Fuel Processing Technology 90 (2009) 1148–1156
Contents lists available at ScienceDirect
Fuel Processing Technology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / f u p r o c
Development of a modeling approach to predict ash formation during co-firing of coal and biomass V. Doshi a,⁎, H.B. Vuthaluru b, R. Korbee c, J.H.A. Kiel d a
School of Engineering, Monash University Sunway Campus, Jalan Lagoon Selatan, Bandar Sunway, Selangor, Malaysia Curtin University of Technology, Kent Street, Bentley 6104, Perth, Western Australia, Australia HRL Technology, Ipswich, Queensland, Australia d ECN Biomass, Coal and Environmental Research, P.O. Box 1, 1755 ZG Petten, The Netherlands b c
a r t i c l e
i n f o
Article history: Received 24 June 2008 Received in revised form 6 May 2009 Accepted 11 May 2009 Keywords: Biomass Speciation Co-firing Ash formation Ash deposition
a b s t r a c t The scope of this paper includes the development of a modelling approach to predict the ash release behaviour and chemical composition of inorganics during co-firing of coal and biomass. In the present work, an advanced analytical method was developed and introduced to determine the speciation of biomass using pH extraction analysis. Biomass samples considered for the study include wood chips, wood bark and straw. The speciation data was used as an input to the chemical speciation model to predict the behaviour and release of ash. It was found that the main gaseous species formed during the combustion of biomass are KCl, NaCl, K2SO4 and Na2SO4. Calculations of gas-to-particle formation were also carried out to determine the chemical composition of coal and biomass during cooling which takes place in the boiler. It was found that the heterogeneous condensation occurring on heat exchange surfaces of boilers is much more than homogeneous condensation. Preliminary studies of interaction between coal and biomass during ash formation process showed that Al, Si and S elements in coal may have a ‘buffering’ effect on biomass alkali metals, thus reducing the release of alkali–gases which act as precursors to ash deposition and corrosion during co-firing. The results obtained in this work are considered to be valuable and form the basis for accurately determining the ash deposition during co-firing. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Co-firing of biomass with coal is a short term option to reduce greenhouse-gas emissions from existing coal-fired boilers used in power generation. Combustion of biomass, a renewable and sustainable energy source with coal in coal-fired boilers, is a practical approach to partial replacement of fossil fuel in conventional power generating utilities. Furthermore, combusting biomass in pulverised fuel (pf) fired power plants is an economically viable option to maximise the use of existing coal-fired boilers. Biomass fuels are considered environmentally friendly as biomass consumes the same amount of CO2 from the atmosphere during growth as it is released during its combustion [1]. Large amounts of wood and other biomass residues remain unused so far and could possibly be made available for use as a source of energy [2]. Although co-firing of biomass and coal represents a cheaper and low risk sustainable energy option, however, several technical issues associated with co-firing, namely difficulty in fuel handling and storage, decrease in overall combustion efficiency, ash-related issues, pollutant emissions, and carbon burnout still require further attention. The main
⁎ Corresponding author. Tel.: +60 355146247; fax: +60 355146207. E-mail address:
[email protected] (V. Doshi). 0378-3820/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.fuproc.2009.05.019
focus in this work is on the fireside issues, as the focal problem during biomass combustion emanates mainly from the high alkali content of its ash which can significantly worsen ash fouling and slagging. This is expected to reduce the area available for heat transfer in the furnace and convective sections of a power utility [3–6]. The aim of this investigation is to develop an approach to improve the understanding of ash transformation and formation behaviour during the co-firing of both fuels. The scope of this paper includes the study of ash-related issues during co-firing of biomass with coal, with emphasis given on development of a conceptual model to predict ash formation during co-firing of biomass and coal. Numerous models and indices are commercially available to predict ash formation for coal combustion. Unfortunately, these traditional empirical indices used for coals are insufficient for predicting ash deposition in biomass fuels as the coal based indices do not take into account the heterogeneous nature of biomass. In addition, interaction occurs between coal and biomass when the fuels are fired together, leading to inaccurate predictions if available coal indices were used instead [3]. The Energy Research Centre of Netherlands (ECN) has carried out numerous investigations on the impact of co-firing of biomass with coal on fouling using a Lab Scale Combustion Simulator (LCS) to simulate the conditions prevailing in an actual pf boiler [7–9]. The studies indicated that the fouling problems caused during biomass combustion occurred
V. Doshi et al. / Fuel Processing Technology 90 (2009) 1148–1156
mainly through alkali species which form a sticky deposit layer. Therefore, the scope of the paper is centered on developing a predictive model capable of determining the ash formation that occurs during co-firing of biomass with coal. 2. Ash formation during coal combustion Besides organic matter, between 5 and 30 wt.% of pulverised coal consists of inorganic matter. The volatile organic matter is vaporised with the gases when the heating process begins, which is followed by the burning of char. The incombustible inorganic matter in coal consists of discrete mineral grains and organically bound inorganic matter that forms ash after combustion. There are more than 100 types of mineral grains that can be found in coal and some of the major ones are quartz, clay minerals like kaolinite, carbonates like calcites and sulfides such as pyrites. The minerals exist in several forms such as discrete mineral grains, flakes or different-shaped in combined form [3]. These minerals exist either within the coal matrix known as included minerals or outside the coal matrix and are called excluded minerals [10]. The organically associated inorganic matter is mainly alkali and alkaline matter bonded with the oxygen functional groups in coal. This material however is a very small fraction of inorganics in high rank coals and are often ignored in the study of ash prediction models [10–12]. Fig. 1 illustrates the association of the organic and inorganic materials observed in a coal particle. Ash formation in coal combustion occurs mainly from fragmentation and coalescence of the mineral matter in the coal. However, it is noted by researchers that a very small percentage (typically 1 wt.%) of ash in coal is formed from vaporisation of the inorganic matter present in the coal [13]. This reaction occurs mainly from the vaporisation of fine mineral particles, refractory oxides and the organically bound matter during devolatilisation and char burnout [14]. The released material then nucleates homogeneously to form aerosols or condense with other available fly ash particles. Ash formation of the included and excluded mineral particles in coal occurs in a different manner. As included minerals are close to each other in a char particle, these grains become molten and enable coalescence and agglomeration to occur during char burnout. Fragmentation occurs when the high heating rates in combustion chambers causes pressure build-up within minerals. The discharge of this build-up causes the minerals to fragment. Included minerals seldom experience fragmentation as was seen in minerals such as silicates and clay minerals. On the other hand, excluded minerals mainly form ash through the fragmentation of minerals. This results in several fragments of ash particles from one char particle, changing the particle size distribution. The fragmentation occurs based on the type of coal being used as the phenomenon depends on char particle size, structure and ash loading. In terms of particle size distribution of the overall ash, the larger ash particles (N2mm) are formed by mechanisms such as coalescence and shedding and the fine ash particles (b2mm) are formed from the vaporisation and condensation mechanisms [16]. Understanding of the mechanisms of ash formation of coal has evolved with the progress of analytical tools and methods. The knowl-
1149
edge gained has brought about the development of several types of models that predict ash formation and deposition of various types of coal during combustion. These models provide a comprehensive approach for predicting the behaviour of minerals in coal by including processes such as transformation and distributions of minerals in the coal [17]. The main input for coal ash formation models comprises of conventional analysis of the coal including, proximate analysis for moisture, density and ash content, elemental and oxide analysis. Using these data, the models are able to predict the distribution of the included and excluded minerals based on the size and density of the coal particles [13]. Studies on high rank coal have found that the fraction of ash formed through vaporisation is very small and therefore most models do not include this process [16]. Three different sub-models are used for the study of coalescence behaviour, namely, full coalescence, partial coalescence or no coalescence model. The first sub-model suggests that all the mineral particles in one char particle comes together to produce one ash particle whilst the third model advocates that each mineral grain in a char particle produces separate ash particles. However, based on the structure and morphology of char particles Li Yan concluded that, the partial coalescence model was found to be the closest model to the results obtained [15]. Monroe in his thesis states that the extent of coalescence can be measured by the shell thickness of the char cenosphere, the size of the coal and mineral particles and the mineral volume fraction [18]. In terms of fragmentation of minerals in coal, the Poisson distribution is used to determine the number of ash fragments envisaged during fragmentation of minerals [19]. Several ash formation models to determine the speciation of inorganic matter in coal exist and are available commercially. Therefore for the development of a coal combustion model, an advanced model called the Ash Effect Predictor has been adopted for the identification of coal speciation [15,16,18]. The focus of this paper, however, is directed towards establishing a reliable method to determine biomass speciation and its interaction with coal during combustion. The next section addresses the biomass combustion and speciation aspects during combustion. 3. Biomass combustion Biomass can be categorized into four main groups, which are woody biomass like wood bark and waste wood, herbaceous biomass like straw and grass, agricultural by-products such as animal manure and fourthly, refuse-derived fuels and waste products [20]. For biomass that comes under the plant category, its main structural composition mostly consists of cellulose, hemicellulose, lignin and the inorganic matter. The dry part of biomass is mainly formed from C, H and O, whilst a minor portion of it contains N, S and P, which are vital for the plants metabolism and physiology [4]. The combustion behaviour of biomass is found to be similar to combustion of low rank coals. As most available coal models are better suited for high rank coals, it is necessary to develop a model for biomass. 3.1. Biomass speciation In addition to organic matter, biomass also consists of inorganic matter that contributes to ash formation processes. Using various literature resources [2,21,22], the speciation of the inorganics in biomass has been compiled and classified into three main groups including:
Fig. 1. Organic and inorganic matters in a coal particle.
▪ Salts that are ionically bound. ▪ Inorganics that are bound organically to the carbonaceous material. ▪ Included minerals that are present in the fuel structure and adventitious minerals or foreign material from harvesting of biomass.
1150
V. Doshi et al. / Fuel Processing Technology 90 (2009) 1148–1156
Table 1 Speciation of biomass based on literature review summary [2,21,22,24,27]. Element
Ionic salts
Organically associated inorganics
Minerals
Na
Sodium nitrate Sodium chloride Potassium nitrate Potassium chloride Calcium nitrate
Calcium pectate
Calcium oxalate Calcite
K Ca
Mg
Si
Calcium chloride Calcium phosphate Magnesium nitrate Magnesium chloride Magnesium phosphate Amorphous silica
S N
Sulfur tetraoxide-2ion
P
Phosphate -3ion
Cl Al Mn
Chloride ion
Fe
Chlorophyll Magnesium pectate Phytolite Quartz Sulfolipids Amino acids, protein Sulfolipids Nucleic acids
Phytates Phytic acid Kaolinite
Organic structures of proteins and carbohydrates Chelates Organic sulfates
Phytoferritin Iron oxide
behaviour of elements most often gives an idea of its chemical speciation even before carrying out the chemical equilibrium calculations. For example for elements with constant leaching behaviour at all pH, it is most likely that the speciation will mostly consist of salts or free ions with a small amount of dissolved organics. For elements that have optimum leaching range at the acidic or pH regions such as Ca, Mg, Al and Si, it is likely that a larger speciation of either minerals or solid organics exists. The concentration of the highest peak is fed the chemical speciation model, known as LeachXS. In this research work, the database of LeachXS has been updated with thermodynamic properties of speciation that are relevant to the biomass samples being used. The equilibrium model can be used to predict the free ions, salts, organically bound matter and minerals that are seen in the inorganic matter within biomass. In this way, a large section of the speciation (inorganics) in biomass can be determined, as the extraction method has efficient leachability, enabling a good prediction for its speciation. It was seen that the results obtained from the biomass speciation using leaching and chemical equilibrium methods are comparable to the classification done in Table 1. Further details on the experimental procedure and the modelling work for biomass speciation can be found elsewhere [26]. 3.2. Ash formation in biomass
The main chemical species that are found in each of these groups is summarised in Table 1. An advanced analytical technique is required to predict the inorganic speciation in the various types of biomass. To date, the only known method of analysing the speciation of the alkali salts and the organically bound inorganics in biomass is done using the chemical fractionation (CF) method. Using CF technique, it is possible to have an idea of the manner in which the various ash forming matters are associated with the fuel by using increasingly strong leaching chemicals to extract the various constituents based on their solubility [4,23,24]. In thermochemistry studies, Baxter makes the assumption that the ash forming elements leached out with the less aggressive solutions will represent the volatile species in biomass during combustion and form the finer fraction of ash [22]. Those leached by the more aggressive solutions will correspond with the less volatile matter, constituting the inert or coarse fraction of the ash. However, Korbee et al. in their work suggest that further refinement of CF method is required for better and more detailed classification of the chemical composition and the bonding of biomass inorganics [7]. The present CF method classifies the inorganics in fuel rather broadly as the solubility is based on three different solutions, namely water, an acid and a base solution. Improvement has been carried out by the author to find a more effective technique to predict the speciation of biomass at a wide range of pH. It was found that a combination of pH extraction test and chemical equilibrium calculation could be used to determine the speciation of biomass. The pH extraction test provides the geochemical fingerprint of every element within biomass by observing its leaching pattern. The maximum quantity leached is used as an input into the chemical equilibrium model to determine the likely speciation of the biomass. Taking the steps mentioned above into further detail, it is firstly vital to establish the leachability curves of the main inorganic matter using pH extraction test. The recommended method to determine the leaching is the Standard European Leaching Method [25] whereby the leaching pH range is fixed between the range of pH 2 and 12. This wide range in pH ensures that all elements are maximally leached out, minimising residues and thereby allowing accurate predictions. The elements analysed are K, Na, Ca, Mg, Si, Al, Cl, S, N and P. The maximum leached value for every element is based on the highest peak of its individual leachability curve. For some elements the amount leached is not related to the pH of the leaching solution, thereby producing a constant graph. These elements are K, Na, Cl and S. The leaching
The ash formation likely to occur in different varieties of biomass fuels can be predicted from the composition of inorganic matter in the fuel. The manner of association and bonding of the inorganic matter with the rest of the fuel can be used to interpret the behaviour of biomass during combustion. Although biomass generally has lower ash content compared to coal, the composition of the ash in biomass is very different. Coal ash mainly comprises of alumino-silicates with clay and quartz, whilst biomass ash mainly consists of quartz and inorganic salts of phosphates, sulphates and chlorides [2]. Korbee et al., in their paper mention that biomass ash mainly contains elements from alkali metals (Na and K), and alkaline earth metals (Mg and Ca), in addition to smaller amounts of Si, P, S, Cl, Mn and Fe [27]. Another structure common in biomass is bio-minerals such as calcium oxalate or phytolith, which are rigid microstructure body that give structural stability to plants. Alkali and alkaline earth metals (K, Na, and Ca), which are present largely in biomass, with other elements like chlorine and sulfate, can cause unfavourable reactions that lead to problems like ash fouling and slagging. The high oxygen and organic volatile matter content of biomass make it more reactive than coal. This is because devolatilisation will occur at lower temperatures, increasing the potential for releasing larger amounts of vapours from inorganic constituents during combustion. Potassium (K) is a macronutrient for plant and a facilitator for osmotic processes. In coal, K is mostly in the form of clays, whilst in biomass it is mostly in the form of ionic salts. More than 90% by wt of potassium is found to be either water-soluble or ion exchangeable as observed during chemical fractionation tests [28]. K is volatile and combines with Si to form low melting silicates in fly ash particles. Na is a minor component (2% by wt) in biomass. It behaves similiarly to K, but at a very small fraction. Na is mainly attached as inorganic salt and is released as a gaseous specie that undergoes gas phase reactions with Cl and S to form chlorides and sulfates. A significant gas-phase reaction forms Na2SO4, which has a low melting point and may further contribute to fouling [29]. Excess Na will lead to the formation of NaOH and Na2CO3. Ca is a common constituent of cell wall and other organic component of cell structure. Through advanced analytical techniques, it is largely found as ion exchangeable and acid soluble material. Ca tends to react mostly with chlorides and sulfates, but is less volatile compared to K. Combusted Ca oxalates transform into oxides and carbonates. The anion chloride plays a key role in transformations of inorganic materials by playing a shuttle role for
V. Doshi et al. / Fuel Processing Technology 90 (2009) 1148–1156
1151
Fig. 2. Condensation of the vapour phase after biomass combustion.
inorganic materials and by serving as a charge balancer. It reacts with alkali material to form volatile and stable alkali chlorides. KCl is the most stable at high temperature (above 1000 °C), followed by KOH. At high temperature chlorides react with metals to form corrosion problems, whilst at lower temperature corrosion occurs through formation of acid gases. On the other hand, sulphur is present as free ions or bound organically with alkali and alkaline metals to form low melting compounds which lead to slagging and fouling at 800 to 900 °C. Most of the sulphur oxidizes during combustion, whilst the rest reacts with K and Ca to form sulfates, which condense on fly ash or deposit on walls. 3.3. Ash formation mechanism during biomass combustion The mechanisms that are involved in ash formation of biomass are vaporisation, condensation and coagulation/agglomeration. During biomass combustion, significant quantities of alkali metals and chlorides are vaporized to form gases such as HCl (g), KCl (g), K2SO4 (g), Na2SO4 (g) and NaCl (g) [30]. The composition of the chemical species released as gas phase can be determined using high temperature equilibrium calculations which are discussed in Section 3.3.1. When the gases are cooled down in the heat exchange section of boilers, they condense and form a large part of the fine ash fraction [31]. The condensation mechanism or also known as the gas-to-particle conversion occurs when compounds in the flue gas of a combustion unit become supersaturated, when the gas moves from the radiant zone towards the heat exchangers [32]. As the gas cools it moves towards equilibrium by condensing into liquid or solid form. According to Strand et al., there are two competing routes for the condensing vapour [33]. The vapour undergoes gas-to-particle conversion to form aerosols by either homogeneous nucleation or heterogeneous condensation on existing particles entrained in the flue gas as shown in Fig. 2. The droplets and aerosols begin to form larger particles through coagula-
tion and agglomeration and finally accumulate as ash particles of various particle size and chemical composition. 3.3.1. Gas phase composition High temperature thermochemical equilibrium calculations can be used to simulate the inorganic gas species and particulate matter that are released by biomass during combustion. This method is also known as Global Equilibrium Analysis and has been used to study the release of elements into the gas phase through volatilisation [34,35]. The composition of this gas phase is then used as an input for gas-toparticle calculations and to determine the condensed phases in the system after combustion. During combustion of biomass, the fuel is heated to a temperature between 1200 and 1500 °C. The vapour released by the combustion process in the radiant section, then goes through a convective section where cooling begins to take place. The equilibrium model calculates the gas composition and distribution of the main inorganic vapours expected at the set temperature range. FACTSAGE is a combination of two linked models in the area of computational thermochemistry, namely FACT-Win and ChemSage that consist of a series of information, database, calculation and manipulation modules that enable the users to access and operate pure substances and solution databases in relation to a variety of thermochemical calculations. This calculation is based upon the concept of Gibbs free energy minimization [36,37]. The model calculates the concentrations of chemical species when specified compounds react, to reach a state of chemical equilibrium [38]. The input supplied to the model is the elemental analysis of the fuel and the model then determines the most thermodynamically stable species that are likely to be present at a certain temperature. FACTSAGE then calculates the formation of the gaseous compounds, liquid phases (salt and liquid melt), solid solution phases and pure condensed phases. In this paper, the cases examined with FACTSAGE are wood chips, wood bark and a blend sample of 80% coal and 20%
Fig. 3. Elemental analysis of various fuels and blend.
1152
V. Doshi et al. / Fuel Processing Technology 90 (2009) 1148–1156
Table 2 Composition distribution of gaseous and solid phases derived from FACTSAGE at various temperature ranges. Temperature range
N 1000 °C
Wood chips
KOH KCl
65% 15%
NaOH NaCl KOH K2SO4 KCl NaOH Na2SO4 NaCl KCl NaCl NaOH
70% 15% 85% 10% 5% 90% 5% 5% 10% 35% 5%
Wood bark
Coal (80%) and biomass (20%)
1000–600 °C
b600 °C
KOH KCl K2SO4 NaOH NaCl KOH K2SO4 KCl NaOH NaCl
Solid phase
45% 10% 20% 40% 15% 50% 43% 7% 90% 10%
Solid phase
Fig. 5. Heterogeneous condensation on surfaces and foreign particles.
Solid phase
Solid phase
biomass. The blend test is expected to give greater insights into the interaction between coal and biomass during combustion in a pf boiler. The ultimate analysis of the samples is shown in Fig. 3. The inputs generally considered are C, H, O, S, Cl, S, P, Al, Ca, Fe, K, Mg, Mn, Na, Pb, Si, Ti, Zn and N, whilst the main elements relevant to ash formation are K, Na, Ca, Cl and S. The temperature range considered was from 500–1500 °C with a step of 50 °C. The FACTSAGE database contains more than 300 gaseous, liquid, and solid phases, including salt melt and silicate melt; however, this paper only looked at the pure gases and solid phase that are stable at the designated temperature range. The results predicted for the distribution of the gases and solids in the biomass samples analysed is summarised in Table 2. It can be seen that the gas distribution for K is mainly in the form of KOH, KCl and K2SO4. Some quantities of the KCl dimer were also observed. However, for temperatures below 800 °C, most of the constituents at equilibrium exist as of solids. Similar observations were seen for the Naelement whereby at temperatures higher than 800 °C, the expected gases are NaOH, NaCl and Na2SO4, which are in agreement with several researchers [39–41]. On the other hand, hardly any gases were seen to form with the Ca and Mg elements during combustion. The equilibrium data analysed from FACTSAGE illustrate that the solids obtained are mainly silicates and phosphates. Among the anions that were looked into are chlorine, sulphates, phosphates, nitrates and carbonates. For sulphates, mainly SO2 gas and K2SO4 solids were found to be stable in this temperature range, whereas for phosphates only solids were present at equilibrium. In the case of co-fired coal and biomass sample, it was observed that the increase of Si and Al from coal decreased the amount of volatile K, thus reducing KCl formations. This was also proven by other researchers using experimental and equilibrium studies on co-firing cases [29,30,41,42]. 3.3.2. Gas-to-particle condensation Homogeneous nucleation occurs when a gas nucleates to form a new phase, solid or liquid, without the aid of a surface [43]. It can be seen in Fig. 4 how during homogeneous nucleation, the gas molecules converge to form a nucleus which then forms a particle. This nucleus
can then become a site for further condensation by other gas molecules. Saturation ratio calculation is generally used to predict the possible occurrence of homogeneous nucleation for a certain body of condensing gas. Saturation ratio is described as the ratio of the partial pressure of the condensing gas to the vapour pressure of the condensing gas. Homogeneous nucleation is expected to occur in systems when the saturation ratio of a vapour species is much higher than unity. McNallan et al. summarise that the occurrence of homogeneous nucleation depends on the cooling rate of the combustion and the number and size of existing sites available for condensation [12]. If surface areas of particles are already present in the system, then it is more likely that heterogeneous condensation will occur. Heterogeneous condensation occurs when the vapour substances condense onto a surface of the same species or onto a surface of a foreign species as shown in Fig. 5. Heterogeneous condensation occurs when the body of gas is supersaturated, that is, when the saturation ratio is just above one. This means it is more thermodynamically possible for heterogeneous condensation to occur rather than homogeneous nucleation, as the free energy required for the former is much lower than the amount required for the latter [44]. In order for heterogeneous condensation to occur primary seed particles are required as a site for condensation. In biomass combustion these seed particles are commonly low volatility compounds such as oxides of Ca, Mg and Si or non-volatilized minerals that have experienced break-up and fragmentation [40]. Fly ash produced from homogeneous nucleation is of sub-micron size, whilst inorganics that are produced through heterogeneous condensation are usually larger in size. However, fine ash particles produced from homogeneous nucleation can further collide and coagulate to form clusters of larger ash particles [45]. Coagulation occurs when spherical aerosol particles collide and stick to form another, larger spherical particle. The mass during coagulation remains constant but a change is observed in the particle size distribution. Agglomeration is observed when colliding particles form an irregularly shaped particle [46]. Particle collision and coagulation lead to a reduction in the total number of particles and an increase in the average size while also reducing the surface area for further condensation. To enable the gas-to-particle calculations in this system, the gas phases involved are predetermined from the high temperature thermochemistry method. For the three biomass studied, namely wood chips, wood bark and straw, the main chemical species in the gas phase are KCl, K2SO4, NaCl and Na2SO4. Fig. 6 summarises the procedure to determine the possibility of homogeneous or heterogeneous condensation occurrence during biomass combustion. If the saturation ratio of the gas is much larger than one, the possibility of homogeneous condensation arises, which then suggests that the calculation of the rate of homogeneous nucleation should be included. However, if the saturation ratio is only slightly larger than one, the step proceeds
Fig. 4. Formation of particle by homogeneous nucleation [43].
V. Doshi et al. / Fuel Processing Technology 90 (2009) 1148–1156
1153
Fig. 7. Comparison of heterogeneous and homogeneous condensations for (a) K2SO4 and (b) KCl. Fig. 6. Homogeneous and heterogeneous condensation occurrences during fuel combustion.
towards the calculation of heterogeneous condensation of the gases on the surface of existing ash particles in the system. The saturation ratio values obtained for the gas phases are summarised in Table 3. The rate of homogeneous nucleation, Jhom measures the number of nuclei formed per unit volume in a heat exchange system. The calculations are based upon the Classical Theory of nucleation whereby the basic assumption is taken from the Capillary Approximation [47]. The classical theory assumes that a cluster can be represented as a droplet and its properties such as density and surface tension are the same as those of bulk liquid. It therefore neglects the inhomogeniety of the density and the curvature correction of surface tension [48]. The homogeneous nucleation rate expression is calculated as follows [44]: Jhom =
2σ πM
1 = 2
2
2
3
mN −16π v1 σ exp 3 ðkT Þ3 ðln SÞ2 S
! ð1Þ
where: v1 M N
volume of one molecule of the condensing gas mass of a molecule of condensing gas number concentration of condensing gas in the system
Table 3 Homogeneous and heterogeneous temperature occurrences for the main gas phases in biomass. Species
Temperature when S N 1 was achieved
Temperature when SNNN1 was achieved
KCl K2SO4 NaCl Na2SO4
700 °C 1100 °C 700 °C 1100 °C
500 900 700 900
°C °C °C °C
σp S
surface tension of particle saturation ratio of condensing gas
For the heterogeneous condensation rate of gas phase, Christensen et al. suggest a calculation expressing the molecular rate of condensation, Jhet, of a gas on a particle of diameter dp [39]. Jhet = 2πDi dp;j Nj
pi − pi;S CFS Rgas T
ð2Þ
where: Di pi pi,S dp,j Nj CFS
diffusion coefficient of condensing gas in flue gas partial pressure of the condensing gas in the system vapour pressure of the condensing gas diameter of each particle available for condensation; molecular weight of gas species Fuchs–Sutugin correction factor
It has to be noted here that for the case of direct condensation on tube surfaces rather than condensation on particles, the vapour pressure, pd has to be calculated at the temperature of the tube and not the gas temperature as done in Eq. (2) [43,49]. It can be seen in Fig. 7 that for both homogeneous and heterogeneous condensations, the alkali sulphates start to condense before the alkali chlorides. The condensation process for alkali chlorides starts at 700 °C, whilst for the alkali sulfates the condensation process starts earlier that is at 1200 °C. This also means that the alkali sulphates may become available particles for the alkali chlorides to condense on heterogeneously. Similar observations are reported in literature by several researchers during their experimental and modelling work [50–52]. The pathway that was discussed for ash formation of biomass in Section 3 has been summarised and displayed in Fig. 8.
1154
V. Doshi et al. / Fuel Processing Technology 90 (2009) 1148–1156
Fig. 8. Schematic flow diagram of ash formation mechanisms and approach taken in modeling the ash formation pathway.
Fig. 9. Schematic of a conceptual model showing the co-firing advisory tool functions.
V. Doshi et al. / Fuel Processing Technology 90 (2009) 1148–1156
4. Development of a conceptual co-firing model 4.1. Coal and biomass interaction during combustion It has been observed through experiments carried out by Biagini et al. [53] that coal and biomass do not interact during the devolatilisation stage of combustion. They found that there was no interaction between both the fuels even though their volatile matter, devolatilisation rate and thermal reactivity were different [53]. However, other researchers disagree with this idea suggesting some amount of interaction occurring between inorganics in coal with that of biomass during the condensation mechanism. Equilibrium calculations have shown that mineral matter in coal such as kaolinite and clays ‘capture’ some of the gas phase alkali metals to condense with the minerals [42,54]. In a pilot investigation carried out to study the interaction between several fuels (three types of coal and five types of biomass), it was observed that sulphur from the coals did interact with K and Na from biomass [5]. This shows that interaction does occur between volatile matter in biomass with Al, Si and S elements in coal. It has been previously observed that KCl is the cause for corrosion problems during ash deposition whilst the low melting temperature of K-silicates leads to severe ash deposition [30]. On the other hand, K-aluminosilicates have a higher melting point which leads to the formation of a less troublesome deposition. From the high temperature thermochemistry calculations reported in this paper, it can be seen that the gas phases released are mainly KCl, NaCl, K2SO4 and Na2SO4 during pure biomass combustion. These are the gases that generally undergo condensation and are responsible for formation of submicron ash. Only a small amount of the remaining material was stable in the form of alkali–silicates and aluminosilicate. However, when a similar analysis was carried out for coal (80% mass fraction) and biomass (20% mass fraction) blend, it was seen that the stable phases were mainly the alkali–aluminosilicates (N80%). This indicates that coal has a ‘buffering’ effect on biomass alkali metals and the chemical composition of coal may help to reduce the release of alkali–gases that can cause deposition and corrosion issues. 4.2. Conceptual model layout A schematic of the ash formation model during co-firing and the interaction between the two fuels is described in Fig. 9. The conceptual model developed for ash formation during co-firing of coal and biomass has been divided into several sections. Ash formation depends upon the possibility and extent of fragmentation or coalescence of the coal particle. The same holds true for the behaviour of the mineral matter in a biomass during combustion. However, the ash formation behaviour for the vaporised and organically bound inorganics within biomass is completely different. As advanced models are already available for coal combustion (Section 2), the work in this model focuses on the vaporised and organically bound inorganics within biomass. Firstly, for biomass, it is vital to have detailed knowledge of the chemical speciation of the inorganics in the fuel as addressed in Section 3.1. Secondly, high temperature chemical equilibrium calculations are used to predict the gas phase of biomass, both in the radiant and convective sections of a boiler. The results of the biomass speciation are taken as the input for this high temperature modelling as discussed in Section 3.3.1. The model provides further knowledge about the composition and phases of ash released during combustion and the interaction between ash-forming elements of biomass with coal. Finally, the condensation or gas-to-particle formation of the gas phase has been defined in Section 3.3.2. Although the scope of this work does not include the actual particle size distribution calculations of the condensed ash, it was observed that homogeneous nucleation leads to the production of new particles from the gas phase, thus increasing the number and mass of the fine particulate phase formation. Heterogeneous condensation, on the other hand, does not change the
1155
Table 4 Comparison between data obtained from ash formation model and lab-scale combustion (LCS) experimental data. Biomass
Elements
Ash formation model results
Experimental data from lab-scale combustion unit
Wood chip and wood bark
K, Na
Largely consists of free and dissolved organics Little mineral matters 50% consists of free and dissolved organics 50% mineral matters Largely consists of mineral matter Largely consists of free and dissolved organics Contains no mineral matters
Largely released from fuel during combustion
Ca, Mg
Si, Al Cl, S
50% of element released f rom fuel during combustion No or little release of element from fuel during combustion All released from fuel during combustion
number of particles already present in the system, but it does increase the particle size by the transfer of condensed gas phase onto the surface area of present particles. Although the results of the work done on validation of this conceptual model, using a lab scale combustion unit is published elsewhere [26], a brief summary of the comparison of the results is reported in Table 4 for wood based biomass. It is evident from the summary that using the ash formation model it was possible to predict the quantity and quality of elements within biomass that are most likely to be released as ash and thereby condense as fine ash. Additionally the model can also predict the elements that are most likely to form minerals and thereby settle as coarse ash. 5. Conclusion The motivation to develop a co-firing advisory tool is to deepen the understanding of ash formation during the co-firing of biomass with coal. It is known that the ash forming matter in coal and biomass is responsible for several ash-related issues. The tool that is described in this paper is a conceptual model, which has four main functions: 1) Introduction of a novel method to determine the speciation of biomass — A pH based extraction method was adopted together with a chemical equilibrium model that determines the speciation of biomass inorganic matter. 2) Classifications of inorganics within the coal and biomass that are of concern during ash formation — Inorganics in fuel were classified into three main groups that represent the bonding of the inorganics to the carbon matrix. The first two groups are the ionically and organically bound inorganics, which are mainly found in biomass and the third group is the minerals found in both coal and biomass. 3) Interaction of inorganics in biomass with those of coal present during combustion — High temperature thermochemical analysis showed that coal may have a ‘buffering’ effect on biomass alkali metals, thus reducing the release of alkali–gases that can cause deposition and corrosion issues during co-firing. 4) Condensation of inorganics during the cooling process in heat exchange areas — High temperature gas released during the combustion of inorganics mainly condenses through the heterogeneous condensation mechanism as alkali sulfates and chlorides. Combined with the existing methods for coal characterisation and the work presented in this paper, it was possible to begin the pathway towards producing an efficient ash formation model for co-firing of coal and biomass. Additional work can be done to determine the particle size distribution of the condensed ash through agglomeration and coagulation processes. This work can also be further extended and combined with ash deposition work to get a complete overview on the fate of the inorganics during co-firing of coal and biomass. The
1156
V. Doshi et al. / Fuel Processing Technology 90 (2009) 1148–1156
resulting comprehensive model can be used to provide technical advice to organisations that plan to utilise biomass feedstock in existing coal fired power station boilers. Acknowledgements The authors are grateful to Paul de Wild, Rob Comans and Andre van Zomeren of the Biomass Team of ECN, for their valuable support in the experimental and modelling work. The authors also wish to thank the reviewers for their constructive and professional comments on the earlier version of the manuscript, which led to a great improvement of this paper. References [1] M. Sami, K. Annamalai, M. Wooldrigde, Co-firing of coal and biomass fuel blends, Progress in Energy and Combustion Science 27 (2001) 171–214. [2] S. van Loo, Handbook of Biomass Combustion and Co-Firing, Twente University Press, Enschede, 2002. [3] R. Bryers, Fireside slagging, fouling and high temperature corrosion of heat transfer surface due to impurities in steam-raising fuels, Progress in Energy Science Combustion 22 (120) (1996) 29. [4] B. Jenkins, L. Baxter, T. Miles, T.J. Miles, Combustion properties of biomass, Fuel Processing Technology 54 (1998) 17–46. [5] A. Robinson, H. Junker, L. Baxter, Pilot-scale investigation of the influence of coalbiomass co firing on ash deposition, Energy and Fuels 16 (2002) 343–355. [6] M. Pronobis, Evaluation of the influence of biomass co-combustion on boiler furnace slagging by means of fusibility correlations, Biomass and Bioenergy 28 (2005) 375–383. [7] R. Korbee, S. Eenkhoorn, P. Heere, J. Kiel, Prediction of ash and deposit formation for biomass co-combustion, Report, Energy Research Centre, Netherlands, 2002. [8] B.O. Skrifvars, R. Backman, Thermodynamic stability calculations in predicting corrosion behaviour at elevated temperature, Materials Science Forum 369–372 (2001) 923–930. [9] F. Frandsen, S. Lith van, R. Korbee, P. Yrjas, R. Backman, I. Obenberger, T. Brunner, M. Joller, Quantification of the release of inorganic elements from biofuel, Fuel Processing Technology 88 (11–12) (2006) 1118–1128. [10] S. Benson, E. Steadman, C. Zygarlicke, T. Erickson, Applications of advanced technology to ash-related problems in boilers, Plenum Publishing Corporation, Portland, 1996. [11] A. Sarofim, A. Padia, J. Howard, The physical transformation of the mineral matter in pulverized coal under simulated combustion conditions, Combustion Science Technology 16 (1977) 187–204. [12] M. Mc Nallan, G. Yurek, J. Elliott, The formation of inorganic particulates by homogeneous nucleation in gases produced by the combustion of coal, Combustion and Flame 42 (1980) 45–60. [13] L. Yan, R. Gupta, T. Wall, The implication of mineral coalescence behaviour on ash formation and ash deposition during pulverised coal combustion, Fuel 80 (2001) 1333–1340. [14] B. Buhre, J. Hinkley, R. Gupta, T.F. Wall, P. Nelson, Submicron ash formation from vaporisation of minerals in coal, 12th International Conference on Coal Science, Cairns, Australia, 2003. [15] L. Yan, R. Gupta, T. Wall, A mathematical model of ash formation during pulverised coal combustion, Fuel 81 (2002) 337–344. [16] H. Wu, T. Wall, G. Liu, G. Bryant, Ash liberation from included minerals during combustion of pulverised coal: the relationship with char structure and burnout, Energy and Fuels 13 (1999) 1197–1202. [17] R. Gupta, Coal research in Newcastle—past, present and future, Fuel 84 (10) (2005) 1176–1188. [18] L. Monroe, (1989). An experimental study of residual fly ash formation in combustion of a bituminous coal. Doctor of Philosophy. Chemical Engineering, Cambridge, Massachusetts Institute of Technology. [19] T. Yamashita, T. Teramae, H. Tominaga, Fly ash formation behaviour in pulverised coal combustion, 19th Annual Pittsburgh Coal Science Conference, Pittsburgh, 1998. [20] A. Williams, M. Pourkashanian, J. Jones, Combustion of pulverised coal and biomass, Progress in Energy and Combustion Science 27 (2001) 587–610. [21] H. Marschener, Mineral Nutrition of Higher Plants, 2nd EditionAcademic Press,1995. [22] L. Baxter, T. Miles, B. Jenkins, D. Dayton, T. Milne, R. Bryers and L. Oden, (1996). Alkali deposits found in biomass power plants. Sandia National Laboratories, US Department of Energy. [23] S. Benson, P. Holm, Comparison of inorganic constituents in three low-rank coals, Industrial and Engineering Chemistry Product Research and Development 24 (1985) 145–149. [24] M. Zevenhoven, B. Skrifvars, P. Yrjas, M. Hupa, L. Nuutinen, R. Laitinen, Searching for improved characterization of ash forming matter in biomass, 16th International Conference on Fluidized Bed Combustion, Nevada, 2001.
[25] CEN/TC 292, European Committee of Standardisation, Characterisation of waste leaching behaviour tests—influence of pH on leaching with continuous pH-control TS14997, 2005, [WWW] http://www.cen.eu/cenorm/index.htm (20 October 2005). [26] V. Doshi, (2007). Investigation into ash related issues during co-combustion of coal and biomass: development of a co-firing advisory tool. Doctor of Philosophy. Curtin University of Technology. Australia. [WWW] http://adt.curtin.edu.au. [27] R. Korbee, J.H.A. Kiel, M. Zevenhoven, B. Skrifvars, P. Jensen, F. Frandsen, Investigation of biomass inorganic matter by advanced fuel analysis and conversion experiments, Report BioAerosol Project, Energy Research Centre of Netherlands, 2003. [28] L. Baxter, T. Miles, B. Jenkins, D. Dayton, T. Milne, R. Bryers, L. Oden, The behaviour of inorganic material in biomass-fired power boilers, Fuel Processing Technology 54 (1998) 47–78. [29] J. Knudsen, P. Jensen, L. Weigang, F. Frandsen, K. Dam-Johanes, Sulfur transformations during thermal conversion of herbaceous biomass, Energy & Fuels 18 (2004) 810–819. [30] Y. Zheng, P. Jensen, A. Jensen, B. Sander, H. Junker, Ash transformation during cofiring coal and straw, Fuel 86 (2007) 1008–1020. [31] M. Zevenhoven, J. Blomquist, B. Skrifvars, R. Backman, M. Hupa, The ash chemistry in fluidised bed gasification of biomass fuels, Fuel 79 (2000) 1353–1361. [32] S. Freidlander, Smoke, Dust and Haze: Fundamentals of Aerosol Dynamics, Second EditionOxford University Press, New York, 2000. [33] M. Strand, M. Bohgard, E. Sweitlicki, A. Gharibi, M. Sanati, Laboratory and field test of a sampling method for characterization of combustion aerosols at high temperatures, Journal of Aerosol Science and Technology 38 (2004) 757–765. [34] M. Blander, T.A. Milne, D.C. Dayton, R. Backman, D. Blake, V. Kuhnel, W. Linak, A. Nordin, A. Ljung, Equilibrium chemistry of biomass combustion: a round-robin set of calculations using available computer programs and databases, Energy and Fuels 15 (2) (2001) 344–349. [35] B.O. Skrifvars, R. Backman, Thermodynamic stability calculations in predicting corrosion behaviour at elevated temperature, Materials Science Forum 369–372 (2001) 923–930. [36] C.W. Bale, P. Chartrand, S.A. Degterov, G. Eriksson, K. Hack, R. Ben Mahfoud, J. Melançon, J.A. Pelton, S. Petersen, FactSage thermochemical software and databases, Calphad 26 (2002) 189–228. [37] E. Jak and D. Saulov, Prediction of ash phase equilibria using FACT models. Research Report 54. CRC for Coal in Sustainable Development, Australia. [38] E. Jak, P. Hayes, C.W. Bale, S.A. Decterov, Application of FactSage Thermodynamic Modeling of Recycled Slags (Al2O3-CaO-FeO-Fe2O3-SiO2-PbO-ZnO) in the Treatment of Wastes From End-of-Life-Vehicles, International Journal of Materials Research 98 (2007) 872–878. [39] K. Christensen, M. Stenholm, H. Livbjerg, The formation of submicron aerosol particles, HCl and SO2 in straw-fired boilers, Journal of Aerosol Science 29 (1998) 421–444. [40] M. Joller, T. Brunner, I. Obernberger, Modelling of aerosol formation during biomass combustion in grate furnaces and comparison with measurements, Energy & Fuels 19 (2004) 311–323. [41] S. Jimenez, J. Ballester, Formation of alkali sulphate aerosols in biomass combustion, Fuel 86 (2007) 486–493. [42] X. Wei, U. Schnellb, K. Hein, Behaviour of gaseous chlorine and alkali metals during biomass thermal utilisation, Fuel 84 (2005) 841–848. [43] M. Jacobson, Fundamentals of atmospheric modelling, Cambridge Press, United Kingdom, 1999. [44] J. Seinfeld, S. Pandis, Atmospheric Chemistry and Physics, Wiley Interscience, Canada, 1998. [45] C. Zygarlicke, D. McCollor, K. Eylands, M. Hetland, M. Musich, C. Crocker, J. Dahl, S. Laducer, Impacts of co firing biomass with fossil fuels, U.S. Department of Energy, National Energy Technology Laboratory, Pittsburgh, 2001. [46] G. Couch, Understanding slagging and fouling during PF combustion, IEA Coal Research, London, 1994. [47] P. Peeters, (2002). Nucleation and condensation in gas-vapor mixtures of alkanes and water. Doctor of Philosophy. Eindhoven University of Technology, The Netherlands. [48] M. Iwamatsu, Homogeneous nucleation of spherical droplets and bubbles, Chinese Journal of Physics 33 (2) (1994). [49] O. Knacke, O. Kubaschewski, K. Hesselmann, Thermochemical properties of inorganic substances, 2nd EditionSpringer-Verlag, Berlin, 1991. [50] I. Obernberger, Ash related problems in biomass combustion plants, 2005 [WWW] http://alexandria.tue.nl/extra2/redes/obernberger2005.pdf (August 20, 2005). [51] P. Glarborg, P. Marshall, Mechanism and modelling of the formation of gaseous alkali sulfates, Combustion and Flame 141 (1) (2005) 22–39. [52] S. Jimenez, J. Ballester, Influence of operating conditions and the role of sulfur in the formation of aerosols from biomass combustion, Combustion and Flame 140 (4) (2005) 346–358. [53] E. Biagini, F. Lippi, L. Petarca, L. Tognotti, Devolatilisation rate of biomasses and coal-biomass blends: an experimental investigation, Fuel 81 (2002) 1041–1050. [54] D. Dayton, D. Belle-Oudrey, A. Nordin, Effect of coal minerals on chlorine and alkali metals released during biomass/coal cofiring, Energy & Fuels 13 (1999) 1203–1211.