Journal of Environmental Treatment Techniques

2019, Volume 7, Issue 4, Pages: 689-695

J. Environ. Treat.

Tech. ISSN: 2309-1185

Journal web link: http://www.jett.dormaj.com

Modelling and Optimization of Biomass Supply

Chain for Bioenergy Production

Siti Fatihah Salleh , Mohd Fadzil Gunawan , Mohd Fikri Bin Zulkarnain , Abdul Halim

Shamsuddin , Tuan Ab Rashid Tuan Abdullah

Institute of Energy Policy and Research (IEPRe), Universiti Tenaga Nasional, Putrajaya Campus, Jalan Ikram-UNITEN, 43000 Kajang, Selangor,

Malaysia

Department of Chemical Engineering, Universiti Sains Malaysia Engineering Campus, 14300 Nibong Tebal, Pulau Pinang

Institute of Sustainable Energy (ISE), Universiti Tenaga Nasional, Putrajaya Campus, Jalan Ikram-UNITEN, 43000 Kajang, Selangor, Malaysia

Received: 01/05/2019

Accepted: 08/07/2019

Published: 03/09/2019

Abstract

Biomass utilization in the generation of bioenergy and biofuels is promoted as a sustainable solution to decarbonise the energy

sector. Effective management of biomass supply chain is pivotal to successful deployment of this energy-rich resource in the power

sector. Using a case study in Malaysia where biomass feedstock sourcing problem is prevalent, a biomass supply chain system was

proposed. Three important terminals were identified, namely; i) biomass supply point (BSP), where raw biomass materials are

collected, ii) biomass processing facility (BPF), as the collection hub cum centralized conversion centre, and iii) biomass demand

centre (BDC), where the processed biomass meets the end-use sectors. A new modelling approach was proposed to optimize the new

BPF location with the objective to minimize total road travelling distance between each terminal. The results were compared with the

other commonly used methods, namely centre of gravity and fuzzy clustering. The positive contributions of the proposed model in

minimizing the overall logistic cost and CO emissions due to fuel consumption were discussed.

Keywords: Biomass, Palm Oil Mills, Bioenergy, Decarbonization, Supply Chain.

03 million tons of biomass, including agricultural waste,

Introduction

forest residues and municipal waste (2). The agricultural

waste represents 91% of the biomass amount, of which over

Biomass is an energy resource derived from living matter

such as field’s crops and trees. Agricultural, forestry wastes

and municipal solid wastes are also considered as biomass.

Being in the tropical sun-belt, Southeast Asian countries

enjoy a year-round sunlight which provides a favourable

condition for agricultural activities. Global innovations in

renewable energy technology and growing awareness on

sustainable agricultural practice have prompted many

countries in this region to adopt low carbon, bio-based energy

production from their agricultural wastes. Among the

common biomass feedstocks are bagasse, palm oil mill waste,

and paddy husk.

7% is derived from palm oil mill residues.

The land area used for oil palm plantation in Malaysia

keeps growing, covering 5.8 million hectares of land in 2017

3). The total amount of processed fresh fruit bunches (FFB)

(

was about 101 million metric ton. After a series of processes

which involves the removal of the oil fruits from the branches

and oil extraction, about 72% of the FFB mass is left as

biomass residues in the form of empty fruit bunches (EFB),

mesocarp fibres (MF), palm kernel shells (PKS), and also

palm oil mill effluents (POME).

These energy-rich biomass resources can be used to

generate electricity, or converted to other marketable

products such as bio-based chemicals, biofuels, animal feed,

wood products and pellets. The estimated electricity

generation potential of palm oil mill biomass comprised of

EFB, PMF and PKS is between 2,400 – 7,460 MW (2, 4),

while it is between 410 - 483 MW for biogas from POME (2,

In 2016, bioenergy power generation in this region tripled

from 1.6 GW in 2000, to 7.2 GW (1). This promising

progress is mostly contributed by Indonesia, Malaysia and

Thailand who are experiencing rapid urbanization and

industrialization. Malaysia for example, produces more than

) (considering 7,200 operation hours of power plant).

In view of its high availability, biomass could play a

Corresponding author: Siti Fatihah Salleh, Institute of Energy

Policy and Research (IEPRe), Universiti Tenaga Nasional,

central role in gearing up the share of renewable energy

power generation. It can act as the base load for the national

or regional grid, slowly taking over the role of coal power

Putrajaya

Campus, Jalan Ikram-UNITEN, 43000 Kajang,

Selangor, Malaysia; E-mail: siti.fatihah@uniten.edu.my and

sitifatihah.salleh@gmail.com.

Journal of Environmental Treatment Techniques

2019, Volume 7, Issue 4, Pages: 689-695

plants while decarbonizing the energy industry along the

way. In order to capture this potential, a cost-effective

biomass supply chain management is crucial. Establishing a

centralized biomass collection center and conversion facility

could be an essential component in the supply chain to reduce

the processing cost and logistic cost.

Many tried to resolve problems mostly related to the triple

bottom line of sustainability; namely i) economy, through

minimization of cost associated with transportation, storage

and inventory, redistribution and disaster losses (6), ii)

environment, through minimization of air pollution such as

CO emission (7), and iii) social, through maximization of

social benefits such as job opportunities (8).

.1 Biomass Supply Chain System

Effective management of biomass supply chain is

The center of gravity (COG) method is a well-established

tool to find the geographical location a new facility which

have to transport goods between the demand and supply

points (9). This method ensures that the optimal location

coordinates would minimize the required weighted travelling

distances among them. In some cases, the method has to be

modified such as implemented by Almetova et al. (2016) to

cater unevenness in logistic flows in cargo network (10).

Furthermore, in order to a solve more complex, multi-

objective, multi-echelon facility location problem, mixed-

integer linear programming model is most commonly adopted

critically important to ensure bioenergy project viability and

economic feasibility. Three important terminals along the

supply chain have been identified, namely; i) biomass supply

point (BSP), ii) biomass processing facility (BPF), and iii)

biomass demand centre (BDC) or the market as shown in

Figure 1.

(

11–14). Some used fuzzy theory to manage situation of

uncertain parameters (11, 15), while others used Fuzzy C

Means (FCM) clustering algorithm as a complementary tool

to assign geographical clustering of distribution points (16,

7) or to fine tune the facility location optimization result

Figure 1: Supply chain in bioenergy production system.

(

18). In FCM method, an objective function called C-means

functional is minimized through

clustering algorithm (19).

a fuzzified k-means

Generally, BSP is where all the biomass is produced and

will be collected from, such as the palm oil mills. Meanwhile,

BPF serves as a collection hub and conversion facility. This

is where the collected biomass will be stored and processed to

improve its quality and durability before shipping. Finally,

the processed biomass will be transported to BDC, where

biomass is consumed by end-use sectors whether industrial,

transport or residential. For industrial, it could be a power

plant or a biorefinery. BDC could also function as a hub for

national or international trade of biomass since there is a high

demand for biomass pellet from South Korea, Japan & China,

which may reach up to 16 million ton by 2020. International

import and export of biomass will further promote

diversification of its product portfolio and market

competitiveness.

This study is a preliminary work to optimize BPF

location provided that BSP and BDC locations are known and

fixed, using a new modelling approach, which is the

minimized least-square regression (LSR) method. This

model enables simultaneous minimization of the road

transportation distance between each terminal, with

consideration of the number of trips required to transport all

the biomass. The result will then be compared with the

conventional COG and FCM methods. An empirical analysis

was implemented via a case study in Perak, Malaysia, based

on the current palm oil mill biomass availability data and

existing road network.

Methodology

.1 Facility Location Optimization Techniques

.2 New Facility Location Optimization

Three methods were used to determine the suitable

The economic viability of bioenergy production system

location of BPF that reduces the overall logistic cost, which

are; i) Centre of gravity (COG), ii) Fuzzy clustering or Fuzzy

C Means (FCM) and iii) Least-square regression (LSR). The

objective function is minimization of the total travelling

distance between the terminals, on the premise that

minimizing it will result in minimized total logistic cost and

depends heavily on the cost of biomass feedstock supplies.

Biomass sourced from agricultural industry in general is not

expensive, the purchase price for raw EFB biomass for

example, could be as low as 1.75 USD/tonne. However, the

logistic cost to transport it from each BSP to BDC could be

large and take a heavy toll on total biomass procurement cost.

In a life cycle assessment of EFB consumption for green

chemical production in Malaysia, Reeb et al. (2014)

discovered that transportation cost caused substantial

financial burdens, responsible for about 61% of the total

delivered cost.

This is where BPF should come into play, by serving as a

collection cum redistribution hub, hence ensuring efficient

logistic chain and matching of demand and supply. The

optimization of BPF location in between BSP and BDC is

therefore critically important to minimize the logistic cost.

Various mathematical approaches have been applied by past

researchers to determine the strategic location of new facility.

CO emissions.

Unlike FCM which is suitable for multiple facility

location problem, a geographical boundary needs to be pre-

assigned prior to COG and LSR analysis. Three decision

factors were considered: (a) state boundaries, (b) state/federal

road transportation network, (c) biomass processing capacity.

The distance from each BSP to BPF was set to be not more

than 100 km. In addition, the biomass supply was assumed to

always match the demand.

Journal of Environmental Treatment Techniques

2019, Volume 7, Issue 4, Pages: 689-695

.1.1 Centre of Gravity (COG)

The centre of gravity (COG) method indicates the ideal

Compared to COG method, this model enables

simultaneous minimization of the road transportation

distances between each terminal, with consideration of the

number of trips required as the weightage. The model was

solved using the Solver toolbox of Microsoft Excel by

minimizing the set objective (ꢖ_ꢝ_ꢞ_ꢝ_ꢟ_ꢠ) and setting the variable

cells as coordinates of BPF.

location in the grid-map that would ensure minimum

weighted travelling distances. In biomass supply chain, raw

biomass from each BSP will be transported to a new

collection hub, which is the BPF. Therefore, the geographical

coordinates of the existing i-BSP terminal (x , y ), as well as

the amount of biomass, w (ton/day) produced by each BSP

were used as the weightage according to the following

formula;

2.3 Mode of Transportation, Road Network, and Travelling

Distance

Assuming that the biomass will be transported in a 20

tons-capacity truck (each supply will be fully loaded with 20

tons of raw biomass or less) the travelling distance will

depend on the land transportation network. In reality, a

highway that directly connects each terminal does not always

exist. Therefore, the distance between two points is not linear,

but depends on the actual existing road network. Given the

coordinates of the potential BPF location optimized by each

model, the road transportation distance was determined using

information provided by Google Maps, then multiplied by the

number of trips required to calculate the total travelling

distance.

∑

^ꢐ^ꢑꢒ ^∑^ꢐ^ꢓꢒ

ꢀ

ꢁꢂꢃꢄꢅꢆꢇꢈꢄꢉꢂ(ꢊꢌꢍ)ꢎꢏ

)

∑ꢐ

ꢋ

∑

ꢐ

where ꢊ_ꢔandꢂꢍ_ꢔare the geographical coordinates of each

BSP, (ꢊꢌꢍ) are the geographical coordinates of BPF and ꢕ

is the amount of biomass (ton/day) transported from BSP to

BPF.

ꢋ

.2.2 Fuzzy Clustering or Fuzzy C Means (FCM)

Fuzzy C Means (FCM) clustering analysis was performed

using a developed program in Fuzzy toolbox of Octave,

which is available from Octave Software (GNU Octave,

https://www.gnu.org/software/octave/). The input and output

data were managed through a Microsoft Excel database.

Given the data set X which includes geographical X and Y

coordinates, the number of clusters 1 < c < N, the process is

described as follows:

.4 Logistic Cost and the Resultant CO Emission

Assuming that the truck travels 40 mile per hour, the

logistic cost was determined as the consolidated costs of three

components which are truck driver wage, truck fuel cost and

truck operation cost as listed below in Table 1;

.Input: keyed in the sample data set (two dimensional)

contains geographical coordinates of x and y of each palm

oil mill;

Table 1: Logistic Cost

Component

Truck Driver Wage

Unit

.Fuzzy C Means algorithm call command – initialized

18.35

2.1

USD/hr

USD/gal

lon

the modelling process, number of clusters of c was given,

which are two,

Fuel Cost

.Output command: prompted the software to show the

Truck Operation Cost

Estimations were made based on general case study on biomass

transportation by 23 ton truck trailer (20) .

200

USD/hr

results

.2.3 Least-Square Regression (LSR) Method

LSR basically calculates a line of best fit to a set of data

In addition, assuming the delivery truck travels 5 miles

per gallon diesel, 2 x 10 metric tons of CO₂would be

emitted per each gallon (21).

pairs, minimizing the sum of squares of the vertical distances

between the data points and the objective function. The

required distance between each terminal may be represented

in the form of following equation:

.5 Case Study Area

In order to locate and estimate the distance of each palm

ꢗ

oil mills, a map was created by using Google Maps software.

Perak, which is one of the major states of oil palm plantations

in Peninsular Malaysia was selected as the region for our case

study. In Perak, there are currently 29 operating palm oil

mills with fresh fruit bunch (FFB) processing capacity of at

least 20 tonne/hour as listed in Table 2, totalling to 1,494 ton

of processed FFB/hour. In this study, the biomass of interest

is EFB, which accounts for 22% of the FFB by weight.

The FFB processing capacity of each identified palm oil

mill was used to estimate the amount of FFB processed daily

using the following equation;

ꢖ ꢎꢏꢊꢘꢙꢊꢚꢛ ꢜꢏꢍꢘꢙꢍꢚꢛ

where ꢖ is the computed distance between each terminal.

Using the Solver toolbox of Microsoft Excel software, the

overall travelling distance, ꢖ_ꢝ_ꢞ_ꢝ_ꢟ_ꢠbetween each terminal

BSP to BPF, and BPF to BDC) were minimized using the

following equation;

(

^ꢖꢝꢞꢝꢟꢠ ꢎꢡꢏꢖ ꢇ ꢛꢜꢏꢖ ꢇ ꢛ

ꢔꢋ ꢔꢋ

ꢋꢢ ꢋꢢ

where ꢖ_ꢔ_ꢋis the total travelling distance required to transport

^bⁱ^o^m^a^s^s^f^r^o^m^B^S^P^t^o^B^P^F^,^ꢖꢋꢢ is the total travelling distance

required to transport processed biomass from BPF to BDC,

ꢇꢄꢉ

ꢀꢣꢄꢅꢤꢥꢥꢤꢦꢂꢁꢁ ꢧ

ꢨꢂ

ꢦꢆꢍ

ꢎꢂꢁꢁ ꢂꢩꢣꢄꢅꢤꢥꢥꢈꢉꢪꢂꢅꢆꢩꢆꢅꢈꢇꢍꢂꢧ

ꢇꢄꢉ

ꢫ

^w^hⁱ^l^e^ꢇꢔꢋ and ꢇ ꢂare the number of trips required to transport

ꢋꢢ

ꢨꢊꢂꢬꢭꢂꢧ

ꢨꢊꢂꢮꢯꢰ

ꢦꢆꢍ

the biomass from each terminal.

Journal of Environmental Treatment Techniques

2019, Volume 7, Issue 4, Pages: 689-695

ꢝꢞꢴ

ꢱꢁ ꢂꢀꢤꢃꢃꢤꢇꢥꢂꢩꢣꢄꢦꢲꢅꢇꢈꢄꢉꢳ

ꢷꢂꢎ

ꢓꢵꢟꢶ

The effective operating hour of each mill was assumed to

be 12 h per day and there will be 0.1% losses of biomass

during the palm oil production process.

ꢝꢞꢴ

ꢀꢣꢄꢅꢤꢥꢥꢤꢦꢂꢁꢁ ꢂꢳ_ꢸ_ꢟ_ꢓꢷꢂꢊꢂꢮꢯꢭꢭꢂꢊꢂꢮꢯꢹꢂꢊꢂꢺꢂꢊꢂꢬꢭ

For power generation, pretreatment of EFB is necessary

to reduce its moisture content as well as to increase its energy

density before firing. Considering that in BPF, the collected

raw EFB will be transformed to EFB pellets which involves

drying and pelleting could be produced daily was estimated

based on the amount of processed FFB, which is 22% per

tonne FFB on wet basis.

In this study, the potential BDC is the existing thermal

power station, which is also the largest operating coal power

plant in Malaysia, Stesen Janakuasa Sultan Azlan Shah

(SJSAS) situated near the . Currently, SJSAS only consumes

pulverized bituminous sub-bituminous coal mostly

imported from Indonesia to fire up its boilers. Coal releases

huge amount of pollutants upon combustion which include

carbon dioxide (CO ), methane (CH ), and nitrous oxide

Table 2: List of palm oil mills and FFB processing

(

N O) and particulate matter. Substituting a fraction of coal

capacity

with biomass could reduce the amount of pollutants

emissions significantly without much alteration needed in the

power generation system.

FFB

Processing

Capacity

No Palm Oil Mill

(

ton/hr)

Results & Discussion

Kilang Kelapa Sawit Lekir

Kilang Sawit Changkat Chermin

Pantai Remis Palm Oil Mill Sdn. Bhd.

KKS United Int. Enterprises (M) Bhd

Kilang Sawit Felcra Nasaruddin

Sri Intan Oil Palm Mill

KKS Peladang & Perusahaan Minyak

Awan Timur Palm Oil Mill

Topaz Emas Sdn Bhd

Temerloh Mill Sdn Bhd

Tian Siang Palm Oil Mill

Selaba Palm Oil Mill

KKS Ganda

Perak Agro Mills Sdn Bhd

KKS TRP

100

120

. 1 BPF Location Optimization

Figure 2 shows the locations of BSPs, BPFs and BDC.

Since the palm oil mills (BSPs) are spread out far away from

each other up to 200 km apart, it is simply not economical to

have only one BPF centre. A single large scale biomass

facility will require longer distances to transport biomass

feedstock from multiple locations, thereby increases the

logistic cost and overall cost of acquiring feedstock. The

distance between each terminal should not be more than 100

km, and preferably less than 30 km because the transportation

cost will be greatly affected by vicinity. Furthermore,

assuming that the scale of each BPF should only be between

00,000- 500,000 metric ton/year, which are the common

large scale pellet plants capacity in the United State, there

should be at least two BPF centres (BPF1 and BPF2) to

process EFB from all identified BSPs in the northern region

and southern region of the Perak state.

KKS Southern Perak

Felcra Processing & Engineering

Kilang Minyak Sawit Tanjung Tualang 40

Gabungan Perusahaan Minyak

Langkap Oil Palm Sdn. Bhd.

KKS Perak Motor Co. Sdn Bhd

SYNN Palm Oil Sdn Bhd

Tian Siang Oil Mill

Central Palm Oil Mill

ST Palm Oil Mill

120

KKS Yee Lee Palm Oil Industries Sdn

Bhd

KKS Tali Ayer (Hilltop Palm Oil)

KKS Chersonese

KKS Trolak

Elphil Palm Oil Mill

It was assumed that during the EFB pelletization process,

only 70% of the mass will remain due to some losses of

moisture content and biomass. Therefore, assuming the

centralized BPF operates 6 days per week in every month, the

amount of EFB pellets produced in a year would be;

Figure 2: BPF location optimization for EFB supply from palm oil

mills to a coal-biomass cofiring power plant.

Journal of Environmental Treatment Techniques

2019, Volume 7, Issue 4, Pages: 689-695

Visually, the locations of BPFs as proposed by the three

methods-COG, FCM and LSR are not far from each other,

just around 5 to 10 km apart (except for the case of LSR-

proposed location of BPF1, which is visibly far from the

locations proposed by the other two models). However, when

the number of trips was taken into account, there is a huge

difference in the travelling distance along the biomass supply

chain, as shown in Table 3. Comparative results show that the

overall travelling distance is significantly minimized by the

LSR method for both BPF1 and BPF2. As a consequence, a

lot of savings can be gained in terms of logistic cost and CO₂

emission due to road transportation as shown in Table 4. By

optimizing the travelling distance between two supply-and-

demand points, which are BSP to BPF, followed by BPF to

BDC, the proposed LSR method has outperformed the other

two methods with the lowest total logistic cost and CO₂

emission of USD 46,570 and 3,170 kg CO₂respectively.

However, the purpose of this study is not to prove which

method is the best, instead it serves to show that there exist

multiple non-programming and handy approaches available

to solve facility location problem. Each method has its own

uniqueness and benefits as shown in Table 5. FCM clustering

analysis could also work hand in hand with COG or LSR in a

hybrid manner to solve a multiple facility location problem.

FCM could be used to assign the geographical clustering

first, then followed by COG or LSR to optimize the facility

location. This sequential approach was also proposed by

Esnaf and Küçükdeniz, 2009 (16). These methods could be

applied when quick response is needed such as for emergency

disaster relief distribution.

3.2 EFB Pellet Production

Table 6 shows the EFB production potential. A total of

3,179 ton of raw EFB can be collected and converted to EFB

pellets amounted to 640,943 ton or 0.64 Mtpa per year.

Currently, SJSAS power plant has 5 power generation units

with total installed capacity of 4,100 MW and 9.5 Mtpa of

annual coal consumption (22). Assuming that the plant

exercise 3-5% biomass cofiring regime, and calorific value

ratio of EFB pellet to coal is 0.76, the power plant would

need about 0.38-0.63 Mtpa of biomass supply annually,

which means that there will be a 100% local demand for the

EFB pellets produced.

Table 3: The total land transportation distance necessary to transport the biomass from terminal-to-terminal.

Overall Travelling

distance, ꢻ_ꢿ_ꣀ_ꢿ_ꣁ_ꣂ

Travelling distance between BSP

Region

Method

Travelling between BPF to BDC, ꢻꢽꢾ (km)

to BPF, ꢻ_ꢼ_ꢽ(km)

(

km)

COG

2,287

2,390

3,245

3,182

5,531

5,572

Northern Perak

(BPF 1)

LSR

COG

2,920

4,013

3,615

3,584

2,178

4,594

4,930

4,066

5,098

8,607

8,545

7,650

Southern Perak

BPF 2)

(

LSR

Table 4: Logistic cost and CO

Total Fuel Consumption

emission due to road transportation.

Total Logistic Cost

emission

(kg)

Method

(gallon)

(USD)

COG

1,757

51,647

3,516

3,511

3,170

1,754

1,584

51,571

46,570

LSR

Table 5: Description of COG, FCM and LSR methods used in this study and lessons learned.

Method

Suitable Application

Parameters

Advantage(s) / Disadvantage(s)

The simplest and well-established method to solve new facility

location problem.

 A precursor step is needed in order to define the geographical

boundary prior to solve multiple facilities location problem.

Can solve multiple facility location problem simultaneously and

instantaneously.



Centre of gravity

COG)

Single/Multiple facility

location problem



w_i

, y

(



Only applicable for multiple facility locations only (n facility

must =>2)

Fuzzy C Means

FCM)

Multiple facility

location problem



, y

(



Unmodified FCM clustering algorithm could not take into

account the relative importance of BSP production capacity during

location optimization.



Can be used to solve multi-level distribution network.

A precursor step is needed in order to define the geographical

Least-square

regression (LSR)

Single/Multiple facility

location problem



ꢖꢔꢋꢌꢇꢔꢋ

ꢖ_ꢋ_ꢢ_ꢂꢌꢇ_ꢋ_ꢢ

boundary prior to solve multiple facilities location problem.

Journal of Environmental Treatment Techniques

2019, Volume 7, Issue 4, Pages: 689-695

Table 6: Estimation of EFB pellet production based on FFB processing capacity of existing palm oil mills.

Total EFB Pellets Production

ton/year)

Biomass/Facility

Total FFB Processed (ton/day)

Total EFB production (ton/day)

(

BPF1

BPF2

Total

6,048

8,791

14,839

1,291

1,888

3,179

260,354

380,589

640,943

.3 Limitation of Current Study

It should be noted that this paper is a preliminary study to

Energy Market Analysis: Southeast Asia. 2018.

Ozturk M, Saba N, Altay V, Iqbal R, Hakeem KR, Jawaid M,

Ibrahim FH. Biomass and bioenergy: An overview of the

development potential in Turkey and Malaysia. Renewable and

Sustainable Energy Reviews. 2017 Nov 1;79:1285-302.

Malaysia Palm Oil Board (MPOB). Palm Oil Production. 2018;

test the performance of LSR method to solve a two-level

distribution network–and compared it with the other

conventional methods of facility location problem. Therefore,

this study does not consider many other cost components

such as inventory cost, labor cost, loading cost, and vehicle

purchase and maintenance cost. Furthermore, it does not

include risk assessment or sensitivity analysis on real

problems such as seasonal biomass production therefore

variety of biomass volume supply, demand changes, road

condition, suitability of the proposed locations with the

current city planning development, and proximity to water

supply. In addition, the case study applied a conservative

assumption that 100% of EFB collected will be converted to

pellets, and that there is no other co-product of interest

existed. As such, the results shown should be interpreted with

caution and modification of the current work would be

necessary to produce more accurate results.

[

Online].

Available:

http://bepi.mpob.gov.my/index.php/en/statistics/production/177-

production-

2017.htmlhttp://bepi.mpob.gov.my/index.php/en/statistics/produ

ction/177-production-2017.html. [Accessed: 30-May-2018].

Shafie SM, Mahlia TM, Masjuki HH, Ahmad-Yazid A. A review

on electricity generation based on biomass residue in Malaysia.

Renewable and Sustainable Energy Reviews. 2012 Oct

;16(8):5879-89.

Loh SK. The potential of the Malaysian oil palm biomass as a

renewable energy source. Energy conversion and management.

2017 Jun 1;141:285-98.

6. Diabat A, Aouam T, Al-Araidah O. The uncapacitated fixed-

charge facility location problem with a multi-echelon inventory

system. In2009 International Conference on Computers

Industrial Engineering 2009 Jul 6 (pp. 803-807). IEEE.

Tang X, Zhang J. The multi-objective capacitated facility

location problem for green logistics. In2015 4th International

Conference on Advanced Logistics and Transport (ICALT) 2015

May 20 (pp. 163-168). IEEE.

Lin CC, Liu WY, Huang GL. Fuzzy Multi-objective Forest

Biomass-to-biofuel Facility Location Problem with Social

Consideration. Energy Procedia. 2019 Feb 1;158:4067-72.

Zhuraboev KA. Improving the delivery of perishable goods,

Current problems of the transport complex of Russia. Interuniv.

coll. sci. Pap. Nosov Magnitogorsk State Tech. Univ.

Magnitogorsk, 2012;2:60–68.

Conclusion

This research has successfully applied LSR method to

optimize BPF facility locations with the goal of minimizing

the cost of supplying the required biomass to a power plant.

Comparative results show that the transportation cost-

performance can be improved significantly by the LSR

method. LSR could outperform the other methods because it

is the only method that take into consideration multi-level the

travelling distance between each terminal. Moving forward, a

more robust model optimization model and algorithm is

needed to produce an integrated solution.

The analysis conducted in this study also sheds light on

the expected volume of EFB supply and logistic cost, while

the economic value derivable from the resources will require

an establish market and relevant policies. For instance,

mechanism for Competitive Generation Market, for multiple

generators to trade with a single buyer; electricity wholesale

market, where retailers can purchase energy from power

plants; and Electricity Retail Market, where consumers can

choose a retailer. Under a competition market, it is expected

the EFB resources can play a competitive substitute or new

resource to meet greener energy mix.

10. Almetova Z, Shepelev V, Shepelev S. Cargo transit terminal

locations according to the existing transport network

configuration. Procedia Engineering. 2016 Jan 1;150:1396-402.

1. Lin C. Fuzzy Multi-objective Forest Biomass-to-biofuel Facility

Location Fuzzy Multi-objective Forest Location and Cooling

Problem with Social Consideration Problem with Social

Consideration Assessing the f. 2019.

2. De Meyer A, Cattrysse D, Van Orshoven J.

A generic

mathematical model to optimise strategic and tactical decisions

in biomass-based supply chains (OPTIMASS). European Journal

of Operational Research. 2015 Aug 16;245(1):247-64.

3. Riedl WF. Heuristic for multi-echelon facility location prob-

lems with non-linear inventory considerations. 2014.

4. Zhang F, Johnson D, Johnson M, Watkins D, Froese R, Wang J.

Decision support system integrating GIS with simulation and

optimisation for a biofuel supply chain. Renewable Energy. 2016

Jan 1;85:740-8.

Acknowledgement

This research work was supported by UNITEN R&D

Sdn. Bhd. (URND) through the TNB Seeding Fund [grant

number U-TR-RD-18-24].

15. Cebi S, Ilbahar E, Atasoy A. A fuzzy information axiom based

method to determine the optimal location for a biomass power

plant: A case study in Aegean Region of Turkey. Energy. 2016

Dec 1;116:894-907.

6. Esnaf Ş, Küçükdeniz T. A fuzzy clustering-based hybrid method

for a multi-facility location problem. Journal of Intelligent

Manufacturing. 2009 Apr 1;20(2):259-65.

References

International Renewable Energy Agency (IRENA). Renewable