What email address or phone number would you like to use to sign in to Docs.com?
If you already have an account that you use with Office or other Microsoft services, enter it here.
Or sign in with:
Signing in allows you to download and like content, and it provides the authors analytical data about your interactions with their content.
Embed code for: Água e pegada de carbono de produtos lácteos selecionados 15 2016
Select a size
Water and carbon footprint of selected dairy products: A case study in Catalonia Vasileia Vasilaki a, Evina Katsou a, Sergio Pons a b, Joan Colon b, c, * a Department of Mechanical, Aerospace and Civil Engineering, Brunel University, Kingston Lane, UB8 3PH Uxbridge, Middlesex, UK b BETA Tech. Center. (TECNIO Network), U Sciente Tech., University of Vic-Central University of Catalonia, de la Laura 13, 08500 Vic, Spain c AXIA Tecnologías Ambientales S.L., 17811 Santa Pau, Girona, Spain a r t i c l e i n f o Article history: Received 10 May 2016 Received in revised form 2 August 2016 Accepted 6 August 2016 Available online 8 August 2016 Keywords: Dairy industry Yogurt production process Water footprint Carbon footprint LCA abstract This study assesses the LCA-based Water Footprint (WF) and Carbon Footprint (CF) of various types of yoghurts in the Spanish dairy plant of La Fageda. Primary data have been used to allocate impacts to the core processing stages. . The total amount of water consumption and greenhouse gas emissions for the production of 1 kg of yoghurt in La Fageda plant are 204 L H2O and 1.94 kg CO2eq respectively. The results indicated that raw milk and milk-based ingredients are the main contributors to all impact categories examined; their contribution to CO2eq ranged from 80 to 96%. Energy consumption and packaging ma- terials have signi ﬁ cant contribution to freshwater ecotoxicity, acidi ﬁ cation and global warming potential (GWP) impact categories ranging from 30 to 99% when raw milk is excluded from the analysis. In terms of the direct impacts of the plant, Cleaning in Place (CIP) and cleaning operations are responsible for 70% of the water requirements, while refrigerators, pasteurisation and packaging account for 70% of the energy consumption in the facility. The water and carbon footprint varies depending on the production process and the region. The sensitivity analysis illustrates that high precipitation and application of different techniques for raw milk production increases the contribution of the direct impacts of the plant from 2% to 15% in terms of water use. © 2016 Elsevier Ltd. All rights reserved. 1. Introduction The food industry is a major consumer of water; it is ranked third for water consumption and wastewater discharge after the chemical and re ﬁ nery industries (Olmez and Kretzschmar, 2009). According to Falkenmark (2008), fresh water availability is ex- pected to be one of the most signi ﬁ cant constraints towards future food production. Given that the food sector is also responsible for 30% of the Greenhouse Gas (GHG) emissions deriving from human activities (Garnett, 2011), mitigating actions must be considered in order to secure sustainability of food systems (Smith et al., 2013; Hallstr€om et al., 2014; West et al., 2014). Dairy products are an irreplaceable aspect of human diet and one of the largest and most important sub-sectors of the Food industry in Europe (Wijnands et al., 2007). However, their production is associated with severe environmental impacts including GHG emissions and water and energy consumption (Milani et al., 2011). In 2007, the dairy sector globally contributed by 4% to the GHG emissions (FAO, 2010). Ac- cording to the Commission Directorate (2006), the average direct use of water for yoghurt manufacturing, ranges from 0.8 to 25 L/kg processed milk. The application of Best Available Techniques (Prasad and Pagan, 2006) can lower water use to 0.1 L/L milk. The establishment of target e standardized environmental in- dicators is a key element for the assessment and mitigation of the environmental impacts and the unsustainable freshwater uti- lisation from the food sector (Fang et al., 2014). Two main ap- proaches have been reported in recent literature for the determination of the water footprint (WF) of a product (McGlade et al., 2012). The ﬁ rst approach was developed by the Water Foot- print Network (WFN) (Hoekstra et al., 2011) following a volumetric concept. It focuses on the identi ﬁ cation of the water-related hot- spots of a product's life cycle (Manzardo et al., 2014). The second one was introduced by the Life Cycle Assessment (LCA) community (Kounina et al., 2013) and focuses on the assessment of environ- mental indicators (i.e. eutrophication, ecotoxicity, human toxicity * Corresponding author. BETA Technology Centre: “U Science Tech”, University of Vic-Central University of Catalonia, Carrer de la Laura nº13, 08500 Vic, Barcelona, Spain. E-mail address: email@example.com (J. Colon). Contents lists available at ScienceDirect Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro http://dx.doi.org/10.1016/j.jclepro.2016.08.032 0959-6526/© 2016 Elsevier Ltd. All rights reserved. Journal of Cleaner Production 139 (2016) 504e516 and water scarcity). The LCA based WF has been extensively used for the assessment of different alternatives towards the mini- misation of product environmental impacts (Feng et al., 2014). However, the most widely accepted LCA methods disregard water- based impacts and water use (Canals et al., 2009). The application of different WF approaches can be confusing, resulting in different outcome even when the same data are used (Danielsson et al., 2015). Ridoutt et al. (2009) mentions that volu- metric methods, such as the WFN methodology (Hoekstra et al., 2011) are often misleading, since they do not consider the envi- ronmental impacts of water use. Zonderland-Thomassen and Ledgard (2012) propose to consider additional impacts related to water degradation and depletion A response to the ‘gap’ is given by the UNEP-SETAC WULCA project (Boulay et al., 2014) and recent ISO, aiming to set up a standardisation tool and guide towards a consistent assessment of the water-related environmental impacts locally, regionally and globally. The ISO follows an LCA-based approach. Another popular eco-indicator, speci ﬁ cally focused on the reduction of GHGs is the Carbon Footprint (CF), which is de ﬁ ned as the total GHG emissions of a product's whole life cycle (Pandey et al., 2011). ISO/TS 14067 (2013) develops standards and guide- lines for the standardisation of the carbon footprint methodology. The majority of research studies examine the environmental impacts of the dairy sector following the LCA methodology (Flysj€ o et al., 2011a). Among them, a signi ﬁ cant number of studies focus on the comparison of organic and conventional farming activities (Cederberg and Mattsson, 2000), considering site speci ﬁ c case studies (Castanheira et al., 2010), milk supply chain (Fantin et al., 2012) and on other milk based products, such as cheese, whey (Kim et al., 2013) and yoghurt (Gonzalez-García et al., 2013). The farm stage has been again the centre of interest in several stand- alone studies on carbon footprint. For example, Flysj€o et al. (2011b) examined the allocation methodologies for multi-output farming systems, whereas a recent study performed by de Leis et al. (2015) compared different farming production systems. Regarding the carbon footprint of dairy products, milk has mainly been examined (Flysj€o et al., 2014). Verge et al. (2013) assessed various dairy products in Canada and concluded that the carbon footprint de- pends on the energy requirements of the processing stages. The majority of the environmental studies in the dairy sector have identi ﬁ ed that climate change (GWP) is the main impact category that is connected to the dairy sector; particularly with the farm stage and feed production. On the other hand, current research regarding the water footprint of dairy products has almost exclu- sively focused on the farm stage. Recently, the WF for the produc- tion of conventional and organic milk of Brazilian dairy farming was assessed (Palhares and Pezzopane, 2015), following the WFN methodology (Hoekstra et al., 2011). The authors concluded that the volumetric water degradation measured by the grey water footprint is not a distinctive indicator of water pollution. A com- bination of different methods for WF assessment for dairy farming in New Zealand has been applied in the work of Zonderland- Thomassen and Ledgard (2012). The authors demonstrated that freshwater availability, depletion and degradation are essential components in order to assess the comprehensive WF pro ﬁ le of dairy farming. The LCA-based WF is becoming an increasingly popular concept. Following the ISO 14046 Standards (2014), Jing et al. (2014) assessed the water availability footprint of dairy sys- tems in Northeast China. In contrast to similar studies following the WFN methodology, the ﬁ ndings of their impact-oriented analysis, show that dairy products can be produced without contributing signi ﬁ cantly to the water scarcity of the region. De Boer et al. (2013) evaluated the impacts related with freshwater use in Dutch dairy farms and compared them with the respective ones related to broccoli production in Spain. The authors found that standalone water use volumetric assessments are not suf ﬁ cient to characterise the impacts from water use. The assessment of dairy products WF throughout their whole supply is limited mainly due to lack of primary data. This limitation has been revealed for the LCA-based water footprint studies, assessing the water-related environmental impacts of products. Only few environmental studies deal with fermented milk products (Gonzalez-García et al., 2013; White et al., 2008), such as yoghurt, whereas the impacts of yoghurt production on water resources and on greenhouse gas emissions have not yet been thoroughly Abbreviations ARc Fraction of cream BAT Best Available Techniques CF Carbon Footprint CIP Clean-in-place Fc g fat content in cream/100 g of skimmed milk produced F&B Food and Beverage Fsm g fat content in 100 g skimmed milk GHG Greenhouse Gas ISO International Organization for Standardisation LCA Life Cycle Assessment LCI Life Cycle Inventory Mc Production (%) of cream/kg skimmed milk Pc g of protein content in cream per 100 g of skimmed milk produced Psm g protein in 100 g of skimmed milk PCR Product Category Rules WF Water footprint WFN Water Footprint Network WTA Withdrawal to availability WULCA Water use in life cycle assessment (working group) WWTP Wastewater treatment plant Impact categories WU Water Use (L H2O) WC Water Consumption (L H2Oeq) WS Water Scarcity (L H2Oeq) FETP Freshwater eutrophication potential (kg Peq) FEP Freshwater ecotoxicity (1,4-Dbeq) AP Acidi ﬁ cation potential (kg PO43eq) HH Disability adjusted life years (DALY) GWP Global warming potential (kg CO2eq) Yoghurt types SNF Set natural full yoghurt SFL Set ﬂ avoured full yoghurt SKN Skimmed natural yoghurt SKFL Skimmed ﬂ avoured yoghurt GN Greek-style natural yoghurt GFL Greek-style ﬂ avoured yoghurt LN Liquid natural yoghurt LF Liquid ﬂ avoured yoghurt IN Ice-cream natural yoghurt IFL Ice-cream ﬂ avoured yoghurt V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516 505 examined. Even in the LCA focused studies of milk products, pro- cessing is considered as ‘a black box’ (Gonzalez-García et al., 2013); this means that materials and energy ﬂ ows have not yet been allocated to speci ﬁ c processes. The allocation of impacts to pro- duction processes is the ﬁ rst step for the improvement of the environmental pro ﬁ le of the industrial sector and the establish- ment of environmental policy (Kouchaki-Penchah et al., 2016a). Additionally, the impacts from life-cycle water use have not been properly addressed in LCA studies. A recent review on LCA studies for milk production published by Baldini et al. (2016) showed that out of 29 studies considered, only 2 addressed impacts from freshwater eutrophication and 4 from freshwater ecotoxicity, while the hotspots for water use have not yet been suf ﬁ ciently examined. The quanti ﬁ cation and evaluation of the WF of dairy products following the ISO and LCA methodology remains a gap in current literature (Tillotson et al., 2014). The main objective of this study is to quantify the water and carbon footprint of a yoghurt processing plant and to assess the water and carbon related environmental impacts. The degradative and consumptive water use of the dairy plant is assessed, providing an analysis of the whole spectrum of the water-related environ- mental impacts in compliance with the LCA-based ISO 14046. Therefore, the water and carbon hotspots are identi ﬁ ed following a ‘cradle to gate’ approach in order to propose measures for the mitigation of targeted environmental impacts. The selection of yoghurt is based both on the nutritive signi ﬁ cance of the product for the human diet and on the lack of environmental studies, concerning yoghurt manufacturing. 2. Materials and methods 2.1. Dairy plant Real data from a yoghurt manufacturing plant are used in order to perform a complete assessment of the impacts of different types of yoghurts on freshwater resources deprivation and depletion taking into consideration both ecosystems and human health on a product and process level. La Fageda is the second largest yoghurt producer in Catalonia. The dairy industry is located in La Garrotxa, in the middle of the Volcanic Natural Park Zone. The company produces various dairy products, managing the whole supply chain from agricultural land and feed production up to the milk pro- cessing and distribution. Core processing stages include produc- tion, packaging and storage of yoghurts, dairy desserts, ice-creams and marmalades. The yoghurt types examined in this work include: the set type fermented and cooled in the package, the stirred type fermented in tanks and cooled before packaging, the drinking type, where the coagulum is liquidated before packaging and the frozen type fermented in tanks and refrigerated like ice cream. The case study of La Fageda focuses on cultured milk products and on set, Greek-style, liquid and ice-cream yoghurt that the company produces. The milk is produced, processed into the ﬁ nal product, packaged onsite and then distributed to several areas in Catalonia. In 2014, the company produced approximately 6274 t of fermented milk products and 7851 t of dairy products in total. 2.2. LCA-based water and carbon footprint assessment 2.2.1. Functional unit Product Category Rules (PCR) setting standards towards envi- ronmental impact assessment for yoghurt were identi ﬁ ed and followed. According to yoghurt's valid PCR (EPD, 2014) the selected functional unit is 1 kg of yoghurt produced in the plant. 2.2.2. System boundaries and processing stages The LCA ‘cradle to gate’ approach was adopted (Finnveden et al., 2009) in order to calculate the water and carbon footprint of selected dairy products taking into consideration the upstream and the core processes of the dairy plant (Fig. 1) as proposed by yo- ghurt's PCR (EPD, 2014). Given that the study focused on the environmental pro ﬁ le of the diary plant the downstream processes were out of the scope of the study. The upstream processes include raw milk, other auxiliary products and packaging materials pro- duction. The core processes mainly refer to the yoghurt production stages including the waste management scheme of the company. Fig. 2 shows a simpli ﬁ ed ﬂ owchart of milk processing (core processes) in the targeted dairy plant. The raw milk is produced on site in La Fageda dairy farm and it is received through pipes in the processing plant. About 15% of the raw milk undergoes a separation process, where skimmed milk (less than 1.5% fat) is separated from the cream (40% fat content). The unseparated whole milk (4% fat) is used for the production of whole set yoghurts and skimmed milk for the production of skimmed yoghurts, while whole milk is mixed with cream in the Greek yoghurt production line. Then all streams are mixed with additives including ﬂ avours, sweeteners and dry powder milk to increase yoghurt consistency. Following the mixing the homogenisation process takes place at high pressure at temperature (65 C) to stabilise the mixture and prevent the fat suspension. Pasteurisation is a heat treatment process at 95 C that destroys the microorganisms in the milk and denatures whey proteins improving the yoghurt's texture. In order to reduce energy consumption, the heated milk that exits pas- teurisation is recirculated back to the system to pre-heat raw milk and enhance homogenisation. After recirculation, the milk mixture is cooled at 10 C using ice water. Then, the milk is heated again at 45 C with steam before the addition of bulk starter that enhances the fermentation process. As illustrated in Fig. 2, the fermentation process differs depending on the type of yoghurt. All products are preserved in refrigerators until distribution. The electricity network provides the energy required for the processing machines in la Fageda. Cleaning operations in La Fageda are divided into automatic cleaning in place for the tanks, pipes and packaging machines, cleaning in place for the pasteurizer and manual cleaning opera- tions. Cleaning in place is an automatic process, where rinsing water and detergent solutions are circulated through tanks, pipes and process lines. Other processes that are indirectly connected with yoghurt manufacturing include the treatment of the well water that is abstracted and puri ﬁ ed in order to comply with the required water quality standards and then used in the processing stages and the wastewater treatment plant at La Fageda premises. Both aforementioned processes are considered as ‘black box’ for the analysis and only total input and output materials and energy are taken into consideration. 2.2.3. Assumptions Avoided impacts regarding the recycling of the produced waste in the plant are taken into consideration for the impact categories of the life-cycle of yoghurt. Environmental credits from recycled materials, mainly packaging waste in the plant, are attributed to the waste management process. Data for the ﬁ nal disposal of waste and the type of treatment were supplied by the company. In addition, the cut-off rule has been applied according tothe ISO 14040 and ISO 14044 standards (2006). Thus, materials and processes that ac- count for less than 1% to the impact categories (i.e. impacts related to the infrastructure of the dairy plant) are not considered. Given that the plant manufactures dairy products that are not examined in this study, only 50% of the impacts related to the of ﬁ ce operations (i.e. electricity consumption) are included in the V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516506 Fig. 1. Supply chain of yoghurt-system boundaries. Fig. 2. Flow chart of the processing stages of yoghurt products in La Fageda. V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516 507 assessment. Due to the lack of available information, the trans- portation of materials from their place of production to Spain was excluded from the analysis. 2.2.4. Allocation of burdens Each yoghurt type in La Fageda is a unique product in the pro- duction stage thus, no burden allocation has been applied. The majority of the production lines are independent of other lines (i.e. iced yoghurts follow different production line from stirred yo- ghurts) enabling the individual calculation of electricity and water consumption for each product. However, when a machine is used for the production of more than one yoghurt group, electricity and direct water consumption have been mass allocated within each group. Heating requirements (woodchips, propane) and auxiliary products that are used for more than one product groups (i.e. chemicals) as well as water treatment and water related emissions have been also mass allocated. Additionally, in La Fageda, part of the raw milk is separated into cream and skimmed milk. Skimmed milk is used for the production of skimmed yoghurts, while cream is used in Greek and iced yo- ghurts. Mass of protein and fat allocation is considered in order to re ﬂ ect the physical relationships between the mass of protein and fat of the partitioning of co-products in the same production line as proposed by yoghurt's PCR (EPD, 2014). Consequently, 53% of the impacts of raw milk and of energy requirements for separation are allocated to skimmed milk for set yoghurt and 47% to cream for Greek yoghurt (Eq. (1)). ARc ¼ ðFc þPcÞMcð%Þ ðFc þPcÞMcð%ÞþðFsm þPsmÞ100ð%Þ (1) where ARc is the fraction of cream, Fc (g) is the fat content in cream per 100 g of skimmed milk produced, Pc (g) is the protein content in cream per 100 g of skimmed milk produced, Mc (%) is the produc- tion of cream per 1 kg skimmed milk produced and Fsm and Psm (g) are the fat and protein content of 100 g of skimmed milk, respectively. 2.3. Water footprint assessment 2.3.1. Impact assessment methodology The impact assessment methods for the water footprint were selected to cover impacts associated with quantitative and degra- dative concerns, as required by ISO 14046. Therefore, the water comprehensive footprint pro ﬁ le used in this study consists of 6 midpoint impact category indicators divided in (i) water availability indicators (water use, water consumption, water scarcity) and water degradation indicators (freshwater eutrophication, fresh- water acidi ﬁ cation and freshwater ecotoxicity). In addition, one endpoint impact category indicator (Human Health water scarcity) has been applied. Methods associated with quantitative water includes (i) water use (expressed in m3 of water) de ﬁ ned as any water withdrawal, water release or other human activity within the drainage basin, (ii) water consumption (expressed in m3 of water) de ﬁ ned as water removed, but not returned to the same drainage basin due to evaporation, transpiration, integration to a product or release into a different drainage basin and (iii) water scarcity (expressed in m3eq of water) which was modelled according to the method proposed by P ﬁ ster et al. (2009). The annual water stress index (WSI) developed from P ﬁ ster et al. (2009), was based on a withdrawal-to- availability ratio, and was used as a characterisation factor. The Spanish average WSI value of 0.7147 m3eq/m3 was applied to pro- cesses that occurred throughout Spain (e.g., electricity, milk pro- duction, etc.) or in unknown locations. Methods associated with water degradation were chosen in order to cover the most common impact pathways. A detailed description of these methods can be found in the literature (Pr e, 2016). Brie ﬂ y, freshwater eutrophication is modelled based on the ReCiPe methodology (Goedkoop et al., 2009) and expressed as ki- logram Peq. Freshwater acidi ﬁ cation is characterized with the IMPACT 2002 þ methodology (Jolliet et al., 2003) and expressed as kilogram SO2eq. Freshwater ecotoxicity is also modelled based on the ReCiPe methodology (Goedkoop et al., 2009) and expressed as Kilogram 1,4-DBeq. At the endpoint, water availability impact assessment methods model the impact pathways from user deprivation (agriculture, domestic, and/or ﬁ sheries) to human health in disability-adjusted life year (DALY) and was calculated from methodology proposed by P ﬁ ster et al., 2009. 2.3.2. Water consumption calculation methodology Different approaches for the calculation of water consumption have been applied for (i) raw milk, milk related products and other ingredients, (ii) other secondary ﬂ ow (energy, packaging, transport, etc.) and (iii) direct water consumption. Blue water consumption for milk, milk related products and other ingredients were calculated based on the country's spe- ci ﬁ c data reported by Mekonnen and Hoekstra (2010a, 2010b). Blue water consumption of all the other secondary ﬂ ows was calculated using water balances in Ecoinvent 3.1 following the methodology proposed by P ﬁ ster et al. (2015). Brie ﬂ y, net freshwater consumption of the secondary ﬂ ows was calculated by deducting water outputs to freshwater bodies from fresh- water inputs. Direct blue water consumption was modelled using the con- sumption factors proposed by Flury et al. (2013) expressing the fraction of water consumed in relation to the water used in the production process. Consumptions factors of 0.05 for cooling water, 0,1 for the other production processes (e.g. cleaning wa- ter) and 1 for water embodied in products were used. 2.4. Carbon footprint determination ISO/TS 14067 (2013) aims to develop standards and guidelines for the standardisation of the carbon footprint methodology. It is also known as Global Warming Potential (GWP) and includes the assessment of all GHG emissions throughout a product's supply chain and their transformation into CO2-equivalents based on their global warming potential (R€ o€os et al., 2013). The transformation into a common unit enables the calculation and the weighting of a product's life cycle emissions that contribute to the greenhouse gas effect (Weidema et al., 2008). The impact category of the Global Warming potential assesses emissions contributing to the Greenhouse effect and thus to global warming. It was extracted from Recipe methodology for a time horizon of 100 years in kg CO2eq. In the Recipe methodology, the GHG emissions attributable to climatechangewere modelled based on a simpli ﬁ ed version of the Fund model (Tol, 2002). The latter is one of the best available models and it is based on highly acknowledged studies (Goedkoop et al., 2009). The Carbon Foot- print of the ﬁ nal product consists of the aggregated CO2 equivalents emitted in the atmosphere through the whole life cycle of yoghurt production. 2.5. Global inventory 2.5.1. Primary data The challenges for the development of the inventory have been V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516508 emphasised by several researchers (Finnveden et al., 2009). Pro- cessing is often considered as a black box (Gonzalez-García et al., 2013) in LCA studies of milk products, mainly due to the limited availability of reviewed data. Meticulous consideration is required when selecting proxy data for the inputs and outputs of the pro- duction processes since differences in the technology and regula- tions can signi ﬁ cantly affect the results (Kouchaki-Penchah et al., 2016b). The development of the life-cycle inventory was based on primary data obtained through visits in the factory and data pro- vided by process engineers and operators. The data collection period lasted one year, from mid 2014 to mid 2015. Real data for products composition, auxiliary materials, energy and water con- sumption, solid waste and wastewater management and transport distances were obtained for this period. The average global in- ventory collected for the yoghurt processing plant is illustrated in Table 1, where all ingredients are summed up and average amounts are used per kg of yoghurt produced. However, in order to assess the impact categories of different yoghurt groups (i.e. set natural yoghurts) differentiated data for the speci ﬁ c amounts of raw in- gredients, packaging materials, energy and water requirements have been provided by the dairy facility (Table 2). The average transportation distances for the raw ingredients (i.e. 40 km), the packaging materials (i.e. 8 km) and the waste management scheme (i.e. 67 km) were selected based on the location of the suppliers and used in all the dairy products studied in this work. 2.5.2. Secondary data The material and energy inventories in Tables 1 and 2 include also background data for all the upstream processes. In the current study, secondary data of processes regarding raw milk, ingredients, different types of energy, packaging materials, waste treatment techniques and transportation were obtained from the highly acknowledged Ecoinvent 3.1 database. The only exception is skimmed milk powder, which was modelled assuming skimmed milk as the main ingredient, while, processing electricity and heat requirements were de ﬁ ned according to the study of Van Zeist et al. (2012). Also, as stated in section 2.3.2, the amount of blue water used and consumed for raw milk and ingredients/additives pro- ductionwere derived from Mekonnen and Hoekstra (2010a, 2010b) due to limited availability of regional data concerning agricultural and milk ingredients. 3. Results and discussion 3.1. Water and carbon footprint of yogurt's production process Analysis of the different yoghurt types processed in La Fageda, indicated signi ﬁ cant differences in the impact categories for each yoghurt type. Table 3 shows the environmental results for the impact categories that were assessed (WU: water use, WC: water consumption, WS: Water Scarcity, FETP: freshwater eutrophication potential, FEP: freshwater ecotoxicity potential, AP: acidi ﬁ cation potential, HH: human health, GWP: global warming potential), in relation to the functional unit (1 kg of yoghurt at dairy plant gate) for the examined yoghurt types. In most cases the differences are related to the different concentrations of additives in each yoghurt type and the different packaging materials used. The product that has the lowest water-related environmental impacts and GHG emissions is the liquid natural yoghurt, while the most environ- mental intensive product is the Greek-style yoghurt and the ice- cream yoghurt (Table 3). This is mainly because liquid natural yoghurt does not contain milk powder, cream or fruit additives that are major contributors to the water pro ﬁ le of the ﬁ nal products. The average total water consumed for the production of 1 kg of fermented product in a Spanish dairy plant is 203.7 L H2Oeq.Given that in the case of La Fageda, more than 99% of water consumption is sourced from indirect water inputs it is important to consider both direct and ‘virtual water’ consumption of dairy products; in- direct water inputs, correspond to an average Spanish case. The direct contribution of La Fageda processing techniques includes the energy and water consumption in the core processes and ranges from 0.2% (representing WS for Greek natural yoghurt) to 7.2% (representing WU for skimmed natural yoghurt) for the majority of the impact categories, as shown in Table 3. Table 1 Global average inventory for the production of 1 kg of yoghurt in La Fageda. Input from technosphere Input from environment Materials Water 3.5 L Raw milk (from dairy farm) 0.89 kg Milk powder 0.02 kg Output to technosphere Additives Yoghurts Sugar 0.05 kg Set yoghurt 0.88 kg Strawberry concentrate 1.2E-03 kg Greek-style yoghurt 0.06 kg Lemon concentrate 9.6E-04 kg Liquid yoghurt 0.06 kg Banana concentrate 9.0E-04 kg Ice-cream yoghurt 0.005 kg Cleaning agents Energy Nitric acid 2.1E-03 kg Electricity (Dairy factory) 0.15 kWh Sodium hydroxide 4.2E-03 kg Transport Packaging materials Truck 3.5e16 t (additives factorydinput) 40 km Polystyrene cups 0.03 kg Truck 3.5e16 t (packagingdinput) 8 km Ice-cream cups 3.7E-04 kg Truck 3.5e16 t (dairy factorydwaste) 67.3 km Cardboard 0.02 kg Glass Bottle 0.02 kg Waste to Treatment Bottle labels 9.8E-05 kg Plastics to recycling 1.0E-02 kg Metal caps for bottles 2.9E-04 kg Cardboard to recycling 4.6E-03 kg Seals for bottles 4.0E-05 kg Glass to recycling 9.8E-03 kg Lids for yoghurt cups 2.8E-03 kg Municipal waste undifferentiated 1.1E-02 kg Cartons 0.02 kg Water emissions (after treatment) Fuels COD 75 mg O2/L Propane gas 0.008 kg Phosphorous 0.007 mg/L Biomass 0.09 kg NKT 28 mg/L V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516 509 The embodied water in raw milk is responsible for up to 90e95% of the total water consumption in the ﬁ nal products. For example, 95% of the total water consumption in liquid natural yoghurts is sourced solely from milk. The same pattern has been observed by several authors (Hoekstra et al., 2011; Zonderland-Thomassen and Ledgard, 2012; Palhares and Pezzopane, 2015). Raw milk and cream are the major contributors of the majority of impact categories for all types of examined yoghurts (on average 86% WS, 87% FEP, 82% Table 2 Speci ﬁ c inventory of ingredients, additives, packaging materials and energy and water use for the production of 1 kg of each speci ﬁ c selected product in La Fageda. SNF: set natural full, SFL: set full ﬂ avoured, SKN: skimmed natural, SKFL: skimmed ﬂ avoured,GN: Greek-style natural,GFL: Greek-style ﬂ avoured, LN: liquid natural,LF: liquid ﬂ avoured, IN: ice-cream natural, IF: ice-cream ﬂ avoured. NSF (kg/kg yoghurt) SFL (kg/kg yoghurt) SKN (kg/kg yoghurt) SKFL (kg/kg yoghurt) GN (kg/kg yoghurt) GFL (kg/kg yoghurt) LN (kg/kg yoghurt) LFL (kg/kg yoghurt) IN (kg/kg yoghurt) IFL (kg/kg yoghurt) Ingredients Raw milk (from dairy farm) 0.97 0.87 ee0.72 0.76 0.9 0.9 0.18 0.12 Skimmed milk (after separation) ee0.93 0.93 e e e e e e Milk powder 0.004 0.026 0.03 0.03 0.004 0.026 e 0.002 0.02 0.02 Cream eeee 0.185 0.185 ee0.11 0.11 Set natural yoghurt e e e e e e e e0.5 0.47 Additivesa Sugar e 0.086 eee0.08 0.1 0.1 0.13 0.11 Strawberry concentrate e 0.002 eee0.002 e 0.05 e 0.05 Lemon concentrate e 0.002 e 0.006 eee0.05 e 0.005 Banana concentrate e 0.002 e e e e e e e e Packaging materials Polystyrene cups 0.037 0.037 0.037 0.037 0.04 0.04 eeee Cardboard 0.038 0.038 0.038 0.038 0.038 0.038 ee0.058 0.058 Lids for yoghurt cups 0.003 0.003 0.003 0.003 0.003 0.003 eeee Cartons 0.035 0.035 0.035 0.035 0.034 0.034 0.17 0.17 0.022 0.022 Glass bottle e e e e e e0.36 0.36 ee Bottle labels e e e e e e0.002 0.002 ee Metal cups for bottles e e e e e e0.005 0.005 ee Seals for bottles e e e e e e0.0007 0.0007 ee Container e e e e e e 0.027 0.027 Ice-cream cups e e e e e e 0.007 0.007 Energy Electricity (Dairy factory) 0.15 0.13 0.14 0.14 0.14 0.15 0.13 0.13 0.14 0.15 Fuels Propane gas 0.009 0.008 0.009 0.009 0.009 0.009 0.008 0.08 0.009 0.009 Biomass 0.1 0.09 0.09 0.09 0.1 0.1 0.09 0.09 0.1 0.1 Input from environment Water 3.6 3.2 3.4 3.4 3.3 3.5 3.5 3.5 5.8 5 a Apart from sugar, only one additive is included as ingredient in each type of dairy product (e.g. SFL lemon yoghurt only have sugar and lemon concentrate as additives). Table 3 WF pro ﬁ le of 1 kg ﬁ nal product in La Fageda. T: Total Impact Assessment, D: % contribution of La Fageda processing plant, SNF: set natural full, SFL: set full ﬂ avoured, SKN: skimmed natural, SKFL: skimmed ﬂ avoured, GN: Greek-style natural, GFL: Greek-style ﬂ avoured, LN: liquid natural, LF: liquid ﬂ avoured, IN: ice-cream natural, IF: ice-cream ﬂ avoured. Impact Category WU L H2O WC L H 2Oeq WS L H2Oeq FETP kg Peq FEP 1,4-DBeq AP kg SO2eq HH DALY GWP kg CO2eq SNF T 218 199 142 2.47E-03 0.10 0.017 5.23E-09 1.89 D 5.1% 0.4% 0.3% 0.7% 3.2% 1.2% 3.1% 5.0% SFL T 234 214 152 2.52E-03 0.10 0.018 1.61E-08 1.98 D 4.7% 0.3% 0.3% 0.7% 3.1% 1.1% 1.0% 4.8% SKN T 153 137 97 1.71E-03 0.07 0.018 4.19E-09 1.41 D 7.2% 0.5% 0.4% 1.1% 4.5% 1.1% 3.9% 6.7% SKFL T 162 146 103 1.71E-03 0.07 0.018 9.52E-09 1.42 D 6.8% 0.5% 0.4% 1.1% 4.5% 1.1% 1.7% 6.6% GN T 346 326 232 4.02E-03 0.15 0.028 7.49E-09 2.92 D 3.5% 0.2% 0.2% 0.5% 2.2% 0.7% 2.4% 3.5% GFL T 356 334 239 4.11E-03 0.16 0.028 7.72E-09 2.98 D 3.4% 0.2% 0.2% 0.5% 2.2% 0.7% 2.3% 3.5% LN T 220 187 133 2.38E-03 0.09 0.029 1.41E-08 2.22 D 4.8% 0.4% 0.3% 0.7% 3.1% 0.7% 1.1% 4.0% LF T 222 190 135 2.38E-03 0.09 0.022 1.50E-08 2.23 D 4.7% 0.4% 0.3% 0.7% 3.1% 0.9% 1.0% 4.0% IN T 287 252 178 3.16E-03 0.12 0.029408 1.72E-08 2.65 D 5.4% 0.4% 0.4% 0.7% 3.0% 1.3% 1.1% 4.1% IFL T 307 273 190 3.12E-03 0.12 0.029648 2.23E-08 2.67 D 5.0% 0.4% 0.3% 0.7% 3.1% 1.2% 0.9% 4.0% Average 223 204 145 2.46E-03 0.10 0.018 1.10E-08 1.94 V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516510 FETP). Raw milk and cream are responsible for 91% and 81% of the water consumption of Greek-style and ice-cream yoghurt respec- tively. In addition, raw milk, which is the main ingredient of yoghurt, accounts on average for 63% of total GHG emissions of liquid yoghurt's supply chain, which increases for set yoghurt; 73% contribution was noticed. Similarly,1 kg of Greek-style yoghurt and ice-cream, raw milk and cream are responsible for 81% and 64% of the total GHG emissions. 3.2. Identiﬁcation and assessment of yogurt processing hotspots Even though the majority of the results identify the farming stage as the most important environmental hotspot, it is important to assess the impacts of the processing stages of yoghurt and pro- pose actions for the improvement of their environmental pro ﬁ le. Until now, most studies have focused on the farm stage of dairy products which has resulted in the signi ﬁ cant reduction of the environmental impact of milk production in many areas (Zonderland-Thomassen and Ledgard, 2012; Jing et al., 2014; Palhares and Pezzopane, 2015). In the UK for example, the water embodied in 1 kg of milk is 24 L, which is signi ﬁ cantly lower than the respective one for the case of Spain, where 1 kg of milk is equivalent to 198 L H2O consumed. This happens because the Spanish dairy farms are irrigated due to low annual precipitation (Lecina et al., 2010). The relative contributions to the impact cate- gories of the subsystems, excluding inputs of raw milk and conse- quently cream, are shown in Fig. 3. Raw ingredients and especially milk powder dominate in almost all impact categories (i.e. WC: 58e89%, FETP: 48e85%, FEP: 24e67%, AP: 17e58%). The absence of milk powder in liquid yoghurts is responsible for the differences in their water pro ﬁ le, where the impacts of the raw ingredients in all impact categories are negligible. Additionally, the packaging materials are identi ﬁ ed as ‘hotspot’ in the majority of the categories. Approximately 10.4e11 L of H2O are abstracted from freshwater resources for the packaging of 1 kg of cupped yoghurt respectively (20%e38% relative contribution) and 2.8e3LH 2O is not returned to the same catchment (8%e15% relative contribution to the water consumption). The environ- mental impacts of packaging materials are also strongly connected to water quality aspects that are a hotspot especially for liquid yoghurts; eutrophication (89%) and ecotoxicity (63%). The envi- ronmental impacts of packaging in yoghurt manufacturing have also been identi ﬁ ed as an environmental hotspot in other research works (Gonzalez-García et al., 2013). This is mainly sourced from the polystyrene used for the packaging of Greek and set yoghurt and the glass used for liquid yoghurt. Another important contributor to the environmental impacts is the energy requirements that are responsible for 16e35% of the Freshwater Ecotoxicity and 10e29% of Global Warming potential. In Spain 69% of the electricity mix consists of carbon-free sources (del Río Gonzalez, 2008); the latter indicates that higher contribution to GHGs emissions is expected for coal-dependant countries. Avoided impacts due to the waste management scheme applied in the plant, and the recycling of the plastic, glass and carton waste, range from 0.02% for the water scarcity impact category to 0.7% for the global warming potential. 3.3. Direct yoghurt processing impacts Direct water use and degradation, along with the energy re- quirements are mainly responsible for the environmental impacts of yoghurt processing. Even though the magnitude of water-related environmental impacts of yoghurt is mainly affected by the indirect water ﬂ ows (Fig. 3), direct water consumption in the plant also plays an important role. Dairy plants are characterized by high- energy demand due to the intensive and extensive range of heat- ing and cooling processes. Fig. 4 presents a breakdown of water, and electricity use as well as GWP in the plant for the production of 1 kg of Greek-style yoghurt. Direct water use accounts for 3.5 L of water abstracted from the local well and used for the production 1 kg of Greek-style yoghurt in La Fageda plant. Ice-cream production requires the highest direct water use volume, which is equal to 5.76 L of water. As shown in Fig. 4, out of 3.5 L H2O used for the production of 1 kg of Greek-style yoghurt, ~86% is directly connected to the processing stages. CIP and cleaning operations are the main contributors for direct water use consuming 2.4 L H2O per kg of yoghurt, while 0.33 L H2O and 0.23 L H2O are required for the production of 1 kg of yoghurt, due to cooling and heating respectively. Satisfactory results are obtained in the plant in terms water management and consumption. This is due to the water recirculation system of the CIP and milk recircu- lation instead of steam or cooling water, which is used during pasteurisation-homogenisation. According to the European Commission Directorate (2006) water consumption for the pro- cessing of fermented milk products, varies from 0.8 to 25 L/kg of processed milk. The estimated value of 3.5 L H2Oeq/kg yoghurt is below the average amount of 3.8 L H2Oeq/kg yoghurt that is pub- lished by Envirowise (undated). The average electricity requirements for the production of 1 kg of fermented products in La Fageda are 0.15 kWh. A biomass and a propane gas boiler are also used, consuming on average 0.21 kWh and 0.07 kWh respectively for the production of 1 kg of fermented products. The benchmark value that has been reported for the European dairy sector is 0.04e0.69 kWh per kg of liquid milk processed (1.08 kg of fermented product equivalent on average) from the electricity network and 0.05e0.46 kWh from fuels (Commission Directorate, 2006). In accordance with the literature ﬁ ndings (Rad and Lewis, 2014), pasteurisation is one of the most energy intensive steps (22% contribution). Moreover, cold tunnels, refrigerators and packaging result in high electricity consumption (26% and 22% respectively). When this work was carried out, several important modi ﬁ cations to the facility were being imple- mented; due to these modi ﬁ cations cold rooms were not properly insulated and doors remained open for hours intensifying the en- ergy use. All yoghurts types, present a similar pro ﬁ le in terms of electricity distribution. The application of biomass boiler in the plant is an ef ﬁ cient strategy to reduce GHG emissions of the pro- cessing stages by reducing the dependency on the national elec- tricity grid. Given that CF of the core processes in La Fageda is highly dependent on the energy consumption (electricity, fuels) of each process, the CF pro ﬁ le is similar to the pro ﬁ le of the electricity consumption. Overall, 0.1 kg CO2eq are emitted per kg of Greek ﬂ avoured yoghurt produced. Again, the most important hotspots are the refrigerators (25%), the pasteurisation (22%) and the pack- aging (22%). A study published by Thoma et al. (2013) on the GHGs emissions from milk production in the United States, followed the IPCC methodology and calculated 0.076 kg CO2eq/kg milk at the factory gate. 3.4. Packaging and ingredients Further examination of the contribution of each packaging material in the cupped Greek-style yoghurt shows that polystyrene cups constitute a ‘hotspot’ for almost all impact categories (Fig. 5a). The average cup weight in La Fageda is approximately 4 g; thus the production of 1 kg of Greek-style yoghurt requires 5.8 L H2O (51% relative contribution of all packaging materials) and carries an ‘embodied’ water load of 2.1 L H2Oeq (69% relative contribution). Polystyrene is another major contributor for human health impacts V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516 511 (77%). Additionally, it is estimated that outof 0.27 kg CO2eq life cycle emissions of packaging 1 kg cupped yoghurt, 53% are sourced from polystyrene based cups. The packaging of liquid yoghurt consists of glass bottles, aluminium based metal caps and labels made of polyethylene- based packaging ﬁ lm. The GHGs emissions originating from the manufacturing of bottled glasses are relatively high (0.38 kg CO2eq per kg of yoghurt). However, glass is a packaging material readily recyclable, theoretically inde ﬁ nitely, in all markets with high recycling rates in Europe (70% in 2013 as reported by the European Fig. 3. Relative contributions to the impact categories of the subsystems excluding raw milk inputs for all the examined fermented products. (a) Water use, (b) water consumption, (c) Water Scarcity Index, (d) freshwater eutrophication, (e) freshwater ecotoxicity, (f) aquatic acidi ﬁ cation, (g) human health, (h) GWP: global warming potential. V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516512 Container Glass Federation statistics; FEVE, 2015). Hence, glass bottles are not considered ‘environmental hotspots’ in the pack- aging family. Even though the main focus of the current work is not on the raw milk production and milk based products, it is important to identify the allocation of the impacts associated with yoghurt's raw ingredients (Fig. 5b). It was found that milk, cream and milk powder are a hot spot of Greek-style yoghurt's ingredients with indirect water load of 330.2 L H2O/kg yoghurt, accounting for 99% of the water consumption of the total raw ingredients. The highest contribution of non-dairy ingredients in terms of water consump- tion was 5.7% for set yoghurt. 3.5. The importance of the location - sensitivity analysis The assessment of the impact categories for the target yoghurt processing plant in Spain revealed that raw milk and milk-based raw materials are the main contributors to environmental im- pacts related to degradative and quantitative water use and GHG emissions (Fig. 6); this is in line with the ﬁ ndings of Gonzalez- García et al. (2013) and is mainly attributed to the farm stage of milk production. All the milk-based and agricultural products in Spain exhibit relatively high impacts concerning water use, con- sumption and scarcity; especially when considering that Spain is a country that has one of the lowest available water per capita among the European countries and it is considered as awater-stressed area (European Environmental Agency e EEA, 2008). However, the annual accumulated precipitation in Spain varies from 150 to 2500 mm, which signi ﬁ cantly affects the water avail- ability across the country (De Castro et al., 2005). Thus, it is important to consider real data, on a watershed lever, for the WF assessment of the upstream processes of dairy products and particularly for raw milk production. Garrotxa, where the targeted dairy plant is located, is characterized by humid Mediterranean climate, with average precipitation of 1000 mm uniformly distributed during the year (Isamat et al., 2008). As a result, cattle feeding and raw milk production in La Fageda relies much more on precipitation, rather than irrigation and the actual water use and consumption of raw milk production is expected to be less than the theoretical one used in this work that takes into account an average value for Spain. Hence, it is important to examine the effect of water availability of the plant in the ﬁ nal water pro ﬁ le of yoghurts. A sensitivity analysis was performed considering an identical dairy facility located in the UK with a reported average yearly precipitation <1000 mm (ranging from <600 to 3000 mm). The results of this new scenario (Scenario 2) were compared against the respective ones obtained for the baseline scenario, where the dairy plant is located in Spain. The characterisation factors of the impact cate- gories for Scenario 2 were UK-based (e.g. WSI ¼ 0.395 m3eq/m3). Additionally, the differences in the water pro ﬁ le of dairy products that are produced in water abundant and water scarce regions with different electricity mixes was examined. The total water con- sumption for the production of 1 kg of ﬂ avoured Greek-style yoghurt in an average UK dairy plant (Scenario 2), with the same ingredient sources, processing techniques, energy requirements and waste management schemes as in an average Spanish plant are 67.7 L H2Oeq instead of 356.1 L H2Oeq (Scenario 1). In the speci ﬁ c case of La Fageda, where the annual precipitation is closer to the average value of UK, total water consumption similar to this second scenario could be expected. As shown in Fig. 6, high precipitation (Scenario 2) and different techniques of raw milk production can increase the relative contribution of the direct impacts of the plant from 2% to 15% in terms of water use, from 0.2% to 1.6% in terms of water consump- tion and from 0.2% to 1.2% in terms of water scarcity impact cate- gory, even when raw milk is included in the analysis. Consequently, the contribution of the raw ingredients in Greek-style yoghurt re- duces the total water use from 93% to 62% in the Spanish dairy plant for the production of 1 kg of product. As a result, the contribution of the energy use and packaging materials to the total water use for the production of 1 kg of Greek-style yoghurt increases from 2% to 3% (Spanish scenario) to 14% and 15% (UK scenario) respectively. The same applies for the total water consumption, where the contribution of the packaging materials increases from 0.9% to 6.5% in scenario 2. Finally, the same trend is observed for the Water Scarcity Index, where in scenario 2 the contribution of the pack- aging materials increases from 0.6% to 8%. 3.6. Measures for the reduction of the environmental impact The total volume of water abstracted from the local well in La Fageda is 3.5 L H2O per kg of yoghurt that is equivalent to 21,950 m3 annually. This is mainly used for cleaning operations, as well as, for the cooling and heating requirements of the plant. The facility Fig. 4. Distribution of (a) water use (a) and (b) electricity consumption per processing stage of Greek-style yoghurt. V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516 513 operates a wastewater treatment plant (WWTP), where the wastewater is treated and returned to the same catchment; thus, the total direct water consumption during the production of 1 kg of yoghurt is 0.3 L H2O, which translates to 2000 m3 of water consumed on an annual basis. Even though the water consumption is low compared to available benchmarks (Envirowise, undated), the reduction of water consumption in the plant is essential; this is mainly due to costs associated with the treatment of the raw well water and wastewater. The good performance of La Fageda in terms of water use is mainly due to the installed CIP system, which is programmed to recycle the ﬁ nal rinse water following a pre-rinse water cycle. In addition, the system recirculates the streams that contain cleaning agents, resulting in the generation of small amounts of wastewater in every cleaning cycle, as proposed by the ‘Best Available Tech- niques - BAT’ (Korsstr€om and Lampi, 2008; Prasad and Pagan, 2006). During pasteurisation, the pasteurised hot milk is used for the pre-heating of the incoming milk and vice versa, achieving Fig. 5. Relative contributions of (a) the ingredients and (b) of packaging materials of for Greek-style ﬂ avoured yoghurts. Fig. 6. (%) (a) Water use, (b) water consumption and (c) water scarcity for a dairy plant located in Spain. V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516514 water and energy savings in the facility. A potential measure to further reduce the water requirements is to reuse speci ﬁ c waste streams; directly or after suitable pre-treatments. For example, the cooling processing water can be re-used for cleaning purposes, especially for the cleaning of the milk tank; s, several dairy com- panies, such as Arla Foods, are already using this minimisation technique. Concerning La Fageda plant, the application of this measure can result in water savings up to 0.33 L H2Oeq per kg of yoghurt; this amount is equivalent to 9% of the total water use. Additionally, there are several techniques that can be applied for the reduction of wastewater and milk loss; thus, reducing waste- water treatment related costs and ‘virtual’ water losses. Typically, milk losses in manual systems can be up to 2% of the total volume being processed (Singh et al., 2014). In La Fageda this translates into 125 t of milk loss annually, which is equivalent to 27,875 m3 of virtual water loss in the drain. The estimation has been done by considering the water that is included in the ﬁ nal product. Tech- nologies that spot the product-water transition points can signi ﬁ - cantly contribute to the reduction of milk losses in wastewater. At the beginning of the production process in the plant, the remaining water in pipes is pushed by milk ending to WWTP. This remaining water in the pipes though, is followed by a ‘mixing zone’ consisting of both water and product loss. The opposite occurs during the CIP, where the milk remaining in the pipes is pushed with water. In La Fageda, these transition points are visually identi ﬁ ed, which results in signi ﬁ cant product losses. Systems for automation and control, such as conductivity transmitters and optical sensors can be installed in existing pipelines with minor modi ﬁ cations. The latter can reduce the biochemical oxygen demand (BOD) in the generated wastewater by 30% and the product loss by 50% (Korsstr€om and Lampi, 2008). In parallel, the use of the aforementioned equip- ment enables the collection of the product water mixtures, which can be further processed. Recommendations for further processing include the drying of milk and cream water mixtures into powder, whereas mixtures containing fermented products can be used as animal feed after being concentrated in ultra ﬁ ltration units. Ac- cording to the European Commission Directorate (2006), recovery of products can also be achieved through the installation of pigging systems, especially for the cream pasteurisation line and the remaining water in the pipes after cleaning operations. The ‘pigs’ can be ﬁ tted in the pipe and when pushed (by water, compressed air or even the product), they empty the pipe removing the wall- deposits. The packaging stage has gathered increasing interest in the past years, while several strategies towards the reduction of its envi- ronmental impact focus on the weight of packaging material and the selection of recyclable and renewable materials (Prasad and Pagan, 2006). The main ‘hotspot’ in terms of packaging was iden- ti ﬁ ed in set and Greek yoghurt in La Fageda; polystyrene cups and aluminium lidding is used. The use of polystyrene cups increases the environmental impacts of the ﬁ nal product (e.g. 0.14 kg CO2eq are emitted per kg of cupped yoghurt produced). The application of innovative technologies in yoghurt packaging materials can reduce the environmental impacts related to packaging in the dairy sector. For instance, bioplastics have already been introduced in yoghurt packaging (Essel, 2012). This initiative reduced the product's packaging carbon footprint by 25% achieving certi ﬁ cation from both the International Sustainability & Carbon Certi ﬁ cation (ISCC) Association and the Institute for Agriculture and Trade Policy (IATP). 4. Conclusions The current work assesses the Water and Carbon Footprint of yoghurt following a ‘cradle to factory gate’ approach. Production of 1 kg of yogurt at a Spanish dairy plant requires 204 L of water consumption and emits 1.94 kg CO2eq. Raw milk and milk in- gredients are the main hotspots in the analysis, contributing more than 80% to all impact categories. Focusing on the core processes of the dairy plant, energy consumption and especially refrigerators, pasteurisation and packaging signi ﬁ cantly contribute in freshwater ecotoxicity, acidi ﬁ cation and GWP impact categories. On average, 3.5 L of direct water are used in the core processes for the pro- duction of 1 kg of yoghurt, 70% of which are connected to the cleaning requirements. The use of characterisation factors for water availability at a national level for the upstream processing of yoghurt manufacturing can lead to under or overestimation of the contribution of the dairy plant to the total environmental pro ﬁ le of dairy products. Characterisation factors on a watershed level are essential and should be a key activity of future works. The ﬁ ndings of this work revealed that the yoghurt's environmental pro ﬁ le can be enhanced through the reduction of product losses and waste generation in the processing stages. However, there is still limited information on the virtual water losses related to the waste generated from the food industry. The developed method can be transferred to other dairy plants, allowing the application of water and carbon footprints as well as LCA in the yogurt's production process. Acknowledgments This work was funded by the SAVE-VIRTUAL-WATER project; Pump-priming award of Brunel University London. The authors would like to thank La Fageda and Laura Millastre for providing all the necessary data to conduct the research. References Baldini, C., Gardoni, D., Guarino, M., 2016. A critical review of the recent evolution of life cycle assessment applied to milk production,. J. Clean. Prod. http:// dx.doi.org/10.2016/jclepro.2016.06.078. Boulay, A.-M., Bare, J., Benini, L., Berger, M., Klemmayer, I., Lathuilliere, M., Loubet, P., Manzardo, A., Margni, M., Nú~nez, M., 2014. Building Consensus on a Generic Water Scarcity Indicator for LCA-based Water Footprint. Preliminary Results. WULCA. Canals, L.M., Chenoweth, J., Chapagain, A., Orr, S., Anton, A., Clift, R., 2009. Assessing freshwater use impacts in LCA: Part Idinventory modelling and characterisa- tion factors for the main impact pathways. Int. J. Life Cycle Assess. 14, 28e42. Castanheira, E., Dias, A.C., Arroja, L., Amaro, R., 2010. The environmental perfor- mance of milk production on a typical Portuguese dairy farm. Agric. Syst. 103, 498e507. Cederberg, C., Mattsson, B., 2000. Life cycle assessment of milk productionda comparison of conventional and organic farming. J. Clean. Prod. 8, 49e60. Commission Directorate, 2006. General JRC Institute for Prospective Technological Studies Integrated Pollution Prevention and Control. Available at: eippcb.jrc.ec. europa.eu/reference/BREF/fdm_bref_0806.pdf (accessed 03.03.16.). Danielsson, L., Nilsson, A., Rydberg, T., Hall, S., 2015. EcoWater report - comparing water footprint methods. In: SvenskaMilj€oinstitutet (Swedish Environmental Research Institute) (Ed.), The Importance of a Life Cycle Approach in Assessing Water Footprint - IVL. De Boer, I.J., Hoving, I.E., Vellinga, T.V., Van de Ven, G.W., Leffelaar, P.A., Gerber, P.J., 2013. Assessing environmental impacts associated with freshwater consump- tion along the life cycle of animal products: the case of Dutch milk production in Noord-Brabant. Int. J. Life Cycle Assess. 18, 193e203. De Castro, M., Martín-Vide, J., Alonso, S., 2005. The climate of Spain: past, present and scenarios for the 21st century. In: A Preliminary General Assessment of the Impacts in Spain Due to the Effects of Climate Change. Spanish Ministry of Environment, Madrid, 162. de Leis, C.M., Cherubini, E., Ruviaro, C.F., da Silva, V.P., do Nascimento Lampert, V., Spies, A., Soares, S.R., 2015. Carbon footprint of milk production in Brazil: a comparative case study. Int. J. Life Cycle Assess. 20, 46e60. del Río Gonzalez, P., 2008. Ten years of renewable electricity policies in Spain: an analysis of successive feed-in tariff reforms. Energy Policy 36 (8), 2917e2929. Envirowise, undated. Benchmarking Water Use in Dairies. Available at: www.wrap. org.uk/sites/ ﬁ les/wrap/EN740_ ﬁ nal.pdf. (accessed 01.04.16.). EPD-Environmental Product Declaration, 2014. Yoghurt, Butter and Cheese, Product Category Rules According to ISO 14025:2006. Valid Until: 2016-09-17 Available at: www.environdec.com/en/EPD-Search/?query¼yoghurt (accessed 10.07.16.). Essel, R., 2012. Meta-analysis of Life Cycle Assessments for PLA (Munich). European Environmental Agency e EEA, 2008. Water Scarcity. Available at: www. V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516 515 eea.europa.eu/themes/water/featured-articles/water-scarcity (accessed 07.02.16.). Falkenmark, M., 2008. Water and sustainability: a reappraisal. Environment 50 (2), 4e16. Fang, K., Heijungs, R., de Snoo, G.R., 2014. Theoretical exploration for the combi- nation of the ecological, energy, carbon, and water footprints: overview of a footprint family. Ecol. Indic. 36, 508e518. Fantin, V., Buttol, P., Pergref ﬁ , R., Masoni, P., 2012. Life cycle assessment of Italian high quality milk production.A comparison with an EPD study. J. Clean. Prod. 28, 150e159. FAO, 2010. Greenhouse Gas Emissions from the Dairy Sector: a Life Cycle Assess- ment. Available at: www.fao.org/docrep/012/k7930e/k7930e00.pdf (accessed 13.12.15.). Feng, C., Gao, X., Tang, Y., Zhang, Y., 2014. Comparative life cycle environmental assessment of ﬂ ue gas desulphurization technologies in China. J. Clean. Prod. 68, 81e92. FEVE, 2015. Glass Recycling Hits 73% in the EU. Available at: www.feve.org/FEVE- Glass-Recycling-Stats-Year-2013/ (accessed 19.02.16.). Finnveden, G., Hauschild, M.Z., Ekvall, T., Guinee, J., Heijungs, R., Hellweg, S., Koehler, A., Pennington, D., Suh, S., 2009. Recent developments in life cycle assessment. J. Environ. Manag. 91, 1e21. Flury, K., Frischknecht, R., Jungbluth, N., Mu~noz, I., 2013. Recommendation for Life Cycle Inventory Analysis for Water Use and Consumption. ESU Services. Flysj€o, A., Cederberg, C., Henriksson, M., Ledgard, S., 2011a. How does co-product handling affect the carbon footprint of milk? case study of milk production in New Zealand and Sweden. Int. J. Life Cycle Assess. 16, 420e430. Flysj€o, A., Henriksson, M., Cederberg, C., Ledgard, S., Englund, J.-E., 2011b. The impact of various parameters on the carbon footprint of milk production in New Zealand and Sweden. Agric. Syst. 104, 459e469. Flysj€o, A., Thrane, M., Hermansen, J.E., 2014. Method to assess the carbon footprint at product level in the dairy industry. Int. Dairy J. 34, 86e92. Garnett, T., 2011. Where are the best opportunities for reducing greenhouse gas emissions in the food system (including the food chain)? Food Policy 36, S23eS32. Goedkoop, M., Heijungs, R., Huijbregts, M., De Schryver, A., Struijs, J., van Zelm, R., 2009. ReCiPe 2008da Life Cycle Impact Assessment Method Which Comprises Harmonised Category Indicators at Themidpoint and the Endpoint Level. VROMeRuimte en Milieu. Ministerie van Volkshuisvesting, Ruimtelijke Orde- ning en Milieubeheer. www.lcia-recipe.net. Gonzalez-García, S., Castanheira, E.G., Dias, A.C., Arroja, L., 2013. Environmental life cycle assessment of a dairy product: the yoghurt. Int. J. Life Cycle Assess. 18, 796e811. Hallstr€om, E., R€ o€os, E., B€orjesson, P., 2014. Sustainable meat consumption: a quantitative analysis of nutritional intake, greenhouse gas emissions and land use from a Swedish perspective. Food Policy 47, 81e90. Hoekstra, A.Y., Chapagain, A.K., Aldaya, M.M., 2011. The Water Footprint Assessment Manual. Isamat, E.B., Bosch, J.F., Carrera, J.M.M., Ramos, B.P., 2008. Agrobiodiversity con- servation in the Garrotxa Volcanic Zone Natural Park, Spain: experience and recommendations for future directions. In: Protected Landscapes and Agro- biodiversity Values. Protected Landscapes and Seascapes, vol.1. IUCN & GTZ, 33. ISO 14040, 2006. Environmental Management-life Cycle Assessment-principles and Framework. ISO 14044, 2006. Environmental Management-life Cycle Assessment- Requirements and Guidelines. ISO 14046, 2014. Environmental Management-water Footprint-principles, Re- quirements and Guidelines. ISO/TS 14067, 2013. Greenhouse Gases-Carbon Footprint of Products-Requirements and Guidelines for Quanti ﬁ cation and Communication. Jing, H., Ling, D., Ridoutt, B.G., HaiLin, Z., Fu, C., Schenck, R., Huizen, D., 2014. Water availability footprint of dairy production in northeast China. In: Proceedings of the 9th International Conference on Life Cycle Assessment in the Agri-food Sector (LCA Food 2014). San Francisco, California, USA, 8-10 October, 2014. American Center for Life Cycle Assessment, pp. 587e594. Jolliet, O., Margni, M., Charles, R., Humbert, S., Payet, J., Rebitzer, G., 2003. Presenting a new method IMPACT 2002 þ: a new life cycle impact assessment method- ology. Int. J. Life Cycle Assess. 8, 324e330. Kim, D., Thoma, G., Nutter, D., Milani, F., Ulrich, R., Norris, G., 2013. Life cycle assessment of cheese and whey production in the USA. Int. J. Life Cycle Assess. 18, 1019e1035. Korsstr€om, E., Lampi, M., 2008. Best Available Techniques (BAT) for the Nordic Dairy Industry. Available at: eldri.ust.is/media/skyrslur2002/BAT_mjolkuridn_2001- 586.pdf (accessed 26.02.16.). Kouchaki-Penchah, H., Shari ﬁ , M., Mousazaeh, H., Zarea-Hosseinabadi, H., 2016a. Life cycle assessment of medium-density ﬁ berboard manufacturing process in Islamic Republic of Iran. J. Clean. Prod. 112, 351e358. Kouchaki-Penchah, H., Shari ﬁ , M., Mousazaeh, H., Zarea-Hosseinabadi, H., Nabavi- Pelesaraei, N., 2016b. Gate to gate life cycle assessment of ﬂ at pressed parti- cleboard production in Islamic Republic of Iran. J. Clean. Prod. 112, 343e350. Kounina, A., Margni, M., Bayart, J.B., Boulay, A.M., Berger, M., Bulle, C., Frischknecht, R., Koehler, A., Canals, L.M., Motoshita, M., 2013. Review of methods addressing freshwater use in life cycle inventory and impact assess- ment. Int. J. Life Cycle Assess. 18, 707e721. Lecina, S., Isidoro, D., Playan, E., Aragües, R., 2010. Irrigation modernization and water conservation in Spain: the case of Riegosdel Alto Aragon. Agric. Water 97, 1663e1675. Manzardo, A., Ren, J., Piantella, A., Mazzi, A., Fedele, A., Scipioni, A., 2014. Integration of water footprint accounting and costs for optimal chemical pulp supply mix in paper industry. J. Clean. Prod. 72, 167e173. McGlade, J., Werner, B., Young, M., Matlock, M., Jefferies, D., Sonneman, G., Aldaya, M.M., P ﬁ ster, S., Berger, M., Farell, C., 2012. Measuring water use in a green economy. In: International Resource Panel (UNEP) (Ed.), A Report of the Working Group on Water Ef ﬁ ciency. UNEP. Mekonnen, M.M., Hoekstra, A.Y., 2010a. The green, blue and grey water footprint of crops and derived crop products. In: UNESCO-IHE (Ed.), Value of Water, Research Report SeriesTwente Water Centre, University of Twente, Enschede, The Netherlands, 47. Mekonnen, M.M., Hoekstra, A.Y., 2010b. The green, blue and grey water footprint of farm animals and animal products. In: UNESCO-IHE (Ed.), Value of Water, Research Report SeriesTwente Water Centre, University of Twente, Enschede, The Netherlands, 48. Milani, F.X., Nutter, D., Thoma, G., 2011. Invited review: environmental impacts of dairy processing and products: a review. J. Dairy Sci. 94, 4243e4254. Olmez, H., Kretzschmar, U., 2009. Potential alternative disinfection methods for organic fresh-cut industry for minimizing water consumption and environ- mental impact. Food Sci. Technol. 42, 686e693. Palhares, J.C.P., Pezzopane, J.R.M., 2015. Water footprint accounting and scarcity indicators of conventional and organic dairy production systems. J. Clean. Prod. 93, 299e307. Pandey, D., Agrawal, M., Pandey, J.S., 2011. Carbon footprint: current methods of estimation. Environ. Monit. Assess. 178, 135e160. P ﬁ ster, S., Koehler, A., Hellweg, S., 2009. Assessing the environmental impacts of freshwater consumption in LCA. Environ. Sci. Technol. 43, 4098e4104. P ﬁ ster, S., Vionnet, S., Lenova, T., Humbert, S., 2015. Ecoinvent 3: assessing water use in LCA and facilitating water footprinting. Int. J. Life Cycle Assess. 21, 1349e1360. Prasad, P., Pagan, B., 2006. Eco-ef ﬁ ciency and dairy processing. Aust. J. Dairy Technol. 61, 231. Pre, 2016. SimaPro Database Manual Methods Library. https://www.pre- sustainability.com/download/DatabaseManualMethods.pdf. Rad, S.J., Lewis, M.J., 2014. Water utilisation, energy utilisation and wastewater management in the dairy industry: a review. Int. J. Dairy Technol. 67 (1), 1e20. Ridoutt, B.G., Eady, S.J., Sellahewa, J., Simons, L., Bektash, R., 2009. Water foot- printing at the product brand level: case study and future challenges. J. Clean. Prod. 17, 1228e1235. R€ o€os, E., Sundberg, C., Tidåker, P., Strid, I., Hansson, P.-A., 2013. Can carbon footprint serve as an indicator of the environmental impact of meat production? Ecol. Indic. 24, 573e581. Singh, N.B., Singh, R., Imam, M.M., 2014. Wastewater Management in dairy in- dustry: pollution abatement and preventive attitudes. Int. J. Sci. Environ. Technol. 3, 672e683. Smith, A.L., Skerlos, S.J., Raskin, L., 2013. Psychrophilic anaerobic membrane bioreactor treatment of domestic wastewater. Water Res. 47, 1655e1665. Thoma, G., Popp, I., Nutter, D., Shonnard, D., Ulrich, R., Matlock, M., Soo-Kim, D., Neiderman, Z., Kemper, N., East, C., Adom, F., 2013. Greenhouse gas emissions from milk production and consumption in the United States: a cradle-to-grave life cycle assessment circa 2008. Int. Dairy J. 31, 3e14. Tillotson, M.R., Liu, J., Guan, D., Wu, P., Zhao, X., Zhang, G., P ﬁ ster, S., Pahlow, M., 2014. Water Footprint Symposium: where next for water footprint and water assessment methodology? Int. J. Life Cycle Assess. 19, 1561e1565. Tol, R.S., 2002. Estimates of the damage costs of climate change. Part 1: benchmark estimates. Environ. Resour. Econ. 21, 47e73. Van Zeist, W.J., Marinussen, M., Blonk, H., Broekema, R., Kool, A., Ponsioen, T.C., 2012. LCI Data for the Calculation Tool Feedprint for Greenhouse Gas Emissions of Feed Production and Utilization: Animal Products. Blonk Consultants and WUR Livestock Research, Gouda, the Netherlands. Verge, X.P.C., Maxime, D., Dyer, J.A., Desjardins, R.L., Arcand, Y., Vanderzaag, A., 2013. Carbon footprint of Canadian dairy products: calculations and issues. J. Dairy Sci. 96, 6091e6104. Weidema, B.P., Thrane, M., Christensen, P., Schmidt, J., Løkke, S., 2008. Carbon footprint. J. Ind. Ecol. 12, 3e6. West, P.C., Gerber, J.S., Engstrom, P.M., Mueller, N.D., Brauman, K.A., Carlson, K.M., Cassidy, E.S., Johnston, M., MacDonald, G.K., Ray, D.K., 2014. Leverage points for improving global food security and the environment. Science 345, 325e328. White, C.H., Kilara, A., Hui, Y.H., Chandan, R.C., 2008. Manufacturing Yogurt and Fermented Milks. John Wiley & Sons. Wijnands, J.H., van der Meulen, H.A.B., Poppe, K.J., 2007. In: Of ﬁ ce for Of ﬁ cial Publications of the European Communities (Ed.), Competitiveness of the Eu- ropean Food Industry: an Economic and Legal Assessment. E.U. Zonderland-Thomassen, M.A., Ledgard, S.F., 2012. Water footprintingeA compari- son of methods using New Zealand dairy farming as a case study. Agric. Syst. 110, 30e40. V. Vasilaki et al. / Journal of Cleaner Production 139 (2016) 504e516516 ronmental impacts of dairy processing and products: a review. J. Dairy Sci. 94, 4243e4254. Olmez, H., Kretzschmar, U., 2009. Potential alternative disinfection methods for organic fresh-cut industry for minimizing water consumption and environ- mental impact. Food Sci. Technol. 42, 686e693. Palhares, J.C.P., Pezzopane, J.R.M., 2015. Water footprint accounting and scarcity indicators of conventional and organic dairy production systems. J. Clean. Prod. 93, 299e307. Pandey, D., Agrawal, M., Pandey, J.S., 2011. Carbon footprint: current methods of estimation. Environ. Monit. Assess. 178, 135e160. P ﬁ ster, S., Koehler, A., Hellweg, S., 2009. Assessing the environmental impacts of freshwater consumption in LCA. Environ. Sci. Technol. 43, 4098e4104. P ﬁ ster, S., Vionnet, S., Lenova, T., Humbert, S., 2015. Ecoinvent 3: assessing water use in LCA and facilitating water footprinting. Int. J. Life Cycle Assess. 21, 1349e1360. Prasad, P., Pagan, B., 2006. Eco-ef ﬁ ciency and dairy processing. Aust. J. Dairy Technol. 61, 231. Pre, 2016. SimaPro Database Manual Methods Library. https://www.pre- sustainability.com/download/DatabaseManualMethods.pdf. Rad, S.J., Lewis, M.J., 2014. Water utilisation, energy utilisation and wastewater management in the dairy industry: a review. Int. J. Dairy Technol. 67 (1), 1e20. Ridoutt, B.G., Eady, S.J., Sellahewa, J., Simons, L., Bektash, R., 2009. Water foot- printing at the product brand level: case study and future challenges. J. Clean. Prod. 17, 1228e1235. R€ o€os, E., Sundberg, C., Tidåker, P., Strid, I., Hansson, P.-A., 2013. Can carbon footprint serve as an indicator of the environmental impact of meat production? Ecol. Indic. 24, 573e581. Singh, N.B., Singh, R., Imam, M.M., 2014. Wastewater Management in dairy in- dustry: pollution abatement and preventive attitudes. Int. J. Sci. Environ. Technol. 3, 672e683. Smith, A.L., Skerlos, S.J., Raskin, L., 2013. Psychrophilic anaerobic membrane bioreactor treatment of domestic wastewater. Water Res. 47, 1655e1665. Thoma, G., Popp, I., Nutter, D., Shonnard, D., Ulrich, R., Matlock, M., Soo-Kim, D., Neiderman, Z., Kemper, N., East, C., Adom, F., 2013. Greenhouse gas emissions from milk production and consumption in the United States: a cradle-to-grave life cycle assessment circa 2008. Int. Dairy J. 31, 3e14. Tillotson, M.R., Liu, J., Guan, D., Wu, P., Zhao, X., Zhang, G., P ﬁ ster, S., Pahlow, M., 2014. Water Footprint Symposium: where next for water footprint and water assessment methodology? Int. J. Life Cycle Assess. 19, 1561e1565. Tol, R.S., 2002. Estimates of the damage costs of climate change. Part 1: benchmark estimates. Environ. Resour. Econ. 21, 47e73. Van Zeist, W.J., Marinussen, M., Blonk, H., Broekema, R., Kool, A., Ponsioen, T.C., 2012. LCI Data for the Calculation Tool Feedprint for Greenhouse Gas Emissions of Feed Production and Utilization: Animal Products. Blonk Consultants and WUR Livestock Research, Gouda, the Netherlands. Verge, X.P.C., Maxime, D., Dyer, J.A., Desjardins, R.L., Arcand, Y., Vanderzaag, A., 20