New Decision Model Ranks Recycled Materials as Top Fix for Construction’s Carbon Problem

The buildings we live and work in are quietly responsible for an enormous share of the climate problem. Manufacturing building materials and constructing the structures themselves account for roughly ten percent of the world’s energy-related carbon dioxide emissions, while the built environment as a whole is linked to about thirty-seven percent. Life cycle assessments add a crucial detail: upstream cradle-to-gate activities — the extraction and processing of raw materials before anything reaches a construction site — generate sixty to seventy percent of a building’s embodied carbon. Yet when firms try to act on these numbers, they face a frustrating paradox. The interventions with the greatest ecological potential, such as carbon capture, hydrogen-based production and heavy digital integration, demand capital expenditures that only the largest companies can absorb, leaving small and medium enterprises locked out of meaningful decarbonisation.

A new open-access study in Cleaner Engineering and Technology by Himasai Kiran Reddy Durgam and Prasanna Venkatesan Ramani tackles this dilemma with an unusually rigorous computational framework. The researchers describe a hybrid Z-Number Reliability-Weighted Analytic Hierarchy Process combined with a true Euclidean TOPSIS engine, applied to the upstream construction supply chain. Their goal was not simply to rank green strategies but to resolve what they call the paradoxical tension between ecological yield and economic impedance — the persistent conflict between what is best for the planet and what a firm can actually afford to implement. The result is a data-driven action priority matrix that sorts twenty-four supply chain interventions into strategic quadrants without any subjective scenario weighting.

The methodological foundation rests on four theoretical lenses. The natural resource-based view holds that embedding environmental imperatives into planning can create competitive advantage, and it governs the framework’s ‘Quick Wins’ quadrant, where relational procurement strategies achieve maximum ecological return at minimum capital cost. Resource dependence theory explains the ‘Major Projects’ quadrant, where organisations must pursue high-impact strategies such as renewables and on-site recycling despite severe infrastructure and capital dependencies. Institutional theory contextualises the ‘Thankless Tasks’ quadrant, where expensive Industry 4.0 applications such as blockchain operate as compliance traps — adopted for external legitimacy rather than genuine decarbonisation. Paradox theory then ties these competing logics together mathematically, using Euclidean spatial geometry to harmonise sustainability benefits against implementation barriers.

The reliability weighting is where the framework departs most sharply from conventional practice. Traditional multi-criteria decision-making methods, including fuzzy extensions of the Analytic Hierarchy Process, implicitly assume that every expert is equally logically consistent — a flaw the authors call fatal in complex supply chain modelling. Drawing on Zadeh’s Z-numbers, which pair a judgement with a measure of its reliability, the team instead extracted each respondent’s individual Consistency Ratio and converted it into a deterministic reliability scalar. Each expert’s crisp eigenvector was then penalised by the square root of this scalar before aggregation, so that highly consistent respondents exerted mathematically superior influence over borderline-inconsistent ones. Of 207 civil engineering professionals surveyed, 202 matrices passed the strict inclusion threshold of a Consistency Ratio below 0.10, with a mean of 0.0551.

Structural stability was validated through a ten-thousand-iteration non-parametric bootstrap resampling algorithm, which the authors deliberately chose over classical static-perturbation Monte Carlo simulations. Rather than injecting artificial uniform noise, the bootstrap resampled the pool of 202 expert matrices with replacement, capturing true inter-rater heterogeneity. The results were striking. Under standard unweighted aggregation, unreliable respondents artificially deflated the priority of raw material interventions to a mean of 0.3756; the Z-number penalty shifted the purified consensus weight to 0.3852 with a tightly constrained standard deviation of plus or minus 0.0050. The gap between the top-ranked criterion, raw material extraction and processing, and the second-ranked criterion, production, exceeded twenty standard deviations — a separation implying a near-zero probability of rank inversion within the observed sample.

The downstream TOPSIS stage introduced what the authors term techno-economic triangulation. Instead of asking managers to estimate capital expenditure in subjective surveys — a task they are rarely equipped to perform — the framework split the evaluation into two axes. The sustainability benefit axis was populated exclusively by the empirical expert data, while the implementation impedance axis was calibrated deterministically from recognised techno-economic and life cycle databases, including Ecoinvent version 3.9, the European Reference Life Cycle Database, the International Energy Agency’s Iron and Steel Technology Roadmap and the Annual Technology Baseline. Cost and complexity scores were mapped onto a discrete one-to-nine scale and averaged into a unified impedance score, deliberately assuming symmetrical elasticity between financial cost and technical difficulty to avoid injecting prior bias.

With both axes normalised into a synchronised zero-to-one Cartesian space, the true Euclidean TOPSIS engine computed exact geometric distances to the positive ideal solution of maximum benefit at zero impedance and to the negative ideal solution, yielding an absolute closeness coefficient for each of the twenty-four alternatives. The verdict was unambiguous. Choosing recyclable materials topped the ranking with a closeness coefficient of 0.8508, leveraging high sustainability yields against a minimal impedance score of just 2.0. Sourcing sustainably followed at 0.7583, with low-emission extraction suppliers third at 0.6902. At the other extreme, blockchain traceability collapsed to 0.2240 and retrofit machinery to 0.2213, their extreme capital and complexity penalties — impedance scores of seven or higher — mathematically eclipsing their localised environmental benefits.

To translate these vectors into policy, the algorithm partitioned the interventions using statistical medians rather than manually assigned boundaries, locking the impact threshold at a closeness coefficient of 0.4362 and the difficulty threshold at an impedance score of 4.50. Ten alternatives, dominated by procurement and relational strategies, landed in the Quick Wins quadrant of high impact and below-median difficulty. Two interventions — expanding renewables and on-site recycling — qualified as Major Projects, offering above-median impact but demanding substantial infrastructure scaling and phased, multi-stakeholder financing. Three fell into Fill-Ins, while nine alternatives, predominantly over-engineered technological deployments, were quarantined as Thankless Tasks. A follow-up bootstrap of the TOPSIS matrix confirmed that the top ranks, including the leading procurement strategies, carried a near-zero probability of rank inversion under bootstrapped expert variance.

The practical implications are blunt. The authors argue that enterprise capital should be aggressively diverted away from speculative digital technology and channelled into physical raw material procurement, where recyclable inputs and sustainable sourcing deliver immediate, localised decarbonisation without cannibalising core capital budgets. Institutional mandates to deploy blockchain or other Industry 4.0 tracking architectures before physical supply chains are optimised, they contend, should be suspended, because such tools function primarily as compliance mechanisms that drain capital from critical physical decarbonisation. The framework does carry acknowledged limits: the respondents were drawn predominantly from the civil engineering domain, the impedance axis is insensitive to live financial volatility, and the model performs static rather than dynamic optimisation. Future work, the authors suggest, should test priority invariance across sub-sectors and integrate adaptive stochastic optimisation with geospatial routing. For now, the message to construction firms is clear — the cheapest, most robust carbon cuts may lie not in the newest technology, but in simply choosing better materials.

Subject of Research: A reliability-weighted multi-criteria decision-making model for prioritising decarbonisation actions in upstream construction supply chains

Article Title: Z-Number reliability-weighted AHP–true Euclidean TOPSIS model for green upstream construction supply chains: A cradle-to-gate action priority matrix

Article References: Durgam, H. K. R., & Ramani, P. V. (2026). Z-Number reliability-weighted AHP–true Euclidean TOPSIS model for green upstream construction supply chains: A cradle-to-gate action priority matrix. Cleaner Engineering and Technology, 34, Article 101328. https://doi.org/10.1016/j.clet.2026.101328

Image Credits: AI Generated

DOI: 10.1016/j.clet.2026.101328

Keywords: green supply chain, construction, decarbonisation, AHP, TOPSIS, Z-numbers, multi-criteria decision-making, embodied carbon, circular economy, recycled materials, bootstrap resampling, action priority matrix