Stock modelling of HVAC systems in Japanese commercial building sector using logistic regression

•The selection of HVAC systems was modeled as logistic regression model.•The stock of HVAC systems in Japanese commercial stock was estimated.•Electricity-driven cooling systems occupies 64% of the stock. The stock of heating, ventilating and air-conditioning (HVAC) systems has not been well understood and its heterogeneity has been often ignored in urban-scale energy demand modelling and analysis although its significant influence on energy demand has been widely recognized. This study estimated the HVAC system stock being used in the Japanese commercial building sector. First, logistic regression analyses were conducted to develop a model to estimate the probability at which HVAC systems are selected when a building is constructed or renovated. The model was applied to building stock segments classified by its predictor variables. The total floor area served by HVAC systems was estimated as the product sum of the estimated HVAC system selection probability and the total floor area of each stock segment. The results showed that share of centralized HVAC systems has been decreasing. The electricity-driven HVAC systems was estimated to occupy 64% for cooling, 23% of which adopted heat storage systems especially in large-scale buildings. The results can be utilized to estimate the energy demand for HVAC of the commercial building stock while addressing the heterogeneity in the HVAC system stock, which would contribute to improving the accuracy of urban-scale energy demand models.