Conference Agenda

This study employs a unique high-resolution dataset from a week of controlled cool-down experiments in a thermally lightweight, unfurnished test dwelling conducted during the winter heating period. Dynamic thermal simulations using EnergyPlus were compared to minute-resolution measurements during six-hour nighttime heating shut-offs. The simulations over-predicted the rate of cooling during the first two hours, most relevant to demand response, underestimating the time to lose 1°C by 30 minutes on average. Applying a high zone air capacitance multiplier (>10) improved agreement but was unrealistic, underscoring the need for improved modelling methods to capture short-term cool downs and reliably forecast thermal flexibility.

The global push toward building electrification places new stresses on electrical feeder infrastructure, demanding Urban Building Energy Models (UBEMs) with improved peak load and load duration prediction capability. This study introduces a deterministic, data-driven framework that transforms 3.8 million parcel-days of hourly electricity data from all eligible residential users in a mid-sized county in the northeastern United States into interpretable annual schedule archetypes. 13 typical daily load shapes and 13 seasonal rule archetypes are identified, forming a deterministic “calendar of the city.” The reconstructed Load Duration Curves (LDCs) are validated at building groups level which represents feeder-level prediction capability. The proposed method exhibits a 13.0% reduction in NRMSE in top 1% peak load forecasting and 50.5% reduction in full LDC forecasting compared with Prototype Building Model schedules typically used in building energy modeling tasks. This deterministic “calendar of the city” annual schedule library enables UBEMs to represent collective behavioral uncertainty with statistical certainty, enhancing feeder-level load duration forecasting and informing grid capacity planning under building electrification scenarios.

The traditional approach to designing, specifying, deploying, and testing building HVAC controls in large commercial buildings is labor-intensive and error-prone, leading to high cost and poor operational performance. Recent efforts have enabled a comprehensive, digital workflow for model-based control design, deployment, and verification using open-source technology and software standards. This paper presents a case study that applies this workflow to retrofit a demand-flexible "zone ratcheting" control sequence onto a commercial control platform connected to a virtual building. The study demonstrates that automated semantic binding successfully mapped the control logic to the correct building points, eliminating the need for manual programming. Once deployed, the sequence functioned as intended: it correctly ratcheted the cooling setpoint and transmitted the corresponding signals to the virtual building. Ultimately, this work aims to catalyze transformation in the building industry by enhancing grid flexibility, efficiency, and comfort through improved control sequences using a scalable, digitalized control deployment that allows end-to-end quality control, leading to scalable, reliable and cost-effective adoption of advanced control.

Large commercial buildings struggle to manage peak demand while transitioning to all-electric systems. Current solutions in market such as Air-to-water heat pumps (AWHPs) often need to be paired with Thermal Energy Storage (TES) for efficient load management, however, the design and controls for TES integration with HVAC is also complex. Time Independent Energy Recovery (TIER) offers an integrative solution by combining heat recovery chillers, TES, and AWHPs to address these limitations. Unlike traditional TES applications that focus primarily on shifting or reducing cooling peaks, TIER leverages TES for heat recovery, improving overall energy efficiency.

This work focuses on enhancing the modeling capabilities of EnergyPlus to quantify TIER’s performance. We address key limitations such as the seasonal constraints of heat recovery chillers and over-simplified existing hot water TES tank object. We address this gap by utilizing the plant loop heat pump object to model heat recovery chillers and developing reversible flow hot water tank. These enhancements will facilitate the modeling of complex HVAC systems, and are critical to market transformation and statewide adoption of novel technologies.