oxid. Stage -O1

1/T (K-1)

Dev. Stage A-N1

Dev. Stage B-O2

Oxid. Stage -O1

Xylan

0.0012 0.0016 0.0020 0.0024-20

-18

-16

-14

-12

-10

-8

0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

0.0008 0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

0.0008 0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

ln (g

( αα αα)/T

2 )

Dev. Stage - O1

Oxid. Stage - O1

Rem. Stage - O1

Dev. Stage - O1

Oxid. Stage - O1

Rem. Stage - O1


20 ºC/min Dev. Stage - O1

Oxid. Stage - O1

Rem. Stage - O1

1/T (K-1)

Dev. Stage - O1

Oxid. Stage - O1

Rem. Stage - O1

Fir wood

0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

0.0008 0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

0.0008 0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

0.0008 0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

ln (g

( αα αα)/T

2 )

Dev. Stage - O1

Oxid. Stage - O1

Rem. Stage - O1

Dev. Stage - O1

Oxid. Stage - O1

Rem. Stage - O1



Oxid. Stage - O1

Rem. Stage - O1

1/T (K-1)

Dev. Stage - O1

Oxid. Stage - O1

Rem. Stage - O1

Eucalyptus wood

0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10

0.0008 0.0012 0.0016 0.0020 0.0024-22

-20

-18

-16

-14

-12

-10 Dev. Stage - O

1

Oxid. Stage A - O1

Oxid. Stage B - O1

ln (g

( αα αα)/T

2 )

Dev. Stage - O1

Oxid. Stage A - O1

Oxid. Stage B - O1

Rem. Stage - O1



Oxid. Stage A - O1

Oxid. Stage B - O1

Rem. Stage - O1

1/T (K-1)

Dev. Stage - O1

Oxid. Stage A - O1

Oxid. Stage B - O1

Rem. Stage - O1

Pine bark

Chapter 3

152

Table 3.5.-Estimated kinetic parameters for the combustion of the main components of lignocellulosic biomass (cellulose, xylan and lignin)

BiomassSample HeatingRate (ºC/min)

E (kJ/mol)

A (1/min)

r 2 E (kJ/mol)

A (1/min)

r 2 E (kJ/mol)

A (1/min)

r 2 E (KJ/mol)

A (1/min)

r 2

Stages Dev. Stage Oxid. Stage - - Mechanism O1 O1 - -

Cellulose 10 164 3.5·1013 0.9978 187 1.21012 0.9932 - - - - - -

20 166 1.3·1014 0.9937 193 2.1·1012 0.9930 - - - - - -

40 171 1.3·1014 0.9910 159 3.4·109 0.9905 - - - - - -

80 173 8.3·1013 0.9908 181 8.6·1013 0.9906 - - - - - -

Stages Dev. Stage A Dev. Stage B Oxid. Stage Imp.Stage Mechanism N1 O2 O1 O1/ O2

*

Xylan 10 128 4.8·1012 0.9975 107 7.6·109 0.9954 146 4.3·109 0.9943 188 1.7·1010 0.9911

20 125 1.9·1012 0.9947 106 3.6·109 0.9945 142 4.3·109 0.9946 252 1.1·1015 0.9761

40 131 8.8·1012 0.9931 105 2.3·109 0.9935 129 4.3·107 0.9951 - - -

80 131 3.7·1012 0.9925 104 1.0·109 0.9954 80 1.1·104 0.9900 - - -

Stages Dev. Stage Oxid. Stage A Oxid. Stage B Rem. Stage Mechanism D3 O1 O1 O1

Lignin 10 96 6.6·109 0.9941 55 1.8·105 0.9931 60 8.2·105 0.9918 - - -

20 84 5.6·106 0.9930 89 2.3·106 0.9965 119 1.2·107 0.9946 - - -

40 83 5.4·106 0.9901 96 8.6·106 0.9976 74 3.6·103 0.9906 - - -

80 70 2.4·105 0.9925 88 1.4·106 0.9958 52 6.1·101 0.9913 599 6.9·1026 0.9932

Chapter 3

153

Table 3.6.- Estimated kinetic parameters for the combustion of different types of lignocellulosic biomass samples (fir wood, eucalyptus

wood and pine bark)

BiomassSample HeatingRate (ºC/min)

E (kJ/mol)

A (1/min)

r 2 E (kJ/mol)

A (1/min)

r 2 E (kJ/mol)

A (1/min)

r 2 E (kJ/mol)

A (1/min)

r 2

Stages Dev. Stage Oxid. Stage Rem. Stage - Mechanism O1 O1 O1 -

Fir wood 10 75 1.3·106 0.9933 126 3.4·108 0.9932 377 9.6·1021 0.9942 - - -

20 82 6.4·106 0.9938 124 2.1·108 0.9908 393 2.6·1022 0.9948 - - -

40 89 4.6·107 0.9905 167 1.2·1011 0.9926 336 3.5·1018 0.9902 - - -

80 88 2.4·107 0.9959 107 1.6·106 0.9911 484 5.1·1022 0.9991 - - -

Stages Dev. Stage Oxid.Stage Rem. Stage - Mechanism O1 O1 O1 -

Eucalyptus wood

10 87 2.1·107 0.9933 179 5.9·1012 0.9930 335 1.6·1019 0.9930 - - -

20 85 1.7·107 0.9918 165 2.7·1011 0.9913 335 6.1·1018 0.9913 - - -

40 90 6.9·107 0.9926 135 1.1·109 0.9947 330 9.6·1017 0.9904 - - -

80 92 6.9·107 0.9921 84 5.3·104 0.9934 314 3.3·1016 0.9923 - - -

Stages Dev. Stage Oxid. Stage A Oxid. Stage B Rem. Stage Mechanism O1 O1 O1 O1

Pine bark 10 95 1.3·108 0.9943 152 1.5·1011 0.9954 188 4.2·1012 0.9916 - - -

20 97 2.8·108 0.9935 177 1.9·1013 0.9908 139 8.1·108 0.9916 610 1.5·1035 0.9930

40 97 4.1·108 0.9953 188 2.2·1014 0.9903 126 5.6·107 0.9910 504 2.5·1028 0.9924

80 103 6.7·108 0.9907 140 3.4·1010 0.9902 107 7.6·105 0.9945 274 5.4·1014 0.9906

Chapter 3

154

(Ni) and diffusion (Di) achieved a regression coefficient above 0.98. These results

agreed well with those reported in literature [2; 7;19]. The general disadvantage of

dynamic thermal analysis is that in many cases more than one function g(α) fits the

experimental results. Consequently, selection of the responsible mechanism and

estimation of the real kinetic parameters can be difficult as previously reported [7].

Furthermore, separate stage models fail to predict the transition region between two

mass losses processes and do not consider chemical processes.

Regarding the Dev. stage, the cellulose combustion followed a first reaction order

mechanism. That of lignin followed a D3 mechanism whereasxylanfollowed a N1and

O2onesforDev. stage Aand Dev. stage B, respectively. The oxidation stage followed

aO1 functionfor cellulose and xylanwhereas the lignin combustion fitted betterinto a

O3 and O1for Oxid. stage Aand Oxid. stage B, correspondingly. The Imp.stage for the

xylanoxidation was only found at 10 and 20 ºC/min. Different mechanism were found

to be meaningful for each heating rate. O1model showed the best r2at 10

ºC/minwhereas O2 model was the one that better described the Imp stage at 20 ºC/min.

Finally, the Rem. stage forlignin oxidation at 80 ºC/min followed an O1mechanism.

Anyway,model function of first order (O1) yielded the best correlation coefficient

for lignocellulosic biomass combustion process (Table 3.6). These results agreed well

withthose reported elsewhere[19] and indicate that the composition of biomass do not

influence the overall reaction mechanism of lignocellulosic biomass oxidation.

Up to now, the effect of the heating rate on biomass thermal decomposition

kinetics is still unresolved [17]. Different authors have proposed that the heating rate

has minimal impact on the frequency factor, which is mainly related to the structure of

the material [7; 17]. Anyway, the activation energy is the main characteristic attributed

to the reactivity of a biomass sample [2; 7]. Figure 3.5 shows the estimated activation

energies at different heating rates for the different stages of the combustion process

ofbiomass samples.Concerning the kinetic parameters obtained for the main

components oflignocellulosic biomass (Figure 3.5a and Table 3.5), several differences

can be observed. Activation energies of biomass main components can beranked as

Chapter 3

155

follows: E(cellulose)>E(xylan)>E(lignin). This order determined that the

decomposition of cellulose is the rate-determining step of the biomass combustion

process[17]. On the other hand, the activation energy valueswere hardly affected by

the heating rate.

Figure 3.4.-Comparison of activation energy for the combustion process of: a) main

components of lignocellulosic biomass (cellulose, lignin and xylan) and b) lignocellulosic

biomass (eucalyptus wood, fir wood and pine bark) at 10, 20, 40 and 80 ºC/min.

The study of the Oxid.stage showed a similar trend as that commented for the

previous stage. Activation energies for cellulose and xylan were higher than that for

lignin. However, the higher the heating rate, the slightly lower the activation energies

were.Finally, the activation energies for the Rem. stage werehigher than those

obtained for the previous ones. This fact can be attributed to the high energy required

to decompose inorganic matter [26]. The effect of the heating rate on the kinetic

parameters for lignocellulosic biomass samples was similar to that commented for

0 20 40 60 80 10050

100

150

200

250

0 20 40 60 80 1000

50

100

150

200

250

300

0 20 40 60 80 10050

75

100

125

0 20 40 60 80 10050

100

150

200

250

0 20 40 60 80 100200

300

400

500

600

700

Dev. Stage

EA

(Kj/m

ol)

Heating Rate (ºC/min)

Xylan A Xylan B Cellulose Lignin A

Lignin B

Oxid. Stage


EA

(Kj/m

ol)


Dev. Stage Fir wood Eucalyptus Wood Pine bark A Pine bark B


Oxid. Stage


Rem. Stage

Chapter 3

156

their main components (Figure 3.4.b and Table 3.6). Activation energies for the Dev.

stage did not differ for different heating rates [1]. However, for the Oxid.stagethe

higher the heating rate, the lower the activation energy was observed. Shen et

al.[1]related this finding with the occurrence of gradient temperatures within the

particle, which increased at higher heating rates. As above commented and expected,

the value of activation energies for the Rem. stage were the highest one.

In order to corroborate the kinetic analysis, the reconstruction of the weight loss

curves was performed. Considering n separate reactions, the kinetic rates of thermal

decomposition of a material can be easily derived from Eq. (4) as follows:

dα. dT� = A. β� e 34 56� f.(α.) (7)

whereαi, Ai, Ei and fi(αi) are the degree of conversion, the pre-exponential factor,

activation energy and model functions obtained for each stage of the combustion

process, respectively.

A VBA-Excel application was developed to solve this model based on the Runge-

Kutta-Fehlberg method for the evaluation of the set of ordinary differential equations.

Figure SS1 shows the experimental data (solid line) compared to the predicted one

(dotted line) for the combustion process of biomass samples at a heating rate of 10

ºC/min, obtained by substituting the calculated activation energy and pre-exponential

factor for each stage into Eq. (7). It can be observed that the proposed model

adequately reproduces the experimental values, obtaining a low error for all cases.

Finally, in order to ensure the reliability of the proposed models, the discrimination of

kinetic parameters was done applying the F-test and the t-test at the 95% confidence

level [23]. The resulting parameters obtained from the linear regression are

summarized in Table 3.7 and 3.8. In terms of statistical results, F-test considered the

regression to be suitable in all cases since the corresponding values to the Fc/Ftest ratio

were larger than one. The t-test was also used for evaluating each parameter in the

Chapter 3

157

model. The values of tc/t-test ratio were also larger than one, showing the statistical

significance of the proposed models and their corresponding parameters.

Table 3.7.- Estimated statistical parameters for the combustion of biomass main components

Sample Heating rate (ºC/ min) Stage Step tc (EA) tc (k0) t-Test Fc (·10-3) F-test

Cellulose

10 Dev. 1 48.8 2.47 1.96 44.8 3 Oxid. 2 74.3 2.61 1.96 362 3

20 Dev. 1 96.4 2.08 1.96 204 3 Oxid. 2 64.4 1.97 1.96 325 3

40 Dev. 1 64.2 4.54 1.96 10.3 3 Oxid. 2 69.3 273 1.96 337 3

80 Dev. 1 629 13.6 1,98 1314 3 Oxid. 2 82.3 2.60 1,99 311 3

Xylan

10

Dev. A 1 106 4.38 1.96 1118 3 Dev. B 2 35.6 2.12 1.96 154 3 Oxid. 3 112 4.20 1.96 578 3 Imp. 4 61.1 2.11 1.96 405 3

20

Dev. A 1 115 2.43 1.97 409 3 Dev. B 2 34.7 2.21 1.96 193 3 Oxid. 3 192 8.04 1.96 1879 3 Imp. 4 54.4 2.27 1.96 313 3

40 Dev. A 1 108 2.22 1.97 361 3 Dev. B 2 35.7 2.03 1.96 195 3 Oxid. 3 120 5.58 1.96 731 3

80 Dev. A 1 63.3 3.40 1.99 181 3 Dev. B 2 32.8 2.30 1.98 314 3 Oxid. 3 68.2 4.53 1.96 349 3

Lignin

10 Dev. 1 119 8.25 1.96 857 3

Oxid. A 2 70.5 3.89 1.96 168 3 Oxid. B 3 65.3 3.39 1.96 310 3

20 Dev. 1 77.3 5.82 1.96 458 3

Oxid. A 2 23.9 2.52 1.96 191 3 Oxid. B 3 91.1 4.19 1.96 419 3

40 Dev. 1 60.7 4.53 1.96 268 3

Oxid. A 2 93.3 4.43 1.96 612 3 Oxid. B 3 107 8.37 1.96 757 3

80

Dev. 1 39.1 2.87 1.96 145 3 Oxid. A 2 104 5.83 1.97 978 3 Oxid. B 3 80.8 8.28 1.96 618 3 Rem. 4 52.4 3.58 1.96 435 3

Chapter 3

158

Table 3.8.- Estimated statistical parameters for the combustion of lignocellulosic biomass

Gas evolved analysis

The main products derived from the combustion of lignocellulosic biomass and its

main components were evaluated by TGA-MS analysis. On the basis of a preliminary

scan, a list of key molecular ions was compiled and gathered inTable 3.9.

Sample Heating rate (ºC/ min) Stage Step tc (EA) tc (k0) t-Test Fc (·10-3) F-test

Eucalyptus wood

10 Dev. 1 131 6.16 1.96 537 3 Oxid. 2 53.1 3.78 1.96 24.5 3 Rem. 3 105 2.19 1.96 673 3

20 Dev. 1 108 5.27 1.96 354 3 Oxid. 2 98.8 2.13 1.96 669 3 Rem. 3 65.6 2.02 1.96 261 3

40 Dev. 1 112 5.33 1.96 370 3 Oxid. 2 108 2.45 1.96 892 3 Rem. 3 75.2 2.96 1.96 474 3

80 Dev. 1 79.2 4.01 1.96 185 3 Oxid. 2 70.1 4.42 1.96 605 3 Rem. 3 99.1 2.44 1.97 659 3

Firwood

Dev. 1 56.7 1.12 1.96 79.1 3 10 Oxid. 2 96.7 4.08 1.96 567 3 Rem. 3 124 2.42 1.98 523 3 Dev. 1 146 3.29 1.96 1826 3

20 Oxid. 2 82.2 3.43 1.96 435 3 Rem. 3 125 2.26 1.97 1147 3 Dev. 1 146 3.29 1.96 1825 3

40 Oxid. 2 69.6 2.74 1.96 425 3 Rem. 3 91.9 2.37 1.96 543 3 Dev. 1 99.8 4.96 1.96 396 3

80 Oxid. 2 74.5 5.15 1.96 685 3 Rem. 3 57.3 2.04 1.96 433 3

Pine bark

10 Dev. 1 150 6.54 1.96 716 3

Oxid. A 2 71.8 2.46 1.96 483 3 Oxid. B 3 79.5 2.44 1.96 542 3

20

Dev. 1 160 7.05 1.96 825 3 Oxid. A 2 72.5 2.17 1.96 486 3 Oxid. B 3 66.9 2.62 1.96 366 3 Rem. 4 78.6 3.18 1.96 567 3

40

Dev. 1 183 8.14 1.96 1030 3 Oxid. A 2 235 10.7 1.96 1832 3 Oxid. B 3 72.1 3.18 1.96 412 3 Rem. 4 89.6 3.38 1.98 1195 3

80

Dev. 1 126 5.58 1.96 471 3 Oxid. A 2 63.5 2.44 1.98 467 3 Oxid. B 3 147 6.70 1.96 716 3 Rem. 4 13.2 3.49 2.02 110 3

Chapter 3

159

Table 3.9.-Molecular ions and probable parent molecules detected in the combustion of lignocellulosic biomass and its main components.

(m/z)

Key molecular ions/Ion fragment Probable parent molecule Cellulose Lignin Xylan Fir wood Eucalyptus wood Pine bark

2 H2+ H2 X X X X X X

15 CH3+ CH4 X X X X X X

16 O+, CH4+ CH4 - X X X X X

18 H2O+ H2O X X X X X X

26 CN+, C2H2+ C2H2 X X X X X X

27 HCN+, C2H3+ HCN (nitriles) X X X X X X

28 C2H4+, CO+ CO X X X X X X

29 C2H5+ C2H5 (Ethylderivates) X X X X X X

30 C2H6+, CH2NH2

+ CH4N (Primary amines) X X X X X X

44 CO2+ CO2 X X X X X X

45 C2H5O+, C2H7N

+, CHS+ C2H5O (hydroxyderivates) X X X X X X

46 NO2+, C2H5OH+ NO2 X X X X X X

47 CH3S+, CCl+, C2H5OH+ CH3Cl X X - X X X

48 CH3SH+, CHCl+, SO+ SO - X - X X X

50 C4H2+, CH3Cl+, CF2

+ CH3S X X - X X X

51 C4H3+, CHF2

+ C4H3(aromatics) X - - X X X

52 C4H4+ C4H4 (aromatics) X - - X X X

53 C4H5+ C4H5 (aromatics) X - - X X X

54 C4H6+, C2H4CN+ C4H6 (aromatics) X - - X X X

55 C4H7+, C3H3O

+ C4H7 (aromatics) X - X X X X

56 C3H6N+, C4H8

+ C4H8 (alquenes) X - X X X X

57 C4H9+, C3H5O

+, C3H2F+ C3H5O (cyclopentanol) X - X X X X

58 C3H8N+ C3H8N (amines) X - X X X X

60 COS+ COS X X X X X X

64 SO2+ SO2 - X X - - -

68 C5H8+, C4H4O

+, C3H6CN+ C4H6O(cyclohexenones) X - - X X X

70 C5H10+, C4H6O

+, C4H8N+ C4H6O (cycloalkanones) X - - - - -

72 C4H8O+, C4H10N

+, C6+ C4H8O(alkanones) X - - - - -

84 C5H10N+ C5H10N (pyrolidines) X - - - - -

95 C5H3O2+ C5H3O2(furycarbonil-derivates) X - - - - -

96 C7H12+ C7H12 (alicyclics) X - - - - -

Chapter 3

160

Mass spectrometry analysis for the oxidation of xylan, cellulose and lignin are

shown in Figure 3.5. MS curves were move up and down in order to obtain a

clearevolved gas profile. Magnifying picture of the curves was placed in the upper

right corner. MS spectra of biomass main components could be divided into different

stages related to their degradations steps studied in the TGA/DTG curves and

described in previous sections.Table 3.10 summarizes the most representative MS ions

detected, their integrated peak intensities in the whole temperature range and the

temperature where the maximum emission peak was found. Cellulose sample

combustion showed the major emission peak for all productsat the devolatilization

stage at temperatures around 350 ºC which is in good agreement with its maximum

DTG peak.In this stage, the degradation of glycosyl units in cellulose produced H2O,

CO and CO2 leading to the formation of char residue.Xie et al.[29]stated that during

the devolatilization of cellulose the complete decomposition of

glycosidicstructuresproceeded. Furthermore, rapid depolymerization of cellulose

turned into the breakdown of the molecule,producing a variety of low molecular

weight products[5]. Inthe char combustion stage only H2O, CO, CO2, NO2,

C2H5O+(hydroxyderivates) and C2H5were detected at 510 ºC. CO and CO2 evolution

from cellulose sample oxidation is believed to be a consequence of the loss of

carbonyl and carboxyl groups previously formed by the oxidation of hydroxyl groups

[5]. This fact can also be corroborated by the presence of C2H5O+ in this temperature

range.In a similar way, xylanevolution was resolved in two stages. This way, volatiles

were released between 200 and 450 ºC whereas only CO2, CO, CH3+,C2H5O

+, HCN

and NO2 were found at higher temperatures. In this case, the maximum evolution rate

for most compounds was found at temperatures between 260 and 270 ºC. The

formation of two peaks for CO, CO2 and H2O in this stage showed the good

correlation with its DTG profile. However, a second group of compounds (H2, CH3,

C2H2, and HCN) had their maximum emission peak at 466 ºC, coincidental to the

shoulder in the CO2 curve. This second peak could be due to char elimination and

rearrangement reactions. Finally, CO, CO2, C2H5O+and NO2decomposed at higher

temperatures. Lignin sampleMS spectra showed a more diversified profile, volatile

Chapter 3

161

compounds evolved within the whole temperature range as expected from the TG-

DTG analysis, which could beattributed to the presence of aromatic compoundsin the

raw material thatremained after their oxidation in air at 240 ºC [29], widening the

temperature range for volatiles evolution.As abovementioned, lignin sample was the

first biomass component to start decomposing. Most volatiles started evolving in the

130-200 ºC temperature range, finding the main emission peak for H2O, CH3Cl,

CH3S+,SO and SO2at 340 ºC.The low temperatures at which oxygenated compounds

evolved pointed out that aliphatic HOgroups were easily removed from lignin[30]. In

this stage, CO, CO2 and H2O were released by C-C bond scission [31]. Furthermore,

the decomposition ofsulphoxide and sulphone groups facilitated lignin

depolymerization[30]. C2H2, CH4N and CH3+ emission peaks occurred at 392, 397 and

427 ºC, respectively.The evolution of these compounds was associated to dehydration

and demethylation reactions [30].H2O, C2H2, HCN, COS, CO and CH4maximum

peaks took place between 483 and 501 ºC. H2 was later released at 529 ºC being

attributed to the hydrogen splitting from aromatic rings [30]. Finally, CO2, NO2 and

C2H5O+peak was found between 553 and 575ºC where the devolatilization of char

took place. A last CO2 peak was found at high temperatures related to carbonates

decomposition. This peak was not observed in cellulose and xylansamples MS profiles

which is due to their low inorganic content. This fact confirmed that the peak observed

in the DTG profiles at high temperatures can be attributed to the inorganic matter

present in the formed ash. Finally, it can be observed that cellulose sample MS spectra

showed the highest intensity peak for all the observed products in the devolatilization

stage being in agreement with its high DTG peak. Furthermore, the low CO2 peak in

the oxidation stage compared to that of lignin and xylan samples was due to the low

amount of fixed carbon. On the contrary, lignin and xylan samples showed their

highest CO and CO2 peaks in the oxidation stage.

Chapter 3

162

Figure 3.5.-Mass spectra of the combustion of the main components of lignocellulosic biomass (cellulose, xylan and lignin) at 40 ºC/min.

200 400 600 800 1000

CH3

+

NO2

H2O

Temperature (ºC)

CO2

C2H

5

CH4N

C2H

5O+

HCN

C2H

2

CO

C4H

7

200 300 400 500 600

Inte

nsity

(a.

u)

Temperature (ºC)

H2

C3H

8N

CH4

COS

C4H

8

C3H

5O

200 400 600 800 1000

SO2

CH3

+

NO2

H2O

Temperature (ºC)

CO2

C2H

5

CH4N

C2H

5O+

HCN

CO

200 300 400 500 600

Inte

nsity

(a.

u)

Temperature (ºC)

H2

C2H

2

CH4

COS

CH3S+

C4H

8O+

200 400 600 800 1000

C3H

5OC

4H

6O+COS

CH3Cl

CH3

+C

4H

8

C4H

6

C3H

8NNO

2

H2O

Inte

nsity

(a.

u.)

Temperature (ºC)

CO2

C2H

5

CH4N

C2H

5O+

HCN C2H

2

COC4H

7

C4H

5

+ C4H

4O

200 300 400 500 600

Inte

nsity

(a.

u)

Temperature (ºC)

H2

C5H

12

+

C5H

3O

2

+

C5H

10N+

CH3S+

C4H

8O+

Chapter 3

163

Table 3.10.-Maximum peak temperatures and integrated peak areas for biomass main components and lignocellulosic biomass

Cellulose Lignin Xylan Firwood Eucalyptuswood Pine bark Compound Tp

* (ºC) Int ** (min/mg)

Tp* (ºC) Int*

(min/mg) Tp

* (ºC) Int* (min/mg)

Tp* (ºC) Int*

(min/mg) Tp

* (ºC) Int* (min/mg)

Tp* (ºC) Int*

(min/mg) H2 356 0.2 529 2.0 468 2.8 363 0.3 305 0.9 351 0.2

CH3+ 356 11.7 427 26.6 468 13.8 342 5.7 313 6.7 313 2.8

CH4 - - 501 0.6 305 0.3 437 40.6 431 55.5 433 48.2 H2O 353 696 488 586 262 552 337 592 323 580 321 878 C2H2 356 2.3 483 1.8 466 2.6 342 1.6 323 1.5 316 1.1 HCN 356 1.6 483 2.1 466 3.7 340 2.5 324 2.1 313 1.5 CO 354 310 496 281 265 176 422 287 413 230.3 415 657

C2H5 351 27.7 392 18.4 262 18.5 343 19.2 313 21.6 311 20.6 CH4N 351 16.5 397 11.6 257 7.8 340 13.8 311 14.7 313 12.6 CO2 354 403.5 575 652 530 445 459 768 416 589 434 953

C2H5O+ 354 4.7 560 2.1 530 6.2 461 9.3 416 7.8 431 10.4 NO2 357 1.8 553 0.8 268 2.6 461 3.4 414 2.6 434 2.2

CH3S+ 354 0.01 334 0.2 - - 489 0.1 291 0.1 306 0.1 SO - - 362 0.7 - - 315 0.03 250 0.05 250 0.03

CH3Cl 352 0.03 344 0.4 - - 323 0.3 311 0.9 311 0.1 C4H3

+ 360 0.03 - - - - 356 0.1 319 0.1 370 0.01 C4H4

+ 360 0.1 - - - - 318 0.2 309 0.3 309 0.08 C4H5

+ 355 0.2 - - - - 333 0.1 324 0.1 306 0.04 C4H6 355 0.3 - - - - 349 0.1 319 0.2 335 0.06 C4H7 355 0.3 - - 271 0.2 349 0.1 312 0.3 314 0.07 C4H8 355 0.2 - - 268 0.2 346 0.2 317 0.2 317 0.08

C3H5O 352 0.03 - - 263 0.1 336 0.1 330 0.1 350 0.03 C3H8N 357 0.4 - - 296 0.3 333 0.3 322 0.3 307 0.1 COS 360 0.04 482 0.2 258 0.1 339 0.1 304 0.3 307 0.08 SO2 - - 334 1.1 - - - - - - - 0.03

C4H4O 352 0.3 - - - - 346 0.1 325 0.1 335 0.03 C4H6O+ 358 0.01 - - - - - - - - - - C4H8O+ 355 0.02 - - - - - - - - - - C5H10N+ 361 0.02 - - - - - - - - - - C5H3O2

+ 358 0.01 - - - - - - - - - - C7H12

+ 361 0.02 - - - - - - - - - -

Chapter 3

164

Chapter 3

165

Figure 6.-Mass spectra of the combustion of lignocellulosic biomass (eucalyptus wood, fir wood and pine bark) at 40 ºC/min.

200 400 600 800 1000

COS

CH3+

C3H

8N

NO2

H2O

Temperature (ºC)

CO2

C2H

5

CH4N

C2H

5O+

C2H

2

CO

C4H

7

HCN

200 300 400 500 600

C4H

8

C4H

4

+

Inte

nsity

(a.

u)

Temperature (ºC)

CH4

H2

CH3Cl

C4H

6

SO2

C4H

4O

C4H

5

+

SO C4H

5

+

200 400 600 800 1000

COS

CH3

+CH3Cl

C4H

8

C3H

8N

NO2

H2O

Inte

nsity

(a.

u.)

Temperature (ºC)

CO2

C2H

5

CH4N

C2H

5O+

C2H

2

CO

C4H

7

C4H

4

+

HCN

200 300 400 500 600

Inte

nsity

(a.

u)Temperature (ºC)

H2

CH3S+CH

4

C4H

6

C3H

5O

C4H

5

+

C4H

3

+

SO

C4H

4O

200 400 600 800 1000

COS

CH3

+

CH3Cl

C4H

8

C4H

4

C3H

8N

NO2

H2O

Temperature (ºC)

CO2

C2H

5

CH4N

C2H

5O+

C2H

2

CO

C4H

7

C4H

4

+

HCN

200 300 400 500 600

Inte

nsity

(a.

u)

Temperature (ºC)

H2

CH3S+

SO2

C3H

5O

C4H

5

+

C4H

3

+

SO

C4H

4O

CH4

Chapter 3

166

Figure 3.6 shows the MS spectra for firwood, eucalyptus wood and pine bark.The

gas products distributionwas pretty similar andmay be divided into three stages. As

commented above, most compounds evolved during the devolatilization stage. Sulfur

compounds were the first to be detected as SO and CH3S+. An intermediate emission

peak was found, where CO and CH4 evolved. H2 was also emitted between these two

stages. Finally, CO, CO2, C2H5O+, and NO2 were the main peaks detected during the

char oxidation stage. The same pattern was followed by all the samples. However, the

emission peaks temperatures changed. As expected from biomass samples DTG

profiles, peaks for eucalyptus wood and pine bark took place at lower temperatures

than those for fir wood. This fact could be due to compositional differences among

them (volatile matter, cellulose and hemicellulose content).

H2O, CO and CO2 were the main products obtained during biomass combustion

(Table 3.10). CO and CO2 evolved over the whole temperature range with a higher

proportion of CO2.CO is assumed to be formed by the creation of molecular oxide

complexes that further rearrange turning into the evolution of CO [32].However,

CO2presented a more complex formation process [32], as its production can be

catalyzed or inhibited by the formation of carbon intermediates[33]. Pine bark released

the highest amount of CO2 which is attributed to its high lignin content. However, the

CO evolution from combustion of lignocellulosic biomass cannot be directly

correlated to that for biomass components since CO signal has a great contribution of

the CO2 one. Furthermore, there is also a minor contribution of ethylene to the ion

corresponding to m/z=28[34].H2O was released in three steps. Firstly, the water

released in the low temperature range was associated to the dehydration of the sample.

Then, the higher peak for water was formed in the main devolatilization stage, being

associated to the evolved aliphatic OH groups [15; 33]. Finally, a shoulder appeared in

the MS spectra, which was related to the water formed by the oxidation of H2 and

calcium carbonate decomposition [33]. The higher amount of H2O evolved from pine

bark sample combustion was mainly due to its higher initial moisture content.Light

hydrocarbons and especially CH4 and C2H5 were also predominant. CH4had two main

origins related to devolatilizationand charring processes [6].It was also interesting to

Chapter 3

167

note the release of nitrogen compounds (N compounds) in form of HCN, NO2 and

CH4N. However, it has to be careful with relative amount of amines as the associated

ion (m/z=30) can be related to NO and C2H6compounds. Thus, N compounds emission

should be related to NO2 rather than NO and CH4N. Two emission zones were found

for N compounds. One stage related to the release of amines and NO2, which is

attributed to the decomposition of proteins [12].In the second one,NO2 was only

detected and attributed to the oxidation of the retained N in the

char[35].Furthermore,Darvel et al.[12]reported a possible explanation for HCN

formation (C(H)1 + C(N)1 → HCN + Cfas, where C(H)1 and C(N)1 arelocalized surface

species and Cfas is a "free active site"). Finally, the rest of the products were found in

low content. Among them, special attention to the release of sulfur and chloride

compounds should be paid since they behave as hazard pollutants. Sulfur compounds

were released as SO, CH3S and COS (SO2 signal was hardly detected during biomass

combustion, which can be explained by the easy fragmentation of this ion within the

mass spectrometer) whereas thechloride fraction was detectedin form of CH3Cl.

3.3.2. Combustion of marine biomass

Thermogravimetric Analysis (TGA)

Figure 3.7 shows TGA-DTG profiles for the microalgae Scenedesmusalmeriensis

(SC), Nannochloropsisgaditana(NG) and Chlorella vulgaris (CV) at a heating rate of

40 ºC/min. Table 3.11 shows the most relevant combustion characteristics. Ignition,

peak and burnout temperatures are the most characteristic parameters when evaluating

the combustion performance of a material [36; 37]. Peak temperature (Tp) refers to the

temperature where maximum loss weight rate (dw/dT)maxis reached. Peak temperature

and its corresponding rate is a measure of combustibility and reactivity, respectively.

Thus, the lower the Tp, the easier the ignition of a material is. The ignition temperature

(Ti) is the temperature at which a sudden decrease in weight loss on the DTG curve is

observed. Ti was calculated as the intersection between the tangent line to the point

which decomposition started and the tangent line to the maximum weightloss rate. The

burnout temperature (Tb) is the temperature where the process is finished.

Chapter 3

168

Figure 3.7.-Thermogravimetric curves for the combustion process of Scenedesmusalmeriensis (SC), Nannochloropsisgaditana (NG) and

Chlorella vulgaris (CV) microalgae.

150 300 450 600 750 900 10500

20

40

60

80

100

150 300 450 600 750 900 10500

20

40

60

80

100

150 300 450 600 750 900 10500

20

40

60

80

100

Sub

-ste

p I+

II

Fin

al v

olat

iliza

tion

Sub

-ste

p III

Sub

-ste

p III

Sub

-ste

p II

Sub

-ste

p I

Sub

-ste

p III

Sub

-ste

p II

Sub

-ste

p I

STAGE ISTAGE II

Wei

ght (

wt.%

)

Temperature (ºC)

SC STAGE I

0.0

0.1

0.2

0.3

0.4

0.5

Wei

ght l

oss

rate

(w

t.%/º

C)

STAGE I STAGE II

Wei

ght (

wt.%

)

Temperature (ºC)

0.0

0.1

0.2

0.3

0.4

NG

Wei

ght l

oss

rate

(w

t.%/º

C)

Sub

-ste

p III

STAGE II

Wei

ght (

wt.%

)

Temperature (ºC)

CV

0.0

0.1

0.2

0.3

0.4

0.5

Wei

ght l

oss

rate

(w

t.%/º

C)

Chapter 3

169

Table 3.11-TGA-DTG characteristics for the combustion process of Chlorella vulgaris (CV),

Scenedesmusalmeriensis (SC) and NannochloropsisGaditana (NG) microalgae

Biomass samples

CV SC NG

1st

peak

2nd

peak

3rd

peak

1st

peak

2nd

peak

3rd

peak

4th

peak

1st

peak

2nd

peak

3rd

peak

4th

peak

Tdo (ºC)* 172 125 142

T i (ºC) * 265 276 237

Tb (ºC) * 725 696 716

Tpo (ºC) * 172 359 514 125 205 402 479 142 396 478 826

Tpf (ºC) * 359 514 725 205 402 479 696 396 478 716 998

Tp (ºC) * 304 378 607 175 311 453 555 284 430 573 955

(dw/dT)max*

(dwt.%/ºC)

0.44 0.24 0.19 0.07 0.48 0.12 0.22 0.40 0.11 0.18 0.05

Residue (%) 16.1 19 5.9

CCF ( · 107)* 3.0 3.4 2.6

*Sh: Shoulder; Tdo: Initial decomposition temperature; Ti: Ignition temperature; Tb: Burnout temperature; Tpo: Initial peak

temperature; Tpf: Final peak temperature; Tp: Peak temperature; (dw/dT)max: Maximum weight loss rate; Combustion

characteristic factor

The main decomposition stages are represented in Figure 3.7 by a solid line

whereas the minor stages are represented as sub-steps and plotted by a dotted line. The

thermal decomposition of microalgae under air atmosphere is usually described by

two stages [38; 39]. These stages were identified in the DTG profile by the formation

of pronounced peaks. The first stage comprised the devolatilization of the samples and

it was extended until temperatures around 500 ºC. This stage was characterized by a

major loss weight corresponding to the release of organic compounds leading to the

formation of char [38]. The second stage, which took place at temperatures above 500

ºC, consisted of the combustion of the formed char and presented a variable Tb

depending on the reactivity and the amount of char formed. However, this

classification cannot result conclusive since different peaks can be observed during the

first stage corresponding to the decomposition of microalgae main components.

Chapter 3

170

The first decomposition stage can be subdivided in three more sub-steps for

samples SC and NG and in two more sub-steps for sample CV. The first sub-step was

represented by a peak for sample SC and as a shoulder for sample NG, which is

related to intrinsic lipid decomposition, such as aldehydes and ketones [40]. In

addition, the decomposition of carbohydrates started in this temperature range (170-

180 ºC) [41]. This sub-step was not found for sample CV and might be due to its

lower lipid and carbohydrate content. The second sub-step was detected by a peak at

284, 304 and 311 ºC for samples NG, CV and SC, respectively, and was associated to

carbohydrates and proteins decomposition [42]. The maximum weight loss rate was

observed in this sub-step for all samples. Sample SC had the highest (dw/dT)max (0.48

wt.%) followed by samples CV (0.44wt.%) and NG (0.40wt.%). Carbohydrates,

proteins and ash content could influence this order. In this sense, the higher their

content in the raw material, the higher the (dw/dT)maxwas. Regarding the ash content,

this was much higher for samples SC and CV than for sample NG. In this sense, alkali

metals present in the ash could catalyze the combustion process increasing the

volatiles yield [43]. Finally, a third sub-step, common to all samples, was observed

close to the char oxidation stage. This peak appeared at lower temperatures for sample

CV (378 ºC) than for samples NG (430 ºC) and SC (453 ºC). In this stage, the final

decomposition of lipids took place and it was mainly associated to the break-down of

hydrocarbon chains of fatty acids [38; 40; 41]. In sample CV, this peak appeared at

lower temperature and with a higher weight loss rate. This fact can be related to the

higher protein content in this sample. In this regard, Kebelmann et al. [40] found a

shoulder for the thermal decomposition of microalgae proteins close to the main

decomposition stage.

The second decomposition stage took place between 478 and 725 ºC. For sample

SC, this stage took place at lower temperatures and with a slightly higher weight loss

rate, if compared to that of samples CV and NG. This fact pointed out that SC

devolatilization led to the formation of a bigger amount of char. This is in agreement

with Ross et al. [41] who reported that high levels of K in the sample promoted the

formation of char. Additionally, a last decomposition step was observed for sample

Chapter 3

171

NG between 826 and 998 ºC in a similar way as reported by Wang et al. for the

combustion of seaweeds [26]. This step is mainly related to volatile metal loss and

carbonate decomposition.

Concerning the general burning profile of the microalgae samples, it can be

observed that sample SC was the first to decompose (125 ºC) whereas samples NG

and CV decomposed at 142 and 172 ºC, respectively. On the other hand, sample SC

was the most difficult sample to ignite. Finally, the sample CV showed the highest Tb

(725 ºC) compared with samples NG (716 ºC) and SC (696 ºC).

Samples SC and CV left a high amount of ash (19 and 16 wt.%, respectively)

compared to sample NG (5.9 wt.%). This fact restricts the use of sample SC and CV

for direct combustion and gasification due to the catalyst/inhibiting effect of the ash.

Thus, a pre-treatment based on water, acid or alkali washing to reduce the influence of

minerals may be needed [44].

The combustion characteristic factor (CCF) can be used to preliminary assess the

microalgae combustion performance [45]. This factor is based on the energy required

to burn a material in terms of low Ti and Tb values and high (dw/dt)max and is

expressed as follows:

889 = (:;:� )<=> ∙ (:;:� )<?=@ �A$ ∙ �B

� (7)

where (dw/dt)max is the maximum burning velocity (%/min); (dw/dt)mean is the average

burning velocity (%/min); Ti is the ignition temperature (K) and Tb is the burnout

temperature (K).

CCF values for all samples are shown in Table 3.11. In all cases, these values were

bigger than 2 indicating the good general burning performance [37]. Sample SC

required less energy than the other samples to perform the combustion. However,

these data must be used only as a reference since they do not give any information

about the heat released during the combustion process.

Chapter 3

172

Differential scanning calorimetry (DSC)

In order to complete the information obtained by TG analyses, the marine biomass

(samples CV, SC and NG) was also investigated by the DSC technique. Experimental

DSC curves are presented in Figure 3.8. DSC main temperatures and heat of

combustion (Hcomb) are included in Table 3.12 DSC analysis of lignocellulosic

biomass has been studied by different authors [24; 37;46]. However, at the best of our

knowledge, the DSC analysis of microalgae combustion has not been explored yet.

Figure 3.8.-DSC curves for the combustion process of Nannochloropsisgaditana (NG),

Scenedesmusalmeriensis (SC) and Chlorella vulgaris (CV) microalgae.

Table 3.11.-DSCcharacteristics for the combustion process of Chlorella vulgaris (CV),

Scenedesmusalmeriensis (SC) and NannochloropsisGaditana (NG)microalgae.

Biomass samples

CV SC NG

1stPeak 2ndPeak 1stPeak 2ndPeak 1stPeak 2ndPeak

T (ºC)* 212-516 516-798 152-428 428-802 156-426 426-774 Tp (ºC)* 491 617 366 570 336 601

Hcomb (kJ/g)* 7.9 7.8 8.8 * T: temperature interval where a thermal event takes place; Tp: peak temperature; Hcomb: Heat released during combustion

120 240 360 480 600 720 840 9600

5

10

15

20

25

30 NG SC CV

Hea

t Flo

w (

W/g

)

Temperature (ºC)

Chapter 3

173

When studying the biomass combustion by the DSC technique, two different

exothermic regions are generally observed [37; 46]. The first region is associated to

the combustion of light volatile matters, which provides reactivity of biomass fuels.

This peak is short and lower, so less heat was released. The second one represents the

combustion of fixed carbon [46]. As can be seen in Figure 3.8, the first exothermic

region for sample CV was represented as a wide shoulder rather than a peak.

Concerning the heat released during the combustion of the different microalgae

samples, it can be observed that samples with wider combustion peaks in the stage of

fixed carbon oxidation showed a higher heat release than samples whose combustion

was mainly developed in the stage of volatiles release. The first exothermic region

appeared in all samples at a similar temperature interval (300 ºC). However, the

second exothermic region covered a higher temperature range for sample NG (500 ºC)

than for samples SC and CV (370 and 280 ºC, respectively). Biomass samples used in

this work were ranked according to the combustion heat as follows: NG > SC > CV.

This trend did not agree well with that obtained if the combustion characteristic factor

(CCF) is considered. In the latter case, sample NG presented the lowest CCF value. As

aforementioned, the CCF measures how easy a combustible is burnt in terms of low

energy requirement to carry out the combustion process (low Ti and Tb). However, it

does not give information about the exothermic reactions taking place during the

combustion. DSC analysis helps to obtain a more realistic approach to the combustion

process of biomass and determine the amount of energy contained in the char.

Therefore, the devolatilization of sample NG led to the formation of the most

energetic char. Furthermore, the high ash content in samples CV and SC might affect

the char oxidation process [43].

Kinetic analysis

The kinetic model used in this work was derived from the pseudo multi-component

separate-stage models (PMSM) approach. In this type of models, the biomass sample

is composed of multiple pseudo components [19]. In this regard, the kinetic

parameters can be determined assuming single separate reactions for the different

Chapter 3

174

stages of thermal conversion. Microalgae combustion is usually described by two

main stages: devolatilization stage (Dev. stage) and char oxidation stage (Oxid. stage).

However, as aforementioned this classification may result unclear due to the fact that

it does not consider different thermal events that take place during these stages. Thus,

an additional sub-classification was carried out as in latter sections. In this regard, the

different event occurring during the Devolatilization stages are named to as sub-step 1,

2 and 3. Furthermore, the last stage for sample NG combustion, related to volatile

metal loss and carbonate decomposition, was named to as rem. step. Therefore, Eq. (6)

was separately used to each of the stages above commented.

The model representing the form of g(α) (Table 3.3), which delivered the highest

correlation coefficient, was considered to be the function representing the mass loss

kinetics for the samples under study. Figure 3.9 shows the plots of ln[g(α)/T2] versus

1/T that provided the best linearity at 40 ºC/min. Table 3.15 summarizes the main

kinetic parameters for the biomass samples here studied. It can be seen that all the

stages fitted well to a straight line. All samples showed the best regression coefficient

for the model O1, which is the most used mechanism for the kinetic calculation of

biomass thermal decomposition [17].

All microalgae samples showed a similar kinetic behavior. The main differences

can be attributed to the different stages considered. In this regard, sample CV did not

show the sub-step 1. Sample NG showed a slightly higher activation energy (116.8

kJ/mol) than sample SC (93.6 8 kJ/mol), which can be explained by the low

temperature required for sample SC to decompose. In a similar way, little difference in

the Ea values was observed for the sub-step 2. Sample NG showed the lowest Ea value

(62.9 kJ/mol), pointing out that it was the most reactive sample and the easiest to

ignite. This way, the combustion reaction is more continuous than that observed for

the other microalgae [47]. Regarding sub-step 3, the low amount of lipid in sample CV

and the closure of sub-step 2 and 3 would explain the lower Ea value obtained for this

sample. Sample SC showed the highest Ea value, which may be due to its lipid

composition. In this regard, Kebelmann et al. [40] reported different thermal behaviors

Chapter 3

175

between the lipids extracted from different types of microalgae being attributed to

different fatty acids compositions. Concerning the second stage of combustion, Ea

values were almost the same for all samples. This is in agreement with the results

obtained by Yu et al. [47] for the combustion of different seaweeds. Finally, sample

NG was the only one that showed a last sub-step at high temperatures. The Ea in this

stage was markedly higher than that for the previous stage, pointing out to the high

energy required for metals and carbonates volatilization.

Figure 3.9.-Plot of ln(g(α)/T) vs 1/T for the combustion process of Scenedesmusalmeriensis

(SC), Nannochloropsisgaditana (NG) and Chlorella vulgaris (CV) microalgae.




:�C:� =

�A �� C �� A(�A) (8)

whereαi, Ai, Ei and fi(αi) are the degree of conversion, the pre-exponential factor,





Figure 3.10 shows the experimental data (solid line) compared to the predicted one

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4-20

-18

-16

-14

-12

-10

-8

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4-20

-18

-16

-14

-12

-10

-8

-6

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4-20

-18

-16

-14

-12

-10

-8

-6CV

ln (g

( αα αα)/T

2 )

1/T (1/K) * 103

Sub-step 2 Sub-step 3 Oxid-step

NG

1/T (1/K) * 103

Sub-step 1 Sub-step 2 Sub-step 3 Oxid-step Rem-step

SC

1/T (1/K) * 103

Sub-step 1 Sub-step 2 Sub-step 3 Oxid-step

Chapter 3

176

(dotted line) obtained by substituting the calculated activation energy and pre-

exponential factor for each stage into Eq. (8). It can be observed that the proposed

model adequately reproduces the experimental values. The mean error between the

experimental and theoretical curves was calculated and shown in Figure 3.10. The

obtained error was lower in all cases than 3.1 %.

Table 3.14.-Estimated kinetic parameters for the combustion of Nannochloropsisgaditana

(NG), Chlorella vulgaris (CV) and Scenedesmusalmeriensis(SC) microalgae

Figure 3.10.-Comparison between experimental and theoretical results for the combustion

process of Scenedesmusalmeriensis (SC), Nannochloropsisgaditana (NG) and Chlorella

vulgaris (CV)microalgae.

400 600 800 1000 12000.0

0.2

0.4

0.6

0.8

1.0

400 600 800 1000 12000.0

0.2

0.4

0.6

0.8

1.0

400 600 800 1000 12000.0

0.2

0.4

0.6

0.8

1.0

Mean error (%): 2.7

Sc

(1- αα αα

)

Mean error (%): 3.1

NG

Temperature (K)

Mean error (%): 1.9

CV

Theoretical Experimental

Stage 1 Sub-step 1 Sub-step 2 Sub-step 3

Biomass Ea

(kJ/mol) A

(1/min) r 2 Ea

(kJ/mol) A

(1/min) r 2 Ea

(kJ/mol) A

(1/min) r 2

NG 116.8 4.7·1013 0.9901 62.9 3.1·105 0.9911 157.8 4.1·1011 0.9933 CV - - - 80.9 1.9·104 0.9915 135.27 5.2·107 0.9901 SC 93.6 1.2·1011 0.9915 71.3 1.4·106 0.9907 178.9 1.3·1013 0.9915

Stage 2

Oxid. step Rem. step

NG 113.2 3.4·106 0.9912 326.5 1.0·1013 0.9907 CV 124.9 5.2·107 0.9921 - - - SC 126.1 3.9·107 0.9908 - - -

Chapter 3

177

Evolved gas analysis

The main products derived from the combustion of the marine biomass (samples

CV, SC and NG) were evaluated by TGA-MS analysis. In this regard, a preliminary

scan was performed in order to identify the main gaseous products released during the

combustion of microalgae samples. The most prominent ions were detected at (m/z)=

2, 15, 18, 27, 28, 29, 30, 41, 44, 45, 46, 48, 50, 58, 60 and 64 corresponding to the

following compounds: H2, CH4, H2O, HCN + C2H4, CO, C2H6, NO + CH4N (primary

amines), C3H5+ (alkenes), C2H5O + CHO2 (esters and ethers + carboxylic groups),

CO2, NO2, SO, CH3Cl, C3H6O (ketones), COS and SO2, respectively. Special attention

must be taken into account when reporting some ions due to they could belong to

multiple compounds. Thus, ions with m/z 27, 30 and 45 are related to the evolution of

different compounds.

Mass spectrometry analyses for the different types of microalgae here considered:

Chlorella vulgaris (CV), Scenedesmusalmeriensis (SC) and

Nannochloropsisgaditana(NG) are shown in Figure 3.15 and 3.16. MS spectra of

microalgae could be divided into different stages related to their degradations steps

studied in the TGA/DTG curves and described in previous sections. Table 3.11

summarizes the most representative MS ions detected, their integrated peak intensities

in the whole temperature range and the temperature where an emission peak was

found.

Microalgae combustion is a chemical process where the organic matter contained

in them is oxidized to release heat [26]. However, the oxidation of microalgae

involves many complex reactions, both in parallel and series, such as thermal

cracking, condensation and depolymerization due to the complex composition of

microalgae. The thermal decomposition of microalgae can be considered step-wise

were carbohydrates, proteins and lipids are decomposed. Thus, the study of gases

evolving during combustion is of high importance from a fundamental point of view

in order to gain further understanding of the complex reactions occurring during

combustion.

Chapter 3

178

Figure 3.11.-Gas evolution profile of CO, CO2, H2O, CH4, C3H5+ (alkenes), C2H5, H2 and C3H6O for the combustion process of

Scenedesmusalmeriensis (SC), Nannochloropsisgaditana (NG) and Chlorella vulgaris (CV) microalgae.

200 400 600 800 1000

200 400 600 800200 400 600 800 1000 200 400 600 800 1000

200 400 600 800

200 400 600 800

200 400 600200 400 600 800

Temperature (ºC)

C2H

6

Temperature (ºC)

NG SC CV

CH4

Inte

nsity

(a.

u.)

Temperature (ºC)

CO

Temperature (ºC)

CO2

Temperature (ºC)

H2

Temperature (ºC)

H2O

Temperature (ºC)

C3H

6O (ketones)

Inte

nsity

(a.

u.)

Temperature (ºC)

C3H

5

+ (alkenes)

Chapter 3

179

Figure 3.12.-Gas evolution profile of NO, NO2 SO, SO2, COS, CH3Cl, HCN and CH4N for the combustion process of Scenedesmusalmeriensis

(SC), Nannochloropsisgaditana (NG) and Chlorella vulgaris (CV) microalgae.

200 400 600 800 1000200 400 600 800200 300 400 500 600 700 200 400 600 800 1000

200 400 600200 400 600 800 1000 200 400 600 800 1000200 400 600 800 1000

Temperature (ºC)

C2H

5O + CHO

2

Temperature (ºC)

COS

Inte

nsity

(a.

u.)

Temperature (ºC)

SO

Temperature (ºC)

SO2

Temperature (ºC)

NG SC CV

CH3Cl

Temperature (ºC)

CH4N + NO

Temperature (ºC)

HCN + C2H

4

Inte

nsity

(a.

u.)

Temperature (ºC)

NO2

Chapter 3

180

Table 6.-MS characteristics for the combustion process of Chlorella vulgaris (CV), Scenedesmusalmeriensis (SC)

andNannochloropsisgaditana (NG) microalgae

SC NG CV (m/z)

Gas Tp

1*

(ºC)

Tp2*(º

C) Tp3

*(ºC)

Tp4*(º

C) Int* Tp1

*(ºC)

Tp2*(º

C) Tp3

*(ºC)

Tp4*(º

C) Int* Tp1

*(ºC)

Tp2*(º

C) Tp3

*(ºC)

Tp4*(º

C) Int*

2 H2 320

517 - - 6.9·10-7

280 552 - - 2.6·10-6

307 552 - - 1.6·10-6

12 C+ 336

552 - - 2.6·10-5

340 590 - - 3.2·10-5

380 611 811 - 2.8·10-5

14 CO+, N+ 325

539 - - 6.1·10-6

338 515 - - 7.4·10-6

379 554 693 - 6.3·10-6

15 CH4 325

506 - - 1.3·10-5

278 333 510 645 2.2·10-5

380 529 - - 1.2·10-5

18 H2O 172

309 523 - 9.1·10-4

282 431 563 - 1.2·10-3

298 375 567 - 7.7·10-4

27 C2H3+,

HCN 347

440 519 - 1.5·10-5

334 449 603 - 1.5·10-5

378 592 899 - 2.7·10-5

28 CO 338

555 - - 2.3·10-4

339 456 602 851 2.5·10-4

380 610 818 - 2.8·10-4

29 C2H6, CHO+

183

309 453 534 271 336 448 592 1.3·10-5

190 291 384 611 9.7·10-6

30 C2H6, CH2O, NO,

CH4N

181

361 598 - 2.7·10-5

275 362 539 680 3.9·10-5

344 562 683 901 1.8·10-5

Chapter 3

181

41 C3H5+

(alkenes) 329

450 - - 2.5·10-6

331 437 - - 4.5·10-6

379 425 501 - 2.6·10-6

44 CO2 329

555 - - 5.3·10-4

193 339 592 870 6.9·10-4

380 610 818 - 4.3·10-4

45 C2H5O+;

CHO2 340

554 - - 6.1·10-6

203 344 587 895 7.6·10-6

372 609 899 - 4.4·10-6

46 NO2; C2H5OH

337

573 - - 1.9·10-6

193 339 591 - 2.5·10-6

392 608 830 - 1.5·10-6

48 SO 279

443 - - 4.1·10-6

275 386 - - 4.2·10-7

322 429 - - 2.8·10-7

50 CH3Cl 306

445 - - 1.2·10-7

279 421 - - 5.3·10-7

303 382 - - 1.5·10-7

58 C3H6O(ketones)

326

- - - 1.2·10-7

326 424 - - 1.9·10-7

395 573 - - 5.5·10-8

60 COS 388

- - - 8.6·10-8

413 - - - 1.6·10-7

262 - - - 1.4·10-7

64 SO2 28

1 375 582 - 4.4·1

0-7 275 383 597 - 4.1·1

0-7 312 390 641 - 2.8·1

0-7

Chapter 3

182

The gaseous emissions followed a similar pattern in all microalgae samples. CO,

CO2 and H2O were the main components produced and evolved over the whole

temperature range. CO and CO2 main peaks took place at temperatures between 555

and 610 ºC being coincident with their DTG peak. The CO and CO2 emissions

detected in this temperature range were due to the fixed carbon oxidation [26].

Emission peaks at lower temperatures were associated to the decomposition of

carboxyl groups in protein and saccharides [48]. Furthermore, an additional peak was

found for samples CV and NG at temperatures above 800 ºC, which was attributed to

the decomposition of mineral matter, as carbonates, in the ash [26; 49]. On the other

hand, H2O emissions reached their maximum evolution rate at 282, 298 and 309 for

samples NG, CV and SC, respectively. The H2O produced at this step was mainly

associated to the oxidation of oxygen containing functional groups (especially

hydroxyl groups). H2O peaks at lower temperatures were attributed to the loss of

cellular water and external water bounded by surface tension [44]. Finally, the water

produced at higher temperatures was associated to the evolution of H2. In this regard,

H2 was produced by the dehydrogenation of the char [50] and reached their maximum

yield at 517 and 552 ºC for samples SC and NG and CV. In addition, H2O emissions

peaks were found at slightly higher temperatures. This fact pointed out that H2O at this

stage was produced by the oxidation of produced H2. H2O peaks were found at lower

temperatures than CO2 ones, indicating that microalgae samples decomposition started

via dehydration of the algae components followed by combustion [50]

Light hydrocarbons (HC), especially CH4, were the main secondary products

detected. The origin of HC is attributed to the decomposition of carbohydrates and

lipids. In this sense, Marcilla et al. [51] reported that the main source of methyl groups

was the decomposition of lipids. Their results agreed well with those reported in this

study as maximum peaks for CH4 were obtained at 506, 510 and 529 ºC.

Carbohydrates decomposition also led to the formation of HC, as it can be observed

from the C2H5 and C3H5 (alkenes) emissions between 180 and 450 ºC. Emission peaks

for HC were also observed at higher temperatures as reported by Bae et al. [52]. The

evolution of oxygen containing hydrocarbons such as ketones (C3H6O) and carboxylic

Chapter 3

183

acids, esters and ethers (C2H5O + CHO2) was attributed to the breaking up of carbonyl

groups from fatty acids[53]. Ketones were mainly detected between 326 and 396 ºC

whereas carboxylic acids, ester and ethers evolved in two steps. The C-C scission of

these compounds may turn out in direct CO emissions, or they can be later combined

with oxygen to form CO2.

The evolution of N-compounds took place forming different emission peaks.

Nitrogen compounds (N-compounds) evolved as CH4N (primary amines), NOx and

HCN. The first peak between 200 and 400 ºC was mainly associated to the

decomposition of proteins. The second peak was related to the ignition of N-

containing compounds between 400 and 500 ºC. In this temperature range, the

presence of primary amines and HCN was less likely and their signal may have an

important contribution of other compounds related to their ions such as HC and NO.

The presence of NO at temperatures above 400 ºC has been reported by different

authors [26; 48]. The last peak at temperatures above 500 ºC corresponded to the

oxidation of the remaining nitrogen in the char. The maximum yield of NO2 was

reached in this stage. The release of NOx is of high importance as they are the primary

components of photochemical smog [26].

Chloride and sulfur compounds were released in lower proportions than nitrogen

compounds. Chloride compounds were mainly detected as CH3Cl and emitted between

200 and 400 ºC, the evolution of this compound was much higher for the sample NG

than for samples SC and CV which is in good agreement with their compositional

analysis (Table 1.2). Concerning the release of sulfur compounds, SO and SO2 were

the main products detected. Their release is mainly associated to the decomposition

and oxidation of sulphated polysaccharides[48]. SO and SO2 maximum peaks were

found at similar temperatures for all samples (270-300 ºC). Furthermore, a second

emission peak was found in the lipid decomposition temperature range (350-480 ºC)

for both compounds due to the degradation of organic sulfides in organic residues

[48].These peaks were characterized by the apparition of SO2 at lower temperatures

than SO. Finally, only SO2 was produced at higher temperatures being quite consistent

Chapter 3

184

with the combustion of fixed carbon temperature range. Furthermore, COS was

produced, which can be originated by the partial oxidation of organic sulfur or the

reaction of SO2 with carbon complexes [54].

Samples SC and NG showed peaks at temperatures above 700 ºC for different

compounds (HCN + C2H4, C2H5, CH4N + NO, NO2, CO and CO2). The evolution of

compounds in this temperature range is mainly associated to the catalytic effect of

some compounds in the ash. Furthermore, the volatilization of some mineral matter

and carbonates may take place as above commented. However, this fact was unusual

as no appreciable weight loss was detected at this temperature for these samples

(<0.05 wt.% for both samples). The interactions of the ash in the combustion of

microalgae were out of the scope of this work. However, the problematic associated to

their presence in combustion processes can be considered of high importance and

further studies will be carried out in order to achieve a deeper insight of their behavior.

3.3.3. Combustion of Canadian biomass

Thermogravimetric analysis

Figure 3.13 shows the thermogravimetric (TGA) and derivative thermogravimetric

(DTG) profiles for different types of biomass here considered: two woody crops

(black spruce and Pinusbanksiana mixtures (BP) and willow (W)), and three

herbaceous non-perennial energy crops (common reed (CR), reed phalaris (RP) and

switchgrass (S)). Table 3.17 shows the main relevant combustion characteristics. The

thermal decomposition of biomass under oxidative environment is usually described

by two stages[19; 37;43]. Firstly, the devolatilization of the sample takes place at low

temperatures (160-400 ºC), leading to char formation. Then, the oxidation of the

sample occurs at temperatures higher than 400 ºC. Generally, each stage is attributed

to the decomposition of the biomass main components (hemicellulose, cellulose and

lignin)[3; 5].Hemicellulose and cellulose are assumed to decompose during the

devolatilization stage[55]. Hemicellulose usually appears as a shoulder in the DTG

curve in the devolatilization stage at low temperatures, whereas cellulose oxidation

Chapter 3

185

produces the main DTG peak in this stage. On the other hand, lignin decomposes in a

wider temperature range being the main responsible of biomass char formation[56].

Finally, the formed char is burnt at high temperatures.

Figure 3.13.-Thermogravimetric curves for the combustion process of: a) non-woody

perennial crops (common reed (CR), reed phalaris (RP) and switchgrass (S)) and b) woody

biomass (black spruce and Pinusbanksiana mixtures (BP) and willow (W))

For woody crops, sample W started decomposing at lower temperatures than

sample BP (175 ºC and 197 ºC for samples W and BP, respectively). On the other

hand, for herbaceous crops, sample RP decomposed at lower temperatures (160 ºC)

than samples S (195 ºC) and CR (206 ºC). Concerning the devolatilization stage,

herbaceous crops showed a more prominent shoulder in the hemicellulose

decomposition region than woody biomass. This shoulder was more visible in the

combustion of sample RP, indicating the presence of higher hemicellulose content.

The ignition temperature was found to be for all samples between 290 and 315 ºC but

for sample RP this temperature was lower (273 ºC). The maximum weight loss rate

0

20

40

60

80

100

200 400 600 800 10000.0

0.3

0.6

0.9

1.2

0

20

40

60

80

100

200 400 600 800 10000.0

0.3

0.6

0.9

1.2

CCF (x107)

2.672.353.16

Wei

ght (

%)

RP CR S

CCF (x107)

Wei

ght l

oss

deriv

ativ

e (

wt.%

/ ºC

)

Temperature (ºC)

4.5

5.16

W BP

Temperature (ºC)

Chapter 3

186

was found for samples BP and S, which pointed out to a higher content in cellulose.

The char oxidation stage started at similar temperatures for all samples (375-391 ºC).

However, it ended at a slightly higher temperature for herbaceous crops, 540 ºC,

compared to 507-534 ºC temperature range for the woody crops. An additional mass

loss weight was found in the combustion of sample W. This fact was attributed to the

burning of the remaining semi-coke. The highest value of Tb was found for sample W,

which is related to the presence of high lignin content. Finally, the amount of residue

(ash) that remained after the combustion process is of great interest. Sample CR left

the highest amount of residue (18.2wt.%) whereas BP yielded a very low one

(1.6wt.%). The ash composition is dominated by metal oxides, especially silica,

calcium oxide and potassium oxide. An high ash content can contribute to the

development of the combustion process due to its catalytic effect[43]. In the opposite,

a high ash content contribute to operational problems due to the occurrence of fouling

and slagging phenomena.

The relative amount of biomass main components is important when determining

the quality of a biomass fuel. In this regard, a high content in hemicellulose and

cellulose turns out in a low Ti and a high (dw/dT)max. On the other hand, a higher

content in lignin produces a high amount of residue to be burnt. Wang et al.

(2009)[37] described the combustion characteristic factor (CCF), that can be used as a

criterion for fuel combustion performance according to the mentioned parameters (the

higher the CCF value is, the easier to ignite a sample is), as follows:

889 = (:;:� )<=> ∙ (:;:� )<?=@ �A$ ∙ �B

� (8)

where (dw/dt)max is the maximum burning velocity (%/min); (dw/dt)mean is the average

burning velocity (%/min); Ti is the ignition temperature (K) and Tb is the burnout

temperature (K).

Chapter 3

187

Table 3.14.-TGA-DTG characteristics for the combustion process of black spruce and Pinusbanksiana mixtures (BP), willow (W),switchgrass

(S), common reed (CR) and reed phalaris (RP)

Woodycrops Herbaceous non perennialcrops

BP W S CR RP

1st

peak

2nd

peak

1s

tpeak

2nd

peak

3rd

peak

Sh* 1st

peak

2nd

peak

Sh* 1st

peak

2nd

peak

1st

peak

2nd

peak

3rd

peak

Tdo (ºC)* 197 177 195 206 160

T i (ºC) * 316 292 314 295 273

Tb (ºC) * 507 695 543 544 549

Tpo (ºC) * 197 391 183 385 611 195 195 380 206 206 375 160 263 381

Tpf (ºC) * 391 507 385 534 695 316 380 543 307 375 544 263 381 549

Tp (ºC) * 353 451 336 410 660 316 339 410 306 332 382 250 326 435

(dw/dT)max*

(dwt.%/ºC)

1.21 0.38 0.82 0.33 0.02 0.67 1.18 0.27 0.55 0.92 0.27 0.24 0.66 0.26

Residue (%) 1.6 8.6 6.7 18.2 11.7

*Sh: Shoulder; Tdo: Initial decomposition temperature; Ti: Ignition temperature; Tb: Burnout temperature; Tpo: Initial peak temperature; Tpf: Final peak temperature; Tp: Peak temperature;

(dw/dT)max: Maximum weight loss rate

Chapter 3

188

CCF values for all samples are plotted in Figure 3.13. In all cases, these values

were bigger than 2 indicating the good general burning performance[37]. However,

these data must be used only as a reference since they do not give any information

about the heat released during the combustion process.

Differential scanning calorimetric analysis

Woody crops (samples W and BP) and herbaceous non-perennial energy crops

(samples S, CR and RP) were also investigated by the DSC technique. Experimental

DSC curves are presented in Figure 3.14. This way, it is possible to identify the kind

of mass loss event explained in the TG analyses[57]. DSC main temperatures and heat

of combustion (Hcomb) are included in Table 3.18.

Figure 3.14.-DSC curves for the combustion process of: a) non-woody perennial crops

(common reed (CR), reed phalaris (RP) and switchgrass (S)) and b) woody biomass (black

spruce and Pinusbanksiana mixtures (BP) and willow (W))

When studying the biomass combustion by the DSC technique, two different

exothermic regions are generally observed[37; 46]. The first region is associated to the

combustion of light volatile matters, which provides reactivity of biomass fuels. The

second one represents the combustion of fixed carbon[46]. As can be seen from Figure

3.14, most biomass samples followed this trend. However, both exothermic regions in

120 240 360 480 600 720 840 9600

5

10

15

20

25

30

35

120 240 360 480 600 720 840 9600

5

10

15

20

25

30

35

40 RP CR S

Hea

t flo

w (

W/g

)

Temperature (ºC)

W BP

Temperature (ºC)

Chapter 3

189

sample CR overlapped turning into only one peak at an intermediate temperature

between those corresponding to volatile matters release and fixed carbon combustion,

respectively. This fact can be explained attending to the low fixed carbon content in

sample CR (Table 3.1). Furthermore, the high ash content might reduce the char

combustion temperatures due to its catalytic metal content[43].

Concerning the heat released during the combustion of the different biomass

samples, it can be observed that samples with more prominent DSC peak in the stage

of combustion of fixed carbon showed a higher heat release than samples whose

combustion was mainly developed in the stage of volatiles release. Biomass samples

used in this work were ranked according to the combustion heat as follows: W> RP

>CR>S >BP. This trenddid not agree well with that obtained if thecombustion

characteristic factor (CCF) is considered. In the latter case, sample BP presented the

highest CCF. As aforementioned, the CCF measures how easy a combustible is burnt

in terms of low energy requirement to carry out the combustion process (low Ti and

Tb). However, it does not give information about the exothermic reactions taking place

during combustion. DSC analysis helps to obtain a more realistic approach to the

combustion process of biomass and determine the amount of containing energy in the

char. Therefore, the devolatilization of samples W and RP led to the formation of the

more energetic char. The combination of this fact and a high CCF value might

determine the combustion quality of a biomass[37; 46].

Kinetic analysis

The kinetic model used in this work was derived from the pseudo multi-component

separate-stage models (PMSM) approach. In this type of models, the biomass sample

is composed of multiple pseudo components[19]. In this regard, the kinetic parameters

can be determined assuming single separate reactions for the different stages of

thermal conversion. As abovementioned, biomass combustion was clearly defined by

two main stages: devolatilization stage (Dev. stage) and char oxidation stage (Oxid.

stage). However, additional decompositions occurred for some samples. For example,

the Dev. Stage for herbaceous crops was represented by two peaks. In order to

Chapter 3

190

differentiate them, the Dev. stage for them was divided into two stages: Dev. stage A

and Dev. stage B. This way, an additional weight loss step took place for sample W

combustion due to the remaining char burning. This stage was named as Rem. stage.

Therefore, eq. (6) was used separately to each of the stages above commented.

The model representing the form of g(α)(Table 3.3), which delivered the highest

correlation coefficient, was considered to be the function representing the mass loss

kinetics for the samples under study.Figure 3.15 shows the plots of ln[g(α)/T2] versus

1/T that provided the best linearity at 40 ºC/min. Table 3.19 summarizes the main

kinetic parameters for the biomass samples here studied. It can be seen that all the

stages fitted well to a straight line (r2> 0.99).All samples showed the best regression

coefficient for the model O1, which is the most used mechanism for the kinetic

calculation of biomass thermal decomposition[17; 19]. However, a good correlation

was also obtained by the model D3 for the Dev. stage 1 in herbaceous crops. This stage

corresponded to the hemicellulose decomposition temperature range. High correlation

coefficients for diffusion mechanisms during the devolatilization stage of

coal/biomass blends were also observed by Gil et al. (2010)[2]. Yorulmaz and Atimtay

et al. (2009)[7] reported that different models can be suitable to describe the biomass

combustion process by thermal analysis. Further research combining dynamic and

isothermal studies should be carried out in order to elucidate the exact mechanisms of

the oxidation process.

Calculated activation energies for the multiple-step model for different types of

woody and herbaceous biomass are listed in Table 4. Activation energies obtained for

the first stage of woody biomass during the oxidation process were quite similar (91

kJ/mol and 101 kJ/mol for samples BP and W, respectively). These values are in good

agreement with those obtained by different authors[1; 2]. Higher values were obtained

for the oxidation stage. Thus, sample BP showed the highest one (143 kJ/mol),

pointing out that the char obtained from this sample is less reactive. Finally, the Rem.

stage in the oxidation of sample W showed a high value of the activation energy (372

kJ/mol). This fact could be attributed to the little amount of remaining char-semi-coke,

Chapter 3

191

which required a high energy to be decomposed[58]. On the other hand, herbaceous

crops showed a more diversified activation energy distribution. Sample RP had the

lowest activation energy value for Dev. stage 1, which is related to its high content in

hemicellulose and lignin[17]. Higher activation energies in the Dev. stage 2 were

obtained for samples RP and S, which is in agreement to the high value required for

cellulose decomposition. However, sample CR showed for this stage a reduction in the

activation energy. This might be due to the high content in ash, which catalyzes the

biomass oxidation lowering the energetic requirement to let the combustion process

progress. Finally, the values of the activation energy for the oxidation stage were

lower than those observed in previous stages.

Figure 3.15.-Plot of ln(g(α)/T) vs 1/T for the combustion process of non-woody perennial

crops (common reed (CR), reed phalaris (RP) and switchgrass (S))and woody biomass (black

spruce and Pinusbanksiana mixtures (BP) and willow (W))

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4-26

-24

-22

-20

-18

-16

-14

-12

-10

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4-26

-24

-22

-20

-18

-16

-14

-12

-10

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4-26

-24

-22

-20

-18

-16

-14

-12

-10

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4-26

-24

-22

-20

-18

-16

-14

-12

-10

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4-26

-24

-22

-20

-18

-16

-14

-12

-10

W

1/T (K -1) * 103

RP

1/T (K -1) * 103

CR

1/T (K -1) * 103

Dev. Stage 1 D3

Dev. Stage 2 O1

Oxid. stage O1

Dev. Stage 1 D3

Dev. Stage 2 O1

Oxid. stage O1

ln (g

( αα αα)/T

2 )

1/T (K -1) * 103

Dev. Stage 1 D3

Dev. Stage 2 O1

Oxid. stage O1

S

Dev. Stage O1

Oxid. Stage O1

Rem. Stage O1

Dev Stage O1

Oxid. Stage O1

BP

ln (g

( αα αα)/T

2 )

1/T (K -1) * 103

Chapter 3

192




:�C:� =

�A �� =C �� A(�A) (8)

whereαi, Ai, Eai and fi(αi) are the degree of conversion, the pre-exponential factor,





Figure 3.16 shows the experimental data (solid line) compared to the predicted one

(dotted line) obtained by substituting the calculated activation energy and pre-

exponential factor for each stage into Eq. (8). It can be observed that the proposed

model adequately reproduces the experimental values. The mean error between the

experimental and theoretical curves was calculated and shown in Figure 3.16. The

obtained error was small in value for all samples (lower than 3.4%).

Chapter 3

193

Figure 4.-Comparison between experimental and theoretical results for the combustion process

of non-woody perennial crops (common reed (CR), reed phalaris (RP) and switchgrass (S))

and woody biomass (black spruce and Pinusbanksiana mixtures (BP) and willow (W))

Evolved gas analysis

Although contaminant emissions associated to biomass are lower than those in

fossil fuels, they must be taken into account due to the high development of biomass

conversion technologies[59]. In this regard, a preliminary scan was performed in order

to identify the main contaminants released during the combustion of woody and

herbaceous biomasses. The most prominent ions related to contaminants were detected

at (m/z) = 18, 27, 28, 30, 36, 44, 46, 48, 50, 64 and 78 corresponding to the following

compounds: H2O, HCN, CO, NO, CO2, NO2, SO, CH3Cl, SO2 and C6H6, respectively.

Figure 3.17 and 3.18 shows the mass spectra obtained for the combustion process

of woody and herbaceous crops. MS curves in Figure 3.18 were moved up and down

in order to clarify the results due to the fact that they mostly overlapped. Table 3.20

400 600 800 1000 12000.0

0.2

0.4

0.6

0.8

1.0

400 600 800 1000 12000.0

0.2

0.4

0.6

0.8

1.0

400 600 800 1000 12000.0

0.2

0.4

0.6

0.8

1.0

400 600 800 1000 12000.0

0.2

0.4

0.6

0.8

1.0

400 600 800 1000 12000.0

0.2

0.4

0.6

0.8

1.0

Mean error (%): 1.86Mean error (%): 0.95

RP

Temperature (K)

S

Temperature (K)

Mean error (%): 3.35

Temperature (K)

W

CR

Temperature (K)

Theoretical Experimental

Mean error (%): 1.82Mean error (%): 2.98

(1- αα αα

)

BP

(1- αα αα

)

Temperature (K)

Chapter 3

194

summarizes the most representative MS ions detected, their integrated peak intensities

in the whole temperature range and the most relevant temperatures.

The mass spectra showed two emission peaks for most detected products. The first

one that took place at lower temperatures, corresponded to the so-called oxidative

pyrolysis or devolatilization of the sample. The second one corresponded to the

oxidation of the char. CO, CO2 and H2O were the main compounds formed in the

combustion process of woody and herbaceous biomass. CO and CO2 evolved for all

samples in the whole oxidation temperature range. Both compounds showed emission

peaks at similar temperatures in the first emission temperature range, whereas in the

char oxidation stage the CO2 peak was detected at higher temperatures. The maximum

CO and CO2 yield was observed in the Oxid. stage. CO formation during the first

stage was associated to decarbonylation reactions, secondary reactions between

volatiles and rearrangement of the char skeleton[33]. Furthermore, the CO evolution

during the char oxidation stage was mainly due to the formation of active carbon sites

in the char which were later oxidized, releasing CO and leaving oxygen atoms

attached to carbon surface[32]. The CO2 evolution during the first decomposition

stage followed a similar path than that for CO. However, the later appearance of CO2

compared to CO during the second stage, pointed out that a fraction of the produced

CO reacted with oxygen increasing the CO2 yield. Additionally, Li and Brown et al.

(2001)[32] established different pathways for CO2 evolution during char combustion

involving the formation of different carbon-oxygen complexes. On the other hand,

H2O spectra was released into one step between 334 and 356 ºC. In this stage, the

water formed is known as pyrolityc water[34]and is produced due to hydroxyl groups

bond scission. A small shoulder could also be observed at slightly high temperatures.

The H2O shoulder is mainly attributed to the oxidation of hydrogen produced at

temperatures higher than 400 ºC due to char cracking reactions and the occurrence of

the reverse water-gas shift reaction[33; 34].

Chapter 3

195

Figure 3.17.-Gas evolution profile of CO, CO2, H2O, NO and NO2 for the combustion process of non-woody perennial crops (common reed

(CR), reed phalaris (RP) and switchgrass (S)) and woody biomass (black spruce and Pinusbanksiana mixtures (BP) and willow (W))

200 400 600 800 1000 200 400 600 800 1000 200 400 600

Inte

nsity

(a.

u.)

Temperature (ºC)

CO

Temperature (ºC)

CO2

Temperature (ºC)

BP W S CR RP

H2O

200 400 600 800 1000 200 400 600 800 1000200 400 600

Inte

nsity

(a.

u.)

Temperature (ºC)

NO

Temperature (ºC)

HCN

Temperature (ºC)

NO2

Chapter 3

196

Figure 3.18.- Gas evolution profile of C6H6, SO, SO2, CH2Cl, CCl2 and CH3Cl for the combustion process of non-woody perennial crops

(common reed (CR), reed phalaris (RP) and switchgrass (S)) and woody biomass (black spruce and Pinusbanksiana mixtures (BP) and willow

(W))

200 400 600

200 400 600 200 400 600

200 400 600 800 1000

200 400 600 800 1000

200 400

Temperature (ºC)

CCl2

Temperature (ºC)

SO

Temperature (ºC)

BP W S CR RP

SO2

Inte

nsity

(a.

u.)

Temperature (ºC)

CH2Cl

Inte

nsity

(a.

u.)

Temperature (ºC)

C6H

6

Temperature (ºC)

CH3Cl

Chapter 3

197

Table 5.-MS characteristics for the combustion process of samples BP, W, CR, RP and S

(m/z) 18 27 28 30 44 46 48 49 50 64 78 82 Biomass Comp H2O HCN CO NO CO2 NO2 SO CH2Cl CH3Cl SO2 C6H6 CCl2

P1* T (ºC)* 127-425 188-408 194-372 188-449 194-377 198-385 - 240-545 209-334 198-464 - 247-373

BP Tp (ºC)* 354 356 357 353 354 354 - 358 320 352 - 337 P2

* T (ºC)* 425-556 408-540 372-566 449-565 377-678 385-593 - - 334-429 - - 373-524 Tp (ºC)* 425 420 440 481 458 468 - - 355 - - 357 Int *(A ºC/mbar mg) 1.7·10-4 2.1·10-6 3.6·10-4 2.2·10-5 6.3·10-4 3.3·10-6 - 1.1·10-8 1.2·10-7 4.9·10-8 - 7.6·10-8

P1* T (ºC)* 113-408 185-391 167-373 169-397 152-365 163-373 233-411 260-553 222-438 217-409 294-532 245-524

W Tp (ºC)* 336 338 342 335 343 334 342 337 325 301 376 341 P2

* T (ºC)* 408-599 391-586 373-641 397-568 365-706 373-687 - - - - - - Tp (ºC)* 408 402 402 441 448 449 - - - - - - Int *(A ºC/mbar mg) 1.1·10-4 1.9·10-6 1.9·10-4 1.4·10-5 4.3·10-4 1.9·10-6 1.8·10-8 2.4·10-8 7.3·10-8 4.3·10-8 3.6·10-8 4.6·10-8

P1* T (ºC)* 157-408 196-389 184-365 210-483 181-360 198-373 235-414 184-400 193-445 207-486 - 253-516

S Tp (ºC)* 341 341 343 339 341 340 295 305 306 320 - 340 P2

* T (ºC)* 408-534 389-558 365-608 483-612 360-720 373-555 - - - - - - Tp (ºC)* 408 395 405 526 420 403 - - - - - - Int *(A ºC/mbar mg) 1.8·10-4 3.6·10-6 3·10-4 1.7·10-5 6.5·10-4 3.3·10-6 6.7·10-8 2.9·10-8 2.2·10-7 1.1·10-7 - 7·10-8

P1* T (ºC)* 139-594 208-387 180-353 180-470 174-348 169-375 213-399 184-400 161-416 169-323 235-435 234-353

CR Tp (ºC)* 336 337 339 335 340 335 307 306 304 295 343 341 P2

* T (ºC)* - 387-543 353-632 470-666 348-733 375-615 - - - 323-532 - - Tp (ºC)* - 386 384 518 397 390 - - - 349 - - Int * ( A ºC/mbar mg) 3.5·10-4 6.5·10-6 6·10-4 3.8·10-5 2·10-3 9.2·10-6 2.3·10-7 2.1·10-8 1.4·10-6 2.3·10-7 3.9·10-8 3·10-8

P1* T (ºC)* 112-261 170-381 157-373 177-413 121-367 193-380 191-405 216-362 181-423 193-488 - -

RP Tp (ºC)* 248 325 334 329 334 329 310 318 317 294 - - P2

* T (ºC)* 261-605 381-629 373-632 413-615 367-734 380-639 - - - - - - Tp (ºC)* 334 418 419 517 459 459 - - - - - - Int *(A ºC/mbar mg) 3.5·10-5 1.7·10-8 1.4·10-4 9.1·10-6 3.1·10-4 1.5·10-6 5.2·10-8 2.6·10-8 1.5·10-7 5.2·10-8 - -

*T: Temperature; *Tp: Peak temperature; * Int: Integrated peak area; *P1 First peak; *P2: Second peak

Chapter 3

198

Two emission regions were also observed for nitrogen compounds. NO and NO2

spectra showed two well developed peaks whereas HCN showed a clear peak in the

Dev. stage and kept evolving as a shoulder up to temperatures around 400 ºC. The first

emission peak for HCN, NO and NO2 took place at similar temperatures (330-356 ºC).

In this stage, the nitrogen containing volatiles are mainly attributed to the

decomposition of proteins[60]. Protein decomposition leads to the formation of

volatile cyclic amides which, due to cracking reactions, produces HCN among other

components[60]. The second peak stood out for the early apparition of NO2 followed

by the NO emission at higher temperatures for all samples except for W. These results

agree well with those reported by Darvell et al. (2012)[12] who found two stages for

the combustion of different biomass char model samples. These stages were

characterized by the release of NO at higher temperatures and the presence of sharper

peaks. Peak areas for NO and HCN showed a higher amount of these compounds

compared to NO2. This fact is due to ions selected as NO and HCN compounds, (m/z)

= 30 and 27, respectively belong to other compounds such as light hydrocarbons,

which are also common products from biomass combustion.

Sulfur compounds (SO and SO2) were found in a lower proportion than nitrogen

ones, which is in agreement with biomass samples ultimate analysis (Table 3.1). Both

compounds were detected in all samples, but in sample BP, which had the lowest

sulfur content in the original material and only SO2 was observed. SO and SO2 were

released during the first stage at temperatures between 300 and 340 ºC. As observed

by Otero et al. (2002), the SO2 peaks occurred at lower temperatures than CO2 peaks

in a narrow temperature range and with a shallow shape. Chloride compounds were

mainly detected as CH3Cl for all studied samples, which showed the clearest spectra,

whereas Cl- ions were more diffused due to the fact that they were found close to the

sensitivity level of the mass spectrometer. In the same way, C6H6 was only observed

in samples W and CR, being an indicative of lignin decomposition.

Sample CR showed the highest yield for volatiles, which were released at lower

temperatures than observed for other biomasses. As above mentioned, the high content

Chapter 3

199

in ashes of sample CR, catalyzed the volatile release, increasing products yields and

shifting them to lower temperatures[61].

3.4. conclusions

Combustion behavior and gas formation from the oxidation process of fir wood,

eucalyptus wood, pine bark and three individual components of lignocellulosic

biomass (cellulose, hemicellulose and lignin) were analyzed by TGA-MS.Biomass

combustion took place into two main stages:devolatilization stage (Dev. stage) and

oxidation stage (Oxid. stage). Most products detected in the combustion of

lignocellulosic biomass were released during the Dev. stage whereas only NO2,

C2H5O+, CO and CO2 were detected at the Oxid.stage. Nitrogen compounds were

released as CH4N, HCN and NOx. Lignocellulosic biomass combustion was fitted to a

first order reaction model (O1). The combustion behavior of marine biomass was

carried out by TGA-DSC-MS. Three different types of microalgae

(Nannochloropsisgaditana(NG), Scenedesmusalmeriensis(SC) and Chlorella vulgaris)

were selected due to their chemical composition. Combustion of microalgae took

place into two main stages: devolatilization stage and oxidation stage. However, up to

three sub-steps could be identified during the microalgae combustion attributed to the

decomposition of carbohydrates, proteins and lipids. The ignition characteristic

showed that samples CV and SC required less amount of energy to develop the

combustion process. However, NG sample released a higher amount of heat during the

combustion. The kinetic analysis of microalgae combustion showed that the most

representative mechanism for representing the process was a first order reaction model

(O1). The excellent fitting between the experimental and theoretical curves (maximum

mean error was 3.1%, for NG sample) confirmed the selection of model O1. CO, CO2

and H2O were the main products released during combustion. Other compounds

detected during the combustion of microalgae were light hydrocarbons (especially

CH4); nitrogen compounds (mainly released as NO, NO2 and HCN); sulfur

compounds (SO, SO2 and COS); hydrogen and other oxygen containing hydrocarbons

Chapter 3

200

(ketones, esters, ethers and carboxylic acids). Nitrogen compounds were found in

higher proportions than sulfur ones.

Combustion behavior and gas formation from the oxidation process of two woody

crops (black spruce and Pinusbanksiana mixtures (BP) and willow (W)), and three

herbaceous non-perennial energy crops (common reed (CR), reed phalaris (RP) and

switchgrass (S)) were studied by TGA-DSC-MS. Samples W and RP showed the best

burning profile by combining a high combustion characteristic factor (CCF) and a

high release of combustion heat (Hcomb). The kinetic analyses of the oxidation process

was performed using pseudo mulit-component separate-stage models (PMSM). The

combustion process was divided into three stages: Devolatilization stage (correlated

with the hemicellulose and cellulose content in the samples), Oxidation stage

(influenced by the initial amount of lignin in the samples) and Remaining burning

(associated to the final char burning and devolatilization of inorganic matter). The

high ash content of CR sample enhanced the amount of volatiles released during the

combustion process lowering its activation energy. The good fitting of experimental

curves with theoretical ones validated the proposed model (mean error below 3.4 %).

H2, CO and CO2were the main product obtained from energy crops combustion

process. Furthermore, NOx were detected in a higher proportion than other pollutants

such as SOx, chloride compounds (CH3Cl) or aromatic ones (C6H6).

3.5. References

[1] Shen, D.K., Gu, S., Luo, K.H., Bridgwater, A.V., Fang, M.X. 2009. Kinetic study

on thermal decomposition of woods in oxidative environment. Fuel, 88(6), 1024-30.

[2] Gil, M.V., Casal, D., Pevida, C., Pis, J.J., Rubiera, F. 2010. Thermal behaviour

and kinetics of coal/biomass blends during co-combustion. Bioresource Technology,

101(14), 5601-8.

[3] Kai, X., Yang, T., Huang, Y., Sun, Y., He, Y., Li, R. 2011. The effect of biomass

components on the co-combustion characteristics of biomass with coal. Zhangjiajie,

Hunan. pp. 1274-8.

Chapter 3

201

[4] Stenseng, M., Zolin, A., Cenni, R., Frandsen, F., Jensen, A., Dam-Johansen, K.

2001. Thermal analysis in combustion research. Journal of Thermal Analysis and

Calorimetry, 64(3), 1325-34.

[5] Cheng, K., Winter, W.T., Stipanovic, A.J. 2012. A modulated-TGA approach to

the kinetics of lignocellulosic biomass pyrolysis/combustion. Polymer Degradation

and Stability, 97(9), 1606-15.

[6] Barneto, A.G., Carmona, J.A., Alfonso, J.E.M., Blanco, J.D. 2009. Kinetic models

based in biomass components for the combustion and pyrolysis of sewage sludge and

its compost. Journal of Analytical and Applied Pyrolysis, 86(1), 108-14.

[7] Yorulmaz, S.Y., Atimtay, A.T. 2009. Investigation of combustion kinetics of

treated and untreated waste wood samples with thermogravimetric analysis. Fuel

Processing Technology, 90(7-8), 939-46.

[8] Cordero, T., Rodríguez-Maroto, J.M., Rodríguez-Mirasol, J., Rodríguez, J.J. 1990.

On the kinetics of thermal decomposition of wood and wood components.

Thermochimica Acta, 164(C), 135-44.

[9] Phukan, M.M., Chutia, R.S., Konwar, B.K., Kataki, R. 2011. Microalgae

Chlorella as a potential bio-energy feedstock. Applied Energy, 88(10), 3307-12.

[10] Naik, S., Goud, V.V., Rout, P.K., Jacobson, K., Dalai, A.K. 2010.

Characterization of Canadian biomass for alternative renewable biofuel. Renewable

Energy, 35(8), 1624-31.

[11] Hood, E.E., Nelson, P., Powell, R. 2011. Plant Biomass Conversion. Wiley.

[12] Darvell, L.I., Brindley, C., Baxter, X.C., Jones, J.M., Williams, A. 2012.

Nitrogen in biomass char and its fate during combustion: A model compound

approach. Energy and Fuels, 26(11), 6482-91.

[13] NPRI. Environment Canada. National Pollutant Release Inventory-2010 NPRI.

[14] Miranda, T., Román, S., Montero, I., Nogales-Delgado, S., Arranz, J.I., Rojas,

C.V., González, J.F. 2012. Study of the emissions and kinetic parameters during

combustion of grape pomace: Dilution as an effective way to reduce pollution. Fuel

Processing Technology, 103, 160-5.

Chapter 3

202

[15] Jakab, E., Till, F., Várhegyi, G. 1991. Thermogravimetric-mass spectrometric

study on the low temperature oxidation of coals. Fuel Processing Technology, 28(3),

221-38.

[16] Heikkinen, J.M., Hordijk, J.C., De Jong, W., Spliethoff, H. 2004.

Thermogravimetry as a tool to classify waste components to be used for energy

generation. Journal of Analytical and Applied Pyrolysis, 71(2), 883-900.

[17] White, J.E., Catallo, W.J., Legendre, B.L. 2011. Biomass pyrolysis kinetics: A



[18] Coats, A.W., Redfern, J.P. 1964. Kinetic parameters from thermogravimetric

data Nature, 201(4914), 68-9.

[19] Liu, N.A., Fan, W., Dobashi, R., Huang, L. 2002. Kinetic modeling of thermal

decomposition of natural cellulosic materials in air atmosphere. Journal of Analytical

and Applied Pyrolysis, 63(2), 303-25.

[20] Zhou, L., Wang, Y., Huang, Q., Cai, J. 2006. Thermogravimetric characteristics

and kinetic of plastic and biomass blends co-pyrolysis. Fuel Processing Technology,

87(11), 963-9.

[21] Sánchez-Silva, L., Gutiérrez, N., Romero, A., Sánchez, P., Valverde, J.L. 2012.

Pyrolysis and combustion kinetics of microcapsules containing carbon nanofibers by

thermal analysis-mass spectrometry. Journal of Analytical and Applied Pyrolysis, 94,

246-52.

[22] Li, X.G. 1999. High-Resolution Thermogravimetry of Cellulose Esters. Journal

of Applied Polymer Science, 71(4), 573-8.

[23] Sanchez-Silva, L., López-González, D., Villaseñor, J., Sánchez, P., Valverde,

J.L. 2012. Thermogravimetric-mass spectrometric analysis of lignocellulosic and

marine biomass pyrolysis. Bioresource Technology, 109, 163-72.

[24] Haykiri-Açma, H. 2003. Combustion characteristics of different biomass

materials. Energy Conversion and Management, 44(1), 155-62.

Chapter 3

203

[25] Gani, A., Naruse, I. 2007. Effect of cellulose and lignin content on pyrolysis and

combustion characteristics for several types of biomass. Renewable Energy, 32(4),

649-61.

[26] Wang, S., Jiang, X.M., Han, X.X., Liu, J.G. 2009. Combustion characteristics of

seaweed biomass. 1. combustion characteristics of enteromorpha clathrata and

sargassum natans. Energy and Fuels, 23(10), 5173-8.

[27] Williams, P.T., Besler, S. 1996. The Influence of Temperature and Heating Rate

on the Slow Pyrolysis of Biomass. Renewable Energy, 7(3), 233-50.

[28] Gadalla, A.M.M. 1984. Kinetics of the decomposition of hydrated oxalates of

calcium and magnesium in air. Thermochimica Acta, 74(1-3), 255-72.

[29] Xie, X., Goodell, B., Zhang, D., Nagle, D.C., Qian, Y., Peterson, M.L., Jellison,

J. 2009. Characterization of carbons derived from cellulose and lignin and their

oxidative behavior. Bioresource Technology, 100(5), 1797-802.

[30] Amen-Chen, C., Pakdel, H., Roy, C. 2001. Production of monomeric phenols by

thermochemical conversion of biomass: a review. Bioresource Technology, 79(3),

277-99.

[31] Liu, Q., Wang, S., Zheng, Y., Luo, Z., Cen, K. 2008. Mechanism study of wood

lignin pyrolysis by using TG-FTIR analysis. Journal of Analytical and Applied

Pyrolysis, 82(1), 170-7.

[32] Li, C., Brown, T.C. 2001. Carbon oxidation kinetics from evolved carbon oxide

analysis during temperature-programmed oxidation. Carbon, 39(5), 725-32.

[33] Shen, D., Hu, J., Xiao, R., Zhang, H., Li, S., Gu, S. 2013. Online evolved gas

analysis by Thermogravimetric-Mass Spectroscopy for thermal decomposition of

biomass and its components under different atmospheres: Part I. Lignin. Bioresource

Technology, 130, 449-56.

[34] Widyawati, M., Church, T.L., Florin, N.H., Harris, A.T. 2011. Hydrogen


oxide. International Journal of Hydrogen Energy, 36(8), 4800-13.

Chapter 3

204

[35] Darvell, L.I., Jones, J.M., Gudka, B., Baxter, X.C., Saddawi, A., Williams, A.,

Malmgren, A. 2010. Combustion properties of some power station biomass fuels.

Fuel, 89(10), 2881-90.

[36] Idris, S.S., Rahman, N.A., Ismail, K. 2012. Combustion characteristics of

Malaysian oil palm biomass, sub-bituminous coal and their respective blends via

thermogravimetric analysis (TGA). Bioresource Technology, 123, 581-91.

[37] Wang, C., Wang, F., Yang, Q., Liang, R. 2009. Thermogravimetric studies of the

behavior of wheat straw with added coal during combustion. Biomass and Bioenergy,

33(1), 50-6.

[38] Tang, Y., Ma, X., Lai, Z. 2011. Thermogravimetric analysis of the combustion of

microalgae and microalgae blended with waste in N2/O2 and CO2/O2 atmospheres.




3189-96.

[40] Kebelmann, K., Hornung, A., Karsten, U., Griffiths, G. 2013. Intermediate

pyrolysis and product identification by TGA and Py-GC/MS of green microalgae and

their extracted protein and lipid components. Biomass and Bioenergy, 49, 38-48.

[41] Ross, A.B., Jones, J.M., Kubacki, M.L., Bridgeman, T. 2008. Classification of

macroalgae as fuel and its thermochemical behaviour. Bioresource Technology,

99(14), 6494-504.

[42] Shuping, Z., Yulong, W., Mingde, Y., Chun, L., Junmao, T. 2010. Pyrolysis

characteristics and kinetics of the marine microalgae Dunaliella tertiolecta using

thermogravimetric analyzer. Bioresource Technology, 101(1), 359-65.

[43] Bridgeman, T.G., Darvell, L.I., Jones, J.M., Williams, P.T., Fahmi, R.,

Bridgwater, A.V., Barraclough, T., Shield, I., Yates, N., Thain, S.C., Donnison, I.S.

2007. Influence of particle size on the analytical and chemical properties of two

energy crops. Fuel, 86(1-2), 60-72.

Chapter 3

205

[44] Li, D., Chen, L., Chen, S., Zhang, X., Chen, F., Ye, N. 2012. Comparative

evaluation of the pyrolytic and kinetic characteristics of a macroalga (Sargassum

thunbergii) and a freshwater plant (Potamogeton crispus). Fuel, 96, 185-91.

[45] Wang, S., Jiang, X.M., Han, X.X., Liu, J.G. 2009. Combustion Characteristics of

Seaweed Biomass. 1. Combustion Characteristics of Enteromorpha clathrata and

Sargassum natans. Energy & Fuels, 23(10), 5173-8.

[46] Kok, M.V., Özgür, E. 2013. Thermal analysis and kinetics of biomass samples.

Fuel Processing Technology, 106(0), 739-43.

[47] Yu, L.J., Wang, S., Jiang, X.M., Wang, N., Zhang, C.Q. 2008. Thermal analysis

studies on combustion characteristics of seaweed. Journal of Thermal Analysis and

Calorimetry, 93(2), 611-7.

[48] Wang, S., Jiang, X.M., Wang, N., Yu, L.J., Li, Z., He, P.M. 2007. Research on

pyrolysis characteristics of seaweed. Energy and Fuels, 21(6), 3723-9.



Application to nannochloropsis sp. Thermochimica Acta, 484(1-2), 41-7.



and Bioenergy, 35(7), 3199-207.



Application to nannochloropsis sp. Thermochimica Acta, 484(1–2), 41-7.

[52] Bae, Y.J., Ryu, C., Jeon, J.K., Park, J., Suh, D.J., Suh, Y.W., Chang, D., Park,

Y.K. 2011. The characteristics of bio-oil produced from the pyrolysis of three marine

macroalgae. Bioresource Technology, 102(3), 3512-20.

[53] Li, D., Chen, L., Zhang, X., Ye, N., Xing, F. 2011. Pyrolytic characteristics and

kinetic studies of three kinds of red algae. Biomass and Bioenergy, 35(5), 1765-72.

[54] Jones, J.M., Harding, A.W., Brown, S.D., Thomas, K.M. 1995. Detection of

reactive intermediate nitrogen and sulfur species in the combustion of carbons that are

models for coal chars. Carbon, 33(6), 833-43.

Chapter 3

206

[55] Tapasvi, D., Khalil, R., Várhegyi, G., Skreiberg, O., Tran, K.Q., Grønli, M.

2013. Kinetic behavior of torrefied biomass in an oxidative environment. Energy and

Fuels, 27(2), 1050-60.

[56] Yang, H., Yan, R., Chen, H., Zheng, C., Lee, D.H., Liang, D.T. 2006. In-depth


cellulose and lignin. Energy and Fuels, 20(1), 388-93.

[57] Leite, B.S., Andreuccetti, M.T., Leite, S.A.F., d'Angelo, J.V.H. 2012. TG and

DSC analyses of eucalyptus black liquor as alternative methods to estimate solids

content. Journal of Thermal Analysis and Calorimetry, 1-6.

[58] Fang, M.X., Shen, D.K., Li, Y.X., Yu, C.J., Luo, Z.Y., Cen, K.F. 2006. Kinetic

study on pyrolysis and combustion of wood under different oxygen concentrations by

using TG-FTIR analysis. Journal of Analytical and Applied Pyrolysis, 77(1), 22-7.

[59] Miranda, T., Román, S., Montero, I., Nogales-Delgado, S., Arranz, J.I., Rojas,

C.V., González, J.F. 2012. Study of the emissions and kinetic parameters during

combustion of grape pomace: Dilution as an effective way to reduce pollution. Fuel

Processing Technology, 103(0), 160-5.

[60] Hansson, K.-M., Samuelsson, J., Tullin, C., Åmand, L.-E. 2004. Formation of

HNCO, HCN, and NH3 from the pyrolysis of bark and nitrogen-containing model

compounds. Combustion and Flame, 137(3), 265-77.

[61] Fahmi, R., Bridgwater, A.V., Darvell, L.I., Jones, J.M., Yates, N., Thain, S.,

Donnison, I.S. 2007. The effect of alkali metals on combustion and pyrolysis of

Lolium and Festuca grasses, switchgrass and willow. Fuel, 86(10-11), 1560-9.

Chapter 4:

GASIFICATION OF LIGNOCELLULOSIC BIOMASS CHAR OBTAINED FROM

PYROLYSIS: KINETIC AND EVOLVED

GAS ANALYSES

The pyrolysis and gasification process of three types of

lignocellulosic biomass (Eucalyptus wood, fir wood and pine bark)

and biomass main components (cellulose, xylan and lignin) were

studied by thermogravimetric-mass spectrometric analysis (TGA-

MS). Pyrolysis was used to obtain a solid fuel (char) that was later

gasified with steam. The reactivity profiles of the studied biomass

samples showed a clear catalytic effect at high conversion values

which was directly correlated with their ash composition.

Gasification apparent rates were obtained by a preliminary

kinetic analysis. Three standard models were used to reproduce the

gasification process. Only cellulose and pine bark samples

obtained an accurate fitting being attributed to their low ash

Chapter 4

209

content and subsequently low catalytic activity. A semi-empirical

model, based on the catalytic effect observed, was proposed which

highly improved the obtained fitting. H2, CO and CO2 were the

main products obtained. Furthermore, secondary products such as

CH4 pointed out the existence of methanation reactions. NOX were

also observed indicating that part of the initial nitrogen in the

sample was retained in the char after the pyrolysis process.

4.1. INTRODUCTION

The production of clean and sustainable fuels arethe main challenges to tackle

upcoming energy crises and global warming [1]. Among all renewable energy sources,

biomass fuels are gaining particular attention as a potential alternative to increase

energy independence on fossil fuels and reduce environmental pollution [2, 3].

Thermochemical conversion of biomass is the most promising route for biomass

utilization. These processes mainly include the direct combustion, to generate heat and

electricity, pyrolysis and gasification, to produce liquid and gaseous fuels which are

suitable for feeding efficient gas engines and gas turbines [4, 5].Pyrolysis plays an

important role in these processes,being the first chemical step in both gasification and

combustion processes. Generally, pyrolysis can be considered as a two-stage process

involving the devolatilization of biomass and the slow heterogeneous conversion of

char [5]. The char generated during pyrolysis is a high energy-density solid fuel

suitable for combustion and gasification processes.

Gasification is of special interest due to the fact that it is compatible with new

applications in the area of biomass conversion coal to liquid and superior

environmental performance especially with regard to CO2 capture and sulfur removal.

Furthermore, it is economical in a wide range of capacities (5 kWe onwards) [1, 6].

Gasificationcan be defined as the conversion of biomass to a gaseous fuel by heating

in a gasification medium such as air, oxygen or steam involving a complex set of

reactions [7, 8]. Char gasification is an important step during thermochemical

Chapter 4

210

conversion of lignocellulosic biomass because it often represents the rate-controlling

phenomenon in the gasifier[9]. The progress of char gasification is a function of

several factors such as particle size, porosity, gasifying agent chemical composition,

gasifying agent partial pressure, reactor temperature, pore structure, number of active

sites and ash content, among others[10].

Thermogravimetric analysis (TGA) has been commonly used for the study of the

thermochemical conversion of biomass [11]. Compared to pyrolysis and combustion,

there are few works reported in literature concerning the gasification of biomass. Most

works have been focused on the study of coals. In this regard,Shabbar et al. [1]

performed a thermodynamic analysis of bituminous coals. Furthermore, Tay et al. [7]

studied the effect of different gasifying agents on the char structure of Victorian

brown coal during the gasification. On the other hand, few studies have been reported

on the gasification of biomass. Mohammed et al. [12]evaluated the thermal

characteristics and kinetics of empty fruit bunches. Additionally, different authors

have investigated the carbon dioxide gasification of biomass chars [13, 14].

During the process of thermochemical conversion of biomass, the composition of

the gas emissions should be determined before industrial application. Complementary

techniques to TGA must be used in order to obtain qualitative information of biomass

transformation during the analyses. Very few studies have been found in literature

concerning gas evolution from biomass gasification. Yang et al. [15] studied the steam

gasification of tobacco by TGA coupled with gas chromatography (GC). Furthermore,

Yoon et al. [16] used TGA-GC to perform an kinetic analysis of thewoody biomass

gasification. In this sense, the use of TGA coupled with mass spectrometry (TGA-MS)

can give a deeper insight of the gasification process being able to afford real-time and

sensitive detection of evolved gases during the thermal analysis [17] .

The aim of this work was to study the pyrolysis and gasification of different types

of lignocellulosic biomass (fir wood, eucalyptus wood and pine bark) and their main

components (cellulose, hemicellulose and lignin) by means of TGA. The pyrolysis of

biomass samples was carried out to obtain a solid fuel (char) which was later gasified

Chapter 4

211

using steam. In addition, a preliminary kinetic analysis of the gasification process was

performed in order to obtain the apparent gasification rates. Finally, the gases released

during the gasification process were analyzed by MS.

4.2. EXPERIMENTAL

4.2.1. Materials

Cellulose, xylan and lignin were purchased from Sigma-Aldrich. Xylanwas used as

representative of that of the hemicellulose. These chemicals are as follow: cellulose

(microcrystaline cellulose with 50 µm average particle size), lignin (alkali lignin in

brown powder form with 50 µm average particle size) and xylan (xylan processed

from beechwood with 100 µm average particle size). The selected terrestrial

biomasses (fir wood, eucalyptus wood and pine bark) were taken from the region of

Castilla-La Mancha (Spain) on the basis of a preliminary analysis [18]. These samples

were dried in an oven for 5 h, milled and sieved to an average particle size between

100-150 µm.

The proximate analysis, ultimate analysis and composition of biomass samples are

shown in Table 1. The metal content in samples was determined by Inductively

Coupled Plasma Spectrometry (ICP) (Table 1). The content of hemicellulose, lignin

and xylan inlignocellulosic biomass samples was calculated according to the method

reported elsewhere [19] (Table 1).

Chapter 4

212

Table 1.-Characterization of lignocellulosic biomass samples (Cellulose, xylan, lignin, fir wood, eucalyptus wood and pine bark)

UltimateAnalysis (wt. %) ProximateAnalysis (wt. %) Biomass Composition (wt. %)db,ab

C H N S O Moisture

(%) Ash (%)

Volatilematter (%)

Fixedcarbon (%)

Cellulose (%)

Lignin (%)

Hemicellulose (%)

Extractives (%)

Cellulose 42.18 6.15 0.01 0.06 51.61 3.0 0.8 90.7 6.0 100 - - -

Lignin 62.09 5.88 0.51 0.54 30.98 1.1 3.7 55.8 39.3 - 100 - -

Xylan 38.41 6.18 0.01 0.11 55.30 6.4 2.8 71.6 19.2 - - 100 -

Eucalyptus Wood

41.62 4.88 0.38 0.03 53.09

2.6 6.8 73.8 16.8

52 17 24 7

Fir Wood 50.12 6.14 0.44 0.00 43.45 2.6 3.4 74.4 19.5 38 24 30 8

Pine Bark 52.71 5.52 0.01 0.08 41.70 4.4 2.7 61.6 31.3 13 31 37 19


Al Ca Cr Cu Fe K Mg Na Ni Pb P Hg V Si Ti

Cellulose 367 2711 16 66 106 575 255 1476 980 37 6869 411 10 237816 84

Lignin 500 868 13 48 126 1069 219 7197 758 35 6100 350 36 181504 150

Xylan 213 4343 7 77 77 456 124 13828 382 17 3326 184 - 65856 81

Eucalyptus Wood

43 4116 16 131 33 5078 1062 1431 51 47 7819 408 9 247228 18

Fir Wood 557 10921 22 71 717 1880 1774 1807 27 41 8608 492 22 353166 62

Pine Bark 946 2726 19 74 385 1254 776 2764 463 73 7360 524 26 474344 70

db - dry basis, ab- ash free basis

Chapter 4

213


4.2.2.1. Thermogravimetric analysis for the combustion process

The pyrolysis and gasification of biomass components was carried out in a TGA

apparatus (TGA-DSC 1, METTLER TOLEDO). The experimental setup used for the

gasification experiments was described in a previous study[20]. Steam was generated

in a bubbler system.Ar was bubbled through degassed water heated to 33 ºC.

Assuming the Ar-H2O mixture was saturated, a current with approximately 5% Vol. of

H2O in Ar was obtained. The pyrolysis of the sample was carried out by preheating the

sample at 105 ºC and then kept at 105 ºC for 10 min to remove the moisture content.

Subsequently, the sample was heated from 105 to 1000 ºC at 40 ºC/min under a 200

Nml/min of Ar. The temperature was kept at 1000 ºC for 10 min to ensure the

completion of the pyrolysis reaction. The sample was then cooled down to the

gasification temperature (900 ºC). The gasification step was carried outunder

isothermal conditions until the entire char was consumed. Previous studies were

carried out according to Sanchez-Silva et al. in order to avoid the effects of heat and

mass transfer limitations[20]. In this sense, initial sample weight was kept at 20 mg,

the particle size was kept in the 100-150 µm range and a constant flow rate of 200 and

50Nml/min were used for pyrolysis and gasification experiments, respectively.

4.2.2.2. TGA-MS analysis of the Gaseous Products


carried out in a thermogravimetric analyzer (TGA-DSC 1; METTLER TOLEDO)




circumvent condensation of exhausting gases. In order to identify ions with m/z in the

range 0-300, a preliminary broad scan was performed at a heating rate of 40 ºC/min.

Although a quantitative analysis was not performed in this work, a comparison of

the intensity peak areas between different samples (semiqualitative analysis) was

Chapter 4

214

carried out by using a normalization procedure. The ion intensities were normalized to

the total ion current to eliminate systematic instrumental errors caused by the

fluctuation of carrier gas flow, the shift in the sensitivity of the mass spectrometer

method and the sample weight. The evolved gas peak areas are useful for comparing

relative amounts of products from different samples [21].

4.2.2.3. Kinetic analysis

In this work, the char gasification was considered as an overall reaction, and a

general kinetic expression can be written as follows [22]:

�� = �(�, �) ∙ �(�) (1)

where k is the apparent gasification reaction rate, which includes the effect of

temperature, T, and the effect of the gasifying agent concentration, Pw, and f(α)

described the changes in the physical or chemical properties of the sample as the

gasification proceeds[22].

Three different modelswere used in this work, the volumetric model (VM) (Eq. 2), the

shrinking core model (SCM) (Eq. 3) and the random pore model (RPM) (Eq. 4):

�(�) = (1 − �) (2)

�(�) = (1 − �)�� (3)

�(�) = (1 − �) ∙ �1 − � ∙ ln(1 − �) (4)

whereα is the degree of conversion and Ψ is a parameter related with the initial pore

structure of the sample (α=0).

For the parameter estimation a VBA-Excel application was developed to solve this

model [23-25]. The Runge-Kutta-Fehlberg methodwas used in the evaluation of the

set of ordinary differential equations. Furthermore, the statistical significance of the

estimated parameters based on the F-test and t-test was performed according to the

procedure described elsewhere [18].

Chapter 4

215


4.3.1. Thermogravimetric analysis

In this work, the pyrolysis of biomass samples was carried out to obtain a solid fuel

(char) which was later gasified using steam. The pyrolysis of biomass samples was

studied previously and described elsewhere [18]. Thus, a comprehensive evaluation of

the pyrolysis process of lignocellulosic biomass was not the objective of the present

study. However, some remarks of this process are described in order to get a better

understanding of the char formation mechanism.

4.3.1.1. Pyrolysis and gasification of lignocellulosic biomass main components

Figures 1 shows the TGA/DTG profiles for the pyrolysis and gasification of

biomass main components (cellulose, xylan and lignin). Table 2 summarizes the most

relevant pyrolysis characteristics for all biomass samples (biomass main components

and lignocellulosic biomass).The pyrolysis of biomass main components took place

between 200 and 700 ºC as it can be seen from their DTG profile (Figure 1.b).Xylan

sample was the first one to decompose showing two peaks at 262 and 306 ºC. On the

other hand, cellulose sample decomposition took place in one stage between 220 and

500 ºC. The cellulose showed the highest weight loss rate (88 wt.%/min) beingthe

sample which released thehighest amount of volatiles. Finally, lignin decomposed

over the whole temperature range (215-700 ºC). Lignin thermal decomposition is a

slow carbonization process producing the highest amount of char (43 wt.%) compared

to that formed in xylan (28 wt.%) and cellulose (8 wt.%) pyrolysis.

Figure 1.c shows the DTG profiles for the steam gasification of the char produced

from biomass main components pyrolysis. It can be seen that the gasification of

biomass chars started as soon as the gasifying agent reached the surface of the char

particle. Lignin and xylan samples produced the most reactive char and it took around

26 and 29 min to be totally gasified whereas the char produced from cellulose sample

was decomposed at a lower rate (~120 min). These facts agreed well with those

Chapter 4

216

0 25 50 75 100 125 1500

20

40

60

80

100

0 3 6 9 12 150

20

40

60

80

0 20 40 60 80 100 1200

1

2

3

4

5

6Pyrolysis Steam Gasification

Pyrolysis

Wei

ght (

wt.%

)

Time (min)

Cellulose Lignin Xylan

Steam Gasification

200

400

600

800

1000

1200

Temperature

Tem

pera

ture

(ºC

)

Wei

ght l

oss

rate

(w

t. %

/ min

)

Time (min) Wei

ght l

oss

rate

(w

t. %

/ min

)

Time (min)

900

1000

1100

Tem

pera

ture

(ºC

)

200 300 400 500 600 700 Temperature (ºC)

DTG DTG

reported by Lv et al. [9] who observed that the gasification of cellulose under dynamic

conditions took place at higher temperatures and lower rates than for lignin.

Figure 1.-Thermogravimetric (TG) and derivothermogravimetric (DTG) curves for the

pyrolysis and gasification processes of biomass main components (cellulose, xylan and lignin).

a) TGA curves for pyrolysis and gasification; b) DTG curves for pyrolysis; c) DTG curves for

gasification

Chapter 4

217

Table 2.- Pyrolysis characteristics for cellulose, lignin, xylan, fir wood, eucalyptus wood and pine bark at 40 ºC/min

Primary components of biomass Lignocellulosic biomass

Cellulose Xylan Lignin Fir wood Eucalyptus wood Pine bark

1st peak 1st peak

2nd peak

1st peak Sh* 1st peak*

Tail * Sh* 1st peak*

Tail * Sh* 1st peak*

Tail *

To (ºC)* 298 208 215 172 165 180

Tp ºC)* 373 262 306 376 327 368 453 304 358 510 293 364 519

(dw/dt)max*

(dwt.%/min)

86.8 52.8 32.3

20.3 15.1 24.2 4.4 15.8 27.9 2.9 11.1 15.5 3.1

Char (wt.%) 8.1 25.3 41.6 25.4 23.7 35.1

*Sh: Shoulder in the DTG curve; Tail: Tail in the DTG curve; To: Initial temperature; Tp: Peak temperature;

(dw/dt)max: Maximum weight loss rate

Chapter 4

218

0 25 50 75 100 125 1500

20

40

60

80

100

0 3 6 9 12 150

5

10

15

20

25

30

0 20 40 60 80 100 1200

1

2

3

4

5

6

Pyrolysis Steam Gasification

Pyrolysis

Wei

ght (

wt.%

)

Time (min)

Pine bark Fir wood Eucalyptus wood

Steam Gasification

200

400

600

800

1000

1200

Temperature

Tem

pera

ture

(ºC

)

Wei

ght l

oss

rate

(w

t. %

/ min

)

Time (min) Wei

ght l

oss

rate

(w

t. %

/ min

)

Time (min)

900

1000

1100

Tem

pera

ture

(ºC

)

200 300 400 500 600 700 Temperature (ºC)

DTG DTG

4. 3.1.2. Pyrolysis and gasification of lignocellulosic biomass

Figures 2 shows the TGA/DTG profiles for the pyrolysis and gasification of

lignocellulosic biomass (fir wood, eucalyptus wood and pine bark). The shape of the

DTG curve was similar for all samples.

Figure 2.-Thermogravimetric (TG) and derivothermogravimetric (DTG) curves for the pyrolys

and gasification processes of lignocellulosic biomass (eucalyptus wood, fir wood and pine

bark). a) TGA curves for pyrolysis and gasification; b) DTG curves for pyrolysis; c) DTG

curves for gasification

Chapter 4

219

Firstly, a shoulder can be observed at temperatures around 300 ºC which is

attributed to hemicellulose decomposition. This shoulder was more sharped for pine

bark sample which is in agreement with its high hemicellulose content (Table 1).

Secondly, the maximum weight loss rate was observed at358, 364 and 368 ºC for

eucalyptus wood, pine bark and fir wood samples, respectively. This stage is ascribed

to cellulose decomposition. Eucalyptus wood sample showed the highest weight loss

rate (27 wt.%/min) compared to that for fir wood (24 wt.%/min) and pine bark (16

wt.%/min) samples due to its high cellulose content. Finally, the maximum peak was

followed by a wide tail which is essentially related to the lignin decomposition leading

to char formation [5]. As expected, the pine bark sample, that was the sample with the

highest lignin content, produced a higher amount of char (35 wt.%). On the other

hand, eucalyptus and fir wood samples, with a similar lignin content, generated a

similar char yield (25 and 24 wt.%, respectively). According to these evidences, the

mechanism of lignocellulosic biomass pyrolysis can be divided into two main stages:

devolatilization of raw biomass, where hemicellulose and cellulose mainly

decompose, and the slow carbonization of the remaining biomass, associated to lignin

decomposition turning into the production of the final char.

Figure 2.c shows the DTG profiles for the steam gasification of the char produced

from lignocellulosic biomass samples pyrolysis. Unlike their pyrolysis behavior, the

gasification of lignocellulosic biomass chars could not be described according to their

initial chemical composition. The eucalyptus wood sample, with the highest cellulose

content, was the one that needed the least time to be gasified. Additionally, the char

produced from pine bark sample pyrolysis, which was expected to decompose at lower

times than eucalyptus and fir wood samplesdue to its high hemicellulose and lignin

content, was the one thattook longer to be gasified. These results suggested that the

char formation from lignocellulosic biomass may be affected by the presence of other

components in the complex matrix of the wood. Therefore, the process cannot be

explained by considering the proportional interactionsbetween their main components.

Chapter 4

220

In this regard, the morphology of the formed char is a factor usually employed to

compare the reactivity of different chars [15].

Char conversion is more complicated than solid devolatilization as it is a

heterogeneous process where the surface is the location of the chemical reactions [5].

It is recognized that the heterogeneous rates of char conversion are determined by the

fundamental components, represented by surface area, surface accesibility, carbon

active sites and catalytic active sites created by indigenous or added inorganic matter,

and the local gaseous reactant concentration. Consequently, the reactive depends on

three chief characteristics of the sample: chemical structure, porosity and inorganic

constituents [5].Furthermore, the concentration of the gasifying agent also plays an

important role in the process. The two first factors might be influenced by the initial

cellulose, hemicellulose and lignin content in lignocellulosic samples. According to

Lv et al.[26], biomass rich in lignin component produced a high surface area and

porous charwhich makes easier the diffusion of the reactive agent turning intohigh

gasification rates. On the contrary, biomass with a high cellulose content produced a

fibrous structure char, lowering the char reactivity. However, this trend was not found

in the experimental results obtained in this study. Therefore, assuming that the

reactive gas concentration was kept constant in all experiments, these results pointed

out that the catalytic activity of the indigenous inorganic matter in the biomass played

a significant role in the gasification of the studied biomass samples.These results agree

well with those reported byXie et al. [27]who observed that specific surface areas and

porosities of lignin and cellulose char prepared at temperatures higher than 700ºC did

not have a meaningful role in the oxidative mass loss process. These results are

discussed in more detailed in the next sub-section.

4.3.1.3. Char Reactivity

The reactivity of char is an important parameter when evaluating the gasification

process. Several definitions were used to evaluate the char reactivity, however the

more extended one refers to the intrinsic reactivity (Ri) and it can be described as

follows [5, 22, 28, 29]:

Chapter 4

221

�� = −1 �� ∙ �� = 11 − �� ∙ �� (10)

wherexiand wiare the conversion and weight of charat any time, respectively. The

reactivity is dependent on the temperature and gas composition and varies with the

conversion degree [5, 30]. Thus, a representative value of reactivity must be presented

in order to make reliable comparisons. In this work, the reactivity at 50 % char

conversion is taken to be representative (R50) 28, 29][30, 31]. R50 values and the time

to achieve 100 % char conversion are summarized in Table 3. As aforementioned the

reactivity of biomass main components was ranked as: Xylan> Lignin>Cellulose. On

the other hand, lignocellulosic biomass samples was: Eucalyptus wood > Fir wood >

Pine bark.Thus, this order is not correlated with the biomass samples chemical

composition.

The gasification rate (ri) is also used to describe the gasification reaction and was

calculated by Eq. (11) [32]:

�� =��

(11)

Figure 3 shows the typical reactivity and gasification rates versus conversion plots on

a comparative basis for biomass main components (cellulose, xylan and lignin) and

lignocellulosic biomass (fir wood, eucalyptus wood and pine bark) samples. It can be

observed that reactivity increase slowly up to conversion values of 0.8. However, for

xylan, lignin, eucalyptus and fir wood samples a sudden rise of reactivity took place

beyond 0.8 conversion, whereas pine bark and cellulose samples showed a lower

rise.This behavior can be explained by a high activity of the inorganic matter

contained in biomass samples[33]. As the gasification proceeds the carbon material is

consumed and the metal to carbon ratios increase which strengthen the catalytic effect

[33, 34].Furthermore, gasification rates versus conversion plots also corroborated this

fact. Pine bark and cellulose samples showed a decreasing trend and no maximum was

obtained whereas xylan, lignin, eucalyptus and fir wood samples showed a maximum.

This fact points out that the catalytic activity of indigenous inorganic matter

Chapter 4

222

increasethe gasification rate of xylan, lignin, eucalyptus and fir wood samples. This

reactivity profiles are similar to those reported by Blasi et al.[35] for the air

gasification of wheat straw, olive husks and grape residues.

Figure 3.-Reactivity versus conversion profiles for a) lignocellulosic biomass main

components (cellulose, xylan and lignin); b) lignocellulosic biomass (eucalyptus wood, fir

wood and pine bark).

Numerical indices such as the alkali index (A.I.) have been defined to describe the

catalytic efficiency of the overall influence of catalytically active species within the

ash [36]. This index is calculated as the ratio of the sum of the fraction of the basic

compounds (catalytic nature) in the ash (CaO, MgO, K2O, Na2O and Fe2O3) to the

fraction of the acidic compounds (non-catalytic nature) (Al2O3 and SiO2):

. " = #ℎ(��.%) ∙ ('()*+�)*,-)*.(�)*/0�)�)(12�)�*3�)�)

(12)

0.0

1.5

3.0

4.5

6.0

7.5

0.0 0.2 0.4 0.6 0.8 1.00.0

0.5

1.0

1.5

0.0

1.5

3.0

4.5

6.0

7.5

0.0 0.2 0.4 0.6 0.8 1.00.0

0.5

1.0

1.5

2.0

2.5

Gas

ifica

tion

rate

(1/

min

)

Rea

ctiv

ity (

1/m

in)

Conversion (X)

Cellulose Xylan Lignin

Gas

ifica

tion

rate

(1/

min

)

Rea

ctiv

ity (

1/m

in)

Conversion (X)

Eucalyptus wood Fir wood Pine bark

Chapter 4

223

0 15 30 45 60 75 90 105 1200.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 300.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 30 350.0

0.2

0.4

0.6

0.8

1.0

0 10 20 300.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 300.0

0.2

0.4

0.6

0.8

1.0

0 15 30 45 60 75 90 105 1200.0

0.2

0.4

0.6

0.8

1.0

Con

vers

ion

(X)

Cellulose

Con

vers

ion

(X)

Time (min)

Lignin

Con

vers

ion

(X)

Xylan

Fir wood

Time (min)

Eucalyptus wood

Experimental VM SCM RPM

Pine bark

Eucalyptus wood showed higher A.I than fir wood and pine bark (Table 3) which

contributed to its higher reactivity. These results agreed well with the results obtained

for the gasification of different types of coal [29, 36-38]. Thus, the gasification of

lignocellulosic biomass char is more influenced by the mineral matter in the ash than

their initial chemical composition.

4.3.2. Gasification kinetic analyses

Figure 4 shows the experimental fitting to the three models used in this work.

Figure 4.-Comparison of the proposed models (volumetric model (VM), shrinking core model

(SCM) and random pore model (RPM)) with experimental data.

Chapter 4

224

Table 4 summarized the obtained gasification constants and the statistical

significance of the models. In general, it can be observed that volumetric model (VM)

obtained the worst fitting to the experimental curves, whereas shrinking core model

(SCM) and random pore model (RPM) achieved a better fit. VM assumes a

homogeneous reaction throughout the particle and a linearly decreasing reaction

surface area with conversion [22, 28]. On the other hand, SCM and RPM describe the

evolution of the solid structure with conversion [34]. SCM assumes that a porous

particle consists of an assembly of uniform nonporous grains and the reaction takes

place on the surface of the grains assuming spherical shape of the porous. This model

predicts a monotonically decreasing reaction rate and surface area because the surface

area of each grain is receding during the gasification which agrees with profiles

described for cellulose and pine bark (Figure 3). Finally, the RPM considers the

overlapping of pore surfaces which reduces the area available for reaction[28]. This

model is able to predict a maximum for the reactivity as the reaction proceeds, as it

considers the competing effects of pore growth during the initial stages of gasification,

and the destruction of the pores due to the coalescence of neighboring pores.As it can

be seen from Figure 4, none of the models accurately predicted the biomass samples

behavior for conversion values greater than 0.8 except for cellulose and lignin

samples. Furthermore, it is clearly observed how these models clearly underpredicted

the conversion values. These results agree well with literature [22, 28, 34]. This

behavior is mainly due to the fact that these models fail to predict the catalytic activity

of the ash and are only valid under the chemical controlled regime [22]. The good

fitting of cellulose and pine bark sample to the proposed models, especially SCM,

corroborated the low activity behavior of their indigenous inorganic matter.Anyway,

curves predicted with SCM models obtained a slightly lower error than SCM for the

gasification of the biomass samples used in this work. In order to ensure the reliability

of the proposed models, the discrimination of kinetic parameters was done applying

the F-test and the t-test at the 95% confidence level [18] (Table 4). In terms of

statistical results, F-test considered the regression to be suitable in all cases since the

corresponding values to the Fc/Ftest ratio were larger than one. The t-test was also used

Chapter 4

225

for evaluating each parameter in the model. The values of tc/t-test ratio were also

larger than one, showing the statistical significance of the proposed models and their

corresponding parameters.

A semi-empirical model was proposed in order to obtain a model that accurately

reproduce the gasification process of biomass samples (xylan, lignin, eucalyptus

wood and pine bark). An expression representing the activity of biomass ashes at high

conversion values was added to the SCM. SCM was chosen due to two facts. Firstly,

the RPM includes an additional adjustable parameter (Ψ) which is difficult to be

measured and secondly, the error obtained was not too different than that obtained for

RPM.The proposed model include two parameters, an activation constant (ka) and

activation order (na) and is described as follows:

�� = �(�, �) ∙ (1 − �)

45� + �( ∙ �78 (13)

Chapter 4

226

Table 4.- Estimated parameters for the proposed models (VM, SCM and RPM) and their statistical analysis

Biomass samples Model Parameters tc ttest Fc Ftest Error (%)

Cellulose

VM 1.71 76 24723 3.84 6.8

SCM K (min -1)(·102) 1.48 55059 1.96 175522 3.84 1.1

RPM 1.21 69809 32677 3.00 5.1

Ψ 1.9 13991

Lignin

VM 5.58 2251 1602 3.84 31.7

SCM K (min -1)(·102) 4.48 98519 1.96 7025 3.84 22.5

RPM 2.14 38039 20194 3.00 8.9

Ψ 20.7 30254

Xylan

VM 8.81 2265 36796 3.84 19.0

SCM K (min -1)(·102) 7.32 89 1.96 110675 3.84 10.5

RPM 2.49 204 217283 3.00 8.8

Ψ 33.5 25350

Fir wood VM 3.31 60942 1102 3.84 17.7

SCM K (min -1)(·102) 3.98 10221 1.96 9177 3.84 11.8

RPM 2.29 20953 32749 3.00 8.8

Ψ 7.3 15236

Eucalyptus wood VM 6.26 4715 6744 3.84 18.9

SCM K (min -1)(·102) 7.23 64 1.96 44861 3.84 12.9

RPM 3.79 4458 73100 3.00 9.3

Ψ 10.2 27593

Pine bark

VM 2.08 1343 11357 3.84 13.7

SCM K (min -1)(·102) 1.67 10011 1.96 132563 3.84 2.6

RPM 1.46 301 32128 3.00 1.2

Ψ 1.8 294

Chapter 4

227

0 5 10 15 20 25 300.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 200.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 30 350.0

0.2

0.4

0.6

0.8

1.00 5 10 15 20

0.0

0.2

0.4

0.6

0.8

1.0

Con

vers

ion

(X)

Lignin

Time (min)

Eucalyptus wood

Con

vers

ion

(X)

Time (min)

Fir wood


Xylan

Figure 5 shows the experimental versus theoretical curves obtained by the

proposed model (Eq. (13)). It can be observed that the additional term satisfactorily

predicted the gasification process over the whole conversion interval. Furthermore, the

errors obtained were all below ± 1% (Table 5). Thus, the addition of this simple term

seemed to be sufficient for obtaining excellent predictions of the gasification rate.

Additionally, the calculated parameters were statistically significant. Correlations of

the proposed parameter (na) were searched with physical properties of the studied

biomass samples in order to explain differences of the reaction rates between

biomasses in a similar way than that reported by Dupont et al., [34]. Assuming the

inorganic matter in the ash remains constant after the gasification process at the

reaction temperature, different parameters as the catalytic inorganic elements content

(Ca, K, Mg, Ca+K+Mg or the combination of them), A.I. and ash content were tested.

Figure 5.-Experimental results versus theoretical results obtained by the proposed semi-

empirical model.

Chapter 4

228

Table 5.- Estimated parameters for the proposed semi-empirical model and their statistical

analysis.

The parameter nashowed a really good linearity with K content with a correlation

coefficient higher than 0.97 (Figure 6). Thus, the proposed model can be generally

described as:

�� = �(�, �) ∙ (1 − �)

45� + �( ∙ �(9.4:;<'(=*5.;∙>9

?�) (14)

This correlationship of the proposed parameters with active catalytic species might

help to understand the role of the ashes in the gasification process. However, future

work must be carried out in order to understand the process from a phenomenological

point of view.

Biomass

samples

Parameter tc ttest Fc Ftest Error

(%)

Lignin k a 1.54·10-2 141 1.96 5282310 2.37 0.1

na 0.94 221

Xylan ka 3.55·10-2 100 1.96 37838 2.45 0.1

na 0.67 36

Fir wood ka 2.56·10-2 3702315 1.96 1431060 2.37 0.3

na 0.76 80

Eucalyptus

wood

ka 4.92·10-2 152 1.98 199429 2.45 0.7

na 0.43 13

Chapter 4

229

0 5 10 150.2

0.3

0.4

0.5

0.6

0.7

0.8

n a

[Ca]

Figure 6.-Correlation between the activation order (na) and the amount of Calcium of biomass

samples.

4.3.3. Gas evolution analyses

Figure 7 and 8 show the mass spectra for the main products obtained for the

gasification process of lignocellulosic biomass and biomass main components. Figure

9 schematizes the gas yield calculated by integrating the data measured by MS. H2,

CO and CO2 were the main products obtained (Figure 7). The evolution of these

products took place within the whole process.The MS profile of these compounds

correlate well with the reactivity one showingmaximums that are correlated with the

higher activity of the mineral content in the ashes.In general, the high amount of

mineral matter result in higher gas yields [16]. This way, eucalyptus wood, fir wood,

lignin and xylan samples showed the highest gas yields whereas pine bark and

cellulose samples had the lowest ones. H2 and CO were obtained in higher proportions

than CO2 for all samples but fir wood one. The high amount of H2 and CO pointed out

that char gasification reactions (C+H2O ↔ CO + H2; C+2H2O ↔ CO + 2H2)were

predominant. Furthermore, the low amount of CO2 obtained for some samples (xylan,

cellulose and pine bark) may indicate the existence of the Boudouard reaction (C +

CO2 ↔ 2CO). As abovementioned, the amount of CO was very low for the fir wood

sample. This fact, can be attributed to the existence of gas-phase reactions as the

water-gas shift reaction (CO + H2O ↔ CO2 + H2) that can be catalyzed by the high

calcium content of this sample.

Chapter 4

230

0 40 80 120

0.00

0.05

0.10

0.15

0.20

0.25

H2

CO

CO2CO2

H2

Time (min)

CO

Cellulose

0 20 40 60 80 100 120

0.0

0.1

0.2

0.3

H2

CO

CO2CO2

H2

Time (min)

Inte

nsity

(A

/(m

bar

mg)

)*10

-4

CO

Pine bark

0 10 20 30 40

0.0

0.5

1.0

1.5

CO2

CO2

H2

Time (min)

Inte

nsity

(A

/(m

bar

mg)

)*10

-4

CO

Fir wood

0 10 20 30 40

0.0

0.5

1.0

1.5

H2

CO

CO2CO2

H2

Time (min)

CO

Xylan

0 10 20 30 40 50

0.0

0.5

1.0

1.5

CO

H2

CO2

Lignin

CO2

H2

Time (min)

CO

0 10 20 30

0

1

2

H2

CO

CO2CO2

H2

Time (min)

Inte

nsity

(A

/(m

bar

mg)

)*10

-4

CO

Eucalyptus wood

Figure 7.-Gas evolution profile of H2, CO, CO2, H2O, NO and NO2 for the gasification process

of cellulose, xylan, lignin, pine bark, eucalyptus wood and fir wood samples.

Chapter 4

231

0 40 80 120

0.0

0.4

0.8Cellulose

NO

CH4

NO2

Time (min)0 20 40 60 80 100 120

0.00

0.05

0.10

Pine bark

COOH

CH4

NO2

Time (min)

Inte

nsity

(A

/(m

bar

mg)

)*10

-6

0 10 20 30 40 50 60

0.0

0.3

0.6

0.9

Fir wood

COOH

NO

CH4

NO2

C2H

2

Time (min)

Inte

nsity

(A

/(m

bar

mg)

)*10

-6

0 10 20 30 40

0

1

2

3

4

Xylan

COOH

NO

CH4

NO2

Time (min)

0 10 20 30 40 50 60

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

COOHNO

CH4

NO2

H2S

HSC

2H

2SO

2SO

Time (min)

Lignin

0 20 40

0.0

0.5

1.0

Eucalyptus wood

COOH

NO

CH4

NO2

C2H

2

Time (min)

Inte

nsity

(A

/(m

bar

mg)

)*10

-6

Figure 8.-Gas evolution profile of secondary products produced in the gasification process of

cellulose, xylan, lignin, pine bark, eucalyptus wood and fir wood samples.

Chapter 4

232

It can be observed that apart from H2, COand CO2, light hydrocarbons such as CH4

and C2H2were obtained in high proportions (Figure 8). Thus, secondary reactions as

methanation (C+H2 ↔ CH4) and thermal cracking (CnHm ↔ Cn-xHm-y + H2 + CH4 + C)

were taking place[39]. Eucalyptus wood gasification showed the highest yield of CH4.

This fact could be explained by the high potassium content of the sample which is

stated in literature as an active methanation catalyst [40].Furthermore, carboxylic

acids (COOH) were also obtained, pointing out the existence of these compounds in

the macromolecular structure of the produced char in a similar way than for lignites

and sub-bituminous coals [36]. Nitrogen oxides were also detected in all the samples.

On the other hand, sulfur compounds such as HS, H2S and SOx were only present in

the gasification process of lignin sample. The origin of nitrogen and sulfur compounds

is due to the dissociation of water at the char surface into hydrogen atom and a

hydroxyl radical which is an extremely active oxidizing agent [41].

Figure 9.- Gaseous product yields for the gasification process of cellulose, xylan, lignin, pine

bark, eucalyptus wood and fir wood samples. a) H2, CO and CO2; b) Secondary products.

Table 6 summarized the elemental analyses of chars produced from pyrolysis and

the final residue (ashes). It can be observed that after the pyrolysis process, N was

retained in the char for all samples whereas S was only detected in the char produced

from lignin sample pyrolysis. Finally, no C was found in the ashes, indicating that the

complete gasification of the samples.

CH4 C2H2 NO SH H2S C2H5O NO2 SO SO20.00

0.02

0.04

0.06 Cellulose Lignin Xylan Fir wood Eucalyptus wood Pine bark

SO2NO

2SOCOOHH

2SHSC

2H

2NO

CH4H2 CO CO2

0

1

2

3

4

CO2COH

2

Gas

yie

ld(A

min

/(m

bar

mg)

)·10

-3

Chapter 4

233

Table 6.-Elemental analysis of biomass char and ash

daf: dry ash

basis, *Oxygen

was calculated

by difference,

N/A: Not

available

4.4. CONCLUSIONS

Thermal characteristics and gas formation during the pyrolysis and gasfication of

eucalyptus wood, fir wood, pine bark and biomass main components (cellulose, xylan

and lignin) were analyzed by TGA-MS. The presence of indigenous inorganic matter

in the gasification process of biomass samples played an important role compared

withtheir initial chemical composition. The reactivity of biomass samples was

Ultimate analysis (wt.%)daf Char

C H N S O*

Cellulose 91.28 0.44 0.07 - 8.25

Lignin 68.02 0.49 0.88 0.16 30.44

Xylan 82.32 0.55 0.58 - 16.57

Fir wood 79.68 0.63 1.22 - 18.51

Eucalyptus wood 69.69 0.61 0.90 - 28.91

Pine bark 84.58 0.46 0.35 - 14.76 Ash

C H N S O*

Cellulose - - - - N/A

Lignin - - - - N/A

Xylan - - - - N/A

Fir wood - - - - N/A

Eucalyptus wood - - - - N/A

Pine bark - - - - N/A

Chapter 4

234

correlated with their alkali index and was ranked as follows: Xylan> lignin > cellulose

and Eucalyptus wood > fir wood > pine bark. The high relevance of inorganic matter

was proved by the inaccuracy of the results obtained by three standards models (VM,

SCM and RPM) which fail to predict the effect of catalytic active species. A semi-

empirical model was proposed in order to accurately model the gasification process.

The proposed model showed errors below 1 %. Furthermore, the models used in this

work were statistically validated. The high production of H2 and CO showed the

predominance of solid-gas reactions. On the other hand, gas phase reactions as water-

gas shift had a higher relevance in the gasification of fir wood due to its high calcium

content. Methanation reactions also took place especially for eucalyptus wood sample

and was correlated to the catalytic effect of potassium.

4.5. REFERENCES

[1] Shabbar S, Janajreh I. 2013. Thermodynamic equilibrium analysis of coal

gasification using Gibbs energy minimization method. Energy Conversion and

Management; 65: 755.

[2] Meng X, de Jong W, Badri FS, Benito P, Basile F, Verkooijen AHM. 2012.

Combustion study of partially gasified willow and DDGS chars using TG analysis and

COMSOL modeling. Biomass and Bioenergy; 39: 356.

[3] Seo DK, Lee SK, Kang MW, Hwang J, Yu T-U. 2010. Gasification reactivity

of biomass chars with CO2. Biomass and Bioenergy; 34(12): 1946.

[4] Ollero P, Serrera A, Arjona R, Alcantarilla S. 2003. The CO2 gasification

kinetics of olive residue. Biomass and Bioenergy; 24(2): 151.

[5] Di Blasi C. 2009. Combustion and gasification rates of lignocellulosic chars.

Progress in Energy and Combustion Science; 35(2): 121.

[6] Kirubakaran V, Sivaramakrishnan V, Nalini R, Sekar T, Premalatha M,

Subramanian P. 2009. A review on gasification of biomass. Renewable and

Sustainable Energy Reviews; 13(1): 179.

Chapter 4

235

[7] Tay HL, Kajitani S, Zhang S, Li CZ. 2013. Effects of gasifying agent on the

evolution of char structure during the gasification of Victorian brown coal. Fuel; 103:

22.

[8] McKendry P. 2002. Energy production from biomass (part 2): Conversion

technologies. Bioresource Technology; 83(1): 47.

[9] Umeki K, Moilanen A, Gómez-Barea A, Konttinen J. 2012. A model of

biomass char gasification describing the change in catalytic activity of ash. Chemical

Engineering Journal; 207-208: 616.

[10] Ahmed II, Gupta AK. 2011. Particle size, porosity and temperature effects on

char conversion. Applied Energy; 88(12): 4667.

[11] White JE, Catallo WJ, Legendre BL. 2011. Biomass pyrolysis kinetics: A


Analytical and Applied Pyrolysis; 91(1): 1.

[12] Mohammed MAA, Salmiaton A, Wan Azlina WAKG, Mohamad Amran MS.

2012. Gasification of oil palm empty fruit bunches: A characterization and kinetic

study. Bioresource Technology; 110: 628.

[13] Khalil R, Várhegyi G, Jäschke S, Grønli MG, Hustad J. 2009. CO2

gasification of biomass chars: A kinetic study. Energy and Fuels; 23(1): 94.

[14] Mani T, Mahinpey N, Murugan P. 2011. Reaction kinetics and mass transfer

studies of biomass char gasification with CO2. Chemical Engineering Science; 66(1):

36.

[15] Yang Y, Jin S, Lin Y, Huang S, Yang H. 2012. Catalytic gasification of

tobacco rob in steam-nitrogen mixture: Kinetic study and fuel gas analysis. Energy;

44(1): 509.

[16] Yoon HC, Pozivil P, Steinfeld A. 2012. Thermogravimetric pyrolysis and

gasification of lignocellulosic biomass and kinetic summative law for parallel

reactions with cellulose, xylan, and lignin. Energy and Fuels; 26(1): 357.

[17] Huang YF, Kuan WH, Chiueh PT, Lo SL. 2011. Pyrolysis of biomass by

thermal analysis-mass spectrometry (TA-MS). Bioresource Technology; 102(3): 3527.

Chapter 4

236

[18] Sanchez-Silva L, López-González D, Villaseñor J, Sánchez P, Valverde JL.

2012. Thermogravimetric-mass spectrometric analysis of lignocellulosic and marine

biomass pyrolysis. Bioresource Technology; 109: 163.

[19] Yang H, Yan R, Chen H, Zheng C, Lee DH, Liang DT. 2006. In-depth


cellulose and lignin. Energy and Fuels; 20(1): 388.

[20] Sanchez-Silva L, López-González D, Garcia-Minguillan AM, Valverde JL.

2013. Pyrolysis, combustion and gasification characteristics of Nannochloropsis

gaditana microalgae. Bioresource Technology; 130: 321.

[21] Jones JM, Harding AW, Brown SD, Thomas KM. 1995. Detection of reactive

intermediate nitrogen and sulfur species in the combustion of carbons that are models

for coal chars. Carbon; 33(6): 833.

[22] Fermoso J, Arias B, Pevida C, Plaza MG, Rubiera F, Pis JJ. 2008. Kinetic

models comparison for steam gasification of different nature fuel chars. Journal of

Thermal Analysis and Calorimetry; 91(3): 779.

[23] Sánchez-Silva L, Gutiérrez N, Romero A, Sánchez P, Valverde JL. 2012.

Pyrolysis and combustion kinetics of microcapsules containing carbon nanofibers by

thermal analysis-mass spectrometry. Journal of Analytical and Applied Pyrolysis; 94:

246.

[24] Jiménez-Borja C, Delgado B, Dorado F, Valverde JL. 2013. Experimental

data and kinetic modeling of the catalytic and electrochemically promoted CH4

oxidation over Pd catalyst-electrodes. Chemical Engineering Journal; 225: 315.

[25] Valverde JL, De Lucas A, Carmona M, González M, Rodríguez JF. 2004. A

generalized model for the measurement of effective diffusion coefficients of

heterovalent ions in ion exchangers by the zero-length column method. Chemical

Engineering Science; 59(1): 71.

[26] Lv D, Xu M, Liu X, Zhan Z, Li Z, Yao H. 2010. Effect of cellulose, lignin,

alkali and alkaline earth metallic species on biomass pyrolysis and gasification. Fuel

Processing Technology; 91(8): 903.

Chapter 4

237

[27] Xie X, Goodell B, Zhang D, Nagle DC, Qian Y, Peterson ML, et al. 2009.

Characterization of carbons derived from cellulose and lignin and their oxidative

behavior. Bioresource Technology; 100(5): 1797.

[28] Mandapati RN, Daggupati S, Mahajani SM, Aghalayam P, Sapru RK, Sharma

RK, et al. 2012. Experiments and kinetic modeling for CO2 gasification of indian coal

chars in the context of underground coal gasification. Industrial and Engineering

Chemistry Research; 51(46): 15041.

[29] Zhang L, Huang J, Fang Y, Wang Y. 2006. Gasification reactivity and kinetics

of typical Chinese anthracite chars with steam and CO2. Energy and Fuels; 20(3):

1201.

[30] Gómez-Barea A, Ollero P, Fernández-Baco C. 2006. Diffusional effects in


investigation. Energy and Fuels; 20(5): 2202.

[31] Mitsuoka K, Hayashi S, Amano H, Kayahara K, Sasaoaka E, Uddin MA.

2011. Gasification of woody biomass char with CO2: The catalytic effects of K and

Ca species on char gasification reactivity. Fuel Processing Technology; 92(1): 26.

[32] Huang Y, Yin X, Wu C, Wang C, Xie J, Zhou Z, et al. 2009. Effects of metal

catalysts on CO2 gasification reactivity of biomass char. Biotechnology Advances;

27(5): 568.

[33] Marquez-Montesinos F, Cordero T, Rodríguez-Mirasol J, Rodríguez JJ. 2002.

CO2 and steam gasification of a grapefruit skin char. Fuel; 81(4): 423.

[34] Dupont C, Nocquet T, Da Costa JA, Verne-Tournon C. 2011. Kinetic

modelling of steam gasification of various woody biomass chars: Influence of

inorganic elements. Bioresource Technology; 102(20): 9743.

[35] Di Blasi C, Buonanno F, Branca C. 1999. Reactivities of some biomass chars

in air. Carbon; 37(8): 1227.

[36] Hattingh BB, Everson RC, Neomagus HWJP, Bunt JR. 2011. Assessing the

catalytic effect of coal ash constituents on the CO 2 gasification rate of high ash,

South African coal. Fuel Processing Technology; 92(10): 2048.

Chapter 4

238

[37] Skodras G, Sakellaropoulos GP. 2002. Mineral matter effects in lignite

gasification. Fuel Processing Technology; 77-78: 151.

[38] Sakawa M, Sakurai Y, Hara Y. 1982. Influence of coal characteristics on CO2

gasification. Fuel; 61(8): 717.

[39] Widyawati M, Church TL, Florin NH, Harris AT. 2011. Hydrogen synthesis

from biomass pyrolysis with in situ carbon dioxide capture using calcium oxide.

International Journal of Hydrogen Energy; 36(8): 4800.

[40] Nanou P, Van Rossum G, Van Swaaij WPM, Kersten SRA. 2011. Evaluation

of catalytic effects in gasification of biomass at intermediate temperature and pressure.

Energy and Fuels; 25(3): 1242.

[41] Chaudhari ST, Dalai AK, Bakhshi NN. 2003. Production of hydrogen and/or

syngas (H2 + CO) via steam gasification of biomass-derived chars. Energy and Fuels;

17(4): 1062.

Chapter 5:

CHARACTERIZATION OF DIFFERENT HEAT

TRANSFER FLUIDS AND DEGRADATION

STUDY BY USING A PILOT PLANT DEVICE OPERATING AT REAL CONDITIONS

A pilot plant was designed to evaluate the degradation of heat

transfer fluids (HTF) for their application in concentrating solar

power plants (CSP). Firstly, the characterization of sixHTFs was

carried out: two ionic liquids ([BMIM][BF4] and [EMIM][BF4]),

two molten salts (Hitec XL and solar salt), a commercial HTF

(Mobiltherm 605) and an oil extracted from

NannochloropsisGaditanamicroalgae(NG oil). Mobiltherm 605

was selected for tuning the pilot plant due to its similarity to HTFs

used in CSP, low cost and easy acquisition. The operating

conditions were set according to thermogravimetric analysis. Thus,

three isothermal experiments were carried out at 140, 160 ant 180

ºC for 15 days. Mobiltherm 605 viscosityincreased with time

indicating that polymerization of hydrocarbon chainstook place.

Two mathematical models were developed to assess the HTF

behaviour in the pilot plant.

A mathematical model for the estimation of the most

representative parameters (viscosity, heat capacity and overall heat

Chapter 5

240

transfer coefficient) of HTF performance was proposed.

Furthermore, an activation/deactivation model was proposed to

predict the variation of the estimated parameters with time. This

model was validated with experimental viscosity measurements

(average error of about 3 %). Finally, the statistical significance of

the model wasproved.

5.1. INTRODUCTION

Nowadays, the electricity power is mainly generated bycentral power plants or

distributed generation systems. Renewable energy resources is one of the solutions to

the dependence on the petroleum import, the energy efficiency and the

conflictsusually arisen from the building of new large power plants based on fossil

fuels [1].

Among all renewable energy sources, the solar energy is by far the most abundant

one. There are two methods to extract electricity from solar radiation: photovoltaic

(PV) and concentrating solar power (CSP) [2]. According to a study of the

International Energy Agency,CSP is three times cheaper than PV[3].

Solar energy is not very dense. It is necessary to concentrate it to produce usable

exploitable temperatures for the production of energy[4].There are two viable

technologies to concentrate solar energy, those that concentrate iton a line and those

that concentrate it on a point. Within the first group, the most developed techniques

are the parabolic trough and linear Fresnel reflector technologies. In the second group,

two technologies stand out: parabolic dishes and solar tower concentrated power

plants[5].Among all of them, parabolic trough is the most mature technology, having a

great promotion worldwide, which has turned out into great cost and performance

improvements.

Chapter 5

241

Parabolic trough plants consist of large fields of parabolic trough collectors, a heat

transfer fluid/steam generation system, aRankine steam turbine/generator cycle and

optional storage and/or fossil-fired backup systems[6]. However, in spite of the great

experience accumulated into this technology, there are still several technique gaps to

make it competitive with traditional fossil-fuel based technologies. The main efforts

have led to the improvement ofthe collector heat transfer [7, 8], the efficiency of the

mirror [9], the integration of a thermal storage system [10]and the substitution of the

heat transfer fluid [11, 12].

The substitution of the current heat transfer fluid (HTF) appears as one of the most

prominent pathways to reduce costs due to the fact that very large quantities of HTF

are needed, entailing high capital investment costs[12]. The present generation of

commercial fluids used are organic synthetic oils composed by eutectic mixtures of

diphenyl oxide and biphenyl[13]. This synthetic oil currently offers the best

combination of low freezing point (12ºC) and upper temperature limit (393ºC) [12].

However, these oils are toxic and high flammable products, resulting into a direct

danger to the plant operators. Furthermore, their use is limited by their degradation

temperature (<400ºC), limiting the efficiency of the thermodynamic cycle for power

generation. Additionally, they have high vapour pressures, exceeding atmospheric

pressure, making difficult its use as thermal storage media as it would require

impractically large pressure vessels[14].

In the recent years, different heat transfer fluids have been proposed. The fluids

that have brought more attention are based on molten salts, mainly inorganic nitrate

salts [14]and ionic liquids[13, 15]due to their excellent thermal properties. Van

Valkenburg et al. [13] established a list of the properties that the new generation of

HTF must satisfyto be considered adequate candidates.Wide liquid temperature range,

high heat capacity, high density, high thermal and chemical stability,low vapour

pressure and non-harmfulness are required[13, 15].

One of the problems when a HTF is evaluated is the difficulty of predicting its

durability as an effective medium as energy carrier in a solar plant. However, no

Chapter 5

242

previous studies in literature about predicting the life cycle of the HTF at large scale

have been reported. Thus, it seems necessary to design an experimental setup in order

to evaluate the thermal performance of HTF. Furthermore, it could be helpful to

establish mathematical models that reproduce the behaviour of the thermal fluids

under different operating conditions.

The aim of this work was to carry out the thermal and physical characterization of

the different thermal fluids that have come up lately as feasible alternatives to be used

in parabolic trough plants, such as molten salts, ionic liquids and microalgae oil.

Furthermore, a comprehensive comparison among the studied HTF was done.

Moreover, the design, assembly and tuning of a pilot plant based on CSP working

principle have been carried out in order to evaluate the degradation of the most

suitable HTF. Finally, mathematical models to predict the thermal behaviour of HTF

under selected operating conditions have been proposed. The suggested models were

validated and their statistical significance was proved.

5.2. MATERIALS AND METHODS

5.2.1 Materials

1-Butyl-3methylimidazolium tetrafluoroborate ([BMIM][BF4]) and 1-ethyl-3-

methylimidazolium tetrafluoroborate ([EMIM][BF4]) were purchased from Sigma

Aldrich.

The molten salts mixtures were prepared using NaNO3, KNO3 and Ca(NO3)2-

tetrahydrate. Reagent grade salts were provided by Sigma Aldrich. Hitec (60%

NaNO3, 40% KNO3, Ca(NO3)2-tetrahydrate), Solar Salt (60% NaNO3, 40% KNO3,

Ca(NO3)2-tetrahydrate) and Hitec XL (60% NaNO3, 40% KNO3, Ca(NO3)2-

tetrahydrate) were synthesized according to Bradshaw et al. [12]procedure.

A commercial heat transfer fluid, MOBILTHERM 605, was purchased from

EXXON MOBIL.

Chapter 5

243

Oil extracted from the microalgae NannochloropsisGaditana (NG microalgae) was

provided by the University of Almeria (Spain).NG microalgae raw material was

purchased from AlgaEnergy Company (Spain). NG microalgae belong

toEustigmatophytes microalgae species with an average composition of 17.5 % lipids,

12.6% fatty acids and 24.1 % of proteins.The extraction process was carried out over

lyophilized biomass, which was previously grounded and sieved to obtain a particle

size between 100 and 300 µm. The extraction was done continuously in a distillation

column joined with a stirring tank, using hexane as the extractor agent.

5.2. Equipment and procedures

5.2.1. Thermogravimetric Analysis (TGA)

Thermal gravimetric analyses were carried out on a Mettler Toledo–TGA-DSC 1.

The initial mass sample was kept between 4 and 15 mg for all tests. Each sample was

analyzed at least twice in order to ensure the reproducibility of the measurements.

Fast scans were performed in the temperature range 40-700 ºC in order to

determinate the degradation temperature (Td) at a heating rate of 10ºC/min under a N2

flow of 60 ml/min. Isothermal TGA experiments were also used for long-term stability

scans. The weight loss of the HTF was evaluated as a function of time under

isothermal conditions at temperatures close to the calculated Tdby fast scans

measurements.

5.2.2 Differential Scanning Calorimetry (DSC) and Modulated DSC (MDSC).

The melting point and heat of fusion of studied samples were measured by using a

differential scanning calorimeter (DSC), TA Instruments model Q100.

Thesemeasurementswerecarried out byvaryingthetemperature in the range from -50 to

300 ºCwith a heating rate of 10 ºC/min under 60 ml/min N2flow.

Chapter 5

244

For themeasurementsofspecificheatcapacity (Cp), thesamplesweresubjectedto a

heatingrampfrom 30 ºCto 300 ºCusing a modulation amplitude of ± 0.5 ºC, with 100

secondsperiodandanunderlyingheating rate of 0.5 ºC/min.

5.2.3. Viscosity

Kinematic viscosity was measured using a Canon–Fenske capillary viscometer.

The viscosity measurements were carried out in the temperature range of 25-100 ºC.

The solution temperature was controlled by a thermostat in a circulating bath

(TAMSON TV2000) monitored by a thermometer. The stopwatch with a resolution of

0.1 s was used to measure the flow times.

5.2.4. Density

The density of different heat transfer fluids was measured by means of a Coriolis

mass flow measuring system, Promass 80. This device allows the measurement of the

density in a large temperature range (-50 – 350 ºC). The calibration was carried out

with air and water at 23.3 ºC and 23.6 ºC in a laboratory certified by ISO/IEC 17025.

5.2.5. Heat Storage

Sensible heat storage is easily calculated from the heat capacity, density and the

temperature change chosen (Eq. 1).

�� = � ∙ �� ∙ (�� − ��) (1)

whereEsis the sensible heat storage (J/m3), ρ is the density of the fluid (kg/m3), Cp is

the heat capacity (J/(kg ºC)), Tout and Tin are the inlet and outlet temperatures of the

solar field, respectively. For this work a rise in temperature of 100 ºC was selected due

to the fact that it is often used in solar applications[13].

5.3. Design of the HTF degradation pilot plant

A pilot plant was designed and constructed for the thermal performance study of

different HTF. The plant with an outer size of 1.15 m x 0.8 m x 2 m is composed of

Chapter 5

245

N2

FI

H-1

V-1

P-1

O-1

Sampling

DI

TI

PI

TIC

TIC

TI

TI

TIFIMITI

TICPI

PI PI

P-67

six functional units: a feedstock vessel, a pumping system, a tubular oven, a heat

exchanger unit, a coriolis mass flow meter and an automatic control system. The

installation is shown schematically in Fig. 1.

Figure 1.Schematic diagram of the HTF degradation pilot scale plant.

The central element of the equipment was a vertical cylindrical vessel (V-1) of 8

litres capacity (90 mm of inner diameter and 1.2 m height). The fluid was heated in V-

1 by an internal electrical resistance, able to reach temperatures up to 400 ºC. The

HTF flowed through the piping system by means of a centrifugal pump Sterling

(ZTND) (P-1) located at the bottom of the deposit. The flow was regulated by a valve

Chapter 5

246

system that allowed part of the oil to be recirculated to V-1. A valve for sampling was

placed at the bottom of V-1 prior to the inlet to P-1.

P-1 impulsedthe thermal fluid through a tubular oven (Fisher 3Kw) (O-1), which

was used to heat the fluidagain, before reaching the heat exchanger unit (H-1). H-1

was made of copper. Table 1 lists the main mechanical details of the heat exchanger.

O-1 was capable of heating the fluid up to 500 ºC.

Table 1. Heat exchanger geometric characteristics: inner diameter (Din), outer diameter (Dext), equivalent length (Leq) and roughness (ε) of the internal pipes.

Din (m) (·103) 7.92

Dout (m) (·103) 9.52

L eq(m) 10.48

ε (·104) 0.15

The fluid was later cooled down by air in a heat exchanger. The air was provided

by a centrifugal fan (Sodeca CMA-426-2M) with a maximum air flow of 850 m3/h.

After the cooling, the HTF passed through a Coriolis mass flow measurement unit

(Promass 80) that continuously registered the flow rate and density. There was a by-

pass to the Coriolis unit in order to let the systemreach the steady state, avoiding

device breakdowns due to high temperatures. Finally, the fluid was redirected to V-1,

closing the loop.

A nitrogen generator (model ZEFIRO 3, CINCL®) was used to provide a

continuous flow of nitrogen (99.999 %) to guarantee an inert atmosphere in the pilot

plant. Nitrogen was introduced into the system through a pipe located in the upper side

of V-1.

The piping system and the vessel used were all made of stainless steel and

thermally insulated in order to minimize heat losses.

Chapter 5

247

The installation had several controllers and indicators. K-type thermocouples were

used to measure temperatures: one was put inside V-1, other two were placed at the

inlet and outlet of P-1; finally, another one was situated into O-1. Finally, other four

thermocouples were used to measure both, the fluid and air temperature, at the inlet

and the outlet of H-1.There were several pressure taps located all over the installation

to measure pressure drop in different sections of the circulation loop. A P18L

transducer was used to measure the relative humidity of air.

Three proportional-integral-derivative (PID) controllers were used to control de

plant. The first one was used to set the temperature of V-1, regulated by the internal

electrical resistance placed inside the vessel. The second one was used to control the

temperature of O-1 by adjusting the power supplied to the oven through the 3 kW

resistance. Finally, a controller was used to set the temperature at the exit of H-1 by

varying the air flow.

Furthermore, the experimental setup was connected to a computer allowing a

remote control of it. Using specific software developed by Adepro engineering

company (Spain), temperatures, differential pressures, fluid density and fluid flow

were processed and recorded every five minutes. Finally, data obtained were

processed with a VBA-excel application to evaluate the main physic properties and its

evolution withtime, as commented next.

5.4. Mathematical model of the thermal performance of HTF in the pilot plant

With data collected from the pilot plant, a pseudo-state model was developed able

to calculate the overall heat transfer coefficient in the heat exchanger,the specific heat

capacity and viscosity of the HTF coming from the heat exchanger with time on

stream. For this purpose, a VBA-Excel application was developed. Likewise, a non-

linear procedure for estimating the parameter in the model that accounted for the

decay/growth with time on stream of the process variables under study was considered

and solved according to the corresponding VBA-Excel application.

Chapter 5

248

Details of the partial calculations are shown in the following sections

5.4.1. Overall heat transfer coefficient determination

The heat flow transferred between a cold and a hot fluid into a heat exchanger is

defined as follows:

� = �� ∙ � ∙ ∆�� (2)

whereQ is the heat transferred(W), A is the heat transfer surface area (m2), ∆Tml is the

log mean temperature difference(ºC) and UF is the overall heat transfer coefficient

(W/(m2·ºC)).

The heat exchanger performance may be evaluated using Eq. (2),being the overall

heat transfer coefficient (UF) defined as follows:

�� = � (� ∙ ∆��)� (3)

Eq. (3) is only valid when the operation is carried out under a steady state

condition. In other words, the flow of the hot and cold streams and their inlet

temperatures must be virtually constant.

5.4.2. HTF specific heat capacity calculation

The heat duty gained or lostwas estimated for both fluids as:

�� = �� ∙ �� ∙ (� − ) (4)

and

�� = �� ∙ �� ∙ (�� − �) (5)

where QHTF and Qair are the heat duty (W)for HTF and air, respectively; CHTF and

Cairare the corresponding specific heat capacity in (J/(kg·ºC)); mHTFand mairare the

corresponding mass flow rates (kg/h);Ti, To, ti and to are the inlet and outlet

temperatures (ºC) for the hot (HTF) and the cold (air) fluids, respectively.

Chapter 5

249

Under steady state conditions,the heat duties for both fluids are balanced according

to the first thermodynamic law(Qair= QHTF= Q).

This way, Eqs. (4) and (5) are equivalent:

�� ∙ �� ∙ (�� − �) = �� ∙ �� ∙ (� − ) (6)

andCHTFcan be calculated from:

�� = �� (�� ∙ (� − ))� (7)

On the other hand, air heat duty can also be defined as:

�� = �� ∙ ∆ �� (8)

wheremairis the dry air mass flow (kg/s) and ∆Hairis the enthalpy of the air

(J/kg)calculated as the difference between the enthalpies of the inlet and outlet air

stream, Ho and Hi, respectively:

∆ �� = − � (9)

These enthalpies can be calculated as a function of absolute humidity in the air as

follows:

= 4184.1 ∙ (0.24 ∙ + ()*+ ∙ (595 + 0.46))(10)

� = 4184.1 ∙ (0.24 ∙ � + ()*+ ∙ (595 + 0.46�))(11)

whereXABSis the absolute humidity in the air surrounded the pilot plant (%).

5.4.3. Viscosity determination

The viscosity determination was performed according to the Newton’s algorithm.

As a first approximation, the model supposed a valueof viscosity for the HTF. Then,

the fluid mean velocity in the air exchanger tube was calculated (Eq. (12)).

Chapter 5

250

/ = 0.1∙�234∙523467∙89:;

(12)

whereρHTF is the working fluid density (kg/m3) and Din is the heat exchanger inlet

diameter tube (m).

Friction factors were calculated for the laminar flow by the Poiseuille equation (Eq.

(13)) whereas the Chen equation was used for the turbulent regime(Eq. (14)):

< = => 16� (13)

?@A = −4 ∙ log(

?E.F0G1 ∙ (

H89:) −

1.0I1JKL ∙ log[ ?

J.OJ1F ∙ PH89:Q

?.?0RO + 1.O10GKLS,UVUW] (14)

Pressure drop of the fluid in the heat exchanger was calculated by the Fanning

equation (Eq. (15)):

Y = (Z234[\Z2349)5]^ + 2 ∙ < ∙ /J ∙ _

89: (15)

wherePHTFiand PHTFoare the working fluid pressure at the inlet and outlet of the heat

exchanger and L is the equivalent length of the pipe where the fluid passed through

(m).F represented a factor that should be zero or close to zero when the iterative

process represented by the Newton’s algorithm is completed.

5.5. Deactivation/activation model.

The variationof the calculated variables (µ, Cp and U) with time on stream was

evaluated in terms of deactivation/activation depending on their decay/growth over

time, respectively. These expressions are well-known in describing deactivation of

catalysts in chemical reactions [16].

Four different deactivation/activation modelswere considered:

`� = >ab\cd∙� (16)

Chapter 5

251

`� = 1 (1 + ef ∙ �)� (17)

`� = �cd (18)

`� = 1 (1 + �cd)� (19)

whereai is the estimated variable (µ, cp and U), kw is the deactivation/activation

constant and t is time.

5.5.1. Parameter estimation.

A VBA-Excel application was developed to solve this model [17], which was

based on the Marquardt-Levenberg algorithm for nonlinear regression [18-20].

The weighted sum of the squared differences between the observed (Exp) and the

calculated (Pr) variables was minimized according to the following equation described

elsewhere [18]:

gg = ∑ ∑ (i�Z� − i�jk�)lJ�lm?��m? (20)

wherei represents the number of equations to be fitted, j the specific experimental data

and n and m the total number of equations and experiments, respectively.

AF-test is a statistical test in which a F-distribution under the null hypothesis is

established. The procedure was based on the comparison between the tabulated F-

value (F-test) and Fc, which was defined elsewhere[19]:

Yn =∑ ∑ (o9pqr );stuW:9uW

∑ ∑ (o9pqvo9wxy)t;(:∙svy)

stuW:9uW (21)

wherep represents the total number of parameters.

If Fc is larger than F(p, n-p, 1-α), assuming a value of α = 0.05, 95% confidence

level, the regression was considered to be meaningful, although there is no guarantee

that the model is statistically suitable since the meaningfulness of each parameter in

the model must be also evaluated.

Hence, a complementary test, named t-test, was used. The t-test considers that the

statistical hypothesis test follows a Student’s t distribution and allows to verify if the

Chapter 5

252

estimate of the parameter bfi differs from a reference value (generally zero). Thus, a

parameter is meaningful (at α = 0.05) each time that the following inequality occurs:

�n� = z{|9z@}({|)99 > �(� − b, 1 −

�J) (22)

where[V(bf)] ii represents the diagonal i th term of the covariance matrix used in the last

step of the n-linear regression procedure.


5.3.1. Heat transfer fluids characterization for their use as thermal fluids in

parabolic trough plants

The heat transfer fluids chosen for their study were: two ionic liquids (1-Butyl-

3methylimidazolium tetrafluoroborate ([BMIM][BF4]) and 1-ethyl-3-

methylimidazolium tetrafluoroborate ([EMIM][BF4])), two molten salts (Hitec XL

(60% NaNO3, 40% KNO3, Ca(NO3)2-tetrahydrate) and solar salt (60% NaNO3, 40%

KNO3, Ca(NO3)2-tetrahydrate), a commercial HTF (Mobiltherm 605) and a new oil

extracted from the microalgae NannocholorpsisGaditana (NG oil). These different

kinds of fluids were selected due to their excellent properties and their different

nature. Ionic liquids and molten salts have exceptional thermal properties since it is

possible that they can be used as thermal storage media [12, 15]. On the other hand,

algae are one of the most attractive biomass feedstock to produce high-valuable

products. In this context, the oil extracted from them,which has been widely studied as

a transport fuel [21], could be used as a valuable HTF.At the best of our knowledge,

the use of the oil extracted directly from algae as a heat transfer fluid has not been

reported yet.

Table 2 shows the main properties that the new generation of HTFs must meet to

be used in the parabolic trough technology field [22].

Chapter 5

253

Table 2.Heat transfer fluids requirements according to National Renewable Energy Laboratory [22].

Heat transfer fluid requirements

Storage density > 1.9 MJ/m3

Freezing point ≤ 0 ºC

High temperature stability ≥ 430 ºC

Vapor presure < 1 atm

Material compatibility Carbon and stainlesssteel

Viscosity Similar toTherminol VP-1

The commercial HTF Therminol®VP-1 was taken as the reference material.

Density, degradation temperature, melting point, heat capacity, heat storage density

and viscosity of the HTFsconsidered in this study together withTherminol® VP-1 are

summarized in Table 3.

Chapter 5

254

Chapter 5

255

Table 3.Main properties of the studied HTF for their use in parabolic trough solar plants.

Property

Heat Transfer Fluids (HTF)

[EMIM][BF 4] [BMIM][BF 4] Solar Salt HiTec XL

Mobiltherm 605

NG oil Therminol® VP-1

Density (g/cm3) (100 ºC) 1.25 1.17 - - 0.83 1.28 1.06

Degradationtemperature (ºC) 430 330 550 650

220 300 257

Meltingpoint (ºC) 14 < -50 230 120 -12 10 12

Heat capacity (J/(g ºC)) (100 ºC) 1.36 1.66 1.49 1.44 2.31 2.27 1.78

Heat storage density, sensible ΔT=100ºC (MJ/m3) 170.4 194.2 282.9* 296.3*

191.7 290.5 188,7

Viscosity (cP) (100ºC) 36.07 119.78 - - 4.40 >>Therminol® VP-

1 2.63

* Molten salts storage density calculation was made according to data collected from literature [11]

Chapter 5

256

3.1.1. Thermal stability.

As was mentioned above, fast-scans do not provide reliable information about a

material long-term thermal stability [23]. However, it is often reported as an

appropriate value for establishing comparison among materials.

According to values listed in Table 3, molten salts were the most thermal stable

compounds.NG oil, HTF Therminol® VP-1, Mobiltherm 605 and [BMIM][BF4]

showed the lowest degradation temperature (Td) whereas the value of Td of

[EMIM][BF 4] kept in the limit of the required value.

Molten salts are eutectic mixtures, which imply that they are really stable in the

fluid state. Furthermore, the studied molten salts are composed by metallic nitrates

conferring a great thermal resistance[12]. On the other hand, ionic liquids thermal

decomposition depends on the nature of the anion rather than that of the cation, and

specially decreases withincreasinghydrophilicity of the anion[24]. Thus,

[EMIM][BF4] showed higher Td value due to the higher hydrophilicity of the

corresponding anion. Besides, ionic liquids can be easily contaminated (water, sodium

ion, silver ion and chloride),which may have an important effect on its thermal

behaviour [13].Finally, Mobiltherm 605 is a paraffinic mineral oil which Tdis

typicallyrangingfrom 150 to 315 ºC [25].

These results were corroborated by long-term experiments. Figure 2 shows the

long-term stability experiments for the ionic liquid [BMIM][BF 4]. The corresponding

Tdmeasured by fast scan turned out to be 330 ºC. However, it can be seen that after 5

hours at 300 ºC the liquid had lost almost the 20 wt. % of its initial weight.

Chapter 5

257

50 100 150 200 250 300 350 400 450 5000

20

40

60

80

100

0 1 2 3 4 520

40

60

80

100

Wei

ght (

%)

Temperature (ºC)

Wei

ght (

%)

Time (h)

C؛ 300 C؛ 350C؛ 375

Figure 2. Fast scans and long-term thermal stability experiments for the ionic liquid

[EMIM][BF 4].

Chapter 5

258

Table 4 shows the degradation rates (%/h) of all HTFs here considered. The

degradation of the fluids took place at temperatures below their corresponding Td.

These results agreed well with those reported in literature [13, 23]. This way, the

thermal stability of the studied HTFsfollowed the trend:Hitec XL > Solar

Salt>[EMIM][BF4]>[BMIM][BF 4] >NG oil>Mobiltherm 605.

Table 4.-Degradation rates of HTFs at different temperatures during isothermal experiments.

HTF Degradation rate (wt.%/h)

Molten salt T= 300 ºC T= 400 ºC T= 500 ºC

Solar salt 0.018 ± 0.001 0.234 ± 0.001 0.445 ± 0.001

Hitec XL 0.001 ± 0.001 0.011 ± 0.001 0.079 ± 0.001

Ionicliquids T= 300 ºC T= 350 ºC T= 375 ºC

[EMIM][BF 4] 4.321 ± 0.001 10.75 ± 0.001 22.456 ± 0.001

[BMIM][BF 4] 0.325 ± 0.001 1.609 ± 0.001 5.231 ± 0.001

Microalgae oil T= 200 ºC T= 235 ºC T= 270 ºC

NG oil 0.021 ± 0.001 1.228 ± 0.001 8.973 ± 0.001

Commercial fluid T= 125 ºC T= 175 ºC T= 225 ºC

Mobiltherm 605 0.842 ± 0.001 6.902 ± 0.001 19.303 ± 0.001

Chapter 5

259

5.1.2. Melting point/freezing point.

Melting temperatures (Tm) were evaluated instead of freezing temperatures due to

the fact that some of the fluids tend to super-cool. The targeted value of Tm would be

below 0ºC, which would allowthese fluids to be used in cold weather regions. Among

the selected HTFs,only [BMIM][BF4] and Mobiltherm 605 maintained their liquid

state at these operating conditions. Molten salts had the highest Tm, being their use in

parabolic trough plants difficult unless freeze protection could be used to keep them in

the liquid state during the process [26]. On the other hand, the ionic liquid

[EMIM][BF 4] and the NG oil showed an acceptable value as long asa minimum

heating is provided to the plant.

5.1.3. Heat capacity

Heat capacity measurements were made by MDSC. Heat capacity affects directly

to the storage capacity of the thermal fluids. All fluids were within an acceptable

range close to Therminol® VP-1 heat capacity. Molten salts and NG oil showed the

lowest values being Mobiltherm 605 the fluid with the highest one.

5.1.4. Density

Density is a necessary value for the calculation of sensible heat storage. The higher

the density, the higher the capacity of the compound to storage heat.The molten salts

density at the targeted temperature (100 ºC) was not evaluated since they are still

solids at this temperature. The ionic liquids density has been reported in

literature[13].NG oil density is higher than that ofMobiltherm 605 andTherminol® VP-

1but it is lower than those of the ionic liquids.

3.1.5. Storage capacity

Storage capacity is calculated as a function of the HTF density, heat capacity and

usable liquid temperature range. These properties determine the use of HTF as thermal

storage media and heat transfer fluids for solar power plants [15]. As it can be

observed, all the studied HTFshad a value of the storage capacity above that of

Chapter 5

260

Therminol® VP-1. Therefore, all the candidates could be considered as useful thermal

storage media.

3.1.6. Viscosity

The parabolic trough-based technology is a flowing system.Thus,the viscosity of

the HTF used in this process is of special interest in order to reduce pumping,

operation and maintenance costs. The viscosity of the reference HTF (Therminol® VP-

1) is 2.48 cSt measured at 40 ºC.NG oil had the highest viscosity of all HTFs

considered. Molten salts were not liquids at ambient temperature, which make them

not suitable for this technology unless freeze protection methods are used. Ionic

liquids under study were closer to the targeted viscosity although still seems too high

to be considered viable candidates without causing problems in the HTF loop. Finally,

Mobiltherm 605 showed the closest viscosity value to that of Therminol® VP-1.

On the other hand, the best HTF should be an inexpensive and nontoxic liquid with

excellent thermo-physical properties and a long service life [25]. The obtained

resultsindicated that the thermal fluids with better propertieswere the ionic liquids, and

more specifically the [EMIM][BF4]. However, there are still several drawbacks to deal

with for its implementation in large scale parabolic trough plants. Availability and

costs are two main issues that a fluid must meet. Ionic liquids and algae oil are still in

the development phase, which imply high production costs and low availability[15].

Furthermore, its preparationprocedures are difficult and tedious. On the other hand,

molten salts are inexpensive and are easily produced although theirhigh melting points

add complexity and larger operating costs [12]. Therefore, the commercial HTF

Mobiltherm 605 was selected in the following study due to its great availability and

similar properties to the commercial fluid used in parabolic trough solar plants. The

main characteristics of Mobiltherm 605 are shown in Table 5.

Chapter 5

261

Table 5.-MOBILTHERM 605 properties.

Properties MOBILTHERM 605

Viscosity (µ) ASTM D 445

40 ºC (cSt)

100 ºC (cSt)

30.4

5.4

Freezing point, ºC, ASTM D97 -12

Flash point (ºC), ASTM D92 230

Maximumoperationtemperature (ºC) 316

Density (kg/l), ASTM D4052 0.86

5.3.2. Pilot plant assembly and tuning.

5.3.2.1. Selection of the operating conditions.

The degradation of a HTF mainly depends on two variables: operation time and

working temperature. The selected procedure to establish the most suitable operating

conditions to evaluate the HTF degradation in the pilot scale plant was based in fast

scans and long-term stability experiments performed by the TGA technique. In

addition, an operation time of 15 days was chosen to evaluate the thermal degradation

of the HTF in the pilot plant.

Figure 3 shows the fast scans and the long-term stability experiments carried out

with Mobiltherm 605. It can be observed that at temperatures above 175 ºC,

Mobiltherm 605 started to decompose. Thus, the temperature set in the vessel (V-1)

was kept in the temperature range between 125-180 ºC where slight degradation was

appreciated. Secondly, the thermal fluid was heated up in the tubular oven (O-1)at 180

ºC in order to keep the temperature of the thermal fluid and accelerate its degradation

Chapter 5

262

50 100 150 200 250 300 350 400 450 5000

20

40

60

80

100

0 1 2 3 4 50

20

40

60

80

100

Wei

ght (

%)

Temperature (ºC)

Wei

ght (

%)

Time (h)

C؛ 125 C؛ 150 C؛ 175 C؛ 200 C؛ 225

in the heat exchanger (H-1). To sum up, three isothermal experiments were carried out

in the pilot plant at working temperatures of 140, 160 and that at180 ºC during 15

days.

Figure 3.Fast scans and long-term thermal stability experiments for the commercial

heat transfer fluid Mobiltherm 605.

Chapter 5

263

130

135

140

145

150

155

160

0 3 6 9 12 150.00

0.01

0.02

0.03

0.04

Tout

T

empe

ratu

re (

ºC)

Tin

ρρ ρρH

TF (k

g/m

3 )

mH

TF (

kg/s

)

Time (days)

740

760

780

800

820

840

5.3.2.2. Pilot Plant Results

Figure 4 shows the most relevant properties measured in the pilot plant: density

(ρ)(kg/m3), mass flow (m)(kg/s), and the inlet and outlet (from the heat exchanger)

temperature of the HTF (Ti and To, respectively) (ºC) for the isothermal experiment at

140 ºC. It can be seen that the results provided by the pilot plant were steady and no

high fluctuations in the data collected were observed.

Figure 4.Most relevant properties measured with the pilot plant: density (ρ), HTF mass flow

(mHTF) and the HTF temperature at the inlet and outlet of the heat exchanger (Ti and To,

respectively).

Chapter 5

264

As described in section 2.4, three main parameters were estimated by means of the

corresponding mathematical model: viscosity (µ) in kg/(m·s), heat capacity (Cp)

J/(kg·ºC) and overall heat transfer coefficient (UF) in J/(m2·ºC·s).These parameters

were normalized respecting their initial value.

The viscosity is an important parameter when evaluating the operative life of a

HTF as it has a direct impact on its heat transfer capacity. Furthermore, viscosity is

important to design heat transfer applications because the pressure drop and the

resulting pumping power depend on its value[27]. The HTF heat capacity (Cp) and the

HTF overall heat transfer coefficient (UF) were also estimated to assess the heat

transfer efficiency and performance of the fluid[28], which are directly related to the

storage capacity of it[15].

There are many researchers focusing on the development of high performance heat

transfer fluids [29]. However, there are not published any standard procedure to

predict the life cycle of a HTF. The limits values have been mainly set byboth users

and manufacturers experience. Nevertheless, there are more and more standards

appearing concerning to oil in-service predictive maintenance. For instance, ASTM D

4378-03[30]and ASTM D 6224[31] are related to In-Service monitoring of mineral

Turbine Oils for Steams and Gas turbines and lubricating oil for auxiliary power

plants equipment, respectively.Thus, it is generallyaccepted a variation in viscosity of

± 15 %respect to its initial value, before an oil in-service could be considered out of its

usable range.

Figure 5 shows the estimated parameters normalized respecting their initial value

(solid line) together with the deactivation model prediction (dotted line), described in

section 2.5,for thethree experiments performedat140 ºC, 160 ºC and 180 ºC. It can be

appreciated the good fitting reached by the deactivation/activation model. The results

Chapter 5

265

0.0

0.4

0.8

1.2

1.6

2.0

0.0

0.4

0.8

1.2

1.6

2.0

0 3 6 9 12 150.0

0.4

0.8

1.2

1.6

2.0

UF/U

Fo

(UF/U

Fo)

th

U.F

. /U

.Fo

160 ºC

Cp/Cpo

(Cp/Cpo)th

Cp/

Cp o

µ/µ

µ/µ

µ/µ

µ/µ οο οο

Time (days)

µ/µo

(µ/µο)

th

0.0

0.4

0.8

1.2

1.6

2.0

0.0

0.4

0.8

1.2

1.6

2.0

0 3 6 9 12 150.0

0.4

0.8

1.2

1.6

2.0

UF/U

Fo

(UF/U

Fo)

th

U.F

. /U

.Fo

180 ºC

Cp/Cpo

(Cp/Cpo)th

Cp/

Cp o

µ/µ

µ/µ

µ/µ

µ/µ οο οο

Time (days)

µ/µo

(µ/µο)

th

0.0

0.4

0.8

1.2

1.6

2.0

0.0

0.4

0.8

1.2

1.6

2.0

0 3 6 9 12 150.0

0.4

0.8

1.2

1.6

2.0

UF/U

Fo

(UF/U

Fo)

th

U.F

. /U

.Fo

140 ºC

Cp/Cpo

(Cp/Cpo)th

Cp/

Cp o

µ/µ

µ/µ

µ/µ

µ/µ οο οο

Time (days)

µ/µo

(µ/µο)th

obtained showed that the change in viscosity after the fifteen days of operation was

higher as the temperature increased, varying from 1.02 at 140 ºC to 1.06 at 180 ºC.

Figure 5.Estimated normalized experimental properties versus theoretical ones (th).

As aforementioned, Mobiltherm 605 is a paraffinic mineral oil formed by

hydrocarbon chains.According to ASTM D 4378-03 [30] and ASTM D 6224-02 [31],

the main factors affecting the service life of these type of oils would be the

contamination by secondary fluids, such as lubricants and water; the oil oxidation and

the oil thermal degradation. In this work, the thermal degradation in inert atmosphere

(provided by a constant flow of pure N2) was only considered. Concerning to the

thermal degradation of the HTF, two effects could take place[25]. Firstly, the viscosity

of the fluid can increase. At high temperature the fluid breaks down into smaller

molecules due to thermal cracking, leading to a decrease in viscosity for the paraffinic

oil. Secondly, the polymerization of hydrocarbon chains might occur,leading to an

Chapter 5

266

increase in viscosity. Therefore, opposite effects could be taken place, producing no

net changes in viscosity even though the HTF degradationwas occurring.

The viscosity of Mobiltherm 605 increased during operationat all the temperatures

essayed. This fact points out that the polymerization path was the predominant effect.

In addition, at the highest temperature tested the maximum variation of the viscosity

turned out to be ~ 6 %,indicating the high thermal stability of Mobiltherm 605.

The heat capacity (Cp) and overall heat transfer coefficient (UF) followed opposite

trends. Cp increased with time, being the final value of 1.02, 1.03 and 1.12 for 140,

160 and 180 ºC, respectively. On the other hand, UF decreased with time. However, no

influence of temperature was found and thefinal value of this parameter was 0.95.

The activation/deactivation constants and the resulting parameters obtained by the

non-linear regression procedure described in section2.5.1, are summarized in Table 6.

As aforementioned, the discrimination of kinetic parameters was done applying the F-

test and the t-test at the 95% confidence level. In terms of statistical results, F-test

considered the regression to be suitable in all cases since the corresponding values to

the Fc/Ftest ratio was larger than 1. The t-test was also used for evaluating each

parameter in the model. As shown in Table 6, the values of tc/t-test ratio were also

larger than 1, showing the statistical significance of the proposed models and their

corresponding parameters.

Chapter 5

267

Table 6.-Estimated deactivation constant (Kw) at different temperatures

(140, 160 and 180 ºC) for MOBILTHERM 605.

Temperature = 140 ºC

Normalized parameters

K w tc t-test Fc F-test

µ/µo -1.22·10-3 4

Cp/Cpo -3.13·10-3 10 1.96 56702 2.37

UF/UFo 2,12·10-3 6




µ/µo -1.32·10-3 4

Cp/Cpo -4.33·10-3 13 1.96 58102 2.37

UF/UFo 2.86·10-3 9




µ/µo -4.85·10-3 25

Cp/Cpo -2.56·10-3 13 1.96 311942 2.37

UF/UFo 2.85·10-3 14

Chapter 5

268

0.0

0.4

0.8

1.2

1.6

2.0

0.0

0.4

0.8

1.2

1.6

2.0

0 3 6 9 12 150.0

0.4

0.8

1.2

1.6

2.0

µ/µ

µ/µ

µ/µ

µ/µ οο οο

(µ/µο)estimated

(µ/µο)exp.

140 ºC

160 ºC

µ/µ

µ/µ

µ/µ

µ/µ οο οο

(µ/µο)estimated

(µ/µο)exp.

180 ºC

µ/µ

µ/µ

µ/µ

µ/µ οο οο

Time (days)

(µ/µο)estimated

(µ/µο)exp.

5.3.3.Model validation

With the aim of validating the theoretical values given by the

activation/deactivation model (th), experimental viscosity measurements were carried

out with a Fenskeviscometer. For these measurements, 100 ml samples were taken

dailyaccording to the standard ASTM D 4378-03[30].

Figure 6 shows the experimental versus theoretical viscosity values as a function of

time.It can be seen how the theoretical valuesfitted the experimental results accurately.

Table 7 summarizes the absolute and relative errors among the theoretical and

experimental values. The following relative errors were obtained: 0.59 %, 1.30 % and

2.33 % at 140 ºC, 160 ºC and 180 ºC, respectively. These results validate the

procedures and model used in this work.

Figure 6.-Estimated normalized experimental viscosities versus experimental ones.

Chapter 5

269

Table 7.-Absolute and relative error of the experimental measurements of viscosity versus the theoretical ones.

140 ºC

Days 0 2 4 6 8 10 12 14 15 Mean error (%)

Absolute error 0 -0.0078 -0.0084 -0.0153 0.0022 0.0014 0.0101 -0.0056 -0.0031

Relative error (%) 0 0.78 0.83 1.52 0.22 0.14 1.00 0.55 0.31 0.59

160 ºC

Absolute error 0 0.0151 0.0215 0.0019 0.0262 0.0269 0.0086 0.0052 -0.0130

Relative error (%) 0 1.51 2.15 0.19 2.60 2.66 0.85 0.51 1.27 1.30

180 ºC

Absolute error 0 0.0093 0.0099 0.0186 0.0114 0.0492 0.0456 0.0593 0.0077

Relative error (%) 0 0.94 0.99 1.86 1.12 4.90 4.54 5.92 0.72 2.33

Chapter 5

270

5.4. CONCLUSIONS

A pilot plant was designed to evaluate the degradation of HTfs to be used in

concentrating solar power plants (CSP). Six different HTFs were characterized:two

ionic liquids (1-Butyl-3methylimidazolium tetrafluoroborate ([BMIM][BF4]) and 1-

ethyl-3-methylimidazolium tetrafluoroborate ([EMIM][BF4])), two molten salts (Hitec

XL (60% NaNO3, 40% KNO3, Ca(NO3)2-tetrahydrate) and solar salt (60% NaNO3,

40% KNO3, Ca(NO3)2-tetrahydrate), a commercial HTF (Mobiltherm 605) and a new

oil extracted from the microalgae NannochloropsisGaditana (NG oil).

Mobiltherm 605 was selected for the assembling and tuning of the pilot plant due

to its great availability and similar properties to the commercial fluid used in parabolic

trough solar plants. The pilot plant behaviour was stable and no high fluctuations of

data collected were detected. Three isothermal experiments were carried out at 140,

160 and 180 ºC for 15 days. The viscosity was selected as the key parameter to follow

the HTF degradation. Mobiltherm 605 viscosity increased with time, indicating that

the polimerization of hydrocarbon chains took place. The variation of viscosity was 6

% at 180 ºC pointing out the high thermal stability of Mobiltherm 605.

Two mathematical models were developed to estimate the most representative

parameters (µ, Cp and UF) with time on stream and predict the behaviour of the

parameters during operation, respectively. The model was validated with experimental

measurements of viscosity obtaining an error lower than 3%. Finally, the statistical

significance of the model was also proved.

Chapter 5

271

5.5. REFERENCES

[1] Sripakagorn A, Srikam C. Design and performance of a moderate temperature

difference Stirling engine. Renew. Energy. 2011;36(6):1728-33.

[2] Du Marchie Van Voorthuysen EH. The promising perspective of Concentrating

Solar Power (CSP). Int. Conf. on Future Power Systems, 2005.

[3] "World Energy Outlook 2004", International Energy Agency, Paris 2004.

[4] El Gharbi N, Derbal H, Bouaichaoui S, Said N. A comparative study between

parabolic trough collector and linear Fresnel reflector technologies. Conference A

comparative study between parabolic trough collector and linear Fresnel reflector

technologies, vol. 6. p. 565-72.

[5] Mills D. Advances in solar thermal electricity technology. Solar Energy.

2004;76(1-3):19-31.

[6] Price H, Lüpfert E, Kearney D, Zarza E, Cohen G, Gee R, et al. Advances in

parabolic trough solar power technology. J. Sol. Energy Eng., Transactions of the

ASME. 2002;124(2):109-25.

[7] Al-Ansary H, Zeitoun O. Numerical study of conduction and convection heat

losses from a half-insulated air-filled annulus of the receiver of a parabolic trough

collector. Sol. Energy. 2011;85(11):3036-45.

[8] Padilla RV, Demirkaya G, Goswami DY, Stefanakos E, Rahman MM. Heat

transfer analysis of parabolic trough solar receiver. Appl. Energy. 2011;88(12):5097-

110.

[9] Sansoni P, Fontani D, Francini F, Giannuzzi A, Sani E, Mercatelli L, et al. Optical

collection efficiency and orientation of a solar trough medium-power plant installed in

Italy. Renew. Energy. 2011;36(9):2341-7.

[10] Laing D, Bahl C, Bauer T, Lehmann D, Steinmann WD. Thermal energy storage

for direct steam generation. Sol. Energy. 2011;85(4):627-33.

Chapter 5

272

[11] Kearney D, Herrmann U, Nava P, Kelly B, Mahoney R, Pacheco J, et al.

Assessment of a molten salt heat transfer fluid in a parabolic trough solar field. J. Sol.

Energy Eng., Transactions of the ASME. 2003;125(2):170-6.

[12] Bradshaw RW, Cordaro JG, Siegel NP. Molten nitrate salt development for

thermal energy storage in parabolic trough solar power systems. Proc. ASME 3rd Int.

Conf. Energy Sustain., vol. 2. p. 615-24.

[13] Valkenburg MEV, Vaughn RL, Williams M, Wilkes JS. Thermochemistry of

ionic liquid heat-transfer fluids. Thermochim. Acta. 2005;425(1–2):181-8.

[14] Blake DM, L.; Hale, M.J.; Price, H.; Kearney, D.; Herrmann, U. New Heat

Transfer and Storage Fluids for Parabolic Trough Solar Thermal Electric Plants. Proc.

11th ColarPACES Int. Symp. Conc. Sol. Power Chem. Energy Technologies. Zurich,

Switzerland 2002.

[15] Wu B, Reddy RG, Rogers RD. Novel ionic liquid thermal storage for solar

thermal electric power systems In: Kleis SJ, Bingham CE editors. Proc. Sol. Forum

2001. Solar Energy: The Power to Choose, Washington, DC. 2001; p. 445-51.

[16] Levenspiel O. Chemical reaction engineering: An introduction to the design of

chemical reactors. In: Wiley, editor. Michigan1962.

[17] Valverde JL, De Lucas A, Carmona M, González M, Rodríguez JF. A generalized

model for the measurement of effective diffusion coefficients of heterovalent ions in

ion exchangers by the zero-length column method. Chem. Eng. Sci. 2004;59(1):71-9.

[18] De Lucas A, Sánchez P, Dorado F, Ramos MJ, Valverde JL. Kinetic model of the

n-octane hydroisomerization on PtBeta agglomerated catalyst: Influence of the

reaction conditions. Ind. Eng. Chem. Res. 2006;45(3):978-85.

[19] Froment GF, Bischof, K.B. Chemical Reactor Analysis and Design. New

York1990.

[20] Marquardt DW. An algorithm for least-squares estimation of nonlinear

parameters. J Soc Ind Appl Math. 1963;11:431-41.

[21] Ross AB, Biller P, Kubacki ML, Li H, Lea-Langton A, Jones JM. Hydrothermal

processing of microalgae using alkali and organic acids. Fuel. 2010;89(9):2234-43.

Chapter 5

273

[22] “Thermal storage for Solar Thermal Parabolic Trough Electric Power Systems”,

National Renewable Energy Laboratory, Request for Proposal Number RCQ-0-30910,

March 27, 2000.

[23] Kosmulski M, Gustafsson J, Rosenholm JB. Thermal stability of low temperature

ionic liquids revisited. Thermochim. Acta. 2004;412(1-2):47-53.

[24] Singh G, Kumar A. Ionic liquids: Physico-chemical, solvent properties and their

applications in chemical processes. Indian J. Chem. – Sect. A Inorg., Phys., Theor.

Anal. Chem.. 2008;47(4):495-503.

[25] Oyekunle LO, Susu AA. Characteristic properties of a locally produced paraffinic

oil and its suitability as a heat-transfer fluid. Pet. Sci. Tech. 2005;23(11-12):1499-509.

[26] Moens L, Blake DM, Rudnicki DL, Hale MJ. Advanced thermal storage fluids

for solar parabolic trough systems. Int. Sol. Energy Conf., 2002; p. 277-83.

[27] Chandrasekar M, Suresh S, Chandra Bose A. Experimental investigations and

theoretical determination of thermal conductivity and viscosity of Al2O3/water

nanofluid. Exp. Therm. and Fluid Sci. 2010;34(2):210-6.

[28] Bergman TL. Effect of reduced specific heats of nanofluids on single phase,

laminar internal forced convection. Int. J. Heat and Mass Transfer. 2009;52(5–

6):1240-4.

[29] Saeedinia M, Akhavan-Behabadi MA, Razi P. Thermal and rheological

characteristics of CuO-Base oil nanofluid flow inside a circular tube. Int. Commun.

Heat and Mass Transfer. 2012;39(1):152-9.

[30] ASTM 4378-03 AD. In-Service Monitoring of Mineral Turbine Oils for Steam

and Gas Turbines. 2003.

[31] ASTM 6224-02 AD. In-Service Monitoring of Lubricating Oil for Auxiliary

Power Plant Equipment. 2003.

Chapter 6:

GENERAL CONCLUSIONS AND

RECOMMENDATIONS

This chapter lists the main conclusions derived from the research performed in this

Doctoral Thesis. In addition, some recommendations are suggested to be taken into

account in further studies.

6.1. CONCLUSIONS

- Pyrolysis, combustion and gasification characteristics of NG microalgae were

analyzed by TGA-MS. High mass loadings cause heat-transfer problems, whereas

small particle sizes led to less diffusion resistance. Gas flow did not affect pyrolysis

and combustion. Gasification temperature had a direct impact on char conversion and

reactivity. Reactivity increased with decreasing sample weight and increasing

porosity. Low gas flow decreased char conversion. Pyrolysis and combustion main

products were generated in the second degradation step. N-compounds evolution was

associated with the microalgae proteins degradation. SO2 release during combustion

could be related to sulphated polysaccharides decomposition. H2 production was

enhanced by steam concentration.

- Thermal characteristics and gas formation during pyrolysis of Fir Wood,

Eucalyptus Wood, Pine Bark, NG microalgae and three individual components of

lignocellulosic biomass (hemicellulose, lignin and cellulose) were analyzed by TGA-

Chapter 6

274

MS. Pyrolysis of lignocellulosic biomass was divided into four zones: moisture

evolution, hemicellulose decomposition, lignin and cellulose degradation and lignin

decomposition. NG microalgae showed the highest thermal stability. The main

products (CO2, light hydrocarbons and H2O) were generated between 200 and 450 ºC.

H2 was produced at high temperatures (>700 ºC). Kinetic model satisfactorily

predicted the pyrolysis of biomass. Furthermore, the statistical significance of the

model was proved.

- Combustion behavior and gas formation from the oxidation process of fir wood,

eucalyptus wood, pine bark and three individual components of lignocellulosic

biomass (cellulose, hemicellulose and lignin) were analyzed by TGA-MS. Biomass

combustion took place into two main stages: devolatilization stage (Dev. stage) and

oxidation stage (Oxid. stage). Most products detected in the combustion of

lignocellulosic biomass were released during the Dev. stage whereas only NO2,

C2H5O+, CO and CO2 were detected at the Oxid. stage. Nitrogen compounds were

released as CH4N, HCN and NOx. Lignocellulosic biomass combustion was fitted to a

first order reaction model (O1).

- Combustion of microalgae took place into two main stages: devolatilization stage

and oxidation stage. However, up to three sub-steps could be identified during the

microalgae combustion attributed to the decomposition of carbohydrates, proteins and

lipids. The ignition characteristic showed that samples CV and SC required less

amount of energy to develop the combustion process. However, NG sample released a

higher amount of heat during the combustion. The kinetic analysis of microalgae

combustion showed that the most representative mechanism for representing the

process was a first order reaction model (O1). The excellent fitting between the

experimental and theoretical curves (maximum mean error was 3.1%, for NG sample)

confirmed the selection of model O1. CO, CO2 and H2O were the main products

released during combustion. Other compounds detected during the combustion of

microalgae were light hydrocarbons (especially CH4); nitrogen compounds (mainly

released as NO, NO2 and HCN); sulfur compounds (SO, SO2 and COS); hydrogen and

Chapter 6

275

other oxygen containing hydrocarbons (ketones, esters, ethers and carboxylic acids).

Nitrogen compounds were found in higher proportions than sulfur ones.

- Samples W and RP showed the best burning profile by combining a high

combustion characteristic factor (CCF) and a high release of combustion heat (Hcomb).

The kinetic analyses of the oxidation process was performed using pseudo mulit-

component separate-stage models (PMSM). The combustion process was divided into

three stages: Devolatilization stage (correlated with the hemicellulose and cellulose

content in the samples), Oxidation stage (influenced by the initial amount of lignin in

the samples) and Remaining burning (associated to the final char burning and

devolatilization of inorganic matter). The high ash content of CR sample enhanced the

amount of volatiles released during the combustion process lowering its activation

energy. The good fitting of experimental curves with theoretical ones validated the

proposed model (mean error below 3.4 %). H2, CO and CO2 were the main product

obtained from energy crops combustion process. Furthermore, NOx were detected in a

higher proportion than other pollutants such as SOx, chloride compounds (CH3Cl) or

aromatic ones (C6H6).

- Thermal characteristics and gas formation during the pyrolysis and gasfication of

eucalyptus wood, fir wood, pine bark and biomass main components (cellulose, xylan

and lignin) were analyzed by TGA-MS. The presence of indigenous inorganic matter

in the gasification process of biomass samples played an important role compared with

their initial chemical composition. The reactivity of biomass samples was correlated

with their alkali index and was ranked as follows: Xylan > lignin > cellulose and

Eucalyptus wood > fir wood > pine bark. The high relevance of inorganic matter was

proved by the inaccuracy of the results obtained by three standards models (VM, SCM

and RPM) which fail to predict the effect of catalytic active species. A semi-empirical

model was proposed in order to accurately model the gasification process. The

proposed model showed errors below 1 %. Furthermore, the models used in this work

were statistically validated. The high production of H2 and CO showed the

predominance of solid-gas reactions. On the other hand, gas phase reactions as water-

Chapter 6

276

gas shift had a higher relevance in the gasification of fir wood due to its high calcium

content. Methanation reactions also took place especially for eucalyptus wood sample

and was correlated to the catalytic effect of potassium.

- A pilot plant was designed to evaluate the degradation of HTfs to be used in

concentrating solar power plants (CSP). Six different HTFs were characterized: two

ionic liquids (1-Butyl-3methylimidazolium tetrafluoroborate ([BMIM][BF4]) and 1-

ethyl-3-methylimidazolium tetrafluoroborate ([EMIM][BF4])), two molten salts (Hitec

XL (60% NaNO3, 40% KNO3, Ca(NO3)2-tetrahydrate) and solar salt (60% NaNO3,

40% KNO3, Ca(NO3)2-tetrahydrate), a commercial HTF (Mobiltherm 605) and a new

oil extracted from the microalgae Nannochloropsis Gaditana (NG oil).

Mobiltherm 605 was selected for the assembling and tuning of the pilot plant due

to its great availability and similar properties to the commercial fluid used in parabolic

trough solar plants. The pilot plant behaviour was stable and no high fluctuations of

data collected were detected. Three isothermal experiments were carried out at 140,

160 and 180 ºC for 15 days. The viscosity was selected as the key parameter to follow

the HTF degradation. Mobiltherm 605 viscosity increased with time, indicating that

the polimerization of hydrocarbon chains took place. The variation of viscosity was 6

% at 180 ºC pointing out the high thermal stability of Mobiltherm 605.

Two mathematical models were developed to estimate the most representative

parameters (µ, Cp and UF) with time on stream and predict the behaviour of the

parameters during operation, respectively. The model was validated with experimental

measurements of viscosity obtaining an error lower than 3%. Finally, the statistical

significance of the model was also proved.

6.2. RECOMMENDATIONS

LIST OF PUBLICATIONS AND CONFERENCES

List of publications and conferences

279

PUBLICATIONS

López-González, D., Valverde, J.L., Fernandez-Lopez, D., Sanchez-Silva, L. (In

press). Thermogravimetric-mass spectrometric analysis on combustion of

lignocellulosic biomass. Bioresource Technology.

López-González, D., Valverde, J.L., Sánchez, P., Sanchez-Silva, L.

2013.Characterization of different heattransfer fluids and degradation study by using a

pilot plant device operating at real conditions. Energy. 54, PP. 240-250

Sanchez-Silva, L., López-González. D., Garcia-Minguillan, A.M., Valverde, J.L.

2013. Pyrolysis,combustion and gasification characteristics of

Nannochloropsisgaditana microalgae. BioresourceTechnology, 130, pp. 321-331

Sanchez-Silva, L., López-González, D., J. Villaseñor, J., Sánchez, P., Valverde, J.L.

2012.Thermogravimetric-mass spectrometric analysis of lignocellulosic and marine

biomass pyrolysis. Bioresource Technology, 109, pp. 163-172.

CONFERENCES

Keynotes A. de Lucas-Consuegra, J.L. Endrino, J. González-Cobos, D. López, J.A. Díaz, J.L. Valverde.ANQUE’s ICCE. Sevilla (Spain), June 2012. Oral presentations J.A. Díaz, J. González-Cobos, D. López-González, A. Romero, J.L. Valverde. ANQUE’s ICCE. Sevilla (Spain), June 2012. L. Sánchez-Silva, D. López-González, J. González-Cobos, J.A. Díaz, J. Villaseñor, P. Sánchez, J.L. Valverde. ANQUE’s ICCE. Sevilla (Spain), June 2012.

List of publications and conferences

280

Posters: 2 contributions.

Tesis Doctoral VALORIZACIÓN DE BIOMASA DE ORIGEN …

Documents

Transcript of Tesis Doctoral VALORIZACIÓN DE BIOMASA DE ORIGEN …