Decision-making tools for evaluating the impact of materials selection on the carbon footprint of buildings

Abstract

The objective of the research described in this article is to improve measurement, prediction and optimization of sustainable building material performance by integrating a decision-making framework for sustainable material selection of building materials with a building information modeling (BIM) tool. Integration of a BIM model with a decision-making tool and sustainable material selection addresses the difficulties of making decisions earlier in the design/build process and allows for specific sustainability trade-off analyses to be conducted, using the actual building conditions and characteristics. It is intended to improve the way building material data is utilized in a building throughout its life cycle, and to model the impact of design, maintenance, operations and occupant behavior modification decisions made in an effort to improve the building’s contribution to a sustainable infrastructure. Pertinent information contained within a BIM model is extracted, and utilized in decision making related to material selection and to the development of ‘what-if’ scenarios. Decision makers will be able to evaluate options for improving their building materials’ environmental sustainability performance. The research provides a new means for sharing data amongst various building modeling programs using a decision analysis model and a new tool for making design decisions related to sustainable building design.

Figure 1.  Conceptual construction of a system dynamics model.

Reproduced with permission from [29].

Figure 2.  Relationship of building information modeling, application programming interface, Java and decision-making (system dynamics) model.

The square shapes represent commercial software, the white rectangles with rounded corners represent existing means of interfacing with these models, and the parallelograms represent the interim steps in moving data from one model to the other.

API: Application programming interface; BIM: Building information modeling; CIMSteel: Computer integrated manufacture of constructional steelwork; CIS/2: CIMSteel integration standards (release 2); gbXML: Green building extensible markup language; IFC; Industry Foundation Classes; ODBC: Open DataBase connectivity.

Reproduced with permission from [44].

Figure 3.  Sustainability metric trade-off using wall thickness as the variable; an example.

Figure 4.  Changing wall thickness based on an external decision-making process; an example.

Reproduced with permission from [44].

Figure 5.  Sustainability metric optimization using recycled content and regional material as the variables; an example.

LEED: Leadership in energy and environmental design.

Figure 6.  Building information modeling-application programming interface code for implementing leadership in energy and environmental design credit calculations; an example.

Reproduced with permission from [44].

There are numerous and significant sustainability-related design challenges facing the building industry. Buildings in the USA consume the majority of electric power and natural gas [1], use a significant portion of water, and are responsible for the majority of waste output and over a third of GHG emissions [2]. The building industry consumes 40% of the world’s raw materials [2,3] (more than any industry apart from food production [4]). A drive to reduce these numbers is changing the way buildings are designed, built, operated and maintained. However, it is difficult to assess the relative improvements in sustainability of one decision versus another [5]. Adding to this difficulty is the fact that many key decisions must be made in the design phase, when the ability to influence project cost is greatest, but when much information about the final design (and future actual performance) is unavailable [6]. “The challenge is thus finding a method to use detailed simulation tools even during the early stages of design when values for many of the variables for the building’s technical sub-systems are not yet available” [7], and to provide the designer with accurate quantitative predictions of the building’s future performance. Since building owners and developers do not have unlimited resources to design and construct their buildings, this creates a dilemma of how best to apply limited building budgets when designing for sustainability. More owners are now requiring or requesting leadership in energy and environmental design (LEED) or other 3rd party certification of their projects. However, to maximize the true sustainability benefits of certification, designers need to know which combination of credits provides the optimal choice of design variables for the building’s sustainability, while keeping their project within its budgetary constraints. The objective of the study summarized in this article is to improve measurement, prediction and optimization of sustainable building material performance by laying the groundwork for integration of decision making into the building information modeling (BIM) design process.

Background

▪ Sustainability indicators

The term ‘sustainable construction’ means “creating a healthy built environment using resource-efficient, ecologically based principles” [3]. However, measuring the sustainability of a building remains problematic. Numerous protocols are currently in use to assess sustainability, including, for example:

	▪ The global reporting initiative, which uses indicators for material use, energy consumption, water use, emissions and waste, and many other environmental and social issues to develop ecological footprint sustainability reports [8];
	▪ Yale University’s Office of Sustainability, which uses three main categories of sustainability metrics: (i) use of natural resources; (ii) systems and processes, which includes procurement, waste management, land use, food, transportation and building design; and (iii) culture, which includes social justice issues on campus [9];
	▪ The European Union’s sustainable development strategy, which includes an extensive list of items for measuring sustainability [10];
	▪ The World Economic Forum is sustainability performance index, which deals with two main categories of indicators: (i) reducing environmental stresses on human health; and (ii) protecting ecosystem vitality [11];
	▪ The American Society for Testing and Materials (ASTM) framework for sustainable design of buildings [12].

In addition, numerous certification and rating systems are available throughout the world for sustainable buildings, including:

	▪ The US Green Building Council’s (USGBC) LEED program [2];
	▪ The Living Building Challenge [13];
	▪ BRE Environmental Assessment Method (BREEAM) [14];
	▪ Green Globes [15];
	▪ Building Owners and Managers Association (BOMA) Building Environmental Standards [16].

While each of these systems requires different performance goals, the categories of environmentally sustainable design features addressed by LEED are illustrative of the general categories of most of these rating systems. LEED, run by the US Green Building Council (an NGO), awards points for various design features, including the major categories of:

	▪ Sustainable sites;
	▪ Water efficiency;
	▪ Energy and atmosphere;
	▪ Materials and resources;
	▪ Indoor environmental quality;
	▪ Innovation and design process [2].

The environmental sustainability of the building is then rated based on a threshold level of points achieved. It is apparent from examination of these sustainability indicators and certification systems that material selection issues occupy important places in nearly every one of them.

The key difficulty in determining the sustainable performance of a building is the fact that the term ‘sustainability’ encompasses a wide range of concepts and the three fundamental components that define ‘sustainability’ (i.e., environmental quality, societal well-being and economic stability) are often in conflict and very difficult to integrate into a single sustainability rating. This difficulty leads to different interpretations of the environmentally sustainable performance of a building [5]. Lacking any acceptable definition of environmentally sustainable performance, discussions of sustainability often deteriorate into discussions of economics, in terms of payback times, life cycle costs and durability. While these are important considerations, they generally miss the point of LEED, or any other rating system, which is ultimately to provide a building that has as small a negative impact on the natural environment and human health and productivity as possible, while being economically viable. Thus, it is important to have a means of making decisions that, while considering costs, can also give an indication of the true level of environmentally sustainable performance of the building.

From another perspective, taking the existing system of determining environmentally sustainable performance as a fixed point, it is important to understand the eco-effectiveness of design features, as well as their cost–effectiveness. In other words, since buildings are built on a finite budget, it would be extremely useful to be able to determine how, given that fixed, finite budget, a designer could maximize a building’s environmentally sustainable performance. This means using sustainability indicators, and the relationships among them and the building systems, to determine an optimal scheme of investing this limited budget in sustainable design features. In this article, the focus is on making trade-off decisions related to material selection and use in the context of reducing the carbon footprint of a building.

▪ Decision making & system dynamics

The premise of this research is that current methods of measuring, predicting and optimizing the sustainable performance of a building can be improved through the use of new, integrated decision-making tools. The current systems for designing buildings rely on a number of disjointed analyses to determine whether discrete requirements have been met by various systems (e.g., heating, ventilation and air conditioning, plumbing and lighting) or design features (e.g., landscaping, renewable energy generation and parking). Often, these specialized analyses optimize for a given performance in terms of the properties addressed in the analysis. These disconnected optimizations may miss system-level interactions that the proposed model is intended to capture. The system-level building design method under consideration in this article allows these currently unconnected analyses to be integrated and optimized in a systemic fashion, based on the carbon footprint of the building. This optimization could also be based on either the sustainability indicators or building rating systems discussed above.

Improving the sustainability performance of buildings is a difficult problem, owing to the difficulty in assessing the relative improvements in sustainability of one decision versus another [5], and the problem of trying to predict future building performance during the design phase, when the ability to influence project cost is greatest, but when much of the other information about the final design is unavailable [6]. Another challenge is that of providing the designer with early feedback on the building’s future performance. The building industry is also very fragmented [17], and the large number of systems and components that make up a building (e.g., structural, mechanical and electrical systems) and the complex interactions among these components make representation and simulation of a building in a single, integrated model very difficult [18]. Many of the sub-systems are designed, constructed, operated and administered by separate entities that may or may not interact and share information [19,20]. Interactions of building occupants with the various building systems add yet another layer of complexity to the building system [21]. Currently, most building models address individual systems [22], including modeling programs that address a single aspect of building performance, such as Green Building Studio™ for energy analysis, FLUENT^® for air flow modeling, CONTAM for indoor air quality, BEES^® for building product selection, SunCast for passive solar design, DAYSIM for day lighting analysis, ATHENA™ for environmental impact of structural materials, Watergy™ for water use and related energy use, and RETScreen^® for renewable energy and passive solar design. The US Department of Energy lists 345 building software tools for energy analysis, alone [101]. A major goal of the building simulation community is the integration of these types of models into a single building model and the better incorporation of the results of these models into decision making [23], which is essential to designing a truly environmentally sustainable building.

System dynamics (SD) is one tool with potential to support building sustainability decisions. It is a modeling method developed from systems thinking ideas [24]. Systems thinking is a holistic approach to problem solving based on the general systems theory [25], a philosophy of science and engineering based on the idea of combining the knowledge gained through analysis and the understanding gained through synthesis to address the root causes of problems [26]. The SD method applies and extends systems thinking concepts to construct computer simulation models, consisting of “an interlocking set of differential algebraic equations developed from a broad spectrum of relevant measured and experiential data” [27], and represented by a diagram, to examine system structure and the effects of altering key variables over time [102]. Figure 1 illustrates this concept of diagramming system structure with three variable types:

	▪ State or stock variables, which track a dependent variable;
	▪ Flow variables, which regulate the changes in the stock variables;
	▪ Auxiliary variables, which modify the flow variables.

Feedback loops can be represented by the relationship between a stock variable and an associated flow variable. While systems thinking is a way of thinking about problems [28], SD uses systems thinking principles to develop models to represent the problems.

Although the SD method has been used in a wide variety of applications, its use in building design has been limited [29]. The SD modeling method is applicable to building system simulation since it is ideal for situations where the system to be modeled is extremely complex or highly dynamic (in time and/or in space) [30]. Its focus is on the basic structure of the system, allowing for incorporation of ‘soft’ factors that can help to capture human behavior of the building occupants [26], and for other highly uncertain variables to be usefully included. Incorporation of these soft factors will be important in a building system model to address the occupants’ perceptions of and reactions to changes in a building’s design and operation. These reactions will determine, to some extent, how building occupants behave (e.g., whether they keep the windows closed when the air conditioning is running). SD allows quantification of system behavior without necessarily requiring a high level of numerical accuracy in the model, as long as the model structure is well-defined [24]. The SD method facilitates the search for leverage points through the use of sensitivity analyses [31], and allows simulation experiments to be conducted on virtual buildings or retrofits [30]. The main difficulties encountered in applying the SD method arise from difficulties in identifying truly dynamic feedback relationships within buildings’ sub-systems.

▪ Building information modeling

The Associated General Contractors of America (AGC) defines a BIM model, as ‘a data-rich, object-oriented, intelligent and parametric digital representation of the facility, from which views and data appropriate to various users’ needs can be extracted and analyzed to generate information that can be used to make decisions and improve the process of delivering the facility” [32]. BIM, however, is more than just the digital representation. It actually represents a shift in the traditional process of building delivery. This process shift is also known as ‘integrated project delivery’ or ‘integrated practice’ [33] and is integral to the current industry trend towards fully integrated and automated project processes [34]. integrated project delivery can be understood as an application of systems thinking to the building industry. The motivation behind implementing integrated project delivery is to integrate the different project participants and processes to make optimal decisions for the project. This optimization is accomplished by envisioning the project as an integrated system composed of various sub-systems, each of which is the responsibility of a different project participant.

The National Building Information Modeling Standard is being developed as a national standard for the use of BIM in building design and construction [35], the US General Services Administration has begun to require BIM on its building projects [36], and the AGC has developed ‘The Contractor’s Guide to Using BIM’ [32]. The US states of Wisconsin and Texas now require BIM on most state projects [37,38]. Efforts to standardize the practice of BIM are also underway within the US armed forces, the European Commission, Scandinavian countries and Singapore [39]. The National Building Information Modeling Standard discusses the role of interoperability, or “seamless data exchange and sharing at the software level among diverse applications, each of which may have its own internal data structure” as being essential in the building delivery process [35]. Importantly, the use of BIM information to “make decisions and improve the process of delivering the facility” cited in AGC’s definition [32] is still limited by traditional divisions of labor and a lack of information sharing among various project participants. Design decisions are often not addressed in a way that considers the project as a larger integrated system. The decisions made using BIM information are, thus, not optimized across the entire project. The research described in this article seeks to implement an interoperable framework for sustainable design and the BIM process, by integrating decision making with BIM design, visualization and analysis.

Integration of BIM with an SD/decision-making tool

One means of using BIM to improve decision making in building design is simply to reduce the amount of work involved in evaluating multiple options early in the design process. Currently, it is possible to export BIM models directly and completely, with minimum data loss, to third party software programs such as Green Building Studio or Ecotect for specialized analysis of energy use or day lighting, or to various structural analysis software programs. Revit^® mechanical, electrial and plumbing and the IES Virtual Environment program are capable of a bidirectional interaction [40]. These program links provide information for use in decision making, but do not provide any framework for actually making the decisions or for optimizing the design features used in these independent third-party analyses. More specifically, they also have very limited capacity for making trade-off analyses among components of multiple different building sub-systems. In addition, these programs do not make the BIM model dynamic. When information is passed from the BIM model, the data is used to construct a model in the analysis software, run the analysis and return results of the modeled performance of the design. If any optimization takes place in of these analysis programs, any modifications are not translated automatically into the BIM model. For example, Green Building Studio has a ‘design advisor’ function that automatically makes suggestions to improve the design, but these are not dynamically integrated into the BIM model [40]. Making a BIM model dynamic is a relatively straightforward process using the BIM program application programming interface (API) for transferring data in two directions from and to a BIM model. This technique is used successfully to couple Revit Structure with structural analysis and design software programs such as ETABS, RAM and RISA, where, for example beam sizes can be updated for example. Creating a two-way dynamic BIM–SD link could lead to more sustainable and cost-effective buildings, and more streamlined, efficient and thus, cost-effective building design processes. These benefits can be achieved by creating a framework to make optimal decisions regarding sustainable design features early in the design process.

The methodology employed in this study was to move data between a SD decision-making software model and a BIM software model. Once data has been transferred from the BIM model to the SD model, it can be used for making decisions regarding the building’s design. Using appropriate sustainability indicators, these decisions may be optimized within the decision-making software to provide for a sustainable building. These decisions may then be actualized by returning the optimized building data to the BIM model, and modifying BIM components as appropriate. The model currently is focused on sustainability indicators, specifically on material selection decisions, and not on any structural or mechanical, electrial and plumbing design considerations.

Specifically, this study addresses the creation of a direct link between the AnyLogic™ [41] SD modeling program and AutoDesk’s Revit Architecture BIM modeling software [103]. This link would be used to automatically populate variables in the AnyLogic™ model with data from the Revit model, and to then update the Revit model by using results from the SD simulation to modify objects in the BIM model directly.

The first step was to identify the required inputs to the SD program that are based on the building geometry or other building data (e.g., room occupant load, component cost, airflow and lighting specifications). This process was initiated by identifying the highest level sustainability metrics (e.g., energy, water and material use) and then determining which building components (e.g., heating, ventilation and air conditioning system, plumbing system and façade) contributed to each metric. Once inputs were identified, the next step was to incorporate these variables as parameters in the BIM model. The prototype link between the BIM and SD programs was accomplished through Java and Visual Basic (VB) applications written to take advantage of the programs’ APIs, as illustrated in Figure 2.

The system described here is designed to allow data to flow in both directions. First, data may move from the BIM model to the SD decision-making model, to specify initial conditions in the SD model, and provide a realistic basis for decision making. Second, the data may move from the SD model back to the BIM model, allowing objects in the BIM model to be updated automatically, based on decisions made using the SD model.

In moving from the BIM model to the SD model, one of two procedures may be followed. In the first procedure, data may be output from the BIM model, via its API into a VB or C# external application. The next step, using a freely available software module, is to translate this VB or C# data into a Java format. Java applets are run as plug-ins to the AnyLogic software program, and the data may be incorporated into the SD model through these applets. In the second procedure, data may be exported from the BIM model in some standardized format, such as Industry Foundation Class files, CimSteel Integration Standards files, or green building eXtensible Markup Language files. These files may then be directly translated to Java files using the translator module, which may be linked via the Java capabilities of AnyLogic to the decision-making model.

Once BIM data has been integrated into the decision-making SD model, it may be used to populate the SD model with appropriate building data. The data is then combined with the structured decision model. Appropriate building sustainability indicators are then used as metrics to determine optimal combinations of building characteristics through sensitivity analyses completed within the SD software program. Other decision-making software and models may be used, but the SD modeling technique carries the added benefit of allowing users to choose sustainability indicators in a fashion important to the user. This flexibility could be used to choose indicators that simulate a LEED rating system, or any other rating system the designer chooses to investigate (e.g., BREEAM, Green Globes and Living Building Challenge). In addition, the various indicators may be weighted by the user to accommodate the priorities of the designer or owner or to reflect the local environmental and climatic conditions in which the building is to be sited.

Once an optimal arrangement of the project variables has been reached within the decision-making model, these decisions can be implemented in the BIM model by following the data-transfer steps in reverse order. Data representing the building characteristics may be exported from the SD model, via its Java applets. This data may then be translated via the third-party Java-VB translator module into a VB file, which is then read in through the BIM program’s API, allowing the changes made to building components in the decision-making model during the optimization process to be reflected in the geometry and parameters of the objects within the BIM model.

▪ Wall thickness optimization; an example

External decision-making models can be used to weigh any number of trade-off scenarios with respect to different performance metrics in the building design. Many parameters of a building design can influence several different sustainable aspects of a building such as the energy consumed, water efficiency, material used and so forth. For example, in terms of the contribution to the building’s carbon footprint resulting from building material selection issues, the amount of concrete used in the structure will be involved in a trade-off between the embodied energy used to construct the building and energy required to heat and cool the building in operation. More concrete in the structure will result in more material that will need to be extracted, manufactured and transported to the site but will also reduce the amount of space that needs to be heated and cooled as well as provide thermal mass and some insulation. The decision-making model can then optimize the thickness of the walls or floors to get optimal performance from the concrete with respect to the building sustainability metric, embodied versus operational energy use in this example, as determined by the designer. A conceptual representation of the relationships that make up this optimization process is shown in Figure 3. Figure 4 shows how the thickness of an exterior wall can be changed inside the BIM model based on a decision-making model outside of the BIM program. It should be noted the structural aspects of the wall are ignored here. On the left, the exterior walls are highlighted and their properties are displayed, including the thickness of each component. The code sample in the center of the figure is used by the API to change the wall thickness based on external input from a decision-making model. After the change is made the properties of the wall are shown on the right of the figure, highlighting the increased thickness of the wall. The BIM model then updates its database based on the change.

▪ Optimization of recycled content & regional material LEED credits; an example

The second example illustrating the use of BIM models for quantifying building material trade-offs is an optimization of the LEED credits for recycled content and regional material. This example uses concrete as an example of a material for which this type of optimization must be made. The relationships affecting the trade-offs between these two indicators are illustrated in Figure 5. Both regional materials and recycled materials impact the building carbon footprint. Both also affect the LEED credit total in this example. In addition, both affect the total cost of the concrete in the building. In this highly simplified example, we have two variables:

	▪ Regional materials concrete amount;
	▪ Building recycled content percentage.

Those two variables affect three main outcomes: (i) building Carbon footprint; (ii) concrete cost; and (iii) LEED credit total.

If these values can be extracted from the BIM model, it will be possible to make an informed decision, after which the process described under the previous example can be used to modify the BIM model according to the results of the optimizationxprocess.

Figure 6 shows an example of the parameters required to calculate the LEED credit for regionally produced materials. The credit requires that the distance from the project that the material was harvested and manufactured is known as well as the percentage of the material that was harvested and manufactured within a 500 mile distance. On the left is an API code segment that is used to add parameters to different components in a BIM model. This can be done for all families (e.g., walls, doors and windows) in a model. On the right-hand side, the properties of a door are displayed showing the newly added parameters under ‘green building properties’ required to calculate the credit.

Outcomes from using BIM for sustainable material selection

Much of the data that would be used by a decision-making model to aid in design of a building is also data that is required to be calculated in order to apply for LEED credits. While LEED has its shortcomings, one of its greatest contributions has been to encourage designers to consider and put more focus on the sustainability of the building and its different systems. LEED identifies the most important factors for designers to consider with respect to different sustainability metrics of a building and therefore the information needed in any decision-making model could also be used to calculate LEED points for a building during the design. One immediate benefit from this would be to reduce the time and effort needed to calculate and apply for LEED credits, which can be a time consuming activity for any design firm. When designers reuse building elements from project to project, they could add parameters to these elements in their BIM models to store the data about that product that would be needed to perform LEED point calculations. Designers could also link third-party models, such as energy models, other types of decision models and LEED point calculators, directly and dynamically to a BIM model. This would make the LEED accreditation process more efficient for designers. A more important long term benefit from this would be to change the design approach such that the sustainability of the building is considered earlier in the design process, in much the same way that BIM has begun to change the building industry towards considering potential problems that were previously not addressed until well after the design was compete and construction had begun (e.g., ‘clash’ analysis).

If the designer can see the LEED credits being accumulated during design, that information can be used to drive decisions as well as other aspects of the building such as constructability or cost. This will allow designers to actually design the building to meet sustainability goals, whether specified by LEED or not. The current LEED points are not weighted perfectly based on the overall importance each point has with respect to the building’s sustainability, but as LEED evolves this will improve. LEED specifies many benchmarks that are required to achieve certain points and although a designer may strive to achieve the specific point they often cannot be sure how close they are until the design is complete. This may result in several points being missed narrowly, which may have been achieved with minor design changes or several requirements being greatly exceeded to the point that the added benefit from the design contributes minimally to the overall building sustainability. This will become more useful as LEED evolves a point scale that better reflects the importance of each metric, possibly to the extent that it can become the driving decision-making model.

A number of challenges related to sustainable building decision making are addressed by the outcomes of this research study. First, the decision-making model linked to the BIM model will allow designers to focus on the greatest environmental impacts and concerns, and to choose LEED points to be pursued on a basis other than lowest cost. Next, this decision-making model will allow designers to capture the interactions of various aspects of building material performance, their impacts on each other and the trade-offs among them. The decision-making model will be a step towards the integration of incommensurable environmental performance indicators. Use of the dynamic link to the decision-making model will allow environmental ratings to be based directly on an accurate BIM model, and will help to more closely approximate actual performance of the building. This will minimize the inconsistencies resulting from multiple independent building models. Finally, the linked BIM and decision-making models will give designers access to real-time carbon footprint analyses during the design, allowing sustainability to drive design decisions, and to be better optimized with project cost.

System dynamics decision making is not designed to replace building analysis software, such as structural analysis, day lighting analysis, energy analysis and so on. Instead, it is intended that the SD model would use the outputs of these various analyses to determine leverage points in the sustainable design of the building. In other words, the SD model will help set priorities for sustainable design and assist engineers/architects in performing trade-off analyses among various components of a building system. To illustrate this, in the section on wall thickness above, the SD model would not be used to determine the strength of the concrete, the structural requirements for load-bearing walls, or the optimal wall area for day lighting. Rather, the SD model will help the designer to choose where to focus efforts. If the SD analysis reveals that the carbon footprint for the building is impacted more strongly by thermal mass considerations than by embodied energy considerations, then this gives the designer important information with which to continue the design process.

Conclusion

This article has developed a workflow for the linking of data within a BIM model to a SD decision-making model. This process can be extended to other building models (e.g., the energy model), bringing more complete data to the decision model. The article has also illustrated this work flow through two building material problems, based on criteria present in most sustainable building rating systems. The first example dealt with the question of embodied energy versus thermal mass considerations in concrete walls. The second dealt with the issues involved in using recycled content or local materials in the concrete walls. The intent of the model developed in this paper is to determine leverage points in the building system where the greatest amount of impact can be brought to bear on the building design.

While unquestionably having a positive impact in the building industry, LEED (and similarly many of the other rating systems) faces three important challenges at this time. First, while the theory would have building owners and designers addressing LEED points that represent the greatest environmental impact or the most importance to them, the reality of the situation may be that many designers address the lowest-cost points first, in an attempt to achieve LEED certification for the lowest total cost, with less regard for actual environmental benefit [42]. Second, further difficulties arise from the fact that LEED addresses a large number of different and conflicting aspects of the sustainability performance of buildings, creating the possibility that one aspect will be neglected in favor of others, while a LEED rating can imply, to those not involved in the certification process, a high level of performance in all aspects of sustainability [43]. Finally, LEED has historically based its ratings on predicted performance, often resulting in a gap between the predicted (rated) performance and the actual performance of the constructed building [42].

Improvements in these three challenge areas lie in the development of decision-making tools for sustainable building design. Decision tools developed using appropriate sustainability indicators will allow LEED points to be chosen on a basis other than lowest cost, and will assist in integrating incommensurable performance indicators. Basing the LEED analysis during design directly on an accurate BIM model will lead to predictions that should more closely approximate actual performance and will minimize inconsistencies that can occur when creating multiple models of the same building.

Incorporation of sustainable material considerations and rating systems into a BIM model will not only help to facilitate these improvements, but will also allow design professionals to improve their processes, which may help to reduce the costs of sustainable, LEED-certified projects. Reduced costs for sustainable buildings will encourage more owners to build such buildings, while at the same time the decision model linked to the BIM model will make these buildings more optimally sustainable than current LEED projects. The software application described above for linking a BIM model with an SD decision model will be the basis for alternative evaluation and optimization within the decision-making tool.

The examples presented in this article are intended to show different potential applications of the decision analysis model and its integration with a BIM model. Using only these two ‘simple’ examples, we can begin to illustrate the complications involved in implementing a sustainability analysis of building materials. More work is clearly needed to extend these models to more aspects of sustainable material selection and design.

Future perspective

As the decision model evolves, it will be possible to incorporate more advanced sustainability criteria in the decision model, moving first to more stringent rating systems, such as the Living Building Challenge, and eventually to develop a ‘deeper’ sustainability rating system. These deeper sustainability decision criteria will be available to the BIM model through the same interactive process described above. In the long term, the process described in this paper could be used to design a building with a real-time sustainability feedback to the designer. This would allow the designer to make real-time decisions based on the carbon footprint of the building as it is designed.

Building information modeling (BIM)

An object-oriented digital, parametric representation of a facility, including 3D views and extractable data that can be customized by users. BIM can also be used to refer to a process, similar to integrated project delivery, where all project participants work together for the benefit of the overall project.

System dynamics (SD)

A modeling method that allows a system (in this case a building) to be represented as a feedback system using a set of differential algebraic equations representing the various sub-systems of the building and represented by a diagram consisting of stock, flow, and auxiliary variables. An SD model allows examination of the long-term behavior of complex systems.

Systems thinking

A holistic philosophy of science and engineering based on the principle that to fully understand a system, multiple, small individual pieces must be studied to understand the fundamental causes of the model’s behavior, and the macroscopic behavior of the system must be studied to understand the underlying behavior of the system.

Application programming interface

A specific set of rules that allows multiple software programs to communicate with each other. It is an interface between the various programs and uses its rules to govern the interactions among them.

Executive summary

Challenges

	▪ Sustainability-related challenges are increasingly important in building design. ▪ Buildings consume the majority of electric power and natural gas in the US. ▪ Buildings use a significant portion of US water. ▪ Buildings account for the majority of US waste output. ▪ Buildings contribute over one third of US greenhouse gas emissions. ▪ Buildings consume approximately 40% of the world’s raw materials.
	▪ It is difficult to assess the relative improvements in sustainability of one decision versus another.
	▪ Key decisions must be made in the design phase before the final design is available.
	▪ More owners are now requiring or requesting 3rd party certification of their projects than in the past.
	▪ Designers need to know which combination of rating credits provides the optimal choice of design variables for the building’s sustainability, while keeping their project within its budgetary constraints.

Decision Making

▪ System Dynamics (SD) is a feedback-based decision-making methodology for modeling complex non-linear systems that focuses on the interrelationships among key variables. ▪ SD modeling applies and extends systems thinking concepts to construct computer simulation models. ▪ SD models are composed of a set of differential algebraic equations developed from relevant measured and experiential data. ▪ SD models allow the user to search for leverage points in the model through sensitivity analyses. ▪ SD modeling allows a user to perform simulation experiments on virtual buildings or retrofits before physical construction begins. ▪ The main challenge in using SD to model a building as a system arises from difficulties in identifying dynamic feedback relationships within buildings’ sub-systems.

Building Information Modeling (BIM)

	▪ BIM is both a technology and a process.
	▪ BIM is an object-oriented, digital, parametric representation of a facility.
	▪ BIM also refers to a process where all project participants work together for the benefit of the overall project.
	▪ BIM allows the various project participants to make optimal decisions for the project by integrating the project processes.

BIM –SD Integration

	▪ A dynamic link between BIM and SD models using API links has been established.
	▪ The link allows for the dynamic transfer of data from a BIM model to a SD model.
	▪ The building design factors are optimized using the SD model.
	▪ The data is then returned from the SD model to update the BIM model.

Examples

	▪ An illustrative example of wall thickness optimization is presented.
	▪ A second illustrative example of recycled content and regional material optimization is also presented.
	▪ Integration of BIM and SD models can help optimize LEED design analysis.

Outcomes

	▪ BIM and SD model integration can help move consideration of sustainability issues earlier in the building design process.
	▪ The integration of the decision-making model into the building design can help make 3rd party certification systems more efficient and impactful.
	▪ The SD model allows a realistic analysis of the interactions and effects of building material decisions.
	▪ The SD model is intended to assist in making trade-offs among different options in building design and to help set design priorities.

Financial & competing interests disclosure

A pending patent, US 2009/0292509 A1 Method and System for Assessing Response of a Building System to an Extreme Event by authors Benjamin P Thompson and Lawrence C Bank, is related this article. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript.