Abstract
Buildings account for nearly 40% of global energy consumption and one-third of CO2 emissions, intensifying demand for low-impact thermal insulation. Natural and recycled fibrous materials can meet much of this demand, but their performance depends on variables interacting across molecular, micro, and macro scales. This review and multiscale synthesis draws on approximately 130 studies (2005–2026) identified through Scopus, Web of Science and Google Scholar, encompassing data obtained by guarded hot plate, heat flow meter, and hot box methods. Thickness raises thermal resistance linearly without changing intrinsic conductivity (λ), bounded by a cost optimum. Density and porosity show a U-shaped relationship with λ, balancing gas-phase conduction and radiation against solid-phase conduction through inter-fiber contacts. Moisture is the most damaging variable, with sharp performance degradation typically reported beyond moisture contents of 5–6% for lignocellulosic systems, unless hydrophobic treatments are applied. At the molecular scale, crystallinity index, microfibril angle, and degree of polymerization create intrinsic anisotropy: flax and hemp (CrI ≈ 80–90%) conduct axial heat more readily than cotton (65–75%). Recycled cotton, PET, and textile blends reach λ = 0.034–0.044 W/(m·K), with cradle-to-gate embodied-CO2 reductions on the order of 5–10× relative to mineral wool reported in several life-cycle assessments, depending on functional unit definition and system boundaries. Effective insulation balances competing transport mechanisms rather than maximising any single property; priority next steps are multiscale modelling, long-term in-situ hygrothermal data, and integrated multifunctional assessment.