The team’s method was as unusual as its subject. Borrowing the Faraday Discussions format from physical chemistry, where opposing viewpoints are debated openly before an audience, and inspired by the practice of adversarial collaboration, the authors deliberately assembled researchers who hold conflicting positions. For each topic they drafted a deliberately provocative resolution, presented arguments for and against it, and then identified common ground. The stakes are high: conference polling at a 2024 plenary session of the International Symposium on Sustainable Systems and Technology revealed nearly even splits on all six debate resolutions among more than 100 participants, and discussions at recent workshops occasionally grew heated, particularly over uncertainty techniques. The authors stress that controversy here signals an active, evolving field rather than an immature one.

The first controversy concerns a deceptively simple question: should LCA be applied to emerging technologies at all? The case for early assessment is compelling. Design flexibility is greatest before commercialization, so environmental considerations can be built in before impacts become locked in. Screening tools such as stoplight diagrams have already flagged hazardous reagents and critical precursors in early-stage battery materials before laboratory work began. But the opposition warns that LCA demands a definable system and adequate process data. When feedstocks, process yields, or deployment scenarios are highly uncertain, results may mislead decision-makers, overstate benefits, entrench unsustainable pathways, or cause promising options to be abandoned prematurely. The network’s common ground is pragmatic: life cycle thinking is valuable at every stage, but it is not equivalent to a comprehensive LCA, and studies should be classified into tiers of rigor, much as techno-economic analyses are scaled by technology maturity.

Uncertainty emerged as perhaps the sharpest fault line. Emerging technologies face scenario uncertainty about future conditions, parameter uncertainty about technological options, and model uncertainty from subjective modeling choices, all layered on top of the variability that plagues even mature assessments. One camp argues that formal probabilistic methods such as Monte Carlo simulation should be applied even at low technology readiness levels, because presenting a single point estimate falsely suggests certainty that no honest analyst can justify. Confidence intervals, they contend, help decision-makers judge the likelihood of meeting performance thresholds. The other side counters that the same data gaps driving the uncertainty also undermine any attempt to characterize it, so a detailed probability distribution can paradoxically convey overconfidence. Probabilistic models capture known unknowns, they argue, while emerging technologies are dominated by unknown unknowns such as unproven scale-up paths and shifting market functions. Simple bounding exercises and scenario analyses may serve better.

On scale-up, the question is whether and how to project how a laboratory-bench process will look at industrial scale. The literature offers a growing toolkit: process simulations, learning curves, molecular structure models, proxy data from analogous technologies, and structured frameworks such as UpFunMatLCA for projecting lab-to-fab development of functional materials. A recent systematic review of 78 studies identified 14 distinct scaling methods, and data-driven advances are emerging, including experimentally derived learning rates used to scale life cycle inventory data for electrochemical ammonia production. Supporters argue that without scaling, comparisons with mature commercial systems are meaningless, since lab setups differ fundamentally from industrial plants in energy efficiency, material purity, and resource utilization. Opponents point to cautionary tales: perovskite solar cell assessments have often scaled up lab-favored materials that are unlikely commercial candidates, and early assessments of quantum computers underestimated impacts by ignoring error correction overhead and cryogenic cooling. Common correction factors, such as assuming industrial energy use is a fraction of lab demand, risk oversimplifying nonlinear, unpredictable transitions.

Comparison with incumbent technologies divides the field almost perfectly. Advocates note that any technology entering an existing market must outperform the incumbent to deliver environmental benefits, so relative comparisons are indispensable for benchmarking progress and guiding research and development priorities. Even with high uncertainty, relative differences between alternatives can be informative. Critics respond that comparator selection is fraught: whether a marginal or average system is chosen, and how optimistically future performance is assumed, can flip an emerging technology from beneficial to detrimental. Outdated background data compound the problem. Many photovoltaic assessments still benchmark against multi-crystalline silicon using inventory data stretching back to 2005, even though the technologies those guidelines describe are expected to account for less than 20 percent of the market by 2025. The consensus position demands careful functional equivalence, updated comparator data, and explicit communication that final go or no-go decisions rest with decision-makers, not analysts.

Standardization sparked equally pointed debate. The ISO 14040 and 14044 standards grant broad latitude, explicitly stating that there is no single method for conducting LCA and that data may be a mixture of measured, calculated, or estimated values. That flexibility, critics of the status quo argue, lets an ISO-compliant label create a false veneer of comparability, particularly at low technology readiness levels where modeling assumptions dominate results. Sector-specific guidance exists for carbon capture, biomass carbon removal, nanomaterials, and photovoltaics, but coverage is inconsistent and fragmented. Supporters of new standards point to practices such as borrowing lithium-ion battery lifetimes as proxies for novel chemistries, which likely biases results. Opponents warn that overly prescriptive rules could entrench context-dependent value choices, restrict the expert judgment that data-poor assessments require, and create false confidence in compliance labels. Both sides agree, however, that standardized reporting of technology readiness level, data provenance, scenario assumptions, and uncertainty treatment would substantially improve transparency.

The sixth controversy, stakeholder engagement, extends the debate beyond methodology into ethics. The authors argue that high-stakes assessments of emerging technologies ideally involve an extended peer community, ranging from subject matter experts who can supply data absent from conventional inventories to community members whose priorities may differ sharply from those of analysts. In one cited example, community members prioritized air quality over climate impacts when evaluating a fuel switch from natural gas to biomass, an insight that reshaped the study’s relevance. Engagement can also expose the need for geographically localized emission factors rather than regional averages. Yet the opposition is sober: practitioners often lack training and resources for meaningful participation, identifying appropriate stakeholders for technologies with undefined use cases invites selection bias, and poorly designed engagement risks becoming tokenistic or exploitative, imposing burdens on marginalized communities. Funders may also resist delays that push results beyond the decision-making window.

What emerges from the exercise is less a rulebook than a map of the field’s fault lines, together with a set of shared priorities. Across all six topics the network converged on framing studies to the decision context, setting minimum reporting expectations for data and study quality, and stating explicitly the limits of transferability for scenario-based projections. Disagreements persist on when to formalize standards and how extensively uncertainty can be treated for low-maturity technologies, and the authors identify recurring trade-offs between flexibility and comparability, early engagement and stakeholder burden, and sophistication and overconfidence. Their central message to practitioners, funders, and policymakers is that early-stage LCA results should be treated as provisional insights that guide research and development rather than definitive predictions, and that clear communication of assumptions, limitations, and uncertainty is not an optional courtesy but the foundation upon which credible, sustainability-informed technology decisions must rest.

Subject of Research: Best practices and methodological controversies in life cycle assessment of emerging technologies

Article Title: Controversy and consensus: common ground and best practices for life cycle assessment of emerging technologies

Article References: Woods-Robinson, R., Abeynayaka, A., Carbajales-Dale, M., Chen, H., Cheng, A., Cooney, G., Kirchofer, A., Kumar, M., Liddell, H. P. H., Peterson, L., Posen, I. D., Moni, S., Sleep, S., Wachs, L., Zargar, S., & Bergerson, J. (2026). Controversy and consensus: common ground and best practices for life cycle assessment of emerging technologies. Journal of Industrial Ecology, 30(4), 1495-1519. https://doi.org/10.1007/s44498-026-00100-5

Image Credits: AI Generated

DOI: 10.1007/s44498-026-00100-5

Keywords: life cycle assessment, emerging technologies, uncertainty, scale-up, standardization, stakeholder engagement, sustainability, technology readiness level, comparative LCA, ex-ante assessment, industrial ecology, environmental decision-making