Applied Thermal Engineering 114 (2017) 1378–1386

Contents lists available at ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

Research Paper

Advanced power generation using biomass wastes from palm oil mills q
Muhammad Aziz a,⇑, Tedi Kurniawan b, Takuya Oda a, Takao Kashiwagi a
a
Advanced Energy Systems for Sustainability, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-ku, Tokyo 152-8550, Japan
b
Faculty of Mechanical Engineering, Universiti Malaysia Pahang, 26600 Pekan, Pahang Darul Makmur, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: This study focuses on the energy-efficient utilization of both solid and liquid wastes from palm oil mills,
Received 23 February 2016 particularly their use for power generation. It includes the integration of a power generation system using
Revised 24 October 2016 empty fruit bunch (EFB) and palm oil mill effluent (POME). The proposed system mainly consists of three
Accepted 5 November 2016
modules: EFB gasification, POME digestion, and additional organic Rankine cycle (ORC). EFBs are dried
Available online 7 November 2016
and converted into a syngas fuel with high calorific value through integrated drying and gasification pro-
cesses. In addition, POME is converted into a biogas fuel for power generation. Biogas engine-based
Keywords:
cogenerators are used for generating both electricity and heat. The remaining unused heat is recovered
Power generation
Empty fruit bunch
by ORC module to generate electricity. The influences of three EFB gasification temperatures (800, 900
Palm oil mill effluent and 1000 °C) in EFB gasification module; and working fluids and pressure in ORC module are evaluated.
Gasification Higher EFB gasification leads to higher generated electricity and remaining heat for ORC module. Power
Digestion generation efficiency increases from 11.2 to 24.6% in case of gasification temperature is increased from
Organic Rankine cycle 800 to 1000 °C. In addition, cyclohexane shows highest energy efficiency compared to toluene and n-
heptane in ORC module. Higher pressure in ORC module also leads to higher energy efficiency. Finally,
the highest total generated power and power generation efficiency obtained by the system are 8.3 MW
and 30.4%, respectively.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction environmental hazard. As a result, POME treatment has become

The production of palm oil and its products has increased
rapidly in the last decade due to high demand following the eco-
nomic growth, especially in developing countries including China,
India, and Southeast Asian countries [1]. This has led to a massive
expansion of palm plantations leading to some associated environ-
mental problems, especially wastes [2]. In palm oil production,
only about 10% of the whole palm tree is currently utilized in
any significant way [3]. The wastes include both liquid and solid.
Generally, palm oil mills require a huge amount of water, and in
turn, dispose of large quantities of wastewater called palm oil mill
effluent (POME). POME is the largest byproduct of the process, with
no specific commercial application. It contains about 95% water
and 5% solids. POME is acidic and has both high biological oxygen
demand (BOD) and chemical oxygen demand (COD). Therefore, the
discharge of POME without sufficient treatment represents an
q
A part of this paper was presented at the 4th International Conference on
Microgeneration and Related Technologies (Microgen IV), October 28–30, 2015,
Tokyo, Japan (Original paper title: "Advanced Utilization of Biomass Wastes from Palm
Oil Mill for Power Generation” and Paper ID.: 0085).
⇑ Corresponding author.
E-mail address: aziz.m.aa@m.titech.ac.jp (M. Aziz).
the primary source of environmental pollution, which includes
high emissions of methane (32.9 kg-CH4 t-POME
1) and its solid
nutrients to water and soil [4].
Solid wastes are also generated during the milling process,
which consist primarily of empty fruit bunch (EFB, 23%), fiber
(12%), and palm kernel shell (5%) [5]. Among them, EFB generally
has the lowest economic value for reuse. It has a relatively high
moisture content of 60–70 wt% on a wet basis (wb) [6], low bulk
density, and non-uniform shape [7]. It requires additional treat-
ment, such as cutting and grinding, before being used. EFB is a
byproduct of the fresh fruit threshing process to detach the palm
fruits.
The treatment of wastes at palm oil mills is very important from
both environmental and economic standpoints. Some developing
countries demand strict environmental controls. Therefore, palm
oil production must be environmentally sustainable if exports are
to be feasible. Accordingly, new environmental friendly technolo-
gies are vital and urgently required. Alternatives for use of solid
and liquid wastes include energy harvesting, industrial material,
and fertilizers. However, high transportation and labor costs and
greenhouse gas (GHG) emissions have become barriers to their uti-
lization [8].

http://dx.doi.org/10.1016/j.applthermaleng.2016.11.031
1359-4311/Ó 2016 Elsevier Ltd. All rights reserved.
 M. Aziz et al. / Applied Thermal Engineering 114 (2017) 1378–1386
Currently, both POME and EFB are treated in a very inefficient
way. POME is treated conventionally through anaerobic and aero-
bic processes to be converted to biogas or to compost before its
residual liquid is discharged into rivers. Although a small part of
produced biogas is used for power generation, in majority it is
flared with the aim of reducing the GHG impacts. In addition, at
present, EFB is commonly disposed in plantations for nutrient recy-
cling without any significant utilization. Unfortunately, this prac-
tice may result in eutrophication and an increase in toxicity in
the soil [9].
Because EFB and POME represent the largest portion of wastes
from palm oil mills, effective utilization of these byproducts is
essential for increasing economic benefits as well as reducing envi-
ronmental impacts. To the best of our knowledge, there is lack of
research on efforts to integrate and effectively utilize both EFB
and POME for small-scale in situ power generation with high
energy efficiency. Aziz et al. proposed an integrated system con-
sisting of drying, gasification, and a combined cycle with relatively
high energy efficiency [10]. However, their system was not suitable
for small-scale in situ power generation because EFB must be col-
lected from multiple mills to address its high demand as fossil fuel.
Luk et al. evaluated an integrated drying and boiler-based power
generation using EFB [11]. However, their proposed system
showed large amounts of exergy destruction, leading to lower
energy efficiency. In addition, the co-digestion of EFB and POME
has been evaluated previously to simplify the treatment steps
and reduce production costs [12]. However, because EFB has very
low biodegradability due to its rich lignocellulosic composition,
such utilization was not promising.
This paper proposes an integrated small-scale in-situ energy
harvesting process (power generation) utilizing EFB and POME
with high energy efficiency, especially utilization of EFB. The objec-
tive of this study is to model an integrated system for power gen-
eration utilizing both EFB and POME through process modeling and
to evaluate its feasibility in terms of energy efficiency. Therefore,
by applying the proposed system, it is expected that palm mill
can increase its energy productivity in efficient way as well as
higher economic performance and lower environmental impact.
Because palm mills are generally located in rural and remote areas
with poor electricity supply, power generation from palm oil mill
wastes is expected to have the potential to improve the electricity
supply to surrounding areas increasing the access to electricity.
Flue gas
EFB Dried EFB
Drying Gasification
Hot gas
POME
Digestion
ORC
Flue gas
Fig. 1. Conceptual diagram of the integrated small-scale power generation for EFB and POME.
1379
2. Proposed power generation system
The concept of the proposed integrated power generation sys-
tem using both EFB and POME is shown in Fig. 1. This kind of in-
situ power generation can be categorized as small scale due to
the limited flow rate of fresh fruit bunch (FFB) which is generally
ranging from 30 to 100 t-FFB h
1. Furthermore, it is predicted that
these capacities are able to produce the electricity up to 20 MW.
The system consists of two main modules: (1) EFB gasification
and POME digestion modules, and (2) an additional organic Rank-
ine cycle (ORC) module. The solid, dashed, and dotted lines repre-
sent material, heat, and electricity flows, respectively.
The EFB gasification module involves three continuous pro-
cesses: (1) drying, (2) gasification, and (3) power generation using
a gas engine. Initially, wet EFBs exhausted from the mill are col-
lected and dried to reduce their moisture content, therefore
increases the calorific value. The heat source required for drying
is the hot flue gas exhausted from gas engines in both EFB gasifica-
tion and POME digestion modules. The hot and dried EFBs then
flow to the gasification module, where EFBs are converted to syn-
gas consisting of H2, CO, CO2, and CH4. Gasification is used because
of its higher carbon conversion efficiency than other conversion
technologies such as pyrolysis [8]. The produced syngas is then
cleaned and used as fuel for an internal combustion engine gener-
ating both electricity and heat (cogeneration). The exhausted heat
from the gas engine typically ranges from 300 to 600 °C, which cre-
ates the potential for use as an energy source. In this study, the
generated heat is used primarily for EFB drying.
The POME digestion module involves two primary processes:
(1) anaerobic digestion and (2) power generation using the gas
engine. During anaerobic digestion, POME is converted to biogas,
which generally consists of about 35–60% methane. POME can be
easily converted to methane with about 98% biodegradability
[12]. Through anaerobic digestion, lipids included in POME are
degraded to glycerol and further converted to methane. The
sequence of anaerobic digestion includes hydrolysis, acidogenesis,
and methanogenesis [13]. The produced biogas is further cleaned
and used as fuel in power generation using the gas engine. Because
the composition and calorific value of the produced syngas and
biogas from each module are different (H2-rich and CH4-rich), in
this study, different gas engines are installed at each corresponding
module. However, use of a single gas engine having the capability
to use produced gases with different compositions is preferred in
consideration of lower costs, integrated operation, and higher total
efficiency (lower heat loss).
Syngas
Gas Engine
Electricity
Biogas
Gas Engine
1380 M. Aziz et al. / Applied Thermal Engineering 114 (2017) 1378–1386
In case there is still any remaining heat from the flue gas, espe-
cially from the EFB gasification module, an additional ORC system
can be employed to increase the total generation efficiency. The
ORC module consists of a pump, preheater, superheater (evapora-
tor), and condenser.
2.1. EFB gasification module
Fig. 2 shows the schematic process flow diagram of the EFB
gasification module. Initially, raw and wet EFBs from palm mills
are cut and shredded to enlarge their surface areas and achieve
uniform size, thereby enhancing moisture and heat transfer and
uniformity during drying. The shredded EFBs then enter a rotary
dryer for moisture removal after being preheated using the flue
gas exhausted from dryer. A rotary dryer is used because of its
large heat transfer area, potential for continuous operation, easier
handling, and excellent drying control [14,15]. The dryer is rotated
by a motor, resulting in excellent particle mixing. Peripheral fins
are installed inside to lift and shower the solid particles, promoting
close contact between the solids and hot gas. In addition, the dryer
has a sloped rotating cylinder to facilitate particle movement from
the feeding inlet to the discharge outlet.
All of the hot flue gas from POME digestion module is used as a
heat source for drying. In addition, the hot flue gas from EFB gasi-
fication is primarily used to cover the heat for drying, while in case
that there is any remaining heat, it is used as the heat source in
ORC module. The heat exchange inside the dryer can be set to
co-current or counter current depending on the flow direction of
both EFBs and hot flue gas. Because the hot flue gas has direct con-
tact with EFBs, a convective heat transfer occurs, leading to rapid
heat and moisture transfer. The evaporated moisture and flue gas
are discharged together, flowing into cyclones for further
separation.
After being ground to smaller size, the hot-dried EFBs then flow
into the gasifier for conversion after being preheated initially using
the raw syngas from gasification. For gasification, a fluidized bed
gasifier is used because of its high level of heat transfer, conversion
rate, and running variable load [16]. In addition, air is used as a
gasification medium. Auto-thermal gasification is adopted in this
study in which the heat is supplied mainly by the preheated air
S10
Wet EFB
Cutter,
S1 shredder
Flue gas S5
S6
Preheater
S2
S4 Separator
Rotary dryer
S3
Belt
Dried EFB
Motor
Grinder
S16
S15
S22
Hot gas from POME digestion
Fig. 2. Schematic process flow diagram of the EFB gasification module.
and EFB, and combustion of carbon including char. The produced
syngas is exhausted from the gasifier and its heat is recovered to
preheat the air for gasification and dried EFB. Next, the raw syngas
is cleaned and cooled to remove particles, sulfur, and tar conden-
sate before being used as fuel for the gas engine.
2.2. POME digestion module
Fig. 3 shows the schematic process flow diagram of the POME
digestion module. POME is usually discharged from the mill at
temperatures of 70–90 °C. The exhausted POME flows to cooling
and acidification ponds, where its temperature is reduced to
approximately 35 °C and its pH is adjusted. The retention time in
these ponds is approximately 1–2 days. Following pond retention,
POME flows to an anaerobic digestion pond, where biogas is pro-
duced in a period of approximately 60 days. The produced biogas
is exhausted from the digester and split into two streams. The pri-
mary stream moves to the gas cleaning system and another stream
moves to the flaring system. Flaring is important because the
amount of produced biogas fluctuates. In the absence of flaring,
facilities for gas storage are required. As a result, the rate of biogas
flowing to the gas engine is controlled and the unused biogas is
flared, converting CH4 to CO2.
The biogas flowing to gas cleaning is further scrubbed for H2S
and moisture removal (drying). The clean and dried biogas finally
flows to the gas engine as fuel for generation of both electricity
and heat. The exhausted flue gas is collected together with a part
of exhausted flue gas from EFB gasification module and utilized
for drying. Furthermore, the remaining liquid wastes from diges-
tion flow consecutively to aerobic and maturation ponds. These
ponds are very important to further reduce the organic content
of the liquid wastes. The required retention time is approximately
14 days, during which BOD might be reduced to less than 100 ppm.
At that point, it can be used as a liquid fertilizer in plantations or
discharged into rivers.
2.3. Use of the ORC module
ORC is considered to be an appropriate technology to recover
low-quality heat and convert it to electricity [17]. Use of ORC can
Raw syngas Gas cleanup
Clean syngas
S12
S7 Gas cooler
Compressor
Particulate
and sulfur
S19
S9
Air (for gasification)
Cyclones
S21 S13
S8
Gasifier Electricity
Gas engine
S11
S18
S14
Compressor
Hot gas
Air
S17
S20
Remaining hot gas (to ORC module)
 M. Aziz et al. / Applied Thermal Engineering 114 (2017) 1378–1386 1381

S27
Gas cleanup Clean syngas
Electricity
S26 Gas engine
POME Particulate S32
and sulfur

Biogas

to EFB drying
S23
S25

Hot gas
S31

system
Flaring
Acidification S24 Anaerobic Control valve
ponds ponds
S22
S28

Aaerobic S29 Maturation Liquid fertilizer S30

ponds ponds (to plantation)

Fig. 3. Schematic process flow diagram of the POME digestion module.

further minimize the loss of exergy, particularly when the remain-
ing waste heat has no specific use. ORC has a working principle
similar to that of an ordinary steam Rankine cycle. However, it uses
organic fluids, including hydrocarbons and refrigerants, which
have a lower specific heat of vaporization than water. Therefore,
it is feasible to convert low-temperature heat into electricity. In
this study, ORC is used in case there is any remaining heat from
the flue gas, particularly from the EFB gasification module.
Fig. 4 shows a schematic diagram of the ORC module that can be
integrated with the above two modules. Although the dual-loop
ORC system is considered to have a higher generation efficiency,
the basic single-loop ORC system is used because of its simpler
structure and lower initial cost [18]. The remaining hot flue gas
from the EFB gasification module is used as the primary heat
source in the ORC module, superheating the working fluid. As a
result, the compressed working fluid will evaporate, producing
high-pressure vapour from the working fluid. Next, it flows to
the turbine for expanding and generating torque to rotate the gen-
erator, producing electricity. Furthermore, because the working
fluid exhausted from the turbine still has a relatively high temper-
ature, a preheater is installed to recover heat. This heat, in turn, is
used to preheat the working fluid.
In ORC, selection of the working fluid is very important because
its thermodynamic properties will influence efficiency [19]. In this
study, some organic working fluids such as toluene, octane, cyclo-
hexane, and heptane are considered suitable [18–20]. In addition,
the stability of the working fluid must be maintained to avoid
any degradation under specific temperature and pressure [19].
S34
Turbine
3. Process modeling and simulation
Table 1 shows the composition of EFB, including the proximate
and elemental analyses used in this study based on previous stud-
ies [21,22]. Generally, EFB has a high content of volatile chemicals
(about 80% wt% db), resulting in a high potential for conversion to
syngas [23]. Furthermore, Table 2 shows the drying and gasifica-
tion conditions for EFB and the specifications for digestion of
POME. The specifications for use of the gas engine are based on
the standard model for a biogas engine manufactured by Yanmar
Co., Ltd., Japan [24].
The average flow rate of FFB fed to palm oil mills ranges from 45
to 60 t-FFB h
1. Considering that the ratio of produced EFB from
FFB is 20%, the flow rate of FFB and wet EFB used in this study
are fixed at 50 and 10 t h
1, respectively. In addition, the inlet tem-
perature of the hot gas entering the rotary dryer is assumed at
450 °C, which is based on detailed specifications of engines and
other literatures [25–27].
Process modeling is performed using a commercial process sim-
ulator SimSci Pro/II (Schneider Electric Software, LLC.). Some addi-
tional assumptions are made: (1) the maximum target moisture
content in drying is 5 wt% wb, (2) the gasifier consists of a mixer
and a conversion reactor, (3) the minimum temperature in the heat
exchanger is 10 °C, (4) POME is obtained from crude palm oil (CPO)
production at a rate of 532 kg t-FFB
1 [1], (5) the air contains
79 mol% nitrogen and 21 mol% oxygen, (6) the ambient tempera-
ture and pressure are 25 °C and 101.33 kPa, respectively, and (7)
heat is exchanged in a counter-current mode, in both the dryer
and heat exchanger. Moreover, considering that the work required
for shredding of EFB with the capacity of 10 t-EFB h
1 is about
75 kW [28] and the work consumed by a ball mill to grind the dried
EFB with capacity of 4.2 t-EFB h
1 is 30 kW [29], the total con-

Table 1
Flue gas S33
S35 Composition of EFB.
S20
Superheater Component Value
Pump
S39 S38 Proximate analysis
Volatiles (wt% db) 87.08
Preheater Fixed carbon (wt% db) 9.27
Ash (wt% db) 3.65
S36
Elemental analysis
S37 C (wt% db) 44.92
Water S41
H (wt% db) 6.21
S40 N (wt% db) 1.17
Condenser S (wt% db) 0.03
O (wt% db) 43.99
Moisture content (wt% wb) 60
Fig. 4. Conceptual flow diagram of the ORC module.
1382 M. Aziz et al. / Applied Thermal Engineering 114 (2017) 1378–1386

Table 2 LHV ¼ HHV
0:212 H
0:0245 MC
0:008 O; ð7Þ

Power generation conditions for drying, gasification, biogas digestion, and other
components. The kinetics of EFB drying are complex. The heat of hot flue gas
from gas engine is used as the energy to overcome some forces
Component Value
including capillary, sorption, and chemical bonding forces. The
EFB drying
amount of evaporated water from the wet EFB that is exhausted
Particle diameter (mm) 5
Min. temperature approach (°C) 10
together with flue gas can be expressed as follows:
Number of flights (–) 12  
MC i MC f
Rotation speed (rpm) 10 mw;ev p ¼ ms
; ð8Þ
1
MC i 1
MC f
EFB gasification
Temperature (°C) 800, 900, 1000 where mw,evp, ms, MCi, and MCf are flow rate of the evaporated water
Mean particle diameter (mm) 0.3
(kg h
1), flow rate of dry solid EFB (kg h
1), initial moisture content
Fluidizing particles (–) Silica sand
Fluidizing particles diameter (mm) 0.3
(wt%), and final moisture content (wt%), respectively.
Considering that there is no heat loss from the system, the
POME digestion
POME density (103 kg m
3) 0.98
energy balance inside the rotary dryer during drying can be
Produced biogas (Nm3 t-POME
1) 25 approximated as follows:
CH4 percentage (%) 55
CH4 calorific value (MJ Nm
3)
ms MC i
50 mg hg;in þ ms hs;in þ hw;in ¼ mg hg;out þ ms hs;out
CH4 density (kg Nm
3) 0.66 1
MC i
Ratio of used biogas to rated capacity (%) 80 ms MC f
þ hw;out þ mw;ev p hv ; ð9Þ
Power generation (gas engine) 1
MC f
Power generation efficiency (%) 32
Total efficiency (%) 84 where mg, hg, hs, hw, and hv are gas flow rate (kg h
1), gas specific
Flue gas temperature (°C) 450 enthalpy (J kg
1), specific enthalpy of solid EFB (J kg
1), water-
specific enthalpy (J kg
1), and vapour-specific enthalpy (J kg
1),
respectively. In addition, subscripts in and out represent inlet and
sumed electricity inside the plant, excluding the electric motor for outlet, respectively.
the rotary dryer and air compressor for fluidization, is 10% of which The heat supplied by the hot gas can be expressed as follows:
is totally generated. In addition, in case the heat from the flue gases
ms MC f
of gas engine is insufficient to provide the heat required for drying, Q¼ C w ðT f
T i Þ þ ms C EFB ðT f
T i Þ þ mw;ev p C w ðT f
an additional heater to produce syngas can be installed to super- 1
MC f
heat the flue gas before entering the rotary dryer.
T i Þ þ DH e v p ; ð10Þ
The specific heat capacity of the solid EFB at constant pressure,
CEFB, is approximated as the compound of fixed carbon and volatile where Cw, Tf, Ti, and DHevp are heat capacity of water (J kg K
1),
1

final temperature, inlet temperature, and latent heat of water evap-

matters. It can be written as follows [30]:
oration (J kg
1), respectively.
C EFB ¼ FC f þ V 1 C v 1 þ V 2 C v 2 ; ð1Þ In addition, the heat that is exchanged inside the rotary dryer
C f ¼
0:218 þ 3:807  10
3 T
1:7558  10
3 T 2 ; ð2Þ can be expressed as follows [34]:

C v 1 ¼ 0:728 þ 3:391  10 T;
3
ð3Þ Q ¼ U v VðDtÞm ; ð11Þ

1
C v 2 ¼ 2:273 þ 2:554  10 T;
3
ð4Þ where Q, Uv, V, and (Dt)m are rate of heat transfer (J s ), volumetric
heat transfer coefficient (J s
1 m
3 K
1), dryer volume (m3), and
where F, V1, and V2 are mass fractions (wt%) of fixed carbon, primary true mean temperature difference (K), respectively. Furthermore,
volatile matter and secondary volatile matter, respectively. As the considering that the number of flights, nf, inside the dryer is 12,
total mass fraction of volatile matter is greater than 10%, the mass the rate of heat transfer inside the dryer can be expressed as follows
fraction of secondary volatile matter, V2, is assumed to be 10%. In [35]:
addition, Cf, Cv1, and Cv2 are the heat capacities (kJ kg
1 K
1) of each
ðnf
1Þ 0:6
fixed carbon, primary volatile matter and secondary volatile matter, Q ¼ 0:228LD G Dtlm ; ð12Þ
respectively. In addition, the heat capacity of ash, Cash, is approxi- 2
mated as following equation. where L, D, G, and Dtlm are dryer length (m), inside diameter (m),

0:5
2
3
gas flow rate per unit area of dryer cross section (kg h
1 m
2),
C ash ¼ K 0 þ K 0 T þ K0T þ K0T ; ð5Þ and logarithmic temperature difference inside the dryer (K),
where Ko, K1, K2, and K3 are the coefficients for the constituent of respectively.
ash. In this study, as the amount of ash is relatively small and also In addition, the electricity consumed by the motor to rotate the
for simplification during calculation, ash is mainly constituted by rotary drum, Wmot, can be calculated as follows [36]:
CaO. Therefore, the values of Ko, K1, K2, and K3 are 58.79, N

1.34  10
2,
11.47  10
5, and 10.34  10
7, respectively [31]. W mot ¼ ½4:75Dw þ 0:1925ðD þ 2Þwr þ 0:33wr ; ð13Þ
735; 499
In addition, higher heating value (HHV, MJ kg
1) of EFB is calculated
based on the following correlation which was developed based on where N, w, and wr are rotation speed (rpm), material load (kg), and
Dulong’s formula [32]. total rotating load (kg), respectively.
In EFB gasification, in which a fluidized bed-type gasifier is
HHV ¼ 0:3941 C þ 1:1783 H þ 0:1005 S
0:1034 O used, the total pressure drop inside the bed during fluidization,

0:0151 N
0:021 A; ð6Þ DPf, can be approximated as follows [37]:
DP f ¼ DP b þ DP d ; ð14Þ
where C, H, S, O, N, and A are mass fractions of carbon, hydrogen,
sulphur, oxygen, nitrogen, and ash, respectively. In addition, the DPb ¼ ð1
emf Þðqp
qg ÞHg=c; ð15Þ
above HHV can be converted to lower heating value (LHV) by using DPd ¼ 0:4DPb ; ð16Þ
the following equation [33].
 M. Aziz et al. / Applied Thermal Engineering 114 (2017) 1378–1386 1383

where DPb, DPd, emf, qp, qg, H, g, and c are pressure drop across the Table 4
bed (kPa), pressure drop across the distributor (kPa), void fraction at Specifications of the ORC module and working fluids.

minimum fluidization velocity (–), particle density (kg m
3), gas Component Value
density (kg m
3), bed height (m), acceleration due to gravity Working fluid (–) Toluene, cyclohexane, n-heptane
(m s
2), and conversion factor (1 kg m N
1 s
2), respectively. Turbine efficiency (%) 90
To enhance uniform particle mixing and heat transfer inside the Turbine inlet pressure (MPa) 1–4
bed, silica sand is used as fluidizing particles, which also act as a Pump efficiency (%) 87
Minimum vapour quality (%) 90
heat carrier. Therefore, the minimum fluidization velocity of all
particles is approximated as the minimum fluidization velocity of Toluene properties
Boiling point (°C) 110.60
silica sand. The cooled and cleaned syngas is considered to be at
Critical temperature (°C) 318.64
ambient temperature before it is compressed for input to the gas Critical pressure (MPa) 4.11
engine.
Cyclohexane properties
To measure the effect of gasification temperature on total Boiling point (°C) 80.74
energy efficiency, three gasification temperatures are observed: Critical temperature (°C) 280.45
800, 900, and 1000 °C. The composition of produced syngas, Critical pressure (MPa) 4.08
excluding the remaining N2 and O2, from different gasification n-Heptane properties
temperatures is based on the experimental results of Mohammed Boiling point (°C) 98.42
Critical temperature (°C) 266.98
et al., as shown in Table 3 [38]. Cold gas efficiency is calculated
Critical pressure (MPa) 2.50
based on the work of Kihedu et al. [39]. Generally, gasification
yields increase following the increase in gasification temperature.
Therefore, a higher LHV of produced syngas is achieved. Note that,
efficiency is used. The generation efficiency of ORC, gORC, can be
because air is used as a gasification medium, unreacted gas (espe-
expressed as follows:
cially N2) is exhausted together with the syngas from the gasifier.
As the work of Mohammed et al. is a non auto-thermal system, Wt
Wp
in system approximation, a part of produced syngas is assumed gORC ¼ ; ð17Þ
Q exh
to be combusted (reacted with air) to provide additional heat
required to achieve the gasification temperature, in addition to where Wt, Wp, and Qexh are work generated by turbine, work
the heat brought by preheated air and EFB. As the result, the real required by pump, and heat amount of flue gas flowing to ORC mod-
cold gas efficiency is lower than one in Table 3. The achieved cold ule, respectively.
gasification efficiencies from the system are 30.6%, 54.2%, and
83.3% for gasification temperatures of 800, 900 and 1000 °C, 4. Results and discussion
respectively. According to Kihedu et al. [39], cold gas efficiency of
biomass in auto-thermal system can achieve up to about 91% using 4.1. Integrated EFB gasification and POME digestion modules
air or air-steam as reactant. They explained that while CO2 is pro-
duced at the lower part of the gasifier because of partial oxidation, Fig. 5 shows the relationship between the net generated power,
CO and H2 are generated in the middle and bottom parts of the power generation efficiency, and EFB gasification temperature. It is
gasifier. important to note that the net generated power includes both gen-
Regarding the ORC cycle, three working fluids are observed: erated power from POME digestion and EFB gasification modules.
toluene (C7H8), cyclohexane (C6H12), and n-heptane (C7H16). They In addition, generated power from the POME digestion module is
are selected for this study based on thermal stability, minimal not influenced by the gasification temperature and is constant
degradation rate over time, availability, operating conditions, and under the same FFB flow rate during palm oil milling. Approxi-
cost [40,41]. Table 4 shows the specifications for the ORC module mately 1.18 MW of electricity is generated from biogas produced
used in this study. Therefore, it is expected that high efficiency by the POME digestion module. As the gasification temperature
can be achieved. Turbine inlet pressure is observed from 1 MPa increases, the net generated power increases, particularly the gen-
to pressure below the critical pressure with an interval of erated power from the EFB gasification module. Numerically, the
0.5 MPa. In addition, the flow rate of the working fluid is adjusted, net generated power increases from 3.05 MW (gasification temper-
thereby achieving maximum generated power. Note that the ature of 800 °C) to 7.46 MW (gasification temperature of 1000 °C).
amount of heat and composition of flue gas will be different for The power generation efficiencies are 11.2%, 16.3%, and 24.6% for
each EFB gasification temperature. each corresponding gasification temperatures of 800, 900, and
To observe the performance of each working fluid in converting 1000 °C.
the thermal energy of exhausted flue gas to electricity, generation

8 40%
Net generated power (MW)

Power generation efficiency

Generated power

Table 3 6 Power generation efficiency 30%

Gasification yield at three gasification temperatures [38].

Component Temperature (°C) 4 20%

800 900 1000
Total yield (%) 68.24 80.05 91.7 2 10%
LHV (MJ m
3) 11.86 13.84 15.55
Cold gas efficiency (%) 34.7 58.8 88.5
0 0%
Gas composition 800 900 1000
H2 (vol%) 17.23 27.42 38.02
Gasification temperature (oC)
CO (vol%) 33.35 33.08 36.36
CH4 (vol%) 11.74 14.29 14.72
Fig. 5. Relationship between net generated power and EFB gasification
CO2 (vol%) 37.68 25.21 10.90
temperature.
1384 M. Aziz et al. / Applied Thermal Engineering 114 (2017) 1378–1386

As the gasification temperature increases, both the LHV and gas 1.5 Toluene (electricity) Cyclohexane (electricity) 30
n-Heptane (electricity) Toluene (efficiency)
yield from EFB gasification increase (Table 3). On the other hand, Cyclohexane (effciency) n-Heptane (efficiency)

Generation efficiency (%)

the amount of produced char, tar and liquid yields decreases

Generated power (MW)

1.2 24
accordingly [38]. At high gasification temperature, the produced
gases from pyrolysis zone can undergo further secondary reaction 0.9 18
such as tar cracking and shift reactions leading to higher gas yield
[42]. As a result, higher generation of both power and heat is 0.6 12
achieved. Therefore, a higher gasification temperature is preferred
for EFB gasification in terms of total energy efficiency.
0.3 6
Furthermore, EFB drying can be facilitated by using the flue gas
heat exhausted from gas engines in both the EFB gasification and
0 0
POME digestion modules. No additional input of heat is required 0 1 2 3 4 5
from outside. Conversely, a large amount of remaining heat from
Turbine inlet pressure (MPa)
the flue gas is produced, as shown in Fig. 6. Again, the total remain-
ing heat from cogeneration systems in both modules increases fol- Fig. 7. Relationships among generated power, generation efficiency, and turbine
lowing the increase in gasification temperature. When the inlet pressure in the ORC module (gasification temperature 900 °C).
gasification temperature is 800 °C, the remaining heat from cogen-
eration in both modules produces only about 1.1 MW. The pro-
1.5 30
duced energy increases significantly to about 8 MW when the
gasification temperature is increased to 1000 °C.

Generation efficiency (%)

Generated power (MW)
A large amount of remaining heat can be used as the heat source 1.2 24

for other processes creating a cogeneration. The excess heat can be

utilized to produce steam which can be supplied to the mills 0.9 18
including palm oil and palm kernel oil mills as both milling pro-
cesses generally require a large amount of steam during their pro- 0.6 12
cesses, particularly during EFB sterilization and palm kernel drying
[1]. On the other hand, this surplus heat also can be utilized to gen- 0.3 Toluene (electricity) Cyclohexane (electricity) 6
erate additional power using ORC module. In addition, the flue gas n-Heptane (electricity) Toluene (efficiency)
Cyclohexane (effciency) n-Heptane (efficiency)
having lower temperature from the dryer can be supplied to an 0 0
anaerobic digestion pond to enhance biogas production as well 0 1 2 3 4 5
as increase the rate of COD removal [43,44]. Turbine inlet pressure (MPa)

Fig. 8. Relationships among generated power, generation efficiency, and turbine

inlet pressure in the ORC module (gasification temperature 1000 °C).

4.2. Integration of the ORC module

Fig. 6 shows that the remaining heat exists in a relatively large
amount when the gasification temperatures reach 900 and
1000 °C. Therefore, the ORC module can be used under these gasi-
fication temperatures. Figs. 7 and 8 show the relationships among
the generated power, generation efficiency, and turbine inlet pres-
sure when the gasification temperatures reach 900 and 1000 °C,
respectively. Moreover, Tables 5 and 6 shows the properties of rep-
resentative streams resulted from process calculation for EFB gasi-
fication temperatures of 900 and 1000 °C, respectively (working
fluid and pressure are cyclohexane and 4 MPa, respectively). In
general, as the turbine inlet pressure increases, generated electric-
ity and generation efficiency increase accordingly. In addition,
under the observed heat source temperature (450 °C), cyclohexane
shows the highest generated power and efficiency, followed by
toluene and n-heptane.
8
Remaining heat (MW)
An EFB gasification temperature of 1000 °C results in higher
generated power by the ORC module due to higher flue gas heat
exhausted from the EFB gasification module. Numerically, when
cyclohexane is used as a working fluid, the additional electricity
generated from the ORC module increases from 0.42 MW to about
1.23 MW when the EFB gasification temperature increases from
900 °C to 1000 °C. Further, generation efficiency is very similar
for each working fluid under the same turbine inlet pressure. The
highest generation efficiency is shown by cyclohexane, which is
about 15.5% in the case of a turbine inlet pressure of 4 MPa. Fur-
thermore, toluene and n-heptane show the highest efficiencies of
about 13.7% and 12.3%, respectively, based on the input heat (the
received heat by ORC module from EFB gasification module).
Finally, the water used to condense the working fluid (which is
preheated in the condenser to relatively high temperatures of
about 50–80 °C, depending on flow rate), can be used to warm
the digestion pond.
According to the category of working fluids for ORC system [45],
toluene and cyclohexane are categorized as an isentropic fluid hav-

ing a nearly vertical vapour saturation curve on temperature-

6 entropy diagram. As the result, vapour remains saturated through-
out turbine expansion without condensation. Furthermore, cyclo-
4 hexane has higher specific heat capacity compared to other
fluids. On the other hand, n-heptane is grouped as dry fluid having
2 positive slope on temperature-entropy diagram. Therefore, the sat-
urated vapour becomes superheated after isentropic expansion. In
addition, as both isentropic and dry fluids need no superheating,
0
700 800 900 1000 1100 the impingement of liquid droplets in turbine during expansion
Gasification temperature (oC) can be avoided.

Fig. 6. Relationship between remaining heat and EFB gasification temperature.

 M. Aziz et al. / Applied Thermal Engineering 114 (2017) 1378–1386 1385

Table 5
Stream properties from the calculation results for gasification temperature of 900 °C (working fluid and pressure are cyclohexane and 4 MPa, respectively).

Stream Temp. (°C) Pressure (kPa) Flowrate (kg h
1) Stream Temp. (°C) Pressure (kPa) Flowrate (kg h
1)
S1 25 101.33 10,000 S20 450 108.33 37,660
S7 254 101.33 4212 S22 450 108.33 12,751
S8 900 110.33 3557 S23 80 101.33 26,600
S10 293 107.33 3557 S26 35 120.33 633
S14 450 110.33 86,308 S34 281 3994.33 15,181
S15 450 108.33 21,034 S35 170 107.33 15,181
S16 450 106.33 33,785 S36 82 104.33 15,181
S17 25 101.33 1605 S37 30 101.33 15,181
S19 890 137.33 1605 S39 145 3997.33 15,181

Table 6
Stream properties from the calculation results for gasification temperature of 1000 °C (working fluid and pressure are cyclohexane and 4 MPa, respectively).

Stream Temp. (°C) Pressure (kPa) Flowrate (kg h
1) Stream Temp. (°C) Pressure (kPa) Flowrate (kg h
1)
S1 25 101.33 10,000 S20 450 108.33 79,824
S7 283 101.33 4212 S22 450 108.33 12,751
S8 1000 110.33 4075 S23 80 101.33 26,600
S10 485 107.33 3557 S26 35 120.33 633
S14 450 106.33 148,308 S34 280 3994.33 36,165
S15 450 108.33 21,034 S35 167 107.33 36,165
S16 450 106.33 33,785 S36 82 104.33 36,165
S17 25 101.33 1477 S37 30 101.33 36,165
S19 990 137.33 1477 S39 142 3997.33 36,165

By using ORC, the highest total generated power and power
generation efficiency are 8.3 MW and 30.4% which can be reached
under gasification temperature of 1000 °C and cyclohexane as
working fluid in ORC module (FFB flow rate of 50 t h
1). In Indone-
sia, the total production of CPO in 2014 is about 31 Mt [46]. Assum-
ing a ratio of produced CPO to FFB of 1:5, total additional power
generated from palm oil mills throughout the country reaches
about 4.5 GW. This potential is expected to increase national
energy security as well as improve access to electrical power
throughout the country.
5. Conclusions
An enhanced utilization of solid and liquid wastes from palm oil
milling processes for power generation is proposed and evaluated.
The proposed system consists of EFB gasification, POME digestion,
and additional ORC modules. A cogeneration system, producing
both electricity and heat, produces syngas and biogas from both
modules. In addition, the surplus and unused heat from the system
is converted to electricity through an additional ORC module.
Based on simulations, it is clear that the proposed system is
very feasible in terms of potential power generation. A higher level
of EFB gasification at 1000 °C is recommended because it can yield
significantly higher net power generation. Under this condition,
the obtained total generated power and power generation effi-
ciency are 8.3 MW and 30.4%, respectively. Application of the pro-
posed system is expected to increase the economic performance of
the mills. In addition, because palm oil mills are distributed
throughout remote and rural areas lacking access to the electricity
grid, this application can further support the supply of electricity
and increase electrification across the country.
Acknowledgements
The authors would like to express their deep thanks to Energy
Technology Center (B2TE), Agency for the Assessment and Applica-
tion of Technology (BPPT), Indonesia, for the assistance in provid-
ing some data required in this study.
References
[1] M. Aziz, T. Oda, T. Kashiwagi, Design and analysis of energy-efficient integrated
crude palm-oil and palm kernel oil processes, J. Jpn. Inst. Energy 94 (2015)
143–150.
[2] C. Ofori-Boateng, K.T. Lee, An oil palm-based biorefinery concept for cellulosic
ethanol and phytochemicals production: sustainability evaluation using
exergetic life cycle assessment, Appl. Therm. Eng. 62 (2014) 90–104.
[3] M. Aziz, D. Budianto, T. Oda, Computational fluid dynamic analysis of co-firing
of palm kernel shell and coal, Energies 9 (2016) 137.
[4] F. Schuchardt, K. Wulfert, D. Darnoko, T. Herawan, Effect of new palm oil mill
processes on the EFB and POME utilization, J. Oil Palm Res. 20 (2008) 115–126.
[5] D. Budianto, M. Aziz, Chayadi, T. Oda, Numerical investigation of co-firing of
palm kernel shell into pulverized coal combustion, J. Jpn. Inst. Energy 95
(2016) 605–614.
[6] M. Aziz, T. Oda, T. Kashiwagi, Innovative steam drying of empty fruit bunch
with high energy efficiency, Drying Technol. 33 (2015) 395–405.
[7] P. Chaivatamaset, P. Sricharoon, S. Tia, B. Bilitewski, The characteristics of bed
agglomeration/defluidization in fluidized bed firing palm fruit bunch and rice
straw, Appl. Therm. Eng. 70 (2014) 737–747.
[8] S.S. Harsono, P. Grundman, L.H. Lau, A. Hansen, M.A.M. Saleh, A. Meyer-Aurich,
et al., Energy balances, greenhouse gas emissions and economics of biochar
production from palm oil fruit bunches, Resour. Conserv. Recycl. 77 (2013)
108–115.
[9] M.C. Law, E.L. Liew, S.L. Chang, Y.S. Chan, C.P. Lwo, Modelling microwave
heating of discrete samples of oil palm kernels, Appl. Therm. Eng. 98 (2016)
702–726.
[10] M. Aziz, P. Prawisudha, B. Prabowo, B.A. Budiman, Integration of energy-
efficient empty fruit bunch drying with gasification/combined cycle systems,
Appl. Energy 139 (2015) 188–195.
[11] H.T. Luk, T.Y.G. Lam, A.O. Oyedun, T. Gebreegziabher, C.W. Hui, Drying of
biomass for power generation: a case study on power generation from empty
fruit bunch, Energy 63 (2013) 205–215.
[12] S. O-Thong, K. Boe, I. Angelidaki, Themophilic anaerobic co-digestion of oil
palm empty fruit bunches with palm oil mill effluent for efficient biogas
production, Appl. Energy 93 (2012) 648–654.
[13] A. Ahmad, R. Ghufran, Z.A. Wahid, Bioenergy from anaerobic degradation of
lipids in palm oil mill effluent, Rev. Environ. Sci. Biotechnol. 10 (2011) 353–
376.
[14] M. Aziz, T. Oda, T. Kashiwagi, Advanced energy harvesting from macroalgae –
innovative integration of drying, gasification and combined cycle, Energies 7
(2014) 8217–8235.
[15] F.J. Gómez-de la Cruz, P.J. Casanova-Peláez, J.M. Palomar-Carnicero, F. Cruz-
Peragón, Modeling of olive-oil mill waste rotary dryers: green energy recovery
systems, Appl. Therm. Eng. 80 (2015) 362–373.
[16] M. Aziz, Power generation from algae employing enhanced process integration
technology, Chem. Eng. Res. Des. 109 (2016) 297–306.
[17] M. Khaljani, R. Khosbakhti Saray, K. Bahlouli, Comprehensive analysis of
energy, exergy and exergo-economic of cogeneration of heat and power in a
combined gas turbine and organic Rankine cycle, Energy Convers. Manage. 97
(2015) 154–165.
1386 M. Aziz et al. / Applied Thermal Engineering 114 (2017) 1378–1386
[18] H.C. Jung, S. Krumdieck, T. Vranjes, Feasibility assessment of refinery waste
heat-to-power conversion using an organic Rankine cycle, Energy Convers.
Manage. 77 (2014) 396–407.
[19] M.A. Siddiqi, B. Atakan, Alkanes as fluids in Rankine cycles in comparison to
water, benzene and toluene, Energy 45 (2012) 256–263.
[20] R. Chacartegui, D. Sanchez, J.M. Munoz, T. Sanchez, Alternative ORC bottoming
cycles FOR combined cycle power plants, Energy 86 (2009) 2162–2170.
[21] M.A.A. Mohammed, A. Salmiaton, W.A.K.G. Wan Azlina, M.S. Mohammad
Amran, A. Fakhru’l-Razi, Y.H. Taufiq-Yap, Hydrogen rich gas from oil palm
biomass as a potential source of renewable energy in Malaysia, Renew.
Sustain. Energy Rev. 15 (2011) 1258–1270.
[22] M.A.A. Mohammed, A. Salmiaton, W.A.K.G. Wan Azlina, M.S. Mohammad
Amran, Gasification of oil palm empty fruit bunches: a characterization and
kinetic study, Bioresource Technol. 110 (2012) 628–636.
[23] A. Demirbas, Combustion characteristics of different biomass fuels, Prog.
Energy Combust. Sci. 30 (2004) 219–230.
[24] Yanmar gas engine CP series, Yanmar Co., Ltd. <https://www.yanmar.com/jp/
energy/renewable_energy/digestion_gas/products/> (accessed February
2016).
[25] H. Li, Q. Chen, X. Zhang, K.N. Finney, V.N. Sharifi, J. Swithenbank, Evaluation of
a biomass drying process using waste heat from process industries: a case
study, Appl. Therm. Eng. 35 (2012) 71–80.
[26] A. Uusitalo, V. Uusitalo, A. Gronman, M. Luoranen, A. Jaatinen-Varri,
Greenhouse gas reduction potential by producing electricity from biogas
engine waste heat using organic Rankine cycle, J. Clean. Prod. 127 (2016) 399–
405.
[27] J. Wu, Y. Ma, Experimental study on performance of a biogas engine driven air
source heat pump system powered by renewable landfill gas, Int. J. Refrig. 62
(2016) 19–29.
[28] Zhengzhou Azeus Machinery Co., Ltd. EFB Shredder. <http://www.
biopelletmachine.com/product/sawdust-making-machine/EFB-shredder-
crushing-machine.html> (accessed August 2016).
[29] Kim Dragon Company. Ball Mill. <http://www.kvdc1.in/product/grinding/wet-
type-overflow-ball-mill.html> (accessed August 2016).
[30] G. Tsatsaronis, M. Winhold, Exergoeconomic analysis and evaluation of
energy-conversion plants—II. Analysis of a coal-fired steam power plant,
Energy 10 (1985) 81–94.
[31] R.G. Berman, T.H. Brown, Heat capacity of minerals in the system Na2O-K2O-
CaO-MgO-FeOFe2O3-A12O3-SiO2-TiO2-H2O-CO2: representation, estimation,
and high temperature extrapolation, Contrib. Mineral. Petrol. 89 (1985),
168-163.
[32] S.A. Channiwala, P.P. Parikh, A unified correlation for estimating HHV of solid,
liquid and gaseous fuels, Fuel 81 (2002) 1051–1063.
[33] WBCSD Cement Sustainability Initiative (CSI)/ECRA GmbH. Lower and Higher
Heating Values (LHV and HHV). <http://www.cement-co2-protocol.org/en/
Content/Internet_Manual/tasks/lower_and_higher_heating_values.htm>
(accessed August 2016).
[34] M. Krokida, D. Marinos-Kouris, A.S. Mujumdar, Rotary drying, in: A.S.
Mujumdar (Ed.), Handbook of Industrial Drying, CRC Press, Boca Raton,
2006, pp. 151–173.
[35] C.O. Miller, B.A. Smith, W.H. Shuette, Factors influencing the operation of
rotary dryers. Part 2. The rotary dryer as a heat exchanger, Trans. AICHE 38
(1942) 841–864.
[36] Y. Liu, M. Aziz, Y. Kansha, S. Bhattacharya, A. Tsutsumi, Application of the self-
heat recuperation technology for energy saving in biomass drying system, Fuel
Process Technol. 117 (2014) 66–74.
[37] Y. Liu, M. Aziz, C. Fushimi, Y. Kansha, K. Mochidzuki, S. Kaneko, et al., Exergy
analysis of biomass drying based on self-heat recuperation technology and its
application to industry: a simulation and experimental study, Ind. Eng. Chem.
Res. 51 (2012) 9997–10007.
[38] M.A.A. Mohammed, A. Salmiaton, W.A.K.G. Wan Azlina, M.S. Mohammad
Amran, A. Fakhru’l-Razi, Air gasification of empty fruit bunch for hydrogen-
rich gas production in a fluidized-bed reactor, Energy Convers. Manage. 52
(2011) 1555–1561.
[39] J.H. Kihedu, R. Yoshie, I. Naruse, Performance indicators for air and air–steam
auto-thermal updraft gasification of biomass in packed bed reactor, Fuel
Process Technol. 141 (2016) 93–98.
[40] E. Prabhu, Solar through organic Rankine electricity system (STORES) stage 1:
Power plant optimization and economics Subcontract Report NREL/SR-550-
39433, National Renewable Energy Laboratory (NREL), 2006.
[41] J. Facao, A.C. Oliveira, Analysis of energetic, design and operational criteria
when choosing an adequate working fluid for small ORC systems, in: Proc.
ASME 2009 Int Mech Eng Congress Exposition (IMECE 2009), Nov 13–19, 2009,
Florida, USA, 2009.
[42] X.W. Dai, Z.Q. Zhou, C.Z. Wu, Y. Chen, Fast pyrolysis of biomass in CFB reactor,
Energy Fuels 14 (2000) 552–557.
[43] J.Y. Jeong, S.M. Son, J.H. Pyon, J.Y. Park, Performance comparison between
mesophilic and thermophilic anaerobic reactors for treatment of palm oil mill
effluent, Bioresour. Technol. 165 (2014) 122–128.
[44] H.Q. Yu, H.H.P. Fang, G.W. Gu, Comparative performance of mesophilic and
thermophilic acidogenic upflow reactors, Process Biochem. 38 (2002) 447–
454.
[45] J. Bao, L. Zhao, A review of working fluid and expander selections for organic
Rankine cycle, Renew. Sustain. Energy Rev. 24 (2013) 325–342.
[46] Indonesia Investments. Palm Oil. Available from <http://www.indonesia-
investments.com/business/commodities/palm-oil/item166> (accessed June
2015).