Aeration Strategies in Composting
Topics covered
Aeration Strategies in Composting
Topics covered
ABSTRACT. The effects of aeration strategies (one−directional airflow, one−directional airflow with recirculation, reversed−
direction airflow, and reversed−direction airflow with recirculation) on the composting process under intermittent aeration
with a 15 min on and a 45 min off cycle for all tests were experimentally investigated. Intermittent aeration was used in these
studies because it has been shown to have higher efficacy (more energy efficient and less odor emission) than continuous aera-
tion for the one−directional airflow case. Paper mill sludge and broiler litter were composted using four 208 L stainless steel
reactor vessels modified to allow studies on four different aeration strategies with a C:N ratio of 25:1. Compost mass was
divided into five layers by polypropylene screens. Detailed analysis of the experimental data showed that aeration strategies
yielded different temperature, moisture, CO2 , O2 , and decomposition profiles. One−directional airflow yielded the highest
temperature, moisture, and decomposition gradient across the composting bed. The experimental data for one−directional
airflow with recirculation showed that the compost temperature profiles were smaller than for the no−recirculation case and
were similar to what was reported in the literature. Reversed−direction airflow had the smallest moisture gradient and
maximum moisture retention. Reversed−direction airflow with recirculation showed temperature and moisture profiles that
were a superimposition of the results for the two independent aeration strategies. Oxygen levels, due to the on/off cycle pattern
and decomposition rates, limited performance and it was concluded that each system was not operated at the optimum point.
Results of these studies will be used to validate finite−difference computer models, which can then be used to find the optimum
operating points for each system.
Keywords. Aeration strategies, Broiler litter, Composting, Paper mill sludge.
C
omposting, a process used to accelerate the biolog- Air is supplied during composting for three purposes:
ical decomposition and stabilization of organic (1) to meet the stoichiometric oxygen demand of decomposi-
substrates, is a way of treating agricultural and in- tion, (2) to remove the water from the compost (maintain
dustrial wastes to make them environmentally be- compost porosity, dry compost), and (3) to remove the heat
nign. Maximum decomposition with minimal odors is generated and maintain compost temperatures within a range
accomplished when the appropriate nutrients (C, N, P, S, wa- favorable for the process. Airflow rates selected to meet these
ter, etc.), pH, and oxygen are supplied to the working organ- needs should be optimized to save energy while maximizing
isms. The factors of C:N ratio, pH, and moisture content are the composting rate. Some factors that enter into the
controlled at the beginning of the process, while other factors optimization are aeration control strategy (continuous,
such as temperature and oxygen levels are controlled during on−off, recycled, etc.), kinetics, and marketing requirements
the process. In most modern composting systems, the process (compost stability) of the compost. Numerous texts discuss
control relies on forced ventilation to supply air in place of the factors that affect aeration requirements, e.g., Haug
natural ventilation. (1993) and Keener et al. (1993, 2002).
One−directional airflow (ODA), the most common meth-
od used during composting, leads to temperature gradients,
moisture gradients, and subsequently non−uniformity of
Article was submitted for review in November 2003; approved for decomposition within the compost matrix. Generally, the
publication by the Structures & Environment Division of ASAE in August
2004.
inlet layer is vulnerable to desiccation due to dry air entering
Use of a commercial trade name or company does not imply the system, while the outlet layer may be exposed to
endorsement of the products but is included only to assist the reader. condensed water resulting from cooling of saturated air
The authors are Kamil Ekinci, Assistant Professor, Faculty of leaving the system. To overcome this problem, two airflow
Agriculture, Agricultural Machinery Department, Suleyman Demirel management strategies have been suggested. They are
University, Cunur Isparta, Turkey; and Harold M. Keener, ASAE
Member Engineer, Professor, David L. Elwell, ASAE Member Engineer, one−directional airflow with air recirculation (ODA−R) and
Senior Researcher, and Frederick C. Michel, Jr., ASAE Member reversed−direction airflow (RDA). ODA−R aerates the
Engineer, Assistant Professor, Department of Food, Agricultural, and compost with a mixture of the hot and humid exhaust gases
Biological Engineering, Ohio Agricultural Research and Development from the compost bed and fresh ambient air, i.e., make−up air.
Center, The Ohio State University, Wooster, Ohio. Corresponding author:
Harold M. Keener, Department of Food, Agricultural, and Biological
This approach was introduced in the 1970s for tunnel systems
Engineering, OARDC, 1680 Madison Avenue, Wooster, OH 44691; phone: (Beyer et al., 1998). RDA changes the airflow direction
330−263−3856; fax: 330−263−3670; e−mail: keener.3@osu.edu. periodically to minimize moisture gradient in the composting
The broiler litter was received in two batches from Case at four levels: bottom (11 cm above floor), section 1 (31 cm),
Farms (Shreve, Ohio) and consisted of a mixture of sawdust section 2 (51 cm), and top (61.5 cm). Measurements for inlet
and manure accumulated during the 49 d growth period for and exhaust air were taken from plenum and exhaust
broilers. The broiler litter was mixed by shovels to obtain (71.5 cm). The composting reactors were equipped with
homogeneity in terms of properties and stored in 20 L closed three tee−ports to facilitate temperature and oxygen measure-
containers until the experiments began. Broiler litter 1 and ment at three levels (11, 31, and 51 cm) and three small
broiler litter 2 had an initial pH of 8.3 and 8, N of 3.0% and straight ports to facilitate temperature measurement at three
3.9%, C/N of 11.6 and 9.7, and ash of 17.8% and 19.7%, levels (at plenum, 61.5, and 71.5 cm).
respectively (table 1). Temperature readings from type K thermocouples were
Two batches of the mix of paper mill sludge and broiler recorded every 10 min with a Digi III Kaye Data Logger and
litter were prepared for experiments (table 1). A spreadsheet an MFE tape recorder. Dew point temperature of ambient air
was used to formulate the mixes with a C:N ratio of 25:1 and was measured with a dew point meter (model 6000, E&EG)
moisture content of 44% (w.b.) (Ekinci, 2001). The pH (7.8 and were recorded every 10 min. Oxygen and carbon dioxide
and 7.6) and C:N ratio (24.3:1 and 23.5:1) of the initial gases were withdrawn from the 11 cm and 31 cm levels of the
compost mixes for mix 1 and mix 2 were similar. However, composting reactors (the same ports as the temperature
some differences between initial moisture and ash content of measurements) for oxygen analysis (model 4000, MSA) and
the mixes occurred because of variation of the compost infrared carbon dioxide analysis (model 3000, MSA). The
materials. Initial ash content for mixes 1 and 2 were 57% and gases were pumped to the analyzer every 80 min per location.
61%, and initial moisture content were 47% and 45%, Oxygen and carbon dioxide gases for each location were
respectively. analyzed at the same time, but there was a 10 min sequenced
time difference between the two locations (11 cm and 31 cm
EXPERIMENTAL EQUIPMENT levels). The top layer oxygen concentration was analyzed
Four 208 L (55 gal) stainless steel reactor vessels with an using a portable oxygen analyzer (model 320P, Teledyne)
inner diameter of 57 cm and height of 68.5 cm (Elwell et al., three times a day.
2001) were modified to allow studies on four different Airflow was supplied to each reactor with an induced draft
aeration strategies. Each had a perforated stainless steel floor blower (model 5C087, Dayton) through 47.6 mm ID PVC
with a hole size of 3.2 mm (diameter) and 40% opening to pipe. The fan was controlled with an on/off timer (model
provide a plenum for air distribution under the compost. Each T1905, Intermatic) with a 15 min on and 45 min off schedule
reactor was wrapped with 50 mm of polystyrene and had a lid for all experiments. Stainless steel, sharp, square−edged, thin
covered with a 50 mm thickness of polystyrene. The reactor plate, concentric orifice plates in the piping located 20 diam-
rested on a plywood platform (10 cm height) that was eters downstream from fans or elbows were used to measure
insulated below with 50 mm thickness of polystyrene. the airflow. The pressure drop across orifice(s) was measured
Sampling of the compost mass was done for each 20 cm depth by a pressure cell (model C264, Setra) and recorded every
at 71.5 cm
at 61.5 cm
at 51 cm
Ports for O2
at 31 cm and CO2
K−Type measurement
Thermocouples at 11 cm
plenum
Fan Orifice
Gate valve
Heating system
Orifice plates H O T
Exhaust air O T
Gate valve
Fresh air inlet
Styrofoam box
10 min. Airflow was calculated from the pressure readings specific humidity, and oxygen concentration) were compared
using equations in Spink (1967). The next sections discuss to check the performance of the insulated box. The mixed
the design of the four composting systems investigated. stream pipe was insulated separately to maintain its tempera-
ture.
One−Directional Airflow Relative humidity of the mixed air stream (“H” in fig. 2)
Layout of the system for ODA with intermittent aeration
was measured with a relative humidity probe (model
is shown in figure 1. Airflow supplied to the compost reactor
HMP135, Vaisala) every 10 min. The relative humidity probe
was controlled using a ball valve directly downstream from was calibrated with saturated salt solutions at room tempera-
the fan and upstream of the orifice used to measure airflow.
ture (ASTM, 1997).
One−Directional Airflow with Recirculation Condensation traps were built to collect condensate water
The composting reactor for ODA−R consisted of intermit- from the system. Condensation happened when the tempera-
tent aeration and an air recirculation system (fig. 2). The ture of the air inside of the box was less than the exhaust
recirculation system had an exhaust stream pipe, an ambient temperature. Therefore, it was critical to control the box
stream pipe, and a mixed air stream pipe. The exhaust gas temperature at the exhaust temperature.
stream and ambient air stream were mixed and drawn through
Reversed−Direction Airflow
the blower. The exhaust stream pipe had a gate valve to adjust The setup for RDA used a ball valve to set airflow and an
the recirculation ratio and a concentric orifice plate to
orifice plate to measure airflow, similar to ODA. However,
calculate the amount of airflow returned. The ambient air
direction of airflow was changed periodically in this
stream pipe had a concentric orifice plate to calculate the treatment using two 3−way valves and one 2−way gate valve
amount of airflow drawn in. The mixed air stream pipe had
(fig. 3). For upward airflow, 1, 2, 7, and 6 were open while
a gate valve to adjust the airflow and a concentric orifice plate
4, 3, and 5 were closed. For downward airflow, 1, 3, 5, 7, and
to calculate the amount of airflow delivered to the reactor. 4 were open while 2 and 6 were closed.
The air recirculation ratio was calculated and recorded when
airflow was on. Temperatures of the exhaust air stream, Reversed−Direction Airflow with Recirculation
ambient air stream, and mixed air stream were measured The RDA−R system reversed the direction of airflow
within the pipe system using type K thermocouples. Temper- while recirculating the exhaust air. There were three 3−way
ature readings from the thermocouples were recorded every valves and one 2−way gate valve to recirculate exhaust air
10 min with a Digi III Kaye Data Logger and an MFE tape
recorder. Oxygen concentration using a portable oxygen
analyzer was measured three times a day in the ambient air
stream and mixed air stream (“O” in fig. 2), as well as the
exhaust air stream.
In order to maintain adiabatic mixing of the exhaust air
stream and the ambient air stream, the recirculation system
was enclosed in a polystyrene box with a wall thickness of
5 cm. The box was heated by five 75 W light bulbs.
Temperature inside the box was controlled by a controller
6
(model UDC 2000, Honeywell) using the reactor exhaust 5
temperature as a reference, based on the feedback from a type 1 7
3
K thermocouple inserted into the box. Additionally, a 2
circulation fan was placed inside the box to distribute the heat 4
homogenously. The box’s purpose was to maintain the
exhaust temperature in the pipe from the composting system. Figure 3. Composting system for RDA.
Calculated and measured variables (mixed temperature,
and to reverse the direction of airflow (fig. 4). For upward Compost temperatures were measured at the 1st, 3rd, 4th,
airflow, 1, 2, 7, 6, 8, and 9 were open while 3, 4, 5, and 10 were and 5th layer in the vessel. A second−order polynomial
closed. For downward airflow, 1, 3, 5, 7, 4, 10, and 9 were equation was used to determine the average temperature in all
open while 2, 6, and 8 were closed. The temperatures from the layers including average 2nd layer temperature.
the pipe system at different locations were measured when The instantaneous respiratory quotient (RQ) was calcu-
recirculating air with reversed−direction airflow. lated by dividing the volume of CO2 produced by the volume
of O2 consumed, as described by Schulze (1960). Dry matter
EXPERIMENTAL PROCEDURE loss (DML) was calculated using the initial and final dry mass
An earlier study by Ekinci et al. (2001) on composting of of compost, as described by Keener et al. (1997).
paper mill sludge with broiler litter at C:N of 25:1 showed
that the optimum moisture for composting was 44% (w.b.).
Therefore, a mix of paper mill sludge with broiler litter at C:N RESULTS AND DISCUSSION
of 25:1 and moisture at 44% (w.b.) was used in this study. CHEMICAL CHANGES
Paper mill sludge (6% moisture, w.b.) was mixed with broiler
The final chemical properties of the compost are present-
litter (31% w.b.) and water was added to raise the mix up to
ed in table 2 for run 1 and in table 3 for run 2. Calculated
40% (w.b.) moisture content. To overcome the hydrophobic DML, based on division of the final dry mass of the compost
nature of the dried paper mill sludge, the mixture was held for
into the initial dry mass, is also given in tables 2 and 3.
at least one day to allow for soaking of water into the
Final pH was between 7.3 and 8.7 for all layers in both runs
material. It was then mixed by shovel on a concrete pad. and changed little from the initial pH of 7.7, which was
Because water had not penetrated adequately into the paper
favorable for microbial activity during composting. Final
mill sludge, a grinder was used to break up the lumps, and 4%
carbon content was 22.9% to 24.9% for the five layers in each
more water was added to the mixture. Thereafter, a test and showed little change from 25.9% for run 1 and 25.1%
mixer−wagon was used to make the mix homogeneous. The
for run 2. Results indicated that utilization of carbon in the
mixture was sampled and analyzed for initial moisture
compost mirrored DML and that percent change in carbon
content, pH, ash, and C:N ratio (table 1). Three samples were was small. Initial N content for run 1 and run 2 was 1.07% for
collected from different locations in each of the compost
all treatments. At the end of the experiments (after 7 d), the
piles. Chemical analyses of samples were performed by
nitrogen content had decreased in 35 of the 40 compost
OARDC’s analytical laboratory. layers, indicating a nitrogen loss from the system in excess
The compost mix was divided into four batches, one for
of that utilized by microbes. Dry matter ash content increased
each treatment. Each batch was then subdivided into five
2% to 7% in all samples.
parts and added to a reactor in layers, with each layer
separated by a polypropylene screen. Each layer’s initial
AIRFLOW
weight and height were measured with an electronic scale
The average airflow with the standard deviation and
(model 2400 series, Fairbanks) and metric ruler. After
loading all of the compost mass into the reactors, total initial recirculation ratios and reversal times are given in table 4 for
run 1 and run 2. The airflow used for ODA−R and RDA−R
weight and initial height were measured.
was almost five times higher than for ODA and RDA. High
Composting time was 7 d for each experiment. At the end,
individual layer weights and heights were measured and airflow is an important feature of air recirculation to maintain
and minimize the temperature gradients in the composting
recorded. In addition, three compost samples were collected
bed via recycling. However, based on the recirculation ratio
from each layer for final moisture content, which was
determined by drying in an oven at 105°C for 24h. Two of 0.90 to 0.91, make−up air was about 0.38 kg air kg
compost −1 d−1, which was 40% of ODA or RDA. This
compost samples per layer were taken for final pH, ash, C, N,
impacted oxygen levels during composting, as will be shown
and C:N determinations.
Table 3. Final chemical properties and DML of compost for run 2.[a]
C Content N Content C:N Ash Content Moisture Content DML
Treatment Layer[b] pH (%)[c] (%)[c] Ratio (%)[c] (%)[d] (%)
ODA Bottom 7.58 ±0.06 22.95 ±0.11 1.08 ±0.02 21.16 ±0.20 65.98 ±0.11 32.63 ±0.52 8.80
2 8.10 ±0.11 23.07 ±0.38 0.98 ±0.05 23.70 ±1.55 65.28 ±0.60 33.48 ±1.23 7.40
3 8.32 ±0.05 23.33 ±0.11 0.94 ±0.03 24.76 ±0.79 64.67 ±0.37 38.29 ±1.58 7.30
4 8.25 ±0.04 22.93 ±0.32 0.91±0.03 25.26 ±0.48 64.64 ±0.46 40.41 ±1.78 6.40
Top 8.18 ±0.11 22.91 ±0.40 0.94 ±0.00 24.43 ±0.48 65.37 ±0.51 43.25 ±1.17 14.40
ODA−R 1 8.14 ±0.01 22.88 ±0.05 1.00 ±0.01 22.88 ±0.25 66.01 ±0.06 35.38 ±0.97 9.01
2 8.37 ±0.04 23.52 ±0.05 0.90 ±0.01 26.19 ±0.24 63.30 ±0.48 42.52 ±0.38 11.10
3 8.42 ±0.06 23.71 ±0.16 0.90 ±0.01 26.39 ±0.01 63.25 ±0.30 42.46 ±0.64 6.70
4 8.43 ±0.05 23.61 ±0.01 0.91 ±0.02 25.90 ±0.50 63.55 ±0.09 39.34 ±0.06 4.30
5 8.31 ±0.18 23.23 ±0.19 0.89 ±0.02 26.22 ±0.84 64.94 ±0.24 41.90 ±0.87 9.70
RDA 1 7.75 ±0.40 23.13 ±0.00 1.01 ±0.07 22.85 ±1.53 64.91 ±0.67 37.27 ±1.30 5.00
2 8.07 ±0.13 23.51 ±0.15 0.95 ±0.01 24.84 ±0.09 64.22 ±0.43 38.49 ±0.33 4.60
3 8.23 ±0.04 23.34 ±0.11 0.92 ±0.00 25.42 ±0.09 63.84 ±0.17 39.77 ±1.04 3.80
4 8.15 ±0.04 23.50 ±0.22 0.94 ±0.02 25.13 ±0.79 63.86 ±0.44 39.81 ±0.40 6.30
5 8.18 ±0.21 23.17 ±0.16 0.97 ±0.03 23.96 ±0.99 64.58 ±0.31 38.83 ±1.42 6.60
RDA−R 1 8.45 ±0.04 23.40 ±0.11 0.87 ±0.01 26.92 ±0.30 63.63 ±0.48 41.55 ±1.24 7.20
2 8.61 ±0.07 23.78 ±0.34 0.87 ±0.02 27.26 ±0.21 62.77 ±0.42 44.50 ±0.58 8.00
3 8.72 ±0.04 23.74 ±0.15 0.88 ±0.03 27.07 ±0.71 63.47 ±0.24 41.33 ±0.73 3.70
4 8.62 ±0.01 23.70 ±0.38 0.88 ±0.03 27.10 ±1.45 63.71 ±0.82 44.47 ±1.31 12.30
5 8.69 ±0.23 23.27 ±0.06 0.87 ±0.01 26.60 ±0.08 64.53 ±0.52 38.58 ±2.19 7.40
[a] Table reports avg. ±SD (n = 2 for all, except n = 3 for moisture content).
[b] “Bottom” represents layer 1, while “top” represents layer 5.
[c] Dry basis.
[d] Wet basis.
later. The direction of airflow was changed every 12 h for experiment and, during this period, the compost mix started
both RDA and RDA−R. When airflow direction was to heat up. This procedure in run 2 did not lead to compost
changed, the amount of airflow supplied was not changed. heating, and ambient temperature was higher than the initial
compost temperature.
COMPOST TEMPERATURE PROFILES During run 2, the research building power went off for 2 h
The initial compost temperature was higher than ambient on the fourth day of composting and caused a failure of the
temperature in run 1. The reason was that water was allowed central heating system’s equipment, which could not be
to penetrate into the particles for 24 h prior to the start of the repaired immediately. This power outage affected building
temperature, which dropped from 21°C to approximately from the high−temperature region to the low−temperature
16°C during the remainder of the experiment. The compost (near the wall) region. This exhaust air circulation cooled the
temperature profiles were affected by this drop in ambient top compost layer during the experiment.
temperature. Temperature histories as a function of layer (T1, Condensation occurring at the surfaces in the reactor’s
T2, T3, T4, and T5) are given in figures 5 through 8 for all headspace was another factor. Condensed water returned to
treatments. the fifth layer at a lower temperature than air entering from
layer 4. Additionally, the conduction loss in the fifth layer
One−Directional Airflow was higher than in the fourth layer. This combined conduc-
A temperature gradient, which was a result of the
tion loss and condensation cooled the top layer during the
evaporative cooling process, existed during composting
experiment. The maximum temperature difference between
(fig. 5). The lowest temperature occurred at the bottom layer,
the first layer and fourth layer was 21.6°C for run 1 at 1.47 d
since it was exposed to the ambient conditions (low humidity
of composting and 24.11°C for run 2 at 4.45 d of composting.
and temperature), whereas the highest temperature existed in
Compost temperature rose to 70°C within 1.5 d in the third
the fourth layer. Theoretically, the highest temperature
and fourth layers for both runs.
should have occurred in the fifth layer. However, the fifth
layer temperature was much lower than the theoretical value One−Directional Airflow with Recirculation
because of convection air current and condensation within High rates of recirculated airflow created more uniform
the headspace of the reactor. temperature profiles in the composting bed when compared
An analysis of convection current occurring in the fifth with the case of non−recirculation and low airflow (ODA)
layer and headspace was performed to explain the physical (fig. 6). The average pulsed airflow used was 304 kg air d−1
phenomena. Result of the analysis showed that compost (4.12 kg air kg compost−1 d−1) for run 1 and 307.74 kg air d−1
temperature close to the wall was lower than the compost (4.05 kg air kg compost−1 d−1) for run 2. The compost
center temperature. Therefore, exhaust air circulated itself temperature rose to the 65°C to 70°C range within 2.0 d for
ÓÓ
Ó Ó Ó
ÓÓ
ÓÓ ÓÓ ÓÓÓ
80 80
Compost Temperature (oC)
(a) (b)
Compost Temperature (oC)
70 70
ÓÓ
ÓÓ
60 60
ÓÓ
50 50
ÓÓ
Ó
40 40
T1 T2 T3 T4 T5 T1 T2 T3 T4 T5
30 30
20 20
Room Temperature Room Temperature
10 10
0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Time (days) Time (days)
Figure 5. Compost temperature as a function of time for ODA: (a) run 1 and (b) run 2.
80 80
(a) (b)
Compost Temperature ( oC)
Compost temperature (o C)
70 70
60 60
50 50
T1 T5 T4 T2 T3 T1 T2 T3 T4 T5
40 40
30 30
20 20
Room Temperature Room Temperature
10 10
0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Time (days) Time (days)
Figure 6. Compost temperature as a function of time for ODA−R: (a) run 1 and (b) run 2.
Figure 7. Compost temperature as a function of time for RDA: (a) run 1 and (b) run 2.
both runs. The maximum temperature difference between study of Harper et al. (1992), heat loss from the wall
layers was 8.15°C at 2.95 d for run 1 and 12°C at 4.56 d for increased. This phenomenon increased the temperature
run 2. gradients between layers.
The lowest temperature in the compost occurred in the
first layer. This was due to the coupling effects of evaporative Reversed−Direction Airflow
Figure 7 shows temperature profiles for reversed−direc-
cooling and heat loss in this layer (Hogan et al., 1989).
tion airflow for run 1 and run 2 based on a 12 h schedule. In
Theoretically, the highest temperature should have occurred
in the fifth layer. However, the fifth layer temperature was the temperature graphs, primary y−axes for layers 2, 3, and
4 and secondary y−axes for layers 1 and 5 are used to present
much lower than the fourth layer because of conduction heat
the data more clearly. When airflow was upward, the first
loses and condensation within the headspace of the reactor.
layer was exposed to ambient temperature, and its tempera-
Although ODA−R had the lowest temperature gradients
ture decreased while the top layer temperature increased.
among the systems tested, there was a discrepancy between
When the airflow was downward, the last layer was exposed
the findings in this work and as reported in the literature. The
to ambient temperature, and its temperature decreased while
temperature differences of 2°C reported in the literature by
the first layer temperature increased. This pattern occurred
Harper et al. (1992) were probably measured from the first
during the experiments with a fixed 12 h reversal schedule.
and last layer. The temperature differences between the first
The temperature fluctuations in the second and the fourth
and the last layer in these studies were 6.6°C at 6.9 d for run 1
layers were still high in magnitude because of the evaporative
and 4.87°C at 0.25 d for run 2. This indicates that the location
cooling process when the airflow direction changed. The
where temperature difference is reported is important.
third layer temperature was least affected by changing the
Harper et al. (1992) reported that the fresh air intake rate for
direction of airflow. This means that temperature profiles
the air recirculation system was 2.28 (average) kg air kg
converged from both sides toward the middle of the reactor.
compost −1 d−1, whereas the fresh air intake rate used in this
The temperature increased to the 70°C to 73°C range within
study was 0.4 kg air kg compost−1 d−1, considering a
1.5 d for run 1 and run 2. The maximum temperature
recirculation ratio of 0.90. High fresh airflow with a high
difference was 50°C at 1.96 d for run 1 and 53°C at 1.96 d
recirculation ratio minimizes temperature gradients. Addi-
for run 2.
tionally, the common characteristic of a small−scale reactor
is that the surface area to volume ratio is higher than that of Reversed−Direction Airflow with Recirculation
large−scale system, such that the potential for heat loss from Figure 8 shows the temperature histories for recycling of
the wall of the reactor increases (VanderGheynst, 1994). exhaust gases while changing the direction of airflow. The
Since the surface area to volume ratio of 5.92 calculated for airflow, reversal time, and recirculation ratio are given in
the composting vessel in this study was higher than that table 4. This system is a combination of ODA−R and RDA,
(2.21) calculated for the tunnel composting system in the and its response is the summation of ODA−R and RDA
80 160
80 140
Compost temperature (oC)
70 140
Compost temperature (o C)
120
Compost temperature (oC)
70 120
60 100 60
T4 100
80 T2 T3
T2 T3 T4 50 80
50
60 40 60
40
T1 T5 40 40
30 30
20 T1 T5 20
20 20
0 0
10 −20 10 −20
(a) (b)
0 −40 0 −40
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Time (days) Time (days)
Figure 8. Compost temperature as a function of time for RDA−R: (a) run 1 and (b) run 2.
25 40
25 40
(a)
O2 35 (b)
35
CO2 concentration (v/v)
20 O2
CO2 concentration (v/v)
O2 concentration (v/v)
20
O2 concentration (v/v)
30 30
15 25 25
15
20 20
10 15 10 15
CO2 10 10
5 5 CO2
5 5
0 0 0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Time (days) Time (days)
Figure 9. Oxygen (O2) and carbon dioxide (CO2) concentrations by percent as a function of time for the (a) first layer and (b) third layer for ODA
(run 1).
concentration (v/v)
CO2 concentration (v/v)
O2 concentration (v/v)
50 50
O2 concentration (v/v)
20 O2 20
45 45
O2
40 40
15 35 15 35
30 30
10 25 10 25
20 20
15 15
5 5 10
CO 2
10
5 5
0 0 0 0
0 1 2 3 4 5 CO
62 7 0 1 2 3 4 5 6 2
CO 7
Time (days) Time (days)
Figure 10. Oxygen (O2) and carbon dioxide (CO2) concentration by percent as a function of time for the (a) first layer and (b) third layer for ODA−R
(run 1).
tory quotient (Keener et al., 1993) gave an RQ value for the RDA−R
first layer of around 1 for run 1, but <1 for the third layer. This Oxygen and carbon dioxide profiles for RDA−R for run 1
indicated that oxygen limitation was more pronounced in the are given in figure 12. Like the ODA−R test, RDA−R was
third layer than in the bottom layer. The RQ values, which oxygen limited due to insufficient make−up air. During the
were greater than 1, indicated that when airflow was on, the experiment, approximately 90% recirculation ratio was
compost reactor was flushed, i.e., oxygen concentration maintained, and since the airflow was kept constant during
increased and carbon dioxide concentration decreased. This the experiment, the system never reached the well−oxygen-
occurred following the off time periods, especially, for ated phase. Reversing the direction of airflow did not change
layer 3 over days 1 to 3 when O2 was approximately 0%. As the O2 and CO2 pattern during the process. The results
a result of this process, it was impossible to measure true showed that the oxygen concentration was not generally
instantaneous values of oxygen and carbon dioxide con- greater than 16%, and instantaneous RQ values were less than
sumed and produced from the system. ODA.
ODA−R
DECOMPOSITION AND ASH CONTENT PROFILES
Results for the ODA−R test showed that the composting
Temperature, moisture content, and oxygen concentration
process usually suffered from lack of oxygen in the innermost
layer (3rd) during the first two days (fig. 10). Maximum affected the resulting ash content and DML profiles within
the composting vessel (tables 2 and 3). Ekinci et al. (2004)
oxygen concentration for this layer was generally less than or
expressed decomposition of paper mill sludge with broiler
equal to 15%. The calculated instantaneous RQ values were
less than ODA. The problem with this experiment was that litter as functions of temperature and initial compost
moisture using a Gaussian model. Result of that study
insufficient make−up air was maintained by using a 90%
showed that the maximum decomposition rate occurred at
recirculation ratio, and since the make−up air was kept
constant during the experiment, the system never reached the 58°C and 44% (w.b.) compost moisture.
well−oxygenated phase. ODA
Since gradients existed in the system in terms of
RDA
Oxygen and carbon dioxide profiles for RDA for run 1 temperature, moisture, and oxygen concentration, decom-
position gradients occurred in the ODA system (tables 2 and
given in figure 11. Evaluation of the O2 and CO2 profiles for
3). Results showed the bottom layer (fig. 5) was close to the
RDA showed that O2 and CO2 profiles followed the airflow
direction and instantaneous RQ values were greater than for optimum temperature but had low moisture levels due to
rapid drying. This low moisture inhibited decomposition.
ODA. Oxygen limitation occurred in the third layer.
The highest decomposition for run 1 occurred at the second
25 50 25 50
(a) 45 (b) 45
concentration (v/v)
O2
concentration (v/v)
20 40 20 40
O2 concentration (v/v)
O2 concentration (v/v)
35 O2 35
15 30 15 30
25 25
10 20 10 20
CO 2 15 15
5 10 5 10
CO 2
CO 2
5 5
CO 2
0 0 0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Time (days) Time (days)
Figure 11. Oxygen (O2) and carbon dioxide (CO2) concentrations by percent as a function of time for the (a) first layer and (b) third layer for RDA
(run 1).
O2 concentration (v/v)
O2 concentration (v/v)
CO2 concentration (v/v)
20 20
30 30
O2
15 25 15 O2 25
20 CO2 20
10 15 10 15
10 10
5 5
5 5
CO2
0 0 0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Time (days) Time (days)
Figure 12. Oxygen (O2) and carbon dioxide (CO2) concentrations by percent as a function of time for the (a) first layer and (b) third layer for RDA−R
(run 1).
layer since neither ambient air conditions nor compost systems can eliminate these gradients within certain limits.
moisture and compost temperature limited the composting- However, oxygen limitation was the biggest problem in the
process. The second layer in run 2 for ODA did not respond system because of the insufficient make−up air, which led to
in the same way, and it had a relatively low decomposition this system yielding the lowest average decomposition rate.
rate. High temperatures existed in the third and the fourth The general observation from this study was that the
layers for both runs, which exhibited limited decomposition. boundary conditions positively affected the decomposition
The fifth layer was under the influence of convection currents process in the first and last layers (table 2 and 3). The lowest
and condensation effects. The fifth layer compost tempera- decomposition occurred in the middle layers because of
ture was close to the optimum temperature. While this layer severe oxygen limitation.
lost water and kept drying, the condensation process led to the
accumulation of water at the top such that 2 to 3 cm of the top
layer was much wetter than the lower zone within the top
layer. The result was that this layer yielded the highest DML
CONCLUSIONS
Four aeration strategies (ODA, ODA−R, RDA, and
for run 2. In conclusion, decomposition gradients were
RDA−R) were investigated to determine their effects on
established among the layers because of their exposure to
composting process variables. The compost mix studied was
different process conditions. The standard deviations of the
paper mill sludge and broiler litter with a C:N ratio of 25:1.
DML for the five layers for ODA were 2.1% for run 1 and
Airflow was intermittent with a 15 min on and a 45 min off
3.2% for run 2.
cycle for all tests. Detailed analysis of experimental data
ODA−R showed that aeration strategies yielded different tempera-
The objective of ODA−R was to eliminate the temperature ture, moisture, and decomposition profiles. Results showed
gradient but not necessarily the moisture gradient. The end that condensation, convection currents, and conduction heat
(7 d) moisture profiles were uniform except for the first layer. losses occurred within the headspace and affected the
The problem with this system was oxygen limitation (see temperature moisture profiles and decomposition in the top
earlier discussion) for both run 1 and run 2. The measured compost layer.
oxygen level in the air recirculation pipe system did not Conclusions for each system are:
exceed 17%. Therefore, decomposition rates in ODA−R S One−directional airflow (ODA) yielded the highest
were lower than in ODA, even though temperature and temperature, moisture, and decomposition gradients
moisture did not limit the process. The lower rates were due across the composting bed. When composting materi-
to low O2 levels. The standard deviations of DML were 1.3% als in this system, the material should be definitely re-
for run 1 and 2.7% for run 2 (tables 2 and 3). mixed to produce high−quality compost. To optimize
the composting rate, fan on−off times and airflow
RDA
should be optimized. This requires computer modeling
The first and fifth layers were exposed to ambient air
of the composting process.
periodically in the RDA system and were affected by the S One−directional airflow with air recirculation (ODA−
boundary conditions. The lowest DML occurred at the third
R) was not operated at the optimum point for oxygen
layer because of high temperature elevations. As can be seen
concentration and suffered from the lack of oxygen
from figure 7, the third layer temperature increased above concentration during the experiment. This situation
72°C. The standard deviations of DML were 3.5% for run 1
was reflected in the temperature and decomposition
and 1.2% for run 2. In conclusion, there was no advantage of
profiles. If the system had been operated at the opti-
RDA over ODA for the test conditions used. However, this mum point, the result could be quite different. The rea-
statement could be quite different if the system was operated
son is that the system yielded temperature profiles that
under optimum conditions.
were similar to those reported in the literature but
RDA−R yielded different oxygen concentration profiles. This
Theoretically, RDA−R should eliminate the gradients means that there is possible justification for the usage
such as temperature, moisture, oxygen, and decomposition. of this system over ODA if oxygen concentrations had
As far as temperature and moisture are concerned, the not been limiting.