Effect of Harmonic Current

i
FOREWORDS
Dean of Faculty of Engineering, University of Indonesia

The Quality in Research (QIR) Conference is the annual event organized by the Faculty of
Engineering, University of Indonesia. Since started in 1998, it has become an excellent forum of
discussion for all researchers from research institutions and universities all over the nation of
Indonesia. The 1
st
and 6
th
QIR Conferences had been successfully organized as the high quality
national conferences, and starting from the 7
th
QIR conference, has been organized to invite
international research papers.
The 10
th
Quality in Research International Conference having a theme of Research for future better
life is to provide an international forum for exchange of the knowledge, information, experience and
result as well as the review of progress and discussion on the state of the art and the future trend
various issues and the developments in the multi-fields of scientific and technology. The main
purposes of this conference are to provide a forum for free discussion of new ideas, development and
applications, including techniques and methods to stimulate and inspire pioneering work, to provide
opportunities for students and young engineers to meet their experienced peer and to provide a
meeting that will enforce progress, stimulate growth and advance the state of knowledge in the multi-
fields of science and technology.
We would like to express our heartiest to thank to all authors and participants for their active
participations in the 10
th
on Quality in Research (QIR) International Conference 2007, and also to all
the paper reviewers, member of the technical committees, and member of the organizing committees,
for their support to the success of this conference. Last but not the least; we would also like to invite
all participants to the next on Quality in Research (QIR) Conference.

Faculty of Engineering, University of Indonesia
Dean,

Prof. Ir. Rinaldy Dalimi, Ph.D

ii

FOREWORDS
Chairman of 10th International Conference on QIR 2007

The 10
th
Quality in Research International Conference will provide an international forum for
exchange the knowledge, information, experience and recent researches of various fields. With a
strong support and presentations from academic, industry and entrepreneurs, the conference will
provide an ideal platform to learn various fields and understand technological trends in the region.
The 10
th
Quality in Research (QIR) International Conference has a theme of Research for Better
Future Life being the third time to go internationally, has invited limited papers from other nations
such as Korea and Malaysia. The conference is organized in parallel sessions focusing on the 8 (eight)
research areas such that many researchers and peer groups may focus their discussion on the relevant
topics. All submitted papers had been reviewed by the technical committees and had been arranged
into 8 (eight) sub-themes according to the following fields:
Energy, Process and Environmental Engineering and Management: Energy and
environmental issues, combustion technology, fluids mechanics and thermal fluid machinery,
thermodynamics and heat transfer, geotechnical and environmental engineering, etc.
Industrial, Manufacturing, Material Engineering, and Management: Production
Engineering, Supply Chain Management, Innovation System, Maintenance System, Quality
Management System, Human Factors Engineering, Organizational System, Fabrication and
Industrial Automation, Manufacturing System: Control Management and Information
Technology, etc
Biomaterial, Biomedical Engineering and Biotechnology: Biomedical numerical modeling,
Biomaterial, Biosensor, Biocompatibility, Biomechanics, Biotechnology, Biomedical
Instrumentation, Biomedical Imaging
Design and Infrastructure Engineering and Management: Product design and
development, composite: Materials and applications, structural dynamics, mechanics of
materials, Construction Management, Public Infrastructures and Services, Structural
Engineering, etc
Special Session on Electronics Engineering
Information and Computation Technology
Sustainable Architecture
Nanomaterials and Nanotechnology: Nano structured material, Nanotechnology,
Nanocomposite, MEMS, Self Assembled Monolayer, Thin Film, etc
The main purposes of this conference are to provide a forum for free discussion of new ideas,
development and applications, including techniques and methods to stimulate and inspire pioneering
work, to provide Opportunities for students and young engineers to meet their experienced peer and to
provide opportunities for students and young engineers to meet their experienced peer and to provide
a meeting that will enforce progress, stimulate growth and advance the state of knowledge in the
multi-fields of science and technology.

Depok, 4 December 2007
The Organizing Committee,
Chairman,

iii
Gunawan Wibisono, Ph.D

Steering Committee

1. Prof. Dr-Ing. Axel Hunger, Universitaet, Duisburg-Essen, Germany
2. Prof. Dr. Carlo Morandi, Universida Degli Studi de Parma, Italy
3. Prof. Dr. Iwao Sasase, KEIO University, Japan
4. Prof. Kim Kyoo-ho, Yeungnam University, Korea
5. Prof. Dr. Ir. Irwan Katili, University of Indonesia
6. Prof. Dr. Ir. Bambang Suryawan, MT, University of Indonesia
7. Prof. Dr. Ir. Dadang Gunawan, M.Eng, University of Indonesia
8. Prof. Dr. Ir. Johny W Soedarsono, DEA, University of Indonesia
9. Prof. Ir. Gunawan Tjahjono, M.Arch, Ph.D, University of Indonesia
10. Prof. Dr. Ir. M. Nasikin, University of Indonesia
11. Isti Surjandari, Ph.D, University of Indonesia
12. Prof. Dr. Ir. Budi Susilo Soepandji, University of Indonesia
13. Prof. Dr. Ir. Sutanto Soehodho, University of Indonesia
14. Prof. Dr. Ir. Sulistyoweny Widanarko, Dilp. SE. MPH, University of Indonesia
15. Prof. Dr. Ir. I Made Kartika Dipl. Ing, University of Indonesia
16. Prof. Dr. Ir. Tresna P. Soemardi, University of Indonesia
17. Prof. Dr. Ir. Sardy, M.Eng, M.Sc, University of Indonesia
18. Prof. Dr. Ir. Bagio Budiardjo M.Sc, University of Indonesia
19. Prof. Dr. Ir. Djoko Hartanto, M.Sc, University of Indonesia
20. Prof. Dr. Ir. Eddy Siradj, M.Eng, University of Indonesia
21. Dr. Ir. Kemas Ridwan K, University of Indonesia
22. Prof. Dr. Widodo Wahyu P, DEA, University of Indonesia
23. Ir. Boy Nurtjahyo M.,MSIE, University of Indonesia
24. Dr. Ir. Dedi Prihadi DEA, University of Indonesia
25. Ir. Hendri D.S. Budiono, M.Eng, University of Indonesia
26. Dr. Ir. Sigit Pranowo Hadiwardoyo, DEA, University of Indonesia
27. Dr. Ir. Herr Soeryantono, University of Indonesia
2 28 8. . Prof. Rinaldy Dalimi, Ph.D, University of Indonesia

Chairman of the Conference

Ir. Gunawan Wibisono, M.Sc, Ph.D

Technical Committee

1. Ir. Gunawan Wibisono, M.Sc, Ph.D
2. Dr. Yosia Irwan
3. Dr. Engkos Kosasih
4. Purnomo Sidi P, Ph.D
5. Abdul Muis, Ph.D
6. Badrul Munir, Ph.D
7. Tania Surya Utami, MT
8. Ir. Beatrianis, M.Si

Foreword From The Dean of Faculty Engineering, UI
Foreword From Chairman of 10th International Conference on QIR 2007
The Committee of 10th International Conference on QIR 2007
Paper No. Title and Name of Author(s)
EPE-01 The Effect of Harmonics Current to Performance of The Over Current Relay
by: Syafrudin and Dahaman
EPE-02 Effect of The Harmonic Current Component to Active Power Losses on Power Transformer
by: Syafrudin and Dahaman I
EPE-03
Optimize placing of passive filter and reduce harmonic in power system, Case study at PT ISPAT
INDO
by: Iwa Garniwa, Budi S, Adrianto
EPE-04
Convergence Approximation Of ABR Formulation For Josephsons Tunneling in UO
2
Chain
Reaction at 45.7 MW Candu Nuclear Reactor
by: Moh. Hardiyanto
EPE-05 Energy Conservation on Steel Industri By Increasing Efficiency and Electric Power Quality
by: Iwa Garniwa, Budi S, Fauzan H J
EPE-06 Energy Conservation on Hotel Building (Case Study in Bandung)
by: Rudy S and Aji Nur W
EPE-07 Indonesia Energy Revolution Scenario and Projection using MESAP/Planet Simulation Model
by: Rinaldy D, Bayu I, and Sven Teske
EPE-08 Ocean wave power plant with air pressure
by: Massus Subekti and Vivian K L
EPE-09 The Time for Indonesia to Optimize Geothermal Energy
by: Massus Subekti
EPE-10
Incorporating energy Commodity Price Volatility in economic Cost of Supply Analysis for
Electricity Expansion Plan
by: Emil E. Dardak
EPE-11
Effects of Chimney Depth and Downjet Height in A Coal briquette Stove on the Co Emission and
Ignition Time
by: Dijan S., Yulianto S.N., and Dian N. Kusuma
EPE-12 Scale up Reactor for Bio-gasoline Production from Crude Palm Oil
by: Bambang HS., Anondho W., and M. Nasikin
EPE-13 Flame Lift-up on A Bunsen Burner; A Preliminary Study
by: C Prapti Mahandari and I Made Kartika D
EPE-14
Study of Bio-Diesel of Coconut and Corn Ethyl Ester Use By processing With The Processor
Series Type in Diesel Engine Performance
by: B. Sugiarto, Sanggul H. Siregar, and Yanuar C.
EPE-15 On The Measurement of Smoke Production rate of Tropical wood
by: Yulianto S. N.
EPE-16 Catalytic Ozonation of Endosulfan in Water With Activated carbon
by: Enjarlis, Setijo Bismo, Slamet, and Roekmijati WS
TABLE OF CONTENTS
EPE-17
Hydrocarbon Selectivity of Fisher Trosch Synthesis from syngs with ratio of H2 to Co=1.0over Co-
Fe/AI2O3 bimetalic catalysts: effect of Fe:Co ratio
by: Dewi Tristantini and Borje Gevert
EPE-18
The deactivation behaviour of HZSM-5 catalys at various temperature of acetone conversion to
mono-aromatic hydrocarbon
by: Setiadi, Prilly F J, and Toshinori K
EPE-19 CO2 Absorpation Through Hollow Fiber Membrane Gas-Liquid Contactors
by: Sutrasno K
EPE-20
Enzymmatic and acid hydrolysis of bagasse for ethanol production by simultanious sacharification
and fermentation
by: Samsuri, Gozan, Nasikin, and Prasetya
EPE-22 Biogrease using modified palm oil as base oil and thickener lithium soap
by: Sukirno, M. Nasikin, B. Heru, Rizqon F, Marius, Dizzi
EPE-21
A general Application of A Direct Method For Multivariable MPC Control Strategy in Chemical
Process
by: A. Ahmad, A. Wahid
EPE-23
Model for Oxidation and Combustion of Ethanol in Wide range of Pressures, Temperatures and
Equivalence Ratios
by: Yuswan Muharam and Soraya S
EPE-24 Water and air Circulation System in Trimaran Fishing Veseel Live fish holds
by: Sunaryo, Anita Muslih
EPE-25 Local Protein Structure Comparison
by: Ford Lumban Gaol and Belawati W
EPE-26
Antibacterial activity Analysis of Dellenia indiaca's Kernel Extract for Escherichia coli: A
Preliminary Study
by: Rita Arbianti, Tania S U, and Ifa P
EPE-27 Antioxidant Activity Analysis for Polar Extract of Dillena Indica'S Kernel
by: Tania Surya Utami, rit Arbianti, Lina FR
EPE-28
Understanding The Metaspaces of Street Vendors in The Cities: Temporality, Strategies and
Tactics
by: Yandi Andri, Paramita A.
EPE-29 Bringing Sustainability Into The Curriculum of Architectural design Studio
by: Paramita A and Yandi Andri.
EPE-30 LED Technology in Architectural Ligthing of Building'S Faade
by: Siti Hanjarinto, Gregorius A.G. A.
EPE-31 Form in The new Architecture Perspective
by: Yandi Andri Yatmo and Kristanti Dewi Paramita
EPE-32
Utilization Effectiveness of Jakarta Municipal Asset (Case Study: East Jakarta Municipal Youth
Center)
by: Azrar Hadi and Ii Karunia
EPE-33
Water Quality Index Indicator of Water Pullutin. A Case Study in west Tarum Channel as Sources
of Water Supply
by: Djoko M. Hartono, Sulistyoweni W., Dwita Sutjiningsih M.
EPE-34
Non Linear Dynamic Identification of Waste Heat Boiler Unit Based on Adaptive Neuro Fuzzy
Inference System (ANFIS)
by: Yuliati
EPE-35 Load Pattern and Realibility Assessment of Power Generation System
by: Suparman, Zuhal, and Rinaldy D.
ISSN: 1411-1284

Proceeding 10
th
Intl QIR 4-6 Dec 2007 EPE-01 1/4

The Effect of Harmonics Current to Performance of
The Over Current Relay

Syafrudin
(1)(2)
and Dahaman Ishak
(1)

(1)
School of Electric and Electronic Engineering Universiti Sains Malaysia
(2)
Fac. of Engineering Electric Department, University of Tanjungpura, Jl. A. Yani, Pontianak
email : syaf_ee@yahoo.com , Syaf@eng.usm.my and dahaman@eng.usm.my

Abstract- The over current relay protection on the
power system is a device that can measure and
responds to abnormal condition and control circuit
breaker to isolate the faulty section of the system.
The over current relays are designed to operate
with sinusoidal current waveform. Now, the
current waveform on the power system have been
distorted non-sinusoidal and lot of containing
harmonic current component, caused by non-liner
loads. For harmonic current condition of system
the performances of that over current relay to be
abnormal.
In this paper, the effect of current harmonic
distortion percentage to over current relay
protection performances is discussed. In this
experiment used an inverse time over current relay
type of induction disc relay. The non-sinusoidal
load currents that consist of different harmonic
spectra were applied to inverse time over current
relay and these nonlinear-load currents were
recorded by a power analyzer instrument
FLUKE 42. According to the experiment results,
the pick up current and the operating time of the
inverse time over current relay increase
proportionally to the total harmonic distortion
(THD) value of the non-sinusoidal current. It is
concluded that, this type of relay cannot protect
the system reliably due to the harmonic
components of current.

Keywords- Non-linear load, Harmonic current,
inverse time over current relay.

I. INTRODUCTION

istorted current waveform content harmonics current
are introduced into power distribution system due to
non-linear loads or system components presenting
non-linear behavior. The increasing use highly non-
linear electric equipment (non-linear loads), the
advancement of semiconductor technology and the
use of more power electronic equipment such as
converters for adjustable speed drives (ASD), switch
mode power supplies, power semiconductor
controller, computers, photocopy machines, TV set
and etc have caused an increasing concern over the
generation of harmonic current and its effects on the
power system. Harmful effect of the harmonic current
on power distribution system like increasing losses,
interferences to communication system, induce
abnormal noise, overheat and burning in various
devices[1].
Harmonic current on power distribution system
also can cause some problems in over current
protection system that provided complete protection
and reliability for the power system. Most of the relay
manufacturers design their relays for pure sinusoidal
current waveform and voltages[2]. The well known
characteristics of relays are not valid under distorted
waveform. Each harmonic frequency component
could produce an independent and cumulative effect,
causing the pick up value of the relay to change
depending on the magnitudes of harmonic
components. Therefore, the relays can not protect the
line or transformer reliably when harmonics are
involved.
In this paper is purposed to determining of the
inverse time over current relay performances for non-
linear load current that containing of different total
harmonic current distortion. In this experiment, non-
sinusoidal wave current that containing harmonic
were applied to the inverse time over current relay and
these currents, pick up current and the operating time
were recorded by a power analyzer instrument.

II. PERFORMANCE OF THE INVERSE TIME
OVER CURRENT RELAY

Inverse time over current relay is used for over
current and short circuit protection of line or
transformer in power distribution systems. In the
Fig.1 shown the construction an electromagnetic
inverse time over current relay type induction disc
unit. All of the operating energy for the relay is
applied to the primary pole coil. The current flowing
in the primary coil will produce a primary flux
U
.
The primary flux will induce an emf in the secondary
coil. The emf in the secondary coil will cause a
current to flow through the winding of the lower
magnet. The secondary current lags behind the
secondary emf. This current creates a magnetic field
in the lower magnet
L
. Both
L
anf
U
will act on
the induction disc and cause it to rotate.
D
ISSN: 1411-1284

Proceeding 10
th

Fig 1. (a) construction of induction over current relay, (b)
induction disc and contacts (c) moving contact position.

The torque created by the magnetic field is
counteracted by the tension of a spiral spring. When
the turning torque overcomes the force of the spring,
the relay will operate. It is determines the minimum
operating current of relay.
The standard current-time characteristic curve of
inverse time over current relay is shown in Fig.2.
Where the extremely inverse relay saturates at four
multiples of pickup, very inverse at two multiples and
the moderatetly inverse at pickup (IEEE standard
inverse time, 1996:10)[3].

Fig.2 Time-Current curve of inverse time over current relay

The non-sinusoidal current or distorted current consist
of any numbers harmonic current component. Each
frequency is order multiple of the fundamental
frequency. If this current flow into the relays, than
each current harmonic component contributing
increasing frequency. Increasing the frequency of the
input current results in little change in the current that
is produce in the lag coil circuit. However, the flux in
this coil will decrease in inverse proportion to the
frequency increase. Similarly, the flux in the other
coil decreases because of the lowered magnetomotive
force across it. BY decreasing the magnetizing current
for frequency increment and constant lag coil circuit,
the effect is the main coil and the secondary coil
fluxes draw closer in phase. This slow down the disc
rotation, cause the pick up to increase and ultimately
causes the efficiency of the electromagnet to
deteriorate to the point of non-operation. Harmonics
current have serious effect on inverse time over
current relay pic up current value and operation time.

III. EXPERIMENTAL RESULTS

The experiment system configuration as illustrated
in Fig.3.

Fig.3 Experiment system configuration

Current of non-linear loads was isolated by a current
transformer (CT) that has a transformation rate 5A to
5A. All the harmonic amplitude levels and the total
harmonic distortion (THD) recorded by Fluke power
analyzer 42.
The operating time of the inverse time over current
relay is inversely proportional to the current.
Technical information of this relay is given in Table I
below:

Table I. Technical Information of Relay
Frequency 50 Hz
Accuracy of pick up current 5%
Resetting current as % of setting 92%
Contact stop time (after de-
energizing)
0.08s
Consumption at rated current 5VA
Over current adjustment ranges 1.00A-1.25A-1.50A-
2.00A-2.50A-3.00A-
4.00A

As non-linear load for generate harmonic currents is
used a single phase full wave rectifier included with
large variable capacitor C, see Fig 3. For generate
various values of total harmonic current distortion can
do with various values of capacitor C, and with varies
resistive load of rectifiers can be controlled magnitude
of rectifier nput current distorted or current flow
through relay.
The waveform of the non-linear load that contain
harmonic current is shown in Fig 4. This waveform
can be expressed in mathematic form as :

ISSN: 1411-1284

Proceeding 10
th

=
+ =
,.... 7 , 5 , 3
1 1
) ( sin 2 ) ( sin 2 ) (
h
h h S
t h I t I t i

(1)
where:
i
S
(t)

is distorted current non-sinusoidal
I
1
is current fundamental component
I
h
is current harmonic component orde h
h is harmonic orde ( h = 3, 5, 7, ..)
is angular frequency ( f 2 = )
f is line system frequency
1
is phase angle of current fundamental component
h
is phase angle of current harmonic component orde h

In the rms value form, the current in (1) can be
expressed as:

=
+ =
,.... 7 , 5 , 3
1
h
h S
I I I

(2)

And total harmonic distortion of current is :
1
~
2
2
I
I
THD
h
h
i
=
=
(3)
In this experiment, the rms value and THD of the
current (non-linear loads) are adjusted by changing
the value of capacitor C and resistor of rectifier. The
experiment result as effect of harmonic current to
inverse time over current relay is given in Table II. In
each mode is set for varies of the THD current values
for a constant value pick up current setting of relay.
The true rms current value I
s
, total harmonic distortion
value (THD
i
), the rms value fundamental current I
1

and operating time of relay were measured for each
mode. For all modes, the pick up current of the relay
was adjustable to 1 Ampere.

Table II. Pick up current values and THD
i
values of
relay
Mode THD
i
(%) I
1
(A) I
S
(A)
0 6.00 1.07 1.12
1 32.32 1.14 1.26
2 48.21 1.22 1.34
3 68.99 1.32 1.69
4 71.05 1.38 1.70
5 86.21 1.41 1.93

Table II. Operating time of relay for current is same 2 A
Mode THD
i
(%) I
S
(A) I
S
(A)
0 6.00 2.00 4.71
1 32.32 2.00 6.32
2 48.21 2.00 6.99
3 68.99 2.00 8.82
4 71.05 2.00 9.76
5 86.21 2.00 14.96

According to the experiment results, the pick up
current of inverse time over current relay was affected
by the presence of harmonic current (THD) and the
pick up value was increase by non-sinusoidal currents.
The increment depends on the total harmonic
distortion (THD) value of the non-linear load current.
With respect to the experimental results in Table III,
the pick up current increased as long as THD value of
current was increase. When THDi was above 86%,
the relay which was set to 1 A, operated at 1.9 A.
This shows that the relay can not perform a proper
protection function and causes damage or heating-up
depending on the rms value of the current and process
time in power system components such as
transmission lines, motors and transformers.
The has different operating time value for the
same relay current value in different modes as shown
in Table III. Since the relay induction disc rotates
rather slowly because of non-linear load current
harmonic components, rel;ay operating time will
increase as long as THD value of current increase.
Power system components that are protected by this
relay will be damaged because of the increase in the
operating time.

IV. CONCLUSIONS
Base on experiment results, the inverse time over
current relay induction disc type can not operate
effectively with non-sinusoidal currents that consist of
several harmonic components (THD). Relay pick up
current will increase as long as THD value of the
current increase as shown in Table II. As given in
Table III, the THD of current changes in spite of rms
value of relay. Non-sinusoidal current remains
constant and relay operating time increase as long as
THD value of current increase. Distortion increment
leads to increase in operating time of relay as shown
in Table III. Thus, power system components can
exposed to heat and finally can damaged.

Acknowledgment
The authors wish to thank University Sains Malaysia
School of Electric and Electronic Engineering
MOSTI that have providing the research grant (FRGS
research Grant No 203/PELEC/6071144) and others
facilities which have utilized in this research. The
authors would like to thanks all power group
members-LPKEE-USM.

ISSN: 1411-1284

Proceeding 10
th

REFERENCES

[1] Bishop M.T and Baranowski, J.F, Evaluating
harmonic-Induced Transformer Heating , IEEE
Transactions on Power Delivery, Vol. 10, No.1 305-
310. April 1996.
[2] Elham B.Makram and Regan B Hainnes, Effect of
Harmonic Distortion in Reactive Power Measurement,
IEEE Transaction Ind. Appl, Vol.28, No.4, July/August
1992.
[3] Power System Relaying Committee of IEEE, PES, IEEE
Std.C37112,IEEE Standard Inverse-time
Characteristic equation for over current relays, 1996
[4] Elmore. W.A. Kramer, CA and Zohholl. S.E, Effect of
waveform Distortion on Protective Relays, IEEE
Trans. On Inds. App, Vol 29, No.2, March/April, pp
404-411.

ISSN: 1411-1284

Proceeding 10
th

Effect of Harmonic Current Component to Active Power
Losses on Power Transformer

Syafrudin
1,2
and Dahaman
1

1.
School of Electric and Electronic Engineering Universiti Sains Malaysia
2
Electric Department of Engineering faculty University of Tanjungpura
email : syaf@eng.usm/my

Abstract Power Transformer on the electric power
distribution system used as changer of voltage levels and
also as connector between of the high voltage lines with
the low voltage lines or the consumer voltage
220V/380V. in the recent year, most of loads that
connected on the power distribution system have non-
linear characteristic (non-linear loads) and containing
high current harmonic component, In this paper, the
effect of current harmonic to active power losses on
power transformer is presented. In this case, assumed
that the transformer voltage is sinusoidal without
containing harmonic component and wave form current
transformer is distorted non-sinusoidal and containing
harmonic that generated by non-linear loads. In the
experimental, the transformer active power losses
determined from different value of input and output
power of transformer for linear load with sinusoidal
current waveform without harmonic and also for non-
linear loads with current waveform distorted with
harmonic content THD. This experiment results show
that the transformer active power losses will be increase
proportionally to percentage of the total harmonic
current distortion THD.

Keywords Transformer power losses, non-linear loads,
total harmonic current distortion THD,

I. INTRODUCTION

n the electric power distribution system, power
transformer functioned as a voltage changer from the
high voltage lines or middle voltage lines 20KV-11KV to
low voltage lines or consumer voltage 380V/220V, and also
functioned as loads connected. Generally, loads on the
modern electric power distribution system have the non-
linear current characteristic. It is caused by non-linear loads
such as electrics equipments in the static converter to
electric motor speed drive, computer, system illumination
and lighting use electronic ballast, AC and DC controlled
power supplies, battery charger, etc.

Those non-linier loads cause line current waveform
distorted non-sinusoidal and containing current harmonics
component. Several of the research result expressed that
more than 60% of the loads are connected on the power
distribution system is non-linear loads with total harmonic
current distortion THD content more than 76%, estimated
that this presented will be increase year to years [ 1].
Thereby the loads current that flow trough on power
transformer also distorted non-sinusoidal and containing
current harmonic component. The distortion current
containing harmonic component have high effective value
compared to the current without harmonic sinusoidal [ 2].
Considering the active power losses of the power
transformer consist of the copper power losses in primary
and secondary winding are function of the loads current and
core losses that function of frequency consist eddy current
lose and hysterisis lose, they losses is represented in
thermal form [ 3]. Therefore, the transformer active power
losse depend on total harmonic distortion of loads current
that flow on this transformer.
In this paper, A conducted an experiment to determining of
the effect current harmonic component to active power
losses at one power transformer. Transformer active power
losses are very importance to be known because it is used
for determining of a transformer operation capacities.
Implementation of the transformer active power losses is in
heat form, so that the high power lose can be causing the
transformer over heat and dammed.
Transformer active power losses determined with get
different values between active power input and output that
transformer. This experiment do for transformer with linear
load and also non-linear load. To show the percentage of
total current harmonic effect to active power losses of this
power transformer, so experiment do for any non-linear
load with same active power and varies current harmonic
content.

II. MODELING AND ANALYSIS

A. Harmonic Current Distortion
Generally, ac current waveform which distorted due to
non-linier load is in non-sinusoidal form and periodic wave[
4]. Base on Fourier series analysis, the non-sinusoidal
current waveform will be consisted of one current
fundamental component and any harmonic current
O
ISSN: 1411-1284

Proceeding 10
th

component that have multiple fundamental frequency. In
mathematic form can be expressed as :
=
+ =
,.... 7 , 5 , 3
1 1
) ( sin 2 ) ( sin 2 ) (
h
h h S
t h I t I t i
(1)
where:
i
S
(t)

is distorted current non-sinusoidal
I
1
is current fundamental component
I
h
is current harmonic component orde h
h is harmonic orde ( h = 3, 5, 7, ..)
is angular frequency ( f 2 = )
f is line system frequency
1
is phase angle of current fundamental component
h
is phase angle of current harmonic component orde h

In the rms value form, the current in (1) can be expressed
as:

=
+ =
,.... 7 , 5 , 3
1
h
h S
I I I
(2)
Percentage of current harmonic content or total harmonic
current distortion THD is defined as bellow :
1
... 9 , 7 , 5 , 3
2
I
I
THD
h
h
h
i
=
= (3)

B. Modeling of Transformer Equivalent Circuit
A power Transformer consist of primary winding,
secondary winding and steel core. Power transformer
equivalent circuit can be draw in figure Fig.1 bellow:

Fig 1. Circuit equivalent of a power transformer
where:
V
p
is primary voltage or input voltage
V
s
is secondary voltage or output voltage
I
p
is primary current or input current
I
s
is secondary current or output current (load current)
R
p
is primary winding resistance
R
s
is secondary winding resistance
X
p
is primary winding reactance
X
s
is secondary winding reactance
X
m
is magnetizing reactance
R
C
adalah resistansi rugi-rugi inti

If the load of transformer is non-linear loads, than the load
current distorted non-sinusoidal and content harmonic
current component expressed as:
=
+ =
,.... 7 , 5 , 3
1
h
hs s s
I I I
(4)
And active power losses of transformer on secondary
winding is :
s
h
h
hs s s s s
R I I R I P
2
... 9 , 7 , 5 , 3
1
2
+ = =

=
(5)
Primary current I
p
consist of secondary current and no-load
current of transformer I
o
, is :

o
h
hs s P
I I I I +
+ =

= ,.... 7 , 5 , 3
1
(6)
While the no-load active power losses consist of or active
power core lose, active power by eddy current and hysterisis
and magnetizing lose. All that losses depend on material of
core and frequency of system. Where the active power lose
of hysterisis is:
=
=
h
h
h hy
f B K P
,... 7 , 5 , 3 , 1
6 . 1
max 1
(7)
and active power lose of the eddy current is :
=
=
,... 7 , 5 , 3 , 1
2
2
max
h
h ed
f B K P (8)
where : K
1
and K
2
is material constants of transformer

Hence, the active power losses on primary side can be
expressed as :
p
h
h
hp p p p p
R I I R I P
2
... 9 , 7 , 5 , 3
1
2
+ = =

=
(9)
Thereby, from equation (5),(7),(8) and (9) hence the
total active power losses a power transformer can be
expressed as bellow:
ed hy S p
P P P P P + + + = (10)
This equation showing that the current harmonic
component is very strong influence to active power losses
value of a power transformer. In this paper, that active
power losses due to harmonic current component of the
power transformer is determined.

III. EXPERIMENTAL RESULTS AND DISCUSSION
A. Experiment
Experiment circuit for determining of active power
losses due to harmonic current on a power transformer is
shown in figure Fig 2. In this experiment do tested to single-
phase power transformer 2000VA, 220V/220V, 50 Hz with
it parameters are bellow :
R
p
= 6.8 , R
s
= 6.8 , L
p
= 0.56 mH , L
s
= 0.56 , X
m
=
1.87 mH.
In this experiment, single phase-transformer is supplied
by single-phase voltage source 120 V, 50 Hz. The fist step,
transformer is given pure resistive or zero THD current with
V
p
R
p
I
p X
p
R
C X
m
N
1
N
2
X
s
R
s
I
s
V
s
I
o
ISSN: 1411-1284

Proceeding 10
th

a certain power load P
s
. In order the load current flow
trough transformer is sinusoidal without harmonic content.
Then, the power input P
p
and output P
s
of that transformer
are recorded. Active power losses can get from different
values of that power input and power output. The next step,
load changed by non-liner load, it is a single phase full
wave rectifier with smoothing capacitors such as in figure
Fig.3. With the same active power of load, recorded of input
and output THD current and power P
p
and P
s
. Active power
losses of the transformer get from different value of P
p
and
P
s
. This experiment repeated again for several THD current
values. Overall of the experiment results in measurement
data are given in Table I, II, III and IV. For example, in
Figure Fig.4 and 5 also given output currents waveform or
load current with their current harmonic spectrum of this
power transformer. In the Fig.6 show a graph of THD
current content vs transformer active power losses.

Fig 2. Experiment circuit

L N
C

(a)
0
0.005 0.01 0.015 0.025 0.030 0.035 0.040 0.020
V
I

(b)
Gambar 3. (a) Penyearah satu fasa gelombang penuh sebagai
beban non-linier (b) bentuk gelombang arus dan tegangannya
input

Table I. Measurement Data for Linear loads
(Current THD=0%)
Input Output
V
p
I
p
P
p
THD
ip
V
s
I
s
P
s
THD
is

120
V
15.09
A
1.805
KW
0.40 % 116V 10.99
A
1.671
KW
0.40 %

Table II. Measurement Data for Linear loads
(Current THD=32.4%)
Input Output
V
p
I
p
P
p
THD
ip
V
s
I
s
P
s
THD
is

120
V
17.65
A
1.838
KW
31.2% 109V 16.87
A
1.671
KW
32.4 %

Table III. Measurement Data for Linear loads
(Current THD=42.1%)
Input Output
V
p
I
p
P
p
THD
ip
V
s
I
s
P
s
THD
is

120
V
20.31
A
1.853
KW
40.0 % 92V 13.25
A
167
KW
42.1 %

Table IV. Measurement Data for Linear loads
(Current THD=68.0%)
Input Output
V
p
I
p
P
p
THD
ip
V
s
I
s
P
s
THD
is

120
V
24.56
A
1.886
KW
64.30% 86V 15.82
A
167
KW
68.0%

Table V. Measurement Data for Linear loads
(Current THD=81.0%)
Input Output
V
p
I
p
P
p
THD
ip
V
s
I
s
P
s
THD
is

120
V
28.39
A
1.907
KW
76.4% 79V 16.71
A
1.67
KW
81.0%

(a)

Power Analyzer 1
Power Analyzer 2
Trafo
Beban
Non-Linier
Sumber
tegangan
Is
Vs
ISSN: 1411-1284

Proceeding 10
th

(b)
Fig 4. Output current waveform without harmonic THD=0% and
their spectrum harmonic.

(a)

(b)
Fig 5. Output current waveform without harmonic THD=0% and
their spectrum harmonic.

THD vs Active Power Losses
0
0.05
0.1
0.15
0.2
0.25
0 20 40 60 80 100
%THD Current
A
c
t
i
v
e

P
o
w
e
r

L
o
s
s
e
s

[
k
W
]

Fig 6. Graphic of THD load current vs active power losses
transformer

B. Discussion
Total active power losses of power transformer can be
found from different values of active power input Pp and
output Ps of that transformer. From the Table I,
measurement data to linear load where current flow through
transformer without harmonic {THD=0%), power input of
transformer is P
p
= 1.320 kW and power output is P
s
=
1,099 kW. Hence, active power losses of transformer can
calculated as P = P
p -
P
s
= 0.221 kW. In the same way,
base on data in Table II the active power losses for THD
current of 32.4% is P = 0.134 kW and then from data in
Table III, IV and V for THD current 42.1%, 68.0% and
81.0% respectively, the active power losses of transformer
is with 0.167 kW, 0.182 kW, 0.215 kW and 0.236 kW
respectively.
Normally, power losses a transformer not more than 10% of
total transformer rating. In this case, transformer active
power losses have achieving more than that. Therefore,
increasing THD current of load (non-linear load) can cause
the transformer over heating in the power distribution
system even though the transformer still under rating
operation.

IV. CONCLUSIONS
In this paper have been conducted investigate to a power
transformer to determining of harmonic current effect to
increase the active power losses.
The transformer active power losses can be determine with
different values of input and output active power.
In this experiment, the active power losses increase of the
power transformer is proportional to total harmonic current
distortion THD of transformer loads.
Finally, in order to operate the power transformer for
several operating under THD percentage of non-linear
loads, it is has proper design in terms of harmonic standard,
proper design in terms of loading capacity heating and
insulation lifetime.
ISSN: 1411-1284

Proceeding 10
th

Acknowledgment
The authors wish to thank University Sains Malaysia
School of Electric and Electronic Engineering MOSTI
that have providing the research grant (FRGS research
Grant No 203/PELEC/6071144) and others facilities which
have utilized in this research. The authors would like to
thank all power group members-LPKEE-USM.

REFERENCES

[1] M.Salih TACI and M.Hadi Sarul, The effect of The Harmonic
Components Upon Transformer Active Losses In case of Non-
Sinusoidal Source, IEEE, New York, 2000.
[2] Taci,M.S , Determining The harmonic Effects of Non-Linear
Loads on Parallel Connected Transfoemer in Terms Of Power
Factor, IEEE Power Quality98, Hydrabad-India.
[3] Bishop M.T and Baranowski, J.F, Evaluating harmonic-
Induced Transformer Heating , IEEE Transactions on Power
Delivery, Vol. 10, No.1 305-310. April 1996.
[4] Elham B.Makram and Regan B Hainnes, Effect of Harmonic
Distortion in Reactive Power Measurement , IEEE
Transaction Ind. Appl, Vol.28, No.4, July/August 1992.
ISSN: 1411-1284
Proceeding 10
th
OPTIMIZE PLACING OF PASSIVE FILTER AND
REDUCE HARMONIC IN POWER SYSTEM
Case Study at PT. ISPAT INDO

Iwa Garniwa MK, Budi Sudiarto; Adrianto

Electrical Engineering Department, University of Indonesia
Depok 16424, Indonesia

Abstract-One of the solutions to reduce harmonic
distortion is using passive filter. The most common
type of passive filter is the single-tuned notch filter.
This is the most economical type and is frequently
sufficient for the application. This paper contents
optimizing of placing a passive filter in a power
system so it can be obtain the maximum and most
efficient filterazation to reduce losses in medium
voltage (MV). How to reduce harmonic and improve
power factor using single tuned notch filter is also
studied in this paper. The object of this research is an
power system which has 3 panel of low voltage (LV)
and the each of them has a different non linear loads.
First panel, TR 20, has a maximum THDi (Total
Harmonic Distortion) equal to 7.28 % and maximum
THDv equal to 2.46 % (both in phase 3). Second
panel, TR 21, has a maximum THDi equal to 62.41
% (phase 1) and has maximum THDv equal to 9.5 %
(phase 3). Third panel, TR 23, has a maximum THDi
equal to 58.37 % (phase 1) and maximum THDv
equal to 8.26 % (phase 3). The MV from the
simulation has THDi equal to 34.14 % and THDv
equal to 1.57 %.

I. Introduction
Current from the harmonic load causes extra
heat, isolation failure, operation failure, etc. One of
the solutions to solve this problem is using passive
filter at the source of the harmonic load. Using this
passive filter, harmonic distortion expected to
decrease until the limit of tolerance so that the power
system can work properly.
Research methodology that is been used in
this paper starts with literature study of design
passive filter so it can be implemented to non linear
load. The result of the design filter will be simulated
using ETAP Power Station 4.0.0. Output of the
simulation will be analyzed and then we can make
the best advice to the company which the data is
being used. The company as the object research is
PT.Ispat Indo. PT. Ispat Indo is the biggest steel
company in East Java with the output production
billet and wire rod as the final product. The data were
taken using power and harmonic analyzer equipment.

II. Harmonic and Filter
Harmonic distortion is caused by nonlinear
devices in the power system. A nonlinear devices is
one in which the current is not proportional to the
applied voltage.

Fig 1. Distortion current cause by non linear loads

The sum of sinusoidal waves which create
non sinusoidal wave can be solving using Fourier
analyses.
) 2 sin( 2 ) (
1
0 n
n
n
n
ft n Y Y t Y + =

=
=
(1)
Where
Y
o
= Amplitude of DC current component
Y
n
= RMS value from harmonic component
F = fundamental frequency
n= phase angle from harmonic component
ISSN: 1411-1284
Proceeding 10
th

Fig 2. Fourier series representation of a distorted
waveform

Total Harmonic Distortion (THD)
expressing distortion level generated by all harmonic
components and defined as follow:
1
1
2
max
M
M
THD
h
n
n
>
= (2)
Where:
THD = Total Harmonic Distortion
M
n
= RMS value of current or voltage n orde
M
1
= RMS value of current or voltage at
fundamental frequency

There are two kinds of filter, active filter
and passive filter. Passive filter is common use in
industry. The weakness using passive filter are
depends on impedance, system frequency, and
component tolerance. Two purposes using filter are:
1. To reduce harmonic in AC network.
2. To provide reactive power for the harmonic
load or other loads.
There are two kind of passive filter, series
filter and shunt filter. Shunt filters that is commonly
use are single tuned filter and order two damped
filter. These two filters are the simplest way of design
and the cheapest to implementation.
Passive filter contents inductance,
capacitance which combines to control harmonic.
Filter passive are commonly use and not expensive.
Using passive filter is possible to interact with the
power system which can disturb it. It is important to
check all of the possibility interaction when filter is
being designed.
The shunt filter works by short-circuiting
harmonic currents as close to the source of the
distortion. This keeps the currents out of the power
system. This filters connected to the bus bar where
reduce of harmonic voltage determined and create a
filter bank.

Fig 3. Passive filter in a power system

There are two parameters need to be
considerate when determine value of R, L, and C:
a. Quality factor (Q)
b. Deviation relative frequency ()

III. Measurement and Data
Program we use to simulate the power
system is ETAP 4.0.0.

Fig 4. Power system design using
ETAP 4.0.0

The data has been recorded using 9625
Power Measurement and transferred to the personal
computer using 9625 Power Measurement Support
Software. The data of MV we have it from the
simulation software (ETAP 4.0.0).
Table 1. Data from the panels when
the THDi maximum
TR 20 TR 21 TR 23 TM
THDi 7.28% 62.41% 58.37% 34.14%
Current 1344 313.3 A 372.5 A 92 A
THDv 2.37% 3.70% 3.42% 1.57%
ISSN: 1411-1284
Proceeding 10
th
Power
Factor
0.68 0.68 0.7 0.66

(a)

(b)
Fig 4. (a) Current distortion wave in TR 20, 21, and
23
(b) Current distortion spectrum in TR 20, 21, and 23

Using IEC 61000-3-6 and 61000-3-4
standard, the limit of THDv is 8 %. From the data,
we can determine the panels which need a passive
filter are TR 21, TR 23, and MV. These panels have
THDv and distortion current over the limit.

IV. Analyze and Solution
By using the equation below, we can
determine the losses in power system:
R I P
losses
2
= (3)
Where
I = total current
R = Impedance
P = power loss
If the impedance of the system constant, we can
decrease the losses by decreasing the current.
Single tuned notch filter is use to reduce
harmonic. The basis parameters of design filter as
follow:
Table 3. TR 21 Filter harmonic data
THDi Maximum (%) 76 Phasa 1
THDv with highest THDi
(%) 4.46
THDv with highest
harmonic order (%) 3.3 Order 5
PFCC kVAR ON 166

Table 4. TR 23 Filter harmonic data
THDv with highest THDi
(%) 4.75
THDv with highest
harmonic order(%) 3.56 Order 5
PFCC kVAR ON 189

Table 5. TM Filter harmonic data
THDv with highestTHDi
(%) 1.47
THDv with highest
harmonic order (%) 1.23 Order 5
PFCC kVAR ON 878

The steps to design a filter based on the
table above as follow:
1. Election of frequency filter.
From the measurement, the highest
harmonic level happened at order 5. Tuning of
order degraded by a little below order in fact so
that n = 4.8. This degradation is conducted as
tolerance to filter component and also variation
of system impedance.
2. Enumeration of capacitor size measure and
resonance frequency.
The formula of current at the fundamental
frequency for capacitor bank as follow:
actual
actual
FLcap
kV
kVAR
I
3
= (4)
ISSN: 1411-1284
Proceeding 10
th
Formula to look for capacitive impedance as
follow:
rated
rated
C
M
kV
X
var
2
= (5)
For TR 20, nominal voltage phase to phase is
430 V. For TR 21 and 23 is 730 V and for MV
is 11 kV.
Inductor impedance determined with
formula following:
2
n
X
X
C
R
= (6)
Because there is reactor, the fundamental
current increase which can expressed by:
( )
R C
bus
FLfilter
X X
V
I
=
3
(7)
Because the current increased, the reactive
power compensation increase and we can
determine using:
FLfilter bus plied
I V kVAR * * 3
sup
= (8)

Table 6. Filter specification
THDi Maximum Filter
Spesification
TR 21 TR 23 TM
System
frequency 50 50 50
Rating of
Capasitor Bank 166 189 878
Rating of Bank
Capacitor
Current 131.25 150.45 46.11
Tuning
harmonic filter 4.8 4.8 4.8
Tuning system
frequency 240 240 240
Capacitor
Impedance 3.211 2.801 137.742
Reactor
Impedansi 0.139 0.122 5.978
Filter Full Load
Current 137.19 157.3 48.2

There are 4 variations when placing the
filter. The results of simulation using ETAP 4.0.0 in
TM as follow:
Table 7. Filter at TR 21

Before After

Difference
THDi (%) 34.14 28 6.14
THDv (%) 1.57 1.26 0.31
Current
(A) 92 84 8
Power
factor 0.66 0.71 -0.05

Table 8. Filter at TR 23
Before After

Differense
THDi (%) 34.14 27 7.14
THDv (%) 1.57 1.21 0.36
Current
(A) 92 83 9
Power
factor 0.66 0.72 -0.06

Table 9. Filter at TR 21 and TR 23
Before After

Difference
THDi (%) 34.14 22 12.14
THDv (%) 1.57 0.98 0.59
Current
(A) 92 76 16
Faktor
Daya 0.66 0.77 -0.11

Table 10. Filter at TM
Before After

Difference
THDi (%) 34.14 27 7.14
THDv (%) 1.57 0.98 0.59
Current
(A) 92 62 30
Power
factor 0.66 0.97 -0.31

ISSN: 1411-1284
Proceeding 10
th
From the result of simulation above, we can
determine that the biggest difference in decreasing
current is when the filter located at TM. The current
was decreased 30 A to the final value 62 A. This
happen because harmonic distortion was reduced by
the passive filter which the filter designed based on
the entire load (TR 20, TR 21, and TR 23). This
makes the filter design more precision. Reactive
power that is supplied by the capacitor bank also
based on the entire load so its enough to increase the
power factor value to 0.97 which can avoid the
penalty. This also helps make the losses in TM
decrease. The THDv was decreased to 0.98 % from
the original value 1.57 %. The THDi also decreased
to 27 % from the original value 31.14 %. These data
shows us that passive filter can reduce harmonic,
increase power factor, and reduce the power loss.
V. Conclusion
1. Single tuned filter can be use to reduce
harmonic distortion, increase power factor, and
reduce power loss.
2. Optimize placing of passive filter can make
passive filter more efficient and effective.
3. Placing the passive filter at the MV in the object
of this research is the best place to minimize the
power loss.
4. The results using filter at MV are:
THDi = 27 %
THDv = 0.98 %
Current = 62 A
Power factor = 0.97

VI. Reference
1. Roger C. Dugan..[et al]. 2002. Electric Power
System Quality. New York : McGraw-Hill
2. Cahyadi, Rifky. 2003. Upaya Menghemat
Energi dengan Mereduksi Harmonisa. Depok :
Skripsi Departemen Teknik Elektro Universitas
Indonesia
3. Setiabudy, Rudy. 2006. The Design Of Passive
Filter To Overcome Harmonic Distortion.
Depok : Proceedings. IJJSS 2006
4. Zunaedi, Totok. 1999. Eliminasi Harmonik Dan
Kompensasi Daya Reaktif di PT. Engenys Steel.
Depok : Skripsi Departemen Teknik Elektro
Universitas Indonesia.

ISSN: 1411-1284

Proceeding 10
th

CONVERGENCE APPROXIMATION OF ABR FORMULATION
FOR JOSEPHSONS TUNNELING IN UO
2
CHAIN REACTION
AT 45.7 MW CANDU NUCLEAR REACTOR

Moh. Hardiyanto
Department of Educational Physics FTMIPA - UNINDRA
Large Hadron Collider Laboratory, CERN Lyon, France
E-mail : moh_hardiyanto_iti@yahoo.com

Abstract - The convergence of free covariant
equation in Einsteins space with quantum condition
is studied using the ABR (Abrikosov-Balseiro-
Russell) formulation in convergence approximation
for Josephson tunneling is important role for
determine of neutrino particle existing, especially
after Cerenkovs effect for 45.7 megawatts nuclear
reactor based on UO
2
chain reaction. This
approaching will be solved the problem for determine
the value of interstellar Electrical Conductivity (EC)
on UO
2
chain reaction, then the post condition of
muon has been known exactly. In this research shown
the value of EC is 4.32 eV at 378 tesla magnetic field
for 2.1 x 10
4
currie/mm fast thermal neutron floating
in 45.7 megawatts adjusted power of Canadian
Deuterium Uranium (CANDU) nuclear reactor. The
pictures had resulted by special Electron-Scanning-
Nuclear-Absorbtion (ESNA) shown any possibilities
of Josephsons tunneling must be boundary by muon
particles without neutrino particle existing for 350
456 tesla magnetic field on UO
2
more enrichment
nuclear fuel at CANDU, whereas this research has
purpose for provide the mathematical formulation to
boundary of muons moving at nuclear research
reactor to a high degree of accuracy and with Catch-
Nuc, one of nuclear beam equipment has a few
important value of experimental effort.
Keywords - ABR formulation, EC value, UO
2

chain reaction, magnetic field value

I. INTRODUCTION
Its is known that for muon-hadron
scattering, the close-coupling equations [1] have
been used extensively in the sophisticated
computation in Abrikosov-Balseiro-Russell (ABR)
formulation [2]. In the close-coupling equations,
the complete wave functions are expanded in target
states on Einsteins space. In this paper the target
states are constructured in the finite L
2
basis space
following the results of Abellian system and next
formulated by ABR without Diracs condition. The
previous studied the convergence of the
approximation target states for discrete and
continuum cases in quantum condition for UO
2

chain reaction at 45.7 megawatts (MW) adjusted
power in Canadian Deuterium Uranium (CANDU)
nuclear research reactor. The advanced studied its
application to the ABR equations which results in
pseudo state close-coupling approximations for
bound-free transition in convergence structure for
Josephsons tunneling. For first step using by
Einsteins space for eliminary construction of ABR
formulation in UO
2
chain reaction.
In this paper, the Abellian system was
conducted for covariant equation to Einsteins
space and ignores the Diracs condition for 300
410 tesla magnetic fields on fast thermal neutron,
to study the convergence behavior of the free
covariant equation part. Its possible analytical
computation is also addressed after finding the
asymptotic behavior.

II. Pseudostate Close-Coupling Approximations
The close-coupling equations in ABR
formulation require to study is given by
[ ]

=

= +
1
0
k
k jk j j
f V f E H (1)
where
2
0
2
1
= H (2)
the total energy of the system connecting the initial
and final state f i satisfies
2 2
1
2
1
2
1
f f i
k k E + = + = (3)
j
f are channel functions related to the
complete wave function as
) ( ) ( ) , (
2
1
1 2 1
r f r r r
j
j
j
= (4)
in which the + (-) superscript refers to singlet
(triplet) scattering and ) (
1
r
j
are muon eigen
functions.
The channel potential
jk
V is given by

+ =
jk jk jk
W U V (5)
where
ISSN: 1411-1284

Proceeding 10
th

+
> =

2 1
2 1
1
2
2 1 1
2 1 2 1
) (
1
) (
1 1
) (
r r
r r
r
r
r r r
r r dr dr U
k
k
j jk
(6)

and
i i j j
i
i jk jk
E W ) ( + =

(7)
with
1 =
i
1
4
=
i
(8)
If replace the target states by a set of pseudo states
(as consequences of finite basis set) in order to
solve the equations approximately. We has to
modify the close-coupling equations for Einsteins
space through for normal condition after
Cerenkovs effect at 315 tesla magnetic field since
the exchange term in the form (7) is valid only for
exact target states. The form appropriate to pseudo
states is given by
Nj Nk Nj Nk jk
r H r H W >< + >< =
) ( ) (
2 1
< > +
N
i
Ni i Ni jk
E
1
) ( (9)
where
i N
label the pseudo states and H(r
1
) is the
target non-linier for Hamiltonian operator in
Abellian.
In calculating the pseudo state close-
coupling approximation one must consider the
error present in the quadrature rule approximation
and the close-coupling potentials when pseudo
states are employed. The first error will not be
discussed here. The second error will be discussed
in association with the direct-potential component
of the channel potentials since they are regarded is
being responsible for the dominant scattering
process especially at higher energy.

III. Convergence Approximation of ABR
Formulation
In the momentum re-presentation, the
derivative of ABR formulation for muon state in
UO
2
chain reaction is given by
V G V V T
B 0 2
+ = (10)
Here G
0
is the diagonal matrix of free channel
Greens functions. This second ABR formulation
approximation has been used by Einsteins space in
Abellian system, especially the fast thermal neutron
floating at 350 tesla magnetic field to test the
suitability of pseudo states expansions with initial
and first states chosen either the ground or 2s, 2p
excited states.
The actual Josephson tunneling element
from (10) require to study is
[ ]
+
=
dr r e r
k k V
j
r i
i ij
j i ij
) ( ) (
2 ) , (
2

(11)
where
j i
k k = . The indices i and j can be
either discrete or continuous after Cerenkovs
effect on 2.1 x 10
4
currie/mm and the range of 315
390 tesla magnetic field. For free covariant
equation in quantum condition for UO
2
enrichment
reaction potentials one uses

= dr r e r I
p
r i
q pq
) ( ) ( 2 ) (
. 2
(12)
where
p
and
q
are initial and final continuum
states having the momenta p and q respectively,
given by
=
|
|
\
|
+
=
0
.
) ( ) 1 2 (
2
1
) (
q
N
q
i
q
q
qr
r q
P r U e i
qr
r
q q
q q
l
l l
l l
l

(13)
and normalized to a function in
3
) 2 /( q . The
approximate target wave function
N
q
p
U is given by
q q
q q
n n
q
N
n
N
q
x p
n
n
q B U
l l
l l
l
) (
2 2 (
) 1 (
) (
1
1
0
+
=
+ +
+
>=
(14)
where

) 2 / (
2 / ) 1 (
2
) 1 (
) 1 ( 2 ) (
+ =
e x
i q B
q
q
q
p
q
p
l
l
(15)
and
..... 0 .
) ( ) ( ) (
1 2
2 /
1
=
=
+
+
q
n
r
n
n
r L e r r
q q q
l
l l l

(16)
solving the integral (15) after quite lengthy
derivation one finally obtains
) (
. .
, .
1
2
3
) 1 (
) ( 32 ) (
q
p
q p q
q p
q p
i
L m
m m M
L
pq
c i
pq V
l
l
l l
l l

+ +

) (
0 00
) 1 2 ( 4
) 1 2 )( 1 2 (
2 / 1

+
+ +
L
pq q p q p
q p
q p
I LM m m
L
L
x
l l
l l
l l
(17)

where
ISSN: 1411-1284

Proceeding 10
th

) ( ) ( ) (
) (
) 2 2 ( ) 2 2 (
) (
1
1 , 1
0 .
2
q m p n
M N
m n
q
q p
L
pq
x P x P q B
p B I
p p
p
q p
l l
l
l l
l
l l
+

=
+ +
+ +
=

)
`
+
+
) ( ) 2 ( Re
1 1
i i i x
v L
) . : 2 2 . 2 2 :
2 ( ) 2 (
2
+ +
+ +
+ + + +
i i
m n v F v x
q p
q p q p
l l
l l l l
(18)

IV. METHODOLOGY
This research using by high degree
accuracy mathematical approaching and a few
experimental efforts with special Electron-
Scanning-Nuclear-Absorbtion (ESNA) and Catch-
Nuc, for see the results of the mathematical
approximation, in case is ABR formulation in
Josephsons tunneling for UO
2
chain reaction at
350 456 tesla magnetic field and determine of
Electrical Conductivity (EC) value in 2.1 x 10
4
fast
thermal neutron floating.
A few mathematical equations in quantum
condition will be explain in this section.

ABR Formulation non Abellian
The metric tensor is defined through the
line element ds as follows:

dx dx g ds =
2
(19)
Hence, different metric shall lead to different
properties of space-time. For a space time where
the metric is defined through
( ) [ ]
|
|
\
|
+ +
=
) sin (
1
2 2 2 2
2
2
2 2 2
d d r
kr
dr
t R dt ds
(20)
the space satisfies the properties that it is
homogeneous and isotropic. Such properties agree
with the nuclear structure in microscopic principles
and are also supported by experimental data.
Accordingly the above metric, called the ABR
Formulation without Abellian system, becomes a
standard model of UO
2
structure. In the above
expression R(t) denote the scale factor and k is a
constant. The universe is closed if k > 0, open if k <
0 and flat if k = 0 and R(t) sometimes are rescaled
in such a way that the value of k assume one of the
three values -1, 0, or +1. The Cartesian form of the
above line element.
+ +
+
+
+
=
2
2
2
2
2
2
2
2
2
2
2 2
1
1
1
1
1
1
) (
kr
kz
dz
kr
ky
dy
kr
kx
dx
t R
dt ds
(21)
|
|
\
|
x
g
x
g
x
g
g
2
1
(22)

Combining (19) and (21), one finds the explicit
forms of non-zero components of the Riemannian
curvature tensors as:
. 00
) (
) (
ab
a
b
t R
t R
R =
+ =
2
0
0
1
) ( ) (
kr
x kx
t R t R R
b a
ab b a
(23)

|
\
|
+ =
A A dA A Tr
k
S
M
3
2
4
(24)
) 2 ( ) 1 (
0 0
' ) , 2 , 1 ( '

E E K

=
(25)
( ) =

3
1
2
24
1
M
wZ
dg g Tr S
(26)

=
1
0
) ( ) ( da ax F ax x A

(27)
=
1
0
) ( ) ( x R d x x

(28)

0
1 '
) (
2
1
) (
'
) 2 (
^
E q g
n
s
Q Q
s
s
l
l
=

|
|
\
|
=

(29)

V. RESULT AND DISCUSSION
It is shown in the final equation that beside the
double series n and m which are finite, the ABR
function F
2
is formally defined by a double series
expansion with radius of convergence severely
restricted. Because the occurrence of n and m
which are finite in the arguments, there is no
problems arise about its convergence. The
remaining is to compute (22) numerically in order
to see how rapid the convergence is since the
expression of Riemannian curvature tensor is
similar with the expression of
F in the non
Abellian gauge. Such a formula (18) may also be
obtained in the general theory of relativity, the fast
ISSN: 1411-1284

Proceeding 10
th

thermal neutron break into UO
2
matrix is show in
Fig. 1.

Fig.1. The floating of muon moving
in UO
2
matrix at 378 tesla
magnetic field
(Courtesy of CANDU nuclear reactor, Canada)

To obtain this formula, let us impose the ABR
gauge condition on the Christoffel symbol as
follows:
0 ) ( = x x
a
(30)
If we make the transformation of

x x
where : [ ] 1 , 0 is a parameter could obtain:
) (
) ( ) (
x
x x x R x
a
+
=
(31)
and if use the relation:
). (
) (
) (
) ( ) (
) (
x F x
x
x F
x
x
x F
d
x d
x F
d
d
=
(32)
)) ( ( ) ( x
d
d
x R x
= (33)

Using the above formula, we shall derive
the Christoffel symbol and contrast them with the
Christoffel symbol previously derive from the ABR
formulation on Abellian system and the re-
polarization of UO
2
matrix will give in Fig. 2.

in UO
2
matrix at 378 tesla magnetic field and fast
thermal neutron at 2.1 x 10
4
currie/mm floating

We have seen that the ABR gauge which
satisfy the Josephsons tunneling effect can be
deduced from (23). We shall refer the symbol of
FS
for the ABR gauge in Christoffel symbol

to distinguish with that derived from the
Robertson-Walker Riemannian tensor, the fast
4
currie/mm floating will
shown up at Fig. 3, especially without the neutrino
particle.
+
+
+
+
+
+
+
= + =
+
+
)! 4 (
) 2 ( ) 1 (
)! 3 (
) 1 ( 2 ) 1 (
)! 2 (
) 1 (
) ( 0 ) , ( ) , (
2
1
) (
n
n
n
n
n
r F kn t r B t r E
n
n n
n
(34)

in UO
2
matrix at 378 tesla magnetic field and fast
4
currie/mm floating without
neutrino particle

Based on Fig. 3, if integrate the equation by parts,
we will obtain the series form of the equation as
=
+
=
1
0
0
)! 2 (
) 1 (
) (
n
n
n n
x
d
d
n
x R x d

(35)
Let us now choose a scale factor as:
R(t) = e
mt
. This is an asymptotic value of the
Friedmann model of UO
2
nuclear structure with
3 / = m , where is the UO
2
constant.
Using by special ESNA, we get the
Josephsons tunneling with Abellian operator such
as the Fig. 4.

ISSN: 1411-1284

Proceeding 10
th

Fig.4. The Josephsons tunneling
in UO
2
matrix at 378 tesla magnetic field

The singularity for the first turns out to be worse
compared to the later since the standard one has
only a singularity factor from EC value at
interstellar of muons moving for 2.1 x 10
4
fast
thermal neutron floating at range 350 456 tesla
magnetic field without neutrino particle existing
after Cerenkovs of ( ) V e kr 32 . 4 1 / 1
2
= .
At least on the pedagogical point of view one then
becomes aware that the ABR formulation in UO
2

chain reaction for a given Riemannian curvature
tensor are not unique.

VI. CONCLUSIONS

Investigations and research using by
ABR formulation in convergence approximation
and ESNA also Catch-Nuc equipments based on
UO
2
chain reaction by Josephson tunneling at 45.7
MW CANDU nuclear research reactor has a few
result, expressed:
a. The strength of fast thermal neutron
floating is 2.1 x 10
4
currie/mm.
b. The values of Electrical Conductivity
(EC) for interstellar muons moving is
4.32 eV at 378 tesla magnetic field.
c. These equations have a high degree
accuracy, so they could be determine of
muons moving after Cerenkovs effect
for 45.7 MW adjusted power.
ACKNOWLEDGMENT
The author wish to thank Canadian
Deuterium Uranium (CANDU) Nuclear Reactor
team for contributions to this work.

REFERENCES
[1] Bransden,B.H.Scott,T.,Shingal,R.
and Roychoudhury,R.K., Journal
Phys.B 15, 2003
[2] Bray I. and Stelbovics.A.T., Phys.Rev. A. 46,
2002
[3] Burke. P.G and Seaton M.J.,
Math. Journal of Phys.10 , 2002
[4] CANDU nuclear reactor, UO
2
Interference
Structure, Canada, 2003
[5] Flinder, and Stelbovics.A.T.,Test of Close-
Coupling in the poet Model, Procedings of the
Sixth International Symposium on Correlations
and polarization in Electronic and Atomic
Collisions and (e,2e) Reactions, Univ. of
California, USA (2001).
[6] Helna, Duprix, Structure of ABR
formulation in Diracs condition, Universit
du Qubec, Canada, 2003
[7] Madison,D.H.and Callaway.J.J.,
Journal Phys B 20, 2002
[8] Stelbovics,A.T., and Abrikosov G.,
American J.Phys.43, 2003

About Author:
Lecturer of Department of Educational Physics
and Researcher at Large Hadron Collider
(LHC) Laboratory Lyon - France, and now
take a doctoral program at Universite du
Quebec, Canada.

ISSN: 1411-1284

Proceeding 10
th

Energy Conservation on Steel Industry
by Increasing Efficiency and Electric Power Quality

Iwa Garniwa; Budi Sudiarto; Fauzan Hanif Jufri

Faculty of Engineering, University of Indonesia
e-mail : iwa@ee.ui.ac.id ; budi@ee.ui.ac.id ; fauzantolenk@yahoo.com

Abstract Electrical energy conservation is
electric energy usage with high efficiency by
minimize losses at all production process, start
from generation, transmission, distribution, and
consumption.
Industry sectors included steel industries are
large energy consumers because industry sectors
use about 30% primary energy. This mean energy
conservation on industry sector gives a significant
role for thrift the energy.
This paper did some research for energy
conservation opportunity on demand side focused
in electrical energy conservation, especially in
compact energy industries, that is steel industries.
This research takes the some opportunity by
increasing efficiency on processes, power systems,
and electric power quality on electrical systems.
Energy consumption efficiency for every process
can be explained from Specific Energy
Consumption (SEC), that is a value meant ratio
between total energy consumption to total product
yield.
The recommendations yield from this paper
will be followed by feasibility study from technical
sides and economic sides, using Life-Cycle Costing
Analysis. The opportunities can be high cost or
investment, or low cost or no cost.
The data on this paper get from survey to one
of steel industries, that is PT SERMANI STEEL,
which produce galvanized iron sheet, and PT
SURABAYA WIRE which produce nails and wire
rod.

Keywords Energy Conservation, steel industry,
electric power quality, energy efficiency

I. INTRODUCTION

ndustry sectors are large energy consumers because
they use about 30% primary energy. This mean
energy conservation on industry sectors give a
significant role for thrift the energy. Besides that,
industry sector consist of large consumers which a
small number, so energy conservation can be easier
than do at other consumers. Energy sectors can be
classified based on its energy consumption, those are
compact energy, for example steel industries, cement
industries, and melting factory.
Processes on steel industries use much energy.
Those are used for electric motor, convert electrical
energy to mechanical energy, and for arc furnace,
convert electrical energy to heat energy. Electrical
energy usage can be reduced by increase or modify
another factors used in processes. As an example,
increasing power factor, efficient loading on motors
and transformers, or increasing the fluid pressure on
furnace, etc.

II. BASIC THEORY [1]

Energy intensity in producing products defined as
Specific Energy Consumption (SEC), it is ratio
between energy consumed to yield products to
amount of products yield. The unit is kWh/Ton. By
conserve energy, we hope that the SEC will decrease,
which mean decrease the energy consumption without
decrease the product yield.

2.1 Improving power factor
The reactive power has to be supplied by the
utility even though it is not actually registered by the
power meter (as real power used). The magnitude of
this reactive power increases as the power factor
decreases. To account for the loss of energy due to the
reactive power, most utilities have established rate
structures that penalize any user that has low power
factor. Therefore, significant savings in the utility
costs can be achieved by improving the power factor.
This improvement can be obtained by adding a set of
capacitors to the entire electrical system. See figure 1,

Figure 1. Improving power factor by using capacitor

The size of these capacitors (Q compensator) is
typically measured in KVAR (the same unit as the
reactive power) and can be determined, as shown in
figure 1, using the power triangle analysis :
I
ISSN: 1411-1284

Proceeding 10
th

tan tan
1 1
tan cos tan cos
P Q Q P
C ex comp ex comp
P P PF PF
C ex comp

| |
= = |
\
| |

= |

\
(1)
Where P stands for the real power (measured in kW),
while Qe and Qc stand for the reactive power before
and after the retrofit, respectively.
The thrift yield from improving power factor can
be defined as [2] :
2
% 1 100%
ex
reduction
comp
PF
Loss
PF

| |

= |
|

\

(2)

2.2 Reducing harmonic distortion
Harmonic distortion (based on IEC 702-07-43) is
waveform distortion which frequency multiplied from
its fundamental frequency in integer number.

Figure 2. Harmonic distortion waveform

To quantify the level of distortion for voltage and
current a dimensionless number, referred to as the
total harmonic distortion (THD), is determined
through a Fourier series analysis of the voltage and
current waveforms and is respectively defined as:
2 2
1
1
1
h
R M S M M T H D
h
h

= =
>
(3)
Harmonic distortion will increase the RMS value
both current and voltage, so that will increase power
loss in the systems. Besides that, harmonic distortion
will reduce equipment lifetime, three phase neutral
current overload, etc.

2.3 Replacement motor with energy-efficient motor
Induction motor commonly used in industrial
sectors has an inductive characteristic, produced by
its magnetic field to rotate the rotor.
One parameter is typically important to identify an
electric motor during full-load operation, the
conversion efficiency of the motor (). This efficiency
expresses the mechanical power as a fraction of the
real electric power consumed by the motor.

mechanical
electrical
P
P
= (4)

Due to various losses (friction, core losses due to
the alternating of the magnetic field and resistive
losses through the windings), the motor efficiency has
typical values ranging from 75% to 95%, depending
on the size of the motor. In the above definition, PM
stands for the mechanical power output of the motor,
expressed in kW or horsepower (HP), which is the
most important factor in selecting a motor.

Table 1. Typical motor efficiencies (Hoshide, 1994)

The improved efficiency of the high/premium-
motors is mainly due to better design with use of
better materials to reduce losses, which however
comes with a higher price (about 10 to 30% more than
standard-efficiency motors). This fact partially
explains why only one-fifth of the motors sold in the
US are energy-efficient.

2.4 Use Adjustable-Speed Driver (ASD)
The use of adjustable-speed drivers (ASDs) offers
numerous advantages to drive systems' energy
efficiency. They can offer the potential for higher
efficiencies, lower operating costs, easier control and
minimal maintenance. The most frequently used type
of adjustable-speed drivers is the frequency controlled
ASDs offering the greatest advantages. They consist
of two types: direct and indirect inverters. The first
ones convert frequency and voltage in one step,
whereas the latter use an intermediary DC link. ASDs
can take the saving about 50% in reheating furnace

2.5 Reducing peak of power demand
This method used by controlling the demand in
consumer side. For reducing peak of demand power, it
can be determined by demand factor and load factor.
If demand factor is low, it means that the connected
load is not efficient. Demand factor describe as
maximum demand
demand factor =
connected load
(5)

Life-Cycle Cost Analysis [3]

LCC = I + R - R + E + OM&R
cpl es
(6)
[ ] NS = E + W + OM&R - I + R - R
A:BC 0 cpl es

(7)
E + OM&R
SIR =
A:BC
I + R - R
0 cpl es
(8)
( ) ( )
1
AIRR = 1 + r SIR -1
N
(9)
ISSN: 1411-1284

Proceeding 10
th

( )
E + OM&R - R + R
t t cpl es
t t

0
1 1
I
t
d t

+ =
y
(10)

Where,
LCC = present value of LCC
I = present value of investment
R
cpl
= present value of capital replacement cost
R
es
=present value of residual cost
E = present value of energy cost
OM&R = present value of operation, maintanance,
repairement cost
:
NS
A BC
= Net Saving
E =
( )
E - E
BC A
= energy cost saving


3.1 PT SERMANI STEEL
The SEC of PT SERMANI STEEL before energy
conservation done is :
KES
Bulan kWh Lembar kG
kWh/Lbr kWh/kG
September 2006 58.600 333.768 1094.322 0,18 0,05
Oktober 2006 63.400 206.265 667.272 0,31 0,10
November 2006 50.900 321.493 1.027.163 0,16 0,05
Desember 2006 61.400 314.480 998.718 0,20 0,06
Januari 2007 60.300 371.483 1.135.617 0,16 0,05
Februari 2007 69.200 315.667 998.460 0,22 0,07
Maret 2007 59.000 310.368 1.053.246 0,19 0,06
April 2007 67.100 188.735 647.882 0,36 0,10
Mei 2007 47.400 354.857 1.135.782 0,13 0,04
Juni 2007 69.900 226.372 722.879 0,31 0,10
Juli 2007 53.600 342.927 766.234 0,16 0,07
Agustus 2007 47.400 119.191 417.582 0,40 0,11
Rata-rata 59.016.67 283.801 888.763,1 0,23 0,066

Table 2. SEC of PT SERMANI STEEL before energy
conservation

The energy conservations which passed by LCC
analysis are :
energy
saving
investment
average
saving
pay back
periode
(kWh/month) (Rp) (Rp/month) (month)
Reducing
connected load
0 13,935,000 2,550,428 3
Improving
power factor
625.68 20,000,000 486,084 20
energy
conservation

Table 3. Electrical energy conservation of PT SERMANI
STEEL

Using ASDs and replacing the motor did not pass
the economic requirement, so they are not feasible for
current time. However, if there will be a plan for
reconstruction and has big enough money for
investment, it will be best to use ASDs and premium-
efficiency motor. While harmonic distortion is just
about 2 % 3 %.
The new SEC after energy conservation are :
before energy
conservation
after energy
conservation
saving
Product yield (Ton) 888.76 888.76
Product Cost (Rp) 54,458,719.00 51,422,207.00 3,036,512.00
Energy Conumption (kWh) 59,016.00 58,390.12 625.88
Rp/Ton 61,274.94 57,858.37 3,416.57
SEC (kWh/Ton) 66.40 65.70 0.70
SEC (kWh/kG) 0.0664 0.0657 0.000704

Table 4. SEC of PT SERMANI STEEL after energy
conservation

From table 4, SEC of PT SERMANI STEEL
decrease 0,7 kWh/ton a month.

3.2 PT SURABAYA WIRE
The SEC of PT SURABAYA WIRE before
energy conservation done is :
Bulan kWh kG
KES
(kWh/kG)
Agustus 2006 160.500 2.187.689 0,07
September 2006 206.100 3.286.985 0,06
Oktober 2006 244.500 2.128.311 0,11
November 2006 173.790 3.703.988 0,05
Desember 2006 284.940 2.908.885 0,10
Januari 2007 167.040 2.417.828 0,07
Februari 2007 190.950 2.158.409 0,09
Maret 2007 173.220 2.941.461 0,06
April 2007 159.690 1.714.668 0,09
Mei 2007 88.620 1.790.362 0,05
Juni2007 139.110 2.593.985 0,05
Juli 2007 181.560 2.216.294 0,08
Rata-rata 180.835 2.504.072 0,07

Table 5. SEC of PT SURABAYA WIRE before energy
conservation

The energy conservations which passed by LCC
analysis are :
energy
saving
investment
average
saving
pay back
periode
(kWh/month) (Rp) (Rp/month) (month)
reducing harmonic
distortion
14,291,280 25,000,000 3,488,358 7
improving power
factor
4,087,459 40,000,000 654,139 24
energy
conservation

Table 6. Electrical energy conservation of PT SURABAYA
WIRE

Demand factor of PT SURABAYA WIRE is
90,1% so that the connected load could not be
reduced. Usage of ASDs and replacement of
premium-efficiency motor do not pass the LCC
Analysis for current time.
The new SEC after energy conservation are :
before energy
conservation
after energy
conservation
saving
Product yield (Ton) 2,504 2,504 -
Product Cost (Rp) 136,364,807 126,256,500.55 10,108,306.45
Energy Conumption (kWh) 180,835 162,456.26 18,378.74
Rp/Ton 54,457.22 50,420.48 4,036.75
SEC (kWh/Ton) 72.22 64.88 7.34
SEC (kWh/kG) 0.0722 0.0649 0.0073

Table 7. SEC of PT SURABAYA WIRE after energy
conservation

ISSN: 1411-1284

Proceeding 10
th

From table 7, SEC of PT SURABAYA WIRE
decrease 7,34 kWh/ton a month.

IV. CONCLUSIONS

T Th he er re e a ar re e f fi iv ve e m me et th ho od ds s t to o c co on ns se er rv ve e e el le ec ct tr ri ic ca al l
e en ne er rg gy y i in n i in nd du us st tr ri ia al l p po ow we er r s sy ys st te em m, , t th he ey y a ar re e i im mp pr ro ov vi in ng g
p po ow we er r f fa ac ct to or r, , r re ed du uc ci in ng g h ha ar rm mo on ni ic c d di is st to or rt ti io on n, , u us si in ng g
A Ad dj ju us st ta ab bl le e- -S Sp pe ee ed d D Dr ri iv ve er r, , r re ep pl la ac ci in ng g m mo ot to or r w wi it th h h hi ig gh h- -
e ef ff fi ic ci ie en nc cy y o on ne e, , a an nd d c co on nt tr ro ol ll li in ng g p pe ea ak k o of f p po ow we er r
d de em ma an nd d. . T Th he ey y c co on ns si is st t o of f h hi ig gh h c co os st t m me et th ho od d a an nd d l lo ow w
c co os st t m me et th ho od d. . H Hi ig gh h c co os st t m me et th ho od d n ne ee ed d a an n a ad dv va an nc ce e s st tu ud dy y
t to o t ta ak ke e t th he e d de ec ci is si io on n, , w wh hi il le e l lo ow w c co os st t n no ot t. .
S Sp pe ec ci if fi ic c E En ne er rg gy y C Co on ns su um mp pt ti io on n ( (S SE EC C) ) o of f P PT T
S SE ER RM MA AN NI I S ST TE EE EL L d de ec cr re ea as se e a ab bo ou ut t 0 0. .7 7 k kW Wh h/ /t to on n a a
m mo on nt th h a af ft te er r d do oi in ng g e en ne er rg gy y c co on ns se er rv va at ti io on n. . T Th he e m me et th ho od ds s
u us se ed d a ar re e r re ed du uc ci in ng g d de em ma an nd d c ca ap pa ac ci it ty y ( (c co on nn ne ec ct te ed d l lo oa ad d) )
a an nd d i im mp pr ro ov vi in ng g p po ow we er r f fa ac ct to or r u us si in ng g c ca ap pa ac ci it to or r b ba an nk k. .
T To ot ta al l e en ne er rg gy y s sa av vi in ng g i is s 6 62 25 5 k kW Wh h a a m mo on nt th h, , a an nd d a av ve er ra ag ge e
s sa av vi in ng g i in n 1 10 0 y ye ea ar rs s o ob bt ta ai in ne ed d i is s R Rp p. . 3 3. .0 03 36 6. .5 51 12 2, ,- - a a
m mo on nt th h. .
P PT T S SU UR RA AB BA AY YA A W WI IR RE E d de ec cr re ea as se e S SE EC C a ab bo ou ut t 7 7, ,3 34 4
k kW Wh h/ /t to on n i in n a a m mo on nt th h a af ft te er r i im mp pr ro ov vi in ng g p po ow we er r f fa ac ct to or r a an nd d
r re ed du uc ci in ng g h ha ar rm mo on ni ic c d di is st to or rt ti io on n a as s s st te ep ps s i in n c co on ns se er rv ve er r
e el le ec ct tr ri ic ca al l e en ne er rg gy y. . T To ot ta al l e en ne er rg gy y s sa av vi in ng g i is s 1 18 8. .3 37 78 8. .7 73 39 9
k kW Wh h a a m mo on nt th h w wi it th h t th he e a av ve er ra ag ge e s sa av vi in ng g g ge et t i in n 1 10 0 y ye ea ar rs s
i is s R Rp p. . 4 4. .1 14 42 2. .4 49 97 7, ,- - a am mo on nt th h

REFERENCES

[1] Centre for Renewable Energy Sources, Energy Audit
Guide, Part B : Systems Retrofit for Energy Efficiency,
European Commission, Directorate General for
Employment and Social Affairs, European Social Fund,
Athens, 2000
[2] Dugan, Roger C, et al, Electrical Power Systems
Quality, McGraw Hill, New York, 2003
[3] Fuller, Sieglinde K, and Petersen, Stephen R, Life-
Cycle Costing Manual, for The Federal Energy
Management Program, NIST Handbook 135, US
Departement of Commerce, Washington, 1995

ISSN: 1411-1284
Proceeding 10
th
Energy Conservation On Hotel Building
(Case Study in Bandung)

Rudy Setiabudy, Aji Nur Widyanto

Electrical Engineering Department, University of Indonesia
Depok 16424, Indonesia
Email : rudy@ee.ui.ac.id ; widyanto@ee.ui.ac.id

ABSTRACT-Energy conservation interpreted as
exploiting energy the efficient, rational and
effective without lessening energy needed,
productivity and comfortable. Energy
conservation can be done by energy audit. The
scope of an energy audit, the complexity of
calculations, and the level of economic evaluation
are all issues that may be handled differently by
each individual auditor and should be defined
prior to beginning any audit activities. An energy
audit can be simply defined as a process to
evaluate where a building uses energy, and identify
opportunities to reduce consumption. energy
consumption in hotels is among the highest in the
non-residential building sector in terms of absolute
values (for example, 280 kWh/m
2
in Greece, 420
kWh/m
2
in France). In this case study, there are
two opportunity of thrift able to be done by this
Hotel, installation filter harmonic and degradation
of energy with pay back period equal to 6 months.
The total opportunity can save the cost per month
equal to Rp 18.221.200,-

key words : Energy, Audit, Conservation

I. INTRODUCTION
ourisms relationship with the environment is
complex and may involve many activities that can
have adverse environmental effects, especially
related to greenhouse gas emissions and climate
change. However, tourism also has the potential to
raise awareness about environmental values and can
serve as a tool to finance the protection of natural
areas and increase their economic importance. The
two sides of this relationship can be seen in the many
municipalities of the Mediterranean basin that are, to
a large extent, dependent on tourism and therefore
also dependent on the quality of their natural
resources to attract visitors. In these communities,
energy consumption in hotels is among the highest in
the non-residential building sector in terms of
absolute values (for example, 280 kWh/m
2
in Greece,
420 kWh/m
2
in France).

In line with the increasing of the development
followed with growth of economics of Indonesia,
hence directly will bring the big challenge for
supplying energy, more than else development
followed with improvement of usage energy with
infinitude. Especially in town - metropolis, at this
time a lot of woke up by the office block, the
shopping centre and also hotel require energy big
enough. Governmental therefore have more than two
decade ago, blazed the way Konservasi Energi
Nasional Programs as indivisible shares from
Kebijakan Energi Nasional. Energy conservation
interpreted as exploiting energy the efficient, rational
and effective without lessening energy needed,
productivity and comfortable.

II. BASIC THEORY
Energy audits can mean different things to
different individuals. The scope of an energy audit,
the complexity of calculations, and the level of
economic evaluation are all issues that may be
handled differently by each individual auditor and
should be defined prior to beginning any audit
activities. An energy audit can be simply defined as a
process to evaluate where a building uses energy, and
identify opportunities to reduce consumption. There is
a direct relationship to the cost of the audit, how much
data will be collected and analyzed, and the number of
conservation opportunities identified. Thus, a first
distinction is made between cost of the audit which
determines the type of audit to be performed. The
second distinction is made between the type of
facility. For example, a building audit may emphasize
the building envelope, lighting, heating, and
ventilation requirements.

An organized approach to auditing will help us collect
useful information and reduce the amount of time
spent evaluating your facility. By splitting the audit
process into three distinct components, pre-site work,
the site visit, and post-site work, it becomes easier to
allocate our time for each step and leads to a more
comprehensive and useful audit report. The following
sections describe the tasks associated with each step
of the audit process.

Pre-Site Work
Pre-site work is important in getting to know basic
aspects of the building. This preparation will help
ensure the most effective use of our on-site time and
minimize disruptions to building personnel. A
thorough pre-site review will also reduce the time
T
ISSN: 1411-1284
Proceeding 10
th
required to complete the on-site portion of the audit.
The pre-site review of building systems and operation
should generate a list of specific questions and issues
to be discussed during the actual visit to the facility.
Pre-site Tasks
1. Collect and review two years of utility
energy data.
2. Obtain mechanical, architectural, and
electrical drawings and specifications for the
original building as well as for any additions
or remodeling work that may have been
done.
3. Draw a simple floor plan of the building on
8-1/2 x 11 or 11 x 17 inch paper.
4. Calculate the gross square footage using
outside building dimensions multiplied by
the number of stories.
5. Use audit data forms to collect, organize and
document all pertinent building and
equipment data
6. Develop a building profile narrative that
includes age, occupancy, description, and
existing conditions of architectural,
mechanical, and electrical systems.
7. Calculate the Energy Use Index (EUI) in
Btu/sqft/year and compare it with EUIs of
similar building.

The Site Visit
With pre-site work completed, we should have a basic
understanding of the building and its systems. The site
visit will be spent inspecting actual systems and
answering specific questions from your pre-site
review. Plan to spend at least a full day on-site for
each building. The amount of time required will vary
depending on the completeness of the pre-site
information collected, the complexity of the building
and systems, and the need for testing of equipment.
Small buildings may take less time. Larger buildings
can take two days or more. Here are some steps to
help you conduct an effective audit:
1. Have all necessary tools available on site.
2. Prior to touring the facility, sit down with the
building manager to review energy
consumption profiles and discuss aspects of
the facility we aren't able to see such as
occupancy schedules, operation and
maintenance practices, and future plans that
may have an impact on energy consumption.
3. Confirm the floor plan on our drawing to the
actual building and note major changes.
4. Fill out the audit data sheets.
5. Look at the systems relating to the ECMs
and O&Ms on your preliminary list.
6. Take pictures as you walk through the
building.

Post-Site Work
Post-site work is a necessary and important step to
ensure the audit will be a useful planning tool. The
auditor needs to evaluate the information gathered
during the site visit, research possible conservation
opportunities, organize the audit into a comprehensive
report, and make recommendations on mechanical,
structural, operational and maintenance
improvements. Post-site work includes the following
steps:
1. Immediately after the audit, review and
clarify your notes.
2. Review and revise your proposed ECM and
O&M lists.
3. Process your photos and paste or import
pictures on 8-1 /2 x 11 inch pages.
4. Organize all charts, graphs, building
descriptions, audit data sheets, notes and
photos into a 3 ring binder.

Electrical System Distribution Audit
The inefficient operation of electrical distribution
systems stems mainly from a low power factor. Power
factor correction is cost-effective when utility
penalties are imposed. Low power factors can be
improved with power factor correction devices and
high-efficiency motors. Additional energy can be
saved by installing energy-efficient transformers and
replacing existing motors with small and/or higher
efficiency motors, or by installing variable-speed
motor drives.
a. Power Factor
The total power requirement of a load is
made up of two components, namely, the
resistive part and the reactive part. The
resistive portion of a load can not be added
directly to the reactive component since it is
essentially ninety degrees out of phase with
the other. The pure resistive power is known
as the watt, while the reactive power is
referred to as the reactive volt amperes. To
compute the total volt ampere load it is
necessary to analyze the power triangle
indicated below:

P(Watt)
QL (VAr)
Qc(VAr)
S = P + QL

Fig 1. Power Triangle

P P
cos
S
2 2
Q P
= =
+

(1)
ISSN: 1411-1284
Proceeding 10
th

Q = Volt Amperes Reactive (VAR)
P = Watts
S = Volt Amperes (VA)
Cos = Power Factor
Amperes Factor

b. Harmonic
Distortion wave (IEC 702-07-43) is
transformation a signal which not be
intended and generally not realize with
fundamental signal reference 50 / 60 Hz. The
prime of Wave distortion for example
harmonic, DC component and inter
harmonic. Harmonic component or habit
referred as harmonic is wave having original
number fold frequency to fundamental
frequency. Equation of Frequency harmonic
is:
h = n x F Hz (2)
with :
h: -n order harmonic frequency.
F: fundamental frequency of system (50Hz/6
Hz)
n: harmonic order.

Fig. 2 Forming process of Harmonic Distortion
wave

Wave which is distortion formed by
result of merger fundamental wave with a
few harmonic wave which also in form of
sinusoidal. Value of harmonic amplitude of
wave is some percentage of amplitude
fundamental waves.
Harmonic spectrum distribution all
harmonic component amplitude as function
from its harmonic order and illustrated to
use histogram. At Fig. 2, can be seen
example of harmonic spectrum. It can be
said that spectrum represent current
comparison or harmonic frequency voltage
to fundamental frequency voltage or current.

Fig. 3. Harmonic Spectrum

Besides can be explained graph, wave
which distortion by harmonic earn also
explained mathematically use Fourier
analysis. Periodic wave which not in form of
sinusoidal can be expressed as amount of
expressed fundamental frequency harmonic
series with the following Fourier analysis
[4]:
0 0 0
1
( ) ( cos sin )
n n
n
f t a a n t b n t
=
= + +

(3)
with :
0
0
1
( )
T
a f t dt
T
=

(4)
= fundamental value of f(t) to one
period
that is from 0 till T

0
0
2
( )cos
T
n
a f t n tdt
T
=

(5)
= 2 x average value of f(t) cos n to one
period wave

0
0
2
( )sin
T
n
b f t n tdt
T
=

(6)
= 2 x average value of f (t) sin n to one
period wave
n = harmonic index

The Fourier equation can be used to
break distortion wave which have become
waving harmonic wave and basis. This
matter becomes basis in analyzing harmonic
at electric power system. On the contrary if
we know voltage amplitude value or current
every order, hence can be obtained by total
current or voltage as follows:

1
1
2
rms
V V =
and
1
1
2
rms
I I =

(7)
max
max
2
2 2 2 2
1 2 3
1
1 1
...
2 2
h
rms h h
h
V V V V V V
=

= = + + + +

(8)
ISSN: 1411-1284
Proceeding 10
th
max
max
2
2 2 2 2
1 2 3
1
1 1
...
2 2
h
rms h h
h
I I I I I I
=

= = + + + +
(9)

Total Harmonic Distortion or THD
expressing distortion level generated by all
harmonic components and defined as
follows:

2
2
1
n
n
n
M
THD
M
=
=
=

(10)
with :
THD = Total Harmonic Distortion
M
n
= RMS value of current or voltage
n
orde
M
1
= RMS value of current or voltage
at
fundamental frequency

III. Measurement and Data
This hotel subscribe to electrics from PLN
equal to 1385 kVA (tariff faction B3). From tables
and curve consume the electrics during one year,
average of usage kWh is around 429.180 kWh per
month. At figure.1 seen the usage energy of electricity
in April is highest usage around 383.046 kWh and the
lowest usage in March around 265.038 kWh. For
lowest billing happened in March (210 million rupiah)
with usage energy of electrics equal to 265.038 kWh.

Table 1. electric bill
Pemakaian Pemakaian
AUGUST 2,005 309,400 73,360 382,760 252,722,665
SEPTEMBER 2,005 305,320 72,460 377,780 259,218,405
OCTOBER 2,005 278,530 64,616 343,146 231,867,505
NOVEMBER 2,005 233,770 54,718 288,488 211,957,895
DECEMBER 2,005 298,000 68,748 366,748 291,198,465
JANUARY 2,006 291,822 65,250 357,072 284,333,365
FEBRUARY 2,006 254,590 59,334 313,924 255,776,595
MARCH 2,006 214,024 51,014 265,038 210,742,695
APRIL 2,006 311,342 71,704 383,046 258,970,710
MAY 2,006 283,096 64,280 347,376 266,925,980
JUNI 2,006 297,348 69,440 366,788 283,009,865
JULI 2,006 262,924 62,176 325,100 271,952,190
BULAN
PEMBAYARAN
TAHUN BIAYA TOTAL
LWBP WBP

Konsumsi Energi
-
50,000
100,000
150,000
200,000
250,000
300,000
350,000
400,000
450,000
Agst. 05 Sept. 05 Okt. 05 Nov. 05 Des. 05 Jan. 06 Febr. 06 Mar. 06 Aprl. 06 Mei. 06 Jun. 06 Jul. 06
Bulan
E
n
e
r
g
i (k
W
h
)
-
50,000,000
100,000,000
150,000,000
200,000,000
250,000,000
300,000,000
350,000,000
B
ia
y
a
(R
p
)
Energi Biaya

Fig. 3 Graphic of electric energy and bill

Measurement consumes electrics energy and
harmonic levels have been done at electrics panel in
company making of bearing. Then, the data has been
recorded by 9625 Power Measurement copied to
computer personal and accessed to use 9625 Power
Measurement Support Software.
Obtained by Measurement data for as follows:
a. Harmonic of Voltage
Value of Total Harmonic Distortion (THD) of
Voltage at the panel as follows:

Table 2. Value of Total Harmonic Distortion (THD) of
Voltage
VOLTAGE HARMONIC

U
1
(%) U
2
(%) U
3
(%)
Average 1.82 1.95 1.95
Maximum 2.35 2.59 2.36
Minimum 1.51 1.61 1.63

By using maximum Voltage THD equal to 3 %,
seen that voltage harmonic level at electrics panel
have in tolerance range.
Hereunder represent each phase of distortion
voltage wave at the electrics panel which have
harmonic. Fig. 4 hereunder represents waveform
noted by measuring instrument.

Fig. 4. voltage wave

2. Current and Level of Harmonic Current
Value of Total Harmonic Distortion (THD) of
Current at the panel as follows:

Table 3. Value of Total Harmonic Distortion (THD) of
Current
CURRENT HARMONIC

I
1
(%) I
2
(%) I
3
(%)
Average 14.04 14.04 15.15
Maximum 20.51 20.51 24.40
Minimum 8.33 8.33 8.26

By using maximum Current THD equal to 10
%, seen that current harmonic level have passed
tolerance range. Fig. 5 represents each phase of
harmonic distortion current wave which have and

ISSN: 1411-1284
Proceeding 10
th

Fig. 5. Current wave

Compared to voltage wave, influence harmonic
distortion current wave stronger destroying. This
matter can be seen from level of harmonic current (%
THD) larger than level of harmonic voltage (% THD).

3. Power Factor
Value of Power Factor at the panel as follows:

Table 4. Value of Power Factor
PF
1
PF
2
PF
3

Average 0.99 0.99 0.98
Maximum 1.00 0.99 0.99
Minimum 1.00 0.98 0.97

By using minimum power factor equal to 0,85,
seen that ower level have in tolerance range.

4. Total Power
Fig. 7 represents of total power which have and

Konsumsi Energi
-
100.000
200.000
300.000
400.000
500.000
600.000
700.000
0
:0
0
1
:0
0
2
:0
0
3
:0
0
4
:0
0
5
:0
0
6
:0
0
7
:0
0
8
:0
0
9
:0
0
1
0
:0
0
1
1
:0
0
1
2
:0
0
1
3
:0
0
1
4
:0
0
1
5
:0
0
1
6
:0
0
1
7
:0
0
1
8
:0
0
1
9
:0
0
2
0
:0
0
2
1
:0
0
2
2
:0
0
2
3
:0
0
Waktu
E
n
e
r
g
i

(
W
)
Harian Hari Kerja Hari Libur

Fig. 6. Total power curve

From graphic, seen the result measurement
of electrics load PLN, the maximal load almost come
near 770,5 kW with mean power factor equal to 0,98
hence used energy at this Hotel almost come near
786,22 kVA or to load contract PLN (1385 kVA)
have weared the load equal to 56,77 %.

IV. Analyze and Solution
Identify energy conservation based on specification
data of electrics installation and result data of
measurement electric energy characteristic, hence
some opportunities of thrift able to be done as follows
:
a. Installation Filter Harmonik
From data and measurement, maximum
level of current harmonic have passed the
tolerance (< 10 %). To overcome the
mentioned needed by usage filter harmonic.
This Harmonic current height will cause
some losses at equipments operation,
overheating, neutral overloading, degradation
lifetime equipments and improvement of
consumption kwh ( current) so that to solve
can be done with installation filter harmonic,
thereby deductible current harmonic till
under 5 %.
Based on calculation, Cost effective for
external time of peak load (
LWBP)kWh/month = Rp.8.136.000,- and
Cost effective for time of peak load (
WBP)kWh/month = Rp. 2.275.200,-. So the
total Cost effective kWh/month = Rp.
8.136.000 + Rp. 2.275.200 = Rp.
10.411.200,- or equal Rp. 124.934.400,- per
annum.

b. Degradation subscribe from PLN
Demand factor is the number show
comparison between maximum electricities
(peak load) with electricity attached
(Subscribed kVA). Pursuant to record
measurement which have been done in
September 2006, we get calculation of peak
load/burden operate for is 786,22 kVA. With
subscribe capacity from PLN equal to 1385
kVA hence level of Demand Factor to month
of September 2006 is
Demand Factor =
786 kVA
1385 kVA
= 56,77 %
From data energy consumption (kWh a
yearlong, seen the highest consumption kwh
happened in April, is so that needed to do
approach of peak load during the month of
September to month April as follows :
383046
786, 72 797, 69
377780
x = =
kVA
Relate at calculation of peak load
(kVA), hence to determine customer capacity
(KVA) can be countable by adding reserves
KVA equal to 20 % from highest load
(kVA), so that = 120 % x 797,69 kVA =
957,22 kVA
ISSN: 1411-1284
Proceeding 10
th
Degradation subscribe from 1385
kVA become 1110 kVA, hence can be done
by the cost-saving kVA equal to :
Cost effective kVA/month
=( 1385 1110) x cost/kVA
= 275 x Rp. 28.400
= Rp. 7.810.000
Cost effective kVA/annum
= 12 x Rp. 7.810.000
= Rp. 93.720.000

Degradation subscribe from 1385 kVA
become 1110 kVA need an investment
around Rp 53.695.000,- so the simple pay
back period :
53.695.000
7.810.000
= =
6,875 month 7 month
V. Conclusion
1. Demand factor of electrics in this Hotel only
56,77 %
2. Peak load oh this hotel equal to 786,22 kVA
3. There are two opportunity of thrift able to
be done by this Hotel, installation filter
harmonic and degradation of energy with
pay back period equal to 6 months.
4. The total opportunity can save the cost per
month equal to Rp 18.221.200,-

References

1. Fan Wang. 2001. on Power Quality and
Protection. Gteborg : Chalmers University of
Technology
2. Roger C. Dugan..[et al]. 2002. Electrical Power
Systems Quality. New York : McGraw-Hill
3. Theraja B.L et.al. 1997. A Text Book of
Electrical Technology. New Delhi : S. Chand &
Company Ltd
4. Thumann, Albert. William J. Handbook of
energy audits, Younger6th ed.

ISSN: 1411-1284
Proceeding 10
th

Indonesia Energy Revolution Scenario and Projection
using MESAP/Planet Simulation Model

Rinaldy Dalimi
1
, Bayu Indrawan
2
, Sven Teske
3

1
Electrical Engineering Department , University of Indonesia
2
Engineering Center, University of Indonesia
Tel. 08194831323, fax. 021-7863506, email : bayu@bayuu.web.id
3
Climate & Energy Unit Greenpeace International

Abstract Two scenarios up to the year 2050 are
outlined in this report. The Reference Scenario is
based on the reference scenario published by the
International Energy Agency in World Energy
Outlook 2006, extrapolated forward from 2030
have been calculated using the MESAP/PlaNet
simulation model. The Energy Revolution Scenario
has a target for the reduction of worldwide per

capita carbon dioxide emissions to less than 1.3
tonnes per year by 2050.
Indonesia Energy Revolution Scenario
describes a development pathway which turns the
present situation into a sustainable energy supply.
Exploitation of the existing large energy efficiency
potential will reduce primary energy demand from
the current 6,900 PJ/a (2004) to 9,500 PJ/a in 2050.
This compares with a demand of 16,000 PJ/a in the
Reference Scenario.
The electricity sector will have the strongest
growth in renewable energy utilisation. By 2050,
more than 60% of electricity will be produced
from renewable energy sources. A capacity of 112
GW will produce 367 TWh/a of electricity. In the
heat supply sector, the contribution of renewables
will continue to grow, reaching more than 65% by
2050. By 2050 over 40% of primary energy
demand will be covered by renewable energy
sources.
Keywords - Indonesia Energy Revolution Scenario,
Reference Scenario, Alternative Scenario,
MESAP/PlaNet simulation model, Renewable Energy.

I. INTRODUCTION
he climate change imperative demands nothing
short of an Energy Revolution. At the core of this
revolution will be a change in the way that energy is
both produced and distributed. The five key principles
behind this shift will be to implement clean,
renewable solutions, especially through decentralised
energy systems, to respect the natural limits of the
environment, to phase out dirty, unsustainable energy
sources, to create equity in the use of resources and to
decouple growth from the consumption of fossil fuels.
Decentralised energy systems, where power and heat
are produced close to the point of final use - avoiding
the current waste of energy during conversion and
distribution - will be central to the Energy Revolution,
as will the need to provide electricity to the two
billion people around the world to whom access is
presently denied.

II. BASIC THEORY

The development of future global energy demand is
determined by three key factors:
Population development: the number of
people consuming energy or using energy
services.
Economic development, for which Gross
Domestic Product (GDP) is the most
commonly used indicator. In general, an
increase in GDP triggers an increase in energy
demand.
Energy intensity: how much energy is
required to produce a unit of GDP.
Both the Reference and Energy Revolution scenarios
are based on the same projections of population and
economic development. The future development of
energy intensity, however, differs between the two,
taking into account the measures to increase energy
efficiency under the Energy Revolution Scenario.


Combining the projections on population
development, GDP growth and energy intensity
results in future development pathways for final
energy demand in Indonesia. These are shown in
Figure below for both the Reference and Energy
Revolution Scenarios. Under the Reference Scenario,
total energy demand will more than double from the
current 4,500 PJ/a to 11,300 PJ/a in 2050. In the
Energy Revolution Scenario, we expect a much
T
ISSN: 1411-1284
Proceeding 10
th

slower increase to 6,500 PJ/a in 2050, which is about
45% more than today, but reducing by 40% the
projected consumption under the Reference Scenario.
The accelerated increase of energy efficiency, which
is a crucial prerequisite for achieving a sufficiently
large share of renewable energy sources in energy
supply, is beneficial not only for the environment but
also from an economic point of view. Taking into
account the full service life, in most cases the
implementation of energy efficiency measures saves
costs compared to additional energy supply. The
mobilisation of energy saving potential leads directly
to a reduction in costs. A dedicated energy efficiency
strategy therefore helps to compensate in part for the
additional costs required during the market
introduction phase of renewable energy sources.

IV. CONCLUSIONS

The Energy Revolution scenario shows a pathway to a
more sustainable future for Indonesias energy sector.
Its main features are improvements in energy
efficiency and a higher share of renewable energy.
This report demonstrates that renewable energy is not
a dream for the future it is real, mature and can be
deployed on a large scale. Decades of technological
progress have seen renewable energy technologies
such as wind turbines, solar photovoltaic panels,
biomass power plants and solar thermal collectors
move steadily into the mainstream.

REFERENCES

[1] Baur, Jorg, MESAP - Methodological Concept of
PlaNet, Institut fr Energiewirtschaft und Rationelle
Energieanwendung, Stuttgart, 1998

[2] Sven Teske, Energy Revolution A Sustainable
World Energy Outlook, Greenpeace International,
2007.

3a4737f339
Reference Scenario
0
2.000
4.000
6.000
8.000
10.000
12.000
2004 2010 2020 2030 2040 2050
Transport Other Sectors Industry
P
J
/
a
0
2.000
4.000
6.000
8.000
10.000
12.000
2004 2010 2020 2030 2040 2050
P
J
/
a
Energy Revolution Scenario
P
J
/
a
0
2.000
4.000
6.000
8.000
10.000
12.000
2004 2010 2020 2030 2040 2050
P
J
/
a
0
2.000
4.000
6.000
8.000
10.000
12.000
2004 2010 2020 2030 2040 2050
ISSN: 1411-1284

Proceeding 10
th

Ocean Wave Power Plant With Air Pressure

Massus Subekti*, Vivian Karim Ladesi

Fac. of Engineering, University of Indonesia,
Telp. 08881547426, email: massus.subekti@ui.ac.id; massussubekti@yahoo.com
Fac. of Oceaninc, Sepuluh November Institut of Technology
Telp. 08563100083, email: karimladesi@yahoo.com

Abstrak- This Paper study about device a power plant
of sea wave with pressure air, this generating
advantage among other things source energi
abundance especially the coastal area and archipelago
area, high closeness energy and the potential area
which unlimited. done wave observation of area to be
used for reference, then design the effective model to
convert sea wave energy become observation data
input so that will be got turbine energy which couple
with generator
Result of research indicate that by using
calculation Boyansi, level of force float (F
1
) at
secondary ponton have diameter 30 cm is 0.14 kN or
140 N. With l1 = 4 m and l2 = 2m, hence got force at
cylinder equal to 280 N. with Wide of Output Surface
7,85.10
-5
m2, Force of cylinder output is 0.27 N.
Correct territorial water for the use model this
conversion is coast of between 10 m until 100 m from
coast lip which adapted by contour sea floor of wave
and coast break condition which possible happened,
for this case is coastal area Lampung. Level of Energy
which convertion very hinge from high of waves that
happened, so that in the end will have an in with the
level of yielded electricity.

Keywords Ocean Wave, Power Plant

I. INTRODUCTION

n this time, most of energy used come from fossil fuel,
that is oil fuel, gas, and coal, used fossil fuel can
destroying the environment wich induce for green house
beside that, fossil fuel is nonrenewable and unsustainable
source energy category.
This problems increase complex when a lot of society
not benefit from electric current, especially on purilies,
isle, seaboard and not yet been reached by netork of
electric are, because the location is far from electric
power station. This condition of couse can pursue
economic growth in this are. For this problem, alternative
souce energy which reach to this area.
One of alternative souce energy the society neds is
exploiting sea wave energy. The profit from this anergy
is enough for supplying high density of energy, unlimited
of potential area, neednt fuel, no pollution, renewable
enrgy, more produces of energy, and cheap cost.

II. BASIC THEORY

Ocean Wave
Wave is one form of transition or switch simple
energy wich that energy received from original source
kinds of sea wave based on generator:
1. Bodies moving on surface make a wave at sea
surface with low periode from low energy
2. Swells on the sea surface from blown wind in the sea
surface
3. Tsunami from disruption of earthguake
4. The tide causing flooding from month and sun
gravitation field
One of its, the important and use in this research is
swells
Many types of wave in the world and difficult to
describe by matematis because unlinear, tree dimension
and its for wich random. But, there are some approach
can be used, among the others is Airy theory, stokes,
mich, Gersner, konidal, and Single, with each other
having different definition.
SWELL (linear wave) is a certain of wave as relativie
have great length and small height. SWELL is represent
wave was energy. Although high wave (H) more small
than long wave (), energy from high SWELL relative.
For the comparison between height wave to length
wave(H/) 1/50.
[1]

.

I
ISSN: 1411-1284

Proceeding 10
th

Figure 1. Sket of wave

Float Building Theory
The observed rom based low of static of float building
theory, because of that the se water fluid not get out from
Archimedes principle where afloat buiding gotten
immersed to water will experience force comparable with
water volume removed, this mean weight of object float
at this fluid same as water volume remove. level of force
that happened at object following equqtion:

F = g V
[2]
(1)
Where : F = Force float (N)
g = Gravitation
V = Volume
= Specific gravity fluid
While, g = 10,06 kN/m
3
atau 1,026 t/m
3
(2)

So that the level of force will influence by volume of
object float.
At This research, the secunder pontoon ude is circle
form, if the assumed is pontoon have diameter (d) so
pontoon volume is

Vponton = 4/3 (d/2)
3
(3)

Wind Energy Convert
Large of energy was transferred by wind energy to
turbine rotor shaft depend on air density, wide turbine
rotor are and air speed, and can be formulated:

E = m v
2 [2]
(4)
dimana : E = energy (joule)
m = mass of air (kg)
v = wind speed (m/detik)
Sum up mass which through of place by wide A m
2

and move with speed m/detik, can be formulated:

m = A v q
[2]
(5)
where : A =Wide (m
2
)
q = air density (kg/m
3
)
large of power which tranfered by turbine heve
efficiency can be formulated:
[3]

P = E (per et of time)
= A q v
3

= r
2
q v
3
(kW)
[3]
(6)
dimana : P = Forve (kW)
= efficiency of turbine rotor
r
2
= wide area of turbien rotor (m
2
)
r = radius of turbien rotor (m)
q = air density (kg/m
3
)
v = wind speed (m/det)


Diagram Alur
Research step path in the form of diagram can be
described by following :

Figure 2. Groove diagram of the research

Detailedly step of research:
1. Literature Study
Collecting circumstantial and more complete
literature-literatur about theory and method which
can be utilized in research
2. Models Scheme of conversion Energy
Determining of model convert energy which is
assuming effective and efficient
3. Determining size measure model
Determining wich size measure model have
making to be conducted a calculation
4. Determination result of and the compatible
location
Simulation/ Case study And place which is suited
for application

S
H/
C
H
h
x
z
SWL : Still water Level
Literature Study
START
Scheme Model.
Size measure Model
Result of.
FINISH
Yes
No
?
ISSN: 1411-1284

Proceeding 10
th

L = 3m
P = 5 m
T = 2m
Pon
ton
Sek
Model and the Size Measure Convert
1. Convert Energy sketch

Figure 3 Convert energy sketch ( seen from front)

2. Diagram Block

Figure 4. Block of diagram of generating device

3. Principle
Energy yielded by secondary ponton from sea wave
channelled to pistone through lever. Piston will
depress air come into air tube 1& 2. the high
pressured air used to turn around turbine which have
been tied on a generator. then the Air thrown

4. Primary Ponton

Figure 5. Primary ponton dimension

Figure 6. High of Primary Ponton from sea floor

Figure 7. Place Arrangement of primary Ponton
( seen from the)

5. Secondary ponton & Activator Lever

Figure 8. Dimension of activator lever.

By using calculation Boyansi (Eq.1), level of force
float (F
1
) at secondary ponton have diameter 30 cm (V
=1.4137x 10
-5
is 0.14 kN or 140 N. With l1 = 4 m and l2
= 2m, hence got force at cylinder equal to 280 N

Primary
Ponton
7 m

F
1

F
2

4
2
piston
G
Generator
Air turbin
Air
save 2
Air Save 1
ISSN: 1411-1284

Proceeding 10
th

F
2

6. Cylinder

Figure 9. Dimension of cylinder

Diameter of piston (
piston
) = 32 cm
Wide of piston surface = 0,08 m
2

Diameter of cylinder output = 1 cm
Wide of Output Surface = 7,85.10
-5
m
2

Force of cylinder output (F
3
) = 0.27 N
Long of Cylinder = 0,5 m
Air volume @ cylinder = 0,04 m
3

Air volume 20 cylinder = 0,8 m
3

6. Air Save
Diameter = 0,25 m
Long = 1 m
Volume Air Save 1 &2 = 0,098 m
3

Diameter output = 5 cm
Surface output = 19.63 cm
2

= 1.96.10
-3
m
2

IV. CONCLUSIONS

1. Ocean Wave Power Plant enough to be used
especially in archipelago region which not yet been
touched by an electrics network
2. Correct territorial water for the use model this
conversion is coastal area with distance 10 m until
100 m from coast lip which is adapted for contour of
sea wave floor condition and coast break which
possible.
3. Level of energy converted very depended from high
of coastal wave that happened, so that in the end will
have an in with the level of yielded electricity

REFERENCES

[1]. Triatmodjo, Bambang, Teknik Pantai, Beta Offset,
Yogyakarata, 1999
[2]. Giancoli, Daoglas, Fisika 1, 5
th
edition, Erlanggga,
Jakarta, 1998
[3]. Abdul Kadir, Energi, Sumber daya, Inovasi, Tenaga
listrik dan Potensi ekonomi, Edisi Kedua, UI Press,
Jakarta, 1995
[4]. Hans Jk, Ocean Energy Reecovery, Honolulu.
Hawai 1989
[5]. Lewis, Edwar V.. Principles Of Navar Architecture,
Printed in The USA, Jersey City,1988
[6]. www.oceanpowertechnologies.com/pdf/senate_hearing_pa
per.pdf
[7]. www.Wavegen.co.uk

: 32cm
30cm
:1cm
F
3

ISSN: 1411-1284

Proceeding 10
th

T
The Time for Indonesia to Optimize Geothermal
Energy

Oleh: Massus Subekti*
* Fac. of Engineering University of Indonesia
Telp. 021-4712137, email: massus.subekti@ui.ac.id; massussubekti@yahoo.com

Abstract- world oil price almost 100 US dollar per
barel progressively push world society to call
away the attention new source energy utilize to
anticipate period rare of petroleum. One of source
energy which nowadays start to become world
attention is geothermal, besides big enough
potential and the environmental friendliness,
convert goethermal become electrics own
efesiency which high enough.
exploiting of Goethermal directly have reached
27.825 MWt in 71 state in the year 2004. While for
the power station have reached 8.200 MWe in 21
state in the world. among other things the United
States of America, French, Germany, and
Indonesian.
Indonesia represent state with biggest potency
geothemal in the world that is 40 % or equal to
27.000 MWe, but only 807 MWe or equal 3%
have been exploited for power station. Some
constraint like weakness still the infrastructure,
low price sell, weakness clarity punish,
governance bureaucracy which the condition still
KKN, subsidize fuel oil and also weakness
awareness socialize, making growth PLTP
Indonesia walk very tardy.
Some step which ought to be done among other
things gift of incentive to development PLTP,
covering: reduction PBB, area retribution, PPH,
PPN, Iease import and the licensing amenity.
Revise legislation about geothermal in order to
give clarity punish to all investor, reduction step
by step subsidize to BBM in order to the price sell
of geothemal more kompetitif, facility of research
activity, seminar and discussion in around
technological development of PLTP in Indonesia,
and which do not less important planning
inwroughtly and continual in development of
national energy in order to there continuity in sill
of energy of primary in Indonesia

Keywords Geothermal, Indonesia

I. INTRODUCTION

The price of petroleum which keeps
increasing has forced the world citizen to start
the change their attention towards the latest energy
resources in order to anticipate the extinction of
petroleum era. One of the energy resources which
become the attention of the world is geothermal since
not only the energy that is produced can be conversed
into electric energy, but its potential is also sufficient
and friendly for the environment.
Several countries who have already applied
geothermal energy such as the United States,
England, France, Italy, Sweden, Swiss, Germany,
New Zealand, Australia, and Japan

II. BASIC THEORY

According to etymological term, Geothermal is
originated from two words, Geo which means earth
and thermal which means heat. Therefore
geothermal means the heat that comes from the earth.
According to the regulation in UU No. 27/ 2003
about geothermal, it is stated that geothermal is the
resource of heat energy contained in hot water, steam
and rocks along with the mineral substances and
other gases which genetically cannot be separated in
geothermal system and its function needs a mining
process.
The movements of the earth layers which collide
with each other causes the radioactive process in
earths layers depth and make a resulting magma
formation with reservoir more than 2000
o
C. Every
year rain and snow trickles absorb into the earth
layers, and are held in rocks layers which are already
exposed to heat waves and magma. These rocks
layers are called geothermal reservoir which range
between 200
o
300
o
C. The water cycle which
happen every year cause the reservoir rocks layers as
the place to create geothermal energy which can be
produced repeatedly in a very long time.
In the earths skin, sometimes the water stream
can be very close with hot rocks where the
temperature can reach until 148
o
C. The water does
not turn into steam because there is no contact with
air. When the hot water comes up into the earth
surface because of a space or a crack on the earths
skin, the hot water will come up, which is usually
called hot spring. This hot spring is commonly used
for hot spring pools, and many of these become a
tourism object.
[1]


A. The History of Geothermal Energy in
Indonesia
ISSN: 1411-1284

Proceeding 10
th

The exploration of geothermal in Indonesia was
started in 1920, Netherland built 5 shallow wells with
depth of +66 meters in West Java. Out of 5 wells
which were built, there is only 1 well left which still
produce superheated steam, that is the locomotive
well in Lumajang. In 1930, the inventory of
geothermal was once stopped because of economic
depression. Then, in 1974 Indonesia started its
geothermal exploration again.
[2]

B. The Potency of Indonesian Geothermal
Indonesia is a country with the largest geothermal
potential in the world as much as 40% from the world
reserve which equals with 27.000 MWe. This
potential is spread especially in Sumatra, Java, Bali,
Sulawesi, Nusa Tenggara Barat, and Nusa Tenggara
Timur. From that potential, only 807 MWe or + 3%
which is already exploited for geothermal power
plant. (See Table 1)

Table 1. The Geothermal Potential in Several
Countries
[3]

Exploitation
No. Country Potential
(MWe) (%)
1 Iceland 5800 202 3.5
2 USA 22990 2534 11
3 Indonesia 27000 807 3
4 Philippines 4335 1931 44.5
5 Japan 20000 535.25 2.7
6 Mexico 6000 953 15
7 New Zealand 3650 435 11.99

Based on the location of Indonesia geothermal
potential, West Java is at the highest rank with 40
locations and followed by NTT with 18 locations,
Aceh with 17 locations, and North Sumatra, West
Sumatra, and North Sulawesi with 16 locations in
each. (Table 2)

Table 2. The Locations of Geothermal Potential in
Indonesia
[4]

No Province sum No Province sum
1 Aceh 17 14 East Java 11
2 North Sumatara 16 15 Bali 5
3 West Sumatra 16 16 NTB 3
4 Riau 1 17 NTT 18
5 Jambi 8 18 Nort Sulawesi 5
6 Bengkulu 4 19 Gorontalo 2
7 Bangka Belitung 3 20 Central Sulws 14
8 South Sumatra 6 21 South Sulws 16
9 Lampung 13 22 SouthEast Sul 13
10 Banten 5 23 Nort Maluku 9
11 West Jawa 40 24 Maluku 6
12 Central Jawa 14 25 Papua 2
13 Yogyakarta 1 26 West Kalmt 3
Total Lokasi 251
Source: The Directorate General of Minerals, Coal, and
Geothermal, ESDM

Based on the category of geothermal potential, the
total amount of Indonesia geothermal resources is
14.080,5 MWe, with 9.467,5 MWe of which is
speculative and the rest of 4.613 MWe is hypothetic
resource. Whereas the total reserved potential is
13.060 MWe. Java is the island with the largest
geothermal potential with estimated reserve reaching
for 1.837 MWe with installed capacities of 785
MWe. (Table 3)

Table 3. Indonesia Geothermal Energy Potential
[4]

Resource(M
We)
Reserve (MWe)
Location
Specul
ative
Hypo
thetic
Estim
ated
Feasi
ble
Verifi
ed
Capac
ities
Sumatera 5,63 2.353 5.433 15 389 2
Jawa 2.362,5 1.591 2,86 603 1.837 785
Bali-Nusa
Tenggara
175 427 871 - 14 -
Sulawesi 925 125 721 110 65 20
Maluku 275 117 142 - - -
Kaliman 50 - - - - -
Papua 50 - - - - -
9.467,5 4.613 10.027 728 2.305 Sum
251 lokasi 14.080,5 13.060
807
MWe
27.140,5
Source: The Directorate General of Minerals, Coal, and
Geothermal, ESDM

From the existed potential, there is only 2.7%
which has been exploited while 3.1% is being in the
exploration phase, 30% is still in the geological
survey phase, and the rest of 63,7% is still being in
the preliminary survey phase. (Chart 1)
[5]

Chart 1. Percentage of Indonesia Geothermal
Energy Potential

C. Indonesia Geothermal Power Plant
Indonesia Geothermal Power Plant Capacity
The exploitation of Indonesia geothermal energy
is enhancing together with the increase of electric
energy necessities. This improvement of usage is
shown from the increasing electric energy capacity
usage from geothermal energy as much as 32,25 MW
in 1982; it increased into 142,42 MW in1990;
increased into 587,5 MW in 1998 and 807 MW in
2004. (Chart 2)
[6]

Percentage of Indonesia Geothermal
Energy Potential
Ekspl oi t asi
2. 7%
Eksplorasi
3.1%
Survei Geologi
30%
Survei Awal
63.7%
ISSN: 1411-1284

Proceeding 10
th

Source : various source
Indonesia Geothermal Power Plant
Capacity
32.25
142.42
807
587.5
0
200
400
600
800
1000
1982 1990 1998 2004
Year
MW

Chart 2. Indonesia Geothermal Power Plant Capacity

Current Indonesia Geothermal Power Plant
At present, there are 7 Geothermal Power Plants
(GPP) which have been operating; 140 MW in
Kamojang, 330 MW at Gunung Salak, 145 MW in
Darajat, 110 MW in Wayang Sindu, 60 MW in Dieng
(Java) and 2 MW in Sibayak (North Sumatra) and 20
MW in Lahendong (North Sulawesi) (Table 4)

Table 4. Current Indonesia Geothermal Power Plant
No Region Company
Capacity
(MW)
1 Kamojang Pertamina 140
2 Gunung Salak Chevron G.Salak 330
3 Darajat Chevron Darajad 145
4 Wayang Windu Star Energy 110
5 Dieng Geodipa 60
6 Sibayak Pertamina 2
7 Lahendong Pertamina 20
T O T A L 807
Source: Department of Energy and Mineral Resources

Exploitation Plan of Geothermal Power Plant
(GPP)
At this moment there are 3 places which are still
in the phase of exploitation plan, they are; Lahendong
II GPP in North Sulawesi with 20 MW capacities
which is planned to activate in the end of 2007, and
Wayang Windu II GPP in West Java with 110 MW
capacities which is planned to operate in 2008, and
Bedugul II GPP with 175 MW capacities which
cannot be confirmed when to operate yet, due to the
contradiction from the residents of Bali itself. (table
5).
[7], [8]

Table 5. Geothermal Power Plants in Exploitation
Plan
No Name
Capacities
(MW)
Year
1 Lahendong II
[7]

(Nort Sulawesi)
20 2007
2 Wayang Windu II
[8]

(West Java)
110 2008
3 Bedugul (Bali) 175 ?
Source: various sources
Geothermal Power Plants in Exploration Phase
At present, there are 14 GPP which are still in the
exploration phase, such as; Seulawah GPP with 160
MWe and Jaboi GPP with 50 MWe in Nanggroe
Aceh Darussalam, Sekincau GPP with 238 MWe in
Lampung, Cisolok GPP with 45 MWe, Tangkuban
Perahu GPP with 220 MWe, Gunung Tampomas
GPP with 20 50 MWe in West Java, Ungaran GPP
with 50 MWe in Central Java, Ngebel Gunung Wilis
GPP with 120 MWe in East Java, Mataloko GPP
with 65 MWe in NTT
[9]

Table 6. List of Geothermal Power Plants in
Exploration Phase
[10]

No Region Company
Capa
city
(MWe)
1 Seulawah NAD 160
2 Jaboi NAD 50
3 Sekincau Lampung 238
4 Cisolok West Java 45
5 Tangkuban Perahu West Java 220
6 G. Tampomas, sumedang West Java 20-50
7 Ungaran Central Java 50
8 Ngebel, gunung Wilis East Java 120
9 Mataloko NTT 65
10 Atadei NTT 40
11 Marana south-east
Selawesi
40
12 Suwana Gorontalo 65
13 Jailolo North Maluku 75
14 Songa-Wayaua North Maluku 140
T O T A L 1358
Source: www.esdm.go.id

Indonesia Geothermal Energy Development Plan
Referring to the Blue Print of National Energy
Management 2005 2025, the government has
planned mixed energy which in 2003 the geothermal
energy will be raised from 1.4% until 3.8% from all
national energy consumption. In order to achieve that
target, the government has made the Road Map of
Indonesia Geothermal Energy Development 2004
2025 by enhancing the capacity of GPP from year to
year. In 2008, it is aimed that the capacities will be
installed for 2000 MW whereas in 2012, 2016, and
2020 it is aimed to reach 3442 MW, 4600 MW and
6000 MW. Therefore, in 2025, the capacity is
intended to reach for 9500 MW.

Constraints in Indonesia Geothermal
Development
A number of steps that should be taken for
Indonesia geothermal development, such as;
1. The distance between the location of geothermal
resources and the capacity load makes the need
of massive infrastructural supports is important.
2. The flaws in infrastructure, especially the
transmission system that connects Geothermal
Power Plant with capacity load.
ISSN: 1411-1284

Proceeding 10
th

3. A significant opening investment. The
geothermal development needs enormous
investment especially in exploration phase which
affects the aspects of expense and value from the
whole project, and the decide of the outcome
steam price. Geothermal Energy Association
(GEA), the Energy Department of the United
States of America noted that making an
exploration of geothermal well with the capacity
of 55MWe will need investment which costs
US$ 5.8 million. This includes the cost as much
as US$ 432 thousand for preliminary study, US $
1.2 million for exploration, and US$ 4.2 million
for properness study.
[11]

4. The fuel subsidy that makes the selling price
becomes less competitive.
5. Very low selling price which is not comparable
with the expended investment.
6. The requirement of advanced technology, which
results in the use of geothermal exploration
technology from out of the country.
7. The weakness in legal transparency that can
create the possibility of contract violation.
8. Bureaucracy constraints. It is difficult to pass
government bureaucracy in who is not
corruption-free yet.
9. The weakness in community awareness and
government role in socializing the importance of
geothermal power plant development.

The Strategies of Geothermal Power Plant
Development in Indonesia
A number of steps that should be taken for
Indonesia geothermal development, such as;
1. The distribution of incentives toward GPP
development that include the decrease in PBB,
region retribution, PPH, PPN, Import Tax and
permission ease.
2. The revision of legalizations about geothermal
usage in order to give legal transparency to the
investors.
3. The distribution of buying guarantee after the
steam is discovered. Based on the consideration
that geothermal is not an exportable commodity,
but it can be useful merely for domestic needs.
4. The fiscal policy of free-tax policy in order to
import the operational goods (PDRI) for the
necessity of geothermal effort is necessary to be
given in order to support axisting contract
project so the production capacity can be added
in accordance with the contract commitment.
5. The periodical decrease of fuel subsidy so that
the selling price of geothermal will be more
competitive.
6. Facilitating researches, seminars and discussions
about the development of geothermal power
plant technology in Indonesia.
7. Supporting the improvement of geothermal usage
for direct significance.
8. Integrated and continuous planning in improving
national energy, consequently, there will be a
synergy in the development of primary energy in
Indonesia

IV. CONCLUSIONS

If it is viewed from the existing potential, the
factor of technological progress, the increase of
worlds oil price and the efforts to reduce the glass
house effects, the future of geothermal energy in
Indonesia is very promising. On the other hand,
considering the high cost of investment needed, it
seems that the development will not be the same as
expected. It needs not only concrete steps from the
government but also complete support from the
community so that the prospect of geothermal energy
in Indonesia is compatible with the expectation.

REFERENCES

[1]. http://www.pertamina.com/index.php?option=com_co
ntent&task= view&id=3015&Itemid=340
[2]. http://www.kompas.com/kompas-
cetak/0307/14/teropong/407920.htm
[3]. www.dim.esdm.go.id/makalah/2-8%20ITB-
Nenny%20M%20S.pdf
[4]. Sugiharto Harsoprayitno, Peluang Panas Bumi
Sebagai Sumber Energi Alternatif Dalam Penyediaan
Tenaga Listrik Nasional, Direktorat jenderal Mineral,
Batubara dan Panas Bumi, Dep ESDM
[5]. http://www.majalahtrust.com/ekonomi/sektor_riil/116
6.ph
[6]. www.dim.esdm.go.id
[7]. www.kompas.com
[8]. rafflesia.wwf.or.id.pdf
[9]. www.esdm.go.id
[10]. Blue Print Pengelolaan Energi Nasional 2005-2025
[11]. http://www.republika.co.id/koran_detail.asp?id=29903
2&kat_id=&kat_id1=&kat_id2=

ISSN: 1411-1284

Proceeding 10
th

Incorporating Energy Commodity Price Volatility in
Economic Cost of Supply Analysis for Electricity
Expansion plan

Emil Elestianto Dardak*

* Fac. Of Economics, International Programme, INDONUSA Esa Unggul University
Tel. (62-21)5674090 email : emil.dardak@indonusa.ac.id

Abstract Rising prices of tradable energy
commodity is a key risk in the provision of electricity
infrastructure. Indonesia is undertaking major
expansion of electricity through construction of coal-
fired power plants. The current abundance of coal
supply, including the technology to utilize low calorie
coal, is the main rationale behind the selection of
coal. At the same time, the electricity utility (PLN), is
faced with the need to pursue least cost expansion,
and is given no clear incentive to take into
consideration energy mix and security in its
expansion plan. Thus, whenever coal-fired power
plant can be pursued as the financially least cost
expansion, any other generation options, even those
promising hedge against commodity price volatility,
such as renewable energy, lose out to coal. This
paper seeks to conduct a cost comparison between
coal-fired power plant and geothermal power as a
renewable energy, in which the aspect of coal price
volatility is taken into consideration. Economic cost
of supply curve will be used to conduct the analysis.

Keywords energy, electricity, renewable energy, cost
of supply, price volatility

I. INTRODUCTION

nergy security is becoming increasingly pertinent
along with the significant hike in fossil fuel price,
particularly oil. A Presidential Decree no.5 has
mandated an energy mix with more than 10%
contribution coming from new and renewable energy.
However, we see coal dominating the immediate
expansion plan with the 10,000MW crash program. At
the same time, renewable energy has grown very little.

Among the most significant barriers to the
implementation of renewable energy is the cost of
provision. Global Environmental Facility has in the past
provided incremental cost financing to provide
incentives for development of renewable energy and to
cover such incremental cost vis--vis conventional fuel
based generation. With the ratification of Kyoto
Protocol, a global mechanism named Clean
Development Mechanism was established to reward
reduction in carbon emission achieved by substituting
less clean energy technology with cleaner technology.

Nevertheless, only 4 years away from the end of the
Kyoto Protocol in 2012, Indonesia has not been able to
utilize this mechanism to upscale its renewable
generation. Geothermal, is available in huge potential in
Indonesia, with an estimated size of 27GW. However,
only 4 percent of this potential has been developed and
only 2 projects have or will soon benefit from carbon
credits, which are Lahendong and Darajat.

Therefore, it can be argued that the provision of carbon
credit cannot enhance viability of renewable energy to a
level that makes it competitive to substitute less clean
fossil fuel technology. On the other hand, the volatility
of fossil fuel price has led many to question about the
reliability of such generation for the longer term.

This is where renewable energy is believed to possess a
comparative advantage by nature, as its price is
relatively stable. Taking geothermal as an example,
geothermal promises stable price over the long term.
However, least cost expansion planning conducted by
the state power utility in Indonesia reveals bare
minimum geothermal generation, as reflected in the
five-year Electricity Provision General Plan. The
competitiveness of renewable energy is clearly low if
reference is made to this least cost planning, however,
this contradicts the notion that renewable energy can
become competitive with its stability of price. It is
argued that the lack of quantification of such
commodity price volatility hedge benefit is the main
reason for its relative absence in the terms of large scale
contribution for future power generation plan.

It is however understandable to argue that the site-
specific characteristics of geothermal project costs have
made it difficult to incorporate in the least cost
E
ISSN: 1411-1284

Proceeding 10
th

expansion planning model. In addition, other factors
may have hindered the development of geothermal,
which includes the level of risk, time taken from
reconnaissance to reach commissioning, and other
possible factors that may affect the comparison between
geothermal and other generation options. Nevertheless,
it is important that quantification of the benefit of hedge
against volatility can be made to allow for such
incorporation in the expansion planning for power
utilities.

II. BASIC THEORY

A study by Wiser and Bolinger (2004) look into the
volatility of fossil fuel prices and how this enhances the
competitiveness of renewable energy with its relatively
certain cost of supply. Their study focuses on the
volatility of gas-fired power plants in the USA, and on
how to appropriately compare the levelized cost of
fixed-price RE to the levelized cost of variable-price
gas-fired generation. Wiser and Bolinger also point out
the common current practice in analytic studies and
utility planning, as previously mentioned in the case of
Indonesian power utility. The practice is to compare the
cost of these resources based on an inherently uncertain
and notoriously inaccurate fuel price forecast.

Wiser and Bolinger suggest that a more appropriate
approach be the comparison of the levelized cost of RE
to the levelized cost of gas-fired generation based on a
guaranteed price of natural gas that can be locked in
with forward, futures, or swap contracts. They test the
reliability of such contracts as reference for long term
price forecasts based on futures, forward and swap gas
prices that were priced in November 2000 November
2003. The finding shows that price forecasts are
consistently made below the forward price. Hence, to
lock in a coal price, ones must pay a premium. At the
same time, price forecasts have been inaccurate, and
often underestimated the actual price of fuel
commodity. Thus, to obtain the benefit of hedge against
price volatility, utility must refer to a forward price if
non renewable energy or fossil fuel based generation is
selected.

Figure 1 shows the comparison for contract price and
spot price for the Northern Appalachian prices between
January 2000 and September 2002. Contract prices are
generally higher than spot prices, whereby contract
prices are more stable and are priced at a premium that
range between $1 and $2 per short ton. In terms of
percentage the premium ranges from 4% to 9%, with the
exception of minor cases where spot price is higher than
contract price. If we take an average of those two
numbers, the base premium for locking in coal price is
6.5%.

Figure 1 Comparison of Contract and Spot Prices for
Coal

In case of renewable energy such as geothermal, the
cost of fuel is not applicable since the source of energy
is contained in the reservoir where it is non-tradeable
and has no fluctuations of price. There is possible
fluctuation in operations and maintenance cost due to
inflation, but since such consideration is not made in
coal power calculation, this is not going to affect the
comparison results. There are indeed make up wells
requirements to maintain steam supply, but this can be
accounted for in the initial capital costing.


A comparison of cost for coal power and levelized cost
for renewable energy (geothermal) was made. Coal
power cannot actually be levelized as cost inevitably
increases along with long term fuel price increase.
Geothermal energy has inherent risk and thus a low case
and high case scenario were used in this calculation.

The assumptions used for calculating cost of coal power
plant: i) Capacity: 300MW, ii) Heating Value: 0.4
ton/MWh, iii) operations and maintenance: $30/KW,
iv)capital cost: $1,000/KW, iv) lifetime: 25 years, and
v) annual operation hours: 7,000 hours.

The assumptions used to calculate geothermal power
(low case) is i) Capacity: 300MW, ii) operations and
maintenance: $18/KW, iii) capital cost: $1,800/KW, iv)
lifetime: 25 years, and v) annual operation hours: 8,760
hours.

The assumptions used to calculate geothermal power
(high case) is i) Capacity: 300MW, ii) operations and
maintenance: $18/KW, iii) capital cost: $1,800/KW, iv)
lifetime: 25 years, and v) annual operation hours: 8,760
hours.

ISSN: 1411-1284

Proceeding 10
th

In the calculation of coal power plant cost, the coal price
used is the forward coal price that is charged at an
average premium of 6.5% of the spot price. The
forecasted spot price is indexed to inflation at an
average of 6% per year. Based on this calculation, we
obtain the results where every year, coal power price
must change due to change in coal fuel price.

Using the above assumptions at the discount rate of
12%, the cost for first year when spot price of coal is
estimated at $45/ton is 4.7 cents. It is important to note
that forward price is used as volatility of price within
the year can be hedged using such contract, and this
makes the comparison with renewable energy more
robust. The cost towards the end of the lifetime at 25
years, reaches 10.6 cents due to the increase in spot coal
price at $182/ton.

If we compare this with spot price based calculation, the
first year cost would be 4.6 cents, and the cost towards
the end of the lifetime reaches 10.1 cents, which is
around 5 percent lower than the price based on forward
contracts for coal that is priced at 6.5 percent premium.

In the case of geothermal, the low case gives a levelized
cost of 3.71 cents per kWh. Levelized cost is possibly
calculated in the case of geothermal since there is no
associated fuel price increase relevant to the costing.
Variations in capital cost is based on geological
conditions, whereby depth of well and productivity are
among key factors that affect the cost of upstream
activities, in addition to the number of wells and success
ratio of drilling. The high case for geothermal gives a
levelized cost of 6.19 cents per kWh.

If geothermal discount rate is assumed at 16% to reflect
the higher risk for the project, the low case changes to
5.94 cents and the high case increases to 7.86 cents per
kWh.

If coal power plant is calculated conventionally, using
spot price at a fixed level, the levelized cost would be
4.61, in which a low case geothermal would be
competitive while a high case geothermal would not be
competitive when discount rate is set at 12%, which is
equal to coal.

A mid-range between the two cases for geothermal
would also not be competitive to this levelized cost of
coal. It is true that in the first year cost for coal in the
previous calculation of coal based on contract price,
coal is still more competitive than geothermal with the
exception of low case. But along with the increase in
coal price, the cost per kWh increases and makes
geothermal more competitive than coal.

When the discount rate is set at 16%, a conventional
calculation of coal would make both geothermal power
not competitive. If we compare both cases, in the case
of 12% discount rate, a high case geothermal would
only be viable when the project reaches its 11
th
year
when the coal power price reaches 6.2 cents. In the case
of 16% discount rate, a low case geothermal would only
be viable when the project reaches its 10
th
year when
coal power price reaches 6 cents, and the high case
geothermal would only be viable when the project
reaches 18
th
year, when coal power price reaches 8
cents.

The challenge now is in finding the levelized cost for
coal power plant. Since the price cannot be levelized,
there needs to be a benchmark for cost over the lifetime
of the project. To use the first year figure would
underestimate the cost per kWh, while using the final
year figure would overestimate since such price would
only be valid for the final year of operation.

As a solution, the benchmark price is calculated based
on the average per kWh price over the lifetime of the
project. The price is also normalized with the discount
rate to reflect the present value of higher power price
that reflects higher purchasing power increase along
with economic growth as indexed by inflation. Inflation
is assumed at an average of 6 percent per annum.

Interestingly, if we use inflation to normalize future
price, the cost of coal power becomes lower as the
increase in price is less than the discounted rate. Thus,
on the basis of future paying ability increasing
proportionate to inflation, an increase in coal price may
be acceptable and the ascending price structure may
allow for cheaper initial power price that may fit with
the existing purchasing ability.

The price of 1.06 cents in the 25
th
year actually becomes
very low when present value is calculated at 6%
discount rate, whereby the normalized price is 2.5 cents.
Thus, when the benchmark price is calculated at the
average per kWh price over the lifetime of the project,
the cost is at 3.25 cents.

IV. CONCLUSIONS

Renewable energy provides the benefit of energy
security and hedge against price volatility. Price
volatility can be contained through locking in of price
using forward contract instruments. Such contract for
coal would create premium at an average of 6.5% of the
spot price. Thus, comparison between renewable energy
and coal price must be made using this contract price,
particularly since utility seek to avoid volatility in
ensuring security of energy supply.
ISSN: 1411-1284

Proceeding 10
th

Conventional method of assuming constant price and
trying to arrive at a levelized cost for coal fired power
plant is certainly underestimating the true cost of
provision over the lifetime of the project. Thus,
comparing renewable energy with coal becomes
difficult and biased towards coal that is currently priced
lower in the market.

The use of inflated price to reflect long term price
increase has fundamentally changed the results.
Conventional levelized cost would give 4.61 cents per
kWh, while a levelized cost would give 4.7 cents in the
first year going upwards to more than 10 cents in the
end of the 25
th
year.

Competitiveness of geothermal increases if coal price
increase is taken into account. However, if coal price
increase is normalized with inflation figures to account
for increasing ability to pay, the increase caused by coal
price can be offset by the increase in purchasing power.
This notion requires careful examination and may
undermine the impact in relation to environment and
potential scarcity due to variability of uses other than
power generation. The key finding of this analysis
would remain the fact that geothermal power promises
hedge, which can significantly alter the competitiveness
of coal.

REFERENCES

[1] Wiser, R. and Bolinger, M. (2004), The Value of
Renewable Energy as Hedge against Fuel Price
Risk, World Renewable Energy Congress
Proceeding, Colorado
[2] Globalcoal.com, Futures in Coal
[3] Platts Coal Database

EFFECTS OF CHIMNEY DEPTH AND DOWNJET
HEIGHT IN A COAL BRIQUETTE STOVE ON
THE CO EMISSION AND IGNITION TIME

Dijan Supramono*, Yulianto Sulistyo Nugroho**, and Dian Nurlita Kusuma*

*Department of Chemical Engineering, Faculty of Engineering, University of Indonesia
**Department of Mechanical Engineering, Faculty of Engineering, University of Indonesia
Depok 16424, West Java, INDONESIA
Phone: 021-7863516, email : dsupramo@che.ui.edu

Abstract - The utilisation of Indonesian coal as fuel
has increased steadily from year to year in response
to more expensive oil. Concurrently, the utilisation
of coal in briquette shapes has also increased,
especially by small and medium industries and low-
income households.
However, their increased utilisation encounters
some disadvantages such as long ignition time
(ignition delay) and high pollutant emissions. In
terms of the CO emission, it has been found that the
current briquette stoves still have high CO emission
in the order of magnitude 100 ppm, which could
reach 1000 ppm. These figures are well above the
threshold emission value of 25 ppm for human
health.
Effects of the chimney depth of the coal briquette
stove and downjet height installed inside the
chimney on the CO emission have been investigated.
Chimney is a zone above the briquette bed in the
coal briquette stove. The depth of the chimney may
affect the residence time of combustion gases above
the briquette bed and longer chimney zone is
hypothesed to allow CO to have sufficient time to
convert to CO
2
. The downjet height is measured
from the mouth of the coal briquette stove to the
surface of briquette bed surface. The downjet above
the coal briquette bed creates a recirculation above
the briquette bed which retains combustion gases for
some time in the chimney zone and enables CO to
have sufficient time to convert to CO
2
. At the same
time, unreacted hydrocarbons have also longer
residence time to convert to CO
2
. The conditions
which accommodate the conversion of CO to CO
2

are encouraged which gives two benefits, lower CO
emission from the briquette stoves and higher coal
heat release from the briquette. Exothermic reaction
from carbon in coal to CO
2
is approximately three
times of that from coal to CO.
In this investigation, data of combustion
temperature and CO emission above the coal
briquette bed versus time were collected in a fire
calorimeter during the combustion of 0.8 kg coal
briquettes starting from their ignition. The depth of
chimney and the height of the downjet were varied
respectively from 10 cm to 30 cm and from 5 cm to
20 cm. The air for combustion flowed naturally,
which is also called natural draft, from the bottom
side of the stove upwards.
The analysis of the data found that the deeper
chimney gives shorter ignition time and higher
maximum temperature. However, the deeper
chimney experienced steeper temperature reduction
after the maximum temperature was achieved. When
the maximum temperature was achieved, the CO
concentration dropped. It was followed by the
increase of CO concentration after the combustion
gas temperature dropped. CO concentration
dropped again after it reached its maximum value.
Averagely the CO concentration is higher in the
deeper chimney. In terms of the effect of downjet
height, it was found that the higher downjet
increases ignition time and maximum temperature,
but reduces the average CO concentration in the
combustion gases. In conclusion, the deeper
chimney and shorter downjet benefit to the shorter
ignition time and adversely increase the CO
concentration in the combustion gases.
.
Keywords: Coal briquettes , Stove, Chimney, Downjet

I. INTRODUCTION

Statistical data published by Ministry of Energy and
Mineral Resources showed that coal briquette
production was predicted to increase twice in the next
four years [1]. Although the utilisation of the coal
briquettes has good prospect, the high CO and
hydrocarbon emissions and long ignition time are still
major problems. Some preliminary experiment on the
related topic in the researchers laboratory found that
the current briquette stoves still have high CO emission
which could reach 700 ppm [2]. These figures are well

beyond the threshold CO emission in working areas
stipulated by Minister of Workforce in 1997, i.e. 25
ppm. Human exposed to such a high CO emission for
long time will have health problems because gas CO can
bond haemoglobin of the blood which hinders the
transport of oxygen in the blood.
The effort carried out by Tekmira Bandung to
reduce CO emission was by creating proper ventilation
in the room where briquette stove using exhaust fan. By
doing so, the gas of the product of coal briquette
combustion is expelled out of the room and fresh air is
drawn towards the room so that the CO content in the
air drops approaching the threshold value [3]. It is well
known that the ventilation system in the kitchens in
most of Indonesian households is rarely equipped with
exhaust fan. Without the fan, the convection of CO to
the exterior air is slow and this gas can remain inside the
kitchens for some time. Therefore, the introduction of a
method/methods to improve the design of coal briquette
stove so that the design itself controls the emission of
CO and hydrocarbons. This research is concerned with
how to reduce the CO gas emission which leads to
achieving or approaching the threshold value of CO
emission.
Emission CO originates from oxidation reactions of
volatile matter as well as carbon. In order to obtain
highly complete reaction towards CO
2
formation 3
requirements should be fulfilled, i.e. sufficient residence
time for conversion of CO to CO
2
, sufficient oxygen to
complete oxidation reaction, and high temperature to
improve reaction kinetics. The first requirement should
be considered in the briquette combustion because the
highest temperature of the briquette material only
reaches about 700
o
C as a consequence of the heat
absorption of the briquette combustion by high heat-
capacity briquette material [3]. This temperature is
much lower than that achieved by liquid or gas fuel
combustion, which could reach 1200
o
C. These low
temperatures are unfavourable for the conversion of CO
to CO
2
which requires high temperatures (more than
1000
o
C) [4]. In order to offset this disadvantage, coal
briquette combustion requires much longer residence
time to obtain high completion of oxidation reaction of
CO to CO
2
, which is the controlling reaction of coal
briquette combustion. The second requirement is related
to the sufficient contact between hydrocarbons of
volatile matter and CO gas and oxygen to react towards
the CO
2
formation. This is related to the mixing pattern
between briquettes and air. The excessive oxygen makes
the heat hardly increase the combustion gases
temperature high enough to proceed the conversion of
CO to CO
2
and incurs partial oxidation. Products of this
partial oxidation are oxygenated compounds such as -
CH
2
O, -CHO and CO gas [5]. The first and second
requirements can be enhanced if the reactions proceed at
high temperatures. Therefore, in order to reduce the
emissions of CO gas and hydrocarbons, in all zones
where the contact between combustion gases and
oxygen in coal briquette stove occurs, the contact time
is sufficiently long, the contact between oxygen and
partial oxidation products is performed well, and the gas
phase temperature is kept high.
II. EXPERIMENTAL
In this research, the first and second requirements
were to be investigated by varying of chimney depth
and downjet height (see Figures 1 and 2). Combustion
reactions wer carried out using natural draft for air
supply. Chimney is a zone above the briquette bed in the
coal briquette stove. The chimney depth is measured
from the mouth of the coal briquette stove to the surface
of briquette bed. The hypothesis of the research is that
the depth of the chimney affects the residence time of
combustion gases above the briquette bed and the
deeper chimney allows CO to have sufficient time to
convert to CO
2
.
The downjet is a jet installed above the coal
briquette bed which directs fresh air jet downwards. The
downjet above the coal briquette bed creates a
recirculation above the briquette bed which retains
combustion gases for some time in the chimney zone
and allows CO and unreacted hydrocarbons to have
sufficient time to convert to CO
2
. The downjet also
supplies additional oxygen for the reactions of CO and
unreacted hydrocarbons. The conditions which
accommodate the conversion of CO to CO
2
are
encouraged which gives two benefits, lower CO
emission from the briquette stoves and higher coal heat
release from the briquette. Exothermic reaction from
carbon in coal to CO
2
is approximately three times of
that from coal to CO.
The depth of chimney was chosen 10, 20 and 30 cm
by adjusting the height of grate and downjet height 5,
10, 15 cm by adjusting the length of the downjet nozzle
at chimney depth of 30 cm. For each value of chimney
depth or downjet height, the CO concentration in the
flue gas of the stove was measured. The velocity of the
air from the nozzle downjet was measured averagely
0.017 m/s.
The measurement of CO concentration was
performed concurrently with the measurement of
temperature of combustion gases immediately above the
briquette bed to obtain information of the ignition time.
Using this information, effects of the chimney depth and
downjet height on the ignition time, which is a measure
of how difficult the coal combustion works, were
obtained.
The equipment and material used in the experiment are
as follows:
1. Fire calorimeter
The fire calorimeter facilitates temperature and CO
concentration measurements. The calorimeter was built

with reference to ASTM standard E1354-1997 and ISO
5660. Figure 3. shows schematic design of fire
calorimeter. A thermocouple was used to measured flue
gas temperature sucked from the briquette stove.
2. Gas analyzer Quintox
A gas sampling gun was inserted into the ducting to
measure CO and residual oxygen concentrations in the
flue gas and connected to a Gas Analyzer, Quintox
KM9106. Using data logger of the gas analyser, CO and
residual oxygen concentrations can be measured at
preset sampling frequency.

3. Coal briquette
There were 2 coal briquettes used in the experiment,
ignition-promoting briquettes and cooking briquettes.
The first briquettes were used to accelerate the ignition
delay of the cooking briquette combustion. They were
prepared by blending the subbituminous coal powder
and 15% weight of ethyl acetate and 10% cooked starch
as briquette glue. The cooking briquettes were supplied
from South Sumatera. The weight ratio of the first to the
second briquettes was 1 : 10. The experiment was
initiated by burning ignition-promoting briquettes for 10
minutes. The burnt briquettes were then laid on the top
of the cooking briquettes and this was the beginning of
the cooking briquette combustion. The combustion
allowed to occur while CO and oxygen concentration
data were logged and flue gas temperatures immediate
above the briquette bed and in the calorimeter ducting
were recorded.

Figure 1. Stove with chimney zone

Figure 2. Stove with downjet installed above the briquette bed

III. RESULTS AND DISCUSSION
1. Chimney Depth
Profile of the flue gas temperature above the briquette
bed against combustion time is shown in Figure 4. At all
values of chimney depth, the profiles are similar in
which the temperature increases to reach a maximum
value. The times when the temperatures start to increase
which represents the initial time of the ignition are
different for different chimney depths. The result
presented in Figure 4 shows that the deeper the
chimney, the earlier the ignition starts and the deepest
chimney gives more abrupt increase of the flue
temperature. At other depths, the increase is more
gradual. In terms of the maximum flue gas temperature,
the deeper the chimney, the maximum temperature is
higher. The ignition is triggered by the release of
volatile matters from the briquette surface. Because the
ignition starts at different times among briquettes in the
stove in which the briquettes close to the ignition-
promoting briquettes (upper layers cooking briquettes)
ignite earlier, the ignition times described in Figure 4
combine different ignition times of the briquettes in the
stove. Briquette bed with the deepest chimney gives
better heat propagation from the ignition-promoting
briquettes to the cooking briquettes compared to others,
implied by abrupt increase of the flue gas temperature.
The earlier ignition time in the stove with deeper
chimney shows that the stove generates earlier heat due
to more oxidation reaction between oxygen and
briquettes. There is a competition between the
favourable condition of more natural draft due to
buoyancy driving force [6] and unfavourable condition
of more natural draft due to back pressure [7] in the
chimney at higher temperature. It seems that back
pressure is more dominant to determine the magnitude
of natural draft. This is implied by the effect of chimney
depth on CO concentration as shown in Figure 5.

Figure 5 shows that. CO molecules start to evolve
when the ignition starts to occur and its concentration
drops when the ignition reaches its maximum
temperature. It shows that CO is converted to CO
2
if
high temperature is available in the flue gas as shown by
corresponding temperature data in Figure 4. The
combustion reaction of coal briquettes is initiated by
fast reaction of coal to gas CO and slow reaction of CO
to CO
2
. The kinetics of the late reaction is enhanced if
high temperature prevails. The CO curve rises again
after the maximum temperature passes for some time
followed by the gradual reduction of CO concentration
as mass of briquettes also reduces. Visually, Figure 5
shows that deeper chimney gives more CO production.
This is in line with what has been mentioned in the
previous paragraph, in which deeper chimney prevents
the natural draft than shallower chimney. Consequently,
CO concentration in the flue gas produced by the deeper
chimney is more than that by shallower chimney. The
average calculation of the CO emission found that the
CO concentration average of the stove at chimney
depths of 30cm, 20cm and 10 cm respectively are 456,
287 and 235ppm. High CO concentration measured in
this experiment suggests that further investigation is
required to reduce this amount. One way to achieve this
may be by preheating the line of the downjet so that the
air exiting the downjet is hot enough to function as
additional reactant in the briquette surface and it is
expected that CO concentration can go lower.

Figure 3. Schematic design of fire calorimeter

2. Downjet Height
Figure 6 shows the effect of downjet height on the flue
gas temperature right above the briquette bed. The
graphs show that the stove with higher position of the
downjet from the surface of the briquette bed, the later
the ignition starts to occur. The velocity of air exiting
the downjet reduces when the air flows further down
away from the downjet nozzle [8]. The location of
downjet near the surface of the briquette bed benefits
the convection of the heat from the ignition-promoting
briquettes to the cooking briquette beneath the earlier
briquettes. Therefore, the stove with lower position of
the downjet gives earlier ignition. However, once
ignition occurs, the heat is absorbed by the downjet air
(quenching). The lower the position of the downjet, the
effect of absorption is stronger, so that the maximum
ignition temperature is lower. This gives consequences
on the CO emission where lower maximum ignition
temperature, which corresponds to the lower average
flue gas temperature, results in higher CO emission as
shown in Figure 7. The averages of CO emission as the
downjet height is varied 10, 15 and 20 cm respectively
are 747, 433 and 301ppm. The quenching whould have
been avoided if the downjet air had been preheated
such as by passing the line of downjet through the
combusting briquette bed.

IV. CONCLUSIONS
The experiment has been carried out using coal
briquettes to determine effect of chimney depth and
downjet height on the ignition time and CO emission.
The air supplied to the briquettes using natural draft.
The experiment gives conclusions that
1. The deeper chimney gives earlier ignition time and
higher maximum temperature. However, due to the
backpressure resulting from the higher combustion
gas temperature in the more confined high chimney,
such favourable temperature even gives adverse effect
on the CO emission.
2. The lower downjet gives earlier ignition to the coal
briquettes. However, once the ignition has
commenced, the cool air exiting the downjet at lower
position gives adverse effect on the combustion
oxidation.
3. Further investigation is needed to improve the effect
of the downjet on the CO emissions. One possible is
to preheat the air in the downjet line before ejecting
the air into the chimney zone.

Figure 4. Effect of chimney depth on combustion gas
temperature

Figure 5. Effect of chimney depth on CO emission in flue gas

Figure 6. Effect of downjet height on combustion gas
temperature

CO Emissions at Varying Downjet Heights
0
200
400
600
800
1000
1200
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Time (s)
C
O

(
p
p
m
)
Jet 10cm
Jet 15cm
Jet 20cm

Figure 7. Effect of downjet height on CO emission in the flue
gas

V. REFERENCES

[1]. DESM, Planning of production and utilisation of coal
briquettes 2006-2010, Departement of Energy and
Mineral Resorces Indonesia, Desember 2005
[2]. Wibowo, T. and Supramono, D., Effects of shape, size and
carbonisation of coal briquettes on the ignition time, heat
release and pollutant emissions, Student Final Project,
2007.
[3]. Balia, L., Research and development of coal briquettes in
Indonesia, Research Centre of Mineral and Coal
Technology, Departement of Energy and Mineral
Resorces Indonesia, 1996
[4]. Makino, A. Drag coefficient of a slowly moving carbon
particle undergoing combustion, Combustion Science and
Technology, vol. 81, pp. 169-192, 1992.
[5]. Turns, S.R., An Introduction to combustion, concepts and
applications, McGraw-Hill, 2nd edition, 2000
[6]. ASHRAE Handbook of Fundamentals, Chapter 26 by
American Society of Heating, Refrigeration and Air-
Conditioning Engineers (ASHRAE). Atlanta, GA: 2001.
[7]. B. Lewis, and G. von Elbe, Combustion, Flames and
Explosions of Gases, Academic Press, Inc, 1987.
[8]. Evans, D.G., Downjet combustion of coal at low
burning rates, Inst. Fuel Conf., Newcastle, N.S.W., 1966,

ISSN: 1411-1284

Proceeding 10
th

Scale Up Reactor for Bio-gasoline Production from
Crude Palm Oil

Bambang Heru Susanto, Anondho Wijanarko and Mohammad Nasikin

Department of Chemical Engineering, the University of Indonesia
Kampus UI Depok, Indonesia 16424

Abstract-The process of synthesizing bio-gasoline
from crude palm oil through cracking reaction was
conducted using a liquid phase batch reactor at
atmospheric pressure. The reaction temperatures
were between 300
o
C and 320
o
C, while the reaction
time were between 1 and 2 hours for each
temperature. In laboratory scale, at 300
o
C, H-zeolite
that was prepared from natural zeolite Klinoptilolite
type, was used with the catalyst/reactant (CPO)
weight ratio 1:75 to produce 10,3% bio-gasoline in
yield that have bio-gasoline fraction around 93.5%
and octane number of 122. Catalyst was prepared by
an ion exchange method using NH
4
NO
3
aqueous
solution. Addition of 5% bio-gasoline to gasoline
increases the gasoline performance similar to the
Pertamax.
This bio-gasoline production method was
scale up to the near commercial scale and for first
step, the dimension of the reactor was increased to
10 L. The reactor dimension scale up was done
base on hydrodynamic calculation with the
assumption of constant ratio of mixing power per
volume. For better homogenized temperature and
composition in the reactor, this scale up uses a
helical ribbon impeller in this mixing reactor, the
value of mixing time is around 10 second. For
producing biogasoline that can also be used as an
additive gasoline, the optimum temperature of this
reaction is similar to the result in laboratory scale.
This reaction produced 12.2% bio-gasoline in yield
that had octane number of 106 and gasoline
fraction around 34.3% (dominated with aromatic
and olefin product). The octane number of this
product was lower to the laboratory scale product,
and for using as an additive of gasoline to
Pertamax grade gasoline, it need an increasing
volume of additive up to 10%.

Key word: scale up, H-zeolite, bio-gasoline, crude
palm oil (CPO)

Introduction
The possibilities of producing fuels
from renewable resources such as biomass are
recently one the major issues in global research
activities (Idem et al., 1997; Twaig et al., 1999;
Katikaneni et al., 1996). These kinds of
technologies are more environmental friendly
and contribute effectively toward sustainable
development (Yusoff, 2004). Vegetable oils have
been the main subject of these researches (Twaig
et al., 1999). Vegetable oils mostly consist of
triglyceride, a hydrocarbon molecules formed
when a propane molecule bonded with three fatty
acids at each of the carbon atoms. This molecule
has a quite similar structure with the
hydrocarbons in crude oil. Therefore it is a very
potential alternative for the crude oil in
producing hydrocarbon fuels. Among the
vegetable oils, palm oil is known to be one of the
most potential alternatives because of its high
amount of production, especially in tropical
country like Indonesia and Malaysia (Yusoff,
2004).
Bio-gasoline was one of renewable
hydrocarbon fuel product from vegetable oil
such as crude palm oil (CPO), and have
characteristic and composition is similar to the
commercial gasoline and may be used as
commercial additive to increase the octane
number of premium gasoline and an attempt to
produce these bio-gasoline from palm oil
through cracking reaction like commonly done in
crude oil refinery recently have not been done.
This was one of the reasons to do this research.
This research was done to study the possibility of
applying the cracking reaction, in the process of
liquid fuel production from palm oil, using
cracking catalyst in a different reaction condition
than the common condition applied in crude oil
refineries. In refineries, the cracking reaction is
commonly done in gas phase reactor with high
pressure. In this research, the reaction was done
in liquid phase reactor with the reaction
temperature below the initial boiling point of
palm oil and in atmospheric pressure.
Some researches have been done to produce
liquid fuels from palm oil. Bhatia, et al have
proved that catalytic cracking of palm oil using
HZSM-5 catalyst at 350
o
C could produce light
products such as methane, ethane, gasoline,
kerosene, diesel oil, and BTX (benzene, toluene,
xylene) (Twaig et al., 1999). He also proved that
liquid fuels can also be produced from palm oil-
based fatty acid mixture which is the waste from
palm oil industry and also from used cooking oil
(Ooi et al., 2004a; Ooi et al., 2004b). Other
attempts with different researchers, different
methods, and different catalysts have also been
done to utilize the palm oil in producing liquid
fuels efficiently (Idem et al., 1997; Twaig et al.,
1999; Katikaneni et al., 1996; Yusoff, 2004; Ooi
et al., 2004a; Ooi et al., 2004b; Twaig et al.,
2003a; Twaig et al., 2003b; Twaig et al., 2004,
Wijanarko and Nasikin, 2007). This research was
ISSN: 1411-1284

Proceeding 10
th

design to produce liquid fuels from CPO by
catalytic cracking using H-Zeolite at 300
o
C for
producing the renewable fuel in gasoline range.

Material and Method
The catalyst used for this experiment
was a natural H-zeolite, that was prepared from
Indonesian Lampoong natural zeolite, mostly
consist of clinoptilolite type zeolite, by ion
exchange method. For 25 kg H-zeolite
preparation, natural zeolite was processed around
50 h at 500 rpm mixing rate using ammonium
nitrate solution (1 N) as the source of proton.
thereafter this produced NH
4
-zeolite form was
dried at 110
o
C around 4 h and then calcined at
520
o
C for 5 hours to omit NH
3
species for
produce H-Zeolite. Then, produced H-zeolite
was characterized by BET to find the catalyst
pore diameter.
The laboratory scale process of
synthesizing bio-gasoline from crude palm oil
through cracking reaction was conducted using a
liquid phase batch reactor with stirrer and
heating jacket at atmospheric pressure, the
reaction temperatures were between 300
o
C and
320
o
C, while the reaction time were between 1
and 2 hours for each temperature. In laboratory
scale, at 300
o
C, it used catalyst/reactant (CPO)
weight ratio 1:75. The reactor was also equipped
with a reflux tunnel in order not to make it over-
pressured without losing any light hydrocarbon
products formed in the reaction. The support of
the catalyst was prepared by an ion exchange
method using NH
4
NO
3
aqueous solution. The
density and boiling point of the reaction products
were measured to evaluate the reaction. The
products were also analyzed using FTIR
spectrometry and GCMS. To obtain the bio-
gasoline, sample of the reactor product was
distilled in an atmospheric distillation column.
The distillations were done twice to get light
hydrocarbons with as high concentration as
possible. An addition of 5% bio-gasoline that
was obtained from previous research result was
capable to increase gasoline performance similar
to commercial Pertamax. The schematic
illustration of this laboratory scale process was
shown in Fig. 1.
The pilot scale process of synthesizing
bio-gasoline from crude palm oil through
cracking reaction was conducted using a liquid
phase batch reactor with double helical ribbon
stirrer and heating jacket at the similar condition
to laboratory scale condition. To obtain the
products, sample of the reactor product was also
purified with the similar process of laboratory
scale process. The products were also analyzed
using FTIR and GCMS. The reactor design
illustration of this pilot scale process including
its detail sizing was shown in Fig. 2.

Result and Discussion
To evaluate the cracking of triglyceride
and to ensure whether cracking really occurred,
the FTIR spectra of one of the reactor products is
compared to the FTIR spectra of palm oil before
the reaction. To do the comparison, the
absorbance peak of C=O bond in the range of
1740-1745 cm
-1
is used as reference because the
amount of this bond did not significantly
changed during reaction (Silverstein and Webster,

Fig. 1. Schematic illustration of the experimental steps.

1997). Fig. 3 shows that relative to the reference
peak, the peak at 1161 cm
-1
in the spectra of
reactor product is dissapeared which was
identified in the spectra of palm oil (not shown).
It means there was CO group during the
reaction was cracking. The peak at 3000 cm
-1

that was dominant in the spectra of palm oil (not
shown), have a significant decreasing after
catalitic cracking, and this indicated that the
decreasing of carbon atom in CPO molecular
structrure was occured. Although at interval 800

ISSN: 1411-1284

Proceeding 10
th

Fig. 2. Batch Reactor Design using Helical Ribbon
Impeller Type (measurement unit in cm)

1000 cm
-1
and also at 700 cm
-1
was seem as a
little peak in FTIR result of bio-gasoline product,
it was indicated that formation of double bond
fungsional group and aromatic group during
cracking reaction was occured. The other results
in pilot scale and also result in smaler case were
same to the above explanation. To identify the
hydrocarbon content of the bio-gasoline, GC-MS
analysis was conducted. GC-MS analysis was
also used to identify the hydrocarbon content of
palm oil before the reaction for comparison. The
results of the analysis are shown in Table 1. Bio-
gasoline was one of renewable hydrocarbon fuel
product from vegetable oil such as crude palm
oil (CPO), and have characteristic and
composition is similar to the commercial
gasoline and may be used as commercial additive
to increase the octane number of premium
gasoline. The yield of cracking reaction in
producing gasoline-grade hydrocarbon is
defined as follow:
% 100 %) ( =
P
Y
Vol Yield
(1)
where P is the palm oil feed volume and Y is the
gasoline-grade hydrocarbons volume in the
reactor products.
The process of synthesizing bio-gasoline from
crude palm oil through cracking reaction was
conducted using a liquid phase batch reactor at
atmospheric pressure, the reaction temperatures
were between 300
o
C and 320
o
C, while the
reaction time were between 1 and 2 hours for
each temperature. In laboratory scale, at 300
o
C, a
H-zeolite was used with the catalyst/reactant
(CPO) weight ratio 1:75 to produce 10,3% bio-
gasoline in yield that have gasoline fraction
around 93.5% and octane number of 122. An
addition of 5% bio-gasoline that was obtained
from laboratory scale research result was capable
to increase premium gasoline performance
similar to commercial Pertamax.

Fig. 3. FTIR result of Bio-gasoline Product at 300
o
C.

This bio-gasoline production method
was developed from laboratory scale up to the
near commercial scale and for first step, the
dimension of the reactor was increased to 10 L
and it would be increase to higher volume. The
reactor dimension scale up was done base on
hydrodynamic calculation with the assumption
of ratio of mixing power per volume is constant.
For better homogenized temperature and
composition in the reactor, this scale up found
that using helical ribbon impeller in this mixing
reactor, the value of mixing time is around 10
second.

ISSN: 1411-1284

Proceeding 10
th

Tabel 1.
GC-MS Analysis Results Comparison of the CPO and bio-gasoline.

% Berat
Scale Up 10 L
Komponen Skala Labarotarium
300
o
C 310
o
C 320
o
C
Parafin/olefin
C1 C4 6,13 9,44 5,9 4,32
C5 6,10 0,38 0,22 0,23
C6 1,76 0,41 1,29 1,13
C7 0,84 4,30 5,27 6,78
C8 52,63 1,12 11,62 7,05
C9 3,09 5,40 9,42 4.57
C10 7,25 16,87 12,33 14,77
C11 15,68 1,53 16,98 6,41
C12 - 2,21 3,73 2,15
C13 - 3,28 4,06 2,41
C14 - 2.84 4,05 4,88
C15 - 14.10 6,71 7,5
C16 1.55 23,2 5,39 21,65
C17 - 5,45 1,81 1,81
C18 4,97 4,65 2,01 4,32
C19 - 0,15 0,09 3,69
C20 - 2,61 - -
Aromatik
C6 - - 0,13 0,46
C7 - - 0,88 0,62
C8 - - 0,43 0,67
C9 - - 2,21 0,14
C10 - 0,15 1,35 0,45
C11 - 0,39 2.39 1,59
C12 - 0,44 1,21 0,55
C13 - 0,76 0,64 0,66
C14 - 0,32 - 1,19
C15 - - - -
C16 - - 0,18 -
Physical Properties
Octane Number 122,4 106 98 98
Density (g/cm
3
) 0,77 0.84 0.81 0.84
Viscosity (poise) 0,054 0.0048 0.0167 0.0359
RVP (kPa) 48,32 - - -
Distillation Temp. (
o
C) 255 280 280 280
Yields (%) 10,3 12,2 14,2 8,4

Note : Grey boxes are bio-gasoline fractions ( around 93,48; 34,28; 70,28; dan 49,42% were respectively bio-
gasoline fraction in laboratory and pilot scale that was done at 300; 310; dan 320
o
C) .

For producing an additive to premium gasoline,
the optimum temperature of this reaction is
similar to the result in laboratory scale, it was
found at 300
o
C (catalyst/reactant weight ratio
1:75 and reaction time 1 hour). This reaction
produced 12.2% bio-gasoline in yield that had
octane number of 106 and gasoline fraction
around 34.3% (dominated with aromatic and
olefin product). The octane number of this
product was lower to the laboratory scale
product, and for using as an additive of premium
gasoline to Pertamax grade gasoline, it need an
increasing volume of up to 10%.

Conclusion
Using H-zeolite catalyst, this process
had succeeded in synthesizing gasoline-grade
hydrocarbons (bio-gasoline) through cracking
reaction even in atmospheric pressure. The bio-
gasoline produced contains C
8
to C
12
and also
aromatics, which also composing the gasoline
from crude oil, with the yield and octane number
of 11.93% and 122.4 in laboratory scale and
12.2% and 106 in pilot scale at the same process
condition. Addition of 5% bio-gasoline to
gasoline was capable to increase premium
gasoline performance similar to commercial
ISSN: 1411-1284

Proceeding 10
th

Pertamax. However, for using as an additive
gasoline to Pertamax grade gasoline, it needs an
increasing volume of additive up to 10%.

Reference
Idem, R.O., Katikaneni, S.P.R., Bakhsi, N.N.,
1997. Catalytic Conversion of Canola oil to
Fuels and Chemicals: Role of Catalyst
Acidity, Basicity, and Shape Selectivity on
Product Distribution. Fuel Process. Tech.
51, 101 - 125
Katikaneni, S.P.R., Adjaye, J.D., Idem, R.O.,
Bakshi, N.N., 1996. Catalytic Conversion of
Canola Oil over Potassium-Impregnated
HZSM-5 Catalysts: C2-C4 Olefin
Production and Model Reaction Studies. Ind.
Eng. Chem. Res. 35, 3332 - 3338
Ooi, Y.S., Zakaria, R., Mohamed, A.R., Bhatia, S.,
2004a. Catalytic Conversion of Palm Oil-
Based Fatty Acid Mixture to Liquid Fuel.
Biomass Bioenerg. 27, 477 478
Ooi, Y.S., Zakaria, R., Mohamed, A.R., Bhatia, S.,
2004b. Catalytic Cracking of Used Palm Oil
and Palm Oil Fatty Acids Mixture for The
Production of Liquid Fuel: Kinetic
Modelling. Energ. Fuels 18, 1555
Silverstein, RM., Webster FX., 1997.
Spectrometric Identification of Organic
Compounds, 6
th
Ed. John Wiley & Sons,
Inc., New York
Twaiq, F.A., Zabidi, N.A.M., Bhatia, S., 1999.
Catalytic Conversion of Palm Oil to
Hydrocarbons: Performance of Various
Zeolite Catalysts. Ind. Eng. Chem. Res. 38,
3230 - 3232
Twaiq, F.A., Zabidi, N.A.M., Mohamed, A.R.,
Bhatia, S., 2003a. Catalytic Conversion of
Palm Oil Over Mesoporous Aluminosilicate
MCM-41 for The Production of Liquid
Hydrocarbon Fuels. Fuel Process. Tech. 84,
105 - 109
Twaiq, F.A., Mohamed, A.R., Bhatia, S., 2003b.
Liquid Hydrocarbon Fuels from Palm Oil by
Catalytic Cracking over Aluminosilicate
Mesoporous Catalysts with Various Si/Al
Ratios. Micropor. Mesopor. Matl., 64, 95 -
96
Twaiq, F.A., Mohamed, A.R., Bhatia, S., 2004.
Performance of Composite Catalysts in
Palm Oil Cracking for the Production of
Liquid Fuels and Chemicals. Fuel Process.
Tech. 85, 1283 1287
Wijanarko A. and M. Nasikin, Biogasoline
production from palm oil by cracking using
NiMo/Zeolite Catalyst, Japan-Indonesia
Bilateral Symposium on Sustainable
Engineering, Yogyakarta, Indonesia, May
14
th
, 2007
Yusoff, S., 2004. Renewable Energy from Palm
Oil Innovation on Effective Utilization of
Waste. J. Cleaner Production. 14, 87 93

ISSN 1411-1284
Proceeding 10
th

Flame Lift-up on A Bunsen Burner;
A Preliminary Study

Cokorda Prapti Mahandari
*
and I Made Kartika Dhiputra
Engineering Faculty University of Indonesia Kampus UI Baru 16424

Tel. 330188, fax. 3918115
email:pmahandari@yahoo.com
*

email:dhiputra_made@yahoo.com

Abstract-Some research on ring stabilizer for lean
premixed turbulent flame has been done. In those
researches the ring mounted flush with the exit
burner port. This basic research also employed a
ring which was incorporated above the tube by a
ring adjuster. Propane and air flow rate were
metered, mixed and introduced to the bottom of
Bunsen burner tube. Propane flow rate kept
unvarying and air flow rate was increased
gradually. On some value of air flow rate, flame
would jump to the ring and it did not attach to the
exit tube any longer. This phenomenon was called
flame lift-up.
This experiment was done on several variation
of ring position. Air Fuel Ratios (AFR) for which
the flame lift-up phenomena took place was
calculated. Furthermore experiments without a
ring have also conducted to compare the flame
characteristic mainly the AFR for blow-off
condition.
It was found that flame lift-up correlate
interchange to the AFR. For the impact of ring
position, it was found that there was a particular
position that made flame lift-up stable very best.
AFR for flame lift-up was in reversed to the
position of the ring. Experiments without ring
obtain that value of AFR for flame lift-up was
above the blow off limit of flame without ring. In
addition, from the Fuidge diagram it was found
that ring did operated as a flame stabilizer.

Keywords: Flame, lift-up, Bunsen, ring

I. INTRODUCTION

nderstanding of the propagation and stability of
premixed flames not only is important for
fundamental combustion research but is also relevant
to the development of new combustion technologies
such as those associated with low-NOx emission, lean
burn, micro-scale combustion, and material synthesis.
In practical burners there are two important design
criteria that are avoided, lift-off and flashback. Lift-
off is the condition where the flame is not attached to
the burner tube or port but it is stabilized at some
distance from the port. Flash back is the occurrence
when the flame enters and propagates through the
burner tube without quenching. Flame lifting, as the
result of lift-off condition, is generally unfavorable for
a number of reasons. First it increases the escape of
unburned gas or incomplete combustion. Secondly,
above the lifting limit, ignition becomes intricate.
Furthermore accurate control of the position of a lifted
flame is difficult to manage so that it resulted in a
poor heat transfer characteristic. Lifted flames can
also be noisy [1].
However in some situations lifted flames are
encouraged as it has been shown that lifted flames can
reduce NOx levels in furnaces [2]. Damage on the
nozzle tip of burner that had occurred very often also
attracted the application of flame lifting. Concerning
on flame stability of lean premix combustion, lift-off
phenomena was took place at the most lean mixture.
There are two mainstream researches on lift-off
conditions: keeping the flame stable at lift-off
condition and reattach the flame to the burner.
Maintaining flame on lift-off condition was achieved
by employ bluff-body, ring stabilizer etc [3-5].
Reattach the flame to the burner was achieved using a
pulsed high voltage discharge [6]. This research was
in the effort of maintaining the lift-off condition stable
by introducing ring stabilizer. Research on ring
stabilizer so far was putting a ring exactly on the tube
of the burner [5]. This research was also using a ring
that introduce above the Bunsen burner tube. This
work however has to be considered as a first initiative.

A quite similar phenomenon with lift-off is
introduced. As lift-off condition keeping the flame
above the tube burner or port freely, this phenomenon
was retained the lifted flame using ring. The
occurrence when the flame jump to the ring was
named lift-up. As a preliminary study, flame lift-up
would be investigated on a Bunsen burner. The tube
was 14 mm diameter and 38 cm height. A ring with
outside diameter 30 mm, inside diameter 10 mm and
thickness 6 mm was placed above the tube burner.
Flow rate of fuel, in this case was propane and air
were metered using rotameters whenever lift-up
happened. The rotameter had been calibrated using a
wet gas meter. 8 position of ring from the exit tube
U
ISSN 1411-1284
Proceeding 10
th

were investigated. Air Fuel Ratio and burning load
was calculated than they were plotted on graphics to
get the Fuidge diagram. Understanding this
phenomenon would inspire the application of flame
lift-up in the future.

II. BASIC THEORY

There are many definitions of combustion. It
involves fuel and oxidizer. Fuel is any substance that
releases energy when oxidized. Oxidizer is any
oxygen-containing substance that reacts with fuel for
example air. Complete combustion can be achieved
by mixing the fuel and the oxidizer. To characterize
the fuel-air mixture some relevant relations are
needed. Air Fuel Ratio or AFR is defined as [7]:
fuel fuel
air air
fuel
air
n M
n M
m
m
AFR = = (1)
It is expressed in terms of mass (m), molecular weight
(M) and number of mol (n). The equivalence ratio of a
air-fuel mixture is defined as

actual
AFR
AFR
tric stoichiome
= (2)
One method to get cleaner combustion is lean
combustion or <1. Working with lean mixtures is
advantageous from a pollution point of view as low
flame temperature implies low potential for NO
generation. However the flame may become too slow
and hence prone to extinction or blow-off. Blow-off is
the condition when the burning velocity is less than
the mixture velocity.
Simple design principle of the Bunsen burner has
been incorporated in many gas appliances such as
cooking stoves and gas burner. Premixed flame from
Bunsen burners are relatively clean and give more
intense combustion with higher effective temperature.
Burning load (BL) is defined as the power of
combustion per area of burner. The equation for
burning load is

A
x HV V
BL
fuel
=
(3)

BL is burning load, W/m
2

V
f
is fuel flowrate, m
3
/s
HV is heating value of fuel, J/m
3

A is area of burner, m
2

Flame will be unstable if the reactant flow rate is
too high. This then results in the flames inability to
maintain its position, and blow-off occurs. The
entrainment of air from the surroundings will then
weaken the mixture. Flame stability can be best
described by a Fuidge diagram. Such a diagram is
reproduced in Figure 1 [8].

Figure 1. Fuidge Diagram

III. EXPERIMENTAL RESULT

Flame lift-up phenomenon was initially
discovered on ring stabilizer experiment. Ring was
placed above the Bunsen burner tube. Fuel flow rate
was fixed and air flow rate was varied until flame
jumped from the exit tube to the ring and sit on the
ring as is shown on Figure 2.

Figure 2. Flame lift-up on Bunsen Burner

This flame was completely different from the
flame on Johnson, et al research. They put a ring
exactly on the exit tube [5].

ISSN 1411-1284
Proceeding 10
th

AFR vs ring position on flame Lift-up
25.00
30.00
35.00
40.00
45.00
50.00
55.00
0 5 10 15 20 25 30 35 40 45
Ring position (mm)
A
F
R

b
y

v
o
l
u
m
e
5 mm
10 mm
15 mm
20 mm
25 mm
30 mm
35 mm
40 mm

Figure 3. Variation of AFR flame lift-up on ring position

AFR vs Burning Load flame lift-up
0
10
20
30
40
50
60
5725 6327 6930 7532 8135 8738
Burning Load, kJ/m2.s
A
F
R

b
y

v
o
l
u
m
e
'30 mm'
'40 mm'
'20 mm'
'blow-off'
'blow-off without ring'

Figure 4. AFR vs Burning Load flame lift-up

AFR and burning load when propane lift-up
happened was calculated based on Equation 1 and
Equation 3. The experiment result is presented on
Figure 3 and Figure 4. AFR
st
by volume of propane or
C
3
H
8
is 23,81 [8]. It means that the combustion was
ISSN 1411-1284
Proceeding 10
th

on lean combustion as the minimum AFR was around
40.

Figure 3 presents AFR of flame lift-up for each
position of ring from the exit tube. 8 position of ring
which were 5mm, 10 mm, 15mm, 20 mm, 25 mm, 30
mm, 35 mm and 40 mm from the exit tube were
done. It is shown that the AFR for flame lift-up was
depend on the position of the ring. As ring position is
higher, the AFR decrease slightly. It means that flow
rate of air must be higher to bring the flame sit on
the ring. Therefore flame lift-up is considered to be an
equilibrium condition of burning and mixture
velocity. Within the same position there was a certain
range of AFR. Among the eight positions it was on 5
mm and 40 mm that lift-up found to be quite difficult.
The wider AFR range implies that flame stability area
was also wider. Ring position of 20 mm and 25 mm
was the most stable among others. It indicates there
are other reasons influence the lift-up phenomenon. It
could be the turbulence and the separation flow of
propane-air.

Figure 4 described the Fuidge diagram for lift-up
flame. Compare to Fuidge diagram for methane [7],
the AFR for propane was very much higher. Note that
AFR
st
methane itself was lower than AFR
st
of
propane. In case of propane, it was 23.18 while
methane was only 9.524. The lift-up curve was above
blow-off curve without ring. In this condition ring
was indeed considered as a flame stabilizer.

IV. CONCLUSION

As a preliminary study, the stability of flame lift-
up phenomenon was investigated. It was found that
lift-up will occur on lean combustion of premixed
flame. From Fuidge Diagram flame, it was revealed
that flame lift-up area is above and blow-off area
without ring. Position of ring has been discovered
influenced very significant.
Further study on geometric of ring as a blockage
area should be deliberated. Temperature between the
exit tube and the ring becomes very striking to
explore if this phenomenon will be implemented
practically.

REFERENCES

[1] Turns, Sthepen R An Introduction to
Combustion Concepts and Applications Mc
GrawHill International, Singapore, 1996
[2] Chao, Y.C., Yuan, T., and Tseng, C.S., Effects of
flame lifting and acoustic excitation on the
reduction of NOx emissions, Combust Sci. and
Tech., vol. 114, pp. 49-65, 1996
[3] Andres A. Chaparro and Baki M. Cetegen,
Blowoff Characteristics of Bluff-body stabilized
conical premixed flames under upstream velocity
modulation, Combustion and Flame Vol 144,
Issues 1-2, pp. 318-335, 2006
[4] A. Kempf, R.P Linddtedt and Janicka, Large
Eddy Simulation of bluffbody stabilized
nonpremixed flame, Combustion and Flame Vol
144 Issued 1-2, pp. 170-189, 2006
[5] M.R. Johnson, L.W Kostiuk, R.K. Cheng A Ring
Stabilizer for Lean Premixed Turbulent Flames
Combustion Group, Energy & Environment
Division, Lawrence Berkeley Laboratory,
Berkeley, California, 94720
[6] K. Criner, A. Cesson, J. Louiche, P. Vervisch,
Stabilization of Turbulent Lifted Jet Flames
Assited by the Pulsed High Voltage Discharge
Combustion and Flame 144, pp. 422-425, 2006
[7] A C McIntosh, Laboratory Manual, MSc Course
in Combustion and Energy, University of Leeds,
1997
[8] http://www.geoff-
jones.co.uk/downloads/chemistryassignmentrepor
t
laminarpremixedflamestabilityandflammabilityan
dburnerdesign_msc.pdf, 11 November, 2007
[9] Drysdale, Dougal, An Introduction to Fire
Dynamics, John Wiley & Sons, England, 1998

ISSN: 1411-1284

Proceeding 10
th

Study of Biodiesel of Coconut and Corn Ethyl Ester
Use by Processing with the Processor Series Type
In Diesel Engine Performance

Bambang Sugiarto, Sanggul H. Siregar, Yanuar Chandra
Fac. of Engineering, University of Indonesia, Jl. Salemba Raya 4, Jakarta 10430
Fac. of Engineering, Tel. 330188, fax. 3918115
email : bangsugi@eng.ui.ac.id

Abstract Biodiesel is the one of the alternative
energy which can be renewed and environmental
friendly. Indonesia has a big potency to develop
and use biodiesel as a Diesel fuel because there is
much kind of plantation resources in it. The
process of biodiesel can be conducted with process
of transesterification. In this experiment Biodiesel
was processed from Coconut Oil and Corn oil by
using processor series type. The performance test
was conducted on Diesel Engine Research and Test
Bed with Nissan type SD 22 engine without any
modification.
The fuel are mixing between Diesel fuel and
biodiesel was variated at biodiesel contain 5 %, 10
% and 20 %. The engine speed changing are 1300,
1500, 1700 and 1900 rpm while the throttle valve
open in 30 %, 40 %, 50 % and 60 %. The testing
result showed that the opacity value decrease when
using these biodiesels.
The result also showed that biodiesel from
coconout oil and corn oil can increase the thermal
efficiency, brake horse power and decrease specific
fuel consumption of Diesel engine test especially in
variation of throttle valve open and variation of
speed engine charge. Generally, from two kinds of
biodiesel (coconut and corn ethyl ester), coconut
ethyl ester with contain 20 % mix with Diesel fuel
has the best resulted.

Keywords : Biodiesel, transesterification, emission.

I. INTRODUCTION

ncreasing of resident growth accompanied with
the expanding of industries and transportation
which is the biggest user of fuel fossil. Other side, the
an effect of this situation will make sources of fuel
fossil sharply decrease and in a moment will finished
[3]. Ironically, the situation is not followed with the
degradation of fuel usage in many countries.
One way to solve this problem is improvement and
exploiting of biodiesel. To obtain biodiesel, the
vegetable oil or animal fat is subjected to a chemical
reaction termed transesterification [6]. In that
reaction, the vegetable oil or animal fat is reacted in
the presence of a catalyst (usually a base) with an
alcohol (ethanol or methanol) to give the
corresponding alkyl esters (or for methanol, the
methyl esters). Figure 1 depicts the transesterification
reaction [2].

Figure 1 : Transesterification Reaction

Biodiesel can be produced from a great variety of
feedstocks. These feedstocks include most common
vegetable oils (e.g., soybean, cottonseed, palm,
peanut, rapeseed/canola, sunflower, safflower,
coconut, corn) and animal fats (usually tallow) as well
as waste oils (e.g., used frying oils). The choice of
feedstock depends largely on geography. Depending
on the origin and quality of the feedstock, changes to
the production process may be necessary.

Biodiesel is miscible with diesel oil in all ratios. In
many countries, this has led to the use of blends of
biodiesel with petrodiesel instead of neat biodiesel.
Blending with diesel fuel can be denoted by acronyms
such as B5, B10 and B20, which indicates a blend of
5%, 10 % and 20% biodiesel with diesel oil [5].
In this experimental, ethanol (pyroxylic spirit) is used
for producing biodiesel of Coconut Oil and Corn Oil.

Biodiesel has many advantages as follows :
It is renewable.
It is energy efficient.
It displaces petroleum derived diesel fuel.
It can be used in most diesel equipment with no or
only minor modifications.
It can reduce global warming gas emissions.
It can reduce tailpipe emissions, including air
toxics.
I
ISSN: 1411-1284

Proceeding 10
th

It is nontoxic, biodegradable, and suitable for
sensitive environments.

Blending of B5, B10 and B20 used in this
experimental without any engine modification.
Properties of fuel used :
Standard Corn
Ethyl
Ester
Coconut
Ethyl
Ester
Viscosity at 40
o
C 2.3-6 mm
2
/s 4.869 2.656
Density at 15
o
C 0.85-0.90 0.8886 0.868
Flash Point Min 100
o
C 180 104
Cloud Point Max 18
o
C 11 -5
Pour Point -
o
C - 3 10
Water content Max 0.05 % 0.02 0.02
LHV kJ/kg 37500 39500
Cetane Number Min 48 50.27 63
Free glycerol Max 0.02 % 0.0135 0.002
Total glycerol Max 0.24 % 0.3235 0.043
TAN, mg KOH/gr Max 0.8 0.2168 0.17
Ester content Min 96.5 % 99,1569 99.1676
Source : BPPT, Measured by ASTM D

Experimental Procedure
In the first case, tests of engine performance and
exhaust gas opacity on diesel oil alone were
conducted as a basis for comparison. In the second
phase, testing with B5, B10 and B20 of Coconut ethyl
ester and Corn ethyl ester. The experiment were
conducted at variation of speed 1300, 1500, 1700 and
1900 at uniform throttle valve open 40 % and
variation of throttle valve open 30%, 40%, 50% and
60% at the rated speed 1500 rpm.

The engine specification used in this experimental
are given in the table below :
Model SD-22 (Nissan Motor Co, Ltd)
Type Water cooled 4 cycle, diesel
engine
Number of cylinder 4
Diameter of
cylinder
83
Stroke 100 mm
Capacity 2163 cc
Compression ratio 22 : 1
Output power 47 PS
Speed 3200 rpm (max)

II. BASIC THEORY

The Process of Biodiesel
In this experimental, biodiesel of Coconut Oil and
Corn Oil was produced by transesterification process
with ethanol (spritus) by using processor series type
as shown in the figure 2 :

Figure 2. Processor series type BDP-10FG-BV

Processor specifications :
MODEL BDP-10FG-BV
TUBE
Material
Thick

Fiber glass
1,5 mm
Production Metode Batch
Production Capacity 10 liter/batch
ElLECTRICAL
Voltage
Frequency
Consumption

220V AC
50 Hz
350 Watt
DIMENSION
High
Width
Length

140 cm
25 cm
52 cm
Duration of process 21 hours

Engine Performance
Some parameters noted during experiment and
used as a raw data. Result of the testing datas shown
with the parameters [8] [1] of fuel consumption (FC),
specific fuel consumption (SFC), brake horse power
(BHP), thermal efficiency (
th
) and exhaut gas opacity
opacity of B5, B10, B20 of Coconut ethyl ester and
Corn ethyl ester. Furthermore, it will be compared
with diesel fuel as a comparator.

Following will be elaborated its calculation method
[8] :
Fuel Consumption
Fuel Consumption per set of time as shown with
equation below :

t
V
FC
g
. 3600
= [L/hr]

(1)

Brake Horse Power (BHP)
ISSN: 1411-1284

Proceeding 10
th

Measurable motor energy output that is transmissed
through the shaft.
60 . 1000
. . . 2 T n
BHP

= [kW]

(2)

Specific Fuel Consumption (SFC)
Fuel Consumption per Brake Horse Power
BHP
FC
SFC = [L/kW.hr]

(3)

Thermal Efficiency (
th
)
Thermal efficiency express the fuel energy
effectivity to product some work.
100
. 632
x
Q
BHP
f
th
= [%]

(4)

Opacity of exhaust gas (k %)
As known, the combustion process of raw materials
in diesel engines gives out few components of
exhaust gas and smoke. The measurement of the
emission of exhaust gas includes measurement of
the opasity of exhaust gas.
Rate of smoke concentration with the percentage of
light being accepted by concentration sensor.
Measured in percentage scale.


Specific Fuel Consumption
a. Specific fuel Consumption at the rated speed (1500
rpm) and variation throttle open (30%, 40%, 50%
and 60%)
Rated speed 1500 rpm
0.1
0.2
0.2
0.3
0.3
0.4
0.4
0.5
0.5
0.6
0.6
20 30 40 50 60 70
% Throttle Valve Open
S
F
C
[
L
/
H
P
.
h
r
]
Diesel Oil B5CCNT
B10CCNT B20CCNT
B5CORN B10CORN
B20CORN

THROTTLE DIESEL
OPEN OIL RESULT % RESULT % RESULT %
30 % 0.398 0.373 -6.4 0.379 -4.8 0.397 -0.3
40 % 0.257 0.259 0.6 0.256 -0.4 0.258 0.4
50 % 0.202 0.243 20.6 0.243 20.8 0.229 13.6
60 % 0.241 0.252 4.4 0.253 4.9 0.249 3.0
RESULT % RESULT % RESULT %
Increase 0.503 26.2 0.367 -7.8 0.369 -7.2
Decrease 0.243 -5.4 0.245 -4.5 0.254 -1.1
0.237 17.6 0.229 13.6 0.230 13.9
0.230 -4.7 0.254 5.2 0.237 -1.8
B5 CORN E. E B10 CORN E. E B20 CORN E. E
B5 COCONOUT E. E B10 COCONOUT E. E B20 COCONOUT E. E

The Specific Fuel Consumption of blends is almost
same at throttle valve open 30 % and then it decreased
at 40% of throttle valve open by average 23,5 %. The
highest decrease was happened on B5 Corn ethyl ester
amount 6,4 % at 30% of throttle valve open. The sfc
decreased at 50% and 60% of throttle valve open. The
sfc decreased with increase in bhp and increased at
50% and 60% of throttle valve open due to effect of
heating value of the blends.

b. Specific fuel Consumption at the rated throttle
valve open (40 %) and variation speed (1300, 1500,
1700 and 1900 rpm)
Throttle Valve Open (40%)
0.1
0.2
0.3
0.4
0.5
0.6
1100 1300 1500 1700 1900 2100
Rotational Speed[rpm]
S
F
C

[
L
/
H
P
.
h
r
]
Diesel Oil B5CCNT
B10CCNT B20CCNT
B5CORN B10CORN
B20CORN

RPM DIESEL
OIL RESULT % RESULT % RESULT %
1300 0.225 0.232 3.1 0.237 5.0 0.234 4.0
1500 0.250 0.262 4.6 0.257 2.6 0.262 4.5
1700 0.266 0.251 -5.8 0.268 0.8 0.292 9.9
1900 0.472 0.379 -19.6 0.489 3.6 0.488 3.6
RESULT % RESULT % RESULT %
Increase 0.230 2.0 0.231 2.3 0.235 4.4
Decrease 0.230 -8.3 0.250 -0.1 0.256 2.4
0.289 8.7 0.268 0.7 0.285 7.1
0.502 6.3 0.423 -10.2 0.490 3.9
B20 CORN E. E B10 CORN E. E
B5 CORN E. E

B5, B10 and B20 of the both biodiesels have
Specific Fuel Consumption which generally higher
than Diesel Oil SFC at up to 1700 rpm but it generally
decreased at 1900 rpm. B5 Coconut ethyl ester was
higher decrease (19.6%) at 1900 rpm. Further, all of
blend is generally show similar behavior at rated of
speed.

Brake Horse Power (BHP)
a. Brake Horse Power at the rated speed (1500 rpm)
and variation of throttle open (30%, 40%, 50% and
60%)
Brake Horse Power
2
6
10
14
18
22
26
20 30 40 50 60 70
B
H
P

[
H
P
]
Diesel Oil B5CCNT
B10CCNT B20CCNT
B5CORN B10CORN
B20CORN

THROTTLE DIESEL
30 % 5.351 5.296 -1.0 5.425 1.4 4.835 -9.7
40 % 18.637 16.423 -11.9 20.298 8.9 19.227 3.2
50 % 21.183 21.405 1.0 21.589 1.9 20.187 -4.7
60 % 22.143 22.143 0.0 - 21.774 -1.7 21.036 -5.0
Increase RESULT % RESULT % RESULT %
Decrease 3.580 -33.1 5.351 0.0 - 8.933 66.9
20.002 7.3 17.880 -4.1 31.562 69.4
20.888 -1.4 20.789 -1.9 33.766 59.4
21.940 -0.9 21.774 -1.7 34.659 56.5

The highest increase of bhp occurred at B20 of
Corn ethyl ester amount 69.9% at 40% of throttle
valve open and and B20 of Corn ethyl ester amount
66.9 % at 40% of throttle valve open. Decrease of
bhp was dominant at 60% of throttle valve open by
ISSN: 1411-1284

Proceeding 10
th

average 2.3%. Bhp increased with increase increase in
engine load.

b. Brake Horse Power (BHP) at the rated throttle valve
open (40 %) and variation speed (1300, 1500, 1700
and 1900 rpm)
Brake Horse Power
2
4
6
8
10
12
14
16
18
20
22
24
1100 1300 1500 1700 1900 2100
B
H
P

[
H
P
]
Diesel Oil B5CCNT
B10CCNT B20CCNT
B5CORN B10CORN
B20CORN

RPM DIESEL
1300 17.911 18.391 2.7 17.591 -1.8 16.552 -7.6
1500 20.851 21.036 0.9 20.667 -0.9 19.836 -4.9
1700 11.920 13.384 12.3 12.422 4.2 10.958 -8.1
1900 4.791 6.544 36.6 4.862 1.5 4.675 -2.4
Decrease 17.591 -1.8 17.591 -1.8 17.495 -2.3
20.335 -2.5 20.408 -2.1 20.482 -1.8
11.084 -7.0 12.548 5.3 11.293 -5.3
4.534 -5.4 5.703 19.0 4.558 -4.9

The highest increases of bhp were occurred of B5
of Coconut ethyl ester (36.6%) at speed 1900, B5
Coconut oil (12.3%) at speed 1700 rpm, B5 of
coconut ethyl ester (0.9%) at 1500 rpm and B5 of
Coconut ethyl ester (2.7%) at speed 1300 rpm. But,
for other blends decrease was occurred at speed 1300
rpm by average 2.2%. Bhp increased to a maximum
(up to 1500 rpm) and decreased due to decreased
engine load by effect of friction increase with engine
speed and becomes dominant at higher speeds.

Thermal Efficiency (th)
a. Thermal efficiency at the rated speed (1500 rpm)
and variation of throttle open (30%, 40%, 50% and
60%)

Thermal Efficiency
5
10
15
20
25
30
35
40
20 30 40 50 60 70

t
h
[
%
]
Diesel Oil B5CCNT
B10CCNT B20CCNT
B5CORN B10CORN
B20CORN

THROTTLE DIESEL
30 % 17.670 19.007 7.6 18.807 6.4 18.184 2.9
40 % 27.386 27.395 0.0 - 27.849 1.7 27.981 2.2
50 % 34.903 29.123 -16.6 29.280 -16.1 31.523 -9.7
60 % 29.148 28.102 -3.6 - 28.150 -3.4 29.051 -0.3
Decrease 5.777 -67.3 8.635 -51.1 8.933 -49.4
32.277 17.9 28.853 5.4 31.562 15.2
33.706 -3.4 33.545 -3.9 33.766 -3.3
35.403 21.5 35.136 20.5 34.659 18.9

The increase of thermal efficiency is most dominant at
40 % of throttle valve open by average 8.48% and
decrease is most dominant at 50% of throttle valve
open by average 7.28%. The drop in thermal
efficiency with in portion of blends must be attributed
to the combustion process.

b. Thermal efficiency at the rated throttle valve open
40 % and variation of speed (1300 rpm, 1500 rpm,
1700 rpm and 1900 rpm)

Thermal Efficiency
0
10
20
30
40
1100 1300 1500 1700 1900 2100

t
h

[
%
[
Diesel Oil B5CCNT
B10CCNT B20CCNT
B5CORN B10CORN
B20CORN

RPM DIESEL
1300 31.215 30.473 -2.4 30.115 -3.5 30.794 -1.3
1500 28.100 27.028 -3.8 27.742 -1.3 27.600 -1.8
1700 26.444 28.267 6.9 26.570 0.5 24.696 -6.6
1900 14.915 18.673 25.2 14.587 -2.2 14.782 -0.9
Decrease 30.854 -1.2 31.061 -0.5 30.970 -0.8
30.922 10.0 28.621 1.9 28.428 1.2
24.539 -7.2 26.722 1.1 25.576 -3.3
14.151 -5.1 16.907 13.4 14.864 -0.3

The better thermal efficiency was most dominated
by B5 Coconut ethyl ester at speed 1900 rpm by
amount 25.2%. Further, for B5 Coconut ethyl ester
decreased 6.9% and B10 Corn ethyl ester increased
1.1% at speed 1700 rpm. But, it generally decreased
by average 1.7% for all of blends.

Exhaust Gas Opacity (%)
a. Exhaust gas opacity at the rated speed (1500 rpm)
and variation throttle open (30%, 40%, 50% and
60%)
ISSN: 1411-1284

Proceeding 10
th

Opacity (%)
UniformSpeed at 1500 rpm
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
20 30 40 50 60 70
k
Diesel Oil B5CCNT
B10CCNT B20CCNT
B5CORN B10CORN
B20CORN

THROTTLE DIESEL
30 % 0.096 0.069 -28.3 0.055 -42.9 0.055 -42.9
40 % 0.180 0.096 -46.7 0.139 -22.8 0.100 -44.7
50 % 0.200 0.122 -38.8 0.122 -39.1 0.101 -49.4
60 % 0.219 0.148 -32.4 0.141 -35.6 0.110 -49.8
Decrease 0.065 -31.9 0.052 -46.1 0.044 -54.5
0.143 -20.6 0.113 -37.5 0.115 -36.4
0.142 -29.1 0.127 -36.6 0.117 -41.4
0.142 -35.2 0.159 -27.6 0.140 -36.3

All of blends have lower opacity than diesel oil
which B20 Corn Oil has highest decrease (54.5%) at
30% of throttle valve open, B5 Coconut Oil opacity
decreased 46.7% at 40% of throttle valve open, B20
Coconout Oil opacity decreased 49.4% at 50% of
throttle valve open and B20 Coconut Oil opacity
decreased 49.8% at 60% of throttle valve open.
Furthermore, all of blends decreased opacity by
average 46.7% as compared diesel fuel. The
biodiesels/ethyl ester fuels produces less smoke,
probably because the biodiesel contains oxygen, high
cetane number which helps the good combustion in
the cylinder.

b. Exhaust gas opacity (%) at rated throttle valve open
(40 %) and variation speed (1300, 1500, 1700 and
1900 rpm)
Opacity (%)
UniformThrottle Valve Open - 40 %
0.000
0.050
0.100
0.150
0.200
0.250
0.300
1100 1300 1500 1700 1900 2100
Rotational speed(rpm)
k
Diesel Oil B5CCNT
B10CCNT B20CCNT
B5CORN B10CORN
B20CORN

RPM DIESEL
1300 0.182 0.127 -30.0 0.131 -27.8 0.104 -42.7
1500 0.174 0.140 -19.8 0.132 -24.4 0.102 -41.4
1700 0.112 0.066 -41.1 0.074 -34.4 0.069 -38.8
1900 0.108 0.075 -30.2 0.079 -26.5 0.079 -26.5
Decrease 0.118 -35.0 0.127 -30.0 0.107 -41.0
0.124 -29.0 0.137 -21.3 0.139 -20.1
0.065 -42.4 0.068 -39.3 0.054 -51.8
0.073 -32.6 0.073 -32.6 0.058 -46.0
B5 CORN E. E
B20 CORN E. E B10 CORN E. E

All of blends have percentage of opacity lower than
diesel oil opacity. B20 Coconout ethyl ester opacity
has highest decrease (42.7%) at speed 1300 rpm, B20
Coconut ethyl ester opacity decreased 41.4% at speed
1500 rpm, B20 Corn ethyl ester opacity decreased
51.8% at speed 1700 and B20 Corn ethyl ester opacity
decreased 46% at speed 1900 rpm. Furthermore, all of
blends decreased opacity by average 39.7% as
compared diesel fuel.

IV. CONCLUSIONS
1. Effect of biodiesel blends for the specific fuel
consumption were almost similar at up to 1500 rpm
but it become increased at speed 1700 to 1900 rpm
which increase by average 16.7%.
2. The highest increase of bhp occurred for B20 Corn
ethyl ester (66.9%) at 30% of throttle valve open
and decreased at 60% of throttle open. The bhp
change was resulted by engine load.
3. The thermal efficiecy increased by average 8.48%
for the blends at 40% throttle valve open and most
dominant decreased at 50% of throttle valve open
(7.28%). While, for the rated speed, the thermal
efficiency increased at 1700 rpm and 1900 rpm
which B5 Coconut ethyl ester was higher (25.2%)
at 1900 rpm and B10 Corn ethyl ester increased
1.1% at 1700 rpm. The drop in thermal efficiency
related to combustion process.
4. Using both biodiesels (Coconut ethyl ester and
Corn ethyl ester), exhaust opacity were lowest (as
compared to diesel fuel), where it decreased by
average 46.7% at the rated speed (1500 rpm) and
39.7% at 40 % throttle valve open. The both
biodiesels contain oxygen, high cetane number
which helps the good combustion in the cylinder.
5. B20 of Corn ethyl ester represent the best
composition for all blends.

REFERENCES

[1] Willard W. Pulkrabek, 2004, Engineering Fundamentals
of The Internal Combustion Engine-Second Edition,
Pearson-Prentice Hall, New Jersey 07458.
[2] Zulkarnain, 2007, Studi Pengaruh Karakteristik
Biodiesel Minyak Goreng Curah (Sawit) dan Minyak
Jagung Dengan Pereaksi Spritus dan Prosessor Jenis
Sususn Terhadap Performa Mesin, Tesis, UI, Jakarta.
[3] Kepmen ESDM No. 0002, 2004, Kebijakan Energi
Terbarukan dan Konservasi Energi, Jakarta.
[4] Heywood, John B, Internal Combustion Engine
Fundamental, McGraw-Hill International Edition.
[5] US department of Energy, Energy and Renewable
Energy, Biodiesel Handling and Use Guideline, 2004
[6] Gerhard Knothe. Et al, The Biodiesel Hand Book,
National
Center for Agricultural Utilization Research
Agricultural
Research Service U.S. Department of Agriculture
Peoria,
Illinois, U.S.A, 2005
[7] Gerhard Knothe. Et al, The Biodiesel, The Use of
Vegetable
ISSN: 1411-1284

Proceeding 10
th

Oils and Their Derivative Diesel Fuels; National Center
for
Agricultural Utilization Research Agricultural Research
Service U.S. Department of Agriculture Peoria, Illinois,
U.S.A.
[8] Nissan D. Motor CO. Ltd, Guide and Manual Book of
Diesel
Engine Test Bed, Japan.

ISSN: 1411-1284

Proceeding 10
th

Abstract This paper examines the smoke production rate
of tropical woods (i.e. Kayu Jati and Kayu Kamper) in the
Cone Calorimeter. The samples were exposed to a constant
incident heat flux of 27 kW/m
2
The smoke opacity
measurements obtained from the Cone Calorimeter were
used to derive the rate of smoke production (RSP) of the
materials. Since, the mass of the sample also simultaneously
measured, the smoke extinction are per unit mass of sample
consumed can also be deduced. This quantity, the so called
SEA (smoke extinction area) is an important parameter in
considering the fire detection and reduced visibility the
occupants of a building in fire. This paper also discusses the
SEA under pyrolysis and flaming conditions. The outcomes
are in good agreements with those reported in the
literatures.

Keywords Smoke opacity, smoke production rate, fire
safety engineering, tropical wood.

I. INTRODUCTION

The early fire behavior of building products is
important for many aspects of fire safety engineering. The
heat release rate is a fundamental variable of fire with
which almost all other emission properties are highly
correlated [1]. Another important descriptor of a fire is the
smoke production rate. Smoke is produced in almost all
fires and presents a major hazards to life. The production
of smoke and its optical properties are often measured
separately based on related standard methods.
Wood and its derived products remain the most
popular materials for furniture used in domestic premises.
In the event of fire, wood based products contribute
significantly in the initiation and growth of fire and the
level of fire severity. The capability to predict the burning
rate of wood in modern times has become increasingly
important as fire safety engineering moves toward a
performance-based approach to building design [1,2].
A fire calorimeter or also known as a cone
calorimeter has been developed in Department of
Mechanical Engineering University of Indonesia. The
fire calorimeter is an integrated small-scale instrument to
measure important fire safety parameters of wood and
other building products. The instrument consists of a
heater of a truncated cone shape and is capable of
providing heat fluxes to the specimen in the range of 0
30 kW/m2, a digital electronic balance, an electric igniter,
thermocouples and a temperature control system, a flow
rate measurement apparatus, a sample holder and a
ducting systems. Calibration of heat flux as a function of
heater temperature is performed with a heat flux meter. At
a sufficient distance a gas sample is taken using a portable
gas analyzer to record O2, CO and CO2 levels [3].
An extensive research work has been carried out to
study fire safety parameters of tropical woods and other
building materials, such as the effect of sample species,
sample thicknesses, grain orientation, fire spread by
means of time to ignition, heat release rate, and mass loss
rate [1-3]. The design of the fire calorimeter also allows
other measurements to be taken as long as the analyzers
are available.
This paper presents the results of smoke property
measurement using the fire calorimeter. In this work,
smoke opacity measurement was carried out using a
recently purchased smoke meter apparatus available in
Thermodynamics laboratory of Department of Mechanical
Engineering University of Indonesia. The focus of this
paper is to evaluate the smoke production rate of some
tropical wood species. A better understanding of the
pattern of smoke production and yield will be one of the
key parameter towards the implementation of
performance-based fire protection system in building and
plant designs.

II. THEORY

Smoke is generated in both smoldering and flaming
combustions. Smoke from smoldering is originated from
chemical degradation and volatiles evolution of carbon
based material when heated to certain temperatures. It
consists of high molecular weight fractions of condense
aerosol / droplets of tar and high boiling liquids.
Meanwhile, smoke from flaming combustion is almost
entirely consists of solid particles [4].
Smoke emission is one of the basic elements for
characterizing a fire environment. The combustion
conditions under which smoke is produced-flaming,
pyrolysis, and smoldering-affect the amount and character
of the smoke. The smoke emission from a flame
represents a balance between growth processes in the

On the measurement of smoke production rate
of tropical wood

Yulianto Sulistyo Nugroho

Department of Mechanical Engineering, University of Indonesia
Kampus UI Depok 16424, INDONESIA
Ph. +62 (021) 7270032, Fax. +62 (021) 7270033, e-mail: yulianto.nugroho@ui.edu

ISSN: 1411-1284

Proceeding 10
th

fuel-rich portion of the flame and burnout with oxygen.
While it is not possible at the present time to predict the
smoke emission as a function of fuel chemistry and
combustion conditions, it is known that an aromatic
polymer, such as polystyrene, produces more smoke than
hydrocarbons with single carbon-carbon bonds, such as
polypropylene. The smoke produced in flaming
combustion tends to have a large content of elemental
(graphitic) carbon [5].
The smoke emission, together with the flow pattern,
determines the smoke concentration as smoke moves
throughout a building. The effects of the smoke produced
by a fire depend on the amount of smoke produced and on
the properties of the smoke. It is the combination of
obscuration and toxicity that present the greatest threat to
the occupants of a building involved in fire. Most of the
fatalities can be attributed to the inhalation of smoke and
toxic gas. The effect of reduced visibility is to delay
escape and increase the duration of exposure of the
occupants of a building to the product of combustion. In
the early stages of a fire, smoke production rate is relevant
to fire detection.
The smoke properties of primary interest to the fire
community are light extinction, visibility, and detection.
However, the most widely measured smoke property is
the light extinction coefficient. The yield of particulate
smoke from a burning material may be assessed by some
method: (i) filtering the smoke and determining the
weight of particulate matter, (ii) collecting the smoke in a
known volume and determining its optical density, and
(iii) allowing the smoke to flow along a duct, measuring
its optical density where plug flow has been established.
Optical density may be determined by measuring the
attenuation of a beam of light passing through the smoke
[4].
The physical basis for light extinction measurements
is the Lambert-Beer Law. In essence, the law states that
there is a logarithmic dependence between the
transmission of light through a substance and the
concentration of the substance, and also between the
transmission and the length of material that the light
travels through.

Fig. 1 Absorption of a beam of light as it travels through the
smoke container, adapted from [6].
According to the Lambert-Beer Law in the absence of
smoke the intensity of light falling on the photocell is (I
o
),
then in the presence of smoke, the reduced intensity (I)
will be given by
( ) kCL I I
o
= exp (1)
where k is the extinction coefficient, C is the mass
concentration of smoke particles and L is the path length
of the optical beam passing through the smoke. Optical
density, D
e
is then defined in natural logarithm as follows
[4-7]:
kCL
I
I
D
o
e
=
= ln (2)
Smoke extinction coefficient, k, (m
-1
) can be
determined by a smoke meter as follows :

=
I
I
L
k
o
ln
1
(3)
Concentration of smoke can also be expressed as
percentage obscuration or smoke opacity (%), given by:
% 100

o
o
I
I I
(4)
The rate of smoke production, RSP (m
2
/s), can be
determined by

=
a
s
s
T
T
V k RSP
&
(5)

where
s
V
&
is the volume flow rate of exhaust gas (m
3
/s), k
is smoke extinction coefficient (m
-1
), T
s
is the gas
temperature in the exhaust duct (K), and T
a
is the ambient
temperature (K).
In a fire calorimeter, the volume flow rate of exhaust
gas,
s
V
&
(m
3
/s) can be estimated using the following
expression [9]:

51 , 0
3
10 5 , 2
=

p
T
T
V
s
a
s
&
(6)
Since in the fire calorimeter assessment, one also
measure the mass loss rate, thus a quantity called as
specific extinction area, SEA (m
2
/kg) can also be defined
as :
ISSN: 1411-1284

Proceeding 10
th

m
RSP SEA
&
1
= (7)
=
kg
m
kg
s
s
m
2 2
=

m T
T
V k SEA
a
s
s
&
&
1
= (8)
The SEA represents the smoke extinction area per
mass of fuel burned.

III. EXPERIMENTAL

3.1 Experimental apparatus

This work was carried out using a fire calorimeter or
also known as a cone calorimeter developed in
Department of Mechanical Engineering University of
Indonesia. The fire calorimeter is a small-scale
instrument to measure ignition time, heat release rate and
smoke production of wood and other building products.
The instrument consists of a heater of a truncated cone
shape and is capable of providing heat fluxes to the
specimen in the range of 0 30 kW/m
2
, a digital
electronic balance, an electric igniter, thermocouples and
a temperature control system, a flow rate measurement
apparatus, a sample holder and a ducting systems.
Calibration of heat flux as a function of heater
temperature is performed with a heat flux meter. At a
sufficient distance a gas sample is taken using a portable
gas analyzer to record O
2
, CO and CO
2
levels. A portable
smoke meter is used to measure the smoke opacity of the
exhaust gas.
The basic design and working principle of the
apparatus was based on the cone calorimeter standard -
ASTM E 1354 [7]. The schematic diagram of the fire
calorimeter was given in other publication by the author
[3].

3.2 Sample preparation

Two tropical wood samples, i.e. Kayu Jati and Kayu
Kamper are used in this work. The samples are prepared
with sides of 10 mm by 10 mm and 16 mm thicknesses.
The wood samples were prepared with an along grain
orientation (Fig. 1). All specimens are conditioned to
constant mass at room temperature and relative humidity.
The properties of the samples are presented in Table 1.

Fig. 2 The specimens are prepared in along grain orientation.

Table 1 Properties of the samples used

Specimen Density,
(kgm
-3
)
Kayu Kamper 621
Kayu Jati 761

3.3 Procedure

A typical run was initiated by supplying electric
power to the heater which is controlled by an electronic
temperature controller using a chromel-alumel
thermocouple of type K. The fan speed is controlled by an
inverter and set at 13,3 Hz to give a constant volume flow
rate. The temperature is set to a certain value to provide a
desired heat flux level. Once the temperature is stable, and
other equipments, i.e. gas analyzer, smoke meter and
digital electronic balance are ready, the sample is inserted
to sample holder. The sample is protected from initial heat
flux from the heater using a radiation shield. The
experiment and data collection are started by opening the
radiation shield. The transient data of temperatures,
sample mass loss, exhaust gas analysis, smoke opacity
measurement are recorded simultaneously through a
DAQ system connected to a personal computer. In this
study, the measurement are focused on the smoke
produced and mass loss measurements. The samples are
exposed vertically to a constant incident heat flux of 27
kW/m
2
.

IV. RESULTS AND DISCUSSION

Fig. 3 Smoke opacity (%) and mass loss rate (gr/s) of Jati
wood measured using fire calorimeter and smoke meter.

Mass loss rate and smoke opacity curves
0 200 400 600 800 1000 1200
0.00
0.02
0.04
0.06
0.08
0.10
0.12
S
m
o
k
e

o
p
a
c
i
t
y

(
%
)
Mass loss rate (g/s)
M
a
s
s

l
o
s
s

r
a
t
e

(
g
/
s
)
Time (s)
0
5
10
15
20 Kayu Jati
Smoke opacity (%)
100 mm
100 mm 10 30 mm
Incident heat flux perpendicular to grain
Along grain
100 mm
100 mm 10 30 mm
100 mm
100 mm 10 30 mm
Along grain
16 mm
100 mm
100 mm 10 30 mm
Along grain
100 mm
100 mm 10 30 mm
100 mm
100 mm 10 30 mm
Along grain
16 mm
ISSN: 1411-1284

Proceeding 10
th

Fig. 3 shows a plot of sample mass loss rate (g/s) and
smoke opacity (%) against time for Kayu Jati sample. It
clearly shows that the first peak of the mass loss curve
occur soon after the sample was ignited. Similar pattern
was found for Kayu Kamper sample (Fig. 4). It indicates
that both wood species have similar physical responds to
incident heating. Meanwhile, Kayu Jati has greater mass
loss rate than Kayu Kamper.

0 200 400 600 800 1000 1200
0.00
0.02
0.04
0.06
0.08
Kayu Kamper
S
m
o
k
e

o
p
a
c
i
t
y

(
%
)
M
a
s
s

l
o
s
s

r
a
t
e

(
g
/
s
)
Time (s)
0
2
4
6
8
10
12
Smoke opacity (%)

Fig. 4 Smoke opacity (%) and mass loss rate (g/s) of Kayu
Kamper measured using fire calorimeter and smoke meter.

As shown on Figs 3 and 4, there was delay on the
smoke opacity measurement data compared to mass loss
rate curves. This is due to differences in probes locations.
The mass of the sample was measured directly below the
sample, meanwhile the smoke probe was located in the
exhaust duct about 1 m away from the sample. A single
peak was identified for both smoke opacity curves, which
is consistent with the mass loss rate patterns.

Smoke extinction area curves

According to Eq. 8, the values of smoke extinction
area, SEA (m
2
/g) for both samples were plotted on Figs. 5
and 6. Again, the mass loss curves were also given on the
same figures to show any possible correlation between the
two curves. It is observed that the Kayu Jati sample has
greater values of SEA, especially in the earlier time, as
clearly shown on Fig. 7.
On the basis of the mass loss rate which has
proportional correlation with the heat release rate [3], the
curves on Fig. 7 can be divided into two sections. The
first section is representing the induction, moisture and
volatile release phase. This phase is known as pyrolisis
conditions where the smoke is relatively darker compared
to the flaming conditions. As reported in the literature the
smoke conversion factor of wood and plastic materials
under pyrolysis conditions has greater values that during
flaming combustion [5]. The smoke produced in flaming
combustion tends to have a large content of elemental
carbon. In fact Kayu Jati has higher density value than
Kayu Kamper (Table 1).
In practice, greater value of smoke extinction area,
SEA means reduction in visibility. As we all concerns,
visibility of exit signs, doors, and windows can be of great
importance to an individual attempting to survive a fire,
and plays important role on fire safety engineering
consideration during building and plant design.

Fig. 5 Smoke extinction area (SEA) (m
2
/g) and mass loss rate
(g/s) of Kayu Jati calculated according to Eq. 8.

Fig. 6 Smoke extinction area (SEA) (m
2
/g) and mass loss rate
(g/s) of Kayu Kamper calculated according to Eq. 8.

0 200 400 600 800 1000 1200
0.00
0.02
0.04
0.06
0.08
Kayu Kamper
M
a
s
s

l
o
s
s

r
a
t
e

(
g
/
s
)
Time (s)
0.0
0.5
1.0
1.5
2.0
2.5
S
E
A

(
m
2
/
g
)
Smoke extinction area, SEA (m
2
/g)
0 200 400 600 800 1000 1200
0.00
0.02
0.04
0.06
0.08
0.10
0.12
M
a
s
s

l
o
s
s

r
a
t
e

(
g
/
s
)
Time (s)
0
2
4
6
8
10
12
Kayu Jati
S
E
A

(
m
2
/
g
)
Smoke extinction area, SEA (m
2
/g)
ISSN: 1411-1284

Proceeding 10
th

0 200 400 600 800 1000
0
2
4
6
8
10
12
Kayu Jati
Kayu Kamper
S
E
A

(
m
2
/
g
)
Time (s)

Fig. 7 Comparison of SEA values for Kayu Jati and Kayu
Kamper.

V. CONCLUSION

The results of this study show that the smoke opacity
patterns correlate with the mass loss rate of the samples
tested. The rate of smoke production in the earlier stages
of combustion, i.e. pyrolysis condition is relatively higher
than that of flaming condition. In general, the value of
smoke extinction area (SEA) of Kayu Jati is greater than
Kayu Kampers. In practice, greater value of smoke
extinction area, SEA means reduction in visibility. This is
of our concern on maintaining the visibility of exit signs,
doors, and windows during a fire accident.
This work shows that the fire calorimeter apparatus
can be used effectively in measurement of smoke
properties.

ACKNOWLEDGMENT

This work has been funded by the Ministry of
National Education Republic of Indonesia through Hibah
Bersaing XIII (2005-2007) Research Fund. Thanks to Mr.
M. Andy Fardiansyah and Mr. Candra Gupta Marihot for
preparation of the samples and the smoke opacity
measurements.

REFERENCES

[1] Quintiere, J. G., A Theoretical Basis for
Flammability Properties, Fire and Materials, 30,
175-214, 2006.
[2] Nugroho, Y.S., Soesanto, and E. Puspiartono, Effect
of Sample Orientation and Thickness on Piloted
Ignition of Timber, Proceedings of the Intl.
Conference on Fluid and Thermal Energy Conversion
2006, Jakarta, Dec. 10 14, 2006, ISSN 0854 9346.
[3] Nugroho, Y.S., Junaedi, A., and Kiswanto, G.,
Pengembangan Kalorimeter Api untuk Karakterisasi
Sifat Bakar Material, Jurnal Teknologi, 4, 294-301,
2005.
[4] Drysdale, D., An Introduction to Fire Dynamics, 2nd
Edition, John Wiley & Sons, John Wiley & Sons,
2003.
[5] Mulholland, GW, (2002), Smoke Production and
Properties, Chapter 13, The SFPE Handbook of Fire
Protection Engineering, 3rd Edition, pp. 2-258 2-
268.
[6] Wikipedia, website: http://en.wikipedia.org/wiki/
Beer-Lambert_law (seen on 21
st
November 2007).
[7] ASTM E 1354-97, Standard Test Method for Heat
and Visible Smoke Release Rates for Materials and
Products Using an Oxygen Consumption
Calorimeter, Annual Book of ASTM Standards,
American Society for Testing Materials,
Philadelphia, 1997.
[8] Myllymaki, J and Baroudi, D, Prediction of smoke
production and heat release rate by convolution
model, Nordtest Technical Report 1297-96, VTT
Building Technology, Finland, 1999.
ISSN: 1411-1284

Proceeding 10
th

Catalytic Ozonation Of Endosulfan
In Water With Activated Carbon

Enjarlis,
a,b
Setijo Bismo,
a
Slamet,
a
Roekmijati W. Soemantojo
a

a
Department of Chemical Engineering, University of Indonesia, Depok16424 Tel. 7863515
b
Department of Chemical Engineering, Institute Technology of Indonesia, Serpong Tangerang 15311 Tel.
7561092 .fax7561092.e-mail: en_jarlis@yahoo.com

Abstract The degradation of endosulfan by catalytic
ozonation with activated carbon as catalyst was
investigated at neutral pH and different
temperatures. The aim of this work is to study the
effect of the addition of activated carbon on the
degradation rate of endosulfan, focusing on the rate
constant (k) and the activation energy (Ea).
Endosulfan was selected here as target organic
pollutant due to its included organochlorine pesticide
(OCPs) may cause serious environmental
concern.The presence of the activated carbon as
catalyst is effect slightly in enhancing ozonation rate
of endosulfanan compared to non catalytic
ozonation. The rate constant (k) of endosulfan were
11.54 x 10
-2
min
-1
at 30
o
C for catalytic ozonation, and
10.27 x 10
-2
min
-1
for non catalytic ozonation. The
activation energy for catalytic ozonation was 6.58
kcal/mol and for non catalytic ozonation was
8,631kcal/mol.

Keywords Endosulfanan, Catalytic ozonation,
Activated carbon.

I. INTRODUCTION

atalytic ozonation with activated carbon as catalyst can
be accelerated ozone decomposition to form OH
radical
, thereby increasing even further the removal efficiency
of this because OH is powerfull, non-selective chemical
oxidant, which acts very rapidly with most organic
compounds [1, 2]. It has been reported that hydroxyl
radicals were generated by combinations of ozone plus
UV radiation, UV radiation plus H
2
O
2
, UV radiation
plus Fentons reagent (photo-Fenton system) [3] and by
mixing a few milligrams of activated carbon in ozone-
containing water [4]. With such context, this study was
investigated the degradation of endosulfan in water by
catalytic ozonation at neutral pH and different
temperatures, which is focused on (i) determining
possible diffusion limitation (ii) determining the rate
constants (k) and activation energy (Ea).
Endosulfan was selected here as target organic pollutant
due to its paradigmatic endocrine discruptors included
organochlorine pesticide (OCPs) may cause serious
environmental concern and health problem in animals,
including humans [5]. According to free radical reaction
theory, OH will attack chlorinated organic compounds
by hydrogen abstraction or electron transfer. Then, the
organic radical will decompose further to chlorinated
intermediates. These intermediates are eventually
oxidized by OH to final products; organic acid and
even carbon dioxide [6].

II. BASIC THEORY

The degradation of contaminant organic in presence
of activated carbon is a combination of competing
homogeneous and heterogeneous reaction: direct
reaction of molecular ozone, and an indirect reaction
involving non selective free-radicals. Both reactions take
place in the solution bulk and on activated carbon
surface [7]. Therefore, the total degradation rate of
contaminant organic in the presence of activated carbon
can be defined as the sum of the homogeneous reaction
rate, (-r
C
)
homogen,
calculated in the absence of activated
carbon, and the heterogeneous reaction rate, (-r
C
)
heterogen
due to the presence of activated carbon []. The total
endosulfan degradation rate can be mathematically
expressed as:
r r r Hete E Homo E Total E . . .
+ =

(1)

As discussed earlier the activated carbon enhanced the
oxidation rate of contaminant by promoting the
reduction of dissolved ozone into hydroxyl radicals that
cause the endosulfan degradation, the rate of equation
given in Tabel 1. Recently, Jans and Hoigne [1] pointed
out that catalytic ozone decomposition can be
cataloqued as another Advance Processes Oxidation
(AOP) with a stoichiometric ratio (ozone
decomposed/hyroxyl radical formed) equal to that of the
other ozone involved in AOPs.

Tabel 1. Rate of equation at degradation of endosulfan
by ozonation in the presence activated carbon

N
o
With activated carbon With out activated carbon
1 (- dC
E
/dt)
hete
= f
1
(C
E
, C
O3,

C
OH
, C
Ac
)
(- dCE/dt)
homo
= f
4
(C
E
,
C
O3,
C
OH
)
C
ISSN: 1411-1284

Proceeding 10
th

2 (- dC
O3
/dt)
hete
= f
2
(C
E
, C
O3,

C
OH,
C
AC
)
(- dC
O3
/dt)
homo
= f
5
(C
E
,
C
O3,
C
OH
)
3 (- dC
OH
/dt)
hete
= f
3
(C
E
, C
O3,

C
OH
, C
AC
)
(- dC
OH
/dt)
homo
= f
6
(C
E
,
C
O3,
C
OH
)


Endosulfan (1,2,3,4,7,7-hexachlorobicyclo-2,2,1
heptene 2,3-bis-hydroxy methane-5,6 sulfite) was
obtained from Chem. Service West Chester with purity
95 %. The endosulfan solution was prepared by
deionozed water obtained from Aquatron Auto Still
Yamato Tipe W-182. The specific surface area of
activated carbon were measured using the multipoint
BET of N
2
adsorption in a Quantachrome Autosorb -6
with surface area 6.87.10
2
m
2
/g. Ozone was produced
by a RS 09805 ozone generator with maximum ozone
production capacity of 0.25 g of O
3
/h. The experimental
instrument consists of an ozone generator, a cylindrical
glass column reactor with an external jacket surrounded
and a water stream which was pumped from
thermostatic bath to maintain the temperature at the
selected value for each experiment. The dimension is of
the reactor is 450 mm high with ID 40 mm which
equiped by inlet diffuser for bubbling the gas mixture,
outlet gas, sampling port and magnetic stirrer. Once the
experiment was started, the air-ozone mixture was fed
into the flasks (KI solution) in order to determine the
ozone concentration in the gas form.

The reactor was filled with 300 ml demineralized-water
and the pH was adjusted to 7. The temperature was set
for 20, 25 and 30
o
C, until a predetermined of water was
saturated with ozone in excess continuously by injecting
ozone gas for 30 min. The process is followed by the
addition 0.5 g of activated carbon (for the catalytic
process) and 4.5 x 10
-5
mol/L endosulfan in solution.
The concentration of endosulfan and ozone presence in
the system was measured at 0, 3, 6, 10, and 15 min of
treatment. The dissolved ozone concentration in aqueous
solution was determined by iodometric methods, which
endosulfan was analyzed by Gas Chromatograph type
4C, colum silicone ov-17 3 meters, ECD (Electron
Capture Detector), Shimadzu, solvent n-hexsane and
mobile gas-phase N
2.
All experiments were carried out
in duplo the presence and the absence of activated
carbon.

The influence of variables on the degradation of
endosulfan were first assessed. Non-catalytic (blank
experiment) and catalytic experiments were conducted
for each experimental series. Since the catalytic
ozonation process involves steps of external and internal
diffusion, variables studied were: activated carbon
particle size and agitation speed. From experimental
result agitation speed equal or higher than 750 rpm and
carbon particle equal or lower 0.2 -0.3 mm did not affect
the ozonation rate. Hence, at these conditions, it can be
assumed that external mass transfer rate and internal
diffusion rate of ozone through pores did not control the
catalyst ozone decomposition [9]. Therefore, agitation
and activated carbon particle particle were kept at 750
rpm and 0.2-0.3 mm size. At these conditions, the
kinetics should only be depended on the chemical
surface ozone decomposition reaction [9]. Next, a series
of experiments were then carried out at different
temperatures to determining Ea from k.

Endosulfan degradation

Fig. 1 shows the influence of the presence of activated
carbon as catalyst for the degradation rate of endosulfan
by catalytic and non-catalytic ozonation. From that
figure, it can be seen that the degradation rate of
endosulfan increased with the presence of activated
carbon compared to the absence of activated carbon.
This is because, the presence of activated carbon can
accelerate ozone decomposition to form OH
radicals
which have a high reactivity to most organic
micropollutants [1, 2].
0,5
1,0
1,5
2,0
2,5
3,0
3,5
4,0
4,5
5,0
0 3 6 9 12 15
time, min
C
E
.
1
0
. -
5

m
o
l
/
L
Fig. 1. Endosulfanan degradation rate by ozonation
processes (x) non-catalytic (blank experiment)
()catalytic ozonation at condition: pH 7; temperature
20
o
C; agitation speed 750 min
-1
;activated carbon particle
0.2-0.3 mm C
O3
6.372 x 10
-3
O
3
mol/L; CE
o
5.0 x 10
-
5
mol/L

Influence of temperature on the degradation of
carbofuran

Fig.2 shows the effect of temperature on degradation of
endosulfan for catalytic ozonation at pH 7. From that
figure, it can be stated that the higher the temperature is,
the higher is the degradation rate of endosulfan. This
effect is due to an increase in the rate constant of the
chemical reaction. Yazgan et al [10] reported that the
change of temperature gave positive effect on the
removal rate of the endosulfan. Although increasing
ISSN: 1411-1284

Proceeding 10
th

temperature causes decreasing in dissolved ozone
concentration (data are not shown), a higher oxidation
rate was obtained at higher temperatures. This
phenomenon can be explained by the increasing
diffusion coefficient at higher temperature [10].
0
1
2
3
4
5
6
0 3 6 9 12 15
time, min
C
E
,
x
1
0
-
5
,
m
o
l
/
L

Fig.2 The effect of temperature on degradation of
endosulfan by catalytic ozonation at condition: pH 7;
agitation speed 750 min
-1
;activated carbon particle 0.2-
0.3 mm C
O3
6.372 x 10
-3
O
3
mol/L; C
Eo
5.0 x 10
-5
mol/L;
temperature 20
o
C, 25
o
C, 30
o
C

Kinetics and Mechanism

The observed endosulfan degradation in the presence of
activated carbon results in a combination of competing
homogeneous and heterogeneous reaction: direct
reaction of molecular ozone, and an indirect reaction
involving non selective free-radicals [7]. The
degradation kinetics of endosulfan by homogeneous
reaction in terms of molecular ozone and hydroxyl
radical that is produced by the decomposition of ozone
at neutral pH can be formulated as follows:
Degradation rate of endosulfan homogeneous reaction:
+ =
OH E
E
O E
E
o
E
C C k C C k
dt
dC
. . . .
2 3 1
hom
.

(2)
Degradation rate of endosulfan by heterogeneous
reaction:

AC OH E
E
AC O E
E
AC E
E
hete
E
C C C k C C C k C C k
dt
dC
. . . . . . .
5 3 4 3
.
+ + =

(3)

Thus, since of ozone concentration in solution in excess
and the constant amount of activated carbon, the
amount of OH and O
3
stayed constant during the
processes (i.e., at a steady state). Therefore, Eq. (1), (2)
and (3) can be rearranged to the pseudo-first-order
equation in Eq. (4) as follows:

( )
E
E
Hete
E
Homo
E
Hete
E
Homo
E
C k k
dt
d
r r
C
Total
. + = + =
(4)

with,
+ =
OH
E
O
E E
Homo
C k C k k . .
2 3 1

(5)

and ,
+ + =
OH
E
O
E
AC
E E
Hete
C k C k C k k . . .
5 3 4 3

(6)

Integrating, with initial condition C
E
= C
Eo
leads to:

t k t k k
C
C
E
Total
E
Hete
E
Homo
E
Eo
. ) ( ln = + =
(7) with, .
E
Hete
E
Homo
E
Total
k k k + =
(8)

where k
E
total
represents the apparent pseudo-first-order
total rate constant for ozonation process in presence
activated carbon. The degradation endosulfan by
homogeneous reaction non catalytic as follows:

t k
C
C
E
Homo
E
Eo
. ln =
(9)

Figure 3 and 4 describe the plot of ln C
Eo
/C
E
) versus
time (t) with slope as k
E

total
and k
E

Homo
in various
temperatures by catalytic and non catalytic oznation,
respectively. Insert of Fig. 3 and 4 shows, typical
Arrehnius plot for ln k

against 1/T to endosulfan in
catalytic and non catalytic ozonation processes,
respectively. This linear regression indicate R
2
= 0.99.
From that figure, it can be stated that the higher the
temperature is, the higher is the degradation rate. This
effect is due to an increase in the constant rate of the
chemical reaction. Similar phenomenon was reported by
many researchers [10, 11].

The degradation of endosulfan in the presence of
activated carbon slightly significantly enhancing the
degradation rate of endosulfan with k
E
total
was 11.54 x
10
-2
min
-1
and Ea 6.58 kcal/mol at 30
o
C compared to
non-catalytic ozonation with k
E
Homo
was 10.27 x 10
-
2
min
-1
and Ea 8.631 kcal/mol. This is because of the
using of activated carbon which small surface area was
(6.87.10
2
m
2
/g) compared to which was used by the
other authors. They used activated carbon with surface
area was larger than 6.87.10
2
m
2
/g [2, 10].

ISSN: 1411-1284

Proceeding 10
th

0
0,4
0,8
1,2
1,6
2
2,4
2,8
3,2
3,6
4
0 3 6 9 12 15
time, min
l
n

(
C
E
o

/
C
E
)
-3,0
-2,5
-2,0
-1,5
-1,0
0,00325 0,0033 0,00335 0,0034 0,00345
1/T (1/.K)
l
n

k

t
o
t
a
l
,

m
i
n
-
1
Fig 3 Constants rate of degradation endosulfan by
catalytic ozonation at condition: pH 7; agitation speed
750 min
-1
;activated carbon particle size 0.2-0.3 mm;
C
O3
6.372.10
-3
O
3
mol/L; C
Eo
5.10
-5
mol/L; temperature
20
o
C, 25
o
C, 30
o
C. Insert: Arrhenius plot for
degradation of carbofuran by catalytic.

On other hand, compared to the degradation of
carbofuran by catalytic ozonation using the same
activated carbon as catalyst we obtained a slower
degradation rate of endosulfan [12]. This is because, the
reaction of endosulfan to ozone and OH
is less reactive
compared to the reaction of carbofuran to ozone and
OH
.
In addition, the presence of activated carbon in
ozonation processes does not really act as a catalyst, but
rather as an initiator and/or a promotor for ozone
transformation into OH
[13]. Finally, from the rate

constants and activation energy given in Table 2
corresponded to the experiments completed at defferent
temperatures in the range of 20-30
o
C by catalytic and
non-catalytic ozonation

0
0,5
1
1,5
2
2,5
3
3,5
4
4,5
0 3 6 9 12 15
time, min
l
n
(
C
E
o
/
C
E
)
-3,3
-2,8
-2,3
-1,8
0,00325 0,0033 0,00335 0,0034 0,00345
1/T (1/.K)
l
n

k
"
,
(
m
i
n
-
1
)
Fig. 4 Constants rate of degradation of endosulfan by
non catalytic ozonation at condition: pH 7; gas flow rate
0.1603 L h
-1
; agitation speed 750 min
-1
; C
O3
6.372.10
-3
O
3
mol/L; CE
o
5.10
-5
mol/L; temperature 20
o
C, 25
o
C,
30
o
C. Insert: Arrhenius plot for degradation of
carbofuran by catalytic.

Table 2. Pseudo-first order constant (k), correlation coefficient (R
2
), activation energy of carbofuran degradation
by catalytic ozonation and non-cataytic ozonation

Catalytic Ozonation Non-catalytic ozonation Temperature
(
o
C) k.10
-2

(min
-1
)
R
2
Ea (kcal/mol) R
2
k. 10
-2
(min
-1
) R
2
Ea (kcal/mol) R
2
30 11.54 0,99 10.27 0,99
25 9.61 0,99 6,586 0,99 7.86 0,99 8.631 0,99
20 7.97 0,99 6.32 0,99
ISSN: 1411-1284

Proceeding 10
th

IV. CONCLUSIONS

The presence of the activated carbon (catalytic
ozonation) is sligtly significant in enhancing the
degradation rate of endosulfan compared to non-
catalytic ozonation processes. The rate constant (k) of
endosulfan were 11.99 x 10
-2
min
-1
at 30
o
C for catalytic
ozonation, and 10.27x10
-2
min
-1
for non catalytic
ozonation. The activation energy (Ea) for catalytic
ozonation were 8.631 kcal/mol and for non catalytic
ozonation 6.586kcal/mol.

REFERENCES

[1] Jans, U., Hoigne, J. Activated carbon and
carbon black catalyzed transformation of
aqueous ozone into OH-radicals Ozone Sci.
Eng, vol 20, pp.67-89, 1998
[2] Beltran, F. J., Rivas, F. J., Fernandez L. A.,
Alvarez P. M., and Montero-de-Espinosa, R.,
Kinetics of Catalytic Ozonation of Oxalic
Acid in Water eith Activated Carbon, Ind.
Eng. Chem. Res. vol.4, pp. 6510 -6517, 2002.
[3] Langlais, B., David A. R., Brink, D. R., Ozone
in Water Treatment Application Engineering,
Cooperative Research Report. Florida. Lewis
Publishing. 1991.
[4] Jans, U., Hoigne, J. Activated carbon and
carbon black catalyzed transformation of
aqueous ozone into OH-radicals Ozone Sci.
Eng., vol 20, pp. 67-89, 1998.
[5] Extoxnet, 2000, Pesticide Information Profiles:
Endosulfan, Extention Toxicology Network,
Oregon State University.
http://ace.orst.edu/info/extoxnet/pips/carbofur.h
tm)
[6] Benitez., F.J., Beltran-Heredia, J., Acero, J.L.,
and Rubio, F.J. Oxidation of several
chlorophenolic derivatives by UV irradiation
and hydroxyl radicals.J. Chem. Technol.
Biotechnol., vol. 76, pp. 312-320, 2001.
[7] Valdes, H., Zaror, C.A. Heterogeneous and
homogeneous catalyti ozonation of
benzothiazole promoted by activated carbon:
Kinetic approach, Chemosphere., vol. 65, pp.
1131-1136, 2006.
[8] Polo Sanchez, M., Ramos Leyva- Ramos,
Rivera-Utrilla., Kinetics of 1,3,6-
naphthalenetrisulphonic acid Ozonation in
presence of activated carbon.Carbon vol. 43,
pp. 962-969, 2005.
[9] Fogler Fogler, H, S. Elements of Chemical
Reaction Engineering. 3
rd
ed; Prentice-Hall:
Englewood Clifft, NJ, 1999
[10] Yazgan, M. S., and Kinaci, C., Beta-
Endosulfan Removal From Water by Ozone
Oxidation, Water Science & Tecnology. Vol.
48, no. 11, pp. 511-517, 2004.
[11] Benitez, F. J., Acero, J. l., and Real, F.J,
Degradation of carbofuran by using ozone,
UV radiation and advanced oxidation
processes, Journal of Hazardous Materials
vol. B89, pp. 51- 65, 2002,
[12] Enjarlis., Setijo Bismo, Slamet and Roekmijati
W. Soematoyo., Kinetics degradation of
carbofuran by ozonation in the presence
activated carbon
Proceeding 14
th
Regional symposium
Chemical engineering. UGM Yokyakarta.
Indonesia, 2007.
[13] Sanchez Polo, M., Gunten, U.V., Rivera-Utrilla,
J.Efficiency of activated carbon to transform
ozone into OH radicals: Influence of
operational parameters, Water Res.vol. 39, pp.
3189-3198, 2005.

ISSN: 1411-1284

Proceeding 10
th

Hydrocarbon selectivity of Fischer-Tropsch synthesis
from syngas with ratio of H
2
to CO = 1.0 over Co-
Fe/Al
2
O
3
bimetallic catalysts : effect of Fe:Co ratio

Dewi Tristantini
a
and Brje Gevert
b

a
University of Indonesia,

Department of Chemical Engineering, 16424 Depok, Indonesia.
b
Chalmers University of Technology,

Department of Chemical and Biological Engineering,
S-412 96 Goteborg, Sweden
a
Corresponding author: Phone: +62.217863516, Fax.:+ 62.217863515, e-mail: detris@che.ui.edu

Abstract
The utilization of bio-syngas produced from low
temperature gasification of biomass i.e. hydrogen
to carbon monoxide ratio (H
2
/CO-ratio) between
1.0-1.5 directly inside a Fischer-Tropsch (FT)
synthesis is an interesting option since it makes a
water gas shift (WGS)-unit and CO
2
-removal unit
prior to the FT-reactor unnecessary. Different
Fe:Co ratio (12 wt% bimetals) Al
2
O
3
-supported
catalysts were performed in Fischer-Tropsch
synthesis with syngas H
2
/CO ratio = 1.0 at 20 bars
and reaction temperature of 483 K
Catalyst characterizations applied in this study
shows that the higher Fe content in the catalyst
decreases CO chemisorbed but did not influence
significantly in the decrease of the BET surface area
and the degree of reduction.
The CO conversion, the hydrocarbon formation
rate and the selectivity to C
5+
were decreased while
the selectivity to CH
4
and the selectivity to CO
2
were
increased with increasing Fe content. The higher
Fe:Co ratio in the catalyst the higher WGS activity
(CO
2
selectivity) though it is still far from the desired
usage H
2
/CO ratio 1.0 (the H
2
/CO ratio in the feed).
The result indicates that the increase of WGS
activity was followed by the decrease of FT activity
probably due to the competition of those reactions on
the same active sites for Fe-Co bimetallic catalysts.

Key words: Bio-syngas; Hydrocarbon; Fischer-
Tropsch Synthesis, Water-Gas-Shift, Cobalt, Iron,
Bimetallic catalysts, Effect Fe:Co.

I. INTRODUCTION

Biomass can be converted to synthesis gas
(syngas), called bio-syngas, via a high or low
temperature gasification process. This bio-syngas
consists mainly of H
2
, CO, CO
2
and CH
4
. It contains
less hydrogen (often named H
2
-poor bio-syngas) in the
mixture compared to syngas from fossil

sources. An option for the production of renewable
automotive fuels from H
2
-poor bio-syngas is the
Fischer-Tropsch (FT) synthesis, in which H
2
and CO is
converted to longer hydrocarbons (HCs) mainly over
Co- or Fe-based catalysts. The purpose of the present
work is to study the catalyst activity and selectivity to
form hydrocarbons for a series of cobalt-iron alumina
supported catalysts using syngas with 1.0 ratio of
hydrogen to carbon monoxide. This process requires
that the WGS occurs internally over the FT-catalyst by
the water produced in the FT-reaction because that the
syngas has a smaller amount of hydrogen than the
reactants stoichiometry needed.

II. BASIC THEORY

When the syngas obtained upon gasification of
biomass used directly inside a Fischer-Tropsch (FT)
synthesis, it requires the water gas-shift occurs
internally over the FT-catalyst by the water produced
in the FT-reaction for that the syngas has a smaller
amount of hydrogen than the reactants stoichiometry
needed.
The WGS reaction occurs simultaneously with the
production of hydrocarbons during FT synthesis over
iron-based catalysts. These two reactions are:
FTS: O H CH H CO
2 2 2
2 + +
(1)
WGS:
2 2 2
H CO O H CO + +
(2)
A high WGS reaction rate could convert a larger
amount of CO (source of carbon) to CO
2
rather than
the desired hydrocarbon products. However, from the
mass balance point of view, the production of CO
2

parallels the production of H
2
potentially increases the
rate of reaction (1) so that CO and CO
2
are
exchangeable through the WGS reaction [1]. Thus the
relative extents of FT synthesis and WGS reactions
need to be optimized for the maximum production of
hydrocarbons.
In our previous studies we reported the effect of
feed gas composition to FT synthesis hydrocarbon
products over Co/Al
2
O
3
catalysts. We applied three
ISSN: 1411-1284

Proceeding 10
th

kind of syngases with H
2
/CO-ratio 2.1, 1.5 and 1.0
using dry feed [2] and with the addition of water [3].
The former study indicated that lower (H
2
/CO-ratio)
decreased the CO conversion and CH
4
selectivity while
the C
5+
selectivity and C
3
-(olefin/paraffin) were
slightly increased.


The FT-synthesis experiments were applied in the
same reactor (pressure 20 bars, initial temperature 483
K) with the initial space velocity 200 cm
3
/min or gas
hourly space velocity (GHSV) 12 SL/g
cat
.h and held
for approximately 12 hours. The GHSV was lowered
to 50 cm
3
/min (or GHSV 3 SL/g
cat
.h) and held for
another 12 hours. The reason for this is that the flow
rate is the minimum flow can be run in the reactor set
up to reach the maximum conversion. The reactor
temperature was increased 10 K every period after
each run 12 hours until the last period at temperature
523 K.
Continuing our work reported in [2,3], the catalysts
we used in this study were 0%, 2.4%, 6%, 9%, and
12% Fe added to 12%, 9.6%, 6%, 3% and 0% Co give
0:20, 5:20, 10:10, 20:5 and 20:0 ratios of Fe to Co. By
remembering that the total bimetal impregnated was
12% weight, we can make simplicity of percentage
ratio of Fe to Co as follow: (a) 0:100 (0% Fe:12% Co),
(b) 20:80 (2.4% Fe:9.6% Co), (c) 50:50 (6% Fe:6%
Co), (d) 80:20 (9.6% Fe:2.4% Co), and (e) 100:0 (12%
Fe:0% Co).

Catalyst characterization
The characterization of catalyst applied in this study is
reported below including temperature program
reduction (TPR), temperature program oxidation (TPO
or oxygen titration), CO chemisorptions, BET surface
area measurements and X-ray diffraction

IV. CONCLUSIONS

3.1. Effect of Fe:Co ratio on the hydrocarbon
selectivity
CO chemisorption
Table 1 shows characterization results refers to this
study. The CO adsorption data show the difference
between the first and second isotherm. It is obvious
that the difference results (the pure chemisorptions
data) show a dramatic decrease of CO adsorbed as the
Fe loading increased. It is known that when Fe is
covering the surface, both the CO uptake and the
reducibility decrease, even though the same adsorption
stoichiometry of CO on Co and Fe could be assumed
[4]. This was the case when Fe was sequentially added
to a Co/TiO
2
catalyst. A probable reason for the lower
CO uptake in that study could be an inefficient
reduction of the Fe
2
O
3
in poor contact with Co
3
O
4
,
since supported Fe
2
O
3
requires higher reduction
temperature than Co
3
O
4
. In the current study, the
dramatic decrease in CO uptake with increased Fe
content (catalysts a d) is not reflected in the degree of
reduction (see Table 1) or FT activity (Table 2). The
reason is believed to be a failure in measuring the
number of active sites in the Fe-containing catalysts,
rather than indicating that bimetallic catalysts would
have much higher site activities than the pure Co
catalyst. A probable explanation for the low CO uptake
is that the Fe is oxidized by traces of H
2
O or O
2
from
the ambient air upon evacuation at high temperature,
and that the evacuation time between the two isotherms
was too long. Schanke [5] performed the evacuation
between a first and a second CO adsorption isotherm
for Fe/Al
2
O
3
catalysts only for two minutes, since
longer evacuation times (1 3 h) almost completely
removed the initially adsorbed CO. Anyhow, the CO
adsorption results indicate that Fe is effectively
covering the Co sites, reaching full coverage of the Co
sites between 20 50 % Fe (catalysts d e). A surface
enrichment of Fe in supported Co-Fe bimetallic
catalysts, and a CO chemisorptions for high Fe content
(above 33 % Fe) Co-Fe bimetallic catalysts similar to
that of pure Fe, is in agreement with Duvenhage and
Covill e [4].
Compared to the current study, they observed
relatively small effects on the CO uptake when
changing Fe:Co ratio on TiO
2
-supported bimetallic
catalysts. Furthermore, they actually observed a 60 %
increase in CO uptake between a 10 wt% Co/TiO
2

catalyst and a co-impregnated 0.1 wt% Fe / 10 wt% Co
/ TiO
2
catalyst. The replacement of Co with Fe in a 10
wt% metal / TiO
2
catalyst actually improved the CO
uptake even when the composition was 5 wt% Fe / 5
wt% Co / TiO
2
, compared to the pure Co catalyst.
However, the FT activity (at a H
2
/CO ratio of 2) was
generally lower the higher the Fe content, even if the
Co loading was kept constant at 10 wt%. This is
explained by a lower site activity of the alloy, and/or of
a surface enrichment of Fe, which is known to have a
much lower activity compared to Co [4]. In another
study with more Fe:Co ratios investigated, Duvenhage
and Coville [6] reported that the FT activity was higher
for 5 wt% Fe / 5 wt% Co / TiO
2
compared to either 2.5
wt% Fe : 7.5 wt% Co or 7.5 wt% Fe : 2.5 wt% Co
supported on TiO
2
. The 5 wt% Fe / 5 wt% Co/ TiO
2

catalyst possessed the highest reducibility but the
lowest CO uptake of the three, resulting in a higher
turnover frequency of that alloy, however far below
that of a pure Co/TiO
2
catalyst.

Table 1. Characterization result of Fe-Co/Al
2
O
3
catalysts

Catalyst
Fe:Co
BET
surface
area,
[m
2
/g
catalyst]
CO
chemisorption
(first - second
isotherm)
[mol CO/g
catalyst]
Degree
of
reduction
[%]
0:100 174 74 47
5:95 156 37 48
15:85 145 35 51
20:80 149 13 50
50:50 153 8 49
80:20 157 7 41
100:0 162 7 40
Al
2
O
3
195 - -

ISSN: 1411-1284

Proceeding 10
th

Table 2. Activities and selectivities for Fe-Co/Al
2
O
3
catalysts in all periods.

The degree of reduction of the catalysts is slightly
increased when incorporating small to moderate
amounts of Fe (catalysts b e) compared to the mono-
metallic Co-catalyst. The same observation has been
made by Duvenhage and Coville [4, 6] who reported
co-impregnated bimetallic Fe-Co/TiO
2
catalysts giving
higher degrees of reduction than the mono-metallic
Co/TiO
2
. Duvenhage and Coville [4] reported that
upon reduction at 623 K in H
2
of the calcined 5 wt% Fe
/ 5 wt% Co / TiO
2
catalyst discussed above (prepared
by incipient wetness co-impregnation with the metal
nitrates), a bcc Fe-Co alloy was formed and all metal in
the catalyst was fully reduced. The reducibility for the
Fe-Co bimetallic catalysts in the current study are
much lower, and also much closer to each other,
possibly indicating stronger interaction between the
metals and the support, and a poorer contact between
Co and Fe.

Effect of GHSV and Fe:Co ratio on the activity
and selectivity
The activity and selectivity of the catalysts in
which the variation of GHSV were done at 483K
(Table 2). It is well known that Fe catalyst has a lower
CO conversion on FT reaction compare to Co catalyst.
Table 2 shows the CO conversion, hydrocarbon
formation rate and selectivity to C
5+
(CO
2
-free),
selectivity to CH
4
(CO
2
-free), selectivity to CO
2
and
C
3
(olefin/paraffin) for different Fe-Co/Al
2
O
3
. As can
be seen in Table 2, the bimetal-containing catalysts (b-
d) show a decrease in hydrocarbon formation rate with
an increase of Fe content. The CO conversion
decreases in the same way as hydrocarbon formation
rate decreases. This is most probably due to the fact
that Fe gave a negative effect to catalyst activity.
Similar results were made for a series of Fe:Co
bimetallic catalyst employed in FT reaction using
syngas with inlet molar H
2
/CO-ratio 2.0-2.1 [7].
Another synchronic result was found for a Fe:Co
bimetallic catalyst employed in benzene hydrogenation
and thiophene desulphurization reactions reaction [4].
It seems that the addition of two FT active metals,
when used together, did not simply give the additive
properties expected from knowledge of the properties
of the individual metals [4, 6]. Some other studies on
the use of Fe/Co catalysts for FT reaction using syngas
from coal or natural gas have been reported [8, 9] and
biomas derived syngas [10,11].
The selectivity to C
5+
also decreases following the
increase of iron content in the catalysts up to 80% (a-
d). This can be understood that iron tend to form
shorter chain of hydrocarbons than the longer one.
The selectivity to C
5+
for the merely iron catalyst (e)
is slightly higher than this for catalyst d. Contrary
with C
5+
selectivity, the selectivity to CH
4
and
selectivity to CO
2
increase very much with the
increase of Fe/Co-ratio in the catalyst. The selectivity
to CH
4
and the selectivity to CO
2
steep up from 11.3%
and 3.9% for 20F-80C (b) to 23.1% and 10.5% for
80F-20C (d), respectively.
These results are opposite with Duvenhages [6]
who got a decrease in CH
4
formation when the
temperature was increased. A trend of decreasing C
5+
-
and increasing of CH
4
- and CO
2
- selectivity
culminate in 80F-20C (d) catalyst and the activity was
lift up back when only Fe metal in the catalyst. The
Fe/Al
2
O
3
(e) catalyst does not show a lower CO
conversion either a lower hydrocarbon formation rate
than those for the one with 80% Fe content (d). The
Catalyst Condition
CO
conversion
[%]
Hydrocarbon
formation rate
[gHC/gcat.h]
Selectivity to C5+
(CO2-free) [%]
Selectivity to CH4
(CO2-free) [%]
Selectivity to
CO2 [%]
(a) 0Fe-100Co Dry feed 5.7 0.20 82.2 8.8 3.4
Reduced GHSV 23.6 0.21 86.6 6.0 1.6
(b) 5Fe-95Co Dry feed 7.5 0.27 78.0 10.1 3.1
Reduced GHSV 25.4 0.23 79.7 8.4 2.3
(c) 15Fe-85Co Dry feed 7.2 0.25 78.2 10.2 3.6
Reduced GHSV 22.0 0.20 78.3 9.3 2.9
(d) 20Fe-80Co Dry feed 5.8 0.20 70.0 13.5 5.1
Reduced GHSV 20.8 0.18 72.8 11.3 3.9
(e) 50Fe-50Co Dry feed 4.4 0.15 64.9 14.8 5.5
Reduced GHSV 13.8 0.12 64.8 14.9 5.5
(f) 80Fe-20Co Dry feed 3.2 0.11 64.9 15.6 8.1
Reduced GHSV 7.7 0.06 48.1 23.1 10.5
(g) 100Fe-0Co Dry feed 4.5 0.16 59.2 16.6 6.1
Reduced GHSV 10.0 0.09 51.9 20.0 8.7
ISSN: 1411-1284

Proceeding 10
th

selectivity to CH
4
is 19.9% and selectivity to CO
2
is
8.7% for Fe-single metal catalyst (e). In short words,
the performance of Fe/Al
2
O
3
to some extent is better
than 80Fe:20Co. This is possibly due to a larger
amount of metallic Fe present in the former which is
indicated by the TPR pattern, where the Fe-peaks are
shifted to lower temperatures for catalyst (d) in the
[9]. It could be also the reason due to the BET surface
area and the CO chemisorptions of Fe catalyst (e) is
little bit higher than those for 80Fe:20Co (d) catalyst.
This result indicates that the Fe-metal is responsible to
form more CO
2
than longer chain hydrocarbons in the
product [12].
The CO hydrogenation rate over (a) (0 % Fe : 100 %
Co) with stoichiometric feed (H
2
/CO = 2.1) is 0.37 g
hydrocarbons (HCs) /g
cat,
h [2, 3]. It is obvious that no
catalyst in the current study, with H
2
/CO = 1.0, comes
close to this value. The WGS equilibrium constant at
483 K is approximately 370, calculated from data in
reference [13]. which means that in order to reach the
WGS equilibrium either the water partial pressure or
the CO partial pressure must be very low at the reactor
exit. For dry experiments with no external water
addition, this means that essentially all water formed in
the reactor must react in the WGS reaction (1.2), giving
a usage ratio of 0.5. Hence, for a feed of H
2
/CO = 1.0,
it is not necessary to reach the WGS equilibrium in
order for the usage ratio to match the feed ratio,
especially not with external water addition. However,
in all FT experiments presented here, the WGS reaction
quote in the gas exiting the reactor has a very low value
(~ 10
-3
10
-1
) compared to the equilibrium constant.
The un-promoted Fe catalyst is obviously not
especially active for the WGS reaction. It is known that
in LTFT synthesis with Fe-based catalysts, the WGS
equilibrium is not reached, due to the low temperature
making the WGS reaction slow. In order to improve
the WGS activity, the promotion of Fe by alkali is
needed [14]. Duvenhage and Coville [4, 6] reported
that the relative WGS activity was tripled for a 5 wt%
Fe / 5 wt% Co / TiO
2
catalyst upon 0.5 wt% K
promotion. However, K is a poison to Co, lowering the
total activity.
A complement to the current study would be to keep
the Co loading constant and to add Fe in different
amounts, since it is obvious that lowering the Co
loading reduces the overall FT activity, even if the
WGS activity is increased by the Fe content. This
would be economically reasonable since iron is about
thousand times less expensive than cobalt. The main
reason for lower activity is of course that the more
active Co is replaced by the less active Fe, and also that
the FT rate over Fe catalysts is negatively affected by
the water partial pressure. Furthermore, not only the FT
reaction, but also the WGS reaction takes place over
the Fe surface. Only for catalyst (d) (20 % Fe : 80 %
Co) the effect of a relatively high WGS activity (due to
a high Fe content) on the total activity (CO conversion)
was positive, if compared with the other bimetallic
catalysts with lower Fe content. In general though, a
higher WGS activity resulted in a lower FT activity
irrespective of water partial pressure in reactor (see
Figure 3), indicating that the WGS- and FT reactions
take place at the same type of sites.

IV. CONCLUSIONS

Fischer-Tropsch synthesis at 20 bars and 483 K with
H
2
-poor syngas (H
2
/CO ratio = 1.0) was performed
over different Fe:Co ratio (12 wt% bimetal) Al
2
O
3
-
supported catalysts. All catalysts had very low activity
for WGS, which resulted in low productivities (per
gram catalyst). Even for external water addition, the
WGS activities were poor. A higher Fe:Co ratio in the
catalyst resulted in a higher WGS activity, however not
lowering the H
2
/CO usage ratio to the desired value of
1.0 (the H
2
/CO ratio in the feed). The results indicate
that the WGS- and FT reaction were competing for the
same active sites (on the Fe-containing catalysts),
giving a decrease in productivity for the catalysts with
highest WGS activity. The higher the Fe content, the
lower was the C
5+
selectivity and C
3
olefin/paraffin
ratio. Water addition increased C
5+
selectivity and C
3

olefin/paraffin ratio and reduced CH
4
selectivity. Fe
was enriched at the catalyst surface, probably covering
the Co sites. In order to achieve a high conversion of a
H
2
-poor syngas, a combined high WGS- and FT
activity is necessary which is difficult when the two
reactions are competing for the same sites.

Acknowledgement
This work was supported by the Swedish Energy
Agency. Odd Asbjrn Lindvg from SINTEF,
Trondheim-Norway is acknowledged for his kindly
help during the experimental part of the work.

REFERENCES

[1] Raje, A., Inga, J.R. Davis, B.H., Fuel 76 (1997) 273.
[2] D. Tristantini, S. Lgdberg, . Borg, B. Gevert, A.
Holmen, Fuel Processing Technol., 88 (2007) 7, 643-649.
[3] D. Tristantini, and B. Gevert, A. proceeding 14
th
RSCE,
UGM, Yogyakarta, 2007. (accepted).
[4] D.J. Duvenhage, N.J. Coville, Appl. Catal. A 153 (1997)
43.
[5] D. Schanke, Hydrogenation of CO over supported iron
catalysts, PhD Thesis, The Laboratory of Industrial
Chemistry, The University of Trondheim, The
Norwegian Institute of Technology, Trondheim, 1986.
[6] D.J. Duvenhage, N.J. Coville, Appl. Catal. A289
(2005) 231.
[7] Guerrero-Ruiz, A., Sepulueda-Escribano, A., Rodriguez-
Ramos, I, Appl. Catal., 81 (1992) 101.
[8] Wender, I., Fuel Process. Technol. 48 (1996) 189.
[9] Soled, S.L., Fiato, R.A., US Patent 4607020 (1986).
[10] Mirzaei, A.A. , Habibpour, R., Kashi, E., Appl. Catal. A
296 (2005) 222.
[11] Jun, K., Roh, H.,Kim, K.,Ryu, J., Lee, K., Appl. Catal.
A. 259 (2004) 221.
[12] Eilers, J.,Posthuma, S. A., Sie, S. T., Catal. Lett., 7
(1990) 253.
[13] H. Hiller, R. Reimert, in: B. Elvers, S. Hawkins, M.
Ravenscroft, J.F. Rounsaville, G. Schulz (Eds.),
Ullmanns Encyclopedia of Industrial Chemistry, vol.
ISSN: 1411-1284

Proceeding 10
th

A12, 5
th
ed., VCH Verlagsgesellschaft mbH, Germany,
1989, pp. 179-181.
[14] A. Holmen, D. Schanke, G. Sundmark, Appl.
Catal. 50 (1989) 211.

ISSN: 1411-1284

Proceeding 10
th

The Deactivation Behaviour of HZSM-5 Catalyst at
Various Temperatures of Acetone Conversion to
Mono-aromatic Hydrocarbons

Setiadi*, Prilly Fatticianita J.*, Toshinori KOJIMA
(1)*

* Department of Chemical Engineering, Faculty of Engineering, University of Indoneisa, Kampus Baru UI
Depok 16424, Tel. 021-7863516, email : setiadi@che.ui.edu, INDONESIA
*
Department of Applied Chemistry, Faculty of Engineering, Seikei University
3-3-1, Kichijoji-Kitamachi, Musashino, Tokyo 180-8633, JAPAN

Abstract- Catalytic stability of HZSM-5 (Si/Al=25)
on various temperatures (573K, 623, 673 and 723 K)
have been studied to obtain the suitable temperature
for the formation of monoaromatic chemicals in
acetone reaction. The reaction was performed in the
near atmospheric pressure with acetone velocity at 4
h
-1
. The results show that the catalyst stability is
strongly depend on the reaction temperatures, and the
catalyst generally underwent a deactivation on each
temperature. However, the best result of acetone
reaction is at 673 K, acetone conversion are closed to
100 % during 17 h and the yield of monoaromatic
chemicals was above 70 % for 13 h of reaction. The
highly decreasing of surface area (total area or
micropore area) of used catalyst, the deactivation is
considered caused by the pore blocking on the
external surface of HZSM-5.

Key words- Reaction Temperatures, Monoaromatic
Chemicals, HZSM-5, Deactivation

I. INTRODUCTION

For design or developing a new catalyst which can
works with a long life time is attractive and
challenging tasks for increasing the operational and
economical benefits of an industrial process. In
principle, life time of a catalyst is correlated to the
extent of reaction and the selectivity to the desired
product in order to meet the aspect of economically
acceptable to the process in the real commercial
industries. Depend on the reaction condition and the
catalyst employed, life time of a catalyst is mostly
ascertained by these factors.
The probability of deactivation process in the
acetone conversion via aldol condensation reaction is
enough high due to the formation of a condensation
product will shut down the reaction over H-ZSM-5
catalyst. Aldol condensation and secondary reactions
of acetone on zeolite H-ZSM-5 were extensively
studied using
13
C NMR spectroscopy (Xu et al., 1994)
and using IR observation of acetone adsorption (Zaki
et. Al, 2000). The formation of the acetone reaction
condensation product is considered to the high
coverage of acetone (high ratio of acetone/active site),
which allows the high mobility of acetone with the
possibility of the bimolecular condensation reaction
product such as mesityl oxide (Panov and Fripiat,
1998). However, the condensation product has a much
lower mobility than acetone. As pointed by Melo et
al. (1997), the condensation product is retained in the
zeolite pores because of their low volatility, of
adsorption on the acid sites or because their size is
close to or greater than the size of pore aperture.
Some reaction techniques were employed to
increase to the catalysts lifetime in acetone
conversion by adding the water (Podrebarac et. al,
1997; Lucas et. al, 2001). However, the systematic
study to correlate the lifetime or stability of catalyst
with reaction temperature is still absent. The
impressive results of the effect of temperature on the
acetone conversion using H-ZSM-5 have been
demonstrated by Chang et al. (1981), Chang and
Silvestri (1977). But, the knowledge about the activity
and stability on each temperature and the typical ratio
of Si/Al of HZSM-5 catalyst were nothing.
For developing a catalyst for the acetone
conversion to aromatic chemicals, we have reported in
the previous study that the activity and stability of
HZSM-5 catalyst with Si/Al ratio=25 is very high
during 10 h of reaction with 100 % conversion. And
673 K is the favourable temperature for the formation
of aromatic chemicals (Setiadi et al., 2003). Based on
the arguments summarized above, this report presents
the results of acetone conversion on the longer times
of stream in more than 10 h of reaction on the various
temperatures.

ISSN: 1411-1284

Proceeding 10
th

Fig. 1 Acetone conversion during time on stream
reaction on various temperatures : 723 K (), 673 (O),
623 K () and 573 K ().
II. EXPERIMENTAL

Fine crystal powder of HZSM-5 Si/Al=25was
used as catalyst for the reaction of acetone
aromatization. One gram of this HZSM-5 was mixed
with quartz sand to avoid the excessive increase of
pressure drop. This mixture was packed in a SUS 316
reactor column with 6 mm i.d and 700 mm length.
Some amount of quartz sand was added at the upper
part of the catalyst bed preheating and vaporising the
acetone loaded into reactor. A stainless steel rod was
inserted at the lower part of the reactor to strengthen
the catalyst bed against the pressure due to the gas
flowing through the bed. To fix completely the
catalyst bed in its position, quartz wool was filled on
each side of the catalyst layer to keep the possible
entrainment of the powder during the experiment
performed.
The reaction temperatures employed were at 723,
673, 623 and 573 K with acetone space velocity 4 h
-1

at near atmospheric pressure. To examine the
stability of HZSM-5 in the acetone conversion on
those temperatures, HZSM-5 fresh catalysts were
used when each experiment was started to run. The
detail description of the experimental method have
been reported in the our previous work (Setiadi et. al,
2003).
The fresh and used catalysts after reaction at each
temperature were characterized by BET method for
total and micropore surface area measurements.
Nitrogen was used as adsorbed gas at the liquid the
liquid nitrogen temperature and these measurements
were carried out by the apparatus of Micromeritics
Gemini model 2360. The determinations of the
micropore surface area were performed by t plot
method and the thickness of adsorbed layers (t) were
calculated by Halsey equation.

III. RESULT AND DISCUSSION

In order to find the optimum temperature for
obtaining the high activity and stability of H-ZSM-5
with Si/Al ratio of 25, the reaction of acetone
conversion was done at various temperatures at 723 K,
673 K, 623 K and 573 K. The results are in Fig. 1
which are the plots of acetone conversions versus
times on stream. The reason of the using of HZSM-5
with the Si/Al ratio 25 is consistent to our previous
result that this ratio exhibited the effective ratio of
HZSM-5 catalyst for the conversion of acetone to
aromatic chemicals (Setiadi et al., 2003). It is clear
shown that if the reaction was performed on the
higher temperature region (more than 673 K), the
activity and stability of HZSM-5 were much better
than on the lower ones. Within 17 hours of reaction,
the results of reactions at 673 and 723 K show that the
time on stream curves of acetone conversions were
stable at 100 % on both temperatures. Unfortunately,
the a-
cetone conversions obtained after 17 h reaction
underwent
a sharp decrease to be around of 20 % especially for
673 K. It indicates that the HZSM-5 catalyst
deactivated after 17 h of reaction.
In the case of the lower temperature region
(lower than 673 K), it obviously shown that the low
stabilities of H-ZSM-5 can be seen and the acetone
conversions were decreased in the shorter times of
stream to be less than 20 % within 12 h reaction (at
623 K) and 4 h reaction (573 K). The tendency of this
deactivation at the lower temperature region, may be
caused by the low ability of HZSM-5 to activate the
acetone molecules than that to the high temperature
region. And the acetone reaction leads to a formation
of multiple condensation products with the higher rate
of formation comparing to a proceed of dehydration
process
Fig. 2 shows the monoaromatic yields during
times on

0
20
40
60
80
100
0 5 10 15 20 25 30
Ti me on stream [h]
A
c
e
t
o
n
e

c
o
n
v
e
r
s
i
o
n

[
%
]
0
20
40
60
80
100
0 5 10 15 20 25 30
Time on stream [h]
Y
i
e
l
d

[
%
m
o
l

c
a
r
b
o
n
]
Fig. 2 Yield of aromatic compound during time on stream
reaction on various temperatures : 723 K (), 673 K (O), 623
K () and 573 K ().
ISSN: 1411-1284

Proceeding 10
th

stream. The similar result was found is that the time
on stream yield of monoaromatic compounds from the
acetone conversion at 673 K were also superior than
that at other temperatures during 13 h. Even though
the stability of acetone conversion obtained by
performing the reaction at 723 K was little bit higher
than 673 K, but this time on stream yield at this
temperature was lower. These results lead to a
conclusion that the temperature reaction of 673 K is
the suitable temperature for conversion of acetone for
aromatic product formation.
In general, the deactivation can not be avoided to
occur on the each temperature region and the
deactivation rate is more rapid on the lower
temperature region. The possibility of a formation
poly condensation products with heavier molecular
weight on each temperature is high and these products
covered the center of active sites or to block the pore
channel of HZSM-5.
To address the possible cause of the decreasing of
catalyst activities, we observed the surface area of
fresh catalyst and the surface of each catalyst after
reaction on each reaction. Table 1 shows the result of
surface area measurement for the fresh and
deactivated catalysts

Table 1 Result of surface area measurement
H-ZSM-5 sample
BET
m
2
/g
Micropore
m
2
/g
1 Fresh catalyst 321.8 209.4
2 Used Catalyst at 573 K 76.0 44.2

on each temperature. It is obviously shown that there
was a dramatic change of the surface area (BET total
surface area or micropore surface) of fresh HZSM-5
and deactivated catalyst. The decreasing of surface
area of fresh HZSM-5 to be deactivated catalyst was
about 76 % for total area and 80 % for micropore area.
And the fact (not shown here) is that there was any
change of physical properties from white color (fresh
catalyst) toward black color (deactivated catalyst).
These findings strongly suggest that the deactivation
is main caused by surface blocking or pore plugging
of HZSM-5. As pointed out by Lucas et al. (1997), the
formation of external coke leads to block of the
channel structure even at low coke contents. The
relative sharply decreases of acetone conversions at
all temperatures as shown in Fig. 1, it could be
correlated to a evident of a high possibility of the
multiple condensation product formation (coke)
which may cover the external surface of HZSM-5 on
the mouth of pore channels.
The deactivation could be caused not only by the
evident of pore blocking by coke formatted during
reaction proceed, but also the process of the
dealumination process. The detection of alumina
external surface could be appropriately performed by
EPMA measurements. Based on the distinction of
curve lines of spatial distribution for alumina obtained
for both catalyst (not shown here), it shows that the
amount of external alumunium atom for used catalyst
is higher than that for fresh catalyst. These result
ssuggest that the dealumination process have also
occurred during the acetone reaction. And also
indicates that dehydration process as one step of
reaction mechanism and the water produced will react
with the framework aluminium atoms. Depend on
the reaction type employed, the chemical species
produces due to the dealumination process can be
more catalytically active or inactive comparing to the
tetrahedral framework aluminium atom. The
excellent report described the dealumination
phenomenon have been demonstrated by Masuda et al.
(1998) which showed that tetrahedral aluminium
atoms are extracted out of the framework and
transformed into partially distorted octahedral
aluminium atoms. These atoms move to the outer
surface of the zeolites crystals. However, those
aluminum atoms are likely not active for the acetone
reaction and to allow in more completely decreasing
of HZSM-5 activity as shown in Fig.1.

IV. CONCLUSIONS

In general, the deactivation process is observed
on all temperatures observed (573, 623, 673 and 723
K) over HZSM-5 (Si/Al-25) and is more rapid if the
acetone reaction was performed on the lower
temperature region. Based on the time on stream
stability experiments, the best temperature for the
formation of monoaromatic chemicals is 673 K. The
stability of the catalyst achieved was close to 100 %
conversion during 17 h and the yield was obtained in
more than 70 % for 13 h of reaction.
The decreasing of catalyst activity should be
connected to the decreasing of the surface area (BET
or micropore) by pore blocking due to the formation
of heavier molecular products from acetone
condensation reaction.

REFFERENCES

1) Chang, Clarence D., W. H. Lang, and W.K. Bell;
"Molecular Shape-Selective Catalysis in Zeolite," in
Catalysis of Organic Reactions edited by William R.
Moser, Marcel Dekker Inc., 73-94 (1981)
2) Clarence D. Chang and Anthony J. Silvestry, "The
Conversion of Methanol and Other O-Compounds to
Hydrocarbons over Zeolite Catalysts," Journal of
Catalysis 47, 249 (1977)
3) Lucas, A., P. Canizares, A. Duran and A. Carrero;
Coke Formation, Location, nature and regeneration
on Dealuminated HZSM-5 type Zeolites, Applied
Catalysis A : General,156 , 299-317 (1997).
4) Lucas, A., P. Canizares, A. Duran,"Improving
deactivation behaviour of HZSM-5 catalysts",
App.Catal. A : General, 206, p.87-93 (2001)
ISSN: 1411-1284

Proceeding 10
th

5) Masuda, T., Yoshihiro Fujikata, Shin R. Mukai, Kenji
Hashimoto; Changes in Catalytic Activity of MFI-
type Zeolites Caused by Dealumination in a Steam
Atmosphere, Applied Catalysis : General, 172, 73-
83 (1998).
6) Melo, L., P. Magnoux, G. Giannetto, F. Alvarez, M.
Guisnet; " Transformation of acetone over a
0.4PtHMFI(60) catalyst. Reaction scheme," Journal of
Molecular Catalysis A : Chemical, 124, 55-161 (1997)
7) Panov A.G.; Fripiat J.J.; "Acetone Condensation
Reaction on Acid Catalysts,"
Journal of Catalysis, 178, no. 1, 188-197 (1998)
8) Panov, A. and J.J. Fripiat; "An Infrared Spectroscopic
Study of Acetone and Mesityl Oxide Adsorption on
Acid Catalyst," Langmuir, 14, 3788-3796 (1998)
9) Podrebarac, G.G., Ng F.T.T, Rempel, G.L, A kinetic
Study of the Aldol Condensation of Acetone using an
Anion Exchange Resin Catalyst, Chemical
Engineering Science, 52, 17, 2991, (September 1997)
10) Setiadi, S., T. Tsutsui, T. Kojima," Conversion of
Acetone to Aromatic Compound with HZSM-5, The
journal of Japan Institute of Energy, 82 (12), (2003)
11) Xu, Teng, Eric J. Munson, and James F. Haw; "Toward
a Systematic Chemistry of Organic Reactions in
Zeolites: In Situ NMR Studies of Ketones," J. Am.
Chem. Soc., 116, 1962-1972 (1994)
12) Zaki, M.I., M. A. Hasan, F.A. Al-Sagheer, and L.
Pasupulety; "Surface Chemistry of Acetone on Metal
Oxides: IR Observation of Acetone Adsorption and
Consequent Surface Reactions on Silica-Alumina
versus Silica and Alumina," Langmuir, 16, 430-436
(2000)

ISSN: 1411-1284

Proceeding 10
th

CO
2
Absorption Through Hollow Fiber Membrane
Gas-Liquid Contactors

Sutrasno Kartohardjono
Faculty of Engineering, University of Indonesia, Kampus Baru UI, Depok 16424
Tel. (6221) 7863516, fax. (6221) 7863515, email: sutrasno@che.ui.edu

Abstract Hollow fiber membrane modules have
been widely used as contactor and filtration
devices to provide high surface area in a small
volume. Membrane-based contactor devices
provide an alternative technology for gas-liquid
contacting operations that overcomes disadvantage
of conventional contactors such as columns and
towers, and in the same time offers substantially
more interfacial area using hollow fiber membrane
contactors. In this study, performance of
hydrophobic microporous polypropylene hollow
fiber membrane contactors is examined to absorb
CO
2
using water and dilute solution of NaOH as
absorbents.
Experiment results show that CO
2
flux through
the membrane contactors using 0.01 M NaOH
solution is 120.000 times higher than using water
as solvent. The CO
2
flux decreases with increasing
contactor-packing fraction for the same liquid
velocity using both solvents. The same effect also
occurred for the mass transfer coefficients
characteristics. Liquid pressure drops increase
with increasing contactor-packing fraction at the
same liquid flow rate due to an increase in friction
between fibers and water. Based on the friction
factor data the fibers surface did not behave like a
smooth pipe within the range of velocity in the
experiments. The friction factors obtained for the
membrane contactors in this study are 4.6 to 14.6
times higher than the theoretical value based on
flow of fluids in the smooth pipe.

Keywords hollow fiber, contactor, mass transfer,
pressure drop

I. INTRODUCTION

ollow fiber membrane modules have been widely
used as contactor and filtration devices to provide
high surface area in a small volume [1]. As a gas-
liquid contactor, unlike more conventional membrane
applications such as microfiltration, ultrafiltration and
reverse osmosis, the driving force for separation is a
concentration rather than a pressure gradient [2,3].
Therefore, only a small pressure drop across the
membrane is required to ensure that gas-liquid
interface is remain immobilized at the membrane
pores [2,4,5].
Membrane-based contactor devices provide an
alternative technology for gas-liquid contacting
operations that overcomes disadvantage of
conventional contactors such as columns and towers,
and in the same time offers substantially more
interfacial area using a suitable membrane
configuration or module such as hollow fiber
membrane modules. In hollow fiber membrane
contactors, gas flow rate and liquid flow rate can be
varied independently of each other without flooding,
loading and foaming, which are characteristics of
dispersion-based gas-liquid contactors. Furthermore,
all of the membrane surface area is available for
contacting regardless of how low the flow rates are.
Membrane-based contactors are easier to be scaled-up
than other conventional mass transfer equipment.
Membrane operations usually scale linearly, so
that an increasing in capacity is achieved simply by
the add of membrane modules subject to the
limitations of supporting equipment such as transfer
pump, piping, etc. In addition, membrane-based
contactors can be operated over a wide range of
capacities by modular design. Small or large
capacities can be obtained by using few or many
modules.
However, membrane based contactors have
several disadvantages such as the additional resistance
of membrane it self to mass transfer performance,
subject to channeling which result in a loss efficiency,
and subject to fouling. Membrane Systems have a
short life and a limitation in the number of
equilibrium stages due to pressure drop constraints.
These relatively few disadvantages are often
compensated by the advantages cited above [6].
Hollow fiber membrane contactors superior to
their surface area per volume compare to conventional
contactors, such as absorption and extraction columns,
typically 50cm
2
/cm
3
, and may be as high as
300cm
2
/cm
3
[7]. These values are about 30 times
more area than what is achievable in gas absorbers
and 500 times what is obtainable in liquid/liquid
extraction columns [6]. Hollow fiber membrane
contactors (HFMC) have been demonstrated in a wide
range of application such as in gas stripping and gas
absorption. In the hydrophobic microporous
H
ISSN: 1411-1284

Proceeding 10
th

membrane gas/liquid contactors, the pores are filled
with gas and as a result the resistance in the
membrane can be neglected to the overall resistances
of the contactor.
This behavior of hydrophobic membrane contactor
has brought many researchers to utilize hydrophobic
microporous hollow fiber membrane contactors to
remove acid gases, mainly CO
2
and H
2
S from natural
gas or CO
2
from flue gas stream [8]. The aims of this
study are: to evaluate the effectiveness of hollow fiber
membrane module as a contactor device in CO
2

absorption process; and to develop a simple model,
which is useful in the membrane contactor design, for
characterizing the mass transfer correlation of CO
2

absorption process in the membrane contactor.

II. BASIC THEORY

The mass transfer of CO
2
from the lumen fiber to
the solvent on the shell-side involves three steps:
transfer of CO
2
in the gas side, diffusion of oxygen
through the membrane, and diffusion of oxygen into
the liquid side. The driving force for CO
2
removal
from the gas side comes from the CO
2
concentration
gradient between the gas and the phases across the
membrane.
It is more convenience to express overall mass
transfer coefficient, K
L
, in the contactors based on the
liquid phase for many gas absorption/stripping
processes. For a microporous-hydrophobic membrane
with gas-filled pores, the relation is,
iL im
i
iG
i
L
k k
H
k
H
K
1 1
+ + =
(1)
where H
i
is the Henrys law constant for the ratio of
the liquid concentration to the gas concentration in
equilibrium. For sparingly soluble gases the relation
is,
iL L
k K
1 1
(2)
indicating that the liquid film controls the mass
transfer.


Hollow fibers membranes used in the experiment
were made of polypropylene supplied by AKZO. The
fibers are 2.7 mm in outer diameter and 0.45 mm in
wall thickness. The hollow fibers modules were made
by potting both ends of the fibers into 1.5 cm nipple
using epoxy (Araldite). The module was connected to
external shell made from PP pipe with a PP T-joint,
and at the other end of the nipple was connected to
vacuum pump using brass reducer 1.5 cm to 0.3 cm.
The diameter of the external shell can be varied using
appropriate connection to PP T-joint where in this
study 1.6 cm shell diameter was used. There were 3
contactors of about 40-cm length and 10, 15 and 20
fibers used in the experiments to give membrane
packing fractions in the contactors of 0.2, 0.3, and 0.4,
respectively.
Schematic of experimental configuration as shown
in Fig.1 was used to measure mass transfer
performance and pressure drop of the membrane
contactors. The liquid in the reservoir, which is water
or 0.1 M NaOH solution, was initially pumped
through the membrane contactor at the shell side.
Meanwhile, CO
2
gas was sent to the contactor through
the lumen of fiber. pH and temperature of the liquid
were automatically measured every 20 s from the pH-
meter probe, and digital manometer measured liquid
pressure differences between inlet and outlet
contactor.

Figure 1. Schematic diagram of experimental set up.

The overall mass transfer coefficients from the
experiments are calculated using Equation (3),
1
*
0
*
ln
C C
C C
A
Q
K
m
L
=
(3)
where C*, C
t
and C
0
are CO
2
concentrations in
equilibrium, liquid inlet and outlet contactor, Q
L
is
liquid flow rate to the membrane contactor, A
m
is
outer membrane fibers surface area, and K
L
is the
observed overall mass transfer coefficient,
respectively. Friction factors of the contactors were
calculated using Equation (4),

2
2
l
v l
Pd
f
e
=
(4)
where f is friction factor of the module, P is liquid
pressure drop, d
e
is equivalent diameter of the
contactor, l is contactor length, is water density, and
v
L
is water velocity, respectively.
CO
2
flux through the membrane contactors, J
CO2
,
and the overall mass transfer coefficients, K
L
, from
the experimental results were plotted versus water
velocity v
l
as shown in Figure 2 and 3, respectively.
As it can be seen from Figure 2, the water can absorb
CO
2
through the membrane contactor as high as
4.5x10
-8
mole CO
2
per square meter membrane area
per hour. Meanwhile, 0.01 M NaOH solution can
absorb CO
2
through the membrane contactor as high
as 5.5x10
-3
mole CO
2
per square meter membrane
area per hour, which is 120.000 times higher than
using water as solvent.
ISSN: 1411-1284

Proceeding 10
th

The CO
2
flux decreases with increasing contactor-
packing fraction for the same liquid velocity. The
same effect also occurred for the mass transfer
coefficient of the membrane contactors. This can be
explained as in a region of low packing fraction
contactor, transverse flow and surface renewal effect
is seem to be more influence to the mass transfer
performance rather than channeling effect, meanwhile
at the higher packing density the channeling effect is
more dominant [1]. From another point of view, the
geometrical dependencies of boundary layer profiles
might also increase the mass transfer coefficient of
contactor with lower packing density. The increase is
directly related to the boundary layer conditions,
where the profiles of boundary layers on a curved
body become thinner with an increasing degree of
curvature, as for example decreasing d
o
fiber with
other conditions constant. Therefore, based on this
theory, the mass transfer coefficient will increase with
decreasing packing fraction of the contactor,
especially at higher liquid velocities. The channeling
occurred because of the existing regions of densely-
and loosely-packed in the contactor which creates
preferential flow around fibers in the contactor
resulted mainly from uneven distribution of fibers and
flow [9]. Membrane contact area might have been
reduced in densely packed regions as fibers were
more adhered to each other, reducing the availability
of contacting surface.

1.E-08
1.E-07
1.E-06
1.E-05
1.E-04
1.E-03
1.E-02
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
V
L
(m/s)
J
C
O
2

(
m
o
l
e
/
m
2
.
s
)
N=10
N=15
N=20
N=10
N=15
N=20

Figure 2. Variation of CO
2
fluxes J
CO2
, in water (filled
symbols) and 0.01 M NaOH solution (open symbols) with
liquid velocities v
L
, in the membrane contactors

To examine the dependence of CO
2
transfer
coefficient on the process parameters using water as a
solvent, correlation are conveniently expressed in
terms of dimensionless Sherwood number, Sh,
Reynolds number, Re, and Schmidt number, Sc.

C B
Sc A Sh Re = (5)

In this study the Schmidt number, Sc, was not varied,
so the 1/3 power-dependence in the literature was
assumed [10], and Equation (5) become,

33 . 0
Re Sc A Sh
B
= (6)

1.E-07
1.E-06
1.E-05
1.E-04
1.E-03
1.E-02
0.1 1
v
L
(m/s)
K
L

(
m
/
s
)
N=10
N=15
N=20
N=10
N=15
N=20

Figure 3. Variation of mass transfer coefficients K
L
, in water
(filled symbols) and 0.01 M NaOH solution (open symbols)
with liquid velocities v
L
, in the membrane contactors

Figure 4 shows the experimental data as Sh plotted
against Re at various packing fraction contactors to
obtain the exponent for Reynolds number, B in
Equation (5) for each contactor. Furthermore,
Equation (6) can be simplified into Equation (7) to see
the dependency of Reynolds number on mass transfer
performance.

b
a Sh Re = (7)

The exponent for Reynolds number, b, was
obtained by the linear regression of Figure 4 and it
was found that the values were ranging from 0.63 to
0.65 and have the average value of b=0.64. This
value indicated that the mass transfer in the contactors
is dominated by turbulent flow [10]. The values of a
for each contactor were obtained from the slope of the
experimental Sherwood number against Re
0.64
where
these values are ranging from 0.0115 to 0.0387. To
obtain the correlation between a and packing fraction
of the contactor,, the values of a were plotted against
contactor packing fraction where the correlation
between a and packing fraction is best fitted by
Equation (8),

72 . 1
0026 . 0

= a (8)

where is packing fraction of the contactor,
respectively.
Equation (8) describes the dependency of
geometry of the hollow fiber membrane contactors
employed in this study where the mass transfer
performance is related to the packing fraction of fibers
in the contactors. An empirical correlation for mass
transfer performance in the hollow fiber membrane
contactors used in the experiments can then be
correlated in the form

ISSN: 1411-1284

Proceeding 10
th

64 . 0 72 . 1
Re 0026 . 0

= Sh (9)

To see the effects of contactor packing fraction on
pressure drop due to friction loss, the friction factors
of the contactor were obtained from Equation (4) and
then were plotted against Reynolds number as shown
in Figure 5. The friction factor increases with
decreasing contactor-packing density for the same
contactor diameter due to an increase in hydraulics
diameter of the contactor as expressed in Equation (4).
The dependency of friction factor on Reynolds
number for fluid flow in the contactor has general
form,
b
a f Re = (10)

where f, a and b are friction factor and empirical
constants from the experimental results, respectively.
The exponent b has a negative value as can be seen
from many literatures [11,12]. As for instance, b is 1
and -0.25 for liquid flow in the laminar region
(Re<2000) and in the turbulent region (Reynolds
number up to 10
5
), respectively [11]. It can be seen
here that the negative value of b becomes smaller as
the flow more turbulent. In this study where the range
of Reynolds number measurement in the contactor is
from 800 to 3400 has b values ranging from 0.27 to
1.01 as shown in Figure 5. As a comparison, It was
found that the friction factor data has value of b = -1.1
for Re<5 and b = -0.5 for 5<Re<100 based on the
experiment using microporous hollow fiber
membrane blood- oxygenator [12].

0.1
1.0
10.0
100 1000 10000
Re
S
h
N=10
N=15
N=20

Figure 4. Variation of Sherwood numbers Sh, with Reynolds
numbers Re, in the membrane contactors

The friction factor ratio of the contactor to the
smooth pipe is shown in Figure 6. These data reveal
that the fibers surface did not behave like a smooth
pipe within the range of velocity in the experiments.
Thus viscous sub layer did not develop due to the
movement of the fibers as quantitatively expressed in
very low viscous drag values compare to inertial drag
values. The random movement of the fibers was more
pronounced at lower velocities, and as a result, the
value of the friction factor is also higher at lower
velocities [4]. The friction factor ratio in Figure 6 is
higher for lower packing densities due to the more
movement of the fibers in low packing fraction
contactor.

0.001
0.01
0.1
1
100 1000 10000
Re
f
N = 10
N = 15
N = 20

Figure 5. Variation of friction factor f, with Reynolds
number Re, in the membrane contactors

The friction factors obtained for the membrane
contactors in this study are 4.6 to 14.6 times higher
than the theoretical value based on flow of fluids in
the smooth pipe as shown in Figure 6. As
comparison, It was reported that the friction factor of
hollow fiber blood oxygenators was about 7 times
higher than the theoretical value at Re<5 and about 31
times higher than the theoretical value at 5<Re<100
[12]. Meanwhile, other researcher reported that the
friction factor of sealed end polypropylene hollow
fiber contactors with packing densities 8.7 to 9.6 %
were about 2 times higher than the values for pipe
flow reported in the literature [4]. It can be seen here
that the friction factor ratio can vary from contactor to
contactor depend on the geometry of the contactors.

0
2
4
6
8
10
12
14
0 1000 2000 3000 4000
Re
f
R
N = 10
N = 15
N = 20

Figure 6. Variation of friction factor ratio f
R
, with Reynolds
number Re, in the membrane contactors

IV. CONCLUSIONS

ISSN: 1411-1284

Proceeding 10
th

Experiments have been conducted to absorb CO
2

gas using water and 0.01 M NaOH solution as
solvents through hydrophobic microporous
polypropylene hollow fiber membrane contactors of
packing fraction 0.2, 0.3 and 0.4. The water can
absorb CO
2
through the membrane contactor as high
as 4.5x10
-8
mole CO
2
per square meter membrane
area per hour. Meanwhile, the 0.01 M NaOH solution
can absorb CO
2
through the membrane contactor as
high as 5.5x10
-3
mole CO
2
per square meter
membrane area per hour, which is 120.000 times
higher than using water as solvent. The CO
2
flux
decreases with increasing contactor-packing fraction
for the same liquid velocity using both water and 0.01
M NaOH solution as solvents. The same effect also
occurred for the mass transfer coefficients of the
membrane contactors.
The mass transfer performance of the membrane
contactors using water as absorbent can be formulated
as a function of contactor geometry and liquid flow
rate. The contactor geometry is represented by
packing fraction, whilst the flow rate is represented by
Reynolds number. An empirical correlation for mass
transfer performance in the hollow fiber membrane
contactors in the range of packing fraction of 0.2 to
0.4 can be correlated in the
form
64 . 0 72 . 1
Re 0026 . 0

= Sh .
Liquid pressure drops increase with increasing
contactor-packing fraction at the same liquid flow rate
due to an increase in friction between fibers and
water. Based on the friction factor data the fibers
surface did not behave like a smooth pipe within the
range of velocity in the experiments. The friction
factors obtained for the membrane contactors in this
study are 4.6 to 14.6 times higher than the theoretical
value based on flow of fluids in the smooth pipe.

REFERENCES

[1] Lipnizki, F., and Field, R.W., Mass transfer
performance for hollow fibre modules with shell-side
axial feed flow: using an engineering approach to
develop a framework, Journal of Membrane Science,
193 (2001), 195-208.
[2] Dindore, V.Y., and Versteeg, G.F., Gasliquid mass
transfer in a cross-flow hollow fiber module:
Analytical model and experimental validation,
International Journal of Heat and Mass Transfer, 48
(16) (2005), 3352-3362.
[3] Sirkar, K.K., Other new membrane processes, in
W.S.W. Ho and K.K. Sirkar (Eds.), Membrane
Handbook, Chapman & Hall, New York, (1992), 885-
899.
[4] Ahmed, T., Semmens, M.J., and Voss, M.A., Energy
loss characteristics of parallel flow bubbleless hollow
fibre membrane aerators, Journal of Membrane
Science, 171 (2000), 87-96.
[5] Lim, S.P., Tan Xiaoyao and Li, K., Gas/vapor
separation using membranes: Effect of pressure drop in
lumen of hollow fibers, Chemical Engineering Science,
55 (4) (2000), 2641-2652.
[6] Gabelman, A., and Hwang, S.T., Hollow fiber
membrane contactors, Journal of Membrane Science,
159 (1999), 61-106.
[7] Wickramasinghe, S.R., Semmens, M.J., and Cussler,
E.L., Better hollow fibre contactors, Journal of
Membrane Science, 62 (1991), 371-388.
[8] Mavroudi, M., Kaldis, S.P., and Sakellaropoulos, G.P.,
A study of mass transfer resistance in membrane gas
liquid contacting processes, Journal of Membrane
Science, 272 (1-2) (2006), 103-115.
[9] Wu, J., and Chen, V., Shell-side mass transfer
performance of randomly packed hollow fibre
modules, Journal of Membrane Science, 172 (2000),
59-74.
[10] Costello, M.J., Fane, A.G., Hogan, P.A., and Schofield,
R.W., The effect of shell side hydrodynamics on the
performance of axial flow hollow fibre modules,
Journal of Membrane Science, 80 (1993), 1-11.
[11] Coulson, J.M., and Richardson, J.F., Chemical
Engineering, Volume 1: Fluid Flow, Heat Transfer and
Mass Transfer, Pergamon Press, Oxford, New York,
Toronto, Sydney, Paris, Frankfurt, 1978, 3rd ed., pp.
40-45.
[12] Wickramasinghe, S.R., Garcia, J.D., and Han, Binbing,
Mass and momentum transfer in hollow fibre blood
oxygenators, Journal of Membrane Science, 208
(2002), 247-256.

ISSN: 1411-1284

Proceeding 10
th

Enzymatic and Acid Hydrolysis of Bagasse for Ethanol
Production by Simultaneous Sacharification and
Fermentation

Samsuri, M.
1,2
, Gozan, M.
1
, Nasikin, M.
1
, Prasetya, B.
2

1
Chemical Engineering Department, Faculty of Engineering,
University of Indonesia, Kampus UI, Depok 16424, Indonesia,
2
Centre of Biotechnology, Indonesian Institute of Science (LIPI),
Jl. Raya Cibinong km 48, Bogor, Jawab Barat, Indonesia

Email address: msyamsuri@ristek.go.id

Abstract-Ethanol was produced from bagasse using
enzymatic hydrolysis, acid hydrolysis and
combination of Enzymatic and acid hydrolysis in
Simultaneous Saccharfication and fermentation.
Cellulase and xylanase were used for enzymatic
hydrolysis and HCl low concentrations (1% and
0,5%) was used for acid hydrolysis.
The results showed that the highest ethanol yield
with enzymatic hydrolysis with enzyme cellulase at
pH 5, about 5,51g/L or about 11, 24% base on
original bagasse. The highest ethanol production
from bagasse by combination of enzyme cellulase
and xylanase at pH 5 about 7,63 g/L or 15,25%.
The highest ethanol yield after hydrolysis with HCl
0,5% is 3,24 g/L at incubation for 24 h. The highest
ethanol yield after hydrolysis with HCl 1% is 3,77
g/L or 7,54% at incubation for 24 h. The highest
ethanol yield after hydrolysis with combination of
cellulase, xylanase and HCl 1% about 8,22 g/L or
16,45%. This yield increasing about 2% if
compared with ethanol yield without acid addition.
If compared with acid hydrolysis by HCl 1%
increasing about 8% based on original bagasse

Key words : Bagasse, SSF, Enzymatic hydrolysis
cellulose, xylanase, Acid hydrolysis HCl

1. INTRODUCTION
Most ethanol produced in the world today is
derived from starch or sucrose (Gong et al., 1999).
Ethanol fermentations are traditionally carried out for
wine or beer production, but ethanol for transportation
and industry are a large potential market and growing
use. Ethanol belongs to clean combustion. Its oxygen
content decreases emissions in mix combustion with
gasoline. Because this ethanol is originally plant
matter, its use as fuel does not contribute to the net
accumulation of carbon dioxide in the atmosphere
(Costello and Chum, 1998). Starch is abundant in crop
materials, but expansion of ethanol production for the
purpose of automotive fuel requires feedstocks that do
not compete for food or fiber (Himmel, et al., 1996).
In the other hand, lignocellulosic material is by
far the most abundant raw material such as in
hardwood, softwood, grasses and agricultural residues.
Further potential raw materials are newsprint, office
papers, municipal solid wastes, etc. Lignocellulose is a
more complex substrate than starch. It consists of a
mixture of carbohydrate polymers (cellulose and
hemicellulose) and lignin. The carbohydrate polymers
are tightly bound to lignin mainly by hydrogen bonds
but also by some covalent bonds (Lee, 1997).
Lignocellulose contains five major sugars, which
varies with the feedstock (Hinman et al, 1989). They
are the hexose; glucose, mannose, galactose and
pentose; xylose and arabinose. Glucose is the most
abundant carbohydrate in lignocellulosic material,
xylose is the second one.
One of the sustainable alternative energy is
relatively cheap production and environment friendly
was development bio ethanol from waste residue
agriculture. It does contain many lignocelluloses like
bagasse (waste residue sugar industry). Indonesia has a
lot of biomass like bagasse. Based on Indonesian
Central Research Sugar Agriculture data, bagasse was
produced 32% from sugarcane. 60% of it, was used by
sugar industry as fuel, material for paper, fungi
industry, etc. Because of that, approximately about
40% of bagasse wasnt use. Polysaccharide of
sugarcane contain more 70%, it divide to 50%-55%
cellulose and 15%-20% hemicellulose. The lignin
contain about 20-23%, beside that it was called ash
compound (Pandey, et al, 2000).
Generaly The process for converting the
lignocellulosic material to ethanol requires: (1)
delignification to liberate cellulose and hemicellulose
from their complex with lignin; (2) depolymerization
of the carbohydrate polymer to produce free sugars;
and (3) fermentation of mixed hexose and pentose
sugar as monosaccharide to produce ethanol. The
accurate analysis during conversion process is also
important for the successful development of feasible
ISSN: 1411-1284

Proceeding 10
th

conversion process. Presently, it is known that among
the key processes described above, the delignification
of lignocellulosic raw materials is the rate-limiting and
most difficult task to be solved (Lee, 1997).
Hydrolysis process is converted polysaccharide to
monosaccharide. Major constituents in enzymatic
hydrolysis and acid hydrolysis from lignocellulosic
material like bagasse are glucose and xylose released
from cellulose and hemicelluloses, respectively.
Ethanol fermentation process utilizing yeast has been
well developed with glucose as a carbon and energy
source. Generally enzymatic hydrolysis using
cellulose or xylanase, and acid hydrolysis is using HCl
or H
2
SO
4
.
Simultaneous Saccharification and Fermentation
(SSF) firstly described by Takagi (Takagi, et al, 1977],
combine enzymatic hydrolysis or acid hydrolysis of
polysaccharide with simultaneous fermentation of the
sugars obtained ethanol. The advantage of this process
is cellulose or hemicellulose were converted to
monosaccharide, it rapidly ferments to ethanol and use
one reactor for entire process so that saved the cost
equipment.
This research was tried to make ethanol from
bagasse by SSF process with enzymatic hydrolysis,
acid hydrolysis and combination of enzymatic
hydrolysis and acid hydrolysis.

2. Methods

2.1 Preparation of sample
Sugarcane bagasse (Saccharum officinarum) were
milled and screened (30-60-mesh). The milled bagasse
was then air-dried to a final humidity 10% and stored
under dry conditions. Bagasse was taked from
Indonesia (from sugar factory in Lampung, Indonesia).
Preparation of material Hydrolysis
HCl 0.5% and 1% was used for acid hydrolysis,
Cellulase and Xylanase was were used for enzymatic
hydrolysis.

2.2 Growth Stock S.cerevisiae
S.cerevisiae AM12 was pre-cultured in Potato
Dextrose Agar (PDA) 2%, Agar (0.25 g), Aquades
(50ml) and incubation for 3-4 days on ambient
temperature, and then was used as yeast for SSF
process.
2.3 Preparing yeast inoculation
S.cerevisiae AM12 fresh from growing stock was
pre-cultured with 50 ml medium (glucose, 10 g/l; yeast
extract, 1.0 g/l; KH
2
PO
4
, 0,1 g/l; MgSO
4
.7H
2
O, 0,1 g/l;
and (NH4)
2
SO
4
, 0,1 g/l) in 50 ml flask. Before
inoculation, medium was autoclaved for 20 minutes,
and then incubation on ambient temperature for 24
hours with use by shaker.
2.4 SSF process
SSF medium 5.5 ml, it containing bagasse (0.25
g), nutrient medium (2.5 ml), 0.5 Na-citrate buffer (pH
4, 4.5, 5) cellulase enzyme (10 FPU = 0.016 g) and 2.5
ml yeast inoculated. Bagasse, nutrient medium and Na-
citrate buffer sterilize for 20 minutes in autoclave.
However, enzyme was added without sterilization.
Nutrient medium is containing 1.0 g/l (NH
4
)
2
PO
4
; 0.05
g/l MgSO
4
.7H
2
O and 2 g/l yeast extract. Cultivation
was took and get in to micro centrifuge tube. The clean
liquid from sample had taken in 24, 48, 72 and 96
hours and doing ethanol concentration test for ethanol
produced.
For strong acid hydrolysis using 0.5% and 1%
(v/v) concentration, chloride acid was added after
bagasse and then enzyme, yeast inoculated (2 ml), Na-
citrate buffer pH 5 (0,5 ml) and nutrient (2 ml).

2.5 Analysis

a. Analysis lignin, holocellulose and -cellulose
Lignin analysis was done by modified klason
lignin method. It is with add sulphuric acid 72% in
sample and agitate until crushed, and then autoclave in
121
0
C for 30 minutes. After that, it filtered with paper
filter, covered by aluminium foil and dry it into oven
for 1 hour weighing the weight.
Holocellulose and -cellulose analysis, it analyze
by wise method. It mixed sample with sodium chlorate,
acetate acid and aquades. Incubation with hot water in
80
0
C, chilled, filtration with aquades and it wash with
acetone. Then, solid residue was dried in oven on
105
0
C for 1-2 nights and weighing the weight.

b. Measuring ethanol concentration
Ethanol concentration was measured by Gas
Chromatography SUPELCOWAX-10 type (Supelco
Inc., 0,53 mm ID., 15 m, 0,5 mm, FID) with
temperature was maintained at 50
0
C. Before tested,
sample was took 50 l and 200 l distilled water
added.

3. Result And Discussion

3.1 Composition Bagasse Analysis

The analysis result show that lignin contain in
bagasse approximately about 24.2% from bagasse total.
-cellulose contain about 52.7% and hemicellulose
contain about 17.5% in bagasse.
0 10 20 30 40 50 60
lignin
-cellulose
Hemicellulose
Extractive chemical
composition (%)

Figure 1. Chemical composition of bagasse

ISSN: 1411-1284

Proceeding 10
th

Totally of hemicellulose and -cellulose that is
opportunity to convert to monosaccharide in hydrolysis
process. Actually about 69% composition of bagasse
is polysaccharide and able to convert to ethanol
production. Theoritically Theoretically if calculated
base of cellulose the maximum ethanol yield about
56,86% base -cellulose and 28,43% base on original
bagasse. If calculated base of hemicellulose maximum
ethanol yield about 56,84 % base on hemicellulose and
11,37 % base on original bagasse. Totally maximum
ethanol yiel base original bagasse about 40%.

3.2 Ethanol production from bagasse with
Enzymatic Hydrolysis

Ethanol production with enzim cellulose
Ethanol production from bagasse using enzyme
cellulase showed in figure 2. The highest ethanol yield
from experiment at pH 5, about 5,51g/L or about 11,
24% base on original bagasse. Optimum pH for
enzyme is 4.8 and yeast is between 4 and 5. So that
SSF process pH is among 4.5-5. This condition is
exactly for enzyme and yeast condition. Beside pH,
temperature is also influence in SSF process.
Temperature for growing yeast media is about 30
0
-
35
0
C (Makanjuola, et al, 1992). But the cellulolytic
enzyme act optimally about 40
0
C. So, the optimum
temperature range for SSF is 35-41
0
C. At lower
temperatures the rate of saccharification decreases
significantly.

0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
0 24 48 72 96
Incubation time (h)
E
t
h
a
n
o
l

c
o
n
c
e
n
t
r
a
t
i
o
n
s

(
g
/
L
)
pH = 4
pH = 4,5
pH = 5

Figure 2. Ethanol production from bagasse with
enzyme cellulose in variation of pH

At this research, SSF temperature was used in
ambient temperature, while yield ethanol was produced
is not too high because cellulase enzyme capability was
decreased to transform cellulose to glucose.
The result also shows that ethanol can be obtained
in 24 hours almost reach optimal concentration in SSF
process. After 24 hours, ethanol concentration not too
significant increase and highest ethanol concentration
was obtained between 72 and 96 hours of incubation
time. From this research, SSF process with pH 5 was
obtained the highest ethanol concentration compare
with pH 4 and pH 4,5.
Based on literature, optimum pH for SSF process
is between pH 4.5 until pH 5 and the highest ethanol
was obtained at pH 5. These phenomena is relevant
with the literature. Based of this research for the other
experiment con was continued on pH 5.
Cellulose are usually a mixture of several
enzymes. At least three major groups of cellulases are
involved in the hydrolysis process; (1) endogluconase
which attacks regions of low crystallinity in the
cellulose fiber, creating free chain-ends; (2)
exogluconase or cellobiohyrolase which degrades the
molecule further by removing cellobiose units from the
free chain-ends; (3) -glicosidase which hydrolyzes
cellobiose to produce glucose. In addition to the three
major groups of cellulase enzymes, there are also a
number of ancillary enzymes that attack
hemicelluloses, such as glucuronidase, acetylesterase,
xylanase V-xylosidase, galactomananase, and
glucomananase (Duff and Murray, 1996). During the
enzymatic hydrolysis, cellulose is degradable by the
cellulases to reducing sugars that can be fermented by
yeast or bacteria to ethanol

Ethanol production from bagasse by combination of
enzyme cellulase and xylanase.

Lignocellulosic material especially bagasse consist
17,5% of hemicellulose. The highest composition of
hemicelluloses is xylan, that can not convert to xylose
by enzyme cellulase. Enzyme xylanase is one of
alternative enzyme able to convert hemicellulose
(xylan) become xylose and xylose able to convert to
ethanol. Ethanol production by enzim cellulase and
xylanase showed in figure 3 and 4, respectively.

The highest ethanol production from bagasse by
combination of enzyme cellulase and xylanase at pH 5
about 7,63 g/L or 15,25%. The yield of ethanol if
compared with yield of ethanol using enzyme
cellubiase without xylanase increase about 3 %. This
phenomena indicated that enzyme xylanase increasing
the ethanol yield. The highest ethanol yield at pH 5,
this phenomena indicated pH 5 is the optimum pH for
process SSF.
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
0 24 48 72 96 120
Incubation time (h)
E
t
h
a
n
o
l

c
o
n
c
e
n
t
r
a
t
i
o
n
s

(
g
r
/
L
) pH = 4
pH = 4,5
pH = 5

Figure 3. Ethanol production from bagasse by enzyme
cellulase and xylanase

ISSN: 1411-1284

Proceeding 10
th

0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
0 24 48 72 96 120
Incubations time (h)
E
t
h
a
n
o
l

c
o
n
c
e
n
t
r
a
t
i
o
n
s

(
%
)
pH = 4
pH = 4,5
pH = 5
Figure 4. Percentage of ethanol production from
bagasse by enzyme cellulase and xylanase base on
original bagasse

3.3 Ethanol production from bagasse with Acid
Hydrolysis

Concentrated acids HCl have been used to
hydrolysis of lignocellulosic materials (in this research
bagasse). The highest ethanol yield after hydrolysis
with HCl 0,5% is 3,24 g/L at incubation for 24 h. The
highest ethanol yield after hydrolysis with HCl 1% is
3,77 g/L or 7,54% at incubation for 24 h. Ethanol
concentration of bagasse after SSF using hydrolysis
with HCl 0,5% and 1% is showed at figure 5,
respectively.

0.00
1.00
2.00
3.00
4.00
5.00
6.00
0 24 48 72 96 120
E
t
a
n
o
l

c
o
n
c
e
n
t
r
a
t
i
o
n
s

(
g
r
/
L
)
HCl
0,5%
Figure 5. Ethanol production from bagasse by acid
hydrolysis (HCl 0,5% and 1%).

Based on data in figure 5, ethanol production
from bagasse by HCl hydrolysis increasing at 24 h
incubation time because HCl is powerful agents for
cellulose and hemicellulose hydrolysis (Sivers and
Zacchi, 2002). The ethanol yield is lower than with
enzyme hydrolysis because HCl was used at the lower
concentrations (only 1% and 0.5%), more higher
concentration HCl the yield of ethanol will be
increasing that is indicated ethanol yield by HCl 1%
higher than 0,5%. But unfortunately concentrated
acids are toxic, corrosive and hazardous and require
reactors that are resistant to corrosion (Sun and Cheng,
2002). In addition, the concentrated acid must be
recovered after hydrolysis to make the process
economically feasible (Sivers and Zacchi, 1995).
Dilute acid hydrolysis has been successfully
developed for pretreatment of lignocellulosic materials.
The dilute sulfuric acid pretreatment can achieve high
reaction rates and significantly improve cellulose
hydrolysis (Esteghlalian et al., 1997). At moderate
temperature, direct saccharification suffered from low
yield because of sugar decomposition. High
temperature in dilute acid treatment is favorable for
cellulose hydrolysis (Millan, 1994). Recently
developed dilute acid hydrolysis processes use less
severe conditions and achieve high xylane to xylose
conversion yields. Achieving high xylane to xylose
conversion yields necessary to achieve favorable
overall process economics because xylane accounts for
up to a third of the total carbohydrate in many
lignocellulosic materials (Hinman et al., 1992). There
are primarily two types of dilute acid pretreatment
processes: high temperature (T greater than 160
o
C),
continuous-flow process for low solids loading (5-10%
weight of substrate/weight of reaction mixture)
(Brennan et al., 1986; Converese et al., 1989), and low
temperature (T less than 160
o
C), batch process for high
solids loading (10-40%) (Esteghlalian et al., 1997).
Although dilute acid pretreatment can significantly
improve the cellulose hydrolysis, its cost is usually
higher than some physico-chemical pretreatment
processes such as steam explosion. A neutralization of
pH is necessary for the down stream enzymatic
hydrolysis of fermentation process (Sun and Cheng.,
2002).

3.4 Ethanol production from bagasse with
combination of Acid Hydrolysis

This research was done by combination of cnzyme
cellulase, xylanase and additional strong acid with low
concentration in SSF process. Additional strong acid
was done together with enzyme. The strong acid was
used is chloride acid (HCl) with concentration 0.5%
and 1% (v/v). Additional strong acid will be affected in
hydrolysis process of cellulose to glucose and can be to
increase ethanol yield.

The results show that the highest ethanol yield
after hydrolysis with combination of cellulase,
xylanase and HCl 1% about 8,22 g/L or 16,45%. This
yield increasing about 2% if compared with ethanol
yield without acid addition. If compared with acid
hydrolysis by HCl 1% increasing about 8% based on
original bagasse. Ethanol production from bagasse
using SSF with combination of enzymatic hydrolysis
(Cellulase and cellubiase) and acid hydrolysis (HCl 1%
and 0,5%) showed at figure 6, respectively.

ISSN: 1411-1284

Proceeding 10
th

0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
0 24 48 72 96 120
E
t
h
a
n
o
l

c
o
n
c
e
n
t
r
a
t
i
o
n
s

(
g
/
L
)
cellulase & xylanase
pH=5
"Cellulase, Xylanase
and HCl 0.5 %"
Cellulase, xylanase
& HCl 1 %"
Figure 6. Ethanol production from bagasse by enzyme
cellulase, xylanase and added HCl low concentrations.

Base on this result, if the acid concentration more
highly, then the hydrolysis capability will more
increase. In acid hydrolysis, the degree of
polymerisation cellulose will be decreased
significantly. It was caused by H
+
concentration
(proton) in strong acid. If the concentration strong acid
more high, then H
+
concentration will be more high
too. Hydrolysis by acid can be able to breakdown -1,
4-glycosidic chain in cellulose. The reaction will start
from proton in acid which rapidly react with glycosidic
oxygen which is connect to two sugar units transform
to conjugation acid. The breakdown of C-O chain and
broken conjugation acid will be transform to
carbonium cyclic ion. After addition water, sugar and
proton will be free (Carrasco, et al, 1994).
Hemicellulose is easier to hydrolysis rather than
cellulose with acid hydrolysis. Hemicellulose also
easier to depolymerisation in high temperature at acid
condition (Palmqvist, et al, 1996) and hemicellulose
sugar such as xylose, arabinose can be free with
hemicellulose hydrolysis. Hemicellulose hydrolysis in
acid condition is possible to form 2-furaldehyde
(furfural) form pentose sugar, it mainly from xylose
and form 5-hydroxymethil-2-furaldehyde (HMF) from
hexoses sugar. Furfural compound and HMF can
delay yeast work because it makes lag phase in ethanol
production and decrease ethanol productivity, but there
is no effect to ethanol was produced. S.cerevisae cans
metabolism furfural and HMF, in order that inhibitor
effect will be disappeared. It caused by membrane
permeability of S.cerevisae.

4. Conclusion

Based on the result of research was obtained, and
then the conclusion:
1. Ethanol was production from bagasse using
enzyme cellulase, xylanase and HCl in low
concentrations.
2. The highest ethanol yield with enzymatic
hydrolysis with enzyme cellulase at pH 5, about
5,51g/L or about 11, 24% base on original
bagasse.
3. The highest ethanol production from bagasse by
combination of enzyme cellulase and xylanase
at pH 5 about 7,63 g/L or 15,25%.
4. The highest ethanol yield after hydrolysis with
HCl 0,5% is 3,24 g/L at incubation for 24 h.
The highest ethanol yield after hydrolysis with
HCl 1% is 3,77 g/L or 7,54% at incubation for
24 h.
5. The highest ethanol yield after hydrolysis with
combination of cellulase, xylanase and HCl 1%
about 8,22 g/L or 16,45%. This yield increasing
about 2% if compared with ethanol yield
without acid addition. If compared with acid
hydrolysis by HCl 1% increasing about 8%
based on original bagasse.

6. References

Carrasco, J.E.; Saiz, M.C.; Navarro, A.; Soriano, P.;
Saef, F. and Martinez, M. (1994) Effect of dilute
acid and steam explosion pretreatments on the
cellulose structure and kinetics of cellulosic
fraction hydrolysis by dilute acids in
lignocellulosic materials. Applied biochemistry
and biotechnology. 45/46, 23-34.
Hatakka, A. (2001) Biodegradation of lignin. In M
Hofrichter and A. Steinbchel (eds.),
Biopolymers, vol. 1. Wiley-VCH, Weinheim,
Germany. p. 129- 180.
Hofrichter, M. (2002) Review: Lignin conversion by
manganese peroxidase (MnP). Enzyme Microb.
Technol. 30:454-466.
Itoh. H., Wada. M., Honda. Y., Kuwahara. M.,
Watanabe. T., 2003. Bioorganosolve
pretreatments for simultaneous saccharification
dan fermentation of beech wood by etanolysis
dan white rot fungi. Journal of Biotechnology
103, 273-280
Lee, J., 1997. Biological conversion of lignocellulosic
biomass to ethanol. Biotechnology 56, 1-24.
Lynd, L.R., Bothast, R.J., Wyman, D.E. 1991. Fuel
etanol from cellulosic biomass. Science 251:
1318-1323.
Martn, C., Galbe, M., Nilvebrant, N.O., Jnsson, L.J.,
2002. Comparison of the fermentability of
enzymatic hydrolyzed of sugarcane bagasse
pretreated by steam explosion using different
impregnating agents. Applied Biochemistry and
Biotechnol. 99, 699-716.
Maiorella, B. L. 1985 Industrial chemical, biochemical
and fuels: Ethanol. In: Comprehensive
biotechnology. The principles, applications and
regulations of biotechnology in industry,
agriculture and medicine. M.-M. Young (ed.),
Pergamon Press Ltd. University of Waterloo,
Ontario, Canada, Vol. 3. Chapter 43, 861-914.
ISSN: 1411-1284

Proceeding 10
th

Makanjuola, D. B., Tymon, A. and Springham, D. G.
1992. Some effects of lactic acid bacteria on
laboratory-scale yeast fermentations. Enzyme
Microb. Technol. 14: 350-357.
Mc Millan, J.D., 1994. Pretreatment of lignocellulosic
biomass. In: Himmel, M.E., Baker, J.O.,
Overend, R.P (Eds), Enzymatic Conversion of
Biomass for Fuels Production. American
Chemical Society, Washington, DC, pp. 292-324
Palmqvist, E., Hahn-Hgerdal, B., Galbe, M., Larsson,
M., Stenberg, K., Szengyel, Zs., Tengborg, C.
and Zacchi, G. 1996a. Design and operation of a
bench-scale process development unit for the
production of ethanol from lignocellulosics.
Biores. Technol. 58: 171-179.
Pandey,A. Soccol,C.R. Nigam,P. And Soccol,V.T.
2000. Biotechnological potential of agro-
indistrial residues. Sugarcane bagasse.
Bioresour Technol. 74: 69-80.
Sun, Y., Cheng, J. 2002. Hydrolysis of lignocellulosic
materials for etanol production: review.
Bioresource Technology 83, 1-11.
Takagi,M., Abe, S., Suzuki, S., Emert, G.H., Yata,
N. 1977. A method for production of etanol
direcly from cellulose using cellulose and yeast.
Proceedings of Bioconversion symposium,
Delhi, 551-571.
Wyman, C.E., 1994. Etanol from ligcellulosic biomass:
Technology, economics, and opportunities.
Bioresource Technology 50, 3-6.
Zadrazil, F. Brunert, H., 1982. Solid state fermentation
of lignocellulose containing plant residues with
Sporotrichum pulverulentum Nov. and
Dichomitus squalens (Karst.) reid. European
Journal of Applied Mirobiology and
Biotechnology 16, 45-51

ISSN: 1411-1284

Proceeding 10
th

A General Application of a Direct Method for
Multivariable MPC Control Strategy in Chemical
Processes

A. Ahmad
1
and A. Wahid
2

1. Department of Chemical Engineering, Universiti Teknologi Malaysia, Johor
2. Department of Chemical Engineering, University of Indonesia, Depok.
Corresponding authors: arshad@fkkksa.utm.my and wahid@che.ui.edu

Abstract A general application of a direct method
for model predictive control strategy is proposed
for multivariable nonlinear control problem in
chemical processes. The aim is to provide a
solution to nonlinear control problem that is
favorable in terms of industrial implementation.
The scheme utilizes multiple linear models to cover
wider range of operating conditions. Depending on
the operating conditions, suitable models of 2x2,
3x3, and 4x4 processes are used in control
computations. Servo and regulatory controls of the
system are examined on constrained and
unconstrained conditions. Comparisons are made
to conventional controllers. The results confirmed
the potentials of the proposed strategy.

Keywords Tuning, model predictive control,
multivariable, nonlinear control, chemical process

I. INTRODUCTION

he cornerstone of MPC is the model [1]. It cause
MPC is called MBPC (model-based predictive
control). MPC uses models in 2 ways: using a reliable
model to predict effect of past control moves on P
future outputs, assuming no future moves, and using
the same model to compute the optimal M controller
moves. Implement first move and repeat procedure.
Predictive control is now one of the most widely
used advanced control methods in industry, especially
in the control of processes that are constrained,
multivariable and uncertain. A large number of
implementation algorithms, included industrial
predictive control applications [2] have a appeared in
the literature.

II. TUNING THE MPC CONTROLLER

Dynamic matrix control (DMC) [3] is the most
popular MPC algorithm used in the chemical process
industry today. Over the past decade, DMC has been
implemented on a wide range of process applications.
A major part of DMCs appeal in industry stems from
the use of a linear finite step response model of the
process and a simple quadratic performance objective
function. The objective function is minimized over a
prediction horizon to compute the optimal controller
output moves as a least-squares problem.
Tuning in conventional control strategy (P, PI, and
PID) is related to obtain an optimum setting of
controller parameters (K
c
, T
i
, and T
d
). Ziegler-Nichols,
Lopez, Ciancone, etc. [4] are some examples of
single-loop tuning in P, PI, and PID controllers.
Huang, et. al. [5] has proposed a direct method for
multi-loop PI/PID controller design based on
FOPDT/SOPDT model of each loop.
MPC controller has certain parameters setting to
achieve its optimum performance. During the time,
trial-and-error efforts have been done to reach this
goal until Shridhar & Cooper [6, 7] proposed a tuning
strategy for unconstraint SISO and multivariable
MPC. Dougherty 8and Cooper [8] proposed a non-
adaptive DMC tuning strategy (see Table 1) based on
all of FOPDT models in systems.
The principles of the multivariable DMC tuning
strategy [8] are:
1) T Sampling time. While a large T refers to a low
computation load, a small T refers to a properly
track the evolving dynamic behavior. Too slow of
a sampling rate will lead to information losses,
and too fast of a sampling rate could lead to
numerically sensitive procedures. Nevertheless,
this method allows values of T other than the
recommended value given in Table 1.
2) P (prediction horizon) and N (model horizon).
Both P and N have the same setting and are
related to the settling time of the slowest (the
largest time constant) sub-process in the
multivariable system. A large P improves the
nominal stability of the closed loop. Meanwhile,
a large N makes controller has long enough time
to avoid the instabilities that can otherwise result
since truncation of the model horizon
misrepresents the effect of controller output
moves in the predicted process variable profile.
3) M Control horizon. M equals to 63.2% of the
settling time of the slowest sub-process in the
multivariable system. This ensures M to be long
T
ISSN: 1411-1284

Proceeding 10
th

enough such that the results of the control actions
are clearly evidenced in the response of the
measured process variable.
4) - Controlled variable weights. The setting of
those parameter are free. However, in the most
case, they are set equal to one.
5) Move suppression coefficients. Its primary
role in DMC is to suppress aggressive controller
actions. When the control horizon is 1 (M = 1),
no move suppression coefficient is needed ( =
0). If the control horizon is greater than 1 (M >
1), then the analytical equation given in Table 1
is used.
Table 1 Non-adaptive DMC tuning strategy [8]
Approximate the process dynamics of all controller output to measured process variable pairs with FOPDT
models:

1. Select the sample time as close as possible to:

2. Compute the prediction horizon, P; and the model horizon, N:

3. Compute a control horizon, M:
M
4. Select the controlled variable weights, , to scale process variable units to be the same.
5. Compute the move suppression coefficients, :

6. Implement DMC using the traditional step response matrix of the actual process and the initial values of
the parameters computed in steps 16.

III. APPLICATION IN GENERAL PROCESSES

Dougherty and Cooper [8] have proved the non-
adaptive DMC tuning strategy in the three 2x2
processes (general transfer function, multi-tank, and
distillation column). Here, we shall illustrate
comparisons between PI/PID and MPC controller
strategy using Table 1 in the same and more complex
processes: 2x2 process, 3x3 process and 4x4 process
[5].

3.1 Wood and Berry distillation column (2x2 process)
Wood and Berry distillation column model [5] is
the most popular model of 2x2 process system. This
model consists of two matrixes: manipulated variable
matrix (2x2) and disturbance matrix (2x1).
Nevertheless, the first matrix (manipulated variable
matrix) is commonly used by researchers [5, 9].
Table 2 shows results of PID and MPC tuning in
Wood & Berry model. Responses of Wood & Berry
model using these settings are shown by Figure 1.
Although, MPC controller is not set in the optimum
setting, MPC controller performances are better than
PID controller performance. MPC controller tuning
has improved its performance by decreasing the
overshoot significantly. In unit step change test of y1,
the y2s overshoot sharply falls from 0.342 to 0.056,
or going down 84%.

Table 2 PID and DMC parameters setting in WB model (2x2 process)
Wood & Berry model
(2x2 process)
PID controller tuning
(Proposed 2 method)
DMC controller
tuning

) (
1 2 . 13
9 . 4
1 9 . 14
8 . 3

) (
) (
1 4 . 14
4 . 19

1 9 . 10
6 . 6
1 21
9 . 18

1 7 . 16
8 . 12
) (
) (
3
8
3 7
3
s F
s
e
s
e
s V
s R
s
e
s
e
s
e
s
e
s Y
s Y
s
s
s s
s s
B
D
+
+
+
+
+
+
=

Loop 1: ,
,
Loop 2: , ,
,

Ts = 1.5
P = 153
M = 35
ISSN: 1411-1284

Proceeding 10
th

0
0.5
1
1.5
y
1
0 10 20 30 40 50 60 70 80 90
-0.02
0
0.02
0.04
0.06
y
2

(a) (b) (c)

Figure 1 Response of multi-loop control for WB process (step change in y1): (a) PID controller performance using Proposed
2 tuning, (b) MPC controller performance using Ts=0.1, P=10, M=2 (default setting), (c) MPC controller performance using
Ts=1.5, P=153, M=35 (tuning setting)

3.2 OR (3x3) process
The Ogunnaike and Ray (OR 3x3) process [5] as
follows is considered (see Table 3). PID controller
tuning uses BLT-4 method instead of proposed
method by Huang et. al. [5]. In this case, responses of
OR (3x3) process are unstable using their proposed
method, though they are good using BLT-4.
As shown by Table 1, the tuning strategy uses
FOPDT (first-order plus dead-time) model to
calculate Ts, P, and M. Because of OR (3x3) model
has one SOPDT (second-order plus dead-time), this
transfer function has to be changed into FOPDT. By
using PRC (process reaction curve) from step change
testing of the SOPDT and applying a method that is
developed by Cecil L. Smith [4], the FOPDT is
obtained.
Figure 2 shows responses of multi-loop control for
OR (3x3) process: (a), (b), and (c) PID controller
performance using BLT-4 tuning (unit step change in
y1, y2, y3 respectively). They have very poor
performances. This disadvantage is improved by MPC
controller.
Next, we are going to compare between responses
using default setting and tuning setting. The results
are shown by Figure 3 and Table 4. Almost
performances by tuning setting are better than by the
default setting. Only two point where the default
setting is better than the tuning setting, they are
overshoot of y2 in y1s step change, and settling time
of y2 in y2s step change.
Table 3 PID and DMC parameters setting in OR model (3x3 process)
OR model (3x3 process) PID controller tuning
(BLT-4 method)
DMC controller
tuning

Loop 1: ,
,
Loop 2: ,
,
Loop 3: ,
,

Ts = 0.71
P = 91
M = 30

ISSN: 1411-1284

Proceeding 10
th

0 50 100 150
0
0.5
1
y
1
0 50 100 150
-1
0
1
y
2
0 50 100 150
-10
0
10
y
3
0 50 100 150
0
0.1
0.2
y
1
0 50 100 150
0
1
2
y
2
0 50 100 150
-10
0
10
y
3
0 50 100 150
-0.01
-0.005
0
y
1
0 50 100 150
-0.02
0
0.02
y
2
0 50 100 150
0
1
2
y
3

(a) (b) (c)
Figure 2 Responses of multi-loop control for OR (3x3) process: (a), (b), and (c) PID controller performance using BLT-4
tuning (unit step change in y1, y2, y3 respectively)
0
0.5
1
1.5
Response: Scenario1
Output: y1
Time (sec): 53.4
Amplitude: 1
y
1
-0.2
0
0.2
Response: Scenario1
Output: y2
Time (sec): 1.6
Amplitude: -0.154
y
2
0 10 20 30 40 50 60 70 80 90 100
-0.04
-0.02
0
0.02
Response: Scenario1
Output: y3
Time (sec): 2.02
Amplitude: -0.0387
y
3
Plant Outputs
0
0.5
1
1.5
Response: Scenario1
Output: y1
Time (sec): 47.5
Amplitude: 1
y
1
-0.2
0
0.2
Response: Scenario1
Output: y2
Time (sec): 2.84
Amplitude: -0.193
y
2
0 10 20 30 40 50 60 70 80 90
-0.04
-0.02
0
0.02
Response: Scenario1
Output: y3
Time (sec): 2.84
Amplitude: -0.0241 y
3

(a) (b)
Figure 3 Responses of multi-loop control for OR (3x3) process (step change in y1): (a) MPC controller performance using
Ts=0.1, P=10, M=2 (default setting); (b) MPC controller performance using Ts=0.71, P=91, M=30 (tuning setting)

Table 4 Performance of MPC controller of OR (3x3 process) model
Default Setting Tuning Setting
Step
Change
Controlled
Variable Settling time Overshoot Settling time Overshoot
y1 53.4 0.08 47.5 0.02
y2 -0.153 -0.193
y1
y3 -0.0386 -0.0241
y1 -0.156 -0.138
y2 21.7 0.01 31 0.01
y2
y3 -0.151 -0.0552
y1 -0.0303 -0.0273
y2 -0.0852 -0.0614
y3
y3 6.52 0 3.18 0

3.3 Alatiqi case 1 (4x4) process
Consider alatiqi case 1 (A1 4x4) process (see
column 1 in Table 5). Alatiqi process has 16 empirical
models that consist of three FOPDT models, two
models having zeros, and eleven SOPDT models. All
of empirical models have to be changed in FOPDT
model to calculate MPC controller tuning. We use
some PRCs (process reaction curves) from step
change testing of the SOPDTs and the models having
zeros, and apply a method that is developed by Cecil
L. Smith [3] btain the FOPDT.
PI controller shows the poor performance (see
Figure 4). Also, the controller performance as shown
by Figure 5 is extremely poor, because the responses
are unstable. So, in this case, the default setting can
not be used in MPC controller. Inevitable, we have to
use the tuning setting in this case.

ISSN: 1411-1284

Proceeding 10
th

Table 5 PID and DMC parameters setting in Alatiqi case 1 model (A1 4x4 process)

Alatiqi case 1 model (A1 4x4 process)
PI controller tuning
(Lee et al. method)
DMC controller
tuning

Loop 1: ,
Loop 2: ,
Loop 3: ,
Loop 4: ,

Ts = 1.62
P = 191
M = 52

0 20 40 60 80 100 120 140 160 180 200
0
0.5
1
y
1
0 20 40 60 80 100 120 140 160 180 200
-0.01
0
0.01
y
2
0 20 40 60 80 100 120 140 160 180 200
-0.5
0
0.5
y
3
0 20 40 60 80 100 120 140 160 180 200
-2
-1
0
y
4
0
1
2
y
1
-0.5
0
0.5
y
2
-0.05
0
0.05
y
3
0 10 20 30 40 50 60 70 80 90
-0.5
0
0.5
y
4
Plant Outputs

(a) (b)
Figure 4 Responses of multi-loop control for alatiqi case 1 (A1 4x4) process: (a) PI controller performance using Lee et. al.
tuning (unit step change in y1); (b) MPC controller performance using tuning setting Ts=1.62, P=191, M=52 (unit step
change in y1)

-5
0
5
x 10
5
y
1
-2
0
2
x 10
4
y
2
-100
0
100
y
3
0 10 20 30 40 50 60 70 80 90 100
-4
-2
0
x 10
4
y
4
Plant Outputs

Figure 5 Alatiqis Responses of MPC controller using
default setting

IV. CONCLUSSIONS

MPC controller offers better control performances
than PI/PID controller, especially in multivariable
processes. Application of MPC controller in the three
more complex models produces the fantastic
performance. To achieve an optimum performance of
MPC controller, non-adaptive DMC controller tuning
can be used. This method has proved that the tuning
setting has the optimum performance. In the complex
processes like WB model (2x2 process) and OR
model (3x3 process), the default setting can be used in
MPC controller. But, in more complex process likes
Alatiqi case 1 model (A1 4x4 process), the default
setting produces very poor performance. So, we must
use the tuning setting.

REFERENCES

[1] D.W. Clarke. Adaptive Predictive Control. A
Rev. Control. Vol. 20, pp. 83-94, 1996
[2] S.J. Qin and T.A. Badgwell, (2003). A survey of
industrial model predictive control technology.
Control Engineering Practice. (11): 733 764
[3] C. R. Cutler & D. L. Ramaker. Dynamic matrix
controla computer control algorithm.
Proceedings of the JACC 1980. San Francisco,
1980.
[4] T. Marlin. Process Control: Designing Processes
and Control Systems for Dynamic Performance.
2nd Edition,McGraw-Hill, New York, 2000.
[5] Huang et. al. A direct method for multi-loop
PI/PID controller design. Journal of Process
Control (13): 769786, 2003.
[6] R. Shridhar & D. J. Cooper. A tuning strategy for
unconstrained SISO model predictive control.
Industrial & Engineering Chemical Research.
(36): 729746, 1997.
ISSN: 1411-1284

Proceeding 10
th

[7] R. Shridhar & D. J. Cooper. A tuning strategy for
unconstrained multivariable model predictive
control. Industrial & Engineering Chemical
Research. (37): 40034016, 1998.
[8] Danielle Dougherty and Doug Cooper. A
Practical Multiple Model Adaptive Strategy for
Multivariable Model Predictive Control. Control
Engineering Practice. (11): 649 664, 2003.
[9] B. Huang., Steven X. Ding, and Nina Thornhill.
Alternative solutions to multi-variate control
performance assessment problems. Journal of
Process Control (16): 457471, 2006.

ISSN: 1411-1284

Proceeding 10
th

Biogrease Using Modified Palm Oil as Base Oil and
Thickener Lithium Soap

Sukirno *, M.Nasikin, B.Heru, Rizqon F, Marius, Dizzi.
* Fac. of Engineering, University of Indonesia, Jl. Salemba Raya 4, Jakarta 10430
Tel. 7863516, fax. 7863515 email : sukirnos@che.ui.edu

Abstract Biogreases without additives, have been
prepared using biolubricant, as the base oil. This
bio lubricant is a modified palm oil which can be
synthesized from palm oil via transesterification and
epoxidation reactions. The thickening agent used in
this grease making study was lithium 12-
hydroxystearate. The making procces of the
biogrease was carried out in a closed flanged flask
or autoclave. Characterizations of the biogrease
were carried out using Penetration Test, Dropping
Test. It was found that lithium biogrease NGLI 2
can be prepared using thickener less that was
needed by mineral oil grease. Performance test by a
Gear Load-carrying test showed that the biogrease
products has better antiwear performance than that
of lithium grease using mineral oil HVI 160 S as the
base oil.

Keywords Biogrease, lithium soap, base oil.
Penetration Test, Dropping Point Test, Gear Load-
carrying Test.

I. INTRODUCTION

ubrication is proccess of interposing a lubricating
material between two surfaces in relative motions in
order to reduce friction
[1]
. Liquid lubricants are the most
known type of lubricanting material, however semi
solid lubricant or grease is widely used in many
lubrication systems due to enomical consideration.
Lubricating greases are, in general, produced by
dispersing a solid thickening agent in a lubricating
fluid. The base oil selected in formulating a grease
should have the same characteristics, as if the
equipment is to be lubricated by the fluid lubricant.
Hence a fundamental difference between a formulated
fluid lubricant and a similarly fully formulated grease,
is the presence in of the thickening agent. These
thickeners are usually soaps, such as lithium, sodium
and calcium salts of long chain fatty acids, and the
more recently introduced thickeners, such as clays,
polyureas
[3]
. The same as lubricating oil, greases also
contain some additives in order to enhance its
performance and to protect the grease and lubricated
surfaces.
[1]
. The most common additives found in
grease are antioxidants to prolong the life of a grease,
anti-corrosion agents to protect metal against attack
from water, sulphides or corrosive elements, extreme
pressure agent to guard against scoring and galling,
antiwear agent to prevent abrasion and metal to metal
contact etc. So base oil, thickener dan addives are
blended to make a grease Typical grease composition
are shown in table 1.

Table l Typical grease composition
(%)
Base oil 75-95
Thickener 5-20
additives 0-10

Actually, the component that perform as lubricant
is the base oil
[3]
. The thickener gives grease its
characteristic rigidity or consistency and is sometimes
thought of as a three-dimensional fibrous network or
sponge that holds the oil in place. During the grease
in use, when the lubricant film between wearing
surfaces thins, the friction heat softens the adjacent
thickener, which expands and releases oil to restore
film thickness. Soap thickeners not only provide
consistency to grease, they affect desired properties
such as water and heat resistance and pumpability.
Fluid lubricants used to formulate grease are
normally petroleum or synthetic oils. Synthetic oils are
higher in cost, so that the majority of greases on the
market are composed of mineral oil. With growing
environmental concerns, vegetable oils are also being
used in applications requiring nontoxic or
biodegradable greases. In America, they are developing
soybean oil. For Indonesia, as the main producer of
palm oil, we should develop palm oil.
In our previous research, we had successfully
prepared a biolubricant via transesterification and
epoxidation reaction of palm oil to a lubricant which
has better oxidation stability than its feed, a RBDPO
olein. This modified palm oil is called called EPOME
Gliserol
[2]
which is biodegradable, and contain no
aromatic, nitrogen, sulfur, and is expected to be used in
foodgrade lubricant formulation. In this present reseach,
the biolubricant is used as base oil for grease making.
The biogrease product is supposed to be better in its
antiwear propertiy, since its has many active
functionality to protect friction surfaces.
The most important feature of a grease is its
rigidity or consistency. A greases consistency is its
resistance to deformation by an applied force. A grease
that is too stiff may not feed into areas requiring
lubrication, while a grease that is too fluid may leak
out. The measure of rigidity or consistency is called
penetration which is measured by penetration test. To
measure penetration, a cone of given weight is allowed
L
ISSN: 1411-1284

Proceeding 10
th

to sink into a grease for 5 seconds at a standard
temperature of 25 C). The depth, in tenths of a
millimeter, to which the cone sinks into the grease is
the penetration. A penetration of 100 would represent a
solid grease while one of 450 would be semifluid The
NLGI has established consistency numbers or grade
numbers, ranging from 000 to 6, corresponding to
specified ranges of penetration numbers. Table 1 lists
the NLGI grease classifications along with a description
the consistency of each classification.

Tabel 2 NLGI Number
NLGI Worked Penetration (Grade)
at 25
0
C (0,1mm)
Virtual Assessment of
Deformability
000 445-475 Very soft
00 400-430 Very soft
0 355-385 Soft
1 310-340 Creamy
2 265-295 Almost solid
3 220-250 Hard
4 175-205 Hard
5 130-160 Very hard, as soap
6 85-115 Very hard, as soap

The purpose of this research is to prepare
biogrease of NGLI 2 which is normally used for gear
and wheel bearing application. In this preliminary study
of making grease, the amount of the thickener is varied
to observe their rigidity or consistency and dropping
points in order to make a graph shown the effect of the
amount of thickener to the biogrease characteristics,
and its lubrication performance in comparison with
similar grease using mineral oil as base oil.

II. EXPERIMENTAL

Materials used
(a) Base oil
The base oil used was a bio lubricant synthesized
from palm oil. It was the product of previous research
and its properties are given in Table 1. The Mineral oil
HVI 160 Pertamina product was also used as base oil
comparison.

Table 1 Properties of base oils
Appearance Yellow pale Yellow
Specific gravity [-] ASTMD-1289 0.91 0.85 0.8
Viscosity @40C cSt ASTMD-445 35 38.9 96
Viscosity @100C cSt ASTMD-446 6.9 8-9 11
Viscosity Index [-] ASTMD-2270 >100 >100 100
Pour point C ASTMD-97 0 8-15 -9
Oxid. Test ( Vis) Bulk Oxidation 5.6 21.5 7.8
Test Method
Biolubricant
(Modified
RBDPO)
RBDPO
M.O
HVI 160
S
Characterization

(b) Thickener
Thickener lithium soap was obtained from the
reaction of fatty acid 12-hydroxystearate (melting point
77.5C).and lithium hydroxide.

Preparation of greases
A mixture of lithium 12-hydroxystearate and
approximately 85% of the bio lubricant to be used in
the manufacture of the grease was heated in a flanged
flask or autoclave, equipped with a stirrer, pressure
indicator, oil heater thermometer, to 120C and stirred,
as shown in figure 1. The lithium hydroxide was added
slowly until solution of the soap occurred (170C).
Heating was continued until the soap melted (200C).
Next, the solution was cooled in a cooler to
approximately 70C with vigorous stirring. atmosphere.
The remainder of the base oil, is then added to give the
final product. Stirring was continued for a further about
1 h, and the product, now a creamy grease, was milled
in Miller..
In this experiment, the amount of the thickener is
varied, and additives has not been added.

Fig.1 Autoclave

Biogrease Product Characterizations
Samples of the greases were tested with
Penetrometer (ASTM D 217), Dropping Point (ASTM
D 566) Water washout (ASTM D 1264).

Biogrease Product Performance Test
Performance tests of biogreases were conducted
using Gear Load-carrying tests. The method used to
assess the load-carrying characteristics of the greases in
this experiment was the gear load-carring test. In this
test, a pair of gear were used. Each test was carried out
at ambient temperature with 10 hours runs at 25 rev/s
and the applied load of 10 kg. The wear debris is
measure by AAS.

ISSN: 1411-1284

Proceeding 10
th

Fig.2 Gear load-carrying test

III. RESULTS AND DISCUSSION

Product Appearance
Figure 3 shows a grease product prepared in
autoclave, as seen under ordinary camera. It is creamy
and the texture is soft and sticky.

Fig.3 Biogrease

Biogrease rigidity
Fig.4 shows the plot between penetration and
composition of the thickener lithium soap obtained
from the biogrease of this experiment. In the figure,
there is also given a plot of a grease using mineral oil,
as a comparison. From the figure can be seen that,
grease consistency increase as the amount of thickener
increase, and also with the same amount of thickener
the biogrease is more rigid than that of petroleum
grease. This means that to produce the same level of
rigidity, the biogrease need smaller amount of the
thickener than that of petroleum grease.

100
150
200
250
300
0 5 10 15 20 25
%Lithium Soap
P
e
n
e
t
r
a
t
i
o
n

(
0
,
1
m
m
)
Lithium Grease HVI 160 Lithium Biogrease

Fig.4 Penetration Value of Li Biogrease

The Biogrease Dropping Point.
Dropping point is an indicator of the heat
resistance of grease. As grease temperature rises, the
grease gets softer, until liquefies and the rigidity is lost.
Dropping point is the temperature at which grease
becomes fluid enough to drip. The dropping point
indicates the upper temperature limit at which grease
retains its structure, not the maximum temperature at
which a grease may be used.
The dropping point measurement of the biogrease
are shown in fig 5. Dropping point increases as the
amount of thickener increases, and for the same amount
of thickener lithium soap, the dropping point of
biogrease is higher than that of mineral oil grease.

100
120
140
160
180
200
220
240
0 5 10 15 20 25
%Lithium Soap
D
r
o
p
p
i
n
g

p
o
i
n
t

C
Li Biogrease Li Grease HVI 160

Fig.5 Dropping Point of Li Biogrease

This experiment also tried to make semibio grease,
where the base oils are mixture of biolubricant and
mineral oil. Figure 3 shows the result of the dropping
point test of semibio grease with various base oil
compositions. The dropping point of semiobio grease
increases as the % of biolubricant increases.

ISSN: 1411-1284

Proceeding 10
th

Li Grease semibio (Biolubricant +HVI 160S)
0
50
100
150
200
250
0 20 40 60 80 100
% Bio lubricant (Modified Palm Oil)
D
r
o
p
p
i
n
g

p
o
i
n
t

(
d
e
r

C
e
l
s
i
u
s
)

Fig.6 Dropping Point of Lithium Semibio Grease

Table 6 lists the results of quality tests of biogrease
in comparison with that of the mineral oil grease.
Lithium biogrease is better than lithium petroleum
grease, in case of their dropping point. But it is inferior
in water washout tests. This test is the ability of grease
to withstand the effects of water with no change in its
ability to lubricate. Water may suspend the oil in the
grease, forming an emulsion that can wash away or, to a
lesser extent, reduce lubricity by diluting and changing
grease consistency and texture. Rusting becomes a
concern if water is allowed to contact iron or steel
components.

Table 2 Characteristic Comparison of BioGrease and
Mineral Oil Grease
Li Soap (%) 15 15 10
Penetration (0,1 mm) 220.7 148 229
NLGI number 3 5 3
Dropping point (
0
C) 221 230 210
Water washout (%) 5 <10* <15*
Li Grease
( HVI 160 S
Pertamina)
Lithium Biogrease

Gear Load-carrying test
The result of performance tests of the biogrease
using gear load-carrying test are shown in Figure 7. In
this figure the amount of wear debris (ppm of Fe) were
plotted against % biolubricant. It can be seen that, as
the composition (%) of biolubricant increases the wear
amount (in ppm of Fe) decreases. It means that
biolubricant can improve the performance of petroleum
grease when is used in grease formulation.

0
50
100
150
200
250
Lithium 15% Lithium 15% Lithium 15% Lithium 15% Lithium 10%
Biolubricant
40% HVI
Biolubricant
60% HVI
Biolubricant
80% HVI
Biolubricant
100% HVI
Biolubricant
100% HVI
p
p
m

F
e

Fig.7 Wear Rate of Li Biogrease using Gear load-
carrying test.

IV. CONCLUSIONS

From this preliminary study of making biogrease
with modified palm oil biolubricant as the base oil,
several conclusions can be summarized as follow:
a. The amount of lithium soap needed is .10 % of
total weight to produce a biogrease NGL 2 using
modified palm oil as the base oil. To reach the
same level of rigidity consistency, the biogrease
without additives needs smaller amount of the
thickener than that of petroleum grease.
b. In gear load-carrying tests, the biogrease without
additives showed better antiwear performance than
that of petroleum grease, with HVI 160 S as the
base oil.

ACKNOWLEDMENTS
This work was funded by RUSNAS Industri Hilir
Kelapa Sawit. The Authors would like to express their
appreciation all who supports to this study.

REFERENCES

1. OBrien, J.A. , Lubricating Oil Additives, Booser, E.R.,
editor., CRC Handook of Lubricating (Theory and
Practice Tribology). Volume 2, Theory and Design, CRC
Press Inc. Boca Raton, Florida, 1983.

2. Fenjerry, Y. "Preparation and Characterization of Epome
Gliserol dan Epome Monoalcohol as Foodgrade
Lubricant", Bachelor Thesis, Chemical Engineering
Department, University of Indonesia, Depok, 2005.

3. H. B. SILVER and I. R. STANLEY, The Effect of The
Thickener on the Efficiency of Ioad-Carryng Additives in
Greases, TRIBOLOGY international June, pp.113-118,
1974.

ISSN: 1411-1284

Proceeding 10
th

Model for Oxidation and Combustion of Ethanol in Wide
Range of Pressures, Temperatures and Equivalence Ratios

Yuswan Muharam and Nonni Soraya Sambudi

Department of Chemical Engineering, Faculty of Engineering, University of Indonesia,
Kampus UI Depok 16424
Phone 021-7863516, fax. 021-7863515, email: muharam@che.ui.edu

Abstract Ethanol as an alternative fuel can
be used either in a mixture or single component. The
advantage of ethanol over gasoline as fuel is the
minimum standard pollutant emitted from its
burning, such as carbonmonoxide. Ethanol oxidation
and combustion processes use 372 elementary
reactions and yield various intermediates and
products.
This computational study has several purposes to
find out, such as ignition delay times, emitted
pollutants and influences of pressure, temperature
and equivalence ratio on ethanol oxidation and
combustion. Hence, a comprehensive and
representative chemical kinetic model for ethanol
oxidation and combustion is needed to fulfill the
purposes of this study. This model is expected for
high validation and represents the real condition of
ethanol oxidation and combustion. The
computational results are compared to the
experimental data from flow reactors and jet stirred
reactors for concentration profiles, and shock tube
for ignition delay times. Good agreement was found
for the validation with experimental data.
Mechanism analyses were doing by applying the
sensitivity and reaction flow analyses. Sensitivity
analysis is used to identify rate-limiting reaction
steps. Reaction flow analysis calculates the
percentage of reaction contribution to the formation
or consumption of chemical species. Sensitivity
analysis shows that OH O+ H O
2
+ is the most
sensitive reaction for ignition delay times.
Ethanol oxidation and combustion were
simulated within the temperature range of 1000-2000
K, the pressure range of 1-15 atm and the
equivalence ratio ( ) range of 0.5-2.0. The simulation
was performed using Chemkin 3.7.1 software.

Keywords ethanol oxidation and combustion, kinetic
model, ignition delay times

I. INTRODUCTION

he usage of fossil fuels suffers pollution problems
that affect the environment. This has become the
main issue of energy resource exploitation. Energy
alternative rises as the solution for less standard
pollutant and sustainable energy. Bioethanol is an
energy alternative which usage is expected to solve the
energy problems.
Based on the National Energy Mix, the Government
has planned the usage of biodiesel and bioethanol in
amount of approximately 2% of overall national fuels
usage in 2010, and will rise up to 5% in 2025 [1].
Previous chemical kinetic modeling studies of
ethanol oxidation were performed by Natarajan and
Bhaskaran [2]. The following modeling works focused
on problems of ethanol ignition delay in shock tubes,
and ethanol laminar flame speeds in burners and jet
stirred reactors [2,3,4,5]. The time needed from the start
of injection and the start of combustion is referred to as
the ignition delay.
Egolfopoulos [4] modeling study demonstrated a
good agreement with Natarajan [2] ignition delay data
and found (
M) OH( H C
5 2
+ M) OH( CH CH
2 3
+ +
) as the
most sensitive reaction for ignition delay. Curran [6] did
not find it in his modeling study at 1100-1900 K and 2-
4.5 bar. Detailed mechanism is generally validated with
experimental data from certain range of operating
condition [7].
So far, the modeling of ethanol oxidation and
combustion had not achieved wide validation range.
Hence, a comprehensive mechanism of ethanol
oxidation and combustion in wide range of operating
condition (pressure, temperature and equivalence ratio)
is needed to be validated with various experimental data
and represents the actual combustion system.

II. BASIC THEORY

Chemical reactions occur when molecules of one
species collide with molecules of another species and,
for some of these collisions, one or more new molecules
T
ISSN: 1411-1284

Proceeding 10
th

will be created. In the chemical reaction the atoms of
reacting molecules are redistributed in the new
molecules. To achieve this, the reacting molecules must
have sufficient kinetic energy so that their chemical
bonds can be broken during the impact and other bonds
can be formed. As the energy of these bonds depends on
the nature of the atoms and on geometrical factors, the
energy content of the products of the collision may be
different from the energy content of the colliding
molecules. This is the basis for heat being released or
absorbed in chemical reactions [8].

II.1 Temperature Dependence of Rate Coefficients
The reaction rate coefficient is given by the
Arrhenius law:

=
RT
E
A k
a
exp'

(1)

where E
a
is the activation energy. Not all molecular
collisions will result in reaction, but only those with
kinetic energy higher than the energy needed to break
the bonds of the reactant molecules. This energy barrier
is the activation energy. Its maximum value corresponds
to the bond energies in the molecule. In dissociation
reactions, the activation energy is approximately equal
to the bond energy being broken. The value of
activation energy may also be much smaller or even
zero [9].
Since most elementary binary reactions exhibit
Arrhenius behavior over modest ranges of temperature,
the temperature dependence can usually be incorporated
with sufficient accuracy into the exponential alone.
However, for the large temperature ranges found in
combustion, non-Arrhenius behavior of the rate
coefficient tends to occur, particularly for processes that
have a small energy barrier. Therefore, it is necessary to
adopt a modified Arrhenius form which expresses the
impact of temperature on the rate coefficient [10],
=
RT
E
AT k
b a
exp
. (2)

II.2 Pressure Dependence of Rate Coefficients
The Lindemann mechanism illustrates the fact that
reaction orders of complex (nonelementary) reactions
depend on the conditions chosen. However, the
Lindemann mechanism itself is a simplified model.
If the rate law of a unimolecular reaction is written
as
[A] d[P]/d
a
k t =
, then the rate coefficient k depends on
pressure and temperature. The theory of unimolecular
reactions yields fall-off curves, which describe the
pressure dependence of k for different temperatures. The
logarithm of the rate coefficient is usually plotted versus
the logarithm of the pressure [9].
Typical fall-off curves are shown in Fig. 1. At very
high pressures
) [M] [M]/(
u a - a u
k k k k k + =
tends to the
limit
k
, i.e., the rate coefficient becomes independent
of the pressure. At very low pressure, the rate
coefficient k is proportional to [M] = p/RT, resulting in a
linear dependence. Similarly, if the effective activation
energy is low, the reaction rate coefficient k will
decrease with temperature [9].

Fig.1. Fall of curves for unimolecular reaction
3 3 6 2
H C H C H C
& &
+
[9]

II.3 Ignition Delay Time
Ignition delay is characteristic for radical chain
explosions. During the ignition delay period, the radical
pool population is increasing at an exponential rate. Yet,
the amount of fuel consumed, and hence the amount of
energy liberated, is too small to be detected. Thus,
important chemical reactions (chain branching,
formation of radicals) take place during the induction
time, whereas the temperature remains nearly constant
[9].
Finally, the radical pool becomes large enough to
consume a significant fraction of the fuel, and rapid
ignition will take place. The precise definition of
induction time depends on the criterion used
(consumption of fuel, formation of CO, formation of
OH, increase of pressure in a constant volume vessel,
etc.) [9].
Due to the temperature dependence of the
underlying elementary reactions, the ignition delay time
depends strongly in the temperature. This is shown in
Fig. 2 for several hydrocarbon-air mixtures [9].

ISSN: 1411-1284

Proceeding 10
th

Fig.2. Calculated (line) and measured (points) ignition delay
times in hydrocarbon-air mixtures [9]


III.1 Mechanism Validation
A mechanism consisting 372 elementary reactions
for ethanol oxidation and combustion has been
developed in this study. The mechanism validation was
performed by comparing calculation results to
experimental measurements for shock tubes, jet stirred
reactors and flow reactors available in literature.

10
100
1000
6 6,5 7 7,5 8
10000/T (K
-1
)
I
g
n
i
t
i
o
n

d
e
l
a
y

t
i
m
e
s
(
m
i
k
r
o
d
e
t
i
k
)

Fig. 3. Comparison between shock tube ignition delay data as
investigated by Natarajan and Bhaskaran [2] and the
numerical calculations using the chemical kinetic model.
Experimental conditions: 2.5% C2H5OH, 7.5% O2, and 90%
Ar, = 1.0,
P = 1 0.2 and 2.0 0.2 atm

As shown in Fig. 3, ignition delay time is shorter at
higher pressure and temperature. This behavior was
reproduced very well by the model.
Figure 4 shows temperature behavior for ignition
delay time in shock tube. Pressurized wave resulted in
shock tubes will increase the system pressure hence
temperature. The rise of pressure and temperature will
improve system reactivity and radical pool. Formation
of large enough radical pool will consume significant
fraction of the fuel and rapid ignition will take place.

Fig.4. Temperature behaviour from ignition delay time in
shock tube

Aboussi [11] investigated ethanol oxidation in jet
stirred reactors. The experimental data were used to
validate the kinetic model by comparing it with
numerical calculations.

0,000001
0,00001
0,0001
0,001
0,01
0,1
0 0,05 0,1 0,15 0,2 0,25 0,3
Residence time (s)
F
r
a
k
s
i

m
o
l
C2H5OH sim C2H4 sim CO sim CO2 sim
C2H5OH exp C2H4 exp CO exp CO2 exp

Fig.5. Ethanol-O
2
-N
2
(0.3%0.9%98.8%) oxidation
in a jet stirred reactor at = 1.0, 1.0 atmosphere,
and a nominal temperature of 1056 K. Measurements and
predictions for C
2
H
5
OH, C
2
H
4
, CO and CO
2

Figure 5 shows the comparison between numerical
calculations for C
2
H
5
OH, C
2
H
4
, CO and CO
2
. Overall,
the modeling results show good agreement with the
species concentration as the residence time was varied.
Norton and Dryer [12] investigated ethanol
oxidation in turbulent flow reactors. The experimental
data were used to validate the kinetic model by
comparing it with numerical calculations using the
chemical kinetic model.

0
0,02
0,04
0,06
0,08
0,1
0,12
0 0,02 0,04 0,06 0,08 0,1 0,12
Residence time (s)
F
r
a
k
s
i

m
o
l
C2H5OHexp O2 exp COexp H2Oexp H2 exp CO2 exp
C2H5OHsim O2 sim COsim H2Osim H2 sim CO2 sim
O
2
0,8
C
2
H
5
OH 2
H
2
O
CO
H
2
CO
2

Fig.6. Comparison between experimental flow reactor
oxidation data for = 1.24 as investigated by Norton and
Dryer [12] and the numerical calculations. The numerical
results were time shifted by 40 msec. Experimental
conditions: 5.81% C2H5OH, 14.07% O
2
, and 80.12% N
2
,
ISSN: 1411-1284

Proceeding 10
th

Reynolds Number = 4900, P = 1atm, and T
in
= 1100 K.
Numerical simulations shown for C
2
H
5
OH, O
2
, CO, CO
2
, H
2
,
and H
2
O.

Fig.6. shows good agreement between experimental
data and numerical results for C
2
H
5
OH, O
2
, CO, CO
2
,
H
2
and H
2
O in temperature of 1100 K and equivalence
ratio of 1.24.

III.2 Mechanism Analyses
III.2.1 Sensitivity Analysis
A sensitivity analysis is used to identify rate-limiting
reaction steps. Numerical calculations for sensitivity
analysis were carried out by changing the rate
coefficient of a particular reaction and calculating the
OH concentration.
-0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 0,8 1,0 1,2
1
= 0,5
= 2,0

Fig.7. Sensitivity coefficients of the important reactions for
ignition delay time in lean and rich fuel mixture, based on
Dunphy dan Simmie[3] experimental condition in shock tube.
P = 3.4 bar, T = 1430 K,
1
= 0.5 (1.25% C
2
H
5
OH, 7.5% O
2

and 91.25% Ar),
2
= 2.0 (2.5% C
2
H
5
OH, 3.75% O
2
and
93.25% Ar)

Figure 7 shows sensitivity coefficients of the
important reactions for ignition delay time. Six most
sensitive reactions for lean and rich fuel mixture are:
(+) OH O+ H O
2
+

(+)
2 3
HO CH + OH O CH
3
+
(+) M HCO+ M CO H + +
(+)
2
O HCO+
2
HO CO+
(+)
M O CH CH
2 3
+ M H HCO CH
3
+ +

(-)
M O CH CH
2 3
+ M O CH CH
2 3
+ +
A sensitivity coefficient might be positive or
negative. A positive sensitivity coefficient indicates a
higher OH concentration and an increased overall
reaction rate, and a negative sensitivity coefficient
indicates a lower OH concentration and a decreased
overall reaction rate of the system.
-0,5 0 0,5 1 1,5
1
CH3+HO2=CH3O+OH
O+OH=O2+H
H+HO2=H2+O2
CH3+CH3(+M)=C2H6(+M)
CH3+HO2=CH4+O2
C2H5OH(+M)=CH3+CH2OH(+M)
C2H5OH(+M)=C2H4+H2O(+M)
C2H5OH+OH=C2H4OH+H2O
C2H5OH+CH3=CH3CHOH+CH4
C2H5OH+HO2=CH3CHOH+H2O2
CH3CH2O+M=CH3HCO+H+M
CH3CH2O+M=CH3+CH2O+M

Fig.8. Sensitivity coefficients of the important reactions for
ethanol oxidation in a flow reactor. Experimental condition:
=0,61 (5.65% C
2
H
5
OH, 27.86% O
2
and 66.49% N
2
), P = 1
atm and T
in
= 1092 K

Figure 8 shows sensitivity analysis results for =
0.61 in flow reactors. Several most sensitive reactions
ethanol oxidation in flow reactor are:
(+) OH O+ H O
2
+
(+)
2 3
HO CH + OH O CH
3
+

(-)
2 3
HO CH +
2 4
O CH +

(-)
2
HO H+
2 2
O H +
(-)
M O CH CH
2 3
+ M O CH CH
2 3
+ +

(+)
M O CH CH
2 3
+ M H HCO CH
3
+ +

III.2.2 Reaction Flow Analysis
A reaction flow analysis calculates the percentage of
reaction contribution to the formation or consumption of
chemical species.

+O2
5,2%
+O2, M
83,7 %
+OH, HO2
97,8 %
+H, OH, O
64,1 %
+OH, O, H
68,3%
+OH
29,1%
+O
24%
+O2
11,1%
+O, O2
40,3%
+OH, H, O
65,1%
+O
27,8%
+M
46,8%
+O2
46%
+O
64,6%
+H
11,2 %
CO
HCO
CH2O
CH2OH CH3
CH3OH
CH2CO
CH4
C2H4
CH2HCO
CH3HCO
CH3CH2O CH3CHOH
CO2
ISSN: 1411-1284

Proceeding 10
th

Fig.9. The pathways of CO
2
formation in flow reactor at P = 1
atm, equivalence ratio () = 0.61 and T = 1092 K

Figure 9 shows carbondioxide formation from its
reaction with OH and HO
2
radical.
OH CO+ H CO
2
+
92.6%
2
HO CO + OH CO
2
+ 5.2%
Carbonmonoxide is resulted from radical formyl
oxidation, formyl decomposition with third body, and
oxidation and decomposition of CH
2
HCO.
2
O HCO +
2
HO CO+ 75.7%
M HCO+ M CO H + + 8%
2 2
O HCO CH + OH CO O CH
2
+ +
3.3%
HCO CH
2
CO CH
3
+
1.9%

IV. CONCLUSIONS

Validation of the mechanism for oxidation and
combustion of ethanol in shock tubes and flow and jet-
stirred reactors shows good agreement. The elementary
reaction OH O+ H O
2
+ is shown as the most
sensitive reaction for ignition delay time at 1430 K, 3.4
and 5 bar, and the equivalence ratios of 0.5 and 2.0. It is
also the most sensitive reaction for ethanol oxidation
rate in flow reactor at 1092 K, 1 atm, and the
equivalence ratio of 0.61.
Ignition delay time in shock tube is shorter at high
pressure and fuel lean condition. Lowest species
concentrations in flow and jet sstirred reactor are found
at low equivalence ratio, low pressure and temperature.

REFERENCES

[1] Prasetyo E. Energi Alternatif Ramah Lingkungan.
Accesed on 22 Mei 2006, from Pikiran Rakyat.
http://www.pikiranrakyat.com/cetak/2005/0705/15/otokir/ut
ama01.htm, 2005.
[2] Marinov N.M . A Detailed Chemical Kinetic Model for
High Temperature Ethanol Oxidation. International
Journal of Chemical Kinetic, pp.183, 1999.
[3] Dunphy M.P, J.M. Simmie. Journal Chemical Society
Faraday Trans 87 pp.16911695, 25492559, 1991.
[4] Egolfopolous F.N, D.X. Du, C.K. Law. Twenty-Fourth
Symposium (International) on Combustion. The
Combustion Institute, pp. 833, 1992.
[5] Dagaut P, M. Cathonnet, J.C. Boettner. Journal Chemistry
Physic 89, pp.867884, 1992.
[6] Curran H.J, et al. Twenty-Fourth Symposium
(International) on Combustion. The Combustion Institute,
pp. 769, 1992.
[7] Held T.J, F.L. Dryer. A Comprehensive Mechanism for
Methanol Oxidation. International Journal Chemical
Kinetic 30, pp.805, 1998.
[8] Muharam Y. Detailed Kinetic Modelling of the Oxidation
and Combustion of Large Hydrocarbons Using an
Automatic Generation of Mechanisms. Dissertation,
University of Heidelberg, Heidelberg, pp.1-27, 2005.
[9] Warnatz J, U. Maas, R.W. Dibble. Combustion, Physical
and Chemical Fundamentals, Modelling and Simulation,
Experiments, Pollutant Formation 3rd ed. Heidelberg:
Springer, 2001.
[10] Glassman I. Combustion 3rd edition. New York:
Academic Press, 1996.
[11] Aboussi B. Dissertation, Orleans, 1991.
[12] Norton T.S, F.L. Dryer. An Experimental and Modeling
Study Of Ethanol Oxidation Kinetics in an Atmospheric
Pressure Flow Reactor. International Journal Chemical
Kinetic 24, pp.319344, 1992.
ISSN: 1411-1284

Proceeding 10
th

WATER AND AIR CIRCULATION SYSTEM IN
TRIMARAN FISHING VESSEL LIVE FISH HOLDS

Sunaryo, Anita Muslih

Naval Architecture Study Program, Mechanical Engineering Department, Faculty of Engineering, University of
Indonesia.
Kampus UI, Depok 16424, tel: 021-7270032, fax: 021- 7270033
naryo@eng.ui.ac.id

ABSTRACT- Indonesia has abundant of natural
resources including fish and aquaculture, but so
far the effort for exploiting it has not optimal yet,
such as in providing sufficient fishing fleet for the
national fishing industry, and the enabling of
traditional fishermen community to improve their
living standard. For this purpose sea fish farming
is introduced to some potential fishermen
communities. One type of the fish being farmed is
grouper, which has high market value to be
exported to the Far-East. But the fish in demand
in the market is live fish, therefore triggered by
this challenge as a continuation of the previous
research, this research is carried out focusing in
the arrangement of the water and air circulation
system in the Trimaran fishing vessel live fish hold,
so that the fish being transported could be kept
alive until they arrive at the destination.
Some requirements should be fulfilled in order to
keep the fish (in this case grouper fish) alive such
as: The density of the fish per m
3
of water in the
fish hold, cleanliness of the water in the fish hold,
salinity of the water, water temperature, mineral
content and concentration in the water, oxygen
and nitrogen contain, and the water circulation
speed.
Since the vessel was designed as a trimaran, all
the propulsion machinery are installed in the side
hulls, and the middle hull is dedicated for the
storing of fish, the water and air circulation system
is arranged in the middle body of the middle hull.
Two pumps are utilized one for circulation fresh
sea water and the other for discharging the
contaminated or dirty water. Air is blown in from
a compressor through distribution pipes to all the
fish holds. For keeping the right temperature in
the fish hold sea water is sucked from the sea chest
and passed through a chiller prior being flown into
the fish hold.
With 16.3 ton fish hold capacity it is estimated
that 0.38 ton of live grouper fish could be
transported.

Key words - trimaran, fishing vessel, fish hold, water
and air circulation

I. INTRODUCTION

Indonesia has abundant of natural resources
including fish and aquaculture, but so far the effort for
exploiting it has not optimal yet and the traditional
fishermen are still using traditional ways of fishing
methods. In order to improve the fishermens standard
of living various efforts have been done among others
is by introducing fish farming for certain types of fish
that have prospective in global market, grouper is one
of such kind of fish. But grouper has to be exported
live, and based on this requirement the research is
focused on the method for transporting the live fish.
The type of boat for carrying the fish is in the form of
trimaran boat, which is considered more stable
compared to the mono hull. In order to keep the fish
alive the water and air circulation in the fish hold
should be well maintained, this include the density of
the fish per m
3
of water in the fish hold, cleanliness of
the water in the fish hold, salinity of the water, water
temperature, mineral content and concentration in the
water, oxygen and nitrogen contain, and the water
circulation speed.
The objectives of the research are: to introduce
the novel design of trimaran vessel, which has its
propulsion engines installed in the side hulls, and the
middle hull is fully used for carrying the live fish, to
design the water and air circulation system in the fish
holds.

II. ARRANGEMENT OF THE BOAT

Grouper has the following requirements in order
to keep them alive:
1. Salinity : 30 - 34 ppt (part per thousand) or
permil,
2. Water temperature : 24 -32 degree Celcius,
3. Water PH :7 9,
4. Speed of water flow : 20 50 cm/second
5. Content of oxygen solution minimum 3 ppm,
6. Nitrogen nitrite : 0 - 0.05 ppm,
7. Ammonia (NH
3
-N) < 0.02 ppm
8. Dark environment because grouper is night
active fish,
9. Density of fish in the water (kg/m
3
),
10. height of the fish hold, and
11. the wide of the hatch.

ISSN: 1411-1284

Proceeding 10
th

To cater the above requirements the boat is
designed as a trimaran boat with the following
specificationsthe side hull are the smaller parts of the
boat where the propulsion engines are located.

1. Middle hull:
The middle hull is the largest part of the boat
where the fish hold are located, the main particulars of
the middle hull are as follows:

Table 1. Middle hull main particulars

There are three fish holds in the middle hull,
which will be filled with water so that the fish can be
kept alive; one pump room located in the middle
where all the machinery for maintaining the fish hold
conditions is located; above the pump room is the
wheel house. Arrangement of the middle hull can be
seen in the figure 1.

Figure
1.
Middle hull

2. Side hull

The size of the side hulls are smaller than the
middle hull, their purpose is to stabilize the boat, in
both of the side hulls the propulsion engines are
installed. The main particulars of the side hulls are as
follows:

Table 2. Side
hulls main
particulars

Figure 2. Side hulls

The arrangement of the fish hold are as follows:

Table 3. Fish hold division:

The capacity of the boat is 380 kg of live grouper
consists of 636 of fish which have economical value.
The sea water needed for the fish is 15,86 m
3
or
16,255 ton.

III. THE WATER AND AIR CIRCULATION
SYSTEM

The water and air circulation system for the fish
holds is generated by the following equipment:
1. Circulation pump 2 HP.
2. General pump 2 HP.
3. Filter
4. Generator
5. Chiller
6. Compressor
Loa 15 m
Lwl 14,55 m
Lpp 14,2 m
D 1,31 m
H 1,67 m
B 2,84 m
Cb 0,47
Cw 0,647
Cm 0,968
Cp 0,486
Loa 10 m
Lwl 9,7 m
Lpp 9,46 m
D 0,87 m.
H 1,12 m.
B 1,9 m
Cb 0,32
Cw 0,55
Cm 0,97
Cp 0,33
Item Hold
1
Hold
2
Hold
3
total
Volume
(m
3
)
2,45 7,5 5,91 15,86
Berat
(ton)
2,51
1
7,687 6,057 16,25
5
ISSN: 1411-1284

Proceeding 10
th

7. circulation pipes
8. suction pipes for dirty water

IV. CONCLUSIONS

After being calculated for the proposed trimaran boat
the water and air circulation system can be concluded
as follows:

1. The total fish hold capacity is 16,225 ton,
consisted of 0,38 ton grouper live fish, and
15,875 ton of sea water.
2. There are two pumps being used one for
circulating the water and general pump for
pumping the dirty water, each is powered by
2 hp electric motor.
3. To maintain the water temperature a chiller is
being used.

REFERENCES

1. Geankoplis,C.J. Transport Processes and Unit
Operations (New Delhi: Prentice Hall of India, 1997).
2. J.O, Traung, Fishing Boat Of The World 2 ( England:
Fishing New Books Ltd,
1960), hal 87 89
3. Kern, D.Q., Process Heat Transfer (Tokyo: International
Student Edition, Mc Graw Hill Book Co, 1965).
4. Mallawa. A, Sudirman, Teknik Penangkapan Ikan
( Jakarta: PT. Rineka Cipta, 2004).

ISSN: 1411-1284

Proceeding 10
th

Local Protein Structure Comparison

1
Ford Lumban Gaol,
2
Belawati Widjaja

1,2
Faculty of Computer Science, University of Indonesia,
Kampus UI Depok, Jawa Barat, Indonesia
Tel. 021-7863419, fax. 021-7863415 email : flg50@ui.ac.id

Abstract Genome sequencing projects are working
to determine the complete genome sequence for
several organisms. The sequencing projects have
produced significant impact on bioinformatics
research by stimulating the development of
sequence analysis tools such as methods to identify
genes in a genome sequence, methods to predict
alternative splicing sites for genes, methods that
compute the sequence homology among genes, and
methods that study the evolutionary relation of
genes, to name a few.
Proteins are the products of genes and the building
blocks for biological function. In this paper, we
identify some basic background on local proteins
structure, protein function and their simulation.

Keywords Genome sequence, Pattern-Based
Structure Alignment, Functional Site Identification,
Structure-based Functional Annotation.

I. PROTEIN STRUCTURE

1.1 Amino Acids
Proteins are chains of -amino acid molecules. An
-amino acid (or simply an amino acid) is a molecule
with three chemical groups and a hydrogen atom
covalently bonded to the same carbon atom, the C
atom. These groups are: a carboxyl group (-COOH),
an amino group (-NH2), and a side chain with
variable size (symbolized as R) [BT91]. The first
carbon atom in a side chain (one that is connected to
the C atom) is the C atom and the second one is the
C atom and so forth. Figure 1 illustrates an example
of amino acids.
Different amino acids have different side chains.
There are a total of 20 amino acids found in naturally
occurring proteins. At physiological temperatures in a
solvent environment, proteins adopt stable three-
dimensional (3D) organizations of amino acid
residues that are critical to their function.

1.2 Four Levels of Protein Structure
The levels are as follows:

Figure 1: Up: a schematic
illustration of an amino acid.
Down: the 3D structure
of an amino acid
(Alanine) whose side
chain contains a single
carbon atom. The atom
types are shown;
unlabeled atoms are
hydrogens. The schematic
diagram is adopted from [BT91] and the 3D structure is
drawn with the VMD software.

Primary structure describes the amino acid
sequence of a protein.
Secondary structure describes the pattern of
hydrogen bonding between amino acids
along the primary sequence. There are three
common types of secondary structures: -
helix, -sheet, and turn.
Tertiary (3D) structure describes the
protein in terms of the coordinates of all of
its atoms.
Quaternary structure applies only to
proteins that have at least two amino acid
chains. Each chain in a multi-chain protein is
a subunit of the protein and the spatial
organization of the subunits of a protein is
the quaternary structure of the protein. A
single-subunit protein does not have a
quaternary structure.

Primary Structure
In a protein, two amino acids are connected by a
peptide bond, a covalent bond formed between the
carboxyl group of one amino acid and the amino
group of the other with elimination of a water
molecule. After the condensation, an amino acid
becomes a amino acid residue (or just a residue, for
short). The C
atom and the hydrogen atom, the

carbonyl group (CO), and the NH group that are
covalently linked to the C
atom are the main chain

atoms; the rest of the atoms in an amino acid are side
chain atoms.
In Figure 2, we show the primary sequence of a
protein with three amino acid residues.
At one end of the sequence (the left one), the residue
contains the full amino group (-NH3) and is the N
terminal of the sequence. The residue at the opposite
end contains the full carboxyl group (-COOH) and is
the C terminal of the sequence. By convention a
ISSN: 1411-1284

Proceeding 10
th

protein sequence is drawn left to right from its N
terminal to its C terminal.
Various protein sequencing techniques can determine
the primary sequence of a protein experimentally.

Figure 2: A schematic illustration of a polypeptide with
three residues: Met, Gly and Ala. The peptide can also be
described as the sequence of the three residues: Met-Gly-
Ala.

Secondary Structure
A segment of protein sequence may fold into a stable
structure called secondary structure.
Three types of secondary structure are common in
proteins.
-helix
-sheet
turn
An -helix is a stable structure where each
residue forms a hydrogen bond with another one that
is four residues apart in the primary sequence. We
show an example of the helix secondary structure
in Figure 3.
A -sheet is another type of stable structure
formed by at least two -strands that are connected
together by hydrogen bonds between the two strands.
A parallel -sheet is a sheet where the two -strands
have the same direction while an anti-parallel -sheet
is one that does not. We show examples of -sheets
in Figure 3.
A turn is a secondary structure that usually
consists of 4-5 amino acids to connect helices or
-sheets.
Unlike the protein primary sequence, protein
secondary structure is usually obtained after solving
the 3D structure of the protein.

Figure 3: A schematic illustration of the -helix and the -
sheet secondary structures. (a) the ribbon representation of
the -helix secondary structure (on the left) and the ball-
stick representation showing all atoms and their chemical
bonds in the structure (on the right). We also show the same
representations for the parallel -sheet secondary structure
(b) and the anti-parallel -sheet secondary structure (c). The
-helix is taken from protein myoglobin 1MBA at positions
131 to 141 as in [Fer99]. The parallel -sheet secondary
structure is taken from protein 2EBN at positions 126 to 130
and 167 to 172. The anti-parallel sheet secondary
structure is taken from protein 1HJ9 at positions 86 to 90
and 104 to 108.

Tertiary Structure and Quaternary Structure
In conditions found within a living organism, a
protein folds into its native structure. The tertiary
structure refers to the positions of all atoms, generally
in the native structure. The process of adopting a 3D
structure is the folding of the protein. Protein 3D
structure is critical for a protein to carry out its
function.
In Figure 4, we show a schematic
representation of a 3D protein structure (myoglobin).
In the same figure, we also show the quaternary
structure of a protein with two chains (HIV protease).

Figure 4: Right: The schematic representation (cartoon) of
the 3D structure of protein myoglobin (PDB id: 1MBA).
Left: The schematic representation (cartoon) of the 3D
structure of protein HIV protease (PDB id: 1MBA). HIV
protease has two chains.

Two types of experimental techniques are
used to determine the 3D structure of a protein.
In X-ray crystallography, a protein is first crystalized
and the structure of the protein is determined by X-ray
diffraction. Nuclear Magnetic Resonance
spectroscopy (NMR) determines the structure of a
protein by measuring the distances among protons and
specially labeled carbon and nitrogen atoms [PR04].
Once the inter-atom distances are determined, a group
of 3D structures (an ensemble) is computed in order to
best fit the distance constraints.

1.3 Structure Classification

Domains
A unit of the tertiary structure of a protein is a
domain, which is the whole amino acid chain or a
(consecutive) segment of the chain that can fold into
stable tertiary structure independent of the rest of the
protein [BT91]. A domain is often a unit of function
i.e. a domain usually carries out a specific function of
a protein. Multi-domain proteins are believed to be
the product of gene fusion i.e. a process where several
genes, each which once coded for a separate protein,
become a single gene during evolution [PR04].

Structures are Grouped Hierarchically
The protein structure space is the set of all possible
protein structures. Protein structure space is often
described by a hierarchical structure called protein
structure classification, at the bottom of which are
individual structures (domains). Structures are
grouped hierarchically based on their secondary
structure components and their closeness at the
sequence, functional, and evolutionary level [PR04].
Here we describe a structure hierarchy, the SCOP
database (Structure Classification of Proteins)
ISSN: 1411-1284

Proceeding 10
th

[MBHC95]. SCOP is maintained manually by domain
experts and considered one of the gold standards for
protein structure classification. For other classification
systems see [OMJ+97].
In SCOP, the unit of the classification is the domain
(e.g. multi-domain proteins are broken into individual
domains that are grouped separately). At the top level
(most abstract level), proteins in SCOP are assigned
to a class based on the secondary structure
components.
The four major classes in SCOP are:
domain class: ones that are composed
almost entirely of -helices
domain class: ones that are composed
almost entirely of -sheets
/ domain class: ones that are composed of
alpha helices and parallel beta sheets
+ domain class: ones that are composed
of alpha helices and antiparallel beta sheets

These four classes cover around 85% of folds in
SCOP. Another three infrequently occurring classes in
SCOP are: multi-domain class, membrane & cell
surface domain class, and small protein domain class.
Proteins within each SCOP class are classified
hierarchically at three additional levels: fold,
superfamily, and family. In Figure 5, we show a
visualization developed by the Berkeley Structural
Genomics Center
http://www.nigms.nih.gov/psi/image
gallery/structures.html in which globally similar
structures are grouped together and globally dissimilar
structures are located far away from each other. This
figure shows segregation between four elongated
regions corresponding to the four SCOP protein
classes: , , / , and + . Further details about
protein structure classification can be found in
[MBHC95].

Figure 5: The top level structural classification of
proteins based on their secondary structure
components.

II. PROTEIN FUNCTION

Proteins are the molecular machinery that perform
the function of living organisms. Protein function can
be described by the role(s) that the protein plays in an
organism. Usually, protein function description is
made at the molecular level, e.g. the role a protein
plays in a chemical reaction. Protein function can also
be described at a physiological level concerning the
whole organism, e.g. the impact of a protein on the
functioning of an organism. We describe protein
function at 3 different levels according to [OJT03]:
Molecular function: A protein's molecular
function is its catalytic activity, its binding
activity, its conformational changes, or its
activity as a building block in a cell. [PR04].
Cellular function A protein's cellular
function is the role that the protein performs
as part of a biological pathway in a cell.
Phenotypic function: A protein's phenotypic
function determines the physiological and
behavioral properties of an organism.
We need to keep in mind that protein function is
context-sensitive with respect to many factors other
than its sequence and structure. These factors include
(but are not limited to) the cellular environment in
which a protein is located, the post-translation
modification(s) of the protein, and the presence or
absence of certain ligand(s). Though often not
mentioned explicitly, these factors are important for
protein function.
In this paper, we concentrate on the molecular
function of a protein. We do so since (1) it is
generally believed that native structure may most
directly be related to the molecular function
[GSB+05], (2) determining the molecular function is
the first step in the determination of the cellular and
phenotypic function of a protein.

III. THE IMPORTANCE OF LOCAL STRUCTURE
COMPARISON

Traditionally global structure comparison is well
investigated in protein structure analysis. Recently the
research focus has shifted towards local structure
comparison. The transition from global structure
comparison to local structure comparison is well
supported by a wide range of experimental evidence.

3.1 Protein Function
It is well known that in a protein there are a few key
residues that if mutated, interfere with the structural
stability or the function of the protein. Those
important residues usually are in spatial contact in the
3D protein structure and hence form a cluster" in the
protein structure. On the other hand, much of the
remaining protein structure, especially surface area,
can tolerate mutations [CS99, SKBW98]. For
example, in a model protein T4 Lysozyme, it was
reported that single amino acid substitutions occurring
in a large fraction of a protein structure (80% of
studied amino acids) tend not to interrupt the function
and the folding of the protein [Mat96].
Biology has accumulated a long list of sites that have
functional or structural significance.
ISSN: 1411-1284

Proceeding 10
th

Such sites can be divided into the following three
categories:
catalytic sites of enzymes
the binding sites of ligands
the folding nuclei of proteins
Local structure similarity among proteins can
implicate structurally conserved amino acid residues
that may carry functional or structural significance
[KWK04].

Similar Structures, Different Function
It is well known that the TIM barrels are a large group
of proteins with a remarkably similar fold, yet widely
varying catalytic function [NOT02]. A striking result
was reported in [NKGP90] showing that even
combined with sequence conservation, global
structure conservation may still not be sufficient to
produce functional conservation. In this study,
Neidhart et al. first demonstrated an example where
two enzymes (mandelate racemase and muconate
lactonizing enzyme) catalyze difierent reactions, yet
the structure and sequence identities are sufficiently
high that they are very likely to have evolved from a
common ancestor. Similar cases have been reviewed
in [GB01].

Different Structures, Same Function
It has been also noticed that similar function does not
require similar structure. For example, the most
versatile enzymes, hydro-lyases and the O-glycosyl
glucosidases, are associated with folds [HG99].
In a systematic study using the structure database
SCOP and the functional database Enzyme
Commission (EC), George et al. estimated 69% of
protein function (at EC subsubclass level) is indeed
carried by proteins in multiple protein superfamilies
[GST+04].
In summary, we want to develop methodology to
recognize protein similarity at the structure pattern
level, and use such method as a complemental method
to ones with similar purpose that are based on
sequence and global structure similarity.

IV. CONCLUSIONS

Protein structure comparison is part of a
bioinformatics research paradigm that performs
comparative analysis of biological data. The
overarching goal is to aid rational experiment design
and thus to expedite biological discovery.
Specifically, through comparison, the paradigm
endeavors to transfer experimentally obtained
biological knowledge from known proteins to
unknown ones, or to discover common structure
among a group of related proteins. Below we review
some of the applications of local structure comparison
including pattern based structure alignment,
functional site identification, structure-based
functional annotation, and protein evolution.
Pattern-Based Structure Alignment
Structure alignment is vital to identifying conserved
residues in protein structure, to studying the evolution
of protein structures, and to facilitating structure
prediction. For example, through structure alignment,
domain experts discover that many diverse sequences
may adopt the same global structure and such
information helps significantly in structure prediction.
Multiple structure alignment has been used to identify
structural commonalities among a group of proteins.
Two approaches have been investigated. The first uses
multiple sequence alignment to solve the difficult
alignment problem and then focuses on identifying
the structural core of proteins. The second approach
applies pairwise structure alignment iteratively in
order to derive multiple structure alignment.
Functional Site Identification
A functional site is a group of amino acids in a protein
that participate in the function of the protein (e.g.
catalyzing chemical reactions or binding to other
proteins). Identifying functional sites is critical in
studying the mechanism of protein function,
predicting protein-protein interaction, and recognizing
evolutionary connections between proteins when there
is no clear clue from sequence or global structure
alignment.
Structure-based Functional Annotation
There is no question that knowing the function of a
protein is of paramount importance in biological
research. Correct function prediction can significantly
simplify and decrease the time needed for
experimental validation. However incorrect
assignments may mislead experimental design and
waste resources. Protein function prediction has been
investigated by recognizing the similarity of a protein
with unknown function to one that has a known
function where similarity can be determined at the
sequence level, the expression level, and at the level
of the gene's chromosome location. In structure based
function annotation, investigators focus on assigning
function to protein structures by recognizing structural
similarity. Compared to sequence-based function
assignment, structure-based methods may have better
annotation because of the additional information
offered by the structure.
In summary, the potential to decrease the
time and cost of experimental techniques, the rapidly
growing body of protein structure and structure
related data, and the large number of applications
necessitate the development of automated comparison
tools for protein structure analysis.

REFERENCES

1. [BT91] CARL BRANDEN and JOHN TOOZE.
Introduction to Protein Structure. Garland Publishing,
1991.
2. [Fer99] ALAN FERSHT. Structure and Mechanism in
Protein Science. W.H. Freeman Comp., 1999.
3. [PR04] GREGORY A PETSKO and DAGMAR
RINGE. Protein Structure and Function. New Science
ISSN: 1411-1284

Proceeding 10
th

Press Ltd, Middlesec House, 34-42 Cleveland Street,
London W1P 6LB, UK, 2004.
4. [MBHC95] A.G. MURZIN, S.E. BRENNER, T.
HUBBARD, and C. CHOTHIA. SCOP: a structural
classification of proteins database for the investigation
of sequences and structures. Journal of Molecular
Biology, 247:536-540, 1995.
5. [OMJ+97] C.A. ORENGO, A.D. MICHIE, S. JONES,
D.T. JONES, M.B. SWINDELLS, and J.M.
THORNTON. CATH - a hierarchic classication of
protein domain structures. Structure, 5(8):1093-
1108,1997.
6. [OJT03] C. ORGENGO, D. JONES, and J.
THORNTON. Bioinformatics: genes, proteins, and
computers. BIOS Scientific Publishers ltd., 2003.
7. [GSB+05] R.A. GEORGE, R.G. SPRIGGS, G.J.
BARTLETT, A. GUTTERIDGE, M.W.
MACARTHUR, C.T. PORTER, B. AL-LAZIKANI,
J.M. THORNTON, and M.B. SWINDELLS. Effective
function annotation through residue conservation.
PNAS, 102:12299-12304, 2005.
8. [CS99] MH CORDES and RT. SAUER. Tolerance of a
protein to multiple polar-to-hydrophobic surface
substitutions. Protein Sci., 8(2):318{25, 1999.
9. [SKBW98] JM SCHWEHM, ES KRISTYANNE, CC
BIGGERS, and STITES WE. Stability effects of
increasing the hydrophobicity of solvent-exposed side
chains in staphylococcal nuclease. Biochemistry,
37(19):6939{48, 1998.
10. [KWK04] E. V. KOONIN, Y. I. WOLF, and G. P.
KAREV, editors. Power Laws, Scale-free Networks
and Genome Biology. Springer, 2004.
11. [NOT02] N NAGANO, CA ORENGO, and JM
THORNTON. One fold with many functions: the
evolutionary relationships between tim barrel families
based on their sequences, structures and functions.
Journal of Molecular Biology, 321:741-765, 2002.
12. [NKGP90] DAVID J. NEIDHART, GEORGE L.
KENYON, JOHN A. GERLT, and GREGORY A.
Petsko. Mandelate racemase and muconate lactonizing
enzyme are mechanistically distinct and structurally
homologous. Nature, 347:692-694, 1990.
13. [GB01] JOHN A. GERLT and PATRICIA C.
BABBITT. Divergent evolution of enzymatic function:
Mechanistically diverse superfamilies and functionally
distinct suprafamilies. Annu. Rev. Biochem.,
70:20946, 2001.
14. [HG99] H. HEGYI and M. GERSTEIN. The
relationship between protein structure and function: a
comprehensive survey with application to the yeast
genome. J Mol Biol, 288:147-164, 1999.
15. [GST+04] RICHARD A. GEORGE, RUTH V.
SPRIGGS, JANET M. THORNTON, BISSAN AL-
LAZIKANI, and MARK B. SWINDELLS. Scopec: a
database of protein catalytic domains. Bioinformatics,
Supp 1:I130-I136,2004.

Antibacterial Activity Analysis of Dillenia indicas
Kernel Extract for Escherichia coli: A Preliminary Study

Rita Arbianti, Tania Surya Utami, and Ifa Puspasari

Chemical Engineering Department, Faculty of Engineering, University of Indonesia
Kampus UI Depok 16424, Phone. 7863516, Fax. 7863515
Email: arbianti@che.ui.edu; nana@che.ui.edu

Abstract The isolation of the crude extract from
Dillenia indicas kernel in n-hexane and chloroform
obtained seven fractions called fraction A G. The
antibacterial activity test with the method of paper
disc for these fractions towards the bacteria
Escherichia coli found that these fractions inhibited
the growth of the bacteria except fraction F which
was eluted at the accommodation of 50 until 51 by
100% chloroform. While fraction D (result of
accommodation of 28 until 37) and fraction G (result
of accommodation 53 until 58 by 100% chloroform)
proved more active inhibit the growth of Escherichia
coli compared to the other, even with the antibiotic
cloramphenicol.

Keywords Extraction, antibacterial activity, Dillenia
indica, Escherichia coli

I. INTRODUCTION

illenia indicas kernel produces specific scent that
can indicate the existence of atsiri oil and terpenoid
as one of its component, since the character of atsiri
oil is also having specific scent [1]. Triterpenoid is
known having ability for antibacterial activity. This is
also proved by some researches were reported that some
of triterpenoid shows antibacterial activity against
bacteria such as Bacillus subtilis, Escherichia coli and
Micrococcus luteus. While on the other literature it was
reported that Dillenia indicas kernel has been used as a
medicine to stomach illness. All these things have put a
strong indication that the ability antibacterial activity is
in Dillenia indicas kernel.

Other research said that part from Dillenia indica,
leaves, is also having ability to retain the growth of some
specific bacteria such as Bacillus cereus [2]. On the other
research it was mentioned that Dillenia papuana species
specifically on the leaves contain some triterpenoid,
oleanen type where this can be used as antibacterial. This
ability is because it has double bonding within carbocillic
and also ketene function from oleanen [3]. This research
is aimed to measure antibacterial activity from Dillenia
indicas kernel which is expected to understand the usage
of Dillenia indicas kernel specific reason.

The objective of this research is to examine
antibacterial activity of chemical compound from
Dillenia indicas kernel and to know the utilization of
this kernel specifically.

II. BASIC THEORY

The testing of antibacterial activity has the same
meaning with determine bacterial sensitivity of
antibacterial. Basically, the method of antibacterial
activity testing could be divided into Diffusion Method
and Dilution Method [4, 5]. In dilution method,
determining of antibacterial activity is based on
diffusivity ability on gelatin slab, which have been
inoculated with tested bacterial. The observation
conducted with existing of inhibited area around the
antibacterial.

While in dilution method, antibacterial mixing with
gelatin medium and the inoculated with tested bacterial.
The observation conducted with bacterial growth in
medium. Antibacterial activity determined by minimum
inhibited concentration of antibacterial in retains
bacterial growth.

In this research, antibacterial activity testing was
conducted by using gelatin diffusion method. Basic
principle of this testing are: The compound that expected
function as antibacterial put in a paper disc on the top of
Petri disc, where inoculated bacterial in gelatin medium
take place. After 24 hours of incubation, inhibited area
formed because of diffusivity of chemical compound
through gelatin, where concentration of molecules
diffusion higher enough to inhibited bacterial growth.
Tested bacterial that used in this research is based on
approach of stomach disease caused.

D


Dillenia indicas conventionally used as medicine of
stomach disease (diarrhea). One of tested bacterial that
known as pathogenic bacterial cause of infection of
intestines is Escherichia coli. The testing is conducted at
25 mg/mL concentration of hexanes crude extract and its
component from isolation. Antibiotic cloramphenicol
used as a positive control with same concentration, while
hexane and chloroform as a solvent as negative control.
The chosen of cloramphenicol is because this
antibacterial has large spectrum and can diffusion well in
gelatin medium [4].

Diffusion of tested solution that has ability as
antibacterial in paper disc, for several times could cause
decreasing of its concentration. When the concentration
is higher enough to inhibit bacterial growth, the growth
would inhibit. Hence could cause the forming of area that
extends over the paper disc with appropriate of
concentration decreasing. The more potential of
antibacterial, the more less of concentration needed.
Hence assumed that the higher diameter of inhibited area,
the more potential this solution as antibacterial.

Result of measurement of inhibited area diameter
can be show in Table 1, while the photograph of
antibacterial testing showed in Fig. 1.

Table 1. Measurement result of inhibited area diameter
Inhibited area
diameter (mm) Sample Eluted
a b c
Crude extract 4 12 7,5 - -
Fraction A 16 23 15 - -
Fraction B 25 26 11,5 - -
Fraction C 28 37 7,5 - -
Fraction D 38 49 27 - -
Fraction E 50 51 11,5 - 8,5
Fraction F 53 58 - - -
Fraction G 8 8 -
Cloramphenicol 26
Hexane -
Chloroform -

Figure 1. Photograph of antibacterial testing with
cakram method

Table 1 shows that all of fraction have inhibted
effect of E.coli growth significantly, except for F fraction
that eluted in 50-51 of 100% chloroform. It can be
caused of existing of chemical compound that do not
have ability to retain bacterial growth. While the
difference between the measurement of fractions depend
on the variation of components in the fraction.
D fraction has a higher inhibited area diameter (27
mm) and competitive enough with commercial antibiotic
cloramphenicol (26 mm) at the same concentration. Thos
fraction, that eluted at 28-37, has good ability as specific
alternative antibacterial of E.coli. While the testing of
used solvent i.e. hexane and chloroform, give the
negative result, that indicates the using of this solvent
have not influence the antibacterial activity.

There are several factors that effected antibacterial
activity, among them is incubation time. The longer
incubation time can influence inhibited area diameter,
because the more time needed to diffusion through
gelatin medium. Hence cause higher concentration in
certain spot at paper disc. Inhibited area also influenced
by ability to diffusion through gelatin medium. The other
factor is accuracy of inhibited area diameter
measurement. For result with better accuracy, firs we
must determine optical cell density of tested bacterial.
Because of this complexity, this research is a preliminary
study i.e. qualitative testing, so the factors mentioned
above are uncontrolled. But overall, the result of this
research shows that hexanes extract of Dillenia indicas
kernel and its isolation components have same result with
literatures.

IV. CONCLUSIONS

The isolation of the crude extract from Dillenia
indicas kernel in n-hexane and chloroform obtained
seven fractions called fraction A G. The antibacterial
activity test with the method of paper disc for these
fractions towards the bacteria Escherichia coli found that
these fractions inhibited the growth of the bacteria except
fraction F which was eluted at the accommodation of 50
until 51 by 100% chloroform.

Fraction D (result of accommodation of 28 until 37)
and fraction G (result of accommodation 53 until 58 by
100% chloroform) proved more active inhibit the growth
of Escherichia coli compared to the other, even with the
antibiotic cloramphenicol.

REFERENCES

[1] Harborne, J.B., Metode Fitokimia: Penuntun Cara
Modern Menganalisis Tumbuhan, diterjemahkan oleh
Kosasih Padmawinata dan Iwang Soediro, Bandung:
Penerbit ITB, 1987.
[2] Sornprasert, Ratapol. et al. The Potential of Crude
Extracts from Matat (Dillenia indica) against Bacteria
Growth, Department of Biology, Faculty of Science and
Technology, Rajabhat Institute, Chandrakasem, Bangkok
10900, Thailand.
[3] Nick, Andre. et al. Antibacterial Triterpenoids from
Dillenia papuana and Their Structure-Activity
Relationships, Phytochemistry, Vol. 40, No. 6, P. 1691-
1695, Elsevier Science Ltd., 1995.

[4] Jawetz, E., J.L Melnick, & E.A. Adelberg, Review of
Medical Microbiology, 11
th
ed., California: Lange
Medical Publication, 1974.
[5] Edwards, I. E. Antimicrobial Drug Action, P. 18-19,
Macmillan Press, 1980.
[6] Prosea, Plant Resources of South-East Asia, Vol. 5,
Bogor: Prosea Foundation, 1995.

Antioxidant Activity Analysis for Polar Extract
of Dillenia indicas Kernel

Tania Surya Utami, Rita Arbianti, and Lina Faty

Chemical Engineering Department, Faculty of Engineering, University of Indonesia
Kampus UI Depok 16424, Phone. 7863516, Fax. 7863515
Email: nana@che.ui.edu ; arbianti@che.ui.edu

Abstract the target of this research is to know
whether compound of polar bioactive of Dillenia
indica kernel own antioxidant activity. First matter
taken is taking polar compound from kernel of
Dillenia indica. This matter is conducted by
extraction use polar solvent, which is ethanol with
some water. Hereinafter, needed by dissociation of
the harsh extract become some fraction by using
column chromatography. Fixed phase of the column
is silica gel, whereas the motion phase is ethanol and
chloroform by the polarity is boosted up by
gradient. The antioxidant activity of the fraction
obtained tested with Carotene bleaching method.
Result of extraction in the form of dilution which
has brown color. After the solvent was evaporated,
we obtained a harsh extract in the form of resin jell
tan like caramel as much 128 gram. Result of
isolation by column chromatography is yield 7
fraction, A until G, with retention time is very
nearby. Result of test of antioxidant activity at
environment temperature shows that good enough
antioxidant activity compared to a negative control
is Fraction B. While test of antioxidant activity
conducted at incubation temperature of 40 and 60
o
C shows Fraction of G and B own good enough
antioxidant activity compared to a negative control.

Keyword Extraction, antioxidant activity, Dillenia
indica, carotene bleaching method

I. INTRODUCTION

ntioxidant is a material with ability to deactivate
free radical, so it can stop the oxidation process
through several mechanism i.e.: reaction with radical
scavenger, singlet oxygen quencher, as a filter for UV
radiation, or ion bonding. Antioxidant is a inhibitor for
lipid peroxide. Antioxidant not only used as protection
for the food from its expiration, but also to defend the
cell, i.e.: lipid membrane failure due to oxidation. This
is because the character of this antioxidant can avoid
chain reaction from the radical formation. This
antioxidant also can be used to as a protector from
some illness which is triggered from lipid oxidation
process in human body, i.e.: heart, stroke, and cancer.

One of the part of this plant that interest for learned is
the kernel. From sample of kernel, it can be seen
obviously that Dillenia indicas kernel is easily to
colour change when it is laid on the open air. This
colour changing is most likely occured due to
antioxidant compound that is available within its
kernel.

The target of this research is to know whether
compound of polar bioactive of Dillenia indicas kernel
own antioxidant activity. First matter taken is taking
polar compound from kernel of Dillenia indica. This
matter is conducted by extraction use polar solvent,
which is ethanol with some water. Hereinafter, needed
by dissociation of the harsh extract become some
fraction by using column chromatography. Fixed phase
of the column is silica gel, whereas the motion phase is
ethanol and chloroform by the polarity is boosted up by
gradient.

II. BASIC THEORY

Antioxidant is a material with ability to deactivate
free radical, so it can stop the oxidation process
through several mechanism i.e.: reaction with radical
scavenger, singlet oxygen quencher, as a filter for UV
radiation, or ion bonding. Antioxidant is a inhibitor for
lipid peroxide. Antioxidant which is added into food
must have criteria [1]:
Efective at low concentration
No difference in colour, taste, aroma, and
nutrient of food
Easy in used
Stable at process condition and storage of the
food
None of toxicity effect in higher dosage

Antioxidant is a inhibitor for lipid peroxide.
Antioxidant not only used as protection for the food
from its expiration, but also to defend the cell, i.e.: lipid
membrane failure due to oxidation. This is because the
character of this antioxidant can avoid chain reaction
A
from the radical formation [2]. This antioxidant also
can be used to as a protector from some illness which is
triggered from lipid oxidation process in human body,
i.e.: heart, stroke, and cancer [1]. Antioxidant
effectivity is depends on several factors [2]:
Iniciator radical
Type of antioxidant
Antioxidant concentration
Interaction with other antioxidant

Antioxidant activity testing for polar extract of
bioactive compound form Dillenia indicas kernel
conducted using Carotene bleaching method. The
testing conducted with 2 condition i.e.: room
temperature and incubation temperature (40, 60, and 80
o
C). 10 mL of soybean oil as sample added with
carotenoid and 0.2 mL polar extract of kernel. Sampel
absorbance measuring with spectrophotometer UV-
visibel. While soybean oil without extract compound as
negative control.


Antioxidant activity testing on room temperature
conducted using carotene bleaching method with
absorbance measuring, for 5 days. Fig. 1 shows
measurement result of absorbance. Fig. 1 shows that
there is decreasing of sample absorbance caused of
oxidation of soybean oil. From figure below, sample
with B and D fraction at 2nd day, have a small shall
compare to other. Which is meant, B and D fraction has
a compound with antioxidant activity.

Perbandingan Penurunan
Absorbansi pada Temperatur
Lingkungan
-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0 2 4 6
Hari
D
e
l
t
a

A
b
s
o
r
b
a
n
s
i
A
B
C
D
E
F
G
Kontrol

Fig. 1. Measurement result of absorbance at room
temperature

Perbandingan Absorbansi pada
Temperatur 40 C
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 2 4
Jam
A
b
s
o
r
b
a
n
s
i
A
B
C
D
E
F
G
Kontrol
Ekstrak Polar

Fig. 2. Measurement result of absorbance at incubation
temperature (40
o
C)

Temperatur 60 C
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0 1 2 3 4
Jam
A
b
s
o
r
b
a
n
s
i
A
B
C
D
E
F
G
Kontrol

temperature (60
o
C)

Fig.2 and Fig. 3 above shows that better
antioxidant activity given by G fraction at incubation
temperature 40
o
C, and B fraction at incubation
temperature 60
o
C. When antioxidant activity testing
conducted at incubation temperature 80
o
C, all of tested
sample does not have antioxidant activity (Fig. 4).

Temperatur 80 C
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0 1 2 3 4
Jam
A
b
s
o
r
b
a
n
s
i
A
B
C
D
E
F
G
Kontrol

temperature (80
o
C)

IV. CONCLUSIONS

Result of extraction in the form of dilution which has
brown color. After the solvent was evaporated, we
obtained a harsh extract in the form of resin jell tan like
caramel as much 128 gram.

From antioxidant activity of the fraction obtained tested
with Carotene bleaching method: result of isolation by
column chromatography is yield 7 fraction, A until G,
with retention time is very nearby. Result of test of
antioxidant activity at environment temperature shows
that good enough antioxidant activity compared to a
negative control is Fraction B. While test of antioxidant
activity conducted at incubation temperature of 40 and
60
o
C shows Fraction of G and B own good enough
antioxidant activity compared to a negative control.

REFERENCES

[1] Gritter, Roy J., 1987, Pengantar Kromatografi,
Bandung: Penerbit ITB.
[2] G., Scott, 1998, Antioxidants, Japan: Bull. Chem. Soc.
[3] Robinson, Trevor, 1995, Kandungan Organik
Tumbuhan Tinggi, Bandung: Penerbit ITB.
[4] Harina, Yuni, 2003, Tesis Magister Kimia, Aktivitas
radical scavenger a-Mangostin dan 1,3,6-trihidroksi-8-
2-(2-hidroksimetil-But-Enil)-9-Xanteron dari kulit buah
Manggis, Depok: UI.
[5] Prosea, 1995, Plant Resources of South-East Asia,
Vol. 5, Bogor: Prosea Foundation.
[6] W, Maruti, 2005, Skripsi Sarjana Teknik, Studi
Pendahuluan Pemanfaatan Ekstrak Daging Buah
Dillenia indica dengan Pelarut Non-polar sebagai
Antioksidan, Depok: Departemen TGP-FTUI.
[7] Boer, Yusnetti, 1999, Antioksidan Kulit Buah Kandis
[Garcinia parvifoloia (Miq.) Miq], Jakarta: Program
Studi Magister Ilmu Kimia FMIPA-UI.
[8] Asri, Isri Nur, 2001, Karya Utama Sarjana Kimia,
Isolasi Dan Studi Aktifitas Antioksidan dari Rimpang
Lempuyang Wangi, Depok: Jurusan Kimia FMIPA-UI.

ISSN: 1411-1284

Proceeding 10
th

Understanding the Metaspaces of Street Vendors
in the Cities: Temporality, Strategies and Tactics

Yandi Andri Yatmo* and Paramita Atmodiwirjo*

* Department of Architecture, University of Indonesia, Kampus UI Depok 16424
Tel. 7863512, fax. 7863514 email : yandiay@yahoo.co.uk

Abstract One of the commonly emerged issues in
many cities is the existence of unplanned and
unexpected urban elements. The phenomenon of
street vendors is a well-known example of such
elements. This paper attempts to understand the
phenomenon of street vendors by considering the
process of its emerging and re-emerging in urban
environment. The discussion of the process
requires an understanding from the point of view
of the street vendors as the subjects. Street
vendors continuously read the existing urban
spaces; their emergence and re-emergence in
urban spaces may be understood as a result of
such reading of the environment. This paper
would discuss the process inherent in the process
of street vendors reading of the environment,
based on the concept of chess as a metaspace
(Bunschoten, 2003). Understanding the
phenomenon of street vendors through the concept
of metaspaces allows us to look beyond the spatial
reality of street vendors in an actual locale in the
cities. It involves the examination of temporality,
as well as strategies and tactics involved within the
metaspaces of street vendors. The concept of
metaspaces becomes an instrument to uncover the
space of knowledge of street vendors within their
process of emergence and re-emergence in the
cities. This may eventually help the development of
planning instrument in dealing with such
phenomenon in the cities.

Keywords street vendors, temporality, metaspaces,
urban environment, strategies

I. TEMPORARY ELEMENTS IN THE CITIES

ost planning and design approaches emphasise
on the needs to create pleasant urban
environment for its inhabitants. Urban environments
should become a set of interrelated networks of
people, places and events, embracing various physical
and socio-cultural aspects. Nevertheless, the presence
of problems and conflicts is unavoidable in urban
environments. This would call the intervention of
planning profession to create order of the
environments.
Cities inevitably involve the existence of
unplanned and unexpected elements, the majority of
which become nuisance and disturbing elements.
Regardless of the efforts to remove these elements
from city spaces, they keep growing throughout the
cities. The phenomenon of street vendors is a well-
known example of such elements.

The phenomenon of temporary elements that keep
emerging and re-emerging in urban environments
clearly suggest the limitation of the planning
profession in determining what happens in the
everyday life of the cities. Planners and architects do
not have full capacity to control and prevent the
possibilities of spatial conflicts and disorder. Borden
et al clearly argued that The city is not the product of
planners and architects [1]. In most cases, the acts of
planners and architects merely represent a part of the
broader system of power, and thus do not necessarily
respond to the everyday phenomenon of urban reality.
Nevertheless, architects and planners cannot neglect
the existence of such unplanned and unexpected
phenomenon, which is related to the instability of
urban environment.

The limited ability of architects and planners in
fully determining the order of the environment has
also been clearly explained by Venturi in discussing
the existence of honky-tonk elements in the cities.
The main justification for honky-tonk elements in
architectural order is their very existence. They are
what we have. Architects can bemoan or try to ignore
them or even try to abolish them, but they will not go
away. Or they will not go away for a long time,
because architects do not have the power to replace
them (nor do they know what to replace them with
[2].

Whilst architects and planners possess limited
capability in determining the spaces of the city, there
must be other players that take significant role in
determining the spaces. Lefebvre suggested his theory
on space as a social (re)production, and explicitly
claimed the role of the society in the production of
space [3]. Hence the cities should be seen as a product
of the society living their everyday life in it.
Furthermore, Lefebvre emphasised the use of the term
M
ISSN: 1411-1284

Proceeding 10
th

subjects that imply their active role in constructing
space, instead of users or inhabitants [1].
A question may emerge on how the process of
space production occurs. Especially in the context of
urban environments with the phenomena of temporary
elements, it becomes necessary to examine the
processes that lead to the emergence of unplanned
elements, things and events in the existing spaces.
Furthermore, there is a need to examine the role of the
subjects in such process of emergence,
disappearance and re-emergence of temporary
elements.

This paper examines the existing problems of
street vendors in Indonesian cities, especially in
Jakarta. In examining the process of emerging and re-
emerging of street vendors as temporary urban
elements, it requires an understanding from the point
of view of the street vendors as the subjects in the
process. As subjects, the street vendors actively
perceive opportunities in certain urban spaces and act
according to that perception. The concept of chess as
a metaspace and Urban Gallery suggested by
Bunschoten [4] would become a framework in
discussing this issue. This concept is taken as an
instrument to uncover the space of knowledge of the
street vendors within their process of emergence in
the cities.

II. STREET VENDORS AND THE PROCESS OF
THEIR EMERGENCE AND RE-EMERGENCE IN
URBAN SPACES

The practice of street vending in Indonesia has
become a growing phenomenon, just like what
happened in other third world cities [5,6]. Street
vendors have always been positioned in a continuing
debate regarding their existence in the cities, between
arguments that support and reject them [5]. One of the
problems of street vendors is related to their existence
in urban spaces. In Indonesia, street vendors have
become the target of eviction in most cities.
Nevertheless, most street vendors keep emerging and
re-emerging.

The prominent characteristics of street vendors are
their mobility and flexibility in their everyday
operation [7]. Although some street vendors may
trade in permanent stalls, some others are non-
permanent vendors who sell their goods in pushcarts,
tents or other temporary structures, trays, boxes, or
even just by laying their goods on a mat. Such variety
in their modes of operation allows them to move from
place to place easily, and trade in various locations at
different times of the day.

Street vendors in Jakarta operate in various
locations. The data indicates that the majority (56.9%)
of street vendors occupy the sidewalks and the streets.
Others occupy green areas, urban parks, parking
areas, shopping and market areas and other public
spaces [8]. Some street vendors utilise the empty lots
that can be found in the city, including the spaces
under the highway flyover. Not all those locations are
legally designated areas for street trading activities.
Only a small number of street vendors (about 16.6%)
operate their trading activities in legal areas
designated to them, while the majority of street
vendors (83.4%) operate in illegal location [8]. The
street vendors activities in illegal locations have
always become the primary reasons to blame them as
a source of disorderliness in urban environment.

Local government in Jakarta have implemented
various policies to regulate the presence of street
vendors, ranging from the registration, legalisation,
designation of certain areas for vending activities and
relocation of the vendors from illegal location into
legal location. In many cases, eviction or street
cleaning operations were conducted against the
vendors who refused to be relocated and resisted to
stay in their current locations.

However, there seems to be a tendency for
inconsistent acts in dealing with street vendors. For
example, in some locations, after the officers
conducted street cleaning operation, they did not take
proper control of the location. This can give way to
the possibility for some street vendors to come back
and operate there. Such inconsistent attitudes have
resulted in repeated street cleaning operations in the
same places. The flexible and mobile characteristics
of street vendors also make it very easy for the street
vendors to dissemble their goods and leave the
location to avoid the street cleaning operations. It is
also very easy for them to come back, reassemble
their stuffs and start trading again only a few hours
after the operation takes place.

The phenomenon of emergence and re-emergence
of street vendors in urban environment have become
inseparable with everyday life of the cities. With such
mobility and flexibility, street vendors becomes a kind
of ephemeral architecture in the city as their
presence can be detached from urban space [9].
Behind the phenomenon of emergence and re-
emergence of street vendors, there may be some
processes in which the street vendors perceive
opportunities in certain urban spaces and act
according to that perception. We will now turn into
the discussion on how the street vendors read the
existing urban spaces and how their emergence and
re-emergence in urban spaces may be understood as a
result of their reading. We would need to first
examine the concept of metaspaces to understand
the process of such reading.

ISSN: 1411-1284

Proceeding 10
th

III. THE METASPACES OF STREET VENDORS

Metaspace is a concept that may explain the
processes that do not directly correspond to the real
experience of space and at the same time relate to it.
Metaspace first was fairly vague term that allowed
us to describe the space and models of games, a space
that was isolated from the daily reality and yet
belonged to it [4].

Bunschoten explained how the concept of
metaspace might be useful to examine the
phenomenon of temporality in urban environment.
Spaces in which we can see temporality are
metaspaces. The metaspace has an innate
ephemeral character. It is the site of the search of
knowledge; it is the container of knowledge-
management processes, but it is also a vehicle for the
search of meaning in the dynamic chaos in which we
live [4]. Hence the metaspace offers a path towards
understanding beyond the actual phenomena of
spaces.

A clear description on how the concept of
metaspace relate to the urban environment is through
the analogy of chess as a metaspace. As Bunschoten
described, A chess game takes place in a metaspace:
the game board, with its mix of simulated battles and
real emotions [4]. As a game, chess involve the
strategies and tactics that are directed to the aim of
surviving and winning the game. It involves the
reading of the situation and it would be interesting
to examine how such reading, strategies and tactics
might be taken as a parallel of the practice of street
vendors to survive in urban environment.

In chess, strategies consist of setting and
achieving long-term goals during the game, while
tactics concentrate on immediate manoeuvre [10]. For
example, strategies involve the evaluation of current
positions, where to place different pieces and the
planning of future play. Meanwhile, tactics become
the means to achieve strategy. Thus these two parts of
thinking strategies and tactics are inseparable.

In setting strategies for a chess game, there are
various aspects that a player should consider
thoroughly. These may include the pieces on board,
pawn structure, the safety of the king, the space, and
the control of squares [10]. The setting up of
strategies and tactics in a chess game nevertheless
involve the examination of current situation as well as
reading the opportunities and possible threats. It
becomes necessary to consider the actors (the pieces)
involved, possible relationship between the actors, as
well as the occupation and control of chess squares as
the setting of the game.

The examination of strategies and tactics as in
chess game becomes relevant as a way to study the
phenomenon of temporality in urban environment.
The practice of street vendors in finding and
occupying the spaces of the city might be considered
as urban tactics. Borden et al. explained, Tactics are
a more proactive response to the city; they are
practices discourses that produce objects. ... They
may be attempts to solve urban problems they may
also be attempts to reconceptualize relations between
the city and the self [1]. What is important in tactics
is the intentionality, which means that tactics are
aimed towards a particular goal. In the case of street
vendors, there is a goal to survive in everyday urban
environment.

Many decisions made by street vendors in their
relationship with the spaces of the cities could be
taken as a parallel of strategies and tactics in chess
game. In their practice, street vendors make decision
to trade in certain places of the cities, at certain times
and in certain ways. Considering the point of view of
street vendors as the subjects in such everyday
practices, the decisions are based on the reading of
the situation, including the actors involved and the
possibility to occupy certain settings. It also involves
the examination of possible practices in time contexts
how to deal with each situation that might be
different from time to time.

The examination of the setting is a crucial aspect
in a chess game. The players need to examine the
status of a particular square before making any
decisions to occupy it whether the square offers
opportunities to settle safely, whether the square is
under threat or under control of the enemys pieces,
and whether the square offers possibilities for further
beneficial move. The practice of street vendors also
involved similar evaluation. Street vendors need to
examine the status of an urban spot, to what extent it
is possible to trade in that location. Street vendors
may consider the possibilities for attracting more
customers, competition with other vendors, conflicts
with formal retail, policing strategies of the
government and the existence of control by certain
groups of society.

Often street vendors make decision to occupy
certain spaces which are ambiguous in their nature.
Such settings do not have clear ownership and identity
and thus become easier to occupy with less threat,
compared to settings with clear ownership. Borden
described such ambiguous setting as zero degree
architecture [11], which often attract skateboarders to
occupy. These include public spaces and streets, as
well as left-over spaces. Similar with street vendors,
In parts this reflects their desire to avoid social
conflict, but it is also an attempt to write anew not
to change meaning but to insert a meaning where
previously there were none [11]. Thus the acts of
street vendors do not necessarily based on their
perception of whether certain spaces are legal or
ISSN: 1411-1284

Proceeding 10
th

illegal trading locations, but more on the perception of
the ownership or identity of the setting.

In some cases of street vendors in Jakarta, there
are settings which are controlled not by government
officials but by certain informal groups of people
called preman or thugs. Such practice is a
representation of the control by informal structures of
society but nevertheless becomes a crucial aspect to
consider in understanding the practice of street
vendors in the cities.
The practice of street vendors is also determined
by the street vendors acts in reading of the
situation, including the existing threat at certain times.
Threats for street vendors mostly come from the
government officers in their acts to relocate or to evict
the street vendors. Street vendors have been used to a
hide and seek strategy to avoid eviction. Usually
they simply run away or stop their trading activities
when the eviction operation is taking place, but
afterwards they will come back to the location and
trade as usual. This is a kind of tactics that they
usually practice in case of eviction. Such tactics
become possible due to the flexible and mobile
characteristics of street vendors.

IV. TOWARDS AN URBAN PLANNING
INSTRUMENT

It becomes clear that some aspects involved in the
process of emergence and re-emergence of street
vendors in the cities might be understood by
examining the street vendors reading of the
situation. The strategies and tactics practiced by street
vendors represent the metaspaces of street vendors.
They become a vehicle to uncover the knowledge of
street vendors.

Following the understanding of the metaspaces of
street vendors, then there is a further challenge in
developing plans to deal with such phenomenon. As
Bunschoten suggested, the further challenge is to use
this concept of a metaspace for the development of a
planning instrument that is simultaneously a new
public space for the interaction and intertwining of
urban actors and their desires and interests [4].
Bunschoten further suggested that the metaspaces
would eventually lead into the concept of Urban
Gallery. When a part of a city is designated a
metaspace, it becomes an Urban Gallery a fluid
form of public space that evolves in time, generating
different definitions of public space and ways of
participating in it [4].

In the case of street vendors, it becomes clear that
the street vendors continuously manoeuvre within
Urban Gallery. There are characteristics of instable,
dynamic and ephemeral involved in the phenomenon
of street vendors in the cities. Here Urban Gallery
might become a device for the management of
transient states [4], thus as a management of such
dynamic and ephemeral nature of the phenomenon.

Street vendors continuously play within
metaspaces through their continuous reading of the
situation and through their strategies and tactics across
spaces and times. To respond to the street vendors
dynamics in metaspaces, there is a need for the
development of planning instrument that is levelling
into the same metaspaces. If street vendors play
within metaspaces, then any efforts to deal with them
should also play in the same ground of metaspaces.
This suggests that the efforts to deal with street
vendors in real spaces would not be successful if the
government do not consider the metaspaces of street
vendors.

An implementation of the planning instrument that
takes into account the dynamics of metaspaces is
through the structure of surveillance that can manage
the players within the metaspaces. It becomes
necessary to create a condition that someone is
watching, in order to redefine permissible spaces
within the society. For power over people,
architecture had wielded the evil technology of the
eye: spectacle and surveillance [12]. In the case of
dealing with street vendors, the government as the
powerful actor needs a way to manage the movement
within metaspaces, in order to control the movement
of street vendors in space and time.

Understanding the phenomenon of street vendors
through the concept of metaspaces allows us to look
beyond the spatial reality of street vendors in an actual
locale in the cities. It involves the examination of
temporality, as well as strategies and tactics involved
within the metaspaces of street vendors. The concept
of metaspaces becomes an instrument to uncover the
space of knowledge of the street vendors within their
process of emergence and re-emergence in the cities.
Eventually, the concept would lead into the needs for
levelling into the same ground of metaspaces in order
to deal with the phenomenon. Street vendors,
government and the society in general become the
actors that play within the metaspaces of the
phenomenon within the Urban Gallery. The urban
gallery is in fact a system that orchestrates the
conversation between actors in the urban domain [4].
Thus a planning instrument should thoroughly
consider the relationship among all actors involved in
the process of emergence and re-emergence of street
vendors.

REFERENCES

[1] I. Borden, J. Kerr, J. Rendell, and A. Pivaro (Eds.), The
Unknown City: Contesting Architecture and Social
Space, MIT Press, Cambridge, London, 2001.
ISSN: 1411-1284

Proceeding 10
th

[2] R. Venturi, Complexity and Contradiction in
Architecture, The Museum of Modern Art, New York,
1977.
[3] H. Lefebvre, The Production of Space, Oxford,
Blackwell, 1991.
[4] R. Bunschoten, Stirring still: The city soul and its
metaspaces, Perspecta, 34, pp.56-63, 2003.
[5] R. Bromley, Street vending and public policy: A global
review, International Journal of Sociology and Social
Policy, vol.20 no.1/2, pp.1-29, 2000.
[6] J. C. Cross, Street vendors, modernity and
postmodernity: Conflict and compromise in the global
economy, International Journal of Sociology and
Social Policy, vol.20 no.1/2, pp.30-52, 2000.
[7] I. Tinker, Street Foods: Urban Food and Employment
in Developing Country, Oxford University Press, New
York, 1997.
[8] Biro Pusat Statistik DKI Jakarta, 141.073 usaha
kakilima di DKI Jakarta mampu menyerap 193.314
tenaga kerja (Hasil Sensus Usaha Kakilima 2001),
Retrieved 8 November 2002, from the World Wide
Web: http://bps.dki.go.id/P2_News/ P28_Kakilima.htm
[9] I. Maharika, On the ephemeral: Tactility and urban
politics, Paper presented at the Indonesian Student
Scientific Meeting, Manchester, 2001.
[10] Wikipedia, Chess, Retrieved 1 November 2007, from
the World Wide Web:
http://en.wikipedia.org/wiki/Chess
[11] I. Borden, Another pavement, another beach:
Skateboarding and the performative critique of
architecture, Dalam Borden, et al (Eds.), The
Unknown City: Contesting Architecture and Social
Space, pp.178-198, MIT Press, Cambridge, London,
2001.
[12] P. Tabor, I am a videocam, Dalam Borden, et al
(Eds.), The Unknown City: Contesting Architecture
and Social Space, pp.122-137, MIT Press, Cambridge,
London, 2001.

ISSN: 1411-1284
Proceeding 10
th

Bringing Sustainability into the Curriculum of
Architectural Design Studio

Paramita Atmodiwirjo* and Yandi Andri Yatmo*

Tel. 7863512, fax. 7863514 email : mitayandi@gmail.com

Abstract The integration of sustainability aspects
into architectural education should be based on a
strong framework. It requires a clear understanding
of sustainability and related knowledge, abilities and
skills that should be acquired through the education
process. There is also a need for thorough
development of delivery methods within the learning
process. The methods should ensure the continuing
practices of sustainable development as a part of
lifelong learning. Without such strong framework,
sustainability would only become a jargon without
clear direction on how they may be translated into
learning contents and learning activities. This paper
addresses the integration of the knowledge of
sustainability into architecture curriculum. The
discussion builds on the analysis of already
established integrated studio curriculum that has
been implemented at the University of Indonesia
within the last few years. It includes the
identification of current contents of the design
studios and the possible opportunities for integrating
sustainability elements into the curriculum. This
paper concludes with the possible mapping of
sustainability contents and activities within the
overall process of design studios throughout the
curriculum.

Keywords sustainability, learning, architectural
education, curriculum, design studio

I. CURRENT MAPPING OF EDUCATION FOR
SUSTAINABILITY

he knowledge of sustainability in many
environment-related professions has recently taken a
new dimension. It no longer becomes merely as
supplementary field of knowledge, indeed it should be
the basis on which everything else should be developed.
It has become a key indicator to asses the
appropriateness of a practice in terms of sustainable
development. This re-positioning of sustainability issues
also unavoidable affects built environment professional
practice and education.
Such re-positioning could be identified by
examining various documents outlining the criteria that
must be fulfilled by every architecture courses
worldwide. UIA/UNESCO Charter for Architectural
Education highlights this issue in the following
statements: We, being responsible for the improvement
of the education of future architects to enable them to
work for a sustainable development in every cultural
heritage That the vision of the future world, cultivated
in architectural schools, should include the following
goals: a decent quality of life for all the inhabitants of
human settlements; a technological application which
respects the social, cultural and aesthetic needs of
people; an ecologically balanced and sustainable
development of the built environment [1].

The needs to integrate sustainability into
architectural education could also be traced in the
validation process for architecture courses in United
Kingdom. In May 2002, Royal Institute of British
Architects (RIBA) together with Architects
Registration Board (ARB) published the new Criteria
for Validation document, which set that in addition to
the existing syllabus, consideration of the development
of the sustainable environment has been identified as
an area to be addressed more seriously in validating
schools of architecture [2].

Similarly in the United States, National Architecture
Accreditation Board (NAAB) Conditions for
Accreditation published in 2006 require architecture
graduates to demonstrate understanding in the area of
Sustainable Design. This area includes Understanding
of the principles of sustainability in making architecture
and urban design decisions that conserve natural and
built resources , and in the creation of healthful
buildings and communities [3].

The statements in those documents clearly illustrate
the needs for all architectural education institutions to
integrate the issues of sustainability within learning
processes. Nevertheless, such integration requires a
clear understanding of sustainability and related
knowledge, abilities and skills that should be acquired
T
ISSN: 1411-1284
Proceeding 10
th

through the process of education of architecture. There
is also a need for thorough development of delivery
methods within the learning process. The methods
should also ensure the continuing practices of
sustainable development as a part of lifelong learning.
Otherwise, sustainability would only become a jargon
without clear direction on how they may be translated
into learning contents and learning activities. Before
translating the issue of sustainability into the everyday
learning processes, there is a need to clarify the possible
scopes, contents and methods in learning about
sustainability in architecture.

II. LEARNING SUSTAINABILITY

Sustainability is a term that embraces various
aspects. The widely accepted definition of sustainable
development is Development that meets the needs of
the present without compromising the ability of future
generations to meet their own needs [4]. For the
purpose of discussing sustainability in architectural
education, it is important that the term is broadly
defined as involving multi dimensions of
environmental, social and economic [5].

Within the discussion of sustainable environment,
there seems to be a tendency to put greater weight of
sustainability in terms of technology. Meanwhile,
sustainability is seldom discussed as a social
phenomenon [6]. This has led into education practices
that tend to associate the learning of sustainability with
certain areas or subjects building technology, building
physics, building services and environmental studies.
Technology is definitely a significant part involved in
the creation of sustainable environment, yet there are
other dimensions of sustainability that need to be
addressed with similar emphasises.

Sustainable, green or ecological architecture have
been defined in various ways, but at least there are some
key issues involved: integrated relationship with nature;
preservation and/or improvement of local ecosystems;
civic action in which environmental physical and social
quality is essential; benchmarks defined by experts; and
the saving and conservation of energy [6]. The wide
range of issues that involved within the term indicates
the need for holistic nature of sustainability thinking
[7]. As a result, architectural education should also
adopt such holistic approach in developing
sustainability contents in its curriculum.

Considering the need for holistic approach, then the
schools of architecture need to promote sustainability
to become part of the mainstream of architectural
education [7]. Such holistic approach is required, rather
than concentrating the teaching and learning of
sustainability in certain subjects of technological or
environmental studies. Such integration into the
mainstream is considered as the best way to ensure the
real values of education for sustainability.

The wide range of dimensions involved in
sustainability has inevitably created further complexities
in the application into everyday learning process. It
becomes clear that there can be no single way to teach
sustainability [7], especially since each dimensions of
sustainability may require certain learning activities and
it would be impossible to equip students with the
knowledge in all dimensions. A possible approach is by
integrating the knowledge of sustainability into design
projects [5]. In this way, students are equipped with
general knowledge of sustainability and at the same
time given opportunities to apply some of their
knowledge into a design project.

The integration of sustainability into architectural
education does not necessarily means a dramatic change
of curriculum contents and learning methods.
Architectural education have been characterised by
studio-based learning or project-based learning, which
clearly distinguished it from other fields [8]. Studio-
based learning is integrative in its nature, with the
requirements to deal with certain issue by taking into
account various issues of human behaviour, social,
historical and physical contexts, aesthetic consideration
and technology. Such integrative nature of design studio
clearly offers possibilities for the integration of
sustainability into learning process.

In fact, an entirely new pedagogy need not be
developed for the learning and teaching of
sustainability [7]. The existing structure of design
studio in most schools of architecture has been set up to
develop the students abilities to deal with problems
from multiple perspectives. What is particularly needed
now is a careful consideration on the extent to which
sustainability dimensions have or have not been
introduced into current curriculum and learning
methods. We will now turn into the discussion on how
the integration of sustainability could be made possible
in the current curriculum structure of architecture at the
University of Indonesia.

III. INTEGRATION INTO DESIGN CURRICULUM:
CHALLENGES AND OPPORTUNITIES

Since 2004, the Department of Architecture at the
University of Indonesia has adopted a curriculum with
integrated design studios as the core of the curriculum.
The integrated design studios becomes the arena for
ISSN: 1411-1284
Proceeding 10
th

students to acquire knowledge in various aspects related
to architectural design, i.e., the knowledge of space and
place, structure and construction, urban studies,
environmental studies, building physics and building
services. Compared to the previous curriculum in which
those areas of knowledge were delivered within
individual subjects, such integrated design studios
facilitate students to develop understanding on how each
aspect might be related to a single design project.

Throughout the curriculum, students go through five
stages of design studios. Each studio has different
emphasis and different levels of complexities, yet they
form a continuous process of learning. Thus the
knowledge acquired within one stage would form the
basis for the subsequent stages.

In everyday learning activities, the emphasis of each
design studio is translated into certain design projects,
most of which are developed by the students based on
certain trigger issues introduced by the studio staffs.
Therefore, the design activities in each studio may vary
in different semesters, and may even vary across
individual students. Thus each student may develop
different contents of knowledge from their peers. This
becomes possible as long as certain performance criteria
are met. The emphasis on each design studio in the
current curriculum is illustrated in Table 1.

Table 1. Emphasis on each design studio
Stages Emphasis
Design
Studio 1
The ecology of self; developing spatial
ideas for self; awareness of structural logic,
environment and human comfort in the
actualisation of spatial ideas.

Design
Studio 2
Developing spatial ideas for a primary
social group; structural logic, environment
and human comfort in the actualisation of
spatial ideas

Design
Studio 3
Advanced exploration of spatial ideas in
urban contexts; technological and
environmental studies to support the
actualisation of spatial ideas

Design
Studio 4
Advanced exploration of spatial ideas
through the development of building
technology, embracing aspects of
environment, materials, construction and
services.

Design
Studio 5
All aspects of spatial design with code
compliances.

The description above suggests that the current
contents of design studio are based on a holistic
approach, with the intention to integrate various bodies
of knowledge related to spatial design. The existing
structure of the design studios clearly allows further
development to integrate more sustainable dimensions
into the learning processes.

The design studios are currently structured as a
continuous learning process, with each stage put more
emphasis on certain aspects. Thorough analysis of the
emphasis and contents in each studio would provide the
illustration on the extent to which sustainable
dimensions have or have not been included.
Furthermore, the description of the emphasis in each
studio may guide the decision to introduce more aspects
of sustainability.

As an example, Design Studio 3 and Design Studio
4 represent two stages of learning with very different
emphasis. Design Studio 3 is focused on the thorough
understanding of the contexts where the project would
be developed. Hence the studio opens possibilities for
further integration of social sustainability including
the awareness of locality, the understanding of social
and cultural contexts, and anticipating the effects of
project intervention to the community structure and
human well-being. On the other hand, Design Studio 4
put more emphasis on the development of ideas based
on the exploration of technology. This studio clearly
allows more technical aspects of sustainability to be
explored, including the understanding of sustainable
materials, appropriate construction strategies, and the
environmental studies of the projects.

Further challenge is on how the learning of
sustainability become fully integrated in most design
studios. In other words, each design studio should not
focus on the social dimension only or technological
dimension only. There is also a need to define how far
the knowledge of sustainability should be delivered in
each stage from merely awareness into the abilities
to fully apply the knowledge in design projects. Some
works are still currently in progress to develop the
integration of further dimensions of sustainability in
some design studios.

The integration of sustainability dimension into the
curriculum of design studio should also be translated
into clear learning activities to acquire certain contents
of knowledge. Various methods that have been
previously practiced allow possibilities for learning
various sustainability dimensions. The methods are
primarily based on experiential learning, in which
students develop knowledge by experiencing and
analysing their experience. This might be done through
practical works on environmental aspects related to
visual, thermal and auditory experience; studying the
ISSN: 1411-1284
Proceeding 10
th

contexts of the projects direct site visit, observation and
contacts with community; historical studies;
experiments in structural logic and construction
strategies; exploration of materials; computer modelling
and simulation; and others.

The variety of learning activities should also
consider the availability and management of learning
resources. It is important to understand the provision of
equipments for practical works and reading materials
that embrace a wide range of sustainability dimensions.
Similarly crucial is the development of staffs
capabilities to facilitate the learning of sustainability
dimensions through exercises in the design studio.

The discussion above suggests that the integration of
sustainability into the curriculum of architecture
would be possible by bringing it into the mainstream of
the curriculum. This means getting it straight into the
learning process of the design studios as the core of
curriculum. The existing structure of integrated design
studio opens many possibilities to expand the contents
with more sustainability elements. Different dimensions
of sustainability may become more apparent in one
stage than in another. Nevertheless, as the different
stages of design studios form a continuous learning
process, it is possible to equip the students with a
comprehensive understanding of sustainability by the
time they complete their study of architecture. The
learning methods that develop students capabilities to
construct their own knowledge would also ensure the
continuous process of life-long learning on
sustainability far beyond higher education study.

REFERENCES

[1] Union Internationale des Architectes, UIA/UNESCO
Charter for Architectural Education, UIA, 2005,
Retrieved 1 November 2007, from the World Wide Web:
http://www.uia-architectes.org/
image/PDF/CHARTES/CHART_ANG.pdf
[2] Royal Institute of British Architects, Criteria for
Validation, RIBA, London, 2002.
[3] National Architecture Accreditation Board, NAAB
Procedures for Accreditation for Professional Degree
Programs in Architecture, NAAB, Washington DC, 2004.
[4] Virginia Environmental Endowment, Education for
Sustainable Development: An Integrated Curriculum,
Blacksburg, Virginia, 2002.
[5] B. Edwards, The culture of sustainability in a school of
architecture, Keynote Symposium Paper, Working
towards a sustainable future: A Symposium on
Sustainability in Architectural Education, CEBE and
RIBA, 2002.
[6] V. Canizaro and K. Tanzer, Introduction, Journal of
Architectural Education, vol.60 no.4, pp.4-14, 2007.
[7] Centre for Education in the Built Environment, Report of
the Sustainability Special Interest Group (Architectural
Education), CEBE, 2003.
[8] D. Hawkes, The architect and the academy,
Architectural Research Quarterly, vol.4 no.1, pp.35-39,
2000.

ISSN: 1411-1284

Proceeding 10
th

LED Technology in Architectural Lighting of
Buildings Faade

Siti Handjarinto and Gregorius A.G.A

Fac. of Engineering, University of Indonesia, New Campus UI Depok 16424
Tel. 7863503, fax. 7270050 email : siti@eng.ui.ec.id, lone_lanrete@yahoo.com

Abstract Artificial lighting, especially exterior
lighting, shows the beauty of building at night,
even make it more beautiful than daytime. With
artificial lighting, an architect can also create
living and dramatic atmosphere on a building or
cityscape at night.
On the other side, in this modern era, the using
of lamps that requires a huge power to illuminate a
building is considered as an energy wasting
because most buildings are used in daytime only.
Light Emitting Diode (LED) is a new lighting
technology that most efficient in energy saving and
operational cost.
The other advantage of LED lamps in
architectural lighting is dynamic lighting. With
dynamic lighting, architectural lighting is no
longer a static lighting, just for illuminating of
buildings facade, but it can makes a building into
a multimedia device.
In this paper, we discuss 3 case studies, Galleria
Department Store, Uniqa Tower, and Dexia
Tower; use different dynamic lighting techniques
to create this multimedia device, i.e. dots, lines,
and surfaces of light. Each technique has its own
advantages and disadvantages, especially in image
resolution and number of used lamps.
In conclusion we propose the solution of this
topic covers the technical matter based on
framework idea of efficient energy and
maintenance artificial lighting of buildings faade
and dynamic lighting trough literacy study.

I. INTRODUCTION

ight makes us feel the visual world. When the
light changes, our experience also changes, this is
the main concept of architectural lighting.
1
Artificial
lighting of buildings facade aims to show its
architectural beauty and to make it can be enjoyed by
citizen at night.
The main issue of facade lighting is how to create
beautiful and low energy cost faade lighting. Beside
beautiful, lighting is also expected to be spectacular,
dynamic, or can be a multimedia device.

1
M. David Egan, Architectural Lighting. (New York:
McGraw-Hill, 2002), 2.
Now days, lamps have been more low energy cost,
compact, long lifetime, and creating a wonderful
architectural lighting technique. The latest artificial
lighting technology is Light Emitting Diode (LED).

II. BASIC THEORY

Solid State Lighting/ Light emitting diode (LED)
ight emitting diode (LED) is a semiconductor
(diode) that emits incoherent monochromatic light
when its given direct current. LED converts
electricity into light directly
2
(see fig 1)

Fig.1. RGB LED (source:www.wikipedia.com)
Inside a diode, interaction between electrons and
holes results a large number of photon, a basic unit of
light. Thus, these photons are focused with plastic
bulb and make a LED glows. (see fig 2)

Fig.2. Structure of a diode (source:www.sslc.ucsb.edu)
Performance characteristic
Luminous Efficacy
The latest InGan white LED now has luminous
efficacy up to 180 lm/w. This means a 5W LED bulb
can replace a 60W incandescent lamp and results the
same amount of lumen (light quantity). (see fig 3)

2
Vincent Laganier, LED Wall of
Light,International Lighting Review 2004: 72.
L
L
ISSN: 1411-1284

Proceeding 10
th

Fig.3. Power comparison between incandescent, fluorescent,
and LED lamps (source:www.sslc.ucsb.edu)

Lifetime and maintenance
LED lamp has a long life time, i.e.: 50,000-
100,000 hours, 5 times longer than fluorescent and 25
times than incandescent. Just like a fluorescent,
LEDs lumen decreases along with its lifetime. LED
lamp is also more durable than incandescent or
discharge lamp because of its solid plastic structure.
(see fig 4)

Fig.4. LED lumen depreciation graph
(source:www.sslc.ucsb.edu)

Color Rendering Index
Spectral distribution graph of LED lamps is
ranged in visible light wave, i.e.: 380-780 nm. (see fig
5) Thus, LED lamps have a good Color Rendering
Index because can display all colors in their actual
color.
LED lamp is also more efficient and safely
because generates less infrared and UV than
incandescent or discharge lamp. (see lamp comparison
in table 1)

Fig.5. Spectral distribution of LED (source:Panasonic
lighting)

Table 1. Lamps Comparison
GLS &
Halogen
Fluorescent HID LED
Dimension
(inch)
3/16-18 6-96 5-44 0.2-1
Wattage 1-
10,000
5-215 18-1800 0.03-5
Lumen/W 9-35 20-100 25-180 18-180
Lifetime
(hr)
750-
4,000
8-20,000 1,500-
24,000
50,000-
100,000
Start time instant rapidly 0-20 min Instant
Dimming able electronic
ballast
additional
circuit
able
CRI 100 95 15-85 67-80

Dynamic Lighting
ynamic lighting is an architectural lighting
concept and design that shows many light
compositions in a time by color changing, switching,
and dimming.

Table 2. Dynamic Lighting
Category Color changing Application
Dots

Orientation
Decorative
Dynamic
Lines

Sign
Decorative
Expressing 2D
elements
S
e
l
f
-
l
u
m
i
n
o
u
s

Surfaces

Uniform fill
lighting
Wall washing
Affect whole
room
LED
spot

Accent lighting
Linear
wash
light
(multiple
spots)

Wall washing
Small room
S
u
r
f
a
c
e

I
l
l
u
m
i
n
a
t
i
o
n

LED
with
optical
lenses

Street lighting
Environment

D
ISSN: 1411-1284

Proceeding 10
th

III. CASE STUDY

Galleria Department Store
Lighting concept
ighting concept of this building is changing
surface, which faades color can changes
dramatically. This concept has been designed by
Rogier van der Heide and Simone Collon from Arup
Lighting, and its architect, Ben van Berkel and
Caroline Bos from UN Studio. To create this
changing surface concept, all surfaces of buildings
faade have to be renovated and redesigned to become
more simple and compact. (see fig 6)

Fig.6. Galleria before (left) and after LED (right)
(source:www.lightingacademy.org)
To illuminate 3278 m2 of facade, 4330 Xilver
RainDrops LED lamps (see fig 7) have been installed
with aluminum frame on its all surfaces. These lamps
which cover all surfaces of buildings faade make
Galleria just like a nest at day time and a box of light
at night time.

Fig.7. Xilver RainDrops

Dynamic lighting
ynamic lighting technique which is applied here
is dots of light. With this technique, each of
lamps can be seen as an independent light source and
can be independently controlled from its color
changing, switching, and dimming by DMX-512
control system.

Fig.8. Galleria at night

The main advantage of this technique is its
flexibility in performing light composition to create a
high resolution image and writing to inform the
product or special event or promotion being held
there. Dots of light makes a faade surface just like a
television monitor with each lamps acts as the pixels.
(see fig 8)
Nevertheless, this dots technique still has a
disadvantage. Dots of light requires a lot of lamps to
cover all surface of a faade. Therefore, it also
requires a lot of expenses.

Persuasive and emotional trigger
In a competitive world of retail, efficient lighting
is very important to perform the product in attracting
the customer. A bright and dynamic ambience can
attract, inspires, and affects the customer to come.

Creating the right trendy and stylish ambience
Color trend is affected by fashion trends and
styles. Galleria Department Store as fashion retail
from high braded fashion must follows this trend and
style.

Uniqa Tower
Lighting concept
aade lighting for this building was designed by
lighting designer LichtKunstLicht as an integral
part of its architecture. The lighting concept is a high
quality and unique lighting and reflecting image of
Uniqa Insurance to the Vienna nightscape. (see fig 9)

Fig.9. Uniqa Tower at night (source:www.uniqagrup.com)

L
D
F
ISSN: 1411-1284

Proceeding 10
th

Faade Uniqa Tower with 7000m2 area consists of
2 layers with 50 cm space between them (hollow
glazed faade). Barcos MIPIX-20 LED lamp with 65
IP was installed on 2900 aluminum frame modules in
this hollow glazed space with 8 cm space between
each lamp. (fig 10)

Fig.10. Barcos MIPIX-20 (left) and its installation in
hollow glazed space (right) (source:www.barco.com)

Dynamic lighting
is lines of light. (see fig 11) LED lamps were
arranged according to aluminum frame modules so we
can see them as lines of light at night time.

Fig.11. Uniqa Tower at night: twists and turns
sequence (source: Professional Lighting Design 51)

The main use of this technique is expressing 2D
elements in a 3D element. Therefore, lines of light is
suitable to illuminate architectural details of a
building. Lines of light also uses fewer lamps than
dots of light because it doesnt cover all surfaces of
faade.
The disadvantage of line technique is resolution of
images resulted is not as good as dots technique
because it depends on its frames.

Creating the right ambience
Up until now, there are 2 programmed lighting
sequences, i.e. the first sequence that has been used
since Christmas 2005 and Twists and Turns
sequence that has been used since April 2006.
Lighting sequences can be easily adjusted and
changed according to the special event like Christmas
or New Year. As a Viennas landmark, lighting
adjustment becomes very important because besides
an office, Uniqa tower as also an integral part of a
greater citys public space, i.e. Danube Canal.
Multimedia device
The integral part of lighting concept of Uniqa
Tower is displaying Uniqa Insurances logo.(fig 12)
This companys logo is displayed with blue light and
turns around the building 2-3 times in 20 minutes.
Therefore, this lighting concept is one way for the
company to promote itself.

Fig.12. Display of Uniqa Insurances logo (source:
Professional Lighting Design 51)

Dexia Tower
Lighting concept
he lighting concept doesnt mean to present the
fix programmed lighting sequence but more to
express architectural characteristic and urban context.
Buildings characteristic, orientation, volume, and
scale were used to be parameters in creating spatial
concept, temporal, and interactive lighting. (see fig
13)

Fig.13. Dexia Tower (source:www.spacecannon.it)
D
T
ISSN: 1411-1284

Proceeding 10
th

LED lamps used to light up the tower are Ath
Luxor Custom (see fig 14) with this following
specification:
1. 4.200 Ath Luxor was used to light up all windows
in 36 floors. Each lamps can be independently
controlled. Lamps were installed inside the
building on the bottom side of windows.
2. 22 Ath Luxor Custom with 40 pieces of 3W
monochromatic blue LED in each fixtures were
used to light up all columns of tower.
3. 60 Ath Luxor Custom were used to light up the
perimeter and main entrance area.

Fig.14. Space Cannons Ath Luxor Cutom LED lamp
(source:www.spacecannon.it)

Dynamic lighting
is surfaces of light. (see fig 15) LED lamps
installed on each window create these windows glow
and can be seen as surfaces of light that cover all
faade of building.

Fig.15. Dexia Tower: blue, red, and yellow
(source:www.dexia-tower.com)

The main advantage of surface technique is its
effectiveness in lighting up a surface (wash light).
Therefore it can cuts initial cost of lighting
installation.
From all techniques of dynamic lighting, surfaces
of light generates the lowest resolution images
because its color changes uniformly on 1 surface.

Creating an interactive urban space
Laboratory for Architecture and Urbanism,
LAb[au], introduced a new interactive lighting
concept with Touch system. In front of tower, Place
Rogier, there is controller station where people can
interact with towers lighting and appearance either
individually or collectively trough multi touch screen.
(see fig 16)
Static inputs (touch, hand, finger, arm, and
duration) define color coordinate (hue) while direction
of dynamic input (hands movement: horizontal,
vertical, diagonal, positive, and negative) define color
saturation (chrome). In this station, there is a monitor
that sows its preview from given inputs on touch
screen.

Fig.16. Touch system, people can interact with tower
individually (left) or collectively (right) (source:www.dexia-
tower.com)
IV. CONCLUSIONS

ight Emitting Diode (LED) is the latest artificial
lighting technology. LED lamp is categorized as
anew lamp family, Solid State Lighting, because
unlike incandescent that burns tungsten filament or
discharge lamp that ionizes gas inside bulb, LED
converts electricity into light directly.
To create a building into a multimedia device,
there must be 2 following conditions:
1. LED is installed facing outside of building.
All of 3 techniques can be used but in
deciding which one and LEDs armature to
be used, we need to consider the buildings
size or scale to its environment or city,
effectiveness of initial cost, and the
impression we want to show.
2. Lighting must be given on all of a large
faade surface. The faades surface must be
uniform and flat, there are no uppermost
architectural details or accent. Therefore,
buildings faade can acts as a gigantic
monitor television. In Galleria Department
Store case study, all of old faade must be
renovated and redesign to realize this
concept.
Although LEDs initial cost is very expensive, its
operational cost is the cheapest among any other
lamps. Therefore, all of total cost in a long term is
D
L
ISSN: 1411-1284

Proceeding 10
th

also cheapest compared than incandescent or
discharge lamps.
By using LED as a lighting technology in
architecture design, it is believe that in the future there
will be a tremendous architectural design variety.

REFERENCES

1) Gegana, Gregorius A. LED Technology in Architectural
Lighting of Buildings Facade. 2007.
2) Egan, M. David. Architectural Lighting. New York:
McGraw-Hill, 2002.
3) Lechner, Norbert. Heating Cooling Lighting: Design
Method for Architects. Canada: John Wiley & Sons,
Inc., 1991.
4) Gary Steffy lighting Design Inc. Time Saver Standard
for Architectural Lighting. United States of
America: McGraw-Hill., 2000.
5) Laganier, Vincent. LED Wall of Light International
Lighting Review 2004: 72.
6) Cornilessen, Sjef. Lighting Design for Healthcare
International Lighting Review 2005: 86 88.
7) de Kruiff, Marike, dan Martin Knoop. Colour in Indoor
Environment International Lighting Review 2006:
100 103.
8) Ruffles, Paul. Out and About Light and Lighting,
August September 2000.
9) DenBaars, Steve. Current Status and Future Potential
of Solid State Lighting. Style Sheet.
http://www.ssldc.ucsb.edu
10) Beesbe, David A. Light Emitting Diode (LED)
Lighting Systems. Style Sheet.
http://www.ssldc.ucsb.edu

ISSN: 1411-1284

Proceeding 10
th

Form in The New Architecture Perspective

Yandi Andri Yatmo* and Kristanti Dewi Paramita**

Tel. 7863512, fax. 7863514 email : yandiay@yahoo.co.uk
** Fac. of Engineering, University of Indonesia, New Campus UI Depok 16424
Tel. 7863503, fax. 7270050 email : kristanti_findig@yahoo.com

BACKGROUND

In daily life, human has always connected with
buildings. This connection enhanced by the
fact that building is a place for human
activities. A building must accommodate the
functional need of the user (as known as
human). On the other hand, in order to develop
this sense of connection, it involves the
emotional side of the user. This emotional side
proves that the building have its essence or
what is said as form.

Buildings form is a design result with high
level of considerations and complexity involve
on it. So many elements acquire in one kind of
design. As Fritz Schumacher said, in designing
there is three phases to pass. The phases
consist of initial phase, middle phase, and the
goal
1
. When planning a building, the first
phase is initial phase. Middle phase is the
second phase, when the architect figures the
building characteristics and do some further
thinking about how it would work in the
future. What is simply interesting is the last
phase, known as the goal. Because it has not
stated as a physical goal, but rather as some
principles, which was express as an informal
type or general type
2
.This type clearly affirm
the idea about the importance of a concept
while designing, whether explicitly or
implicitly.

This concept usually evolves from the architect
expression about the building environment or
the imagination about the building
characteristic
3
. Knowledge about design
location is highly needed to arrange a
contextual-centered analysis. As the time goes
by, the community lifestyle is constantly
changing. These changes affect the community
comprehension about form in architecture.
This is unavoidable, because human is
architectures center
4
. The New Architecture is
also one of the effects of those changes, which
appear as the improvement of architectures
philosophy relate to the developing of urban
community. The development of form in this
philosophy is unique, placed in the language
that buildings provide to its user.

TOWARDS FORM IN THE NEW
ARCHITECTURE

Defining Form

Plato said that the real meaning of form is a quality in
reality
5
. Form is the idea or the essence of any object
in this world. A form is stable and never changes, and
the basis of the real image of the object. It is different
with shape. Shape is something that can be sees by us,
such as content, composition, color or texture
6
. It is
commonly known that form and shape are usually
united in one and develop the whole expression of any
object exist.

In architectural definition, form and shape exist in
every building, just as if they exist in every living
being. Soriano explain furthermore that in the tradition
of humanist aesthetics, form is a bundle of what has
done or contain
7
. Therefore, in other words that form is
a package of meaning from an architecture object. The
characteristics are unique, unusual, and last forever.
Form could be understood permanently because
everyone understands it without bordered by social
class or such categories. It is not universal, but singular,
which means something very extraordinary so that
everyone could recognize it.

Architect has the right to define the buildings form by
making a subjective or personal decision. A subjective
decision is usually made by using the theory or method
that is significant for the design condition. On the other
hand, a personal decision is made by using the
designers experiences or memories. Charles Jencks
also said about this decision process as playing God
8
,
because of its random feature.

Form, Community Condition and Evolution

The history of the humankind elucidates the relation
between buildings form and community condition. As
an example, modern era appear as one of the influence
of the First World War. It is said that after the war,
modernismthat is before is a minority tastehighly
mark out the age
9
. Technically, the Modern building is
more economically rational to build in after war period
because it has less decorative elements that give
minimize the building cost. On the other hand, the
ISSN: 1411-1284

Proceeding 10
th

shape stability of the Modern building is a
representative of the community requirement
in after war condition that still in chaos. It
expresses the need of a strong and powerful
image. James Smith supports this idea by his
statement that Modern design is seen
through the lens of masculinity
10
It has
clearly stated that the essence of the building
has connected with the community recent
condition.

It has said in the above sentences that
community is always changing. Technology
has an important part on this. The development
of technology improves the human ability and
accomplishment to communicate, and later
influence to the development of the humans
living space. This development of humans
living space contributes to the conception of a
new culture in a human life period. Chris Abel
named this new culture as eco culture
11
. One of
the characteristics of eco culture is its form
that arranged for the environment, has a
significant purpose and allocated for the
climate condition
12
.

The humans living space improvement makes
some influence on the flexibility of buildings
development. We could build anything
anywhere. It is decentralized, not centralized
as the previous culture did. This decentralized
pattern also affects the community condition to
become plural, because there is no certain
domination or power like the centralized
pattern. This plural community is not limited
by certain codes or references in the beginning.
With plurality, the feature of the form
expression would be more open and free. As
Gausa said, plural could be defined as an
opening for possibilities. Plural is, by
definition, open: it is notor does not seek to
beunivocal, absolute, complete...
13
Later,
because there is no certain domination, it
allows a building to expose its own form
purely. This is the beginning of a new
perspective in architecture that comes from the
evolution of the community.
The Thoughts of New Architecture

The term of New Architecture is coined by
Jeffrey Kipnis. The appearance of this term
was influenced by the MOMA Decon
exhibition held in 1988. Along with that, the
works of several architects such as Peter
Eisenman, Frank Gehry, Daniel Libeskind,
Bernard Tschumi, Rem Koolhaas and Zaha
Hadid also have impact for the term. Their
contribution in architecture has considered
exotic and then categorized as New Architecture
14
.

The final aim of the New Architecture is to create some
real significance of Architecture in this contemporary
world fully filled with commercialism. Commercialism
drives the market, create barriers appear in codes or
references. If the architecture world only follows these
commercialism codes, it would end as the death of
architecture meaning. This is because the building only
pursues the profit, not the soul. Morris said
architecture is the quintessential art, an expression
of pleasure in work, an art made by people for the
people as a joy for the maker and the user We could
deduce that the profit-oriented building does not
presents any art, any pleasure or even joy, because the
building does not have any soul. They do not seek for
their form. This architecture without meaning would
not produce any bonding with the user, because they do
not react emotionally.

As been said before, technology has a great impact. In
the New Architecture, this improvement really has an
influence, especially on the communication technology
and digital technology. Charles Jencks said that there is
a paradigm to use these new sciences to create a new
dynamic architectural expression
15
. The existence of
the digital revolution made us think differently.
Mandour support this idea by saying that digital
revolution affecting the way we think about
architecture. It allows the expression on a grand scale,
widen the architects range and change the
communication of architecture language
16
.

Unger have a theory on the New Architecture product
characteristics. He said that such an architecture
must be blank, it must be point and be incongruous and
incoherent
17
Blank is because it is not connected
with any classic codes. The term incongruous and
incoherent, lead us to the basic need of seeking for
form in New Architecture, which is the need of
information. It is inescapable that information is highly
required to develop the form for a building or any
architecture product. This is because so many elements
attached each other in one design condition, so
information as many as could be obtained would
support the form defining process effectively.

INFORMATION AND FORM IN THE NEW
ARCHITECTURE

Sensing the Information in the New Architecture

Gausa said that information on the New Architecture is
synthetical, condensed and it consist in layers
18
. The
layers are the key to know about how human could
sense the information. As said before, technology
affects the living space improvement that rise the
human need about mobility. When a human moves,
even though his movement coordinate is limited
ISSN: 1411-1284

Proceeding 10
th

because of the gravity, he could sense his
surroundings widely. With this sense, human
could collect information from his surrounding
without struggling too much. This information
also often determines the way he moves, based
by his purpose or his need for the activity. De
Carteau said that there is some network
appears from the movement and it creates
some compact stories that related one to
another
19
. We could reckon that the movement
influences the network, and it would later
create certain space dimension from those
kinds of circulations.

The difference between the New Architecture
and the previous time is primly in its
movement. In the previous time, they delivered
the information in united form and most of all,
only seek for efficiency. They use this kind of
offer because the information enlightening
design in the building is not with movement
but by contemplation. The contemplation way
is no longer suitable, because this is an era
which everything must be in speed or appear in
instant. The characteristics of the New
Architecture are in its kinetic forces rather than
staticity
20
.

The Influence of Information in Spaces

Nowadays, information is one of the humans
basic needs. When you communicate with
others and exchange information, you do not
have to see them in a physical meeting like
before. You could do it while doing other
activities, like walking, eating or even while
you are buying your groceries. This is
something evolutionary if we compare it with
the habits before, when speak or walk while
eating could be consider not polite or taboo.
These changes on communication behavior
would automatically change the space meaning
from static or fixed become dynamic and
changeable. This meaning comes because you
could use the space while continually moving
or do everything you could. Soriano
designated this kind of space with the term
collective space
21
. He described the space as
a space where the connection between the
property and its own function is no longer
clear
22
. It could be blur or even disappeared.
What are constant about this space are only the
individuals inside of it. The space inside is
built with the imagination of the individuals.
The imagination comes from the information
each person collect, and everyone would react
differently based on their needs or their
personalities.

We could see the example in a building that has a static
function like shopping mall or train station. The main
user of the building is the passengers or the shoppers.
However, in reality, it is not always like that. You
could use the train hall or the shopping corridors for
meet people, socialize and have a chat, or even playing
skateboard with your mates. This is a rough example of
how the building function could change and still
accommodate the need of the external users. This
phenomenon would create a paradox because of the
cycle appears in the spacefrom fixed space to the
dynamic one, all over again. The paradox appears
because there is something contradictive happens in the
space of the New Architecture. Human determine the
architecture, but on the other hand, architecture
determine human moves and imaginations. Tschumi
described this phenomenon as violence in
architecture
23
. The term violence could be analyze as
the intervention between human body and space one
and another.

Information Kinds, the Process, and the Final

The informations main purposes are to educate,
entertain, or to declare something
24
. The information
collected for design process is related with the human
life pattern. It is something about consuming,
producing, or just living as a human
25
. It could not be
served as two-dimensional form, but must contain
space and time elements. The information usually
presented in diagram, as compressions of information
26
.

This information then develops some boundaries. These
boundaries come from the information that establishes
a dynamic plan and certain message of the building.
The message builds building essence or form. The
boundaries come from a set of imaginative movement
from the plan that placed by the message. The message
is so powerful it determined the absence of shape
27
. The
urge to vanish the statement of shape comes from the
act of avoiding any certain composition. Composition
is considered as something subjective and must be
avoided. The expectation is the form could create a
space purely based on the message as the information
we produce.

Surface is a physical product of information, and the
outline of the space. The definition of surface is the
outer part of something
28
. The surfaces of space are
something straightly in touch with human in their daily
activities. In the New Architecture, they produce the
information based on the message to influence the
surfaces. Later, the human will feel that there are some
certain messages from the space while they
experiencing the surfaces. Surface composition could
be tangible or intangible, up to the message. Both
tangible and intangible surface has the same purpose to
give the information to the users about the message.
Then, the surfaces would establish the final form from
ISSN: 1411-1284

Proceeding 10
th

information process, known as the visualized
form. Charles Jencks divided these form
expression in to three parts. It is fractal
architecture, biomorphic building, and iconic
architecture
29
.
Fractal architecture is some expression that
used a repetition method of a disorganized
geometric pattern
30
. Daniel Libeskind said that
the basic idea of this form of expression is to
propose disorientation consist of several
components with its own meaning, just like
what he did in his work Jewish Museum.

Biomorphic architecture or often called as a
form of organi-tech is concentrated in
ecologically sustainable architecture. In the
attempt to solve the urban condition with high
level of pollution, the form expression that is
sustainable for the environment is necessary.
Besides, to prevent the face of the city from
rigorous and cold image, this expression uses
the nature as basis. In the end, the form
appears as an organic form. The organic form
is usually consists from one cell that later
would grow based on the space-related
pattern and the information contained in the
form. The form could also unify with the
surroundings and later would establish a
landscape.

The last one is iconic architecture. Its main
character is super-impressive. This kind of
form has several purposes that have social or
public qualities. These qualities appear
because of the building need to be an icon or a
landmark. The iconic architecture represents
the human needs to be an icon or to be
different from others. It is the part from the
need to be appreciated by others based to
Maslow theory
31
.

In current community with high level of
plurality, the need to be higher or
monumental could no longer be applied by
ornaments that show the social class. To create
a monumentality expression the solution is
designing some singular and extravagant expression. In
the New Architecture, this form should remain eternal
and have a high level of locality concept, as a bonding
form with its environment.

Even though there is differences in the designers
media, but there is some process they surpass through
in the production of form. They treat the information
that collected before and later produce the main
expression to verify the final form for the design.

CONCLUSION

The thoughts of the New Architecture appear as a
reaction of the degradation of architecture meaning in
some era. It is an era when architecture is only market
consumption and no longer has any soul in it. The New
Architecture is an evolution emerges together with the
development of the technology that changes human
life.

Human is the basic part of New Architecture. As
human improve, architecture is widen. The New
Architecture introduces the thoughts about building
form as a distinctive issue rather than concentrated only
in composition. Form in architecture is considered as an
answer for human problem in reality that more complex
nowadays than the time before. The final form would
appear with dynamic character on it, because it is
produced in the context of space and time.

Information is very important in the New Architecture
that divides into two sections, the collected information
and the produced information. The collected
information is everything about site, the user and the
surrounding. The produced information is the main
essence of form in the New Architecture. This
information set the message of the building. The
information production could be seen in the surface
manipulations. These manipulations would give
different information to the users and make the users to
develop some perceptions. It is these perceptions that
define the true meaning of form in the building. Users
then could interact with the building. With form as the
essence of architecture, the possibilities of exploration
in a very diverse field are widely open.

REFERENCES

1
Egon Schirmbeck. Idea, form, and Architecture: Design Principles in Contemporary Architecture.
Van Nostrand Reinhold Company. New York. 1987. Page 9.
2
Ibid.
3
What is meant as characteristic is the impression that the society commonly aware of. The example
that usually universally accepted is that the office building must have a professional image, or a
kindergarten must have a cheerful feeling. These characteristics require will have impacts on the
production of the building essence.
4
Peter Noever. Architecture in Transition, Between Deconstruction and New Modernism. Prestel .
Munich. 1989. Page 116.
5
http://en.wikipedia.org/wiki/Forms_in_architecture, last modification 4th January 2007.
ISSN: 1411-1284

Proceeding 10
th

6
Thesaurus:All reference books.
7
Manuel Gausa, et al. The Metapolis Dictionary of Advanced Architecture. City, Technology and
Society in the Information Age. Actar Publisher. Barcelona. 2003. Page.236.
8
Charles Jencks dan Karl Kropf. Theories and Manifestoes of Contemporary Architecture. John Willey
&Sons. Great Britain. 1997.Page.6.
9
http://en.wikipedia.org/wiki/Modernism, last modification 24th April 2007
10
ocw.mit.edu/NR/rdonlyres/Architecture/4-645Fall-2004/7BD9E846-BCA8-4235-9D09-
949505AAE10B/0/responses_5_2.pdf.
11
Chris Abel. Architecture, Technology, and Process. Architectural Press. Oxford. 2004. Page18.
12
Ibid.
13
Society in the Information Age. Actar Publisher. Barcelona. 2003. Page 486.
14
Jeffrey Kipnis. Towards New Architecture. Academy Edition. London. 1993. hal.41.
15
Proceedings of London Conference, "New Science, New Urbanism, New Architecture"
http://www.katarxis3.com/Page3
16
Mohamed Alaa Mandour. ARCHITECTURE FORM IN THE 21ST CENTURY. Architecture
Department, Helwan University, Egypt. Page 195.
17
Jeffrey Kipnis. Towards New Architecture. Academy Edition. London. 1993. Page 42.
18
Society in the Information Age. Actar Publisher. Barcelona. 2003.Page 21
19
Michael De Certeau. The Practices of Everyday Life. University of California Press. California.
2002. Page 93.
20
Society in the Information Age. Actar Publisher. Barcelona. 2003.Page 345.
21
Society in the Information Age.Page 561.
22
Ibid.
23
Martin McQuillan. Deconstruction, A Reader. Edinburgh, University Press. Edinburgh. 1988. Page
230.
24
Society in the Information Age. Actar Publisher. Barcelona. 2003. Page 343.
25
Manuel Gausa, et al.Op Cit.Page 162.
26
Ibid.
27
Ibid.
28
Thesaurus U.S, Microsoft Office 2003.
29
Proceedings of London Conference, "New Science, New Urbanism, New Architecture"
http://www.katarxis3.com/Page56.
30
Thesaurus, all reference books.
31
In Architecture for People, New York, 1980, Byron Mikellides in Appendix part, theres a theory
about human need called Maslow theory, which consist of:
1. physiological needs
2. safety/security needs
3. social needs
4. self esteem/ego needs
5. actualization needs
Iconic architecture represent the fourth need, which is the need to be appreciated or considered
extraordinary by their surroundings.
ISSN: 1411-1284

Proceeding 10
th

Utilization Effectiveness of Jakarta Municipal Asset
(Case Study: East Jakarta Municipal Youth Center)

Azrar Hadi
*)
and Ii Karunia
**)

*)
Lecturer in Department of Architecture and Urban Studies University of Indonesia..
Department of Architecture, Faculty of Engineering, University of Indonesia
Telp. +62-21-7863512, Fax +62-21-7863514, email : azrar_ind@yahoo.com

**)
Urban Asset Management student and staff of Jakarta Municipal..
Jl Merdeka Selatan 8-9 Blok G Lt-13 Jakarta 10110
Telp. 3823255, Fax 3848850, email : mazik65@yahoo.com

Abstract:
This study is worked through about utilization
effectiveness of East Jakarta Municipal Youth Center
(EJMYC) as urban facility that developed to
accommodate young generations activities (sport, art
and culture). The purposes of this study are to know:
1) community perception of location, facility,
accesibility, and neighborhoods of EJMYC; 2) their
visit motivation. This study is more quantitative
research. Operational model of this study to integrate
asset management theory with an eclectic models of
theoritical perspectives Paul A. Bell et al. The data
collected by quiestionnaire and deep interview. Two
hunred and eighty four East Jakarta Municipal
community that classified into student, sport
teacher/trainer and public society as respondent.
Based on the data analysis and in depth interview,
the result of the research as follows: 1) condition of
location, facility, accesibility, and neighborhoods of
EJMYC had not yet appropriate for urban
community needs, but as urban facility the EJMYC
was still need; 3) utilization of EJMYC have still not
optimum, ascertainable from visitor and occupancy
that lower and unfitted with functionality of facilities;
4) under utilization of EJMYC as the impact of their
perception and motivation. In this case the role of
their perception and motivation as explainable: (a)
minimum of visitor indicated that EJMYC had not
yet appropriate for urban community needs; (b)
minimum of visitor indicated that community
motivation is weak; (c) weak motivation of
community drive by bad perception about facilities
condition of EJMYC and demotivated of sport
teacher/trainer; 5) utilization effectiveness of EJMYC
positively influenced by perception and motivation of
community.

Keywords : Asset utilization, Youth Center.

1. INTRODUCTION

ast Jakarta Municipal Youth Center (EJMYC)
is a public facility that developed by Jakarta
Municipal to accommodate young
generations activities (sport, art, and culture).
Its has been operating since 1971. Hasurungan Pakpahan
research proved that the young generations infrequent
utilised EJMYC (Pakpahan, 1984). According to
EJMYC management report, in year 2005, incidental
activities at the sprort center (30,41 %) and auditorium
(31,51 %).
Mark W. Patterson said, asset management is the
science or art of thedirecting the management of real
estate investment to ensure that its value is maximized
and enhanced over the long term for benefit of the
investor (Patterson,1995:3-4). In the mean time, Rambat
Lupiyoadi and A. Hamdani said, there are trees quality
orientation that must be consistent mutually : (1)
consumer perception, (2) products or services), and (3)
process. Contribution from the trees quality orientation
can be contribute to success of company or organization
from customer satisfaction, employee satisfaction, dan
profitability of company or organization (Lupiyoadi dan
Hamdani, 2006).
Management of East Jakarta Municipality Youth
Center (M-EJMYC) as organizer as useful as possible
must be optimalizing existence of EJMYC. Success and
achievment of M-EJMYC with optimalization of EJMYC
as one of centre for community center expected can
improve acceptance of retribution and invite all organizer
of advertisement to install advertisement so that able to
give addition of acceptance for Government of DKI
Jakarta from sector of Iease of Advertisement for
installation of advertisement around the city fasilitas in
the range of EJYMC.
Existence of the ETMYC with all potency represent
matter which require to be managed carefully and
E
ISSN: 1411-1284

Proceeding 10
th

socialized in an optimal manner so that able to give
positive image and added value for society and all
investor and also other party owning enthusiasm to do
cooperation with asset exploiting. This research will to
know: 1) community perception about EJMYC; 2)
community motivation of visit and exploited EJMYC
facility.
This Fact expected can revealing existence and
potency of EJMYC as one of city facility and source of
income for Jakarta Municipal. Issues in this research : 1)
whether location condition, facility condition,
aksesibilitas condition and neigborhods condition of
EJMYC able to support existence of EJMYC as one of
city facility that matching with the community
requirement ?, and 2) whether perception about location
condition, facility condition, aksesibilitas condition and
neigborhods condition of EJMYC can influence to
community motivationt to visit and exploit facility of
EJMYC ?

2. BASIC THEORY

There are four factors effecting product positioning
effectiveness in property marketing : 1) products
characteristics (property proucts/services are usually
high-valued, principally stemming from high production
costs. The risk in selling these products/services are also
high), 2) customers characteristics (proucts/services are
not normally bought frequently. So, they have more
sringent criteria in their purchase decisions. The most
important is their affordability. Second, the facilities to
effect the demand. Third, the product characteristics), 3)
competitive characteristics (channel design is influenced
by competitors channels. Some developers want to
compete in the same manner as their competitors), and
4) company characteristics (the company size determines
the size of its markets and its ability to design marketing
channel. Not least in importance is sales force qualities.
The knowledge, experience,dedication, creativity, and
team work are among the important factors determing
companys ability to withstand market challenge) (Iman,
2002:256-261).
Jeremi Dasso et all said, nany elements are
considered in site selection and analysis property : 1)
location or situs, 2) accessibility, 3) size and shape, 4)
physical characteristics, 5) utilities and services, 6)
applicable public regulations, and 7) cost or value
(Dasso, 1992:222). In the mean in time, the location
focus for service firms should be on the determining the
volume of business and revenue. There are eight major
components of volume of business and revenue fot the
srvice firms : 1) purchasing power of the customer
drawing area, 2) service and image compatibility with
demographic of the customer drawing area, 3)
competition in he area, qulaity of the competition,
uniqueness of the firms andcompetitors location, 6)
physical qualities of facilities and neighboring business,
7) operating policies of the firm, and 8) qulaity of
management (Rendera nd Haizer, 1996:213-214).

Perception is the interpretation of received stimulus.
In other word, perception represent ways in which one
senses, interprets and analyses incoming stimulus.
Stimulus is something that increase bodily or mental
activity. Perception has been used in developing
marketing communication strategies such as the
perceptual mapping, wihich is founded on the premise
that individuals will seek to relate to things in term of
relationship with or similarity to things with which they
are already familiar (Iman, 2002:131-132). Paul A. Bell
et all said, perception of the objective physical conditions
depends on the objective conditions themselves, as well
as on the individual different factors and the attitudinal.
If this subjective perception determines that the
environtment is within an optimal range of stimulation or
is congruent with intended behaviour, the rsults is
homestatic, the adjectif form of homestasis, or an
equalization of desired and actual input. On the other
hand, if the environment is experienced as outside the

Figure -1. An eclectic model of theoritical perspectives

OBJECTIVE
PHYSICAL
CONDITIONS
INDIVIDUAL
DIFFERENCES

SITUATIONLA
FACTORS
SOCIAL
CONDITIONS

CULTURAL
FACTORS
PERCEPTION
OF THE
ENVIRONMEN
AS WITHIN
OPTIMAL RANGE
OF STIMULATION
AS OUTSIDE
OPTIMAL RANGE
OF STIMULATION
HOMEOSTASIS
AROUSAL
STRESS
OVERLOAD
REACTANCE

COPING
ADAPTATION
and/or
ADJUSTEMNT
CONTINOUED
AROUSAL
and/or
STRESS
POSIBBLE
AFTEREFFE
CTS
and/or
CUMULATIV
IF COPING
UNSUCCESSFU
L
IF COPING
SUCCESSFUL
POSIBBLE
AFTEREFFE
CT
and/or
CUMULATIV
ISSN: 1411-1284

Proceeding 10
th

optimal range of stimulation, then one or more of the
following psychological states results : arousal, stress,
information overload, or reactance. The presence of one
or more these states leads to coping strategies. If the
attemped coping strategies are successful, adaptation or
adjustmenoccurs, possibly followed by such aftereffects.
Should be the coping strategies not be successful ,
however, arousal and stress will continue, possibly
heightened by the individuals awareness that straegies
are falling. Potential aftereffects of such inability to cope
include exhaustion,learned helpness, severe performance
decrements, and mental disorder. Finallly, as indicated by
the feedback lops, experiences with the environment
influence perception of he environment for the future
encounters and also contribute to individual differences
for future experiences. Theoritical concepts an eclectic
scheme of environment-behaviour relatishionship as
presented in flowchart in figure 1. (Paul, 2001:131-133).

3. EXPERIMENTAL RESULTS

EJMYC provided with facility : swimming pool,
sports center, auditorium, climb wall, gymnastic room,
courses rooms (computer, music, and english Ianguage),
cultural art gallery, mushola and park really not yet been
exploited with optimum by community. Perceivable
above mentioned condition remember from 284
responder bringing back kuisioner really do not all
facility which is there are in EJMYC known and
exploited by society and also most responder tell seldom
pay a visit to EJMYC (81,3%). Some of th responden
said that many responden which still own motivation pay
a visit to EJMYC (55,7%).
Facility which is at most known by the responder is
swimming pool ( 93%) while facility which less a lot of
known by space of computer courses (20%). Meanwhile,
seen from experience of responder really which is at
most exploited by responder is swimming pool (72%)
while facility which less a lot of exploited by responder
is room of Ianguage courses (8%). Facility which is made
available EJMYC to date still be needed, Tthe mentioned
seen from > 70% responder express all facility in
EJMYC still be needed except room courses (< 70%).
The result of survey about condition indicate that
condition EJMYC have disagree with requirement
socialize by the reason: 1) its location uncovenience (
82,7); 2) its facility unmaintenance ( 75,1); 3)
accessibilty is not good ( 83,1%); and 4) neighborhoods
less support ( 81,7%). Meanwhile, result of survey about
role EJMYC really existence EJMYC still playing a part
of requirement accomplishment able to go into society
(60,6%), training the sports (63,4%) and the achievement
improvement management in have sport (63,1%).
Understandable the mentioned remember its visit to
EJMYC get support from family ( 57,4%), friends (
59,6%), and teacher/trainer (64,8%).
Regression output of Perception variable shown that
only the facility condition which by siginifikan positive
influenced to somebody perception about EJMYC. But
then, location condition, accesibility, old age and the
economic condition positive influenced to community
perception about EJMYC but dont significant, while
environmental condition about, occupation and level of
education negative influenced to community perception
about EJMYC but dont significant. In the other proses,
regression output of Motivation variable shown that
affiliation, family driven, friends driven, and
teacher/trainer driven positive influenced to community
motivation for visit to EJMYC. while skilled requirement
and the requirement have achievement [positive
influence to community motivation for visit to EJMYC
but do not significant.
Unfavourable community perception about physical
condition of EJMYC and lack of motivation pay a visit to
EJMYC give unfavourable continuation effect to
exploiting EJMYC. Anharmonic among perception,
motivate and mount of visit (PB) to EJMYC as
implication from anharmonic of the fact and expectation
of society to benefit boundary at the price of or the
expense paid its as presented in flowchart in figure-2 So,
that ready and M-EJMYC not yet able to realize public
service quality which the efficiency and effective.
Correlation of perception and motivation with number of
visited to EJMYC as presented in graphic in figure-2.
Existence EJMYC as public facilities unable to
fulfill society requirements which immeasurable to
progressively grow unfavourable perception to EJMYC
and lessen community motivationto pay a visit and
exploit facility of EJMYC. The Condition indicate that
EJMYC which initially more destined for society have
age 12-20 year, latterly claimed to able to fulfill
requirement socialize in common with various [his/its]
ISSN: 1411-1284

Proceeding 10
th

activity. With reference to [the] mentioned require to
be strived [by] settlement in so many area of[is inclusive
of among other things settlement adminitrasi budget
management system and retribution which the allocation
utilization of EJMYC.

4. CONCLUSIONS

1. Condition of location, facility, accesibility, and
neighborhoods of EJMYC had not yet appropriate
for urban community needs, but as urban facility the
EJMYC was still need.
2. EJMYC as public facility still need seeing that more
than 60 % of respondent was declare that EJMYC as
a needed and they used for anothers activity.
3. Utilization of EJMYC have still not optimum,
ascertainable from visitor and occupancy that lower
and unfitted with functionality of facilities.
4. Under utilization of EJMYC as the impact of their
perception and motivation. In this case the role of
their perception and motivation as explainable:
a. minimum of visitor indicated that EJMYC had not
yet appropriate for urban community needs,
b. minimum of visitor indicated that community
motivation is weak,
c. weak motivation of community drive by bad
perception about facilities condition of EJMYC
and demotivated of sport teacher/trainer.
5. Utilization effectiveness of EJMYC positively
influenced by perception and motivation of
community.

REFERENCES

[1] Bell, Paul A, et all, Environtmental Psychology :
Principles and Practice, 5
nd
Edition, Thomson, Australia., 1996
[2] Dasso, Jerome et al, Real Estate-12
th
Edition, Prentice-
Hall Inc, Englewood Cliffs -New Jersey,1992.
[3] Iman, Abdul Hamid Mar : An Introduction to Property
Marketing, Universitas Teknologi Malaysia, Skudai Johor
Darul Tazim - Malaysia. 2002.
[4] Lupiyoadi, Rambat dan Hamdani, A, Manajemen
Pemasaran Jasa, Salemba Empat, Jakarta, 2006.
[5] Render, Barry, and Haizer, Jay: Principles of Operation
Management, 2
nd
Edition, Prentice Hall, New Jersey., 1996

Figure -2. Normal P-P Plot of Regression Standardized Residual

0 . 0 0 . 2 0 . 4 0 . 6 0 . 8 1 . 0
O b s e r v e d C u m P r o b
0 . 0
0 . 2
0 . 4
0 . 6
0 . 8
1 . 0
E
x
p
e
c
t
e
d
C
u
m
P
r
o
b
D e p e n d e n t V a r i a b l e : P B
ISSN: 1411-1284
Proceeding 10
th
Water Quality Index as Indicator of Water Pollution.
A Case Study in West Tarum Channel as Sources of
Water Supply

Djoko M. Hartono; Sulistyoweni W; Dwita Sutjiningsih M
Departement of Civil Engineering Faculty of Engineering, University of Indonesia,,Depok
Tel. 7270029, fax. 7270028 email : djokomh@eng.ui.ac.id

Abstract To fulfill the need for water supply, river
is the main alternative of source of water among
other sources of water. However, the surface water
has been setting worst basically in its quality and it
is over its permissible criteria standards as source
water for water supply. The change in the quality of
water source is caused by people activities that make
pollution along the river in form of both domestic
waste industry waste and solid waste. In many cases
pollutant contained in the river are so hard that
water supply treatment plant can not run the
operation well and most of the time the service to the
user being stopped. Besides, global climate change
will affect raw water resources and has to consider
its effect soo. Water Quality Index is one of indicator
on water pollution is using to measure water quality
of surface water from West Tarum Channel which
are used by 3 Water Treatment Plan. Water Quality
Index which has adopted from Storet Method was
calculated and compare with Water Quality Index
adopted from the National Sanitation Foundation
USA with some modification and improvement, has
been measured and the results show the raw water
quality among 3 water intake has in the worst
category.

Keywords: criteria standard, water quality index, water
treatment plant.
I. INTRODUCTION

ater is very important for human being, animals
and plants and without water life on earth would
not exist. Surface water is one among other water
sources as raw water to produces water supply
processing by water treatment plan. The population
growth, rapid urbanization and industrial growth,
increasing of quality of life are some example on
increasing demand for water supply. Besides, quantity
and quality of surface water tend to decrease due to the
fact that impact on changes in nature as well as
pollution by man. The need to determine the chemical,
physical and biological characteristic of natural water
resources is essential in view of their utilization
particularly as a source of municipal water supply.
Water quality index which was developed in the early
1970s, can be used to monitor water quality changes in
a particular raw water for water supply, or it can be
used to compare a raw water supply quality in other raw
water supply in the region or from around the world.
The result can also be used to determine if particular
raw water is considered to be healthy. The study on
raw water quality was conducted at raw water intake of
Buaran, Pulo Gadung and Pejompongan Water
Treatment Plan. Among other water treatment plan to
fulfill demand of water supply in Jakarata, Buaran
Water Treatment Plan, Pulo Gadung Water Treatment
Plan as well as Pejompongan Water Treatment Plan are
using same source of raw water from West Tarum
Channel which is supplied form Jatiluhur Dam.

II. BASIC THEORY

Water quality status is based on State Ministry of
Environment Number 115 year 2003 on which
indicated several level of pollution on raw water
comparing with raw water quality standard.
Determination on water quality indicator can be done
with 3 formulation:
1.STORET Method, which used value system as
mention on Tabel 1.

Tabel 1. Determination Value System for Quality
Status
Parameter Numbe
r of
sample
Value Phisyc
s
Chemis
t
Biolog
y
< 10 Maximum
Minimum
Average
-1
-1
-3
-2
-2
-6
-3
-3
-9
10 Maxsimu
m
Minimum
Average
-2
-2
-6
-4
-4
-12
-6
-6
-18
Source: Ministry of Environment Decree No
115/2003
W
ISSN: 1411-1284
Proceeding 10
th

STORET method has 4 classification:
a. A Class, very good, score = 0, fulfill water quality
standard
b. B Class, good, skor = -1 to -10, low pollution
c. C Class, medium, skor = -11 to -30, medium
pollution
d. D Class, bad, skor -31, heavy pollution

2. Pollution Index Method is used to determine level
of pollution relative to parameter quality permissible.
This formula also based on State Ministry of
Environment Decree Number 115/2003. Pollution index
can be formulated :

Pij =(C
1
/L
1j
,C
2
/L
2j
,.........,C
i
/L
ij
) (1)
Whereas,
P
ij
= pollution indeks
L
ij
=concentration of raw water quality parameter
C
i
= concentration water quality chosen

Value C
i
/L
ij
= 1,0 is critical value.
Evaluation of pollution index can be categorized:
a. 0 Pij 1, in good condition fit in with quality
standard
b.1,0 < Pij 5, low pollution
c.5,0 Pij 10, medium pollutin
d.Pij > 10, heavy pollution

3. Water Quality Index.
One of the earlier attempt in formulating water quality
index was done by National Sanitation and Foundation
USA and Horton, 1970. According to them, water
quality index is

QI =
..... 2 W 1 W
CnWn ......... 2 W 2 C 1 W 1 C
+ +
+ + +
(2)

QI = water quality index
C = rating
W = weight
M1= temperature value
M2= obvious pollution value
Brown et al. and modification by Lohani, 1981, has
used the approach to develop a weighted mean index of
the form:
n
WQI = w
1
q
1
(3)
i=1

whwreas:
WQI= water quality index, a number between 0 and
100
qi = the quality of the i th parameter, a number
between 0 and 100
wi = the unit weight of the i th parameter, a number
between 0 and 1, and
1 1 q 1 w
n
1 i
=
=
(4)
n = number of parameter

The Water Quality Index uses a scale from 0 to 100 to
rate yhe quality of the water, with 100 being the highest
possible score. Once the overall WQI score is known, it
can be compared against the following scale to
determine how healthy or how pollute the water is on
giver time. Water quality cab be scale:
a. 91-100, excellent water quality
b. 71-90, good water quality
c. 51-70, average water quality
d. 26-50, fair water quality
e. 0-25, poor water quality

Raw water quality with rating falling in the excellent
and good range would able to support a high diversity
of aquatic life. In addition, the water would also be
suitable for all forms of recreation, including those
involving direct contact with water. Raw water quality
achieving only an average rating generally have less
diversity of aquatic organisms and frequently have
increased algae growth. Raw water quality falling into
the fair range are only be able to support a limited
number of aquatic life forms, and it is expected that
these water have abundant quality problem. A raw
water quality with poor quality rating would not
normally be considered acceptable for activities
involving direct contact with the water, such as
swimming.
The scale with value 100, will be based on maximum
permissible as mention on Government Decree Number
82-2001 for class 1 of raw quality. The scale with
value 0 will base on class 4 of raw water quality.
Raw water quality parameter will calculate from
dominant parameter occurred during time of
observation. The weight factor will use correlation from
statistic.


By using multivariate analysis among number raw
water parameter observed there are 9 parameter the
most significance parameter among 48 parameter
observed regularly by water supply enterprises. The 9
parameter are , turbidity, Total Dissolved Solid,
ammonia, iron, manganese, BOD, COD, suspended
solid and total coli form.
ISSN: 1411-1284
Proceeding 10
th
From Table 2, as can be seen raw water quality
comparing with water quality standard adopted from
Governor DKI Decree and Government Decree.

Tabel 2. Raw Water Quality Status (Year 2003-
2006)

Raw Water Quality
(Maximum)
No
.
Para
meter
Unit
SK Gub.DKI
No.582/
1995
PP No.
82/2001
B P 1 P 2 PG
1. Turbidity NTU 100 - 15326 8800 9060 13246
2. TDS mg/l 500 1000 312 330 282 241
3. Ammonia mg/l 1,0 0,5 2,28 2,62 2,75 4
4. Iron mg/l 2,0 0,3 13,3 0,95 0,88 29,8
5. Manganese mg/l 0,5 0,1 6,79 5,5 1,48 5,6
6. BOD mg/l 10 2 60 42 41 60
7. COD mg/l 20 10 106 144 101 210
8. S S mg/l 100 50 5290 2834 2540 14560
9. Total coli
x 10
Nol/
100 ml
10 1 201 4800 680 201
Sources: Observation and calculation (2006)

From above table, all of parameter exceed the quality
standard, and the highest number occurred will disturb
water supply treatment operation. The result of
calculation each method can be seen on Table 3, Table
4 and Table 5.

Table 3. Calculation Water Quality Method Storet
(Year 2003-2006)
Year
Raw
Water
Buaran
Raw Water
Pulo
Gadung
Raw Water
Pejompongan
I
Raw Water
Pejompongan
II
2003 -72 -76 -64 -58
2004 -74 -74 -66 -66
2005 -72 -74 -64 -66
Source: Calculation (2006)

From above table can be concluded that during year
2003 until 2006 raw water quality show in heavy
pollution condition ( > -31) as a result strong effort
made in produce water supply with quality standard
given

Table 4. Calculation Water Quality Method Pollution
Index
(Year 2003-2006)
Year
Raw
Water
Buaran
Raw Water
Pulo
Gadung
Raw Water
Pejompongan
I
Raw Water
Pejompongan
II
2003 11,75 11,79 10,91 10,35
2004 10,10 9,90 10,54 9,99
2005 10,70 10,34 11,09 11.03
Source: Calculation

From above table can be concluded that during year
2003 until 2006 raw water quality show in heavy
pollution condition ( > 10) as a result strong effort
made in produce water supply with quality standard
given or treatment process stop.

Table 5. Calculation Water Quality Method WQI
(Year 2003-2006)
Year
Raw
Water
Buaran
Raw Water
Pulo
Gadung
Raw Water
Pejompongan
I
Raw Water
Pejompongan
II
2003 21,21 21,21 22,02 25,08
2004 27,8 15,01 34,46 32,20
2005 15,98 15,01 45,22 46,88
Source: Calculation

From above table, almost all of raw water quality can
be categorize as the worst raw water quality and
indication have less diversity of aquatic organisms and
frequently have increased algae growth.
With that condition, water treatment plan is not allow to
operate. Since, water supply is very essential for human
being, follow up action has done and as a result more
chemical put to treat the water.
Water quality in Water Treatment Pejompongan better
than other water intake, since in the distribution gate in
Cawang , there is some addition treatment, by using
screen, turbulence as a result on water gate and pump
and additional chlorination put into transmission line.

IV. CONCLUSIONS

Population growth and increasing industrial along the
river which are polluting the river and erosion caused
by rain fall made raw water for water supply from
West Tarum Channel become worst. Special treatment
by adding more chemical in processing the water has
already done as a result more sludge produced. In the
near future, modification on water treatment plan is
needed so that water treatment plan can treat more
efficiency .
Many regulation in protecting raw water has made,
however in implementing that regulation need more
serious and coordination in between institution
relevance should be made.

REFERENCES

[1] American Water Works Association, 1999. Water
Quality and Treatment, A Handbook of
ISSN: 1411-1284
Proceeding 10
th
Community Water Supplies, fifth edition,
McGraw Hill, 1999.
.
[2] Bordalo A.A, Nilsumranchit W and Chalermwat K,
2001. Water Quality and Uses of the Bangpakong
River (Estern Thailand), Water Research Volume
35 No. 15 (3635-3642). 2001.

Appendices

Appendices 1. Calculation on Storet Method (2003)
Para
met
er
U
ni
t
PP
82
/20
01

Raw Water Quality

Buaran
Pulo
Gadung
Pejompong
an 1
Pejompong
an 2

Quali
ty
V
al
u
e
Quali
ty
V
al
u
e
Quali
ty
V
al
u
e

Quali
ty
V
al
u
e
Turb
idity
N
T
U
100
1070
8
-
1
1176
8
-
1 8089
-
1 5808
-
1

76
-
1 150
-
1 135
-
1 169
-
1

4295.
57
-
3
3968.
29
-
3
2983.
43
-
3 2659
-
3
TDS m
g
/l
500
288 0 241 0 330 0 282 0

92 0
145.
8 0 138 0
139.
1 0

170.
93 0
193.
04 0
199.
19 0
185.
05 0
Am
moni
a
m
g
0.5
2.28
-
2 4
-
2 2.62
-
2 2.75
-
2

0.31 0 0.89
-
2 0.25 0 0.24 0

1.03
-
6 2.09
-
6 0.98
-
6 1.00
-
6
Iron m
g
/l
0.3
12.6
0
-
2
19.2
4
-
2 0.95
-
2 0.46
-
2

1.17
-
2 3.87
-
2 0.03 0 0.03 0

5.51
-
6 9.88
-
6 0.22 0 0.14 0
Man
gann
ese
m
g
/l
0.1
6.80
-
2 4.03
-
2 0.38
-
2 0.29
-
2

0.09 0 0.12
-
2 0.03 0 0.01 0

1.53
-
6 1.04
-
6 0.15
-
6 0.10 0
BOD m
g
/l
0.1
60
-
2 60
-
2 42
-
2 35
-
2

8
-
2 7
-
2 12
-
2 14
-
2

18.8
6
-
6 18.5
-
6
23.1
4
-
6
21.6
4
-
6
COD m
g
/l
20
78.2
-
2 78.2
-
2 101
-
2 101
-
2

8.1 0 9 0 20 0 20 0

28.8
4
-
6
29.8
4
-
6
42.7
1
-
6
44.2
1
-
6
SS m
g
100
5180
-
2 5180
-
2 675
-
2 1107
-
2

35 0 19 0 105 0 117 0

1546
-
6
1077
.43
-
6
311.
14
-
6
352.
14
-
6
Total
Coli
m
g
/l
100
00 2010
00
-
3
2010
00
-
3
4200
000
-
3
5600
00
-
3

7000
-
3
7700
0
-
3
2800
00
-
3 1800
-
3

15342
9 -9
17928
6 -9
10200
00 -9
17482
9
-
9
Total
-
7
2
-
7
6
-
6
4
-
5
8

Appendices 2. Calculation on Storet Method (2004)
Paramete
r Unit
PP
82/

Raw Water Quality

2001 Buaran
Pulo
Gadung

Pejomponga
n 1

Pejompongan 2

Quali
ty
V
al
u
e
Quali
ty
V
al
u
e
Quali
ty
v
al
u
e
Quali
ty
Valu
e
Turbidity NTU 100
1532
6
-
1 9242
-
1 6930
-
1 5730 -1

70
-
1 138
-
1 116
-
1 101 -1

5140.1
9
-
3
3830.7
1
-
3
3043.1
4
-
3 2771 -3
TDS mg/l 500
312 0 208 0 292 0 256 0

103 0
121.
4 0 107 0 122 0

157.
82 0
156.
46 0
184.
79 0
179.
2 0
Ammonia mg 0.5
1.21
-
2 1.75
-
2 1.67
-
2 1.21 -2

0.34 0 0.21 0 0.36 0 0.26 0

0.73
-
6 0.98
-
6 0.99
-
6 0.79 -6
Iron mg/l 0.3
12.9
59
-
2
29.7
6
-
2 0.16 0 0.38 -2

0.93
1
-
2 2.27
-
2 0.02 0 0.02 0

7.97
-
6
14.4
3
-
6 0.08 0 0.12 0
Mangannes
e
mg/l 0.1
2.77
-
2 5.6
-
2 0.90
-
2 1.48 -2

0.15
4
-
2 0.21
-
2 0.17
-
2 0.01 0

0.85
-
6 1.60
-
6 0.43
-
6 0.45 -6
BOD mg/l 0.1
18
-
2 14
-
2 34
-
2 24 -2

8
-
2 8
-
2 12
-
2 12 -2

11.9
3
-
6
11.5
7
-
6
20.0
7
-
6
17.5
7 -6
COD mg/l 20
34
-
2 160
-
2 82
-
2 96 -2

9.6 0 16.5 0 16 0 18 0

26.7
-
6
55.7
9
-
6
38.7
1
-
6
36.1
4 -6
SS mg 100
5290
-
2 6210
-
2 2834
-
2 2378 -2

26 0 40 0 109
-
2 113 -2

1873.6
1
-
6
1437
.5
-
6
738.
29
-
6
622.
64 -6
Total Coli mg/l 1000
0
2010
00
-
3
2010
00
-
3
4800
000
-
3
5000
00 -3

1550
00
-
3
2010
00
-
3
1300
0
-
3 100 -3

194428.
57
-9
20100
0 -9
13635
71 -9
10308
0 -9
Total
-
7
4
-
7
4
-
6
6 -66

ISSN: 1411-1284
Proceeding 10
th

Appendix 3. Calculation on Storet Method (2005)
Para
meter
Uni
t
PP
82
/

Raw Water Quality

200
1 Buaran Pulo Gadung Pejompongan 1 Pejompongan 2
Quality
Va
lue Quality
Va
lue Quality
Va
lu
e Quality
Va
lue
Turbidit
y
NTU 10
0
11068 -1 13146 -1 8800 -1 9060 -1

1333 -1 2754 -1 49.1 -1 1504 -1

6750.4
3 -3 7042 -3
4598.8
7 -3 5684.36 -3
TDS mg/
l
50
0
179 0 192 0 207 0 179 0

91 0 121 0 107 0 115 0

128.93 0 154.31 0 153.71 0 149.21 0
Amonia mg 0.5
1.28 -2 2.52 -2 1.28 -2 1 -2

0.48 0 0.3 0 0.21 0 0.3 0

0.84 -6 1.21 -6 0.52 -6 0.69 -6
Besi mg/
l
0.3
9.313 -2 24.53 -2 0.06 0 0.18 0

3.696 -2 5.25 -2 0.02 0 0.02 0

7.36 -6 13.61 -6 0.04 0 0.08 0
Mangan mg/
l
0.1
1.103 -2 2.31 -2 0.79 -2 0.62 -2

0.091 0 0.23 -2 0.17 -2 0.17 -2

0.56 -6 0.73 -6 0.42 -6 0.34 -6
BOD mg/
l
0.1
13 -2 20 -2 42 -2 41 -2

9 -2 8 -2 12 -2 9 -2

11 -6 12.64 -6 21.5 -6 19.86 -6
COD mg/
l
20
106 -2 210 -2 61 -2 52 -2

5.6 0 14 0 20 0 20 0

48.56 -6 89.64 -6 36.43 -6 35.43 -6
SS mg 10
0
4590 -2 14560 -2 1019 -2 1101 -2

71 0 85 0 131 -2 166 -2

1803.5 -6 2815.5 -6 569.29 -6 556.29 -6
Total
Coli
mg/
l
10
00
0 201000 -3 201000 -3
480000
0 -3 390000 -3

201000 -3 62000 -3 440000 -3 260 -3

201000 -9 177857 -9
147642
9 -9
86972.8
6 -9
Jumlah -72 -74
-
66 -66

Appendix 4. Calculation on Pollution Index (2003)
Par
a
me
ter
Un
it
PP
82
20
01
Kualitas Air Baku (Maksimum, Minimum, Rata-rata)
Buaran Pulo Gadung Pejompongan 1 Pejompongan 2
Li Ci Ci/
Li
Ci/
Li
ba
ru
Ci Ci/
Li
Ci/
Li
ba
ru
Ci Ci/
Li
Ci/
Li
ba
ru
Ci Ci/
Li
Ci/
Li
ba
ru
Tur
bidi
ty
NT
U
10
0
10
70
8
10
7.
1
11
.1
48
11
76
8
11
7.
7
11
.3
5
80
89
80
.8
9
10
.5
3
58
08
58.
08
9.
82
42
96
42
.9
6
9.
16
5
39
68
.2
9
39
.6
8
8.
99
29
83
.4
3
29
.8
3
8.
37
26
59
26.
59
8.
12
TD
S
m
g/l
50
0
28
8
0.
58
0.
58
24
1
0.
48
0.
48
33
0
0.
66
0.
66
28
2
0.5
6
0.
56
17
1
0.
34
0.
14
19
3.
04
0.
39
0.
39
19
9.
19
0.
40
0.
4
18
5.
05
0.3
7
0.
37
Am
mo
nia
m
g
0.
5
2.
28
4.
56
4.
29
4 8 5.
5
2.
62
5.
24
4.
59
2.
75
5.5 4.
7
1.
03
2.
06
2.
56
9
2.
09
4.
17
4.
1
0.
98
1.
97
2.
47
1.
00
2.0
1
2.
51
Iro
n
m
g/l
0.
3
12
.6
04
42
.0
1
9.
11
19
.2
4
64
.1
3
10
.0
3
0.
95
3.
17
3.
5
0.
46
1.5
3
1.
92
5.
51
18
.3
7
7.
32
9.
88
32
.9
3
8.
58
0.
22
0.
73
0.
73
0.
14
0.4
5
0.
45
Ma
ng
an
ese
m
g/l
0.
1
7 67
.9
5
10
.1
6
4.
03
40
.3
1
9.
02
0.
38
3.
77
3.
88
0.
29
2.8
9
3.
3
1.
53
15
.3
6.
92
1.
04
10
.4
5
6.
09
0.
15
1.
53
1.
92
0.
10
1.0
2
1.
04
BO
D
m
g/l
0.
1
60 60
0
14
.8
9
60 60
0
14
.8
9
42 42
0
14
.1
1
35 35
0
13
.7
2
19 19
0
12
.3
9
18
.5
18
5
12
.3
3
23
.1
4
23
1.
43
12
.8
2
21
.6
4
21
6.4
3
12
.6
7
CO
D
m
g/l
20 78
.2
3.
91
3.
96
78
.2
3.
91
3.
96
10
1
5.
05
4.
51
10
1
5.0
5
4.
51
28
.8
1.
44
1.
79
29
.8
4
1.
49
1.
86
42
.7
1
2.
14
2.
65
44
.2
1
2.2
1
2.
72
SS m
g
10
0
51
80
51
.8
9.
57
51
80
51
.8
9.
57
67
5
6.
75
5.
14
11
07
11.
07
6.
22
15
46
15
.4
6
6.
94
10
77
.4
3
10
.7
7
6.
16
31
1.
14
3.
11
3.
46
35
2.
14
3.5
2
3.
73
Tot
al
Col
i
m
g/l
10
00
0
20
10
00
20
.1
7.
51
20
10
00
20
.1
7.
51
42
00
00
0
42
0
14
.1
1
56
00
00
0.0
03
0.
00
3
15
34
29
15
.3
4
6.
92
17
92
86
17
.9
3
7.
26
10
20
00
0
10
2
11
.0
4
17
48
29
17.
48
7.
21
Ma
ksi
mu
m
14
.8
9
14
.8
9
14
.1
1
13
.7
2
Su
m
7.
38
7.
52
6.
26
5.
12
PIy 11
.7
5
11
.7
9
10
.9
1
10
.3
5

ISSN: 1411-1284
Proceeding 10
th


Para
me
Ter
Unit PP
82
20
01
Kualitas Air Baku (Maksimum, Minimum, Rata-rata)
Li Ci Ci/L
i
Ci/
Li
bar
u
Ci Ci/
Li
Ci/
Li
bar
u
Ci Ci/L
i
Ci/L
i
bar
u
Ci Ci/L
i
Ci/L
i
bar
u
Turbi
dity
NTU 10
0
153
26
153
.3
11.
92
924
2
92.
42
10.
82
693
0
69.
3
10.
2
573
0
57.
3
9.7
9
514
0
51.
4
9.5
5
383
0.7
1
38.
31
8.9
1
304
3.1
4
30.
43
8.4
1
277
1
27.
71
8.2
1
TDS mg/l 50
0
312 0.6
2
0.6
2
208 0.4
2
0,4
2
292 0.5
8
0.5
8
256 0.5
1
0.5
1
158 0.3
2
0.3
2
156
.46
0.3
1
0.3
1
184
.79
0.3
7
0.3
7
179
.2
0.3
6
0.3
6
Amm
onia
mg 0.5 1.2
1
0,0
2
0.0
2
1.7
5
3.5 3.7
2
1.6
7
3.3
4
3.6
1
1.2
1
2.4
2
2.9
1
0.7
3
1,4
6
1.8
2
0.9
8
1.9
5
2.4
5
0.9
9
1.9
8
2.4
8
0.7
9
1.5
9
2
Iron mg/l 0.3 13 43.
20
9.1
7
29.
76
99.
19
10.
98
0.1
6
0.5
3
0.5
3
0.3
8
1.2
7
1.5
2
7.9
7
26.
57
8.1
2
14.
43
48.
09
9.4
1
0.0
8
0.2
5
0.2
5
0.1
2
0.4
1
0.4
1
Man
gane
se
mg/l 0.1 2.7
7
27.
7
8.2
1
5.6 56 9.7
4
0.9
0
8.9
6
5.7
6
1.4
8
14.
8
6.8
5
0.8
5
8.5 5.6
4
1.6
0
16.
00
7.0
2
0.4
3
4.2
7
4.1
5
0.4
5
4.4
6
4.2
4
BOD mg/l 0.1 18 180 12.
27
14 140 11.
7
34 340 13.
65
24 240 12.
9
12 120 11.
39
11.
57
115
.7
11.
31
20.
07
200
.7
12.
5
17.
57
175
.71
12.
22
COD mg/l 20 34 1.7 2.1
5
160 8 5.5
1
82 4.1 4.0
6
96 4.8 4.4
16.
7
0.8
4
0.8
4
55.
79
2.7
9
3.2
2
38.
71
1.9
4
2.4
3
36.
14
1.8
1
2.2
9
SS mg 10
0
529
0
52.
9
9.6
1
621
0
62.
1
9.9
6
283
4
28.
34
8.2
6
237
8
23.
78
7.8
8
187
4
18.
74
7.3
6
143
7.5
14.
38
6.7
8
738
.29
7.3
8
5.3
4
622
.64
6.2
3
4.9
7
Total
Coli
mg/l 10
00
0
2E+
06
201 12.
5
201
000
20.
1
7.5
1
480
000
0
480 14,
4
500
000
50 9.4
9
194
429
19.
44
7.4
4
201
000
20.
1
7.5
1
136
357
1
136
.4
11.
67
103
080
10.
31
6.0
6
Maks
imu
m
12.
5
11.
7
13.
65
12.
9
Sum 6.9
2
7.7
0
5.9
9
5.7
8
PIy 10.
1
9.9 10.
54
9.9
9

ISSN: 1411-1284
Proceeding 10
th


Parame
ter
Satuan PP 82
2001
Kualitas Air Baku (Maksimum, Rata-rata)
Li Ci Ci/Li Ci/Li
baru
Ci Ci/
Li
Ci/
Li
ba
ru
Ci Ci/
Li
Ci/
Li
ba
ru
Ci Ci/L
i
Ci/
Li
ba
ru
Turbidity NTU 100 11068 110.7 11.22 131
46
131
.5
11.
59
880
0
88 10.
72
906
0
90.
6
10.
78
6750 67.5 10.14 704
2
70.
42
10.
23
459
8.8
7
45.
99
9.3
1
568
4.4
56.
84
9.7
7
TDS mg/l 500 179 0.36 0.36 192 0.3
8
0.3
8
207 0.4
1
0.4
1
179 0.3
6
0.3
6
128.9 0.26 0.26 154
.31
0.3
1
0.3
1
153
.71
0.3
1
0.3
1
149
.21
0.3
0
0.3
Ammonia mg 0.5 1.28 2.56 3.04 2.5
2
5.0
4
4.5
1
1.2
8
2.5
6
3.0
4
1 2 2.5
0.84 1.68 1.13 1.2
1
2.4
2
2.9
2
0.5
2
1.0
4
1.0
8
0.6
921
1.3
8
1.6
9
Iron mg/l 0.3 9 306.7 13.43 24.
53
81.
76
10.
56
0.0
6
0.2 0.2 0.1
8
306
.67
13.
42
7.36 24.53 7.94 13.
61
45.
38
9.2
8
0.0
4
0.1
2
0.1
2
0.0
785
0.2
6
0.2
6
Mangan mg/l 0.1 1 11.03 6.21 2.3
1
23.
13
7.8
2
0.7
9
7.8
5
5.4
7
0.6
2
6.2 4.9
6
0.56 5.6 4.74 0.7
3
7.3
0
5.3
1
0.4
2
4.1
6
4.0
9
0.3
419
3.4
2
3.6
7
BOD mg/l 0.1 13 130 11.57 20 200 12.
5
42 420 14.
11
41 410 14.
06
11 110 11.2 12.
64
126
.43
11.
5
21.
5
215 12.
66
19.
857
215 12.
66
COD mg/l 20 106 5.3 4.62 210 10.
5
6.1 61 3.0
5
3.4
2
52 3.0
5
3.4
2
48.6 2.43 2.92 89.
64
4.4
8
4.2
5
36.
43
1.8
2
2.3 35.
429
1.8
2
2.3
SS mg 100 4590 45.9 9.3 145
60
145
.6
11.
81
101
9
10.
19
6 110
1
10.
19
6
1803 18.03 6.28 281
5.5
28.
16
8.2
5
569
.29
5.6
9
4.7
7
556
.29
5.6
9
4.7
7
Total Coli mg/l 10000 201000 20.1
7.51
201
000
20.
1
7.5
2
480
000
0
480 14.
4
390
000
480 14.
4
201000 20.1 7.51 177
857
17.
79
7.2
5
147
642
9
147
.6
11.
84
869
73
147
.64
11.
85
Maksimum 13.43 12.
5
14.
4
14.
4
Sum 6.99 7.6
1
6.2
4
6.9
2
PIy 10.7 10.
34
11.
09
11,
03

Non Linear Dynamic Identification of Waste Heat
Boiler Unit Based on
Adaptive Neuro Fuzzy Inference System (ANFIS)

Yuliati

Electrical Engineering Department, Widya Mandala Surabaya Catholic University.
Jl. Kalijudan 37 Surabaya,
Phone: +62-31-3891264, Email : yuliatheresia@yahoo.com

Abstract- Waste Heat Boiler (WHB) unit is a
crucial plant for a fertilizer plant especially for
steam product .The water level in the its steam
drum shall be maintained at certain limit in order
to give a good steam quality. It is generally
accepted that a boiler unit is a highly nonlinear
and strongly coupled complex system. Due to the
dynamic characteristic, it is difficult to have its
model in order to give suitable control system of
the steam drum water level theoretically.
The success of model-based non linear control
technique is usually conditioned by the
availability of accurate models which reflects the
non linear process complexities. Prior to
implementing any control scheme, its dynamic
properties need to be identified properly. A real
time data driven neuro fuzzy model will be
presented for the water level of boiler unit. Neuro
fuzzy design combine architectural of neural
network and philosophical of fuzzy system aspects
of an expert resulting in an artificial brain, which
can be used as an identifier.
An experiment for identifying the water level
of boiler unit was conducted on real-time plant
operation of a fertilizer plant in Gresik, East
Java, Indonesia. Validation model was also
performed to verify the results of identification.
The performance of the procedure shows how the
neuro fuzzy algorithm produces accurate models
for non linear control design.

Keywords : system identification, waste heat boiler,
real plant, fuzzy logic, neural network.

I. Introduction

There are two types of boilers in the steam
generation i.e. fired boilers and unfired boilers (heat
recovery steam generator) is in fact a heat exchanger
extracting heat from through put gases and raising
water to steam condition. Waste heat boiler unit
representing type of unfired boiler system which
using heat transfer energy form gas turbine exhaust
gas flow. Its about 60% of 70% of the heat that
obtained by combustion in gas turbine, re-exploited
by WHB to produce the steam.
Its generally accepted that a boiler unit is a
highly non linear and strongly coupled complex
system. Its dynamic characteristic makes it very
difficult to model and control the water level of
steam drum theoretically. The water level of High
pressure steam drum in WHB unit should be
maintained within safe limit. A too high of water
level produces wet steam that could damage the
turbine equipment and the turbine can trip off. On the
other hand, too low of water level risks uncovering
the water tubes and exposing them to heat stress and
damage. In both cases, the plant shuts down
unintentionally.
The success of model-based non linear control
technique is usually conditioned by the availability of
accurate models which reflects the non linear process
complexities. Prior to implementing any control
scheme, its dynamic properties need to be identified
properly.
Neural network have been applied successfully
in the identification and control of non linear
dynamic system. The universal approximation
capabilities of multilayer perceptron (MLP) make it
popular choice for modelling non linear systems.
However, neural network architecture is rather
complicated and difficult to be understood by human.
On the other hand, the fuzzy inference system can
initialise and learn linguistic and semi linguistic
(Sugeno) rules therefore it can be considered as
direct transfer knowledge which is the main
advantage of fuzzy inference systems but the
learning capabilities of fuzzy system is not as good
as neural network. It is then to develop to integrate
the strength of both methodologies in order to
achieve the learning capabilities and transfer
knowledge via fuzzy if-then rules, producing an
alternative approach so called neuro fuzzy.
In this paper, a development of non linear
identification for high pressure drum (HP drum) of
boiler unit will be design using neuro fuzzy

approach which the fuzzy inference system can be
tuned with a neural network algorithm during
learning based on collection of input output data
obtained from real-time experiment. By the using
real plant, it is expected that the model obtained
could capture the basic feature of water level of
steam drum system and this technique could also
improve the modelling quality of the system.

II. Waste Heat Boiler Plant

Waste Heat Boiler (WHB) unit plays a crucial
role in a fertilizer plant as a part which produces
steam. The temperature of the exhaust gas from gas
turbine is about 600
0
C used to produce superheated
steam which has 65 kg/cm2 pressure and temperature
460
0
C. Boiler Feed Water (BFW) is feed into High
Pressure Steam Drum using 2 pumps which standby
interchange each other. Control valve is used to
control flow of BFW; the level of HP steam drum is
controlled by 2 types of drum level control i.e. single
element and three element control. If the steam
product is smaller then 30 ton/hr, single element
control used to control level drum. In the other hand,
if the steam product is greater then 30 ton/hr, three
elements drum level control used to control level
drum. It combines between level drum, steam flow
and feed water flow.
WHB unit in PT. Petrokimia Gresik is an unfired
boiler completed by some sub system (Figure 1) i.e.
Super heater, high pressure (HP) evaporator, low
pressure (LP) evaporator, economizer, low header,
HP/LP down comers, HP/LP drum, duct burner, and
firing room.

Figure 1. WHB sub system

III. The Experimental Set-up

The experiment set-up is referred to Piping &
Instrumentation Diagram (P&ID) in PT. Petrokimia
Gresik. There are several conditions that must be
considered to design this experiment as follow:
a. Pressure and temperature of BFW,
Pressure and temperature of HP drum,
steam product temperature and blow down
were kept constant as much as possible.
b. Set point flow steam product is based on
the need of steam in urea plant.
c. Data have been collected in real time from
the existing Distributed Control System
(DCS) with sampling time 1 minute.
d. Each 3000 data for every corresponding
inputs and outputs have been collected. The
first 500 data will be used for
identification, and the 1500 remaining data
for model validation.
e. Due to high risk of its plant operation, the
experiment has been conducted by closed
loop manner without change the real plant
operation.
The record of input output data during the conducted
experiment can be shown in Figure 2 and will be
used for modelling and validation purpose.

(a) Flow BFW

(b) WHB Steam Flow

(c) Level WHB HP Drum

Figure 2. Representation of Input-Output Data

During identification process, it can be defined as
shown in Table 1 for input-output variable for
modelling purpose.

Table 1. Input-Output Variable for Modelling
Variable Tag
number
Description
) 1 (
1
k x
FC.22220 Flow BFW (Ton/hr)
) 1 (
2
k x
FC.22221 Steam Product Flow
(Ton/hr)
) 1 ( k y
LC.22220 Level water in HP steam
drum at sampling instant
k-1 as an input model (%)
) (k y
LC.22220 Level water in HP steam
drum at sampling instant k
as an output model (%)

III. Neuro Fuzzy Dynamic Identification

3.1 ANFIS Structure

ANFIS implements a Takagi- Sugeno Fuzzy
Models have to be a powerful tool for the non linear
dynamic identification system. ANFIS scheme as an
estimator model can be shown in Figure 3.

Figure 3. ANFIS as an Estimator Model

The first order of ANFIS structure with three
inputs and one output can be described by the
following rule base:
Rule i
th
: if ) 1 (
1
k x is A
i
and ) 1 (
2
k x is B
i
and
) 1 ( k y is
i
C then
i i i i i
v k y r k x q k x p f + + + = ) 1 ( ) 1 ( ) 1 (
2 1
,
m i ,..., 2 , 1 = (1)
Where m denotes the number of rules. In the other
hand, the non linear AutoRegressive with eXogenous
input (NARX) model is frequently used for non
linear identification techniques. Therefore, in order
to represent the neuro fuzzy for Multi Input Single
Output (MISO) approach can be defined as
)) 1 ( ),..., ( ), 1 ( ),..., ( ( ) 1 ( + + = +
b a m
n k u k u n k y k y f k y

(2)
The input vector x , of the neuro fuzzy model as:

)] 1 ( ),..., ( ), 1 ( ),..., ( [ + + =
b a
n k u k u n k y k y x
(3)
Where ) 1 ( ),..., ( +
a
n k y k y and
) 1 ( ),..., ( +
b
n k u k u are tapped delay lag (TDL)
process output and input respectively. ANFIS has
five layered architecture which has three inputs and
one output can be shown in Figure 4.

Figure 4. Structure of ANFIS

The architecture employs an adaptive network
consist of adaptive nodes in the 1
st
and 4
th
layer. In
the 1
st

is for fuzzification of the input variable and T-
norm operator are deployed in the 2
nd
layer to
compute the rule antecedent part (premise
parameter). The 3
rd
layer normalizes the rule
strengths followed by 4
th
layer where the consequent
parameters are determined. The output network
computes the overall input as the summation of all
incoming signal. The description and computation in
every layer can be described as shown in Table 2.

Table 2. The Description and Computation of Each
Layer in ANFIS Method
Layer Description
1

Every nodes i is an adaptive node with a
node output :
) (
, 1
x O
Ai i
u = , and ) (
2 , 1
y O
Bi i
= u
The output of this layer is the
membership values of the premise part.
The generalized bell is specified by
three parameters {a,b,c}:
i
b
i
i
Ai
a
c x
x
|
|
\
|
+
=
2
1
1
) ( u
has been chosen as membership function
2

Fixed node labelled which multiplies
the incoming signal and output product
represents the firing strength of a rule:
) ( ) (
, 2
y x w O
Bi Ai i i
u u = = , i=1,2
3

Fixed node labelled N which
normalizes the rule strengths
2 1
, 3
w w
w
w O
i
i i
+
= = , i=1,2

4 Every node i is an adaptive node with a
node function :
( )
i i i i i i i
r y q x p w f w O + + = =
, 4
, i= 1,2
Referred to as consequent parameters.
5 This single node is a fixed node which
computes the overall output as the
summation of all incoming signal:
= = =
i
i
i
i
i i
i i i
w
f w
f w f O
5 ,

3.2 Hybrid Learning Algorithm

The hybrid learning algorithm has two passes for
ANFIS i.e. forward pass and backward pass. Suppose
that S1 is a set of non linear parameters and S2 is a
set of linear parameters in the forward pass, node
outputs go forward until 4 layer and the consequent
parameters are identified by the least square
estimation. In the backward pass, the error signals
propagate backward and then premise parameters are
update by gradient descent.
a. Forward pass
If the values of the premise parameters are fixed,
then the overall output can be expressed as a linear
combination of the consequent parameters, the output
f can rewritten as
2
2 1
2
1
2 1
1
f
w w
w
f
w w
w
f
+
+
+
=
( ) ( )
2 2 2 2 1 1 1 1
r y q x p w r y q x p w + + + + + =
( ) ( ) ( ) ( )
2 2 1 1 1 1 1 1
p x w r w q y w p x w + + + =
( ) ( )r w q y w
2 2 2
+ + (4)
From the above equation, it can be seen that the
consequent parameters are linear respect to the
system output. If N learning data is applied to the
equation (4), it can be represented by
y A =
Where unknown parameter in S2 and y is is N
vector data. The problem is how to find the best
solution of this equation that minimize
2
y A ,
is the least square estimator
( ) y A A A
T T
1
*

= (5)

b. Backward pass
In the backward pass the error signal
propagate from the output end toward the input end
using the simple gradient steepest descent for each
training data entry. By minimizing a cost function
defined by
=
=
T
k
k e E
1
2
) (
2
1
(6)
With ) ( ) ( ) ( k y k y k e = (7)
is an error function at time instant k, with y(k) is the
desired response.
For updating the changes of each parameter, the
gradients are needed to minimize E Using the
relation:
) (
) (
) 1 (
) (
k
k E
k k
L
i
L
i
L
i

+ = (8)
Where is the learning rate and
) (
) (
k
k E
L
i

denotes
the order derivative of the error signals (difference
between an actual output and desired output vector)
respect to the node parameter .

IV. The Results of System Identification

Identification process has been conducted using
Multi Input Single Output (MISO) approach as
shown in Figure 4. For learning process and
modelling purpose, assuming that the plant under
consideration can be represented as a first order
process, then the regressor /identifier have been
chosen as
x= )] 1 ( ), 1 ( ), 1 ( [
2 1
k y k x k x (9)
And after 50 epochs for learning process, the ANFIS
structure model can be obtained as the following: the
number of nodes: 34, the number of linear
parameters: 32, the number of nonlinear parameters:
18, the number of training data pairs: 500, and the
number of fuzzy rules: 8.
The result of system identification and its
corresponding errors of learning process can be
shown in Figure 5 and 6, respectively.

Figure 5. The result of System Identification

Figure 6. Error

It can be seen from these figure that the output
of the identified model can follow the measurement
signals satisfactorily and the error (the difference
between these two signals) are relatively small
enough.
V. Model Validation

Beside a direct observation by comparing the
plot of output measurements and the output model
based on the estimated parameter, quantitative
criteria of the Root Mean Square Error (RMSE) can
be used for validation model for a measure the
goodness of fit the model. It can be defined as
RMSE = | |
=

N
t
k y k y
n
1
2
) ( ) (
1
(10)
If the RMSE 5 . 0 then the developed model can be
accepted as a good approach to represent the model
system. The 1500 remaining data was used for
validation purpose, and the result of model validation
and error of validation model can be shown in Figure
7 and 8, respectively.

Figure 7. The Result of Model Validation

Figure 8. Error of the Model Validation

The RMSE both of the identification process and the
model validation are relatively small enough and it
can be seen in Table 3.

Table 3. Root Mean Square Error
Data from :

Identification Validation
RMSE 0.0324 0.0640

VI. Conclusion

An alternative approach to identify level water
of high pressure steam drum in WHB unit using
neuro fuzzy technique has been proposed. By
comparing between the response models with the
actual output measurements revealed that the
satisfactory model matching were obtained. It means
that the non-linear characteristics of system can be
capture from the real basic feature of the WHB
system. The performance of the procedure shows
how the neuro fuzzy algorithm produces accurate
models for non linear control design. Further
application for level control system design based on
identified model is currently under study.

VII. References

[1]. Johanson, R, System Modelling and Identification,
Prentice-Hall International, Inc, NJ, USA, 1993.
[2]. Jyh-Shing Roger Jang, Chuen-Tsai Sun, Neuro Fuzzy
Modelling and Control, The proceeding of the 1995
IEEE, International Conference on Control
Application, 1995.
[3]. J.S.R. Jang, C.T. Sun, E. Mizutani, Neuro Fuzzy and
Soft Computing: A Computational Approach to
Learning and Machine Intelligent, Prentice-Hall,
International Edition, 1997.
[4]. Narendra,K.S, Parthasarathy,K, Identification and
Control of Dynamical Sistems Using Neural Network,
IEEE Transaction on Neural Network, Vol.1, No.1,
March 1990.
[5]. A.W. Ordys, A.W. Pike, M.A. Johnson, R.M. Katebi,
and M.J. Grimble, Modelling and Simulation of
Power Generation Plants, Springer-Verlag, 1994.
[6]. Standard Fasel Lantjes, Waste Heat Boiler, PT.
Petrokimia Gresik, 1992.
[7]. Xiaoguang Chang, A Recurrent Fuzzy- Neuro Dynamic
Model for a Stoked-Fired Boiler, Laboratory of
Intelligent Technology and Systems Tsinghua
University, China, May 1999.
[8]., .Procidia
TM
Control Solutions boiler
Control Overview, Application Data by SIEMENS,
March 2006.
[9]. Yul Y.Nazaruddin, Joko W, S. Hadisupadmo, Inverse
Learning Control using Neuro Fuzzy Approach for a
Process Mini Plant, International Conference on
Physics and Control, Rusia, August 2003.

ISSN: 1411-1284
Proceeding 10
th

LOAD PATTERN AND REALIBILITY ASSESSMENT OF
POWER GENERATION SYSTEM

Suparman, Zuhal, and Rinaldy D.

Effect of Harmonic Current

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Effect of Harmonic Current

Uploaded by

Copyright:

Available Formats

i

for the ABR gauge in Christoffel symbol

Engineering Faculty University of Indonesia Kampus UI Baru 16424

[13]. Finally, from the rate

atom and the hydrogen atom, the

atom are the main chain

You might also like