Ash Agglomeration and Deposition during Combustion of Poultry Litter

Jun 27, 2013 - ABSTRACT: In this study, we have characterized the ash resulting from fluidized bed combustion of poultry litter as being dominated by ...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/EF

Ash Agglomeration and Deposition during Combustion of Poultry Litter in a Bubbling Fluidized-Bed Combustor Deirdre Lynch,†,‡ Anne Marie Henihan,† Witold Kwapinski,† Lian Zhang,‡ and James J. Leahy*,† †

Department of Chemical and Environmental Science, University of Limerick, Limerick, Ireland Department of Chemical Engineering, Monash University, Clayton, VIC 3800, Australia



S Supporting Information *

ABSTRACT: In this study, we have characterized the ash resulting from fluidized bed combustion of poultry litter as being dominated by a coarse fraction of crystalline ash composed of alkali-Ca-phosphates and a fine fraction of particulate K2SO4 and KCl. Bed agglomeration was found to be coating-induced with two distinct layers present. The inner layer (0.05−0.09 mm thick) was formed due to the reaction of gaseous potassium with the sand (SiO2) surface forming K-silicates with low melting points. Further chemical reaction on the surface of the bed material strengthened the coating forming a molten glassy phase. The outer layer was composed of loosely bound, fine particulate ash originating from the char. Thermodynamic equilibrium calculations showed slag formation in the combustion zone is highly temperature-dependent, with slag formation predicted to increase from 1.8 kg at 600 °C to 7.35 kg at 1000 °C per hour of operation (5.21 kg of ash). Of this slag phase, SiO2 and K2O were the dominant phases, accounting for almost 95%, highlighting the role of K-silicates in initiating bed agglomeration. The remaining 5% was predicted to consist mainly of Al2O3, K2SO4, and Na2O. Deposition downstream in the low-temperature regions was found to occur mostly through the vaporization−condensation mechanism, with equilibrium decreasing significantly with decreasing temperatures. The dominant alkali chloride-containing gas predicted to form in the combustion zone was KCl, which corresponds with the high KCl content in the fine baghouse ash.



INTRODUCTION Background. The European Union has introduced legislation1 encouraging the generation of energy from biomass fuels, because of their carbon-neutral status, in an effort to mitigate climate change. Biomass is defined as the biodegradable fraction of products, waste, and residues from biological origin (including agriculture, forestry, and related industries). Biomass can include woody fuels, herbaceous crops or grasses, dedicated energy crops, and manures.2 Poultry litter is one such manure; it consists of a mixture of bedding material, excreta, waste feed, and feathers. Previous work by the authors has characterized poultry litter as a biomass resource and identified fluidized bed combustion (FBC) as a suitable conversion technology.3 Fluidized-bed technology is one of the most promising methods for biomass combustion, because of its ability to handle fuels of varying composition with good combustion efficiency and low emissions.4−6 FBC has been selected in several combustion trials utilizing poultry litter,7−9 and it is used commercially for the combustion of low-quality fossil fuels, such as peat,10 and the co-combustion of biomass fuels.11,12 However, the variable mineral content of biomass can give rise to operational problems including slagging, fouling, and possible corrosion during combustion applications.13,14 Consequently, much of the current research surrounding biomass fuels is focused on ash formation and the mechanisms of ash deposition and agglomeration.11,15−17 Ash Formation. The primary mechanisms for ash formation during combustion of a biomass fuel involve release and vaporization of the alkali metals and chlorides from the fuel and char particles, interaction of inorganic elements leading to salt formation in liquid and gas phases, condensation (homogeneous or heterogeneous) as the gas phases achieve © 2013 American Chemical Society

equilibrium in the cooler sections of the unit and coagulation or agglomeration of the condensed phases.4,15,18−22 Bed agglomeration is a potential issue during FBC of biomass, because it can result in increased maintenance and operational costs. In severe cases, it can cause total defluidization of the bed.23 Agglomeration of bed material may occur due to the formation of low-temperature melting compounds and eutectics. Where sand is used as a bed material, the reactions between quartz particles and alkali metals in the fuel have been identified as key factors in the formation of sticky alkali-silicate coating layers that lead to agglomeration.24 An extensive review on the mechanisms involved in bed agglomeration by Brus et al.25 and Ö hman et al.,26 identified coating-induced agglomeration as one of the most important routes for agglomeration during combustion of certain biomass fuels. Deposit formation in the heat exchangers and ancillary equipment can lead to reduced heat-transfer capacity and corrosion or erosion issues.4 The amount of chlorine (Cl) in fuels can directly influence the amount of alkali vaporized during combustion, particularly potassium (K), with higher Cl content leading to higher amounts of vaporized alkali species.27 Potassium chloride (KCl) is among the most stable alkali gases present at high temperatures. In the absence of Cl, hydroxide gases (e.g., KOH) are the next most common.27 When these gases cool, they nucleate and condense on the metal surfaces, or on the surface of ash already present.28,29 As the temperature decreases, carbonates are also likely to form because of the reaction of alkali with Received: July 16, 2012 Revised: June 19, 2013 Published: June 27, 2013 4684

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694

Energy & Fuels

Article

Table 1. Ash Composition of the Elements Expressed as Their Most Common Oxides Ash Content (wt %) fuel

ref

SiO2

Al2O3

Fe2O3

CaO

Mgo

P2O5

K2O

SO3

Na2O

TiO2

MnO

rest*

59 60 60 61 62

2.80 53.3 68.18 49.30 0.85 36.9

0.57 2.43 7.04 9.40 0.12 19.8

1.20 1.22 5.45 8.30 0.33 13.6

21.00 12.7 7.89 17.20 15.02 9.8

11.0 10.8 2.43 1.10 11.38 2.1

19.00 6.5 1.56 1.90 40.12 3.48

28.00 5.2 4.51 9.60 22.49 1.1

8.1 n/a 1.19 2.60 n/a 11.9

3.9 1.26 1.20 0.50 9.58 0.1

0.00 n/a 0.55 0.10 0.01 0.3

0.68 n/a 0.27b n/a 0.09 0.1

3.75 6.59 0c 0c 0c 0.82

a

poultry litter miscanthus pine chips spruce wood rapeseed cake peat

Laboratory generated ash at 550 °C. bGiven as Mn: 2090 ppm (0.21% as Mn or 0.27% as MnO). cNormalized to 100%. *Rest is loss on ignition (LOI) for poultry litter, and omitted results in the case of miscanthus and peat, which could include LOI. a

CO2.23 These compounds can then react with gaseous sulfur (SO2/SO3) and form sulfates on the metal surface.4 These alkali chlorides or sulfates are often sticky and can lead to deposit buildup due to inertial impaction of the ash-laden gas.30 Although there is an abundance of information in the literature on the formation mechanisms of ash, agglomeration, and deposition for woody biomass, few studies exist on the mechanisms involved in ash formation during combustion of poultry litter. Poultry Litter Ash. The composition of poultry litter ash and other fuels taken from literature are given in Table 1. We can see that poultry litter differs from other biomass in that it contains low silicon (Si) with significantly high levels of potassium (K), calcium (Ca), phosphorus (P), and magnesium (Mg) (K > Ca > P > Mg). These are added to poultry rations to improve feed conversion;31−34 however, significant quantities remain undigested and are thus excreted. For example, it is estimated that ∼70% of dietary P is excreted by the birds, resulting in a P-rich litter.35,36 The role of P in ash formation, agglomeration, and deposition is not as well-defined as that of K. Some K-rich phosphates are known to have low melting points which, as molten phases, can enhance the sintering of residual ash at combustion temperatures.37 Piotrowska et al.38 found that (i) high-P fuels aggravated bed agglomeration, and (ii) a high K/Ca ratio in P-rich fuels is associated with low-temperature ash sintering and melting. Objectives. There is considerable interest in using poultry litter as a fuel; however, if this is to be achieved, then improving our understanding of the role of K, together with P, Ca, and Mg, with respect to their role in ash formation during combustion, is necessary. Therefore, the objectives of the present study were (i) to characterize poultry litter ash and agglomerates formed during FBC, and (ii) to establish the distribution pattern of the major elements, and investigate agglomeration and deposition mechanisms using laboratory techniques in conjunction with equilibrium analysis.



Table 2. Fuel Analysis, on an As-Received Basis proximate and ultimate volatiles fixed carbona high heating value, HHV moisture ash carbon hydrogen nitrogen sulfur oxygena chlorine a

% % MJ kg −1 % % % % % % % %

poultry litter

wood shavings

39.75 ± 0.03 8.01 ± 1.51 10.18 ± 0.04 43.67 ± 1.52 8.57 ± 0.05 24.93 ± 0.11 3.69 ± 0.05 2.57 ± 0.06 0.30 ± 0.02 15.86 ± 0.12 0.41 ± 0.03

44.13 ± 0.4 5.95 ± 1.19 10.27 ± 0.06 49.69 ± 1.2 0.23 ± 0.02 26.35 ± 0.23 2.85 ± 0.03 0.06 ± 0.01 n/d 20.82 ± 0.24 n/d

By difference.

temperature under N2, weighed, homogenized, and subsampled for analysis. Combustion Trials/Ash Generation. Combustion trials using a 200 kWth bubbling FBC burning 100% poultry litter (as received) were conducted. A schematic of the system is given in Figure 1 and process conditions are given in Table 3. The system consisted of the FBC unit, two tubular air−water heat exchangers connected directly after the FBC, followed by a cyclone and a baghouse filter. This system generated five separate process ash (PA) streams: bed ash (BA), which was a mixture of bed material (sand), coated bed material, agglomerated bed material, and ash; ash from heat exchangers 1 and 2 (denoted as HE 1 and HE 2, respectively), cyclone ash (CYC), and baghouse ash (BH). Once steady-state conditions had been achieved,39 large steel containers were placed beneath each of the ash outlet points for a period of 1 h. The theoretical ash was calculated based on the feed rate and ash content from proximate analysis. The amount of BA was calculated by difference, since it was not feasible to separate the sand and ash fractions, because of agglomeration. The bed consisted of ∼140 kg of sand (98.29% SiO2, 1.29% Fe2O3, and 0.31% Al2O3) and was not replenished for the duration of the experiment. PA samples were taken on site, homogenized, and subsampled for further analysis. Generated ash (GA) was produced by heating samples of poultry litter to 550 °C in a muffle furnace for a minimum of 3 h. Ash Analysis. The PA samples composition and morphology were analyzed at different magnifications, using a JEOL JSM-840A scanning electron microscopy system, coupled with energy-dispersive X-ray spectroscopy (SEM-EDX). Subsamples of the ash were taken and mounted onto carbon tabs and sputter-coated with gold to prevent charging. Samples of the BA were mounted in epoxy resin, cured, and then ground and polished to reveal the cross section of the particles. These were mounted on a carbon tab using carbon paste to secure them to the mount, and sputter-coated with gold. SEM-EDX and line scans were used to identify the composition of the coating layers. Particle size analysis was conducted using a Malvern Mastersizer particle size analyzer, and using conventional BS sieving methods for the sand and BA fraction. A Rigaku 2100 X-ray fluorescence (XRF) spectrophotometer was used to identify major and minor elements in the PA, GA, and char

METHODOLOGY

Fuel Analysis. Poultry litter primarily consists of a mixture of manure (excreta) and bedding material. Wood shavings (Picea abis and Pinus sylvestris) formed the basis of all bedding material analyzed during this research. The most significant physical and chemical properties were characterized and reported in previous work.3 The proximate and ultimate analysis of poultry litter and wood shavings used in the experiments reported here are shown in Table 2. Char was generated from the litter by pyrolysis in a quartz fixed-bed batch reactor (950 mm (l) × 45 mm (id)), in a furnace fitted with a thermal controller (Eurothermal 2416), under nitrogen at atmospheric pressure. The reaction vessel was purged with N2 prior to addition of the litter. It was heated at a rate of 20 °C min−1 to 30 °C min−1 until the desired temperature (400, 600, and 800 °C) was reached and held for 30 min. These temperatures reflect the temperature range associated with the combustion process. The resultant char was cooled to room 4685

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694

Energy & Fuels

Article

Figure 1. Schematic of the process. samples; and a Rigaku RINT X-ray diffractometer (XRD) was used to identify crystalline species in the PA and GA. XRD analysis was conducted using a Cu Kα edge tube operated at 40 kV and 40 mA. Data was collected from 3° 2θ to 90° 2θ at a slow scanning speed (0.002° s−1). The crystalline species were identified by matching the peaks to the standards in the International Centre for Diffraction Data (ICDD) file. The samples for XRF and XRD analysis were ground to a fine powder, dried at 105 °C until constant in weight, and kept in a desiccator prior to analysis, because of their hygroscopic nature.

Table 3. Process Conditions process conditions

unit

value

fuel feed rate (range) PL particle size (range) sand particle size (range) weight of sand used bed pressure (range) bed height bed temp (range) free board temp (range) inlet bag filters temp exit temp fluidising air velocity secondary air velocity water vapor (flue gas) excess air

kg h−1 mm mm kg mbar mm °C °C °C °C m s−1 m s−1 % %

60.2 (55−65) 8.0 (6−8) 0.75 (0.55−0.95) 140 22 (17−23) 200 655 (619−688) 934 (898−994) 141 117 0.5 1.11 12.7 134.13

Figure 2. Flow diagram of thermodynamic equilibrium calculations.

Thermodynamic Equilibrium Modeling. Thermodynamic calculations were conducted using FactSage 6.140 to predict the gaseous and particulate inorganic species liberated during the combustion of poultry litter and their subsequent reactions downstream of the combustor. The program uses the method of minimization of the total Gibbs free energy of the system, and thermodynamic data taken from the Fact database included real gases, pure liquids and solids, as well as nonideal phases. Data from the FACT-SlagF (FSlag-liquid) model was also included. Slag−liquid oxide species that were accounted for in this model included the following: Na2O, SiO2, CaO, Al2O3, K2O, Na2SO4, K2SO4, CaSO4, NaCl, CaCl2, Na2CO3, and CaCO3. In this study, we compared the calculated values to experimental results in order to determine the amount of each element entrained by the flue gas and carried to the subsequent process section, rather than reaching equilibrium. The calculation flow diagram presented in Figure 2 was proposed for this purpose. In theory, once the litter is fed into the FBC, it will decompose rapidly (under pyrolysis conditions) to form char and volatiles. The inorganic elements will also partition into two major fractions: the first fraction is in the form of gaseous vapors or fine particulates released with the volatiles (typically Cl, K, Na, and S),

while the second fraction remains in the char matrix, (Si, Ca, P, Mg, Fe). Theoretically, the volatile fraction will be entrained by the flue gas and exit the reactor quickly, while the elements in the char are assumed to form the bed ash. However, in reality, in a fluidized-bed regime, it is likely that some of the char-bound elements will be swept away from the char surface and entrained by the flue gas. Similarly, some of the morevolatile elements may remain in the bed bound with stable species and could react with the char and bed material. The elemental composition of poultry litter (Table 2) and the combustion conditions (Table 3) were used as the input for the model. The calculated output from each step was compared with the experimental ash output, in terms of ash composition, to elucidate the unequilibrated fraction of each element. The fraction of each element in the solid and liquid compounds was calculated back to its most common oxide, for comparison with XRF data. For example, solid K3Na(SO4)2 was calculated to its equivalent weight of K2O, Na2O, and SO3. The input conditions for each step of the model are given in Table 4. Further details on the calculation procedure are given in File No. 1 in the Supporting Information (S.I.1). 4686

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694

Energy & Fuels

Article

Table 4. FactSage Input Conditions of Operation Input Conditions (kg h−1)



component

phase

input to bed

input to HE1

input to HE2

input to CYC

input to BH

N2 O2 CO2 H2O NO2 SO2 HCl (P2O3)2 KCl NO SO3 SiO2 Al2O3 Fe2O3 CaO MgO K2O SO3 Na2O

gas gas gas gas gas gas gas gas gas gas gas solid (quartz) solid (gamma) solid (hematite) solid (lime) solid (periclase) solid solid solid

1270 339 58.4 25.3 5.97 0.45 0.12 0.64

1276.8 342.4 58.52 25.38 0.008 0.20 0.05 0.41 0.115 0.31 0.03 0.035 0.04 0.13 1.48 0.225 0.81 0.60 0.05

1276.8 342.4 58.08 25.38

1276.8 342.4 57.64 25.38 0.0002

1276.8 342.4 57.2 25.38 0.0001

8.14 0.05 0.13 1.54 0.35 1.60 0.40 0.143

RESULTS AND DISCUSSION Characterization of Poultry Litter Ash. The results of particle size analysis of each of the ash samples (excluding BA) are shown in Figure 3. The results of the BA fraction will be discussed in subsequent sections. The BH had a single submicrometer peak with an average particle size of 0.29 μm, while the CYC and HEs exhibited multimodal distribution consisting of fine (2 μm) fractions, because of the mechanisms of ash formation during the different stages of combustion.41 The stages can be broadly categorized as evaporation of moisture, devolatilization (decomposition), and char burnout and fragmentation.13,15,42 During decomposition, volatile species present in the fuel (alkali metals, SO2/SO3, and chlorides) are released into the vapor phase. When these vapors cool, they condense to form the fine fraction of the ash,15 whereas char burnout and fragmentation generally give rise to the coarse fraction of the ash due to coalescence of the nonvolatile components contained within the fuel matrix (included minerals).13,15 The abrasive and turbulent nature of the FBC leads to more effective char attrition, but this may also lead to elutriation of the fine char particles. The results of the ash mass balance showed that, with a feed rate of ∼60.6 kg h−1, an expected 5.21 kg of ash was produced. Of this ash, the BA accounted for 63.34%, HE1 accounted for 21.98%, HE2 accounted for 10.23%, and the CYC and BH accounted for 4.2% and 0.23%, respectively. The compositions of these ash fractions, determined using XRF, are shown in Table 5. The results of the BA fraction are normalized to 3% SiO2 in order to eliminate the dilution effect of the bed material. Char and GA composition are also included for comparison. First, we can see evidence of elutriation of bed media in the increased concentration of SiO2 in the ancillary equipment, particularly HE 1 and CYC. Silicon, aluminum, and iron are all depleted in the BH, indicating that they are predominantly distributed in the coarser ash fractions. Corresponding levels of Ca, P, and K are present in the BA, HEs, and CYC fractions (19.3−26.0 wt %; 17.1−27.9 wt %, and 21.15−33.9 wt %, respectively), indicating the presence of stable Ca-phosphates and/or K−Ca-phosphates.

0.003 0.17 0.008 0.009

0.0003 0.08

0.03

0.03 0.12 1.19 0.13 0.64 0.83 0.03

0.03 0.115 1.06 0.07 0.52 0.79 0.02

0.02 0.11 1.00 0.05 0.47 0.78 0.01

The baghouse fraction is characterized by almost 70 wt % K2O, which may be a result of either delayed condensation of volatile K-species,43 or alternatively, condensation could occur earlier, but the small particle size of the condensed species, coupled with the short residence time, may have kept them airborne longer.44 XRF analysis is expressed as the elements most common oxide, but it is likely that the K in the BH is present as chlorides, sulfates, or carbonates.15,45 The morphology of the PA streams was analyzed using SEMEDX to determine the characteristic composition of the individual ash particles. Figures 4a−d show the results of this analysis. The BA fraction will be discussed separately. The EDX analysis of each stream was averaged over a minimum of four points or spectra, and the average is presented in Table 6 on a carbon- and oxygen-free basis. In accordance with XRF analysis, the EDX results for PA show that it is dominated by K, Ca, P, and Mg. Both HE 1 and HE 2 had broadly similar SEM images consisting of clusters of irregularly shaped particles, together with some spherical crystals. The CYC fraction contained both rodshaped and spherical crystals, dominated by K, Ca, Mg, and P, with small amounts of Na, S, and Cl. In the BH, we can see loosely agglomerated clusters with submicrometer- or nanoparticle-sized nuclei, indicating the predominance of the vaporization−condensation mechanism for ash formation here. Rod-shaped crystals coated in a thick gauzelike layer are also present. The EDX data of the BH reveals a clear dominance of K, Cl, and S. This confirms the high K concentration found in the XRF analysis, and indicates that gaseous K2SO4 and KCl will condense at the lower temperatures found in the baghouse. Since the presence of S as SO2/SO3 in the flue gas limits the formation of KCl,46 condensation of the K2SO4 is likely to occur prior to KCl for both homogeneous and heterogeneous mechanisms.15 XRD was used to identify the crystalline components in the GA and PA streams, the results of which are shown in Figure 5. The HEs and CYC exhibit largely similar patterns, containing K2SO4, MgO, KNaCa2(PO4)2, and Ca3(PO4)2. HE 2 contains quartz, indicating the presence of elutriated bed material. The BH fraction consists primarily of K2SO4 and KCl, verifying the SEM-EDX results, indicating that both KCl and K2SO4 form 4687

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694

Energy & Fuels

Article

Figure 3. Particle size distribution of PA.

Table 5. XRF Analysis of Poultry Litter Ash Samples Composition (wt %) generated ash char (800 °C) bed ash bed ashb heat exchanger 1 heat exchanger 2 cyclone baghouse a

SiO2

Al2O3

Fe2O3

CaO

MgO

P2O5

K2O

SO3

Na2O

TiO2

MnO

LOIa

2.8 2.7 52.65 3.0* 6.35 3.5 5.4 1.5

0.6 0.7 0.5 1.0 0.95 0.9 1.0 0.2

1.2 1.7 1.0 1.99 1.03 1.3 1.2 0.3

21.0 26.0 9.3 19.3 25.6 26.0 24.7 7.3

11.0 7.1 3.0 6.2 8.6 11.0 10.0 2.7

19.0 16.0 8.2 17.1 27.5 22.0 27.9 7.75

28.0 33.0 16.3 33.9 21.15 23.0 22.8 69.8

8.1 4.0 6.4 13.3 5.1 6.5 5.45 6.9

3.9 2.8 1.8 3.8 2.6 3.1 2.9 1.2

0.00 0.00 0.02 0.04 0.05 0.00 0.05 0.00

0.68 0.79 0.14 0.31 0.43 0.84 0.41 0.11

3.75 5.21 0.69 0.11 0.64 1.82 0.00 2.24

Loss on ignition. bCalculated to 3% SiO2.

Figure 4. SEM images of PA: (a) HE 1, (b) HE 2, (c) CYC, and (d) BH.

a significant part of the fine particulate matter. The absence of KCl in the BA indicates that K-sulfates are dominant over

K-chlorides at higher temperatures.15 Although K-carbonates were not observed, they could be present as amorphous species. 4688

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694

Energy & Fuels

Article

Table 6. EDX Analysisa of Poultry Litter Ash Samples

fresh sand, to 1550 μm for the BA mixture after combustion. Over 70% of the total BA mixture was in the size range of 1000− 2000 μm, with ∼20% in the range of 600−1000 μm, indicating that the majority of individual sand particles have become coated with a layer of ash, signifying the important role of coatinginduced agglomeration. The particle size of the residual bed ash may be further increased by coalescence of mineral inclusions or chemical reactions involving other species.47−49 Visual observation of the BA fraction revealed two distinct types of agglomerates: friable deposits of partially sintered ash and bed media shown in Figure 6a, and highly or totally sintered ash, shown in Figure 6b. In addition to the two types of large agglomerates, it was observed that the individual sand particles became coated with ash and showed some evidence of cluster formation. This was confirmed by SEM, which revealed coarse particles of raspberry-like clusters, measuring ∼1 mm × 0.6 mm (see Figure 6c). EDX analysis revealed the coating on the bed sand grains consisted of Ca, K, P, and Mg, most likely in oxide form, with sulfates also present. SEM observation on the cross section of BA was carried out, and the results are shown in Figure 7a. From this, it is evident that some reaction with the surface of the sand had occurred, leading to a two-stage coating-induced mechanism. The initial layer is noncontinuous, but solid and glassy in appearance where it does appear on the surface of the sand, measuring between 0.05−0.09 mm. The outer layer appears to be composed of loosely bound, fine ash particles originating from the surface of the char or inside the char matrix. This is consistent with the scientific literature, where two distinct layers are generally differentiated; an inner coating dominated by the attack or reaction of ash constituents with the bed material, and an outer coating consisting of granular material, formed by the adhesion of ash particles to the inner layer.25,26,50,51 A line scan was taken on the magnified imagelocation denoted by a white arrow in Figure 7a (inset)and the compiled EDX analysis of this scan is shown in Figure 7b. This shows that the bulk of the initial layer is dominated by K and Si, indicating that molten K-silicates have formed and initiated the agglomeration mechanism. Moreover, since P is not present in the inner layer, it is likely the K here is derived primarily from organically bound K in the litter, which was readily evaporated and penetrated the pores of the coarse quartz particle. Similar reactions may be responsible for the presence of Na, Ca, and S in the inner layer, or these could be as a result of small mineral inclusions in the molten layer. Chemical reactions on the surface of the bed material may have taken place to strengthen interparticle bonds and form molten glassy phases which developed into tightly sintered structures.23 The neck is also dominated by Si and K, while the presence of Ca and S are also more apparent compared to the inner layer. The content of elements other than Si and K begin to increase from the neck into the outer layer, especially P and Ca as well as Mg, S, Fe, and Na, which originated from the char particles. These data indicate that P is not involved in the initial layer formation, but does form a significant part of the outer layer, because of the stability of the original phosphates. The importance of K-silicates in the agglomerate initiation is evident, and is in accordance with literature.11,15,52,53 Thermodynamic Equilibrium Analysis. The first step in the thermodynamic analysis predicted the speciation of the major elements in the bed between 600 °C and 1000 °C in 50 °C steps and included transitional stages. The results of this step are summarized in Figures 8a−c. (Input conditions are given in Table 4.)

Composition (wt %)

a

element

HE 1

HE 2

CYC

BH

Na Mg Si P S K Ca Mn Cl

10.58 19.01 3.69 23.78 2.12 17.38 22.88 0.57

5.08 4.37

3.15 16.61 3.89 16.33 3.57 26.52 23.01 4.59 2.32

2.10 1.14 0.14 1.36 11.15 55.29 2.88

11.23 6.67 45.58 14.69 12.38

25.95

Normalized to carbon-free and oxygen-free conditions.

Figure 5. XRD analyses of GA and PA. α − SiO2; β − K2(SO4); γ − KCl; δ − MgO; ε − KNaCa2(PO4)2; η − Ca3(PO4)2; ω − K3Na(SO4)2; ϕ − Ca5(PO4)3(OH).

GA was found to contain K2SO4 and KCl, in addition to KNaCa2(PO4)2 and Ca5(PO4)3(OH), verifying the presence of original Ca-phosphates in the litter, including K−Na−Caphosphates, which appear to have underwent limited transformations during the combustion process. Results of the PA analysis show that the combustion of poultry litter results in ash dominated by crystalline coarse ash composed of alkali-Ca-phosphates, with fines of K2SO4 and KCl. Bed Agglomeration. Particle size analysis of the BA mixture showed an increase in average particle size from 750 μm for the 4689

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694

Energy & Fuels

Article

Figure 6. SEM images of BA: (a) friable deposits, (b) highly sintered ash, (c) BA particles.

Figure 7. (a) Cross-sectional analysis of BA; (b) line scan EDX data of the coating.

In Figure 8a, we can see that gaseous HCl and NO are dominant, accounting for 97% of the total phase at lower temperatures, and 64% at higher temperatures (concentrations of N2, O2, H2O and CO2 are excluded from the calculation to improve recognition of the minor gases). Gaseous KCl and SO2 are the next most abundant, with the highest concentration of gaseous KCl at 750 °C, and SO2 being predicted to increase at temperatures of >850 °C. Gaseous NaCl is also present in significant quantities and highest at 850 °C. The dominance of K and Cl containing gases is due to their volatility, and their presence in the fine fraction of the ash is primarily due to the vaporization−condensation mechanism of ash formation.15

The onset of bed agglomeration can lead to areas of partial defluidization, which gives rise to localized hot spots where temperatures rise above the softening temperature of certain elements and compounds, allowing fuel particles to stick together, leading to clinker formation.54 In this context, the quantity of slag phase predicted rose from 1.8 kg at 600 °C to 7.35 kg at 1000 °C, based on average hourly operation of 60.6 kg of litter with 5.21 kg of ash. In Figure 8b, we can see that SiO2 is dominant at all temperatures, rising from 70%−75% of total slag, while K2O is the next most abundant, ranging from 25%−20% with increasing temperature; the remaining 5% consists primarily of Al2O3, K2SO4, and Na2O. This indicates the importance of 4690

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694

Energy & Fuels

Article

Figure 8. Speciation of major elements from 600 °C to 1000 °C: (a) gas phase, (b) slag phase, and (c) solid phase.

compounds, and this was confirmed by XRD analysis, which identified the presence of Ca5(PO4)3(OH) (hydroxyapatite) in the GA and Ca3(PO4)2 (whitlockite) in the BA samples. Hydroxyapatite is only predicted to form within the range of 600−730 °C, probably because of reactions between the calciumrich phosphates and water vapor in the flue gas.37 Overall, the model predicts that >99% of all calcium is present in solid crystalline form. FactSage 6.1 does not include a slag model accounting for the high concentrations of phosphate found in poultry litter, and this may result in a deviation between observed behaviors and the predictions of the chemical equilibrium calculations.56 However, it was found that P was present in the GA as originally bound Ca-phosphates, so it is likely that limited transformations

K-silicates in the bed agglomeration mechanism. In accordance with the FactSage prediction, the line scan (Figure 7b) indicated the presence of significant amounts of Si and K, together with small amounts of S, Na, and Ca, in the initial melt. The formation of low melting Ca-based silicates was not predicted in the calculations due to the low bed temperatures experienced. The lowest eutectic melting point of Ca is ∼1265 °C,55 so the presence of calcium in the melt (Figure 7b) is likely due to original Ca-compounds in the char, e.g., hydroxyapatite, which became included in the K/Na melt phase. In Figure 8c, we can see that the solid phase is dominated by SiO2 due to the contribution from the bed material. K2SO4 is a significant part of the solid phase and decreases significantly after 935 °C. P is only predicted to be present in calcium-containing 4691

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694

Energy & Fuels

Article

occurred.57,58 Thermodynamic calculations also assume infinite residence time, perfect mixing and do not account for reaction kinetics or transport limitations.52 As a result, the calculated equilibrium results are not always comparable with real data. To validate the projections, the predicted results were compared with experimental data. The amount of BA predicted to form was greater than that determined by the ash mass balance. Therefore, we can calculate x2 (Figure 2) as being the difference between the equilibrium result for char combustion and the process BA result. Having calculated x1 as the difference in elements between the original poultry litter and the ash from the char, i.e., the volatilebound elements, we can combine with x2 to form the input for the heat exchangers section, and so on for the remaining sections. The extent to which equilibrium is reached for each of the major elements is presented in Table 7. The highest degrees of

baghouse, where Ca-species were anticipated to dominate. This could be due to entrainment of fine particulate K2SO4 from the bed to the baghouse. XRD results show the presence of KCl in the baghouse, while FactSage predicted it to have fully condensed in the heat exchangers. Again, this illustrates the difficulty in condensation of these gases at lower temperatures. The formation of 18 different crystalline structures was predicted, and XRD analysis confirmed the presence of 6. It is possible that the cooling rate of gases was too rapid to allow for preferential crystal formation, and amorphous structures that contain the same elements dominate instead.44



CONCLUSIONS The feed requirements of poultry results in a litter which is high in potassium, phosphorus, calcium, and magnesium, which increases the propensity for ash-related problems during combustion. The major conclusions achieved from this study are as follows: (1) The coarse ash was found to be dominated by alkali-Caphosphates, while the fine fraction was dominated by K2SO4 and KCl. (2) Bed agglomeration was found to be coating-induced, with two distinct layers present. The inner layer (0.05−0.09 mm thick) was formed due to the reaction of gaseous potassium with the sand (SiO2) surface, forming Ksilicates. Further chemical reaction on the surface of the bed material strengthened the coating, forming a molten glassy phase. The outer layer was composed of loosely bound, fine particulate ash originating from the char. (3) Deposition in the subsequent regions was found to occur mostly through the vaporization−condensation mechanism, with equilibrium decreasing significantly with decreasing temperatures. The dominant alkali chloride containing gas predicted to form in the combustion zone was KCl, which corresponds to the high KCl content in the fine baghouse ash. (4) Thermodynamic equilibrium calculations showed that slag formation increased with increasing temperature, and that SiO2 and K2O were the dominant species in the slag phase, accounting for almost 95%. The remainder consisted mostly of Al2O3, K2SO4, and Na2O.

Table 7. Fraction of Equilibrated Elements % Equilibrium element

BA

HE 1

HE 2

CYC

BH

Al2O3 Fe2O3 CaO MgO K2O Na2O P2O5 SO3

32.00 24.24 19.24 27.97 33.41 42.19 30.61 26.92

27.50 8.99 19.72 43.46 27.55 46.73 58.83 6.56

17.28 5.67 11.59 46.13 19.13 47.89 53.45 4.19

8.33 2.61 5.11 32.35 9.69 33.33 59.80 1.51

0.12 0.03 0.07 0.71 1.81 1.24 2.27 0.08

overall equilibrium achieved are in the BA and HE 1, with the subsequent sections showing a progressive decrease. This is because equilibrium is much more difficult to achieve at lower temperatures. In the bed, we see that both Na and K achieve the highest equilibrium (42.19% and 30.61%, respectively), while Ca is the lowest at 19.24%. This can be explained by a preferential presence of Na and K in the gas phase while Ca is mainly present as solid oxide particles in the flue gas. The homogeneous reactions for Na and K in the gas phase are also much faster than that for Ca in the solid phase. Sulfur also achieves 26.92% in the BA, but is below 10% in the subsequent sections. P achieves the highest equilibrium level in the HEs and CYC fractions, confirming the stability of phosphates in the ash. In the baghouse, the maximum equilibrium is just over 2%, again indicating the difficulty in achieving equilibrium at low temperatures. Table 8 presents the crystalline species predicted to form using FactSage, with the species identified by the XRD results highlighted in bold. It correctly identified K2SO4 as a significant species in the bed, heat exchangers, and cyclone; however, contrary to the XRD results, it was not predicted to form in the



ASSOCIATED CONTENT

S Supporting Information *

The procedure for thermodynamic equilibrium modeling throughout the overall process is detailed in this section. This material is available free of charge via the Internet at http://pubs. acs.org.

Table 8. Crystalline Species Predicted by FactSagea

a

BA

HEs

CYC

BH

SiO2 K2SO4 Ca3(PO4)2 Na2Ca3Si6O16 MgOCaOSi2O4 Fe2O3 K3Na(SO4)2

K2SO4 Ca5(PO4)3(OH) MgO KCl (MgO)(Fe2O3) MgAl2O4

K2SO4 Fe2O3 MgCO3 MgAl2O4 Ca5(PO4)3(OH)

CaMg(CO3)2 KNO3 CaSO4 K3Na(SO4)2 Ca5(PO4)3(OH) CaCO3 Fe2O3 KAl9O14

Species highlighted in bold font are those which were identified using XRD. 4692

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694

Energy & Fuels



Article

(15) Doshi, V.; Vuthaluru, H. B.; Korbee, R.; Kiel, J. H. A. Development of a modeling approach to predict ash formation during co-firing of coal and biomass. Fuel Process. Technol. 2009, 90 (9), 1148− 1156. (16) Frandsen, F. J.; van Lith, S. C.; Korbee, R.; Yrjas, P.; Backman, R.; Obernberger, I.; Brunner, T.; Joller, M. Quantification of the release of inorganic elements from biofuels. Fuel Process. Technol. 2007, 88 (11− 12), 1118−1128. (17) Zevenhoven-Onderwater, M.; Backman, R.; Skrifvars, B. J.; Hupa, M.; Liliendahl, T.; Rosen, C.; Sjostrom, K.; Engvall, K.; Hallgren, A. The ash chemistry in fluidised bed gasification of biomass fuels. Part II: Ash behaviour prediction versus bench scale agglomeration tests. Fuel 2001, 80 (10), 1503−1512. (18) Llorente, M. J. F.; Laplaza, J. M. M.; Cuadrado, R. E.; García, J. E. C. Ash behaviour of lignocellulosic biomass in bubbling fluidised bed combustion. Fuel 2006, 85 (9), 1157−1165. (19) Yan, L.; Gupta, R. P.; Wall, T. F. The implication of mineral coalescence behaviour on ash formation and ash deposition during pulverised coal combustion. Fuel 2001, 80 (9), 1333−1340. (20) Wang, L.; Hustad, J. E.; Skreiberg, Ø.; Skjevrak, G.; Grønli, M. A Critical Review on Additives to Reduce Ash Related Operation Problems in Biomass Combustion Applications. Energy Procedia 2012, 20 (0), 20−29. (21) van Lith, S. C.; Jensen, P. A.; Frandsen, F. J.; Glarborg, P. Release to the Gas Phase of Inorganic Elements during Wood Combustion. Part 2. Influence of Fuel Composition. Energy Fuels 2008, 22 (3), 1598− 1609. (22) Knudsen, J. N.; Jensen, P. A.; Dam-Johansen, K. Transformation and Release to the Gas Phase of Cl, K, and S during Combustion of Annual Biomass. Energy Fuels 2004, 18 (5), 1385−1399. (23) Baxter, L. L.; Miles, T. R.; Jenkins, B. M.; Milne, T.; Dayton, D.; Bryers, R. W.; Oden, L. L. The behavior of inorganic material in biomassfired power boilers: Field and laboratory experiences. Fuel Process. Technol. 1998, 54 (1−3), 47−78. (24) Barišić, V.; Åmand, L.-E.; Coda Zabetta, E. The role of limestone in preventing agglomeration and slagging during CFB combustion of high phosphorus fuels. Presented at World Bioenergy; Jönköping, Sweden, 2008. (25) Brus, E.; Ö hman, M.; Nordin, A. Mechanisms of Bed Agglomeration during Fluidized-Bed Combustion of Biomass Fuels. Energy Fuels 2005, 19 (3), 825−832. (26) Ö hman, M.; Pommer, L.; Nordin, A. Bed Agglomeration Characteristics and Mechanisms during Gasification and Combustion of Biomass Fuels. Energy Fuels 2005, 19 (4), 1742−1748. (27) Jenkins, B. M.; Baxter, L. L.; Miles, T. R., Jr.; Miles, T. R. Combustion properties of biomass. Fuel Process. Technol. 1998, 54 (1− 3), 17−46. (28) Jiménez, S.; Ballester, J. Influence of operating conditions and the role of sulfur in the formation of aerosols from biomass combustion. Combust. Flame 2005, 140 (4), 346−358. (29) Jiménez, S.; Ballester, J. Formation of alkali sulphate aerosols in biomass combustion. Fuel 2007, 86 (4), 486−493. (30) Baxter, L. L. Ash deposition during biomass and coal combustion: A mechanistic approach. Biomass Bioenergy 1993, 4 (2), 85−102. (31) European Commission. The Welfare of Chickens Kept for Meat Production (Broilers) 21/03/2000, 2000. (32) Gaál, K. K.; Sáfár, O.; Gulyás, L.; Stadler, P. Magnesium in Animal Nutrition. J. Am. College Nutrit. 2004, 23 (6), 754S−757S. (33) Gous, R. M. Nutritional limitations on growth and development in poultry. Livestock Sci. 2010, 130 (1−3), 25−32. (34) Oliveira, J.; Albino, L.; Rostagno, H.; Páez, L.; Carvalho, D. Dietary levels of potassium for broiler chickens. Rev. Brasil. Ciência ́ Avicola 2005, 7, 33−37. (35) Bolan, N. S.; Naidu, R.; Anderson, C. Management of Phosphorus in Organic Amendments for Sustainable Production and Environmental Protection; Occasional Report No. 25, Fertilizer and Lime Research Centre, Massey University, Palmerston North, New Zealand, 2012; p 6.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS Funding for this work has been provided by the Department of Agriculture, Fisheries and Food, under RSF No. 07 559, and also through the Irish Research Council for Science, Engineering and Technology (IRCSET). The authors would like to acknowledge the EUOSSIC program, funded by the EU which facilitated this collaboration. The authors would also like to acknowledge the assistance provided by Ms. Ellen Lavoie of the MCEM, Monash University for her assistance in preparing the cross section of the bed ash particles.



REFERENCES

(1) On the promotion of the use of energy from renewable sources and amending and subsequently repealing Directives 2001/77/EC and 2003/ 30/EC (Text with EEA relevance), Council Directive (EC) 2009/28/EC, April 23, 2009. (2) McKendry, P. Energy production from biomass (Part 1): Overview of biomass. Bioresour. Technol. 2002, 83 (1), 37−46. (3) Lynch, D.; Henihan, A. M.; Bowen, B.; Lynch, D.; McDonnell, K.; Kwapinski, W.; Leahy, J. J. Utilisation of poultry litter as an energy feedstock. Biomass Bioenergy 2013, 49 (0), 197−204. (4) Khan, A. A.; de Jong, W.; Jansens, P. J.; Spliethoff, H. Biomass combustion in fluidized bed boilers: Potential problems and remedies. Fuel Process. Technol. 2009, 90 (1), 21−50. (5) Anthony, E. J. Fluidized bed combustion of alternative solid fuels: Status, successes and problems of the technology. Prog. Energy Combust. Sci. 1995, 21 (3), 239−268. (6) Werther, J.; Saenger, M.; Hartge, E. U.; Ogada, T.; Siagi, Z. Combustion of agricultural wastes. Prog. Energy Combust. Sci. 2000, 26, 1−27. (7) Abelha, P.; Gulyurtlu, I.; Boavida, D.; Seabra Barros, J.; Cabrita, I.; Leahy, J. J.; Kelleher, B.; Leahy, M.; Henihan, A. M. Combustion of poultry litter in a fluidised bed combustor. Fuel 2003, 82 (6), 687−692. (8) Henihan, A. M.; Leahy, M. J.; Leahy, J. J.; Cummins, E.; Kelleher, B. P. Emissions modeling of fluidised bed co-combustion of poultry litter and peat. Bioresour. Technol. 2003, 87 (3), 289−294. (9) Zhu, S.; Lee, S. W. Co-combustion performance of poultry wastes and natural gas in the advanced Swirling Fluidized Bed Combustor (SFBC). Waste Manage. 2005, 25 (5), 511−518. (10) Tuohy, A.; Bazilian, M.; Doherty, R.; Gallachóir, B. Ó .; O’Malley, M. Burning peat in Ireland: An electricity market dispatch perspective. Energy Policy 2009, 37 (8), 3035−3042. (11) Piotrowska, P.; Grimm, A.; Skoglund, N.; Boman, C.; Ö hman, M.; Zevenhoven, M.; Boström, D.; Hupa, M. Fluidized-Bed Combustion of Mixtures of Rapeseed Cake and Bark: The Resulting Bed Agglomeration Characteristics. Energy Fuels 2012, 26 (4), 2028−2037. (12) Tarelho, L. A. C.; Neves, D. S. F.; Matos, M. A. A. Forest biomass waste combustion in a pilot-scale bubbling fluidised bed combustor. Biomass Bioenergy 2011, 35 (4), 1511−1523. (13) Zevenhoven-Onderwater, M. F. J. Ash Forming Matter in Biomass Fuels, Ph.D. Thesis, Åbo Akademi University, Turku, Finland, 2001 (ISBN 952 12 0813 9). (14) Teixeira, P.; Lopes, H.; Gulyurtlu, I.; Lapa, N.; Abelha, P. Evaluation of slagging and fouling tendency during biomass co-firing with coal in a fluidized bed. Biomass Bioenergy 2012, 39 (0), 192−203. 4693

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694

Energy & Fuels

Article

(36) Hobbs, P., Phosphorus recovery from unprocessed manure. In Phosphorus in Environmental Technology: Principles and Applications; Valsami-Jones, E., Ed.; IWA Publishing: London, 2004. (37) Lindström, E.; Sandström, M.; Boström, D.; Ö hman, M. Slagging Characteristics during Combustion of Cereal Grains Rich in Phosphorus. Energy Fuels 2007, 21 (2), 710−717. (38) Piotrowska, P.; Zevenhoven, M.; Hupa, M.; Giuntoli, J.; de Jong, W. Residues from the production of biofuels for transportation: Characterization and ash sintering tendency. Fuel Process. Technol. 2013, 105, 37−45. (39) Methods for Assessing thermal performance of boilers for steam, hot water and high temperature heat transfer fluidsPart 1: Concise procedure, British Standard BS 845-1; British Standards Institute: London, 1987. (40) Bale, C. W.; Bélisle, E.; Chartrand, P.; Decterov, S. A.; Eriksson, G.; Hack, K.; Jung, I. H.; Kang, Y. B.; Melançon, J.; Pelton, A. D.; Robelin, C.; Petersen, S. FactSage thermochemical software and databasesRecent developments. Calphad 2009, 33 (2), 295−311. (41) Wu, H.; Wall, T.; Liu, G.; Bryant, G. Ash Liberation from Included Minerals during Combustion of Pulverized Coal: The Relationship with Char Structure and Burnout. Energy Fuels 1999, 13 (6), 1197−1202. (42) Wu, H.; Bryant, G.; Wall, T. The Effect of Pressure on Ash Formation during Pulverized Coal Combustion. Energy Fuels 2000, 14 (4), 745−750. (43) Nielsen, H. P.; Frandsen, F. J.; Dam-Johansen, K.; Baxter, L. L. The implications of chlorine-associated corrosion on the operation of biomass-fired boilers. Prog. Energy Combust. Sci. 2000, 26 (3), 283−298. (44) Li, L.; Yu, C.; Huang, F.; Bai, J.; Fang, M.; Luo, Z. Study on the Deposits Derived from a Biomass Circulating Fluidized-Bed Boiler. Energy Fuels 2012, 26 (9), 6008−6014. (45) Hupa, M. Ash-Related Issues in Fluidized-Bed Combustion of Biomasses: Recent Research Highlights. Energy Fuels 2011, 26 (1), 4− 14. (46) Elled, A. L.; Davidsson, K. O.; Åmand, L.-E. Sewage sludge as a deposit inhibitor when co-fired with high potassium fuels. Biomass Bioenergy 2010, 34 (11), 1546−1554. (47) Coda, B. Studies on Ash Behaviour during Co-Combustion of Paper Sludge in Fluidized Bed Boilers. University of Stuttgart, Germany, Stuttgart, 2004. (48) Laumb, J. D.; Folkedahl, B. C.; Zygarlicke, C. J. Characteristics and Behavior of Inorganic Constituents. In Combustion Engineering Issues for Solid Fuel Systems, Miller, B. G.; Tillman, D. A., Eds.; Academic Press: Amsterdam, Boston, 2008; Chapter 4, pp 133−170. (49) Quann, R. J.; Neville, M.; Sarofim, A. F. A laboratory study on the effect of the coal selection on the amount and composition of combustion generated submicron particles. Combust. Sci. Technol. 1990, 74, 245−265. (50) Grimm, A.; Ö hman, M.; Lindberg, T.; Fredriksson, A.; Boström, D. Bed Agglomeration Characteristics in Fluidized-Bed Combustion of Biomass Fuels Using Olivine as Bed Material. Energy Fuels 2012, 26 (7), 4550−4559. (51) Grimm, A.; Skoglund, N.; Boström, D.; Ö hman, M. Bed Agglomeration Characteristics in Fluidized Quartz Bed Combustion of Phosphorus-Rich Biomass Fuels. Energy Fuels 2011, 25 (3), 937−947. (52) Nordgren, D.; Hedman, H.; Padban, N.; Boström, D.; Ö hman, M. Ash transformations in pulverised fuel co-combustion of straw and woody biomass. Fuel Process. Technol. 2013, 105, 52−58. (53) Tillman, D. A. Introduction. In Combustion Engineering Issues for Solid Fuel Systems; Miller, B. G.; Tillman, D. A., Eds.; Academic Press: Amsterdam, Boston, 2008; Chapter 1, pp 1−32. (54) Basu, P., Combustion and Gasification in Fluidized Beds; Taylor & Francis: Boca Raton, FL, 2006. (55) Harb, J. N.; Munson, C. L.; Richards, G. H. Use of equilibrium calculations to predict the behavior of coal ash in combustion systems. Energy Fuels 1993, 7 (2), 208−214. (56) Lindberg, D.; Backman, R.; Chartrand, P.; Hupa, M. Towards a comprehensive thermodynamic database for ash-forming elements in biomass and waste combustionCurrent situation and future developments. Fuel Process. Technol. 2013, 105, 129−141.

(57) Crannell, B. S.; Eighmy, T. T.; Krzanowski, J. E.; Eusden, J. D., Jr,; Shaw, E. L.; Francis, C. A. Heavy metal stabilization in municipal solid waste combustion bottom ash using soluble phosphate. Waste Manage. 2000, 20 (2−3), 135−148. (58) Gupta, S.; Dubikova, M.; French, D.; Sahajwalla, V. Effect of CO2 Gasification on the Transformations of Coke Minerals at High Temperatures. Energy Fuels 2007, 21 (2), 1052−1061. (59) Zevenhoven-Onderwater, M.; Backman, R.; Skrifvars, B.-J.; Hupa, M. The ash chemistry in fluidised bed gasification of biomass fuels. Part I: Predicting the chemistry of melting ashes and ash−bed material interaction. Fuel 2001, 80 (10), 1489−1502. (60) Vassilev, S. V.; Baxter, D.; Andersen, L. K.; Vassileva, C. G. An overview of the chemical composition of biomass. Fuel 2010, 89 (5), 913−933. (61) Piotrowska, P.; Zevenhoven, M.; Davidsson, K.; Hupa, M.; Åmand, L.-E.; Barišić, V.; Coda Zabetta, E. Fate of Alkali Metals and Phosphorus of Rapeseed Cake in Circulating Fluidized Bed Boiler Part 1: Cocombustion with Wood. Energy Fuels 2010, 24 (1), 333−345. (62) Theis, M.; Skrifvars, B.-J.; Hupa, M.; Tran, H. Fouling tendency of ash resulting from burning mixtures of biofuels. Part 1. Deposition rates. Fuel 2006, 85 (7−8), 1125−1130.

4694

dx.doi.org/10.1021/ef400744u | Energy Fuels 2013, 27, 4684−4694