This may be the author’s version of a work that was submitted/acceptedfor publication in the following source:

Toth, Bianca, Salim, Flora, Drogemuller, Robin, & Frazer, John(2011)Support for energy-oriented design in the Australian context.In Bodart, M & Evrard, A (Eds.) Conference Proceedings of the 27th In-ternational Conference on Passive and Low Energy Architecture (PLEA)(Volume 2).Presses universitaires de Louvain, Belgium, pp. 47-52.

This file was downloaded from: https://eprints.qut.edu.au/47226/

c© Copyright 2011 please consult authors

This work is covered by copyright. Unless the document is being made available under aCreative Commons Licence, you must assume that re-use is limited to personal use andthat permission from the copyright owner must be obtained for all other uses. If the docu-ment is available under a Creative Commons License (or other specified license) then referto the Licence for details of permitted re-use. It is a condition of access that users recog-nise and abide by the legal requirements associated with these rights. If you believe thatthis work infringes copyright please provide details by email to qut.copyright@qut.edu.au

Notice: Please note that this document may not be the Version of Record(i.e. published version) of the work. Author manuscript versions (as Sub-mitted for peer review or as Accepted for publication after peer review) canbe identified by an absence of publisher branding and/or typeset appear-ance. If there is any doubt, please refer to the published source.

http:// www.plea2011.be/

https://eprints.qut.edu.au/view/person/Toth,_Bianca.htmlhttps://eprints.qut.edu.au/view/person/Drogemuller,_Robin.htmlhttps://eprints.qut.edu.au/view/person/Frazer,_John.htmlhttps://eprints.qut.edu.au/47226/http://www.plea2011.be/

Support for Energy-Oriented Design in the Australian Context

Bianca TOTH1, Flora SALIM2, Robin DROGEMULLER1, John FRAZER1 1School of Design, Queensland University of Technology, Brisbane, Australia

2Spatial Information Architecture Laboratory, RMIT University, Melbourne, Australia

ABSTRACT: There is a need for decision support tools that integrate energy simulation into early design in the context of Australian practice. Despite the proliferation of simulation programs in the last decade, there are no ready-to-use applications that cater specifically for the Australian climate and regulations. Furthermore, the majority of existing tools focus on achieving interaction with the design domain through model-based interoperability, and largely overlook the issue of process integration. This paper proposes an energy-oriented design environment that both accommodates the Australian context and provides interactive and iterative information exchanges that facilitate feedback between domains. It then presents the structure for DEEPA, an openly customisable system that couples parametric modelling and energy simulation software as a means of developing a decision support tool to allow designers to rapidly and flexibly assess the performance of early design alternatives. Finally, it discusses the benefits of developing a dynamic and concurrent performance evaluation process that parallels the characteristics and relationships of the design process.

Keywords: energy, simulation, parametric, modelling, performance.

1. INTRODUCTION Minimising energy consumption in buildings is

critical to achieving carbon reduction targets, both in Australia and worldwide. While there are no precise Australian statistics outlining the energy consumption of the building sector as a whole, recent research in the US exposes this market division as the single largest emitter of greenhouse gases, responsible for almost half of all CO2 produced [1]. Operational energy consumption is the major contributing factor in this statistic, accounting for approximately 90% of building sector emissions (with building construction and materials making up the remaining 10%), and estimated to be responsible for greater than 40% of all energy-related carbon emissions in the US, and 33% globally [1,2].

Our homes and workplaces play a large role in consuming the energy that produces these emissions. In Australia, 11% of the nation’s total energy consumption can be attributed to the operation of residential buildings, while the operation of commercial buildings is responsible for a further 6% [3]. Heating, ventilation and air-conditioning (HVAC) systems are the major consumers of energy in both building types, with their ongoing operation accounting for 39% of usage in households and 66% in commercial buildings [4]. This makes HVAC an obvious target for energy reduction strategies. Since these systems are primarily necessary to moderate and compensate thermal loads that are determined by actual building design, minimising building loads through architecture that works with, rather than against, natural energy flows, offers considerable potential to improve building energy efficiency.

More integrated design processes that incorporate simulation as a decision support tool for investigating the complex relationships between architecture and services are therefore needed. This

is particularly important in the early design stage as the decisions made at this time determine around 80% of the environmental impacts and operational costs of a building [5]. Presently, however, no tools exist to seamlessly integrate performance evaluation into early design in the context of Australian practice.

If this problem is to be addressed, barriers to the integration of energy simulation in early design in Australia must first be investigated, and initiatives for overcoming them proposed. The following section examines the roles of process, representation and technology in this issue of integration, in order to establish a holistic strategy for developing an energy-oriented design system in the Australian context.

2. BARRIERS TO INTEGRATION 2.1. Evaluation Frameworks

Methods for integrating simulation into the design process depend heavily on establishing effective strategies for exchanging information between design and analysis applications. In early design, it is crucial that these interactions are rapid and flexible to support the iterative investigation of design alternatives. However, as will be discussed below, the more common strategies for achieving integration lack these fundamental requirements.

Data model interoperation is the most widely adopted approach for exchanging information, and relies on programs sharing data at the level of the product model [6]. This is accomplished through either model exchange, via a neutral file format that serves as a generic common representation, or model sharing, where a single data management system contains the entire building description from which domain-specific applications can extract required information [7]. As can be seen in Figures 1 and 2, in both cases design and analysis models are developed separately. Issues of data redundancy and inconsistency are often experienced, and

significant amounts of manual remodelling are required to locate, translate and update data between applications [8]. Neither scenario allows interactive data exchanges during the design process, and therefore cannot support the rapid transformations needed to undertake performance evaluation in early design.

Figure 1: Data model interoperation: model exchange.

Figure 2: Data model interoperation: model sharing.

Within more traditional engineering frameworks, a data and process model integration approach is commonly employed to effect interaction [9]. In this scenario, illustrated in Figure 3, a single application provides the faculty to simulate different domains [9]. While this approach does have the advantage of requiring only one model, removing inconsistencies and simplifying data management, it does not generate an open design environment [9]. Programs that employ this strategy tend to provide comprehensive simulation capabilities across a range of engineering domains, but do not typically extend to accommodate architectural design environments. And in cases where attempts have been made to incorporate analysis capabilities into modelling software, the amount of user control over the inherent simulation environment is quite limited. Furthermore, the user is restricted to the features and options offered by the particular program, which may generate unwanted constraints [9]. The creation of the digital model becomes a difficult task, and the expertise required to generate it relegates the modelling to just one or a few people [7]. Since this approach does not enable the shared development of design solutions, it lacks the flexibility required for early design.

Figure 3: Data and process model integration.

A data and process model cooperation approach is currently emerging as an alternative strategy for information exchange [10]. In this scenario, programs are effectively coupled by providing the facility to link to other applications at run time [7]. Generally, one program controls the evaluation process and invokes

other applications as required, automatically generating the necessary analysis models and performing analyses, as illustrated in Figure 4 [7]. This allows information to be cooperatively exchanged during the design process in a manner that is able to be customised to meet the needs of the user [9].

Figure 4: Data and process model cooperation.

The cooperative approach suggests that there must be more to the design process than simply data exchanges supported by a common building representation. It points instead to a flexible design framework that allows simulation programs to be called at the right time depending on the design decision being explored. This research adopts the cooperative approach as the basis for an energy-oriented design system, as it is able to support the rapid and flexible interactions necessary for early design exploration.

2.2. Design and Analysis Representations

Information exchanges between domains are also subject to complex issues of different representation paradigms. Architectural design models use solid geometries to faithfully illustrate the tangible qualities of building components, while energy analysis models require centreline surface geometries to examine the building as a series of filters whose behavioural properties affect energy transfers between spaces [11]. This difference leads to problems and inaccuracies when models are translated between the two domains, as can be seen in Figure 5. In this example, a ‘gap’ is generated between adjoining walls in the analytical model, which results in the two physically separate spaces on either side of the wall being treated as a single space, leading to errors in simulation calculations.

Figure 5: The difference in representation between an architectural design model and an energy simulation model.

Building Information Models (BIMs) also pose further problems in that their detailed data structures are often too complex for early design [12]. These models contain an overabundance of data that becomes redundant in a simulation environment, while simultaneously lacking the information necessary to carry out analyses [13]. A simplified design representation schema that focuses on a greater degree of integration with the analytical domain is therefore needed [14].

Parametric design environments present the opportunity to develop more suitable representation schemas. Their inherent flexibility provides a means of defining building constructs that are semantically compatible, at both a geometric and behavioural level, with the requirements of an energy simulation model. Additionally, by structuring components through declared parameters, models can be manipulated intuitively to generate new design options without manual rebuilding of the design model for each scenario [15]. This enables large numbers of options to be created in short spaces of time, which is ideal for early design exploration [16].

There are two parametric applications that are commonly used in Australian practice and research – Rhinoceros (with the Grasshopper plug-in) and GenerativeComponents (GC). Given that GC has a well-tested extensibility and a long-standing capacity for compiling new user features, and inbuilt capacity to integrate with more conventional BIM technologies, this research selects this modelling application to be used in the development of a cooperative energy-oriented design system.

2.3. Energy Simulation Software

Beyond the questions of representation and evaluation frameworks, there are issues with simulation applications that must also be taken into account. In theory, the capacity of simulation to handle dynamic and iterative design investigations makes it an effective means of evaluation for early design. However, recent research has revealed that in practice, simulation is ranked amongst the lowest methods of decision support used by designers [17]. Experience, intuition and rules-of-thumb are preferred, despite lacking the ability to evaluate climate and design-specific idiosyncrasies that is inherent in simulation [17].

There are a number of reasons why energy simulation has not been taken up as a decision support tool in early design. Traditionally, these tools have been used by services engineers late in the design process, primarily for verification purposes, and as a result, energy simulation requires detailed information (which may not be available in the early design phase) about a building’s construction and services before an analysis can be performed [18]. In addition to the complexity of the software acting as a deterrent to its use, the interfaces for these tools are typically cumbersome, non-visual and unintuitive [19]. The translation of 3D models from architectural software is not well supported by existing data mappings between design and analysis domains, and there is an inability to reuse non-geometric data between projects [20]. Inputs and outputs are largely

numeric, and, consequently, the translation of model descriptions and simulation results is a non-trivial task that often constrains the designers’ ability to understand and decide between design alternatives [21]. As a result, energy analyses are time-consuming and complicated to carry out.

Despite the proliferation of energy simulation applications in the last ten years, there are still no ready-to-use design support tools that specifically cater for the Australian climate and building regulations. In addition to the more general obstacles discussed above, many simulation programs are unable to adequately simulate the latent heat associated with Australia’s high humidity climates, nor the mechanical equipment used to accommodate these environmental conditions [22].

In order to determine the most appropriate software for the development of an energy-oriented design system in the Australian context, a review of a number of energy simulation tools was undertaken. Three criteria were considered as follows: 1. The ability to simulate Australian climactic

conditions and associated HVAC equipment. 2. A demonstrated capacity for software extension

and customisation through the provision of an open and well-documented application programming interface (API) or scripting interface that makes it accessible remotely.

3. The use of verified and validated methods of calculation.

The results of this review are shown in table 1 below.

Table 1: Fulfilment of selection criteria of energy simulation tools.

Simulation application

Australian context

Customisation (API/scripting)

Validated calculations

DesignBuilder 1 1 DeST 2 ? DOE-2 Ecotect Energy-10 EnergyPlus eQUEST ESP-r GBS* Hevacomp 1 1 IES VE OpenStudio 1 1 Tas Trace

* GBS – Green Building Studio. 1 Application is interface/plug-in to EnergyPlus, not a

simulation engine in its own right. 2 Program is only available in Chinese.

To elaborate, the capacity for customisation is necessary so that the simulation application can be adapted to suit the requirements of users in early design, such as a more visual user interface, the automated reuse of standardised data, and custom translations of models from design modelling software. Validated calculation methods might seem contradictory to the development of a strategy with a primary focus on early design, however are

necessary to ensure a degree of reliability in energy consumption and ratings estimates. Without a verified simulation tool, engineers are unlikely to continue to refine the simulations produced by the application throughout the design process, as the inaccuracy produced by unverified tools is unsuitable for system specification and control optimisation tasks that occur later in project development.

As can be seen in the results table, DOE-2 and EnergyPlus (E+) are the only two applications that satisfy all three criteria. Given that E+ is a modular software based on the most popular features of DOE-2, but extended to include additional functionality, it has been selected as the tool most suitable to the development of an early energy design system for the Australian context.

3. PROPOSED DESIGN SYSTEM In response to the barriers identified in the

previous section, this research proposes DEEPA (Dynamic Energy-Efficient Parametric Architecture), an energy-oriented design system that couples GC and E+ to create a decision support tool for early design. The structure of this system (which is currently under development) is illustrated in Figure 6. This openly customisable modelling environment establishes a dynamic and cooperative performance evaluation process that parallels the characteristics and logical relationships of the design process and permits smooth transitions between domain-specific representations. It achieves this by linking the two applications through a server-based performance specification database that assigns building geometry the behavioural attributes required to perform energy simulation. This enables analytical models to be automatically generated from design representations without the need for expert interpretation and translation, so that the modelling environment is able to directly invoke the simulation

process. Evaluation occurs in close to real time, with results being pushed to a web application that displays design options and performance outcomes side-by-side.

The system consists of four key components, which are discussed in the following subsections.

3.1. Custom Plug-In for Parametric Software

Additional features must be embedded inside GC to ensure that design models are able to correctly generate corresponding analytical representations and that the simulation process can be invoked. Energy analysis requires building geometry to be expressed as a series of zones defined by closed sets of planar surfaces, to which behavioural properties must be attributed [23]. Since parametric design environments do not recognise the constructs of ‘zone’ and ‘surface’, new representations need to be created to ensure semantic compatibility. Each ‘surface’ is based on the geometry of polygon, with the addition of an inbuilt ‘construction’ property to allocate conductance values. A ‘zone’ adopts a solid geometry to represent its internal space, and requires the additional properties of ‘activity’ and ‘HVAC system’ to determine internal gains and services loads respectively. Within GC, these behavioural properties are simply tags assigned to the geometry. Once the energy simulation procedure is invoked from within the modelling software however, the plug-in sends a query to the database to retrieve the relevant attribute data for each tag, as well as additional information concerning weather and building scheduling. This information, along with the geometric data from the model, is then forwarded to E+ to be analysed.

 

3.2. Performance Specification Database

The performance specification database is essential for ensuring that the necessary behavioural

Figure 6: Structure of the DEEPA system.

properties can be attributed to the building geometry without overcomplicating input requirements for the user. This separates the numerical representation of data from the design model, allowing designers to focus on the manipulation of form and space. A Graphical User Interface (GUI) is used for populating, viewing and editing this database, with the data visually organised into individual tab separators for each of the following: Construction Types: to define the thermal

properties of the materials and their combination and position within the building.

Activities: to define the internal gains for occupancy, lighting and equipment, as well as ventilation rates, for each activity being housed.

HVAC Systems: to define the climate control systems being used in the building.

Schedules: to define the hours of occupancy and systems operation.

Environment: to define the weather data to be used in the simulation calculations.

The database is hosted on a web server, but can also run as stand-alone. Users can work in connected or disconnected modes, depending on the availability of internet connection. In addition to this, the simulation process can also be invoked from within the database, so that building attributes can be further refined once a design is selected.

3.3. Server-Side Energy Simulation

E+ performs analysis on a text-based representation of the building data known as an Input Data File (IDF), which is created when the simulation procedure is invoked. The results of the analysis are then generated as CSV and HTML files, so that they can be displayed directly in the results interface. At the same time, the geometry from the parametric model is stored in the database for on-demand visualisation of the three-dimensional data. This ensures that a snapshot of the design is captured for every simulation that is performed, to assist in keeping track of the design options and their respective performance evaluations.

3.4. Web Application for Results Visualisation

As well as displaying the results of the energy analysis, this web application is also embedded with a Java applet that displays the stored geometry, so that design options and performance outcomes can be viewed side-by-side by multiple users. Simplified simulation results are also returned to GC to provide the designer direct access to the performance outcomes within the design environment.

4. KEY BENEFITS This system moves away from the trend of linking

energy analysis to BIM-based design environments, and the use of data-oriented integration strategies, and turns instead to parametric modelling software for the rapid and flexible process-based exploration that it offers early design. Although it makes use of verified methods of simulation, this tool does not seek to provide precise estimations of operational energy consumption, but rather to provide a reliable

means of comparing early design alternatives, so that more informed decisions can be made. By ensuring that a high level of consistency is maintained in the structuring of analytical models, designers will be able to directly observe the impacts of their decisions by comparing the performance of different design alternatives.

Four key benefits arise from establishing a system that focuses on process integration: Collaboration: Architects and engineers are able

to work in parallel, with the architects undertaking the modelling of different design alternatives, using input from the performance specification database that is manipulated and refined by the engineers to ensure accuracy in the results. This integrates these typically separate tasks of design and specification to produce a holistic understanding of the building, while mimicking the existing workflows of each discipline. With the design and analysis outcomes being published to a common web application, different disciplines are able to review options simultaneously and make decisions collaboratively.

Iteration: By integrating energy simulation into a parametric modelling environment, design options can be produced and assessed rapidly, allowing more alternatives to be considered.

Customisation: Users have the freedom to define their own construction types, activities, HVAC systems and building schedules within the property specification database. This is a key characteristic of the system, as one of the primary downfalls of simplified energy analysis applications that attempt to accommodate early design is that the default properties data is largely hidden from the user and usually only suits the climate and context in which the program was developed. In addition to this, the actual coupling link is customisable and can be extended as required to include further capabilities such as code-checking.

Scalability: The system accommodates various usage scenarios, from a single user working on a local computer to multiple users accessing the database, server, and results, and can swap between modes of operation at any stage in the design process.

It is envisaged that as well as permitting the performance of early design alternatives to be evaluated, this system will also facilitate the sharing of design intelligence across disciplines, so that a more holistic understanding of the factors affecting energy consumption can be developed.

5. CONCLUSION While recent advances in both modelling and

simulation software have given rise to opportunities for direct links between design and analysis applications, there is still a need for decision support tools that integrate energy simulation into early design in the Australian context. This paper has presented the structure for DEEPA, an energy-oriented design system that is currently under development which will address this need.

In the long run it is envisaged that this system will be extended to link seamlessly into more conventional BIM-based modelling systems for design development in the later design stages, and to include other simulation environments, such as structure and daylighting. Additionally, while this tool is being developed for the Australian context, it is expected that its inherent flexibility will result in it being readily adaptable to other locations and climates. But more important than these technological developments are the impacts for process integration that a system like this points to. It is anticipated that the introduction of flexible and concurrent design and analysis environments, such as the one presented in this paper, will open up a dialogue between architects and engineers that is of as great a benefit as the performance evaluation capabilities that these systems provide.

6. ACKNOWLEDGEMENTS The research reported here has been funded by

the Australian Research Council. We acknowledge their support and that of the Queensland Government Project Services in Brisbane for the opportunities that they provide by extending their workplace to foster this research.

7. REFERENCES [1] Architecture 2030, 2010, Problem: The Building

Sector, viewed 2 November 2010, .

[2] Intergovernmental Panel on Climate Change, 2007, Climate Change 2007: Mitigation of Climate Change, Cambridge University Press, New York.

[3] Australian Bureau of Statistics, 2008, 2008 Year Book Australia (Catalogue No.1301.0), Australian Bureau of Statistics, Canberra.

[4] Ambrose, M 2009, ‘Energy-efficient planning and design’, in Technology, Design and Process Innovation in the Built Environment, eds P Newton, K Hampson & R Drogemuller, Spon Press, London and New York, pp.238-249.

[5] Bogenstätter, U 2000, ‘Prediction and optimization of life-cycle costs in early design’, Building Research & Information, vol.28, no.5, pp.376-386.

[6] Augenbroe, G, de Wilde, P, Moon, H & Malkawi, A 2004, ‘An interoperability workbench for design analysis integration’, Energy and Buildings, vol.36, no.8, pp.737-748.

[7] Citherlet, S, Clarke, J & Hand, J 2001, ‘Integration in building physics simulation’, Energy and Buildings, vol.33, no.5, pp.451-461.

[8] Schlueter, A & Thesseling, F 2009, ‘Building information model based energy/exergy performance assessment in early design stages’, Automation in Construction, vol.18, no.2, pp.153-163.

[9] Hensen, J 2004, ‘Towards more effective use of building performance simulation in design’, in Design and Decision Support Systems in Architecture and Urban Planning, eds J Van Leeuwen & H Timmermans, Eindhoven University of Technology, Eindhoven, pp.291-306.

[10] Malkawi, A 2004, ‘Developments in environmental performance simulation’, Automation in Construction, vol.13, no.4, pp.437-445.

[11] Steel, J, Drogemuller, R & Toth, B 2010, ‘Model interoperability in building information modelling’, Software and Systems Modelling, In press.

[12] Parthenios, P 2005, ‘Conceptual Design Tools for Architects’, Ph.D. thesis, Harvard University.

[13] Mahdavi, A 2004, ‘Reflections on computational building models’, Building and Environment, vol.39, no.8, pp.913-925.

[14] Kolarevic, B 2003, ‘Computing the performative in architecture’, in Proc. of the 21st eCAADe Conference, Graz, pp.457-464.

[15] Aish, R & Woodbury, R 2005, ‘Mulit-level interaction in parametric design’, in Smart Graphics, eds A Butz, B Fischer, A Krüger & P Olivier, Springer, Berlin/Heidelberg, pp.151-162.

[16] Gane, V & Haymaker, J 2009, ‘Design scenarios: Methodology for requirements driven parametric modeling of high rises’, in Proc. of the 9th CONVR Conference, Sydney, pp.79-89.

[17] Pedrini, A & Szokolay, S 2005, ‘The architects approach to the project of energy efficient office buildings in warm climates and the importance of design methods’, in Proc. of the 9th IBPSA Conference, Montreal, pp.937-944.

[18] Maasen, W, De Groot, E & Hoenen, M 2003, ‘Early design support tool for building services design model development’, in Proc. of the 8th IBPSA Conference, Eindhoven, pp.761-768.

[19] Punjabi, S & Miranda, V 2005, ‘Development of an integrated building design information interface’, in Proc. of the 9th IBPSA Conference, Montreal, pp.969-976.

[20] Nicholas, P & Burry, M 2007, ‘Import as: Interpretation and precision tools’, in Proc. of the 12th CAADRIA Conference, Nanjing, pp.249-257.

[21] Prazeres, L, Kim, J & Hand J 2009, ‘Improving communication in building simulation supported projects’, in Proc. of the 11th IBPSA Conference, Glasgow, pp.1244-1251.

[22] Crawley, D, Hand, J, Kummert, M & Griffith, B 2005, Contrasting the capabilities of building energy performance simulation programs, US Department of Energy, Washington.

[23] Dong, B, Lam, K, Huang, G & Dobbs G 2007, ‘A comparative study of the IFC and gbXML informational infrastructures for data exchange in computational design support environments’, in Proc. of the 10th IBPSA Conference, Beijing, pp.1530-1537.