fuel processing technology - startsidausers.abo.fi/mzevenho/portfolj/publikationer/refereed...

Fuel Processing Technology 105 (2013) 37–45

Contents lists available at SciVerse ScienceDirect

Fuel Processing Technology

j ourna l homepage: www.e lsev ie r .com/ locate / fuproc

Residues from the production of biofuels for transportation: Characterization and ashsintering tendency

Patrycja Piotrowska a,⁎, Maria Zevenhoven a, Mikko Hupa a, Jacopo Giuntoli b, Wiebren de Jong b

a Åbo Akademi University, Process Chemistry Centre, Biskopsgatan 8, Åbo (Turku), Finlandb Delft University of Technology, Process and Energy Department, Energy Technology Section, Leeghwaterstraat 44, Delft, The Netherlands

⁎ Corresponding author. Tel.: +358 2 215 3507.E-mail address: [email protected] (P. Piotrowska).

0378-3820/$ – see front matter © 2011 Elsevier B.V. Alldoi:10.1016/j.fuproc.2011.09.020

a b s t r a c t
a r t i c l e i n f o
Article history:Received 7 January 2011Received in revised form 6 September 2011Accepted 29 September 2011Available online 1 November 2011

Keywords:Ash forming matterAsh sinteringFuel characterizationBiofuel residues

The continuously growing demand for fuels in the transportation sector has led to the increased use of liquid bio-fuels, with a corresponding increase in the amounts of production residues.With energy recovery, these residuescould become a source of biomass fuels. The current work investigated the behaviour of four potential biomassfuels during combustion. Fermented sewage sludge (FSS) is the residue frombiogas production via anaerobic fer-mentation; distillers' dried grains with solubles (DDGS) are a residue from barley bioethanol production; andrapeseed cake (RC) and palm kernel cake (PKC) are both oil extraction residues from first generation biodieselproduction plants. Since the main challenges during biomass combustion are related to ash behaviour, thispaper focuses on the ash-formingmatter in the fuels and its effect on the ash sintering tendency. Three differenttypes of tests were applied to study ash sintering: compression strength, microsample sintering, and standardash sintering. The latter was carried out using hot stage microscopy and computer aided image analysis, andthe resultswere comparedwith those from the compression strength andmicrosample tests. To obtain addition-al information about the nature and extent of the sintering, ash samples before and after sintering were exam-ined with a scanning electron microscope equipped with energy dispersive X-ray (SEM/EDX) technology anda stereomicroscope incorporating a digital camera. Tomake it possible to study the ash-formingmatter, chemicalfractionationwere performed accompanied by fuel proximate analyses bymeans of thermogravimetric analyzer.The results clearly showed that the lowest initial sintering temperatures were below 800 °C for RC and DDGS inall the sintering tests, while PKC and FSS did not exhibit any significant sintering below 900 °C. This could beinterpreted tomean thatwhen the combustion temperature is below 900 °C, no significant ash-related problemswill occur. The chemistry of the fuel ashes investigated is dominated by phosphorus, and the shift between alkalimetal and alkaline earth metal phosphates governs the sintering behaviour. The FSS sintering tendency is dom-inated by the iron oxide chemistry.

rights reserved.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

One of the key targets set by the European Council is an increase inthe renewable energy share to 20% of European Union (EU)energy consumption by 2020 [1]. This can be achieved only if a portionof the fossil fuels are replaced by biomass. As suggested by the EuropeanEnvironment Agency (EEA) data for 2007, the present proportion of re-newables in primary energy consumption is 8% [2], which implies thatmember countries will have to make extraordinary efforts to meet thetarget. Because energy demand is constantly growing, new fuels are re-quired, so in recent years, increased attention has been paid to energyrecovery from biomass or waste-derived fuels.

The transportation sector can become an important source of newbiomass fuels. The constant escalation of energy demand in the trans-portation sector, oil import dependence, and climate change are the

most significant factors driving legislation to enforce an increase inthe proportion of biofuels in the transportation fuel market. EuropeanDirective 2009/28/EC imposes a reference value of 10% for the use ofbiofuels in this sector by the year 2020 [1]. According to EEA data [3],the proportion of biofuels in the EU-27 transportation sector reached3% in 2007; in 2006, it was below 2%, so the biofuels market is growingand must do so rapidly. The increased amounts of production residueswill then be available for energy recovery.

The literature has reported extensively on the combustion of bio-mass and related operating problems [4–11]. The main challenges dur-ing biomass combustion are related to ash behaviour and before newbiomass fuels can be put into use, ash properties need to be determined.

The aimof this paper is to investigate the ash sintering tendency of po-tential biomass fuels in terms of their behaviour during combustion. Foursolid residues from the production of liquid and gaseous biofuels are de-scribed: fermented sewage sludge (FSS) from biogas production via an-aerobic fermentation; dried distillers' grains with solubles (DDGS) frombarley bioethanol production; rapeseed cake (RC); and palm kernelcake (PKC). The last two are oil extraction residues from first generation

http://dx.doi.org/10.1016/j.fuproc.2011.09.020

mailto:[email protected]

http://dx.doi.org/10.1016/j.fuproc.2011.09.020

http://www.sciencedirect.com/science/journal/03783820

Fig. 1. Sintering tendency classification after Moilanen A. [23].

Fig. 2. Phases in the ash melting process (original shape=shape and size at 550 °C) according to CEN/TS 15370–1:2006 (E).

0

10

20

30

40

50

60

70

80

RC DDGS PKC FSS

wt-

% d

ry f

uel

volatile matter fixed carbon ash

Fig. 3. Proximate analyses.

38 P. Piotrowska et al. / Fuel Processing Technology 105 (2013) 37–45

biodiesel production plants. The same fuels were described in detail byGiuntoli et al. [12], who investigated the fate of fuel nitrogen.

The sintering tendency of biomass fuels was studied in the laborato-ry by several authors, e.g., [13–16], while Bartels et al. conducted a re-view of available laboratory methods for bed sintering prediction interm of fluidized beds [17]. Steenari et al. [13] investigated rapeseedcake, among other fuels. However, the samples had lower sodium con-tent than those used in the present work. The sintering of sewagesludge ash was studied byWang et al. [16], who found that, for sludgestreated with Fe2(SO4)3 at a wastewater treatment plant, sinteringstarted at 1000 °C. This result will be compared with those obtained inthe current study for fermented sewage sludge. The sintering tendencyof the two other fuels presented here has not been reported elsewhere.

Ash sintering was studied using three different tests: standard ashsintering, compression strength, andmicrosample sintering. The resultsfrom the fuel proximate analysis and chemical fractionation are pre-sented along with those from ash sintering tests as will be demonstrat-ed in Section 3. Ashed fuel samples were analyzed using a scanningelectronmicroscope equippedwith energy dispersive X-ray (SEM/EDX)technology and a stereomicroscope incorporating a digital camera.

2. Materials and methods

2.1. Fuel and ash characterization

The FSS used in this work came from a plant operating an anaero-bic fermentation process for biogas production (Ab Stormossen Oy).The DDGS originated from a barley bioethanol production plant(Abengoa), while both the PKC and RC are residues from oil seed bio-diesel production (E. On and Emmelev A/S, respectively).

The fuel proximate analysis was performed bymeans of thermogravi-metric analysis (TGA). The sampleswere pre-dried andground to obtain amean particle size less than 1.0×10−3 m. First, each fuel sample washeated to 105 °C at a heating rate of 0.33 °C/s and kept at a constant tem-perature until the mass stabilized to allow complete drying. Next, thetemperature was increased to 550 °C at the same heating rate to deter-mine the volatile content. When the mass stabilized, only char remained.During both stages, an inert gas (N2) flow was maintained at the level of1.7×10−6 m3/s. To determine the ash content, the N2 was replaced by

synthetic air (a mixture of 20%vol. O2 and 80%vol. N2) to allow completecombustion of the organic residue. This method was adapted fromMayoral et al. [18]. To assess the repeatability of the tests, they werecarried out in triplicate. Five repetitions were performed for the FSSsample.

Additionally, chemical fractionation was applied to all the fuels to de-termine the chemical compounds which contain the ash-formingmatter.Chemical fractionation is based on selective consecutive leaching bywater (H2O), 1 M ammonium acetate (NH4Ac), and 1 M hydrochloricacid (HCl) [19–22]. Subjecting the untreated fuel samples to the increas-ingly aggressive solvents produces a series of four fractions. The untreatedsamples, the three liquid fractions, and the remaining solids were ana-lyzed by an external laboratory according to Swedish standards.

Standard ash was prepared according to CEN/TS 15403 to determinethe ash content. A fuel sample was ground to a mean particle size lessthan 1.0×10−3 m (more than 90%) and, for FSS, less than 0.25×10−3 m(more than 80%). Then, it was placed in a laboratory furnace in such away that the sample loading did not exceed 1.0 kg/m2. The furnace washeated to 250 °C at a rate of 0.083 °C/s. The samplewas left at this temper-ature for 1 h to allow devolatilization before ignition. Afterwards, the fur-nace was heated at 0.083 °C/s to 550 °C. The sample was maintained atthis temperature for 2 h, removed from the furnace, cooled in ambientair for approximately 5 min, and then transferred to a desiccator whereit was stored for later use in a sealed container.

image of Fig.�3

0

30000

60000

90000

120000

150000

180000

x10-6

kg

/kg

dry

fu

elx1

0-6 k

g/k

g d

ry f

uel

x10-6

kg

/kg

dry

fu

elx1

0-6 k

g/k

g d

ry f

uel

Fermented Sewage Sludge

Rest fraction Leached in HClLeached in NH4Ac Leached in H2OUntreated Fuel

0

3000

6000

9000

12000

15000Palm Kernel Cake


0

3000

6000

9000

12000

15000Distillers Dried Grains with Solubles


0

3000

6000

9000

12000

15000

Cl S P Si Al Mg Ca Na K Mn Fe Ti Zn




Rapeseed Cake


Fig. 4. Chemical fractionation results. In the graph of fermented sewage sludge (top left), the scale on the Y-axis is different to accommodate the higher ash concentration.

0

5

10

15

20

25

30

35

40

Cl

SO

3

P2O

5

SiO

2

Al2

O3

MgO

CaO

Na2

O

K2O

MnO

Fe2

O3

ZnO


ash

co

mp

osi

tio

n a

s o

xid

es[w

t-%

]as

h c

om

po

siti

on

as

oxi

des

[wt-

%]

ash

co

mp

osi

tio

n a

s o

xid

es[w

t-%

]as

h c

om

po

siti

on

as

oxi

des

[wt-

%]

Fe2O351 wt-%

0

5

10

15

20

25

30

35

40

Cl

SO

3

P2O

5

SiO

2

Al2

O3

MgO

CaO

Na2

O

K2O

MnO

Fe2

O3

ZnO

Palm Kernel Cake

0

5

10

15

20

25

30

35

40

Cl

SO

3

P2O

5

SiO

2

Al2

O3

MgO

CaO

Na2

O

K2O

MnO

Fe2

O3

ZnO

Dried Distillers Grains with Solubles

0

5

10

15

20

25

30

35

40

Cl

SO

3

P2O

5

SiO

2

Al2

O3

MgO

CaO

Na2

O

K2O

MnO

Fe2

O3

ZnO

Rapeseed Cake

Fig. 5. Composition of fuel standard ash prepared at 550 °C according to CEN/TS 15403, analyzed with SEM/EDX. Bar plots present the average of three samples and triangles showvalues for each sample.

39P. Piotrowska et al. / Fuel Processing Technology 105 (2013) 37–45

Table 1Summary of microsample sintering test results.

Temperature [°C] RC DDGS PKC FSS

550 0 0 0 0750 * * 0 0850 * **950 ** Molten * **1050 *** n/a ** ***

n/a–data not available.


2.2. Sintering tendency

The ash sintering tendency of the fuels was investigated using threedifferent tests: microsample sintering, pellet sintering, and standard ashsintering according to CEN/TS 15370. As discussed in Section 2.1, pre-pared standard ash was used in the pellet sintering and standard ash sin-tering tests, while the original fuels were used for microsample sinteringtest. The sintered ash particles were analyzed by means of SEM/EDX.

Themicrosample sintering test of fuel samples was investigated usinga thermogravimetric analyzer (TGA) and an optical microscope equippedwith a digital camera. The fuels were ground to a mean particle size lessthan 1.0×10−3 m to ensure the homogeneity of the samples. Approxi-mately 15.5±0.3×10−6 kg of the fuel samplewas placed in the hot reac-tor at a stabilized temperature of 550 °C, 750 °C, 850 °C, 950 °C, and1050 °C. The airflow was kept constant at 0.83×10−6 m3 /s. Three sam-ples were tested for each fuel at each temperature (except 1050 °C).The sample was removed after 25 min, and a picture was taken with adigital camera through the optical microscope at three different magnifi-cations: 1.6, 3.2, and 5.0. Based on the pictures, the ash sampleswere clas-sified according to the four stages of sintering introduced by Moilanen et

Mag = 30x WD = 13 mm

RC 1050°C

Mag = 30x WD = 13 mm400 µm


RC 950°C

3

43 5

6

RC 750°C

20 µm

20 µm

20 µm

Mag = 1000x WD

400 µm

400 µm

Mag = 1000x WD

Mag = 1000x WD

Fig. 6. SEM/EDX point analyses of rapeseed cake (RC) after the microsamp

al. [23] (Fig. 1). Unsintered ash residue that resembles the original fuelparticles is marked “0”. There are two classes of partly sintered ash parti-cles. The first, labelled with one asterisk, describes a sintered sample inwhich individual particles of the original unsintered sample can still beobserved, so that the ash sample forms a sort of cake. In the second, la-belled with a double asterisk, individual particles of the original samplecan no longer be distinguished. Completely sintered ash particles are la-belled with three asterisks.

The pellet sintering test is based on compression strength measure-ment tests described in detail by Hupa et al. [24] and Skrifvars [25]. Themethod comprises three main stages: pelletizing, heat treatment, andcrushing. The standard ash was pelletized into 8.0×10−3 m-high cylin-derswith a diameter of approximately 8.3×10−3 m. The pressure appliedto all thepelletswas the same (approx. 40×105 Pa); itwas determinedbyfinding theminimummeasurable pressure necessary to pelletize the ash.After preparation, the sampleswere subjected to heat treatment for 4 h atvarious temperatures in the range 350°–950 °C in a tube furnace with aconstant synthetic air flow of 3.3×10−6 m3/ s. Three pellets were pre-pared for each temperature. After heat treatment, the pellets were re-moved and cooled to room temperature in desiccators. Beforecompression strength measurements were performed, the size of eachpellet was measured to make it possible to calculate the compressionstrength per mm2. The compression strength was measured with a stan-dard crushing device.

Standard ash sintering test was performed according to the proceduredescribed in CEN/TS 15370. Ash pellets with a diameter equal to theirheight (3.0×10−3 m) were subjected to heat treatment at a controlledheating rate up to 1500 °C; the heating rate between 500 °C and1500 °Cwas 0.083 °C/s. Images of thepellet duringheatingwere recordedand evaluated automatically with an image analysis routine written in

0

10

20

30

40

50

60

70

80

90

100

P Si Al Mg Ca Na K Fe

spot 1 spot 2 spot 3 spot 4 spot5 spot6

0

10

20

30

40

50

60

70

80

90

100


ED

X a

nal

ysis

[m

ol-

%]

spot 1 spot 2 spot 3 spot 4spot5 spot 6 spot 7

0

10

20

30

40

50

60

70

80

90

100


ED

X a

nal

ysis

[m

ol-

%]

ED

X a

nal

ysis

[m

ol-

%]

spot 1 spot 2 spot3 spot 4 spot 5

1 2

54

12

= 13 mm

= 13 mm

= 13 mm

le sintering test. Results are shown on oxygen and carbon free basis.


DDGS 850°C

0

10

20

30

40

50

60

70

80

90

100


spot 1 spot 2 spot 3 spot 4spot 5 spot 6 spot 7

12

43

5

7

6

Mag = 30x WD = 13 mm400 µµm

DDGS 750°C

Mag = 500x WD = 13 mm

5

0

10

20

30

40

50

60

70

80

90

100


ED

X a

nal

ysis

[m

ol-

%]

ED

X a

nal

ysis

[m

ol-

%]

spot 1 spot 2 spot 3 spot 4 area 5

20 µµm

400 µµm 20 µµm Mag = 1000x WD = 13 mm

Fig. 7. SEM/EDX point analyses of dried distillers' grains with solubles (DDGS) after the microsample sintering test. Results are shown on oxygen and carbon free basis.


MATLAB, and they were used to identify the phases in the ash meltingprocess, as described in Fig. 2. The transitions were determined basedon the shrinkage of the pellet area. The initial shrinkage temperature cor-responds to a 5% shrinkage of the cross-sectional area. The initial defor-mation temperature, defined in the standard as a 15% change in theshape factor, could not be observed for the square-like cross sections, sothese are not included in the results (Section 3.2.). Three replicates wereperformed for each fuel.

3. Results and discussion

3.1. Fuel analyses

The proximate analyses performed for each fuel were averaged, and asummary of the results appears in Fig. 3. For all four fuels, the proportionof volatile matter was similar (approximately 80%), and so was the fixed

PKC 850°C

Mag = 30x WD = 13 mm Mag = 1000x WD

PKC 950°C

Mag = 30x WD = 13 mm400 µm 10 µm

40 µm

12

567

Mag = 2000x WD

Fig. 8. SEM/EDX point analyses of palm kernel cake (PKC) after the microsam

carbon content (approximately 20%) on a dry and ash-free basis. Howev-er, there is a significant difference between fermented sewage sludge andthe three other fuels in terms of ash content: that of FSS is much higher,nearly 50%of the dry fuel. The origin of these samples accounts for thedif-ference. While RC, DDGS, and PKC come from plant seeds, FSS originatesfrom humanwaste. Very good reproducibility was achieved for both bio-mass residues and biowaste (±2% for FSS and ±1% for the other fuels).

The chemical fractionation results are shown in Fig. 4. This analysisgives some indication of the constituents with which the ash-formingelements in the fuel are combined. For example, the ash-formingmatterof FSS occurs mainly in the form of iron phosphates, which are likelypresent because iron sulphate is used during wastewater treatmentprocesses for the precipitation of phosphorus. The presence of alumini-um silicates in the form of zeolites may also be indicated; the alumini-um in the hydrochloric acid soluble fraction most probably originatesfrom the dealumination of zeolites in the HCl solution.

= 13 mm

0

10

20

30

40

50

60

70

80

90

100


ED

X a

nal

ysis

[m

ol-

%]

spot 1 spot 2 spot 3 spot 4 spot 5 area 6

0

10

20

30

40

50

60

70

80

90

100


ED

X a

nal

ysis

[m

ol-

%]

spot 1 spot 2 spot 3 spot 4spot5 spot6 spot7

= 13 mm

ple sintering test. Results are shown on oxygen and carbon free basis.

0

10

20

30

40

50

60

70

80

90

100


ED

X a

nal

ysis

[m

ol-

%]

area 1 area 2 area 3 area 4

FSS 950°C

Mag = 30x WD = 13 mm400 µm

FSS 850°C

Mag = 30x WD = 13 mm400 µm

0

10

20

30

40

50

60

70

80

90

100


ED

X a

nal

ysis

[m

ol-

%]

area 1 area 2 area 3spot 4 spot 5 area 6

Mag = 500x WD = 13 mm20 µm

Mag = 500x WD = 13 mm20 µm

2

3

6

1

Fig. 9. SEM/EDX point and area analyses of fermented sewage sludge (FSS) after the microsample sintering test. Results are shown on oxygen and carbon free basis.


Compared to the ash-formingmatter of the other fuels, that of FSS isnot highly soluble. The ash-formingmatter of DDGS ismainly leached inwater, indicating the presence of highly soluble salts of metal phos-phates and chlorides. In addition, some mostly insoluble silicon isdetected. The origin of the siliconmight be explained by contaminationwith sand. Some silicon could also have accumulated in the barley grainduring plant growth, most probably in soluble form. The presence ofsulphur in the unleached residual fraction indicates that it is covalentlybound to organic matter. In PKC, insoluble aluminum silicates, probablycontaminants or residues from the production process, occur togetherwith highly soluble potassium, phosphorus, and chlorine. Low

048

12162024283236404448

0 200 400 600 800 1000

Co

mp

ress

ion

str

eng

th[N

/mm

2 ]

Co

mp

ress

ion

str

eng

th[N

/mm

2 ]

Temperature [°C]


048

12162024283236404448

0 200 400 600 800 1000

Temperature [°C]

Rapeseed Cake

0,00,51,01,52,02,53,03,54,04,55,0

0 200 400 600 800 1000

com

pre

ssio

n s

tren

gth

[N/m

m2 ]

Temperature [°C]


Fig. 10. Pellet sintering results: compression stre

concentrations of silicon are found in RC, inwhich the ash-formingmat-ter consists of alkali metals, phosphorus, sulphur, and chlorine; they ap-pear mainly in the fraction leached by water, indicating the presence ofsimple salts like alkali metal chlorides, sulphates, and phosphates. Asubstantial amount of sulphur was found in the residual fraction, sug-gesting that RC also contains covalently bound sulphur. Acid and am-monium acetate soluble phosphorus is assumed to occur in the formof phytic acid salts together with alkaline earth metals.

Depending on how it is combined with other constituents, ash-form-ing matter undergoes different transformations in combustion, such asoxidation (e.g., the oxidation of organic sulphur to SO2); chemical

048

12162024283236404448

Palm Kernel Cake

Co

mp

ress

ion

str

eng

th[N

/mm

2 ]

0 200 400 600 800 1000

Temperature [°C]

0 200 400 600 800 1000

Distillers Dried Grains with Solubles

Palm Kernel Cake

Co

mp

ress

ion

str

eng

th[N

/mm

2 ]

048

12162024283236404448

Temperature [°C]

0,00,51,01,52,02,53,03,54,04,55,0

com

pre

ssio

n s

tren

gth

[N/m

m2 ]

0 200 400 600 800 1000

Temperature [°C]

ngth vs. temperature during heat treatment.


reactions of, for example, reactive specieswith bedmaterial [7]; and ther-mal decomposition (e.g., the calcination of calcium carbonates or loss ofstructural water from minerals or crystals). Consequently, how an ele-ment is combined in the fuel typically differs from how it is combinedin the ash. Nonetheless, an understanding of the ash-forming matter infuels serves as a starting point for understanding its subsequent transfor-mations. Fig. 5 shows the composition of fuel ash prepared at 550 °C (de-scribed in Section 2.1). Despite the visible differences in the ashes, there isa common feature: the high phosphorus content. The FSS ash is dominat-ed by iron, probably in the form of oxide, while the phosphorus could bein the form of calcium phosphates. The DDGS ash has the highest potassi-um and silicon content.When potassium is available for reactionwith sil-ica, it is likely that potassium silicateswill form.However, the high affinityof phosphorus for potassium is expected to result in the formation of po-tassium phosphate [26]. The main element in both PKC ash and RC ash isphosphorus, most probably in the form of different phosphate salts,whereas in the fuels, it is organically bound. While PKC ash has high sili-con and aluminium content, RC ash has the highest concentration of sodi-um and chlorine (approximately 9 wt.% and 1 wt.%, respectively). The ashcomposition of the fuels probably determines their sintering tendency, asdiscussed in Section 3.2.

0

200

400

600

800

1000

1200

1400

RC

DD

GS

PK

C

FS

S

RC

DD

GS

PK

C

FS

S

RC

DD

GS

PK

C

FS

S

Shrinkage Hemisphere Flow

Tem

per

atu

re [

°C]

Fig. 11. Standard ash sintering test results.

3.2. Sintering tendency

For each fuel, the sintering tendency as determined by the micro-sample tests is shown in Table 1. The images used for the evaluationare included in an appendix. The first indication of FSS ash sinteringwas observed at 850 °C. At 950 °C, particles of the original unsinteredsample could not be clearly observed, while at 1050 °C, the samplewas completely sintered (labelled with 3 asterisks).

Based on Table 1, PKC exhibits the lowest sintering tendency of allthe fuels considered in this work, whereas DDGS exhibits the highestsintering tendency. Already at 950 °C, the molten sample cannot be re-moved from the crucible. For RC, cake-like sintering could first be ob-served at 750 °C. However, further progression of the sintering couldbe observed at 950 °C. Since the tests were performed withΔT=100 °C, it is not possible to determine whether the change oc-curred at lower temperatures. Since the results are based onmicroscop-ic visual inspection, they might be somewhat subjective.

Figs. 6–9 present the SEM/EDX analyses of ash samples after themicrosample sintering tests. It can be observed that the sintered FSS ashis dominated by iron oxides, whereas no large sintered particles can beseen for the palm kernel cake in the temperature range investigated.The SEM/EDX analyses of PKC samples indicate a greater presence of alu-minum silicate particles and silicate particles than of alkali metal phos-phate particles. The two biomass residues with the highest sinteringtendency, RC and DDGS, have the highest phosphorus concentration. At950 °C, DDGS already formed a molten amorphous phase. This sampleconsists mainly of potassium phosphate salts, with hardly any earth alka-line. The concentration of (Ca+Mg) in RC is much higher than in DDGS,so the further progression of sintering could be observed at 950 °C. TheSEM/EDX analyses indicate that the sintering tendency of these twofuels is driven principally by the presence of alkali metal phosphate parti-cles, as opposed to earth alkaline phosphate particles. Themelting behav-iour of alkali phosphates is different from that of earth alkali phosphates,which is why the shift between the two types of phosphates is of greatimportance during the combustion of phosphorus-rich fuels. Themeltingbehaviour of the twokinds of phosphatewas discussedby Lindströmet al.[27], who studied the slagging characteristics of phosphorus fuels withand without limestone addition. She concluded that lime, as a source ofcalcium, contributes to the formation of high-temperature melting calci-um potassium phosphates. The formation of high melting phosphateswas also investigated by Barišić et al. [28], who reached a similar conclu-sion. Steenari et al. [13] improved the sintering tendency of rapeseed cakewith the addition of kaolin as well as of limestone. The high K/Ca ratio for

phosphorus-rich fuelswas associatedwith low temperature ash sinteringand melting.

The results of the pellet sintering tests are shown in Fig. 10. For eachfuel, the dependence of the compression strength on the temperature isplotted. The curves have different shapes, indicating the differences inthe sinteringmechanisms. The leftmost points at 25 °C refer to standardash pelletized and crushed with no prior heat treatment. This is the ref-erence compression strength, corresponding to the unsintered area onthe graphs. To determine the initial sintering temperature, the lastunsintered temperature point was chosen. Changes in the compressionstrength below 1 N/mm2 are considered to be negligible and are as-sumed to have no significant influence on the sintering of the ash pel-lets, and thus on boiler operation. The DDGS sintering curve starts toincrease slowly after 450 °C, with a sharp rise at 650 °C. The sinteringcurve for RC shows an increase after 400 °C, and after 600 °C, there isa dramatic increase. For both FSS and PKC, no heavy sintering was ob-served. The sintering of FSS begins after 900 °C. The small changes inthe shape of the curve before 900 °C could be the result of ash transfor-mation, which probably does not have a major impact on the sinteringof ash during boiler operation. Interestingly, PKC exhibits hardly anysintering tendency, as shown by the low peak at 900 °C.

Fig. 11 presents the standard ash sintering test results. The character-istic temperatures are defined in termsof the shapes in Fig. 2 (Section2.2).Very good reproducibilitywas obtained for the triple replicates. The initialshrinkage temperature is taken to be an indication of the onset of sinter-ing of the ash since the shrinkage is the result of a decrease in porosity,which could be caused by partial melting, viscous flow, or chemical reac-tions. The lowest initial shrinkage temperature, that recorded for RC, wasapproximately 650 °C. For DDGS, it was a little higher, approximately720 °C, and for PKC and FSS, it was much higher, approximately 900 °Cand 920 °C, respectively. The trend exhibited by the four fuels is in agree-mentwith that for the pellet sintering tests, with the initial sintering tem-perature increasing in the following order: RCbDDGSbPKCbFSS.However, the pellet sintering tests suggest that the onset of sinteringtakes place at lower temperatures for RC andDDGS. The sintering temper-atures for FSS and PKC are in good agreement with those obtained withpellet sintering. Basedon themicrosample sintering tests, the onset of sin-tering for PKC and FSS is expected to occur between 850 °C and 950 °C,which is in good agreement with the results from the other types of sin-tering tests performed on standard 550 °C ash. The initial sintering tem-perature for RC and DDGS (Table 1) is 750 °C; however, themicrosample sintering tests were not performed between 550 °C and750 °C, so it is not possible to determine whether sintering occurred atlower temperatures. Nevertheless, the results for DDGS are inconsistent:the hemisphere and flow temperatures from the ash sintering tests aresignificantly higher than those obtained with the microsample sinteringtests. The two fuels with the lowest sintering temperatures, DDGS andRC, have the highest concentrations of water-soluble alkali metals andphosphorus. The water-soluble fraction is considered to be the reactivepart of the fuel [7]. The composition of palm kernel cake is similar to


that of RC. However, their sintering tendencies differ significantly, whichcould be attributed to the smaller ratio of water-soluble alkali metals toearth alkaline metals in PKC; another possibility is the influence of theAl-Si in the fuel sample. As pointed out by Steenari et al. [13], phosphatespecies can interact with the silicate phases, but the interaction needs tobe further clarified. The chemistry of FSS is dominated by iron oxidesand, as concluded byWang et al. [16], the appearance and disappearanceof Fe2O3 in ash was the leading factor affecting the sintering of sewagesludge ash when Fe2(SO4)3) was used at a wastewater treatment plantto precipitate phosphorus.

4. Conclusions

Thiswork analyzed four residues frombiofuel production: fermentedsewage sludge (FSS); distillers' dried grains with solubles (DDGS); rape-seed cake (RC); and palm kernel cake (PKC). Chemical fractionationrevealed the diversity of the fuels. The major element in all of them isphosphorus, but the combinations in which the ash-forming matteroccurs are different for each fuel. The inorganic matter in FSS is domi-nated by iron phosphates and zeolites, while in DDGS, it is dominatedby potassium phosphates and silica. The composition of RC is complex,most likely dominated by alkali chlorides, alkali phosphates, and phyticacid salts. In addition to phosphates, PKC contains aluminum silicatesand iron. Differences in the ash sintering tendency arise from differencesin the ash-forming matter content and the way it is combined. Thelowest initial sintering temperatures were observed for RC and DDGS,

FermentedSewageSludge

T:550°C;sintering:0 T:850°C;sintering:

RapeseedCake

T:550°C; sintering:0 T:750°C; sintering:*

DistillersDried

Grains withSolubles


PalmKernel


*

and for the three sintering methods, it is below 800 °C, which suggeststhat these fuels should not be combusted alone, without precautionsagainst ash-related problems, no matter what the combustion techno-logy. The low initial sintering temperatures can probably be attributedto the very high concentration of alkali metal phosphates in the ash. ThePKC and FSS samples do not exhibit significant sintering below 900 °C,which could be interpreted to mean that when the combustion tempera-ture is below 900 °C, no significant ash-related problems will occur.

Acknowledgements

The authors gratefully acknowledge the financial support provid-ed by the Graduate School in Chemical Engineering (Finland) andby ChemCom, a project funded mainly through TEKES and coordinat-ed by the Process Chemistry Centre at Åbo Akademi University, Turku,Finland. Samples were provided within the framework of the SAFECproject, a joint effort of FosterWheeler Oy and TEKES, and by EmmelevA/S, Ab Stormossen Oy (Ms. Johanna Penttinen-Källroos), E. On,and Abengoa, to whom the authors would also like to express theirgratitude. Dr. Bengt-Johan Skrifvars is thanked for sharing his exper-tise in compression strength tests. A program in Matlab, crucial forthe evaluation of hot stage microscope data, was generously providedby Johan Fagerlund at the Laboratory of Thermal and Flow Engineer-ing. Special thanks are due to Peter Backman, Piia Leppäsalo, JaanaPaananen, and Linus Silvander for their assistance with various as-pects of the experimental work.

Appendix A

Micro-sample sintering test results (crucible diameter is 4 mm).

T:1050°C;sintering:***T: 950°C;sintering: **

T:950°C; sintering:** T:1050°C;sintering:***

T:850°C; sintering:** T:950°C;molten

T:950°C; sintering:* T:1050°C; sintering:**

Cake


References

[1] Directive 2009/28/EC of the European Parliament and of the Council of 23 April 2009on the promotion of the use of energy from renewable sources and amendingandsubsequently repealing directives 2001/77/EC and 2003/30/EC Official Journal ofthe European Union 2009.

[2] http://www.eea.europa.eu/data-and-maps/figures/final-energy- consumption-by-sector.

[3] http://www.eea.europa.eu/data-and-maps/figures/share-of-biofuels-in-transport-fuels-1.

[4] B.M. Jenkins, L.L. Baxter, T.R. Miles Jr., T.R. Miles, Combustion properties of biomass,Fuel Processing Technology 54 (1998) 17–46.

[5] L.L. Baxter, T.R. Miles, T.R. Miles Jr., B.M. Jenkins, T. Milne, D. Dayton, R.W. Bryers,L.L. Oden, The behavior of inorganic material in biomass-fired power boilers: fieldand laboratory experiences, Fuel Processing Technology 54 (1998) 47–78.

[6] J. Werther, M. Saenger, E.-U. Hartge, T. Ogada, Z. Siagi, Combustion of agriculturalresidues, Progress in Energy and Combustion Science 26 (2000) 1–27.

[7] M. Zevenhoven-Onderwater, M. Öhman, B.-J. Skrifvars, R. Backman, A. Nordin, M.Hupa, Bed agglomeration characteristics of wood-derived fuels in FBC, Energy &Fuels 20 (2006) 818–824.

[8] M. Öhman, A. Nordin, B.-J. Skrifvars, R. Backman, M. Hupa, Bed agglomerationcharacteristics during fluidized bed combustion of biomass fuels, Energy & Fuels14 (2000) 169–178.

[9] K.O. Davidsson, L.-E. Åmand, B.-M. Steenari, A.-L. Elled, D. Eskilsson, B. Leckner, Coun-termeasures against alkali-related problems during combustion of biomass in a circu-lating fluidized bed boiler Chem, Engineering Sciences 63 (21) (2008) 5314–5329.

[10] F.J. Frandsen, Utilizing biomass and waste for power production—a decade of con-tributing to the understanding, interpretation and analysis of deposits and corrosionproducts, Fuel 84 (2005) 1277–1294.

[11] K.O. Davidsson, L.-E. Åmand, B. Leckner, B. Kovacevik, M. Svane, M. Hagström, J.B.C.Pettersson, J. Pettersson, H. Asteman, J.-E. Svensson, L.-G. Johansson, Potassium,chlorine, and sulfur in ash, particles, deposits, and corrosion during wood combus-tion in a circulating fluidized-bed boiler, Energy & Fuels 21 (2007) 71–81.

[12] J. Giuntoli, W. de Jong, A.H.M. Verkooijen, P. Piotrowska, M. Zevenhoven, M. Hupa,Combustion characteristics of biomass residues and biowastes: fate of fuel nitrogen,Energy & Fuels 24 (10) (2010) 5309–5319.

[13] B.-M. Steenari, A. Lundberg, H. Pettersson, M. Wilewska-Bien, D. Andersson, Investi-gation of Ash Sintering during Combustion of Agricultural Residues and the Effect ofAdditives, Energy & Fuels 23 (2009) 5655–5662.

[14] A.A. Tortosa Masiá, B.J.P. Buhre, R.P. Gupta, T.F. Wall, Characterising ash of biomassand waste, Fuel Processing Technology 88 (2007) 1071–1081.

[15] M.J. Fernandez Llorentea, J.E. Carrasco Garcıa, Comparing methods for predictingthe sintering of biomass ash in combustion, Fuel 84 (2005) 1893–1900.

[16] L. Wang, G. Skjevrak, J.E. Hustad, M. Grønli, Sintering Characteristics and MineralTransformation of Sewage Sludge Ashes, Proceedings of Impacts of Fuel Qualityon Power Production and Environment, Saariselkä (Finland), 2010.

[17] M. Bartels,W. Lin, J. Nijenhuis, F. Kapteijn, J.R. van Ommen, Agglomeration in fluidizedbeds at high temperatures:mechanisms, detection andprevention, Progress in Energyand Combustion Science 34 (2008) 633–666.

[18] M.C. Mayoral, M.T. Izquierdo, J.M. Andrés, B. Rubio, Different approaches to proximateanalysis by thermogravimetry analysis, Thermochimica Acta 370 (2001) 91–97.

[19] M. Zevenhoven, P. Yrjas, R. Backman, B.-J. Skrifvars, M. Hupa, The Åbo Akademidatabase-fuel characterization, Proceedings of the 18th International Conference onFluidized Bed Combustion, Toronto (Ontario, Canada), 2005, pp. FBC2005–FBC78093.

[20] B.-J. Skrifvars, J.-P. Blomquist,M.Hupa, R. Backman, Predicting theash behavior duringbiomass combustion in FBC conditions by combining advanced fuel analysis withthermodynamic multicomponent equilibrium calculations, Proceedings of the 15thAnnual International Pittsburgh Coal Conference, Pittsburgh (PA, USA), 1998.

[21] S.A. Benson, P.L. Holm, Comparison of inorganic constituents in three low-rank coals,Industrial & engineering chemistry product research and development 24 (1985)145–149.

[22] J. Werkelin, Ash forming elements and their chemical forms in woody biomassfuels, Academic Dissertation, Åbo Akademi University, Turku, Finland, 2008,ISBN 978-952-12-2125-5.

[23] A. Moilanen, E. Kurkela, J. Laatikainen-Luntama, Ash behaviour in biomassfluidised-bed gasification, in: Gupta, et al., (Eds.), Impact of mineral impuritiesin solid fuel combustion, Kulwer Academic/Plenum Publishers, New York, 1999,pp. 555–567.

[24] M. Hupa, B.-J. Skrifvars, A. Moilanen, Measuring the sintering tendency of ash by alaboratory method, Journal of the Institute of Energy (1989) 131–137.

[25] B.-J. Skrifvars, Sintering Tendency of Different Fuel Ashes in Combustion and Gasifica-tion Conditions, Academic Dissertation, Åbo Akademi, Turku, Finland, 1994, Report94–6.

[26] D. Boström, G. Eriksson, C. Boman, M. Öhman, Ash transformations in fluidized-bedcombustion of rapeseed meal, Energy & Fuels 23 (2009) 2700–2706.

[27] E. Lindström, M. Sandström, D. Boström, M. Öhman, Slagging Characteristics duringCombustion of Cereal Grains Rich in Phosphorus, Energy & Fuels 21 (2007) 710–717.

[28] V. Barišić, L.-E. Åmand, E. Coda Zabetta, The role of limestone preventing the agglom-eration and slagging during CFB combustion of high-phosphorous fuels, Presented atWorld Bioenergy 2008, Jönköping (Sweden), 2008, May, 27–29.

http://www.eea.europa.eu/data-and-maps/figures/final-energy

http://www.eea.europa.eu/data-and-maps/figures/share-of-biofuels-in-transport-fuels-1

http://www.eea.europa.eu/data-and-maps/figures/share-of-biofuels-in-transport-fuels-1

fuel processing technology - startsidausers.abo.fi/mzevenho/portfolj/publikationer/refereed...

Documents