Why Value Engineering Keeps Squeezing Sustainability Out of Big Infrastructure Design

When engineers sit down to design a railway line or a nuclear facility, they often turn to a decades-old method called value engineering, a framework built on a deceptively simple idea: maximize the value of a project by relating its functions to the resources needed to deliver them. Born in 1947 as a way to boost product competitiveness by cutting resource use, the method has since spread across industries and evolved to include broader definitions of value, from customer satisfaction to stakeholder needs. Yet a new study published in Environmental and Sustainability Indicators suggests that this venerable tool has a structural blind spot. When researchers examined real projects in the French railway and nuclear sectors, they found that the criteria engineers use to judge value overwhelmingly favor technical and economic concerns, leaving social and ecological dimensions marginalized at precisely the moment when decisions lock in a project’s long-term footprint.

The research, conducted by Alexis Lalevée and Claudine Gillot in collaboration with an engineering company specializing in complex infrastructure, analyzed ten value engineering projects carried out between 2013 and 2020. These are the kinds of projects that shape the lives of multiple generations: transport and energy systems with time horizons approaching a century. The team collected end-of-study reports from each project and classified every value criterion into four dimensions: technical, economic, social and ecological. Two independent coders performed the classification, one a researcher working from the written reports and the other a value engineering expert who had firsthand knowledge of the projects and the intentions of their stakeholders. Disagreements between the two coders were resolved through discussion, and the expert’s industrial classification was ultimately retained to ensure alignment with actual practice.

The numbers are striking. Of the 88 value criteria identified across the ten projects, 45 percent were primarily technical or functional, and 32 percent were economic. In other words, 78 percent of the criteria that determined what counted as valuable in these multi-generational infrastructure projects served technical and economic ends. Ecological criteria accounted for just 7 percent of the total, and social criteria for 16 percent. The imbalance varied widely between projects: in one project, only a single criterion out of eleven escaped the techno-economic mold, while in another, only two out of fifteen criteria were techno-economic. Nuclear projects proved especially skewed, with technical criteria making up nearly 66 percent of their value criteria and ecological criteria a mere 2.63 percent.

To understand how this imbalance arises, the researchers traced the value engineering process as it unfolds in practice, using a well-documented illustration case about connecting two cities. The method begins with stakeholder identification, in which the funding authority assembles a working group of roughly fifteen parties, including citizens and companies. Individual interviews then characterize stakeholder needs, which are translated into value criteria, some easily quantified and others harder to express. Functional analysis follows, ranking the project’s functions by importance; in the illustration case, passenger transport accounted for 45 percent of the project’s importance and limiting journey times another 25 percent. Potential solutions, from railways and roads to canals and cycle routes, are then explored, filtered through five types of feasibility, and compared against the defined value criteria before a report goes to the decision-maker.

The problem, the study argues, lies in the method’s deep functional bias. Value engineering defines value as a relationship between functions and resources, and its functional analysis translates needs into functions described as transformations of energy within a system. Sustainability concerns get absorbed into this machinery only by being converted into functions themselves. There is no such thing as a biodiversity value criterion in the literature; instead, there is a function of protecting biodiversity. Environmental integration typically happens through Life Cycle Assessment, the industry’s dominant tool for measuring ecological impact, which shares the same functional logic. The researchers contend that this translation may oversimplify complex socio-ecological systems and that treating ecology as something made useful through functional units is itself a debatable assumption in an era of planetary resource depletion.

A second limitation is subjectivity. Value criteria are co-constructed in workshops where stakeholders with different expertise, hierarchical positions and interests negotiate what matters. The word safety means one thing in the railway sector, where it aligns with common-sense notions of security, and something quite different in the nuclear sector, where it refers to the integrity of goods and people to ensure continuity of operations, while security covers external risks like intrusion. Criteria such as image and awareness, modularity of the installation, or circular economy blur across social, economic and technical dimensions depending on who is defining them. Unlike formal multi-criteria decision-making methods such as the Analytic Hierarchy Process or MACBETH, which provide structured protocols for weighting criteria through pairwise comparisons, value engineering has no standardized method for resolving conflicts or ensuring consistency. Stakeholders with technical or economic expertise may simply prioritize the dimensions they know best, quietly marginalizing social and ecological concerns.

The researchers found a telling exception that points toward a remedy. In one project, stakeholders were given the opportunity to articulate their individual perspectives before a collective consensus was established, and that same project showed a notably higher integration of sustainability issues. This correlation, the authors suggest, hints that preserving the individual dimension within collective value construction could be key to broadening what counts as valuable. It also raises a deeper question about whether the very notion of a shared collective value erases individual concerns that are essential for capturing socio-ecological issues, a phenomenon documented in social psychology research on group processes.

The study also identifies a cognitive barrier. Social and ecological issues remain abstract concepts for most stakeholders, who lack the knowledge and skills to characterize and measure them within a framework that, by definition and by habit, carries technical and economic connotations. The authors propose that frameworks like the Doughnut model, which visualizes a safe and just operating space between social foundations and ecological ceilings, could help stakeholders build the understanding needed to define genuinely sustainable value criteria. Previous work by the lead author has explored adapting the Doughnut for engineering training, though the researchers caution that such theoretical proposals remain untested in real-world industrial applications.

The authors are candid about the limits of their own study. The analysis focused exclusively on early design phases, when solutions are still unclear, so the findings may not generalize to later project stages where sustainability criteria could evolve or gain significance. The research relied on a single value engineering expert for project selection and final classification validation, introducing potential selection and classification bias, although the elements in the project reports themselves suggested the techno-economic dominance observed. The underlying project data were confidential and had to be anonymized, with only generic criterion terms like modularity, safety and cost available for discussion. The authors recommend repeating the experiment with multiple experts from diverse backgrounds, including environmental and social scientists, and integrating structured multi-criteria methods to standardize criteria weighting.

The takeaway for the infrastructure sector is that value engineering, for all its adaptability and its genuine strength in bringing stakeholders together, cannot be relied upon to deliver sustainability on its own. The method’s functional bias, its unstructured subjectivity and the absence of any common reference framework for defining sustainable criteria combine to push socio-environmental concerns to the margins of early design decisions, the very decisions that shape a project’s impact for a century. The researchers argue that hybrid approaches, combining the collaborative power of value engineering workshops with the formal weighting techniques of multi-criteria decision-making, could rebalance the scales. Testing such approaches in industrial settings, they conclude, is the necessary next step if the tools that design our railways and power plants are to serve not just budgets and technical performance, but the societies and ecosystems that must live with the results.

Subject of Research: Limits of value engineering assessment criteria for integrating sustainability in early design phases of major infrastructure projects

Article Title: The limits of value engineering assessment criteria for integrating sustainability issues – Lessons from early design phases of major civil engineering infrastructure projects

Article References: Lalevée, A., & Gillot, C. (2026). The limits of value engineering assessment criteria for integrating sustainability issues – Lessons from early design phases of major civil engineering infrastructure projects. Environmental and Sustainability Indicators, 32, Article 101547. https://doi.org/10.1016/j.indic.2026.101547

Image Credits: AI Generated

DOI: 10.1016/j.indic.2026.101547

Keywords: value engineering, sustainability, infrastructure design, life cycle assessment, multi-criteria decision-making, stakeholders, railway, nuclear, functional analysis, value criteria, Doughnut model, early design phases