Calibrated Weibull survival analysis for reconstructing Malaysia’s active road transport fleet and energy demand

Abstract

Malaysia faces significant environmental and fuel security challenges due to rising vehicle registration from demographic and economic growth. Accurately estimating the active vehicles (N) is crucial for assessing fuel consumption (FC), emissions, and developing effective urban transportation strategies. However, our understanding of N by vehicle classification and annual FC by category remains limited. This study proposes a benchmark-constrained Weibull survival model that calibrates category-specific lifespans and a common shape parameter to observed fleet benchmarks (2017 by category and 2021 total). The calibrated survival functions are integrated with utilization-based average distance travelled (ADT) adjustments to reconcile FC estimates with national energy statistics, thereby jointly addressing fleet reconstruction and energy demand estimation. Applied to seven vehicle classes from 1997 to 2023, the model yields survival scales (T*) of about 12 to 23 years with a common shape parameter (b*) of 2.123 and reproduces observed active fleets, projecting N of 20.27 million in 2017, 21.11 million in 2019, 21.71 million in 2021, and 22.90 million in 2023. A Monte-Carlo perturbation analysis confirms that N remains stable under moderate parameter uncertainty. The estimated N was then used to quantify 2019 FC by fuel type. After ADT adjustments, results align closely with the National Energy Balance (NEB), with petrol and diesel deviating by −0.06% and + 0.03%, respectively. The study provides a transparent and transferable framework for deriving N, age structure, and fuel demand in data-limited contexts, offering empirical basis for transport energy planning, fuel security assessment, and low carbon scenario analysis.