Conference Agenda

This paper introduces a data-driven method for assessing economizer performance in air handling units (AHUs) using a combination of physics-based modeling and statistical scoring. The method first employs a thermodynamic AHU model to estimate thermal energy consumption under both actual (monitored) and reference (idealized) economizer operation. The difference between these energy estimates, defined as the energy deviation, reflects the operational performance of each AHU. A probability distribution is then fitted to the deviation data for each system. The cumulative probability is then used as the basis for a normalized performance indicator across systems. Validation across 15 AHUs demonstrated that the proposed method consistently identified units with control faults or inefficient operation. The approach is easily scalable across large numbers of systems, making it an effective tool for large-scale building energy management and maintenance prioritization.

Photovoltaic (PV) cell temperature has been known to impact electricity production. Simulation models that disregard the impact of cell temperature on PV efficiency often overpredict or underpredict power output. This study investigates the impact of cell temperature on PV efficiency across 13,265 locations globally through more than half a million simulations. The simulation results showed that the Simple Model, which does not take cell temperature into account in PV simulation, had -5% to +10% of differences in annual electricity production for the vast majority of the locations. The Simple Model also showed an average Normalized Mean Absolute Error (NMAE) of up to 8% and a maximum NMAE of more than 30% in monthly electricity production. Based on the simulation results, a pair of regression models were trained for more accurate results than the Simple Model yet simple enough to be used for back-of-the-envelope calculations. The Modified Simple Model showed reduced prediction errors of 2.34% for the tropic zone and 1.89% for non-tropic zones when compared to the errors of 6.21% and 3.28% respectively by the original Simple Model.

The accurate prediction of indoor airflow is essential for effective indoor environment control. Computational fluid dynamics (CFD) has long been a valuable tool for such analyses; however, its high computational cost continues to limit practical applications. Several approaches, such as fast fluid dynamics (FFD) and simplified turbulence models, have been proposed to improve computational efficiencies. Nevertheless, these methods rely on simplified or approximated formulations, which often increase numerical diffusion and consequently reduce accuracy. Recently, machine learning-based surrogate models have been introduced to accelerate CFD predictions, yet the still require substantial computational costs for generating large training datasets. To address these challenges, this study proposes impulse-based fluid dynamics (IFD), which is implemented in OpenFOAM for the first time. The proposed IFD reformulates the incompressible Navier-Stokes equations using an impulse transformation developed in the computre graphics field. For quantitative evaluation in indoor environments, two representative indoor airflow cases were selected and compared with conventional CFD and FFD solvers. The results demonstrate that IFD achieves faster computational speed and higher predictive accuracy than existing methods, while maintaining robustness even with reduced mesh resolution. Therefore, this study highlights the potential of the IFD framework as a computationally efficient alternative to high-cost CFD simulations for indoor airflow analysis.

Gray-box thermal network models are essential for building energy prediction and model predictive control, but traditional system identification (SYSID) methods often struggle with unmeasured disturbances such as occupancy, lighting, and infiltration, particularly in small and medium commercial buildings. Existing approaches that incorporate these disturbances as lumped terms or black-box functions can improve identification but often require high-quality data, risk overfitting, and compromise the physical interpretability of gray-box models. To address these limitations, we propose a periodic disturbance (PD) approach that leverages the weekly periodic patterns exhibited by many internal gains in buildings. Rather than minimizing total innovation error, our method models unmeasured disturbances as periodic functions with a weekly cycle, subtracts the periodic mean of innovations, and minimizes the residuals. This enables SYSID to focus on reducing random noise and measurement errors rather than fitting periodic components with the physical model. We validated the PD approach using a calibrated building model derived from FLEXLAB data augmented with stochastic unmeasured disturbances. The PD method achieved consistently superior performance across diverse test scenarios (RMSE: 0.3-2.6C) compared to conventional (CONV), ID, and OD methods (RMSE: 2.2-9.1C), while it still shows slight discrepancies during the heating season. Future work will focus on experimental validation with actual buildings, hybrid modeling strategies to improve future predictions, and multi-zone extensions for buildings with diverse occupancy schedules.

This paper presents a pairwise comparison framework for validating early-design analytical building performance models against detailed simulation tools. Unlike traditional validation approaches based on hourly or monthly output matching, the proposed framework evaluates decision consistency—whether an early-design method and a detailed simulation engine provide the same relative recommendations when comparing alternative design scenarios. This perspective better reflects early-design workflows, where directional guidance is more important than precise numerical prediction. The framework is demonstrated using a large design space comprising 324 building construction and operation scenarios evaluated across 25 cities representing ASHRAE climate zones 0–8, yielding 52,326 pairwise comparisons per climate. Across this testbed application, the framework produces stable and interpretable consistency patterns using both static temperature thresholds and adaptive thermal comfort models. Beyond quantifying agreement, the framework enables systematic identification of scenario pairs for which early-design and detailed models diverge, supporting diagnostic interpretation and targeted model refinement. The proposed approach addresses a gap in existing validation standards, which provide limited guidance for evaluating low-fidelity, abstracted models intended for early design decision-making, and is applicable to a broad class of early-design analytical tools used in education and professional practice.

Existing commercial buildings account for about 35% of U.S. energy use (IEA 2023). Improving their performance requires targeted retrofits supported by accurate, structured, and consistent audit data to guide investment decisions, program-level decision-making, and performance modeling. Energy audits are inherently data-intensive, requiring collection and reporting of detailed building characteristics, systems, schedules, utility data, and energy efficiency measures (EEMs). Inconsistent data formats and reporting mechanisms can impede analysis, hinder the use of streamlined workflows, resulting in process inefficient, added cost and delays.

Audit Template, a web-based platform, supports consistent audit data collection and reporting. Designed to facilitate automated analysis and data analytics, through machine-readable interoperable data formats, it has been adopted by multiple energy audit programs. This paper introduces the Audit Template, articulates the technical advancements that enable comprehensive analysis of existing buildings, and examines the broadened capabilities it affords to both auditors and program administrators.