LIGNITE DRYING TECHNOLOGIES FOR POWER GENERATION: PHYSICOCHEMICAL PROPERTIES, PYROLYSIS AND COMBUSTION OF DRIED LIGNITE PRODUCTS A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy By George Favas Faculty of Engineering and Industrial Science Swinburne University of Technology 2008
383
Embed
Lignite drying technologies for power generation physiochemical … · 2017. 1. 3. · LIGNITE DRYING TECHNOLOGIES FOR POWER GENERATION: PHYSICOCHEMICAL PROPERTIES, PYROLYSIS AND
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
LIGNITE DRYING TECHNOLOGIES
FOR POWER GENERATION:
PHYSICOCHEMICAL PROPERTIES,
PYROLYSIS AND COMBUSTION
OF DRIED LIGNITE PRODUCTS
A thesis submitted in fulfilment of the
requirements for the degree of
Doctor of Philosophy
By
George Favas
Faculty of Engineering and Industrial Science
Swinburne University of Technology
2008
Abstract ii
ABSTRACT
Three technologies have been under serious consideration for reducing the moisture
content of Latrobe Valley lignites – namely, hydrothermal dewatering (HTD),
mechanical thermal expression (MTE) and steam drying (SD). This thesis compares
the relative effectiveness of these technologies in terms of water retention in the
product, solids recovery from the process and organic carbon in the by-product
water.
For the same Loy Yang lignite, containing 59% water (on a wet basis), it is shown
that the temperature at which 50% of this water is removed is, for MTE: 125°C, for
SD: 210°C and for HTD: 350°C.
With MTE, the processing temperature required to achieve high moisture reductions
was significantly reduced by simultaneous application of mechanical pressure. For
Loy Yang lignite the optimum applied pressure was identified as approximately
5.1MPa. Further increases had relatively little effect on the moisture content of the
product pellet.
The relatively high temperatures required to achieve 50% water removal by HTD
also leads to poor solids recovery (86%) due to decarboxylation and devolatilisation
processes that concomitantly occur under these conditions. Poor solids recovery is
also observed when SD is carried out at high temperatures (>250°C).
Abstract iii
Much of the sodium removed from drying was in the form of NaCl. In MTE,
processing temperature was more effective in removing sodium than applied pressure
(within the experimental parameters tested). In contrast, SD processing temperatures
above 280°C had no effect on the proportion of sodium leached out from the lignite.
The slurrying of the lignite with water in HTD facilitated removal of sodium,
magnesium, calcium and chlorine during the process compared to MTE and SD. This
considerably higher removal of inorganic material, particularly sodium, is an
advantage in that it will reduce fouling and slagging propensities during combustion
of the product.
Rapid pyrolysis of the HTD product gave significantly lower yields of methane,
ethene, carbon monoxide and carbon dioxide when compared to corresponding yields
of the raw lignite, MTE and SD products. The lower yields for the HTD product
were attributed to volatilisation during hydrothermal dewatering and also to the
reduction of catalytic reforming cations. The significant volatile yield loss during
hydrothermal dewatering could be disadvantageous in some industrial processes,
which convert the carbon matter to lower molecular weight fractions (eg gasification,
liquefaction).
The combustion reactivity of MTE, HTD and SD products did not demonstrate
significant differences from the raw lignite. The combustion reactivity of the
thermally dried products decreased with increasing severity of the conditions.
However these changes were quite minimal suggesting that the conventional boiler
Abstract iv
systems currently in operation in the Latrobe Valley would be more than adequate in
combusting thermally treated products from any of the three drying processes.
The effectiveness of catalytic inorganic species clearly outweighed pore volume
effects for affecting the combustion reactivity of the lignites. The chemical form and
concentration of the inorganic component in the hydrocarbon macromolecular matrix
can influence the rate of combustion at 400°C. The removal of inorganic components
from the coal by washing or by HTD or MTE has a much larger impact on reducing
the combustion reactivity and reducing the peak temperature of the sample when
compared to significant reductions in porosity resulting from thermal drying.
Additional combustion experiments on a diverse suite of well-characterised lignites
sourced from the Latrobe Valley confirmed that catalytic inorganic species clearly
outweighed other physico-chemical effects for affecting the combustion reactivity of
the lignites. The combined catalytic effects of iron, magnesium and calcium
accounted for more than 95% of the variation in the combustion reactivity of the raw
coals. In addition, the combustion reactivity of Latrobe Valley lignite samples was
strongly correlated to the sum of the univalent charges of the AAEM species plus the
univalent charge of the acid-extractable iron.
Finally, the surface area and pore volumes of carbons/chars have negligible effect on
the combustion reactivity and the inorganic components in the carbon
macromolecular structure are the major influences in the combustion rate of the
sample at 400°C.
Acknowledgements v
Acknowledgements
The first acknowledgement in every thesis should always be given to the supervisor.
In this case I would like to thank Dr Francois Malherbe and Professor Ian Harding
for the opportunity to undertake this degree.
I would also like to thank the CRC for Clean Power from Lignite, the executives Dr
David Brockway, Mr Malcolm MacIntosh and the late Dr Peter Jackson, for
supplying the raw lignite samples and permission to operate their MTE batch cell.
Also, I would like to thank the School of Chemistry at Monash University and Dr
Alan Chaffee for the use of their analytical equipment. Also, I would like to thank Dr
Chun-Zhu Li for his support at the commencement of this research project, and
Professor Roy Jackson and Dr Marc Marshall for their encouragement and guidance
throughout the years.
A very special acknowledgment goes out to my wife, Mrs Marija Favas. Her
persistence, inspiration and ability to whip up a meal during all hours of the day have
been the sole source of time management for this project. I would also like to thank
Marija for her persistence and understanding, in particular the late nights and early
mornings during the write up of this thesis.
Acknowledgements vi
Last but surely not least, to my twin boys, Ivan and Christian, who were born
December 2007 and have already provided much joy and happiness to my family.
Declaration vii
DECLARATION
To the best of my knowledge this thesis contains no material which has been
accepted for the award of any other degree or diploma in any university and
contains no material previously published or written by another person except
where due reference is made. Also, where the work is based on joint
research or publications, discloses the relative contributions of respective
Actual mass (g) of ash, acid extractable inorganics, chlorine and total sulphur in raw and processed lignite. ......................................................................................................... 277
APPENDIX I .................................................................................................................... 278
Molar values of acid extractable inorganics, chlorine and total sulphur in raw and processed lignite................................................................................................................. 278
For Yallourn and Morwell samples, the raw lignite was passed through 125μm and
90μm sieves (without milling) and the fraction 90<μm<125 was collected (mean
particle size ~100μm) for combustion reactivity tests.
2.5 Water washing and acid washing of lignites The proportion of water soluble salts (mainly free ionic metals) and water insoluble
ions (metals bound to the lignite as phenolates and carboxylates) were determined by
comparing analyses of the original lignite and the same lignite washed with distilled
water. For lignite water washing, lignite [100g (wb)] and distilled water (500mL)
were slurried and stirred in a conical flask for 24h and then filtered. The lignite was
Chapter 2 Experimental 23
again slurried with 500mL distilled water, stirred for 24h and filtered. Washing was
repeated 3 times and the lignite then stored in airtight containers prior to analysis.
For acid washing, the same procedure as above was performed however the washing
was carried out with 0.1M HCl solution. The lignite slurry was washed and filtered 3
times with HCl solution followed by an additional 3 washes with distilled water to
remove residual acid solution. The water washed and acid washed samples were then
sieved to 90<μm<125, stored in airtight containers prior to combustion reactivity
tests.
2.6 Autoclaves Three batch autoclave systems were used in this study:
1. For hydrothermal dewatering experiments, a 70mL 316 SS autoclave heated in a
stationary position in a fluidised sandbath.
2. For steam drying experiments, a 500mL 316 SS autoclave heated in a stationary
position.
3. For mechanical thermal expression, a batch 300mL 316 SS autoclave heated in a
stationary position using a heating jacket.
Chapter 2 Experimental 24
2.7 Hydrothermal dewatering A schematic diagram of a 70mL autoclave is shown in Figure 2.1. 40.00g (weighed
to an accuracy of 0.01g) of 1 part lignite : 3 parts water (1:3 lignite:water ratio) was
loaded into the autoclave. The autoclave was sealed and leak tested with hydrogen
(~8MPa) using a thermal conductivity leak detector “Leak Seeker 96” and the
autoclave was then evacuated and kept under vacuum for 5min so to remove any
trapped gases. The autoclave was mounted onto a stainless steel cradle and lowered
into the sandbath, the temperature of which had been preset to the values summarised
in Table 2.2, varying with the desired autoclave temperature. The autoclave
temperature was monitored by a thermocouple placed inside the thermocouple well
(thermowell) of the autoclave.
Table 2.2 Set-up temperatures for 70mL autoclave HTD reactions
Autoclave
temperature
(°C)
Sandbath preset
temperature
(°C)
Equilibration
Point *
(°C)
180 220 199
200 240 218
230 270 248
250 300 281
280 340 306
300 360 325
320 390 354
350 430 378
* Equilibration point is the temperature of the sandbath which gives the desired temperature inside the autoclave
Chapter 2 Experimental
25
The air pressure used to agitate the sandbath was set at 250kPa and after the
autoclave had been lowered, the sandbath temperature setting was then lowered to
the equilibration point given in Table 2.2.
The temperature of the autoclaves and the temperature of the sandbath were
monitored using a Shimaden SR41 and Eurotherm controller, respectively. The
desired autoclave temperature (Table 2.2) was reached within 10min with a
maximum overshoot of 1°C and then maintained for an additional 10min at the
processing temperature. After the 10min holding time, the reactor was quenched in
cold water and further cooled in a stream of compressed air until the autoclave
temperature fell to 10°C above the room temperature (see Section 2.8 for workup
procedure).
2.8 HTD and SD work-up procedure After cooling, a stream of compressed air was used to clear all sand from around the
sealing surfaces of the reactor. The gas in the reactor was then slowly vented and the
solid/liquid products were thoroughly scraped out and carefully collected by washing
with distilled water. For experiments involving Total Organic Carbon (TOC) analysis
of the product water, the water from the reaction was filtered with no further dilution
and immediately collected and sealed in an airtight container, before more water was
added to wash the entire solid product out of the autoclave. The solid products were
collected and dried in an oven at 105°C for 3h under a stream of nitrogen gas.
26
1.25”
4.50”
0.75”
1.00”
1.75”
0.625”
1.50”
1.437”
Angle 60O
0.060”
0.040”1.437”
1.850”
1.696”
0.375”
1.00”
0.125”
0.0625” Hole
0.406” 0.427”
1.00”
1.850”
1.00”0.1875”
Pressure Ring
Top View
Side ViewAutoclave Head
Autoclave Body 2.625”
0.50”
1.00”
2.125”
1.4375” 2.625”
Autoclave Head
Thermowell and ClosureSide View
Top View
Figure 2.1 Schematic diagram of the 70mL autoclave
Chapter 2 Experimental
Chapter 2 Experimental 27
2.9 Mechanical thermal expression MTE experiments were conducted using a batch MTE unit (Figure 2.2)
Figure 2.2 Schematic diagram of the MTE batch unit
Raw lignite (~600μm particle size, 100g wb) was placed inside a compression-
permeability cell consisting of a piston and die assembly with a 50mm internal bore
diameter (Figure 2.3).
Figure 2.3 Components of the MTE batch cell
Chapter 2 Experimental 28
A new flexible O-ring was used for each experiment so to ensure that no steam
escaped the cell during each test. The cell was filled with a known weight of distilled
water (~100g) to expel air from the lignite sample, sealed and heated up to the
desired temperature (125 to 250°C). Once the desired temperature was reached, a
computer controlled Instron Universal Testing Apparatus (Model 5569) was used to
automatically increase the force on the MTE cell at a compression rate of
50mm min-1 until the final pressing pressure was reached (typically 2.5MPa, 5.1MPa,
12.7MPa or 25.0MPa). The pressing pressure was maintained for 20min before
cooling down the sample and collecting the pressed product. Water expelled from the
sample during the experiment was allowed to pass through porous brass sintered
plates (Figure 2.2) to a lower collection chamber of the cell and also to the hollow
piston chamber.
2.10 MTE workup procedure On completion of the MTE run, the applied pressure on the MTE cell was released
and the cell dismantled. The MTE pellet was collected, weighed immediately and
stored inside two air tight bags.
The MTE product water was then collected and stored in a separate bottle. The entire
system was then thoroughly rinsed with distilled water and the rinsing water also
collected in a separate bottle. Both the product water and rinsing water were then
filtered through a glass fibre filter and stored below 5°C for subsequent use. Analysis
of the product water and rinse water were conducted separately and the resulting
values combined to give a total concentration.
Chapter 2 Experimental 29
2.11 Steam drying experiments A container with a 75μm mesh top and bottom was charged with 25.00g wb
(weighed to an accuracy of 0.01g) of raw lignite and then attached to the
thermocouple rod of a 500mL autoclave. Distilled water (50.00g, weighed to an
accuracy of 0.01g) was poured into the autoclave and the thermocouple with the
container was then lowered to a position above the water level (Figure 2.4). The
autoclave was sealed, leak tested and then heated up to the desired temperature (130
to 350°C) which was then held for 30min at the processing temperature (Figure 2.5).
After the 30min holding time had elapsed, the reactor was quenched in cold water
and further cooled in a stream of compressed air until the autoclave temperature fell
to 10°C above room temperature. The container inside the autoclave was removed,
weighed and the solid lignite products dried in an oven at 105°C for 3h under a
stream of nitrogen gas.
Chapter 2 Experimental 30
Figure 2.4 Schematic of the 500mL batch autoclave used for steam drying lignite
Chapter 2 Experimental 31
0
50
100
150
200
250
300
350
0 5 10 15 20 25 30 35 40 45 50
Reaction time (min)
Aut
ocla
ve te
mpe
ratu
re (°
C)
350°C
330°C
320°C
310°C
300°C
280°C
250°C
230°C
200°C
180°C
150°C
130°C
Figure 2.5 Temperature heat up profile for the steam drying experiments.
Table 2.3 Set-up temperatures for 500mL autoclave SD reactions Autoclave
temperature
(°C)
Sandbath preset
temperature
(°C)
Equilibration
point
(°C)
130 170 145
150 200 170
180 220 200
200 240 220
230 310 250
250 350 280
280 380 305
300 415 330
310 435 340
320 450 345
330 470 355
350 500 380
Note: Compressed air bubbling through the fluidized sandbath was often adjusted during the experiment in order to achieve a 15min heat up time and to prevent overshoot of the autoclave temperature.
Chapter 2 Experimental 32
2.12 Pore size distribution analysis Pore size distributions on the lignite samples were performed on a Micromeritics
AutoPore porosimeter. Approximately 0.2g of pre-dried raw or dried products
(~0.2g) was loaded into the penetrometer and the mass recorded on a balance (four
figure accuracy). Low pressure and high-pressure mercury intrusions were measured
in the range of 0.76 to 29.99psia (5kPa to 207kPa) and 29.99 to 60,000psia
(0.2 to 414MPa), respectively. Each pressure step was allowed to equilibrate for
15sec before the intrusion measurement was recorded. On completion of the high-
pressure analysis, the mercury was emptied into a plastic polyethylene bottle and the
penetrometer cleaned using a 50:50 mixture of methylethylketone and toluene.
2.13 Determination of lignite moisture content The moisture contents of the raw lignite and the thermally treated products were
performed in duplicate and were determined similarly to the Australian Standard
Method (AS2434.1) [110]. For HTD and SD products, a representative sample
weighing approximately 5g was weighed and dried under nitrogen at 105°C for 3h.
For MTE products, the solid pellet (approximately 40g db) was also dried at 105°C
under nitrogen but for 6h. The dried products were then placed inside of desiccator,
sealed, and the air evacuated to prevent moisture reabsorption from the air. The
samples were allowed to cool down inside the desiccator for 15min before being
weighed to determine their respective moisture contents.
Chapter 2 Experimental 33
2.14 Wet and dry pellet density measurements Wet and dry density measurements of the MTE pellet were conducted using a
calliper and calculated using the equations 2.1 to 2.4 given below:
31.0mm (b)
50.0mm (a)
42.0mm (c)
Y
Figure 2.6 Schematic of a wet MTE pellet
YY
XX
x 21 (%) shrinkage Vertical
wet
dry
wet
dry⎟⎟⎠
⎞⎜⎜⎝
⎛+= Equation 2.1
100 x aa
(%) shrinkage Horizontalwet
dry= Equation 2.2
drywet pellet of Mass -pellet of Mass pellet in water of mass Total = Equation 2.3
x100pellet wet of volumeTotal
pelletin water of volumeTotal(%)porosity Wet = Equation 2.4
Chapter 2 Experimental 34
2.15 Slow and rapid pyrolysis of raw and thermally treated products
Slow and fast pyrolysis experiments of the raw lignite and thermally treated products
were carried out using a quartz fluidised-bed/fixed-bed reactor (Figure 2.8) heated
with an external tubular furnace. The reaction system consisted of a particle feeder
and a quartz reactor system. Rapid heating of the raw lignite and the thermally
treated products was achieved by injecting a stream of particles directly into a
heated, fluidised bed of zircon sand. Details of the particle feeder, quartz reactor and
the experimental protocol are described below.
2.15.1 Pre-treatment of raw and thermally dried products Prior to a pyrolysis experiment, the raw lignite or thermally treated products were
vacuum dried overnight at room temperature. The moisture contents of the vacuum
dried lignite particles were determined in duplicate before being placed into the
particle feeder. Tests involving passing argon into the particle feeder for 20min gave
no additional loss in moisture from the vacuum dried particles.
2.15.2 Particle feeder A schematic diagram of the particle feeder system is shown in Figure 2.7. The design
of the particle feeder was based on similar systems published elsewhere [111, 112]. The
particle feeder consisted of three tubes; an outer 1/2” glass tube, a 3/8” stainless steel
tube (gas stream tube) which was positioned inside the 1/2”glass tube and a 1/4”
The glass tube contained a rubber seal and a Viton-A® O-ring inside a male and
female nut which provided a gentle seal against any leakage of the argon gas stream
and also allowed the movement of the glass tube along the outer of the 3/8” stainless
steel tube. A stepper motor, controlled by a Thurlby Thandar Instruments TG215
Chapter 2 Experimental 35
2MHz function generator, moved the glass tube upwards which in turn, brought the
particle bed closer to the ends of the stainless steel tubes. The passing of argon
through the ends of the stainless steel tubes created a dilute particle dispersion above
the particle bed. Part of this particle dispersion was blown into the 1/4” tube and fed
into the quartz reactor. The rate of the particles being blown into the quartz reactor
was controlled by the speed of a stepper motor and the flow of argon passing through
the feeder. For all of the fast heating pyrolysis experiments, the output frequency of
the function generator was set at 4Hz and the flow rate of the argon carrier gas was
maintained at 1.0L min-1. Under these conditions, the stepper motor moved the glass
tube upwards at 3mm min-1 and the particle feed rate into the quartz reactor was
~100mg min-1. A vibrator attached to the bottom of the glass tube holder also
facilitated the agitation of the reservoir particle bed and enhanced the dispersion of
particles. In addition, the vibrator minimised channelling, prevented particle
deposition on the glass tube walls and prevented tube blockages on the inlet of the
1/4” carrier tube.
Chapter 2 Experimental 36
Lignite or thermally treated particles
” SS tubing1/4
” SS tubing1/4
3/8” Union Tee
Lignite particlesentrained in gas
(to quartz reactor)
” SS tubing3/8
O-ring
Rubber seal
Gas in
Female nut
Glass tube” 1/2
Male nut
VibratorScrew connectedto stepper motor
Gla
s s tu
be m
ove m
ent
(Gas stream tube; tube position fixed)
(Particle carrier tube; tube position fixed)
Figure 2.7 Schematic of the lignite particle feeder
Chapter 2 Experimental 37
2.15.3 Quartz reactor design Quartz fluidised bed reactors have been used by a number of groups in studying the
pyrolysis and gasification of lignites [113-116]. The design of the quartz reactor system
for this study was based on a similar system previously used in coal pyrolysis [100].
Quartz instead of high temperature grade stainless steel was selected as the reactor
material because the impurities in stainless steel (such as Cr, Mn, Ni and S) could
potentially catalyse [117, 118] and interfere with pyrolytic reactions as these metallic
species are also known to be very good petrochemical catalysts (see Appendix A for
the chemical analysis of high temperature grade stainless steel (SS253MA)). In
contrast, quartz is relatively inert [119, 120], it can withstand temperatures above
1000°C and is relatively inexpensive. Furthermore, similar quartz reactors [113, 121]
heated in an external furnace have reported rapid particle heating rates of the order
104°C s-1.
The quartz reactor for this study consisted of a 40mm diameter chamber with two
quartz frits (porosity 0) spaced 130mm from one another (Figure 2.8). The upper frit
prevented the elutriation of char particles and permitted the removal of evolved
volatile material out of the quartz reactor during pyrolysis whilst the bottom frit
reduced the incidence of gas channelling through the sand bed. Furthermore, a
second argon stream containing the coal particles facilitated in fluidising the sand
bed. This second stream was injected directly into the chamber via a ¼” tube located
approximately 20mm above the bottom glass frit. To achieve rapid particle heating
rates, this ¼” particle injection tube was surrounded by a water cooled 1/2”
condenser.
Chapter 2 Experimental 38
Fluidisinggas
Coolingwater
in
Sandbed
Quartzfritz
Lignite or thermally treated
particles entrainedin gas
Thermocouple
Volatiles
Water cooledprobe
Coolingwaterout
Collection tube(collection of
charred particles)
Volatiles andgas out
130m
m
40mm60mm
480m
m
20m
m
Figure 2.8 Fluidized quartz reactor design
Chapter 2 Experimental 39
2.15.4 Slow pyrolysis experiments For the slow heating experiments, 160g of acid washed dried zircon sand was loaded
into the quartz reactor and the reactor weighed. The raw lignite or thermally treated
products (3.00g db) were also fed into the reactor at room temperature. Ultra high
purity (>99.99%) argon (total flow rate of 3.0L min-1) was used to fluidised the sand
bed and to thoroughly mix the lignite material inside the reactor. The argon was fed
into the reactor at two separate points, from the bottom and from a tube inside the
condenser arm (Figure 2.8). The flow of argon into the reactor was controlled using
two Aalborg Mass Flow Controllers (MFCs). A flow of water (1L min-1) was also
used to cool the gas inlet inside the condenser arm and prevent subsequent blockages
during the experiment.
The quartz reactor was then lowered into a Ceramic Engineering tubular furnace
(Figure 2.9) and slowly heated (40-50°C min-1) from room temperature to the desired
pyrolysis temperature. The quartz reactor was maintained at temperature for 15min
before being lifted out from the furnace and cooled down. The pyrolysis temperature
was measured with a K-type thermocouple positioned on the upper frit of the quartz
reactor (Figure 2.8). The maximum temperature overshoot allowed was 0.5°C. The
tubular furnace contained three separate heating zones which provided uniform
heating along the quartz reactor chamber (Figure 2.9).
Chapter 2 Experimental 40
Fluidising gas
Coolingwater
Coal particlesfrom feeder
Thermocouple
Volatiles and gas out
Furnace walls
Quartz reactor
Pulley
Reel
Furnace Thermocouple 1
Furnace Thermocouple 2
Furnace Thermocouple 3
Figure 2.9 Schematic diagram of the pulley system used to lower and raise the quartz
reactor from the tubular furnace.
Chapter 2 Experimental 41
2.15.5 Fast pyrolysis experiments For fast heating experiments, a particle feeder was used to inject a stream of raw
lignite and thermally treated particles directly into the reactor at pyrolysis
temperature (Figure 2.7). The schematic of the experimental rig for the fast pyrolysis
tests is shown in Figure 2.10. Similar to the slow heating experiments, 160g of acid
washed dried zircon sand was loaded into the quartz reactor and the reactor weighed.
The flow of argon and cooling water were also kept the same as described above.
The quartz reactor was then heated to the desired temperature and allowed to
equilibrate for 15min before injecting the lignite material (~3.0g db) from the particle
feeder and into the hot reactor. The particles were fed from a particle feeder (see
Section 2.15.2) at a rate of ~100mg min-1.
On completion of the experiment, the quartz reactor was cooled down to room
temperature, the water within the condenser was evaporated using a Bunsen burner
and any residual tar material that was deposited within the thermocouple and exit
tubes was burnt off. The quartz reactor was then reconnected to the experimental rig
and heated to 150°C. A stream of argon passed through the sand/char bed to drive
off absorbed moisture for 60min and then cooled to room temperature. The weight
loss during pyrolysis (i.e. the total volatile yield) was determined by direct
measurement of the reactor weights before and after each experiment.
Chapter 2 Experimental 42
Arg
on
P
P P
Feeder
Reactor
Exhaust
P
MFC
T
Mass Flow Controller
Pressure gauge
Thermocouple
Valve
MFC
MFC
Figure 2.10 Schematic of experimental rig
Chapter 2 Experimental 43
2.15.6 Char collection A schematic of the method for collecting the char particles is shown in Figure 2.11.
The char particles were retrieved from the quartz reactor via the collection tube. To
collect the char particles, the particle injection and thermocouple tubes were plugged
and a stream of argon was passed through the fluidising gas tube and out of the
collection tube. The heavy zircon sand made it easy to separate the residual char
from the sand due to differences in density. The flow rate of the argon was increased
until the char was separated from the sand and collected in a cellulose thimble. The
char material was then placed in a plastic container and stored in a freezer for
subsequent analysis.
Tube plugged
Argon in
Argon out
Sand and charred particles
Charred particlesin collection tube
Figure 2.11 Schematic of the method for collecting the char particles from the quartz
reactor.
Chapter 2 Experimental 44
2.16 Volatile matter and fixed carbon determination Proximate analysis using a Setaram Labsys thermogravimetric analyser (TGA) was
conducted on raw and thermally treated products. Approximately 25 to 30mg of oven
dried lignite (105°C, Argon, see Section 2.13) of less than 250μm particle size was
weighed to ±0.00001g into the TGA’s ceramic crucible. The TGA was purged with
argon gas to provide an inert atmosphere. The sample was initially heated at 20°C for
30sec. The temperature was then increased at a rate of 30°C per min up to 110°C and
maintained at this value for 15min so as to remove any reabsorbed water. The sample
was then heated at a rate of 50°C per min to 950°C and held at 950°C for 15min. The
purging gas in the TGA was then switched over to air to allow combustion of the
lignite material. The analysis was completed after a flat baseline was obtained. The
ash yield of the sample was determined from the mass remaining in the crucible.
2.17 Combustion reactivities Combustion reactivities for the lignite samples (125 to 250μm particle size) were
measured using a TGA. The sample was initially heated at 20°C for 30sec in air. The
temperature was then increased at a rate of 30°C per min up to 400°C (non-
isothermal) and maintained until complete conversion was achieved (isothermal).
Combustion experiments were also conducted on Morwell and Yallourn MTE
products (150°C/5.1MPa) up to 400°C (90 to 125μm and 250-500μm particle size,
respectively) and also up to 450°C for the 250-500μm particle size only. The
combustion reactivity (Rx) was calculated from the DTG data (dW/dt) according to
the formula: dt
dWW
RX ⋅−=1
where W is the weight of the sample (dry ash free) at any given time (t).
Chapter 2 Experimental 45
2.18 Ash determinations Ash determinations on raw and thermally treated lignites were conducted using a
Carbolite (type ESF 12/2) furnace. Two representative samples of the lignite to be
ashed (~1g) were pre-dried at 105°C under a flow of nitrogen (see Section 2.13). The
dried samples were taken out of the oven, immediately placed in silica crucibles of
known mass and then weighed to ±0.0001g. A third empty crucible was weighed at
the same time. The three crucibles were heated without lids in a muffle furnace at
400°C for 3h. The temperature of the oven was then raised to 600°C (to avoid loss of
sodium), and was held for 6h. The crucibles plus ash were cooled for 30min in a
desiccator and weighed. The mass of the ash was calculated, correcting for the
change in mass of the crucibles during heating from the change in mass of the empty
crucible.
2.19 Total inorganic contents The total sodium (Na), calcium (Ca), magnesium (Mg), iron (Fe), aluminium (Al),
potassium (K), silicon (Si) and titanium (Ti) were determined according to the
Australian Standard method (AS 1038.14.1) [122]. The acid extractable metals Na, Ca,
Mg, Fe and Al were determined by a method based on the Australian Standard
method (AS 2434.9) [123]. An oven dried (see Section 2.13) lignite sample was wetted
with ethanol in a beaker, then 60mL of a solution of 0.003M H2SO4 was added and
the mixture brought to boil, allowed to simmer uncovered for 15min and covered
(with a watch glass) for a further 45min. The mixture was then filtered and the filter
cake was further washed with hot 0.003M H2SO4.
Chapter 2 Experimental 46
The Al, Fe, Mg, Ca and Na contents of the filtrate were determined by Atomic
Absorption Spectrophotometry (AAS). The instrument was calibrated against
solutions of known concentration. The sample absorption was then used to obtain the
sample concentration by reading off the calibration curve. Analyses were done in
triplicate with regular re-calibrations and blank solution checks.
2.20 Product water analysis
2.20.1 Cation and anion analysis Cation and anion analysis were performed at Monash University Water Studies
Centre (WSC) on the product water obtained from the drying processes. A Perkin
Elmer flow injection system (FIAS 200) coupled to a Perkin Elmer atomic
absorption spectrometer was used to analyse for potassium (WSC test method 11A),
sodium (WSC test method 12A), calcium (WSC test method 10A) and magnesium
(WSC test method 13A). Direct AAS was used for aluminium analysis using WSC
test method 14.
Chloride was determined colourimetrically by reacting mercury (II) thiocyanate
(Hg(SCN)2) and iron (III) to form the complex ion Fe(SCN)2+, which is red in
colour. The absorbance of the coloured complex, measured spectrophotometrically,
is proportional to the Cl– in the sample (WSC test method 7).
Sulphate was detected by mixing barium chloride and methyl thymol blue solution
with the wastewater sample. Sulphate present in the wastewater precipitated as
barium sulphate. Excess barium in the solution was detected by the formation of a
Chapter 2 Experimental 47
blue coloured chelate at high pH, the concentration of which depends on the
concentration of sulphate in the solution. As the sulphate concentration increased, the
amount of barium available to form the chelate was reduced and the absorbance
decreased (WSC test method 8).
2.19.1 Total organic carbon (TOC) The TOC present in the water produced from the drying processes was determined
according to WSC test method 57A, using a Shimadzu TOC-5000 analyser.
The product water sample was injected into the TOC analyser and heated to 700°C in
the presence of high purity oxygen. The oxidation of the organic carbon material to
carbon dioxide (CO2) was detected and recorded by the instrument. A calibration
curve was produced and the peak area was proportional to the concentration of the
carbon in the sample.
2.21 Surface area and micropore volume CO2 and N2 adsorption isotherms were conducted on the lignites and thermally
treated products to determine the surface areas and pore size distributions,
respectively. The adsorption isotherms were measured using a Micromeritics Tristar
surface area analyser. The glass sample tube, equipped with a Transeal™ valve, was
weighed on a five-figure balance. Samples were placed in the sample cell (0.15 to
0.25g) and the entire set-up reweighed on the same balance. The samples were then
heated at 105°C for 6 hours under vacuum. The individual glass tubes were cooled
down to room temperature and the entire set-up reweighed to determine the mass of
the dried sample.
Chapter 2 Experimental 48
The sample tubes were then connected to the Tristar analyser and the entire surface
area analyser system placed under vacuum for 4 hours.
Typical run conditions for CO2 adsorption were :
Bath temperature (°C) 0
Adsorption cut off pressure (P Po-1
) 0.040
Desorption cut off pressure (P Po-1
) 0.020
Adsorbate dose : (cm3g
-1 STP) 1.000
Desorbate dose : (cm3g
-1 STP) 1.000
Equilibration sampling time (s) 180
The surface area and micropore volume of the carbon samples were calculated via
the Dubinin-Radushkevich equation from the CO2 adsorption isotherm. The carbon
dioxide isotherms were completed within a 6h period in order to avoid swelling
effects of the lignite from long exposure to CO2 [124-126].
Typical run conditions for N2 adsorption were:
Bath temperature (°C) -196
Adsorption cut off pressure (P Po-1
) 0.995
Desorption cut off pressure (P Po-1
) 0.700
Adsorbate dose : (cm3g
-1 STP) 5.000
Desorbate dose : (cm3g
-1 STP) 5.000
Equilibration sampling time (s) 30
The BET method [127] was applied to calculate the surface area from the N2
adsorption isotherm.
Chapter 2 Experimental 49
2.22 Helium density Density determinations were made by helium displacement using a Micromeritics
Accupye 1330 pycnometer. In contrast to mercury, the helium penetrates the
micropore structure of the sample and the density determined is therefore the true
density of the sample. Regulated pressure from the helium cylinder was manually set
at 135kPa. Calibration of the pycnometer was conducted on each day of the analysis
using Micromeritics stainless steel ball standards. Conditions for calibrating the
instrument using the standard and performing an analysis of the sample was:
Cell volume 11.753 cm3
Number of runs / purges 10
Equilibration rate 35 Pa min-1
Maximum fill pressure 133kPa
The pycnometer automatically measured the room temperature and zeroed the
pressure transducers during the analysis. Reported results are the average and
standard deviation of ten sequential determinations for each sample.
Chapter 2 Experimental 50
2.23 Pyrolysis-gas chromatography Pyrolysis gas chromatography (py-gc) was carried out on the raw Loy Yang lignite
and on the thermally treated products. An Agilent 6890N series gas chromatograph
connected to a SGE Pyrojector II was used. The gas chromatograph was equipped
with a flame ionisation detector (FID) and a thermal conductivity detector (TCD).
Two columns, a 30m GS-GasPro and a 30m HP-5 obtained from Agilent
Technologies (see Table 2.5) were used.
For each experiment, a solid pelletizer was used for injecting the carbon products
into the projector. Each pelletizer was weighed to an accuracy of ±0.00001g using a
Sartorius ME balance. The raw lignite and thermally treated products (particle size
90-125μm) were pre-dried in a vacuum oven at 80°C for a minimum of 24h prior to
loading ~1.0mg of sample into the pelletizer. The pelletizer assembly was then
placed in a separate vacuum oven at 30°C for a minimum of 3h and then re-weighed.
The pelletizer assembly was then connected to the pyrojector and the air inside the
pelletizer removed by purging helium through the assembly for a minimum of
30min, prior to injection into the pyrojector furnace.
For Loy Yang lignite and the thermally dried products, pyrolysis was performed at
600°C, 700°C, 800°C and 900°C. Several raw lignites from the Latrobe Valley were
also pyrolysed at 900°C. The py-gc conditions are given in Table 2.4. After each
experiment, the pyrojector furnace was cooled down and the entire pyrojector
assembly dismantled. The charred carbon within the furnace liner was removed and
Chapter 2 Experimental 51
the quartz furnace liner and transfer tube cleaned by burning off residual organic
matter in an external ceramic furnace at 400°C.
Table 2.4 Run conditions for pyrolysis-gas chromatography
Column GS-GasPro HP-5
Pyrojector pressure (psi) 25.0 25.0
Initial oven temperature 30°C 30°C
Time at initial oven temperature 4.0min 4.0min
Temperature program Rate : 3°C/min to 260°C 3°C/min to 320°C
Final temperature 260°C 320°C
Time at final temperature 10min 30min
Injector temperature 250°C 250°C
Injector pressure (psi) 22.75 22.75
Injector split splitless splitless
FID detector temperature 320°C 450°C
FID Hydrogen flow (mL min-1) 40.0 40.0
FID Air flow (mL min-1) 450 450
FID Helium makeup flow (mL min-1) 10.0 10.0
TCD detector temperature 320°C 320°C
TCD reference flow (mL min-1) 20.0 20.0
TCD Helium makeup flow (mL min-1) 7.0 7.0
Carrier gas He He
Carrier gas flow 5.5mL/min 5.5mL/min
Chapter 2 Experimental 52
Table 2.5 Capillary Column properties
Column GS-GasPro HP-5
length 30m 30m
ID mm 0.32mm 0.53mm
Film layer 10μm 0.15μm
Standard gas mixtures from BOC gases (Table 2.6) were used to identify and
quantify the concentration of the chromatogram peaks. Also, the water peak in the
TCD gc-trace was identified and quantified by injecting water vapour in the gas
chromatograph containing the GS-GasPro column. The water vapour was generated
with a Setaram Wetsys humidifier system.
Calibration curves for each of the standard gas mixtures were performed by varying
the injection volume of the gas sampling valve. Three injection volumes were
performed for each calibration curve, 0.200mL, 0.500mL and 1.000mL, respectively.
Table 2.6 Standard gas mixtures
Standard gas mixture 1 (in helium)
Concentration (%)
Standard gas mixture 2 (in methane)
Concentration (%)
Methane 4.11 ± 0.08 Carbon monoxide 5.02 ± 0.5
Ethane 1.40 ± 0.03 Carbon dioxide 20.2 ± 0.2
Ethylene 1.60 ± 0.03 Hydrogen 29.0 ± 0.2
Propane 1.10 ± 0.02
Propylene 2.35 ± 0.05
iso-butane 0.889 ± 0.018 Standard gas mixture 3
Concentration (%)
n-butane 0.636 ± 0.013 Acetylene 100
Chapter 2 Experimental 53
Liquid standards obtained from Sigma-Aldrich (Table 2.7) were also used to identify
and quantify chromatogram peaks. Injection of liquid samples was performed using
an Agilent 7683B autoinjector on the gas chromatograph.
Table 2.7 Liquid standards
n-hexane methanol benzene
1-hexane ethanol toluene
n-heptane n-propanol o-xylene
n-octane n-butanol m-xylene
n-nonane n-pentanol p-xylene
n-decane n-hexanol phenol
C8-C20 alkane solution n-heptanol hydroquinone
C21-C40 alkane solution n-octanol acetone
formaldehyde n-nonanol methylacetate
acetaldehyde n-decanol ethylacetate
Chapter 2 Experimental 54
2.24 Pyrolysis–gas chromatography–mass spectrometry Pyrolysis – gas chromatography – mass spectrometry (py-gc-ms) analysis was
utilised to identify conclusively the chemical nature of the peaks observed in
chromatograms. Py-gc-ms was performed only for the raw Loy Yang lignite. An
Agilent 6890N series gas chromatograph connected to a SGE Pyrojector II and an
Agilent 5973N Mass Selective Detector (MSD) were used. The analysis conditions
are shown in Table 2.8. The column was a HP-5, (30 meter in length, 0.53mm
internal diameter (id)). All gc runs were conducted with a 10:1 split mode.
Table 2.8 Run conditions for gas chromatography-mass spectrometry
Column HP-5
Initial oven temperature 40°C
Time at initial oven temperature 2.0min
Temperature program Rate : 4°C/min to 320°C
Final temperature 320°C
Time at final temperature 15min
Injector temperature 250°C
Detector temperature 320°C
Carrier gas He
Carrier gas flow 1.0mL/min
Chapter 2 Experimental 55
2.25 Scanning Electron Microscope (SEM) - Energy Dispersion X-rays (EDX)
Qualitative and semi-quantitative elemental chemical analyse of the samples were
performed using a JEOL JSM-6490LA SEM connected to an Energy Dispersive X-
ray (EDX) probe. The analysis conditions for the SEM-EDX are shown in Table 2.9.
The SEM-EDX was used to scan the surface of raw lignites or the surface of ash, to
detect the presence of metallic species.
Table 2.9 Run conditions for SEM-EDX
SEM dwell time 0.1 msec
Sweep count 100
Working distance (WD) 11 mm
Voltage 20.0 kV
Probe Current 1.00 nA
PHA mode T3
Energy Range 0 - 20 keV
2.26 X-ray diffraction XRD analyses were conducted on raw Latrobe Valley lignites to detect the presence
of ionic salts. X-ray diffraction analysis was conducted using a Bruker AXS D8 X-
ray Diffractometer (XRD) at Swinburne University of Technology. The XRD
experiments were carried out using a Cu Kα radiation source in the 2θ scan mode.
Chapter 2 Experimental 56
2.27 Errors The errors for the measured quantities are given in the Tables. They are derived from
three sources.
• For those quantities measured in accordance with an Australian Standard, the
errors are based on the reproducibility quoted in the standard.
• For other quantities measured by outside laboratories, the errors are those
given by the analyst.
• For other quantities, the errors are based on variation found in multiple
determinations for the sample.
Chapter 3 Lignite drying technologies 57
CHAPTER 3 LIGNITE DRYING TECHNOLOGIES
The key to reducing greenhouse gas emissions and improving the efficiency of
existing lignite power stations is the implementation of more efficient drying
technologies. This chapter will investigate the HTD, MTE and SD technologies
under different processing conditions. Factors, which may affect the combustion
reactivity of the dried products such as pore structure and inorganic compositions,
will also be investigated in this chapter. Furthermore, the combustion reactivity of
the products is investigated in Chapter 5.
3.1 Effect of HTD conditions on retained moisture A major problem associated with moisture determinations of HTD products is the
difficulty in distinguishing between the surface water encapsulating the particles and
the residual moisture inside the particles. Equilibrium moisture holding capacities
(MHC) are commonly used to determine the moisture contents of products at
different humidities [128], but these would be misleading in a continuous
(commercial) process where a filtration step is likely to be incorporated prior to
combustion. The moisture content of the filter cake should provide a better indication
of the extent of dewatering from the HTD process than MHC.
Chapter 3 Lignite drying technologies 58
The filter cake moisture contents from the HTD process at different temperatures are
shown in Table 3.1.
Table 3.1 Effect of temperature on the filter cake moisture content from the HTD process.
H2O
Water retained in the filter
cake
Proportion water removed
Solids recovery
% wb g/g db %db % db ±0.5* ±0.05+ ±0.55+ ±0.2 Loy Yang raw 59.7 1.48 - -
180°C 59.5 1.47 0.7 99.7
200°C 58.7 1.42 4.0 99.5
230°C 57.9 1.38 7.1 98.2
250°C 55.9 1.27 14.4 98.0
280°C 53.6 1.16 22.0 95.2
300°C 49.6 0.98 33.5 92.0
320°C 43.2 0.76 48.6 86.3
350°C ^ 34.7 0.53 64.1 74.5
All experiments repeated in triplicate (^duplicate experiments) * Error determined from reproducibility of results + Error determined from summation of errors in analytical results
It can be seen that increasing the treatment temperature reduced the moisture content
of the HTD product material; however, moisture reductions (outside the error limits)
were only observed for processing temperatures of 200°C and above (Figure 3.1). At
200°C, the moisture content of the HTD product was reduced by 4% db and, at
230°C, by 7% db. Such low moisture reductions are inadequate and unacceptable in a
commercial process. It was only when the HTD temperature was raised to 320°C that
the moisture in the product material was reduced to approximately half of that in the
parent lignite.
Chapter 3 Lignite drying technologies 59
High HTD temperatures give drier products; however, such high temperatures also
lead to high levels of organic material in the product water (see APPENDIX F). At
320°C, the quantity of TOC removed to the product water was about 100 mg (see
APPENDIX F). This represents approximately 1.0% of the organic matter originally
charged to the HTD reactor and highlights the fact that substantial product water
remediation facilities would be required as part of any commercial development of
this process. Furthermore, high processing temperatures also lead to a significant
reduction in solids recovery in the final product. At 350°C, only 75%db of the
original mass could be recovered (Table 3.1).
The reduced moisture content for HTD products is thought to be largely due to
structural changes (both chemical and physical) that take place during processing and
that are enhanced at higher temperatures. Some of the structural changes that occur
include:
1. Decarboxylation: An increase in temperature will increase the extent of
decarboxylation [129-131]. Decarboxylation, reduces the concentration of polar sites
and, thus, reduces the capacity of the product to retain water [132, 133].
2. Migration of waxes: At higher processing temperatures the lignocellulosic
structure of the lignite becomes less rigid. This may facilitate the migration of
loosely bound lignite components, such as waxes [134] to the surface, thus increasing
the hydrophobicity of the product material [54, 59, 135].
3. Changes in spatial arrangement: The migration of water and waxes out of the
original structure may cause irreversible changes in the spatial arrangement of the
atoms leading to a reduction in porosity [136].
Chapter 3 Lignite drying technologies 60
A HTD processing temperature of 320°C has been advocated by several previous
workers [105, 136]. Generally, the physico-chemical changes that occur for lignites at
this temperature also facilitates the formation of higher concentration lignite-water
slurries than can be achieved at lower processing temperatures [59]. So, in HTD
processing, temperature is a key factor. Other process variables, such as reaction
time, heatup time, agitation, autoclave size, slurry concentration and lignite mean
particle size (>50mm) have relatively little effect on the (large) pore volumes that
remain in the product material [136] and hence on the retained moisture in the product.
3.2 Mechanical Thermal Expression (MTE)
3.2.1 Effect of processing temperature The effect of temperature on the retained water in the MTE beneficiated product is
shown in Table 3.2 and in Figure 3.1. The relationship between processing
temperature and retained moisture is almost linear. It is also evident (Table 3.2) that
more than half of the water present in the raw lignite was removed under very mild
MTE conditions (125°C and 5.1MPa). Increasing the temperature to 150°C led to
64% of the water being removed (i.e. 0.5% water removed from the MTE pellet per
1°C increase from 125°C to 150°C). Further increases in processing temperature
from 150°C to 250°C resulted in additional water being removed from the MTE
pellet however at a much reduced rate (i.e. 0.12% water reduction per 1°C increase
from 150°C to 250°C; Figure 3.1).
Chapter 3 Lignite drying technologies 61
Table 3.2 Mass recovery and moisture content of raw lignites and MTE products treated at different temperatures and 5.1MPa mechanical pressure
H2O
Water retained in product
Proportion water removed
Solids recovery
% wb g/g db %db % db ±0.5* ±0.05+ ±0.55+ ±0.2 Loy Yang raw 59.7 1.48 - -
125°C 42.0 0.72 51.2 99.6
150°C 34.8 0.53 63.9 99.6
180°C 32.3 0.48 67.8 99.6
200°C 29.5 0.42 71.7 99.6
250°C 22.4 0.29 80.5# 99.0
All experiments repeated in triplicate * Error determined from reproducibility of results + Error determined from summation of errors in analytical results # Error is less than 1.5 (see APPENDIX A for all errors)
Several workers have also reported similar linear trends between processing
temperature and retained moisture in their batch MTE systems [15, 137]. Guo [15]
reported a rate of moisture reduction of about 0.12% per 1°C increase from 150°C to
300°C at 12MPa applied pressure. A similar moisture reduction rate was obtained in
this study from 150°C to 250°C (Table 3.2 and Figure 3.1).
Chapter 3 Lignite drying technologies 62
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0 100 200 300 400
Temperature (°C)
Ret
aine
d m
oist
ure
(g/g
db)
Raw lignite HTD MTE SD
50% moisture reduction line
raw lignite
Figure 3.1 Effect of temperature on the retained water in the HTD product; MTE
pellet with 5.1MPa of applied mechanical pressure; and in the SD product.
In contrast, Kealy et al. [106] reported that increasing the processing temperature
beyond 150°C gave only marginal improvements in reducing the moisture content of
the final MTE product. These apparent different conclusions can be rationalized by
realizing that Kealy et al. [106] had deduced this at lower applied pressures (i.e. at
2.5MPa and 5MPa) whereas other workers (except for this study) had based their
conclusions at 12MPa. On the contrary, Guo [15] did report that at low applied
mechanical pressures (eg 3MPa), an increase in processing temperature from 180°C
to 200°C gave products with similar moisture contents whereas, further increasing
the processing temperature to 230°C gave an adverse effect with a significant
increase in the product moisture content (more than 10%wb). Guo [15] explained this
phenomenon as a significant reduction in the effective compression pressure caused
by the sharp increase in the saturation vapour pressure of water. Based on the results
Chapter 3 Lignite drying technologies 63
from this study there is no particular optimum temperature for the MTE process; but
it is clear that the proportion of water removed at any given temperature is far higher
by MTE than by HTD.
3.2.2 Effect of mechanical pressure The effect of applied mechanical pressure on the retained water in the MTE
beneficiated product is shown in Table 3.3 and in Figure 3.2. Almost half of the
moisture in the lignite was removed at relatively low mechanical pressures (2.5MPa)
at 150°C. An increase in mechanical pressure from 2.5MPa to 5.1MPa at 150°C
reduced the moisture content of the product significantly by an additional 15%db,
whereas a further increase in mechanical pressure had relatively little effect.
Table 3.3 Mass recovery and moisture content of raw lignites and MTE products treated at 150°C and at different mechanical pressures.
H2O
Water retained in product
Proportion water removed
Solids recovery
% wb g/g db %db % db ±0.5* ±0.05+ ±0.55+ ±0.2 Loy Yang raw 59.7 1.48 - -
2.5 MPa 43.0 0.76 48.9 99.6
5.1 MPa 34.8 0.53 63.9 99.6
12.7 MPa 32.0 0.47 68.2 99.6
25.0 MPa 29.5 0.43 71.0 99.5
All experiments repeated in triplicate (^duplicate experiments) * Error determined from reproducibility of results + Error determined from summation of errors in analytical results Over the range of the variables tested for MTE, there appears to be an optimum
pressure (approx 5.1MPa) beyond which further increases have relatively little effect
on the moisture content of the final pellet. In regards to processing temperature, an
optimum condition could not be deduced in this study.
Chapter 3 Lignite drying technologies 64
0
5
10
15
20
25
30
0.0 0.2 0.4 0.6 0.8 1.0
Retained moisture (g/g)
App
lied
mec
hani
cal p
ress
ure
(MP
a)
Figure 3.2 Effect of applied mechanical pressure on the retained water in the MTE pellet at 150°C
(Data points are represented by symbols, lines are interpolated)
Preliminary MTE studies on some Indonesian [138] and German [139] lignites under
different MTE processing conditions suggest that there is no single set of optimum
MTE condition for all lignites. In general, these studies suggest that lignites with
high moisture contents require milder MTE processing (i.e. lower temperatures and
mechanical pressures); conversely lignites with lower inherent moisture contents,
require more severe MTE processing conditions.
The operating regime can be compared with the processing conditions employed in
the 25 tonne per hour demonstration plant at Niederaussem, which was (typically)
200°C and 12MPa mechanical pressure A processing temperature of 200°C is
relatively mild when compared to effective HTD temperatures (refer above).
Similarly, mechanical pressures of 12MPa are also mild when compared to simple
the 25 tonne per hour demonstration plant at Niederaussem, which was (typically)
200°C and 12MPa mechanical pressure A processing temperature of 200°C is
relatively mild when compared to effective HTD temperatures (refer above).
Similarly, mechanical pressures of 12MPa are also mild when compared to simple
Chapter 3 Lignite drying technologies 65
mechanical expression, where mechanical pressures up to 50MPa have been
employed at ambient temperature [12].
Over the temperature range examined in this study (< 250°C), the changes in the
organic structure in the MTE products are minimal. Not surprisingly, the overall
mass recovery was also very high with more than 99% of the original mass being
recovered after the MTE process. It is only at temperature above 250°C that
substantial decarboxylation takes place [104] and migration of waxes may occur [136].
3.3 Steam drying (SD) - effect of SD conditions on retained moisture
Prior studies on steam drying [140, 141] have suggested that increasing the processing
temperature will lower the lignite’s moisture content, increase the rate of drying,
increase the degree of decarboxylation and devolatilisation, and increase the
concentration of organic material in the condensate. However, very little quantitative
information is available regarding changes to the lignite’s chemical and physical
properties. Similar to prior studies, the water content of the SD product decreased
with increasing temperature and the relationship was approximately linear over the
range of 130°C to 320°C (Table 3.4, Figure 3.1). Between 320°C and 350°C, the
further reduction in the water content was slight since, overall, the residual water
content was already very low.
The advantage that comes with increasing the temperature to reduce the moisture
content is counterbalanced by the increased loss of volatile material (see solids
recovery values in Table 3.4). Temperatures higher than 250°C were accompanied by
Chapter 3 Lignite drying technologies 66
increasingly large losses of volatile matter and reduced solids recoveries.
It should be noted that in SD, unlike HTD and MTE, the residence time of particles
at processing temperature and pressure can have a significant effect on the proportion
of moisture removed. Also, in situations where steam removal has been continuous
(as distinct from the batch approach used here) the levels of retained moisture have
typically been much lower than the values reported here (see Bongers et al. [140]).
Table 3.4 Effect of temperature on the filter cake moisture content from the SD process.
H2O
Water retained in product
Proportion water removed
Solids recovery
% wb g/g db %db % db ±0.5* ±0.05+ ±0.55+ ±0.2 Loy Yang raw 59.7 1.48 - -
130°C 59.6 1.48 0.7 99.9
150°C 55.8 1.26 15.0 98.8
180°C 50.5 1.02 31.6 98.5
200°C 44.9 0.81 45.2 98.4
230°C 38.4 0.62 56.1 97.6
250°C 27.4 0.38 74.7 95.5
280°C 15.4 0.18 87.7 91.6
300°C 11.5 0.13 91.3 88.5
310°C 6.8 0.07 95.1 85.8
320°C 4.7 0.05 96.7 83.5
330°C 2.2 0.02 98.5 80.6
350°C 1.4 0.01 99.0 72.0
All experiments repeated in triplicate * Error determined from reproducibility of results + Error determined from summation of errors in analytical results
Chapter 3 Lignite drying technologies 67
3.4 Pore structure of HTD, MTE and SD products Changes in the product’s physical properties during drying (e.g. pore
reduction/product hardness) may play a significant role in the milling process prior to
combustion. Measuring the internal pore volume of thermally treated products can
also provide useful information into the product’s maximum slurry concentration in a
coal water mixture [59, 105, 142], and into the product’s moisture holding capacity [59].
Low porosity products have been produced in both the MTE [142] and from the HTD
process [59]. Currently, very little information is available on whether significant
changes in the product’s pore volume can affect its combustion reactivity. Low
porosity HTD products from high porosity lignites have been reported to take
considerably longer to reach complete combustion [59].
The internal pore structure of the products as determined by mercury porosimetry,
surface area analysis and helium pycnometry are shown in Table 3.5. The surface
area of the lignite products was measured by carbon dioxide gas adsorption and
calculated using the Dubinin equation [107]. Differences between carbon dioxide gas
adsorption and nitrogen BET gas adosprtion on lignites are further discussed in
APPENDIX B. For MTE, increasing the processing temperature resulted in a gradual
increase in the product surface area (about 20m2g-1 from 125°C to 250°C) (Table 3.5
and Figure 3.4). Similarly, for SD, the surface area of the products marginally
increased from 130°C to 230°C but at higher temperatures, the surface area fell at an
average rate of ~0.6 m2g-1 per 1°C increase (i.e. from 231m2g-1 at 230°C to 164m2g-1
at 350°C). In contrast, the surface area of the HTD products also decreased
significantly with increasing processing temperatures above 250°C but at an average
Chapter 3 Lignite drying technologies 68
rate of ~0.9 m2g-1 per 1°C increase (i.e. from 229m2g-1 at 250°C to 138m2g-1 at
350°C). The lower surface areas for the HTD products compared to SD products
could be the result of loosely bounded components within the lignite such as waxes
and tars [143] which become mobilized at the higher temperatures. These mobilized
components may in turn permeate into the micropore region of the particle and
subsequently occupy some of the volume within this pore region. As a consequence,
the product’s surface area is also reduced (see Section 3.1). For MTE, an increase in
applied mechanical pressure from 2.5MPa to 12.7MPa resulted in a gradual increase
in the product’s surface area from 205m2g-1 to 219m2g-1. However, further increases
in mechanical pressure beyond 12.7MPa resulted in a decline in the product’s surface
area (Figure 3.3). This inverted hyperbolic relationship between surface area and
applied mechanical pressure is likely to be the result of physical (rather than
chemical) changes that occur at the different mechanical pressures. The rise of the
product’s surface area with increasing mechanical pressure could be the result of the
larger pores (not measured by CO2 adsorption) being compressed into the micropore
region (Figure 3.3). Whereas, the fall in the product surface area with higher applied
mechanical pressures may be attributed to micropores collapsing by the additional
force of the mechanical press.
It should be pointed out that the micropore region measured by CO2 adsorption only
covers a very small portion of the overall pore size distribution (i.e. 1.0 to 0.25nm
pore radius) of the lignite products. Subsequently, the pore regions from mercury
porosimetry, surface area and helium pycnometry, respectively, (Table 3.5) were
combined to give a better representation on the overall pore size distribution of the
Chapter 3 Lignite drying technologies 69
thermally treated products (see APPENDIX C). The volume occupied by ‘carbon’
(impermeable to He) marginally decreased with increasing processing temperature
(MTE, HTD and SD) and with increasing applied mechanical pressure (MTE only)
(Table 3.5). This decrease could be attributed to some organic carbon being leached
out from the lignite during the drying process.
200
205
210
215
220
225
0 5 10 15 20 25 30
Mechanical pressure (MPa)
Sur
face
are
a (m
2 g-1)
Figure 3.3 Effect of applied pressure on the surface area of MTE products
Increasing the MTE temperature from 125°C to 250°C (fixed pressure) did not
change either the micropore or large pore volume significantly. However, for HTD,
there was a marked reduction in the large pore (intra-particle) volume of the product
with increasing temperature. The micropore volume and carbon density were also
reduced (Table 3.5). Increasing the MTE pressure (fixed temperature) substantially
reduced the large pore volume but not the micropore volume or carbon density. Over
the range of conditions investigated, the effect of mechanical pressure was greater
than that of temperature.
Chapter 3 Lignite drying technologies
70
Table 3.5 Effect of processing conditions on the large pore (mercury porosimetry), micropore (CO2 surface area) and carbon density (helium pycnometry) of MTE, HTD and SD products, respectively.
Hg porosimetry
CO2 Adsorption at 273K
He pycnometry
Large pores (1000 to 1.5nm
pore radius) (cm3g-1) ±0.02
Dubinin surface area (m2g-1)
±2
Micropores (1.0 to 0.25nm pore
radius) (cm3g-1) ±0.001
Carbon density (>0.08nm pore
radius) (g cm-3) ±0.0005
MTE 125°C / 5.1MPa 0.28 212 0.057 1.4001
MTE 150°C / 5.1MPa 0.27 214 0.057 1.3995
MTE 180°C / 5.1MPa 0.27 219 0.059 1.3990
MTE 200°C / 5.1MPa 0.25 224 0.060 1.3984
MTE 250°C / 5.1MPa 0.24 233 0.062 1.3956
MTE 150°C / 2.5MPa 0.31 205 0.055 1.3987
MTE 150°C / 5.1MPa 0.27 214 0.057 1.3995
MTE 150°C / 12.7MPa 0.27 219 0.059 1.3947
MTE 150°C / 25.0MPa 0.23 210 0.056 1.3992
HTD 200°C 0.61 234 0.063 1.3974
HTD 250°C 0.62 229 0.061 1.3907
HTD 280°C 0.56 218 0.058 1.3861
HTD 300°C 0.52 182 0.049 1.3805
HTD 320°C 0.46 161 0.043 1.3774
HTD 350°C 0.39 138 0.037 1.3696
SD 130°C 0.55 223 0.060 1.3952
SD 150°C 0.55 226 0.061 1.3953
SD 180°C 0.56 226 0.061 1.3945
SD 200°C 0.65 232 0.062 1.3923
SD 230°C 0.67 231 0.062 1.3919
SD 250°C 0.62 214 0.057 1.3901
SD 280°C 0.52 200 0.054 1.3836
SD 300°C 0.51 199 0.053 1.3816
SD 310°C 0.50 194 0.052 1.3810
SD 320°C 0.48 188 0.050 1.3764
SD 330°C 0.48 176 0.047 1.3745
SD 350°C 0.46 164 0.044 1.3723
71
Chapter 3 Lignite drying technologies
Figure 3.4 Relationship between surface area and processing temperature for (1) MTE, (2) HTD, (3) SD and (4) all three drying processes
120 140 160 180 200 220 240 260 280 300210
215
220
225
230
235
1 MTE
Surfa
ce A
rea
(m2 g-1
)
Processing Temperature (°C)150 200 250 300 350 400
140
160
180
200
220
2402 HTD
Surfa
ce A
rea
(m2 g-1
)
Processing Temperature (°C)
100 150 200 250 300 350 400
140
160
180
200
220
240
3 SD
Processing Temperature (°C)
Surfa
ce A
rea
(m2 g-1
)
100 150 200 250 300 350 400
140
160
180
200
220
240
4 MTE HTD SD
Surfa
ce A
rea
(m2 g-1
)
Processing Temperature (°C)
Chapter 3 Lignite drying technologies
72
For SD, increasing the temperature from 130°C to 180°C had no apparent effect on
pore volume but an increase to 230°C gave a significant increase in porosity
(Figure 3.5). This increase is likely attributable to changes in the lignite’s rigidity at
these temperatures, resulting in less shrinkage when the material is dried to zero
moisture for pore volume determination. It is unlikely that the pore volume of the
initial steam dried product is actually higher at 230ºC than at 180ºC (see Bongers et
al. [140]). A further increase in processing temperature above 230°C gave a small but
gradual decrease in the intra-particle large pore and micropore volumes and the
carbon density (Table 3.5). The trends at processing temperatures above 230°C are
similar to those for products of HTD but above 280ºC, the rate of decrease of large
pore and micropore volumes with temperature was greater for the product of HTD
than for the products of SD.
The volume of the retained water exceeded the available internal pore volume of the
products of MTE and HTD (Figure 3.6a) possibly because for both types of products,
some water may be adsorbed on particle surfaces. For products of SD (Figure 3.6b),
the volume of water retained did not appear to be related to the product’s pore
volume. It is believed that this is an artefact of the experimental protocol. At high
temperatures surface water may evaporate during cooling of the autoclave. At lower
temperatures very little evaporation would be expected, but excess water may be
retained on the particle surfaces (as for products of HTD) and/or the measured
porosity may be lower than its true value as a result of shrinkage during the drying to
zero moisture which precedes porosity measurement. As noted above, the effect of
temperature on pore volume suggests that the product of SD at lower temperatures is
MTE 150°C / 25.0MPa 0.017 0.017 0.021 0.004 0.025 0.12 0.021 SD 130°C 0.009 0.004 0.007 0.001 0.005 0.028 a Note that 100g wb raw lignite is equivalent to 40g of dried lignite SD 150°C 0.008 0.004 0.007 0.001 0.005 0.028 b the error is ±0.004g for Na, Ca, Mg, Al, FeNP and Cl whereas error in S tot is ±0.01g. SD 180°C 0.007 0.004 0.007 0.001 0.006 0.028 c NP = non-pyritic SD 200°C 0.006 0.004 0.007 0.001 0.006 0.028
a For HTD and SD, the error is ±0.001g for Na, Ca, Mg, Al, FeNP and Cl whereas error in S tot is ±0.005g; * the error for S tot is ±0.004g, b NP = non-pyritic.
Chapter 3 Lignite drying technologies
Chapter 3 Lignite drying technologies 77
Within the limits of error, it cannot be concluded whether any aluminium, non-
pyritic iron or sulphur was removed from any of the three drying processes
(Table 3.6 and Table 3.7). A decrease of aluminium and iron during MTE and HTD
is not often encountered for Latrobe Valley lignites [139, 151, 152] and for higher ash
South Australian lignites [139, 153, 154], however, some Indonesian coals have reported
reductions in aluminium and iron contents from both HTD [155] and MTE [138].
3.5.1 Sodium in dried lignite products Sodium species present during high temperature lignite combustion play an
important part in boiler fouling and are often a major cause of corrosion in turbine
blades [156]. Subsequently, the removal of sodium during drying can also significantly
improve the quality of the final product.
Both MTE and HTD gave significant reductions in sodium under all of the
experimental conditions tested whereas for SD, significant sodium reductions were
only found at temperatures above 230°C. Even at the mildest MTE processing
temperature, the level of sodium was reduced by 45% and further increases in
temperature only resulted in marginal additional sodium being removed (compare
Table 3.6 and Table 3.8). An increase in applied mechanical pressure from 2.5MPa
to 25MPa had no significant effect in removing more sodium.
For HTD processing at 200°C, the proportion of sodium removed was significant at
45%, despite the negligible water removal. This sodium reduction is probably due to
sodium salts washed out of the lignite from the additional water added to make up
the lignite slurry for the HTD process. The level of sodium was reduced by an
additional 22% with an increase in processing temperature up to 250°C (the
Chapter 3 Lignite drying technologies 78
maximum temperature used in the MTE process), and further increases in HTD
processing temperature resulted in more sodium being removed. At 350°C, the level
of sodium in the HTD product was negligible. Within the limits of error, the removal
of sodium in the MTE process was as efficient as for the HTD process at the same
temperature.
In most Latrobe Valley lignites the sodium is predominantly in the form of sodium
chloride [151] (also see APPENDIX E) but some of the cations present are bound to
carboxylate groups in the lignite [85]. Since the reduction in sodium levels in the HTD
process continued to increase with temperature till all the sodium had been removed,
sodium originally in carboxylate form was also leached out at sufficiently high
temperatures. It is likely that this sodium carboxylate had been converted to
carbonate/bicarbonate following degradation of the carboxylic groups and thus
became leachable.
The sodium level of the SD product also fell with increasing temperature, which is
not consistent with previous studies [90]. The major difference between the studies
described by Allardice [90] and this work is the method by which the steam is supplied
to the system. In most SD systems, a constant flow of steam is passed through the
lignite, so that all the removed water is converted to vapour. Hence it will not remove
any appreciable quantities of soluble inorganics. In the closed system used in this
study, liquid water will be present in equilibrium with the vapour phase, so that
soluble inorganics may be leached from the lignite into the liquid phase at the bottom
of the autoclave.
Chapter 3 Lignite drying technologies 79
Table 3.8 Proportions of inorganics, and chlorine removed by drying
Note: the proportions of aluminium, non-pyritic iron and total sulphur removed for all three drying processes are within the error limits of this analysis and therefore are not included in this table; NS = Not Significant; NA = Not Applicable (the concentration of the inorganic ion was too low to be detected, i.e. <0.01wt%db)
Chapter 3 Lignite drying technologies 80
3.5.2 Calcium and magnesium in dried lignite products All of the calcium and magnesium in Latrobe Valley lignites occur as exchangeable
cations associated with carboxyl groups [85]. In the MTE process, a significant
reduction in magnesium was only found at the higher processing temperature (i.e.
250°C / 5.1MPa). Similarly, for HTD and SD, significant reductions in the level of
magnesium were found at 280°C and 300°C, respectively (Table 3.8).
The behaviour of calcium and magnesium are expected to be very similar during the
drying processes however because the level of calcium was half that of magnesium,
the error associated in the analysis was higher. Significant reductions in calcium
were only present in HTD at processing temperatures ≥320°C however within the
limits of error, no significant differences were evident between Ca and Mg under
these conditions.
Similar to sodium, the reduction in magnesium and calcium levels are likely due to
extensive degradation of carboxylate groups and an increase in water removal at
these higher temperatures.
3.5.3 Inorganic reduction in dried products versus water removal The relationship between sodium removal and the proportion of water removed in
the dried products is given in Figure 3.7. For MTE and SD, the fraction of sodium
removed from the raw lignite was less than the fraction of water removed. In
contrast, for HTD, a significantly higher proportion of sodium was removed than
water (Figure 3.7). The higher proportion of sodium removed can be attributed to
Chapter 3 Lignite drying technologies 81
some washing of the lignite by the distilled water added to the HTD and MTE
systems (see Experimental). Direct comparisons between the proportion of water and
the proportion of total sodium removed from the lignite should be treated with
caution, as not all of the sodium in the lignite is soluble. Some of the sodium in Loy
Yang lignite is insoluble and is bound to the lignite as carboxylates and phenolates. It
is more meaningful to compare the proportion of water-extractable sodium removed
(calculated from distilled water washing of the lignite) with the proportion of water
removed.
0 10 20 30 40 50 60 70 80 90 100
-20
0
20
40
60
80
100 SD
% s
odiu
m re
mov
ed (d
b)
Proportion of water removed (%)
0
20
40
60
80
100 HTD
% s
odiu
m re
mov
ed (d
b)
0
20
40
60
80
100 MTE
% s
odiu
m re
mov
ed (d
b)
Figure 3.7 Relationship between drying process on product moisture and sodium
contents in the dried products. Note, the dotted line has been used in each graph for illustrative purposes only (i.e. 1:1 ratio) and should not be taken as the trendline for
all the given values.
Chapter 3 Lignite drying technologies 82
Figure 3.8 gives the proportion of water-soluble sodium leached out from the lignite
during the MTE, HTD and SD processes. Comparison of the data from the raw and
water-washed lignite (see APPENDIX E) showed only minor differences suggesting
that much of the inorganic material in these samples was strongly bound (Table E.1).
The level of sodium in Loy Yang lignite was reduced by 68 ± 11%db and chlorine by
45 ± 16%db. No significant differences were found in the levels of Ca, Mg, Al, FeNP
and Stot with water washing.
0 20 40 60 80 100 120 1400
20
40
60
80
100
120
140
% w
ater
-ext
ract
able
sodi
um re
mov
ed (d
b)
Proportion of water removed (%) HTD MTE SD
Figure 3.8 Relationship between the proportions of water and soluble sodium removed by MTE, HTD and SD processing.
Much of the sodium removed from the three drying processes under mild processing
temperatures (i.e. <250°C) was in the form of NaCl. The extent of decarboxylation at
temperatures below 250°C was relatively low and therefore sodium carboxylate
Chapter 3 Lignite drying technologies 83
degradation would also be marginal. For MTE, the proportion of water-soluble
sodium leached out was greater than the proportion of water removed which is also
likely attributable to some washing of the lignite by the distilled water added to the
MTE cell to displace the air in the system (see Section 2.9). In contrast, for SD, the
proportion of water-soluble sodium removed at temperatures <280°C was directly
proportional to the amount of water removed. However at processing temperatures
greater than 280°C, the proportion of sodium was less than the proportion of water.
As described in Section 3.5.1, the liquid water will be present in equilibrium with the
vapour phase and some soluble inorganic material may be leached out from the
lignite however it appears that with higher processing temperatures (above 280°C),
more water is being removed from the lignite as steam, thus reducing the ability for
sodium to be leached out from the lignite in liquid form. Furthermore, it is expected
that at temperatures above 250°C, significant decarboxylation occurs [130, 157] and that
additional sodium (previously in the form of a carboxylate) would be removed from
the lignite however as depicted in Figure 3.7, this is clearly not the case. This further
supports that at higher processing temperature, more water from the lignite is
removed as steam. In HTD, the proportion of sodium removed from the lignite was
significantly greater than the proportion of water removed (Figure 3.8). At 250°C,
99%db of the water-soluble sodium was removed. Increasing the HTD processing
beyond 250°C resulted in water-soluble sodium reductions of greater than 100% (up
to 140% for the 350°C sample). These very high sodium reductions can be attributed
to detachment of sodium – carboxylate bonds at the higher processing temperatures
and due to some washing of the lignite while in slurry form.
Chapter 3 Lignite drying technologies 84
3.5.4 Chlorine in dried lignite products Chlorine salts are very soluble in water. From the water washing experiments, less
than half of the chlorine was leached out from the lignite after repeated distilled
water washings (Table E.1). This suggests that a significant proportion of the
chlorine in Loy Yang lignite is strongly bound within the coal matrix. Chlorine anion
interactions within coals have been reported elsewhere [158-164].
Vassilev et al. [164] postulated that water-insoluble chlorine can be ionically and
covalently bound to the organic portion of the coal. Huggins and Huffman [165]
reported that the interaction between the maceral surface and the chlorine anion was
relatively strong however no evidence was found for any organic chlorine in the Loy
Yang lignite.
For MTE, the error associated in calculating the proportion of water-soluble chlorine
was high therefore no solid conclusions can be reached from the data given in
Figure 3.9. The change in water-extractable chlorine with temperature for HTD and
SD broadly parallelled in the three drying processes to that of sodium, in agreement
with the proposal that the water-soluble chlorine was predominantly in the form of
NaCl. For SD, above 280ºC there was no significant increase in the proportion of
chlorine removed with increasing processing temperature. However for HTD, the
proportion of water-soluble chlorine leached out was significantly greater than the
proportion of water removed suggesting that as for sodium, chlorine was washed out
of the lignite in the water added to make up the slurry for HTD.
Chapter 3 Lignite drying technologies 85
0 20 40 60 80 100 120 140 160 1800
20
40
60
80
100
120
140
160
180
% w
ater
ext
ract
able
ch
lorin
e re
mov
ed (d
b)
Proportion of water removed (%) HTD MTE SD
Figure 3.9 Relationship between the proportions of water and soluble chlorine
removed by MTE, HTD and SD processing.
The proportion of water-extractable chlorine versus the proportion of water-
extractable sodium is shown in Figure 3.9 for MTE, HTD and SD products. For
HTD, higher processing temperatures (eg >280°C) resulted in more than 100% of
the water-soluble chlorine being removed. This high proportion of water-soluble
chlorine was likely associated to some of the more strongly bounded chlorine anions,
which are not normally washed out with water, being removed from the lignite.
Similarly, SD at higher processing temperatures, chlorine anion dissociation could
account for the proportion of water-soluble chlorine removed also exceeding the
100% mark. It is expected that at these higher processing temperatures (eg 320°C),
some of the chlorine could be removed as a gas (HCl or Cl2) along with the steam. In
contrast, appreciable amounts of Na are not leached from the lignite when the water
is removed as steam instead of as a liquid [90].
Chapter 3 Lignite drying technologies 86
0 20 40 60 80 100 120 140 160 1800
20406080
100120140160180
% w
ater
-ext
ract
able
ch
lorin
e re
mov
ed (d
b)
HTD
0 20 40 60 80 100 1200
20
40
60
80
100
120
% water-extractable sodium removed (db)
SD
% w
ater
-ext
ract
able
ch
lorin
e re
mov
ed (d
b)
0 20 40 60 80 1000
20
40
60
80
100MTE
% w
ater
-ext
ract
able
ch
lorin
e re
mov
ed (d
b)
Figure 3.10 Relationship between the proportion of sodium and chlorine present in the dried solid products from the MTE, HTD and SD processes.
Analysis of the product water support the findings deduced from the solid products
(APPENDIX F). In addition, traces of sulphate in the product water suggested that a
small portion of the total sulphur was water-soluble. The MTE process was most
favourable in regards to the amount of water removed from the lignite and the low
concentration of organic matter being leached out into the product water. The SD
process also gave high moisture reductions but at higher temperatures than MTE and
Chapter 3 Lignite drying technologies 87
with significantly higher TOC in the product water. However for HTD, the fact that
significant moisture reductions can only be achieved at much higher processing
temperatures than MTE and SD and that its product water was highly concentrated
with organic matter is a major hindrance in its commercialisation as a drying process.
3.6 Conclusions The extent of water removal from MTE, HTD and SD processing of the same sample
of Loy Yang lignite was conducted. For comparative purposes, the effectiveness of
each drying process can be evaluated by examining the parameter at which 50%
water removal is achieved. For HTD, the processing temperature had to be increased
to 320°C to achieve a 50% moisture reduction. At this high temperature, the
proportion of mass recovered was only 86%db because of decarboxylation and
devolatilisation reactions. Under these conditions, the net wet specific energy
(NWSE) of the HTD product had increased by 57% when compared to the raw
lignite (i.e. 8.3 MJ kg-1 to 13.1 MJ kg-1). In contrast, for MTE, the processing
temperature required to achieve high moisture reductions was significantly reduced
by simultaneous application of mechanical pressure. For Loy Yang lignite the
optimum applied pressure was identified as approximately 5.1MPa, which is
relatively mild when compared to the Niederaussem MTE plant. Further increases
had relatively little effect on the moisture content of the product pellet. At 125°C and
5.1MPa of applied mechanical force, the moisture content was reduced by 50%
compared to the parent lignite (NWSE 13.4 MJ kg-1). Similarly, at 150°C and
2.5MPa applied mechanical force also resulted in a 50% moisture reduction (NWSE
15.5 MJ kg-1). Under both MTE processing conditions, lignite mass recoveries were
very high (>99%db).
Chapter 3 Lignite drying technologies 88
In SD, the temperature necessary to remove half of the water from the parent lignite
was significantly lower than HTD but higher than MTE. A SD temperature of
approximately 215°C was necessary to achieve a 50% moisture reduction. The
lignite mass recovery in steam drying at 215°C was also high at 98%db. The NWSE
for the 230°C SD product was 14.4 MJ kg-1, an increase of 74% compared to the raw
lignite.
The internal pore structure of thermally dried products was characterised using
mercury porosimetry, surface area gas adsorption and helium pycnometry. The MTE
process was more effective in reducing the large pore volume of the products than
HTD and SD. An increase in mechanical pressure up to 12MPa increased the
micropore volume, which was likely attributed to the compression of larger pores
into the micropore region. Furthermore, additional increases in mechanical pressure
beyond 12MPa resulted in a decline in the micropore volume, which was probably
due to the micropores themselves being compressed under these high pressures. In
addition, the micropore volume and the carbon density of MTE, HTD and SD
decreased at higher processing temperatures. The micropore volume being small,
these changes did not greatly affect the overall internal pore volume.
The retained moisture in the MTE pellet correlated well with the dry density however
the relationship was not equally proportional on a gram per gram basis. Similarly to
MTE, the volume of retained water in the HTD products exceeded the available dry
pore volume. However for SD products, the pore volume and retained water did not
Chapter 3 Lignite drying technologies 89
follow a linear relationship thus suggesting the water retained is not related to the
product’s pore volume.
Much of the sodium removed from drying was in the form of NaCl. In MTE,
processing temperature was more effective in removing sodium than applied pressure
(within the experimental parameters tested). In contrast, SD processing temperatures
above 280°C reduced the ability of sodium to be leached out from the lignite.
Furthermore, the lignite slurry used in HTD facilitated significantly higher levels of
sodium, magnesium, calcium and chlorine to be removed from the process compared
to MTE and SD. This considerably higher removal of inorganic material, in
particularly sodium, is an advantage to the process in regards to reducing fouling and
slagging propensities during combustion.
The detachment of carboxylate bonds from the lignite with increasing processing
temperature could explain the higher than expected inorganic matter being removed
from the lignite during drying. Furthermore, at higher processing temperatures, some
of the more strongly bounded chlorine anions (not associated with the sodium ion)
were removed.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 90
CHAPTER 4 PYROLYSIS OF RAW LIGNITE AND
DRIED PRODUCTS
4.1 Introduction Coal pyrolysis is a very complex process and plays a part in nearly all coal
conversion processes, such as combustion, gasification and liquefaction. The word
“pyrolysis” is of Greek origin and means “to decompose by heat”. In simple terms,
pyrolysis is the thermal fission of a sample in the absence of oxygen into molecules
of lower mass. Most coal structures consist of structural units joined together by
either weak or strong linkages to form a three-dimensional macromolecular network.
During pyrolysis, the weakest bridges break, producing molecular fragments that are
released as tar, light hydrocarbons and non-organic gases (carbon monoxide, carbon
dioxide, water etc).
In this chapter, several pyrolysis techniques are utilised to investigate the pyrolysis
behaviour of the Loy Yang lignite and its thermally treated products. The slow
pyrolysis of the lignite/lignite products was investigated in two systems; a
thermogravimetric analysis system and a quartz fluidized-bed/fixed-bed reactor
system. The flash pyrolysis of the lignite/lignite products was also investigated in
two systems; the quartz reactor system used for the slow pyrolysis experiments and a
pyrolysis-gas chromatography system. The quantification and characterisation of
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 91
volatised components from the pyrolysis of the lignite and its thermally treated
products as a function of temperature is also described.
4.2 Proximate analysis Thermogravimetric analysis (TGA) has been extensively used as a tool for
investigating the pyrolysis mechanisms of samples over a range of gases and
pressures. A number of groups have also used TGA for determining proximate
analysis in both bituminous and low-rank coals [166-173].
A typical TGA thermogram that was used to determine the proximate analysis of the
raw lignite and its dried products is shown in Figure 4.1. The procedure followed has
been detailed by Earnest [174] and is fully described in Section 2.15. Furthermore, the
TGA thermograms for the MTE, HTD and SD products are given in Appendix G.
The TGA thermograms show that some mass was lost when the lignite sample was
heated to 110°C in an inert atmosphere (Ar), predominantly due to the loss of
reabsorbed water from the oven dried lignite sample. A second, larger weight loss
(due to the loss of volatiles) was observed when the sample was heated at 50°Cmin-1
to 950°C and maintained at this temperature for 15min. After the 15min, the gas
atmosphere was changed from argon to air, allowing the sample to burn until a
constant weight was obtained. The remaining sample weight represented the ash
yield of the lignite. The ash, volatile and fixed carbon yields, determined from the
proximate analysis, for the raw lignite, HTD, MTE and SD products are given in
Table 4.1.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 92
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
Time (seconds)
Wei
ght L
oss
(%)
0
100
200
300
400
500
600
700
800
900
1000
0 600 1200 1800 2400 3000 3600 4200 4800 5400
Temperature (ºC
)
Weight Loss (%)Temperature
-50.11 % : Volatiles
-6.30 % : Water
-48.96 % (db)
Air
-50.14 % (db)
-48.92 % : Fixed Carbon
-0.89 % : Ash-0.90 % (db)
Figure 4.1 TGA thermograms used in determining proximate analysis of a lignite sample
Within the limits of error, processing temperatures beyond 250°C gave significant
reductions in volatile yield for the HTD and the MTE products when compared to the
raw lignite whereas, significant reductions in the SD products were noted at
temperatures beyond 230°C. For HTD and MTE, proximate analysis on a 230°C
treated product was not conducted but it is expected to be similar to the SD product
at 230°C. Furthermore, the effect of applied mechanical pressure at 150°C in MTE
had a negligible effect on the volatile yield of its products (Table 4.1).
The devolatilisation rate was significantly increased at temperatures beyond 250°C
(Figure 4.2). It is expected that an increase in processing temperature will lead to an
increase in C, accompanied by a reduction in O and volatile matter, an elimination of
CO2 via decarboxylation of the lignite structure [175] and the removal of some small
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 93
molecular weight hydrocarbons in the product water stream [144, 176]. Furthermore, the
higher volatile yield in the HTD products compared to the SD products at processing
temperatures beyond 250°C could be attributed to enhanced entrapment of mobile
waxes/tars etc, which are part of the volatile fraction in the HTD product. These
mobile volatile constituents are likely to re-solidify on the surface of the HTD
product during cooling however for the SD products, these fractions are likely to be
volatilised from the product and re-condensed into the product water. This is also
supported by the higher organic carbon concentrations measured for the SD product
waters compared to HTD product waters (see Table F.2).
In Chapter 3, it was concluded that for HTD, a temperature of 320°C was necessary
to achieve a 50% moisture reduction but at the expense of a significantly lower mass
recovery. For HTD at 320°C, this lower mass recovery corresponds to ~6wt%daf
volatile yield loss in the final product (Table 4.1). This significant volatile yield loss
could be disadvantageous in some industrial processes, which convert the carbon
matter to lower molecular weight fractions (eg gasification, liquefaction).
In contrast, for SD, an estimated temperature of 215°C would dedicate a 50%
moisture reduction however, advantageously, without the sacrifice in volatile yield
(Table 4.1). Similarly, the operating temperature in the MTE process was relatively
low and subsequently the volatile yield in the final products was not affected
(Table 4.1).
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 94
Table 4.1 Proximate analysis of raw lignite, HTD, MTE and SD products Ash
isomerisation, molecular rearrangement, hydrogen transfer etc. [184]. These reactions
are often concurrent and produce opposite thermal effects [184, 185].
Comparison between the two SD products reveal a significant decrease in the
amplitude of the first exothermic peak for the 350°C SD product when compared to
the 130°C SD product (see Figure 4.3 for numbering of the peaks). Elder and
Harris [186] reported that primary carbonisation commences at ~350°C with the initial
release of carbon dioxide and hydrogen. This suggests that for the 350°C SD product,
some carbonisation had occurred during the SD process which can account for the
lower amplitude in the first exothermic peak. All the other heat flow peaks within
this pyrolysis period (as numbered in Figure 4.3) are very similar between the two
SD products.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 97
Note: Numbers on the convoluted peaks in
A deconvoluted plot of the heat flow curve (in the region 250°C to 950°C) for the
130°C SD product is shown in Figure 4.4. Much research has been conducted in
attempt to relate the pyrolysis behaviour of carbonaceous materials at different
temperatures. It is generally accepted that the weight loss of peat and lignite below
350°C is the result of decarboxylation of acid groups , the dehydration of
hydroxylated aliphatic structures and the generation of low molecular
weight alcohols . Francioso et al. [182] reported the first exothermal peak generated
below 350°C was due to these effects.
The second exothermic peak (Figure 4.4 and Figure 4.3) has been previously linked
with polymerisation and molecular rearrangement [184]. Similarly, Ding et al. [183]
convoluted peaks in
A deconvoluted plot of the heat flow curve (in the region 250°C to 950°C) for the
130°C SD product is shown in Figure 4.4. Much research has been conducted in
attempt to relate the pyrolysis behaviour of carbonaceous materials at different
temperatures. It is generally accepted that the weight loss of peat and lignite below
350°C is the result of decarboxylation of acid groups , the dehydration of
hydroxylated aliphatic structures and the generation of low molecular
weight alcohols . Francioso et al. [182] reported the first exothermal peak generated
below 350°C was due to these effects.
The second exothermic peak (Figure 4.4 and Figure 4.3) has been previously linked
with polymerisation and molecular rearrangement [184]. Similarly, Ding et al. [183]
-20
-10
0
10
20
30
40
50
60
70
80
90
100
Time (seconds)
Hea
t Flo
w (m
W)
0
100
200
300
400
500
600
700
800
900
1000
0 600 1200 1800 2400 3000 3600 4200 4800
Tem
pera
ture
(ºC
)
Steam dried 130°CSteam dried 350°CTemperature
2
3
4
1
Figure 4.3 Comparisons between the heat flow of 130°C and 350°C SD products, 30mLmin-1air flow, 30.0mg sample.
heat flow curves correspond to the de deFigure 4.4. Figure 4.4.
[182, 187, 188]
[182, 187, 188]
[187]
[182, 187, 188]
[182, 187, 188]
[187]
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 98
reported that exothermic peaks >450°C are due to condensation and carbonization
reactions during pyrolysis.
1200 1400 1600 1800 2000 2200 2400 2600 2800
-10
0
10
20
30
40
43
2
1
Time (seconds)
Hea
t flo
w (m
W)
200
300
400
500
600
700
800
900
1000
Tem
pera
ture
(ºC
)
Figure 4.4 Deconvoluted curve fit of the DSC from the 130°C SD product. Black line shows the raw DSC curve, green lines are the peaks identified from devolution and
the red line is the sum of the green peaks. The temperature profile is in blue.
Exothermic peaks 3 and 4 in Figure 4.4 are more likely the sum of concurrent
xothermic and endothermic reactions during pyrolysis. Martinez-Alonso et al. e
ported that devolatilisation and cracking reactions are endothermic and that
[184]
re
polymerization and condensation reactions are exothermic. The effects of rapid
decomposition of the organic mass (exothermic) overlaps the volatile evolution
effects (endothermic) [189]. These effects could not be distinguished from the heat
flow curves.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 99
4.3.2 Combustion DSC at 950°C of SD products Switching the gas stream from argon to air in the TGA marked the commencement
of the combustion stage. Comparing the low temperature (130°C) and high
temperature (350°C) SD products in Figure 4.3, show similar exothermic behavioural
combustion phase. This spike is attributed to the residual carbon in the sample
igniting (see Appendix G). The variables that affect combustion and ignition are
further discussed in Chapter 5. Interestingly, a higher SD processing temperature
resulted in longer periods to complete the combustion of the sample (Figure 4.3).
This longer combustion period is likely attributed to the higher char to volatile ratio
for the 350°C SD product as a result of its treatment at the higher processing
temperature (see Table 4.1). That is, after devolatilisation, the SD products that had
subsequently, would require longer periods to completely combust.
4.4 Quartz reactor pyrolysis experiments
raw lignite and
e thermally treated products as a function of temperature and heating rates. Similar,
n applied to measure the
volatile yields of carbonaceous materials under different heating profiles and
pyrolysis temperatures [190]. The overall behaviour of the raw lignite and the
thermally treated products as a function of pyrolysis temperature and heat-up mode is
discussed in this section.
patterns. Both products were found to increase the rate of heat given off with
increasing time followed by a large exothermic spike, which marked the end of the
been treated at higher temperatures contain a higher proportion of fixed carbon and
Pyrolysis experiments were conducted in a fluidized-bed/fixed bed quartz reactor for
the purpose of identifying differences in volatile yield between the
th
fluidized-bed/fixed bed quartz reactor systems have also bee
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 100
4.4.1 Slow pyrolysis experiments The slow pyrolysis refers to the experiments where the lignite sample was loaded
into the fluidised-bed/fixed reactor and then heated at 40-50°C min to the indicated
temperature with a 15min
-1
holding time. Char yields produced from the slow heat-up
pyrolysis of the raw lignite and the thermally treated products as a function of
increasing pyrolysis temperature were discussed in the previous sections of this
chapter from the TGA-DSC curves (see Section 4.3). In this section, slow heat-up
pyrolysis of the raw lignite and thermally treated products is re-examined using a
quartz reactor system. A major difference between pyrolysis in the TGA and
pyrolysis in the quartz reactor system is the experimental sample size (i.e. in the
quartz reactor 100 times more sample was pyrolysed). Also, with the quartz reactor,
the char yield (and hence volatile yield) at each incremental temperature increase is
examined and compared to the thermally treated products.
The data in Figure 4.5 indicate that for the raw lignite, the majority of the sample
weight loss took place at temperatures lower than 400°C with a volatile yield of more
°C to 600°C resulted in
n additional volatile yield of 18wt%db. The primary pyrolysis products evolved
[188]
[191]
temperature is shown in Figure 4.5. The physicochemical changes to the lignite with
than 20wt%db. Increasing the pyrolysis temperature from 400
a
below 600°C for lignites are water and carbon dioxide . Murray reported that for
a Latrobe Valley lignite, 70% of the carboxyl groups were lost by 300°C and 92% by
600°C . Further increases in temperature to 900°C at the slow heating rate
resulted in less than 7wt%db additional weight loss for the raw lignite.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products
101
the raw lignite (Figure 4.5b).
all difference is attributed to some volatiles having been lost (approximately
, gave the largest difference in volatile yield when compared to the raw
E and SD products (Figure 4.5c). This difference is primarily due to the
The char yield profile of the MTE 150°C / 5.1MPa treated product as a function of
temperature did not show any significant difference compared to the raw lignite
(Figure 4.5a). These similarities support the TGA proximate analysis results which
also showed no significant difference in char yield between the raw lignite and the
MTE 150°C / 5.1MPa treated product (Table 4.1).
For the SD 230°C product, the char yield profile from slow heat-up pyrolysis as a
function of temperature was marginally different to
This sm
2wt%db) during the SD treatment process (see Section 3.3; Table 3.4). The quartz
reactor slow pyrolysis char yield results also support the TGA proximate results in
Table 4.1.
The char yield profile for the HTD 320°C product as a function of slow pyrolysis
temperature
lignite, MT
significant amount of volatiles lost (Section 3.1) during the hydrothermal dewatering
process at 320°C.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products
45
50
55
60
65
70
75
80
85
350 450 550 650 750 850 950
Pyrolysis temperature (ºC)
Cha
r yie
ld (%
daf)
LYLA raw lignite; slow heat up
MTE 150ºC / 5.1MPa; slow heat up
45
50
55
60
65
70
75
80
85
350 450 550 650 750 850 950
Pyrolysis temperature (ºC)
Cha
r yie
ld (%
daf)
LYLA raw lignite; slow heat up
SD 230ºC; slow heat up
45
50
55
60
65
70
75
80
85
350 450 550 650 750 850 950
Pyrolysis temperature (ºC)
Cha
r yie
ld (%
daf)
LYLA raw lignite; slow heat up
HTD 320ºC; slow heat up
a b c
102
Figure 4.5 Comparisons of char yields from the pyrolysis of raw LYLA lignite with (a) MTE (b) SD (c) HTD treated products as a function of temperature in the fluidized-bed/fixed-bed reactor operated at the slow heating rate mode.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products
103
Furthermore, the char yield profile as a function of temperature of the raw lignite
versus the water-washed lignite did not show in any notable differences (Figure 4.6).
45
50
55
60
65
70
75
80
85
350 450 550 650 750 850 950
Pyrolysis temperature (ºC)
Cha
r yie
ld (%
daf)
LYLA raw lignite; slow heat up LYLA water washed; slow heat up
Figure 4.6 Comparisons of char yields from the pyrolysis of raw and water washed LYLA lignite operated at the slow heating mode, as a function of temperature in the
fluidized-bed/fixed-bed reactor.
4.4.2 Fast pyrolysis experiments In Section 4.2, proximate analysis determinations were performed using a TGA with
a slow heating rate of 50°C min-1. A limitation to the TGA is that very fast heat up
rates (eg. > 100°C s-1) cannot be achieved. Furthermore, in slow heating rate
experiments, most of the H2O and CO2 are released from the lignite by the time the
reactor has reached 500°C [130]. That is, the presence of H2O and CO2 at elevated
temperatures (>700°C) could facilitate in the gasification of the char [192]. Instead,
fast heat up rates (i.e. flash pyrolysis) of lignite particles, can be accomplished with a
Chapter 4 Pyrolysis of Raw Lignite and Dried Products
104
fluidized-bed/fixed-bed quartz reactor system. Similar quartz reactors [113, 121] heated
in an external furnace have reported rapid particle heating rates in the order of
104°C s-1.
Pyrolysis at the fast heating rate and slow heating rate showed no significant
differences in the char yield profile as a function of temperature for the raw lignite
(Figure 4.7). In addition, no significant differences in the char yield profile were
found between the slow heating rate and fast heating rate for the MTE, SD and HTD
products (Figure 4.8).
45
50
55
60
65
70
75
80
85
350 450 550 650 750 850 950
Pyrolysis temperature (ºC)
Cha
r yie
ld (%
daf)
LYLA raw lignite slow heat up LYLA raw lignite fast heat up
Figure 4.7 Comparisons of char yields from the pyrolysis of raw LYLA lignite operated at the slow and fast heating rate mode, as a function of temperature in the
fluidized-bed/fixed-bed reactor.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products
105
Similarly, Zeng et al. [193] reported that varying the pyrolysis heating rate had a
marginal effect on the char yield of Loy Yang lignite when pyrolysed to 900°C in a
wire mesh reactor. Furthermore, Takarada et al. [194] reported that the coexistence of
steam at ~700°C, which is removed during the slow heat up experiments, did not
cause significant additional volatile yield during the pyrolysis/gasification of
Yallourn lignite. In contrast, pyrolysis of biomass conducted by Kweon et al. [190]
using a similar fluidized bed/fixed-bed reactor to this study, reported that
temperatures greater than 700°C, the fast heating rate gave noticeably lower char
yields than the corresponding slow heating rate experiments. The differences in
volatile/char yields with biomass were explained by the onset of in situ reforming of
the char from the inherent biomass AAEM species. The data in Figure 4.7 and Figure
4.8 suggest that in situ reforming of Loy Yang lignite and of the thermally dried
products is negligible as a function of pyrolysis temperature. Furthermore, the Loy
Yang lignites used in this study and by Zeng et al. [193] had low AAEM contents
which may account for the negligible in situ reforming of the char during flash
pyrolysis.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products
45
50
55
60
65
70
75
80
85
350 450 550 650 750 850 950
Pyrolysis temperature (ºC)
Cha
r yie
ld (%
daf)
MTE 150ºC / 5.1MPa; slow heat up
MTE 150ºC / 5.1MPa; fast heat up
45
50
55
60
65
70
75
80
85
350 450 550 650 750 850 950
Pyrolysis temperature (ºC)
Cha
r yie
ld (%
daf)
SD 230ºC; slow heat up
SD 230ºC; fast heat up
45
50
55
60
65
70
75
80
85
350 450 550 650 750 850 950
Pyrolysis temperature (ºC)
Cha
r yie
ld (%
daf)
HTD 320ºC; slow heat up
HTD 320ºC; fast heat up
a b c
106
Figure 4.8 Comparisons of char yields from the pyrolysis of (a) MTE (b) SD (c) HTD treated products as a function of temperature in the fluidized-bed/fixed-bed reactor operated at the slow and fast heating rate mode
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 107
4.5 Tar yields For many coals, tar represents the major initial volatile species released during
pyrolysis. Previous works have reported that the yields and nature of tars depend not
only on coal type but also on pyrolysis conditions, including particle heating rate,
reactor residence time, nature of gaseous atmosphere and pressure [85, 195-199] . Work
with Loy Yang lignite have also reported that high heating rate favours the release of
(larger) aromatic ring systems during pyrolysis [199-201].
The tar yield produced from the pyrolysis of raw LYLA lignite operated at the slow
and fast heating rate mode, as a function of temperature is shown (Figure 4.9). Tar
from coal pyrolysis is normally released at relatively low temperatures [121]. For the
LYLA lignite, pyrolysis at 500°C produced 5wt%db of tar in both the slow and fast
heating rate modes (Figure 4.9). Increasing the pyrolysis temperature to 600°C
resulted in an additional ~18wt%db tar yield. Significant differences in the heating
modes were only found at pyrolysis temperatures above 700°C. For the slow heating
mode, the tar yield marginally increased from 700°C and 900°C, whereas for the fast
heating mode, the tar yield decreased substantially from 18wt%db to 2wt%db,
respectively. These differences in tar yields between the slow and fast heating modes
for pyrolysis temperatures above 700°C is likely attributed to secondary vapour-
phase reactions resulting in the formation of hydrocarbon gases and modified tar [121].
Mochida et al. [202] suggested that the volatile fractions can modify the pyrolysis
process because of its ability to act as a solvent for the system and/or as a hydrogen
donor. Similarly, Bermejo et al. [203] also reported that low molecular weight
components can modify the rheological properties of the tar (pitch) during pyrolysis.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 108
In relation to the fast heating experiments (Figure 4.9), the coexistence of
devolatilised low molecular weight components and tar in the same reactor system
could potentially facilitate in the breakdown of the tar at pyrolysis temperatures
above 700°C. In contrast, for the slow heating rate experiments, most of the low
molecular weight components devolatilised from the lignite, are carried out of the
reactor and subsequently cannot partake in any secondary vapour-phase reactions
with the tar at the higher pyrolysis temperatures. Furthermore, the volatilisation of
alkali and alkaline earth metal (AAEM) species at pyrolysis temperatures above
700°C could also facilitate in the catalytic breakdown of the tars. The behaviour of
AAEM species during pyrolysis for the slow and fast heating rate modes is discussed
in Section 4.6.
500 600 700 800 9000
5
10
15
20
25
30
Tar y
ield
wt%
of c
oal (
db)
Pyrolysis temperature (°C)
Slow heatup Fast heatup
Figure 4.9 Comparisons of tar yields from the pyrolysis of raw LYLA lignite
operated at the slow and fast heating rate mode, as a function of temperature in the fluidized-bed/fixed-bed reactor.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 109
4.6 Inorganic contents Pyrolysis of coal is considered the initial steps of all thermochemical utilisation
processes such as combustion, gasification, carbonisation and liquefaction. The
release of volatile inorganic species in pyrolysis is of significant importance to these
power generation processes. In conventional coal combustion systems, the inorganic
components in the lignite are largely responsible for slagging and fouling [85] whereas
in advanced gasification/reforming based power generation systems, the volatilised
inorganic components are also one of the major causes the erosion and corrosion of
the gas turbine components [94, 95].
The behaviour of sodium, calcium, magnesium and chlorine from the pyrolysis of
raw LYLA lignite operated at the slow and fast heating rate mode, as a function of
temperature is shown in Figure 4.10. The behaviour of sodium, calcium, magnesium
and chlorine was also conducted on the pyrolysed thermally dried products (except
for HTD because of its low inherent inorganic content after the process) and the
overall trends between the samples were found to be relatively similar. Subsequently,
to simplify the discussion in the behaviour of these ionic species only the raw lignite
will be discussed in this thesis.
For sodium, pyrolysis at 500°C resulted in 5wt%db loss from the original lignite and
an increase in pyrolysis temperature to 600°C resulted in an additional ~12wt%db
volatilisation in both the slow and fast heating rate modes. The volatilisation of
sodium at temperatures up to 600°C from a Loy Yang lignite has been attributed to
the release of low molecular mass carboxylates during pyrolysis [96, 98, 100]. The
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 110
pyrolysis heating rate mode only showed significant differences in the volatilised
sodium at temperatures of 700°C and above. In the slow heating mode, the sodium
volatilisation marginally increased by 5wt%db from 700°C to 900°C, whereas, in the
fast heating more than half of the original sodium in the lignite was volatilized at
700°C and at 900°C, negligible sodium was detected in the remaining char.
The volatilisation of sodium at pyrolysis temperatures of 700°C and above, in the
fast heating rate mode (Figure 4.10), appears to be correlated to the catalytic
breakdown of the tars (Figure 4.9). The cations in Latrobe Valley lignites are known
to play an important role in the final tar yields during rapid pyrolysis [113, 201]. Tyler
and Schafer flash pyrolysed lignites in a fluidised bed reactor and reported that the
removal of cations present in lignites markedly increased the tar yield [114].
Furthermore, the volatilisation of sodium from Latrobe Valley lignites at
temperatures higher than 700°C have reported to be linked to volatile-char
interactions causing the reforming/cracking of volatiles on the char surface [97, 204]
and also causing the condensation of aromatic ring structures in the char [205, 206]. The
tar (precursors) produced from Loy Yang lignites are highly aliphatic which, if
retained in the solid phase, crack to form mainly gases during pyrolysis [207].
In contrast, in the slow heating mode, the small loss of sodium from 700°C to 900°C
did not result in proportional reductions in tar yield. A likely explanation is that most
of the volatilised sodium in the slow heating rate mode was carried out of the reactor
and subsequently could not partake in any secondary catalytic vapour-phase
reactions with the tar at the higher pyrolysis temperatures.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 111
On the contrary to sodium, calcium and magnesium volatilisation was not affected by
differences in pyrolysis heating modes (Figure 4.10). The majority of volatised
calcium and magnesium had occurred at 500°C with losses of 7wt%db and
13wt%db, respectively, and further increases in pyrolysis temperatures only
marginally increased the volatilisation of these cations. A number of researchers
have attributed the reduction in the tar yield by divalent cations to the cross-linking
effects of the divalent ions, bringing the (carboxylic) groups closer in the coal
structure. These marginal increases in volatile calcium and magnesium at pyrolysis
temperatures greater than 600°C (Figure 4.10) cannot account for the significant
reductions in the tar yields, in particularly when the proportion of calcium and
magnesium retained in the char did not change between the different heating rate
modes. Wu et al. [97] also reported that volatilisation of calcium and magnesium had
little effect on radical-char interactions during the flash pyrolysis of a Latrobe Valley
lignite in a fluidised bed reactor.
The volatilisation behaviour of chlorine followed an entirely different trend to the
volatilisation of sodium in both heating rate modes and at the different pyrolysis
temperatures (Figure 4.10). At 500°C, more than half of the initial chlorine in the
lignite was volatilised whereas under the same corresponding conditions, the
volatilisation of sodium was minimal at 5wt%db. Prior work with Loy Yang lignite
has found that about 10% of the chlorine in a NaCl-loaded sample was volatilised at
temperatures as low as 200°C [100].
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 112
In the slow heating mode, the chlorine volatilisation remained relatively unchanged
from 500°C to 700°C whereas from 700°C to 900°C, an additional 25wt%db of the
chlorine was volatilised. At 900°C, about 80% of the initial chlorine in the lignite
was volatilised whereas under the same corresponding conditions, only 30wt%db of
sodium was removed from the char.
In contrast, the chlorine volatilisation trend in the fast heating mode was opposite to
chlorine behaviour in the slow heating mode and also contrastingly different to the
trend reported for sodium in the fast heating mode (Figure 4.10). An increase in
pyrolysis temperature above 500°C in the fast heating mode gave significantly less
volatilised chlorine. At 700°C, 30wt%db of the chlorine was volatilised which
corresponds to 25wt%db less volatilised chlorine than at 500°C. An increase in
pyrolysis temperature above 700°C in the fast heating mode did not result in any
further changes in the volatilised chlorine. The significantly different behaviours
between the volatilisation of chlorine and that of sodium clearly suggests that not all
of the sodium and chlorine, if any, were volatilised as NaCl molecules. In addition,
no significant increases in sodium or chlorine volatilisation were found around the
melting point of NaCl (800.7°C) which further supports the conclusion that sodium
and chlorine did not volatilise as NaCl. Similar conclusions have also been reported
elsewhere [100].
The decrease in the volatilisation of chlorine in the fast heating mode over the
temperature range from 500°C to 700°C is likely attributed to the recombination of
chlorine containing species such as HCl, with the nascent char. Quyn et al. [100]
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 113
postulated that the retention of chlorine coincided with the onset of massive bond
breakages in the pyrolysing coal/char matrix and that some forms of chlorine could
have recombined with some of the newly formed free radical sites inside the char at
high temperatures. Whereas, in the slow heating mode, the chlorine containing
species were swept out of the reactor and subsequently could not react with the char
in the reactor at high temperature.
500 600 700 800 900
0
20
40
60
80
100
500 600 700 800 900
0
20
40
60
80
100
500 600 700 800 900
0
20
40
60
80
100
dc
a b
Na
vola
tilis
atio
n, %
Na
in li
gnite
Temperature (°C)
Slow heatup Fast heatup
Cl v
olat
ilisa
tion,
%C
l in
ligni
te
Temperature (°C)
Slow heatup Fast heatup
Mg
vola
tilis
atio
n, %
Mg
in li
gnite
Temperature (°C)
Slow heatup Fast heatup
500 600 700 800 900
0
20
40
60
80
100
Ca
vola
tilis
atio
n, %
Ca
in li
gnite
Temperature (°C)
Slow heatup Fast heatup
Figure 4.10 Comparisons of volatilized (a) sodium, (b) chlorine, (c) magnesium and (d) calcium, from the pyrolysis of raw LYLA lignite operated at the slow and fast
heating rate mode, as a function of temperature in the fluidized-bed/fixed-bed reactor
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 114
4.7 Pyrolysis-gas chromatography Analytical pyrolysis methods including pyrolysis-mass spectrometry (py-ms),
pyrolysis-gas chromatography (py-gc) and pyrolysis-gas chromatography / mass
spectrometry (py-gc-ms) have been commonly used to characterize both low-rank
lignites [147, 208, 209] and high-rank coals [210, 211] and also in characterising the
pyrolysed products from coal derived fractions such as coal macerals [211, 212] and
humic acids [147, 213, 214]. The fragments volatilised from coal pyrolysis are separated
by gas chromatography. The obtained pyrogram constitutes a fingerprint of the
starting macromolecule and gives information on the relative amount of its
monomeric components [215]. The pyrolysis-gas chromatography of the raw Loy
Yang lignite at temperatures of 600°C, 700°C, 800°C and 900°C is shown in
Figure 4.11.
Curves for the individual hydrocarbon gas yields (methane (CH4), ethane (C2H6),
ethene (C2H4), ethylene (C2H2), propane (C3H8), propene (C3H6), water (H2O),
carbon monoxide (CO) and carbon dioxide (CO2) from the pyrolysis of Loy Yang
lignite are shown in Figure 4.12. The two major hydrocarbon components volatilised
during the pyrolysis the raw Loy Yang lignite were methane and ethene.
Furthermore, the yields of both methane and ethene increased as a function of
pyrolysis temperature from 600°C to 900°C. At 900°C, more than 6wt%db of
methane and about 4wt%db of ethene was volatilised from the raw lignite.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 115
C H3 8
500000
0
xylenes
toluene
phenol
C5
C6
C H3 8
C H3 6
benzene
benzene
8000000
4000000
0
Abun
danc
e
600 C
10 20 30 40 50 60 70 80
500000
1000000
0
6000000
3000000
0
80000
0
C6
C7toluene
xylenes
phenol
C5 C7
C8
C4
C H3 8
C H3 6
benzene
300000
0
toluenexylenes
phenol
C5C7
C8
C6
C H3 6
C H3 8
5000000
2500000
0
0
toluene
phenolC5 xylenesC H2 2
C H3 6
benzene
600000
Abun
danc
eAb
unda
nce
Abun
danc
e
Time (min)
900 C
800 C
700 C
Figure 4.11 Gc trace of pyrolysed Loy Yang raw lignite using a GS-GasPro column
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 116
propane, propene (c) carbon monoxide, carbon dioxide and water, from the py-gc of Loy Yang raw lignite as a function of temperature using a GS-GasPro column.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 117
The increase in yield of methane with increasing pyrolysis temperature is likely
attributed to the thermal cracking of the char [216], tar and volatilised hydrocarbon
gases. An example of the thermal cracking process can be illustrated with ethane
(Equation 4.1 to Equation 4.3).
C2H6 → CH3● + CH3
●Equation 4.1
C2H6 + CH3●→ C2H5
● + CH4 Equation 4.2
CH3● + H●→ CH4 Equation 4.3
Similarly, ethene, propene and propylene increased yields with increasing pyrolysis
temperature while their correspondent saturated hydrocarbons, ethane and propane,
reached maximum yields at 700°C and 800°C, respectively, before declining with
further increases in pyrolysis temperatures (Figure 4.12). The decline in yields of
ethane and propane and the subsequent increases in correspondent olefin yields are
likely attributed to concurrent chemical reforming processes at the higher pyrolysis
temperatures. The reforming of alkanes into olefins is well documented in the
literature and can occur via numerous different reaction pathways including steam
cracking [217, 218], catalytic dehydrogenation [219-223] or oxidative dehydrogenation [223-
231], oxidative coupling [232, 233] or conversion into aromatic hydrocarbons (that is, for
alkanes greater than ethane) [234-236]. The light hydrocarbons produced during rapid
pyrolysis of the lignite are aliphatic in nature [206, 207, 237, 238] and prone to thermal
cracking at elevated temperatures. At the higher pyrolysis temperatures, significant
amounts of free radicals (especially H radicals) would be generated from the mass
bond breaking of the pyrolysing char. The generation of these free radicals with the
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 118
char would facilitate numerous different parallel-consecutive reforming/cracking
reaction pathways with the volatile species (Figure 4.13).
Catalytic and thermal cracking
+ H O (Steam reforming)2
+ H (Hydroreforming, hydrocracking)2
+ CO (Dry reforming)2
TARCO
H2
CH4
CO2
Figure 4.13 Illustration of the concurrent primary and secondary pyrolysis reactions
in the breakdown of tar
The volatilisation of sodium during the rapid pyrolysis of the Loy Yang lignite at
temperatures above 700°C (Section 4.6) is also likely to play an important role in the
reforming/cracking of volatiles [113, 201]. Machocki and Denis reported that Na/CaO
catalysed the reforming of ethane [233]. Several authors have also postulated that the
sodium present in the lignite facilitated the catalytic reforming/cracking of volatiles
on the char surface [201, 239]. In addition to sodium, the presence of calcium,
magnesium, aluminium and iron on the surface of the char may also facilitate the
catalytic cracking of the tar and the reforming of saturated hydrocarbons at
temperatures above 700°C. Wornat and Nelson [240] pyrolysed raw and calcium
exchanged Yallourn lignite in a fluidised bed reactor and reported that the tar yields
were lower for the calcium form sample than the raw lignite, indicating that calcium
catalysed the conversion of tar. In addition, these authors reported that the catalytic
influences of calcium affected the yields of aromatic hydrogen and unsaturated
hydrocarbon substituents in the tar compared to the raw lignite. Vernaglia et al. [241]
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 119
also found that the aromatic/aliphatic composition of tar to change with the type of
cation (H, Na, and Ca).
The rapid pyrolysis of the Loy Yang lignite also produced significant yields of
carbon monoxide, carbon dioxide and water (Figure 4.12). For carbon dioxide, the
yield reached a maximum of more than 9wt%db at 700°C and further increases in the
pyrolysis temperature resulted in a marginal decline in yield. In contrast, the carbon
monoxide and water yields from the pyrolysis of the lignite increased with increasing
temperature. At 900°C, more than 11wt%db of carbon monoxide and more than
4wt%db of water was volatilised from the raw lignite.
Prior to pyrolysis, all of the water from the raw lignite was removed and therefore
the water measured (Figure 4.12) was generated from the pyrolysis of the lignite.
Doolan and Mackie [242] proposed that phenols decompose to produce carbon
monoxide, via a free radical route involving phenyl and hydroxyl radicals, with the
phenolic oxygen finishing up as a water molecule. Also, Shin et al. [243] pyrolysed
catechols and hydroquinones, which are abundant in lignites [244], and reported that
these model compounds eventually converted to carbon monoxide, carbon dioxide
and water and suggested the catalytic cracking reaction pathway shown in
Figure 4.14.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 120
Figure 4.14 Shin et al.’s [243] proposed catalytic cracking reaction pathway of catechols and hydroquinones to produce carbon monoxide, carbon dioxide and water.
The considerable amount of water produced from the primary pyrolysis of the lignite
is likely to be consumed by secondary pyrolysis reactions [245-247]. Steam reforming
(also known as steam gasification and/or the water gas shift reaction) has been
extensively investigated for lignites [192] and is normally simplified by the chemical
reaction shown as Equation 4.4.
C + H2O → CO + H2 (ΔH° 298K = 205.9 kJ/mol) Equation 4.4
Many authors have reported that the water gas shift reaction is the major chemical
process responsible for the concurrent reduction in tar yield (see Section 4.5) and the
increase in carbon monoxide yield (Figure 4.12). Matsuo et al. [197] postulated that
pyrolytic water produced from the pyrolysis of an iron-impregnated lignite in a drop-
tube reactor, played a vital role in the catalytic steam reforming of the tar on the char
surface, which was accompanied by a significant reduction of the tar yield and
considerable increases in yields of hydrogen and carbon monoxide. Rapid pyrolysis
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 121
experiments with a Yallourn lignite using a drop tube furnace have also found that
the primary tar can undergo rapid steam reforming with the pyrolysis-derived water
at residence times as short as 2 sec [245] and that longer residence times of about 15
sec can almost fully convert the tar (i.e. 99%) to gaseous hydrocarbons, hydrogen
and carbon monoxide/dioxide [248].
The pyrolysis of the Loy Yang lignite at 900°C in the quartz reactor under the fast
heating mode (see Section 4.5) gave a tar yield of 2wt%db. The presence of tar at the
high pyrolysis temperature may be attributed to a short contact time between the
pyrolysis-derived water and the tar. In the quartz reactor, the pyrolysis-derived water
would be swept out of the reactor and subsequently unable to react with the tar in
secondary steam reforming reactions at high temperature. Furthermore, the
significant yield of water (more than 4wt%db) during the pyrolysis of the Loy Yang
lignite at 900°C, in the pyrojector, also suggests that the contact time between the tar
and volatilised hydrocarbons was short. That is, according to the water gas shift
reaction, the volatilised water would be consumed along with the organic carbon in
the tar or with volatilised hydrocarbons to form additional carbon monoxide and
hydrogen.
The increase in carbon monoxide yield with increasing pyrolysis temperature could
also be attributed to the carbon dioxide reforming of the tar and volatilised
hydrocarbons (Equation 4.5). The marginal decrease in carbon dioxide yield and the
corresponding increase in carbon monoxide yield at pyrolysis temperatures above
700°C (Figure 4.12) suggests that carbon dioxide reforming may have also played a
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 122
role in the production carbon monoxide. Carbon dioxide reforming of hydrocarbons
(also known as the Boudouard reaction) at high temperatures to produce carbon
monoxide and hydrogen (syngas) has been studied by several authors [249, 250] and the
Haghighi et al. [253] investigated the reaction mechanism of carbon dioxide reforming
of methane over a bed of coal char at temperatures between 800°C and 950°C and
reported the production of syngas with a maximum H2/CO ratio of one. In addition,
Mark et al. [254] modelled the reaction kinetics of carbon dioxide reforming methane
in the temperature range of 700°C to 850°C and concluded that the reaction was
approximately first order. The high yields of carbon dioxide and methane produced
during the pyrolysis of the raw Loy Yang lignite (Figure 4.12) would therefore
rapidly facilitate the conversion of these gases into carbon monoxide and hydrogen.
The yields of benzene, toluene, xylene (BTX) and phenol from the pyrolysis of the
raw Loy Yang lignite as a function of temperature are shown in Figure 4.15. At
700°C, the yields of the BTX and phenol were relatively similar at 0.01%wtdb. An
increase in pyrolysis temperature gave increases in yields for all the single ringed
aromatic hydrocarbons which is likely due from the thermal cracking / catalytic
reforming of the tar at the elevated pyrolysis temperatures [245]. The BTX yields
measured for the Loy Yang lignite are of approximately an order of magnitude lower
than the yields reported by Muira et al. [255] in the pyrolysis of Morwell lignite at
750°C. Similarly, Takarada et al. [256] investigated the pyrolysis of Yallourn lignite
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 123
over the temperature range 600°C to 1000°C using a fluidized bed reactor and also
reported at least an order of magnitude higher BTX yields compared to the Loy Yang
lignite used in this study. The higher BTX yields of the Morwell and Yallourn
lignites could be attributed to a number of factors including: differences in the extent
of secondary reactions of volatile matter in the pyrolyser reactor systems; differences
in the lignin-derived macromolecular structure; and differences in the typically
higher catalytic cation components in Morwell and Yallourn lignites (see Chapter 6).
The formation and volatilisation of aromatic compounds at different pyrolysis
temperatures is further discussed in the following Section.
600 650 700 750 800 850 900
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Pro
duct
yie
ld w
t% o
f coa
l (db
)
Temperature (°C)
Benzene Toluene Xylenes Phenol
Figure 4.15 Quantification of volatilised (a) benzene, toluene, xylenes and phenol
using a GS-GasPro column.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 124
4.7.1 Py-gc-ms with a HP-5 chromatography column The GS-GasPro chromatography column used in the gas chromatogram performed
well in separating low molecular weight hydrocarbons (Figure 4.11) and non-organic
gases however a limitation to this column was that long chain aliphatic and high
molecular weight triterpenoid materials which have been reported to be pyrolysis
components of Latrobe Valley lignites [208] were not eluted. Subsequently, pyrolysis
gas chromatography experiments were also conducted with the raw Loy Yang lignite
using a HP-5 chromatography column (Figure 4.16).
Figure 4.16 Gc trace of pyrolysed Loy Yang raw lignite at 600°C using a HP-5 column
The pyrogram of the raw Loy Yang lignite at 600°C, using a HP-5 chromatography
column (Figure 4.16), can be separated into four main regions:
• Light weight hydrocarbon gases
• Lignin derived phenolic region
• Lipid derived long chain aliphatics
• High molecular weight triterpenoid region.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 125
The pyrograms from the pyrolysis of the raw Loy Yang lignite as a function of
temperature, using a HP-5 chromatography column is shown in (Figure 4.18). The
sum product yield of the phenolic region and the long chain aliphatic region (C4 to
C30 alkane / alkenes) from the pyrolysis of the raw lignite at 700°C using the HP-5
chromatography column was less than 1wt%db (Figure 4.17). Similarly,
Chaffee et al. [208] had estimated that less than 1wt%db of what is eluted is alkanes
and alkenes from pyrolysis-gc of lignite at 700°C. In contrast, the sum product yield
of the phenolic region and the long chain aliphatic (C4 to C30) region at 800°C
increased significantly to 3wt%db, and at 900°C, the sum product yield decreased to
about 1wt%db (Figure 4.17). The decrease in the sum product yield (long chain
aliphatic region plus phenolic region) is attributed to the thermal cracking/catalytic
reforming of the volatiles to form light hydrocarbon and inorganic gases (see
Section 4.7).
600 650 700 750 800 850 9000
1
2
3
4
5
6
7
Prod
uct y
ield
wt%
of c
oal (
db)
Temperature (°C)
Aliphatic and aromatic hydrocarbons (C4 - C30)
Figure 4.17 Quantification of volatilised alkanes/alkenes (C4-C30) using a HP-5 column, from the py-gc of Loy Yang raw lignite as a function of temperature.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 126
Figure 4.18 Gc trace of pyrolysed Loy Yang raw lignite using a HP-5 column
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 127
The relative abundances and product distribution of the pyrolysis products changed
significantly with increasing pyrolysis temperature. Inspection of the relative
abundances from the 600°C pyrogram showed a low product distribution in the
phenolic region and a high product distribution of the long chain aliphatics. The sum
product yields of the phenolic region and the long chain aliphatic region (C4 to C30
alkane / alkenes) at 600°C and 700°C remained relatively unchanged (Figure 4.17)
however a comparison of the product distribution yields at 700°C showed an increase
in the phenolic region and a decrease of the long chain aliphatic region. Furthermore,
at 600°C and at 700°C, the peak for pristine (a vitamin E-derivative [257]) is easily
identified in the pyrograms between the peaks corresponding to C17 and C18 n-
alkanes/alkenes (Figure 4.17) whereas, the pyrogram at 800°C did not show the
presence of pristine nor many of the long chain aliphatic alkane/alkenes or the
characteristic triterpenoid peaks.
Christiansen et al. [258] flash pyrolysed a Columbian coal and reported that the
aliphatic hydrocarbons which were eluded at 750°C were not detected at 1025°C and
that the major components at 1025°C were single and multi-ringed aromatics.
Similarly, the pyrogram at 900°C in Figure 4.18 showed a significant increase in
single and multi-ringed aromatics (i.e. based on the retention times of selected
aromatic standards). Christiansen et al. [258] also pyrolysed a mixture of unbranched
alkanes at different temperatures and found the degradation of aliphatic material
under high pyrolysis temperatures, led to the formation of aromatic compounds and
that the degree of transformation increased with increasing temperature.
Furthermore, the formation of aromatic compounds from long chain polymers was
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 128
demonstrated by Purser and Wolley [259] who reported the presence of benzene,
toluene, xylene and styrene from the pyrolysis of polypropylene. The reductions of
the long chain aliphatic alkanes/alkenes region and the triterpenoid region and the
subsequent increase in aromatic/phenolic region with increasing pyrolysis
temperature (Figure 4.18) are likely attributed to the thermal cracking/catalytic
reforming of these hydrocarbons (see Section 4.7).
4.7.2 Volatile yield balance The mass balance of the different volatile fractions from the pyrolysis of the raw
lignite measured from the quartz reactor experiments and from the pyrograms, as a
function of temperature is shown in Table 4.2. In Section 4.4.2, the volatile yield
was found to be unaffected by the heating rate of the lignite and subsequently, it is
reasonable to assume that a similar volatile yield would result from the pyrolysis of
the raw lignite using the pyrojector. Also, the tar yield values from the quartz reactor,
fast heatup pyrolysis experiments have been included in Table 4.2 as guide for
estimating approximate non-tar volatiles.
The pyrolysis-gas chromatography setup used in this study shows that an estimated
70% of the non-tar volatiles were accountable (Table 4.2). This is a vast
improvement from earlier work by Chaffee et al. [208] which estimated that only 9-
15% of the pyrolysis material is eluted from the column from pyrolysis-gas
chromatography of lignite up to 700°C. Chaffee et al. [208] explained that most of the
volatiles from the pyrolysis of lignites could not be detected in their system because
of low-molecular weight compounds being eluded too rapidly from the
chromatographic system so their contribution to the pyrolysate could not be assessed
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 129
and that the presence of gases such as carbon monoxide, carbon dioxide, water and
hydrogen could not be detected with a flame ionisation detector (FID) [208].
Advantageously, the gas chromatogram used in this thesis was equipped with a FID
detector (for detection of hydrocarbons) and TCD detector (for detection of fixed
gases) and the application of GS-GasPro and HP-5 chromatography columns proved
effective in measuring a large proportion of the non-tar volatiles. The drawback
however was the inability to measure hydrogen gas which is a major non-tar volatile
gas from the pyrolysis of the lignite. Interestingly, the proportion of unaccounted
pyrolysates increased with increasing pyrolysis temperature which could be
attributed to the increase in hydrogen production from the catalytic reforming of
volatiles (see Section 4.7). Furthermore, inorganic nitrogen (eg NH3, NOx etc) and
sulphur (SOx) gas components were also eluded too rapidly from the
chromatographic system used in this study however the concentrations are expected
to be low because of the lignite’s low nitrogen (0.66wt%daf) and sulphur
(0.27wt%daf) contents. Alternative chromatography columns could overcome these
limitations thus increasing the overall mass balance accountability of the non-tar
volatiles.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 130
Table 4.2 Mass balance of volatile yield quartz reactor versus pyrolysate yield measured in the programs
Figure 4.20 Quantification of volatilised (a) methane, (b) ethane, (c) ethene, (d) ethylene (e) carbon monoxide, (f) carbon dioxide, from the py-gc of Loy Yang raw
lignite, MTE, SD and HTD treated products as a function of temperature using a GS-GasPro column.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 135
4.10 Conclusion For a drying process, a high level of water removal from the raw lignite is desirable
however significant losses in volatile material could be disadvantageous for some
industrial processes which convert the carbon matter to lower molecular weight
fractions (eg gasification, liquefaction etc). For MTE and SD, the operating
parameters for achieving a 50% moisture reduction from the raw lignite were
relatively mild (<250°C) and advantageously, the volatile fraction in the lignite was
not affected. In contrast, for HTD, the drying temperatures necessary for achieving a
similar moisture reduction were much higher (i.e. 320°C), and subsequently a
significant amount of volatile yield was lost from the raw lignite after processing.
Pyrolysis experiments with raw Loy Yang lignite using a quartz fluidized-bed/fixed
bed reactor system showed that the heating rate had no effect on the char yield of the
product as a function of temperature. Similarly, no significant differences in the char
yield profile were found between the slow heating rate and fast heating rate for the
MTE, SD and HTD products. In contrast, heating rate did affect the composition of
the volatile products in particular the tar yield and volatilisation of sodium and
chlorine. When the lignite particles were rapidly pyrolysed above 700°C, marked
increases in sodium volatilisation coincided with significant reductions in tar yield
whereas, for slow heating particles, sodium volatilisation and tar yield remained
relatively unchanged. The volatilisation of sodium in the fast heatup process at
temperatures above 700°C, is believed to be linked to volatile-char interactions
causing the reforming/cracking of the tar. In contrast to sodium, the volatilisation
chlorine during fast heatup pyrolysis was found to decrease at temperatures above
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 136
600°C which was postulated to be the result of recombination of chlorine with newly
formed radical sites inside the carbon macromolecular structure at high temperatures.
Furthermore, the concentrations of the alkaline metals calcium and magnesium were
not affected by differences in pyrolysis heating modes.
Rapid heat-up was also achieved with a pyrolysis-gas chromatography system. The
two major hydrocarbon components detected from the rapid pyrolysis of Loy Yang
lignite were methane and ethene which increased in yield as a function of
temperature from 600°C to 900°C. The yields of ethylene, propene and propylene
increased yields with increasing pyrolysis temperature while their correspondent
saturated hydrocarbons, ethane and propane, reached maximum yields at 700°C and
800°C, respectively, before declining with further increases in pyrolysis
temperatures. In addition, the increase in single aromatic hydrocarbon as a function
of increasing pyrolysis temperature also coincided with the breakdown of terpenoid
material and long-chain aliphatic compounds.
The pyrolysis-gas chromatography system which was equipped with a FID detector
(for detection of hydrocarbons) and TCD detector (for detection of inorganic gases)
and the application of GS-GasPro and HP-5 chromatography columns proved
effective in measuring a large proportion of the non-tar volatiles. An estimated 70%
of the non-tar volatiles were accountable which is a vast improvement over previous
studies. In contrast, quantification of non-tar volatiles using a py-gc-ms system,
which has been the instrument of choice in prior studies, could only account up to
15% of the non-tar volatiles.
Chapter 4 Pyrolysis of Raw Lignite and Dried Products 137
Rapid pyrolysis of the thermally treated products in the py-gc system showed that the
methane, ethene, carbon monoxide and carbon dioxide yields for the HTD product
were significantly lower when compared to the methane yields of the raw lignite and
MTE and SD products. The lower yields for the HTD product is likely attributed to
volatilisation during hydrothermal dewatering and also from the reduction of
catalytic reforming cations. The significant volatile yield loss during hydrothermal
dewatering could be disadvantageous in some industrial processes, which convert the
carbon matter to lower molecular weight fractions (eg gasification, liquefaction).
Chapter 5 Combustion of Raw Lignite and Dried Products 138
CHAPTER 5 COMBUSTION OF RAW LIGNITE AND
DRIED PRODUCTS 5.1 Introduction Combustion is generally the process where fuel and oxygen burn together at
sufficiently high temperature to evolve heat and combustion products. The events
that lead to combustion when a coal particle is progressively heated in air can be
separated into three main stages [265]:
• Devolatilisation of the coal particles and the consequent charring of the
particles
• Combustion of the volatile matter in the gas phase
• Combustion (or burning) on the solid surface of the residual char particle
Thermogravimetric tests for measuring combustion reactivities for carbonaceous
materials are widely reported in the literature. Generally, the applied techniques fall
into two categories (i) isothermal, where the sample is maintained at constant
temperature and (ii) non-isothermal, where the sample is heated at a constant rate.
The choice of technique for evaluating the combustion reactivity of the sample is
important. Isothermal measurements are often conducted by heating the sample to
the desired temperature in an inert gas before switching the gas stream to an oxygen-
containing supply. This technique is similar to the proximate analysis tests described
in the Section 4.2 however the combustion temperatures used in isothermal
measurements are milder ranging from 350°C to 500°C [266]. Isothermal
Chapter 5 Combustion of Raw Lignite and Dried Products 139
measurements taken at different temperatures can also give additional kinetic
information (e.g. Arrhenius plots of log [reaction rate] against inverse absolute
temperature can be applied to calculate the activation energy of the samples [266-268]).
The isothermal technique has been found to be ideal in measuring the combustion
reactivity of high temperature treated char products from pyrolysis and/or
gasification (eg >500°C) however this technique cannot be applied to raw
lignites [269] or to the thermally dried products described in Chapter 3. That is, it is
difficult to use isothermal techniques to investigate oxidation-combustion kinetics of
raw lignite or thermally treated products (Chapter 3) because of the high reactivity
nature of the sample upon exposure to oxygen. Furthermore, heating the samples to
an isothermal temperature greater than the processing temperature used to generate
the thermally treated products (Chapter 3) will make the raw lignite and thermally
treated products more alike (eg similar volatile/char ratios, elemental composition,
oxygen group functionality, etc) and therefore, the information obtained would be
meaningless.
Alternatively, the non-isothermal approach is more applicable for investigating
differences between the raw lignite [270-273] and thermally treated products. Previous
workers who have used the non-isothermal approach have heated the sample at a
constant rate up to 900°C. Under such conditions, complete conversion normally
occurs before the sample has reached 900°C [266, 268]. Non-isothermal measurements
in the TGA are relatively fast (the sample can be heated to 900°C within 1h).
Furthermore, in a non-isothermal measurement, the peak temperature can be
Chapter 5 Combustion of Raw Lignite and Dried Products 140
determined (i.e. the temperature at which the rate of weight loss from the sample is at
a maximum; a high peak temperature is indicative of a less reactive fuel [268]). In
addition, the ignition temperature [274] and the burnout temperature (i.e. the rate of
weight loss is less than 1% per min) can also be deduced [275]. A disadvantage of the
non-isothermal approach is the inability to accurately assess the catalytic activity of
alkali and alkaline earth metal species (AAEM) at a given temperature [111].
In this study, a combination between isothermal and non-isothermal techniques was
used to evaluate the reactivity differences between the raw lignite and its thermally
dried products. That is, the samples were heated at a constant rate to the desired
combustion temperature (non-isothermal) and maintained until complete conversion
was achieved (isothermal). The experimental parameters, which may affect the
combustion reactivity of the dried products in the TGA, are investigated in following
sections.
Chapter 5 Combustion of Raw Lignite and Dried Products 141
5.1.1 Effect of particle size Thirteen discrete particle size intervals were investigated using the raw lignite. For
comparative purposes, 30.0mg of sample was used in each combustion test.
Figure 5.1 suggests that particle size of the sample had very little effect on the
combustion rate. This behaviour is in direct contradiction to what other workers have
reported in their combustion investigations (i.e. with increasing particle size, the
combustion reactivity is decreased). This contrasting difference can be rationalised
by understanding the experimental protocol that was applied for the combustion tests
in this study.
The combustion process in a TGA can be described by the following several
steps [276]:
• The diffusion of gaseous reactants and products (mass transfer) from the bulk
of the gas phase to the internal surface of the reacting solid particle
• Adsorption of gaseous reactants on and desorption of reaction products from
the solid surfaces
• Chemical reaction between the adsorbed gas and solid
It is expected that with an increase in particle size, the diffusion of gaseous products
through the pores of the particle will take longer and will therefore reduce the
combustion rate of the particle. Various combustion models have attempted to
include gaseous diffusion through pores and between particles [277-283].
Morgan et al. [284] reported that particle size and coal properties were responsible for
different coal-burning profiles obtained from TG/DTG analysis. Morgan et al. [284]
Chapter 5 Combustion of Raw Lignite and Dried Products
142
also mentioned that oxygen uptake and particle reactivity increased with decreasing
particle size. In addition, particle sizes greater than 125μm gave significantly
different curve profiles compared to smaller particle sized samples. Gold [285]
concluded that the temperature and the magnitude of the exothermic peak were
affected by the heating rate, sample mass and particle size.
Based on the results of previous workers mentioned above, a likely explanation for
the similarity between combustion curves given in Figure 5.1 is the result of gaseous
diffusion limitations of reactants and products through the internal and external
surface of the particles. This diffusion limitation is attributed to the large sample
mass used in each TGA experiment. The effect of sample mass on the combustion
rates is further elucidated in the following section.
Chapter 5 Combustion of Raw Lignite and Dried Products
< 63 microns63 to 75 microns75 to 90 microns90 to 106 microns106 to 125 microns125 to 150 microns150 to 180 microns180 to 250 microns250 to 355 microns425 to 500 microns600 to 710 microns850 to 1000 microns 1180 to 1400 micronsTemperature
Figure 5.1 Combustion of raw, discrete particle sized lignite. 30mLmin-1 air flow, 30mg sample
Chapter 5 Combustion of Raw Lignite and Dried Products 144
5.1.2 Effect of sample mass loading on combustion rates The amount of sample placed inside the TGA crucible had a significant effect on the
combustion rate obtained (Figure 5.2). The rate of combustion increased with a
reduction of sample loading. The combustion reactivity of dried products will be
dependant on the transport of oxygen and heat to the surface of the particle where
oxidation-combustion takes place. In the TGA, the diffusion of oxygen to the surface
and the removal of product gases (eg CO2) from the surface can significantly affect
the observed combustion rate. That is, the carbon dioxide released during combustion
must diffuse through the pore structure of the particle, through the empty spaces
between the particles and then to the upper surface of the char bed where it is carried
away from the furnace. Internal diffusion limitations have been reported by Hemati
and Laguerie [286] in their investigation on steam gasification kinetics of charcoal in a
TGA. Ollero et al. [283] also reported that the internal diffusion resistance of a 3mm
deep char bed was quite significant thus affecting the measured gasification rates in
the TGA. In addition, Ollero et al. [283] suggested that if the removal of CO2 was not
fast enough, the temperature of the inside layers of the bed may be lower than that of
the surface of the bed. This endothermic effect (i.e. lower combustion temperature)
and the slow external diffusion of CO2 from the sample bed can explain the slow
combustion rates observed with an increase of sample loaded in the crucible.
In Figure 5.2 the combustion rate of the 5.8 and 6.7mg sample was similar therefore
suggesting that the internal diffusion limitations of air and product gases were
significantly reduced. In order to eliminate diffusion limitations of gases, a smaller
crucible and smaller sample masses were used in each of the combustion tests
Chapter 5 Combustion of Raw Lignite and Dried Products
145
described from here onwards. For comparative purposes, it was also decided to apply
a monolayer bed of particles and a constant sample mass of 2.0mg. These changes
would allow comparative differences in combustion reactivity of the dried products
to be more meaningful.
Chapter 5 Combustion of Raw Lignite and Dried Products
Figure 5.12 Combustion of MTE products (90-125μm particle size) in the TGA at 400°C. Conditions: 30mLmin-1 air flow, 2.0mg sample
(1) Effect of processing temperature, (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (2) Effect of applied mechanical pressure, (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion
Chapter 5 Combustion of Raw Lignite and Dried Products 165
2000 3000 4000 5000-100
-80
-60
-40
-20
0
raw, 200°C, 250°C, 280°C, 300°C, 320°C, 350°C
1a
Wei
ght l
oss
(%)
Time (sec)
1000 2000 3000 40000
20
40
60
80
100
120 1b
Time (sec)
Hea
t Flo
w (m
W)
0.0 0.2 0.4 0.6 0.80.00
0.02
0.04
0.06
0.08
0.10
0.12 1c
Spe
cific
reac
tivity
(min
-1)
Conversion (daf)
Figure 5.13 Combustion of raw and HTD products (90-125μm particle size) in the
TGA at 400°C. Conditions: 30mLmin-1 air flow, 2.0mg sample
Chapter 5 Combustion of Raw Lignite and Dried Products 166
2000 3000 4000 5000-100
-80
-60
-40
-20
0 1a
Wei
ght l
oss
(%)
Time (sec)
2000 3000 4000 50000
20
40
60
80
100
120
Raw coal, SD 130°C, SD 250°C, SD 300°C, SD 350°C
1b
Hea
t Flo
w (m
W)
Time (sec)
0.0 0.2 0.4 0.6 0.8 1.00.00
0.05
0.10
0.15
0.201c
Conversion (daf)
Spe
cific
reac
tivity
(min
-1)
Figure 5.14 Combustion of raw and SD products (90-125μm particle size) in the
TGA at 400°C
Chapter 5 Combustion of Raw Lignite and Dried Products 167
5.5 Combustion reactivity of MTE Morwell and Yallourn lignites
The combustion reactivity of Loy Yang lignite and of the dried products from HTD,
MTE and SD have been extensively discussed in previous sections of this chapter.
The physico-chemical changes arising from drying Loy Yang lignite with the MTE,
HTD or SD process, resulted in only small differences in the combustion reactivity
when compared to the parent lignite. To further examine the effects of thermal
drying and physico-chemical properties affecting the combustion reactivity of
lignites, it was decided to extend this study using two additional Latrobe Valley
lignites from the Morwell and Yallourn open cut mines, respectively. The MTE
process was decided to be the best process to further evaluate the combustion
reactivity of thermally dried coal. The MTE process was chosen because of the
process’s very high lignite mass recovery at relatively low processing temperatures,
the significant reductions in the large pore region and the considerable reductions in
the inherent inorganic species in the final product.
The pore volumes of MTE processed Loy Yang, Morwell and Yallourn lignites at
150°C/5.1MPa are shown in Figure 5.15. Comparing the MTE products from the
three lignites show that the MTE Morwell product gave the lowest large pore volume
(0.18cm3g-1) whereas identical pore volumes (0.27cm3g-1) were measured for MTE
Loy Yang and MTE Yallourn. Despite the three lignites having marginally different
large pore volumes, the combustion reactivity for the Loy Yang MTE product was
contrastingly different to the Morwell and Yallourn MTE products (Figure 5.16). The
Loy Yang MTE product clearly demonstrated a significantly longer duration for
complete combustion-oxidation to be achieved. This difference could not be
Chapter 5 Combustion of Raw Lignite and Dried Products 168
explained by the sodium contents in the MTE products because all three had similar
concentrations (Table 5.2).
0.71 0.70 0.70
0.06 0.06 0.06
0.27
0.180.27
0.00
0.20
0.40
0.60
0.80
1.00
1.20
Loy
Yan
g
Mor
wel
l
Yal
lour
n
Por
e vo
lum
e (c
m3 g-1
)
Vol. occupied by carbon Micropore volume Large pore volume
Figure 5.15 Pore volumes of MTE processed Loy Yang, Morwell and Yallourn lignites at 150°C/5.1MPa
Chapter 5 Combustion of Raw Lignite and Dried Products 169
Figure 5.16 Combustion of Loy Yang, Morwell and Yallourn MTE products at 150ºC/ 5.1MPa
(1) 90-125μm, combustion 400°C (a) weight loss vs time (b) heat flow vs time (c) specific reactivity vs conversion
Chapter 5 Combustion of Raw Lignite and Dried Products
170
Despite the MTE Morwell product having a lowest large pore volume (Figure 5.15),
the MTE Yallourn product was found to be more reactive at 400ºC (Figure 5.16).
This contrasting difference was more evident in the 250-500μm MTE products
(Figure 5.17) whereas at 450ºC, the Yallourn MTE product was still marginally more
reactive than the Morwell MTE product. The order of reactivity of the three lignites
can be explained if iron catalysed the combustion reactions (Figure 5.17). Indeed,
multiple linear regression on the inter-relationships between inorganic components
and the combustion reactivity of the coals does confirm that iron is more effective
than the AAEM cations in catalysing the combustion-oxidation process (see Section
6.6.2. Furthermore, the effect of iron and AAEM cations on the combustion
reactivity of lignites is further discussed in Section 6.7 and Section 6.8, respectively.
Table 5.2 Ash, acid extractable inorganics and chlorine in the raw lignites and MTE products (wt% db)a, b. MTE conditions: 150°C/5.1MPac
Lignite Ash Na Ca Mg Al FeNPd Cl
Loy Yang raw lignite 0.9 0.09 0.04 0.07 0.01 0.06 0.07
Loy Yang MTE 0.9 0.05 0.04 0.06 0.01 0.06 0.05
Morwell raw lignite 2.2 0.08 0.32 0.21 0.02 0.36 0.06
Morwell MTE 2.2 0.05 0.31 0.19 0.02 0.35 0.05
Yallourn raw lignite 2.0 0.07 0.14 0.18 0.01 0.57 0.05
Yallourn MTE 2.0 0.04 0.15 0.17 0.01 0.58 0.05 a The error is ±0.01wt% for concentrations of 0.01 - 0.10wt% and ±0.02wt% for concentrations greater than 0.10wt% for all elements. b Actual mass values (g) for the raw and MTE products is given in Appendix C. c MTE products are shown in bold so to allow easier comparison between the inorganics of the different lignites and their specific reactivities shown in Figure 5.16 to Figure 5.18. d NP = non-pyritic.
Chapter 5 Combustion of Raw Lignite and Dried Products
Figure 5.17 Combustion of Loy Yang, Morwell and Yallourn MTE products at 150ºC/ 5.1MPa
(1) 90-125μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (2) 250-500μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (3) 250-500μm particle size, combustion at 450°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion
Chapter 5 Combustion of Raw Lignite and Dried Products
172
Despite the expected large differences in the large pore volumes between the MTE
products and the raw lignites (see Chapter 3), only marginal differences were evident
between the combustion reactivity of the raw lignite and its products (Figure 5.18).
MTE processing had a marginal impact on removing acid extractable inorganics and
chlorine ions from the Latrobe Valley lignites (Table 5.2). Subsequently, differences
in the combustion reactivity arising from the removal of catalytic inorganic
components from the coal were not evident when compared to the parent lignite.
Advantageously, these marginal differences suggests that the conventional boiler
systems currently in operation in the Latrobe Valley would be more than adequate in
combusting thermally treated products and water washed lignites from their
respective mines.
173
Chapter 5 Combustion of Ra
(1) Loy Yang 90-125μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (2) Morwell 90-125μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (3) Yallourn 90-125μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion
w Lignite and Dried Products
Figure 5.18 Combustion of Loy Yang, Morwell and Yallourn raw lignite, MTE products (150ºC/ 5.1MPa) and water washed lignite
2000 3000 4000-100
-80
-60
-40
-20
0
1a
Wei
ght l
oss
(%)
Time (sec)
2000 3000 4000-100
-80
-60
-40
-20
0
2a
Wei
ght l
oss
(%)
Time (sec)
2000 3000 4000-100
-80
-60
-40
-20
0
3a
Time (sec)
Wei
ght l
oss
(%)
2000 3000 40000
20406080
100120140160180200220240260280300320
1b
Hea
t Flo
w (m
W)
Time (sec)
1000 2000 3000 40000
20406080
100120140160180200220240260280300320
2b
Time (sec)
Hea
t Flo
w (m
W)
1000 2000 3000 40000
20406080
100120140160180200220240260280300320
3b
Hea
t Flo
w (m
W)
Time (sec)
0.0 0.2 0.4 0.6 0.8 1.00.0
0.2
0.4
0.6
0.8
1c
Spe
cific
reac
tivity
(min
-1)
Conversion (daf)
0.0 0.2 0.4 0.6 0.8 1.00.0
0.2
0.4
0.6
0.8
2c
Conversion (daf)
Spe
cific
reac
tivity
(min
-1)
0.0 0.2 0.4 0.6 0.8 1.00.0
0.2
0.4
0.6
0.8
Raw lignite, MTE 150C/5.1MPa, Washed lignite
3c
Conversion (daf)
Spe
cific
reac
tivity
(min
-1)
Chapter 5 Combustion of Raw Lignite and Dried Products 174
5.6 Conclusions Differences in the combustion reactivity of MTE, HTD and SD products were
measured using a TGA. A monolayer of lignite particles inside the TGA crucible was
found to eliminate gaseous diffusion limitations of oxygen and product gases.
Furthermore, an isothermal combustion temperature of 400°C was found to be
adequate in achieving a steady combustion-oxidation state and subsequently also
reducing the likelihood of particle ignition.
The physico-chemical changes arising from drying Loy Yang lignite in MTE, HTD
or SD resulted in only marginal differences on the combustion reactivity between the
sample products. Differences in the large pore volume of the dried products only had
a minor affect on combustion reactivity. Furthermore, the effectiveness of catalytic
inorganic species clearly outweighed pore volume effects for increasing the
combustion reactivity of the products. The results suggest that non-pyritic iron and
AAEM cations are the major combustion promoters among the metal constituents of
Latrobe Valley lignites.
The minor differences in combustion reactivity between the parent lignite and the
thermally dried products suggests that conventional boiler systems currently in
operation in the Latrobe Valley would be more than adequate in combusting the
dried products from any of the three drying processes.
Chapter 6 Combustion of Raw Lignites
175
CHAPTER 6 COMBUSTION OF RAW LIGNITES
6.1 Introduction The previous two chapters have investigated in detail the pyrolysis and combustion
reactivity of a Loy Yang lignite and of the dried products from HTD, MTE and SD.
Furthermore, the combustion reactivity of a single Morwell and Yallourn lignite was
also investigated in Chapter 5. Restricting a study to a single lignite from each mine
can lead to misleading conclusions, particularly if one of the lignites has unusual
features. It was envisaged that relating the chemical and structural properties of a
series of lignites to reaction results could be used to investigate the parameters which
govern the combustion reactivity of lignites. Correlation of lignite properties with
reaction results also can suggest mechanistic explanations of any trends found.
In this chapter, combustion on a diverse suite of well-characterised lignites sourced
from the Latrobe Valley open cut mines is examined. In addition, the
intercorrelations between physicochemical properties and the combustion reactivity
of the lignites are investigated using multiple regression analysis. Furthermore, the
transformations cationic components during volatilisation and combustion, and the
effectiveness of these species in facilitating the breakdown/oxidation of the organic
components in the coal are discussed.
Chapter 6 Combustion of Raw Lignites
176
6.2 Latrobe Valley lignites – background information The major economic coal deposit of the Gippsland Basin occurs in the Latrobe
Valley, which is located in the southern-eastern part of Australia, approximately
150km east of Melbourne, Victoria. The Latrobe Valley coals are amongst the lowest
rank coals commercially utilized anywhere in the world. The lignite resources in this
area are vast by world standards and are concentrated in exceptionally thick seams
under a relatively thin cover of overburden. Latrobe Valley lignites are of Tertiary
age and are typically soft, high in moisture and low in ash.
Three major open cut mines are in operation in the Latrobe Valley (Figure 1.1). The
first of the major open cut mines which was first operated by the State Electricity
Commission of Victoria was Yallourn. The Yallourn open cut mine has been
providing fuel to the Yallourn power stations since 1924. The average thickness of
the Yallourn seam mined is 60m with a coal to overburden ratio of 3.5:1 [305]. To the
south of the Yallourn open cut mine is the Morwell open cut mine which provides
fuel for the Hazelwood power station and for the Morwell briquette factory. Two
major seams are present in this area, with the Morwell 1 seam reaching a maximum
thickness of 165m beneath the Morwell Township, and the Morwell 2 seam reaching
up to 55m in thickness [305]. The Yallourn and Morwell seams are also mined at the
Loy Yang open cut mine, with up to 230m thickness of continuous low ash coal [305].
The Loy Yang open cut mine is the major source of energy for the largest of the
power stations in the Latrobe Valley.
Chapter 6 Combustion of Raw Lignites
177
A total of ten Latrobe Valley coals from the three fields were selected for this study
to examine the effects of variation in lithotype, atomic H/C ratio, inorganics, porosity
combustion reactivity and pyrolysis. The coals were:
• Loy Yang Open Cut run-of-mine low ash – medium dark lithotype (LYLA)
• Loy Yang Open Cut medium sodium content – medium dark lithotype (LYMNa)
• Loy Yang Open Cut high sodium content – medium dark lithotype (LY Na)
• Morwell Open Cut medium magnesium content – medium dark lithotype (MMTE)
• Morwell Open Cut high magnesium content – medium dark lithotype (MMg)
• Yallourn Open Cut East Field run-of-mine – medium dark lithotype (YMTE)
• Yallourn Open Cut pale lithotype mined from under the former townsite (YT Pale)
• Yallourn Open Cut dark lithotype mined from under the former townsite (YT Dark)
• Yallourn Open Cut East Field high iron content – medium dark lithotype (YEF Fe)
• Yallourn Open Cut East Field dark lithotype (YEF Dark)
Analytical information for these coals is listed in Table 6.1, Table 6.2 and Table 6.3.
6.3 Characterisation of Latrobe Valley lignites The proximate analysis (moisture, ash, volatile matter and fixed carbon), the ultimate
analysis (carbon, hydrogen, nitrogen and sulphur), the moisture holding capacity and
the calorific value for the Latrobe Valley lignites are shown in Table 6.1. The coal
rank as measured by the calorific value of the coals on an ash-free, moist (a.f.m.)
basis (Table 6.1) and according to the Australian Standard [306], classifies the Latrobe
Valley lignites as lower-rank brown coals.
Chapter 6 Combustion of Raw Lignites
178
Brown coal lithotypes in the Latrobe Valley refers to coal-banding visible in air-
dried coal. George [307] characterised the lithotype of brown coals into 5 categories;
pale, light, medium-light, medium-dark and dark and reported that from dark to light,
the moisture content decreased up to 5%, volatile matter increased from 48% to 63%,
specific energy (gross dry basis) increased from 26 to 29 MJ kg-1 and the hardness
decreased in the air-dried state. Similarly, the lithotype of the Yallourn coals in Table
6.1 also demonstrated similar characteristics from dark to pale. The dark lithotypes
(YEFD and YTD) had the highest MHC and correspondingly lowest gross calorific
value (a.f.m.) whereas the MHC of the medium dark lithotypes (YMTE and YEFFe)
was significantly lower than the dark lithotypes. The pale lithotype YTP had the
lowest MHC within the Yallourn coal suite (Table 6.1).
The YTP lithotype was also contrastingly different from all the other Latrobe Valley
coals in the suite by its relatively low oxygen content, high volatile matter and high
atomic H/C ratio. Higgins et al. [308] also reported similar trends with lithotype and in
addition also found an increase in porosity, a decrease in surface area and a decrease
in apparent density from dark to pale lithotypes.
The only two Morwell coals in Table 6.1 were medium-dark lithotypes and in
comparison to the Yallourn medium-dark lithotype coals, the moisture content and
MHC were significantly lower. The Yallourn seam is the top most and hence
youngest of the seams in the Latrobe Valley and the Morwell seam underlies the
Yallourn seam [309]. As coalification proceeds the moisture content of the coal
decreases due to loss of oxygen containing functional groups (–OH, –COOH, –C=O).
Chapter 6 Combustion of Raw Lignites
179
The Morwell and Yallourn medium-dark lithotypes had similar oxygen contents
(25.1 to 26.7wt%d.a.f.) and similar atomic H/C ratio (0.83 to 0.85) thus coalification
differences between the samples were not evident. Alternatively, the burial depth of
the mined coals could be an underlying factor in the differences in the moisture
contents between the Yallourn and Morwell medium-dark lithotype coals.
Holdgate [310] reported that the average moisture content of the seams in the Latrobe
Valley decrease approximately 1% every 20m burial depth due to compression. The
Morwell and Yallourn medium-dark lithotypes did not show any other
distinguishable differences that could explain the variation in moisture contents,
MHC and corresponding calorific values.
The Yallourn and Morwell seams are mined in the Loy Yang open cut mine. The
three Loy Yang coals in the suite, of medium-dark lithotype, have similar physical
and organic characteristics with the only exception of lower ash yield and oxygen
content of the LYLA coal. In terms of proximate and ultimate analysis the Loy Yang
coals (except for LYLA) are not clearly distinguishable between the Yallourn and
Morwell coals of similar lithotype.
180
Table 6.1 Moisture holding capacity, proximate analysis, ultimate analysis and calorific value of the Latrobe Valley coals used in this study Elemental Analysis
Alternatively, Kumar et al. [351] reported that CaO can also be transformed to CaCO3
via the hydration and carbonation pathway
CaO + H2O Ca(OH)⎯→⎯ 2
Ca(OH)2 + CO2 CaCO⎯→⎯ 3 + H2O
According to the reaction schemes above and according to previous investigators, the
form of calcium in the hydrocarbon matrix can influence the rate of the oxidation-
combustion process. The association of dispersed complexed calcium and active
Chapter 6 Combustion of Raw Lignites 207
catalytic ionic salts can facilitate excellent catalytic activity during combustion
whereas other forms such as calcium carbonate and calcium chloride may not be as
effective in improving the rate of combustion.
6.8.2 Magnesium Multiple linear regression analysis on the parameters which affect the combustion
reactivity of the Latrobe Valley samples, identified that the magnesium content had a
stronger influence on the rate of combustion than calcium, with correlation
coefficient constants of 0.74 and 0.64, respectively (see Section 6.6.2). Magnesium
and calcium are alkaline earth metal and therefore their chemistry is comparable.
Analogous to calcium, magnesium ions are also associated to organic oxygen in the
lignites and upon low temperature decarboxylation/devolatilisation/combustion, the
chemical transformations from carboxylate to ionic salt are also expected.
Comparative combustion tests in the catalytic effectiveness of magnesium versus
calcium are very limited, in particular for lignites. Badin [332] speculated that the
catalytic effectiveness of cations was related to the proton-transfer capacity of the
cations in water (pKa values). Badin claimed that cations with a high pKa value were
analogous to the strength of the cations to undergo base-forming reactions by
hydrolysis (or solvolysis) during early stages of coal combustion. Subsequently,
Badin hypothesized that calcium would be more effective than magnesium as a
combustion catalyst and that the order of effectiveness would follow in the order of
Na+ > Ca2+ > Mg2+ >Al3+. Similarly, Yan et al. [352] investigated the effect of mineral
species of the combustion oxidation of coke and reported the catalytic activity of
calcium was greater than magnesium with the order of Fe > Na ≈ K > Ca > Mg > Ti.
Chapter 6 Combustion of Raw Lignites 208
In addition, Sujanti and Zhang [353] investigated the role of inherent inorganic matter
in low temperature oxidation of a Victorian brown coal and reported that calcium
carbonate was more reactive and promoted spontaneous combustion whereas the
carbonated form of magnesium inhibited spontaneous combustion. In contrast,
catalytic activity in the combustion of graphite have found that magnesium had a
higher activity than calcium [354].
The combustion of coal is a complex, multiphase, multicomponent chemically
reacting system involving concurrent heterogenous chemical reactions with organic
and inorganic components in the coal. Unfortunately, research into comparing the
catalytic effectiveness of magnesium and calcium during combustion is very limited.
Some studies (mentioned above) claim calcium is more effective than magnesium
whereas others claim vise versa. In this study, magnesium was statistically calculated
to have more of an influence into promoting the combustion reactivity of the coal
when compared to calcium. These differences are likely stemmed down from the
chemical forms of magnesium and calcium, and their connection and involvement
within the hydrocarbon macromolecular framework of the sample under the different
stages of volatilisation and combustion. Furthermore, the transformation of these
cations during the volatilisation and combustion process, from cationic complexation
with organic oxygen function groups within the coal matrix, to an active catalytic
oxidising agent, would also determine its effectiveness in facilitating the
breakdown/oxidation of the organic components in the coal.
Chapter 6 Combustion of Raw Lignites
209
Table 6.10 AAEM species, iron, chloride and sulphide ions present in the raw lignite, water washed and acid washed products (mmol of univalent charge; equivalent to 1g
of dry raw lignite).
Na+ Fe2+/3+ Ca2+ Mg2+ Cl- S2- Total +ve ions
Total -ve ions
LYLA Raw 0.004 0.004 0.002 0.007 0.002 0.001 0.017 0.003 H2O 0.002 0.004 0.002 0.006 0.001 0.001 0.014 0.002 Acid <0.001 0.001 <0.001 0.001 <0.001 0.001 0.002 0.001 LYMNa Raw 0.020 0.002 0.002 0.013 0.012 0.001 0.037 0.013 H2O 0.004 0.002 0.002 0.010 0.001 0.001 0.018 0.002 Acid 0.001 0.001 <0.001 0.002 <0.001 0.001 0.004 0.001 LY HNa Raw 0.029 0.023 0.009 0.044 0.009 0.001 0.105 0.010 H2O 0.007 0.020 0.009 0.041 0.001 0.001 0.077 0.002 Acid <0.001 0.016 <0.001 <0.001 <0.001 0.001 0.016 0.001 MMTE Raw 0.003 0.019 0.016 0.017 0.002 0.001 0.055 0.003 H2O 0.002 0.019 0.015 0.016 0.001 0.001 0.052 0.002 Acid <0.001 0.013 <0.001 <0.001 <0.001 0.001 0.013 0.001 MMg Raw 0.009 0.002 0.024 0.032 0.003 0.001 0.067 0.004 H2O 0.004 0.002 0.024 0.032 0.001 0.001 0.062 0.002 Acid <0.001 0.002 0.001 0.001 <0.001 0.001 0.004 0.001 YMTE Raw 0.003 0.031 0.007 0.015 0.001 0.001 0.056 0.002 H2O 0.002 0.031 0.007 0.014 0.001 0.001 0.054 0.002 Acid <0.001 0.013 0.001 0.002 <0.001 0.001 0.016 0.001 YT Pale Raw 0.002 0.012 0.006 0.011 0.002 0.001 0.031 0.003 H2O <0.001 0.011 0.005 0.010 0.002 0.001 0.026 0.003 Acid <0.001 0.008 <0.001 <0.001 <0.001 0.001 0.008 0.001 YT Dark Raw 0.002 0.018 0.008 0.015 0.001 0.001 0.043 0.002 H2O <0.001 0.017 0.008 0.014 0.001 0.001 0.039 0.002 Acid <0.001 0.013 <0.001 0.001 <0.001 0.001 0.014 0.001 YEF Fe Raw 0.002 0.038 0.008 0.020 0.003 0.001 0.068 0.004 H2O 0.001 0.036 0.008 0.020 0.002 0.001 0.065 0.003 Acid <0.001 0.025 <0.001 <0.001 <0.001 0.001 0.025 0.001 YEF Dark Raw 0.003 0.031 0.008 0.013 0.001 <0.001 0.055 0.001 H2O 0.001 0.028 0.008 0.012 0.001 <0.001 0.049 0.001 Acid <0.001 0.019 <0.001 0.001 <0.001 <0.001 0.020 <0.001 Example: Charge balances for MMTE raw Total positive charge = Na + Fe + Ca + Mg = 0.003 + 0.019 + 0.016 + 0.017 = 0.055 mmol Total negative charge = Cl + S = 0.002 + 0.001 = 0.003 mmol Note: Alkali metals lithium and potassium univalent charges were <0.001. Also, qualitative analysis using a scanning electron microscope (SEM) – energy dispersion x-ray (EDX) system confirms that all the major inorganic components in the coal have been identified (see Appendix K).
LYHNa
LYHNa
YEFFeYEFFe
MMTE
MMg
MMg
MMTE
YEFDYMTEYMTE
YEFD
YTD
YTPYTP
YTD
LYMNa
LYMNa
LYLALYLA
LYMNaMMg
LYLAYTD YMTE
YEFDMMTE LYHNaYTP YEFFe
R2 = 0.48
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Combustion reactivity value
AA
EM
con
cent
ratio
n (m
mol
cha
rge)
Raw lignites Water washed lignites Acid washed lignites
Figure 6.9 AAEM concentration versus combustion reactivity for the raw, water washed and acid washed lignite samples.
Chapter 6 Combustion of Raw Lignites
210
Chapter 6 Combustion of Raw Lignites
Figure 6.10 AAEM and iron concentration (cation concentration) versus the combustion reactivity value for the raw, water washed and acid washed lignite samples.
211
YTDMMg
LYMNaLYLA
LYMNaMMg
LYLA
LYLA
YMTEMMTE
LYHNa
YEFFe
YEFD
YEFFeYTP
YTD
MMg
MMg
LYHNa
YEFFe
MMTEYEFDYMTE
YTDLYMNa
LYHNa
YEFFe
YTP
YMTEYEFD
R2 = 0.83
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Combustion reactivity value
Cat
ion
conc
entra
tion
(mm
ol c
harg
e)
Raw lignite Water washed lignite Acid washed lignite
Chapter 6 Combustion of Raw Lignites 212
6.8.3 Sodium Sodium was the only alkali metal detected in the raw Latrobe Valley lignites that was
in reasonable concentration that could affect the combustion reactivity. The levels
lithium and potassium were not included in the total cationic concentration because
their levels were low and therefore their influence on the combustion reactivity of the
coal samples would be negligible. Multiple regression analysis expressed that the
sodium content had a weak positive linear correlation with the measured combustion
reactivity with a correlation coefficient of 0.31. According to this statistical analysis,
sodium was not as effective as iron, magnesium and calcium, respectively in
promoting the oxidation-combustion process. In contrast, previous studies have
reported that sodium is more effective than alkaline metals in the combustion
reactivity of the coal [332, 352]. Furthermore, char derived samples from the pyrolysis
of Loy Yang lignite (from Section 4.4) gave a strong correlation between the
concentration of sodium remaining in the char and the combustion reactivity of the
sample, whereas the calcium and magnesium contents which remained relatively
unchanged as a function of pyrolysis temperature, did not influence the rate of
combustion to the same extent (see Chapter 7). The contrasting differences in the
catalytic activity of sodium for the raw coals compared to the charred pyrolysed
samples is likely attributed to the chemical forms of sodium in these samples during
the combustion-oxidation process.
The behaviour and chemical transformations of sodium during pyrolysis and
combustion are different. The pyrolysis of the Loy Yang lignite (see Section 4.6)
found that some of the sodium in the coal volatilised at temperatures as low as
Chapter 6 Combustion of Raw Lignites 213
500ºC. Furthermore, prior work has also reported sodium volatilisation during
pyrolysis can occur as low as 300ºC [355]. In contrast, in a slow heatup combustion
process at 400ºC, the sodium present in the coal sample is unlikely to be volatilised
but instead be bonded continuously within the carbon macromolecular network.
Similar to alkaline earth metals, the transformation of organically bound sodium to
its ionic salt form during the isothermal combustion of the coal samples at 400ºC can
be simplified and summarised by the chemical reactions below:
Stage 1: Detachment of calcium from organic functional groups
(hydrocarbon-COO-Na) (hydrocarbon –Na) + CO⎯→⎯ 2
Stage 2: Formation of ionic salts and ash
(hydrocarbon –Na) + O2 Na⎯→⎯ 2O
Na2O + CO2 ⎯→⎯ Na2CO3
According to the reaction schemes above, the form of the sodium in the raw coal and
its transformation during the oxidation combustion process would significantly
influence the reactivity of the sample. Studies on impregnating sodium containing
compounds into a coal/char have reported that sodium in the form of NaCl would not
be as catalytically effective as other forms of sodium such as Na2CO3 [356, 357]. Upon
combustion, finely dispersed catalytic material would agglomerate into fine particles
would could affect the combustion rate of the sample. In addition, the dispersion of
the sodium in the carbon macromolecular matrix would also govern the catalytic
effects of sodium during combustion [358].
Chapter 6 Combustion of Raw Lignites 214
6.8.4 Aluminium Aluminium salts such as alumina oxide (Al2O3), are excellent catalysts in
liquefaction and gasification of coal which is predominately attributed to its alkaline
properties. In contrast, aluminium cations during combustion have been found to
exhibit low catalytic effectiveness during combustion when compared to reforming
processes [359]. In this study, statistical multiple linear regression identified that the
presence of acid extractable aluminium had a negative, effect of the combustion
reactivity of the coal samples. This negative behaviour could be the consequence of
two separate inhibiting effects: (1) the production of water from the combustion
process; and (2) the obstruction of active catalysts (eg AAEM cations and iron)
resulting from the complexation with aluminium cations.
Several investigations have reported reduced catalytic performance in the
combustion of hydrocarbons due to aluminium cations. Burch et al. [360] and
Aquila et al. [361] investigated the activation of C-H bonds in different hydrocarbons
on the surfaces of metal oxides and metal catalysts and reported that alumina can
deactivate highly active catalysts under oxidizing conditions. This group proposed
that the deactivation by alumina was attributed to dehydroxylation (i.e. the
generation of water from the combustion process) on the surface of the alumina. The
production of water from the chemical breakdown of the coal structure is beneficial
in pyrolysis and gasification however in combustion, the presence of water has a
significant impediment effect. Burch et al. [360] proposed that the presence of water
on the active sites of catalysts could result in the active site being blocked for
catalytic oxidation of C-H bonds. Prior investigations have also proposed that some
form of hydroxyl radical blocks reaction sites on the surface of the catalyst [362].
Chapter 6 Combustion of Raw Lignites 215
Ciuparu et al. [363, 364] also proposed that water inhibited the oxygen exchange
between the catalyst surface and the gas phase, as well as reoxidation of a partially
reduced catalyst with oxygen.
The second inhibiting effect by aluminium cations could be the impediment of active
catalysts within the coal matrix during the combustion-oxidation process. Inhibition
of coal combustion by mineral matter occurs via possible restrictions for access of
oxygen to combustible surfaces. Experiments with aluminium treated Loy Yang
lignite have shown the formation of aluminium phases such as MgAlO4 and sodium
aluminosilicates [365]. Furthermore, aluminium in the coal can combine with calcium
and magnesium compounds during combustion to form Ca3Al2O6, CaAl2O4 and
MgAl2O4 [366]. As previously mentioned, the chemical forms AAEM species and iron
(including associations with aluminium ions), and their involvement within the
hydrocarbon macromolecular structure under the different stages of combustion,
could affect the catalytic performance in facilitating the breakdown/oxidation of the
organic components in the coal.
Of the two inhibiting courses of action for the aluminium cations, the obstruction of
active catalysts (eg AAEM cations and iron) is likely to be the more prominent
consequential effect during combustion. The first inhibiting effect of water
occupying active sites on the aluminium would still catalyse the breakdown of C-H
bonds in the coal but at a reduced rate, and the overall effect for combustion would
be a net positive. However, multiple linear regression exhibited a net negative effect
Chapter 6 Combustion of Raw Lignites 216
for aluminium signifying that the impediment of active catalysts in the coal is the
more significant inhibiting process during combustion-oxidation of the coal.
6.8.5 Chlorine Multiple linear regression analysis on the parameters which affect the combustion
reactivity of the Latrobe Valley samples, identified that the chlorine content gave a
weak linear and net positive effect on the rate of combustion with correlation
coefficient constant 0.20 (see Section 6.6.2). Furthermore, when the counter anion
charges of chloride and sulphide are subtracted from the sum of AAEM + iron
univalent charge, the correlation coefficient constant between univalent charge and
combustion reactivity is further improved to 0.87 (Figure 6.11). The sulphide
contents in the Latrobe Valley samples remained relatively unchanged and therefore
changes resulting from anionic influences on the surface of the catalyst was
predominantly chlorine (other than oxygen).
Chlorine in many ways can act as a catalyst poison and as a catalyst promoter. As a
catalyst poison, alkali and alkaline halides pronouncedly inhibit the combustion with
oxygen [367-369]. Minkoff and Tipper [370] reported that the presence of halide salts
such as NaCl significantly hampered hydrogen-oxygen combustion at 460ºC.
Quyn et al. [358] also reported that the retention of chlorine in char can greatly
decrease the reactivity of the char. The inhibiting effect of chlorine when associated
with a potentially active catalyst (eg NaCl, CaCl2, FeCl3 etc) would result in an
overall negative correlation effect in the combustion-oxidation process.
Chapter 6 Combustion of Raw Lignites
217
As a catalyst promoter, chlorine can facilitate the combustion process by freeing up
active sites within the coal matrix and allowing the transformation of cations into a
more active catalytic form. During combustion, the weakly bonded chlorine within
the coal would be dissociated and volatilised as HCl. In Section 4.6, more than half
of the chlorine in the lignite was volatilised during pyrolysis at 500ºC and prior work
has also reported that about 10% of the chlorine in a NaCl-loaded sample was
volatilised at temperatures as low as 200ºC [100, 355]. The departure of chlorine from
the coal/char matrix would constructively free-up corresponding AAEM cations and
iron to potentially become more active catalysts [371]. Furthermore, with the removal
of acidic vapours such as HCl would lead to the promotion and transformation of
metal species into their highly catalytic basic and oxide chemical forms. Finally,
multiple linear regression exhibited a net positive effect for chlorine signifying that
the volatilisation of chlorine out from the coal matrix and allowing inherent cationic
ions to transform into a more active catalytic form is the more dominant process for
chlorine during combustion-oxidation of Latrobe Valley lignites.
Chapter 8 Combustion of Raw Lignites
218
Figure 6.11 Cation concentration (AAEM and iron concentration) minus anion concentration (chloride and sulphide) versus the combustion reactivity value for the raw, water washed and acid washed lignite samples.
YTDMMgLYMNaLYLA
LYMNaMMg
LYLA
LYLA
YMTEMMTE
LYHNa
YEFFe
YEFD
YEFFeYEFFe
YTD
MMgMMg
LYHNa
YEFFe
MMTEYEFDYMTE
YTD
LYMNa
LYHNa
YEFFe
YTP
YMTEYEFD
R2 = 0.87
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Combustion reactivity value
Cat
ion
min
us a
nion
con
cent
ratio
n (m
mol
cha
rge)
Raw lignite Water washed lignite Acid washed lignite
Chapter 6 Combustion of Raw Lignites 219
6.9 Effect of surface area on combustion reactivity The effect of washing the raw lignites with water or with dilute acid on the surface
area of the product is shown in Table 6.11. Washing the lignites with water to
remove water-soluble salts from the coal matrix had no significant affect on the
surface area of the samples (except for YTP where the increase in surface area was
only marginal). In contrast, washing the raw lignites with dilute acid did result in
marginal increases in the surface area in the acid-washed product. This increase in
surface area is likely associated with the metallic cations present in the raw lignites
which are unlikely to contribute to the adsorption capacity of the coal but instead
contribute to the weight of the sample [372-375].
Table 6.11 Combustion reactivity arbitrary values and surface area for the raw, water and acid washed Latrobe Valley coals.
Combustion reactivity
value
Surface area*
(m2 g-1)
Combustion reactivity
value
Surface area*
(m2 g-1) LYLA Raw 0.06 223 YMTE Raw 0.28 213 H2O 0.05 225 H2O 0.23 214 Acid 0.03 227 Acid 0.13 215 LYMNa Raw 0.09 200 YTP Raw 0.16 190 H2O 0.07 210 H2O 0.14 192 Acid 0.03 233 Acid 0.10 206 LYHNa Raw 0.33 171 YTD Raw 0.18 199 H2O 0.29 171 H2O 0.13 207 Acid 0.11 185 Acid 0.08 206 MMTE Raw 0.27 229 YEF Fe Raw 0.36 203 H2O 0.25 232 H2O 0.31 204 Acid 0.09 240 Acid 0.19 205 MMg Raw 0.24 194 YEFD Raw 0.27 227 H2O 0.23 194 H2O 0.20 228 Acid 0.04 220 Acid 0.14 230 * The surface area error is ±2 m2 g-1
Chapter 6 Combustion of Raw Lignites 220
The surface area of the lignite samples was weakly correlated to the combustion
reactivity of the coal with a correlation coefficient of only 0.06 (Figure 6.12). Similar
results were found using multiple linear regression analysis to determine the physico-
chemical properties affecting the combustion reactivity of the lignite samples (see
Section 6.6.1, Table 6.5).
LYHNa
MMg
YEFD
YTPYTP
YTD
YEFD
YMTE YMTE YMTE
MMTEYEFD
YTDYEFFe YEFFE YEFFe
MMTE
MMgMMgLYMNa
YTPYTDLYMNa
LYLA
LYHNa
LYHNa
MMTE
LYLALYLALYMNa
R2 = 0.06
100
120
140
160
180
200
220
240
260
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Combustion reactivity value
Sur
face
are
a (m
2 g-1)
Raw lignite Water washed lignite Acid washed lignite Figure 6.12 Surface area versus the combustion reactivity of the Latrobe Valley
lignite samples
The effect of washing the raw coals with water and dilute acid also had no significant
differences in the large pore volumes or carbon densities (determined by helium
pycnometry) when compared with the parent coal. The results confirm that the
removal of inorganic components from the coal by washing has a much larger impact
on reducing the combustion reactivity of the sample when compared to significant
reductions in porosity resulting from thermal drying.
Chapter 6 Combustion of Raw Lignites 221
6.10 Conclusions Multiple linear regression was a valuable tool for investigating the parameters
affecting the combustion reactivity of coals. Combustion on a diverse suite of well-
characterised lignites sourced from the Latrobe Valley open cut mines showed that
catalytic inorganic species clearly outweighed other physico-chemical effects for
affecting the combustion reactivity of the lignites. The catalytic effects of iron in
combustion have often been overlooked. Multiple regression analysis clearly
identified a very strong positive linear relationship between the combustion reactivity
and the iron content of the coal and moderate relationships between the combustion
reactivity and the magnesium and calcium contents. The combined catalytic effects
of iron, magnesium and calcium accounted for more than 95% of the variation in the
combustion reactivity of the raw coals. In contrast, the aluminium present in the raw
lignites inhibited the combustion-oxidation process whereas chlorine was found to
marginally improve combustion.
The combustion reactivity of Latrobe Valley lignite samples was also found to be
strongly correlated to the sum of the univalent charges of the AAEM species plus the
univalent charge of the acid-extractable iron. In contrast, only a moderate correlation
with the combustion reactivity of the samples was found when the sum of AAEM
species were only considered as the major catalytic components in the coal.
The chemical form of the inorganic components in the hydrocarbon matrix can
influence the rate of the oxidation-combustion process. The SEM-EDX results
showed that the inorganic components in the raw coals were well dispersed and the
Chapter 6 Combustion of Raw Lignites 222
XRD spectra did not reveal the presence of inorganic crystalline salts on the surface
of the coal. Furthermore, a general feature of the Latrobe Valley coals was that much
of the sodium was easily be removed by water washing the coal and that most of the
Mg, Ca, Al, Fe and remaining Na was also extractable with weak acid thus
suggesting ionic associations in the form of carboxylate/phenolates or simple
carbonates/hydroxides/chlorides rather than being contained in clays or other
refractory silicates. The removal of the inorganic species from the macromolecular
coal matrix by water washing or by acid washing the lignite, had a marked effect on
the combustion reactivity and on the peak temperature of the coal sample.
Furthermore, the effectiveness of catalytic inorganic species clearly outweighed pore
volume effects for affecting the combustion reactivity of the lignites. The removal of
inorganic components from the coal by washing or by HTD or MTE has a much
larger impact on reducing the combustion reactivity of the sample when compared to
significant reductions in porosity resulting from thermal drying.
Chapter 7 Combustion of Chars 223
CHAPTER 7 COMBUSTION OF CHARS
7.1 Introduction The combustion reactivity of char is of particular importance in several next
generation coal power plant systems such as advanced pressurised fluid bed
combustion (APFBC). In an APFBC system, the coal is firstly pyrolysed (or partially
gasified) to produce a fuel gas and a stream of char. The char is then burnt separately
in a pressurised fluid bed boiler to generate steam. The fuel gases from the carboniser
are mixed with the combustion products of the char and burnt with air in a separate
combustor system [376].
The physico-chemical properties of char are contrastingly different to that of the raw
coal. The combustion reactivity of char has been found to be affected with variations
in pyrolysis temperature due to alterations in the hydrocarbon macromolecular
network [377-380] and because of catalytic inorganic component transformations within
the char [381-384]. In addition, an increase in aromaticity and changes in porosity could
also have significant impacts on the combustion reactivity of the char [385-387].
In this chapter, the combustion reactivity of the chars produced from the pyrolysis of
an MTE product (see Chapter 4) is investigated. The influence of catalytic
components present in the char and their effect on the combustion-oxidation process
will be compared with the results reported for the raw lignites, thermally dried
products (see Chapter 5) and from the products of water and acid washing (see
Chapter 7 Combustion of Chars 224
Chapter 6). In addition, the effect of the char surface area and large pore volume on
the combustion reactivity of the char is also examined.
7.2 Char reactivity In the previous chapter, it was concluded that the inorganic components are the
backbone to the differences in the combustion reactivity for the raw, water washed
and acid washed lignite samples. This section continues on the combustion reactivity
theme by exploring the combustion behaviour of char samples generated from the
pyrolysis of the MTE product (150°C/5.1MPa) using the quartz fluidized-bed/fixed-
bed reactor (see Section 4.4). The combustion reactivity values of the pyrolysed
MTE samples as a function of temperature and as a function of heating mode are
shown in Figure 7.1.
The combustion reactivity of the char samples decreased as a function of temperature
with marked reductions at pyrolysis temperatures of 700°C and above. In addition,
the char samples that were produced from the fast pyrolysis of the MTE product gave
significantly lower combustion reactivity values when compared to the slowly heated
pyrolysed char samples. Correspondingly, differences in the remaining cationic
components in the macromolecular structure of the char after pyrolysis are likely to
account for the differences in the combustion reactivity as a function of temperature
and as a function of heating mode.
Chapter 7 Combustion of Chars 225
500 600 700 800 900
0.00
0.01
0.02
0.03
0.04
0.05
0.06
Com
bust
ion
reac
tivity
val
ue
Pyrolysis temperature (°C)
Slow heatup Fast heatup
Figure 7.1 Comparisons of the combustion reactivity of chars from the pyrolysis of
MTE lignite operated at the slow and fast heating rate mode, as a function of temperature in the fluidized-bed/fixed-bed reactor.
The major cationic ion volatilised from the char during pyrolysis of the MTE product
was sodium (see Section 4.6) which could account for the differences in the
combustion reactivity of the chars. The form of the sodium ions present in the char
would significantly affect the combustion reactivity of the char. Chen and Yang [388]
reported that the formation of alkali clusters on the surface of the char were
beneficial to the catalytic effectiveness of these species. Furthermore, alkali catalytic
clusters were more effective than alkali phenolated forms on the char surface [389]. At
500°C, the char would contain a high organic oxygen concentration and the sodium
present in the char would likely be bonded to this organic oxygen. The association of
sodium with organic oxygen function groups in the char would therefore, reduce its
Chapter 7 Combustion of Chars 226
effectiveness in catalysing the combustion-oxidation of the char when compared to
the existing alkali clusters. The similar combustion reactivities between the 500°C
slow and fast pyrolysed char samples suggest similar forms of sodium associations
within the char matrix. Increasing the pyrolysis temperature would result in further
loss of organic oxygen and the formation of sodium clusters on the surface of the
char. The pyrolysis heating rate mode showed significant volatilisation of sodium at
temperatures above 700°C whereas for the slowly heated chars only a small loss of
sodium was reported from 700°C to 900°C (see Section 4.6). As a consequence, the
higher proportion of sodium retained in the slow heated char sample would also
contain a higher proportion of sodium clusters which could account for the higher
combustion reactivities when compared to the fast pyrolysed chars. Indeed, multiple
linear regression confirm that the sodium content remaining in the char after
pyrolysis was the major contributing cation in influencing the combustion reactivity
of the char sample (Table 7.1).
Table 7.1 Intercorrelations among reactivity and inorganic ion species for MTE charred samples
Reactivity Na Cl Mg Ca
Reactivity 1.000
Na .918 1.000
Cl -.138 -.434 1.000
Mg .786 .774 -.067 1.000
Ca .630 .687 .018 .820 1.000
Note: * p < 0.05, ** p < 0.01, *** p < 0.001, N = 10
Chapter 7 Combustion of Chars 227
In comparison, the magnesium and calcium contents in the char were not
significantly affected by differences in heating modes and marginally increased as a
function of pyrolysis temperature from 500°C to 900°C (see Section 4.6). Multiple
linear regression analysis also found a strong linear correlation between the calcium
and magnesium contents versus the combustion reactivity of the char. However, the
magnesium and calcium contents in the chars were not as influential as the sodium
content in effecting the combustion behaviour of the char (Table 7.1). That is,
magnesium and calcium had been found to be more active than sodium during the
combustion of the raw lignite, the water washed and the acid washed samples (see
Section 6.6) but for the char samples, this behaviour was reversed.
In contrast to sodium, the formation of crystalline growth of alkaline metal oxides on
the surface of the char have been linked to the reduction in char reactivity as a
function of pyrolysis temperature [390]. Hence, these marked differences in the
performance of calcium and magnesium are likely associated with a number of
factors including: crystalline growth, catalytic dispersion, chemical transformations
and associations within the hydrocarbon matrix and also the chemical ionic forms
during the different stages of pyrolysis which may consequently reduce their
effectiveness in catalysing the oxidation-combustion of the char.
Chapter 7 Combustion of Chars 228
Furthermore, the chlorine content in the char after pyrolysis had a negative impact on
the combustion reactivity of the char thus instigating the inhibition of active catalysts
such as magnesium and calcium, within the char matrix. Similarly, prior work has
also reported that the retention of chlorine in the char can greatly decrease the
reactivity of the char [358, 391].
Also, acid extractable iron which had been found to be the most effective catalyst in
raw coals (see Section 6.6), could not account for the differences in combustion
reactivities between the slow and fast pyrolysed char samples. The iron contents
between a slow or a fast pyrolysed char sample as a function of pyrolysis
temperature showed no significant differences. For example, a slow and fast
pyrolysed char sample at 900°C had the same iron contents. Instead, the reduction in
the combustion reactivity between the slow and fast pyrolysed char samples as a
function of pyrolysis temperature could also be attributed to the chemical
transformation of the iron into a less effective catalytic form during the slow
pyrolysis of the MTE lignite.
As discussed above and in Section 6.8, the concentration of the inorganic
components within the char matrix would be the main factor influencing the
combustion reactivity of carbonaceous materials. Comparison of ionic charge to char
mass ratio versus the combustion reactivity of the chars from the two different
heating modes is shown in Figure 7.2. The combustion reactivity of the char samples
were found to decrease as the inorganic concentration charge balance (Na+ + Ca2+ +
Mg2+ + Fe2+/3+ - Cl-) to char mass ratio decreased, however the chars produced under
Chapter 7 Combustion of Chars
229
the different heating regimes behaved remarkably different (Figure 7.2). For the
chars collected from fast pyrolysis, the inorganic charge to char mass ratio decreased
considerably at pyrolysis temperatures above 600°C and as a consequence, the
combustion reactivity of these chars were significantly lower than the slowly
pyrolysed chars. Furthermore, the profile for the fast pyrolysed chars shown in
Figure 7.2 suggests that the inorganic components in the char are more effective in
catalysing the combustion-oxidation process than the inorganic forms present in the
slow pyrolysed chars. These differences in the combustion profiles (Figure 7.2)
imply that the pyrolysis heating modes affect the chemical transformations of
catalytic cationic components within the macromolecular structure of the char during
the different stages of pyrolysis.
230
0.00 0.01 0.02 0.03 0.04 0.05 0.060.008
0.010
0.012
0.014
600°C 500°C
700°C800°C
900°C
Con
cent
ratio
n io
n ch
arge
to c
har m
ass
ratio
(mol
uni
vale
nt io
n ch
arge
per
gra
m d
b)
Combustion reactivity value
500°C
600°C
700°C
800°C
900°C
Slow heatup Fast heatup
Chapter 7 Combustion of chars
Figure 7.2 Concentration on charge to char mass ratio versus combustion reactivity of the char from the slow and fast pyrolysis experiments using the fluidized-bed/fixed-bed reactor
Chapter 7 Combustion of chars 231
7.3 Porosity of chars Recapping previous porosity results, the MTE process was found to be the most
effective drying process in reducing the large pore volume in the final product
however despite these large pore volume reductions, only marginal differences were
evident between the combustion reactivity of the raw lignite and its products. Also,
the effect of surface area (or micropore volume) on the combustion reactivity of raw,
water washed and acid washed Latrobe Valley lignites was found to be weakly
correlated to the combustion reactivity of the samples. Furthermore, the effectiveness
of catalytic inorganic species in the coal clearly outweighed pore volume effects in
the combustion reactivity of raw lignites.
In this section, the effect of porosity on the combustion reactivity is re-investigated
for char samples instead of raw lignites. Combustion tests were performed on the
char samples produced from a MTE sample during the slow and fast pyrolysis
experiments in the quartz fluidized-bed/fixed-bed reactor (see Section 4.4). The
purpose of this re-visit is to compare the physico-chemical properties of char
materials and their effects on the combustion reactivity; and also to compare these
correlations with those previously reported for the unpyrolysed samples. In addition,
the effect of the micropore volume (which is associated to the adsorptive capacity of
the sample) on the combustion reactivity of the sample has not clearly been resolved
because surface area differences between the raw lignites were relatively small (see
Table 6.11).
Chapter 7 Combustion of chars 232
The pore volume of Loy Yang MTE coal (150°C/5.1MPa) and the pore volumes of
chars from the pyrolysis of MTE operated at the slow and fast heating rate mode, as a
function of temperature in the fluidized-bed/fixed-bed reactor are shown in Figure
7.3. Interestingly, the char particles showed larger pore volumes than the parent MTE
product despite loosing a significant amount of mass during the pyrolysis process.
For example for the 900°C char, 50% of its original weight (db) was volatilised
during pyrolysis however its large pore volume remained high at 0.42cm3g-1
(Figure 7.3). Similarly, Sainsbury and Hawksley [392] reported that Yallourn chars
prepared above 800°C with rapid heating rates, showed large pores and that the
overall shrinkage of the lignite particle was only small despite losing 25 per cent of
its weight during pyrolysis. Similar findings have also been made for chars from
other Latrobe Valley lignites [393].
0.71 0.70 0.69 0.670.60
0.52 0.48
0.70 0.68 0.660.59
0.52 0.47
0.06 0.05 0.14 0.170.20
0.220.23
0.05 0.14 0.180.21
0.230.24
0.27
0.540.51 0.47
0.460.45
0.42
0.550.51 0.46
0.440.43
0.39
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
LY MTE(parentsample)
400 500 600 700 800 900 400 500 600 700 800 900
Pyrolysis temperature (ºC)
Tota
l vol
ume
(cm
3 g-1)
Vol. occupied by carbon Micropore volume Large pore volume
Fast heat-up pyrolysisSlow heat-up pyrolysis
150°
C/5
.1M
Pa
Figure 7.3 Comparisons of pore volume of Loy Yang MTE coal (150°C/5.1MPa) and chars from the pyrolysis of MTE operated at the slow and fast heating rate mode, as a
function of temperature in the fluidized-bed/fixed-bed reactor.
Chapter 7 Combustion of chars 233
It is yet unclear on whether the micropore volume which is associated with the
absorptive capacity of the material has any positive affect in facilitating the
combustion of the material. Kamishita et al. [394] had postulated that the reduction of
a lignite char reactivity in air was attributed to both a decrease in active surface area
and deactivation of catalytic inorganic components. One would expect that a material
with a high absorptive capacity would facilitate the rate of exchange between oxygen
and vapourous volatiles on the surface of the micropores whereas the presence of a
large pore volume would facilitate the pathway of these gases out of the
macromolecular structure. Furthermore, a high surface area should be beneficial in
the dispersion of active catalysts on the surface of the char and in the reduction of
alkaline metal clusters which have been reported to deactivate the effectiveness of
alkaline catalysts in the char [390]. Subsequently, it is speculated that a high absorptive
capacity and large pore volumes should be beneficial to the combustion process. The
surface area of the char samples from the pyrolysis of MTE lignite operated at the
slow and fast heating rate mode, as a function of temperature in the fluidized-
bed/fixed-bed reactor is shown in Figure 7.4.
The surface area of the char samples increased as a function of pyrolysis temperature
and only small differences were evident between the char samples produced from the
slow and fast heating rate regimes. At 900°C, the rapidly heated pyrolysed char
sample had an adsorptive capacity of almost 900m2g-1 which is comparative to some
activated carbon adsorbents [395-402]. In contrast, the surface areas of the char samples
do not show any positive evidence in enhancing the combustion process. That is, as a
function of pyrolysis temperature, the surface area of the char samples significantly
Chapter 7 Combustion of chars 234
increased whereas the combustion reactivity substantially decreased. The reduction
in combustion reactivity in the char samples with increasing pyrolysis temperature is
clearly influenced by the extent of volatilisation of active catalytic species from the
coal macromolecular structure during pyrolysis whereas surface area and large pore
volumes had no noticeable affect.
400 500 600 700 800 900100
200
300
400
500
600
700
800
900
Sur
face
are
a (m
2 g-1)
Pyrolysis temperature (°C)
Slow heatup Fast heatup
Figure 7.4 Comparisons of the surface area of chars from the pyrolysis of MTE lignite operated at the slow and fast heating rate mode, as a function of temperature
in the fluidized-bed/fixed-bed reactor.
Chapter 7 Combustion of chars 235
To further elucidate the influence of surface area and porosity on the combustion of
chars, combustion tests were also performed on a synthetic activated carbon
containing no inorganic components and with a high surface area (1068m2g-1) and a
significantly large pore volume (Figure 7.5). The total pore volume of the synthetic
activated carbon relative to the char samples from the pyrolysis of the MTE product
as a function of pyrolysis temperature is shown in Figure 7.6.
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.0010.010.11
Pore radius (μm)
Incr
emen
tal i
ntru
sion
(m
L g-1
)
Figure 7.5 Pore size distribution of a synthetic activated carbon from Helsa-werk
Vol. occupied by carbon Micropore volume Large pore volume
Fast heat-up pyrolysisSlow heat-up pyrolysis
Figure 7.6 Comparisons of pore volume of a synthetic activated carbon and the pore volumes of chars from the pyrolysis of a MTE sample, operated at the slow and fast
heating rate mode, as a function of temperature in the fluidized-bed/fixed-bed reactor.
Combustion tests on the synthetic activated carbon at 400°C showed that despite its
high surface area and pore volume, the activated carbon burnt very slowly and gave a
combustion reactivity value of <0.001. That is, at 400°C, the synthetic activated
carbon took more than 24 hours to completely burn a 2mg sample which is
considerably longer when compared to the coal samples combusted in this study (eg
of the coal samples, the Loy Yang Low Ash (LYLA) acid washed sample had the
lowest combustion reactivity value and it took less than 90 minutes to completely
burn). The results definitively support that the surface area and pore volumes of
carbonaceous materials have negligible affect on the combustion reactivity and that
inorganic components in the carbon macromolecular structure are the major
promoters in enhancing the combustion reactivity of the sample.
Chapter 7 Combustion of chars 237
7.4 Conclusions The individual inorganic components in coal/char are the major contributors in
dictating the rate of combustion-oxidation at 400°C. Pyrolysis temperature and
pyrolysis heating mode were found to significantly affect the combustion reactivity
of chars at temperatures above 600°C. The concentrations of individual cationic
components in the macromolecular structure of the char after pyrolysis account for
the differences in the combustion reactivity of the char samples.
The volatilisation of sodium during pyrolysis of the MTE product can explain the
differences in the combustion reactivity of the chars. Multiple linear regression also
confirmed that the sodium content remaining in the char after pyrolysis was the
major contributing cation in influencing the combustion reactivity of the char
samples.
Transformations of inorganic components during pyrolysis play important roles in
determining char reactivity. Magnesium and calcium were more active than sodium
during the combustion of the raw lignite but for the char samples, this behaviour was
reversed. The formation of crystalline growth of alkaline metal oxides on the surface
of the char during pyrolysis is likely responsible for the reduction in the effectiveness
of these cations in the combustion-oxidation of the char. The presence of chlorine in
the char also inhibited the catalytic activity of the inherent inorganic components.
Chapter 7 Combustion of chars 238
The combustion reactivity of the char samples were found to decrease as the
• Micropores (from CO2 adsorption, i.e. 0.25nm< p.r. <2nm [420])
• Large pores (intruded by mercury, i.e. 1.5nm< p.r. <1000nm)
Thus, the total internal pore volume (intra-particle volume) can be taken as the sum
of the micropore and large pore volumes. A representation of the overall pore
volume for raw lignite is given in Figure C.1. As noted above, the porosity of raw
lignites cannot be measured reliably by these methods. The true total internal pore
volume of raw Victorian lignite has been estimated as 1.5mLg-1 db [421] (cf 0.60mLg-1
db for the sum of the micropore and large pore volume of the raw lignite in
Figure 3.5).
Appendix C: Pore size distributions 246
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Raw coal
Tota
l vol
ume
(cm
3 g-1
)
Large pore volume
Micropore volume
Vol. occupied by carbon
Mercury porosimetry1000 to 1.5nm
CO2 surface area1.0 to 0.25nm
Helium pycnometry>0.08nm
Pore volume1000 to 0.08nm
Figure C.1 Pore volume distribution of raw lignite attained from mercury porosimetry, CO2 surface area and helium pycnometry (values are given as the radius
of the pores).
Appendix D: Physical measurements of MTE density
247
APPENDIX D Physical measurments of MTE density
In the MTE process, a pellet is obtained as the final product. The advantage of a
pellet is that accurate physical measurements such as wet and dry densities can be
measured using a calliper (see Experimental) before and after drying. These physical
measurements can provide useful information on the extent of shrinkage during
drying to zero moisture prior to analysis. Furthermore, these measurements can help
validate the total pore volume determinations described in Section 3.4. A schematic
of the MTE pellet is shown in Section 2.14.
The wet density of the MTE pellet is expected to increase with more water being
removed from the lignite because of the increase in the ratio of carbon density (i.e. ~
ρ 1.4gcm-3) to water (ρ 1.0gcm-3). Increasing the MTE temperature (fixed pressure)
resulted in a linear increase in the pellet wet density (Figure D.1a) whereas
increasing the MTE applied mechanical pressure resulted in an exponential increase
in the pellet wet density (Figure D.1b). These trends are also expected because the
wet density is dependant on the retained moisture in the MTE pellets (see
Section 3.2). Correspondingly, and not surprisingly, as the wet density of the MTE
product was increased, the wet porosity of the product also decreased (i.e. porosity
(cm3g-1) = 1/density (gcm-3); Figure D.2).
0
5
10
15
20
25
30
1.08 1.13 1.18 1.23 1.28
Wet density (gcm-3)
App
lied
mec
hani
cal p
ress
ure
(MP
a)
R2 = 0.96
100
150
200
250
300
1.08 1.13 1.18 1.23 1.28
Wet density (gcm-3)
Rea
ctio
n te
mpe
ratu
re (°
C)
a b
Figure D.1 Relationship between wet density of the MTE pellet versus (a) reaction temperature (b) applied mechanical pressure
Appendix D: Physical measurements of MTE density 248
249
Appendix D: Physical meas
0
5
10
15
20
25
30
0 5 10 15 20 25 30 35 40 45 50 55 60
Wet porosity (%)
App
lied
mec
hani
cal p
ress
ure
(MP
a)
R2 = 0.97
100
150
200
250
300
0 5 10 15 20 25 30 35 40 45 50 55 60
Wet porosity (%)
Rea
ctio
n te
mpe
ratu
re (°
C)
a b
urements of MTE density
Figure D.2 Relationship between wet porosity of the MTE pellet versus (a) reaction temperature (b) applied mechanical pressure
Appendix D: Physical measurements of MTE density 250
The pellet dimensions measured for the wet density results are of particular
importance for calculating the extent of pellet shrinkage upon drying (see Section
2.14 for Equations). Removing the retained water from the wet MTE pellets resulted
in the collapse of a significant amount of pores. Under the mildest processing
conditions, 125°C/5.1MPa and 150°C/2.5MPa, approximately 65% of the pore
volumes were retained after drying (Table D.1). A higher proportion of the original
pore volume was preserved with increasing processing temperature and applied
mechanical pressure, thus suggesting an increase in lignite structure rigidity. Within
the experimental parameters investigated, the effect of temperature was more
effective than applied pressure in conserving the internal pore structure during drying
to zero moisture. At 250°C/5.1MPa, approximately 87% of the original pore volume
in the wet pellet was maintained after drying. These results further emphasize the
caution in comparing the pore volumes between raw and mildly processed lignites
because of the significant pore collapse that occurs during complete drying.
Table D.1 Pore volumes of wet and dry MTE pellets and the proportion of the original pellet size upon drying to zero moisture.
The proportion of the original pellet size that was maintained after drying to zero
moisture was highly correlated to the wet density of the final product (r2 value of
0.98; Figure D.3c). In general, with an increase in MTE processing temperature
and/or applied mechanical pressure, the product’s large pore volume decreased (see
section 3.4), its rigidity and hardness increased and as a consequence to these
changes, the extent of shrinkage upon drying decreased. These physical changes to
the pellet’s properties are further supported by the reasonable linear correlation
(r2 value of 0.8; Figure D.4a) of the wet versus dry MTE pellet density and the
excellent correlations between dry density versus retained moisture, and dry density
versus wet porosity of the pellets (Figure D.4b and c, respectively). Unfortunately,
due to time limitations of this study, this correlation has only been proven for the
MTE processed Loy Yang lignite and further work is required to establish whether
these relationships are also true for other coals. Nevertheless, the significance of
these correlations is that the pellet dry density (as measured by the calliper
procedure) can be used to extrapolate the moisture content and wet porosity of the
MTE pellet. Also the meaningfulness of the mercury porosimetry and the helium
pycnometry measurements which was questioned earlier because of the shrinkage
and the internal pore collapse of lignite products when completely dried, can be
elucidated by comparing the physical dry density measurements obtained from
measuring the MTE pellet and the dry density values given by the analytical
techniques.
Appendix D: Physical measurements of MTE density 252
0
5
10
15
20
25
30
90 92
94 96 98 100
Proportio of original size (%)
App
lied
mec
hani
cal p
ress
ure
(MP
a)
R2 = 0.98
100
150
200
250
300
90 92 94 96 98 100
Proportion of original size (%)
Rea
ctio
n te
mpe
ratu
re (°
C)
a b
R2 = 0.98
1.08
1.10
1.12
1.14
1.16
1.18
1.20
1.22
1.24
1.26
1.28
90 92 94 96 98 100
Proportion of original size (%)
Wet
den
sity
(gcm
-3)
c
n
Figure D.3 Relationship between proportions of the original size of MTE pellet after drying to zero moisture versus (a) reaction temperature (b) applied mechanical pressure (c) wet density of the MTE pellet.
Appendix D: Physical measurements of MTE density 253
R2 = 0.99
0
5
10
15
20
25
30
35
40
45
50
55
60
0.8 0.9 1.0 1.1
Dry density (gcm-3)
Ret
aine
d m
oist
ure
(% H
2 O
, wb)
R2 = 0.80
.9 1.0 1.1
Dry density (gcm-3)
b
0.8
0.9
1.0
1.1
1.2
1.3
1.4
0.8 0
Wet
den
sity
(gcm
-3)
a
R2 = 0.98
0
10
20
30
40
50
60
0.8 0.9 1.0 1
Dry density (gcm-3)
Wet
por
osity
(%)
c
.1
Figure D.4 Relationship between the dry and wet density of MTE pellets, processed under different conditions
Appendix D: Physical measurements of MTE density
254
The large pore volumes measured analytically with mercury porosimetry were
considerably lower than the physical pore volumes of the solid MTE pellet (Table
D.2), in particular for the milder processed MTE products (e.g. 125°C/5.1MPa and
150°C/2.5MPa). These lower values are likely due to the unaccounted micropore
volumes (measured by gas adsorption; ~0.06 cm3g-1; Table 3.5) and the pore regions
between 1 to 1.5nm and 0.08 to 0.25nm, respectively. Taking these unaccounted
regions into consideration, the results obtained by analytical methods (except for the
milder processed products) are very close to the dry pore volumes measured with the
solid MTE pellet (Figure D.5).
Based on the very good correlations between the physically measured dry density of
the pellet versus the pellet’s wet density, and against the pellet’s retained moisture
then it may also be possible to translate the total pore volumes measured by mercury
porosimetry and gas adsorption back to the retained moisture content and the wet
porosity of the thermally treated lignite products. However, caution should still be
taken when interpreting the total pore volumes of completely dried products because
unless the extent of shrinkage is known, then calculation of the wet porosity would
be difficult. For the purpose of this study, the total pore volumes of the completely
dried products will be used to compare the differences in the combustion reactivity of
each of the products (see Chapter 3).
Appendix D: Physical measurements of MTE density
255
Table D.2 Comparison between large pore region (measured by mercury porosimetry) and pore volume calculated by measuring the dimensions of the dry
MTE pellet
Large pores (1000 to 1.5nm
pore radius) (cm3g-1) ±0.02
Pore volume
dry (cm3g-1)
Difference
(cm3g-1)
MTE 125°C / 5.1MPa 0.28 0.48 0.20
MTE 150°C / 5.1MPa 0.27 0.38 0.11
MTE 180°C / 5.1MPa 0.27 0.36 0.09
MTE 200°C / 5.1MPa 0.25 0.33 0.07
MTE 250°C / 5.1MPa 0.24 0.25 0.01
MTE 150°C / 2.5MPa 0.31 0.48 0.18
MTE 150°C / 5.1MPa 0.27 0.38 0.11
MTE 150°C / 12.7MPa 0.27 0.37 0.10
MTE 150°C / 25.0MPa 0.23 0.33 0.10
Appendix D: Physical measurements of MTE density
256
256
Figure D.5 Relationship between retained water and pore volume determined from mercury porosimetry and gas adsorption; wet and dry pore volumes determined from the MTE pellet
0.0
0.2
0.4
0.6
0.8
1.0
0.0 0.2 0.4 0.6 0.8 1.0
Retained water (g/g db)
Por
e vo
lum
e (c
m3 g-1
)
Determined from Hg and He pycnometry Wet pore volume (MTE pellet) Dry pore volume (MTE pellet)
milder MTE processed products
Appendix E: Determination of soluble ionic salts and carboxylate ions 257
APPENDIX E Determination of soluble ionic salts and
carboxylate ions
Results from washing the Loy Yang lignite with distilled water and the remaining
ash and inorganic concentrations in the washed product are given in Table E.1. In
addition, Figure 3.8 gives the proportion of water-soluble sodium leached out from
the lignite during the MTE, HTD and SD processes.
Table E.1 Acid extractable inorganics, chlorine and total sulphur in raw and water washed lignite (100g of wet raw lignite, actual mass in g) a.
Process condition Na Ca Mg Al FeNPb Stot Cl
Loy Yang raw 0.036 0.017 0.030 0.004 0.025 0.12 0.030
Loy Yang washed 0.010 0.020 0.028 0.004 0.028 0.11 0.018
Note, the error is ±0.004g for Na, Ca, Mg, Al, FeNP and Cl whereas error in S tot is ±0.02g b NP = non-pyritic
Comparison of the data from the raw and water-washed lignite showed only minor
differences suggesting that much of the inorganic material in these samples was
strongly bound (Table E.1). The level of sodium in Loy Yang lignite was reduced by
68 ± 11%db and chlorine by 45 ± 16%db. No significant differences were found in
the levels of Ca, Mg, Al, FeNP and Stot with water washing.
Appendix F: Product water from MTE, HTD and SD 258
APPENDIX F Product water from MTE, HTD and SD
Inorganic analysis of the product waters Analysis of the product water collected from each of the experiments helps to
validate the conclusions deduced from the solid products. Table F.1 and Table F.2
give the quantities of inorganic species leached from the lignite during the MTE,
HTD and SD processes. The major inorganic species leached out into the product
water were sodium and chlorine and to a lesser extent sulphate. This suggests that
most of the soluble sodium in Loy Yang lignite was likely ionically bound to the
chlorine ion and to a lesser extent to the sulphate ion. A significant reduction in total
sulphur was not measured in the solid dried product thus suggesting that only small
portion of the sulphur in the lignite was water-soluble. Furthermore, the negligible
changes in Ca and Mg contents of the solid lignite products during drying made it
unsurprising that the levels of these elements in the product water were also small.
Potassium salts, similar to sodium salts, are very soluble in water. The level of
potassium was not measured in the raw lignite or solid products however its
concentration in the product water was marginally higher than Ca. Furthermore,
similar to sodium, the level of potassium in the product water significantly increased
with increasing processing temperatures in HTD and SD thus suggesting that a
majority of the potassium was also in carboxylate form.
The total ionic mass in the product water for all three drying processes was very low.
Appendix F: Product water from MTE, HTD and SD
259
For MTE, approximately 60mg of inorganic material was leached out into the
product water from 100g of wet raw lignite (i.e. 40g db). Similarly, in SD at 250°C,
the total inorganic mass leached was in proportion to the amount leached out in the
MTE process (i.e. from 10g db of raw lignite, 15mg of inorganic matter was
measured). However, for HTD, considerably more inorganic matter was dissolved
into the water at the same processing temperature (i.e. 24mg from 10g db raw
lignite). The higher proportion of inorganic material leached out from the HTD
process was because of the addition of water to the raw lignite to make up the slurry
therefore some washing of the lignite had occurred prior to HTD treatment.
In the solid products, an increase in processing temperature beyond 250°C (for HTD
and SD) resulted in both additional water and additional inorganic matter being
removed from the lignite. Furthermore, it is postulated that at these higher
processing temperatures, carboxylate bonds in the lignite detach resulting in
additional inorganic material being removed. Analysis of the product water for HTD
and SD validate the results of the solid products. The levels of sodium and chlorine,
the two most abundant extractable ions in the lignite, were found to increase in
concentration in the product water with increasing temperature.
The lower analytical error in measuring the inorganic species in the product water
compared to the solid product allowed the behaviour of the inorganic matter in
lignite to be better elucidated. The levels of Ca, which did not exhibit significant
differences in the solid products, did however display a small and gradual increase in
concentration in the product water with increasing processing temperature.
E, HTD and SD 260
Table F.1 Amount of inorganics present and pH for the product water from MTE processing 100g of wet raw lignite (i.e. 40g db raw lignite; all values are in mg)
Table F.2 Amount of inorganics present and pH for the product water from HTD and SD processing 10g of dried raw lignite (All values are in mg)
The presence of Fe and Al could not be detected in the product water SD 250°C 3.8 4.8 0.17 0.28 0.39 6.1 2.8 14.5 30
Error in the product water analysis in mg = 5% (mgL-1) x Volume of water SD 280°C 3.9 5.8 0.25 0.34 0.50 8.2 3.0 18.1 63
SD 300°C 3.8 6.2 0.25 0.38 0.51 9.6 2.6 19.5 109
Note: HTD 180°C and 230°C product water measurements on the pH and inorganic SD 310°C 3.8 6.3 0.26 0.39 0.41 10.2 2.2 19.8 122
analysis were not conducted however the total organic carbon of the product water was SD 320°C 3.8 6.4 0.28 0.40 0.41 10.5 2.7 20.7 142
calculated to 1.8mg and 3.8mg, respectively. SD 330°C 3.8 6.5 0.31 0.50 0.35 11.0 2.5 21.2 175
SD 350°C 3.7 6.6 0.33 0.55 0.36 11.7 2.6 22.1 286
a = Total mass only includes the inorganics and not the total organic carbon mass
b = TOC (Total Organic Carbon)
Appendix F: Product water from MT
Appendix F: Product water from MTE, HTD and SD
261
In SD, the level of sulphate in the product water did not increase significantly with
increasing processing temperature beyond 250°C but a small increase was reported
in the HTD water. These differences are also likely attributed to sulphate ions being
trapped in the SD product because of more water being removed from the lignite as
steam, thus reducing its ability to be leached out from the lignite in liquid form.
In Table J.1 and Table J.2, these data are recalculated to determine the maximum
possible concentration of these species in a product water stream (where there is no
added water). Note that the data are tabulated in mgmL-1 (in other words, parts per
thousand). Furthermore, the molarity concentration (M) of the ionic species in the
product water is also given in Figure F.1 and in APPENDIX J.
In Chapter 3, a minimum HTD processing temperature of 320°C was proposed for
achieving approximately 50% moisture reduction. At this temperature, the total
inorganic mass leached into the product water was 4mgmL-1 (~60% more than MTE
and SD; Table F.4) which is still relatively low when compared to concentrations
from higher ash lignites [153]. A higher proportion of inorganic removal from the
lignite (in particular sodium) is advantageous because of fouling and slagging
problems associated during combustion [104]. However, product waters of very high
salt concentration may require some cleaning before being released back into the
environment.
Table F.3 Concentration of inorganics present for the product water from MTE processing 100g of wet raw lignite (i.e. 40g db lignite; all values are in mgmL-1)
Table F.4 Concentration of inorganics present for the product water from HTD and SD processing 10g of dried raw lignite (All values are in mgmL-1)
Example: Charge balances for HTD 350°C Total positive charge = Na + K+ Ca + Mg = 0.45 + 0.01 + 0.10 + 0.08 = 0.64mmol Total negative charge = Cl + SO4
2- = 0.47 + 0.09 = 0.56 mmol
Appendix F: Product water from MTE, HTD and SD
266
0.0 0.1 0.2 0.3 0.4 0.50.0
0.1
0.2
0.3
0.4
Sodium (mmol charge)
SD
Chl
orin
e (m
mol
cha
rge)
0.0
0.1
0.2
0.3
0.4HTD
Chl
orin
e (m
mol
cha
rge)
0.0
0.1
0.2
0.3
0.4
0.5
Chl
orin
e (m
mol
cha
rge)
MTE
Figure F.2 Relationship between chlorine and sodium ions present in the water
(mmol) collected from the MTE, HTD and SD processes
Appendix F: Product water from MTE, HTD and SD
267
pH of the product water The general rules of ionic equilibrium [422] would predict that the pH should increase
with the charge of the weak acid anions per unit volume (i.e. the charge per unit
volume on the weak acid anions which were not determined in the analysis). These
rules are only applicable in the absence of hydrolysable cations, such as Fe3+ or Al3+.
The level of iron and aluminium did not change in the solid products and were also
undetectable in the product water for all three drying processes. The pH of the
product waters should in theory be equal to the difference between the charge on the
cations and that on the strong acid anions (Cl-, SO42-).
These expectations were partly fulfilled for the lower temperature SD product waters
(Figure F.3), in that the pH tended to increase with the weak acid anion charge per
unit volume calculated from the product water analyses of Table F.5. However, for
MTE and HTD and for SD at higher processing temperatures, this theory could not
be sustained. The carbonate (CO32-) and bicarbonate (HCO3
-) concentrations in the
water at the measured pH and CO2 concentration will be too low to explain the
difference between positive and negative ions (APPENDIX K).
Alternatively, the amount and nature of the organic acid dissolved in the product
water is likely to play a contributing role in its pH. Previously, workers have reported
significant concentration of anions such as acetate in hydrothermal processing
wastewaters [423]. Furthermore, the greater acidic functional group content of the low-
rank lignites [424] would be consistent with a greater organic-acidic functional group
content of the corresponding waters.
Appendix F: Product water from MTE, HTD and SD
268
0 1 2 3 4 5 6 7
-0.04
-0.02
0.00
0.02
0.04
0.06
0.08
Product water pH
SD
Con
cent
ratio
n of
wea
k ac
id a
nion
s (m
ol u
niva
lent
ions
/L)
0.00
0.02
0.04
0.06
0.08HTD
Con
cent
ratio
n of
wea
k ac
id a
nion
s (m
ol u
niva
lent
ions
/L)
0.00
0.02
0.04
0.06
0.08
0.10
Con
cent
ratio
n of
wea
k ac
id a
nion
s (m
ol u
niva
lent
ions
/L)
MTE
Figure F.3 Relationship between the concentration of weak acid anions and the pH of the product waters from MTE, HTD and SD
Appendix F: Product water from MTE, HTD and SD
269
Thus the pH of product waters from the drying processes is probably usually
controlled by the concentration and nature of weak organic acid molecules leached
out during processing into the water and the concentration of alkali and alkaline earth
cations in the water. The pH will increase with the concentration of cations and,
other things being equal, with decreasing average strength of the weak acid.
Total Organic Carbon The major disadvantages of higher processing temperatures include (i) the higher
water vapour pressure in the system and (ii) an increase in the amount of loosely
bounded components (such as waxes and tars) from the lignite during processing.
Corresponding to (ii) there is an increase of total organic carbon (TOC) in the
product water. The latter is illustrated in Table F.1 and in Figure F.4 and
Figure F.5.
On a commercial scale, it is desirable that any dewatering process is able to achieve
a high degree of water removal from the raw lignite while minimising the costs
associated in handling by-products from the process. High TOC concentrations are
undesirable since they will increase the cost of the product water treatment. For
MTE, very high moisture reductions were achieved with only relatively small
amounts of organic material being leached into the product water (Figure F.4 and
Figure F.6). For MTE, the total organic carbon in the water increased linearly with
increasing proportion of water removed from the raw lignite (Figure F.4).
For HTD, an appreciable amount of water was only removed from the lignite at
much higher temperatures than for MTE (or SD) and, under such conditions, the
Appendix F: Product water from MTE, HTD and SD
270
product water also contained very high concentrations of organic material
(Figure F.4 and Figure F.6). For HTD, the TOC in the product water increased
exponentially with increasing proportion of water removed from the lignite
(Figure F.5). Analysis of HTD product waters have shown the presence of phenols
and cresols as the major components [423, 425]. Nakajima et al. [426] have also reported
a similar exponential increase in the product water TOC with increasing HTD
processing temperature with Loy Yang lignite. This high concentration of organic
material in the product water is therefore a major disadvantage for the HTD process.
SD also achieved very high moisture reductions at relatively mild processing
temperatures (e.g. 56%db at 230°C) however the concentration of dissolved organic
material in its product water was an order of magnitude higher compared to an MTE
product of a similar moisture content (e.g. 125°C/5.1MPa; 51%db H2O reduction).
Furthermore, for SD, the TOC in the product was found to increase exponentially
above 180°C (Figure F.5) whereas for HTD, an exponential increase was found
above 230°C. This is consistent to previous HTD work [427].
Appendix F: Product water from MTE, HTD and SD
271
0
50
100
150
200
250
300
0.1
1
10
100
Tota
l Org
anic
Car
bon
(mg)
MTE
0.1
1
10
100 MTE
Tota
l Org
anic
Car
bon
(mg)
0
50
100
150
200
250HTD
Tota
l Org
anic
Car
bon
(mg)
Tota
l Org
anic
Car
bon
(mg)
HTD
0 50 100 150 200 250 300 350 400
50
100
150
200
250
Processing temperature (°C)
SD
Tota
l Org
anic
Car
bon
(mg)
0 50 100 150 200 250 300 350 4000.1
1
10
100SD
Processing temperature (°C)
Tota
l Org
anic
Car
bon
(mg)
Figure F.4 Total Organic Carbon of the product water collected from the MTE, HTD and SD processes at different temperatures. (Note: all TOC values have been calculated on a 10g db).
Figure F.5 The log of Total Organic Carbon of the product water collected from the MTE, HTD and SD processes at different temperatures (Note: all TOC values have been calculated on a 10g db).
Appendix F: Product water from MTE, HTD and SD
272
0 20 40 60 80 10
50
100
150
200
250
0
Tota
l Org
anic
Car
bon
(mg)
SD
50
100
150
200
250
Proportion of water removed (%)
HTD
Tota
l Org
anic
Car
bon
(mg)
50
100
150
200
250
300
MTE
Tota
l Org
anic
Car
bon
(mg)
Figure F.6 Relationship between total organic carbon leached out into the product water versus the proportion of water removed from MTE, HTD and SD (all TOC
values have been calculated on a 10g db).
Note: For MTE the curve fitting is linear, for HTD it is exponential growth and for SD, the curve fitting is Lorentzian.
Appendix G: Sample calculations
273
APPENDIX G Sample calculations
Mass recovery (sample calculation for 150°C / 5.1MPa) 105g (wb) of raw lignite was used in the experiment. This corresponds to 42.20g lignite db After MTE 64.49g of lignite (wb) was recovered Dry mass of recovered lignite was 65.2 % of 64.49g = 42.05g lignite (db)
Therefore, % recovery = 20.4205.42 x 100 = 99.6% db
Water removed (sample calculation for 150°C / 5.1MPa) To work out the proportion of water removed from the lignite during MTE processing from the results In 105g of MTE lignite there is 64.49g of lignite and 35.51g of H2O
This corresponds to 35.51 x49.6420.42 = 23.24g of H2O in the original lignite
Original water content in 105g of lignite = 62.80g
% of water removed = 10080.62
24.2380.62 xg
gg − = 63 % H2O
Appendix G: Sample calculations
274
Sample calculation of the conversion of wt%db values to actual mass values
(example sodium conversion for 150°C / 5.1MPa)
Mass of species (g) = 100
valuewt%db x Mass of lignite sample (db)
Mass of sodium in raw Loy Yang lignite = 100090.0
x 42.20 = 0.038g
Mass of sodium in MTE product = 100045.0
x 42.05 = 0.019g
Sample calculation of % sodium removal for 150°C / 5.1MPa
a The error is ±0.01wt% for concentrations of 0.01 - 0.10wt% and ±0.02wt% for concentrations greater than 0.10wt% for all elements except sulphur, for which it is ±0.05wt%. The error in the ash is ±0.1wt%. b NP = non-pyritic.
Appendix G: Sample calculations
276
Table G.2 Proportion of inorganic material removed and associated errors for MTE, HTD and SD
APPENDIX J Table J.1 Molarity concentration of inorganics present for the product water from MTE processing 100g of wet raw lignite (i.e. 40g db lignite; all values are in M)
Table J.2 Molarity concentration of inorganics present for the product water from HTD and SD processing 10g of dried raw lignite (All values are in M)
Weight loss 130°CWeight loss 150°CWeight loss 180°CWeight loss 200°CWeight loss 230°CWeight loss 250°CWeight loss 280°CWeight loss 300°CWeight loss 310°CWeight loss 320°CWeight loss 330°CWeight loss 350°CTemperature
Appendix M: Combustion of MTE, HTD and SD products
285
APPENDIX M Combustion of MTE, HTD and SD products
(1) 90-125μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (2) 250-500μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (3) 250-500μm particle size, combustion at 450°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion
Figure M.1 Combustion of MTE products produced at different processing temperatures ranging from 125°C to 250°C.
(1) 90-125μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (2) 250-500μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (3) 250-500μm particle size, combustion at 450°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion
Appendix M: Combustion of MTE, HTD and SD products
Figure M.2 Combustion of MTE products produced at different applied pressures ranging from 2.1MPa to 25.0MPa.
(1) 90-125μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (2) 250-500μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (3) 250-500μm particle size, combustion at 450°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion
Appendix M: Combustion of MTE, HTD and SD products
287
2000 3000 4000 5000-100
-80
-60
-40
-20
0
raw, 200°C, 250°C, 280°C, 300°C, 320°C, 350°C
1a
Wei
ght l
oss
(%)
Time (sec)2000 3000 4000 5000
-100
-80
-60
-40
-20
0 2a
Wei
ght l
oss
(%)
Time (sec)2000 3000
-100
-80
-60
-40
-20
0 3a
Time (sec)
Wei
ght l
oss
(%)
1000 2000 3000 40000
20
40
60
80
100
120 1b
Time (sec)
Hea
t Flo
w (m
W)
1000 2000 3000 40000
20
40
60
80
100
120 2b
Time (sec)
Hea
t Flo
w (m
W)
1000 2000 30000
20
40
60
80
100
120
140
160
180
3b
Time (sec)
Hea
t Flo
w (m
W)
0.0 0.2 0.4 0.6 0.80.00
0.02
0.04
0.06
0.08
0.10
0.12 1c
Spe
cific
reac
tivity
(min
-1)
Conversion (daf)0.0 0.2 0.4 0.6 0.8
0.00
0.02
0.04
0.06
0.08
0.10
0.12 2c
Spe
cific
reac
tivity
(min
-1)
Conversion (daf)0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
3c
Spe
cific
reac
tivity
(min
-1)
Conversion (daf)
Figure M.3 Combustion of HTD products produced at different processing temperatures ranging from 200°C to 350°C.
(1) 90-125μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (2) 250-500μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (3) 250-500μm particle size, combustion at 450°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion
Appendix M: Combustion of MTE, HTD and SD products
288
HTD and SD products
289
Appendix M: Combustion of MTE,
Figure M.4 Combustion of SD products produced at different processing temperatures ranging from 130°C to 350°C. (1) 90-125μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (2) 250-500μm particle size, combustion at 400°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion (3) 250-500μm particle size, combustion at 450°C (a) weight loss versus time (b) heat flow versus time (c) specific reactivity versus conversion