Abstract The torrefaction of sugarcane bagasse particles was evaluated using standard thermogravimetric equipment (TGA) and in a torrefaction reactor explicitly designed for this work. Different particle sizes (fine, large, and unground bagasse) were torrefied to examine any potential effect of particle type and size on the devolatilization process. The custom-designed reactor allowed measurement of the mass and temperature of the particle and the condensable volatiles captured through a condensation unit held at −12 °C during torrefaction. Final products of the process (liquids, solids, and gases) were measured and characterized to understand and analyze the processes occurring in the sugarcane biomass during torrefaction. A two-step kinetic model was fitted to experimental data on fine particles to get the kinetic parameters. A particle model was developed to understand the dynamics of biomass heating and conversion under isothermal conditions. Results of tests for large particles show an effect of particle size on the process of decomposition of the biomass. The large unground biomass showed the lowest mass loss among evaluated particles. The torrefaction process resulted in significant thermal impact for fine particles, having the highest center temperatures, devolatilization, and changes in the O-H group with FTIR tests.

Thermal decomposition of sugarcane bagasse (SCB), yellow poplar (YP), and biomass polymers (xylan, cellulose and lignin) were studied under isothermal conditions using a specially designed thermogravimetric analyzer (Quartz Wool Matrix, QWM) in the temperature range 240-300 degrees C. Reactivity of biomass was predicted using the measured reactivity of three individual polymers and a simple arithmetic superposition principle. The superposition model was modified to increase the prediction accuracies of torrefaction behavior. To evaluate the impact of torrefaction on combustion, the combustion kinetics were also studied for the torrefied SCB and YP, and for the pyrolyzed YP. Activation energies for combustion reaction using the Flynn-Wall-Ozawa (FWO) method for the non-isothermal events were estimated as a function of the degree of conversion. Reactivity indexes were also determined for each sample when 50% of decomposition was reached. Different stages of thermal degradation of each polymer in the material were identified in the combustion process, which was significant only at the low heating rate. The results confirm that the reaction rate decreases with torrefaction, thereby increasing the reaction index. The char from the torrefied biomass has more thermal stability than the char from raw biomass, which can be confirmed by the increased reaction index.

A two-stage, inclined continuous rotary torrefier with novel flights has been developed in the Biomass Conversion Laboratory at Dalhousie University for improving biomass torrefaction processes. Experimental work on torrefaction of small poplar wood particles (0.5-1.0 mm) in the torrefier was undertaken fijr a deeper understanding of the working of such torrefiers where the volatile gas released was used as the torrefaction medium instead of nitrogen. The rotary torrefier is operated under different operating conditions by varying its rotational speed; tilt angle and temperature. Measured chemical and physical properties of the torrefied products included ultimate and proximate analysis, structural analysis, energy density, mass yield, energy yield, and bulk density. A novel probe was developed to Wiled samples of biomass and measure temperature at different interior points along the length of the rotary torrefaction reactor while the biomass was being progressively torrefied in it. Axial temperature distribution of the rotary torrefier showed a parabolic profile but the fixed carbon content, volatile, and energy density of biomass undergoing torrefaction varied linearly along the length of the torrefier. For torrefaction at 300 C and 5 rpm and 1 of tilt angle the change in heating value was 40%, while the mass yield and energy yield of torrefied biomass were 34% and 48%, respectively. Results showed that temperature is the most irhpottant parameter in this torrefaction process.

Torrefaction is defined as a thermal pre-treatment process performed within a temperature range of 200-300 degrees C, at low-heating rates (<20 degrees C/min) and for residence times between 15-60 min in inert environments. A phenomenological model of the torrefaction process of large biomass particles is developed in this work. Mass and energy balance coupled to a kinetic model take into account two steps of the biomass decomposition. First of the two steps, considers simultaneous production of vapor and solids from raw biomass. The vapor phase comprises a mixture of condensable and non-condensable gases, while the solid phase consists of torrefied biomass. The second step involves decomposition of volatiles into gases and secondary char. The model analyzes torrefaction behavior of both large and small biomass particles, predicting their final solid and gas yields, temperatures distribution, internal pressure and velocity of the gas phase within the particles. The model also predicts maximum conversions for given particle sizes and temperatures during the process. For given set of conditions small particles showed higher (similar to 77%) than that (similar to 52%) for large particles. Maximum interstitial gas velocities inside the large particle (25 mm in diameter and 65 mm in length) was about 1.2 mm/s and pressure gradients of about 2000 kPa and it occurred after 20 min in the process. (C) 2016 Elsevier B.V. All rights reserved.

Torrefaction, though defined as a low temperature (200–300°C) decomposition of biomass in an oxygen free atmosphere, it is hard to obtain such environment in a commercial unit unless one uses expensive means of nitrogen flushing or indirect heating. Oxygen leakage that adversely affects the product quality is unavoidable in commercial directly heated torrefier. Present work attempts to examine the optimum concentration of oxygen in the torrefier that can be tolerated without greatly compromising the product quality. In this work, torrefaction of relatively large pieces (25·4 and 19 mm diameter) of poplar wood was conducted at different oxygen concentration as well as in inert atmosphere while observing its effect on the temperature profile in the biomass interior, mass yield, energy yield and energy density. Results obtained are in agreement with those obtained in previous work on fine biomass particles that mass yield and energy yield decreases with oxygen presence in the torrefier. It, however, notes a slight increase in energy density. The work observes a sharp decline in mass yield beyond about 14% oxygen concentration suggesting that this may be the practical limit of oxygen in a torrefier. Finally this work notes that the presence of modest amount of oxygen could not only be tolerated but it may have some positive effect on the commercial design of a torrefier.

An effort is made here to develop a systematic design method for fluidised bed gasifiers. Three design approaches are considered: equilibrium approach, residence time approach and kinetic model approach. The equilibrium approach, which only predicts the gas composition, has two models: simple stoichiometric model and more comprehensive model. The latter gives better prediction of maximum achievable gas yield for a given set of feedstock and operating conditions. The residence time approach is primarily based on experimentally derived char gasification time and provides a good option for sizing of a gasifier, especially for carbonaceous fuels. The kinetic model, which is generally used for simulation of an existing gasifier, has been used here for the design considering the rate of reactions and effect of geometry and bed hydrodynamics on the gasifier performance. Because of its complexity, the kinetic model is most difficult to implement for initial design of a gasifier. More than three design approaches have been used to theoretically design a gasifier. Design results are compared with those predicted and obtained from experiments.

The paper presents experimental and theoretical investigations into heat transfer to cross- and vertical tubes in the standpipe of a circulating fluidised bed (CFB) boiler. Two separate test rigs (one to study the heat transfer on the wall and the other to cross-tubes) were used for this study of heat transfer to moving packed beds in the standpipe of a CFB. The height averaged heat transfer coefficient on the wall and the surface averaged heat transfer coefficient on the tubes are comparable at a given solid flux through the standpipe. Heat transfer coefficient on the wall reduced from the top to the bottom, and the rate of decrease reduced with height as generally observed on the furnace wall of a CFB riser. Analysis of heat transfer around the tube showed that heat transfer is the lowest at the bottom of the tube due to the presence of an area of stagnant particles resembling a wake. Owing to the increased residence time of particles on the surface of larger diameter tubes, an increase in the tube diameter caused a decrease in the heat transfer coefficient at all flowrates. The models proposed in this paper show good agreement with the experimental results from both cross- and vertical tubes.

Heat transfer coefficients were measured for 227 and 87 micron size particles of silica sand in a 102 mm dia. and 5.5 m tall circulating fluidized bed made of plexiglass. Effects of solid circulation rate, fluidization velocity, bed density and particle sizes were studied. The heat transfer coefficients were calculated from the temperatures measured along the axis of a 25 mm dia 100 mm long cylindrical heat flux meter. KEYWORDS: Heat transferFast bedCirculating Fluidized BedCirculation Rate

Everyday, around the world thousands of fluidised bed boilers are started up. Thus, start-up is an important aspect of a fluidised bed boiler operation. Although there are various start-up methods, oil firing is the most popular and commonly used technique for the boiler start-up. The spiralling price rise of oil necessitates a deeper understanding of this process and thereby development of a means for reduction of the start-up oil consumption. Research work on the start-up process was carried out in a subcompact circulating fluidised bed boiler built jointly by Dalhousie University and Greenfield Research Inc. at Kolaghat Thermal Power Station, W. Bengal, India. This 3.2 MWth unit is used for research and it also has the provision to supply auxiliary steam to the 210 MWe pulverised coal fired boiler. Waste product (Mill Rejects) from the coal pulverising mills of the utility boilers, is used as the design fuel in this unit. An oil burner placed above the furnace grid is used to heat up bed solids to a temperature at which the combustion of Mill Rejects can initiate. Temperature of the bed and other parts of the freeboard is measured during the heating up of the bed. Results showed that parameters such as initial particle size distribution, static depth of bed solids and fluidising air velocity influenced the heating profile of the bed. Based on this information, a method of reduction in oil consumption by premixing fuels with the bed solids was developed. This process accelerated the heating rate of the bed, thus reducing the oil consumption during the start-up. Studies were carried out to establish the optimum percentage of fuel mix in the bed. This paper presents preliminary results of a comprehensive study aimed at optimising consumption of fuel oil per start-up.

Cavity-type inertial separators developed by GRI (Patent no. 2, 159, 949, Canada, 2002) were tested in a semi-industrial size circulating fluidized bed pilot plant operated at room temperature. Three rows of separators were hung from the roof of the pilot plant where one row was kept inside the riser and the others were kept in the primary separation chamber, located between the back-pass and the riser. Parameters measured were axial pressure drops along the height of the riser, vertical solids flux on the separator walls, lateral outwards solids flux in the riser with and without separator and local temperatures on the separator walls. A net downwards solids flux is on the inner wall of the separators; however, no downwards solids flux is on the outer walls of the separators. Heat transfer coefficients on the outer wall are found higher than those on the inner walls of the separator. It is also found that the presence of inertial separators not only provides additional heat transfer surfaces but also indirectly increases the heat transfer coefficients on the riser wall. Copyright (c) 2005 John Wiley & Sons, Ltd.

The working of loop seal is one of the least understood aspects of a circulating fluidized bed boiler although this component is critical to the operation of such boilers. The present work is an experimental study of the flow of solids; through the recycle chamber of a loop seal in a circulating fluidized bed (CFB). The effect of different parameters on the velocity of solid overflow through the recycle chamber is investigated. The sharp-crested weir theory of free surface water flow was used to analyze the solid flow through the loop seal. The analysis gives a good indication of the validity of the above theory for solids overflow over a weir. The results show that the solid flow rate increases with increase in bed height above the weir and the particle size had no effect on the value of the weir coefficient.

The fragmentation behaviour of coal in fluidized beds is studied by using three coals having different swelling indices. Experiments were carried out in a 41-mm-diameter turbulent fluidized bed operated at 800 degrees C. Both inert and oxidizing environments were used in a series of tests carried out in this unit. Quantitative analyses of the fragments produced show that primary fragments are more numerous than secondary fragments. The swelling index of the coal exerted a major influence on the primary fragments. The secondary fragmentation was not influenced to a great extent by the combustion.

Gas–solid separator, a key component of a circulating fluidized bed boiler, controls the recirculation of solids around this type of boiler. As the technology matures, the drive is to have smaller size units handling a greater amount of solids. The impact separator is well suited to meet these demands with a low pressure drop from operating at low gas velocities. Complex hydrodynamics are encountered in an impact separator, where a staggered array of collection elements separates solids from an incoming mixture of gas and solid particles. The research was to examine the performance of this separator experimentally in a scale model, built to provide benchmark data. At an inlet gas velocity of 4.0 m s−1, an overall collection efficiency of 87.5% of particles with a mean particle diameter of 125 µm was realized. The results were compared with those from a similar study. Copyright © 2004 John Wiley & Sons, Ltd.

A semiempirical model has been developed for the prediction of the grade efficiencies of a gas–solid cyclone for circulating fluidized bed (CFB) boilers. The model is based on the equilibrium orbit theory. Experiments were conducted in a large cyclone to gather data on grade efficiencies for model verification. The model gives a good agreement between predicted and experimental values of grade efficiencies.

Using data from large CFB boilers, and taking average solids concentration, size and height of the boiler-furnace as variables, a correlation for fractional wall coverage has been developed. This correlation for wall coverage and several other refinements have been used to modify the cluster renewal model of heat transfer Predicted heat transfer coefficients from this model, for a wide range of CFB boilers, show good agreements with those measured in these boilers.

In the present work experiments are conducted to investigate the effect of operating parameters on heat transfer from bed to U-beam impact separators located in the top region of the riser column. The effect of suspension density and bed temperature on heat transfer from bed to the impact separators (test sections) are investigated. The experimental unit consists of a circulating fluidized bed riser column, which is 0.23 m×0.23 m in bed cross-section, 6.3 m in height with a return leg and back pass. The U-beam impact separators are located in the top region of the riser column. Furnace oil # 2 is burnt in the unit and the experimental investigations are conducted. Water is circulated through the U-beam impact separators. The presence of the impact separators in the top region of riser column helps in solids separation and also to absorb certain fraction of heat liberated in the furnace. The bed to U-beam impact separator heat transfer coefficient increases with suspension density due to increased particle concentration, which results in higher cluster and particle heat transfer. The heat transfer coefficient increases with bed temperature due to increased convection and radiation.

Circulating fluidized bed (CFB) boilers have gained wide scale acceptance in both the process and utility industries in sizes up to 300 MWe. Their ability to burn opportunity fuels such as petroleum coke has carved out a special niche for CFB boilers in the energy market. Presently more than 600 CFB boilers are either in operation or under construction worldwide. Boiler purchasers have a much wider choice of available designs and manufacturers to choose from, making bid selection more difficult. Even with performance guarantees in place, it is prudent for buyers to evaluate proposed designs in order to fully appreciate the various options and to identify potential problems. CFBCAD© is an intelligent software developed by extensive research into design methodologies for CFB boilers and critical analysis of the design of many CFB boilers manufactured by different companies around the world. The model used considers user-inputted fuel specifications and steam conditions, and performs sizing calculations for the furnace and heat transfer surfaces. A variety of heat transfer surface configurations are available for analysis. It has been used to evaluate the design of some operating plants and to try and predict deviations from design parameters.

The paper reports an experimental and theoretical investigation on heat transfer to the walls of the standpipe of a circulating fluidized bed boiler. The heat transfer was studied in a standpipe test section, 1940 mm long and 100 mm square cross section. Solids 3 used in this study were silica-alumina ceramic and sand with densities of 700 and 2564 kg/m(3), and sizes of 130 and 266 mum, respectively. The results show that there are two distinct zones: dense and dilute. For both types of particles, the heat transfer is higher in the dense section. The heat transfer coefficient increases with increase in the solids circulation rate. For a given circulation rate, a finer particle yields higher heat transfer coefficient. The paper also presents a mechanistic model of heat transfer in the standpipe. Heat transfer coefficients predicted by the proposed model agree well with the experimental data.

A techno-economic feasibility study has been conducted to investigate revamping four 58 MWth (15 MWe) pulverized-coal (PC) boilers with circulating fluidized bed (CFB) firing. The steam generators at Amarkantak Thermal Power Station in Chachai, Madhya Pradesh, are owned by Madhya Pradesh State Electricity Board (MPSEB), supplied by Simmering-Graz-Pauker AG of Vienna, Austria, and commissioned in mid-1965. The study reveals that: (i) CFB revamping of the boilers is technically feasible and economically sound; (ii) Performance improvement of the plant is significant in terms of such indices as plant load factor, forced outage and auxiliary oil consumption, among others. The expected performance improvement is due in large part to the elimination of key outage-prone components such as pulverizers (mills) and burners. (iii) There is significant improvement in emissions performance due to the reduction in emissions of NOx and fly ash; (iv) The financial analysis indicates that the CFB revamping option gives the highest return on investment compared to alternatives.

Experiments are conducted to investigate the effect of system pressure, Ca/S ratio and primary air velocity on sulphur capture in a pressurized circulating fluidized bed (PCFB) combustor. The pressure inside the PCFB combustor is varied from 200 to 700 kPa. The Ca/S ratio is varied from 1.6 to 3.0. The primary air velocity ranges from 3 to 7 in s(-1). The bed temperature is maintained at 750 degreesC. The sulphur capture increases with system pressure in the present range of experimental investigations. The sulphur capture also increases with Ca/S ratio up to a certain ratio and then shows a decreasing trend for the given operating conditions. A semi-empirical model is developed for explaining the sulphur capture mechanism in the pressurized circulating fluidized bed combustor under batch combustion conditions. The experimental data are validated with the model predictions and a reasonable agreement has been observed. Copyright (C) 2002 John Wiley Sons, Ltd.