Format And Type Fonts CHEMICAL ENGINEERING TRANSACTIONS VOL. 39, 2014 A publication of The Italian Association of Chemical Engineering www.aidic.it/cet Guest Editors: Petar Sabev Varbanov, Jiří Jaromír Klemeš, Peng Yen Liew, Jun Yow Yong Copyright © 2014, AIDIC Servizi S.r.l., ISBN 978-88-95608-30-3; ISSN 2283-9216 DOI: 10.3303/CET1439257 Please cite this article as: Wan Y.K., Ng R.T.L., Ng D.K.S., Tan R.R., 2014, Life cycle optimisation of a sustainable sago value chain, Chemical Engineering Transactions, 39, 1537-1542 DOI:10.3303/CET1439257 1537 Life Cycle Optimisation of a Sustainable Sago Value Chain Yoke Kin Wan a , Rex T. L. Ng b , Denny K. S. Ng* a , Raymond R. Tan c a Department of Chemical and Environmental Engineering/Centre of Excellence for Green Technologies, The University of Nottingham, Malaysia Campus, Broga Road, 43500 Semenyih, Selangor, Malaysia. b Faculty of Chemical Engineering/Institute of Hydrogen Economy, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia. c Chemical Engineering Department/Center for Engineering and Sustainable Development Research, De La Salle University, 2401 Taft Avenue, 0922, Manila, Philippines. Denny.Ng@nottingham.edu.my In line with global concern on the sustainable development, economic, environmental and social aspects should be considered simultaneously in an entire value chain. In this work, a cradle-to-gate life cycle optimisation (LCO) of sago value chain that involves plantation, harvesting and processing of sago starch as well as transportation of sago starch to customer is developed. In addition, fuzzy optimisation is adapted to minimise both total operating cost and workplace footprint of sago value chain. Note that in this work, the workplace footprint is divided into three levels of risk which are death (D), non-permanent disability (NPD) and permanent disability (PD) risks. These risks and operating cost are considered simultaneously and traded off to identify the optimum pathway with minimum risks as well as operating cost for sago value chain. A realistic case study is solved to illustrate the developed approach. 1. Introduction Sago palm is a species of genus Metroxylon, with scientific name of Metroxylon sagu. It grows in tropical lowland forest in South East Asia and Papua New Guinea (Flach, 1997). Sago palm is considered as ‘starch crop of the 21 st century’, due to its strong ability to thrive in poor soil conditions (Singhal et al., 2008). During the growing cycle of sago palm, the starch accumulates in the sago trunk in the early growing stages. After approximately ten years of growing, the starch can be extracted from the trunk to produce sago starch. Such starch is one of the carbohydrate sources for humans, and it can be utilised as raw material to produce various food products (e.g., noodles, cakes, biscuits, etc.) or converted into value– added products (e.g., ethanol, sugar, kojic acid, etc.) via various technologies and processes (Singhal et al., 2008). To convert sago palm into sago starch, several steps such as plantation, harvesting and processing of sago starch are involved. The sago starch is then either supplied to local or exported to others countries. In the previous research works, life cycle assessment (LCA) is used to quantify and evaluate the environmental impact and economic performance (Kniel et al., 1996). LCA was then extended to develop an optimum LCA performance (OLCAP) framework to identify optimum process pathway (Azapagic and Clift, 1999). Later, work environmental footprint was also introduced in LCA by De Benedetto and Klemeš (2009). However, these previous research works were mainly focused on the development of assessment and optimisation methodology, and no research work is conducted on sago value chain. Therefore, in this work, a cradle-to-gate life optimisation (LCO) approach is developed to identify the optimum pathway for sago value chain with minimum total operating costs and workplace footprint. The workplace footprint is divided into three levels of risk which are death (D), non-permanent disability (NPD) and permanent disability (PD) risks. These risks and operation costs are considered simultaneously in each step of sago value chain. Fuzzy optimisation approach is adapted in this work to solve the multiobjective optimisation simultaneously. A realistic case study of sago value chain is solved to illustrate the developed approach. mailto:Denny.Ng@nottingham.edu.my 1538 2. Problem Statement A set of sago plantation g ∈ G is given with area, Ag. The area of plantation g is converted into sago palm via conversion rate of Vg. Sago palm is then converted into a set of raw material m ∈ M via conversion rate of Vg,m. The raw material is given flowrate and weight of Xg,m and qg,m, respectively. These raw materials are then supplied to a set of sago mills f ∈ F which having capacity of Zm,f,g with flowrate of Xm,f,g to produce a set of products p ∈ P based on the conversion rate of Vm,f,g. The products are then transferred from sago mills f to a set of port j ∈ J with flowrate of Xf,p,j. The port capacity is given as Zj and total flowrate to port j is given as Xp,j. The stored products in port j are then shipped to a set of customer u ∈ U with flowrate of Xp,j,u based on the demand of customers (Du,p). In order to identify the optimum pathway with minimum total operation cost and total life cycle risks of sago value chain, a Mixed Integer Linear Programming (MILP) approach is used for the formulation of the following mathematical models. 3. Life Cycle Optimisation Formulation 3.1 Mass Balances mg mg gg X , , V .V A  mg∀∀ (1) ∑ F f fmgmg XX 1 ,,,   mg∀∀ (2) pfmmg G g fmgpfm X ,,, 1 ,,,, Vq Z . . ≥ ∑  pfm ∀∀∀ (3) ∑ ∑ M m J j jpfpfm XX 1 1 ,,,,    jpfm ∀∀∀∀ (4) ∑∑ P p F f jpfj X 1 1 ,,Z    j∀ (5) ∑ U u ujpjp XX 1 ,,,   jp∀∀ (6) ∑ J j ujppu X 1 ,,, D   pu∀∀ (7) 3.2 Cost Computations ggg HC . HarvHarv UC g∀ (8) Tran , Tran_roadTran ,, UC g,m,ffgfmg nC . d .  fmg ∀∀∀ (9) RMat mg, . UC,, RMat fmgf XC  f∀ (10) Tran ,,, Tran_roadTran ,, UC jpfjfjpf nC . d .  jpf ∀∀∀ (11) T ran ,, 1 1 T ran_PortT ran_Port UC ujp P p U u jj nC . ∑    j∀ (12) 1539 CPT T ran ,, 1 T ran_Sea , T ran_Sea , UC n n C ujp P p ujuj . ∑   uj∀∀ (13) pfmpfmpfm XC ,, Process ,, Process ,, UC .  pfm ∀∀∀ (14) ∑∑∑ ∑∑∑∑∑∑∑∑∑∑ M m F f P p pfm J j U u uj J j j F f P p J j jpf F f f G g M m F f fmg G g g C CCCCCCTotC 1 1 1 Process ,, 1 1 T ran_Sea , 1 T ran_Port 1 1 1 T ran ,, 1 RMat 1 1 1 T ran ,, 1 Harv             (15) 3.3 Risk Computations ggg HR . HarvHarv r g∀ (16) ∑ . d . Y y g,m,ffygyfmg nR T ran ,, T ran_MillT ran_Plant ,, r fmg ∀∀∀ (17)    P p f pf f X R 1 Process,Process r . 000,1 f∀ (18) ∑ . d . Y y jpfjyfyjpf nR T ran ,,,, T ranPortT ran_Mill_ ,, r jpf ∀∀∀ (19) PortPort r . 000,1 j j j X R  j∀ (20) j,u ujp P p Sea uj n n R dr CPT T ran ,, 1 Sea , . . ∑   uj∀∀ (21) ∑∑ ∑∑∑∑∑∑∑∑∑ J j U u uj J j j F f P p J j jpf F f f G g M m F f fmg G g g R RRRRRTotR 1 1 Sea , 1 Port 1 1 1 PortT ran_Mill_ ,, 1 Process 1 1 1 _MillT ran_Plant ,, 1 Harv          (22) 3.4 Fuzzy Optimisation of Objectives λ TotC    LLUL UL TotCTotC TotC (23) λ TotR    LLUL UL TotRTotR TotR (24) λ Max (25) 1540 4. Case Study In this work, a case study of sago value chain in Sarawak, Malaysia is developed. Figure 1 shows the superstructure of sago value chain. It is consists of plantations, raw materials, sago mills, products, ports and customers. Based on the statistic (Department of Agriculture Sarawak, 2014), the sago palm is mainly grown in Mukah and Betong divisions in Sarawak. In this case, the major four districts (e.g., Mukah, Dalat, Saratok and Betong) are identified as plantation areas. In these plantations, the sago palm is harvested to produce sago logs as raw material. These sago logs are then sent to sago mills for sago starch production. As sago mills in Sarawak are located in Mukah, Dalat and Pusa districts, thus, two sago mills in Mukah and three sago mills in Dalat and one sago mills in Pusa are located as the sago mills for analysis. The produced sago starch is then sent to the three ports (e.g., Kuching port, Sibu port and Miri port) in Sarawak for storage before exporting to customers (e.g., Japan, Peninsular Malaysia, Singapore and Thailand) based on the customers’ demand. In this case, Cost of Insurance and Freight (CIF) term is used as shipment term for sago starch delivery. Mukah Dalat Saratok Betong Mukah A Mukah B Dalat A Dalat B Dalat C Pusa Sago Log Sago Starch Kuching Sibu Miri Japan Pen. Malaysia Singapore Thailand Plantation g ϵ G Raw Materials m ϵ M Sago Mills f ϵ F Products p ϵ P Ports j ϵ J Customers u ϵ U Figure 1: Superstructure of sago value chain In order to determine the optimum total operating cost and total life cycle risks, the vital information is given in Tables 1 – 3. Table 1 shows plantation capacity, sago mills processing capacity and ports capacity as well as customers demand. Table 2 shows the unit cost of sea-freight to customers, port’s charges, harvesting in plantations, sago starch processing of sago mills and selling prices of sago log. Table 3 shows the death (D), non-permanent disability (NPD), permanent disability (PD) risk of road accident, sea-freight, forestry and logging, sago starch processing and ports handling. Solving the Equations (1) – (25) in Mixed Integer Linear Programming (MILP) model with commercial optimisation software, LINGO, version 13, with Global Solver, optimum pathway of sago value chain is identified. Results of this case study is summarised in Table 4 and Figure 2. Based on the optimised results shown in Table 4, the optimum pathway with maximum λ of 0.46 is identified. Note that the targeted minimum total operating cost is given as 8.992 x 10 7 MYR/y with minimum risk of 19,370 x 10 -6 D/y, 146,142 x 10 -6 NPD/y and 5,560 x 10 —6 PD/y. The optimum pathway with mass flowrate is shown in Figure 2. Table 1: Plantation capacity, sago mills processing capacity, ports capacity and customers demand Plantation Area Capacity gA (ha) Sago Mills Processing Capacity, pfm ,,Z (t/y) Ports Port Capacity, jZ (t/y) Customer Demand pu,D (t/y) Mukah 2,599 Mukah A 13,200 Kuch 7,000,000 Japan 13,000 Dalat 17,541 Mukah B 8,250 Sibu 450,000 P. M’sia 30,700 Saratok 1,907 Dalat A 7,260 MIri 53,900 Singapo 3,000 Betong 3,776 Dalat B 8,250 Thailand 1,300 Dalat C 8,250 Pusa 3,960 1541 Table 2: Unit cost of sea-freight, road transportation, ports charges, raw materials, harvesting and sago starch processing Japan, Tran_Sea , UC uj (MYR/trip) P. Malaysia, Tran_Sea , UC uj (MYR/trip) Singapore, Tran_Sea , UC uj (MYR/trip) Thailand, Tran_Sea , UC uj (MYR/trip) Ports charges, Tran_Port UC j (MYR/container) Kuch 3,960 1,650 1,485 2,640 1,500 Sibu 3,729 1,980 1,584 2,805 1,300 Miri 3,531 2,310 1,650 2,970 1,200 Log selling price, RMat mg,UC (MYR/log) Harvesting Cost, Harv UCg (MYR/palm) Processing Cost, Process ,,UC pfm (MYR/t) Mukah 10 3.8 Mukah A 296 Dalat 12 4.2 Mukah B 303 Saratok 8 3.0 Dalat A 305 Betong 9 3.6 Dalat B 300 Dalat C 310 Road Transportation cost, Tran_road UC (MYR/km) 4.5 Pusa 278 Table 3: Death (D), non-permanent disability (NPD) and permanent disability (PD) risk of road accident, sea-freight, forestry and logging, sago processing and port handling Road accident risk T ran ry x 10 -14 Forestry and logging Risk Harv gr x 10 -9 D/km NPD/km PD/km D/palm NPD/palm PD/palm Kuching 156 239 234 Mukah 26.9035 231.265 8.54901 Samarahan 2.78 13.9 2.78 Dalat 69.8603 600.525 22.1992 Serian 27.8 19.5 100 Saratok 5.45857 46.9223 1.73455 Simunjan 8.34 2.78 75.1 Betong 10.2128 87.7901 3.24528 Sri Aman 103 97.3 8.34 Betong 27.8 8.34 22.2 Processing Risk, Process fr x 10 -8 Saratok 16.7 2.78 58.4 D/ton NPD/ton PD/ton Sarikei 47.3 91.8 111 Mukah A 2.63 38.7 4.35 Maradong 0.00 0.00 0.00 Mukah B 2.63 38.7 4.35 Sibu 114 0.00 2.78 Dalat A 3.23 47.5 5.34 Dalat 2.78 13.9 2.78 Dalat B 3.23 47.5 5.34 Mukah 27.8 8.34 22.2 Dalat C 3.23 47.5 5.34 Tatau 33.4 8.34 16.7 Pusa 6.57 96.6 10.9 Bintulu 139 13.9 50.1 Miri 186 114 656 Port Handling Risk, Port j r x 10 -8 D/ton NPD/ton PD/ton Sea freight Risk, Sear x 10 -15 Kuching 16.4 100 1.54 D/nm 2.21 Sibu 28.2 172 2.64 Miri 24.7 151 2.31 Table 4: Case study results λ TotC (MYR/y x 10 7 ) TotR D (D/y x 10 -6 ) TotR NPD (NPD/y x 10 -6 ) TotR PD (PD/y x 10 -6 ) Min = TotC - 8.821 20,516 153,251 5,678 Min = TotR D - 9.139 18,035 140,673 5,677 Min = TotR NPD - 9.139 18,403 140,025 5,452 Min = TotR PD - 9.139 18,897 142,638 5,424 Min = λ 0.46 8.992 19,370 146,142 5,560 1542 Mukah Dalat Saratok Betong Mukah A Mukah B Dalat A Dalat B Dalat C Pusa Kuching Sibu Miri Japan Pen. Malaysia Singapore Thailand Plantation g ϵ G Sago Mills f ϵ F Ports j ϵ J Customers u ϵ U 60 ,1 50 ,0 00 k g/ y 14,780,000 kg/y 16 ,1 30 ,0 00 k g/ y 10 ,3 40 ,0 00 k g/ y 36 ,3 00 ,0 00 k g/ y 41 ,2 50 ,0 00 k g/ y 41 ,25 0,0 00 kg /y 19,800,000 kg/y 1,080,000 kg/y 3 ,9 6 0 ,0 0 0 k g /y 8,250,000 kg/y 10,0 00 kg /y 7,260,000 kg/y 8,2 50 ,00 0 k g/y 10,950,000 kg/y 13,280,000 kg/y 17 ,42 0,0 00 kg /y 13 ,0 00 ,0 00 k g/ y 3,000,000 kg/y 1,300,000 kg/y 8, 24 0, 00 0 kg /y Figure 2: Optimum pathway of sago value chain 5. Conclusions This paper presents a life cycle optimisation (LCO) approach which can identify the optimum pathway of sago value chain with minimum total operating cost and total life cycle risks via fuzzy optimisation. This work can be further extended in future with consideration of environmental aspects (e.g., carbon and water footprint). In addition, integration of sago-based biorefinery into sago value chain can also be conducted. Acknowledgement The financial support from Crops for the Future Research Centre (CFFRC) via CFFRCPLUS postgraduate studentships (BioP1-003) is gratefully acknowledged. References Azapagic A., Clift R., 1999, The application of life cycle assessment to process optimisation, Comp. Chem. Eng., 23(10), 1509-1526. De Benedetto, L., Klemeš, J., 2009, The environmental performance strategy map: an integrated LCA approach to support the strategic decision-making process, J. Clean. Prod., 17, 900-906. Department of Agriculture Sarawak, 2014, Sarawak agriculture statistics accessed 09.03.2014 Flach M., 1997, Sago palm Metroxylon sagu Rottb, Promoting the conservation and use of underutilized and neglected crops 13, Institute of Plant Genetics and Crop Plant Research, Gatersleben/International Plant Genetic Resources Institute, Rome, Italy. Kniel G.E., Delmarco K., Petrie J.G., 1996, Life cycle assessment applied to process design: Environmental and economic analysis and optimization of nitric acid plant, Environ. Progress, 15(4), 221-228. Singhal R.S., Kennedy J.F., Gopalakrishnan S.M., Kaczmarek A., Knill C.J., Akmar P.F., 2008, Industrial production, processing, and utilization of sago palm-derived products, Carbohydr. Polym., 72, 1-20. http://www.doa.sarawak.gov.my/