Pet

(1)

Comparative assessment of the environmental pro

�

le of PLA and PET

drinking water bottles from a life cycle perspective

Seksan Papong

aa

,, P

Pomthong Malakul

omthong Malakul

aa,,bb,,

*

, , R

R u

uethai

ethai Trungka

Trungkavashir

vashira

akun

kun

aa

,,

Pechda

Pechda Wen

Wenunun

unun

aa

, Tassaneewan Chom-in

aa

, Manit Nithitanakul

bb

, Ed Sarobol

cc a

a_{National Metal and Materials Technology Center, Thailand Science Park, Pathumthani, Thailand}_{National Metal and Materials Technology Center, Thailand Science Park, Pathumthani, Thailand} b

b_{The Petroleum and Petrochemical College, Chulalongkorn University, Patumwan, Bangkok, Thailand}_{The Petroleum and Petrochemical College, Chulalongkorn University, Patumwan, Bangkok, Thailand} cc_{Department of Agronomy, Faculty of Agriculture, Kasetsart University, Chatujak, Bangkok, Thailand}_{Department of Agronomy, Faculty of Agriculture, Kasetsart University, Chatujak, Bangkok, Thailand}

a r t a r t i c l i c l e e i n f oi n f o Article history: Article history: Received 11 March 2013 Received 11 March 2013 Received in revised form Received in revised form 19 September 2013 19 September 2013 Accepted 20 September 2013 Accepted 20 September 2013 Available online xxx Available online xxx Keywords: Keywords:

Life cycle assessment Life cycle assessment Polylactic acid (PLA) Polylactic acid (PLA) Cassava

Cassava

Polyethylene terephthalate (PET) Polyethylene terephthalate (PET)

a b s t r a c t a b s t r a c t

Bioplastic polymer that is produced from cassava has been considered the most promising alternative to Bioplastic polymer that is produced from cassava has been considered the most promising alternative to conventional plastics as there is an abundant renewable resource in Thailand. The objective of this study conventional plastics as there is an abundant renewable resource in Thailand. The objective of this study was to analyze the life cycle environmental performance of polylactic acid (PLA) drinking water bottles was to analyze the life cycle environmental performance of polylactic acid (PLA) drinking water bottles produ

produced in ced in ThailanThailand d with an with an emphasemphasis on is on differendifferent end-of-life scenariost end-of-life scenarios. . The functional unit was The functional unit was set atset at 1000 units of 250-ml drinking water bottles. The system boundary of the study covered all stages in the 1000 units of 250-ml drinking water bottles. The system boundary of the study covered all stages in the life

life cycle, including cycle, including cultivcultivation and ation and harvesharvesting, cassava starch ting, cassava starch produproduction, transporction, transportation, glucose tation, glucose propro-duction, the polymerization process to produce PLA resin, PLA bottles propro-duction, and disposal process. duction, the polymerization process to produce PLA resin, PLA bottles production, and disposal process. The input

The inputeeoutput data included the use of resources (water, chemicals, materials), energy (electricity,output data included the use of resources (water, chemicals, materials), energy (electricity, fuels), and all emissions based on the functional unit. The life cycle environmental performance of PLA fuels), and all emissions based on the functional unit. The life cycle environmental performance of PLA drinking water bottles was compared with that of polyethylene terephthalate (PET) bottles for the same drinking water bottles was compared with that of polyethylene terephthalate (PET) bottles for the same functional unit. The global warming potential, fossil energy demand, acidi

functional unit. The global warming potential, fossil energy demand, acidi��cation, eutrophication, andcation, eutrophication, and human toxicity were selected in the analysis. The results obtained in this study showed that the human toxicity were selected in the analysis. The results obtained in this study showed that the envi-ronmental performance of cassava-based PLA bottles was better than PET bottles in terms of global ronmental performance of cassava-based PLA bottles was better than PET bottles in terms of global warmi

warming, reduction of ng, reduction of dependdependency on ency on fossil energyfossil energy, , and human and human toxitoxicity. In city. In additaddition, it ion, it was shown thatwas shown that improv

improving cassava starch process by combining with ing cassava starch process by combining with biogas producbiogas production and tion and utilizutilization will lead ation will lead to sig-to sig-ni

ni��cant reduction in global warming potential and eutrophication potential.cant reduction in global warming potential and eutrophication potential.



1. Introduction 1. Introduction

Nowadays

Nowadays, several million tonnes , several million tonnes of plastics are of plastics are produceproduced everyd every year. Plastics can be found in everything from clothing to year. Plastics can be found in everything from clothing to machin-ery. Pl

ery. Plastics are used foastics are used for packaging materir packaging materials and almostals and almost everyevery typetype of

of coconsunsumer mer prprodoductuct, , and and thuthus s the the coconsunsumpmptiotion n of of plplastasticicss continue to rise at

continue to rise at an increasing rate. Virtually all plastics are madean increasing rate. Virtually all plastics are made from petroleum resources, such as oil, coal or natural gas, which from petroleum resources, such as oil, coal or natural gas, which will eventually become exhausted and it may take thousands of will eventually become exhausted and it may take thousands of year

years s for plasticfor plastics s to be to be biodbiodegraegraded (ded (NampNampootoothiri hiri et et al., al., 20120100).). Rene

Renewablwable e matematerialrials s are are matematerialrials s from from naturnatural al resoresourcurces es oror

natural biomass resources such as corn starch, cellulose, cassava natural biomass resources such as corn starch, cellulose, cassava and sugarcane (

and sugarcane (Detzel and Kruger, 2006Detzel and Kruger, 2006;; NIA, 2008 NIA, 2008). Bio-based). Bio-based materials are considered an environmental friendly alternative to materials are considered an environmental friendly alternative to petroleum-based materials. They can be produced without toxic petroleum-based materials. They can be produced without toxic by-products and are biodegradable in nature. In addition, the net by-products and are biodegradable in nature. In addition, the net bal

balancancee ofof carcarbondioxbondioxididee ofof biobiopolpolymeymersrs isis neuneutratrall becbecausausee thethe COCO22

re

relealeased sed dudurinring g the the prproduoductiction on and and disdisposposal al of of biobioplplastasticics s isis balanced by the CO

balanced by the CO22 consumed during plant growth ( consumed during plant growth (Gironi andGironi and

Piemonte, 2011; Uihlein et al., 2008

Piemonte, 2011; Uihlein et al., 2008).).

Ther

There e are are seveseveral ral renerenewabwable-bale-based sed polpolymerymers s or or biopbiopolymolymersers bei

beingng prprododuceucedd witwithh anan aimaim toto minminimiimizeze thethe envenvirironmonmententalal impimpactact produce

produced fd from conventional plastics made rom conventional plastics made from non-renewablefrom non-renewable resource. Biodegradable plastics are very important from an resource. Biodegradable plastics are very important from an envi-ronmental friendly point of view to reduce the impact at the ronmental friendly point of view to reduce the impact at the end-of-life (EOL) phase, because in the best case scenario it is possible of-life (EOL) phase, because in the best case scenario it is possible to recover the energy (combustion) or biomass resources (

to recover the energy (combustion) or biomass resources ( GrigaleGrigale

et al., 2010

et al., 2010). However, at EOL of biopolymers the fate and chemi-). However, at EOL of biopolymers the fate and chemi-cal behavior

cal behavior are not well documented and are believeare not well documented and are believed to be highlyd to be highly va

variariable ble in in the the popotententiatial l impimpactacts s ((BoydBoyd, , 20120111). ). StarStarch-bch-basedased

Abbreviations:

Abbreviations: LCA, Life cycle assessment; PLA, Polylactic acid; PET, Polyethylene LCA, Life cycle assessment; PLA, Polylactic acid; PET, Polyethylene terep

terephthahthalate; late; GWPGWP, , GlobaGlobal l warmwarming ing potenpotential; tial; APAP, , AcidiAcidi��catiocation n potenpotential; tial; EPEP,, Eutrophication potential; HTP, Human toxicity potential.

Eutrophication potential; HTP, Human toxicity potential.

*

* CorrCorrespondesponding ing authoauthor. r. NatioNational nal Metal Metal and and MaterMaterials ials TecTechnolohnology gy CenterCenter,,

Thailand Science Park, Pathumthani, Thailand. Tel.:

Thailand Science Park, Pathumthani, Thailand. Tel.: þþ_{66 2 218 4117; fax:}_{66 2 218 4117; fax:}þþ_{66 2 215}_{66 2 215}

4459. 4459.

E-mail address:

E-mail address: pomt [email protected]@chula.ac.thc.th (P. Malakul). (P. Malakul).

Contents lists available at

Contents lists available at ScienceDirect ScienceDirect

Journal

Journal of

of Cleaner

Cleaner Production

Production

j o u r n a l h o m e p a g e :

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . cw w w . e l s e v i e r . co m / l o c a t e / j c l e p r oo m / l o c a t e / j c l e p r o

0959-6526/$

(2)

polymers were the �rst important group to become commercial products. They are used as a raw material in �lm production and blended with petroleum-based polymers to reduce cost and enhance biodegradability. In addition, polylactic acid (PLA) is pre-pared from lactic acid and is one of the most promising products for packaging applications.

Life Cycle Assessment (LCA) is a useful tool for evaluating and quantifying the energy and environmental consequences associ-ated with a product, process, or service ( ISO, 2006a). Even though LCA is an established method, results may differ depending on the scope, system boundary, country and time. Several research studies have shown that a biopolymer (such as PLA) generates a lower carbon and lower fossil energy consumption than fossil-based polymers such as polyethylene terephthalate (PET), polystyrene (PS) and polypropylene (PP) (Detzel and Kruger, 2006; Vink et al., 2007; Uihlein et al., 2008; Madival et al., 2009; Groot and Borén, 2010; Gironi and Piemonte, 2011). However, there is disagree-ment regarding the life cycle impact of biopolymers as some impact categories indicate biopolymers may have a more negative impact on the environment than conventional plastics due to their weight and their production methods. Such as, a study by Tabone et al. (2010), using LCA methodology, indicated that while some bio-polymers have a lower impact on fossil fuel consumption and greenhouse gas (GHG) emissions than conventional polymers, they could have higher environmental impacts in terms of eutrophica-tion, carcinogens, and ozone layer depletion. This is due to the use of fertilizers, pesticides and the land use change required for the increased agriculture production.

PLA is a sustainable alternative to conventional polymers, because the lactides can be mass produced by the microbial fermentation of agricultural by-products, mainly carbohydrate rich substances ( John et al., 2006). Recent developments show that lactic acid can be converted to polylactic acid through two main routes: �rst, the indirect route via lactide, and second, direct polymerization by polycondensation, producing PLA. Both products are generally referred to as PLA (Wolf, 2005).

This paper aims to evaluate the environmental performance associated with PLA bottles produced from cassava in Thailand in comparison with PET bottles, based on the life cycle approach. The life cycle inventory analysis and impact assessment were carried out based on ISO 14040 for all stages involved in the product sys-tems, which included cassava cultivation and harvesting, starch production, lactic acid production and PLA resin conversion, plastic bottles production, transportation, and disposal.

2. Methodology

The LCA technique used in this study was based on ISO 14040 framework (ISO 2006a) and ISO 14044eguidelines and re-quirements (ISO 2006b), which consist of four steps;goal and scope de�nition, inventory analysis, impact assessment, and interpretation.

2.1. Goal and scope de �nition

The �rst step of an LCA is de�ning the scope and goal of an investigation, which can be established on the analysis and un-derstanding of a product’s life cycle, the improvement of

produc-tion processes, or the use of the results for marketing purposes. The goal of this study was to assess the life cycle environmental per-formance of drinking water bottles made of polylactic acid pro-duced from cassava in comparison with similar PET bottles produced in Thailand. The functional unit (FU) of this study was 1000 units of 250-ml drinking water bottles. The scope of the PLA study includes the cassava cultivation and harvesting, starch

production, glucose production, production of lactic acid, lactides and PLA, water bottle production, and disposal. The system boundary of the PLA system is shown in Fig. 1.

The bio-based polymeric resins were compared on an equal weight basis with petroleum-based resins whereas the bio-based product (drinking water bottle) was compared based on the func-tional unit (1000 bottles). The environmental pro�les of the petroleum-based polymers were gathered from the national life cycle inventory database of Thailand, which represents an average of production sites in Thailand. The system boundary of the PET system is shown in Fig. 2.

2.2. Data sources, assumptions, and limitations

In this study most of the input eoutput data were collected as primary data at the actual sites in Thailand, including cassava plantations and harvesting, cassava starch production, and bottle production plants. The collected data included raw materials used, energy consumption, utilities, and waste generated within the system boundary. The secondary data were used in this study as necessary and were obtained from literature, calculations, the Ecoinvent database and the IPCC method for items such as the production of fertilizers, herbicides, etc. However, this study did not take into account CO2 uptake during the cassava growing for

glucose requirements, nor did it include the impacts of infrastruc-ture such as construction of the process plant, equipment mainte-nance, etc. The background data for this study were gathered from the national life cycle inventory (LCI) databases of Thailand (MTEC, 2011), research reports (MTEC, 2009; DEDE, 2012), and Ecoinvent (2008) databases as described in Table 1.

2.3. Inventory analysis

The inventory data were gathered which include the material and energy inputs, air emissions, waterborne emissions, and solid waste involved in the life cycle of the cassava-based PLA and PET product. All data of the processes were complied and the inventory analysis was performed based on a functional unit of product. De-tails of each stage are described in the following sections.

2.3.1. Cassava cultivation and harvesting stage

The main concentration of the cassava planting is now found in the northeast of Thailand, especially in Nakhonratchasima prov-ince. Cassava has excellent drought tolerance properties and can be planted in almostall soil types.It is mostly grown by a large number of farmers, who own small plots of land. Few organized large-scale plantations have been established in Thailand, as this is prohibited by the land reform act. The cassava harvested area, for the whole country, in 2011 was 1.14 million ha and production yield was 19.30 tonne of fresh roots per ha (OAE, 2012). The cassava farming ac-tivities include land preparation, planting, fertilization, weeding, and harvesting. The foreground data on fuel, lubrication oil, fertil-izers, and herbicides inputs were collected through a�eld survey in 2011, in Nakhonratchasima and Chaiyaphum provinces, the northeastern cultivating areas of the country. With respect to the allocation method for this stage, since cassava stems are mainly used for new planting which is considered as an internal use in the system, the environmental loads of the cassava cultivation and harvesting stage are allocated only to the cassava roots. The carbon dioxide from the air and solar energy for the photosynthesis pro-cess were excluded in this analysis. Emissions to air during prep-aration of cassava �elds of planting and emissions from fertilizers during growth are included. For emissions during cassava growing from nitrogen fertilizer, it was assumed that of the total N applied 10% will be evaporated as NH3, and 1% is assumed to be evaporated

(3)

Crude oil extraction (Middle East)

Oil refinery

Monomer production (Purified terephthalic acid) Electricity Fuel Aux. Chemicals Catalysts Water By-products Emissions to air Wastewater Solid waste Natural gas extraction

(Gulf of Thailand)

Gas separation

Monomer production (Ethylene glycol)

PET production

PET bottle production

PET incineration

Fig. 2. The system boundary for PET bottles system.

Cultivation and harvesting

(16 tonne truck transport: fields to starch mills 50 km)

Cassava starch production

Base case: electricity from the national grid, and heat from fuel oil (45%) and biogas (55%)

Option I: electricity from the national grid and completed replacement of fuel oil by biogas (32 tonne truck transport: starch mills to glucose

plant 270 km)

Glucose production

(Glucose plant located close to PLA pellets factory)

Lactic acid, lactide and PLA production

Base case: electricity from the national grid and steam from natural gas

Option II: electricity and steam production from natural gas based on combined heat and power (CHP) (16 tonne truck transport: PLA pellets plant to bottles

production plants 170 km)

PLA bottles production

(Bottles transport was neglected)

Disposal scenarios

(16 tonne truck transport: users to disposal sites 40 km) Fertilizers Herbicides Diesel Lubricating oil By-products: cassava pulp Emissions to air: CO2, CH4, N2O, SO2, etc.

Wastewater: COD, BOD Solid waste: cassava peel, sand, etc. By-product: Cassava stems Air emissions: CO2, N2O, CH4, etc. Emissions to air: CO2, CH4, SO2, etc.

Wastewater: COD, BOD Electricity Fuel oil Water Electricity Electricity Fuel oil Water Aux. chemical Emission to air Wastewater Solid waste By-products: gypsum Emissions to air: CO2, CH4, SO2, etc.

Wastewater: COD, BOD Solid waste; sludge Electricity

Fuel oil Water Aux. chemical Enzymes

Wastewater: COD, BOD Solid waste: PLA scrap Electricity

Water

(4)

as N2O-N (IPCC, 2006). Some relevant data on this stage and

ac-tivities used in the analysis are shown in Table 2. 2.3.2. Cassava starch production stage

One kilogram of cassava starch requires 3.9e4.5 kg of fresh cassava roots at its starch content is only 25% ( Chavalparit and Ongwandee, 2009). In Thailand, the large-scale processing facil-ities with advanced processing machines and technology have been replacing the primitive and small-scale factories. The cassava starch processing methods could be divided into two processes; tradi-tional and modern. The modern process, typically practices in the large-and-medium-scale factories, relies on a number of pieces of highly ef �cient equipment and machines. The production process may be divided into eight steps as follows: determining the starch percentage; removing sand and impurities in the rotary screener; peeling, cleansing and chopping out root rails; putting the fresh clean cassava into the Rasper and then Decanter to remove the protein; passing the slurry through a screen to remove the �bers; separating the �ne �bers and impurities using a centrifuge; drying

out the starch by passing it through the hot-aired dryer column; and�nally packing the�ne powder into sacks for sale. Inventory data were gathered from three cassava starch factories in Nakhonratchasima and Chachoengsoa provinces and are summa-rized in Table 3. The environmental burdens of the cassava starch production system are allocated between the cassava starch and cassava pulp, based on a mass allocation approach in term of starch content. In the base case scenario based on the current situation of Thai cassava starch industry, this is assumed to require heat generated from fuel oil (45%) and biogas (55%) (NSTDA, 2011), and electricity from the national grid. The improvement option (option I) is the complete replacement of fuel oil by biogas from anaerobic treatment of the mill ef �uents.

2.3.3. Glucose production stage

Commercially, glucose is produced via the enzymatic hydrolysis starch for which many crops can be used as the source of starch such as corn, wheat, cassava, rice, etc. Glucose production from cassava starch consists of three steps: liquefaction, sacchari�cation, and puri�cation. Because information on energy used in glucose production from cassava in Thailand has not been published, this study has gathered the inventory data from the report on the

�nancial and economic viability of bioplastics production in Thailand (Chiarakorn et al., 2011), and Renouf et al. (2008). One kilogram of glucose production requires 0.144 kWh of electricity and 0.0067 L of fuel oil.

2.3.4. Lactic acid, lactide and PLA production stage

Glucose is converted to lactic acid by fermentation, followed by puri�cation. The fermentation process requires energy use (steam and electricity) and contributes substantially to the fossil energy demand of PLA. Sulfuric acid, calcium carbonate, and auxiliary chemicals are required as operating supplies. The PLA manufacturing from lactic acid occurs in two steps. The �rst step is the conversion of lactic acid into the lactide, and then puri�cation by distillation. In the second step the polymerization of lactide to polylactide takes place in the presence of a tin catalyst. Inventory data on the energy use and process chemical demand for the lactic acid, lactide, and polylactide production were extracted from Groot and Borén (2010). Based on 1 kg of PLA, the production requires 0.97 kWh of electricityand 12.74 MJ of steam. This study considered two different scenarios as described below:

Table 2

Inventory data of cassava root cultivation stage.

Flow Unit Amount Type Related activities Inputs

Fertilizers (NePeK): 15e15e15

kg/ha/year 154_{15 Material input Fertilizer application}

kg/ha/y ear 335 Material input Fertilizer application

kg/ha/y ear 484 Material input Fertilizer application

Paraquat kg active ingredient/ ha/year

0.96_{0.27 Material input Weeding}

Glyphosate kg active ingredient/ ha/year

1.440.52 Material input Weeding

Diesel l/ha/year 35_{12 Energy input} _{Soil preparation,}

weeding, and harvesting Outputs

Cassava roots tonne fresh roots/ha/year

19.30 Product output

Cassava stems tonne/ha/year 3.6 Internal �ow Use in new planting

Table 3

Inventory data of cassava starch production stage.

Flow Unit Amount Type Related activities Base case Option I

Inputs Cassava root kg/kg starch 4.33_{0.39 4.33}_{0.39 Material} input Farming Water l/kg starch 18.657.16 18.657.16 Material input Processing water and steam production Fuel oil MJ/kg starch 1.28_{0.67 0} _Energy input

Burning for steam and electricity production Biogas m3/kg starch 0.030.03 0.060.01 Internal �ow

Burning for steam and electricity production Electricity kg/kg starch 0.21_{0.04 0.18}_{0.01 Energy} input In process electricity use Outputs Cassava starch (13% MC) kg/kg starch 1.00 1.00 Product output Allocation by starch content Cassava pulp (dry mass) kg/kg starch 0.39 0.39 By-product Allocation by starch content Table 1

Sources of background data used in this study. Background data Source

Fertilizers production Ecoinvent (2008) Herbicides production Ecoinvent (2008) Crude oil production Ecoinvent (2008) Chemicals production Ecoinvent (2008)

Terephthalic acid production Adjusted from Ecoinvent (2008)a

E thylene glycol production Adj ust ed fro m Ecoinvent (2008)a

PET resin production Adjusted from Ecoinvent (2008)b Road transport by truck MTEC (2011)

Diesel production MTEC (2011) Natural gas production MTEC (2011)

Electricity grid-mixed production MTEC (2009) and DEDE (2012) Steam production Adjusted from Ecoinvent (2008)c Combined heat and power (CHP) system Adjusted from Ecoinvent (2008)c Remarks:

a _{Adjusted by replacement with the data from Thai electricity and heat databases.} b _{Adjusted by replacement with Thai database such as energy sources and}

feed-stock ratio.

(5)

_{Base case e electricity from national grid and steam production}

from natural gas were used to assess the environmental per-formance of the product systems.

 Option II e electricity and steam production from natural gas

based on combined heat and power (CHP) system was used to evaluate the impact on environment of the product systems.

2.3.5. PET resin production

The inventory data of PET resin production are divided into �ve major stages including raw material extraction, primary material production, monomer production, PET production, and related transport. The raw material extraction stage involves crude oil extraction and natural gas extraction, background data being gathered from Ecoinvent (2008) database. Transport of crude oil from the Middle East and South of Asia to the oil re�neries at Rayong province, in the east of the country, by ocean tanker was estimated at 6700 km, whereas natural gas is piped transmission from the Gulf of Thailand to the Rayong gas separation plants. At the oil re�nery, crude oil is processed to produce naphtha and then cracked to paraxylene. At the gas separation, natural gas is pro-cessed to produce ethane which is a feedstock to produce ole�ns. Inventory data of oil re�nery and natural gas separation were gathered from the national LCI database of Thailand (MTEC, 2011). The monomer production stage includes the production of puri�ed terephthalic acid (PTA) and monoethylene glycol (MEG). PTA is produced via oxidation reaction of paraxylene with acetic acid as solvent and cobalt as a catalyst. The production of 1 kg of PTA re-quires 0.66 kg of paraxylene, 0.43 kg of water, 0.47 kWh of elec-tricity and 3.93 MJ of heat (Ecoinvent, 2008). MEG is produced from ethylene via intermediate derivative of ethylene oxide by reaction with water then conversion to MEG. The production of 1 kg of MEG requires 0.72 kg of ethylene oxide, 6.18 kg of water, and 0.39 kWh of electricity, whereas 1 kg of ethylene oxide is produced from 0.83 kg of ethylene, 0.46 kg of liquid oxygen, and 0.33 kWh of electricity (Ecoinvent, 2008). The inventory data of both monomers were adjusted from the Ecoinvent database using the electricity and heat data from Thai databases developed by MTEC (2009). PET resin is produced by reacting PTA with MEG and catalyst. The main pro-duction process steps are raw material preparation, esteri�cation, pre-polycondensation, and polycondensation. Based on 1 kg of PET resin, the production requires 0.87 kg of PTA, 0.35 kg of MEG (Indorama Venture Public Company Limited, 2013), 0.38 kWh of electricity, and 6.3 MJ of heat (Ecoinvent, 2008). Inventory data of PET production in this study were adjusted data from the Ecoinvent database by replacement with Thai database such as energy sources and feedstock ratio.

2.3.6. Conversion of polymer pellets into bottles

Polymer pellets are converted into plastic bottles using a stretch blow molding process. The energy requirement is mainly met by electricity and the level of consumption depends on the polymer type, the polymer mass, as determined by the weight of injected packages, and the machine model and capacity. Inventory data of plastic bottle weights and electricity demand for injection were collected from an actual manufacturing plant located in Bangkok, Thailand. All information on the conversion processes was taken from the year 2011. Based on 1 kg of PLA bottle, the production requires 1.20 kg of PLA resin and 2.69 kWh of electricity, whereas PET bottle production requires 1.12 kg of PET resin and 2.33 kWh of electricity.

2.3.7. Disposal scenarios of used bottles

This study considered various end-of-life scenarios based on four different disposal methods, including composting, land�ll,

recycling, and incineration. The scenarios of used PLA bottles are as follows:

_{Scenario 1 (S1) e 100% composting}

_{Scenario 2 (S2) e 100% incineration with energy recovery}  Scenario 3 (S3) e 100% land�ll without energy recovery

_{Scenario 4 (S4) e 100% land}_�ll with energy recovery

_{Scenario 5 (S5) e 100% chemical recycling}

_{Scenario 6 (S6) e 80% composting} þ 20% land�ll with energy recovery

 Scenario 7 (S7) e 80% compostingþ 20% incineration with

en-ergy recovery

2.3.7.1. Composting. Composting is a bene�cial waste management system, particularly where land�ll sites are limited, and in cities with dense populations. The primary mechanism of degradation of PLA is hydrolysis, catalyzed by temperature, followed by bacterial attack on the fragmented residues (Farrington et al., 2005). In this study, the composting has considered only PLA bottles, the com-posting model and data are based on the comcom-posting plant at Phang, Chiang Maiprovince, in the northern part of the country. We assumed 87% degradation of PLA in the presence of oxygen while the remaining 13% is compost or digestate to be used as soil conditioner. It is considered that 100% of carbon from the degra-dation evolves into CO2, which makes it carbon neutral

(Suwanmanee et al., 2010).

2.3.7.2. Incineration. Incineration refers to a process that com-busted the waste to generate electricity. Electricity production was calculated with a lower heating value (LHV) of polymers, and the electricity production ef �ciency of the waste incineration plant was assumed to be about 30% based on a dedicated incineration facility using European technology that uses only plastic bottles (Dornburg et al., 2006). The facility enables for energy recovery: 28 kWh of the electric energy are considered for the incineration of 1000 PET bottles, while 20 kWh of the electric energy for the incineration of 1000 PLA bottles (data taken are based on low heating value of polymer). In addition, it was assumed that the electricity produced by the incineration of PET and PLA bottles is used to substitute the grid electricity of Thailand.

2.3.7.3. Land �ll. In a sanitary land�ll facility, there is little moisture and/or insuf �cient temperature that could result in the degradation of PLA. A previous study carried out by Kolstad et al. (2012) found that PLA contained in land�ll at 21 _{C, 50e65%}

moisture, for 390 days of digestion did not degrade. Degradation of PLA under anaerobic condition only occurs under speci�c conditions which involve high temperature (>55 C), high

hu-midity, and a suitable mixture with other organic materials (Merrild and Hedal, 2010; Yagi et al., 2009). In addition, the published data based on Bohlmann (2004) suggested that PLA waste will be biodegraded in water after 11 months at 25 _{C in an}

anaerobic environment, which generates methane. In this study, we assumed the worst scenario to estimate GHG emissions ac-cording to Bohlmann (2004), Merrild and Hedal (2010), Yagi et al. (2009), and the theoretical stoichiometric reaction. The potential methane generation from PLA based on the theoretical stoi-chiometric reaction is shown as follows:

C6H8O4þ2H2O/3CO₂þ3CH₄

All carbon in PLA waste was converted to CH4 and CO2 in the

land�ll; 100% anaerobic biodegradation generates 334 g CH4 per

(6)

or other materials converting the waste to carbon dioxide. The estimation of methane emission in this study was calculated as follows:

CH₄ emission ¼ CH₄ generationCH₄ recovery CH₄ oxidation

Incase ofthe land�ll with energy recovery, it wasassumed that 60% of the methane generated was recovered and combusted in the boiler to generate electricity and that the rest (40%) escaped into the atmosphere (Pattharathanon et al., 2012). The electricity pro-duced is used to replace the national-grid electricity based on 30% energy ef �ciency. Emissions from biogas power generation were estimated based on credited replacement in the electricity mix in Thailand. For used PET bottles, this study considered only 1% degradation after 100 years (Gironi and Piemonte, 2011); the en-ergy consumption of land�ll facility was estimated from the Pan-omsarakam land�ll site, Chachoengsoa province.

2.3.7.4. Chemical recycling of plastic waste. Currently, chemical recycling of virgin PLA waste generated during polymerization can be proceeded in the production process, but it is also a possible future option for the recovery of used post-consumer PLA pack-aging materials. Chemical recycling is understood as a hydrolysis process in the �rst step, followed by a puri�cation step with lactic acid monomers being the �nal product of the recycling process. These monomers can be fed back into the PLA polymerization process. The PLA recycling scenario was considered in this study; about 90% of PLAwaste can be converted to lactic acid by hydrolysis at 250_{C with a processing time of 10 e20 min. The energy}

con-sumption for PLA waste conversion to lactic acid is 0.6 MJ per kg PLA (Dornburg et al., 2006). The energy requirement for polymer-ization was estimated based on Groot and Borén (2010) with a conversion yield of polymerization process of 85%. In this analysis, the overall conversion system can produce 0.76 kg recycled PLA per kg PLA bottle waste thus, only 0.76 kg of the recycled PLA can replace virgin cassava-based PLA resin. For used PET bottles, this study assumed that the conversion yield of monomers from PET waste is about 80% (Genta, 2003) and that the conversion yield of the polymerization process to produce new PET granules is about 79% (Ecoinvent, 2008). The materials and energy requirement for polymerization to produce PET was estimated based on the Ecoinvent (2008) database.

2.3.8. Transport

Transport operations are particularly relevant for transportation of polymer pellets to the plastic bottle producer. The production of the fossil-based polymers and PLA pellets is located in the Rayong province, in the east of the country, hence an average transport distance of 170 km by 16 tonne truck has been assumed based on information obtained from the bottle manufacturer. The glucose factory was assumed to be close to the PLA plant. The distance from the cassava�elds to the cassava starch factories wasestimated to be 50 km by 16 tonne truck. The starch transport from Nakhonratch-asima and Chachoengsoa provinces to Rayong province (location of the glucose and PLA plant) was assumed to be 270 km by 32 tonne truck. The waste bottles transport from households to disposal fa-cilities was assumed to be 40 km by 16 tonne truck.The background dataset for the truck transport was based on the national life cycle inventory databases of Thailand that were collected, validated, and evaluated by MTEC (2011). The transportation from the drinking water manufacturers to the distributors and consumers was not included in this study.

2.4. Impact assessment

Basically, the impact assessment phase converts the LCI results into assigned categories. This phase aimed to evaluate the signi� -cance of potential environmental impacts. Different life cycle impact assessment methods wereavailable in the SimaPro software such as eco-indicator 95, CML, EDIP, TRACI, cumulative energy demand (CED), etc. The CML 2 baseline 2000method was chosen in this study. The impact categories considered in the method are global warming, acidi�cation, eutrophication, and human toxicity potential. In addition, the cumulative energy demand method was selected to assess the fossil energy demand category. These impacts categories considered in this study are relevant in the Thailand perspective.

3. Results and discussion 3.1. Cradle-to-gate

3.1.1. Global warming potential (GWP)

In this section, the life cycle impact assessment (LCIA) was analyzed for 1000 drinking water bottles of PLA and PET for the relevant impact categories using the impact assessment model based on the CML 2 baseline 2000. As PLA resin is currently pro-duced in Thailand by Purac (Thailand) so the production of PLA resin based on Purac (Thailand) was used as a base model for this study, with a modi�cationthat cassava was used instead of sugar. In this part, we focused on GWP represented by GHG emissions (kg CO2 eq.) as shown in Fig. 3. For PLA resin production life cycle, the

total GHG emission for cassava-based PLA resin production was based on the base case scenario which was 2.48 kg CO2 eq. per kg

resin. In this scenario, the major GHG emissions (about 57.30%) came from the polymerization process due to energy consumption, including steam and grid electricity. The second part of the GHG emissions came from cassava starch production, accounting for 28.42%, due to CH4 emissions from the wastewater treatment

process. In the cultivation stage, GHG emissions accounting for 6.94%, mainly came from fertilizer utilization and N2O emission

from N-fertilizers. Consequently, the full utilization of biogas from the wastewater treatment of cassava starch production has been proposed as an improvement option (option I) to help reduce GWP. The net GHG for this option was found to reduce to 1.96 kg CO2 eq.

per kg resin. Based on option I, the PLA production stage could be further improved to option II by additional installation of a com-bined heat and power (CHP) system instead of the grid electricity and steam energy from natural gas. For this option, net GHG could be reduced to 1.54 kg CO2 eq. per kg resin.

Concerning GWP of plastic resin, several studies (Detzel and Kruger, 2006; Vink et al., 2007; Groot and Borén, 2010; Gironi and Piemonte, 2011) have shown that PLA resin had lower GWP than its fossil-based resins such as PET, PS and PP which is in good agreement with our study. When comparing our results with a similar study by Groot and Borén (2010), they reported GHG emissions of 0.50e0.80 kg CO2 eq. per kg sugarcane-based PLA

produced in Thailand. This is lower than the value obtained in our study, which was 1.54e2.48 kg CO2 eq. per kg cassava-based PLA.

The main difference was that Groot and Borén (2010) study also took into account CO2 uptake during sugarcane cultivation, while

CO2uptake during cassava cultivation was notincluded in our study

because we considered that CO2 was released into the atmosphere

at the end-of-life of the PLA product, thus net CO2balance was zero.

When compared to the corn-based PLA studied by Vink et al. (2010) and Gironi and Piemonte (2011), they reported the GHG emissions of 1.30 and 1.09 kg CO2 eq. per kg corn-based PLA, respectively,

(7)

cassava-based PLA. The same reason as above could be used to explain this difference since they also took into account the CO2 uptake during

corn growing.

Based on the functional unit de�ned in this study (1000 drink-ing water bottles) it was found that different amountof PETand PLA resins wereused which were16.26 kg and 16.35 kg of PLA resin and PET resin, respectively. Although GHG emission during stretch blow molding of PET bottles was lower than that of PLA, the results showed that the total GHG emissions of cassava-based PLA bottles was lower than that of PET bottles as shown in Fig. 3.

3.1.2. Fossil energy demand

When comparing the energy required for PLA bottles with PET bottles, the results showed that PLA bottles had lower fossil energy consumption than PET bottles (Fig. 4). The fossil energy con-sumption for 1 kg of PLA resin for the base case scenario, option I, and option II was 32.47 MJ, 31.15 MJ, and 26.22 MJ, respectively. In addition, we found that production of cassava-based PLA bottles consumed less energy than production of PET bottles. The fossil energy required to produce resins for manufacturing 1000 bottles was 700e800 MJ for PLA and 2120 MJ for PET. The energy con-sumption during the stretch blow molding process was 267 MJ for PLA and 249 MJ for PET, which was related to the amount of resin required to produce bottles and the speci�c heats of each polymer. The fossil energy consumption obtained in our study was shown to be quite close to the value reported by Groot and Borén (2010). However, in comparison with the corn-based PLA studied by Vink et al. (2010) and Gironi and Piemonte (2011), the value obtained

in our study was lower than that of corn-based PLA because of the intensive use of chemical fertilizers and pesticides and lower yield in corn cultivation.

3.1.3. Acidi �cation potential

The third impact category considered in this study was acidi� -cation potential (AP). The AP of cassava-based PLA resin production for the base case scenario, option I, and option II was 16.16, 15.91, and 14.51 g SO2 eq. per kg resin, respectively. When comparing the

results of this study with Groot and Borén (2010), our study showed lower AP than that of the sugarcane-based PLA. This is mainly due to the greater amounts of SO2 and NO x generated in the sugar

production from sugarcane as compared to cassava starch pro-duction, and SO2 emissions from sulfuric acid production that is

used in the lactic acid production process. However, in comparison with the study of Gironi and Piemonte (2011), they reported the AP impact of 11.52 g SO2 eq. per kg corn-based PLA which was lower

than the value obtained in our study for cassava-based PLA. The main reason is due to the difference of sources of electricity used in the PLA production process. Fig. 5 shows the comparison of AP of the three PLA cases and PET bottles for the functional unit of 1000 bottles. The results revealed that cassava-based PLA bottles have higher AP than PET bottles.

3.1.4. Eutrophication potential

The eutrophication potential (EP) of PLA resin for the base case scenario, option I, and option II was shown to be 9.22, 3.74, and 3.53 g PO4 eq. per kg resin, respectively. In the base case scenario,

0 250 500 750 1000 1250 1500 1750 2000 2250

PLA/Base Case-Bottle PLA/Option I-Bottle PLA/Option II-Bottle PET Bottle

Bottle production Resin transport PLA or PET production Glucose production Raw material transport Starch production Cultivation or raw material production F o s s i l E n e r g y D e m a n d ( M J L H V / F U )

Fig. 4. Fossil energy demand pro�les of PLA and PET bottles.

0 10 20 30 40 50 60 70 80

Bottle production Resin transport PLA or PET production Glucose production Raw material transport Starch production Cultivation or raw material production G W P ( k g C O e 2 q . / F U )

(8)

the results showed that EP impact mainly comes from cassava starch production stage accounting for 71.04%, and secondly from the cassava cultivation stage accounting for 20.73%. While in option I and II, the EP impact mainly comes from cultivation, starch pro-duction, and PLA production stage, respectively. When compared the results of this study with Groot and Borén (2010) and Gironi and Piemonte (2011), two cases of this study (option I and option II) have lower EP than that of the sugarcane- and corn-based PLA whereas the base case of this study has higher EP than the sugar-cane- and corn-based PLA. This is mainly due to higher chemical oxygen demand (COD) generated in the cassava starch production as compared to sugar production from sugarcane and dextrose production from corn. For the comparison at the product stage as shown in Fig. 6, the results revealed that PLA bottles had higher impact than PETbottles,especially in the base case scenario. LowEP impact of PETbottles is mainly due to lowchemical oxygen demand (COD) in wastewater in the PET resin production which is a petrochemical catalysis process.

3.1.5. Human toxicity potential

The human toxicity potential (HTP) of PLAresin for the base case scenario, option I, and option II was shown to be 2.67, 2.52, and 1.34 kg 1,4-DB eq. per kg resin, respectively. In comparison with Groot and Borén (2010), the results of our study have lower HTP than that of the sugarcane-based PLA. This is mainly due to greater amount of harmful emissions (such as NO x, SO2, particulates)

generated from bagasse combustion in the sugar production from sugarcane as compared to cassava starch production. From the comparison at the product stage shown in Fig. 7, the results showed that PLA bottles had lower impact than PET bottles. High HTP impact of PET bottles is mainly due to the greater amount of

harmful emissions from terephthalic acid and ethylene glycol production processes.

3.2. End-of-life scenarios

In this study, four disposal technologies were considered which include composting, land�ll (with and without energy recovery), chemical recycling, and bottle incineration. Based on these four disposal technologies, seven waste management scenarios (S1eS7) were considered, as described in detail in Section 2.3.7, in order to assess the environmental impacts of the disposal phase of bio-plastic wastes and to determine suitable waste management schemes for bioplastics. The basis for the analysisin this part was to treat 1000 PLA bottles (100% PLA waste).

3.2.1. Global warming potential

Fig. 8 shows GWP of the seven waste management/disposal scenarios of PLA in comparison with PET bottles treated by incin-eration, land�ll, and chemical recycling based on 1000 plastic bottles being treated. The results showed that S3 had highest GWP, followed by S4, incineration of PET bottles, and S6, respectively. The negative values for S1, S2, S5, S7, and PET recycling indicate that these scenarios gave a positive effect in terms of CO2 saving. The

end-of-life of PLA for each disposal technology is discussed below. 3.2.1.1. Composting. For composting, the bioplastic wastes are

degraded biologically under aerobic conditions, which results in soil conditioner substance or digestate and CO2 emission. As PLA is

produced from renewable resources, the CO2 emitted is considered

carbon neutral in this study (not counted as GHGemission). The soil conditioner substance (70%) from the composting process is usually 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35

PLA/ Ba se Ca se-Bott le PLA/Option I-B ot tle PLA/ Option II -B ot tle PET B ott le

Bottle production Resin transport PLA or PET production Glucose production Raw ma terial transport Starch production Cultivation or raw material production A P ( k g S O e 2 q . / F U )

Fig. 5. AP pro�les of PLA and PET bottles.

0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18

Bottle production Resin transport PLA or PET production Glucose production Raw material transport Starch production Cultivation or raw material production E P ( k g P O e 4 q . / F U )

(9)

mixed with animal manure (30%) and utilized as organic compost, which can replace the use of organic fertilizer. Thus, the total GWP of the composting process should be compensated by the GWP of the organic fertilizer production. As a result, the net GWP of com-posting technology was 1.04 kg CO2eq. per FU. From Fig. 8, it can be

seen that the GWP of the composting treatment for PLA waste (S1) was shown to be lower than that of the land�ll technology studied (S3 and S4).

3.2.1.2. Incineration. The treatment of PLA waste by incineration could be considered to have additional environmental bene�ts due to the energy generated and recovered that can offset the impacts associated with other energy sources. The remaining part from the combustion of plastics is ash, which needs to be treated by land�ll. The energy generated, as estimated from their LHV, is utilized to produce electricity which is considered as a compensation for the grid-mix electricity. Thus, the GHG of grid-mix electricity of the Electricity Generating Authority of Thailand (EGAT) is used to subtract from the total GHG emission of the incineration process. Consequently, the net GWP of incineration technology was2.81 kg CO₂ eq. per FU as shown in Fig. 8. When comparing

between the incineration with energy recovery of PLA bottles and PET bottles, the results showed that PLA incineration had lower

GHG impact than PET incineration. The high GHG impact of PET incineration is due to CO2 emitted from PET combustion which is

considered as fossil carbon, thus leading to global warming. In contrast, CO2 generated from PLA combustion is considered as

carbon neutral which does not affect global warming (Grigale et al., 2010; van der Harst and Potting, 2013).

3.2.1.3. Land �ll. For land�ll waste treatment, two cases were considered in this study: one with energy recovery (with biogas collection and utilization to generate electricity) and another one without energy recovery. It can be seen from Fig. 8 that the GWP of land�ll without energy recovery of 83.15 kg CO2 eq. per FU was the

highest among seven scenarios covered in this study. The largest amount of GHG generated from land�ll was the result of degra-dation of PLA under anaerobic conditions in the land�ll site, which emitted a large amount of methane (334 g CH4 per kg dry PLA) to

the atmosphere. As a result, the GWP of this treatment technology was shown to be highest among all the treatment scenarios stud-ied. Land�ll of PET bottles contributes very little to GWP since PET bottles are inert materials and do not decompose in land�lls.

Inthe caseof land�ll with energy recovery, relevant information was obtained from actual land�ll site of Bangkok Metropolitan Authority in Chachoengsoa province. It is reported that 60% of 0 5 10 15 20 25 30 35 40

Bottle production Resin transport PLA or PET production Glucose production Raw material transport Starch production Cultivation or raw material production H u m a n T o x i c i t y ( k g 1 , 4 D B e q . / F U )

Fig. 7. HTP pro�les of PLA and PET bottles.

-10 0 10 20 30 40 50 60 70 80 90 Disposal

Used bottle collection G W P ( k g C O e 2 q . / F U )

(10)

methane generated in this land�ll site could be collected (recov-ered) through a grid of pipelines buried underneath the land�ll site and sent to a gas engine and generator in order to produce elec-tricity. It is estimated that the other 40% of methane could escape to the atmosphere. The energy recovered was estimated to be equal to electricity of 0.75 kWh per FU, which was supplied to the grid at site. This helps reduce the need to produce equal amounts of electricity and it is considered to reduce the environmental impact by compensating the GWP resulting from the national electricity grid. Thus, the total GWP for land�ll with energy recovery was decreased to 27.24 kg CO2 eq. per FU as shown in Fig. 8.

3.2.1.4. Chemical recycling. The chemical recycling process used in this study was based on a literature review (Dornburg et al., 2006) where PLA waste was recycled back to monomer (lactic acid) and then re-polymerized to PLA resin. Since the recycled PLA is �nally converted to new resin, this recycling activity leads to a reduction in need to produce fresh PLA resin (from virgin material). Thus, the total GWP of recycled PLA waste should be deducted by the GWP of the production of fresh PLA resin (1.96 kg CO2 eq. per kg resin in

option I case). As a result, the net GWP for recycling PLA waste was shown to be 0.14 kg CO2 eq. per FU. However, it should be noted

that PLA is most likely used as a blend with other plastics and/or additives, thus the potential for recycling the material back to a monomer is low due to the technical feasibility and cost. For PET bottles, it can be seen that PET recycling has a positive effect in term of GHG saving.

A combination of disposal technologies for the treatment of PLA wastes was also studied which were scenarios S6 and S7. S6 com-bined 80% composting and 20% land�ll with energy recovery while S7 wasa combination of 80% composting and 20% incineration with energy recovery. The results showed that the S6 scenario had a higher GWP impact than S7.

3.2.2. Energy resources

Fig. 9 shows the energy consumption of the end-of-life scenarios for 1000 PLA and PET bottles. S2 (100% incineration), S4 (100% land�ll with energy recovery), S5 (80% composting þ 20% land�ll with energy recovery), S6 (80% composting þ 20% land�ll with energy recovery), S7 (80% composting þ 20% incineration with energy recovery), PET recycling, and PET incineration were shown to have negative values of energy consumption, indicating that these scenarios can save energy resources. In contrast, S1, S3 and PET land�ll did not show any bene�ts in term of energy resources.

3.3. GWP from cradle-to-grave

In this part, the GWP of the whole life cycle of bioplastics or

“cradle-to-grave” was considered, which combines all phases

throughout the life cycle of PLA, including resin production, bottles production, transportation, and disposal.

The life cycle GHG emissions of PLA bottles for different waste management scenarios are shown in Fig. 10. The S3 (100% land�ll without energy recovery) had the highest impact of 129e144 kg CO2 eq. per FU, while S2 (100% incineration with energy recovery)

was shown to be the best scenario which had the lowest GWP of 43e58 kg CO2 eq. per FU. S1 (100% composting) had an impact of

46e64 kg CO2 eq. per FU, which was close to S7 (80% composting

and 20% incineration with energy recovery). When comparing the life cycle GHG emissions between PLA and PET bottles, the results showed that PLA bottles had a lower impact than PET bottles in almost all scenarios, except S3 and S4.

4. Conclusions

This study evaluated the environmental performance of PLAand PET bottles for drinking water based on a life cycle perspective. For cradle-to-gate analysis, PET bottles contributed higher values in almost all impact categories, except for eutrophication and

acidi-�cation potential. It is shown that PLA bottles can reduce CO2

emissions, human toxicity and fossil energy demand. On the other hand, PLA causes high impact in terms of eutrophication due to a high COD in cassava starch wastewater generated for the base case scenario. Based on option I and II, EP impact of PLA was highest due to the cultivation stage because cassava planting requires the use of agrochemicals such as fertilizers and pesticides that contributed to eutrophication potential. The results showed that cassava-based PLA resin had a much higher GHG emissions than sugarcane- and corn-based PLA. This can be explained that both the sugarcane- and corn-based PLA took into account the CO2 uptake during the plant

growth, while CO2 uptake during the cassava cultivation of the

cassava-based PLA was not included since CO2 uptake was released

into the atmosphere at the end-of-life of the PLA product. However, the overall GWP can be lowered by improvement options proposed in this study which are improved utilization of wastewater from cassava starch plant to produce biogas for steam and electricity production, and applying a CHP system in the PLA plant. By incorporating these improvement options in the analysis, the GWP performance of cassava-based PLA has shown to be better than

-350 -300 -250 -200 -150 -100 -50 0 50 100 Disposal Used bottle collection F o s s i l E n e r g y D e m a n d ( M J L H V / F U )

(11)

conventional plastics, such as PET, which are used to produce to the same products.

For end-of-life analysis, four disposal technologies, including land�ll, recycling, composting, and incineration, were used in this study to evaluate the environmental performance of PLA bottles using seven different waste management scenarios (S1eS7). The results showed that incineration technology contributed the lowest GHG emission, followed by recycling and composting. The analysis also showed that S3 (100% land�ll without energy recovery) was the worst scenario while S1, S2, S5, and S7 scenarios could reduce CO2. Through the analysis in this study, the appropriated end-of-life

approach to PLA waste management and ways to improve the life cycle environmental performance of PLA could be offered. When combining the end-of-life into the whole life cycle, the cradle-to-grave analysis showed that PLA bottles are more environmental friendly than the PET bottles in terms of GHG emissions. However, this could be achieved through the use of appropriate waste management of bioplastic wastes which includes composting, incineration and recycling. The LCA results in this study indicated the possibility of improvement in lactic acid production and cassava starch production in order to minimize the environmental impact for the development of greener chemicals in the future.

Acknowledgments

This research was supported by the National Innovation Agency (NIA), Ministry of Science and Technology (Thailand), and Kasetsart University. The authors would like to thank all contributors for the data used in this study.

References

Bohlmann, G.M., 2004. Biodegradable packaging life-cycle assessment. Environ. Progr. 23, 342e346.

Boyd, S.B., 2011. Bio-based versus Conventional Plastics for Electronics Housings: LCA Literature Review. The Sustainability Consortium White Paper, 15 pp . Chavalparit, O., Ongwandee, M., 2009. Clean technology for the tapioca starch

in-dustry in Thailand. J. Cleaner Prod. 17, 105 e110.

Chiarakorn, S., Permpoonwiwat, C.K., Nanthachatchavankul, P., 2011. Financial and economic viability of bioplastic production in Thailand. In: European Associa-tion of Environmental and Resource Economists, 18th Annual Conference, 29 Junee2 July 2011, Rome.

DEDE, 2012. Electric Power in Thailand 2012. Department of Alternatives Energy Development and Ef �ciency (DEDE), Thailand Ministry of Energy, Bangkok . Detzel, A., Kruger, M., July 2006. Life Cycle Assessment of Polylactide (LCA). A

Comparison of Food Packaging made from NatureWorks PLA and Alternative Materials. IFEU Report, Heidelberg, Germany, 146 pp .

Dornburg, V., Faaij, A., Patel, M., Turkenburg, W.C., 2006. Economics and GHG emission reduction of a PLA bio-re�nery systemdcombining bottom-up anal-ysis with price elasticity effects. Resour. Conserv. Recycl. 46, 377 e409. Ecoinvent, 2008. Ecoinvent v2.2 Database. Swiss Center for Life Cycle Inventories,

Switzerland.

Farrington, D.W., Lunt, J., Davies, S., Blackburn, R.S., 2005. Poly(Lactic acid) Fibers. Chapter 6: Biodegradable and Sustainable Fibres. Woodhead Publishing Limited, England.

Genta, M., 2003. Development of Chemical Recycling Process for Post-consumer PET Bottles by Methanolysis in Supercritical Methanol. In: Technical Review, vol. 40. Mitsubishi Heavy Industries, Ltd.. Extra No. 1, pp. 1 e4 Available: www. mhi.co.jp/en/technology/review/pdf/e401ex/e395254.pdf

Gironi, F., Piemonte, V., 2011. Life cycle assessment of polylactic acid and poly-ethylene terephthalate bottles for drinking water. Environ. Progr. Sustain. En-ergy 30, 459e468.

Grigale, Z., Simaanovsa, J., Kalnins, M., Dzene, A., Tupureina, V., 2010. Biodegradable packaging from life cycle perspective. Sci. J. Riga Tech. Univ. 21, 90 e95. Groot, W.J., Borén, T., 2010. Life cycle assessment of the manufacture of lactide and

PLA biopolymers from sugarcane in Thailand. Int. J. Life Cycle Assess. 15, 970 e 984.

Indorama Venture Public Company Limited, 2013. Annual Information Report 2013 (Form 56-1). submitted to The Stock Exchange of Thailand, Bangkok.

IPCC, 2006. In: Eggleston, H.S., Buendia, L., Miwa, K., Ngara, T., Tanabe, K. (Eds.), 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Prepared by the National Greenhouse Gas Inventories Programme. IGES, Japan.

ISO, 2006a. ISO 14040: Environmental Management e Life Cycle Assessment e Principals and Framework. International Organization for Standardization, Genève, Switzerland.

ISO, 2006b. ISO 14044: Environmental Managemente Life Cycle Assessment e Requirements and Guidelines. International Organization for Standardization, Genève, Switzerland.

John, R.P., Nampoothiri, K.M., Pandey, A., 2006. Solid-state fermentation for L -lactic

acid production from agro wastes using Lactobacillus delbrueckii. Process Bio-chem. 41, 759e763.

Kolstad, J.J., Vink, E.T.H., Wilde, B.D., Debeer, L., 2012. Assessment of anaerobic degradation of Ingeo polylactides under accelerated land�ll conditions. Polym. Degrad. Stab. 97, 1131e1141.

Madival, S., Auras, R., Singh, S.P., Narayan, R., 2009. Assessment of the environ-mental pro�le of PLA, PET and PS clamshell containers using LCA methodology. J. Cleaner Prod. 17, 1183e1194.

Merrild, H., Hedal, K.N., 2010. Comparative Life Cycle Assessment of Two Materials for the K2233-1C Tray: PET Compared to the PLA Polymer INGEO. FORCE Technology, Denmark.

MTEC, 2009. A Research for Eco-ef �ciency Comparison of Electricity Generation from Renewable Energy and Fossil Fuels by Using Life Cycle Assessment. Final Report,. supportedby TheElectricity GenerationAuthorityof Thailand,Nonthaburi, Thailand. MTEC, 2011. Thai National Life Cycle Inventory Database. Unpublished data: http://

www.thailcidatabase.net/index.aspx.

Nampoothiri, K.M., Nair, N.R., John, R.P., 2010. An overview of the recent de-velopments in polylactide (PLA) research. Bioresour. Technol. 101, 8493 e8501. NIA, 2008. National roadmap for the development of bioplastics industry (2008e 2012). National Innovation Agency (NIA), Ministry of Science and Technology, Thailand

NSTDA, 2011. Strategic Research and Development Plan for Thai Cassava Starch Industry (2012e2016). National Science and Development Agency (NSTDA), Pathumthani.

OAE, 2012. Agricultural Statistics of Thailand: Cassava (in Thai): www.oae.go.th.

-25 0 25 50 75 100 125 150 Disposal Used bottle collection Bottle production Resin transport Cradle to resin production G W P ( k g C O e q . / F U )

(12)

Pattharathanon, T., Towprayoon, S., Wangyao, K., 2012. Greenhouse gas emission and economic evaluation from Municipal solid waste land�ll in Thailand. In: The 2nd International Conference on Biotechnology and Environment Man-agement vol. 42. IACSIT Press, Singapore, pp. 79 e85. http://dx.doi.org/10.7763/ IPCBEE, 16 pp.

Renouf, M.A., Wegener, M.K., Nielsen, L.K., 2008. An environmental life cycle assessment comparing Australian sugarcane with US corn and UK sugar beet as producers of sugars for fermentation. Biomass Bioenergy 32, 1144e1155. Suwanmanee, U., Leejarkpai, T., Rudeekit, Y., Mungcharoen, T., 2010. Life cycle

en-ergy consumption and greenhouse gas emissions of polylactic acid (PLA) and polystyrene (PS) trays. Kasetsart J. Nat. Sci. 44, 703 e716.

Tabone, M.D., Cregg, J.J., Beckman, E.J., Landis, A.E., 2010. Sustainability metrics: life cycle assessment and green design in polymers. Environ. Sci. Technol. 44, 8264e8269.

Uihlein, A., Ehrenberger, S., Schebek, L., 2008. Utilisation options of renewable re-sources: a life cycle assessment of selected products. J. Cleaner Prod. 16, 1306e 1320.

van der Harst, E., Potting, J., 2013. A critical comparison of ten disposable cup LCAs. Environ. Impact Assess. Rev. 43, 86e96.

Vink, E., Glassner, D., Kolstad, J., Wooley, R., O’Connor, R., 2007. The eco-pro�les for

current and near-future NatureWorks(R) polylactide (PLA) production. Ind. Biotechnol. 3, 58e81.

Vink, E.T.H., Davies, S., Kolstad, J.J., 2010. The eco-pro�le for current Ingeo

poly-lactide production. Indust. Biotechnol. 6 (4), 212 e224.

Wolf, O., 2005. Techno-economic Feasibility of Large-scale Production of Bio-based Polymers in Europe. Technical Report EUR 22103 EN. Institute for Prospective Technological Studies, Joint Research Centre (DG JRC), 256 pp .

Yagi, H., Ninomiya, F., Funabashi, M., Kunioka, M., 2009. Anaerobic biodegradation tests of poly(lactic acid) and polycaprolactone using new evaluation system for methane fermentation in anaerobic sludge. Polym. Degrad. Stab. 94, 1397e 1404.