The early phase of design holds significant importance in the architectural design process. Well-considered design decisions during this stage lay the groundwork for achieving energy performance goals effectively. It is therefore crucial to explore a wide range of options to inform the design process. Running and evaluating hundreds of simulations using robust, full-featured energy modeling software has proven time-intensive and cumbersome. Meanwhile, simulation tools that focus on early-design and bulk results have issues with accuracy and longevity in the design process.

To tackle this problem, my colleagues and I developed a Python-based parametric program within IES Virtual Environment to perform parametric simulations using a detailed energy model that can evolve with the design. The tool has the following three main outstanding features:

Automation, streamlining and flexibility. The tool has the ability to assemble and seamlessly run hundreds of simulations consecutively without interruption. The intuitive user interface provides a wide selection of parameter options, including envelope U-factor, SHGC, WWR, etc., and it offers the modeler the flexibility to create customized simulation sets that enable/disable certain parameters as needed. By automatically running every conceivable combination within the parameter space, the script eliminates the necessity for manual intervention, saving substantial time and minimizing the effort and resources typically associated with such analyses.

Accuracy and reliability. Diverging from simplified software alternatives, this tool is built on a sophisticated and full-featured simulation engine. For instance, instead of relying on basic drag-and-drop systems, it utilizes a complex HVAC system network capable of accommodating various mechanical system design features, such as heat recovery, mixed-mode ventilation, advanced sequence of operations, etc. This greatly enhances the precision and reliability of results, providing a more realistic representation of energy performance, thereby providing meaningful information for design decision-making.

Data reporting and visualization. The tool goes beyond simulation by leveraging additional scripts developed in-house. It generates bulk results based on all simulations conducted and presents them through an interactive web-based platform called "The BEM ToolBox". By providing a dynamic and visual representation of complex data (such as a parallel coordinate plot), this dashboard not only facilitates a thorough understanding of the data but also allows modelers to investigate results efficiently. Furthermore, the dashboard can be shared with clients and partners with a public link which prompts efficient collaboration within the project design team.

In conclusion, the integration of automation, accuracy, and data reporting/visualization features of this custom-developed parametric simulation tool is a transformative solution to iterative energy analysis in the early design phase, offering a substantial leap forward in efficiency and effectiveness of building energy performance optimization.

The success of recent machine learning (ML) methods in natural language processing and computer vision is due in part to training on massive, broad data. To facilitate exploring this approach for ML applied to building energy management, we have developed and released the BuildingsBench research platform. The platform provides a large-scale pretraining dataset of building energy consumption time series sourced from the NREL End Use Load Profiles (EULP) dataset, which is a collection of building simulations that emulate the U.S. residential and commercial building stock. The pretraining dataset consists of 1.8 million 1-hour resolution time series (550,000 residential and 350,000 commercial time series across two weather-years). Our platform can be used to study approaches that aim to learn "one model for all buildings", i.e., "foundation models" that learn universal patterns of energy consumption shared by many buildings. We demonstrate that models pretrained on the simulated dataset can achieve strong performance on the downstream task of short-term load forecasting for real buildings. Future iterations of BuildingsBench will expand the benchmark tasks beyond forecasting to include, for example, large-scale building stock surrogate modeling and urban building energy modeling. The BuildingsBench code and data can be accessed from https://github.com/NREL/BuildingsBench.

The concept of embodied carbon refers to the total greenhouse gas emissions associated with the entire life cycle of a product or system, from raw material extraction to manufacturing, transportation, installation, and eventual decommissioning. Understanding and mitigating embodied carbon is essential for making informed decisions to pursue environmentally conscious energy solutions.

Our journey begins by exploring the landscape of PV system installations, delving into the two primary methods – ballasted and roof-mounted. Ballasted installations, where solar panels are secured on a structure using weights, and roof-mounted installations, where panels are affixed directly to the building's roof, present distinct advantages and challenges. While ballasted systems offer flexibility and ease of installation, roof-mounted solutions integrate seamlessly with existing structures, optimizing space and aesthetics.

The heart of this presentation lies in the comprehensive analysis of the embodied carbon associated with these two installation methods. We scrutinize the entire life cycle, evaluating the carbon footprint at each stage. Every step is dissected to quantify the environmental impact, from the extraction of raw materials for manufacturing to the transportation of components and the actual installation process. This analysis serves as a crucial tool for decision-makers, architects, and sustainability enthusiasts aiming to make informed choices when adopting solar energy solutions.

As we unravel the carbon complexities, we also address the implications of our findings on the broader sustainability landscape. Does one installation method significantly outshine the other in terms of environmental performance? How can the industry leverage this knowledge to enhance overall sustainability practices? These questions guide our exploration into the potential implications of our analysis on future PV system implementations.

In conclusion, the presentation seeks to empower stakeholders with a deep understanding of the embodied carbon associated with PV systems, particularly in the context of ballasted versus roof-mounted installations. By shedding light on the environmental nuances of these methods, we pave the way for a more sustainable and conscientious approach to harnessing solar energy, contributing to a greener and cleaner energy future.

We demonstrate a new approach to tackle “BIM2BEM” and present preliminary results. Although Building Information Modeling (BIM) has become the industry norm for data generation and management, automatically transitioning from BIM to Building Energy Modeling (BEM) remains challenging. Current “BIM2BEM” processes still require manual intervention to achieve good accuracy and consistency. In many cases, modelers simply create BEM models from scratch, bypassing the potential benefits of using existing BIM data.

This new approach diverges from the conventional BIM2BEM practices, which require detailed descriptions of a building's physical layout, dimensions, and spatial relationships. Instead, we create non-geometric building energy models from the BIM model that focuses exclusively on the building's energy-related characteristics. The adoption of non-geometric modeling offers multiple notable advantages. Unlike methods reliant on geometric precision, our approach does not demand a 'water-tight' geometric model as a starting point. Rather, it concentrates on extracting essential elements directly relevant to energy simulation. This shift in focus not only streamlines the modeling process but also reduces the complexity and specificity needed in the initial gbXML models, thereby facilitating a more efficient and robust workflow for the “BIM2BEM” process. The approach also provides additional benefits to flexibly update geometry models in an integrated design context.

In 2024, the U.S. Department of Energy (DOE), in collaboration with national laboratories, released an updated version of the Quick Building Assessment Tool (QBAT) to further support the Bipartisan Infrastructure Law (BIL) to Renew America’s Schools. Since its initial launch in 2022, QBAT has empowered building managers to generate energy recommendations for their facilities, irrespective of their expertise or available funding. With a commitment to equity, QBAT has significantly reduced the barriers to comprehending a school's existing energy needs and potential improvements. This tool offers users a simplified process to identify retrofits from DOE's advanced energy retrofit guides and ASHRAE's commercial building energy codes and standards. QBAT models a building’s energy performance through an EnergyPlus model, simulating the user's building based on five easily identifiable characteristics. These characteristics inform the simulation’s parameters based on commercial building survey data, building energy standards, and industry expertise.

In the latest update, QBAT’s recommendation engine provides users full customization of over 30 energy efficiency and decarbonization measures and identifies health, and safety benefits for use in their BIL needs assessment applications. The recommendation engine assesses the applicability of these retrofits by considering various parameters. These include climate zone, use type, equipment and building vintage, existing equipment, and ease of retrofit. Retrofits span a spectrum, ranging from like-to-like system replacements and control measures to more comprehensive and high-performance initiatives, with a focus on decarbonization. In QBAT’s updated measure selection window, users choose retrofits with the assistance of a comprehensive guide for each measure. These guides include supporting documentation that cover implementation details and considerations from the perspectives of energy efficiency, health, and safety. With retrofits selected, the simulation engine performs an annual EnergyPlus simulation for both the existing and upgraded building. The simulation results get aggregated into a PDF report, providing users with documentation they need to review the current and proposed energy performance of their building under standard operating conditions. Lastly and most importantly, this report contains the information applicants need to complete the needs assessment application to participate in the BIL to Renew America’s Schools initiative. Through an easy-to-use interface and simple input parameters, QBAT provides an equitable capability to allow school districts with limited resources to meet the application requirements for the Renew America Schools program.

This presentation will delve into QBAT's simplified modeling approach for defining detailed whole building models using limited user inputs, recent updates, workflow, and comparison to detailed audits. It will also showcase the extent to which this tool has been successful is providing an equitable solution for resource-constrained schools districts.

HVAC control sequences of operation standardized by ASHRAE Guideline 36 have demonstrated, through field validation, the potential to reduce energy use by 12–60% in nonresidential buildings compared to typical practice (Taylor Engineering, TRC, Integral Group, 2022). An underutilized opportunity exists for energy savings in building retrofits and retro-commissioning using optimized sequences of operation. The project team developed a calculator to estimate savings from implementing Guideline 36 sequences of operation in existing buildings. This web-hosted calculator is based on an energy modeling database that includes variations of climate zone, building size and HVAC system configuration.

Stakeholder outreach identified the absence of an offering in efficiency programs that is flexible enough to account for building and system characteristics but simpler than a custom energy modeling approach. The team did extensive testing of energy modeling parameters to determine the 13 parameters with the greatest impact on controls measure performance. Based on feedback from stakeholders, most of the inputs are optional, allowing for greater accessibility and ease of use. The calculator includes an uncertainty analysis that accounts for the added uncertainty from unknown building parameters and returns a dynamically calculated uncertainty range.

This presentation outlines the process the project team completed to select the measures for analysis, refine the input parameters and complete two rounds of parametric energy model simulations of 48,000 and 64,000 EnergyPlus simulations respectively. It describes the statistical methods used to create the back end of the calculator and the team’s uncertainty analysis at each stage in calculator development. The presentation also discusses the automated calibration performed by the calculator to reduce error.

The project team recommends that the framework developed here be refined for use by individual energy efficiency programs. The approach is scalable to different building types, climate zones, efficiency measures and levels of accuracy. The presentation discusses lessons learned and proposed next steps in the process of tailoring the savings estimation calculator to specific markets and goals of programs administrators.

Many property owners and building managers are being tasked with retrofitting their buildings to deliver on company, city, and national policies. These stakeholders want to determine for the buildings in their portfolio, which buildings should be retrofit with which measures, often limited with a capital cost budget. They also seek to convince policymakers and local bureaucrats that decarbonization plans are also beneficial to society. This talk discusses on-going research to develop tools to navigate and provide decarbonization solutions to both audiences by pairing building energy models with multi-objective optimization to generate solutions for property owners. By incorporating the metrics of interest of all stakeholders, we are able to generate optimal retrofit solutions that consider both perspectives. In this preliminary analysis, we consider 5 case study buildings to be retrofit with passive measures, specifically wall insulation, roof insulation and windows, considering objectives of minimizing annualized capital and operational cost, minimizing greenhouse gas emissions, and maximizing jobs generated in the economy. The five buildings are 5 commercial buildings located in New York State from the Comstock building energy modeling database, a database with commercial building energy models designed to represent the United States commercial building stock. Each building considered 3 retrofit options for increased wall insulation, three retrofit options for increased roof insulation, and three retrofit options for windows (single glazing low-e, double glazing and triple glazing). The NGSA III genetic algorithm was paired with these models and retrofit options to generation optimal solutions across the three objectives. The outcomes show that moving along the pareto front from high to low emissions (and low to high cost, and low to high jobs), solutions can change drastically with small changes in cost. For example, at $150,000 Annualized cost (an average of $30,000 per building) the optimal solution is to retrofit the walls and roof of building 3, the roof of building 2 and the roof and walls of building 4. At an annualized cost of $175,000 annualized cost (an average of $35,000 per building), the optimal solution is to retrofit the walls and roof of building 3, the walls of building 1 and the walls of building 5. For a $5,000 difference in cost per building, the solution is to retrofit two completely different buildings. This demonstrates that the optimal solutions to portfolio retrofits are very local and what is best to do really depends on the buildings in the portfolio and the specific amounts of funds that are to be spent and expectations for jobs generated. This highlights the need for (1) the development of such tools to generate these solutions and (2) the need to engage deeply with stakeholders to identify the retrofit solutions of interest.

There is a need for building potable water system (including water heating and hot water distribution) performance goals and design criteria that promote water quality, water conservation, and energy conservation. At present, Building Water System (BWS) goals and design criteria are scattered through codes, standards, and programmatic requirements and, particularly in the case of water quality goals, are generally absent, unactionable, or at worst in conflict with each other. As a result, design engineers, installers, and system operators are left without a basis for designing, constructing, and operating BWSs and might be unaware of basic system requirements, particularly those most connected to public health. Establishing comprehensive goals is hampered by knowledge gaps and the wide diversity of systems and supply waters among BWSs; modeling and simulation will play a key role in filling knowledge gaps, refining the goals and helping design professionals design systems that meet or surpass the goals. Present-day models are not yet capable of the range of simulations required for contributing to improved water system performance and the goals reviewed in this presentation can help modelers identify and prioritize model development.

Building Water Systems are among the most impactful sub-systems in buildings. For most types of buildings, water and energy for heating water are among the largest operating costs. For any building with occupants and particularly for health-care facilities, the liability from health risks due to water quality degradation (e.g., costs associated with hospital-acquired cases of Legionnaires’ disease) arising from poor design, installation, commissioning or operation can be staggering.

This session reports a blue-ribbon panel’s draft recommendations on building potable water system performance goals and design criteria. The goals were developed to promote chemical and microbial water quality, safety, water conservation, and energy conservation and to be measurable and actionable. Modelers can contribute to establishing goals and introducing them to practice by synthesizing research studies on fate and transport, safety and sustainability into application tools and generalizing research findings into knowledge accessible by practitioners.

The key simulation and modeling needs related to building water systems will be covered during discussions at the end of the session. The highest priority needs include

• BWS thermal models that accurately simulate thermal dispersion and predict time-to-tap for hot water delivery,

• Water quality models predicting the accumulation and release of metals under realistic use scenarios and for the wide variety of water supplies buildings use,

• Disinfectant fate and transport models, including demand and decay during stagnation and disinfection byproduct formation in water heaters and pipe networks,

• Microbial growth, release and transport models and

• Fixture and fitting pressure loss over the range of Reynolds numbers encountered during daily uses.

Spandrel thermal transmittance can be difficult to pin down. Modelers typically think of spandrel panels as opaque wall, like wood framing or masonry systems, because of how they are treated in energy codes. Most other design professionals and contractors group them with vision glazing because they are part of the curtain wall system. This can lead to discrepancies in how the framing is addressed. To further complicate the matter, there is no standard procedure for reporting the spandrel assembly separate from the vision glazing. Mishandling the framing effects can make an R-6 panel assembly look like R-20. This lightning talk will distill sage advice from our in-house building envelope commissioning agent into modeler quick tips on accounting for framing and product performance.

Building energy modeling (BEM) has been widely used by researchers, regulators, and engineers to quantify building energy performance. Quality assurance (QA) and quality control (QC) of the model's performance are essential parts of such analysis. Currently, the QA/QC approaches are typically performed manually and in an ad-hoc manner, relying heavily on either simulation inputs and/or reported (summarized) outputs. This process can be tedious, error-prone, and time-consuming when verifying a large number of models. In most instances, QA/QC approaches require a modeler or reviewer to spend a considerable amount of time reviewing inputs, outputs, and/or metrics to ensure that the outcome of a simulation is acceptable. Meanwhile, simulation engines are constantly evolving, introducing bug fixes and new features, as well new bugs. Therefore, time-series output verification (with a combination of summarized inputs and outputs) is the only way to ensure that models behave as expected. Only reviewing model inputs will not guarantee the expected simulated behavior of building system components. ConStrain (formally known as ANIMATE) is an open-source data-driven building performance verification framework that conducts automated output-based verification of building operation requirements, with specific focus on time-series-based verification of control requirements. An earlier concept of this framework has been presented at the 2022 Building Performance Analysis and SimBuild conference. This presentation will introduce and demonstrate recent development efforts of ConStrain, which includes the addition of Application Programming Interfaces (API) for existing functionalities, the development of a workflow specification for defining verification jobs and its corresponding API and Graphical User Interface. These new developments greatly improve the functionality and usability of ConStrain in order to verify the implementation of building system controls in simulation models and actual buildings.