From 2012 to 2014, IMOA participated in an industry-wide effort to review current LCA practice and experience within the metals and minerals industry, and to develop new guidance on how to adopt a more harmonized approach to LCI and life cycle impact assessment (LCIA) methodologies within the industry. The resulting guidance document (PE 2014a) identifies four topic areas as essential for alignment with respect to metal-related LCAs: (1) system boundaries, (2) coproduct modeling, (3) life cycle impact assessment (LCIA), and (4) metals recycling modeling. The following sections present IMOA’s perspective and highlight the issues most relevant to molybdenum.
The cradle-to-gate molybdenum LCIs are intended to be applied in a broader life cycle context. PE (2014a) asserts that products should be evaluated on a full life cycle basis, which includes use phase and metal recovery/recycling at end of life, and should be based on a well-defined functional unit that incorporates a product’s performance, service life, etc. This ensures a more complete picture of a product’s environmental impacts and, by appropriately defining the functional unit, the attributes of the product at use phase (i.e., the reason the product or material is used in the first place) can be justly accounted for.
By way of its formal data request process, IMOA has seen its LCI data applied to cradle-to-grave studies to support both new and conventional applications. For example, as part of an ongoing analysis funded by the US Department of Energy’s Bioenergy Technologies Office, Pacific NW National Lab (PNNL) quantified the relative life cycle impacts of catalyst use for a proposed biofuel-upgrading process, which utilizes nickel and molybdenum (NiMo) catalysts or cobalt and molybdenum (CoMo) catalysts. PNNL used molybdenum chemical LCI data to help assess the relative importance of catalyst management scenarios in the context of the overall biofuels upgrading process. The results, which include a comprehensive greenhouse gas (GHG) analysis of the systems, are not yet public.
Lavery et al. (2013) used IMOA’s LCI data for molybdenum as an activity enhancing dopant in a sponge nickel catalyst produced by gas atomization. The study evaluated the use of the gas atomization catalyst versus its conventional cast and crush production route, in the reaction of butyraldehyde to butanol. Results found that the benefits of lower energy and emissions during the gas atomization catalyst’s use phase outweighed the environmental impacts of the production, manufacturing, and recycling of the materials making up the catalyst. The results specifically identify molybdenum as a contributing factor to the large energy savings during use phase, despite a relatively higher contribution of GHGs and acidification at upstream production.
PE International (2014b) performed a cradle-to-grave LCA on the B-pillar of a passenger vehicle, comparing the previous press-hardened boron steel design to an advanced high strength steel (AHSS) containing molybdenum as an alloying element. The B-pillar is part of the vehicle’s structural body, and its main function is to protect occupants and help maintain the structural integrity of the vehicle during a side impact. The lighter, stronger AHSS required 4 kg less mass than its steel counterpart to perform the same function. Results of the analysis showed that when accounting for the full life of the vehicle, lightweighting via the lower mass of the AHSS material in the B-pillar was responsible for a higher fuel efficiency during use and resulted in net overall savings for global warming potential, acidification potential, eutrophication potential, photochemical smog formation, and primary energy demand.
While the molybdenum LCIs are generally not used purely as cradle-to-gate, there are few instances where downstream life cycle stages can be rationally omitted. For example, a global flooring company used molybdenum chemical LCI data to “build” the molybdenum-bearing flame retardants in their carpet LCAs. Since the flame retardants comprise less than 2 % of carpet mass and a hot spot analysis showed that they were not environmentally relevant inputs, the carpet LCA did not get into the fine detail of the flame retardants at use phase or their specific modeling at the carpet’s end-of-life.
Molybdenum may be mined as a single metal ore or with copper and potentially other metals in multi-metal ores. In this latter case, the copper and molybdenum coproducts need to be appropriately modeled.Footnote 1 The molybdenum LCIs follow the approach recommended in PE (2014a) Table 4, i.e., for base metals, where the preferred method is to use mass allocation for the coproducts, on the basis of the total metal output.Footnote 2 The choice of mass allocation is reasonable: “Mass is a consistent physical property of the metal and allows for a geographic and temporal consistency…” (PE 2014a, Table 4). Furthermore, the mass of outputs remain relatively constant over a number of years, while economic allocation (market value) could fluctuate considerably in a short period of time, leading to LCA results that may not always be representative of the system.
For the molybdenum LCIs, the allocation percentage used was based on the mass of metals in the concentrates.Footnote 3 The allocation percentage is carried upstream through to the mining process as shown in Fig. 3, to account for the actual metal recovered at the concentration process, not the potential metal found in the ore.
Life cycle impact assessment
The molybdenum LCIs do not include LCIA, as the intent is to enable LCA practitioners to apply impact categories that best meet the goal and scope of their studies. IMOA supports the use of the five LCIA categories recommended for use in LCAs involving metals: global warming potential, acidification potential, eutrophication potential, smog formation potential, and ozone depletion potential. These have a considerable level of scientific robustness and consensus in the LCA community. Two methodologies that could be used to calculate these and have substantial agreement among them include CML (Guinée 2002) for European-based studies and TRACI (Bare 2003) for North American studies.Footnote 4
While PE (2014a) describes other categories available to LCA practitioners, including resource depletion potential, abiotic depletion potential, land use, and toxicity potential, these are less scientifically robust than those supported for molybdenum and other metals studies. PE (2014a) Sec 5.2 documents why these categories are not currently recommended for use, especially for metals LCAs. The section on human and ecological toxicity categories deserves highlighting here, as toxicity categories are so often used in LCA due to their inclusion in many popular impact methodologies (see, e.g., PE 2014a, Sec. 126.96.36.199). Furthermore, toxicity-related LCA results are often a focal point in the results—and sometimes a criterion for decision-making—since, anecdotally speaking, the perception of “toxicity” in a studied system can be emotive when it comes to the health of humans or the ecosystem. Toxicity is indeed often a necessary aspect to include when evaluating risks in a study system, but LCA is not an appropriate tool to address this. Information on fate and effects of the chemicals released to the environment are needed to understand toxicity, and since LCA does not provide this information, its use for measuring toxicity impacts is limited to the identification of potential hot spots that would require further investigation with other methods or tools such as traditional risk assessment. When toxicity assessment using LCA based toxicity impact methodologies is performed, significant caution should be exercised; results may be misleading unless limitations of the impact method and/or underlying data are clearly set forth in the interpretation stage. One example of potentially misleading toxicity results for stainless steels was featured in an early version of a technical report that laid the groundwork for the revision of the EU Ecolabel criteria for bed mattresses (JRC 2013). An evaluation of different metal spring materials in the spring mattress presented alarmingly high spikes in both the freshwater and marine ecotoxicity categories for the stainless steel springs (JRC 2013, Figure 14). Due to stakeholder feedback and the lack of available high quality data on stainless steel at the time, the report ultimately dismissed the analysis as inconclusive. Still, this example demonstrates the significant potential for results to be negatively misconstrued—and the merit of a material to be wrongly tarnished as a consequence—when the impact category and/or background data are not properly vetted or insufficient explanation is provided.
PE (2014a) suggests that LCA practitioners use the USEtox model (Rosenbaum 2008) if they perform toxicity assessment as part of the LCA. But even though USEtox is considered by many LCA experts to be the most robust LCA toxicity model currently available, it has significant limitations (PE 2014a, pp. 46–47 and Rosenbaum 2008). A key drawback is that the current USEtox characterization factors are within a factor of 100–1000 for human health and 10–100 for freshwater ecotoxicity (Rosenbaum 2008). Even though USEtox is considered more robust than the other toxicity modeling methods, its low level of precision should be highlighted and taken into account during the interpretation phase of the LCA.
Should toxicity assessment be performed in an LCA study, IMOA advocates the approach given in PE (2014a) Sec. 188.8.131.52: “(a) look for existing risk assessments for the metals; (b) use the current LCA toxicity models with caution; (c) make sure the most recent data/models are being used; and, (d) consider toxicity separately from other environmental indicators.”
MOA supports taking the recycling approach described in PE (2014a) sec. 4 when the molybdenum in molybdenum-bearing products is recovered and recycled into other applications, i.e., as an alloy in a new steel product.