Press Release: August 09, 2020
Energy Consumption is one of the most important factors for reducing the environmental footprint and also for increasing the building energy efficiency to cut down the overall investment. The energy consumption of the building can be estimated from the building’s name plate data and operating schedule of the systems. However, it’s a rough estimate and does not consider other factors like climate and building system interaction. Thus, when opting for building energy efficiency measures, energy modelling can be a very powerful tool.
A study by the American Institute of Architects (AIA) found that buildings with energy modeling saved a considerable amount of energy as compared to buildings without energy modeling. The study found out that without energy modeling, the buildings achieved 37% energy savings while buildings with energy modeling were able to save 52% energy. In other words, energy efficiency projects in buildings were able to save 15% more energy when energy modeling was used.
To understand the energy models, lets look at the 3 main types depending on how each model process information.
White Box: Energy Modeling with Physics
Physics based energy models are the most accurate ones and software like DOE-12 and EnergyPlus use this approach. Since the creating of physics-based models require all the necessary data and equations, white box energy models are quite challenging and demanding. Due to their high level of demand and complexity, the models make the computer simulation slow.
Apart from being highly accurate, white box energy models don’t depend on any historical data which makes it possible to simulate a non-existing building as long as all the physical properties are known. With the right engineering expertise and computing power, white box energy models can provide valuable insights and information for the project owners and developers.
Black Box: Energy Modeling with Data
Unlike white box models that are entirely physics based for predicting behavior, black box models use reverse engineering with already existing data. Since the data required for black box models is already available, the calibrations are easy and can process much faster. Some of the known examples are Artificial Neural Networks (ANN), Support Vector Machines (SVM) and Statistical Regression.
The dependency of pre existing data can also be one of the main limitations at times. Creating a black box model for a non-existing building is impossible since these energy efficiency applications depend on buildings that produce data or on buildings with similar properties. Without a pre-existing data, data-based energy modeling cannot be calibrated.
For existing buildings, black box models can be useful during managing energy efficiency measures since the impact of these measures can be simulated even before implementing them. It allows to analyze the building issues by pinpointing the causes once the model is created.
Grey Box: Hybrid Energy Modeling
Grey Box have the design elements of both white box and black box models since they use physics equation like white box to represent the building behavior, but these equations are simpler. This allows faster building model simulation once it has been calibrated. However, simplified physics equation results in lower accuracy than white box. To compensate for this, grey box models use historical data for calibration. Thus, grey box energy model offers a balance between both, accuracy and speed as well.
For calibration process, black box and grey box energy models are often referred to as “training” the models, since the parameters of the simulation are adjusted till the model’s results match the behavior of the system being modeled.
Using Energy Modeling Effectively
Depending upon the building requirements, the energy efficiency measures can vary to meet the required optimal conditions. Thus, neither of the above mention energy models can be considered as better than the other. When there is no data available for the energy consumption profile of a building, white box model is the only option which can also be used to compare the actual behavior of the building with the physics-based behavior of the building. On the other hand, available data from the existing buildings can provide some valuable insights for the building owners when opting for the required energy modeling measures.
For building in New York City, energy modeling can be very useful considering the stringent emission limit based on the Local Law 97 of 2019.