Driven by ambitious climate goals, the energy infrastructure that provides cooling in buildings is anticipated to undergo rapid change and growth. As space cooling in commercial buildings in the US consumes 11% of the total electricity with a projected growth of 38% by 2050, gaining a deeper understanding of cooling requirements in existing buildings becomes imperative to improve energy efficiency, resilience, and flexibility.
This talk presents simple and scalable data-driven solutions to characterize the cooling requirements in a portfolio of 119 co-located commercial buildings in a warm-summer Mediterranean climate. Our cooling load characterization yields interpretable regressions, which can serve as data-driven benchmarks. These models effectively capture daily variability, explaining over 70% of variance for buildings that collectively represent 85-94% of the portfolio’s overall cooling load in five different years. Our results indicate that the portfolio’s base load cooling ranges from 2.6 to 3.0 MJ/m2/day. Cooling consumption increases by 7.6-9.8% for every 1°C (~1.8°F) rise in average daily outside air temperature. Factoring out geography-driven differences, our key findings reveal substantial usage-driven heterogeneity across buildings. The average estimated base load cooling varies from 0.50 to 4.4 MJ/m2/day across buildings, with the highest loads observed in healthcare and the lowest in residential buildings. Our findings complement electric meter analyses and have implications for building energy retrofits, fault diagnosis, and demand-side management.