• Survival rates of Japanese building stock analyzed using inflow and stock data. • Log-normal fits wooden buildings; Weibull fits non-wooden buildings. • 2021 half-life: 50 years (wooden), 52 years (non-wooden), increasing over 30 years. • Projected building half-lives to 2050 span trend and saturating scenarios. • Results inform long-term building stock management and environmental planning. Understanding long-term changes in building lifespan at the national level is essential for predicting future building-related stocks and flows, assessing energy consumption and environmental impacts, and developing effective countermeasures. This study calculated the survival rate of building stock over time for each inflow and stock year and estimated appropriate survival functions and their key parameter—the half-life (defined as the number of elapsed years at which the survival rate reaches 50%)—based on time-series data on building stock by construction year in Japan. Logistic, normal, log-normal, Weibull, and gamma distributions were considered for survival functions. For stock-year-based analysis, the most suitable survival functions were the log-normal distribution for wooden buildings and the Weibull distribution for non-wooden buildings. As of 2021, the estimated half-life was 50 years for wooden buildings and 52 years for non-wooden buildings, an increase of 15 and 16 years, respectively, over the past 30 years. Future half-life projections toward 2050 were explored under multiple scenarios, including unconstrained and saturating forms, to explicitly reflect structural uncertainty in long-term projections. The findings provide insights for projecting future material stocks and flows in the building sector and for assessing life-cycle environmental impacts, thereby supporting environmental planning and policies.
Kayo et al. (Sun,) studied this question.