Improving The Usability Of Energy Simulation Applications In Processing Common Building Performance Inquiries
Building performance simulation tools must be supplied with large amounts of information. Conventional approaches to gathering such information are often cumbersome and time-consuming. Hence, opportunities for in-depth simulation-supported exploration of design and retrofit options may not be optimally exploited. Thus, efforts are necessary to support users, especially within the design community, in data acquisition for (and communication to) building simulation models. In this context, the present paper reports on a two-fold approach: First, we revisited the requirments in view of necessary building representations (building models) for most common and typical inquiries especially in early stages of design decision making that could be supported by performance simulation. Toward this end, we considered the potential of a previously developed building model (Shared Object Model as deployed in the SEMPER project). Parallel to this step, we explored the types of data input required by two types of existing building performance assessment tools: One tool is an energy simulation application, whereas the second is a calculation tool for the generation of energy certificates for buildings. These two activities resulted in a general building representation detailed enough to address a large set of common early-design performance inquiries. The representation is obviously less broad – and in certain areas less detailed – than the universal Industry Foundation Classes (IFC) model, but is structurally compatible with the IFC model and can be mapped into it. Second, the data input process was aided with a number of measures. For example, recurrent data input instances were captured in terms of default data sets. Such sets include common construction types, frequently used occupancy-related input data (presence, schedules, set-points, etc.), weather conditions, and HVAC system types. Such preprocessed data stes have been shown to simplify and expedite the task of pouplating data models for building performance simulation.
From N. Ghiassi, F Shayeganfar,U. Pont, A. Mahdavi, S. Fenz, J. Heurix, A. Anjomshoaa, T. Neubauer, A. Tjoa: „Improving the usability of energy simulation applications in processing common building performance inquiries„; in:“Simulace Budov a Techniky Prostredi„, O. Sikula, J. Hirs (Hrg.); Ceska Technika – nakladatelstvi CVUT, 1 (2012), ISBN: 978-80-260-3392-9.
„The overall performance of buildings is heavily impacted by design decisions made during early stages of the design process . In these stages where the geometry and semantic properties of the building are subject to constant changes, an evaluation of different design approaches in view of performance can help determine strategies which lead to optimal results. To achieve a relatively precise overview of the performance of a building, performance assessment tools must be supplied with large amounts of information. Such information includes buildings‘ geometry, building components‘ technical properties, occupants‘ presence and actions, micro-climatic data, internal conditions, etc. Conventional approaches to providing a comprehensive building model for this purpose are often cumbersome, costly and time-consuming. Within the design community, where advanced experiences and skills of the professional consultants or researchers in the field of building performance simulation may be lacking, due to usability issues, opportunities for in-depth simulation-supported exploration of design and retrofit options may not be optimally exploited. Given the amount of time and expertise that need to be invested in generating models to analyze and compare the performance of different alternatives, simulation supported design is generally considered not affordable. On the other hand, various options offered by such applications – which in terms call for complicated and comprehensive building representation – may not be required in most common and typical inquiries especially in early stages of design where decision making is based on comparative analysis. Efforts are necessary to support users in data acquisition for (and communication to) building simulation models.“
„Computer Aided Design (CAD) Applications‘ representations of buildings are based on detailed definitions of building geometry. Those definitions thus contain significant amounts of information needed by CAD tools but not used by other types of tools as those tools need only rudimentary definitions of building geometry to operate. Building energy performance simulation tools, as well as many other types of simulation and analysis tools (like acoustics and fire dispersion simulation tools) have their own internal data models of building geometry. Such internal data models represent views of building geometry typically used by the disciplines served by these simulation and analysis tools, and are usually much simpler than the geometry data models of CAD tools . Lack of a clear connection between design applications and performance assessment tools is greatly contributing to the usability issues of these tools. Despite endeavors to secure interoperability between different building related applications, most performance assessment tools, still require manual data input or considerable post-processing due to the specificity of the data format they support. Certain applications such as Ecotect and Green Building Studio are BIM compatible . But the number of BIM users within the design community remains very low as opposed to designers and architects who use 2 dimensional representation capabilities of common CAD applications. The Industry Foundation Classes (IFC) data model, though intended as a universal data exchange model, despite being comprehensive, fails to provide the required data format for performance assessment without a considerable amount of complicated over- work. Moreover, in our experience, the IFC representations of geometry are not unique and may vary, depending on the CAD application and drawing method used, which further complicates the acquisition of the right type of data for assessment purposes. Moreover, the assignment of semantic properties to the building elements is not supported by basic IFC in a manner to account for the relationship between construction elements and the spaces they are attached to. Efforts have been made to automatically derive simulation-compliant models from IFC building representations .“
„To facilitate and up to a certain extent automate the acquisition of data for performance assessment purposes, we considered the potential of a previously developed building model; Shared Object Model (SOM) as deployed in the SEMPER project . Parallel to this, we revisited the requirements in view of necessary building representations (building models) for common and most recurrent performance inquiries. These activities resulted in the development of a fairly comprehensive space-based building data model (Semergy Building Model referred to as SBM), detailed enough to address a large set of common early-design performance inquiries (using normative procedures or simulation) through minor adjustments, and yet retaining a clear structure which helps further simplification of data entry according to the needs and level of expertise of the user. We explored the types of input data required by two types of existing building performance assessment tools: One tool is an energy simulation application (Energy Plus), whereas the second is a calculation tool for the generation of energy certificates for buildings (Archiphysic). The latter is generally speaking an instance of the standard normative procedure in many European countries, as intended by the Energy Performance of Buildings Directive (EPBD). Energy Plus is an advanced calculation engine commonly used for research purposes, run on a text-based Input Data File (IDF) generated through a spread-sheet. Although some user-friendlier interfaces (ex. Design Builder) have been developed to facilitate the use of the application, it remains fairly complicated and confusing for the novice user . IDF is structured in packages of data which can be categorized into four groups based on the nature of the data included and the facilitation method addressed them:
1) Some data packages are crucial to the running of the program and receiving results but are not directly parts of the representation of the building. Such data include simulation parameters, output settings, calculation methods and models used by the program. Although such settings cannot be ignored, for the purpose of common inquiries they can be defaulted to certain values or options to reduce user input.
2) The second group, are essential building related data. Such data packages are parts of the representation of the building (building model) without which the simulation does not run. Such packages include geometry definition, material properties, weather data, etc. Unlike the first category, this information needs to be specified by the user. However, simplification opportunities still exist within these categories. For instance, Geometry can be defined through various methods. (Multiple points coordinates, base point coordinate and two dimensions, areas, various adjacency definition options). In such cases, the most general of the methods was adopted to ensure accuracy as well as simplicity of the stored data.
3) The third category includes data packages which are not essential for the running of the calculation, but have considerable effects on the results. Internal gains, air flows, and HVAC settings are examples of such data packages. Although such data is case specific (similar to the second category), based on some minor input from the user (ex. space functions, HVAC type), geometry-derived values (space volume or area), and average measured values (common infiltration rate, lighting levels, occupants, etc), they can be approximated by simple calculations and thus drastically reduce the effort on the side of the user.
4) The last category includes data packages which are meant for advanced calculations and research purposes such as parametric objects, which can be set aside entirely in case of common performance inquiries.“