Estimating the energy consumption of building systems based on their nameplate data and operating schedule provides only a rough approximation. This method fails to consider local climate variations and the interactions between different building systems. To make informed decisions about energy efficiency measures for building projects, energy modeling emerges as an invaluable tool.
The American Institute of Architects (AIA) conducted a study comparing energy savings in buildings with and without the use of energy modeling. The findings revealed that building owners who implemented energy efficiency measures without modeling achieved a 37% reduction in energy consumption. However, this number increased to 52% when energy modeling was utilized. In essence, energy efficiency projects that incorporated energy modeling realized an additional 15% in energy savings.
Energy models are typically classified into three main types based on how they process information: white-box, black-box, and grey-box models. This classification is commonly used by data scientists and is applicable beyond energy models.
White-Box Models: Physics-Based Energy Modeling
White-box models are grounded in physics and are considered the most accurate. Software such as DOE-2 and EnergyPlus utilize this approach. Developing white-box models is demanding, as it requires comprehensive equations and data. Due to their complexity, these models also demand substantial computing power, resulting in slower simulations.
Despite their complexity, white-box models have the advantage of not requiring historical data. They can simulate a building’s energy performance even before it is constructed, provided its physical properties are known. For real estate developers with access to engineering expertise and computing resources, white-box models offer valuable insights.
Black-Box Models: Data-Driven Energy Modeling
Black-box modeling takes a different approach by relying on existing data to reverse-engineer predictions. These models can be easily calibrated with available data and processed much faster than white-box models. Examples of data-driven modeling methods include artificial neural networks (ANN), support vector machines (SVM), and statistical regression.
The primary limitation of black-box models is their dependence on historical data. These models are only valid for the specific building that produced the data or other buildings with similar properties. Creating a black-box model for a new building is impossible without data for calibration.
Black-box models are particularly useful for managing existing buildings, as they can simulate the impact of energy efficiency measures before implementation. Once developed, these models can analyze building issues and identify their causes.
Grey-Box Models: Hybrid Energy Modeling
Grey-box models combine elements of both white-box and black-box models. They use simpler physics equations to represent building behavior and are thus faster to simulate once calibrated. However, the simplification of physics equations results in a loss of accuracy. To mitigate this, grey-box models are calibrated with historical data, similar to black-box models.
Grey-box models offer a balance between the accuracy of white-box models and the speed of black-box models. The calibration process for black-box and grey-box models is often referred to as “training” the model, where simulation parameters are adjusted to match the system’s behavior.
Using Energy Modeling Effectively
Energy modeling is particularly beneficial for building owners in New York City, especially considering Local Law 97 of 2019, which enforces stringent emission limits starting from 2024. With numerous options for upgrading buildings, owners must identify the best combination of measures to reduce emissions. Ideally, a building retrofit should maximize energy savings and avoided emissions per dollar invested.
The first step in saving energy is to understand how your building uses it. NY Engineers can analyze your consumption with advanced energy modeling techniques. Contact us at (786) 788-0295 or email at info@ny-engineers.com for more information.
