lable at ScienceDirect

Journal of Cleaner Production 69 (2014) 83e90

Contents lists avai

Journal of Cleaner Production

journal homepage: www.elsevier .com/locate/ jc lepro

Reusable plastic crate or recyclable cardboard box? A comparison oftwo delivery systems

Sirkka Koskela a,*, Helena Dahlbo a, Jáchym Judl a, Marja-Riitta Korhonen a,Mervi Niininen b

a Finnish Environment Institute SYKE, Centre for Sustainable Consumption and Production, PO Box 140, FI-00251 Helsinki, Finlandb Stora Enso Oyj, Research Centre Imatra, FI-55800 Imatra, Finland

a r t i c l e i n f o

Article history:Received 10 January 2013Received in revised form9 January 2014Accepted 12 January 2014Available online 27 January 2014

Keywords:PackagingTransportationDeliveryLife cycle assessmentCorrugated cardboardPlastic

* Corresponding author. Tel.: þ358 400 148 811; faE-mail address: sirkka.koskela@ymparisto.fi (S. Ko

0959-6526/$ e see front matter � 2014 Published byhttp://dx.doi.org/10.1016/j.jclepro.2014.01.045

a b s t r a c t

During a product’s entire life cycle the significance of packaging varies in terms of environmental im-pacts. From the perspective of companies which manufacture packaging or packaging has an importantrole in their value chain it can be a relevant issue to focus on in their efforts to improve the environ-mental performance of their activities. The aim of this study was to compare the life cycle environmentalimpacts of a real product (bread) delivery system using either reusable HPDE plastic crates or recyclablecorrugated cardboard (CCB) boxes for product transportation. In this paper we focused on the deliverysystems (not the delivered product) covering the manufacturing of the crates/boxes, their use, the de-livery routes from bakery to retailers and waste management/recycling of the crates/boxes. As a result weconcluded that the recyclable CCB box system was a more environmentally friendly option than thereusable HPDE plastic crate system in all the studied impact categories based on the defined boundariesand assumptions. Transportation played a very important role in the environmental impacts of theanalysed systems. Therefore, changes, e.g. in the weights of products and their secondary packaging orthe transportation distances could affect the results considerably.

� 2014 Published by Elsevier Ltd.

1. Introduction

During a product’s entire life cycle the significance of packagingvaries in terms of environmental impacts. Especially with foodstuff,manufacturing of the product itself is much more resource andenergy intensive than the manufacturing of its packaging(Jungbluth et al., 2000). However, from the perspective of com-panies which manufacture packaging or packaging has an impor-tant role in their value chain, it can be a relevant issue to focus on intheir efforts to improve the environmental performance of theiractivities. Emissions from the production stage of packaging are notthe only aspects to be considered. In delivery systems, upstreamprocesses, transportation in the distribution network and wastemanagement issues must also be taken into account in order toassess environmental impacts holistically.

Many industrialised countries have policy frameworks andmeasures aiming to minimize packaging waste and their environ-mental impacts (e.g. Sonneweld, 2000). The measures vary from

x: þ358 9 5490 2491.skela).

Elsevier Ltd.

strict regulations imposed by governments to voluntary agree-ments between stakeholders. According to the waste hierarchygiven in the EU Waste Framework Directive (2008/98/EC), the firstpriority of waste management is to prevent waste from beinggenerated. Also the European Parliament and Council Directive onpackaging and packaging waste (94/62/EC, amended by theDirective 2004/12/EC) contains provisions on the prevention ofpackaging waste, on the reuse of packaging and on the recovery andrecycling of packaging waste. Reuse of products is undoubtedly agood measure for preventing waste since it can lengthen the life-time of a product significantly. However, when looking at theoverall environmental impacts of the product system where thereusable product is included, the picture is more complex due toe.g. the transportation and washing needed in order to enable theproduct reuse. This emphasizes the need for comprehensive envi-ronmental assessments of product systems in order to supportdecision making when choosing between different types of pack-aging materials and products.

At present, packaging is a necessary part of delivery systems.The basic packaging functions are transportation, storage and dis-tribution (Oki and Sasaki, 2000). In general, the functions of pack-aging materials, such as prevention of contamination, protection

S. Koskela et al. / Journal of Cleaner Production 69 (2014) 83e9084

against damages, preservation of contents and communications tothe customer, are many and varied in extent and complexity (e.g.,Oki and Sasaki, 2000; Pasquilo et al., 2011). In our study we focusedon storage, loading and transportation functions. The transport ofgoods demands particular container properties related to e.g.shape, weight, hygiene, handling and labelling (Stiftung InitiativeMehrweg, 2009). Transportation itself can be a very importantfactor in climate change impacts caused over the life cycles ofpackaging (e.g. Andersson and Ohlsson, 1999; Sim et al., 2007;Pretty et al., 2005; Meisterling et al., 2009). The importance oftransport modelling in LCA has been well known for long(Jørgensen et al., 1996). According to Jørgensen et al. (1996),transport contributes to 5e15% of themajor environmental impactsof a life cycle of a product. According to Madival et al. (2009),transport may contribute significantly to the environmental im-pacts of agricultural products and Gunady et al. (2012) emphasiseespecially the effects of a long distance transport.

In recent decades, there has been an ongoing trend to find newmaterials based on biomass or renewable resources to replace non-renewable materials, e.g. petroleum-based plastics (e.g. Madivalet al., 2009). Several current policies (e.g. “Thematic Strategy onthe Sustainable Use of Natural Resources COM/2005/670, An EUStrategy for Biofuels” COM/2006/34, A resource-efficient Europe e

Flagship initiative under the Europe 2020 Strategy COM/2011/21),which aim to achieve a more resource-efficient economy, supportthe increased use of materials made from renewable resources.Products made from renewable materials are not automatically abetter choice over ones made from non-renewable materials, sincethe whole supply chain from extraction to end-of-life must beconsidered. There can be aspects, e.g. reusing or recycling ofpackaging, which could change the ranking.

Many comparison studies of packaging systems have beenaccomplished (e.g. Ross and Evans, 2003; Lee and Xu, 2004; Singhet al., 2006; Raugei et al., 2009). In all of them reusable plasticcontainers proved to be a better choice compared to single-usepackaging. This bears out the general conception that reuse is al-ways better than recycling. But is it true in every case?

In our study we compared the life cycle environmental impactsof a real life delivery system using HDPE plastic crates or CCB boxesfor transportation of the delivery product. We assessed the impactsof two delivery systems, one using a crate made of non-renewableplastic and a box made of renewable CCB. Both materials haveadvantages and disadvantages in terms of environmental impacts.Plastic crates are durable and washable, hence they can be reusedseveral hundred times (in our study approximately 700 times)before finally being recovered as material for new plastic productsor as energy. CCB boxes can be designed to be strong but light, andalthough they can only be used once, they can be recovered andused in the production of new fibre products or as energy.

Levi et al. (2011) compared plastic containers and corrugatedboxes to each other in Italian fruit distribution. They concluded thatemissions from the manufacturing of corrugated box were greaterthan those frommanufacturing plastic crates and the importance oftransportation was identified in the environmental impacts of thedistribution systems. The study of Stiftung Initiative Mehrweg(2009) presents the results of a comparison of fruit delivery sys-tems in some European countries and Singh et al. (2006) inNorthern American market finding the plastic container systembetter than the CCB system. However, these studies cannot becompared to our study as such due to several differences in themodelling assumptions. The greatest differences existed in e.g. thematerial composition of the crates/boxes, their weights, the num-ber of circulations and transportation parameters. Additionally,end-of-life phases deviated from each other for both plastic cratesand CCB boxes.

The scope of the comparison is not the use of secondary pack-aging needed for the delivery system, but the delivery system ofpackaged bread using CBB and HDPE plastic crates. The aim was tocompare the life cycle environmental impacts of a real deliverysystem using either reusable HDPE plastic crates or recyclable CCBboxes for product transportation. The delivered product was toastbread which is a light weighted packed daily foodstuff delivered tothe whole Finland. The results do not include the processes relatedto bread baking and its upstream, because the delivered product isnot in the focus of this study. Theweight of bread is, however, takenaccount in the impacts of transportation. The study was imple-mented in cooperationwith the leading bakery company in Finland(VAASAN Oy) and with a global manufacturer of biomaterials, pa-per, packaging and wood products (Stora Enso Oyj). Both provideddata and valuable insights from a business perspective for thestudy. A critical review of the study was conducted by the SwedishEnvironmental Research Institute (IVL).

2. Materials and methods

2.1. Life cycle assessment and data sources

In order to achieve more sustainable production patterns, theenvironmental implications of the whole supply chain of products(both goods and services), their use, and waste management (ILCD,2010) must be considered. Life cycle assessment (LCA) studiesthereby help to avoid resolving one environmental problem whilecreating others, avoiding so called “shifting of burdens”. Life cycleassessment (LCA) is a method for integrating the environmentalimpacts of a studied product or a service over the whole valuechain. It is an internationally standardized method (ISO 14040,14044) with comprehensive guidelines (ILCD, 2010). In full LCAall processes and flows are followed from cradle-to-grave (i.e. fromresource extraction to waste disposal) taking into account allrelevant environmental impact categories.

The goal of the study was to compare the life cycle environ-mental impacts of two different product systems for bread deliveryfrom the bakery to consumers. The main difference in the systemswas the type of material used for the delivery crates, either plasticor CCB, which generated differences in, among others,manufacturing and transportation (Fig. 1). The product systems(referred to as plastic crate system and CCB box system) includedthe life cycles of manufacturing of the crates/boxes from virginmaterials and the delivery system of bread. The study tried toestablish which container material would be more favourable froman environmental perspective in this specific distribution system.

The weight of one plastic crate is 1.450 g with inside dimensionsof 560 � 360 � 125 mm. It is made of high-density polyethylene(HDPE). The CCB box weighs 190 g and its dimensions are540� 330�110mm. The bread delivered is toast bread. Theweightof an average loaf of bread is 340 g (2.720 g in one crate/box). Theweight of one plastic bag used for the bread packaging is 2 g (16 g inone crate/box). The different dimensions of a crate/box indicateslightly different capacities. However, all the crates/boxes hold thesame load, 8 loaves of bread, therefore they perform the samefunction in the studied systems.

Collected inventory data consisted of primary data from theparticipating companies, e.g. data related to the manufacturing ofCCB boxes, transportation distances and modes and the washingprocess for crates. The washing of crates, but not the consumptionof tap water (as a resource), was included in our assessment. Thewashing process also requires energy for heating the water and forthe washing process, impacts of which are included in the study.The washing mainly removes dust from the crates and the de-tergents used for washing do not include phosphorus. Generic data

Fig. 1. The studied product systems. Life cycle impacts of bread baking were excluded from the results shown in this paper.

S. Koskela et al. / Journal of Cleaner Production 69 (2014) 83e90 85

from the ecoinvent v. 2.2 database were used for the manufacturingof plastic crates, washing chemicals and energy production. Emis-sions from transportation were calculated according to the FinnishLipasto database (VTT, 2009) and the ecoinvent v2.2 database. Datafor recovery processes inwastemanagement were based on studiespublished by Korhonen and Dahlbo (2007) and Myllymaa et al.(2008).

The impact categories analysed in the life cycle impact assess-ment (LCIA) were selected based on data availability and theirrelevance to the studied product systems: climate change, terres-trial acidification, photochemical oxidant formation, particulatematter formation, and fossil depletion. These impact categoriesindicate well the impacts of transportation and energy use formanufacturing processes. In addition, fossil depletion is connectedto plastic production. Freshwater eutrophication was selectedbecause cardboard production causes nutrient releases into water.Land use issues are very important aspects considering impacts ofrenewable resources. We did not obtain data for land use, thereforeit is not considered in our study. Additionally, consensus of themethods for calculating land use has not been reached.

For calculating the environmental impacts of the systems,characterisation and normalisation factors (European referencearea) were taken from the ReCiPe (2011) midpoint (hierarchist)method, which has been developed for life cycle impact assess-ments (Goedkoop et al., 2009). ReCiPe was selected, because it’sone of the most up-to-date LCIA methods, currently. It is also underthe continuous development.

The data in this study are subject to uncertainties, which arevery common in all LCA studies related to lacking specific data andmodel uncertainties (see e.g. Heijungs and Huijbregts, 2004; Guoand Murphy, 2012; Mattila et al., 2012). The model uncertaintiesare similar in both systems, but uncertainties related to parameterspartly differ. No uncertainty analysis was conducted in this study,but three sensitivity analyses were done. Since transportation was

found to be the main contributor to the overall environmentalimpacts of the system, a sensitivity analysis was performed fortransport distances. Two additional sensitivity analyses concernedthe number of uses of the plastic crates and different allocationmethods for calculating the benefits of CCB recycling.

2.2. Description of the compared product systems

In life cycle comparisons, the boundaries of product systemsmust be consistent (ILCD, 2010). Therefore the plastic crate and CCBbox systems had exactly the same boundaries includingmany equalcomponents such as delivery routes from bakery to retailers andprimary bread packaging (plastic bag). The systems differed in themanufacturing of crates/boxes, their use, transportation impacts indelivery (crate collection and take-back) and waste management/recycling of the crates/boxes (Fig. 1).

The function of the studied systemswas to distribute bread frombakery to consumers. The functional unit of the product systems was8 loaves of bread delivered in one crate/box. However, the studyfocused on the comparison of the two different crate/box materialsand unit processes related to them, hence bread baking with itsupstream were excluded from the results presented in this paper.

2.3. Transport modelling

In plastic crate system the plastic crates were manufactured inFinland and transported to Tallinn, Estonia, where the bread wasbaked. In CCB box system the sheets of CCB were manufactured inLatvia where they were also cut into individual box sheets knownas CCB blanks and transported to Tallinn. At the bakery, the boxeswere assembled from the blanks in the box forming machine.Packed bread was then transported in crates or boxes from Tallinn(via Helsinki) to the main distribution centre in Eastern Finland.From there it was delivered to local distribution centres and then to

Table 1Transport distances and modes in bread delivery and crate/box collection.

Load Route Distance (km) Mode of transport

Hdpe granulate Production site/ crate manufacturing site 140 Lorry 16e32 t (ecoinvent)CCB blanks Production site (Latvia)/ bakery 310 Lorry 16e32 t (ecoinvent)Crate Production site/ bakery 100 Semi-trailer combination (hwy)

80 RORO ship10 Semi-trailer combination (urb)

Bread Bakery/ port of Tallinn 10 Semi-trailer combination (urb)Tallinn/Helsinki 80 RORO shipHelsinki/main distr. Centre 130 Semi-trailer combination (hwy)Main distr. Centre/ local distr. Centre 333 Full trailer combination (hwy)Local distr. Centre/ retail 162 Heavy delivery lorry (deli)

Crate backhaul Same distances and transport modes as for bread deliveryEmpty runs (CCB case) Same distances and transport modes as for bread deliveryObsolete crate collection Retail/ incineration/recycling 150 Heavy delivery lorry (deli)CCB collection Retail/ recycling 250 Heavy delivery lorry (deli)

Note: symbols hwy, urb and deli stand for highway, urban and delivery driving conditions, respectively. The delivery drivingmode has a defined highwaymileage share whichaccounts for 30%.

S. Koskela et al. / Journal of Cleaner Production 69 (2014) 83e9086

retailers. From retailers empty crates were transported back to themain distribution centre for washing and then back to the bakery inTallinn where they were reused. Obsolete plastic crates werecollected and transported to energy recovery and recycling plants.Empty CCB boxes were collected and transported to recyclingplants (Table 1).

2.3.1. Road freight transportTypically, transport impacts in LCA are calculated so that a mass

of transported load (in tonnes) is multiplied by a distance (in kil-ometres) and this value (in tonne-kilometres) is linked to an in-ventory of a unit process representing the transport of 1 t-km usinga specific mode of transport.

LCI of road freight transport unit processes in major LCI data-bases, such as the ecoinvent v2.2 (Ecoinvent, 2010), are inventoriedfor trucks of a specific tonnage, for an average annual mileage andan average load (Spielmann and Scholz, 2005). The average valuesusually represent the situation in Europe. Although commonlyused, these datasets are not suitable for every case study. In someproducts volume is the limiting factor for transportation, known asvolume-limited. These are typically light products large in volume,such as empty plastic crates or other light three-dimensional ob-jects. After the first LCIA calculations it became clear that thestudied system is sensitive to the way transport is accounted for.According to the ILCD Guidelines (2010), more specific transportmodelling had to be implemented.

For this study the Finnish calculation system for traffic exhaustemissions and energy consumption, Lipasto (VTT, 2009) was cho-sen as the most appropriate methodology and database. Lipasto

Table 2Weights of loads for distribution and return transportation.

Vehicle type Palettes per vehicle Specific load (in tonne

HDPE crate system

Distribution R

Semi-trailer combination 26 Palettes 8.75 3Gross vehicle mass 40tPayload capacity 25tFull trailer combination 41 Palettes 13.79 4Gross vehicle mass 60tPayload capacity 40tHeavy delivery lorry Approx. 9 pallets 3.15 1Gross vehicle mass 15tPay load capacity 9t

Note: distribution means the traffic from bakery to retailers (loaded with bread and craterecycling) are transported. The same degree of loading as for return and collection trips inbakery.

enables calculation of the unit emission profile and fuel con-sumption for defined trucks of a specific load, which gives anadvantage over generic LCI databases by providing more case-specific results. Emission data for a 2010 fleet average were usedin the calculation. In order to account for the upstream environ-mental impacts of transportation we combined Lipasto withecoinvent v2.2 (production and distribution of diesel fuel,manufacturing, maintenance and end-of-life of a vehicle and con-struction, maintenance and end-of-life of road infrastructure).

2.3.2. Sea transportThe calculation of specific RORO (roll-on/roll-off) ship emissions

is not applicable for trailers of different loads. Thus there was nodifference when accounting for sea transport between thecompared systems. The only difference is a justified assumptionthat 20% of otherwise empty trucks on the inbound route of CCBbox system are utilised by another product system. These aretherefore excluded from the system boundary. Unit emissions ofthe RORO ship were treated in the same way as those of the vehi-cles. They were combined with ecoinvent processes for fuel pro-duction, barge manufacture, maintenance and end-of-life.

2.3.3. Calculation approachThe Lipasto methodology (VTT, 2009) features an equation (Eq.

(1)) which was used in the calculation.

ex ¼�ea þ

�eb � ea

lc� lx

��� 1lx

(1)

s)

CCB box system

eturn and collection Distribution Return and collection

.02 6.12 Empty

.76 9.66 Empty

.08 2.21 Empty (return) 4.5 (collection)

s/boxes). In the return and collection trips only empty crates (collected CCB boxes forthe plastic crate system was used in crate transport from the production site to the

S. Koskela et al. / Journal of Cleaner Production 69 (2014) 83e90 87

The symbol ex in the equation above represents tonne-kilometreemissions (g/t-km) of a vehicle with user-defined payload lx (t). Thesymbol lc stands for payload capacity (t) of the vehicle; eb repre-sents vehicle-kilometre emissions (g/v-km) of the vehicle whentransporting lc and ea represents vehicle-kilometre emissions (g/v-km) of the vehicle when empty. The values of ea and eb were ob-tained from the Lipasto database. Inventory included emissions ofCO, HC, NOx, PM, CH4, N2O, NH3, SO2 and CO2 as well as fuel con-sumption of the vehicle.

The maximum load of each truck was calculated as the sum ofthe masses of crates, bread and bread packaging. TransportationEUR pallets were assumed to be an integral part of a truck andtherefore were not accounted for as part of the load. It was calcu-lated that one EUR-palette fits 80 crates/boxes. Based on that figure,specific loads of different trucks were calculated (Table 2). Thesewere used for calculating unit emission profiles in urban andhighway driving mode. Unit emissions for delivery driving werecalculated for the last step in the bread distribution. The selecteddriving mode for each route is specified in Table 1.

We calculated LCI of 1 tonne-km for each truck loaded to itsmaximumvolume with the transported product (packed breads) inthe transportation containers (CCB box, plastic crate). We calcu-lated LCIA results by multiplying this LCI by the mass of trans-portation containers only.

2.4. Benefits from reuse, recycling and energy recovery

After a fibre-based product has been used, it can be recycled.There are several ways of calculating the benefits of recycling. Westudied how the impacts of climate change would vary using open-loop allocation (ISO/TR 14049), monetary allocation and the systemexpansion approach to the CCB Box System. The results did notdiffer significantly from each other (less than 1%). We chose to usethe system expansion (or avoided emissions) approach for thiscomparison study.

In Finland, CCB is typically recycled for use in coreboardmanufacturing. In practice, coreboard is always manufacturedfrom recycled fibres, and virgin fibres are used only in minorquantities or not at all. However, in our study, we assumed that ifvirgin fibres were used they would be virgin fluting, which due toits fibre properties is the most suitable for coreboardmanufacturing. Thus the emissions avoided by recycling werecalculated as the difference between the emissions ofmanufacturing coreboard with virgin fluting and the emissions ofmanufacturing coreboard using recycled CCB. The raw materialsfor coreboard used in the calculations were 83% recovered fibreand 17% virgin fibre (data obtained from the participating forestcompany). The virgin fibre input is needed to compensate the lossof fibre strength during recycling.

The plastic crates can also be recovered as material or energy atthe end-of-life phase. When the crates were returned to the maindistribution centre for washing, obsolete and broken crates wereseparated from the reusable ones. Approximately 20% of theobsolete crates were recovered as material in the production ofplastic profiles, which is a process using e.g. HDPE as raw material(based on information from the bakery company). Plastic profilescan be used e.g. in patio constructions instead of impregnatedwood. Therefore the production of impregnated wood wasconsidered to be the process that can be avoided by plastic recy-cling (Korhonen and Dahlbo, 2007). In addition, around 80% of theobsolete crates were recovered as energy using in a boiler. Thetype of boiler used could not be specified; hence we assumed thatthe combustion of one kg of plastic produced 33 MJ of heat (basedon the lowest heating value of plastic waste, Statistics Finland,2011). This heat was assumed to replace separate production of

heat for which an emission factor of 62.77 kg CO2/GJ was used(average emission factor for separate production of heat, Motiva,2004).

In our product system, the benefits of reusing plastic crates wereconsidered by only including the amount of emissions that can beassociated with one circulation of the crate. The number of circu-lations of one crate was calculated from the average lifetime of onecrate (13.75 years), the number of circulations of one crate per year(61.54) and the duration of one circulation (4.87 days) (data ob-tained from the participating bakery company). For the specificroutemodelled in our system the circulation of a crate took one daylonger than for the average route. Taking this into consideration, weestimated 700 circulations for each plastic crate. Hence weassumed that the distribution of 8 loaves of bread delivered in onecrate represented 1/700 of the emissions of manufacturing oneplastic crate. Likewise, the benefits of material and energy recov-ered at the end-of-life of the crate were assessed for 1/700 of acrate.

3. Results and discussion

3.1. Environmental impacts

3.1.1. Overall impactsThis LCA study showed that for delivering 8 loaves of bread in

one container, CCB box system was a more environmentallyfriendly option than plastic crate system in all impact categoriesbased on the defined boundaries and assumptions (Fig. 2, Table 3).

3.1.2. TransportationIn the delivery phase of the studied systems, transportation was

the most significant contributor to all studied environmental im-pacts. For example the overall climate change impact related totransportation was 1.144 kg CO2-eq in plastic crate system and0.757 kg CO2-eq in CCB box system. Distances, modes of trans-portation and particularly load were the most important factors interms of transportation impacts. The greatest differences in theimpacts of transportation between the systems were caused by thedifferent weights of the crates/boxes and by the circulations ofplastic crates. It should be noted that the delivery network in thestudied systems covered the whole of Finland where the distanceswere very long. Local bakeries would decrease the amount oftransportation, but such a decentralised system would havedifferent impacts and the overall outcome cannot be evaluatedwithout a comprehensive analysis. In the use phase (i.e. deliverynetwork) CCB boxes performed better, because plastic crates needwashing after every use which causes impacts on the environment.However, the significance of washing on the total impacts was verylow.

3.1.3. Manufacturing of crates/boxesThe environmental impacts of manufacturing one HDPE plastic

cratewere higher than those of one CCB box, but the fact that crateswere reused hundreds of times decreased the impacts significantly(Fig. 3). As a result of the reuse of crates, the impacts frommanufacturing a plastic cratewere lower than those of a CCB box. Inour study, however, the recycling of CCB to coreboard productionchanged the overall impacts more in favour of CCB.

Decreasing, or increasing, the number of times the plastic cratesare used had only a very small impact on the overall results of thesystem (unless the number of uses is extremely low). For illus-trating this, the results of climate change were calculated with thenumber of uses of the crates ranging from 10 to 800. The resultsshowed that for a range of uses from 10 to 100 times the impacts ofmanufacturing decreased notably, but there after the differences

Fig. 2. Climate change (CC), terrestrial acidification (TA), freshwater eutrophication (FE), photochemical oxidant formation (POF), particulate matter formation (PMF) and fossildepletion (FD) according to the life cycle stages of the systems analysed and not including impacts of bread baking with its upstream. FU ¼ 8 loaves of bread delivered in one crate/box. Note: benefits of recovery of plastic crates by incineration are not visible due to the minor relevance.

S. Koskela et al. / Journal of Cleaner Production 69 (2014) 83e9088

were not significant (Fig. 3). In the range of hundreds of uses theimpacts from crate manufacturing per one use were minor.

Table 3Environmental impacts of the analysed systems.

Impact category Unit Plasticcrate system

CCB boxsystem

Climate change kg CO2-eq 1.18Eþ0 8.76E�1Terrestrial acidification kg SO2 eq 7.90E�3 6.24E�3Freshwater eutrophication kg PO4 eq 4.15E�4 3.79E�4Photochemical

oxidant formationkgNMVOC 1.02E�2 7.40E�3

Particulate matter formation kg PM10 eq 3.14E�3 2.37E�3Fossil depletion kg oil eq 4.29E�1 3.14E�1

Therefore the manufacturing of the crate played a very small role inthe system as a whole.

3.1.4. End-of-life phasesThe benefits of the recovery of plastic crates were very low (per

functional unit) with the defined end-of-life assumptions (20%material recovery, 80% energy recovery based on the current situ-ation) (Fig. 2) and therefore they are not visible as negative valuesin the results. The optimal end-of-life phase for plastic would be100% material recovery to compensate for virgin plastic as has beenshown by, e.g. Lazarevic et al. (2010). However, since only 1/700 of aplastic crate (which is equal to 2.07 g) was allocated per a functionalunit, changes in the benefits would not affect the overall result ofour study.

Fig. 3. Climate change impacts of crate manufacturing (3.99 kg CO2-eq/crate) allocated per number of circulations.

S. Koskela et al. / Journal of Cleaner Production 69 (2014) 83e90 89

The benefits of recycling CCB boxes are shown as negative valuesoriginating from the avoided emissions of using recycled CCBinstead of virgin fibre in coreboard manufacturing (Fig. 2). Also inthe future, demand for recycled corrugated boxes is expected to behigh, due to continuous demand for recycled fibres. Corrugatedboxes from retailers and industry are very often collected sepa-rately, which ensure the high purity and quality of fibres.

CCB recycling into coreboard manufacturing generated thegreatest benefits in the climate change impact category (Fig. 2).Without the benefit (0.106 kg CO2 -eq) the result for CCB box systemwould be 12% higher, but still lower than climate impact of plastic

CC

TA

POF

PMF

FD

0.00E+00 2.00E-13 4.00E-13 6.00E-13

normalised results

corrugated cardboard box (CCB)plastic crate

Fig. 4. Normalised results of impact categories. FU ¼ 8 loaves of bread delivered in onecrate/box.

Fig. 5. Sensitivity analysis of transport dis

crate system. Carbon sequestration in forests is an important part ofthe life cycle of fibre-based products. Taking this into considerationwould increase thebenefits of recycling inCCBBoxSystem.There aremodels attempting to link carbon sequestration in forests to productcarbon footprints but they contain high uncertainties. No scientificconsensus has yet been reached on how to incorporate CO2 removalby forests into product specific assessments. Therefore a conven-tional approachwas applied in this studyandnobenefit fromcarbonsequestration was allocated to CCB boxes.

3.2. Normalised results

Normalisation is a means to present LCIA results with morecomprehensible values than the impact category indicator scoresand this also made it possible to make comparisons between thecontributions of impact categories to the reference system (e.g.Dahlbo et al., 2013). In this study, European reference values wereused for each impact category.

Since weighting between impact categories was not done, thenormalised impact category results can be considered to be of equalimportance from the perspective of environmental protectionwithin the region considered in the normalisation reference values(Europe) and the harmfulness of one category over another cannotbe evaluated (Dahlbo et al., 2013).

Thenormalised results indicated that plastic crate systemmade agreater contribution to the impacts in Europe than CCB box system(Fig. 4). For both systems, the lowest contribution was to climatechange impacts and the highest to fossil depletion. For CCB boxsystem the contribution to the particulate matter formation washigher than to the terrestrial acidification and the photochemicaloxidant formation. In contrast, for plastic crate system the contri-butions to particulate matter formation and photochemical oxidantformation were higher than the contribution to terrestrialacidification.

tances on freshwater eutrophication.

S. Koskela et al. / Journal of Cleaner Production 69 (2014) 83e9090

3.3. Sensitivity analysis on transport distances

Transport clearly dominates the overall results for each of theanalysed impact category. Thus we applied a sensitivity analysis ontransport distances in order to analyse the robustness of our find-ings. We varied all distances except the ferry transport, whichwould in any case remain the same. Two alternative scenarios werecalculated, one for distances half shorter and one for distancestwice longer. Due to linearity of themodel themain conclusions didnot change for most impact categories. Except for freshwatereutrophication (Fig. 5), where the contribution of the CCBmanufacturing became more significant for the shorter distancesscenario and it outbalanced lower transport impacts caused bylighter CCB box, compared to the plastic crate. It is important tonotice that the impacts for shorter distances are more uncertainthan ones for longer distances, because transport equipment wouldlikely differ from the assumptions made in the study (e.g. smallertrucks for local delivery).

4. Conclusions

In this LCA comparison study two bread delivery systems wereexamined. In the first system the product was delivered in HDPEplastic crates and in the second one in CCB boxes. As a result we canconclude that the CCB box system was a more environmentallyfriendly option than the plastic crate system in all studied impactcategories based on the defined boundaries and assumptions.

Our study proved that a conclusion onwhich delivery systemhasmore favourable environmental impacts cannot be made based onthe container material only. Before decision making, the whole de-livery system including all necessary processes must be assessed. Ingeneral, a long-lasting, reusable product is considered to be a betterchoice, but as our study proved a recycled product can also be a goodoption but it requires a profitable and effective recycling system.

Transportation played a very important role in the environ-mental impacts of the analysed systems. However, changes, e.g. inthe weights of products and their secondary package or thetransportation distances could affect the results considerably.Regardless of the size of the distribution area it is important todevelop logistics and also vehicles further in order to decreaseenvironmental impacts of delivery systems.

Acknowledgements

This study was conducted as part of the MMEA research pro-gramme managed by the Cluster for Energy and Environment andfunded by TEKES e the Finnish Funding Agency for Technology andInnovation, Stora Enso Oyj and the Finnish Environment Institute(SYKE). A critical review of the study was conducted by the SwedishEnvironmental Research Institute (IVL).

References

Andersson, K., Ohlsson, T., 1999. Life cycle assessment of bread produced ondifferent scales. Int. J. Life Cycle Assess. 4 (1), 25e40.

Dahlbo, H., Koskela, S., Pihkola, H., Nors, M., Federley, M., Seppälä, J., 2013. Com-parison of different normalised LCIA results and their feasibility in communi-cation. Int J Life Cycle Assess. 18, 850e860. http://dx.doi.org/10.1007/s11367-012-0498-4.

Ecoinvent Database v.2.2, 2010. Swiss Centre for Life Cycle Inventories. http://www.ecoinvent.ch.

Goedkoop, M.J., Heijungs, R., Huijbregts, M., De Schryver, A., Struijs, J., Van Zelm, R.,6 January 2009. ReCiPe 2009. A Life Cycle Impact Assessment Method WhichComprises Harmonised Category Indicators at The Midpoint and The EndpointLevel, first ed. Report I: Characterisation. http://www.lcia-recipe.net (accessed20.05.11).

Gunady, M.G.A., Biswas, W., Solah, V.A., James, A.P., June 2012. Evaluating the globalwarming potential of the fresh produce supply chain for strawberries, romaine/

cos lettuces (Lactuca sativa), and button mushrooms (Agaricus bisporus) inWestern Australia using life cycle assessment (LCA). J. Clean. Prod. 28, 81e87.

Guo, M., Murphy, R.J., 2012. LCA data quality: sensitivity and uncertainty analysis.Sci. Total Environ. 435e436, 230e243.

Heijungs, R., Huijbregts, M.A.J., 2004. A review of approaches to treat uncertainty inLCA. In: Proceedings of the Second Biennial Meeting of IEMSs, Complexity andIntegrated Resources Management, 14e17 June 2004, Osnabrück, Germany,pp. 332e339.

ILCD Handbook. International Reference Life Cycle Data System, 2010. EuropeanCommission, Joint Research Centre. http://lct.jrc.ec.europa.eu/pdf-directory/ILCD-Handbook-General-guide-for-LCA-DETAIL-online-12March2010.pdf(accessed 04.12.12).

ISO 14040, 2006. Environmental Management e Life Cycle Assessment e Principlesand Framework. International Organization for Standardization.

ISO 14044, 2006. Environmental Management e Life Cycle Assessment e Re-quirements and Guidelines. International Organization for Standardization.

Jungbluth, N., Tietje, O., Scholz, R.W., 2000. Food purchases: impacts from theconsumers’ point of view investigated with a modular LCA. Int. J. Life CycleAssess. 5 (3), 134e142.

Jørgensen, A.-M., Ywema, P., Frees, N., Exner, S., Bracke, R., 1996. Transportation inLCA. Int. J. Life Cycle Assess. 1 (4), 218e220.

Korhonen, M.-R., Dahlbo, H., 2007. Reducing greenhouse gas emissions by recyclingplastics and textiles into products. In: The Finnish Environment, 30/2007.http://www.ymparisto.fi/default.asp?contentid¼249187&lan¼en (accessed17.12.11).

Lazarevic, D., Aoustin, E., Buclet, N., Brandt, N., 2010. Plastic waste management inthe context of a European recycling society: comparing results and un-certainties in a life cycle perspective. Resour. Conserv. Recycl 55, 246e259.

Lee, S.G., Xu, X., 2004. A simplified life cycle assessment of re-usable and single-usebulk transit packaging. Packag. Technol. Sci. 17, 67e83.

Levi, M., Cortesi, S., Vezzoli, C., Salvia, G., 2011. A comparative life cycle assessmentof disposable and reusable packaging for the distribution of Italian fruit andvegetables. Packag. Technol. Sci. 24, 387e400.

Madival, S., Auras, R., Singh, S., Narayan, R., 2009. Assessment of the environmentalprofile of PLA, PET and PS clamshell containers using LCA methodology. J. Clean.Prod. 17, 1183e1194.

Mattila, T., Leskinen, P., Soimakallio, S., Sironen, S., 2012. Uncertainty in environ-mentally conscious decision making: beer or wine? Int. J. Life Cycle Assess. 17(6), 696e705.

Meisterling, K., Samaras, C., Schweizer, V., 2009. Decisions to reduce greenhousegases from agriculture and product transport: LCA case study of organic andconventional wheat. J. Clean. Prod. 17, 222e230.

Motiva, 2004. Yksittäisen Kohteen CO2-päästöjen Laskentaohjeistus Sekä Käy-tettävät CO2-päästökertoimet (Guidance and CO2-emission factors for calcu-lating CO2-emissions from a single object) (in Finnish).

Myllymaa, T., Moliis, K., Tohka, A., Rantanen, P., Ollikainen, M., Dahlbo, H., 2008.Jätteiden Kierrätyksen Ja Polton Käsittelyketjujen Ympäristökuormitus Ja Kus-tannukset. Inventaarioraportti (Environmental loads and costs of waste recy-cling and incineration processes. Inventory report). Rep. Finnish Environ. Inst.28/2008.

Oki, Y., Sasaki, M., 2000. Social and environmental impacts of packaging (LCA andassessment of packaging functions). Packag. Technol. Sci. 13, 45e53.

Pasquilo, J., Meneses, M., Castells, F., 2011. The carbon footprint and energy con-sumption of beverage packaging selection and disposal. J. Food Eng. 103, 357e365.

Pretty, J.N., Ball, A.S., Lang, T., Morison, J.I.L., 2005. Farm costs and food miles: anassessment of the full cost of the UK weekly food basket. Food Policy 30 (9), 1e19.

Raugei, M., Fullana-i-Palmer, P., Puig, R., Torres, A., 2009. A comparative life cycleassessment of single-use fibre drums versus reusable steel drums. Packag.Technol. Sci. 22, 443e450.

ReCiPe, 2011 (accessed 04.12.11). http://www.lcia-recipe.net/.Ross, S., Evans, D., 2003. The environmental effect of reusing and recycling a plastic-

based packaging system. J. Clean. Prod. 11, 561e571.Sim, S., Barry, M., Clift, R., Cowell, S.J., 2007. The relative importance of transport in

determining an appropriate sustainability strategy for food sourcing. Int. J. LifeCycle Assess 12 (6), 422e431.

Singh, S.P., Chonhenchob, V., Singh, J., 2006. Life cycle inventory and analysis of re-usable plastic containers and display-ready corrugated containers used forpackaging fresh fruits and vegetables. Packag. Technol. Sci. 19, 279e293.

Sonneweld, K., 2000. The role of life cycle assessment as a decision support tool forpackaging. Packag. Technol. Sci. 13, 55e61.

Spielmann, M., Scholz, R.W., 2005. Life cycle inventories of transport services. Int. J.Life Cycle Assess 10 (1), 85e94.

Stiftung Initiative Mehrweg, 2009. The Sustainability of Packaging Systems for Fruitand Vegetable Transport in Europe Based on Life-cycle Analysis e Update 2009.http://www.pe-international.com/company/newsroom/news/2009/news-2009-detail/article/update-2009-the-sustainability-of-packaging-systems-for-fruit-and-vegetable-transport-in-europe-ba/ (accessed 28.12.12).

Statistics Finland, 2011. Fuel Classification and Emission Coefficients. http://www.stat.fi/tup/khkinv/khkaasut_polttoaineluokitus.html (accessed 17.12.11).

VTT, 2009. LIPASTO e A Calculation System for Traffic Exhaust Emissions and En-ergy Consumption in Finland. http://lipasto.vtt.fi/yksikkopaastot/tavaraliikennee/vesiliikennee/roroe.htm (accessed 17.05.11).