JOURNAL OF CIVIL ENGINEERING AND MANAGEMENT

ISSN 1392-3730 / eISSN 1822-3605

2016 Volume 22(3): 333–345
doi:10.3846/13923730.2014.897975

A PROCESS TO DIVIDE CURVED WALLS IN IFC-BIM INTO SEGMENTED

STRAIGHT WALLS FOR BUILDING ENERGY ANALYSIS

Karam KIM, Jungho YU

Department of Architectural Engineering, Kwangwoon University, 139070 Seoul, South Korea

Received 23 Jan 2013; accepted 03 May 2013

Abstract. The amount of energy that buildings consume is currently attracting significant interest throughout the world,
and considerable attention is being devoted to the development of green buildings to reduce energy consumption; how-
ever, the current process being used to conduct building energy analysis has a significant shortcoming in that it cannot
directly use building information modelling (BIM)-based representational data that includes a curved wall. Currently,
the curved wall must be converted into segmented straight walls (SSWs) in the building model in order for the building
energy analysis (BEA) program to be able to recognize and use the data. In this paper, we have proposed a segmentation
process for curved walls in industry foundation classes (IFC)-based BIM for BEA. The proposed process consists of
three sub-processes: 1) extracting data from the IFC model; 2) dividing the curved wall into several segmented straight
walls; and 3) generating an INP file as a building description for DOE-2.2-based BEA. The proposed process will enable
the engineers who are responsible for BEA to use a BIM-based model directly in the BEA program without having to do
additional work. The proposed process can help ensure that the BEA results are accurate and reliable.
Keywords: BIM, building energy analysis, IFC, curved wall, representation data.

Introduction
tion of a building (Laine, Karola 2007; Woo et al. 2011;
Since the energy consumption of buildings continues to Dziugaite-Tumeniene et al. 2012; Oh et al. 2011; Venckus
increase, the use of building energy analysis (BEA) to et al. 2012; Ali 2004). However, since there is a problem
simulate building energy consumption has become an related to the interoperability of data between BIM-based
important issue worldwide. In the last decade, extensive design software and the BEA system, the 3D model, in
research related to BEA has been conducted using various general, must be checked and remodelled by the BEA
methodologies. Schlueter and Thesseling (2009) proposed engineer (Maile et al. 2007). To that end, in 2000, the
a BEA approach that considered the thermo-economic green building XML (gbXML) file format was developed
concept of exergy in alternative designs. Exergy analysis for the BEA process. Since gbXML supports 3D models,
takes into account the overall available energy (Shukuya, BIM-based BEA systems can be used in several appli-
Hammache 2002). Kim et al. (2011) suggested a BEA ap- cations (Hygh et al. 2012; Pratt et al. 2012). However,
proach that used data mining to determine useful patterns as some BIM-based design programs cannot export the
for improving energy efficiency during the design phase. building model in gbXML directly, the engineer must re-
And Kim and Yu (2013a) proposed a material name- model the building model through the use of other soft-
matching system using ontology to automatically add the ware that supports the gbXML file.
thermal properties of construction materials. One of the most widely-used BEA engines in the
BEA systems have also been developed with build- world is the DOE-2.2 engine; however, since the BEA
ing information modelling (BIM) technology. Since BIM engine cannot support a complex building model, a spe-
technology has been used in architecture, engineering, cific algorithm is necessary to analyze a complex build-
and contracting (AEC) industry, a 3D-based building ing model, and researchers have attempted to analyze
model can be used in the BEA process; using a BIM- complex building models (Sun et al. 2002; Tang et al.
based BEA system allows the engineer to easily and 2006; Monstvilas et al. 2012; Suter, Mahdavi 2004; Kim,
accurately calculate or simulate the energy consump- Yu 2013b). Previous research, however, has been limited

This article has been corrected since first published. Please see the statement of correct (doi:10.3846/13923730.2015.1129176
of the erratum).

Corresponding author: Jungho Yu

E-mail: [email protected]
333 Copyright © 2016 Vilnius Gediminas Technical University (VGTU) Press
www.tandfonline.com/tcem
334 K. Kim, J. Yu. A process to divide curved walls in IFC-BIM into segmented straight walls for building energy analysis

in that a specific system or mathematical algorithm must free-form curves. In addition, the representation data are
be used that lacks data interoperability with other BEA referenced by the “basiscurve” attribute of the IfcCircle
tools. Thus, engineers may be required to use a specific entity from the IFC schema as a curved wall. The DOE-2.2
tool with which they have no experience in the process engine is used in this study as the BEA engine, so the
of implementing a building model that involves complex format of the building description is the INP file format.
geometry.
There are a few approaches to analysing the ener- 1. Required data
gy consumption of a building model that include curved In this paper, we focused on the representation data of
walls. The first approach is to rebuild a building model a curved wall in the IFC model’s data. For this reason,
for the purpose of energy analysis into a BEA program when a curved wall is created by a BIM-based design
(e.g. ECOTECT or eQUEST). In this case, an engineer tool, the representation data is stored in specific data en-
must segment a curved wall into a series of straight tities. Before the data can be used in the proposed pro-
walls, because most of the BEA programs cannot process cess, these data entities must be defined. The following
curved wall data (Miller 2010). The second approach is contexts show the representation data of the specific data
to use an add-on program (e.g. AUTODESK 360 Energy entities in the IFC schema.
analysis or EcoDesigner) on a BIM authorizing program A curved wall has the attributes of the IfcWall en-
(e.g. Revit or ArchiCAD). This is a very simple and suc- tity in the IFC schema. There are many attributes for
cinct approach, but it has limitations in regard to the in- the IfcWall, but the Representation and ObjectPlacement
teroperability of the building model. The third approach data are the most important data related to the proposed
is to use the gbXML file of a building model. As the process. The data for ObjectPlacement are related to the
gbXML file format is designed only for BEA, it is widely 3D-based coordinates of the relative coordinate system
known that using a gbXML file is a de-facto standard for (RCS) that is called the X’-Y’ system. In this paper,
BEA; however, if we can extract all the required data for the Ifc Axis 2 Placement 3D entity is done through the
BEA from an IFC file, it is much more effective to use IfcObjectPlacement entity.
an IFC file than to handle other formats, such as gbXML. There is an angle value (θR) between the X-Y and
To solve these problems, we propose a process for X’-Y’ systems. In addition, the representation data of a
dividing curved walls in IFC into segmented straight walls curved wall, as its shape information, are created by the
(SSWs) for the BEA process. In the process, a curved basis curve (according to the IfcCircle entity in the IFC
wall is divided into a number of SSWs by calculating schema) to relate IfcShaperep Resentation through the If-
the coordinates using trigonometric functions. Since the cProduct Representation entity. Some data are required to
proposed process adopts the IFC file as an input, it is not develop the proposed process, as shown in Figure 1 and
necessary to generate any additional building model for listed below:
the BEA process. The proposed process consists of three
sub-processes: 1) extracting data from the IFC model; 2) 1. X-Y system
dividing the curved wall into several SSWs; and 3) gen- X-Y system refers to the absolute coordinate system
erating an INP file. Further, if the sub-processes are com- (ACS) that defines the location of the building elements
puterized, the process of segmenting a curved wall into on the IFC schema. In this paper, the X-Y system faces
SSWs can be automatically implemented. the X’-Y’ system as an RCS.
In this paper, we have limited curved walls to those
that arc along their footprint, avoiding elliptical and

Fig. 1. Definitions of required data on proposed process

Journal of Civil Engineering and Management, 2016, 22(3): 333–345 335

2. X’-Y’ system 2. Proposed process

The X’Y’ system refers to the RCS that defines the
2.1. Overall process
representation of the curved wall according to the Ifc-
Circle entity. To divide a curved wall into several (i.e. n) SSWs, the
3. Center Point (CP) overall proposed process is shown in Figure 3. The pro-
CP is the center point of the circle that defines cess proposed in Figure 3 is comprised of three sub-
the shape of a curved wall. CP exists on the Y’ axis in processes. First, the related data are extracted from the IFC
the X’-Y’ system, as shown in Figure 1. If the radius file using a specific entity. The extracted data are stored to
of the circle is R, CP is located by (0, R) in the X’-Y’ the BIM object DB to use in the other procedures. Based
system. on the extracted data of the BIM object DB, segmented
4. Trimmed Point (TP1) straight wall points in the RCS (SSWPRn) are created by
TP1 is the starting point of the reference line of a the process of point rotation using trigonometric functions.
curved wall. In general, TP1 is located at the zero point Then, SSWPRn is converted into segmented straight wall
(0, 0) in the X’-Y’ system to define the reference line of points in the ACS (SSWPAn) by a specific mathemati-
a curved wall. cal formula using the values of θR and θA. Segmented
5. Trimmed Point (TP2) straight walls (SSWn) are defined by connecting each
TP2 is an end point of the reference line of a curved SSWPAn. The INP file is generated by the created SSWn s,
wall in the X’-Y’ system. including the attributes of the original curved wall, such as
6. Radius value (R) the materials used, thickness, and building story.
R is the radius of the circular sector of a reference line
2.2. Process of extracting data from IFC
of a curved wall. It is extracted from the IfcCircle entity.
7. Trimmed angle1 (θ1) To extract the required data of the proposed algorithm,
θ1 is the angle between the zero point and TP1 in several data entities are parsed, including IfcWall and If-
the X’-Y’ system: it can be extracted from the IfcPa- cTrimmedCurve. Both entities are related to each other
rameterValue of IfcTrimmedCurve through the Time1 by relationships of reference. For example, the represen-
attribute. tation data of IfcWall are contained in IfcShape Repre-
8. Trimmed angle2 (θ2) sentation as a polygon. If the wall is curved, the polygon
θ2 is the angle between the zero point and TP2 in is expressed by the IfcTrimmedCurve entity as a relation-
the X’-Y’ system. ship of reference. At first, the system searches IfcWall
9. θ entities in the IFC file. If the IfcWall entity has the If-
θ is an angle that is related to the center point of a cTrimmedCurve attribute in the representation data, the
circular sector. In the proposed process, θ is calculated wall is curved. Then, Sense Agreement data value can be
by Eqn (1): extracted from the IfcTrimmedCurve entity. The CP, R,
θR, TP1, θ1, TP2, and θ2 data also can be extracted from
. (1) the IfcTrimmedCurve entity. In addition, the location co-
ordinates, θA, and the depth data can be extracted from
10. Rotation angle value between systems (R)
the representation data of the IfcWall entity. The data
θR is the angle of the element representation be-
extraction sub-process is shown in Figure 4.
tween the X-Y and X’-Y’ systems to locate the elements
on the X-Y system according to the angle of the X and 2.3. Process of dividing curved walls
X’ axes.
2.3.1. Process of creating SSWPRn
11. Rotation angle value with location data (θA)
θA is the angle of element placement. When the ele- First, the angle value θ is calculated through θ2 of Trim2
ment location is defined by the IFC schema, the angles and θ1 of Trim1 as the IfcParameter Value from BIM ob-
of the ObjectPlacement and Representaion attributes of ject DB (Eqn (1)). Using the value of Sense agreement,
a building element must be considered. the direction of θ is defined, and, if the value of Sense
12. Sense Agreement agreement is false, the direction of θ is changed by “– θ”.
Sense Agreement is the standard direction of an an- Then, the number of SSWs provided as input to calcu-
gle to measure the rotation of an angle in the IFC sche- late the coordinates of SSWPRn is the “n” value, which
ma. In general, “False” connotes a clockwise direction, should be larger than 2. Then, the coordinates of TP1 are
and “True” means a counter clockwise direction. determined. If there is no TP1 on the BIM object DB at
13. Location coordinates this time, then TP1 has not been extracted from IFC, and
Location coordinates are the data of coordinates that TP1 is obtained by Eqn (2), using CP of the IfcCircle. In
define where the element is placed on the X-Y system. general, the coordinates of CP are (0, R):
They consist of X, Y, and Z axis values.
(2)
14. Depth
Depth is the wall height data value of the curved To calculate the coordinates of SSWPRn, we rotate the
wall. The depth data are contained as representation data. coordinates of TP1 to the zero point in the X’-Y’ sys-
336 K. Kim, J. Yu. A process to divide curved walls in IFC-BIM into segmented straight walls for building energy analysis

Fig. 2. Data extraction path in IFC

tem. To that end, CP is moved to the zero point using The process of creating SSWPRn using representation
Eqns (3) and (4), and it is designated as CP’. At this data from the IFC model is shown in Figure 6.
time, all of the extracted coordinates are also moved the
same distance: 2.3.2. Process of creating SSWPAn
Since SSWPAn are created in the X’-Y’ system, SSWPAn
(3)
must be converted to the X-Y system as the ACS. Before
.(4) doing so, the angle of rotation between X’-Y’ and X-Y is
defined by Eqn (7):
Next, θ is divided by n to calculate the rotation angle of (7)
TP1′ . Using the rotation angle, TP1′ is rotated by θ/n in The SSWPRn can be rotated with Eqn (7) to convert
the direction of θ n times using Eqns (5) and (6), desig- it to SSWPAn as Eqns (8) and (9):
nated as SSWPRn (Fig. 5). When k becomes equal to n,
the coordinates of SSWPRn are the same as TP2′ : ; (8)

 (5) . (9)

(6) The SSWPA ′n are located in the X-Y system, but

not in the original location of the curved wall. Thus,
Journal of Civil Engineering and Management, 2016, 22(3): 333–345 337

Fig. 3. Overall process

data is replaced by the depth data in the extracted data.
SSWPA ′n must be moved by the function of move as In addition, the width data is replaced by the length
Eqn (10): of SSWn. The coordinates of X, Y, and Z can be filled by
the created SSWPAn. Moreover, the required data for the
Function of move SSWPA'n = INP file must have the units of ft-lb.
.(10)
( Location coordinates ) − (SSWPA′0 ) The data format of the INP file can be structured as
an exterior wall data template used to generate an INP
Using the function of move, all of the SSWPA ′n file. Figure 9 is a template of INP file using the data map-
are moved to the original location of the curved wall as ping table. In this regard, the attributes applied to the can
SSWPA ′n . be added to the template as property data. For example,
a curved wall might contain material information in IFC
2.3.3. Process of creating SSWn file. Then, the material information can be applied to the
To create SSWn between SSWPAn as an element, the created SSWn as an additional applied attribute.
distance between SSWPAn must be known to use the The process of generating an INP file, including the
reference line of the created SSWn. Eqn (11) is used to attributes of the curved wall, is shown in Figure 10.
calculate the distance of SSWPAn (Fig. 8):
3. Verification
. (11)
To verify the proposed process, a sample curved wall was
modeled by ArchiCAD version 14 and named model A.
2.4. Process of generating INP file Contrastingly, the curved wall was remodeled with three
The created SSWn must have specific attributes, such SSWs by the same CAD program as a practical case
as the material and the thickness of the original curved of the BEA process and was named model B (Fig. 11).
wall. To that end, the attribute data are loaded into the Thus, the verification process shows the difference in
created SSW from the BIM object DB. Finally, the terms of the representation data between the proposed
created SSWn, including the attributes of the original process using models A and B. For model A, the value
curved wall, can be entered into an INP file as a build- of the curved wall was 113 degrees; however, for model
ing description of the DOE-2.2 BEA engine. The cre- B, the location points of each straight wall have an angle
ated placement and representation data of the created of 38 degrees.
SSWn can be converted to the INP format according Models A and B were exported to the IFC model us-
to Table 1. According to the DOE-2.2 manual (James ing the ArchiCAD 14 translator. Since model A include
J. Hirsch & Associates 2009), the required data of a a curved wall, the representation data can be extracted,
wall in an INP file is defined by four data. The height as shown in Table 2, by the proposed process (Fig. 12).
338 K. Kim, J. Yu. A process to divide curved walls in IFC-BIM into segmented straight walls for building energy analysis

Fig. 4. Process to extract data from IFC

Fig. 5. Concept of rotated SSWPRn based on CP’

Journal of Civil Engineering and Management, 2016, 22(3): 333–345 339

Fig. 6. Process to create SSWPRn

Fig. 7. Process of SSWPAn

340 K. Kim, J. Yu. A process to divide curved walls in IFC-BIM into segmented straight walls for building energy analysis

Fig. 8. Process to create SSWn

Table 1. Data mapping to generate INP file

Required data for INP file Data from BIM object DB Description
WALL NAME Name of SSWn The name value of a created SSWn.
HEIGHT Depth The height value of an element.
WIDTH The length of SSWn The width value of an element.
X
Y The coordinates of SSWPAn The location coordinates as a start point of an element.
Z

Fig. 9. Template of INP file

Table 3 shows the location points, i.e., the starting

point of each element’s reference line on the ACS, of the
straight walls for model B.
Next, the CP is moved to the zero point in the X’-Y’
system to rotate the TP1. TP1 and TP2 also are moved
6,000 mm in the direction of negative Y’. Then, the TP1′
point is rotated basis on CP′ to create SSWPRn, as
follows:
Journal of Civil Engineering and Management, 2016, 22(3): 333–345 341

Fig. 10. Process of generating INP file

Since the created SSWPRn must be converted to the

X-Y system as the ACS, they were rotated by the angle
of value rotation (i.e. θA + θR), as follows:

Fig. 11. Sample models

Table 2. Data extracted from model A

Data name Data value

Center Point (0, 6000)
Radius(R) 6000
Ref Direction (θR) 90
Ref Direction (θA) 33.56
Sense Agreement True
Trimmed Point1 (TP1) (0, 0)
Trimmed Angle1 (θ1) 0
(5527.707984,
Trimmed Point2 (TP2)
8333.333333)
Trimmed Angle2 (θ2) 112.8853805
The calculated SSWPA ′n must move to the original location
Location Coordinates (0, 10000) of the curved wall in model A. SSWPA ′n were moved as
342 K. Kim, J. Yu. A process to divide curved walls in IFC-BIM into segmented straight walls for building energy analysis

Fig. 12. Points extracted from model A

Fig. 15. Created SSWPAn

Function of move SSWPA ′n (i.e. (0, 10000) – (–5000.156,

3316.390) = (5000.156, 6683.610), as follows:

Fig. 13. Created SSWPRn

Using the SSWPAn, the SSWns can be created as

model A’ (Fig. 16). The length of SSWn is calculated as
follows:

Fig. 14. Created SSWPA ′n

Table 3. Data extracted from model B

Data name Data value

The location point of
(6935.004779, 12362.79036)
element 1
The location point of
(3065.086446, 12362.82144) According to the previous process, a comparison be-
element 2 tween the calculated coordinates of SSWPAn from model A’
The location point of
(0, 10000) and the location coordinates from model B are shown
element 3 in Table 4. As a result, the coordinates of the creat-
Journal of Civil Engineering and Management, 2016, 22(3): 333–345 343

The width data were filled by the calculated length of

SSWn: in this case, the width was about 3,870 mm (12.7 ft).
The coordinates of the location data were filled by
SSWPAn as the value of X and Y axes (Table 5). How-
ever, the Z axis was filled by the evaluation of the original
curved wall as an attribute: in this case, the evaluation
value was zero.

Conclusions
Since BIM technology is systemically being used in the
BEA process, BEA engineers can use a 3D-based build-
ing model in the BEA process, thereby making simula-
Fig. 16. Representation data of model A’ with SSWn tion of the building’s energy usage more accurate and
efficient; however, since the current BEA process can-
ed SSWPAn in model A’ are the same as the location not use representation data and include a curved wall
coordinates of the elements in model B except that they directly, the building model must translate the curved
are rounded off to the nearest integer. wall into a number of SSWs to recognize the represen-
Model A’ with can be converted to an INP file as a tation data in the BEA program. As this, remodelling
building description for the DOE-2.2 engine. In the INP work must be done manually, it opens up the opportunity
format, several representation data of a wall are necessary for human error. In addition, since the results of the re-
for BEA, such as height, width, and the coordinates of the modelling work might differ from worker to worker, the
location (i.e. X, Y, and Z axes) (Kim, Yu 2012). To convert accuracy and reliability of the BEA results are reduced.
the representation data of model A’, the required data must As the number of the SSWs increases, the accuracy and
be filled by SSWPAn and SSWn using the units of ft-lb. reliability also increase; however, since the remodelling
The height data were filled by the attributes of the original work is manual, the ability to determine the number of
curved wall: in this case, the height was 2,800 mm (9.186 ft). SSWs is limited.

Table 4. Comparison between models A and B

SSWPAn Model A’ The element Model B

SSWPA3 (0.000, 10000.000) element 3 (0, 10000)

SSWPA2 (3064.885, 12362.934) element 2 (3065.086446, 12362.82144)

SSWPA1 (6934.864, 12363.116) element 1 (6935.004779, 12362.79036)

Table 5. Representation data to convert INP file

Coordinates from INP format

Location points Element Coordinates value
Model A (ft-lb units)
SSW3 = EXTERIOR-WALL
HEIGHT = 9.19
(0.000, WIDTH = 12.7
SSWPA3 SSW3 10000.000) X = 0.0
CP = (0, 6000)
R = 6000 Y = 10000.0
θR = 90 Z = 0.0
θA = 33.56 SSW2 = EXTERIOR-WALL
Sense Agreement = True HEIGHT = 9.19
TP1 = (0, 0) (3064.885, WIDTH = 12.7
θ1 = 0
SSWPA2 SSW2 12362.934) X = 3064.885
TP2 = Y = 12362.934
(5527.707984,8333.333333) Z = 0.0
θ2 = 112.8853805 SSW1 = EXTERIOR-WALL
Location Coordinates = HEIGHT = 9.19
(0, 10000) (6934.864, WIDTH = 12.7
SSWPA1 SSW1 12363.116) X = 6934.864
Y = 12363.116
Z = 0.0
344 K. Kim, J. Yu. A process to divide curved walls in IFC-BIM into segmented straight walls for building energy analysis

To solve this problem, we proposed a process to Hygh, J. S.; Decarolis, J. F.; Hill, D. B.; Ranjithan, S. R. 2012.
divide curved walls in IFC into segmented straight walls Multivariate regression as an energy assessment tool in
early building design, Building and Environment 57:
for BEA. To that end, we first, defined the extracted data
165–175. http://dx.doi.org/10.1016/j.buildenv.2012.04.021
for the proposed process. Then, a trimmed point was Hirch, J. J. and Associates. 2009. Building energy use and cost
rotated by a specific angle to create the new location analysis program, Vol. 2: dictionary. Lawrence Berkeley
coordinates using a trigonometric function. After that, national laboratory, 1–512.
the straight walls were created between the calculated Kim, H.; Stumpf, A.; Kim, W. 2011. Analysis of an energy
efficient building design through data mining approach,
points as a building element. Finally the attributes of the
­Automation in Construction 20(1): 37–43.
original curved wall were applied to the created SSWn. http://dx.doi.org/10.1016/j.autcon.2010.07.006
To verify the proposed process, two sample models were Kim, K.; Yu, J. 2012. A method for extraction geometry data
used to evaluate the proposed process, i.e. model A, from IFC file for building energy load analysis, Journal
which included a curved wall, and model B, which did of the Architectural Institute of Korea Planning & Design
28(5): 241–248.
not. The results indicated that the coordinates of the cre-
Kim, K.; Yu, J. 2013a. Semantic material name matching sys-
ated SSWPAn in model A’ were almost the same as the tem for building energy analysis, Automation in Construc-
location coordinates of the element in model B. tion 30: 242–255.
There is a few advantages over the three current http://dx.doi.org/10.1016/j.autcon.2012.11.011
approaches, comprised of the following: 1) human- Kim, K.; Yu, J. 2013b. IFCXML based automatic data in-
put approach for building energy performance analy-
dependent tasks that can easily create mistakes in the
sis, Journal of Construction Engineering and Project
BEA process are reduced; 2) data interoperability of a Management 3(1): 14–21.
building model is assured by the use of the IFC file for- http://dx.doi.org/10.6106/JCEPM.2013.3.1.014
mat; 3) reliability and consistency of the BEA result are Laine, T.; Karola, A. 2007. Benefits of building information
increased; and 4) the BEA process can be more efficient models in energy analysis, in The Proceedings of Clima
2007 Wellbeing Indoors. 8 p.
because additional building models only for a BEA pur-
Maile, T.; Fischer, M.; Bazjanac, V. 2007. Building energy per-
pose (such as gbXML) are no longer necessary. formance simulation tools – a life-cycle and interoper-
This process will enable engineers involved in a able perspective. CIFE working paper #WP107. Stanford:
BEA process to use a BIM-based building model direct- Center for Integrated Facility Engineering.
ly in the BEA program without having to do additional Monstvilas, E.; Stankevicius, V.; Karbauskite, J.; Burlingis, A.;
Banionis, K. 2012. Hourly calculation method of build-
work. Consequently, this process can contribute toward
ing energy demand for space heating and cooling based
ensuring the accuracy and reliability of the results of the on steady-state heat balance equations, Journal of Civil
BEA. As n of SSWn increases, the possibility of error Engineering and Management 18(3): 356–368.
in the results of the BEA decreases. In addition, since http://dx.doi.org/10.3846/13923730.2012.689994
remodelling a building model for BEA will no longer Miller, J. 2010. Leveraging BIM for energy analysis. Autodesk
university tutorial resource [online], [cited 11 April 2014].
be necessary, the efficiency and objectivity of the BEA
Available from Internet:
process increases dramatically. http://sustainabilityworkshop.autodesk.com/sites/default/files/
Future work should further evaluate a BIM-based core-page-files/leveragingbimforenergyanalysis.pdf
building model as a practical case, including various Oh, S.; Kim, Y.; Park, C.; Kim, I. 2011. Process-driven BIM-
wall, slab, and roof types. In addition, elliptical and free- based optimal design using intergration of energyplus, ge-
netic algorithm and pareto optimality, in 12th Conference
form curved walls should be considered with the pro-
of International Building Performance Simulation Associ-
posed process. ation, 14–16 November 2011, Sydney, Australia, 894–901.
Pratt, K. B.; Jones, N. L.; Schumann, L.; Bosworth, D. E.;
Acknowledgements Heumann, A. D. 2012. Automated translation of architec-
tural models for energy simulation, in Proceedings of the
This work was supported by the National Research Foun- 2012 Symposium on Simulation for Architecture and Urban
dation of Korea (NRF) grant funded by the Korea gov- Design, 2012, San Diego, CA, USA, Article No. 6, 1–8.
Schlueter, A.; Thesseling, F. 2009. Building information model
ernment (MEST) (NRF-2014R1A2A2A05006437).
based energy/exergy performance assessment in early de-
This research was supported by a grant (15AUDP- sign stages, Automation in Construction 18(2): 153–163.
C067809-03) from Architecture & Urban Development http://dx.doi.org/10.1016/j.autcon.2008.07.003
Research Program funded by Ministry of Land, Infra- Shukuya, M.; Hammache, A. 2002. Introduction to the con-
structure and Transport of Korean government. cept of exergy – for a better understanding of low-tem-
perature-heating and high-temperature-cooling systems.
Espoo: VTT. 61 p.
References Suter, G.; Mahdavi, A. 2004. Elements of a representation
framework for performance-based design, Building and
Ali, M. 2004. Developments in environmental performance
Environment 39(8): 969–988.
simulation, Automation in Construction 13(4): 437–445.
http://dx.doi.org/10.1016/j.buildenv.2004.01.021
http://dx.doi.org/10.1016/j.autcon.2004.03.002
Sun, J.; Tang, G.; Zhang, L.; Li, N. 2002. An efficient solu-
Dziugaite-Tumeniene, R.; Jankauskas, V.; Motuziene, V. 2012.
tion method for predicting indoor environment of build-
Energy balance of a low energy house, Journal of Civil
ings with complex geometric configuration, Building and
Engineering and Management 18(3): 369–377.
Environment 37(10): 915–922.
http://dx.doi.org/10.3846/13923730.2012.691107
http://dx.doi.org/10.1016/S0360-1323(01)00090-7
Journal of Civil Engineering and Management, 2016, 22(3): 333–345 345

Tang, R.; Meir, I. A.; Wu, T. 2006. Thermal performance of northern climate, Journal of Civil Engineering and Man-
non air-conditioned buildings with vaulted roofs in com- agement 18(6): 828–833.
parison with flat roofs, Building Environment 41(3): 268– http://dx.doi.org/10.3846/13923730.2012.720938
276. http://dx.doi.org/10.1016/j.buildenv.2005.01.008 Woo, J.; Diggelman, C.; Abushakra, B. 2011. BIM-based en-
Venckus, N.; Bliudzius, R.; Poderyte, J.; Burlingis, A. 2012. ergy mornitoring with XML parsing engine, in 28th In-
The heating load determination of low energy buildings in ternational Symposium on Automation and Robotics in
Construction, 2011, Seoul, Korea, S16-1: 544–545.

Karam KIM. PhD Student majoring Architectural Engineering at Kwangwoon University located in Seoul, South Korea. He is an
Academic Member of BuildingSMART Korea. His research interests include data interoperability using BIM technology, semantic
web, and IFC schema in construction engineering.

Jungho YU. PhD, Associate Professor of Architectural Engineering at Kwangwoon University located in Seoul, South Korea. He
is a member of various institute related to construction industry including Building SMART Korea. His research interests include
construction IT and decision analysis in construction management.