Simulating future energy consumption in office buildings using an ensemble of morphed climate data

Designers and policy makers use simulations to characterize building energy performance; depending on localized weather files-typically assembled from historically-measured weather data-to project the building's behavior and energy use. However, changes in global climate and advances in climate science research reveal significant differences between historical meteorological trends and the patterns of current and future climate. The increasing variability and uncertainty associated with climate change will affect buildings in complex ways that depend on the interaction of buildings' properties, human behavior, and climatic context. Previous studies developed and tested methods to create future climate files by modifying historical data, typically assuming a single model of global climate and a single emission scenario. The present study takes a more comprehensive approach, using an ensemble of fourteen Global Climate Models and two Representative Concentration Pathways to incorporate the uncertainty of future climate projections into building energy simulations. To understand the effects on buildings over their lifespan, a prototypical large office was tested in three different US cities (Boston, Miami and San Francisco) and three future time windows (2030, 2060 and 2090). Driven by increases in cooling energy, annual primary energy consumption increased by 2090 for all projected climate conditions tested, by up to 10% in the edge-case climate of San Francisco where cooling requirements had hitherto been minor. There was significant variability in results and drivers among the different locations and projections, emphasizing the need for specific modeling to support local design practices.