Study on the fate of metal elements from biomass in a benchscale fluidized bed gasifier

Filter char, bed material and the product gas stream were sampled and analyzed for a total of 21 elements including Al, Ca, Fe, K, Mg, Na and Si defined as major elements ME, and Ba, Cd,

Study on the fate of metal elements from biomass in a bench-scale fluidized

bed gasifier

Hong Cui⇑, Scott Q Turn, Vheissu Keffer, Donald Evans, Thai Tran, Michael Foley

Hawaii Natural Energy Institute, University of Hawaii at Manoa, 1680 East–West Road, POST 109, Honolulu, HI 96822, United States

a r t i c l e i n f o

Article history:

Received 26 September 2010

Received in revised form 13 July 2011

Accepted 15 July 2011

Available online 2 August 2011

Keywords:

Biomass gasification

Syngas

Trace elements

Fluidized bed gasifier

a b s t r a c t

Two types of biomass fuel were gasified in a steam atmosphere using a bench-scale fluidized bed reactor Filter char, bed material and the product gas stream were sampled and analyzed for a total of 21 elements including Al, Ca, Fe, K, Mg, Na and Si defined as major elements (ME), and Ba, Cd, Co, Cr, Cu, Mn, Mo, Ni, P,

Sr, Pb, Ti, V, and Zn defined as trace elements (TE) The effects of the sampling system and gasification system on the measurement were determined Mass balances for ME and TE are reported for individual elements and overall It was found that most ash particles or metal elements can be captured by a silicon carbide candle filter that removes small particles from the gas phase, but some of volatile elements pass through the filter and are present in the gas stream

Ó 2011 Elsevier Ltd All rights reserved

1 Introduction

Biomass gasification is a carbon neutral technology producing

fuel gas or synthesis gas (syngas), which must be cleaned up and

qualified for downstream devices such as gas turbines, solid oxide

fuel cells (SOFC), or catalysts for liquid chemical/fuel synthesis

Contaminants of concern present in the syngas are tar components

[1–8], alkali metals[9–16], sulfur[1,6,17–21], and chloride species

[11,20–26], as well as trace elements[9,13,15,27–43] Some

spe-cific inorganic elements[44], including As, Be, Cl, Cd, Cr, Hg, Na,

K, P, Pb, Sb, Se, V, and Zn, which may be present as vapor in syngas

at high temperature, have been found to have strong effects on

SOFC performance For alkali metals (Na and K), the concentrations

are limited to 0.1 ppm or 50 ppb in fuel gas according to

specifica-tions for the protection of gas turbine hardware[13,45] A more

strict tolerance of <10 ppb was reported and cited for alkali and

HCl concentration in syngas for liquid fuel production by

Fischer–Tropsch synthesis[46]

Most studies on trace element behavior during combustion or

gasification have been based on theoretical calculation[10,32]or

on experiments with coal[28,34,41], sewage sludge[37]or

mix-ture of these fuels[47], and woody or grassy plants[9,14,47] In

general, their behavior under gasification conditions is broadly

similar to that described with combustion studies[27], although

the compounds these elements form would not be the same

between an oxidizing combustion atmosphere and reducing

gasification atmosphere

Diaz-Somoano and Martinez-Tarazona[32]investigated some potentially hazardous and corrosive elements under the gasifica-tion condigasifica-tions based on the thermodynamic equilibrium calcula-tion, and grouped these elements according to their volatile behavior Group A includes those that may be totally condensed

in gasification gas cleaning and emission conditions, e.g., Mn The elements mainly present in the gas phase would be classified in Group C, e.g., Se, Hg, and B The rest of the elements would form

an intermediate Group B, which also have two subgroups Group B1 would include those elements that are totally or partially in the gas phase at the temperature of hot gas cleaning systems (500–800 °C), as is the case of Co and Be; and Group B2 would

be present in the gas phase at the lower temperature (<500 °C), such as Sb, As, Cd, Pb, Zn, Ni, Cr and V Similar classification was also introduced by Meij and te Winkel[38], in the term of ‘‘relative enrichment’’ (RE), and used for describing the trace elements par-titioning in the multi-stream combustion system Based on the RE factor, elements can be grouped into three classes, which are pres-ent in the multiple streams from boiler or the gas cleanup system Class I elements with RE factor around 1 are defined as elements that do not vaporize, such as Al, Ca, Ce, Cs, Eu, Fe, Hf, K, La, Mg,

Sc, Sm, Si, Sr, Th and Ti Class II contains those elements that con-dense with small particulates, and have lesser RE factors Class III elements are generally enriched in the gas phase, e.g., B, Hg, Se, and have the lowest RE factors in the ash stream Obviously, these efforts attempted to reduce the complexity of data analysis and evaluation by grouping or classifying these elements with similar vapor behavior

The mechanisms affecting the partitioning of trace elements in gasification or combustion processes are complex Emissions of 0016-2361/$ - see front matter Ó 2011 Elsevier Ltd All rights reserved.

doi: 10.1016/j.fuel.2011.07.029

Contents lists available atScienceDirect

Fuel

j o u r n a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / f u e l

Please cite this article in press as: Cui H et al Study on the fate of metal elements from biomass in a bench-scale fluidized bed gasifier Fuel (2011),

elements cannot be predicted from the vapor pressure alone, but

also are dependent on reactor design, operating conditions, and

fuel properties, as well as hot gas clean up systems In a typical

bio-mass gasification system, syngas can be filtered by a candle filter at

high temperature, which implies that those elements associating

with fine particulates can be captured However, operating

tem-perature has been found to significantly impact the capture

effi-ciency For example, more Pb was detected in the fuel gas with

increasing filter temperatures, from <2lg/m3 at 450 °C, to

<40lg/m3at 500 °C, and to 70lg/m3at 580 °C[31]

A mass balance closure on trace elements of 80–120% is

sug-gested to be acceptable, but it is still a challenge for individual

ele-ment Some factors have been recognized for more complete mass

balances and mostly depend on the degree to which they may be

recovered and quantified in samples[37]and the limitation of

ana-lytical equipment and methods[48] In previous work[9], alkali

elements were collected and analyzed in the samples collected

from input and output streams, including feedstock, gas streams,

filter char and bed material, in a bench scale fluidized bed gasifier

The analytical results suggested that elemental mass balances

were highly variable for specific elements, but generally acceptable given the small element masses present in the fuel

Recent interest has developed in fast growing plant species and agriculture waste that can be used as biomass feedstocks to replace fossil fuels in tropical area Efforts are under way to investigate and evaluate their gasification behavior and gas contaminants The current project focuses on the fate of trace elements during steam gasification in a fluidized-bed gasifier Scientific and technical knowledge about the fate of trace elements have been previously gained, but mostly from certain types of fuel with higher contents

of trace elements, such as coal and sewage sludge, during combus-tion and gasificacombus-tion Very little attencombus-tion has been devoted to trace element concentrations in product gas from biomass under steam gasification conditions In this study, Leucaena leucocephala and sugarcane bagasse, fiber derived from Saccharum officinarum, were gasified in a steam atmosphere using a bench-scale fluidized bed reactor The fate of trace elements (Ba, Cd, Co, Cr, Cu, Mn, Mo, Ni,

P, Sr, Pb, Ti, V, and Zn) both in the gas phase and solid phase were investigated Major inorganic elements, including Al, Ca, Fe, K, Mg,

Na and Si, primary components of fine particles, were also

Cold tap water in

Hot tap water out

Gas flow

SS tube

Quartz tube

Teflon tube

Sampling train

Pump Meter

To exhaust burner

Ice bath Fig 1 Gas sampling system for trace element collection from gasifier product gas.

(Before) (After) Fig 2 Pictures of gas sampling impingers before and after sample collection.

Please cite this article in press as: Cui H et al Study on the fate of metal elements from biomass in a bench-scale fluidized bed gasifier Fuel (2011),

quantified in the study The gas sampling method, distribution of

elements in gas and solid streams, closure of mass balances, and

ef-fects of multiple sources were presented

2 Experimental

2.1 Biomass feedstocks

Leucaena leucocephala, a fast growing, nitrogen-fixing tree, and

sugarcane (Saccharum officinarum) bagasse, the fibrous residue from

raw sugar production were used as fuels in the gasification tests

Leucaena was harvested from plots on the University of Hawaii

Experiment Station at Waimanalo on Oahu Bagasse was obtained

from the Gay & Robinson sugar factory at Kaumakani on Kauai

Both fuel lots were dried to ambient equilibrium moisture

con-tent of 10% dry basis and then hammer milled to pass a screen of

3 mm Fuel samples were subjected to ultimate, proximate, and

major ash species (Si, Al, Fe, Ti, Ca, Na, K, Mg, P, S, Cl) analyses Energy content of the fuels was determined Fuel samples were also analyzed for Ba, Be, Cd, Co, Cr, Cu, Mn, Mo, Ni, Pb, Sr, V, Zn,

Th, Sb, As, B, F, Gh, Se, Sn, and U

2.2 Fluidized bed gasifier setup

A description of the bench scale biomass gasification system used in the tests can be found elsewhere [1] For convenience, the brief description of the design and operating parameters are described below

0

100

200

300

400

500

600

700

Trapping solution

(a) Major elements

0

20

40

60

80

100

120

140

Trapping solution

(b) Trace elements Fig 4 Contents and partition of metal elements in the sampling blank (SBL)

samples (Based on 100 g of rinse solution and 200 g of trapping solution).

0 500 1000 1500 2000 2500

Rinse solution for probe Condensate water Trapping solution

(a) Major elements

0 100 200 300 400 500 600

Rinse solution for probe Condensate water Trapping solution

(b) Trace elements Fig 5 Partition of major and trace elements in the gasifier blank (GBL) samples.

Elements contribution in syngas stream

Fuel

SBL

GBL

Syngas

Fig 3 Illustration of elements contribution into syngas stream from fuel, sampling system, and gasifier system.

Please cite this article in press as: Cui H et al Study on the fate of metal elements from biomass in a bench-scale fluidized bed gasifier Fuel (2011),

The gasifier reactor is constructed of 310 stainless steel pipe,

with a bed diameter of 89 mm and a freeboard diameter of

152 mm The reactor is externally heated by four, 4 kW heaters

Pressure taps, thermocouples and probe access ports are located

along the height of the reactor Fuel was fed to the reactor from

a sealed fuel hopper via a variable speed metering screw Nitrogen

and steam were used as fluidizing agents for the tests Nitrogen

was added to the windbox below the distributor by a mass flow

controller Steam was directed to the windbox from a steam

generator that received deionized water from a calibrated,

preci-sion metering pump Nitrogen was used as an inert trace gas to

permit calculation of gas yields and to control fluidization The

bed material consisted of alumina beads with diameters in the

range of 210–420lm

Flow exits the reactor, passes through a heated, silicon carbide

filter, 60 mm in diameter, 490 mm in length, and with a pore size

of 3lm (Pall Process Filtration Corporation, Cortland, NY) Filter

elements have been operated at temperatures ranging from 600

to 850 °C A sampling port is located at the exit of the filter vessel

to provide a slip stream of the product gas for sampling

Electronic signals from thermocouples, pressure transducers,

on-line gas analyzers, and gas meters are processed using two,

32 channel multiplexer amplifiers (Model SCXI-1100, National

Instruments, Austin, TX) and a 12 bit, analog to digital converter

board (National Instruments, Austin, TX) controlled by a personal

computer

2.3 Gasifier tests and sample collection Four tests, which were conducted and identified as T090731 (no fuel), T080724 (Leucaena), T080821 (Leucaena), and T080912 (bagasse), are discussed in this paper These tests include a gasifier blank run (without fuel), and three gasifier runs (with fuel) under the same test conditions

Table 3summarizes the samples and detail testing/sampling/ collecting parameters Briefly, the gasifier was operated at 800 °C with a fuel feeding rate of 1 kg per hour and steam feeding rate

of 2 kg per hour Alumina-silicate beads (10 kg) were used as bed material in the gasifier for each test The temperature of the fil-ter was maintained at 650 °C and the sampling port was located

at the filter outlet In addition, syngas composition is also listed in Table 3for each gasifier run The following samples relevant to trace element partitioning were collected for analyses each test: gas samples, bed materials before and after the test, char recovered from the filter, and char present in the bed at the conclusion of a test

2.3.1 Gas samples and sampling/recovery method All gas samples were taken from the outlet of filter chamber Fig 1 illustrates the gas sampling system, modified from EPA Method 29, which includes a sampling probe and a sampling train The sampling probe is a 3/800(0.95 cm) id quartz tube, inserting in a stainless steel tube The outlet of the quartz tube was connected to

a series of gas impingers by flexible polytetrafluoroethylene (PTFE) tubing The stainless steel tube housing the quartz sample probe was jacketed with a 25 mm tube that was flushed with cooling water to reduce the gas temperature prior to entering the impingers The sampling train consisted of four borosilicate glass impingers connected in series and fitted with ground taper joint with O-rings seals The impinger train was placed in an ice bath

to condense and collect trace elements from the gasification gas stream The 1st impinger (1500 mL) is an empty and acts as a mois-ture trap The 2nd and 3rd impingers (500 mL) are each loaded with 100 mL of trapping solution (5% HNO3/10% H2O2) The 4th impinger is empty All impingers were fitted with straight tube inlets to reduce the potential for blockage from tar condensation Three or four sets of impingers including a sampling blank were prepared prior to a gasifier test The weight of each impinger was recorded empty and with trapping solution prior to the test, and again after sampling was complete.Fig 2 shows pictures of the sampling train before and after a typical test Steam from the pro-cess stream was condensed in the first bottle and to a lesser degree

in the second and third bottles that were loaded with trapping solution The coloration of the liquid samples is the result of tar components present in the aqueous solution

Three liquid samples were recovered from each sampling train and stored separately in containers Sample A included the rinse solution (100 mL of acetone and 0.1 N HNO3 solutions) from the quartz probe and the connecting PTFE tube Sample B was the condensate recovered from the first impinger Sample C included the liquids collected from second, third, and fourth impingers All impingers and glassware (inserts and U-tubes) were

Table 1

Fuel property and ash analysis for Leucaena and bagasse.

Proximate analysis (% dry basis)

Ultimate analysis (% dry basis)

Ash analysis (% dry basis)

Table 2

.

Please cite this article in press as: Cui H et al Study on the fate of metal elements from biomass in a bench-scale fluidized bed gasifier Fuel (2011),

rinsed three times with 10 mL portions of 0.1 N HNO3solution and

added to the appropriate sample container

A sampling blank (SBL) and a gasifier blank (GBL) were also

col-lected SBL’s were used to investigate the background of the

mod-ified Method 29 sampling train and involved recovering and

analyzing trapping and rinse solutions from sampling equipment

that was not used for sample collection GBL samples were

gener-ated by sampling from the gasifier operating at typical

tempera-tures with N2and steam flow but without fuel These were used

to investigate the background of the gasifier system.Fig 3

illus-trates the definitions for the three types of gas samples in the

study

2.3.2 Solid samples

Fuel and reactant gas inputs were simultaneously stopped at

the conclusion of a test and the system purged with N2flow Solid

samples remaining in the gasifier bed and filter chamber were

collected after each test The char inventory of the bed was

sepa-rated from the bed material and is termed bed char Note that

the bed char samples invariably contained some bed material

and that ash content was higher than char recovered from the

filter, as shown inTable 3

2.4 Elements group and analysis

To simplify the data analysis and discussion, all metal elements

collected in the samples are classified as major elements (ME) or

trace elements (TE) Major elements in the study include Al, Ca,

Fe, K, Mg, Na, and Si; trace elements include Ba, Cd, Co, Cr, Cu,

Mn, Mo, Ni, P, Sr, Ti, V, Pb and Zn

The liquid samples collected from gas sampling train were ana-lyzed for individual elements using inductively coupled plasma/ atomic emission spectrometry (ICP–AES) Solid samples including char samples and bed material were ashed in a muffle furnace at

600 °C to completely remove carbon and then digested by mixed acids (70% HNO3, 37% HCl, 48% HF and saturated H3BO3solution for bed materials; and 70% HNO3, 48% HF, 30% H2O2and saturated

H3BO3solution for filter char) using microwave digestion methods The resulting liquid samples were subjected to elemental analysis using ICP–AES

2.5 Approach for elemental mass balance The input and output flow rate data can be used together to cal-culate a mass balance in a reactor system The flow rates of each element (Qi,j) were calculated based on the element content (Ci,j)

in the stream and the stream flow rate (qi,j) for each input and out-put stream, according to the following equation:

where Qi,jis the flow of the element i in the stream j through the reactor,lg/h (STP); Ci,jis the concentration of the element i in the stream j through the reactor, lg/m3(STP); qjis the flow rate of stream j through the reactor, m3/h (STP)

The sum of all the input and output flow rates of each element gives the total input (Qi,In) and output (Qi,out) streams of each ele-ment, respectively,

Table 3

Average operating and sampling parameters for the gasifier tests.

(no fuel)

b

Gasifier blank (no fuel in).

Please cite this article in press as: Cui H et al Study on the fate of metal elements from biomass in a bench-scale fluidized bed gasifier Fuel (2011),

Qi;In¼Xn j¼1

Qi;jand Qi;out¼Xm

j¼1

where n is number of the input streams, and m is number of the output streams The steady-state mass balance of each element should be

or the closure of each element mass balance can be represented by means of the ratio between output and input streams of the ele-ment A good mass balance requires the ratio of Qi,In/Qi,outto be close

to 1.0

The input streams and output streams are dependent on the gasification process including gasifier type, heating source and any process conditioning units In the fluidized bed system used

in the study, the input streams include fuel, deionized water (steam) and N2and the output streams include raw syngas (includ-ing steam and N2) and char recovered from the filter vessel and bed

Water and N2 were excluded from the input streams in the mass balance calculation under the assumption of negligible ele-ment contribution from pure water (grade II, deionized water) and compressed bottle N2that were used in the tests

In the fluidized bed, unreacted char remained embedded with bed material that were difficult to separate Compared with the weight of char in the filter, the amount of bed char was commonly 13–15% for Leucaena and 5% for bagasse gasification Thus the bed char was not included in the mass balance calculation

A greater amount (10 kg) of bed material was used in the gas-ifier than the amount of fuel input (1 kg/h feedrate with a typical feeding period less than 7 h) Some elements were retained on the surface of bed material[9], but this sink was not saturated completely during the testing period Thus pre- and post-test bed materials were accounted for in the mass balance calculation Fuel input, pre- and post-test bed materials, syngas, and filter char were included in the mass balance calculations Metal ele-ments are assumed to distribute from input streams into output streams by leaving in the gas phase or binding/retaining with solid particulates, thus

Qi;out¼ Qi;gasþ Qi;charþ Qi;bed0 ð6Þ

where, Qi,Inand Qi,outare the summed flow rate of element i in the inlet and outlet streams, respectively Qi,gas, Qi;bed0and Qi,charcan be calculated as following:

Qi;gas¼mi;gas

Qi;char¼mi;char

Qi;bed 0¼mi;bed0

where, mi,gas, mi,charand mi,bedare amounts (lg) of element i content

in the gas, filter char and bed material samples, respectively; and t

is the sample collecting time (h)

In addition, the concentration of element i based on the gas stream can be expressed as:

Ci;g¼mi

Vg

ð10Þ

where Ci,g is the concentration of element i in the gas stream,

lg/m3; Vgis collected gas volume, m3

3 gas

N2

a )

Please cite this article in press as: Cui H et al Study on the fate of metal elements from biomass in a bench-scale fluidized bed gasifier Fuel (2011),

3 Results and discussion

3.1 Fuel properties

The fuel properties including ash analysis data are summarized

inTable 1 Metal element analysis results based on fuel are

pre-sented separately inTable 2

The two fuels have markedly different ash content Leucaena has

an ash content of 0.86% (dry basis) that is lower than bagasse with

4.17% ash content As a percentage of dry fuel, bagasse has a higher

content for most elements with the exception of Ca, K, Mg, Mo, P,

and Sr

3.2 Gas samples

3.2.1 Metal elements from sampling and gasifier

Two types of blank samples, sampling blanks (SBL) and gasifier

blanks (GBL), were collected.Fig 3illustrates the contributions of

metal elements collected in these blank samples to the syngas

sample

For the sampling blank, the elemental content and partitioning

of major elements (ME) and trace elements (TE) in the rinse and

trapping solutions are shown inFig 4 The rinse step contributed

from 40% to 75% of the major elements and from 13% to 87% of

the trace elements in the total, indicating that, depending on the

sample, either the rinse solution or trapping solutions serve as

pri-mary source of elements comprising the background of the SBL Some elements were found in the rinse solution at higher concen-trations than in the trapping solution, which indicates that specific contaminants came from the rinse process, either from impinger attachments or rinse solutions Patterns were not, however, always repeated In test T080724, for example, the rinse solution contrib-uted eight times the amount of Zn compared to the trapping solution, whereas both solutions contributed equal amounts of

Zn in test T080912

The gasifier blank samples (GBL) would include elements from the SBL and any metal elements derived from the reactant gases (N2and steam, if applied), the bed material, the silicon carbide fil-ter element, and any of the working surfaces of the gasifier/reactor and associated piping.Fig 5shows the contents and partitioning of major and trace elements among the ‘‘rinse solution for probe’’,

‘‘condensate water’’ and ‘‘trapping solution’’ It would seem likely that elements would not be distributed equally among the three fractions, but could be enriched in the ‘‘condensate water’’ as the gasifier represents a potentially large source of elements of interest

3.2.2 Metal elements in the syngas samples Trace element tests were conducted on two dates using Leuca-ena as fuel Two gas samples were collected on each of the test dates yielding four samples in total A single test was conducted using bagasse and yielded two gas samples Average results from the gas samples are presented in Table 4 and are reported in

lg/m3on a dry, N2-free, basis at STP, not corrected by any blanks Cases where the standard deviation of the measurement exceeds

0

2000

4000

6000

8000

10000

12000

14000

16000

Rinse solution for probe Condensate water Trapping solution

(a) Major elements

0

1000

2000

3000

4000

5000

6000

Rinse solution for probe Condensate water Trapping solution

(b) Trace elements Fig 6 Elements partition in the samples collected from gas sampling train during

Leucaena gasification (Note: data for ‘‘condensate water’’ sample in T080724-SET1

was unavailable).

0 1000 2000 3000 4000 5000 6000 7000 8000

Rinse solution for probe Condensate water Trapping solution

(a) Major elements

0 200 400 600 800 1000 1200 1400

080912-SET2 080912-SET1

Rinse solution for probe Condensate water Trapping solution

(b) Trace elements Fig 7 Element partitioning in the samples collected from gas sampling train during bagasse gasification.

Please cite this article in press as: Cui H et al Study on the fate of metal elements from biomass in a bench-scale fluidized bed gasifier Fuel (2011),

the mean are shaded and these occur almost exclusively in the

Leucaena data

Pb and Zn were found as the two most abundant trace elements

in the gas samples, regardless of fuel type Mohjtahedi et al study

indicated that both Pb and Zn were bound with Cl to form

dichlo-rides (PbCl2and ZnCl2) in the volatile phase at high temperature

[49] For Leucaena and bagasse, Pb was 1690.3lg/m3 and

447.9lg/m3 in the dry, N2-free gas, respectively, and Zn was

1151.9lg/m3and 144.1lg/m3in the dry, N2-free gas, respectively

As in the previous paragraph, however, higher fuel concentrations

did not translate to higher gas phase concentrations Bagasse has

greater concentrations of Pb (54.6 mg/kg), Zn (44.6 mg/kg),

and Cl (0.71%) compared to Leucaena’s concentrations of Pb

(0.2 mg/kg), Zn (3.48 mg/kg), and Cl (0.13%), as shown inTables 1

and 2 The other trace elements were detected in the gas phase

in varied concentrations, as shown inTable 4

3.2.3 Metal elements distribution in the gas sampling train

Figs 6 and 7summarize the elements contents and distribution

of major elements and trace elements in the sample parts, ‘‘rinse

solution for probe’’, ‘‘condensate water’’ and ‘‘trapping solution’’

for Leucaena and bagasse, respectively It is found that ‘‘rinse

solu-tion for probe’’ contains more major and trace elements in the first

impinger set than those in the second, which indicates that gasifier

system would be clean after the first gas sampling Also, the

lim-ited data show that no primary zone of the impinger train

consis-tently captured more of the analytes, and that any zone could

contribute from 15% up to 70% of the total element mass

Table 5provides a relative break down of the collected total

ele-ments (ToE), major eleele-ments (ME) and trace eleele-ments (TE) in the

fuel, gas phase, gasifier blank (GBL), and sampling blank (SBL)

For Leucaena, the gas phase ratio of TE/ToE varies between 18%

and 35% and is greater than the TE/ToE in the fuel (7.5%), indicating

that these trace elements are more concentrated in the gas phase

than in the fuel Similarly, the ratio for bagasse is between 14% and 20% and is greater than the ratio of TE/ToE in the fuel (6.3%)

Calculation fromTable 5shows that the total concentrations of elements (ToE includes 21 elements) in the product gas are 12,680 and 6000lg/m3on average for Leucaena and bagasse, respectively Similarly, the concentrations of trace elements (TE includes Ba, Cd,

Co, Cr, Cu, Mn, Mo, Ni, P, Pb, Sr, Ti, V, Zn) in the gas phase are 3086 and 994lg/m3 for Leucaena and bagasse, respectively This indi-cates that high concentrations of elements in the fuel do not nec-essarily yield high concentrations in the gas phase, provided that adequate hot gas filtration is applied Apparently ash composition plays an important role Alkali metals (K + Na) are a principal con-cern amongst the major elements The total concentration of (K + Na) is 955.6lg/m3for Leucaena and 1570.1lg/m3for bagasse These values are greater than those measured at VTT’s pressurized fluidized bed gasification test rig immediately before a gas cooler (750–900 °C)[13]and do not meet gas turbine requirements The contributions of fuel, gasifier, and sampling system were calculated and summarized based on test T080912 (bagasse) and listed inTable 6 Bagasse contributes roughly 70% of major ele-ments and trace eleele-ments in the gas phase The gasifier system contributes 20% and 25% for major elements and trace elements, respectively Sampling systems accounted for more than 10% of the major elements and 2.2% of trace elements in the total elements collected in the gas phase

Fig 8shows the individual element contributions (normalized) from the gasifier system (including sampling system) and fuel to the syngas samples For the major elements, K seems to be contrib-uted by the gasifier system more than by the fuel A possible reason

is that past testing with the reactor system used a variety of fuels, some high in K K may have been accumulated during earlier tests and was then re-emitted during the present tests with relatively low-K fuels For the trace elements, Cr, Cu, Mo, and Ni, are

Table 5

Proportions of major (ME) and trace elements (TE) in the total collected elements (ToE).

a

Elements data were deducted by the sampling blank (SBL) from gas samples.

b

Elements data were deducted by the gasifier blank from gas samples.

Table 6

Contribution of major elements and trace elements in the gas phase from fuel, gasifier, and sampling, %.

Please cite this article in press as: Cui H et al Study on the fate of metal elements from biomass in a bench-scale fluidized bed gasifier Fuel (2011),

elements that could be contributed by reactor materials and some

lubricants used in the gasifier system Reed et al.[31]reported high

levels of Cr, Mn, Ni, and Mo, measured in all of gas samples and

suggested it was attributed to corrosive components and MoS2

anti-seize lubricants used in the plant

T080821

T080912

T090731

-20%

0%

20%

40%

60%

80%

100%

Al Ca Fe K Mg Na Si

-Bagasse (T080912) AVERAGE GBL

-20%

0%

20%

40%

60%

80%

100%

-20%

0%

20%

40%

60%

80%

100%

-Lueceana (T080724&T080821) AVERAGE GBL

-20%

0%

20%

40%

60%

80%

100%

Fig 8 Distribution of individual element from fuel and gasifier system (including

sampling system) in the syngas stream produced in Leucaena and bagasse

gasification.

Please cite this article in press as: Cui H et al Study on the fate of metal elements from biomass in a bench-scale fluidized bed gasifier Fuel (2011),

3.3 Metal element deposition on the bed material The primary elements in the fresh, pre-test bed material are Al,

Fe, Si, and Ti, which account for 99.1% of the total elemental mass Besides these, the element contents in the bed material samples of pre- and post-tests, and the difference between them are listed in Table 7

T090731 is a GBL test in which all bed material elements re-mained at the same level or varied little For the fuel gasification tests, T080724, T080821 and T080912, Ca, K, Mg, Na and P have greater contents in the post-test bed materials than that in the pre-test bed materials, whereas Mn levels were stable Greater content indicates elements present in the fuel adhered to the bed material surface during gasification For the other trace elements, the data are not as clear The present data confirmed previous results [9] that Ca, K, Mg and P are retained by bed material

3.4 Metal element content in the filter char Table 8shows the elemental content of the filter char samples Elements present in higher concentration include the major elements, Al, Ca, Fe, K, Mg, Na, Si and Ti Obviously, filter char samples have varied element composition depending on the fuel type and tests It is noted that Cd was only detected in the gas stream, with non-detectable amounts found in the solid output streams Reed et al.[31]noted that Cd was only quantifiable in the fuel gas during coal gasification, but was concentrated in the hot gas filter fines with addition of sewage sludge pellets in the tests It suggests that Cd concentration in the gas stream is mostly depended on Cd input

3.5 Relative enrichments of elements in the gas phase and filter char Fig 9 shows element distributions in the two main output streams, gas and filter char, for the three tests Cd, Mo, Pb and Zn were partitioned most noticeably to the syngas stream

3.6 Elemental mass balance

In the fluidized bed system, input streams include fuel, water (steam) and N2 stream Output streams include filter char and raw gas (with steam and N2) Bed material serves as an accumula-tion point for elements so both pre- and post-test bed materials were included in the mass balance calculations Any contributions from the water and N2streams can be accounted for in the gasifier blank sample De-ionized water and N2from compressed gas cylin-ders were used for the tests

Table 9presents the ratio of element masses contained in the input and output streams, (Qi,In/Qi,out), in gasification tests of Leuca-ena and bagasse

For Leucaena (T080724 and T080821), most major elements have good mass balance, with the ratios of closure in a range of 0.86–1.33, with the exception of K with closure ratios of 1.37 and 1.66 in two tests Closure for the trace elements, Ba, Co, Cr, Cu

Mn, P Pb, Sr, Ti, V, and Zn, falls within this range Cd and Mo have ratios much less than 1.0, indicating that these elements were not accounted for completely in the output stream and/or were overes-timated in the input stream The closure ratio of Ni is greater than 1.0 indicating that a source of Ni may be present in the system, e.g., the stainless steel reactor

For bagasse (T080912), the mass balance results for most elements were poor

Please cite this article in press as: Cui H et al Study on the fate of metal elements from biomass in a bench-scale fluidized bed gasifier Fuel (2011),