Conference Agenda

With the increasing impacts of climate change, relying solely on historical weather data such as TMY3 is no longer sufficient for designing buildings that remain resilient in the future. The Future Typical Meteorological Year (fTMY) dataset released by the Oak Ridge National Laboratory provides standardized future weather data across multiple RCP scenarios and time periods, removing a key technical barrier. However, practical challenges remain in applying this data within architectural workflows.

This study proposes a replicable climate assessment framework that integrates fTMY datasets into early-stage design through three modules: (1) Hot Days, visualizing extreme heat trends; (2) Cooling and Heating Demand, identifying degree-day shifts and climate zone changes; and (3) Outdoor Comfort, simplifying UTCI-based comfort evaluation.

The framework enables comparative visualization of multiple EPW files, offering accessible insights for designers and clients. It supports clearer communication, cross-disciplinary understanding, and proactive decision-making, advancing climate-responsive design.

With the increasing importance of building thermal evaluations through building simulation, this paper presents a statistics-based method for identifying representative heat and cold events using historical weather data. Applying this method, thermal event weather files were developed for 115 locations across all 50 U.S. states and 16 climate zones. The resulting framework and datasets provide a systematic, data-driven approach for simulating severe temperature conditions. This enables more robust evaluation of building thermal performance, occupant comfort, and health outcomes, while supporting improved energy resource planning and grid reliability under severe weather conditions.

Building performance research increasingly emphasizes thermal resilience—the ability of a building to sustain habitable indoor environments during power outages and extreme weather events. While existing studies propose various simulation-based frameworks, they often diverge in weather data selection, event definition, and metric application, limiting cross-study comparability. This paper systematically examines these methodological differences and evaluates how simulation setups influence resilience outcomes. Through two sets of EnergyPlus-based experiments integrating parametric modeling in Rhino–Grasshopper, we assess (1) the impact of weather data, extreme event selection methods, and outage timing on resilience results, and (2) the sensitivity of three thermal resilience metrics—Passive Survivability Time (PSt), Thermal Autonomy (TA), and Maximum Thermal Stress (MTS)—to architectural parameters across 16 U.S. climate zones. Results show that methodological variations significantly alter the outcome of resilience metrics. Across climate zones, AMY most consistently captures both peak and cumulative extremes. Among event definitions, no single method consistently outperforms the others in identifying unsafe outcomes; different approaches emphasize distinct dimensions of thermal stress, including intensity, duration, and seasonal drivers. Correlation analyses reveal that summer resilience is primarily governed by ventilation-enabled convective and night-flush cooling, whereas winter resilience is shaped by envelope performance and thermal inertia.

Building demand flexibility plays a critical role in balancing supply and demand in low-carbon energy systems, yet profiling flexibility behaviours across diverse buildings remains challenging due to their context-dependent nature. This study presents a data-driven framework to profile time-series demand flexibility behaviours in buildings, aiming for transferability across varying event designs and weather conditions. Using Fast Fourier Transform (FFT)-based decomposition, the method extracts event-independent flexibility patterns and quantifies contextual variations, enabling the derivation of interpretable descriptors. Applied to a synthetic dataset of 2,000 UK office buildings under varied demand response scenarios, the approach identifies representative flexibility behaviour types with potential in transferability to unseen events. This framework is expected to provide a foundation for future work in scalable building classification and time-series prediction of demand-side flexibility.

Energy efficiency (EE) retrofitting often overlooks indoor overheating risks (HR) in non–air-conditioned homes during summer. This study applies NSGA-II optimization to Baltimore rowhouses to balance EE and HR. Relative to the International Energy Conservation Code (IECC) 2021 baseline, the lowest-risk solution reduced HR by 13.5–16.8 % without increasing heating energy use, featuring higher wall and roof R-values, lower window U-factor and SHGC, and higher infiltration rate. Under future (2080–2099) weather scenario, optimal parameters followed a similar trend, but experienced a 140–421 % rise in overheating hours, alongside a 3–6 % decrease in heating demand relative to current conditions. In conclusion, modest adjustments to IECC 2021 could reduce summertime overheating risks without compromising energy performance.

Net zero energy (NZE) buildings are typically designed to tight margins, with on-site renewable energy production matched to a highly efficient building design. But NZE design is usually based on historical climate data, and it is therefore unclear how these buildings will perform in the future and whether they will remain NZE. This study explores the role of future climate change in the design and operation of a case study NZE building. Energy models for a code-compliant baseline and proposed design were simulated under several future climate scenarios. The results show that the proposed building is more resilient to climate change than the baseline, and that it will continue to meet its net zero target, even under the most extreme future predictions. The proposed building will also shift from a winter peak to a summer peak, aligning its energy consumption better with PV generation.

Accurate weather data is essential for reliable building performance simulation, yet most modeling workflows rely on Typical Meteorological Year (TMY) files that do not reflect actual year--to-year weather variability or recent climate trends. To address this gap, this work presents a workflow for generating "Actual Model Year" (AMY) weather files from any available weather station and integrating them into both current-year and future-climate building simulations.

This is generated using a python package created by the Pacific Northwest National Lab, diyepw, which produces fully functional AMY Energy Plus Weather (EPW) files that suitable for use in IES, Energy+, Grasshopper, and other simulation software that can use EPWs to import weather related data. This enables modelers to simulate building performance under the precise weather conditions experienced during a given calendar year, improving calibration accuracy and allowing deeper investigation into extreme events, atypical seasons, or observed operational issues.

Beyond representing current conditions, the workflow also supports future-climate evaluation by taking generated AMYs or other EPWs and morphing them to match climate change scenarios using the Future Weather Generator developed and maintained by the CURA Lab at the University of Coimbra and uses data based on the CMIP climate projections. This generator can also account for the urban heat island effect and it's impact as the climate changes. This creates an AMY for a future year, which can be used to simulate how a building under the same circumstances as a particular year or set of years would behave under the same set of circumstances. These files can be used to answer questions such as:

How would a recent heatwave, wildfire season, or cold spell perform under mid-century climate conditions?

Using AMY files for both present-day and future-shifted scenarios provides two key advantages:

Comparability: AMY-based climate-shifted weather allows direct comparison between current performance and projected future conditions under the same sequence of weather events.

Realism: By grounding future climate assessments in actual historical variability, simulations better represent extreme conditions and operational challenges that matter for resilience planning.

This paper describes the development of the python-based AMY generator and the workflow of using it in combination with Future Weather Generator to create ready to use EPWs for simulating future climate resistant buildings. AMYs can enhance calibration, improve climate resilience assessments, and support more robust design decisions especially when creating multiyear AMYs. This approach offers a practical pathway for modelers to bridge the gap between historical performance evaluation and forward-looking climate analysis.