L U L E A i U N I V E R S I T Y ^ i v ^ g
O F T E C H N O L O G Y
2000:23
D O f T O R Al TT4FSTS
Study of Possibilities and Some Problems of Using Cane Residues as Fuel in a Gas Turbine
for Power Generation in the Sugar Industry
Department of Mechanical Engineering
Division of Energy Engineering
Study of Possibilities and Some Problems of Using Cane Residues as Fuel in a Gas Turbine for Power Generation
in the Sugar Industry
Mohamed Gabra
Doctoral Thesis Division of Energy Engineering Department of Mechanical Engineering
Luleå University ofTechnology 971 87 Luleå, Sweden
Luleå, Sweden 2000
Study of Possibilities and Some Problems of Using Producer
Gas from Gasification of Cane Residues in a Gas Turbine for Increased Power Generation in the Sugar Industry
A B S T R A C T
Use o f cane residues as a source o f energy could help many developing countries to improve their balance o f payments by decreasing fossil fuels imports and increasing the electrical power production. It can be shown that the amount o f electricity that can be produced by using commercially and proven technology could be up to 15 times the on-site needs o f a sugar m i l l . This would make it possible to sell excess power to the national grid.
The cane trash would then be used as a fuel during the off-season period and the back-pressure steam turbines w i t h low adrnission data that are presently being used i n most o f sugar factories would be replaced w i t h extraction/condensing turbines with steam generated i n more efficient and advanced boilers. A substantial further increase o f the electricity output could be achieved by using a combined gas turbine/steam turbine process integrated w i t h a gasifier supplying combustible gas to the gas turbine combustor. The utilisation o f this technology f o r co- generation i n sugar mills is however prevented by the lack o f proven technology f o r utilisation of the combustible gas f r o m the gasification of cane residues as fuel f o r gas turbines. The problem is to f i n d a way to avoid erosion, deposits and coaosion i n the gas turbine caused by ash-forming elements i n the cane residues.
A method f o r avoiding excessive amounts of alkali compounds and particles i n producer gas f r o m gasification o f cane residues powder has been studied and evaluated. Powderised cane residue is gasified i n a cyclone gasifier and the product gas burnt i n an adapted combustion chamber. The cyclone gasifier works as a particle separator as well.
The results o f gasification tests w i t h bagasse and cane trash i n the cyclone gasifier, so far only made at atmospheric pressure, show that the gasification is stable and the system is running smoothly f o r certain ranges of conditions. The heating values o f the producer gas f r o m the gasification o f bagasse and cane trash are sufficient for stable combustion. The total carryover particles concentrations i n the producer gas f r o m bagasse gasification were found to be i n the range o f that is acceptable f o r A B B Stal GT35P gas turbine, but for cane trash were found to be slightly higher than allowable. As regards the fraction o f large particles, above 10 p m , the observations are inconclusive. It appears however that improvement o f the cyclone performance f o r separation o f such particles is desirable. Even though the gasification cyclone separated about 75%-80% o f the alkali compounds supplied w i t h the fuel, the contents o f these compounds i n the producer gas are still higher than specified as acceptable f o r the GT35P gas turbine. The melting temperature o f the f l y ash f o r both bagasse and cane trash ash was found to be much above the inlet temperature o f this gas turbine, but deposition problems can not be excluded since some carryover particles i n the gas seem to have been melted. As an overall assessment, cane trash appears to be a more problematic fuel than bagasse i n this application.
Pre-treatment o f the cane trash, such as washing to reduce the contents o f C l and K might help reducing the problems.
A comparative study o f bagasse gasification i n the cyclone gasifier and a fluidised bed gasifier w i t h quartz sand as bed material has also been conducted. The comparative study showed alkah retention i n the range o f 12% - 4% i n the fluidised bed whereas i n the cyclone the alkali separation was found to be about 70%. The observations indicate that no significant coating o f the bed material particles is formed and that the retention i n the fluidised bed is mainly caused by the ash particles accumulated in the bed. Further studies are needed to investigate the effects o f the bed material composition on alkali retention i n a fluidised bed.
I n memory o f my father Ahmed Mukhtar Gabra to my mother Suad Omer Attia
to my w i f e Nagwa Moh anted
and children Ahmed El-tayb andZein El-abdeen
P R E F A C E
This study was performed at Luleå University o f Technology, Division o f Energy Engineering between A p r i l 1996 and September 2000. The first part of this study was started at Stockholm Environment Institute (SEI) between A p r i l 1994 and A p r i l 1996. The experimental work has been done at Energy Technology Centre i n Piteå (ETC). This thesis is continuation o f my Licentiate thesis on bagasse gasification in a cyclone gasifier at Luleå University of Technology, Division of Energy Engineering i n November 1998. Financial support f r o m the Swedish Intemational Development Cooperation Agency (Sida) is gratefully acknowledged.
I am greatly indebted to m y supervisor Professor B j ö r n Kjellström who introduced me to the world o f biomass, especially the field o f biomass gasification. It gives me great pleasure to acknowledge his valuable discussions, constructive criticism of the work, friendship, never failing support and encouragement. His guidance, kind help and sincere readiness to facilitate whatever was needed have been crucial to the completion o f this study and always w i l l be remembered w i t h sincere respect and faithfulness.
Thanks are due to Professor Anders Nordin (Department o f Chemical Engineering, U m e å University) f o r the useful and f r u i t f u l discussions, Energy Technology Centre i n Piteå, especially to M r s . Ulla Jonsson f o r providing excellent working conditions and to M r . Lennart Nordsvahn f o r his technical and computer assistance. Many thanks are also due to m y co- authors at (ETC) Dr. Marcus Ö h m a n and Tekn. L i c . Esbjörn Pettersson f o r their co-operation and creating a friendly atmosphere at ETC. I am especially grateful to Dr.Rainer Backman (Åbo Akademi, Finland) f o r collaboration, discussions and his k i n d support through this work.
I w o u l d like to thank the staff at the ETC and Division o f Energy Engineering at Luleå University o f Technology f o r good, relaxed working atmosphere and kind help, to M r . Fredrik Granberg at L u l e å University o f Technology for his help i n pelletizing the cane residues. I would like also to thank M r . Johannes Kjellström f o r teaching me how to use the Window drawing program (Micrografx).
I take this opportunity to extend m y appreciation to M r . R. C. Rampling the factory manager (in 1995) o f Kenan Sugar Company Limited Sudan (KSC) and M r . V . Boura, purchasing co- ordinator, o f K S C - London office, f o r their co-operation and assistance i n providing the project w i t h the bagasse and cane trash.
M y sincerely thanks are due to m y mother Suad Omer f o r her tremendous support, to my brothers and sisters f o r their love and continuous support. Finally thanks to my f a i t h f u l w i f e Nagwa f o r her patience, accepting the inconvenience caused by m y long absence, her great love, encouragement and unfailing support made it is possible for me to continue.
Luleå, September 2000 Mohamed Gabra
L I S T O F C O N T E N T S
A B S T R A C T I I P R E F A C E I V L I S T O F C O N T E N T S V
L I S T O F P A P E R S V I I S C I E N T I F I C C O N T I B U T I O N V H I
O V E R V I E W X 1 I N T R O D U C T I O N 1
2 O B J E C T I V E S O F T H E T H E S I S 3 3 B I O M A S S E N E R G Y C O N V E R S I O N 3
3.1 Pyrolysis technology 3 3.2 Biomass gasification technology 4
3.2.1 Overview 4 3.2.2 Types o f gasifier 4 3.2.3 Mechanism o f gasifying a biomass particle i n a cyclone gasifier 5
3.2.4 Gasification o f biomass particles i n a fluidised bed gasifier 6
3.3 Gas quality criteria for gas turbine operation 7 4 A R E V I E W O F F U E L R E S I D U E G E N E R A T E D I N T H E S U G A R
I N D U S T R Y 9 4.1 Overview 9 4.2 Bagasse as residue fuel 10
4.3 Cane trash as residue f u e l 11 5 B I O M A S S G A S I F I C A T I O N T E C H N O L O G Y D E M O N S T R A T I O N F O R
E L E C T R I C I T Y G E N E R A T I O N 13
5.1 Overview 13 5.2 V ä r n a m o biomass gasification demonstration plant 14
5.3 A R B R E biomass gasification demonstration plant 15
5.4 U S A demonstration projects 17 5.4.1 Hawaii gasifier demonstration plant 17
5.4.2 Vermont biomass gasification demonstration plant 18
5.5 Brazilian biomass demonstration program 19 6 E V A L U A T I O N O F D I F F E R E N T PROCESS O P T I O N S F O R
C O - G E N E R A T I O N I N T H E S U G A R I N D U S T R Y 20 7 E X P E R I M E N T A L E Q U I P M E N T A N D P R O C E D U R E S 22
7.1 Test facility w i t h atmospheric cyclone gasifier 22
7.1.1 Operating principle 22
7.1.3 Measuring devices 23 7.1.4 Experimental procedures 24 7.2 Test facility w i t h fluidised bed gasifier 25
7.2.1 Experimental equipment 25 7.2.2 Fuel preparation and feeding 26 7.2.3 Procedure f o r determination o f bed agglomeration 26
7.2.4 Comparative study o f alkali species carryover i n F B G 27 7.3 Determination o f composition and properties o f fuels, char and ash 28
7.3.1 Fuel analysis 28 7.3.2 Ash analysis 29 7.3.3 The standard ash fusion test ( A S T M ) 29
7.3.4 Ash melting tests 29 7.4 Fuel characteristics 29 8 E Q U I L I B R I U M C A L C U L A T I O N S F O R B A G A S S E A S H R E A C T I O N S 31
9 E X P E R I M E N T A L R E S U L T S . 33 9.1 Cyclone gasification experiments 33
9.1.1 Equivalence ratios f o r stable operation 33
9.1.2 Heating value 34 9.1.3 Particles i n the producer gas 34
9.1.4 A l k a h compounds i n the producer gas 35 9.1.5 The melting temperature o f the f l y ash carried w i t h the producer gas 35
9.2 Fluidised bed gasification experiments 36 9.2.1 Studies o f bed agglomeration tendencies for cane residues 36
9.2.2 A l k a l i species carryover 36
10 M A I N C O N C L U S I O N S 38 11 F U T U R E W O R K 40 12 R E F E R E N C E S 41
LIST OF PAPERS AND CONTRIBUTIONS FROM T H E AUTHOR OF THIS THESIS
PAPERI
Gabra, M . and Kjellström, B .
Evaluation o f New Process Options for Co-generation i n the Sugar Industry.
Paper Presented at the Intemational Gas Turbine and Aeroengine Congress &
Exhibition, Birmingham, U K . Organised by American Society o f Mechanical Engineers ( A S M E ) , 96-GT-301, i n June 10-13,1996.
Mohamed Gabra has written this paper. The paper was reviewed before submission by professor B j ö m Kjellström. It received the "Best Paper A w a r d " by the Industrial and Cogeneration Committee o f the Intemational Gas Turbine Institute.
PAPER n
Gabra, M . , Salman, H . and Kjellström, B .
Development o f a Sugar Cane Residue Feeding System for a Cyclone Gasifier.
Biomass and Bioenergy V o l . 15 N o . 2 pp 143-153, 1998.
Mohamed Gabra has written this paper. The experimental w o r k was made i n cooperation w i t h Hassan Salman. The paper was reviewed before submission by Professor Björn Kjellström.
PAPER III
Gabra, M . , Pettersson, E., Backman, R. and Kjellström B .
Evaluation o f Cyclone Gasifier Performance for Gasification o f Sugar Cane Residue - Part 1 Gasification o f Bagasse
Accepted f o r publication i n Biomass and Bioenergy
PAPER I V
Gabra, M . , Pettersson, E., Backman, R. and Kjellström B .
Evaluation of Cyclone Gasifier Performance f o r Gasification of Sugar Cane Residue - Part 2 Gasification o f Cane Trash
Accepted f o r publication i n Biomass and Bioenergy
Mohamed Gabra has written these two papers. The gas analysis measurements and tar analysis have been done by Esbjörn Pettersson. Ash melting tests and the theoretical calculations o f the melting o f the ashes that were collected f r o m the exhaust gas after the boiler and before the chimney, have been done by Rainer Backman. The papers were reviewed before submission by professor B j ö m Kjellström.
PAPERV
Natarajan, E., Ö h m a n , M . , Gabra, M . , Nordin, A . , Liliedahl, T. and Rao A . N .
Experimental Determination of Bed Agglomeration Tendencies o f Some Common Agricultural Residues i n Fluidised Bed Combustion and Gasification.
E. Natarajan has written this paper. Mohamed Gabras contribution is that he provided this smdy w i t h the bagasse and cane trash fuel and by all information and analysis needed about these two types o f fuels. He also participated i n the experimental w o r k and by discussions and remarks i n the preparation o f the manuscript.
P A P E R V I
Gabra, M . , Ö h m a n , M . , Nordin, A . and Kjellström, B .
A l k a l i Retention/Separation during Bagasse Gasification A Comparison between a Ruidised Bed and a Cyclone Gasifier
Manuscript to be submitted to Biomass and Bioenergy (2000)
Mohamed Gabra has written this paper. He has run the gasification tests i n the F B G w i t h the assistance of Marcus Ö h m a n . Anders Nordin has done the SEM/EDS analysis.
Mohamed Gabra has also made the equilibrium calculations o f the phase distribution o f the alkali compounds at different pressures and temperatures. Anders Nordin and Marcus Ö h m a n have contributed i n discussion and i n result analysis. The paper was reviewed before submission by professor B j ö m Kjellström.
Additional publications and internal reports of relevance in the present work, but not included in the thesis are:
1. Gabra, M . and Kjellström, B . A Pre-Feasibility Assessment o f the Potential o f Cane Residues f o r Co-generation i n the Sugar Industry. Working paper, published by the Stockholm Environment Institute, Stockholm i n collaboration w i t h S I D A , series N o . 4 1 , Sweden, 1995.
2. Sugarcane Residual Fuels - A Viable Substitute for Fossil Fuels in the Tanzania Sugar Industry. I n Renewable Energy for Development, August 1995, V o l . 8, No.2.
3. Biomass Gasification Technology for Developing Countries. I n Renewable Energy for Development, M a y 1996, V o l . 9, N o . l .
4. Kjellström, B . , Gabra, M . , Pettersson E. and Salman, H . Cyclone Gasification o f Different Types o f Biomass Powder at Atmospheric Pressure. Research funded i n part by the European Communities, Contract JOR3 CT98 0281, Activity Number 3.1, November 1999.
S C I E N T I F I C C O N T R I B U T I O N
The f o l l o w i n g scientific contributions are the result of this work:
• Reduced uncertainties about the relative merits o f different possibilities to increase electric power generation i n the cane sugar industries.
• Quantification o f the potential of cane trash as fuel.
• Development o f a feeding system f o r bagasse and cane trash that can be used to provide a cyclone gasifier w i t h a reliable and controllable fuel f l o w .
• Demonstration o f range of conditions f o r stable gasification of bagasse and cane trash i n a cyclone gasifier.
• Determination o f alkali content, the carryover particles i n the producer gas and the heating values of the producer gas f r o m bagasse/cane trash gasification i n a cyclone gasifier.
• Determination o f agglomeration temperature o f bagasse and cane trash during combustion and gasification i n a fluidised bed gasifier.
• Detennination o f alkali retention i n a fluidised bed gasifier using bagasse as a feedstock.
• Widening o f the understanding o f the complexity of phenomena that control alkali retention i n a fluidised bed gasifier.
• Further confirmation that chemical equilibrium predictions overestimate the formation o f gaseous and liquid ash species under biomass gasification conditions.
O V E R V I E W
Biomass is the most important source of energy i n developing countries and also plays a significant role i n a number o f industrial countries. The world drives 14% o f its energy f r o m biomass, which is about 25 million barrels o f oil equivalent per day. About 35 % o f the total energy i n the developing countries is f r o m biomass [Hall, 1993].
Sugar cane is one o f the most efficient crops f o r biomass as a source o f energy. It has some o f the most advantageous properties o f biomass. Apart f r o m being used to produce sugar and molasses, sugar cane grows faster and produces more biomass than most o f energy crops. Due to its wide distribution and abundance i n many countries, sugar cane can therefore be an important option f o r the replacement o f fossil fuels.
The potential o f cane residuals as a source o f energy could help many developing countries to improve their balance o f payments by decreasing fossil fuels imports and increasing the electrical power production.
There are two types o f residual fuels generated by the sugar industry, namely bagasse and cane trash. Bagasse is cane residue, which is left over after the extraction o f sugar.
Cane trash is the leaves and tops o f cane plant that are usually burned before harvesting.
This study was started by making an energy audit f o r a small sugar factory i n Tanzania, named TPC (Tanganyika Planting Company), see Jenefors, A et. al, [1993]. The objective o f the TPC study was to provide a basis f o r evaluation o f possibilities to increase electricity generation i n a typical sugar factory w i t h a back pressure steam power plant using data on actual performance.
The first part o f m y study was to carry out such an evaluation. The results are presented in paper I and i n more complete f o r m i n [Gabra and Kjellström, 1995]. I n the second part o f m y study I focused on biomass gasification and connecting gasifiers to advanced electric power system. The reason for this emphasis is that gasifiers can be connected to advanced power systems with high efficiencies and very low emissions. I would like also to contribute i n solving some o f the problems associated w i t h this technology and to improve technical performance o f the existing biomass power industry. The results o f this work are presented i n papers I I - V I .
I n paper I , different options for increasing the electricity generation i n sugar mills by using more advanced steam processes and combined cycle technology, using cane trash and bagasse as a fuel have been analysed. TPC sugar m i l l was selected as a case study for investigation. Introduction o f a combined gas turbine/steam turbine process w o u l d make i t possible to increase the electricity output f r o m 2.5 M W to 30 M W at this plant during m i l l i n g season. B y using cane trash as f u e l during the off-season period, the electricity generation can be increased by a factor o f 20 compared to what is generated at TPC sugar factory during the time of this study i n 1995.
The financial evaluation indicated that the annual profit would range f r o m US$ 3.5 m i l l i o n f o r the advanced steam process with 6.5 years pay-back time, to US$ 4.7 m i l l i o n f o r the combined gas turbine/steam turbine process w i t h 6.8 years pay-back time.
In the comparisons between the different options the study concluded that the highest electric yield could be obtained by using a combined gas turbine/steam turbine integrated w i t h gasifier supplying combustible gas to gas turbine combustor. However, using producer gas f r o m biomass gasification to fuel a gas turbine still has a number o f technical uncertainties. Gas turbines are generally designed to run on very clean fuel such as natural gas or light petroleum.
Sugar cane residue (bagasse and cane trash) has relatively high ash content with elements i n the ash, which may f o r m compounds, which are released during the gasification processes. Some o f these elements, such as alkali elements (K+Na) may f o r m corrosive species that can cause severe corrosion damage to the turbine blades i f not suppressed. A s h particles carried into the gas turbine may also lead to deposition problems or erosion, depending on the size and the properties o f the particles.
Gas turbine hardware consideration w i l l impose constraints o f the level o f particulates and alkali metals i n the producer gas delivered to the gas turbine. The current industrial gas turbine specification l i m i t f o r alkali compounds i n the producer gas entering the combustion chamber is o f the order o f 0.1 ppmw. The particle concentration entering the gas turbine f o r an economic durability gas turbine must be kept below 4.25 ppmw.
For particles size i t is required that no particles larger than 10 p m be present i n the producer gas.
The levels f o r alkah metals and particulates that can be tolerated by a gas turbine are not w e l l established because o f a lack of operating experience. Contaminant limits presently specified by gas turbine manufacturers therefore are probably conservative.
A B B Stal has developed a gas turbine (GT 35P model) especially for coal combustion in PFBC (Pressurised Fluidised Bed Combustion) combustors, which has high capability to cope w i t h high dust loads and high alkali content. It w i l l be assumed that the experiences f r o m coal fired gas turbines can be applied also to gas turbines fuelled with biomass. The experience f r o m the PFBC plants indicate that alkali content o f the producer gas o f about 70 mg/kg (wet-gas) can be used without serious problems i f there are particles i n the burned gas that can act as condensation nuclei. The particle concentration i n the producer gas o f 1400-4200 ppmw is considered to be i n the acceptable range f o r A B B Stal G T 35P model w i t h low aerodynamic blade loading, i f most o f the particles are small (less than about 10pm). For this study A B B Stal G T 35 P model criteria have been used to evaluate the results.
There has been a considerable amount o f bench-scale and pilot plant research conducted i n Europe and U.S to develop methods f o r clean up o f the producer gas f r o m biomass gasification that are suitable f o r operation o f a gas turbine. Most o f these projects are using a fluidised bed gasifier (FBG) and include extensive gas cleaning which leads to a high capital investment.
Electricity f r o m biomass fuels w i l l compete with fossil fuels. The challenge is to f i n d a method to reduce the cost o f gas cleaning that makes gasification o f biomass financially attractive.
In this thesis a cyclone gasifier (CG) process has been studied and evaluated. The
gasification i n a C G and F B G also has been conducted to evaluate which gasification process is most attractive as regards alkali retention/separation.
One o f the most important problems associated w i t h the use o f a biomass fuel i n a C G is the feeding o f the fuel. I n paper I I a new method to improve the feeding system and to provide the C G with a reliable and controllable fuel f l o w o f bagasse/cane trash has been used. The first feeding tests were performed with crushed bagasse and crushed cane trash. The low bulk density and cohesive character o f crushed bagasse/cane trash powder initially created a great number o f difficulties concerning the f l o w o f the fuel f r o m the f u e l feeding bin to the cyclone. To improve the characteristics o f the fuel and to ehminate the feeding problems, the crushed bagasse/cane trash was pelletized and then ground. It was found that this treatment changed the shape o f the slivers o f ground bagasse/cane trash fuel, the bulk density increased and made the fuel more homogeneous. Screw calibration tests f o r ground bagasse/cane trash were performed and the system worked smoothly. The mass f l o w rate was found to increase linearly w i t h increasing screw speed. The fluctuations o f the feeding rate o f ground bagasse/cane trash were found to be within reasonable limits.
Papers I I I and I V cover the results o f cyclone gasification tests at atmospheric pressure w i t h sugar cane residue.
Paper I I I presenting results f r o m bagasse gasification experiments shows that the heating values o f the producer gas are i n the range o f 3.5-4 M J / N m3- (dry gas) that w o u l d correspond to approximately 2-2.5 MJ/kg wet-gas. N o problems i n igniting and burning the producer gas were experienced during the gasification tests. Significant alkah separation has been achieved i n the cyclone stage. However the total alkali contents o f the producer gas were found to be higher than i n A B B Stal PFBC gas turbines and at least an order o f magnitude higher than what is required by most gas turbine manufacturers f o r operation o f a gas turbine. The carryover particles concentrations i n the producer gas were found to be i n the range o f that for PFBC gas turbines, but higher than what is required by most gas turbine manufacturers for operation o f a gas turbine. The investigations o f the particles entrained w i t h the producer gas by using Scanning Electronic Microscope (SEM) give indication that most o f the particles are below 10 p m i n size. The f l y ash-melting tests have not shown any major ash melting up to 1200 °C, but it was found that some o f the particles entrained w i t h producer gas were partially melted.
Paper I V presenting results f r o m cane trash gasification experiments shows that the heating value o f the producer gas is on the average o f 4.7 M J / N m3 (dry gas) that w o u l d correspond to approximately 3.35 MJ/kg (wet gas), which is sufficient f o r stable gas turbine combustion. I t appears that the heating values o f the producer gas f r o m cane trash gasification tests are higher than the heating values o f the producer gas f r o m bagasse gasification tests. The reason is certainly that steam was used as injection medium i n the bagasse gasification tests whereas air was used for injection o f cane trash.
Significant alkali separation has been achieved i n the cyclone stage. However the alkali levels and particle concentrations i n the producer gas were found to be higher than allowable i n a gas turbine. The alkali compounds entrained i n the product gas f r o m cane
trash gasification are considerably higher than observed for gasification o f bagasse. The reason is simply the higher content o f these elements i n the cane trash fuel.
The f l y ash-melting tests have not shown any major ash melting up to 1000 ° C , but the deposition problems can not be excluded since some particles i n the producer gas seem to have been melted and since some gasification o f K and Na-compounds is indicated.
I n paper V the agglomeration temperatures f o r bagasse and cane trash i n combustion and gasification has been experimentally determined i n a bench-scale-fluidised bed gasifier. Both the combustion and gasification experiments were conducted i n two different bed materials, quartz and lime. It was found that the initial agglomeration temperature o f bagasse w i t h both quartz and lime bed materials i n combustion and gasification is more than 1000°C, which is much higher than the gasification temperature selected f o r the study o f paper V I . For cane trash the agglomeration temperatures w i t h quartz and lime bed materials i n combustion and gasification were found to be i n the range o f 800 to 900 °C.
I n the comparative study presented i n paper V I the alkali retention in the F B G was found to be i n the range o f 12% - 4 % whereas i n the C G the alkali separation was f o u n d to be about 70%. The SEM/EDS indicates that no significant coating of the F B G bed material particles is formed. The retention i n the F B G i n this study is mainly caused by the ash particles accumulated i n the bed. The elemental maps analyses found that all the ash is found i n the bed material distributed as solid ash particles, freely moving and dispersed between the bed particles, consisting mainly o f Si, A l and K . It might be assumed that those elements selectively formed some kind o f potassium aluminium silicate w i t h a relatively high melting temperature, which is continuously scavenged and lost f r o m the bed through elutriation. The results f o r retention in the FBG deviate f r o m those presented b y Turn [1988] f o r gasification o f bagasse, which indicate about 4 0 % retention. Different explanations for this discrepancy are possible. The different compositions o f the bed material used i n the t w o studies could have influenced the formation o f coating on the bed particles.
EquUibrium calculations show that about 10-15% o f the K and Na w i l l be i n gas phase at the operating temperature o f the gasifiers and that the remainder should be i n l i q u i d phase. However, no melting could be observed i n any o f the gasifiers. This confirms conclusions drawn by others that equihbrium calculations overestimate the formation o f gaseous and liquid phases o f biomass ashes i n gasifiers.
Integrated experiments with a gas turbine need to be done f o r accurate evaluation o f the possibilities to use the producer gas f r o m the gasification o f cane residues to run a gas turbine without problems o f hard deposits and corrosion on the turbine blades.
1 INTRODUCTION
Biomass is an important energy resource. It is broadly defined as trees and other plants, which have stored chemical energy f r o m sunlight through the process o f photosynthesis. The net production o f biomass by plant photosynthesis has been estimated to be 120 billion dry tonnes per year, having an energy content equivalent to over 5 times the present world energy demand [Brascep, 1992]. Although biomass contributes to about 14% o f the world energy input (ranging f r o m 3% i n the industrialised countries to 38% i n the developing world), much o f which is unaccounted f o r i n commercial energy statistics, it remains a significant but largely ignored resource [Morris and Waldheim, 2000].
The energy i n biomass may be realised either by direct use as i n combustion, or b y upgrading into a more valuable and usable fuel such as fuel gas or a combustible liquid. The potential for producing gaseous and liquid fuels f r o m biomass can only be realised i f new technologies are developed. Improved conversion technologies are needed to produce a variety o f gaseous and liquid fuels at competitive costs.
I n recent years a renewed interest i n biomass gasification has evolved. The basic principles of gasifier technology have been studied and developed since the early 19th century. A technology to use solid fuels for internal combustion engines was developed and extensively used i n European countries during W o r l d War I I . Tractors, cars, railway engines, fishing boats, sawmills, pumping stations and other industrial machinery were run on gas produced f r o m the partial combustion o f charcoal and wood.
Under the combined pressures o f rising o i l prices, balance o f payments difficulties, and unreliable f u e l supplies, many developing countries find themselves i n a position similar to that o f a wartime siege economy. Thus, the incentive to develop alternative fuel sources is strong, and gasification o f biomass is one amongst a range o f possible choices.
Sugar cane has some o f the most advantageous properties o f biomass. Apart f r o m being used to produce sugar and molasses, sugar grows faster and produces more biomass than most of energy crops. Sugar cane residue (bagasse and tops plus leaves - called cane trash) can o f f e r a significant potential for generation o f electricity [Gabra and Kjellström, 1996].
Using the cane sugar industry f o r electricity generation also has a positive environmental impact. The cane sugar industry may contribute to the reduction o f green house gas emission by replacing fossil fuelled thermal power plants w i t h biomass powered plants. Compared to oil and coal fuels, biomass has a much lower sulphur content. This reduces SO2 emissions and lowers the risk o f soil and water acidification. The ash content i n biomass is lower than i n coal, meaning fewer problems with ash disposal and leakage o f heavy metals
Better utilisation o f cane residues i n the sugar industry, can be an important option f o r the replacement o f the fossil fuel and f o r increasing electricity generation i n developing countries, due to its wide distribution and abundance i n many o f the developing countries. A substantial increase o f the electricity output i n the sugar industry could be achieved i f cane trash is used as a fuel during the off-season period and i f the backpressure steam turbines with l o w admission data that are presently being used i n most o f sugar factories are replaced by a combined gas turbine/steam turbine integrated w i t h a gasifier supplying combustible gas to the gas turbine combustor. The utilisation o f this technology f o r co-generation i n sugar mills is however
prevented by the lack of proven technology f o r utihsation o f the combustible gas f r o m the gasification o f cane residues as fuels f o r running of gas turbines. The problem is to f i n d a way to avoid those ash-forming mineral elements i n the solid fuel that cause erosion, deposits and corrosion i n the gas turbine.
There has been a considerable amount o f bench-scale and pilot plant research conducted i n Europe and U.S to develop methods for clean up of the producer gas from biomass gasification that are suitable f o r operation of a gas turbine. Most o f these projects are using a fluidised bed gasifier and include extensive gas cleaning which leads to a high capital investment. Electricity from biomass fuels w i l l compete w i t h fossil fuels. The challenge is to f i n d a method to reduce the cost o f gas production and gas cleaning that makes gasification o f biomass financially attractive.
I n the present w o r k producer gas f r o m gasification o f cane residues i n an atmospheric cyclone gasifier (CG) has been studied and evaluated with respect to its suitability f o r operation of a gas turbine. The cyclone gasifier works as particle separator as well as a gasifier. This process, i f it shows acceptable performance, could open possibilities to reduce the investment required f o r the gasifier and the gas cleaning. Comparative bench-scale tests of bagasse gasification i n an isothermal fluidised bed gasifier (FBG) and in the (CG) have also been done, to evaluate which gasification process is most attractive as regards alkali retention/separation.
2 O B J E C T I V E S OF T H E THESIS
The objectives o f this thesis are:
• T o study different possibilities for increasing the electricity generation i n the sugar industry.
• T o investigate the possibilities of using a producer gas f r o m gasification of cane residues in a gas turbine.
3 BIOMASS E N E R G Y CONVERSION
Biomass conversion technologies can be separated into two basic categories; thermochemical conversion processes (pyrolysis, gasification and direct combustion), and biochemical processes. A summary of the main characteristics o f thermochemical conversion processes can be found i n Bridgwater [1984], [1994] and [1995], Bridgwater and Cottam [1992]. Extensive reviews o f gasification, combustion and biogas production processes can be found i n K j e l l s t r ö m and Svenningson [1990], Strehler [1990], Nordin [1993] and Ellegård [1990].
3.1 Pyrolysis technology
Pyrolysis is a process for thermal conversion o f sohd fuel i n the complete absence o f oxidising agent (air/oxygen), or w i t h such limited supply that gasification does not occur to any appreciable extent. Heat is usually added indirectly i n a variety o f forms. The pyrolysis process takes place at temperatures in the range o f 200 - 800 °C. Most o f the cellulose and hemi- cellulose and part of the lignin w i l l disintegrate to f o r m smaller and lighter molecules, which are gases at the pyrolysis temperature. As these gases cool, some of the vapours condense to f o r m a liquid, which is the bio-oil. The remaining part of the biomass, mainly parts o f the lignin, is left as a solid i.e. the charcoal [Nordin and Kjellström, 1996].
B y changing the rate o f heating and the final temperature it is possible to m o d i f y the proportions o f the gas, liquid and solid product [Zanzi, 1994]. A relatively slow reaction at low temperatures is required to maximise solid char yield. Fast pyrolysis at l o w temperatures up to about 650 °C maximises the yield o f condensable vapours. About 70% o f the dry biomass weight can then be obtained as liquid [Nordin and Kjellström, 1996]. Fast pyrolysis at higher temperamre above 700 °C maximises gas yields. Sufficiently high heating rates o f biomass convert nearly all of the organic solid fuel into volatiles. Rapid pyrolysis at high temperature plays an important role as the initial step i n gasification and combustion [Zanzi, 1994].
Commercial applications o f biomass pyrolysis are either focused on production of charcoal or production o f a liquid product, bio-oil. Charcoal has a market f o r certain industrial purposes and is used as a smokeless fuel. The bio-oil could be used as a substitute f o r fuel o i l and as feedstock f o r production of synthetic gasoline or diesel fuel. Pyrolysis f o r charcoal production is carried out i n a fixed bed or a fluidised bed. Pyrolysis f o r bio-oil production is made i n fluidised bed reactors.
3.2 Biomass gasification technology
3.2.1 Overview
The gasification is a process f o r converting a biomass fuel into combustible gases b y a partial combustion process. The oxidising agent is air or oxygen and often steam is added as a gasifying agent to reduce the reaction temperature and f o r temperature control. This process comes between the pyrolysis process and complete combustion i.e. more oxidising agent is used than f o r pyrolysis but less than for complete combustion.
The combustible fraction of a solid fuel can be divided into volatile and non-volatile fractions.
The overall rate o f gasification o f the biomass particle depends upon individual rates of the processes involved, i.e. drying, release of the combustible volatiles, mixing of the volatiles vapour and the oxidant, combustion o f the volatiles and the gasification o f non-volatile combustibles. The rates o f these individual processes depend upon the size o f a fuel particle, the heat transfer w i t h surroundings and the gas composition i n the vicinity of the particle. The amounts o f fuel gasified w i l l also depend on the residence time, which w i l l be quite different i n different types o f gasifiers.
The combustible gas generated by gasification, sometimes called producer gas, can be used as a fuel i n a variety of applications. The producer gas consists o f carbon monoxide, hydrogen and methane as the main combustible components. There are also other organic compounds i n low concentration. Some o f these are condensible and are generally referred to as " tar v\ The balance consists o f carbon dioxide, water vapour and nitrogen.
The heating value of the producer gas and its potential uses depend on how the gas is made. A producer gas o f low energy is produced by partially combusting a biomass fuel w i t h air or air/steam. A i r contains about 79 % nitrogen. This nitrogen dilutes the product gas. For this reason, the air-blown gasification process generates fuels w i t h lower heating values o f about 3 to 6 M J / N m3. This gas is suitable for operation o f boilers or engines (diesel engines, gas turbines etc).
Oxygen gasification gives better quality gas with medium heating values o f 10 to 20 M J / N m3 but requires oxygen supply w i t h associated problems o f cost and safety. Such a medium heating value gas can also be produced by pyrolysis and by more complex technologies such as t w i n fluidised beds. The medium energy gas can be used for a variety of industrial processes such as synthesis gas to make methanol, gasoline or ammonia.
A i r - b l o w n gasification is the simplest and most widely used technology since the gaseous product is formed at high efficiency without requiring air separation and without heat transfer problems as i n pyrolysis. The chemical reactions occurring during the air-blown gasification process depends on gasifier design, fuel type, particle size distribution and operating conditions.
3.2.2 Types of gasifiers
A wide range o f gasifiers are currently being used and developed for producing combustible gases f r o m biomass. Most o f these f a l l into the two major categories of fixed-bed and moving-
grates. M o v i n g bed gasifiers include fluidised bed, entrained bed, and other types o f reactors.
A l l must include basic features such as an effective feed system, a method to inject air, oxygen, or heat, a means to remove and clean the gas, and a way to remove ash and unreacted char.
Each gasifier type has specific advantages and limitations, which have direct implications for its performance characteristics. A gasifier which is well suited for some applications may be unsuited f o r others. N o single type o f gasifier is appropriate for all applications. The ideal design should be able to produce a clean producer gas f r o m a wide range o f fuels, i t should be capable o f w o r k i n g efficiently; it should be able to respond rapidly to changes i n load and the investment and running cost should be reasonable.
The traditional gasifer types that i n principle can be used i n electricity production f r o m biomass are up-draft, down-draft, fluidised bed and entrained bed. However i n practice the usability of these processes is limited mainly to updraft and fluidised bed gasifiers [Kurkela et al., 1993].
The capacity required and the gas quality are the important parameters f o r selection o f the most suitable type of gasifier f o r electricity production. The maximum capacity of up-draft counter - current gasifier is about 10 M W (fuel) whereas fluidised bed gasifiers can be designed f o r at least f i v e times larger capacities. Cyclone gasification could be another possibihty but this approach has not as yet been used i n commercial installations
The problem o f integrating a fluidised bed gasifier with a combined gas turbine/steam turbine process to supply a producer gas to a gas turbine combustor is that the producer gas is contaminated w i t h a high dust load and corrosive species that can cause severe damage to gas turbine blades. Therefore an extensive gas cleaning must be used before the gas is burned to generate the hot gas that drives the gas turbine. This leads to relatively high specific investment for power plants o f small and medium capacity. A n alternative option is to gasify the biomass fuel i n a cyclone gasifier. The cyclone gasifier works as particle separator as well. For this study a special attention has been focused on a cyclone gasifier and fluidised bed gasifier.
The air-blown gasification process i n a cyclone gasifier and a fluidised bed gasifier w i l l be discussed below.
3.2.3 Mechanism of gasifying a biomass particle in a cyclone gasifier
A cyclone gasifier is characterised by short residence time and heat transfer to the particles mainly by radiation f r o m the cyclone walls and reacting fuel particles, convection f r o m hot gas and to some extent contact w i t h the cyclone walls. The fuel particles are small, generally less than a f e w m m as m a x i m u m size. The residence time has been predicted to less than 1 second, see Fredriksson [1999].
As soon as a biomass f u e l particle enters the hot cyclone, it dries and is pyrolysed, w h i c h implies that a selfsustaining exothermic reaction takes place i n which the natural structure o f the biomass particle breaks down and devolatilisation starts. The volatile combustibles are released and mixed w i t h surrounding air. A diffusion flame is stabilised around the particle where the combustibles and the oxygen f o r m a flammable mixture. Little oxygen can penetrate through the flame into the f u e l particle. During this process, the size o f the particles is only slightly reduced but the particle density is decreasing. The result is a residual solid (char) and a gas mixture composed primarily o f carbon dioxide, carbon monoxide, hydrogen, water vapour, nitrogen and pyrolysis products including tar and hydrocarbons. Biomass consists o f 75 %-85 % volatile
matter compared to half or less this level with coal, so pyrolysis plays a larger role i n biomass gasification.
A f t e r the oxygen around the particle is consumed and the volatile flame is extinguished, the char w i l l be gasified by reactions with the carbon dioxide and water vapour to give a fuel gas composed mainly o f CO, H2, CH4, C2H4 and C2H2. Tar and hydrocarbons may also react or crack to f o r m simpler products. Most of the hydrogen that is produced remains free. However, a portion o f it can combine w i t h carbon to f o r m small amounts o f methane and hydrocarbons [Foley and Barnard, 1983]. I f the amount of steam injected w i t h the fuel is high or the moisture content i n the fuel is high, the so-called water shift reaction can also be important. I n the water shift reaction the carbon monoxide reacts with water to produce carbon dioxide and hydrogen.
This is an unfavourable reaction since i t reduces the calorific value o f the final gas. I n the case o f excess steam present during the gasification process, a considerable proportion o f steam normally passes through the cyclone gasifier without reacting and simply becomes a component i n the final gas stream.
The complete reactions of all tars, hydrocarbons and char depend on the equivalence ratio, the geometry o f the cyclone gasifier and the residence time.
A fraction o f the ash forming compounds may be volatilised and released to the gas phase during the gasification process. A part o f these volatiles may react or condense on the surfaces o f the char/fly-ash particles and change the surface composition and size o f them. The volatilisation fraction depends on the ash composition of the fuel, the particle size, the particle temperamre and the gasification pressure.
Cyclones or centrifugal separators have been used since the 1880s. They are the most widely used technique to clean a gas f r o m solid and also they are used to separate fluids o f different densities, and solid f r o m hquids. Cyclones are simple and inexpensive to make, relatively economical to operate, and are adaptable to a wide range of operating conditions. The design and operation o f cyclones to remove dust f r o m a gas stream have been set on a practical basis by Stairmand, [1951] and ter Linden, [1949]. Also a lot of work has been reported where the cyclone separator technique has been studied, see f o r example Alexander [1949], Lapple [1950]
and Ogawa [1948.]. Some work also has been done to model the hydrodynamics i n the cyclone to predict the separation efficiency and pressure drop, see f o r example Leith and Licht [1972];
Dietz [1981]; Mothes and L ö f f l e r [1988]; Iozia and L e i f h [1990]; Ramachandran et al. [1991];
Perry and Green [1984].
The design o f a cyclone used as gasifier or combustor f o r coal has been studied f o r example by Syred and Beer [1974]. Only a f e w studies on gasification of wood fuels i n cyclones have been reported i n the literature. Experiments with cyclone gasifier using a wood powder as feedstock have been done by Cousins and Robinson [1985], Kallner et al. [1994], Kjellström et al. [1998], K j e l l s t r ö m et al. [1999]; Fredriksson [1999]. No studies on gasification of cane residues i n a cyclone gasifier have been reported.
3.2.4 Gasification of biomass particles in a fluidised bed gasifier
A fluidised bed can accept a wide range of particle sizes. The feedstock is fed continuously and is fluidised together w i t h an inert heat-distributing material e.g. sand, by oxidant and/or steam injected f r o m below. Fluidised bed gasifiers have a good m i x i n g o f the bed and excellent
control of blast air and they operate w i t h intense m i x i n g o f the fuel and the reacting gases. The excellent heat and mass transfer i n fluidised-bed gasifiers lead to relatively uniform temperature throughout the bed. Therefore the temperature can be kept at relatively low temperature level (about 900 °C), to prevent ash melting. This temperature range is below the fusion temperature o f most types o f ash. Slag and clinker formation can therefore be avoided.
This opens the possibihty to handle a wide range o f biomass fuels including the low-density feedstock like cane residues [Nordin and Kjellström, 1996].
The mechanisms f o r conversion o f a biomass particle to a combustible gas are the same i n the fluidised bed and the cyclone. The heating rate of the particle may be higher i n the fluidised bed, which would result i n a higher yield of volatiles. The temperature can be expected to be more u n i f o r m i n the fluidised bed than i n the cyclone gasifier. This may have implications for the risk for local vaporisation or melting of ash species.
The most important difference is the residence time that is quite short in the cyclone, leading to formation o f a char residue. I n the fluidised bed, the particle w i l l remain until it is either small enough to be carried away by the gas f l o w or happens to be leaving the bed together w i t h the material removed to control the bed level and the ash content i n the bed.
3.3 Gas quality criteria for gas turbine operation
Using producer gas f r o m biomass gasification to fuel a gas turbine still has a number of technical uncertainties. Gas turbines are generally designed to run on very clean f u e l such as natural gas or light petroleum.
Sugar cane residue (bagasse and cane trash) has relatively high ash content. The inorganic compounds i n the ash may be released during the gasification processes. I f some o f these compounds are deposited i n liquid f o r m on the metal surfaces i n the turbine, severe corrosion damage can occur. A s h particles carried into the gas turbine may also lead to deposition problems or erosion, depending on the size and the properties of the particles.
The elements that are o f main concern f o r the deposition and corrosion problems are Ca, Si, K, N a and C l . Ca and Si are mainly concern o f the hard deposition i n gas turbines. The presence of alkali elements (K+Na) i n the producer gas above a certain level can cause the corrosion problems. Also Cl has an influence on increasing corrosion problems. The higher the C l content i n the feedstock, the higher the vapour-phase alkali concentrations i n the producer gas [Mojtahedi and Backman, 1989]. A t equilibrium, K C l and NaCl have been predicted to be the predominant alkali gas species under gasification conditions and volatile compounds o f K and N a were found to increase w i t h increasing fuel C l contents [ T u m , 1988].
I n order to avoid deposition and corrosion problems when fuels containing K and Na are used, many gas turbine vendors require that the contents o f K and Na compounds in the gas entering the gas turbine must be below a certain value. The current industrial gas turbine specification l i m i t f o r alkali compounds i n the producer gas entering the combustion chamber is o f the order o f 0.1 ppmw [Scandrett, 1983, Scandrett, and C l i f t , 1984, Singh et al., 1986, Mojtahedi and Backman, 1989 and Hallgren, 1994 ] .
Erosion wear is markedly influenced by the size and concentration o f particles entrained with producer gas. Research i n the U.S. [Hallgren, 1994] has stated that f o r an economic durability
gas turbine, the f u e l gas dust entering the gas turbine concentration must be kept below 4.7 m g / N m3 that w o u l d correspond to approximately 4.25 ppm. For particles size it is required that no particles larger than 10 p m be present i n the producer gas [Mingxian et al., 1996]. It was found that the rate of erosion was insignificant when the impacting particles were below 10pm [Raask, 1984].
The levels o f alkali metals and particles entrained w i t h the producer gas that can be tolerated by a gas turbine are not well established because o f a lack of operating experience. Contaminant limits presently specified by gas turbine manufacturers therefore are probably conservative.
A B B Stal use a different approach for gas turbines adapted to dirty fuels and requires that the melting temperature o f the ash compounds entering into the gas turbine shall exceed the material temperature i n the turbine [Nilsson, 1973]. A B B Stal has developed a gas turbine (GT 35P model) especially f o r coal combustion i n PFBC (Pressurised Fluidised Bed Combustion) plants. This turbine has a capability to cope with high dust loads and high alkah content. The experience f r o m coal fired gas turbines might be used also i n biomass fired gas turbine applications. Gas quality requirements for this turbine are discussed by Strand [2000]. Jansson [1996] has also provided some data that may use f o r estimation of acceptable particle loads. It should be observed that the burned gas f l o w w i l l be about 7 times the product gas f l o w for a product gas temperature of 800 °C and a turbine inlet temperature o f 850 °C. For this study A B B Stal G T 35P model criteria have been used to evaluate the results.
• Effective heating value o f wet product gas above about 2.5 MJ/kg
• Contaminants i n product gas 1. Dust load
• Total below about 3000 mg/kg
• > 8 p m below about 70 mg/kg 2. Potassium and sodiun below about 70 mg/kg
Details about the alkah level and particles that were found i n the producer gas f r o m cane residue gasification i n the cyclone gasifier and the fluidised bed gasifier w i l l be presented i n section 11.
Tar that is formed i n gasification is a complex mixture of organic compounds ranging f r o m light compounds like benzene to heavy polyaromatic hydrocarbons. The temperature at which fuel pyrolysis takes place has a very pronounced effect on tar composition and accordingly tars can be divided into l o w and high temperature tars. L o w temperature tar is formed at temperatures lower than about 650 °C and it is mainly composed o f components that are primary decomposition products of the fuel structure. High temperature tar is composed of mono- and polyaromatic compounds formed mainly i n the secondary reactions of the primary pyrolysis products o f fuel [Simell, 1997]. The amount o f tar produced during the gasification process depends on the type o f the fuel used. For instance cane trash fuel used i n this study produces more tar than the bagasse fuel. Tars constitute an important enough energy component o f the f u e l gas that removing them f r o m the gas would result i n a loss of system efficiency. I f tar condenses on cool surfaces, severe operation problems can result i n different parts o f the system. The approach used i n this study f o r dealing with tar problems is to maintain the temperature of the producer gas above the dew point of tars all the way f r o m the gasifier to the gas turbine combustor. In the cyclone gasification process studied in this thesis, the increase o f heating values o f the producer gas caused by the presence of tar formed i n the cyclone
4 A R E V I E W O F F U E L R E S I D U E G E N E R A T E D I N T H E S U G A R I N D U S T R Y
4.1 Overview
Cane sugar is grown i n 79 countries within the tropic and subtropical belt. Most of these countries are underdeveloped and suffer f r o m the lack of fossil fuels. Table (1) shows a list o f developing countries where cane sugar is grown [Suzor et al., 1991].
T A B L E 1. Developing countries where sugar cane is grown
Central America/Caribbean Asia A f r i c a
Barbados Bangladesh Angola
Belize Burma Cameroon
Costa Rica China Chad
Cuba India Congo
Dominican Republic Indonesia Egypt
E l Salvador Iran Ethiopia
Guadeloupe Iraq Gabon
Guatemala Malaysia Ghana
Haiti Pakistan Guinea
Honduras Philippines Ivory Coast
Jamaica South Vietnam Kenya
Martinique Sri Lanka Malagasy Republic
Mexico Thailand M a l a w i
Nicaragua M a l i
Panama Mauritius
St. Kitts Morocco
Trinidad-Tobago Mozambique
Nigeria
South America Oceanic Rwanda
Senegal
Argentina F i j i Somalia
B o l i v i a Western Samoa South A f r i c a
Brazil Sudan
Chile Swaziland
Colombia Tanzania
Ecuador Uganda
Guyana Zaire
Paraguay Zambia
Peru Zimbabwe
Surinam Uruguay Venezuela
There are two types o f residue available to all sugar factories processing sugar cane, namely bagasse as a process residue and cane trash as a field residue. The amounts o f these residues depend on a number o f factors such as, area o f cultivated land, yield per hectare, land fertility and climate, harvesting and transportation methods.
4.2 Bagasse as a residue fuel
The quantity o f bagasse produced per ton cane processed depends mainly on the fibre content o f the cane, which varies between 12% and 18%. Typically 0.3 to 0.35 ton o f bagasse at a moisture content o f 45%-50% is produced per ton o f cane [Andersson, 1985]. Paper I shows that the fuel resource available i n bagasse at 50% moismre would be about 2.5GJ/ton.
The fuel characteristics o f bagasse mainly depend on moisture and ash content. Bagasse normally contains around 50% moisture and 1 - 5 % ash [Kadyszewski, 1990]. Most ash i n bagasse comes f r o m soil brought with cane stalks and depends on harvesting practices. Some factories wash soil f r o m the cane before it is milled. Data for the composition and heating value o f bagasse and a few data on ash melting characteristics are shown i n table 2. The data f o r bagasse o f Sudanese origin, which has been used i n this study are also included. Most o f the data indicate an ash content of about 2%. The higher ash content reported i n Mexico and Sudan could be explained by different harvesting methods that lead to more contamination by soil.
T A B L E 2. Collection of data for bagasse fuel
Origin Ultimate analysis Higher dry ash free basis heating
% weight value MJ/kg
C 0 H
Cuba 44.5 49.3 6.2 18.6 2.9
Dominican 50.5 43.4 6.1 18.9 2.1 1240-1300
Republic
Hawaii 46.9 46.6 6.5 19.0 1.5
Java 46.9 46.4 6.7 20.1 1.9
Mexico 53.3 39.8 6.9 19.4 11.3
Peru 49.9 44.2 5.9 20.5 1.7
Puerto Rico 45.0 48.6 6.4 20.5 1.8
Tanzania 48.2 45.1 6.7 19.9 2.5
Sudan 45.2 41.8 5.4 18.2 7.4 1200-1500
Thailand 49.2 45.9 4.9 19.4 1.6 900-1040
Source: Gabra and Kjellström, [1995]
Ash Ash fusion content temperature dry basis °C
% weight
The ash content and the ash melting characteristics are very important factors for the assessment o f the possibihties to use bagasse directly as gas tarbine fuel. The ash characteristics are determined by the composition o f inorganic matter i n the ash and the temperature determines the composition and properties o f the inorganic compounds i n the combustion gases. This w i l l determine the risk for deposits on hot surfaces and the properties o f the deposits. Some metals i n the ash like potassium and sodium easily f o r m compounds that are volatile at temperature about 800 °C and cause deposition and corrosion. More details about the ash-forming elements o f bagasse and cane trash used i n this study can be seen i n section 8.
4.3 Cane trash as a residue fuel
Cane trash is the leaves and tops of cane plant that are usually burned before harvesting. Some use the Latin American word barbojo for this residue. Cane tops and leaves are produced i n abundance during the life cycle o f a cane plant. US A I D , [1986a] report that the burning o f fields before harvesting consumes potential boiler fuel that has been estimated to be the equivalent o f 2.5 to 4 barrels o f oil per acre.
Since the tops and leaves have been considered valueless, the volume o f this material has not been accurately measured. A study team i n the Dominican Republic, US A I D , [1989] concluded that on average 0.67 tons o f cane residues are left i n the f i e l d (at 50% moisture) f o r every ton o f cane stalks harvested
The agricultural benefits o f a blanket o f cane trash are widely recognised. These include retaining moisture i n the upper level o f soil, providing weed and erosion control, and enhancing soil fertility by addition o f organic matter. Experience in burned fields f r o m Thailand, Jamaica and Puerto Rico shows, that these benefits can be obtained with only a fraction o f the total volume o f tops and leaf, but the size o f the fraction that should be retained has not yet been clearly established [ U S A I D , 1987a; U S A I D , 1986b and USATD, 1989]. Experiments i n Puerto Rico suggest that 30 to 50% o f the cane trash is sufficient to maintain the organic content o f the soil [ U S A I D , 1989]. Long-term studies were recommended to f u l l y assess the agronomic effects of cane trash removal, but the initial trials indicated that increased weed growth and decreased soil moisture retention were not serious problems. Where ratooning is practised (as i n much o f Puerto Rico and Jamaica), a major concern appears to be with potential damage to the emerging crop during mechanical collection o f cane trash. To avoid damage to the new growth, i t has been recommended that the cane trash be recovered within one week after harvesting. Another concern is soil compacting that may occur with the use of heavy machinery, particularly on wet soil, which can reduce ratoon yields [USATD, 1987b].
The B S T team i n Thailand concluded that the collection o f 50% o f the available trash w o u l d be more than sufficient to sustain off-season power production at new cane power plants [USATD, 1989]. L e f t on the field after cutting the cane trash dried f r o m 50% to about 30% moisture w i t h i n 4 to 6 day's [USATD, 1987b]. The drying i n the field beyond six days is not important i n achieving a lower moisture content [USATD, 1987a]. Air-dried cane trash has a higher calorific value, depending on moisture content and ash content. Burning trash with 30% moisture significantly increases the heat release rate i n the boiler, reduces excess air requirements, and improves boiler performance[USAID, 1989].
Paper I shows that w i t h 50% collection and natural drying to 30% moisture, the fuel resource available i n cane trash would be about 2.7 GJ/ton o f cane. This means that the cane trash represents a fuel resource o f the same magnitude as the bagasse.
Utilisation o f the cane trash as fuel requires that the trash is collected, stored until needed, transported to the sugar m i l l and prepared for combustion. Collection o f cane trash requires that the trash is raked into windrows. This also facilitates drying before collection and baling. The windrows are then collected by a pull-type forage harvester. Storage o f the bales can be outdoors near the cane field. This makes it possible to utilise the same transport system as used for the cane, when the bales are to be brought to the m i l l . The bales must then be shredded or disintegrated into smaller pieces that can be handled by the fuel transportation and feeding system designed f o r bagasse.
A feasibility smdy o f steam turbine power generation i n 1987, f o r Monymusk sugar factory i n Jamaica estimated a cost o f cane trash with 35% moisture content including cutting, baling, transportation and storage o f US$ 12/ton [ U S A I D , 1987b]. I n 1995 a smdy at TPC sugar factory in Tanzania, estimated that the cost o f baled cane trash with 30% moisture content would be US$ 16.8/ton [Gabra and KjeUstrom, 1995].
There are very f e w experiences world-wide regarding harvesting and using the cane trash as a boiler fuel. The harvesting and storage of cane trash for energy has not been done on a large commercial scale. However small-scale field trials have been conducted i n Florida and more extensive field trials w i t h three varieties o f cane have been carried out i n Puerto Rico [USATD, 1988; Chair, 1987 and Varua, 1987]. I n Central Roman sugar m i l l that is located on the southern coast o f the Dominican Republic, the harvesting o f cane trash has been i n operation since the early o f 1980:s. Manual harvesting o f cane trash plus mechanised chopping and recovery has been practised i n the Dominican Republic, since 1982 [Lopez, 1987]. Manual harvesting and delivery o f cane trash as a supplementary boiler fuel has been carried out f o r three seasons at a sugar factory i n the Philippines [USATD, 1986c].
5 B I O M A S S G A S I F I C A T I O N T E C H N O L O G Y D E M O N - S T R A T I O N F O R E L E C T R I C I T Y G E N E R A T I O N
5.1 Overview
This review o f the biomass gasification projects has been based on information provided by Morris and Waldheim, [2000], Ståhl, [1999], National Renewable Energy Laboratory (NREL) of U.S. D O E report [1998] and Gabra and Kjellström, [1995].
N R E L [1998] report that U S A biomass power plants using conventional steam turbines currently generate 7000 megawatts ( M W ) . Industry experts believe biomass gasifiers could help double the efficiency o f the current biopower industry average, now typically about 20%.
In addition to providing clean, economical energy, gasifiers greatly increase the types o f biomass fuels suitable f o r commercial power generation. In the United States biomass gasifiers could increase the generating capacity to 10,000 M W by 2010.
The concept o f advanced biomass power has already moved towards realisation. The interest i n biomass gasification f o r gas turbine operation is growing.
Results f r o m practical tests with wood powder fuelled gas turbine have been presented by Hamrick [1989] and Ragland et al. [1995]. Hamrick [1989] used a pressurised air-staged combustor where the wood was introduced downward in a vertical chamber. Ragland et al.
[1995] burnt the wood fuel i n an L-shaped downdraft, gravel-bed combustor. The test by Hamrick [1989] was hmited to 760 hours o f operation with oak sawdust as fuel. Ragland et al.
[1995] operated the gas turbine on wood chips f o r a total o f 255 hours.
The tests show that it is possible to operate a gas turbine with wood as fuel without visible erosion but with deposits on the turbine blades. The deposits were however possible to remove w i t h walnut shelling. In the test by Hamrick [1989], the deposits became glassy and could only be removed by scraping i f the turbine inlet temperature was higher than 790 °C. A n analysis of the deposits showed an enrichment o f potassium sulphate on the turbine blades compared to the corresponding amount i n the original fuel. Relatively high concentration o f CaO, M g O , K2S 04 and Ca(OH)2 i n the deposits was found i n the experiments by Ragland et al. [1995]. I n the tests by Hamrick [1989] the maximum particulate concentration was measured to 33 ppmw with a m a x i m u m o f 2.6 u m . Hamrick [1989] used one cyclone to clean the gas f r o m particles before entering the turbine. Ragland et al. [1995] did not include any hot gas cleanup and did not present any information regarding contents o f solids i n the gas to the gas turbine.
The Biomass Integrated Gasification Combined Cycle (B-IGCC) technologies are under development i n Brazil, England, United States, Sweden. The plants i n a demonstration scale based on B-IGCC i n these countries are V ä r n a m o i n Sweden, Arbre i n England and Vermont and Hawaii i n United States. A summary o f the process o f these plants as well as the demonstration gasification programme i n Brazil is given below. The specifications available for these plants are given i n table 3.
