A back-end view to climatic adaptation: Partitioning weather-induced cement demand variance in wet humid environment

摘要

Purpose - This study investigates the level of variance in the real time demand for bagged cement, induced in response to the climatic sequence of the humid tropics, to support best practice calls for a weather-responsive supply chain strategy. Design/methodology/approach - Data on the consumption of cement and site works for 100 ongoing building construction sites were gathered for a period of 12 months. The variance partitioning capabilities of the Ordinary Least Squares and Hierarchical Linear Modelling forms of regression analysis are comparatively used to evaluate the sensitivity of cement demand to the meteorological profile of wet-humid climate Findings - The study outcome provides statistical evidence demonstrating that the meteorological profile of wet-humid climate induces a significantly high percentage of the variance in the real-time demand for bagged cement on construction sites. However, nested within this variance, are the fixed effects of the cement footprint of the building architecture inherent in the locality. Particularly, positive changes to reduce the wet trade composition of buildings or compensating changes in technological bias, are necessary to combat weather interference in the humid tropics. Research limitations/implications - The findings are exploratory, and not for the purposes of holistically forecasting cement demand, and can therefore only form part of a more comprehensive decision support system, bespoke to the study area. Practical implications - The study outcome provides a back-end view to climatic adaptation in wet humid settings, making a compelling case for localized climate-risk adaptive supply chain strategies and policies geared towards sustainability in cement usage. Originality/value - The study delineates the confounding impact of weather, distinct from local building architecture and technological bias, thus creating a methodological platform for replication and comparative productivity studies in diverse geographical areas.