Journal of Cleaner Production 268 (2020) 121793

Contents lists available at ScienceDirect

Journal of Cleaner Production

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

Implementing LCA early in food innovation processes: Study on

spirulina-based food products

Thomas a, Ce
Chloe
cile Gre

my-Gros a, b, Aure

lie Perrin a, Ronan Symoneaux a,
Isabelle Maître a, *
a
USC 1422 GRAPPE, Ecole Sup
erieure d’Agricultures (ESA), INRAE, SFR 4207 QUASAV, 55 rue Rabelais, 49007, Angers, France
b
LARIS, EA 7315, University of Angers, SFR MathSTIC, Polytech Angers, 62 Avenue Notre Dame du Lac, 49000, Angers, France

a r t i c l e i n f o a b s t r a c t

Article history: Food transformation companies could employ LCA upstream of product development to eco-design their
Received 15 January 2020 new products, in order to limit the impact of new food products on the environment. But the knowledge
Received in revised form required when using LCA on food products under development is imperfect. This study aims to overcome
14 April 2020
the difficulties of assessing the environmental impacts of imprecise food pre-concepts early on in
Accepted 17 April 2020
Available online 15 May 2020
innovation process, by developing a three-step modular LCA method: (1) creation of scenarios for each
pre-concept within four modules: “ingredients,” “packaging,” “transportation,” and “energy þ water”; (2)
Handling Editor: Yee Van Fan transversal analyses of the module scenarios and the modules themselves to estimate the pre-concepts’
environmental impacts, which were then conveyed by single scores; and (3) dissemination of the results
Keywords: to stakeholders. The method was applied to spirulina-based food pre-concepts. Results showed that the
Innovative product modular approach identified hotspots relevant to the company. The ingredients module contributed the
Ideation most to the impacts but was the most variable and uncertain module. The single scores facilitated
Sustainability decision-making for stakeholders unfamiliar with LCA. But some expertise was required to inform the
Food LCA
company on the origin of impacts and the causes of variations. The method served as an informational
Framework
tool that led the food company to question some of its prevailing assumptions. This study concludes that
Eco-design
it is possible to perform LCA on imprecise pre-concepts using a simplified modular framework. Research
must continue to include agri-food data in LCA databases and to simplify the internalization of LCA
expertise in agri-food businesses.
© 2020 Elsevier Ltd. All rights reserved.

1. Introduction International, 2019) as consumers consider such products as

more environmentally friendly (Azzurra and Paola, 2009). This
In 2015, the United Nations announced its sustainable devel- context enhances agri-food industries to consider their role in
opment goals (SDGs) with a view to increasing the commitment to protecting the environment, and most of them are striving to
sustainability within political and economic systems worldwide. achieve greater sustainability.
The protection of environmental ecosystems, resources and human The 21st century sees the great evolution of sustainability
health have become a global priority. In Europe, food production research (Medicine et al., 2019). Many methods and tools are
accounts for 20e30% of all anthropogenic impacts on the envi- currently available to assess the environmental impacts of objects,
ronment (Notarnicola et al., 2017). Consumers increasingly demand agricultural productions of food transformations. Some are quali-
for healthier and more sustainable products. For example, expec- tative (e.g., checklists and guidelines), others are quantitative and
tations for “clean labels” products grew, and the market share of monocriteria (e.g. Carbon Footprint) or multicriteria (e.g. Life Cyle
organic products has been growing (FiBL and IFOAM Organics Assessment) (Rousseaux et al., 2017). Even though monocriteria
assessment, like water use (Hashemi et al., 2019), water allocation
(Sedghamiz et al., 2018) in crops, or climate change (Pandey et al.,

rieure d’Agricultures (ESA), 55 rue
2011), are often used in first approach, quantitative methods such
* Corresponding author. GRAPPE, Ecole Supe
Rabelais, 49000 Angers, France. as the recognized Life-cycle assessment (LCA) standardized method
E-mail addresses: [email protected] (C. Thomas), r.symoneaux@ (ISO 14040) are more reliable tools to characterize and assess
groupe-esa.com (R. Symoneaux), [email protected] (I. Maître).

https://doi.org/10.1016/j.jclepro.2020.121793
0959-6526/© 2020 Elsevier Ltd. All rights reserved.
2 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793

multiple environmental impacts of products and services. development, validation, launch). During the ideation stage, com-
Past studies that performed LCAs for agri-food products have mercial opportunities are identified, and brainstorming generates
been quite diverse in nature (Perrin et al., 2014) and generally hard possible concepts that are referred to as pre-concepts. A pre-
to compare (Roy et al., 2009). The food transformation sectors like concept can simply be a type of a product that meets some con-
meat (e.g. chicken meat: Skunca et al., 2018; red meat: Colley et al., sumer expectations that are identified, e.g. “a drink that provides
2020) and dairy (Baldini et al., 2017) are more and more studied. health benefits”. At this early stage, also called “fuzzy front-end”,
The studies on the full value chain of food sectors have progressed. information flows are complex and diverse. They nonetheless
Some studies think beyond the factory gate and deal with the determine the direction of future product development (Kim and
environmental impact of the life cycle of products (e.g. vegetables: Wilemon, 2002), in particular when the product is discontinuous
Moreno et al., 2018), meals, restaurants (Caldero
n et al., 2018), (i.e. totally new) (Reid and Brentani, 2004). Pre-concepts are
meals (Stylianou et al., 2017) and even dietary changes (Hallstro €m therefore highly imprecise. Their characteristics (e.g., ingredients,
et al., 2015). The life cycle thinking is believed to be a key concept packaging) are deliberately poorly defined e fuzzy e to enable
for supporting sustainability transition (Sala et al., 2017). some flexibility in the choice of product characteristics upstream of
Some literature mention that LCA can play a role in decision- its development (Takey and Carvalho, 2016). Using an environ-
making contexts (Moro Piekarski et al., 2013), particularly as a mental assessment tool at this fuzzy early stage of the innovation
prospective tool for product and process design during develop- process was never done. The difficulty of carrying out LCA so early
ment (Tillman, 2000), such as in eco-design. LCA can help identify in the innovation process is precisely that LCA requires precise
key actions for reducing the product’s impact during the design of input data while the characteristics of the pre-concept are impre-
innovative products or redesign of existing products (i.e. eco- cise (specified in the Product Environmental Impact Guide:
desing and eco-innovation: Cluzel et al., 2014), in order to answer European Commission, 2012). This challenge adds to the challenges
sustainability challenges (Yannou-Le Bris and Ferrandi, 2016). Eco- regarding the use of LCA in the agri-food context. Furthermore, the
design, as defined by the International Organization for Standard- LCA assessments are built on the basis of existing data, which
ization standard number 14006:2011, is the “integration of envi- makes it difficult to use them for early design purposes (Skerlos,
ronmental aspects into product design and development, with the 2015). However, carrying LCA very early would help to anticipate
aim of reducing adverse environmental impacts throughout a the product attributes with regard to the expected environmental
product’s life cycle”. Eco-design approaches were applied in impacts, prior to the development stage. What the literature lacks,
manufacturing firms from the 1970s (Notarnicola et al., 2015). By and which would be useful to agri-food companies, is an improved
contrast, eco-design methods and tools in agri-food industry were framework for conducting LCAs during the early stages of an
used more recently because of the challenges to apply LCA in this innovation process. The present study intends to fill this gap.
context (Notarnicola et al., 2017). The research question that the present paper addresses is the
The literature presenting application of eco-design in agri-food following: How can LCA be used to assess the environmental im-
industry is scarce and heterogenous. Some studies focus in partic- pacts of imprecise food pre-concepts during the early stage of an
ular on the simplification of the assessment tools for the evaluation innovation process? The objective of this study was to develop a
of agricultural sectors (Renouf et al., 2018). Some studies show that methodological framework for integrating modular LCA into the
modularizing LCA is particularly efficient to reduce the complexity early ideation stage of an innovation process, so that food com-
of assessing food product systems (Brondi and Carpanzano, 2011), panies could consider environmental impacts of the imprecise pre-
product design alternatives (Gabrisch et al., 2019), and identify concepts before developing them. The present paper proposes a
hotspots in relation to consumers’ point of view (Jungbluth et al., framework composed of three steps that was applied to a case
2000). However, practitioners still face challenges in making LCA study in which innovative spirulina-based foods were being
suitable for assessing complex food systems. Three factors can in- developed by a French SME.
fluence the integration of LCA in design processes. First, the fre- This article is structured as follows: firstly, the methodological
quency of creation of new products in the food industry is very framework and the case study are detailed in the materials and
high. Second, the LCA method must be improved to better adapt to methods section. Then, the obtained results are presented accord-
the complexity of food systems (Sala et al., 2017). Data are ing to each of the three steps of the methodology: 1) generate
extremely variable (Roy et al., 2009), and there is a need to consider scenarios for imprecise pre-concepts, 2) analyze modules to iden-
the entire food chain (Sonesson et al., 2010). Third, as a conse- tify hotspots and compare pre-concepts, 3) communicate the re-
quence of the second factor, the incorporation of environmental sults to guide the company decision-making. Results are discussed
resource management into pre-existing organizational models is as they occur. Finally, the conclusion expresses the main result and
perceived as time and energy intensive on the part of business lessons learned from the case study, and draw the need for future
managers and collaborators (Petala et al., 2010). Fourth, multiple research.
functional units are possible (e.g. quantities are most used, other-
wise the nutrient units can also be used: McAuliffe et al., 2019) and 2. Materials and methods
every product, process (Roy et al., 2009) and packaging (Verghese
et al., 2012) has its own particularities, leading to extensive in- 2.1. Proposed methodological framework for assessing the
ventories. Most agri-food companies are therefore not equipped environmental impacts of pre-concepts
with sufficient expertise and knowledge to deal with the diversity
of eco-design tools available (Rousseaux et al., 2017), and guidelines 2.1.1. General approach
are lacking for eco-design practices using assessment tools Product creation begins with the ideation stage, in which the
(Hospido et al., 2010). goal is to generate as many innovative pre-concepts as possible.
While facing those challenges, it seems relevant that the Subsequently, the project team compares and ranks the pre-
research in LCA should consider the environment earlier in the concepts based on a list of criteria. During the subsequent stage-
innovation process of food industries, before the development dconceptualization stagedthe top pre-concepts are transformed
stage where eco-design is usually used, i.e. during what is called the into marketable concepts. The proposed method for assessing the
ideation stage (from the five-stages innovation process drawn upon environmental impacts of pre-concepts before the decision has
the model of Cooper et al. (2004): ideation, conceptualization, three basic steps (Fig. 1).
 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793 3

of each resource can differ (Fig. 2).

2.2. The case study

The method framework was applied to a new product devel-

opment (NPD) process being carried out by GlobeXplore (a French
agri-food SME) with the help of a food company (MidCap), a
technical institute, and a research team, who formed the project’s
steering committee. Communication took place directly with the
company’s director and its head of R&D, whose backgrounds were
exclusively in agri-food; they had no previous familiarity with
procedures for assessing environmental impacts. Results were al-
ways presented to the project’s steering committee. The NPD pro-
cess followed the five stages described by Cooper et al. (2004).
Spirulina is a microalga that has nutritional benefits (Habib
et al., 2008) because of its high protein, vitamin, and mineral
contents. The company wished to incorporate fresh (i.e., non-
dehydrated) spirulina into perishable products sold in the cold
Fig. 1. Three-step method for assessing the environmental impacts of agri-food pre-
aisle of supermarkets. The target market was consumers within
concepts during the ideation stage and disseminating the results to company France. The pre-concepts were products in which spirulina had
stakeholders. been added to a plant-based food (fruits, vegetables, legumes, or
oilseeds). Some pre-concepts were drinks, while others were
products with more texture. They differed in many of their traits
(e.g., appearance, texture, type of ingredients, and packaging). For
The first step is to establish modules focused on key sources of reasons of confidentiality, the authors cannot provide a more
environmental impacts (see section 2.1.2) and then generate detailed description of all the pre-concepts here. To illustrate the
different possible scenarios within each module for each pre- analyses, the authors focused on three of the pre-concepts: spir-
concept. The second step is to carry out LCAs to identify hotspots ulina-based milk, jam, and apple compote; each provided a daily
and inform decision-making. Using transversal analyses, the im- serving of 12 g of fresh spirulina and its associated nutritional
pacts of different scenarios and combinations of scenarios can be benefits.
compared. By analyzing the contributions of the different scenarios In this NPD project, the company needed to rapidly select the
to the environmental impacts of the pre-concepts, hotspots can be top two to three pre-concepts from among the eight pre-concepts
identified. The scenarios with the smallest environmental impacts that had emerged during the ideation stage. The relative perfor-
can then be chosen for each pre-concept, which can then be mance of the different pre-concepts had been assessed by the
compared to each other. The third step is to translate the results of consortium based on information provided by stakeholders during
these analyses into single scores for each pre-concept, which can meetings of the project’s steering committee. Performance esti-
then be communicated to the company. The goal is to help com- mates related to market potential and technical and economic
pany stakeholders (1) understand the environmental impacts of the feasibility were assessed and presented by the company. The
pre-concepts so that they can be prioritized accordingly during the research team presented the LCA method and the results of the
decision-making process and (2) implement appropriate ecodesign transversal/comparative analyses for the pre-concepts. An analysis
approaches during product development. of the exchanges among the consortium members during meetings
helped us assess the potential of the modular LCA proposed here.
2.1.2. Creation of modules and module scenarios for the different
pre-concepts 2.3. Application of life-cycle analysis to the case study
To function reliably, LCA requires detailed inventory data.
However, the characteristics of pre-concepts are, by nature, only 2.3.1. Objectives and functional units
vaguely defined. To deal with this situation, the proposed method In this study, the objective of the LCA was to assess the envi-
utilizes a modular approach. More specifically, agri-food pre-con- ronmental impacts of eight imprecise pre-concepts with a view to
cepts are broken down into different modular elements. informing company decision-making upstream of product devel-
Jungbluth et al. (2000) and Roy et al. (2009) identified the fac- opment. To help identify key differences in the module scenarios
tors that most greatly influenced the impacts of food products: (step 1 of the method), module-specific functional units were
agricultural production, the type of ingredients, food trans- chosen. Ingredients were compared based on mass. Packaging was
formation, packaging, the means of transportation, consumption, compared based on a fixed quantity of packaged product (e.g. 1 L of
and energy þ water usage. These factors thus appeared to be best spirulina-based milk, 200 g of jam). Transportation means and
suited to the modular LCA of agri-food pre-concepts. distances were compared based on payload-distance (tkm/kg of
Since Jungbluth et al. (2000) performed modular analyses that packaged product). Energy was compared in kWh, and water was
focused on each of the factors identified, the research team decided compared in m3. To assess the different pre-concepts (step 2 of the
to create modules according to the chosen factors: ingredients (that method), the functional unit was a serving containing 12 g of fresh
includes production), packaging, transportation, and energy/water. spirulina, which meets 100% of the daily requirement for vitamin
Within each module, there are various possible scenarios. For B12.
example, in the “ingredients” module, different ingredients can be
employed to create the food; in the “packaging” module, the food 2.3.2. Study system
can be packaged in different materials and/or sizes; in the “trans- The study system incorporated almost the entire life cycle of the
portation” module, means and distances of transportation can potential products resulting from the pre-concepts: from the pro-
differ; and in the “energy þ water” module, quantities and sources duction of the ingredients and the packaging materials to the
4 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793

Fig. 2. Defining the modules and module scenarios for the pre-concepts.

treatment of the waste produced. The Fig. 3 shows the study sys- Agribalyse® database (v. 1.3) and the EcoInvent3 database (v. 3.3).
tem, its boundaries, and the stages that fall outside the boundaries. The environmental impacts of the pre-concepts were expressed
This system included the agricultural production of the ingredients, using the midpoint results and the single scores. The 18 categories
the manufacturing of the packaging materials, the energy and of midpoint results obtained using the ReCiPe method are listed in
water used to create the finished products, the transportation of the Table 1.
products, and the treatment of the packaging waste. The different
scenarios for recycling packaging waste in France were defined 3. Results and discussion
based on information published in the annual report by Eco-
Emballages (data from 2016). 3.1. Scenarios for the pre-concepts
Certain stages were left out of the study system to simplify the
assessment of the pre-conceptsdthe elements that differed among To deal with the uncertainty associated with the imprecise pre-
potential products or between the potential products and a stan- concepts, the first methodological step was to generate different
dard of reference were solely included (Hospido et al., 2010). simple scenarios for the possible products. These scenarios were
created within each of the four modules “ingredients”, “packaging”,
 The production of spirulina was treated as a constant parameter. “transportation,” and “energy þ water,” which represented the
The pre-concepts contained the same daily serving of spirulina. main sources of environmental impacts over a food product’s life
Furthermore, the authors did not have access to any inventory cycle.
data for fresh spirulina production.
 The processes for transforming the potential products were 3.1.1. Ingredients module
treated as constant parameters. First, the LCA databases There were several different ingredient options for each pre-
currently contain little information on transformation pro- concept (Table 2). For example, the jam was composed of sugar
cesses, making it impossible to more explicitly incorporate them and a type of fruit, and the milk could use soy, oat, almond, or cow
without heavy data collection work. Second, the same produc- milk. An inventory of the ingredients that could be employed to
tion line (already in place within the factory) would be used make each pre-concept was thus created.
regardless of the product selected, so there would be no differ-
ences in the infrastructure associated with the different pre- 3.1.2. Packaging module
concepts. Third, the mean consumption of electricity and wa- There were several packaging scenarios for each pre-concept
ter by the factory to generate each kg of potential product is based on commercially available packaging. For example, for the
already taken into account within the analysis and allowed us to spirulina-based milk, the types of possible packaging were glass
estimate the resources used during product transformation. bottles, plastic bottles made of high-density polyethylene (HDPE),
 Product use by consumers, including purchasing, storage, or plastic bottles made of polyethylene terephthalate (PET). The
cooking, was unknown. The imprecision and diversity of the bottles could also come in different sizes, from 250 mL to 1 L. An
pre-concepts resulted in a high degree of uncertainty with example inventory of all the different packaging scenarios for the
regards to consumer behavior (e.g., shelf life). In addition, dur- three illustrative pre-concepts is presented in Table 3.
ing the end-of-life stage, only the packaging waste treatment
was accounted for. 3.1.3. Transportation module
 It was difficult to estimate the waste produced by product The transportation scenarios took into account the trans-
transformation (rejection), transport (breakage) or storage portation of the spirulina from its production site to the factory
(expired). where the products would be manufactured and the transportation
of the final products to four purchasing centers in different parts of
France. The loads in the trucks filled with the potential products
2.3.3. Inventory data and the characterization of environmental were calculated based on packaging type and truck carrying ca-
impacts pacity. The trucks were assumed to be empty upon the return trip.
To run the LCAs, SimaPro® software (v. 8.5) was used. The Eu- The company chose to use the same means of transportation for all
ropean ReCiPe methodology was employed with the Hierarchist the pre-concepts: a 28-tonne truck pulling a refrigerated semi-
model (v. 1.13) (Goedkoop et al., 2013) in tandem with the trailer (kept at 4  C). The mean estimated roundtrip payload-
 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793 5

Fig. 3. Stages and boundaries of the life-cycle analysis of the pre-concepts.

6 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793

Table 1 guidelines based on the literature would also benefit the

Categories of impacts associated with the midpoint results obtained with the companies.
ReCiPe method (Hierarchist model; v. 1.13).
This state of affairs underscores that LCA databases for food
Impact categories Units ingredients and transformation processes remain incomplete. Even
Climate change kg CO2-eq if they are expanding, the databases do not include data for all the
Ozone depletion kg CFC-11-eq ingredients of potential interest to the agri-food industry. For
Terrestrial acidification kg SO2-eq example, no information is available for additives such as xanthan
Freshwater eutrophication kg P-eq
gum, which is used to thicken the matrices of plant-based foods. As
Marine eutrophication kg N-eq
Human toxicity kg 1,4-DB-eq a result, analyses may be skewed and thus biased for or against a
Photochemical oxidant formation kg NMVOC given pre-concept when it is compared to others whose life-cycle
Particulate matter formation kg PM10-eq data are more complete and/or accurate.
Terrestrial ecotoxicity kg 1,4-DB-eq
Freshwater ecotoxicity kg 1,4-DB-eq
Marine ecotoxicity kg 1,4-DB-eq 3.2. Transversal analyses, module contribution, and pre-concept
Ionizing radiation kBq U235-eq comparisons
Agricultural land occupation m2a
Urban land occupation m2a
Natural land transformation m2
The transversal analyses (contribution of modalities, compari-
Water depletion m3 son of modalities, contributions of modules to the impacts of pre-
Metal depletion kg Fe-eq concepts) generated diverse knowledge about the environmental
Fossil fuel depletion kg oil-eq impacts of the pre-concepts. These results are presented below.

3.2.1. Module scenarios with the smallest impacts

Table 2
Different ingredient scenarios (in column) for the three example pre-concepts 3.2.1.1. Ingredients module. For the ingredients’ module, the envi-
(spirulina-based milk, jam, and apple compote). ronmental impacts of using different ingredients to make the pre-
concepts were compared and contrasted.
Pre-concepts Milk Jam Apple compote
For example, for the spirulina-based milk (Fig. 4), it was clear
Ingredients Soy milk Kiwi Additional sugar that using an ingredient of animal origin (cow milk) led to greater
Oat milk Pear No additional sugar
Almond milk e e
environmental impacts than using ingredients that were plant
Cow milk e e based, except in the case of almond milk, which had greater im-
Number of options 4 2 2 pacts in the categories water depletion and human toxicity
(because most of the almonds on the market come from California
and are irrigated). Previous research has shown that the con-
distance was 4.7 tkm/kg of packaged product. It was the same for all sumption of animal products has a greater environmental impact
the pre-concepts because they would all use the same distribution than the consumption of plant products, such that a vegetarian diet
chain. reduces carbon emissions by 20e39% (Hallstro € m et al., 2015;
Pernollet et al., 2017). The use of oat milk and soy milk led to the
smallest impacts, except in the case of marine eutrophication
3.1.4. Energy þ water module (because oat crops need to be fertilized).
Energy and water consumption were estimated by taking the There were large differences in the impacts associated with
factory’s annual energy and water consumption in 2016 and different produce types. For example, for the spirulina-based jam,
dividing it by the mean total mass of products manufactured over the use of kiwi resulted in smaller environmental impacts than the
the year. This calculation yielded a mean energy usage of 0.4 kWh of use of pear, except in the categories climate change and fossil fuel
electricity and 0.007 m3 of water for each kg of product depletion, for which the results were similar (Fig. A.1). It could be
manufactured. that pear trees are more vulnerable to pest and diseases and are
By using a modular LCA to assess pre-concepts, it is possible to thus treated more frequently with pesticides. In contrast, other
deal with a certain amount of uncertainty in the data, which could scenarios did not make much of a difference. For the spirulina-
otherwise lead to a faulty interpretation of the results and unreli- based apple compote, there was little difference in the environ-
able recommendations for the company. The ingredients found in mental impacts associated with the two sugar scenarios (no addi-
lesser quantities in food products can generally be ignored unless tional sugar versus additional sugar equivalent to 8% of cooked
they have previously been identified as hotspots for environmental mass) (Fig. A.2). Sugar addition thus appears to have no influence
impacts (e.g., animal proteins, non-seasonal produce). If LCA data- on the environmental performance of the product.
bases lack information on ingredients present in larger quantities, To dissect the origin of these impacts, it was necessary to assess
similar ingredients can be used as substitutes in the analyses. the contributions associated with the different ingredients.
Simple tools, like checklists about key trends in ingredients, could Frequently, companies lack the expertise to perform these types of
be developed for R&D teams. Developing decision-making analyses. Databases often only provide inventories and no

Table 3
Different packaging scenarios shown for three example pre-concepts (spirulina-based milk, jam, and apple compote); Alu. ¼ aluminum; PP ¼ polypropylene (plastic).

Type Jars Bottles Lids

Material Glass Plastic - PET Glass Plastic - HDPE Tinplate screw top Alu. seal Paper seal PP screw top

Size 400 g 90 g 20 g 125 g 500 g 1 kg 250 mL 500 mL 1L 250 mL 500 mL 1L

Milk x x x x x x x x
Jam x x
Apple compote x x x x x
 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793 7

Fig. 4. Comparison of the four ingredient scenarios for 1 kg of the spirulina-based milk pre-concept: (from left to right) cow milk, oat milk, almond milk, and soy milk (the midpoint
results for the impact categories).

information regarding the interpretation of impacts. As a conse- packaging size: one 1-L bottle had less of an environmental impact
quence, it is advisable to provide interpreted results associated than two 500-mL bottles or four 250-mL bottles (Fig. 5). Similar
with databases in supplementary report to make it easier for the results were obtained for the PET bottles (Fig. A.3) and the HDPE
user to understand the variation in the impacts of the product being bottles (Fig. A.4), both with polypropylene (PP) pop top lids. Second,
analyzed. bottles of the same size made with different materials (HDPE, PET,
and glass) were compared. The results showed that an HDPE bottle
had 0e15% of the environmental impact of a glass bottle (Fig. A.5)
3.2.1.2. Packaging module. For the packaging module, the envi- because of its lighter mass and less energy costly waste treatment.
ronmental impacts of using different packaging sizes and materials The PET bottle had an environmental impact in between those of
were compared and contrasted. First, glass bottles of different sizes the HDPE bottle and the glass bottle.
(1 L, 500 mL, and 250 mL) were compared, all with tinplate screw It is generally straightforward to assess packaging options using
top lids; the basis for comparison was the packaging needed to hold LCA in a way that is customized to company circumstances because
1 L of liquid. The results revealed that there was an effect of

Fig. 5. Comparison of the three sizes of glass bottles, 1 L, 500 mL, and 250 mL, based on the packaging needed to hold 1 L of liquid (Hierarchist model; midpoint results).
8 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793

LCA databases contain ample information about the direct impacts environmental impacts because agricultural commodities differ
of packaging (Molina-Besch et al., 2019), and literature exists on significantly in their impacts. Past research has indicated that
how to optimize the packaging design (Park et al., 2014). In contrast agricultural production plays a greater role than packaging and
to the guidelines used by (Verghese et al., 2012), the LCA of pack- transportation in defining the environmental impacts of food
aging allows different packaging scenarios to be compared and to products (Pernollet et al., 2017). The contribution of transportation
assess the environmental impacts associated with the quantity of is determined by product quantity and packaging mass.
packaged product. Discussions about packaging size raise questions Given that pre-concepts are generally imprecise, ingredient
about food waste. Here, the results revealed that larger packaging choice can have a large influence on and modify a potential prod-
had less of an environmental impact that did smaller packaging. uct’s impacts. The end results will be very sensitive to the choice of
However, this finding does not take into account indirect effects ingredients, which can have a negative effect when the company
involving consumer behavior. For example, packages that contain a makes decisions without finalizing the product recipe. However,
quantity of food that exceeds consumer needs may result in the removing the ingredients module because of its degree of uncer-
food product going bad and being thrown away, which increases tainty may result in the severe underestimation of the pre-con-
food waste (Molina-Besch et al., 2019). When designing food cepts’ environmental impacts. These results highlight that despite
products, it is important to keep in mind that reducing impacts by the great uncertainty it carries out, the ingredient module is of
increasing packaging size can increase impacts at the product’s end major importance for the environmental assessment.
of life.
As the market shifts towards new packaging solutions, there will 3.3. Communicating the results of the pre-concept comparisons
be contrasting consequences for LCAs. On the one hand, because
single-material packaging is coming back into favor and is more The pre-concepts were subsequently compared to each other
easily recycled, LCAs will be simplified. On the other hand, because using single scores (Fig. 7). This comparison showed that certain
of the arrival of new types of materials that are absent from the pre-concepts had greater environmental impacts than did others.
databases, LCAs will become more complex. Pre-concepts 4 and 7 had the greatest impacts and were both liquid
products. The transversal analyses showed that their impacts were
3.2.1.3. Transportation and energy þ water modules. The product due to their ingredients and transportation (their size and mass
transformation process was treated as a constant parameter. It was were greater). Pre-concepts 2, 5, and 6 had the smallest impacts. In
characterized using the estimated amount of energy and water particular, pre-concept 6 was composed of almost pure spirulina,
used to produce a kg of product, based on the factory’s figures for for which there was no inventory (i.e., spirulina production was
the previous year. While these mean values convey the likely outside the boundaries of the study system). Consequently, the
magnitude of energy and water consumption, they are not neces- environmental impacts for pre-concept 6 arose exclusively from
sarily reflective of the real usage associated with future products the packaging, transportation, and energy þ water modules. This
because those figures will depend on production volumes and result highlights the limit of the functional unit used which favor-
actual resource consumption by the equipment used. It is crucial to izes pre-concepts involving no or few additional ingredients. With
characterize the impacts of the food transformation process in another functional unit such as calories or added value, the result
greater detail, especially when the pre-concepts would lead to could have been very different.
products that would not necessarily use the same individual op- Pre-concepts are inherently uncertain and could give rise to a
erations. The company could install electricity and water meters on range of possible products. As a result, it is difficult to reliably es-
each piece of industrial equipment to measure the resource con- timate the environmental impacts of a given future product. Here, a
sumption associated with each individual operation. It would single representative scenario per pre-concept was chosen to
therefore be possible to characterize the transformation process for facilitate the comparisons. The goal was to help the company
each new food product and thus limit the factory’s environmental identify which pre-concepts could have the smallest environmental
impacts by reducing unnecessary resource use. In France, projects impacts upstream of product development. Then, as products are
like Agribalyse® v.3 is working to expand LCA databases by adding being developed, it is necessary to adapt inventories along with the
more information on transformation processes in various in- definition of product characteristics.
dustries, including the beef and dairy industries. The LCA results for this case study were shared with the com-
pany during a meeting of the project’s steering committee. It was
3.2.2. Module contributions to pre-concept environmental impacts expressed that the graph with the single scores for the eight pre-
The modules that contributed the most to the pre-concepts’ concepts was the best visual representation of the results and
environmental impacts by comparing the modules’ results for a would allow rapid decisions to be made about the pre-concepts to
given combination of scenarios were identified. More specifically, prioritize. The company’s director explicitly noted “This form of
to represent each module, the scenarios with the smallest envi- conveying the data is the easiest to understand.”, which is in line
ronmental impacts were chosen, as estimated in the transversal with the need for endpoint and scores mentioned by the literature
analyses. for an effective decision-making support (Ka €gi et al., 2016).
It was found that the ingredients module, followed by the The single scores also raised questions, and it was necessary to
transportation module, contributed the most to the environmental discuss the transversal (midpoint) results in detail to clarify the key
impacts of the eight pre-concepts in this case study. In the example sources of environmental impacts so that they could be addressed.
of the spirulina-based kiwi jam pre-concept (Fig. 6), the ingredients This process was more complicated than simply discussing the
module is the main source of environmental impacts (accounting single scores. Consequently, the results of the transversal analyses
for 15e95% of impacts depending on the category). The trans- d notably the contributions of the module scenarios to the envi-
portation module was second greatest source of environmental ronmental impacts of the pre-concepts d were presented during
impacts; it was followed by the packaging module (which exam- the same steering committee meeting. The company was surprised
ined primary, secondary, and tertiary packaging) and the by some of the findings because they seemed counterintuitive,
energy þ water module. The results varied depending on the pre- notably that impacts were greater for glass versus plastic pack-
concept. aging, and that locally sourced ingredient such as apple and cu-
The ingredients module contributed to major differences in cumber have more impacts than imported foods like banana. The
 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793 9

Fig. 6. Environmental impacts of a functional unit of the spirulina-based kiwi jam pre-concept: contributions made by the ingredients, packaging, transportation, and
energy þ water modules (ReCiPe method, Hierarchist model, midpoint results).

Fig. 7. Comparison of the environmental impacts of the NPD project’s eight pre-concepts using single scores (millipoints [mpt]) resulting from the categories human health,
ecosystems, and natural resources (Hierarchist model; endpoint results). Functional.

literature shows that transportation contribute significantly to the versus non-seasonal production) because the absence of this in-
carbon footprint of products (Striebig et al., 2019), but other as- formation can bias the results. Additional indicators are also lacking
sessments suggest that some local situations may be worse than to assess the overall impacts of products (e.g. soil carbon changes in
non-local because of different modes of transport, storage and milk production: Knudsen et al., 2019), biodiversity in particular,
consumption (Edwards-Jones, 2010). Consequently, the LCA results which remains complex to assess (Winter et al., 2017). For example,
caused the company to question some of its previously held as- although single-use glass containers have a greater environmental
sumptions. For example, a common assumption is that organic impact than single-use plastic containers, plastics release com-
products are better for the environment. However, past research pounds during use may pose health risks (Yang et al., 2019).
using LCA has shown that there are only small differences in the Finally, as illustrated in this case study, modular LCA can be used
environmental impacts of organically versus conventionally pro- as an information provider tool for agri-food industries. Even
duced foods: organic foods have lower impacts on a per area basis though results were found useful for the decision of the company to
but not on a per product unit basis (Tuomisto et al., 2012). The prioritize concepts for the following stages of innovation process
effects of consumption of chemical pesticides included into con- (conceptualization, development), the company found that the LCA
ventional foods are not included into LCA. method remains too complex and time-consuming to be used in
At present, the tools (i.e., LCA and the LCA databases) are not business-as-usual. The multiple steps, the need for precise in-
sufficiently developed to deal with eco-design challenges in the ventories, and the required expertise to analyze the results were
food sector. In particular, criteria that differentiate products need to the major issues mentioned by the stakeholders, in accordance to
be accounted for in inventories (for example, organic versus con- Sala et al. (2017).
ventional production, local versus non-local sourcing, or seasonal
10 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793

4. Perspectives uncertain. The comparison of pre-concepts with single scores was

relevant to the non-expert stakeholders in assisting their decisions.
There are several ways in which the modular LCA method pre- However, some expertise was required to inform them about the
sented here could be improved. First, one should include the use origin of impacts and their variations. Finally, the modular LCA
stage of the life cycle, since the use of the products by the con- contributed to change mindset by revealing to the company the
sumers at home may have a strong impact and alter the decision inaccuracy of their assumptions about environmental impacts.
made on the product characteristics (Notarnicola et al., 2017). Further research is needed to add open agri-food data into LCA
Second, further work should incorporate a sensitivity analysis for databases in order to facilitate rapid inventories for agri-food
the pre-concepts that expresses the results of pre-concept com- companies that wish to implement design for environment in
parison and the variability associated with the different possible their innovation process.
module scenarios. For example, Fran Cluzel et al. (2014) suggested
providing baseline scenarios, best-case scenarios, and worst-case Declaration of competing interest
scenarios. Third, the study system’s complexity, the degree of un-
certainty of pre-concepts, and the lack of data in databases does not The authors declare that they have no known competing
make it easy to generate precise scenarios for the ingredients’ financial interests or personal relationships that could have
module. appeared to influence the work reported in this paper.

CRediT authorship contribution statement

5. Conclusion

Chloe
Thomas: Conceptualization, Methodology, Investigation,

The early stage of the innovation process is critical to the

cile Gre
Writing - original draft. Ce
my-Gros: Methodology, Valida-
determination of the characteristics of a future food product, and
tion, Writing - review & editing. Aure
lie Perrin: Methodology,
the environmental criterion should be included to the company’s
Validation, Writing - review & editing. Ronan Symoneaux: Fund-
decisions. The main scientific contribution of this document is the
ing acquisition, Writing - review & editing. Isabelle Maître: Project
proposition of a modular LCA methodology to assess the environ-
administration, Writing - review & editing.
mental impacts of imprecise food pre-concepts in the early stage of
the innovation process. More precisely, the framework proposed in
this manuscript answers the difficulty of life cycle inventory and Acknowledgements
analysis with three steps: creation of modules to simplify the in-
ventories, transversal analyses (contributions and comparisons) to The present study was part of the “Spilife” project, funded by the
identify hotspots to be addressed through eco-design, and
gion Bretagne (France) and initiated by the company Globe-
Re
communication with single scores to the decision makers. Xplore. The present study was carried out with the facilities of the
In the studied case of spirulina-based food products, the four Senso’Veg platform.
modules created were ingredients, packaging, transportation, wa-
ter & energy consumption. The transversal analyses showed that Appendix
the ingredients module contributed the most to the environmental
impacts of the pre-concepts, and was the most variable and

Fig. A.1. Comparison of two ingredient scenarios for 1 kg of the spirulina-based jam pre-concept (based on the midpoint results for the impact categories): kiwi jam (in blue) and
pear jam (in green)
 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793 11

Fig. A.2. Comparison of two ingredient scenarios for 1 kg of the spirulina-based apple compote pre-concept (based on the midpoint results for the impact categories): apple
compote with additional sugar (8% of cooked mass; in blue) and apple compote without additional sugar

Fig. A.3. Comparison of the three sizes of plastic bottles made from PET: 1 L (in dark blue), 500 mL (in green), and 250 mL (in light blue) (Hierarchist model; midpoint results); the
basis for comparison was the packaging needed to hold 1 L of liquid
12 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793

Challenging World, Greece, p. 16.

Fig. A.4. Comparison of the three sizes of plastic bottles made from HDPE: 1 L (in dark blue), 500 mL (in green), and 250 mL (in light blue) (Hierarchist model; midpoint results); the
basis for comparison was the packaging needed to hold 1 L of liquid

Fig. A.5. Comparison of the three materials used to make 1-L bottles: plastic e HDPE (in dark blue), plastic e PET (in green), and glass (in light blue) (Hierarchist model; midpoint
results)

References Baldini, C., Gardoni, D., Guarino, M., 2017. A critical review of the recent evolution of
Life Cycle Assessment applied to milk production. J. Clean. Prod. 140, 421e435.
https://doi.org/10.1016/j.jclepro.2016.06.078. Towards eco-efficient agriculture
Azzurra, A., Paola, P., 2009. Consumers’ behaviours and attitudes toward healthy
and food systems: selected papers addressing the global challenges for food
food products: the case of Organic and Functional foods. In: Presented at the
systems, including those presented at the Conference “LCA for Feeding the
113th EAAE Seminar: A Resilient European Food Industry and Food Chain in a
 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793 13

planet and energy for life” (6-8 October 2015, Stresa & Milan Expo, Italy). International Publishing, Cham, pp. 255e293. https://doi.org/10.1007/978-3-
Brondi, C., Carpanzano, E., 2011. A modular framework for the LCA-based simulation 319-68177-1_12.
of production systems. CIRP J. Manuf. Sci. Technol. Prod. Network. Sustain. 4, Moro Piekarski, C., Mendes da Luz, L., Zocche, L., de Francisco, A.C., 2013. Life cycle
305e312. https://doi.org/10.1016/j.cirpj.2011.06.006. assessment as entrepreneurial tool for business management and green in-
Caldero
n, L.A., Herrero, M., Laca, A., Díaz, M., 2018. Environmental impact of a novations. J. Technol. Manag. Innovat. 8, 44e53. https://doi.org/10.4067/S0718-
traditional cooked dish at four different manufacturing scales: from ready meal 27242013000100005.
industry and catering company to traditional restaurant and homemade. Int. J. Notarnicola, B., Sala, S., Anton, A., McLaren, S.J., Saouter, E., Sonesson, U., 2017. The
Life Cycle Assess. 23, 811e823. https://doi.org/10.1007/s11367-017-1326-7. role of life cycle assessment in supporting sustainable agri-food systems: a
Cluzel, F., Vallet, F., Bertoluci, G., Leroy, Y., 2014. Eco-design vs eco-innovation: an review of the challenges. J. Clean. Prod. 140, 399e409. https://doi.org/10.1016/
industrial survey. In: International Design Conference. Presented at the Design j.jclepro.2016.06.071. Towards eco-efficient agriculture and food systems:
2014, Dubrovnik, Croatie. selected papers addressing the global challenges for food systems, including
Cluzel, François, Yannou, B., Millet, D., Leroy, Y., 2014. Exploitation scenarios in in- those presented at the Conference “LCA for Feeding the planet and energy for
dustrial system LCA. Int. J. Life Cycle Assess. 19, 231e245. https://doi.org/ life” (6-8 October 2015, Stresa & Milan Expo, Italy).
10.1007/s11367-013-0631-z. Notarnicola, B., Salomone, R., Petti, L., Renzulli, P.A., Roma, R., Cerutti, A.K., 2015. Life
Colley, T.A., Birkved, M., Olsen, S.I., Hauschild, M.Z., 2020. Using a gate-to-gate LCA Cycle Assessment in the Agri-Food Sector: Case Studies, Methodological Issues
to apply circular economy principles to a food processing SME. J. Clean. Prod. and Best Practices, 2015th ed. Springer.
251, 119566. https://doi.org/10.1016/j.jclepro.2019.119566. Pandey, D., Agrawal, M., Pandey, J.S., 2011. Carbon footprint: current methods of
Cooper, R.G., Edgett, S.J., Kleinschmidt, E.J., 2004. Benchmarking best NPD estimation. Environ. Monit. Assess. 178, 135e160. https://doi.org/10.1007/
practicesdI. Res. Technol. Manag. 47, 31e43. https://doi.org/10.1080/ s10661-010-1678-y.
08956308.2004.11671606. Park, S.-I., Lee, D.S., Han, J.H., 2014. Chapter 22 - eco-design for food packaging
Eco-Emballages, 2016. Rapport Annuel [WWW Document]. EcoEmballages. URL. innovations. In: Han, J.H. (Ed.), Innovations in Food Packaging, second ed. Food
http://www.ecoemballages.fr/eco-emballages/qui-sommes-nous/rapport- Science and Technology. Academic Press, San Diego, pp. 537e547. https://
annuel. accessed 10.9.19. doi.org/10.1016/B978-0-12-394601-0.00022-9.
Edwards-Jones, G., 2010. Does eating local food reduce the environmental impact of Pernollet, F., Coelho, C.R.V., van der Werf, H.M.G., 2017. Methods to simplify diet and
food production and enhance consumer health? Proc. Nutr. Soc. 69, 582e591. food life cycle inventories: accuracy versus data-collection resources. J. Clean.
https://doi.org/10.1017/S0029665110002004. Prod. 140, 410e420. https://doi.org/10.1016/j.jclepro.2016.06.111. Towards eco-
European Commission, 2012. Product Environmental Footprint (PEF) Guide: efficient agriculture and food systems: selected papers addressing the global
Consolidated Version (No. Ares(2012)873782). European Commission, Italy. challenges for food systems, including those presented at the Conference “LCA
FiBL, IFOAM Organics International, 2019. The World of Organic Agriculture. for Feeding the planet and energy for life” (6-8 October 2015, Stresa & Milan
Gabrisch, C., Cerdas, F., Herrmann, C., 2019. Product system modularization in LCA Expo, Italy).
towards a graph theory based optimization for product design alternatives. In: Perrin, A., Basset-Mens, C., Gabrielle, B., 2014. Life cycle assessment of vegetable
Schebek, L., Herrmann, C., Cerdas, F. (Eds.), Progress in Life Cycle Assessment, products: a review focusing on cropping systems diversity and the estimation
Sustainable Production, Life Cycle Engineering and Management. Springer In- of field emissions. Int. J. Life Cycle Assess. 19, 1247e1263. https://doi.org/
ternational Publishing, Cham, pp. 37e44. https://doi.org/10.1007/978-3-319- 10.1007/s11367-014-0724-3.
92237-9_5. Petala, E., Wever, R., Dutilh, C., Brezet, H., 2010. The role of new product develop-
Goedkoop, M., Heijungs, R., Huijbregts, M., De Schryver, A., Struijs, J., van Zelm, R., ment briefs in implementing sustainability: a case study. J. Eng. Technol. Manag.
2013. ReCiPe 2008: A Life Cycle Impact Assessment Method Which Comprises 27, 172e182. https://doi.org/10.1016/j.jengtecman.2010.06.004.
Harmonised Category Indicators at the Midpoint and the Endpoint Level Reid, S.E., Brentani, U.D., 2004. The fuzzy front end of new product development for
(version 1.08) No. Report 1: Characterisation, first ed. discontinuous innovations: a theoretical model. J. Prod. Innovat. Manag. 21,
Habib, M.A.B., Humtington, T.C., Hasan, M.R., 2008. Review on Culture, Production 170e184. https://doi.org/10.1111/j.0737-6782.2004.00068.x.
and Use of Spirulina as Food for Humans and Feeds for Domestic Animals and Renouf, M.A., Renaud-Gentie
, C., Perrin, A., van der Werf, H.M.G., Kanyarushoki, C.,
Fish. FAO Fisheries and Aquaculture Circular. Jourjon, F., 2018. Effectiveness criteria for customised agricultural life cycle
Hallstro€ m, E., Carlsson-Kanyama, A., Bo € rjesson, P., 2015. Environmental impact of assessment tools. J. Clean. Prod. 179, 246e254. https://doi.org/10.1016/
dietary change: a systematic review. J. Clean. Prod. 91, 1e11. https://doi.org/ j.jclepro.2017.12.170.
10.1016/j.jclepro.2014.12.008. Rousseaux, P., Gremy-Gros, C., Bonnin, M., Henriel-Ricordel, C., Bernard, P., Floury, L.,
Hashemi, M., Zadeh, H.M., Arasteh, P.D., Zarghami, M., 2019. Economic and envi- Staigre, G., Vincent, P., 2017. “Eco-tool-seeker”: a new and unique business
ronmental impacts of cropping pattern elements using systems dynamics. Civil guide for choosing ecodesign tools. J. Clean. Prod. 151, 546e577. https://doi.org/
Eng. J. 5 https://doi.org/10.28991/cej-2019-03091308, 1020-1032e1032. 10.1016/j.jclepro.2017.03.089.
Hospido, A., Davis, J., Berlin, J., Sonesson, U., 2010. A review of methodological issues Roy, P., Nei, D., Orikasa, T., Xu, Q., Okadome, H., Nakamura, N., Shiina, T., 2009.
affecting LCA of novel food products. Int. J. Life Cycle Assess. 15, 44e52. https:// A review of life cycle assessment (LCA) on some food products. J. Food Eng. 90,
doi.org/10.1007/s11367-009-0130-4. 1e10. https://doi.org/10.1016/j.jfoodeng.2008.06.016.
Jungbluth, N., Tietje, O., Scholz, R.W., 2000. Food purchases: impacts from the Sala, S., Anton, A., McLaren, S.J., Notarnicola, B., Saouter, E., Sonesson, U., 2017. In
consumers’ point of view investigated with a modular LCA. Int. J. LCA 5, 134. quest of reducing the environmental impacts of food production and con-
https://doi.org/10.1007/BF02978609. sumption. J. Clean. Prod. 140, 387e398. https://doi.org/10.1016/j.jcle-
Ka€gi, T., Dinkel, F., Frischknecht, R., Humbert, S., Lindberg, J., De Mester, S., pro.2016.09.054. Towards eco-efficient agriculture and food systems: selected
Ponsioen, T., Sala, S., Schenker, U.W., 2016. Session “midpoint, endpoint or papers addressing the global challenges for food systems, including those
single score for decision-making?”dSETAC europe 25th annual meeting, may presented at the Conference “LCA for Feeding the planet and energy for life” (6-
5th, 2015. Int. J. Life Cycle Assess. 21, 129e132. https://doi.org/10.1007/s11367- 8 October 2015, Stresa & Milan Expo, Italy).
015-0998-0. Sedghamiz, A., Heidarpour, M., Nikoo, M.R., Eslamian, S., 2018. A game theory
Kim, J., Wilemon, D., 2002. Focusing the fuzzy fronteend in new product devel- approach for conjunctive use optimization model based on virtual water
opment. R D Manag. 32, 269e279. https://doi.org/10.1111/1467-9310.00259. concept. Civil Eng. J. 4 https://doi.org/10.28991/cej-0309175, 1315-1325e1325.
Knudsen, M.T., Dorca-Preda, T., Djomo, S.N., Pen ~ a, N., Padel, S., Smith, L.G., Skerlos, S.J., 2015. Promoting effectiveness in sustainable design. Procedia CIRP
Zollitsch, W., Ho € rtenhuber, S., Hermansen, J.E., 2019. The importance of 22nd CIRP Conf. Life Cycle Eng. 29, 13e18. https://doi.org/10.1016/
including soil carbon changes, ecotoxicity and biodiversity impacts in envi- j.procir.2015.02.080.
ronmental life cycle assessments of organic and conventional milk in Western Skunca, D., Tomasevic, I., Nastasijevic, I., Tomovic, V., Djekic, I., 2018. Life cycle
Europe. J. Clean. Prod. 215, 433e443. https://doi.org/10.1016/ assessment of the chicken meat chain. J. Clean. Prod. 184, 440e450. https://
j.jclepro.2018.12.273. doi.org/10.1016/j.jclepro.2018.02.274.
McAuliffe, G.A., Takahashi, T., Lee, M.R.F., 2019. Applications of nutritional functional Sonesson, U., Berlin, J., Hospido, A., 2010. In: Sonesson, U., Berlin, J., Ziegler, F. (Eds.),
units in commodity-level life cycle assessment (LCA) of agri-food systems. Int. J. 9 - towards Sustainable Industrial Food Production Using Life Cycle Assessment
Life Cycle Assess. https://doi.org/10.1007/s11367-019-01679-7. Approaches. Environmental Assessment and Management in the Food Industry,
Medicine, N.A. of S., Engineering, and, Engineering, N.A. of, Sciences, D. on E. and P., Woodhead Publishing Series in Food Science, Technology and Nutrition.
Studies, D. on E. and L., Board, W.S. and T, Board, O.S., Programs, N.O. of, Sci- Woodhead Publishing, pp. 165e176. https://doi.org/10.1533/
ences, B. on L., Toxicology, B. on E.S. and, Resources, B. on E.S. and, Systems, B. 9780857090225.3.165.
on E. and E., Technology, B. on C.S. andClimate, B. on A.S. and, Resources, B. on A. Striebig, B., Smitts, E., Morton, S., 2019. Impact of transportation on carbon dioxide
and N., Century, C. on the G.C. and O. in E.E. for the T.-F, 2019. In: Environmental emissions from locally vs. Non-locally sourced food. Emerg. Sci. J. 3, 222e234.
Engineering for the 21st Century: Addressing Grand Challenges. National https://doi.org/10.28991/esj-2019-01184.
Academies Press. Stylianou, K.S., Nguyen, V.K., Fulgoni, V.L., Jolliet, O., 2017. Environmental impacts of
Molina-Besch, K., Wikstro €m, F., Williams, H., 2019. The environmental impact of mixed dishes: a case study on pizza. Faseb. J. 31 https://doi.org/10.1096/
packaging in food supply chainsddoes life cycle assessment of food provide the fasebj.31.1_supplement.lb386 lb386elb386.
full picture? Int. J. Life Cycle Assess. 24, 37e50. https://doi.org/10.1007/s11367- Takey, S.M., Carvalho, M.M., 2016. Fuzzy front end of systemic innovations: a con-
018-1500-6. ceptual framework based on a systematic literature review. Technol. Forecast.
Moreno, J., Pablos, C., Marug
an, J., 2018. Quantitative methods for life cycle Soc. Change 111, 97e109. https://doi.org/10.1016/j.techfore.2016.06.011.
assessment (LCA) applied to the vegetable industry. In: Pe
rez-Rodríguez, F., Tillman, A.-M., 2000. Significance of decision-making for LCA methodology. Envi-
Skandamis, P., Valdramidis, V. (Eds.), Quantitative Methods for Food Safety and ron. Impact Assess. Rev. 20, 113e123. https://doi.org/10.1016/S0195-9255(99)
Quality in the Vegetable Industry, Food Microbiology and Food Safety. Springer 00035-9.
14 C. Thomas et al. / Journal of Cleaner Production 268 (2020) 121793

Tuomisto, H.L., Hodge, I.D., Riordan, P., Macdonald, D.W., 2012. Does organic Winter, L., Lehmann, A., Finogenova, N., Finkbeiner, M., 2017. Including biodiversity
farming reduce environmental impacts? e a meta-analysis of European in life cycle assessment e state of the art, gaps and research needs. Environ.
research. J. Environ. Manag. 112, 309e320. https://doi.org/10.1016/ Impact Assess. Rev. 67, 88e100. https://doi.org/10.1016/j.eiar.2017.08.006.
j.jenvman.2012.08.018. Yang, J., Song, W., Wang, X., Li, Y., Sun, J., Gong, W., Sun, C., 2019. Migration of
Verghese, K., Lockrey, S., Clune, S., Sivaraman, D., 2012. In: Yam, K.L., Lee, D.S. (Eds.), phthalates from plastic packages to convenience foods and its cumulative
19 - Life Cycle Assessment (LCA) of Food and Beverage Packaging. Emerging health risk assessments. Food Addit. Contam. B 12, 151e158. https://doi.org/
Food Packaging Technologies, Woodhead Publishing Series in Food Science, 10.1080/19393210.2019.1574909.
Technology and Nutrition. Woodhead Publishing, pp. 380e408. https://doi.org/ Yannou-Le Bris, G., Ferrandi, J.-M., 2016. Eco-concevoir et e
co-innover : deux
10.1533/9780857095664.4.380. strate
gies essentielles pour re

pondre aux enjeux socie
taux a venir. IAA 19e21.