Natural resources are essential for our society, either for provision, supporting, regulating or cultural services [1
]. However, due to the world population growth, together with the increase in the consumption per capita and poor resource management, we are being led to a sustainability crisis. Natural resources may be classified in several ways, (a) renewable or non-renewable, (b) stocks, funds or flows, (c) biotic or abiotic, among other classifications [2
]. Regarding the latter, biotic resources are those that come from living organisms, while abiotic resources are the result of past biological processes (e.g., crude oil) or chemical processes (e.g., metal).
One of the tools that may assist in sustainable resource management, at the industrial scale, is a Life Cycle Assessment (LCA). Resources are seen in two ways in an LCA: (1) In one way, they are the inputs needed in industrial processes for the production of a product, and in this sense they are evaluated at the life cycle inventory (LCI) stage; and (2) in another way, they are evaluated as an area of protection (AoP), in life cycle impact assessments (LCIAs), i.e., natural resources are one type of environmental impact assessed, and there are different methods to evaluate these impacts.
According to traditional classifications [2
], these methods may be categorized into three groups: (1) Resource accounting methods (RAM), which make a more simplified impact analysis, focused mainly on grouping the resources into single score indicators, as energy or mass; (2) Midpoint resource depletion methods, that go beyond RAM, evaluating impacts related to resource depletion due to its use, as the use-to-availability ratio; and (3) Endpoint resource depletion methods, that go even beyond the previous group, taking into account the consequences of resource depletion, in many cases, through backup technology [4
], e.g., evaluating the extra effort (energy or cost) needed to extract less economically feasible resources.
Due to the lack of consensus in the LCA community regarding impacts on natural resources, there are other approaches of how the LCIA methods evaluate that AoP. Rorbech et al.
] classified LCIA methods into three groups: (1) Methods that account for the consumption of limited resources, which rather evaluate the resource competition and assume that they are exchangeable (as RAM); (2) Methods that evaluate the depletion of resources, which may be subdivided into midpoint and endpoint (and according to the authors would better represent the AoP Resources); and (3) Methods that evaluate the extra effort needed in the future due to actual resource extraction (e.g., Recipe Endpoint), which according to the authors do not represent a specific AoP as Resources, but are midpoint impacts that affect other AoPs (human health and natural environment). Dewulf et al.
] suggested new AoPs for LCA and Life Cycle Sustainability Assessment (LCSA), proposing five perspectives: (P1) the safeguard subject is the resource itself; (P2) the concern is the capacity of this resource to generate provisional services; (P3) the safeguard subject is the capacity of this resource to generate other ecosystem services; (P4) where consequential aspects are considered, e.g., socioeconomic mechanisms are taken into account; and (P5) the concern is human well-being, giving a rather holistic perception by grouping all previous perspectives. The authors also mention that P4 and P5 go beyond classical LCA, and would better fit in LCSA.
Even though there are different ways of grouping LCIA methods in LCA, and how they affect (different) AoPs, in this manuscript we used the rather traditional overview, proposed by ILCD (International Reference Life Cycle Data System) and Swart et al.
] (Figure 1
). In this sense, there are some studies that already critically evaluated different LCIA methods, for instance, Liao et al.
] evaluated thermodynamic-based RAM, and pointed out the Cumulative Exergy Extraction from the Natural Environment (CEENE) [9
] and the Solar Energy Demand (SED) [10
], as the recommended LCIA methods for that approach. In ILCD [3
], where different LCIAs were evaluated in order to make recommendations for the European context, the Abiotic Depletion Potential (ADP) [11
], adapted to the reserve base (Reserve base, according to the USGS (United States Geological Survey), accounts for all reserves that have the actual potential of extraction and may be economically viable in the future. The ADP method originally considered the ultimate reserve, which would be the amount of a certain resource in the Earth’s crust), was recommended for midpoint assessment, while there was no recommendation for endpoint LCIA methods. The Life Cycle Initiative, from UNEP-SETAC (United Nations Environment Programme and the Society for Environmental Toxicology and Chemistry), is an ongoing project to create a worldwide consensus on recommendations of LCIA methods (http://www.lifecycleinitiative.org/activities/phase-i/life-cycle-impact-assessment-programme/
] which may be considered as a step forward to what has been done by ILCD [3
Due to the variety of LCIA methods available in literature and the complexity in choosing one for an LCA study, there is a demand by LCA practitioners for support in decision making in private and public organizations. Therefore, the Brazilian Life Cycle Impact Assessment Network (Rede de Pesquisa em Avaliação do Ciclo de Vida, RAICV, 2014), (Regimento da Rede de Pesquisa em Avaliação de Impacto do Ciclo de Vida. São Bernardo do Campo, 11 November 2014, RAICV) evaluated different LCIA methods, for several impact categories, including Abiotic Resources. Nevertheless, due to certain characteristics from the Abiotic Resources category (e.g., a relative site-generic impact), the results for this category may be applied to other countries as well. As will be seen later, some RAM create characterization factors (CF) for both biotic and abiotic resources; thus, in some cases the recommendation went beyond abiotic resources (for RAM). Therefore, the objective of this manuscript was to evaluate different operational LCIA methods for (abiotic) resources available in literature in order to propose a recommendation .To facilitate the assessment, we applied some of these operational LCIA methods to a case study of ethylene production in Brazil through bio-based and fossil-based routes.
3. Experimental Section
In order to search for different operational LCIA methods available in literature, we used different keyword combinations (e.g., resources and LCA) on web tools, such as Web of Science. Articles published from the last 20 years, until December of 2014, were considered. After that, the articles that referred to the LCIA methods per se
were selected, i.e.
, we excluded case studies that used those LCIA methods. Finally, we evaluated them through two criteria: (1) Scope: in which the amount of elementary flows that could be accounted for was evaluated, i.e.
, the amount of CF available in the LCIA method. The availability of regionalized CF was also considered, but since spatial-differentiation in LCIA is not applicable to resource depletion assessment [36
], this was considered solely for the RAM (not for midpoint and endpoint LCIA methods); (2) Scientific robustness: in which the model behind the LCIA method was evaluated, how it was scientifically proposed (theory used and/or cause-and-effect relation), how clear was the documentation, and if the method was fully operational. Other criteria could be used in our evaluation (e.g., acceptance by LCA community), but we preferred to focus on rather technical criteria. We gave scores between one (the lowest) and five (the highest) for each of these criteria and later we calculated an arithmetic average in order to provide a final score for each of the LCIA methods evaluated.
After the theoretical assessment of the LCIA methods, some of them were applied in a case study of ethylene production. For that, two scenarios were considered:
A traditional FE, which was based on the data in the ecoinvent [55
] dataset named “ethylene, average (RER) production, Alloc Def” (There is no dataset in ecoinvent for Brazilian ethylene). The inputs and outputs in this dataset are arranged as aggregated LCI; thus, it is not possible to clearly identify the life cycle stage of each elementary flow;
A BE, from Brazil, where sugarcane is produced to generate ethanol, that is further dehydrated into ethylene. Therefore, Cavalett et al.
] was used for sugarcane and ethanol data and the Swedish Life Cycle Center (CPM) database [57
] for the ethanol-to-ethylene process unit. Sugarcane and ethanol production considered in reference [56
] is from advanced technologic cultivation and production systems, from the state of São Paulo (Brazil). The ethanol-to-ethylene process unit is based on pilot scale data.
After modeling the life cycle of these two scenarios of ethylene, we performed an LCIA through different resource-based LCIA methods. Then, we evaluated which scenario had the best/worst results and searched for the main hotspots that each of these LCIA methods identified.