J. S. Haberl

Building Energy Use Prediction and System Identification Using Recurrent Neural Networks

Following several successful applications of feedforward neural networks (NNs) to the building energy prediction problem a more difficult problem has been addressed recently: namely, the prediction of building energy consumption well into the future without knowledge of immediately past energy consumption. This paper will report results on a recent study of six months of hourly data recorded at the Zachry Engineering Center (ZEC) in College Station, TX. Also reported are results on finding the R and C values for buildings from networks trained on building data.

Exploring new techniques for displaying complex building energy consumption data

Abstract This paper explores advanced data displays which may help building operators better understand complex energy data by enhancing the display of the data with animation (or time-sequencing). Animated displays such as the ones developed in this paper enhance the usefulness of static graphic displays because time and temperature dependent trends can be immediately seen. This is particularly useful for buildings because many of the energy consuming loads are schedule and temperature dependent. There is an increasing need for new display paradigms that can help building operators visually diagnose complex problems that may otherwise not be detected by efficient energy management and control system (EMCS) algorithms. This need becomes even more important during times of a shrinking labor pool as building operators are being asked to perform more complex control and monitoring tasks. In this paper animated displays have been developed specifically for use in viewing building energy data. Several examples are provided from a large engineering center in central Texas where the animated displays make a faulty flow meter easier to diagnose and allow the operator to visually detect simultaneous heating and cooling.

A framework to integrate object-oriented physical modelling with building information modelling for building thermal simulation

This paper presents a framework for integrating building information modelling (BIM) and object-oriented physical modelling-based building energy modelling (BEM) focusing on thermal simulation to support decision-making in the design process. The framework is made of a system interface between BIM and Modelica-based BEM and the visualization of simulation results for building designers. The interface consists of the following two major features: (1) pre-processing BIM models to add required thermal parameters into BIM and generate the building topology and (2) translating BIM to Modelica-based building energy modelling automatically and running the thermal simulation. The visualization component presents the simulation results in BIM for designers to understand the relationship between design decisions and the building performance. For the framework implementation, we have created a ModelicaBIM library and utilized the Modelica Buildings library developed by the Lawrence Berkeley National Laboratory. We conducted a case study to demonstrate and validate the framework simulation results.

Translating Building Information Modeling to Building Energy Modeling Using Model View Definition

This paper presents a new approach to translate between Building Information Modeling (BIM) and Building Energy Modeling (BEM) that uses Modelica, an object-oriented declarative, equation-based simulation environment. The approach (BIM2BEM) has been developed using a data modeling method to enable seamless model translations of building geometry, materials, and topology. Using data modeling, we created a Model View Definition (MVD) consisting of a process model and a class diagram. The process model demonstrates object-mapping between BIM and Modelica-based BEM (ModelicaBEM) and facilitates the definition of required information during model translations. The class diagram represents the information and object relationships to produce a class package intermediate between the BIM and BEM. The implementation of the intermediate class package enables system interface (Revit2Modelica) development for automatic BIM data translation into ModelicaBEM. In order to demonstrate and validate our approach, simulation result comparisons have been conducted via three test cases using (1) the BIM-based Modelica models generated from Revit2Modelica and (2) BEM models manually created using LBNL Modelica Buildings library. Our implementation shows that BIM2BEM (1) enables BIM models to be translated into ModelicaBEM models, (2) enables system interface development based on the MVD for thermal simulation, and (3) facilitates the reuse of original BIM data into building energy simulation without an import/export process.

Origins of analysis methods used to design high-performance commercial buildings: Whole-building energy simulation

Many commercial buildings today do not perform the way they were simulated. One potential reason for this discrepancy is that designers using building energy simulation programs do not fully understand the analysis methods that the programs are based on and may therefore have unreasonable expectations about the actual system performance or energy use. Therefore, the purpose of this study is to trace the origins of the most widely used building energy simulation programs and the analysis methods of thermal envelope loads used in the software to analyze high-performance commercial buildings in the United States. Such an analysis is important to better understand the capabilities of building energy simulation programs so they can be used more accurately to simulate the performance of an intended design. In this study, a new comprehensive genealogy chart was developed to support the explanations for the origins of the analysis methods of thermal envelope loads used in whole-building energy simulation programs. Two other works (Oh and Haberl 2015a, 2015b) explained the origins of the analysis methods of solar photovoltaic, solar thermal, passive solar, and daylighting simulation programs.

Origins of analysis methods used to design high-performance commercial buildings: Solar energy analysis

This study reviews the origins of the analysis methods used to design high-performance commercial buildings. This study focuses on the origins of the analysis methods used in solar thermal, passive solar, and solar photovoltaic analysis software, developed in the United States and Canada, using a new comprehensive genealogy chart. This historical analysis is important because it gives readers a better understanding of the fundamentals of the analysis methods. The origins of the analysis methods of whole-building energy and daylighting simulation programs were reviewed in other works (Oh and Haberl 2015a, 2015b).

Origins of analysis methods used to design high-performance commercial buildings: Daylighting simulation

This study presents a review of the origins of the analysis methods used to design high-performance commercial buildings. This study includes the origins of the analysis methods used in daylighting analysis software developed in the United States. The analysis of this study can help readers better understand and identify the analysis methods used in daylighting simulation programs. In other works, the origins of the analysis methods of whole-building energy and solar energy analysis software were reviewed (Oh and Haberl 2015a, 2015b).

Function-based flow modeling and animation

This paper summarizes a function-based approach to model and animate 2D and 3D flows. We use periodic functions to create cyclical animations that represent 2D and 3D flows. These periodic functions are constructed with an extremely simple algorithm from a set of oriented lines. The speed and orientation of the flow are described directly by the orientation and the lengths of these oriented lines. The resulting cyclical animations are then obtained by sampling the constructed periodic functions. Our approach is independent of dimension, i.e. for 2D and 3D flow the same types of periodic functions are used. Rendering images for 2D and 3D flows is slightly different. In 2D function values directly are mapped to color values. On the other hand, in 3D function values are first mapped to color and opacity and then the volume is rendered by our volume renderer. Modeled and animated flows are used to improve the visualization of operations of rolling piston and rotary vane compressors. Copyright

Literature review of building peak cooling load methods in the United States

Today, sizing HVAC systems plays an important role in the building design process, requiring an accurate method in order to avoid problems from over- or under-sized systems. To date, there have been five peak cooling load methods published by ASHRAE, including: the Total Equivalent Temperature Difference/Time Averaging Method, the Transfer Function Method, the Cooling Load Temperature Difference/Solar Cooling Load/Cooling Load Factor Method, the Heat Balance Method, and the Radiant Time Series Method. This study provides a thorough review of the five methods with respect to their history and summarizes the method differences that can lead to inaccurate sensible cooling load calculations.

Field-test of the ASHRAE/CIBSE/USGBC performance measurement protocols: Part II advanced level energy protocols

The American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE)/Chartered Institution of Building Services Engineers (CIBSE)/U.S. Green Building Council (USGBC) performance measurement protocols provide a standardized set of protocols for measuring and comparing the operational performance of occupied commercial buildings. The ASHRAE performance measurement protocols have been developed at three levels of cost/accuracy, Basic (Indicative), Intermediate (Diagnostic), and Advanced (Investigative), for the following six performance categories: energy use, water use, thermal comfort, indoor air quality, lighting, and acoustics. This article presents the results of an effort to develop and apply a field test to evaluate the Advanced level of the performance measurement energy protocols in a case-study office building in Central Texas. The data collected include multi-year sub-hourly whole-building electricity data for the total and major end uses, thermal data for chillers and condensers, and coincident on-site weather data, as well as monthly utility bills of the case-study building. The data collected were then analyzed to calculate the corresponding performance metrics based on the Advanced level of the performance measurement protocols energy protocols and compared with the appropriate benchmarks. The problems and issues with implementing the performance measurement protocols Advanced level energy protocols in a case-study building were noted throughout the entire research process. The evaluation revealed four issues, for each of which recommendations were developed to improve the current version of the performance measurement protocols. The results for the performance measurement protocols Basic level applications, including all six areas, were reported in Kim and Haberl (2012a). A companion paper (Kim and Haberl 2017) presents the results of applying the performance measurement protocols Intermediate level energy protocols.