Energy intensity is a key variable in energy-demand analysis, long-term scenario building, and strategic energy planning. Although global energy intensity has declined over recent decades, substantial heterogeneity persists across countries and economic sectors, complicating the interpretation of future pathways. Moreover, many long-term energy and climate scenarios rely on assumptions about sustained improvements in energy intensity that are only weakly justified by historical evidence. This study develops empirically grounded reference pathways for global and sectoral energy intensity based on historical data for more than 110 countries over the period 1990–2021. Countries are grouped using a data-driven clustering approach that captures similarities in the temporal evolution of energy-intensity trajectories across four major sectors: agriculture, services, industry, and manufacturing. For each cluster, historical trends are summarised using transparent exponential fits to derive conditional baseline pathways extended to mid-century. To account for historically observed variability, empirical uncertainty envelopes are constructed from the dispersion of country-level trajectories within each cluster. These envelopes define plausible ranges of energy-intensity evolution conditional on past dynamics and provide empirical constraints on long-term projections. The resulting baselines reveal a systematic deceleration in energy-intensity improvements and only partial convergence across countries and sectors, with several sectors exhibiting early stabilisation. Comparisons with widely used global and sectoral scenarios indicate that many climate-aligned pathways imply rates of energy-intensity decline that lie outside historically observed empirical ranges, even under optimistic assumptions. Rather than offering forecasts or normative prescriptions, this analysis provides transparent empirical benchmarks to support scenario interpretation, strategic planning, and cross-sectoral benchmarking. • Empirically grounded baselines for global and sectoral energy intensity. • Country clusters reveal distinct long-term energy-intensity trajectories. • Historical dispersion defines empirical uncertainty envelopes. • Evidence of deceleration, partial convergence, and sectoral saturation. • Some climate-aligned scenarios exceed historical empirical constraints.
Barrios-Sánchez et al. (Sun,) studied this question.