You are on page 1of 8

BULK DENSITY OF WET AND DRY WHEAT STRAW

AND SWITCHGRASS PARTICLES

P. S. Lam, S. Sokhansanj, X. Bi, C. J. Lim, L. J. Naimi, M. Hoque,


S. Mani, A. R. Womac, X. P. Ye, S. Narayan

ABSTRACT. Bulk density is a major physical property in designing the logistic system for biomass handling. The size, shape,
moisture content, individual particle density, and surface characteristics are few factors affecting the bulk density. This
research investigates the effects of true particle lengths ranging from 6 to 50 mm and moisture contents ranging from 8% to
60% wet basis (wb) on the bulk density of wheat straw and switchgrass. Three types of particle densities of straw and
switchgrass measured were: a hollow particle density assuming a hollow cylindrical geometry, a solid particle density
assuming a solid cylindrical geometry, and a particle density measured using a gas pycnometer at a gas pressure of 40 kPa.
The bulk density of both loose‐fill and packed‐fill biomass samples was examined. The calculated wet and dry bulk density
ranged from 24 to 111 kg m ‐3 for straw and from 49 to 266 kg m ‐3 for switchgrass. The corresponding tapped bulk density
ranged from 34 to 130 kg m ‐3 for straw and 68 to 323 kg m ‐3 for switchgrass. The increase in bulk density due to tapping the
container was from 10% for short 6‐mm particles to more than 50% for long 50‐mm particles. An equation relating the bulk
density of stems as a function of moisture content, dry bulk density, and particle size was developed. After the validation of
this bulk density equation, the relationship would be highly useful in designing the logistics system for large‐scale transport
of biomass to a biorefinery. The bulk density and particle density data of uniform particles would be important, if straw and
switchgrass is used for pulping and paper making.
Keywords. Bulk density, Particle density, Straw, Switchgrass, Packing, Porosity, Fibrous biomass.

B
ulk density is an important characteristic of depends on bulk density and flow characteristics of feedstock
biomass that influences directly the cost of (Woodcock and Mason, 1987). A recent study by Ryu et al.
feedstock delivered to a biorefinery and storage (2006) investigated the effect of bulk density on the
cost (Sokhansanj and Fenton, 2006). It also combustion characteristics of biomass. They found that the
impacts storage requirements, the sizing of the material ignition front speed was inversely proportional to bulk
handling system and how the material behaves during density, while the burning rate tends to decrease linearly. As
subsequent thermo‐chemical and biological processes. a physical property, bulk density (rb ) depends on material
(McKendry, 2002). The engineering design and operation of composition (c), particle shape (f) and size (l, d), orientation
transport equipment, storages, and conversion processes of particles (s), specific density of individual particles (rp ),
particle size distribution (PSD), moisture content (mc), and
applied axial pressure (p) (Peleg, 1983; Lang et al., 1993;
Submitted for review in November 2007 as manuscript number FPE Sokhansanj and Lang, 1996).
7262; approved for publication by Food & Process Engineering Institute ò = f ( c , ò , l , d , f , s , PSD , mc , p ) (1)
Division of ASABE in April 2008. Presented at the 2007 ASABE Annual b p
Meeting as Paper No. 076058.
The authors are Pak Sui Lam, ASABE Member, Graduate Research where l and d are length and diameter of particles. The
Associate, Department of Chemical and Biological Engineering, variables in the right side of equation 1 are not all
University of British Columbia, Vancouver, British Columbia, Canada; independent of each other. For example, moisture content has
Shahab Sokhansanj, ASABE Member, Adjunct professor, Department of
Chemical and Biological Engineering, University of British Columbia, an effect on particle density and surface characteristics.
Vancouver, British Columbia, Canada and Distinguished Research Similarly, particle size and distribution depends on upstream
Scientist Oak Ridge National Laboratory, Environmental Sciences processes such as chopping, drying, grinding, sieving, etc.
Division, Oak Ridge, Tennessee; Xiaotao Bi, Professor, C. Jim Lim, Length (l) and diameter (d) define a shape factor (f) for the
Professor, Ladan J. Naimi, ASABE Member, Graduate Research
Associate, Mozammel Hoque, ASABE Member, Posdoctoral Associate
particle. Other factors such as surface characteristics also
Department of Chemical and Biological Engineering, University of British affect the bulk density and flow characteristics of biomass.
Columbia, Vancouver, British Columbia, Canada,; Sudhagar Mani, The agricultural biomass particles are usually cylindrical
Assistant Professor, Faculty of Engineering, Driftmier Engineering Center, or disk shape depending upon the part of the plant they
University of Georgia, Athens Georgia; Alvin Ray Womac, ASABE originate. The long particles with aspect ratio (l/d) of greater
Member, Professor, X. Philip Ye, Assistant Professor, Department of
Biosystems Engineering and Soil Science, The University of Tennessee, than one are classified as cylindrical while the short particles
Knoxville, Tennessee; Sundar Narayan, Professor, Lambton College, (l/d<1) are classified as disk. The packing behavior of these
Sarnia, Ontario, Canada. Corresponding author: Pak Sui Lam, non‐spherical shapes usually starts from studying the
Department of Chemical and Biological Engineering, University of British mono‐sized particles (Zou and Yu, 1996a; Zhang et al., 2006)
Columbia, Vancouver, British Columbia, Canada, V6T1Z3; phone:
604‐827‐3413; fax: 604‐802‐6003; e‐mail: plam@chml.ubc.ca.
and then gradually moves to the particle mixtures with

Applied Engineering in Agriculture


Vol. 24(3): 351‐358 E 2008 American Society of Agricultural and Biological Engineers ISSN 0883-8542 351
different shapes and volume fractions (Yu and Zou, 1996; The moisture content of switchgrass and straw was
Zou et al., 1997). adjusted by spraying a predetermined amount of distilled
Zou and Yu (1996a) and Yu and Zou (1996) developed water uniformly over the samples in a container. The
equations to estimate the porosity of mono‐sized moistened samples were thoroughly mixed and sealed inside
non‐spherical particles by defining a sphericity factor. In a a plastic bag. The bags were placed inside a sealed plastic
subsequent publication, Zou et al. (1997) reported that the container for 48 h at 4°C to allow the samples to equilibrate
packing of long cylindrical particles was dependent on the to three levels of target moisture content (i.e., 20%, 40%, and
length distribution of particles and the porosity of bulk 60% wb). The final moisture contents were measured using
cylindrical particles could not be estimated from packing air oven method (ASAE Standard S358.2 FEB 03, ASABE
relations for spherical particles. The shape and size effect Standards, 2006).
contribute to the packing structure of the non‐spherical
particles and the shape effect dominates the packing structure BULK DENSITY
with the particle length distribution. They assumed a Bulk density of wheat straw and switchgrass was
well‐defined solid cylindrical geometry in their analysis. The determined according to the ASAE Standard S269.4 DEC 91
applicability of this phenomenon to biomass particles may be (ASABE Standards, 2007) for cubes, pellets, and crumbles.
questionable as straws and switchgrass are hollow‐tube Three cylindrical containers each with a specific inside
structures with non‐uniform diameter along the particle diameter (DT) were used for determination of bulk density of
length. different particle sizes (table 1). Figure 1 shows a plexiglass
container loosely filled with the same size switchgrass
OBJECTIVES particles to measure bulk density. Each measurement was
The objective of this research is to measure the bulk repeated for five refills using the same straw or switchgrass
density of straw and switchgrass with respect to some of the sample. Biomass was poured into the container from a certain
measurable physical attributes (particle length and moisture height to facilitate the free flowing of the samples until the
content) and to develop predictive equations for bulk density container was overflowed. The height of pour for 50‐ and
of biomass. 25‐mm long particles was 500 mm and for 12.5‐ and 6‐mm
long particles was 200 mm. The height of pour was measured
from the bottom of the container. The excess material was
MATERIALS AND METHODS removed by striking a straight edge across the top. The weight
of the material with the container was recorded. The net
SAMPLE PREPARATION
weight of the sample was obtained by subtracting the weight
The wheat straw was collected from Richmond Country
of the empty container. For tapped density, the loosely filled
Farms, Richmond, BC. The bales were of rectangular shape container was tapped on the laboratory bench five times by
with a moisture content of 8.45% wet basis (wb) as received.
dropping the container from a height of 50 and 10 cm for the
The switchgrass was a round bale acquired from a farm in
long and short particles, respectively. Tapping was done by
Manitoba. The moisture content of switchgrass was 7.96% the same individual. Filling and tapping was repeated until
wet basis (wb) as received. The moisture contents were
the container was overflowed. The filled container was
measured using ASAE Standard S358.2 FEB 03 (ASABE
weighed to 0.01‐g precision. Bulk density was calculated by
Standards, 2006) for forages. The procedure specifies drying dividing the mass over the container volume.
5‐ to 20‐g samples of chopped or ground material in a
convection oven at 105°C for 24 h. The moisture content is
PARTICLE DENSITY
determined by subtracting the dried mass from original wet
The straw and switchgrass are hollow cylindrical particles
mass divided by the wet mass.
as shown schematically in figure 2. Cleaned stems of straw
The stems of straw and switchgrass were sorted out
manually. The leafy material and small pieces were removed. and switchgrass were oven dried at 105°C for 24 h. The
volume of a single particle was measured in three ways: by
The stems were cut into sizes of 50.80, 25.40, 12.70, and
geometrical dimensions (assumed as hollow and solid) and
6.35 mm (2, 1, 1/2, and 1/4 in.) using a scissor. The actual
particles size of switchgrass and straw are listed in table 1. by gas displacement. These volume measurements are
described below.

Table 1. Measured dimensions and mass of single particles (n=10).


Length, L (mm) Diameter, Din (mm) Thickness, t (mm) Mass, mp (g)
Type of
Biomass Average SD Average SD Average SD Average SD
56.598 2.708 2.648 0.419 0.146 0.053 0.029 0.010
34.503 2.045 2.914 0.542 0.145 0.035 0.024 0.008
Wheat straw
17.172 1.432 2.952 0.346 0.125 0.025 0.010 0.003
9.078 0.530 3.380 0.485 0.216 0.073 0.008 0.003
51.774 1.990 2.600 0.581 0.340 0.108 0.064 0.019
28.974 3.013 2.855 0.565 0.581 0.239 0.043 0.019
Switchgrass
14.721 1.079 2.916 0.562 0.486 0.152 0.025 0.015
8.042 1.064 2.420 0.468 0.486 0.152 0.008 0.002

352 APPLIED ENGINEERING IN AGRICULTURE


radius of the hollow stem (m), Vp, hollow is the actual volume
of the hollow cylindrical samples (m3), and Vp, solid is the
actual volume of the solid cylindrical samples (m3).

Gas Displacement
A gas multipycnometer (Quantachrome Corp., Boynton
Beach, Fla.) that measures the volume of a group of particles
was used. The measurement is based on the pressure
difference between a known reference volume (VR ) and the
volume of the instrument's sample cell (Vc ). Nitrogen is used
as the gas to fill the reference and sample cells. The pressure
is set at around 40 kPa (near maximum as specified by the
instrument specifications). The pycnometer volume of the
sample (Vp) is calculated from
⎛P ⎞ (6)
V p = Vc − V R ⎢⎢ 1 − 1⎟⎟
⎝ P2 ⎠
where P1 is the pressure reading after pressurizing the
reference cell (kPa), P2 is the pressure after connecting the
Figure 1. A container filled with 14.7‐mm Switchgrass particles for bulk reference cell to the sample cell. The particle density (ρs ) of
density measurement.
the sample is its mass mp divided by the pycnometer particle
volume (Vpvc )
Dimensional Measurement
mp (7)
The length (l), diameter (Dout ), and the thickness (t) of ò =
10 particles in each size group were measured using a s
V pvc
150‐mm Mastercraft digital caliper with a precision of
0.01 mm. The weight (mp ) of each particle was measured The instrument readings on each sample were repeated three
using A&D GR200 digital balance with a precision of 0.1 mg. times.
The average diameter of the samples was taken as the average
value of the diameter in both ends and the middle of the WET AND DRY BULK DENSITY RELATIONSHIPS
samples along the length of the sample. The particle density The relationship between the wet (w) and dry (d) bulk
measurement of each sample was determined from the density of the samples is represented as a mixture equation in
following equations: two forms of equation 8 (Peleg, 1983) and equation 9
mp (2) (Hollenbach et al., 1982)
òp =
V p , hollow 1 1− Mw Mw (8)
= +
ò ò ò
b d w
Dout − Din (3)
t=
2 ò = ò (1 + M ) (9)
b d w
where ρb is the wet bulk density of the samples (kg m‐3) at
V p , hollow = (2rt + t 2 )pl (4)
moisture content of Mw, ρd is the bulk density (kg m‐3)
measured using a bone dried sample. Mw is the moisture
V p , solid = pr 2 l (5) content of the wet samples (decimal wet basis), ρw is the bulk
density of water (1000 kg m‐3).
where rp is the particle density of the samples (kg/m3), mp is
the mass of the samples (kg), t is the thickness of the wall of
the tube (m), Din is the inner diameter of the samples (m),
Dout is the outer diameter of the samples (m), r is the inner RESULTS
PARTICLE VOLUME AND DENSITY
L Table 1 lists the length, diameter, tube thickness, and the
mass of each of the cut stems of straw and switchgrass. The
t average lengths deviated from the target lengths by a few
millimeters. In most cases, the coefficient of variations
(standard deviation divided by average) among 10 samples
in each length group were around 0.10 to 0.15 (10% to 15%).
Dout
Outside diameter of the particles ranged from 2.4 to 2.9 mm
Din except the 9‐mm wheat straw. Switchgrass stems were more
than twice in thickness than the straw and that reflected in the
Figure 2. Dimensions of a stem assuming cylindrical shape (L: length,
Dout: outer diameter, t: wall thickness, Din: inner diameter of the hollow mass of most of the individual particles.
stem).

Vol. 24(3): 351‐358 353


Table 2. Volume and density of individual particles calculated from dimensional measurement and from gas displacement.
Volume (cm3) Particle Density (g cm‐3)
Nominal Lengths
Biomass (mm) Hollow Solid Pycn[a] Hollow Solid Pycn[a] Pycn[b]
50.80 0.074 0.316 0.027 0.497 0.093 1.058 1.275
25.40 0.050 0.239 0.026 0.557 0.106 0.930
Wheat
12.70 0.021 0.118 0.008 0.519 0.084 1.178
6.35 0.023 0.083 0.008 0.395 0.093 1.038
50.80 0.169 0.285 0.102 0.544 0.241 0.630 1.179
25.40 0.193 0.189 0.069 0.377 0.227 0.619
Switchgrass
12.70 0.080 0.101 0.038 0.454 0.244 0.657
6.35 0.036 0.038 0.012 0.366 0.223 0.657
[a] Readings from gas displacement pycnometer at 40‐kPa pressure.
[b] Particle density of the finely ground biomass (sieved through 60 mesh 0.25 mm) measured with pycnometer.

Table 2 lists the calculated volumes from dimensional tub ground wheat straw using 50.8‐mm screen size was about
measurements and gas displacement (pycnometer). The 56 kg m‐3 and was highly varied with particle size
volumes calculated from a hollow cylinder was less than that distribution of the sample. Ebeling and Jenkins (1985)
calculated from a solid cylinder. Single particle volumes reported that loose bulk density of wheat straw was about
measured by pycnometer showed switchgrass have more 20 kg m‐3 and was increased to 65 kg m‐3 when the sample
solids than the straw. The volumes of individual particles was chopped into smaller sizes( 10% moisture content).
measured using pycnometer were generally less than the Switchgrass had a higher bulk density (on average 125 kg m‐3
volumes calculated from dimensional measurements for loose‐fill and 147 kg m‐3 for tapped) than straw (50 kg m‐3
indicating that the gas in pycnometer had filled some of the for loose‐fill and 65 kg m‐3 for tapped). Bulk density
cellular voids. The last column lists the particle density of increased with moisture content of particles. The tapped bulk
powdered switchgrass and straw. The particle density for densities of wheat straws and switchgrass ranged from 34 to
wheat straw appears to be nearer to the particle density of 130 kg m‐3 and 68 to 323 kg m‐3, respectively, for the
finely ground sample. But for switchgrass, the particle moisture content of 8% to 60% for four different particle
density of finely ground sample was almost twice the particle sizes. The percent increase in bulk density due to tapping was
size of the large particles. The applied pressure of 40 kPa on the highest for longer biomass particles than for shorter
gas was not adequate to penetrate all of the voids in the particles. For most tests, tapping switchgrass produced a
switchgrass particle. larger percentage increase in bulk density than wheat straw.
Visually, we could not detect any influence of moisture
BULK VOLUME AND DENSITY content on bulk density change due to tapping.
Table 3 lists the results of mass and bulk volume of dry and
wet biomass and the calculated bulk density for each size RELATIONSHIP BETWEEN WET AND DRY BULK DENSITY
group. The ratio of the container diameter to the length of the Equations 8 and 9 were used for predicting the wet bulk
particles ranged from 5.38 to 24. Zou and Yu (1996b) states density from the dry bulk density. We assumed the measured
that the diameter of the column should be at least 20 times the bulk density at 8% moisture content to be the dry bulk
equivalent volume diameter of cylindrical particle in order to density. We then calculated bulk density of switchgrass and
minimize the edge effect. We noticed larger variations in straw for moisture contents at 20%, 40%, and 60% wb.
measured bulk mass of larger particles than in measured bulk Figure 3 shows sum of square of errors (SSE) between the
mass of smaller particles and this may be due to the edge predicted and measured values. For switchgrass, the SSE for
effect. This needs to be verified by constructing a larger equation 9 was higher than the SSE for equation 8. The SSE
diameter container for measuring the bulk density of larger values were small for larger particles. The SSE values
particles (i.e. to have the ratio of container diameter over the increased for the smaller particles. For wheat straw, the SSE
largest particle length more than 20). If we assume a diameter for equation 9 was slightly lower than the SSE for equation 8,
to particle length ratio of 20 for the 50‐mm length particle, but the SSE increased at high moisture of 60%. For
the container diameter should be at least a meter. equation 8, the SSE values appeared to be more random than
The wet bulk density increases as the particle size those for equation 9. We recommend equation 8 for further
decreases and moisture content increases. The least bulk analysis.
density was 24 kg m‐3 for 8% moisture content of 56.6‐mm The relationship between the dry bulk density and particle
straw particle in a loose fill and that increased to a maximum size can best be described by power law equations in the
of 111 kg m‐3 for 61.2% moisture content with the 9.1‐mm following form,
particle. The tapped density of the straw increased from
ò = ax −b (10)
34 kg m‐3 for 8.5% moisture content to 130 kg m‐3 for 61.2% b
moisture content. The least bulk density of switchgrass for where rb is the dry bulk density and x is the nominal particle
loose fill was 49 kg m‐3 for 8% moisture content of 51.8‐mm size. Table 4 lists constants a and b and R2 for equation 10.
particle and increased to the maximum density of 323 kg m‐3 The estimated values for a and b in table 4 were averaged and
for 59.4% moisture content of 8‐mm particles when tapped. it was observed that the constant a did not change much
Arthur et al. (1982) reported that the average bulk density of

354 APPLIED ENGINEERING IN AGRICULTURE


Table 3. Volume and mass of straw and switchgrass for calculating bulk density.
Loose Fill (n = 5) Tapped Dense (n = 5 )
Moisture Content (%wb) Stem
Length Mean SD Volume Density Mean SD Density
Target Measured SD Mean (mm) Mass (g) Mass (g) (cm3) (kg m‐3) Mass (g) Mass (g) (kg m‐3)
Wheat Straw
8.50 0.1 56.6 287.0 10.0 11880 24 548.1 3.7 46
34.5 383.2 19.1 11880 32 401.0 10.1 34
8
17.2 105.0 1.5 2300 46 139.7 1.7 61
9.1 121.5 1.9 2300 53 168.1 1.9 73
21.6 0.7 56.6 322.0 6.7 11880 27 440.0 7.1 37
19.6 1.0 34.5 415.0 6.0 11880 35 525.4 7.6 44
20
20.7 0.6 17.2 98.2 6.5 2300 43 112.2 0.5 49
18.7 1.3 9.1 119.6 1.7 2300 52 141.6 1.7 62
43.7 1.0 56.6 348.8 3.0 11880 29 526.6 7.3 44
48.1 2.4 34.5 491.6 4.0 11880 41 683.8 3.9 58
40
43.1 12.1 17.2 131.6 1.1 2300 57 161.6 1.1 70
41.4 5.0 9.1 162.6 0.9 2300 71 197.2 2.3 86
58.6 1.1 56.6 634.0 10.0 11880 53 912.8 3.7 77
56.4 1.5 34.5 472.6 22.7 11880 40 728.4 7.9 61
60
64.6 2.7 17.2 192.8 6.0 2300 84 239.8 1.9 104
61.2 3.6 9.1 255.6 2.1 2300 111 300.0 1.6 130
Switchgrass
8.0 0.2 51.8 587.4 4.4 11880 49 804.0 8.2 68
29.0 1012.6 13.6 11880 85 1102.6 0.9 93
8
14.7 286.0 4.6 2300 124 355.6 2.4 155
8.0 50.0 -- 310 161 50.0 -- 161
23.0 1.5 51.8 616.8 5.6 11880 52 815.2 13.3 69
24.0 0.2 29.0 972.0 3.7 11880 82 1252.6 6.4 105
20
19.0 1.9 14.7 275.6 2.3 2300 120 337.2 2.6 147
22.8 2.0 8.0 55.0 -- 390 141 55.0 -- 141
43.0 3.6 51.8 704.4 10.2 11880 59 1063.4 11.3 90
39.2 3.2 29.0 1096.4 12.3 11880 92 1492.4 11.2 126
40
41.5 1.3 14.7 350.0 2.9 2300 152 432.6 2.4 188
36.0 1.3 8.0 67.0 -- 410 165 67.0 -- 196
57.6 0.7 51.8 881.4 40.9 11880 74 1336.0 31.6 112
54.7 0.4 29.0 1753.6 7.3 11880 148 2181.2 13.7 184
60
58.6 1.1 14.7 512.8 5.9 2300 223 662.6 4.3 288
59.4 3.5 8.0 101.0 -- 380 266 101.0 -- 323

between loose and tapped fill for straw and switchgrass. DISCUSSION
Overall estimates for a and b were found by trial and error Biomass samples used in this research had several
procedure as a = 166.24, b = 0.41 for straw, and a = 574.5 and characteristics that would influence the applicability of the
b = 0.54 for switchgrass. For estimating the tapped density, results to real situations: (1) samples were clean and made up
the values of b decreased to 0.33 for straw and 0.46 for solely of stems, (2) samples were cut to exact lengths and the
switchgrass. bulk density of each size groups were measured separately,
Figure 4 is the plot of wet bulk density of biomass and (3) samples were artificially wetted by spraying water on
calculated by combining equations 8 and 10, and using the the sample in order to attain high moisture conditions. The
values of a and b from table 4. lengths studied were limited to a range between 6 mm to just
1 1− M w MW (11) over 50 mm. Uniform size of straw and switchgrass particles
= + results in low bulk density due to large pore spaces among
òb ax −b òw
particles. The higher bulk density due to reorientation of the
Visually, the equation represents the wet bulk density of particles was achieved by tapping the container five times on
biomass reasonably good. As it was shown in figure 3, the the work bench. In practice, this kind of tapping is not
predictions become less accurate for smaller size particles. practical and needs to be replaced by vibration. The
direction, amplitude, and frequency of vibrations to achieve

Vol. 24(3): 351‐358 355


2500
WS Series
WS Parallel
2000
SG Series
SG Parallel

Square of errors 1500

1000

500

0
20% 20% 20% 20% 40% 40% 40% 40% 60% 60% 60% 60%
50 mm 25 mm 12 mm 6 mm 50 mm 25 mm 12 mm 6 mm 50 mm 25 mm 12 mm 6 mm

Figure 3. The square of errors of the values obtained from series and parallel equation for the prediction of wet based bulk density from dry based bulk
density at different moisture content and particle sizes of wheat straw (WS) and switchgrass (SG).

maximum bulk density needs to be explored. The real bulk CONCLUSIONS


biomass is made up of particles of varying sizes and various The following conclusions are drawn from this work:
parts of the plant (leaves, nodes, stems). The question is how 1 The bulk density of switchgrass ranged from 49 to
deviations from a bulk of stems of a single size would affect 266 kg m‐3 for loose fill and from 68 to 288 kg m‐3 for
the bulk density of a group of heterogeneous material. The packed fill after tapping. For straw these values ranged
high moisture content biomass used in these experiments was from 24 to 111 kg m‐3 for loose fill and from 37 to
artificially moistened by adding water. The artificially 130 kg m‐3 for the packed fill.
wetted material will have bulk properties different from 2 Longer particles resulted in a larger percentage increase in
naturally moist biomass. It is important to quantify and packing than the smaller particles by tapping. In some
develop relationships that would reflect the real samples. The cases the increase in bulk density as a result of packing was
present results would be highly valuable where uniform sized more than 50%. Increase in bulk density of small particles
straw and switchgrass particles are used. In paper making (6 mm) was about 10%.
process, the properties of uniform particle sizes are very 3 For estimating bulk density of a moist material from the
important during pulping and fiber making process. Wheat bulk density of dry material we found less bias and
straw and switchgrass are potential, low‐cost feedstock for deviation from actual measured values when using
paper making and preparing uniform sized particles are very 1 1 − M w M W where rb is the wet bulk density of
important to maintain minimum fiber length during pulping = +
process. òb òd òw
biomass at a moisture content of Mw (decimal fraction wet
basis), x is the particle length (mm).

Table 4. Relationship between biomass particle size and bulk and tapped densities.
Bulk Density, ρb (kg/m3) Tapped Density, ρtb (kg/m3)
Moisture
Biomass Content a b R2 RMSE[a] a b R2 RMSE
8 113.79 0.39 0.98 5.08 150.95 0.37 0.99 9.00
20 98.80 0.32 0.97 3.08 97.88 0.25 1.00 3.03
Wheat straw
40 161.64 0.43 0.99 6.63 155.20 0.31 1.00 6.39
60 290.71 0.51 0.99 14.51 261.04 0.37 1.00 14.34
8 502.10 0.57 0.98 11.04 494.82 0.49 0.96 14.45
20 378.10 0.49 0.96 10.58 397.28 0.43 0.94 16.40
Switchgrass
40 479.77 0.52 0.94 16.50 448.51 0.40 0.92 17.37
60 938.03 0.61 0.93 26.18 938.67 0.52 0.93 28.84
[a] The root mean square error.

356 APPLIED ENGINEERING IN AGRICULTURE


350

WS 8% mc WS 20% mc
300
WS 40% mc WS 60% mc

Bulk density (wb) (kg/m3)


250 SG 8% mc SG 20% mc

SG 40% mc SG 60% mc
200

150

100

50

0
0 10 20 30 40 50 60
Nominal particle size (mm)

Figure 4. Measured (symbols) and calculated wet bulk density (solid lines) of switchgrass (SG) and wheat straw (WS) at different sizes and moisture
contents for the loose packing using equation 11 and constants from table 4.

4 Individual particle density was dependent upon the way ASABE Standards. 2007. S269.4. Cubes, pellets and crumbles -
we calculated or measured the volume of the particles. Definitions and methods for determining density, durability and
Assuming that a particle was a solid cylinder, the particle moisture content. St. Joseph, Mich.: ASABE.
density of wheat straw was about 90 kg m‐3, assuming it Ebeling, J. M., and B. M. Jenkins. 1985. Physical and chemical
properties of biomass fuels. Transactions of the ASAE 28(3):
was a hollow cylinder, the particle density was 500 kg m‐3,
898‐901.
and measuring the volume using a pycnometer, the Fasina, O. O. 2006. Flow and physical properties of switchgrass,
particle density was roughly 1100 kg m‐3. The peanut hull and poultry litter. Transactions of the ASABE 49(3):
corresponding values for switchgrass were 230 for solid, 721‐728.
400 for hollow, and 650 kg m‐3 for pycnometer Hollenbach, A. M., M. Peleg, and R. Rufner. 1982. Effect of four
measurements. We concluded that the biomass needs to be anticaking agents on the bulk characteristics of ground sugar. J.
pulverized in order to have an accurate particle density Food Sci. 47(2): 538‐544.
(solid density) measurements. Lang, W., S. Sokhansanj, and F. W. Sosulski. 1993. Bulk volume
5 Further work is required to determine an optimum size for shrinkage during drying of wheat and canola. J. Food Process
the container for bulk density measurement and a Eng. 16(4): 966‐5334
McKendry, P. 2002. Energy production from biomass (part 1):
procedure for loose and pack fills. We also need to
Overview of biomass. Bioresour. Technol. 83(1): 37‐46.
measure the volume change of a given quantity of biomass Mani, S., L. G. Tabil, and S. Sokhansanj. 2004. Grinding
of mix of sizes. This work should be repeated with freshly performance and physical properties of wheat and barley straws,
harvested material of high moisture content. corn stover and switchgrass. Biomass Bioenergy 27(4): 339‐35.
Peleg, M. 1983. Physical characteristics of food powders. In
ACKNOWLEDGEMENT Physical Properties of Food, 293‐321. Westport, Conn.: AVI
This research is made possible through a USDA/DOE Publishing Co., Inc.
Grant to the University of Tennessee and a sub contract to the Ryu, C., Y. B. Yang, A. Bhor, N. E. Yates, V. N. Sharifi, and J.
University of British Columbia. The financial assistance Switenbank. 2006. Effect of fuel properties on biomass
combustion: Part 1. Experiments‐fuel type, equivalence ratio and
from the Oak Ridge National Laboratory in support of the
particle size. Fuel 85(7‐8): 1039‐1046.
logistics project is acknowledged. NSERC's Discovery and Sokhansanj, S., and W. Lang. 1996. Prediction of kernel and bulk
Strategic Grants supported the work of graduate students on volume of wheat and canola during adsorption and desorption.
this and related projects. J. Agr. Eng. Res. 63(2): 129‐136.
Sokhansanj, S., and J. Fenton. 2006. Cost benefit of biomass supply
and pre‐processing, BIOCAP research integration program
synthesis paper. Ottawa, Canada: BIOCAP Canada Foundation.
REFERENCES Yu, A. B., and R. P. Zou. 1996. Modifying the linear packing model
Arthur, J. F., R. A. Kepner, J. B. Dobie, G. E. Miller, and P. S. for predicting the porosity of nonspherical particle mixtures. Ind.
Parsons. 1982. Tub grinder performance with crop and forest Eng. Chem. Res. 35(10): 3730‐3741.
residues. Transactions of the ASAE 25(6): 1488‐1494. Zhang, W., K. E. Thompson, A. H. Reed, and L. Beenken. 2006.
ASABE Standards. 2006. S358.2. Moisture measurement - Forages. Relationship between packing structure and porosity in fixed
St. Joseph, Mich.: ASABE. beds of equilateral cylindrical particles. Chem. Eng. Sci. 61(24):
8060‐8074.

Vol. 24(3): 351‐358 357


Zou, R. P., and A. B. Yu. 1996a. Evaluation of the packing mp Mass of the samples (g)
characteristics of mono‐sized non‐spherical particles. Powder P1 Pressure reading after pressurizing the reference
Technol. 88(1): 71‐79. volume (kPa)
Zou, R. P., and A. B. Yu. 1996b. Wall effect on the packing of P2 Pressure after including the volume of the cell (kPa)
cylindrical particles. Chem. Eng. Sci. 51(7): 1177‐1180.
r Inner radius of the hollow stem (m)
Zou, R. P., X. Lin, A. B. Yu, and P. Wong. 1997. Packing of
cylindrical particles with a length distribution. J. Am. Ceram. t Thickness of the inner wall of the samples (mm)
Soc. 89(3): 646‐652. Vp, hollow Actual volume of the hollow cylindrical samples
Zou, R. P., C. L. Feng, J. Q. Xu, and A. B. Yu. 2005. Predictions of (m3)
the porosity of multicomponent mixtures of wet coarse spheres. Vp, solid Actual volume of the solid cylindrical samples (m3)
Ind. Eng. Chem. Res. 44(22): 8401‐8408. Vpvc Pycnometer particle volume (m3)
Woodcock, C. R., and J. S. Mason. 1987. Bulk Solids Handling: An ρb Wet based bulk density of the samples (kg/m3) at
Introduction to the Practice and Technology. Glasgow, Scotland: moisture content of Mw
Blackie and Son Ltd. ρb,wet_tapped Tapped bulk density (wb) of the samples
(kg/m 3)
ρb,wet_loose Loose bulk density (wb) of the samples (kg/m3)
NOMENCLATURE ρd Dry based bulk density (kg/m3) at bone dry sample
Ap Actual surface area of the particle (m2) ρp Particle density of the samples after drying (kg/m3)
a, b Biomass species constants ρs Particle density measured by pycnometer (kg/m3)
Din Inner diameter of the samples (mm) ρw Bulk density of water (1000 kg/m3)
Dout Outer diameter of the samples (mm)
DT Diameter of the container for bulk density
measurement (mm)
l Particle length (mm)
Mw Moisture content of the wet samples (decimal wet
basis)

358 APPLIED ENGINEERING IN AGRICULTURE

You might also like