Main Challenges Identified in LCSA Studies So Far

The references in Table 3.1 were then analysed on the challenges faced. In Annex 1, these challenges are summarised for the different (groups of) references. Scanning through these references and generalising the challenges identified results in the following interpretation of the main challenges:

• The need for data and methods, particularly the lack of (proper and quantitative) SLCA indicators (Klöpffer 2008; Finkbeiner et al. 2010; Traverso et al. 2012a; Hu et al. 2013; Ostermeyer et al. 2013; Kucukvar and Tatari 2013; Vinyes et al. 2013; Zamagni et al. 2013).

• The need for practical (case study) examples (how to put LCSA in practice?) (Cinelli et al. 2013; Giesen et al.. 2013; Hu et al. 2013; Zamagni et al. 2013).

• How to communicate LCSA results (Finkbeiner et al. 2010; Traverso et al.

2012a, b; Bachmann 2013; Pesonen and Horn 2013)?

• The need for comprehensive methods dealing with all relevant uncertainties related to life cycle-based approaches (Zamagni 2012; Pesonen and Horn 2013; Kucukvar and Tatari 2013 and Kucukvar et al. 2014a, b).

• How to deal with technological, economic and political mechanisms at different levels of analysis (Cinelli et al. 2013; Sala et al. 2013a, b; Zamagni 2012; Zamagni et al. 2013)?

• The need for more dynamic models (Ostermeyer et al. 2013; Onat et al. 2014).

• How to deal with value choices and subjectivity in, particularly, the weighting step (Stamford and Azapagic 2012; Traverso et al. 2012b; Bachmann 2013; Manzardo et al. 2012; Sala et al. 2013a, b; Vinyes et al. 2013)?

• The need for further development of life cycle-based scenario evaluations (Zamagni 2012; Heijungs et al. 2014).

• How to deal with benefits (beneficial impacts), particularly in SLCA (Bachmann

2013)?

• How to avoid double counting (inconsistent application) between LCA, LCC and SLCA (Zamagni 2012; Bachmann 2013)?

• How to deal with different perspectives (producer, customer, societal) on costs in LCC (Finkbeiner et al. 2010)?

• How to (practically) relate (disciplinary) models to different types of life cycle sustainability questions (Guinée et al. 2011; Zamagni et al. 2013; Stefanova et al. 2014)?

From this list of challenges, those most frequently cited concern the 'broadening of impacts' dimension in general, the need for more practical examples of LCSA, efficient ways of communicating LCSA results and the need for more data and methods particularly for SLCA indicators and comprehensive uncertainty assessment. Note that with respect to SLCA, there are many more authors that identified these challenges (e.g. Jørgensen et al. 2008), but their references were excluded due to the limitations of our bibliometric analysis (see above).

The number of indicators that the various studies adopt for addressing the three pillars of sustainability in a life cycle perspective varies from a few (e.g. Moriizumi et al. limit their LCSA of two mangrove management systems in Thailand to just three indicators, one for each dimension of the 'triple bottom line') to several dozen indicators (e.g. Stamford and Azapagic adopted 43 indicators to address the same three pillars in their LCSA on electricity options for the UK). The challenges faced by studies adopting only a few indicators obviously include how to broaden the number of indicators. The challenges for studies adopting dozens of indicators include how to communicate their results to decision-makers and/or how to further weight (evaluate) and aggregate the indicator results, for example, applying (multicriteria) decision analysis.

The topic of 'deepening' is addressed less by the studies listed in Annex 1. Nevertheless, several references mention (e.g. Cinelli et al. 2013; Sala et al. 2013b; Zamagni 2012; Zamagni et al. 2013; Pesonen and Horn 2013; Kucukvar and Tatari 2013; Kucukvar et al. 2014a, b) and some even address (Hertwich et al. 2014) typical 'deepening' topics such as the need for comprehensive uncertainty assessment and methods for dealing with rebound effects. But, again, these references were excluded from Table 3.1 and Annex 1 due to the limitations of our bibliometric analysis (see above). However, we feel that deepening discussions are very important as part of maturing LCSA approaches. We illustrate this by the example of modelling rebound effects in a life cycle perspective, which has been addressed by several authors (Hertwich 2005; Hofstetter et al. 2006; Thiesen et al. 2008; Girod et al. 2011; Druckman et al. 2011; Font Vivanco and Voet 2014).

Hertwich (2005) defines the rebound effect as 'a behavioural or other systemic response to a measure taken to reduce environmental impacts that offsets the effect of the measure. As a result of this secondary effect, the environmental benefits of eco-efficiency measures are lower than anticipated (rebound) or even negative (backfire)'. For example, the positive effect of more efficient cars has largely been offset by an overall shift to larger and heavier cars (see Chap. 18). Similarly, the introduction of high-efficient light bulbs has been combined with an expansion of the number of light points. Recently, Font Vivanco and Voet (2014) performed a review describing the state of the art in incorporating the rebound effect into LCAbased studies and analysed their main strengths and weaknesses. Their literature review identified a total of 42 relevant scientific documents, from which 17 provided quantitative estimates of the rebound effect using LCA-based approaches. It appeared that 'the inclusion of the rebound effect into LCA-based studies is still one of the most relevant unresolved issues in the field; […] only few studies provide quantitative estimates (mostly for carbon dioxide and global warming […])'. Font Vivanco and Voet concluded that 'while a number of LCA-based studies have considered such effects […], no generally applicable guidelines have been developed so far; […] consequently, a panoply of non-consensual definitions and analytical approaches have arisen within the LCA community, and rebound effects have been both unevenly and inconsistently incorporated into LCA-based studies'. The results of this reviews show that, while incorporating the rebound effect into LCA studies is recognised as a very important topic and has received some attention, there is still no generally applicable and/or comprehensive method for dealing with rebound in a life cycle perspective. A similar conclusion is valid with respect to the challenge of incorporating comprehensive though practical uncertainty assessments into LCA (see, e.g. Gregory et al. 2013; Harst and Potting, 2013; Henriksson et al. 2014, 2015; Mendoza et al. 2014).

Finally, the bibliographic analysis showed that there are an increasing number of LCSA studies (e.g. Giesen et al. 2013; Hu et al. 2013; Manzardo et al. 2012; Stefanova et al. 2014; Heijungs et al. 2014) dealing with scenarios.^[1] The studies explore possible configurations of emerging new technologies, product systems or consumption baskets, comparing their potential impacts to alternative technologies, product systems and consumption baskets. Such studies are very relevant, particularly if performed ex ante or parallel to the technology development trajectory, as in that way LCSA is able to advise the technology developer whether developments are on the 'right' track while identifying hot spots for improvement. Considering this increase and the relevance of scenarios for evaluating possible more sustainable futures, we might even consider changing the meaning of the abbreviation LCSA from life cycle sustainability assessment to life cycle-based scenario assessment.

The results from our brief questionnaire among the ISIE-LCSA membership largely support the challenges discussed above while particularly adding challenges as communication with and involvement of stakeholders in the LCSA process, education and standardisation of LCSA methods.

• [1] See also Spielmann et al. (2005), Hertwich et al. (2014) and Koning et al. (2015) that report on scenario-based life cycle modelling but not as part of an LCSA framework. See also Chap. 2 of this book.