You are on page 1of 11

Journal of Cleaner Production xxx (2013) 1e11

Contents lists available at ScienceDirect

Journal of Cleaner Production


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

Optimal process network for municipal solid waste management in Iskandar Malaysia
Sie Ting Tan a, Chew Tin Lee b, *, Haslenda Hashim a, Wai Shin Ho a, Jeng Shiun Lim a
a b

Process System Engineering Centre (PROSPECT), Faculty of Chemical Engineering, Universiti Teknologi Malaysia, UTM Skudai, 81310 Johor, Malaysia Department of Bioprocess Engineering, Faculty of Chemical Engineering, Universiti Teknologi Malaysia, UTM Skudai, 81310 Johor, Malaysia

a r t i c l e i n f o
Article history: Received 31 August 2013 Received in revised form 2 December 2013 Accepted 3 December 2013 Available online xxx Keywords: Municipal solid waste (MSW) Resource processing network Waste treatment technologies Optimisation Mixed integer linear programming (MILP)

a b s t r a c t
Ineffective management of municipal solid waste (MSW) may cause degradation of valuable land resources and create long-term environmental and human health problems. A sustainable and efcient waste management strategy is needed to balance the need for development, the quality of human life and the environment. This study aims to synthesis a MSW processing network to produce energy and value-added products for achieving economic and environmental competitiveness. An optimisation model that integrates four major utilisation technologies was incorporated to facilitate a cost-effective processing network. The model is able to predict the best mix of waste treatment technologies, forecast the production of by-product from waste treatment process, estimate the facility capacity, forecast the greenhouse gas (GHG) emission of the system, and eventually generate an optimal cost-effective solution for municipal solid waste management (MSWM). Four scenarios for MSWM were considered to analyse the economic impact of different waste utilisation alternatives: i) the business as usual (BAU) scenario as a baseline study, ii) the waste-to-energy (WTE) scenario, iii) the waste-to-recycling (WTR) scenario, and iv) the mixed technology (MIXTECH) scenario. The MIXTECH scenario was able to provide the best mix of waste utilisation technologies. The optimal waste allocation in terms of percentage involved landll gas recovery system (LFGRS) (14%), mass burn incineration (3%), material recycling facilities (MRF) (56%), and composting (27%). The optimal scenario would be able to achieve the renewable energy (RE) target, achieve the recycling target and promote composting as the waste reduction alternative for the region being studied. Sensitivity analyses were conducted for the optimal or MIXTECH scenario to examine the effect of the RE target and GHG emission reduction target with respect to the system cost and waste allocation to each technology. The proposed mixed integer linear programming (MILP) model was applied for Iskandar Malaysia (IM) as a case study. 2013 Elsevier Ltd. All rights reserved.

1. Introduction Municipal solid waste (MSW) refers to waste generated from residential, commercial, institution and public parks (Fodor and Kleme s, 2012). Solid waste management (SWM) involves many technologies associated with controlling waste generation, handling and storage, transportation, processing and nal disposal. The hierarchy of SWM was formed since 1970s, several evolution, different versions of solid waste treatment hierarchies exist. One of those affordable hierarchies is suggested by Finnveden et al. (2005),

* Corresponding author. Tel.: 60 7 553 5594; fax: 60 7 5538003. E-mail addresses: ctlee@utm.my, chewtin@gmail.com, ctlee@cheme.utm.my (C. T. Lee). 0959-6526/$ e see front matter 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jclepro.2013.12.005

in the order of reduction of waste amount, reuse, recycle, compost or recovery through incineration and nally landll disposal. It explained that the main objective of SWM is to treat the waste generated. In addition, energy and recyclable material can be recovered as by-products to achieve sustainable waste management that is environmental friendly, economically reasonable and socially acceptable (Tchobanoglous and Kreith, 2002). Rapid urbanisation, population growth and industrialisation contribute towards large-scale increase of MSW in Malaysia. These factors have changed the characteristics and composition of the solid waste generated. The daily waste generation has also shown an upward trend. Waste generation was 16,200 t in year 2001. This amount increased to 19,100 t in 2005, 17,000 t in 2007 and 21,000 t in 2009 (Ahmad et al., 2011). Due to the increased population growth rate, the daily solid waste generated is estimated to be 31,000 t/d by 2020 (Johari et al., 2012).

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005

S.T. Tan et al. / Journal of Cleaner Production xxx (2013) 1e11

Depending on its characteristics, the MSW can be preferentially processed by different approaches. The present waste management methods in Malaysia are highly dependent on landll as only 5.5% of the MSW is recycled and 1% is composted, while the remaining 94.5% of MSW is disposed on the landll site (Periathamby et al., 2009). The practice of waste segregation is random and unofcial in Malaysia. Waste recycling is mainly performed by garbage scavengers at the landll sites. To date, SWM in Malaysia is at the stage of transition and planning towards sustainable and effective approaches. Ineffective management of waste may cause degradation of valuable land resources, increase land costs, and create long-term environmental and human health problems. Sustainable and more efcient waste management strategies are needed to reduce the heavy reliance on landlls. Malaysia aims to establish a holistic framework that considers the trade-off involved in the segregation process and the economic performance of different MSW practices to achieve the national MSW recycling rate (22% of the total MSW) by the year 2020 (Ministry of Housing and Local Government, 2005). The segregation and recycling of waste are essential to improve the performance of waste processing. 1.1. Literature review on waste management model The complexity of SWM includes the prediction of solid waste generation, selection of waste treatment technologies, selection of facility sites, estimation of facility capacity, operation of the facility, scheduling of the system and transportation of the waste. SWM can be modelled through a system perspective (Seadon, 2010). System analysis tools for supporting decision making in waste management were developed since 1970s, these models can be categorised into two groups: (1) system engineering (SE) models and (2) system assessment (SA) models (Pires et al., 2011). The SA models can be used to analyse the performance of an existing waste management system, for example life-cycle assessment (LCA), risk assessment, and material ow analyses (Juul et al., 2013). For instance, Chen and Chang (2010) developed a range of SA models to assess the performance of MSW recycling in Taiwan. The models included the diffusion effect and the organisational learning effect as the key variable for recycling performance, however, other variables such as costing and environmental protection are not considered in the models. The LCA tool is a popular tool to solve the complex issues of SWM. For instance, Othman et al. (2013) reviewed the application of LCA for the assessment of integrated solid waste management for several Asian countries. The study focused on the assessment of environmental impacts of various waste treatment technologies and concluded that recycling, anaerobic digestion and thermal treatments are effective technologies for the Asian countries. Wanichpongpan and Gheewala (2007) used LCA as a decision tool to assess the environmental impact of landll gas-to-energy system in Thailand, they concluded that a centralised landlling facility is environmental and economical benecial as compared to small landlls. Poeschl et al. (2012a, b) used LCA to analyse the biogas production system and utilisation pathways from different input sources including from MSW and feedstock. Liamsanguan and Gheewala (2008) used LCA to assess the holistic impact of integrated solid waste management towards the mitigation of greenhouse gases emission in Phuket, Thailand. These LCA studies tend to assess the waste treatment technologies focussing on the environmental impact and with less consideration on the detailed modelling and optimisation for the economical impact of the processes. While SA models focus on the assessment and analysis of the existing systems, SE models focus on the design and solution of a waste management system. Methods such as multi-criteria decision models (MCDM), simulation models, forecasting models, cost-benet analysis, and optimisation models are widely used in

the SE approach (Pires et al., 2011). The optimisation model developed using SE model emphasises the design of a system by a specic objective function which gives the best solution to the objective function (Juul et al., 2013). Various types of techniques have been implemented as an optimisation model for SWM. These include the linear programming (LP), mixed integer linear programming (MILP), non-linear programming (NLP), multi-objective programming (MOP), stochastic programming, two-stage programming, fuzzy method programming, and hybrid models. An overview of the optimisation models for SWM are summarised in Table 1. The early stage of SE model developed for SWM focused on the cost-effectiveness principle of LP with a single-objective optimisation scheme (Juul et al., 2013). For example, Mnster and Meibom (2010, 2011) designed an energy system using the Balmorel model to optimise the investment cost for different waste-to-energy (WTE) technologies in the northern Europe. In addition, a LP model was developed by Rathi (2007) to investigate the SWM technologies that focused on composting by taking into account both the economic and environmental impacts. Other LP model as developed by Salvia et al. (2002) also addressed the similar issue of waste management and emphasised on the analysis of one particular technology. LP model is typically applied and limited to a single process that does not support the evaluation and selection of multiple technologies. More powerful modelling tools are needed to conduct modelling work for SWM notably for the real case studies that involve a range of uncertainties. For instance, more complex modelling methods including mixed integer linear programming (MILP), non-linear programming (NLP) (Chang et al., 1997; Shadiya et al., 2012), stochastic programming (Guo and Huang, 2009b), fuzzy logic (Yeh and Xu, 2013) and hybrid model (Xu et al., 2010; Li and Chen, 2011; Chang et al., 2012) were developed to assess the complex scenarios of SWM in the real world. MILP is relatively simple and can be applied to consider the complex scenario with uncertainties using the binary selection function that facilitates the selection of multiple technologies and dynamic planning of resource network for SWM. Badran and ElHaggar (2006) proposed a MILP model for the optimal management of MSW at Port Said, Egypt, with the objective of minimising the waste collection and transportation costs. Dai et al. (2011) designed a MILP model to assess waste allocation issue and the expansion of capacity for the waste treatment facility. SantibaezAguilar et al. (2013) determined the optimal supply chain network for waste utilisation using MILP. Ng et al. (2013) developed a MILP model to determine the waste-to-energy network that optimised the cost, waste energy potential utilisation, and the carbon footprint. As presented in Table 1, many models were developed based on various waste management technologies including composting, recycling and disposal to optimise the economical factor. Relatively fewer models have simultaneously considered the economical factor based on energy system and WTE technologies such as that by Mnster and Meibom (2011). Ng et al. (2013) developed a WTE processing network with integrated consideration for economical and environmental factor. However, the model did not incorporate other waste treatment alternatives such as recycling or composting. As a whole, it is of great challenge and interest to integrate both the waste management system and energy system into the modelling works to achieve optimal economical and environmental consideration. 1.2. Research objectives and scopes In general, MSW in Malaysia is typically disposed in a bin or container within the house premise and collected by the respective

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005

S.T. Tan et al. / Journal of Cleaner Production xxx (2013) 1e11 Table 1 Optimisation models for solid waste management. Method Reference Objectives Focus Energy system Linear programming Mnster and Meibom (2011, 2010) Rathi (2007) To maximise economic utility of energy consumers (Balmorel) To integrate the best feasible method of waste management in Mumbai, focus on composting. To minimise total system cost for waste and energy management To minimise the waste collection and transportation costs To minimise cost of waste ow allocation and facility capacity expansion To determine the optimal supply chain network for the waste utilisation To determine optimal processing network for waste-to-energy system. To maximise short-term waste management strategies based on cost, energy, and material recovery To maximise prot, while minimizing waste through source reduction To minimise costs of capacity expansion and waste ows To minimise the sum of the squared differences between individual e-waste products best dimension sustainability score Interval-parameter stochastic robust optimisation, waste ows, revenue from WTE Fuzzy-stochastic-interval linear programming for supporting Multi-objective programming and costbenet criteria on global warming impact in waste management U U Waste management Optimisation on Economy U U Environment

Salvia et al. (2002) Mixed integer linear programming (MILP) Badran and El-Haggar (2006) Dai et al. (2011)

U U U

U U U

Santibaez-Aguilar et al. (2013) Ng et al. (2013) Non-linear programming (NLP) Chang and Chang (1998)

U U U U

U U U U

Shadiya et al. (2012) Stochastic programming Fuzzy logic model Guo and Huang (2009) Yeh and Xu (2013)

U U U

U U U

Hybrid model

Xu et al. (2010)

Li and Chen (2011) Chang et al. (2012)

U U

U U U

regional private concessionaires. The wastes are rstly transferred to transfer stations for compaction in compacting containers before being sent to the wastes disposal sites. Some of the MSW is recycled or composted; however, the predominant treatment methods for MSW are landlls or the open dumpsites. Although there are various models designed to address the MSW issues, the literature indicates the lack of proposal for comprehensive waste management in Malaysia which integrated technology selection for solid waste treatment and waste-to-energy (WTE) treatment, mitigation of GHG emission and optimisation of economical impact. Therefore, this study aims to synthesise a cost-effective processing network for integrated MSW in Malaysia that consider the following factors: (a) Resource allocation (b) Production portfolio (c) Best available technology for an appropriate capacity and time of construction (d) Economical and environmental optimal for integrated solid waste treatment and waste-to energy (WTE) treatment. The proposed processing network of MSW was designed through a MILP model. The MILP model integrated several waste utilisation technologies, including the landll gas recovery system (LFGRS), waste incineration with energy recovery, material recycling facilities (MRF), and composting. The proposed system aimed to maximise the protability of MSW processing network that considered the product demand (energy demand and recycling demand) and the carbon emission reduction target. The proposed

MILP model was applied in Iskandar Malaysia (IM) as a case study. The MSWM system has been tested using different scenarios, namely with or without the consideration of WTE and/or waste-torecycling (WTR) strategies to analyse the feasibility and best potential for a future MSWM system in the IM region. It is envisaged that the developed model can assist the waste management planner to design and schedule a protable yet sustainable MSWM system for the region. Section 2 presents the methodology of this study, a superstructure that primarily contains the resource inputs, technology alternatives and product streams was rst constructed and then the optimisation model and the constraints are presented. Section 3 presents the case study with the data and assumptions. It also provides the conversion yield for each process, the feasibility criteria for directing the segregated waste to a specic technology, the economic data for the technology, resources (wastes and end products) and incentives, and the ow rate of each waste directed into a specic technology. In Section 4, the results covering product portfolio, economical and environmental analysis under different scenarios are discussed and compared. 2. Research methodology 2.1. Superstructure for model development This study presents a utilisation system for MSWM considering a set of representative waste treatment technologies that can potentially be implemented in the IM region. A superstructure is

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005

S.T. Tan et al. / Journal of Cleaner Production xxx (2013) 1e11

developed to illustrate the ow conguration of the MSWM, as shown in Fig. 1. In this study, the waste source, i, will be segregated by process, p, to be further classied into different categories of i (i.e. 1: food waste; 2: paper; 3: yard waste; 4: plastic; 5: glass and ceramic; 6: metal; 7: textile). Assuming that the practice of waste segregation exists, different types of waste will be processed according to their potential value. The system consists of four waste treatment technologies, p, (denoted by a, b, c, d) that covered two types of WTE technologies, p (i.e., LFGRS and incineration) and two WTR technologies, (i.e., composting and recycling through the MRF). For example, the food waste, i, will be allocated to three waste treatment technologies (composting, incineration, and LFGRS). The four key waste treatment technologies would transform the segregated MSW into three key value-added products, i, (denoted by 8: compost; 9: recycled materials, and 10: electricity) to full the demand for products and renewable energy (RE). The product, i, will be distributed to market for selling, e, or electricity to the national grid, f. 2.2. Model formulation The complexity of an MSWM system is described by the number of relationships among the components in the system. The mathematical analysis of the cost-effective processing network for a MSW is conducted by applying the MILP model due to its simplicity and common use for solving the complex SWM issues. The objective and model constraints incorporated several aspects of economics, energy, recycling, waste segregation and waste to value-added product to represent the real scenario in Malaysia. Two-step programming approaches (i.e. simulation and optimisation) are used to evaluate the sustainability of the designed system. The simulation model is used to describe the energy production from waste and landll gas generation, while the optimisation model will provide an optimal solution for the MSWM. The model is developed and computed using the General Algebraic Modelling System (GAMS, version 22.9) (GAMS Development Corporation, 2013), a computer software for solving mathematical programming and optimisation problems. 2.2.1. Objective function The optimisation model is formulated with an objective function and several constraints. The objective function aims to maximise the overall prot (PROFIT) of the MSWM system as described by Eq. (1). This function consists of the revenue from product selling (REV), processing cost (PCOST), total capital cost (CCOST) and the variable cost (VCOST).

REV represents the product revenues from the MSWM system as described in Eq. (2). PROit is the production rate of product i during period t. PRICEit denotes the unit price of product i in period t, which is obtained from the published price and manufacturing quotes.

REV

X
it

PRICEit PROit

(2)

PCOST is the total processing cost of producing the value-added product, as shown in Eq. (3). MATipt denotes the input rate of waste i into process p during period t. UPCOSTpt is the unit processing cost of process p at period t.

PCOST

X
ipt

UPCOSTpt MATipt

(3)

VCOST represents the total variable operating and maintenance costs of the system as described in Eq. (4). In Eq. (4), UVCostpt is the unit variable cost of the corresponding process p at time t.

VCOST

X
ipt

UVCostpt MATipt

(4)

The capital cost CCOST is described by Eq. (5), where YPpz is the binary decision variable for purchasing technology p with capacity z, while ACPCOSTpzt is the annualised capital cost of technology p with capacity z at period t.

CCOST

X
pzt

YPpz ACPCOSTpzt

(5)

2.2.2. Constraints To dene the relationship among the variables and parameters in this model, several linear equality, inequality and matrix manipulation constraints are developed in the following text. 2.2.2.1. Mass balance for resources. Two types of resources are introduced in the MSWM system: the internal resource and the external resource. RESit is the amount of resource i in the system during time period t, as described in Eq. (6). EXRECit is the input rate of the external resource, i, during period t, while SGRESipt is the quantity of resource i generated within the system under process p through period t.

RESit EXRECit

X
p

SGRESipt

cict

(6)

PROFIT REV PCOST CCOST VCOST

(1)

The waste resource RESi that is fed into the respective process, p, can be converted into a product PROit, either as a material product i, an energy product i, or both, with different conversion rate ((MATipt),

Fig. 1. The superstructure of the MSWM system.

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005

S.T. Tan et al. / Journal of Cleaner Production xxx (2013) 1e11

as described in Eq. (7). The potential conversion of resource-toproduct is explained via a superstructure as depicted in Fig. 1.

RESit PROit

X
p

MATipt

cict

(7)

2.2.2.2. Mass balances for processing unit. The mass balance for the processing technology, p, is computed using Eqs. (8) and (9). In Eq. (8), PRESpt is the quantity of the resource being processed by process p at period t. The integration matrices of the corresponding process, including the material process selection matrix (MPSMip) and the process resource conversion matrix (PRCMpi), are described by Eqs. (8) and (9).

2.3.2. Wasteeto-energy (WTE) scenario Under this scenario, the WTE production will be maximised according to the energy demand targeted by the policy makers (Blueprint for Iskandar Malaysia, 2010b). To maximise the preference for energy produced from waste, the incentives (feed e in tariff) for promoting the use of MSW for energy production are included in the model formulation. Under this scenario, the WTE practice including landll gas recovery system (LFGRS) and waste incineration were incorporated. 2.3.3. Waste-to-recycling (WTR) scenario Under the WTR scenario, MSW utilisation through MRF and organic waste recycling via composting technology were introduced. Other technologies relevant to WTE were not considered. 2.3.4. Mixed technology (MIXTECH) scenario The mixed technology (MIXTECH) scenario introduces all potential MSWM practices available in the case study, there are LFGRS, incineration, MRF, and composting. This scenario is designed to achieve the maximum net prot for the proposed MSW processing network without exceeding the product demand. An optimal solution for MSWM is anticipated to full every demand without over- or under-production. 3. Case study and input data 3.1. Case study e Iskandar Malaysia (IM) Iskandar Malaysia (IM) is the third largest metropolis and the most developed region in the Southern Peninsula of Malaysia. IM aims to be transformed into a metropolitan by 2020. IM covers an area of approximately 2217 km2 and has a population of 1.7 million with ve agship zones: Zone A (JB city centre), Zone B (Nusajaya), Zone C (Western Gate Development), Zone D (Eastern Gate Development), and Zone E (Senai-Skudai), as shown in Fig. 2. IM was established in the year 2008 from four different municipalities with an expanding population base and increased economic activity due to its rapid development since 2008. Solid waste generation in IM increased by approximately 30% from 2005 to 2010 and is expected to increase 50% by 2025. More than 95% of the waste is directly disposed in three nal disposal landll sites located around the region, as shown in Fig. 2. Only a small portion of the waste is recycled informally. Of the three nal disposal landll sites in IM, only one site involves a sanitary landll with LFGRS; another two are conventional landlls or dumpsites. Table 2 projects the annual waste generation in IM. In an effort to improve the current SWM in IM, the city council could introduce several measures, such as waste separation at source, upgrading the current landll into a sanitary landll, establishing a new MRF, and adapting WTE facilities to utilise the waste, as outlined in the Blueprint for Iskandar Malaysia (2010a, b). 3.2. Input data 3.2.1. Waste data The MSW in IM are categorised into seven fractions, namely food, yard, paper, plastics, glass and ceramic, metal, and textile wastes. The composition of the MSW is shown in Table 3. Organic waste is the main component of MSW in IM, representing more than 40% of the total waste. The MSW in IM has an average caloric value of 16.68 MJ/kg (Low Carbon Society Blueprint for Iskandar Malaysia 2025, 2012). The value was calculated based on the moisture content, combustible content and ash fraction of 57%, 35%, and 8.2% respectively.

PRESpt

X
i

MATipt MPSMip

cpct

(8)

PRESpt PRCMpi SGRESipt

cicpct

(9)

2.2.2.3. Process sizing. The amount to be processed is governed by CAPpz which is the capacity of the process with size z, as determined in Eq. (10). The amount of processing material, PRESpt, must be less than or equal to the process capacity CAPpz with the binary variable YOPpzt to determine whether the process p should be operated at size z during period t.

PRESpt 

X
z

YOPpzt CAPpz

cpct

(10)

2.2.2.4. Product demands. The market demand for each product uctuates over time. Thus, product, i, must full the product demand PRODEMit at time t, as described by Eq. (11).

PROit  PRODEMit

cict

(11)

2.2.2.5. CO2 emissions. In addition to the economic constraints, the system is also bounded by an environmental limitation: the carbon emission of the waste treatment process CEPpt is the carbon emission from process p at period t, while CERT is the carbon emission reduction target for the system. The carbon emission of the waste treatment process in year 2005, CEPpt2005, is used as the baseline value. To meet the annual carbon emission reduction target, the emission from the process must be equal to or less than the reduction requirement, as shown by Eq. (12).

X X 1 CERT CEPpt2005 CEPpt2005 CEPpt 


p p

ct

(12)

The carbon emission of each process, CEPpt, is denoted by Eq. (13), where EFp is the emission factor for process p.

CEPpt

X
i

MATipt EFp

ct

(13)

2.3. Scenario setting To evaluate the impacts of different waste management options on the utilisation system for MSW, four scenarios were constructed in this study. 2.3.1. Business as usual (BAU) scenario The BAU scenario represents the baseline study for the current scenario in IM. Under this scenario, year 2005 was set as the current practice year where most of the MSW was landlled and a small percentage was recycled. The BAU scenario assumes that no other effort of MSWM is introduced other than the landlling.

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005

S.T. Tan et al. / Journal of Cleaner Production xxx (2013) 1e11

Fig. 2. Five major agship zones (A to E) of Iskandar Malaysia (Iskandar Malaysia, 2013) and the three major landll sites.

Table 2 Projection of annual waste generation in IM from year 2012 to 2025 (Blueprint for Iskandar Malaysia, 2010a). Year 2012 2014 2016 2018 2020 2022 2024 2025

Annual waste 0.766 0.859 0.963 1.080 1.198 1.358 1.523 1.577 generation (Mt)

Table 5. The by-product selling price is presented in Table 6. To increase the development of RE in Malaysia, feed-in tariff policies were designed to offer guaranteed prices for xed periods of time for RE, depending on the types of technologies and capacity (Pusat Tenaga Malaysia, 2010). The feed-in tariff for MSW in Malaysia is presented in Table 7. Table 8 presents the RE demand for IM. 3.3. System boundaries and major assumptions Several assumptions were made for the model developed in this study: (a) The practice of waste segregation assumed with a rate of 100%, the ideal value. Different types of waste were processed according to their respective potential values. (b) The recovery factors used for MRF is assumed to be 90%. (c) The study aimed to synthesise an optimal processing network for waste management. The system did not consider the transfer and transportation cost of MSW. (d) At least one year was required to construct the WTR plants (MRF and composting) and three years for WTE plants (LFGRS and incineration). 4. Results and discussion

3.2.2. Waste treatment technologies and related data The four waste treatment technologies considered in this case study included LFGRS, waste incineration with energy recovery, MRFs, and large-scale composting facilities. The types of waste allocated to each technology are shown in Table 4. The waste treatment system would treat waste effectively by reducing the volume and also generate by-products. For instance, the energy produced from incineration and LFG was assumed to be converted to electricity. The residue of the waste incineration process included bottom ash and y ash would be sold as the by-product, while the treatment cost of the residue was included in the incineration cost. Other by-products from waste treatment technologies included compost and recycled materials. The input for each technology in terms of cost analysis and emission rate is shown in

Table 3 The composition and waste-related data in IM. Types Food Yard Paper Plastic Glass/Ceramic Metal Textile Total/average
a b

Composition (%)a 41.1 2.5 20.9 22.2 3.6 2.0 7.7 100

LHV (MJ/kg)a 5.26 0.48 3.08 5.38 0.01 0.01 2.48 16.68

Carbon Content (%)b 41.47 37.37 42.61 60.93 0 0 60.42 34.68

In this section, the modelling and optimisation results generated for all four scenarios involving different waste treatment
Table 4 Waste allocation to technologies. Food LFGRS Incineration Composting MRF U U U Yard U U U Paper U U U U Plastic U U Glass U U Metal U U Textile U U U

Low Carbon Society Blueprint for Iskandar Malaysia 2025, 2012. Tchobanoglous et al., 1993.

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005

S.T. Tan et al. / Journal of Cleaner Production xxx (2013) 1e11 Table 5 Cost analysis and emission rate of technologies (EIA, 2010). Capital cost USD/t 500 800 250 200 O&M cost USD/t 0.4 0.6 0.2 0.3 Variable cost USD/t 1.3 2.0 0.5 0.8 Operation time d/y 360 292 292 292 GHG emissions kCO2/t 0.2 0.5 0.05 3.6 Table 7 Feed-in tariff for RE from MSW in Malaysia (Pusat Tenaga Malaysia, 2010). RE utilisation Biomass <10 MW 10 MW, 20 MW 20 MW, 30 MW Bonus for MSW Biogas <4 MW 4 MW, 10 MW 10 MW, 30 MW Bonus for landll Year 16 16 16 16 16 16 16 16 RM/kWh 0.31 0.29 0.27 0.10 0.32 0.30 0.28 0.08 Degression (%) 0.5 0.5 0.5 1.8 0.5 0.5 0.5 0.5

LRGRS Incineration Composting MRF

technologies are presented and discussed. Sensitivity analyses were performed based on the optimal solution to evaluate the impact of uctuations in the product price, product demand and GHG reduction demand on the system conguration with regards to the protability of the MSWM system. The data were input to the MILP model for the MSWM system and optimised with the CPLEX solver (version 12.3) from GAMS software (version 22.9) (GAMS Development Corporation, 2013). 4.1. Comparison of different waste management scenarios Table 9 shows the analysis of the MSWM system for all four different scenarios in IM, namely BAU, WTE, WTR, and MIXTECH. The BAU scenario assumed that there was no introduction of alternative waste treatment technologies and maintained the current practices of MSWM. Under the BAU scenario, 93% of waste generated in IM was directly allocated to the landll site, and the remainder (7%) was recycled by the MRF. Meanwhile, there was no energy or fertiliser production from the waste to achieve the maximum value of waste utilisation. A negative annual prot was predicted in the BAU scenario due to the high maintenance and operation cost for managing the landlls without any revenue generated from any value-added products from the waste. The BAU scenario also produces the highest carbon emission (3.285 Mt CO2 eq/y) among all four scenarios. The WTE scenario was designed to anticipate a MSWM system that considered two major WTE treatment technologies, namely the LFGRS and waste incineration. A total of USD 276.52 M/y of net prot was obtained for the MSWM system under the WTE scenario, where 15% and 59% of waste was allocated for incineration and LFGRS respectively. Only 27% of the waste was recommended for recycling by the model. The energy production from the WTE scenario was extremely high (8594.13 GWh) compared to the targeted renewable energy (RE) demand of IM (2285.71 GWh) (Blueprint for Iskandar Malaysia, 2010b). The overproduction of energy in the form of electricity from MSWM increased the investment cost (USD 7692.86 M/y) and resulted in energy waste. Nonetheless, the WTE scenario showed the highest environmental protection with the lowest carbon emission (0.195 Mt CO2 eq/y) among all scenarios. As an alternative to landlling, maximal recycling and composting capabilities were introduced under the WTR scenario without the consideration of WTE. The lowest positive prot (USD 34.67 M/y) was predicted under the WTR scenario compared to the
Table 6 Selling price of by-product. Price Electricity, USD/kWh Compost, USD/t Paper, USD/t Plastic, USD/t Glass/Ceramic, USD/t Metal, USD/t Textile, USD/t 124.40 153.37 38.16 204.16 45.08 229.01 45.08 Reference Hashim and Ho, 2011 Rodionov and Nakata, 2011 MHLG, 2012 MHLG, 2012 MHLG, 2012 MHLG, 2012 MHLG, 2012

WTE and MIXTECH scenarios because the selling of by-products from recycling and composting alternatives was not nancially favourable compared to energy as a by-product. Waste (98%) was suggested to be recycled (56% to MRF, 42% to composting), while only 2% of unrecyclable waste was suggested to be landlled. To achieve the optimal solution of MSWM that would full the targets of recycling and RE in IM, the MIXTECH scenario was proposed. The optimisation result from Table 9 shows that the optimal prot of USD 101.85 M/y was achieved annually under the MIXTECH scenario. The MIXTECH scenario suggested a best mix of waste utilisation technologies to be implemented in IM with the maximum net prot and without exceeding the product demand. The percentage of waste allocation were recommended to be 3, 14, 56, and 27 for incineration, LFGRS, MRF and composting, respectively. The MIXTECH scenario promoted waste recycling and composting due to their lower investment cost while introducing WTE technology (LFGRS and incineration) to full the RE target for IM. The LFGRS exhibited good potential to generate sufcient energy for the IM region. Waste incineration was less attractive due to its high investment cost. The MIXTECH scenario achieved the RE target of IM (2285.71 GWh/y) with lowered GHG emissions (0.992 Mt CO2 eq/y) compared to the other scenarios. 4.2. Optimal planning under MIXTECH scenario The MIXTECH scenario successfully planned an optimal solution for the MSWM system as proposed for the IM region. Three main objectives for effective and sustainable MSWM planning were achieved under the MIXTECH scenario, these included a best mix of waste utilisation technology with i) maximising the prot of the MSWM system, ii) achieving the RE target demand, and iii) promoting recycling and composting. The optimiser suggested a combination of incineration, LFGRS, MRFs and composting to be implemented in IM. The MSWM planning for the period of year 2013 to year 2025 is presented in Fig. 3. As indicated in Fig. 3, the earliest construction of waste treatment technologies would begin in 2012. The results suggested that an incineration power plant of 10 Mt/y should be constructed by 2012 in IM to achieve the RE target. In addition, two LFG plants should be constructed in 2013 and 2018 with capacities of 45 Mt/y and 50 Mt/y. In addition, three recycling plants were suggested to be constructed in 2013, 2014, and 2015 with capacities of 40 Mt/y, 40 Mt/y and 22 Mt/y, respectively. Composting plants with capacities of 34 Mt/y and 36 Mt/y were recommended to be constructed by 2013 and 2020 as the landll sites would become limited due to rapid development of IM in year 2013 and beyond.
Table 8 RE demand from MSW in IM (Blueprint for Iskandar Malaysia, 2010b). Key targets for RE RE from MSW (MW) 2010 e 2015 25 2020 50 2025 50

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005

S.T. Tan et al. / Journal of Cleaner Production xxx (2013) 1e11

Table 9 Analysis of four scenarios for MSWM system in Iskandar Malaysia (IM). Unit BAU WTE WTR MIXTECH
Net profit

Sensitivity analyses of RE demand

Financial planning Net prot M USD/y 43.09 474.04 7692.86 8164.29 0.1100 0.195 9594.13 15 n/a 59 27 n/a 35 n/a 58 7 n/a 34.67 233.37 268.04 0.0034 1.459 250.30 0 n/a 2 56 42 0 n/a 1 9 90 101.85 1787.14 1889.29 0.0400 0.992 2285.71 3 n/a 14 56 27 6 n/a 23 3 67

Total system cost M USD/y 58.53 Total revenue M USD/y 15.44 Unit treatment cost M USD/t 0.0015 3.285 GHG emission Mt CO2 eq/y Energy production GWh/y n/a Percentage of waste allocation to technologies Incineration % n/a Landll % 93 LFGRS % n/a MRF % 7 Composting % n/a GHG emission percentage Incineration % n/a Landll % 98 LFGRS % n/a MRF % 2 Composting % n/a

Total revenue Total system cost Unit treatment cost Compost production GHG Emission Energy production -80% -60% -40% -20% 0% 20% 40% 60%

Change of energy demand 50% 30% 20% -20% -30% -50%

Change in overall economic potential of MSWM

Fig. 4. Sensitivity analyses for RE demand on the overall costing and GHG emissions of the MSWM system.

4.3. Sensitivity analyses Sensitivity analyses were conducted to observe the effects of energy demand and the GHG emission reduction target on the conguration of the MSWM system in terms of waste allocation to each technology. As MIXTECH scenario was found to be the best scenario in this study, sensitivity analyses were conducted solely on this scenario with respect to the change of the RE target and the reduction of GHG emission target. 4.4. Sensitivity analyses on the RE target One of the main products of the MSWM system was RE production in the form of electricity. Fluctuation in energy demands from MSW was expected to affect the conguration, protability and GHG emissions of the MSWM system under the optimal scenario. Sensitivity analyses were conducted on the MIXTECH scenario by adjusting the energy demand within 0e50% increments and decrements (i.e., 20%, 30% 50%, 20%, 30%, and 50%). Fig. 4 shows the change of RE target towards the protability and GHG emissions of the MSWM system. The increase in energy demand of 20%, 30% and 50% provided a positive impact on the overall economic potential of the MSWM system in terms of net prot, total system cost, unit treatment cost and total revenue, and vice versa. However, the net prot did not necessarily follow the incremental trend as at some point, a decrease of net prot was observed with higher energy demand (30% to 50%). This was mainly due to the increase in investment or capital cost of the WTE technology at a larger capacity. The change in energy demand also tended to affect

the GHG emissions. The increase in energy demand by 30% and 50% would reduce the GHG emissions to 25% and 14%, because both of the WTE technologies (LFG and incineration) were implemented with methane gas or energy recovery, which emitted less GHG to the environment compared to the WTR technologies that did not consider methane gas recovery (i.e., during anaerobic composting). As the RE demand decreased to a lower level (i.e., 20%, 20%, 30%, and 50%), an increment of GHG emissions was observed because the waste utilisation by WTR technologies increased when the RE demand decreased. Composting was found to be the major contributor to GHG emissions among all four waste treatment technologies proposed for the system. The composting process which converted the MSW to compost through anaerobic digestion released methane gas as the by-product. The methane gas was not captured for reuse during composting in this study. Consequently, the analyses tended to favour the behaviour of the WTE process due to reduced GHG emissions. Fig. 5 presents the sensitivity analyses correlating energy demand with the change in waste utilisation in term of percentage (%) of waste allocation to each waste treatment technology. As energy

Sensitivity analyses of energy demand on waste utilisation


Percentage of wasre allocation to technologies (%) 100 90 80 70 60 50 40 30 20 10 0 -50% -30% -20% 0% 20% change of energy demand Landfill gas capture Recycling 30% 50%

60 Capacity of plant (Mt) 50 40 30 20 10 0 Incineration LFG Reocvery system MRF Composting plant

Incineration

Composting

Year
Fig. 3. MSWM system under the MIXTECH scenario. Fig. 5. Sensitivity analyses of RE demand for different percentage of waste allocation to the MSWM system.

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005

S.T. Tan et al. / Journal of Cleaner Production xxx (2013) 1e11

Sensitivity analyses of GHG emission

Net profit Change of GHG emission reduction 60% 40% Compost production -350%-300%-250%-200%-150%-100% -50% 0% 50% Change on overall economic potential of MSWM
Fig. 6. Sensitivity analyses for GHG emission reduction on the overall economic potential and the production of compost in the MSWM system.

Total revenue

Total system cost

20%

demand increased, the percentage of waste utilisation by WTE technologies would increase. Under the MIXTECH scenario, as the energy demand increased from the base level ((i.e. 0% or no change) to 30%, waste allocation to LFGRS increased from 14% to 18% while incinerated waste decreased. As the energy demand increased to 50%, the percentage of waste allocated to incineration increased to 20% while landlled waste decreased to 4%. These results signicantly illustrated that the type and capacity of waste allocation could signicantly inuence the cost for MSWM. As energy demand changes within 0e50%, only three technologies (incineration, LFGRS and composting) would result in signicant change of the percentages of waste allocation, recycling is least inuenced by the energy demand. 4.5. Sensitivity analyses on the target of GHG emission reduction To examine the effect of the GHG emission reduction target on the system conguration and protability, a sensitivity analysis was conducted based on the optimal result of the MIXTECH by adjusting the GHG emission reduction targets with increments of 20%, 40%, and 60% from the baseline year of 2005. Referring to Fig. 6, as the GHG reduction target increased from 20% to 40%, no signicant change was observed in the overall economic potential in terms of the total system cost, total revenue, and the net prot because the GHG emissions generated under the MIXTECH scenario were already in the range of lower than 40% of the reduction target in

GHG emissions compared to the baseline. As the reduction target of GHG emissions increased to 60% compared to the BAU scenario, a negative impact of 300% to net prot was observed for the system. A high percentage of the GHG reduction target would require the integration of large portions of WTE energy technology. This integration signicantly increased the investment cost and hence overweighed the increase in the revenue generated. The increment in the GHG reduction target also tended to reduce the production of compost in the MSWM system. When the GHG emission targets increase by 20%, 40%, and 60% from the baseline, the production of compost would be reduced by 17%, 33%, and 42%, respectively. This result indicated that the composting technology emitted higher levels of GHG through the anaerobic digestion process without the recovery of methane. Fig. 7 presents the conguration of waste utilisation for the sensitivity analysis of GHG emission reduction. At the lower level of GHG emission reduction targets (20% and 40%), there was no signicant impact on the conguration of the waste allocation. As the GHG emission reduction target increased to 60%, the percentage of waste utilisation by WTE technologies would increase by 18% for incineration and 9% for LFGRS. In addition, the allocation of waste for composting would decrease from 27% to 17%. This result illustrated again that the composting technology was the key contributor to GHG emissions. 5. Conclusions A multi-period MILP model for optimising the MSWM system was developed for the IM region in this study. This study indicated that waste treatment technologies including incineration, LFGRS, composting, and MRFs could provide attractive economic benets and RE options compared to the existing MSWM system in the IM region. The developed model was tested using different scenario settings to analyse the economic feasibility and potential of an effective MSWM system. The model with the MIXTECH scenario emphasised a cost-effective waste processing network that could provide a maximum net prot of USD 101.85 M/y in IM. The best mix of waste utilisation technologies in terms of % of waste allocation to the following technologies were LFGRS (14%), incineration (3%), recycling (56%) and composting (27%). The best mix of technology selection would be able to achieve the RE target and the recycling target and promote composting as an improved waste reduction strategy for the studied region. The results of the sensitivity analyses explained that the technology selection of the MSWM system was highly inuenced by the costs of technologies, product (RE) targets and GHG emission reduction targets. This model could be extended to include the costs of land area for LFG recovery and composting plants, as both technologies could be economically benecial but may be restricted by the availability of land in a country. Moreover, the cost of transporting waste to the processing plants, the variety of waste treatment technologies and products, environmental factors and the locations of waste treatment plants should be considered in the future. Acknowledgements The authors gratefully acknowledge the Ministry of Higher Education (MOHE) and University Teknologi Malaysia (UTM) for providing the research grant under Vote No. Q.JI3.2525.01H52. The authors also acknowledge the Japan International Cooperation Agency (JICA) under the scheme of Science and Technology Research Partnership for Sustainable Development (SATREPS) for the project entitled Development of Low Carbon Scenarios for Asian Region.

Sensitivity analyses of GHG emission reduction on waste utilisation


Percentage of waste allocation to technologies (%) 100 90 80 70 60 50 40 30 20 10 0 0% 20% 40% Change of GHG emission reduction Landfill gas capture Recycling 60%

Incineration

Composting

Fig. 7. Sensitivity analyses for the change of GHG emission reduction target on the percentage of waste allocation in the MSWM system.

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005

10

S.T. Tan et al. / Journal of Cleaner Production xxx (2013) 1e11

Nomenclature Acronyms BAU Business as usual scenario GAMS General Algebraic Modeling System GHG Greenhouse gas IM Iskandar Malaysia LFGRS Landll gas recovery system LP Linear programming MILP Mixed integer linear programming MIXTECH Mixed technologies scenario MOP Multi-objective programming MRF Material recycling facilities MSW Municipal solid waste MSWM Municipal solid waste management NLP Non-linear programming O&M Operation and maintenance RE Renewable energy RM Ringgit Malaysia SWM Solid waste management USD United State Dollar WTE Waste-to-energy scenario WTR Waste-to-recycling scenario Sets i p t z

YPpz

Binary decision variable for purchasing technology p with capacity z

References
Ahmad, S., Kadir, M.Z.A.A., Shae, S., 2011. Current perspective of the renewable energy development in Malaysia. Renew. Sustain. Energy Rev. 15, 897e904. Badran, M.F., El-Haggar, S.M., 2006. Optimization of municipal solid waste management in Port Said e Egypt. Waste Manag. 26, 534e545. Blueprint for Iskandar Malaysia, 2010a. Integrated Solid Waste Management Blueprint of Iskandar Malaysia. Johor Bahru, Malaysia. www.iskandarmalaysia.com. my/blueprint-for-iskandar-malaysia (accessed 01.12.13.). Blueprint for Iskandar Malaysia, 2010b. Renewable Energy and Energy Efciency Blueprint for Iskandar Malaysia. Johor Bahru, Malaysia. www.iskandarmalaysia. com.my/blueprint-for-iskandar-malaysia (accessed 01.12.13.). Chang, Y.H., Chang, N.-B., 1998. Optimization analysis for the development of shortteam solid waste management strategies using presorting process prior to incinerators. Resour. Conserv. Recycl. Volume 24 (1), 7e32. Chang, N.-B., Chen, Y.L., Wang, S.F., 1997. A fuzzy interval multiobjective mixed integer programming approach for the optimal planning of solid waste management systems. Fuzzy Sets Syst. 89 (1), 35e60. Chang, N.-B., Qi, C., Islam, K., Hossain, F., 2012. Comparisons between global warming potential and costebenet criteria for optimal planning of a municipal solid waste management system. J. Clean. Prod. 20 (1), 1e13. http://dx.doi.org/ 10.1016/j.jclepro.2011.08.017. Chen, Y.-T., Chang, D.-S., 2010. Diffusion effect and learning effect: an examination on MSW recycling. J. Clean. Prod. 18 (5), 496e503. Dai, C., Li, Y.P., Huang, G.H., 2011. A two-stage support-vector-regression optimization model for municipal solid waste management e a case study of Beijing, China. J. Environ. Manag. 92 (12), 3023e3037. Energy Information Administration (EIA), 2010. Updated Capacity Cost Estimates for Electricity Generation Plants (Washington, USA). Finnveden, G., Johansson, J., Lind, P., Moberg, A., 2005. Life cycle assessment of energy from solid waste - Part 1: general methodology and results. J. Clean. Prod. 13, 213e229. Fodor, Z., Kleme s, J.J., 2012. Waste as alternative fuel e minimising emissions and efuents by advanced design. Process Safety Environ. Prot. 90 (3), 263e284. GAMS Development Corporation, 2013. GAMS: General Algebraic Modeling System (Version 22. 9). Washington, USA. www.gams.com (accessed 11.11.2013). Guo, P., Huang, G.H., 2009. Inexact fuzzy-stochastic mixed-integer programming approach for long-term planning of waste management e Part A: Methodology. J. Environ. Manag. 91 (2), 461e470. Hashim, H., Ho, W.S., 2011. Renewable energy policies and initiatives for a sustainable energy future in Malaysia. Renew. Sustain. Energy Rev. 15, 4780e4787. Iskandar Malaysia, 2013. Five Year Progress Report. Johor Bahru, Malaysia. www. iskandarmalaysia.com.my/iskandar-malaysia-brochures (accessed 01.12.2013). Johari, A., Ahmed, S., Hashim, H., Alkali, H., Ramli, M., 2012. Economic and environmental benets of landll gas from municipal solid waste in Malaysia. Renew. Sustain. Energy Rev. 16 (5), 2907e2912. Juul, N., Mnster, M., Ravn, H., Sderman, M.L., 2013. Challenges when performing economic optimization of waste treatment: a review. Waste Manag. 33 (9), 1918e1925. Li, P., Chen, B., 2011. FSILP: fuzzy-stochastic-interval linear programming for supporting municipal solid waste management. J. Environ. Manag. 92 (4), 1198e1209. Liamsanguan, C., Gheewala, S.H., 2008. The holistic impact of integrated solid waste management on greenhouse gas emissions in Phuket. J. Clean. Prod. 16 (17), 1865e1871. Low Carbon Society Blueprint for Iskandar Malaysia 2025, 2012. Johor Bahru, Malaysia. 2050.nies.go.jp/cop/cop18/SPM_LCS%20Blueprint_Iskandar% 20Malaysia.pdf (accessed 01.12.13.). Ministry of Housing and Local Government, 2005. National Strategic Plan for Solid Waste Management (Putrajaya, Malaysia). Ministry of Housing and Local Government (MHLG), 2012. Laboratory of Solid Waste Management. Putrajaya, Malaysia. www.kpkt.gov.my/kpkt/leupload/ hebahan/lab_sisa_pepejal.pdf (accessed 07.07.2013). Mnster, M., Meibom, P., 2010. Long-term affected energy production of waste to energy technologies identied by use of energy system analysis. Waste Manag. 30 (12), 2510e2519. Mnster, M., Meibom, P., 2011. Optimization of use of waste in the future energy system. Energy 36 (3), 1612e1622. Ng, W.P.Q., Varbanov, P.S., Kleme s, J.J., Hegyhati, M., Bertok, B., Heckl, I., Lam, H.L., 2013. Waste to energy for small cities: economics versus carbon footprint. Chem. Eng. Trans. 35, 889e894. http://dx.doi.org/10.3303/CET1335148. Othman, S.N., Noor, Z.Z., Abba, A.H., Yusuf, R.O., Hassan, M.A.A., 2013. Review on life cycle assessment of integrated solid waste management in some Asian countries. J. Clean. Prod. 41, 251e262. Periathamby, A., Hamid, F., Khidzir, K., 2009. Evolution of solid waste management in Malaysia: impacts and implications of the solid waste bill, 2007. J. Mater. Cycles Waste Manag. 11 (2), 96e103. Pires, A., Martinho, G., Chang, N.-B., 2011. Solid waste management in European countries: a review of systems analysis techniques. J. Environ. Manag. 92 (4), 1033e1050. http://dx.doi.org/10.1016/j.jenvman.2010.11.024.

Resource/product Process/technology Period Capacity of plant

Parameter ACPCOSTpzt Annualised capital cost of technology p with capacity z at period t (USD/t) CERT Carbon emission reduction target for the system EFp Emission factor for process p. MPSMip Material process selection matrix PRCMpi Process resource conversion matrix PRICEit Unit price of product i in period t (USD/unit) PRODEMit Product demand for product i at time t (units/y) RESit Amount of resource i in the system during time period t (t/y) UPCOSTpt Unit processing cost of process p in period t (USD/unit) UVCostpt Unit variable cost of the corresponding process p in time t (USD/unit) Variables CAPpz CCOST CEPpt EXRECit MATipt PCOST PRESpt PROit PROFIT REV SGRESipt VCOST YOPpzt

Capacity of the process p with size z Capital cost (USD) Carbon emission from process p in period t (t/y) Input rate of external resource i during period t (t/y) Input rate for material i into process p during period t (units/y) Total processing cost of the resource (USD) Quantity of the resource being processed by process p at period t (units/y) Production rate of product i during period t (units/y) Overall prot (USD) Product revenues of the MSWM system (USD) Quantity of resource i generated within the system under process p through period t (t/y) Total variable operating and maintenance cost of the system (USD) Binary decision variable to decide whether the process p should be operated at size z during period t

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005

S.T. Tan et al. / Journal of Cleaner Production xxx (2013) 1e11 Poeschl, M., Ward, S., Owende, P., 2012a. Environmental impacts of biogas deployment e Part I: life cycle inventory for evaluation of production process emissions to air. J. Clean. Prod. 24 (1), 168e183. Poeschl, M., Ward, S., Owende, P., 2012b. Environmental impacts of biogas deployment e Part II: life cycle assessment of multiple production and utilization pathways. J. Clean. Prod. 24 (1), 184e201. Pusat Tenaga Malaysia, 2010. Malaysias 2011 Proposed Solar, Biomass, Biogas and Hydro Tariffs (Seri Kembangan, Malaysia). Rathi, S., 2007. Optimization model for integrated municipal solid waste management in Mumbai, India. Environ. Dev. Econ. 12, 105e121. Rodionov, M., Nakata, T., 2011. Design of an optimal waste utilization system: a case study in St. Petersburg, Russia. Sustainability 3, 1486e1509. Salvia, M., Cosmi, C., Macchiato, M., Mangiamele, L., 2002. Waste management system optimisation for Southern Italy with MARKAL model. Resour. Conserv. Recycl. 34 (2), 91e106. Santibaez-Aguilar, J.E., Ponce-Ortega, J.M., Betzabe Gonzlez-Campos, J., SernaGonzlez, M., El-Halwagi, M.M., 2013. Optimal planning for the sustainable utilization of municipal solid waste. Waste Manag. 33 (12), 2607e2622.

11

Seadon, J.K., 2010. Sustainable waste management systems. J. Clean. Prod. 18 (16e 17), 1639e1651. Shadiya, O.O., Satish, V., High, K.A., 2012. Process enhancement through waste minimization and multiobjective optimization. J. Clean. Prod. 31 (1), 137e149. Tchobanoglous, G., Theisen, H., Vigil, S.A., 1993. Integrated Solid Waste Management e Engineering Principles and Management Issues. McGraw-Hill, New York, USA. Tchobanoglous, G., Kreith, F., 2002. Handbook of Solid Waste Management, second ed. McGraw-Hill, New York, USA. Wanichpongpan, W., Gheewala, S.H., 2007. Life cycle assessment as a decision support tool for landll gas-to energy projects. J. Clean. Prod. 15 (18), 1819e 1826. Xu, Y., Huang, G.H., Qin, X.S., Cao, M.F., Sun, Y., 2010. An interval-parameter stochastic robust optimization model for supporting municipal solid waste management under uncertainty. Waste Manag. 30 (2), 316e327. Yeh, C.-H., Xu, Y., 2013. Sustainable planning of e-waste recycling activities using fuzzy multicriteria decision making. J. Clean. Prod. 52, 194e204.

Please cite this article in press as: Tan, S.T., et al., Optimal process network for municipal solid waste management in Iskandar Malaysia, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.12.005