Conference Agenda

This study introduces a sensitivity analysis framework to identify key factors influencing energy consumption in single-family homes across California’s 16 climate zones. The workflow includes:

(1) CBECC-Res modeling of eight critical features such as insulation types, fan efficiency, distribution insulation, and major HVAC and water heating systems;

(2) Data ETL using Python libraries for processing and visualization;

(3) Preliminary reporting of feature impacts;

(4) Modeling and sensitivity analysis via PCA, correlation ranking, and SHAP to capture non-linear relationships and prioritize features; and

(5) Final reporting of insights.

Results reveal five dominant features: ceiling insulation, water heating system, space heating system, exterior wall insulation, and attic insulation. Distribution insulation ranks slightly lower but remains influential. These findings highlight the importance of envelope and mechanical systems in energy performance, enabling targeted data collection, accurate modeling, and informed policy decisions for California’s diverse climates.

Building Energy Modeling (BEM) is critical for ensuring compliance with stringent efficiency standards like the California Energy Code (Title 24). Practitioners in California are constrained by compliance software that generates voluminous PDF reports with bland data tables, with more useful data being available only to the most savvy users after considerable effort. While this issue is acute in California due to the mandatory automated baseline and robust ruleset, the challenge of extracting meaningful quality assurance and quality control (QA/QC) insights from dense outputs is a universal pain point across the BEM industry. Gaining actionable insights often requires bespoke solutions that are out of reach for many firms, leading to labor-intensive reviews, missed optimization opportunities, and increased costs.

This presentation introduces a newly developed software framework designed to transform this landscape. Funded by Southern California Edison and the California Building Energy Modeling (CalBEM) community, the project addresses the industry's need for accessible, modern data analysis. The tool parses complex input and output data from Title 24 compliance software (CBECC) and renders it into interactive, user-friendly dashboards. While the current deployment focuses on California compliance, the methodology demonstrates how automated data visualization can streamline QA/QC for any modeling standard where built-in reporting is insufficient.

Attendees will explore the tool’s architecture, which bridges desktop data extraction with a web-based visualization platform. We will detail the technology stack—including open-source tools such as Python, Pandas, and Plotly—and demonstrate how the system empowers modelers to move beyond static page-turning with dynamic filtering and "drill-down" capabilities. A Jupyter notebook was provided alongside the dashboard tools to illustrate the underlying data parsing logic and chart generation and hopefully inspire others to pursue modern tools such as these. Key features to be highlighted include:

Results Reporting: Visualizations that break down site energy use and compliance metrics, clarifying the impact of Proposed versus Baseline (Standard Design) models.

Input Validation: Summaries of critical space-level inputs (e.g., lighting, ventilation) to rapidly flag potential errors.

HVAC Performance: Airside system summaries that isolate capacity and efficiency metrics, helping users pinpoint the specific sources of energy savings or penalties.

Automated QA Flags: Rule-based checks derived from industry standards that automatically highlight anomalies such as zero lighting power or insufficient airflow.

By visualizing the "invisible" data within compliance models, this tool aims to improve model quality, streamline workflows, and democratize access to high-level energy analysis. The presentation will conclude by connecting this work to the broader future of BEM data reporting, specifically referencing insights from the author's recent work with the IBPSA-USA Building Data Exchange project. We will advocate for a shift toward standardized, machine-readable BEM data as the essential foundation for unlocking a robust ecosystem of interoperable QA/QC and reporting tools.

California energy programs increasingly need building prototype models that work consistently across different climate zones, construction eras, and modeling tools, that remain representative of actual building stock. As part of a Southern California Edison (SCE) effort, nonresidential prototypes had to be updated for new construction and four existing vintages, from 1978 through 2022, and aligned with CEUS-2022 data. Producing and calibrating hundreds of models across 16 California climate zones quickly showed that traditional, hand-built prototype workflows were too slow, error-prone, and difficult to scale, prompting the need for a more efficient and transparent approach.

This paper introduces a template-based, engine-agnostic modeling framework that uses standardized scorecards, automated IDF generation, batch simulation, and validation to measured stock data. The goal was to create a workflow that is easy to repeat, easy to review, and flexible enough to support future code development, policy analysis, and resilience studies.

At the core of the framework is a Scorecard Database (Scorecard DB) that brings all prototype inputs together in a format that is readable by people and usable by machines. Each nonresidential prototype includes hundreds of inputs, covering geometry, envelope, internal loads, ventilation and infiltration, HVAC systems, water heating, schedules, and setpoints. Instead of manually entering these values into EnergyPlus models, the Scorecard DB acts as a single source of truth that can be shared across tools and simulation engines. While EnergyPlus v25.1 is used as the reference engine in this work, the approach is intentionally designed so prototypes are not locked to any one platform.

The models are created using a template-based IDF workflow. Prototype-specific IDF templates include placeholders for all values that change by climate zone, vintage, or construction type. Python scripts fill in these placeholders using data from the Scorecard DB and a parametric framework, automatically generating final IDFs for every target case. In total, the framework supports 26 nonresidential prototypes, each modeled across 16 climate zones and four vintages, plus a new construction case, resulting in 2000 of consistent simulations.

Verification and validation are built directly into the process. Automated checks confirm that all placeholders are resolved, units are consistent, and inputs fall within valid ranges before simulations are run. Results are screened for errors, warnings, and unmet hours, then compared to CEUS-2022 end-use data. Where gaps exceed acceptable thresholds, key drivers such as internal loads, ventilation and infiltration, schedules, HVAC performance, and water heating, are adjusted through an iterative calibration loop.

The final outputs include reusable IDF templates, fully populated climate- and vintage-specific models, structured simulation results, gap reports, and documented calibration decisions. By making large-scale prototype modeling faster, clearer, and easier to update, this framework provides a strong foundation for resilient building simulation and future energy code development.

The California Prototypes Development project, funded by Southern California Edison's (SCE) Codes and Standards Program, aims to establish a unified set of prototype building models representing California's building stock. These prototype models—collectively referred to as CalBEM prototypes—enable policymaking agencies in California, primarily the California Energy Commission (CEC) and the California Public Utilities Commission (CPUC) to use an up-to-date representation of the building stock. This paper describes the methodology used for modeling and validating single-family and low-rise multifamily hot water energy consumption in the CalBEM prototypes using data from the 2019 Residential Appliance Saturation Survey (RASS). All modeling was carried out using EnergyPlus™ and the resulting annual unit-energy-consumption (UEC) was benchmarked against the RASS conditional-demand analysis. The paper describes the selection of DHW plant characteristics, development of inputs, schedules, and the methodology used in validation of DHW end-use consumption.