Automation in Construction 18 (2009) 153–163

Contents lists available at ScienceDirect

Automation in Construction
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a u t c o n

Building information model based energy/exergy performance assessment in early

design stages
Arno Schlueter ⁎, Frank Thesseling
Institute of Building Technologies, Building Systems Group, ETH Zürich, Switzerland

a r t i c l e i n f o a b s t r a c t

Article history: Due to the rising awareness of climate change and resulting building regulations worldwide, building
Accepted 12 July 2008 designers increasingly have to consider the energy performance of their building designs. Currently,
performance simulation is mostly executed after the design stage and thus not integrated into design
Keywords: decision-making. In order to evaluate the dependencies of performance criteria on form, material and
Building information model
technical systems, building performance assessment has to be seamlessly integrated into the design process.
Building performance
Energy analysis
In this approach, the capability of building information models to store multi-disciplinary information is
Exergy analysis utilized to access parameters necessary for performance calculations. In addition to the calculation of energy
Design support balances, the concept of exergy is used to evaluate the quality of energy sources, resulting in a higher
Parametric design flexibility of measures to optimize a building design. A prototypical tool integrated into a building
information modelling software is described, enabling instantaneous energy and exergy calculations and the
graphical visualisation of the resulting performance indices.
© 2008 Elsevier B.V. All rights reserved.

1. Introduction lack of integration into the design leads to extensive modifications

The operation of buildings accounts for 40% of global CO2
emissions [1]. These emissions are directly linked to the energy a
building consumes in order to maintain its usability. Heating or
cooling energy has to be supplied to maintain a certain comfort level.
Additional power is needed to operate lighting, appliances and
building service systems.
Due to the increased awareness of energy consumption and related
CO2 emissions, building regulations such as the European Buildings
Directive [1] in Europe, Minergie in Switzerland [2], or programs such
as LEED [3] in the USA have been established over the last years.
Architects and planners are increasingly forced to consider energy
consumption and the environmental impact of their building designs.
Building performance is defined differently among professionals and
researchers. In the context of this paper, we define building
performance as related to energy consumption, the most important
issue concerning CO2 emissions.
It is widely acclaimed that the most important design decisions
concerning building sustainability have to be made in the early design
stages—that is by the architect or building designer. In common
architectural practice however, performance analysis to support
design decision-making is only used for the few buildings facing
engineering challenges or explicitly focussing on sustainability. The
⁎ Corresponding author. Tel.: +41 44 633 3618; fax: +41 44 633 1047.
E-mail addresses: schlueter@arch.ethz.ch (A. Schlueter), thesseling@arch.ethz.ch
(F. Thesseling).
afterwards to meet performance criteria.
This practice also leads to buildings that might be sustainable
considering their energy consumption but not in architectural aspects.
Current developments in computer applications in architecture
have led to impressive results, for example in the fields of advanced
geometry and computer-aided production methods. Both fields
directly address the domain of the architect and are used for
progression in architectural design. Computers are used to enhance
the traditional toolsets. It seems obvious to extend these methods to
implement building performance aspects. However, no tools exist to
seamlessly integrate performance assessment into the design process
or to support the design and decision-making of the architect or
building designer. Holistic performance assessment is not considered
in any kind of computer-aided architectural design (CAAD) environ-
ment that architects use.
This is due to several reasons. In the traditional architectural
workflow, performance assessment is mostly done subsequent to the
architects design. It is done by the expert, in most cases the engineer. A
lot of expert software exists for every type of simulation of specific and
overall performance of buildings and building components. Available
simulation tools are therefore aimed at the expert and make explicit
expert knowledge necessary to input the data needed, run the
simulations and interpret the results. In the early design stages, this
data is often not available. Architects are mostly non-experts
considering performance simulation. As generalists, they do not
know about precisely every parameter necessary to run an expert
simulation. But they know about form, materials and preferred

0926-5805/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.autcon.2008.07.003
154 A. Schlueter, F. Thesseling / Automation in Construction 18 (2009) 153–163
technical systems of their building design. This information, if defined
during the design, can be used as input to evaluate building
performance. The amount of information and the complexity of its
dependencies makes computational methods necessary [4].
Building performance is measured by using mathematical calcula-
tion models. Based on the task, these range from simplified statistical
models to specialized physical simulations. To support decision-
making, the highest level of precision is not necessary. Performance
assessment for the early design stages has to show the tendencies and,
most important, dependencies of decisions.
In our approach we focus on the energy performance of buildings.
We expand the common energy analysis by considering the thermo-
economic concept of exergy. Exergy analysis takes the quality of
energy into account. It enables balancing between building form,
materials and technical systems and makes a holistic view of the
building possible. A building information model is used to store
necessary data and access parameters during the design process.
Energy and exergy calculations are directly integrated into the
building information modelling editor by a prototypical tool, the
Design Performance Viewer (DPV). The tool enables fast estimation of
energy and exergy performance of the specific design, facilitating
necessary parameter input by using non-expert decision criteria.
Results are visualized using selected key performance indices and
graphical visualization.
2. Building information modelling
To be able to consider building performance in the early design
stages, access to all information defining a building such as its form,
materialization and technical systems is necessary. Common, docu-
ment-based CAD planning environments do not support this
integrated view of a building. In machine engineering, the concept
of “semantic datamodels” [5] was established in the 1970s, connecting
logical and physical information. For the needs of the building
industry, this concept was adapted for generic “building description
systems” [6], later called “building product models” [7]. Since 2002,
the term of “building information models” [8] has been widespread.
Building information models enable storing multi-disciplinary infor-
mation within one virtual building representation. A building
information models is a “richer repository” [7] than a set of drawings,
since it has the ability to store different types of information.
These types of information include geometric, semantic and
topological information. Geometric information directly relates to
the building form in three dimensions. Semantic information
describes the properties of components such as u-values of walls.
Topological information captures the dependencies of components.
Or, as Eastman [7] states, a building information model contains the
“form, behaviour and relations of parts and assemblies”.
Building information models have to be distinguished between
proprietary models established by software companies and open,
non-proprietary models such as the Industry Foundation Classes (IFC),
developed by the International Alliance for Interoperability (IAI) [9].
The IFC serve as an “intelligent, comprehensive and universal data
model of buildings” [10]. The IFC data model is already used by a
number of CAD tools as an export and import option [11]. Several case
studies have shown its applicability in the design process [12,13]. All
efforts aim at the IFC becoming a standard for information exchange in
the building industry. An extensive description of generic, non-
proprietary models such as the IFC can be found in [7].
Besides the generic models, software companies have developed
their own internal models to feed their CAD and BIM software. These
models differ in their structure and capabilities to establish a
consistent model. In the building industry, the concept of building
information modelling has gained increasing acceptance over the last
years. Recent developments of BIM editors are suitable to serve as fully
functional planning environments for architects. In our approach we
utilize the model of the proprietary modelling editor “Revit” from
AUTODESK [14]. The software enables easy establishment and editing
of an extendable building information model. Its model database can
be accessed and extended by an application programming interface
(API). An IFC-export and import is possible to support data exchange
between different software packages. A model created in the software
can thus be exported as a generic IFC-compliant building information
model.
One of the goals of building information modelling is to make the
cooperation between stakeholders in the building process more
efficient. This is achieved by storing relevant design information of
every step in the design process. This ability of the model to serve as a
interdependent, multi-disciplinary data repository make new
approaches on integrating performance analysis into design possible.
Parameters defined during the design process can be accessed and
edited during the design process and utilized for the performance
calculation.
Even though switching to BIM-based design environments poses
major challenges to architectural offices, recent market surveys show
that 48% of the architectural offices in the US already use methods of
building information modelling [15]. Today, the biggest obstacle for
architects to adopt BIM methods is the tentative use of BIM by other
industry partners such as engineering firms. To be able to exploit the
benefits of BIM, the adoption of BIM in the early stages of design is
crucial. As one of the first official institutions, the U.S. General Service
Administration requires BIM for the submission of mayor projects for
final concept approval. With the use of BIM, the GSA encourages
“accurate energy estimates in the design process” [16], strengthening
the adoption of BIM from the early design stages on.
3. Performance analysis
Many tools for computational performance analysis have been
developed, yet their application and thus their impact on the design
process has been “rather limited” [4]. Tools for performance analysis
can be divided into two different groups, either based on a statistic
calculation model or a physical calculation model.
3.1. Physical calculation models
Physical calculation models make the precise calculation of
detailed tasks as well as overall energy consumption possible. From
zone loads, daylighting and solar to multizone airflow, highly precise
calculations for every possible engineering task are available. Many
expert tools use physical calculation models for the calculation such as
TRNSYS [17], IES Virtual Environment [18] or EnergyPlus [19]. A
comprehensive contrasting survey of software using physical models
can be found in [20]. Necessary information input to run expert
simulations is extensive, so is knowledge to perform and interpret the
simulation results. As one of the few tools with focus on an more
graphical interface and less effort to conduct a performance analysis,
Ecotect [21] is targeted at the architectural design process.
All external tools require the input of geometry of the design to
define the simulation model. This is mostly done by either importing
the geometry or manually rebuilding it in each software. Importing
and exporting of building geometry is error-prone and tedious,
especially as geometry models established in CAD-software are
often not suitable as simulation models. The simulation results and
possible conclusions remain in the simulation software, a feedback
into the design software is not possible. Changes in design due to
performance criteria have to be done manually in the design software,
the model has to be exported and simulated again. These steps have to
be repeated after every change in design. As long as a fully functional
IFC-based data exchange is not yet available, external tools utilizing a
physical calculation model for performance assessment only apply for
critical design tasks.
 A. Schlueter, F. Thesseling / Automation in Construction 18 (2009) 153–163
3.2. Statistic calculation models
Statistic calculation models are simplified models for the estima-
tion of total the energy demand, heating or lighting energy demand.
Regulations such as the German Energy Savings Regulation EnEV [22]
or the Swiss Minergie [2] use statistical calculation models for
mandatory application in the building process. Instead of calculating
the physical processes within the building, empirically found factors
are applied. Due to their more abstract nature, these models deliver
rather rough estimates, which is accepted to fulfil the regulations.
Compared to the input necessary for physical models, much less
detailed information for the calculations is needed, thus facilitating
the parameter input. The coarser resolution also speeds up computa-
tion of the results. As physical models take up to minutes and hours to
compute, statistic models deliver results in less than seconds. Even if
the results are not highly precise, the delivered performance indices
enable judgement on building performance. Typical statistic calcula-
tion tools are spreadsheet-based or even web based applications such
as IdeaXP [23].
As the goal of this approach was to realize immediate performance
assessment, a statistic calculation model for the performance
calculations was used. In order for the architect to apply in the design
process, the effort for input and editing of parameters was tried to be
kept as small as possible. This work focuses on the integration of
energy and exergy analysis into the architectural design process.
Therefore, the utilized energy and exergy calculation models are only
described briefly to display which parameters are implemented and
which results are calculated. The mathematical model applied to
estimate the energy gains and losses is derived from the German
Energy Conservation Regulation EnEV [21]. This regulation is
mandatory for the calculation of the energy demand of new and
existing buildings in Germany. The exergy calculation model devel-
oped by Schmidt [27] and applied in this approach (see Section 5.) also
uses this mathematical model for the energy calculations.
4. Energy analysis
To be operated to maintain user comfort and functionality, a
building needs a defined amount of energy that has to be supplied. In
order to estimate the amount of energy that is needed, an energy
balance has to be set up. The demand side is calculated, cumulating
energy losses such as transmission and ventilation heat losses of the
building envelope. These losses can be fully or partly compensated by
the energy gains. Different sources of energy gains can be utilized.
Internal energy gains caused by appliances and users as well as solar
gains through openings diminish the amount of heating energy that
has to be supplied. Additional energy input is needed for lighting,
ventilation and for the operation of building systems. Deducting gains
from the overall losses results in the overall energy demand which has
to be delivered (Fig. 1).
Fig. 1. Implemented energy model.
155
Six key performance indices are calculated to display the energy
performance of the building. These indices serve as an estimate to
show the energy performance of the specific building design at the
maximum temperature difference on a specified location. Most of the
necessary input parameters such as geometry and masses, component
properties and dependencies are automatically taken from the
building information model. The parameters for the design of the
heating system are defined in the tool interface as described in 6.2.3.
To simplify user input, some parameters of lesser impact on the overall
performance are defined as static parameters. Contrasting to the
underlying regulation, the heating energy demand is calculated for
steady state conditions. The focus is on the energy demand at a
maximum temperature necessary to layout the heating system, not on
the annual energy demand the regulation aims for. Also contrasting to
the regulation, the energy demand for domestic hot water is
neglected. Following key performance indices are calculated:
4.1. Transmission heat losses of the envelope
All information about the geometry of windows, walls, roofs and
floors is taken directly from the building model as well as the specific
u-values of wall and window objects. Indoor (θi) and outdoor (θe)
temperatures to layout the heating systems are defined by the location
of the building. The total transmission heat loss is the sum of the heat
losses of all envelope surfaces. Heat bridges are not considered.
ΦT ¼ ∑ Fx;i  Ui  Ai   ðθi −θe Þ


½WŠ ð1Þ
The temperature correlation factor Fx,i enables using the same design
temperature difference for the calculation of parts facing different
environmental conditions. This factor is set according to the regulations
to 1.0 for exterior walls and roofs and to 0.6 for walls and floors facing the
ground [22]. Winter gardens, attics and unheated rooms are neglected.
4.2. Ventilation heat losses
A simplified formula captures the ventilation heat losses. The
overall volume V is taken from the building model and multiplied by
the air exchange rate nd. The specific heat capacity of air (0.34 W h/m3
K) is taken into account.
ΦV ¼ ð0:34  nd  V Þ  ðθi −θe Þ ½WŠ ð2Þ
4.3. Solar heat gains through windows
The maximum amount of solar radiation is defined by the
geographic location of the building and orientation of the opening.
In dependency to opening surfaces Aw,i, the solar radiation heating up
the building inside is calculated for every window. The g-value (total
solar transmittance) of the windows defines the energy input of solar
radiation passing through a specific glass.
 
ΦS ¼ ∑ Is;j  Aw;i  g L;i  F F;i F W;i F C;i F S;i ½WŠ ð3Þ
Four correction factors consider possible shading by shading
devices (FC), shading by surrounding buildings (FS), non-orthogonal
solar radiation (FW) and window framing (FF). To simplify parameter
input, these factors are set to standard values according to the
regulations, resulting in a cumulated correction factor of 0.567.
4.4. Internal heat gains
Internal heat gains caused by humans are stored as static
parameter within the occupancy parameter noo of the room. In this
approach, the heat gain per person Φ″i,o was set to a mean values of
156 A. Schlueter, F. Thesseling / Automation in Construction 18 (2009) 153–163
80 W. It is multiplied by the statistic number of occupants. In order to
simplify parameter input, the specific heat gain by electrical
appliances Φ″i,e was set as static value for all rooms. It is multiplied
with the room area. To capture different building types, this static
value can be adapted.
Φi;e ¼ ΦWi;e  An ½WŠ
Φi;o ¼ ΦWi;o  noo ½WŠ
4.5. Specific lighting power/lighting power
The calculation of the specific lighting power is taken from the
Swiss regulation SIA 380/4 [24]. The necessary illuminance Evm of
each room defines the specific lighting power. In relation to the type of
artificial lighting and its specific light efficiency ηV, the specific
lighting power is calculated. The calculation includes factors of usage
and aging ρv, for the efficiency of the lamp ηLo and a specific room
characteristics ηR such as reflectivity and room geometry. These
factors are set to standard values dependent on the type of artificial
lighting in accordance to the regulations.
ðEvm  pv Þ
W=m2
 
pLi ¼
ð6Þ
ηV  ηLo  ηR
Resulting, the necessary lighting power is
Φi;L ¼ pLi  An ½WŠ
The specific lighting power is also added to the internal heat gains.
Additionally, the auxiliary electrical energy for ventilation based on
the internal volume is calculated and added to the total electrical
energy consumption.
4.6. Resulting heat demand
All heat flows including gains and losses are summed up to create
the heating energy balance:
heat demand ¼ sum of heat losses−sum
of heat gains
Φh¼ ðΦT þ ΦV Þ− Φs þ Φi;o þ Φi;e þ Φi;L ½WŠ:
5. Exergy analysis
The common definition of energy utilization refers to the first law
of thermodynamics which states that energy is stored in every device
and process and can neither be consumed nor destroyed; it can only
be transformed [25]. For a more detailed analysis of energy flows in
buildings, this concept is “inadequate for depicting some important
aspects of energy resource utilization” [25]. In other words, it is not
Fig. 2. The heating chain as modelled for the energy calculations ([27], modified).
precise enough. In order to describe energy flows in buildings, the
concept of exergy was introduced by Shukuya [26]. To understand the
thermodynamic concepts of exergy and its counterpart, entropy, the
following description of an environmental control system for build-
ings, such as a heating, can be assumed:
In order to maintain comfortable indoor temperatures, energy
ð4Þ and mass have to be supplied. In the case of a heating system,
assuming steady state conditions, heat transmission occurs from the
ð5Þ warm inside to the cold outside of the environment. If assumed that
the energy is only transformed it should be possible to directly reuse
the energy [27]. However, the potential of the reusable energy is
smaller than at the point of input, the system “discards something”
[27]. The constant flow of energy from the warm inside of a building
to the cold outside results in increasing flow of entropy on the way
through the building envelope [28]. The concept of exergy quantifies
the potential of an energy source to be dispersed. The concept of
entropy helps to quantify the state of dispersion of the energy
source.
Unlike energy, exergy can be lost by decreasing the potential of an
energy source to do work. This can be illustrated with a simple
example [25]: Within an enclosure, a combustion process is used to
heat up water for the heating. An exergy source such as fuel, delivering
a temperature of 900 °C, heats up the water to a maximum
temperature of 60 °C. Only a small amount of the heat delivered is
used to heat the water, the rest diffuses into the surrounding
environment. According to the first law of thermodynamics, the
initial amount of energy within the enclosure is still the same. When
the fuel has been burnt, the potential of the water and the slightly
ð7Þ warmer environment to do further work is much lower than the
potential of the fuel that was used. The potential has been destroyed,
exergy has been lost. Exergy can therefore also be described as the
“valuable part of energy” [27]. The method of exergy analysis can be
used to “guide efforts to reduce sources of inefficiency in existing
systems” [25]. To analyse the systems, an integrated view on the
building is necessary. This integrated view must include building
geometry, construction and technical systems, making multi-dis-
ciplinary information of the building design crucial.
Most of the energy used in buildings is used to maintain
comfortable room temperatures. Heating consumes up to 57% of a
buildings total energy demand [1]. High-potential energy sources
ð8Þ
(exergy) and low potential energy sources found in the environment
can be used to generate heating energy. Low potential sources such as
outside air or geothermal heat are infinitely available. When the
heating chain is designed according to the optimal utilization of a low
potential source, for example to drive a heat pump, exergy obtained
from sources such as fossil fuels can be reduced. Because of the
influence of building form, construction and technical systems on the
exergy consumption of a building, the consideration of exergy results
in more flexibility for the building designer to choose optimization
measures. The common term of “saving energy” therefore has to be
 A. Schlueter, F. Thesseling / Automation in Construction 18 (2009) 153–163 157

Fig. 3. Implementation setup.

redefined; not energy but exergy efficiency is important to reduce CO2 surface and the outside temperature. This factor is then multiplied
emissions. with the heat demand of the building.

5.1. Energy/exergy calculations of the heating chain Exheat ¼ Φh  Fq;heat ½WŠ ð10Þ

In order to calculate the exergy flow through the heating chain, the
5.1.3. Emission subsystem
energy flow has to be estimated first. As heating energy is transformed
First the heat losses of the emission subsystem Φloss,E have to be
from heat generation to emission into the room, losses at each of the
calculated by taking the efficiency of the chosen emission system into
steps occur. The sum of losses defines the overall heating energy
account. Then the exergy load of the emission system is estimated in
demand to be supplied by primary energy which is transformed from
relation to inlet (Tin), return (Tret) and outside (To) temperatures.
fossil or renewable energy sources. In accordance to the DIN 4701-10
[29], the analysis of the heating chain starts at the generation of the Φh þ Φloss;E


T
 
primary energy. When the supplied energy passes through the steps, ΔExemis ¼  ðTin −Tret Þ−To  ln in ½WŠ ð11Þ
ðTin −Tret Þ Tret
the occurring losses are dependant on factors such as construction of
the envelope and choice of heating systems components.
The exergy demand after the emission systems is:
At the building level, all components influencing the heating chain
such as building geometry, construction and choice of systems are Exemis ¼ Exheat þ ΔExemis ð12Þ
considered for the analysis. The subsystems of the heating chain can
be separated into six steps. For each step, additional input parameters
for the building service systems are defined according to Schmidt [27]. 5.1.4. Distribution subsystem
These input parameters are automatically set due to the combination The exergy demand of the distribution subsystem is calculated
of the subsystems the user chooses in the interface. A detailed list of similar to the emission subsystem. First, its heat losses are calculated.
the input parameters can be found in Appendix A. The mean design temperature Tdis is used as inlet temperature, the
The energy and exergy calculations of the heating chain have to be return temperature is the design temperature minus the temperature
performed in the opposite direction of the development of the heat drop ΔTdis.
demand as shown in Fig. 2. The demand of each subsystem must be
Φloss; D Tdis
  
satisfied by the subsystem before. Following, the calculations for each ΔExdis ¼  ΔTdis −To  ln ½WŠ ð13Þ
step are briefly described in accordance to the model developed by ΔTdis Tdis −ΔTdis
Schmidt [27]:
The exergy demand after the distribution systems is:
5.1.1. Envelope subsystem
The total heat demand as calculated in the energy calculations is Exdis ¼ Exemis þ ΔExdis ½WŠ ð14Þ
most important for the first step. The exergy demand of the room is
estimated by multiplying the heat demand with the quality factor of 5.1.5. Storage subsystem
the room Fq,room. This quality factor is estimated by the Carnot The exergy demand of the storage subsystem is calculated similar
efficiency using the outside and inside temperature of the room. Then to the distribution subsystem. First, the heat losses are calculated. This
the exergy demand to be satisfied is: time, the mean design temperature Tsto is used as inlet temperature,
Exroom ¼ Φh  Fq;room ½WŠ ð9Þ the return temperature is the design temperature minus the
temperature drop ΔTsto.

5.1.2. Room air subsystem Φloss;S Tdis þ ΔTdis

  
ΔExsto ¼  ΔTsto −To  ln ½WŠ ð15Þ
The room is heated by a warm surface. The temperature difference ΔTsto Tdis þ ΔTdis −ΔTsto
between the heated surface and the room temperature defines the
exergy content. With the estimation of the surface temperature of the The exergy demand after the storage systems is:
heater, a new quality factor of the heater surface Fq,heat is also
calculated by the Carnot efficiency using the temperature of the heater Exsto ¼ Exdis þ ΔExsto ½WŠ ð16Þ
158 A. Schlueter, F. Thesseling / Automation in Construction 18 (2009) 153–163

Table 1 a fraction factor to divide the fossil part from the renewable part Frenew
Additional parameters to be added to the building model must be given. In this case, Fp is set to 3.0 (Germany). The non-
Component Parameter Type renewable part can be calculated as follows:
Room Illumination Semantic 
Occupancy Semantic
Eprim;tot ¼ ΦGe  Fp þ ðPl þ PV þ ∑Paux Þ Fp;electricity ½WŠ ð20Þ
Lighting Semantic
Wall u-value Semantic If an heat source Eenvironment is utilized to extract heat from the
toRoom Topological environment, the additional renewable part can be calculated:
Window u-value Semantic
g-value Semantic Erenew ¼ ΦGe  Frenew þ Eenvironment ½WŠ ð21Þ
Orientation Topological
toRoom Topological
5.1.8. Results of the exergy calculations
The exergy calculation results are displayed by two performance
5.1.6. Generation subsystem indices which show overall and specific exergy utilization.
The generation subsystem has to satisfy the demand of all
subsystems. A possible seasonal storage utilizes thermal solar power 5.1.8.1. Total exergy load. This index shows the total amount of exergy
with a solar fraction Fs. The requested energy of generation therefore is necessary to supply the building. The total exergy load is dependent

 1 on building construction, geometry and system selection. Optimiza-
ΦGe ¼ Φh þ Φloss;E þ Φloss;D þ Φloss;S ð1−FsÞ ½WŠ ð17Þ
ηG tion can be achieved by balancing between these three different fields.
The total exergy load of the building is:
The generation system is supplied with an energy carrier with a
pre-defined quality factor Fq,S. This factor is based from “statistical 
Extot ¼ ΦGe  Fp  Fq;S þ ðPl þ PV þ ∑Paux Þ Fp;electricity þ Erenew
material and political discussion” [27] and evaluates the qualities of  Fq;renew ½WŠ ð22Þ
different energy sources. In this case it is set to 1.0 (Germany). The
total exergy load therefore is:
5.1.8.2. Total exergy system efficiency. Dividing the remaining exergy
ExGe ¼ ΦGe  Fq;S ½WŠ ð18Þ
that leaves the room through the envelope by the total exergy load of
In addition, the exergy load of other building service components the room results in an exergy system efficiency index. It displays the
such as lighting Pl and ventilation PV is calculated, multiplied by the ratio of exergy that is actually used to heat the room.
quality factor and added to the total exergy demand of the building. 
Exroom

Exeff ¼ ð23Þ
Explant ¼ ðPl þ PV Þ Fq;electricity ½WŠ ð19Þ Extot

5.1.7. Primary energy transformation subsystem 6. Implementation: The Design Performance Viewer (DPV)
The required primary energy inputs are defined by the calculated
overall energy and exergy loads. It can be differentiated between The mathematical models for energy and exergy calculation are
renewable and non-renewable parts. The primary energy factor Fp and implemented into the software code. Necessary parameter values for

Fig. 4. Modeller with tool interface.

 A. Schlueter, F. Thesseling / Automation in Construction 18 (2009) 153–163 159

values have to be adapted, all geometry data is updated

automatically. In addition to the these parameters some naming
conventions have to receive attention, such as the correct naming
of the levels.
The input parameters for the energy/exergy calculations of the
heating chain (see Section 5.1) are automatically set by choosing the
preferred heating chain components in the interface. These system
specific parameters are embedded into the program code and not
stored in the building model.

6.2. Tool interface and visualization

The tool interface addresses the architect and building designer

who has to make conceptual and design decisions already in an
early stage of his design. The amount of necessary input
parameters was to be kept as small as possible without sacrificing
the significance and plausibility of the results. As many parameters
as possible are directly read out of the building information model.
Of the five different interface tabs, only one—the “Systems” tab—
needs user input to choose the subsystems of the heating chain.
The subsystem selections are labelled in a way an architect can
relate to them, such as “boiler” or “radiator”. The resulting input
parameters are automatically set according to the combination of
subsystems as some parameters influence parameters of other
subsystems.
Another important aspect is the visualization of the results. For
fast and intuitive comprehension, it is not sufficient to display the
results how this is commonly done in simulation software—in

Fig. 5. Building Data tab.

the calculations are accessed directly from the building model

database. In order to access the model database, the supplied
application programming interface (API) of the model editor is
utilized. The API can be addressed using a .NET [30] programming
language such as Visual Basic or C#. Applying the concept of object-
oriented programming, the software is organised in classes such as
calculation and visualization classes. The focus of the implementation
was on speed of calculation and fast and intuitive display of results. In
order to serve as a design support tool accepted by the building
designer, the results have to be calculated and visualized in quasi real-
time (Fig. 3).
The building model of a specific design is established in the
editor. Once the necessary parameters are set, the DPV is started as
an external application out of the editor. During the design
process, the user can switch between the calculations and
visualization in the tool interface and the building model of his
design. Alongside the progression of the building design, the
resulting performance can be assessed at any time. Additional
import or export procedures or manual data entry is not necessary
(Fig. 4).

6.1. Model parameterization

All geometry data such as areas and volumes is automatically taken

from model geometry. For the energy and exergy calculations, nine
additional parameters have to be added to the object properties of
rooms, walls and windows (Table 1).
These parameters have to be added once at the beginning of
the modelling process. During the design only the parameter Fig. 6. Performance tab with Kiviat diagram.
160 A. Schlueter, F. Thesseling / Automation in Construction 18 (2009) 153–163

Fig. 7. Building footprint of initial (left) and optimized design (right).

endless rows of numbers. Fast visual feedback is necessary for quick visualize the energy flows by components. Arrows of different
interpretation of the results; suitable graphical visualisation has to direction and strength show losses and gains of individual building
be implemented. components. The energy balance shows which components account

6.2.1. Building Data

The first interface tab “Building Data” (Fig. 5) shows extractions
from the geometry model and provides information important for the
architect such as opening surface ratio or orientation ratio of the
windows. This information is graphically represented by a bar chart.
The tab also displays the calculated average u- and g-values of walls
and windows.

6.2.2. Performance
The “Performance” tab (Fig. 6) displays selected energy perfor-
mance indices and visualizes the calculated results in a Kiviat diagram.
Kiviat diagrams are common in computer performance evaluation
[31] and can also be found in economic [32] and environmental
information visualization. The values of the performance indices are
plotted onto their individual axis. Connecting the nodes creates a
distinct shape, the “building performance footprint”. Changes of the
shape can be easily linked to changes in building design. For fast visual
judgement of the performance of design alternatives, their specific
shapes can be compared (Fig. 7).

6.2.3. Systems Selection

The tab “Systems Selection” (Fig. 8) is the only part of the
interface where user input is required. The building designer can
choose the component of the heating chain he prefers, from
generation of heating energy to storage, distribution and emission.
According to the selections in the pull-down menus, different
system values such as the system efficiency, supply and return
temperatures are automatically set. Additionally, the design tem-
peratures of the heating system, the outside and inside tempera-
tures, can be altered, defining the environmental conditions. These
values are used to define the conditions for the energy and exergy
analysis of the heating chain.

6.2.4. Energy/Exergy
The “En/Ex Balance” tab (Fig. 9) displays results and visualizations
of the energy and exergy flows of the current building and heating
system design. An automatically generated Sankey diagram is used to Fig. 8. System selection tab.
 A. Schlueter, F. Thesseling / Automation in Construction 18 (2009) 153–163
Fig. 9. Energy/exergy balance tab.
to which percentage to the energy loss of the building and which
sources contribute to the energy gains (Fig. 10).
Below the Sankey diagram, the exergy bar graph displays the
exergy performance indices: For each step of the heating chain the
exergy losses are calculated and displayed. It shows how exergy is
Fig. 10. Sankey diagrams of design with different energy supply configurations.
161
utilized throughout the steps of the heating chain. The exergy
efficiency index shows the percentage of exergy that is actually used
to heat the rooms. The total amount of exergy required for the
specific building is displayed as “Demand” under the first bar of the
bar graph.
Finally, the “Data” tab lists all executed energy and exergy
calculations and the extracted parameter values. This tab can be
viewed for the detailed results.
7. Results and discussion
The prototypical tool DPV enables balancing between possible
measures to increase overall building performance. The results from
the calculations of the simplified energy model implemented were
compared to the results of a commercial and certified software [33]
which is used to verify the conforming of the EnEV regulation. The
results show variations below 5%, proving the sufficiency for the
proposed early stage performance assessment. In contrast to the
software used for the comparison, the assessment of the perfor-
mance analysis of a specific building takes only a few seconds using
the DPV. In addition to the calculation of total energy and exergy
demands, the building designer can decide which optimization
measure is most suitable for the concept and context of the building.
Most important, balancing between form, materialization and
technical systems is possible from the beginning on. If, for example,
the façade cannot be altered, a better heating system using a
different energy source can be chosen. If, for another example, a slab
heating is desired, a certain setup of the heating system is necessary:
Due to the smaller heat exchange rate of the slab heating, low heat
losses are required. These can be achieved by choosing a good
envelope insulation and/or a mechanical ventilation. Also, solar gains
can be used to heat up the rooms. In order to increase solar gains,
the opening surfaces should be increased and the g-values of the
glass should be adapted.
Utilizing building information modelling to realize fast energy
and exergy performance assessment opens up the possibility of a
more integrated view on buildings during their early design stages.
The parameterized model enables capturing the complexity result-
ing from manifold dependencies of building components and
environment. Implementing the concept of exergy proved to
make more precise definitions of efficiency in building possible. It
enables balancing between different qualities of energy for
162 A. Schlueter, F. Thesseling / Automation in Construction 18 (2009) 153–163

different purposes. The concept of “low-exergy” [34]—to optimize Appendix A. System selections and input parameters of the
the use of the part of high-potential energy—offers more flexibility heating chain
for the building designer to choose appropriate measures for an
optimization. If, for example, a façade cannot be altered due to System parameters generation
conservatory reasons, the exergy demand can be reduced by Systems
choosing appropriate technical systems utilizing renewable 1. Condensing boiler
2. Heat pump water/glycol
sources. This also applies to new buildings: if a renewable energy System variables Unit
source such as geothermal heat is available, higher transmission Efficiency ηG/COP –
heat losses of the envelope caused by a certain façade design can be Primary energy factor source FP –
counterbalanced. In such a case, the energy demand would be high Quality factor of source Fq,S –
Max. supply temperature θS,max °C
but the exergy demand responsible for the CO2 emissions would
Auxiliary energy paux,ge W/kWheat
stay low. Only a fraction of high-potential energy, the exergy, is Auxiliary energy paux,ge,const W
necessary to drive the heat pump for heat generation. Due to the Part. environmental energy Frenew –
more precise view on the different qualities of energy and the
resulting increase of flexibility for the building designer, the System parameters storage
1. No storage
concept of low-energy should be replaced by the more precise
2. Small/day storage
definition of low-exergy. 3. Seasonal storage
Intuitive parameter input and instantaneous calculations are System variables Unit
important for performance assessment playing an equal role in the Heat loss/efficiency ηG –
Auxiliary energy paux,S W/kWheat
design process. However, these simplifications have to lead to
Solar fraction FS –
correct calculation results. In this approach, the fast calculation is
achieved by using a statistical mathematical model. The model System parameter distribution
does not include dynamic calculations over a certain time period Subsystem Selection
and simplifies certain input parameters. It cannot be used to 1. Boiler position inside/outside/no distribution
2. Insulation no/good/bad insulation
simulate the annual energy demand of the building. However, it
3. Design temperature low b 35 °C/middle b 50 °C/high (other)
shows tendencies and estimates necessary to make design 4. Temperature drop low b 35 K/middle b 50 K/high (other)
decisions and to communicate them. In professional application, System variables Unit
the tendencies shown by the DPV would have to be supplemented Heat loss/efficiency ηD –
Auxiliary energy paux,ge W/kWheat
by simulations offering a higher resolution. Important however is
that tendencies how good a design performs can be discovered and System parameters emission
considered to make design decisions. Further work on the DPV is Systems
dedicated to include dynamic calculations of energy demands by 1. Floor heating
keeping the immediate output of the results. A first case study in 2. Radiator
3. Ceiling heating
context of the “ViaGialla” concept [35], carried out with 27 students
4. Slab heating
of architecture at ETH Zürich has shown that the DPV is suitable to System variables Unit
integrate performance criteria into design decision-making. In two Inlet temperature θin °C
workshops, the students used the DPV from the beginning of their Return temperature θret °C
Auxiliary energy paux,E W/kWheat
designs. The introduced editor made easy establishment and
Max. heat emission pheat,max W/m2
parameterization of the building information model possible. Heat loss/efficiency ηE –
Students were able to evaluate their designs by using the DPV. In
some cases, these evaluations had strong effects on their design,
some students chose to give performance aspects less considera- References
tion. But all students became aware of the dependencies and
possible measures to design energy/exergy efficient buildings. [1] Directive 2002/91/EC of the European Parliament and of the Council of on the
energy performance of buildings. 2002 [cited 2007], Available from: http://ec.
Intuitive visualization of the results to enable easy and quick
europa.eu/energy/demand/legislation/buildings_en.htm.
interpretation proved to be very important and has to be carried [2] Minergie Standards. 2008 [cited], Available from: http://www.minergie.ch/index.
even further. php.
Computer based methods are used in an impressing and [3] The Leadership in Energy and Environmental Design (LEED) Green Building Rating
System. 2007 [cited], Available from: http://www.usgbc.org/DisplayPage.aspx?
exciting way in various fields of architecture such as advanced CMSPageID=222.
geometry and computer-aided production methods. New tools [4] R. Brahme, A. Mahdavi, K. Lam, S. Gupta, Complex Building Performance Analysis
make it possible to design and actually build forms that would not in Early Stages of Design, IBPSA, Rio de Janeiro, Brazil, 2001.
[5] H. Grabowski, M. Eigner, Semantic data-model requirements and realization with a
be possible without the use of computers. These powerful methods relational data-structure, Computer-Aided Design 11 (3) (1979) 158–168.
have to be applied to actually design better building, not only [6] C. Eastman, General purpose building description systems, Computer Aided
better looking ones. Incorporating building performance analysis Design 8 (1) (1976) 17–26.
[7] C. Eastman, Building Product Models: Computer Environments Supporting Design
into the design process utilizes the full potential of computational and Construction, CRC, Boca Raton, 1999.
methods in architecture. To capture the complex dependencies, [8] J. Laiserin, The Laiserin Letter. [cited 2008], Available from: http://www.laiserin.com/.
to view the building as a system makes new approaches in [9] International Alliance for Interoperability (IAI). [cited 2007], Available from:
http://www.iai-international.org/.
architecture possible. These approaches have to be further
[10] V. Bazjanec, Building energy performance simulation as part of interoperable
explored. software environments, Building and Environment 39 (8) (2004).
[11] IFC Compliant Software. [cited 2007], Available from: http://www.iai.fhm.edu/.
[12] M. Dayal, J. Timmermans, Analyse des 3D-Datenaustausches via IFC-Modell am Beispiel
Acknowledgements
komplexer Objektdokumentation, der Automobilindustrie mit dem Ziel der Optimier-
ung von Planungsprozessen, Fraunhofer IRB Verlag, Stuttgart, 2004.
The authors wish to acknowledge Prof. Dr. Hansjürg Leibund- [13] J. Plume, J. Mitchell, Collaborative Design Using a Shared IFC Building Model—
gut, Forrest Meggers and Luca Baldini for their inspiration and Learning from Experience, Elsevier B.V., 2007.
[14] AUTODESK Homepage. [cited 2008], Available from: www.autodesk.com.
support in research and implementation of the energy and exergy [15] Interoperability in the Construction Industry. 2007 [cited 2008], Available from:
models. http://www.aia.org/SiteObjects/files/ipd_SMReport.pdf.
 A. Schlueter, F. Thesseling / Automation in Construction 18 (2009) 153–163
[16] U.S. General Services Administration: 3D–4D Building Information Modeling.
[cited 2008], Available from: www.gsa.gov/bim.
[17] TRNSYS — Simulation environment for the transient simulation of thermal systems
including multi-zone buildings, Homepage. [cited 2008], Available from: http://
www.transsolar.com/__software/docs/trnsys/trnsys_uebersicht_en.htm.
[18] IES Virtual Environment, Homepage. [cited 2008], Available from: http://www.
iesve.com/content/.
[19] EnergyPlus Energy Simulation Software, Homepage. [cited 2008], Available from:
http://www.eere.energy.gov/buildings/energyplus/.
[20] D. Crawley, L. Hand, M. Kummert, B. Griffith, Contrasting the Capabilities of
Building Energy Performance Simulation Programs, Proceedings of the Ninth
International IBPSA Conference, Montereal, Canada, 2005.
[21] ECOTECT Homepage. [cited 2007], Available from: http://www.squ1.com/index.
html.
[22] EnEV — Energieeinsparverordnung (Version 2007). 2002 [cited], Available from:
http://www.gesetze-im-internet.de/enev_2007/BJNR151900007.html.
[23] IdeaXP — Berechnungsprogramm für HLK, Homepage. [cited 2007], Available
from: http://interway.ideaxp.ch/.
[24] SIA — Schweizerischer Ingenieur-und Architekenverein. [cited 2008], Available
from: http://www.sia.ch/d/praxis/normen/normenwerk.cfm.
[25] M. Moran, H. Shapiro, Fundamentals of Engineering Thermodynamics, 4th ed.John
Wiley & Sons,, Hoboken, 2000.
163
[26] M. Shukuya, Energy, Entropy, Exergy and Space Heating Systems, Proceedings of
the 3rd International Conference “Healthy Building ‘94”, vol. 1, 1994, pp. 369–374.
[27] D. Schmidt, Design of low exergy buildings — method and a pre-design tool,
International Journal of Low Energy and Sustainable Buildings 3 (2003).
[28] M. Shukuya, Bioclimatic design as rational design of exergy–entropy process.
Proceedings of PLEA ’98, Lisbon, Portugal, 1998, pp. 321–324.
[29] DIN 4701-10: Energy Efficiency of Heating and Ventilation Systems in Buildings —
Part 10: Heating, Domestic Hot Water, Ventilation. / German National Standard,
Deutsches Institut fuer Normung e.V, Berlin, 2002.
[30] Microsoft .NET, Homepage. [cited 2008], Available from: http://www.microsoft.
com/net/basics.mspx.
[31] K. Kolence, P. Kiviat, Software unit profiles and Kiviat figures, Performance
Evaluation Review 2 (3) (1973) 2–12.
[32] D. Tegarden, Business information visualization, Communications of the Associa-
tion of Information Systems 1 (1999).
[33] WEKA Gebäudeplaner. [cited 2008], Available from: http://www.weka-enev.de.
[34] M. Shukuya, A. Hammache, Introduction to the Concept of Exergy — for a Better
Understanding of Low-Temperature-Heating and High-Temperature-Cooling
Systems, VTT, Espoo, 2002.
[35] H. Leibundgut, Institut für Hochbautechnik, ViaGialla: Wegbeschreibung für
Gebäude in eine nachhaltige Energie-Zukunft, ETH, Eidgenössische Technische
Hochschule Zürich, 2007.