See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/334448038

An IFC data preparation workflow for building energy performance simulation

Conference Paper · July 2019

DOI: 10.35490/EC3.2019.188

CITATION READS

1 949

4 authors:

Kyriakos Katsigarakis Georgios Nektarios Lilis

University College London University College London
24 PUBLICATIONS 67 CITATIONS 42 PUBLICATIONS 238 CITATIONS

SEE PROFILE SEE PROFILE

Georgios I. Giannakis Dimitrios Rovas

Technical University of Crete Technical University of Crete
48 PUBLICATIONS 242 CITATIONS 91 PUBLICATIONS 1,918 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Building as a Service (Ecosystem) View project

OPTEEMAL View project

All content following this page was uploaded by Georgios Nektarios Lilis on 13 July 2019.

The user has requested enhancement of the downloaded file.

 2019 European Conference on Computing in Construction
Chania, Crete, Greece
July 10-12, 2019

AN IFC DATA PREPARATION WORKFLOW

FOR BUILDING ENERGY PERFORMANCE SIMULATION

Kyriakos Katsigarakis1, Georgios Nektarios Lilis1, Georgios Giannakis1 and Dimitrios Rovas2
1Technical University of Crete, Greece
2University College London, London, UK

Abstract research studies focus on BIM-based BEPS models gen-

eration (Andriamamonjy et al. 2018, Thorade et al. 2015,
Accurate building energy performance simulation model
Wimmer et al. 2015, Giannakis et al. 2015).
generation from IFC data files require the input IFC data
to be correct and complete. To ensure such data qual- BIM is an object-oriented digital representation of a build-
ity conditions, methods integrating IFC data correctness ing, which also can capture part of the information re-
and completeness checks, under a common architecture, quired for generating a BEPS input data file. Relevant
are introduced. IFC data correctness is supported by the open-BIM data schemas include the Industry Foundation
use of a dedicated geometric error detection tool, which Classes (IFC) (ISO 16739-1 2018) and the green-building
identifies and reports back to the building designer, errors XML schema (gbXML). Concerning building geometry,
affecting the building’s 2nd -level space boundary topology. IFC supports three dimensional representations, while
IFC data completeness is performed by the use of a ded- gbXML supports only planar geometry. For this reason, a
icated for building energy performance simulation Model plethora of recent studies focuses on developing a method-
View Definition schema. The quality checking results are ology to automatically generate geometry inputs for BEPS
reported in a BIM collaboration file format. from IFC files. Additionally, IFC appears to be a more
suitable choice as its rich content enables interoperabil-
Introduction ity among different software environments and can easily
be updated following a building’s life cycle (Hitchcock &
The accuracy of Building Energy Performance Simula- Wong 2011).
tion (BEPS) results is determined by the quality of its
However, applying an automated data transformation from
input data, mainly comprising the building geometry, its
IFC data to input data of BEPS tools is not a straightfor-
construction’s materials, internal loads, HVAC systems
ward but a tedious task often due to imperfections of the
and components, weather data, operating strategies and
provided IFC files. Although commercial BIM authoring
schedules, and simulation specific parameters (Maile et al.
tools (RevitTM , ArchicadTM , VasariTM ) support exporta-
2007).
tion of IFC files, their quality is not acceptable in the con-
In current practice, to develop a BEPS model, modelers text of BEPS input data generation. Even RevitTM , and
gather and combine 2D drawings such as Architectural and its IFC4 Design Transfer View (DTV) exporter, which
Mechanical Electrical Plumping (MEP) plan views, ma- currently seems to be the most mature tool, is far from
terial data and other information, and manually transform being perfect: while IFC schema can capture data about
them into the specific input data, required by the respective material thermal properties, internal gains (schedules and
BEPS engine. This process has two strong weaknesses: a) densities) and 2nd -level space boundaries (Bazjanac 2010),
it is very time-consuming, often requiring more time than requested from the BEPS perspective, relevant information
is available due to project’s deadlines; and b) it is a non- is missing or incorrect in the exported IFC4 DTV files.
standardized process that produces BEPS models whose
results can significantly vary from one modeler to an- Proposed IFC4 workflow
other according to their experience (Berkeley et al. 2014),
even given the same initial building design information. To examine and eliminate these IFC data imperfections,
A methodology for automated creation of BEPS models in this work an IFC4 data workflow is introduced. This
could make the BEPS modeling process much more ex- workflow aims at delivering error-free IFC4 files ready for
pedient and therefore, less vulnerable to modeling errors. BEPS input data generation. The workflow starts with the
Building Information Models (BIM) are information-rich BIM preparation and the IFC data exportation processes
repositories that could be used to streamline and expe- (first block of Figure 1). Since the lack of training of the
dite the collection of such information. Hence, recent BIM designer contributes to data inaccuracies, a set of de-
sign guidelines have been developed to help the designer tool. For its development, other supporting tools, relevant
to define properly all the requested information during the libraries and an MVD XML for conventional BEPS have
BIM preparation process. After the BIM data are prepared, been created. Other attempts in this domain have been
the IFC data exportation process follows. However, due to primarily focused on developing a MVD XML for con-
issues mentioned above, the exported IFC file is of poor ventional and advanced BEPS (Pinheiro et al. 2018). The
quality (as for the 2nd -level space boundary information) initial processes and the three main stages of the workflow
or lacks valuable information (as for the material thermal are described analytically in the following sections.
properties and internal gains); for the latter, RevitTM ’s
IFC4 DTV exporter has been modified to correctly popu- IFC data preparation and exportation
late such information. To help the BIM designer on creating an error-free (or
error-reduced) BIM, from which the maximum level of
IFC data Design BEPS information can be exported, a document that
preparation process presents a set of design guidelines as they apply for
and RevitTM has been developed – similar guidelines can be
exportation IFC exporter
developed for other BIM authoring tools.
The guidelines are provided as a set of modelling rules be-
IFC
file longing to two broad categories: Static and Dynamic Data.
Static data rules cover the building geometry, construction
1 IFC data materials, glazing information, building services, and so
BCF
geometric GED tool file forth, while dynamic data rules cover the specification
check
of time-varying parameters such as occupancy schedules,
Geometrically internal gains schedules, use of equipment, or occupant
correct IFC file
actions. A subset of these guidelines are briefly presented
2 IFC data below.
enrichment CBIP tool
Note here that concerning HVAC systems data representa-
tion, while IFC is the most widely used openBIM schema,
Enriched
IFC file
it suffers from limitation in the description of HVAC sys-
tem in the context of BEPS. Commonly, HVAC modelling
3 IFC data in BEPS engines require further information than what
completeness MVD tool
check is available in an IFC file. Hence, the presented BIM
design guidelines, are mostly focused on building energy
BEPS-ready demands estimation aspects.
IFC file
Spaces and HVAC zones
Figure 1: Overview of the proposed IFC data workflow Accurate heating- and cooling-loads simulation can only
be accomplished when spaces are created in all internal
After the BIM preparation and IFC exportation processes, building closed areas to account for the entire volume of
the IFC4 data workflow continues in three main stages. the building model. Spaces should be defined for each
During the first stage (IFC data geometric check) the ar- closed area of the building, including unoccupied closed
chitectural geometric content of the IFC file is examined air volumes such as plenums. Following the definition of
using the GED tool (Lilis et al. 2015), for possible er- spaces, a checking that they are properly bounded must be
rors which affect the BEPS model generation process and performed using floor plans and section views, to guaran-
if found they are reported in BIM Collaboration Format tee that a) there are no overlaps between the space volume
(BCF). In the second stage (IFC data enrichment) the IFC and the attached building elements, and b) there are no
file is enriched using the Common Boundary Intersec- void volumes (common case when the space volume is not
tion Projection (CBIP) tool with the necessary for BEPS attached to building elements).
2nd -level space boundary information. Similar enrichment HVAC zones comprise one or more spaces that have com-
tools, such as the SBT tool (Rose & Bazjanac 2015), have mon environmental or design requirements. Spaces that
been developed using graph-based methodologies. In the are on different levels can be added to the same zone.
final stage (IFC data completeness check) several checks Zones are rather important for BEPS as they define the
are performed on the IFC data to ensure acceptable quality granularity of the approximation. Grouping spaces into
for BEPS input data file generation. The checks are per- zones in RevitTM is required to ensure manageable com-
formed using the Model View Defintion (MVD) Checking
plexity of the BEPS model. When a space has not been (densities), but not modify existing schedules, since these
assigned to a zone, a new zone for this space is defined schedules have been defined according to relevant stan-
automatically. dards (ASHRAE 90.1 2010); if a new schedule is required
the modeler must define a user-defined schedule and edit
Material thermal properties its values properly.
In a broad sense, two main material categories exist:
IFC exporter modification
Opaque and Transparent. Knowledge of material thermal
property values is required in the modelling of the various To overcome limitations of the default Revit IFC4 DTV
energy transfers this includes absorbed solar shortwave exporter on exporting information about thermal proper-
radiation flux, conduction heat transfer through walls, con- ties of each building entity’s construction material and
vective transfer to the space (or outside) air and fraction of internal gains (schedules and densities), which have been
sunlight transmitted to building interiors. properly set following the aforementioned guidelines, and
For Opaque materials, knowledge of the following thermal to guarantee correct exportation of such information, the
properties is required: Thickness, Conductivity, Density, default exporter (which is an open source component) has
Roughness, Specific Heat, Thermal absorptance, and Solar been modified to create a bespoke OptEEmAL version
absorptance. Transparent materials require definition of that addresses these shortcomings. The OptEEmAL IFC
the following parameters: U-factor and Solar Heat Gain Exporter supports the latest version of RevitTM (2018.3).
Coefficient (SHGC). In Revit, these material properties
Exportation setup
can be defined by the designer and can be subsequently
accessed through the Revit API. To define and retrieve Assuming that the BIM data insertion according to the
such information, the BIM designer must activate the ther- guidelines and the OptEEmAL IFC exporter’s installa-
mal properties population for each material through the tion have been accomplished, the BIM designer proceeds
"Thermal" tab. with the exportation setup where few of the defaults IFC-
Most IFC-to-BEPS methods currently use construction exporting options are modified properly to retrieve the
and material thermal properties of building elements from requested information.
fixed predefined libraries (Thorade et al. 2015, Wimmer The first modification is the rooms exportation’s deacti-
et al. 2015). However, as mentioned above, this informa- vation. Utilizing the Revit IFC4 DTV exporter and/or its
tion can be explicitly prescribed in the RevitTM model, modified version, both spaces and rooms are exported as
and can also be represented in IFC (e.g. IfcThermalCon- IfcSpace instances. While spaces definition throughout
ductivityMeasure for thermal conductivity). Surprisingly, the models is prerequisite, rooms definition is not: in-
the default Revit IFC-exporter is not able to export such formation that a room object could provide in not usable
information. in the context of BEPS. Hence, to avoid duplication of
IfcSpaces’ geometry representation, the designer must de-
Internal gains and operation schedules activate the rooms’ exportation. To this direction, designer
Building equipment’s operation (artificial lighting and must change the IFC options by editing the rooms RevitTM
electrical equipment), as well as the occupants’ presence category and setting "Not Exported" in the respective "IFC
and the building infiltration rate, contribute to the overall Class Name".
internal gains and can be represented as time-varying func- Setting the rooms as not exported elements, the file is ready
tions, called herein internal gains schedules and assigned to be exported: the IFC4 Design Transfer View is selected
to every building’s space. as MVD, while a) the RevitTM property sets and b) the IFC
The IFC schema includes classes for representing the in- common property sets exportation are activated.
ternal gains schedules (IfcTimeSeries). Objects of these
classes can be attached to individual space instances. How- IFC data geometry check
ever, concerning the existing BIM-authoring tools, these After the IFC file export geometric errors that affect the
schedules can be selected from predefined variants. In IFC data enrichment stage (stage 2 Figure 1), (Lilis et al.
RevitTM , such information can be defined in building 2015) must be detected automatically and corrected man-
and/or space scale as follows: a) define the building type, ually. These errors belong to a subset of all possible errors
which sets a predefined template for the internal gains in IFC data files (detected by commercial IFC data check-
(schedules and densities), and its infiltration class; b) de- ing software such as Solibri Model Checker, TEKLA and
fine the people, artificial lighting and electric equipment others) and a dedicated Geometric Error Detection (GED)
(density and schedules) at each space by selecting the space tool have been developed for their detection.
type. At this point, it is worth-mentioning that the de- Errors that are detected by the GED tool belong to mul-
signer is allowed to edit/modify non-schedule parameters
tiple categories. For example, an architectural construc- tionally, most IFC exporters do not or export this topology
tion (e.g. wall, slab, etc.) that intersects with a building or they export it with errors.
space volume, renders the 2nd -level space boundary sur-
face related to this construction and space undetectable I. Architectural element set II. 2nd-level space boundaries

(see example in part I of Figure 2). Two intersecting wall CBIP External Virtual
opening volumes next to each other, attached to the same Space space
boundaries boundaries
building space (see example in part II of Figure 2), pro-
duce 2nd -level space boundaries which also intersect with Internal space boundaries
III. CBIP tool results
each other (space boundary surface duplication). For all of
2nd-level space boundary partitions
these reasons, methods for detection and semi-automatic of a slab of a real building model.
correction of these inaccuracies have been proposed (Lilis Detected by CBIP tool.
Displayed using Solibri Model
et al. 2015). Viewer.

I. Space - Wall clash II. Wall Opening – Wall Opening clash

Figure 3: 2nd -level space boundary surface generation of
CBIP tool
Space
Space 1 Space 2 Consequently, to produce this topology in a consistent and
Intersection
precise way, CBIP tool has been developed (Lilis et al.
2017), which performs the necessary geometric transfor-
Space 1 Space 2
Space boundary duplication
mation operations on the existing IFC architectural geo-
III. GED tool results
metric content (part I of Figure 3), calculates the 2nd -level
Space - space clash space boundary surfaces (part II of Figure 3) and enriches
of a real building
model. the input IFC file with the required geometric data of these
Detected by GED tool. surfaces and their associations to building space volumes
Reported in BCF.
and material constructions. As illustrated in the example
Viewed using
BIMcollab
of part III of Figure 3, CBIP correctly decomposes the
on AutoDesk’s floor surface into multiple sub-surfaces depending on the
Navisworks. location of the building spaces underneath the slab.
Although the geometric content of the enriched IFC file
Figure 2: Geometric errors detected by GED tool might be adequate for BEPS model generation (the neces-
sary 2nd -level space boundaries are present with the correct
The GED tool reports the detected errors back to the de- associations to related building elements), additional in-
signer in a BCF file format (part III of Figure 2). The BCF formation related to thermal properties of the materials
error report file contains the 3D geometric definitions of and time schedules of entities which change state (peo-
the architectural elements involved in a geometric error as ple, devices, openings, etc.), required for BEPS model
well as the set of surfaces and line segments indicating generation, might be missing. Consequently, an IFC data
at the exact location of the error in the building’s three completeness checking stage, following the model view
dimensional space. definition approach, appears to be necessary, as described
in the following section.
IFC data enrichment
After the designer’s manual corrections of the aforemen-
IFC data completeness check
tioned remaining geometric errors, the exported IFC file For an IFC-to-BEPS model data transformation, presence
is ready to be enriched with the necessary for BEPS ge- of certain IFC data is prerequisite. More specifically, IFC
ometric content, which is essentially the 2nd -level space data have to satisfy the exchange requirements of the BEPS
boundary surface topology (Bazjanac 2010). This topol- input data generation; hence, certain subsets of the IFC
ogy contains surfaces through which thermal energy flows schema are validated by evaluating the IFC entities and
among internal building spaces and the building environ- their attributes in terms of existence, correctness, unique-
ment (air or ground) and are required in order to assess ness and conditional dependencies (Chipman et al. 2014).
total thermal energy balance of the building (see example There are two main approaches of developing an IFC val-
in part II of Figure 3). This surface topology cannot be idation tool. First using an IFC library, the developer can
obtained directly from the geometry of the architectural build static rules by querying the IFC entities and checking
elements (e.g. walls, slabs, etc.) as the building spaces their attributes. This approach has some weaknesses: a)
are not perfectly aligned vertically and horizontally. Addi- it does not support reusability of the defined rules; and b)
it requires advance programming skills to validate the re- are summarized as follows:
lations of the IFC entities under a certain IFC subset. The
second approach is the development of an IFC validation • Public constructor without arguments: instantiates an
tool that supports the Model View Definition (MVD) auto- object of an IFC class;
• Public constructor with arguments: the constructor
mated checking (Zhang et al. 2015). An MVD describes
includes all the defined properties (IFC attributes)
the subset of the IFC schema that is needed for a data-
as well as the properties which are defined in upper
exchange in a specific domain, which can be formalized
levels (IFC abstract classes);
using the mvdXML open standard that has been released by • Public method to initialize the properties of an IFC
buildingSMART International. The IFC validation tools class;
which support MVD checking use the mvdXML format to • Public method to initialize the collections set of the
import the IFC subsets known as templates and the defined inverse properties;
rule sets known as concept to perform an automated vali- • Public getters/setters for each property;
dation of the IFC data. To develop such a tool, in this work • Public method to generate the step line of an IFC
some supporting tools and libraries have been developed, class.
described in the following sections.
Furthermore, the EXPRESS Parsing Tool generates the
EXPRESS Parsing Tool IFC factory class which allows the instantiation of the IFC
objects by using the IFC class names.
The proposed EXPRESS Parsing Tool is a standalone pro-
gram written in Java using the code model library and is IFC Library
suitable to generate IFC java classes from the correspond-
The proposed IFC Library provides the possibility to parse
ing EXPRESS files which have been released by build-
efficiently the IFC-SPF data and to instantiate in-memory
ingSMART International. The tool can parse successfully
the IFC model of the corresponding IFC file. An em-
the most frequently used IFC releases, from IFC2x3 TC1
bedded API is used for the loaded objects manipulation,
to IFC4x1 final. In the first step, the EXPRESS data
providing the developers with useful functionalities such
are transformed into in-memory objects. Consequently,
as adding new IFC instances, modifying existing instances
the defined mapping rule set is applied on top of this in-
or querying the IFC model. The core of this library comes
memory representation to initialize the corresponding Java
from the output of the EXPRESS Parsing Tool and is or-
code model. The mapping rules are summarized in Table
ganized in multiple packages to handle effectively the sup-
1.
ported versions of the EXPRESS schema. The current
version of the IFC library can handle a wide range of
Table 1: Mapping rules between EXPRESS schema and different IFC schemas (IFC2x3 TC1 to IFC4x1 final). Ad-
Java model ditionally, the IFC library can initialize the properties of
the inverse relations by adding the instance of a class to
EXPRESS Model Java Model the set collection of the inverse connected property. The
ENTITY Java Class initialization of the inverse relations is archived by calling
SELECT Java Interface a specific public method as listed next:
ABSTRACT SUPERTYPE Java Abstract class
Java Class Listing 1: Part of IfcMaterialProperties class
public I f c M a t e r i a l P r o p e r t i e s (){
Enumeration class }
Custom list wrapper
public void i n v e r s e ( ) {
Custom array wrapper i f ( t h i s . m a t e r i a l s ! = n u l l &&
TYPE Boolean wrapper class t h i s . m a t e r i a l . g e t H a s P r o p e r t i e s ( ) != n u l l ){
t h i s . m a t e r i a l . g e t H a s P r o p e r t i e s ( ) . add ( t h i s ) ;
Logical wrapper class }
Real wrapper class }
Integer wrapper class
String wrapper class Listing 2: Part of IfcMaterialDefinition class
Binary wrapper class public I f c M a t e r i a l D e f i n i t i o n ( ) {
this . associatedTo =
INVERSE Set Collection new HashSet < I f c R e l A s s o c i a t e s M a t e r i a l > ( ) ;
this . hasExternalReferences =
new HashSet < I f c E x t e r n a l R e f e r e n c e R e l a t i o n s h i p > ( ) ;
this . hasProperties =
The Java code model has a built-in method to generate the new HashSet < I f c M a t e r i a l P r o p e r t i e s > ( ) ;
defined IFC classes and to organize them into packages }
based on the version of the given EXPRESS schema. Each public void i n v e r s e ( ) {
IFC Java class implements some standard methods which }
Using the inverse set collections, it is possible to get the is performed by the CBIP tool. The result of the CBIP
IfcMaterialProperties instance from the inverse method tools execution is the enrichment of the IFC file with rele-
getHasProperties() of the IfcMaterial class. In this exam- vant 2nd -level space boundary information (IfcRelSpace-
ple, the IfcMaterial class is subtype of the IfcMaterialDef- Boundary2ndLevel class population).
inition. The inverse method is called for each IFC instance The 2nd -level space boundary content of the enriched IFC
after the deserialization process. file is checked by the model view checking process, using
the following four checking rules, which are implemented
MVD Checking
under the same concept using multiple template rules:
One of the main goals of the MVD specification is to help
the community develop BIM tools that allow the automatic 1. The existence of the 2nd -level space boundaries is
validation of IFC data against rule sets that are based on checked by examining the presence of the IfcRelSpace-
the mvdXML standard. Mainly, three steps are needed to Boundary2ndLevel class instances in the enriched IFC
achieve automatic validation of IFC data: a) the prepa- file.
ration of the rule set (mvdXML) using buildingSMART
certified tools such as the ifcDoc (see figure 4); b) the 2. For each IfcRelSpaceBoundary2ndLevel populated in-
model checking service that applies the given rules to the stance, the presence of its related building element is
loaded IFC instances; c) the report generation service that examined.
generates reports based on open standards such as the BCF.
3. For each IfcRelSpaceBoundary2ndLevel instance, the
For the creation of valid mvdXML files the ifcDoc software
related building elements type, the value of the Ex-
is used. This software has been developed by buildingS-
ternalOrInternal property, the corresponding space
MART International in order to improve the consistent
boundary (if the space boundary is INTERNAL) and
and computer-interpretable definition of the MVD speci-
the parent space boundary (if the space boundary refers
fication. The user can easily create a custom model view
to an opening) are examined.
and assign new concepts under it. Each concept has an
applicable entity (IFC class), a connection to an existing 4. For each IfcRelSpaceBoundary2ndLevel instance the
template (subsection of the IFC schema) and the applied relating space index, is checked.
rules based on logical operations (IFC entities or attributes
that are checked by type or by value under certain condi- 5. Finally, the last rule checks if there is an IfcRelSpace-
tions). Boundary2ndLevel instance with ExternalOrInternal
The proposed MVD checking tool is a standalone service property value "EXTERNAL_EARTH".
written in pure Java without having external dependencies
to third party libraries. The main goal was to achieve If any missing or invalid information is detected the cor-
a clean end-to-end design where the different parts are responding template rule fails to validate the defined pa-
separated based on their functionalities. Thus, the EX- rameters. These parameters are summarized for different
PRESS Parsing Tool is used to generate the IFC Library space boundary types (internal, external, corresponding
which in its turn is used internally by the model view and parent) and related building elements (wall, slab, door,
checking process. There are two main features that the window, plate, opening and virtual element) in Table 2.
checking process uses from the IFC Library: a) the IFC-
SPF serialization/deserialization feature to load the IFC
Table 2: Space boundary checking rules
data in-memory and to perform queries; and b) the inverse
connections of the IFC entities to validate the existence of Applicable Boundary Corresponding Parent
an IFC schema’s subset starting from an applicable IFC IFC Entity Condition Space Boundary
root entity. The MVD checking tool takes as an input IfcWall Internal Missing -
the IFC-SPF data as well as the MVD-XML data. The IfcSlab Internal Missing -
output of the tool is the boolean result of the validation External - Missing
for each given concept. The serialization of the results IfcDoor Internal - Missing
Internal Missing -
can be represented in a graphical environment or using
External - Missing
open standards such as the BCF that supports workflow
IfcWindow Internal - Missing
communication in BIM processes. Internal Missing -
From a BEPS viewpoint, the partition of building con- External - Missing
IfcPlate
struction (interior/exterior walls, floors, roofs) and open- Internal - -
ing (door, window) surfaces into 2nd -level space boundary IfcOpening External - -
(Bazjanac 2010) surfaces is a prerequisite. This process IfcVirtualElement External - -
 Figure 4: ifcDOC tool

As mentioned above, other IFC data that are required in

Table 4: Opaque material property rules
the context of BEPS include the space occupants pres-
ence, artificial lighting, electrical equipment operation Property name Property type
(which act as thermal sources in terms of BEPS) and re- SpecificHeatCapacity IfcPropertySingleValue
lated operation schedules. These data are exported by the ThermalConductivity IfcPropertySingleValue
modified IFC RevitTM exporter by populating a new Pset MassDensity IfcPropertySingleValue
class, named Pset_Space InternalGainsDesign, assigned
to each IfcSpace class. The properties of this new class
are checked against the property name and the entity type Table 5: Transparent material property rules
as listed in Table 3. Property name Property type
Visual Light Transmittance IfcPropertySingleValue
Table 3: Space checking template rules of Pset SpaceIn- Solar Heat Gain Coefficient IfcPropertySingleValue
ternalGainsDesign Heat Transfer Coefficient (U) IfcPropertySingleValue
Thermal Resistance (R) IfcPropertySingleValue
Property name Entity type
OccupancySchedule IfccIreggularTimeSeries
LightingSchedule IfccIreggularTimeSeries Conclusions
EquipmentSchedule IfccIreggularTimeSeries
HeatGainLighting IfcPropertySingleValue The automated BEPS model generation from IFC data has
HeatGainEquipment IfcPropertySingleValue received lately considerable attention. However, imperfec-
AreaPerOccupant IfcPropertySingleValue tions of the provided IFC files prevent the BIM-to-BEPS
HeatGainPerOccupant IfcPropertySingleValue data transformation process from being widely and auto-
InfiltrationRate IfcPropertySingleValue matically applicable.
The focus of this work has been the proposal of an IFC
BEPS require a number of properties to be assigned to ma- data quality checking workflow towards populating IFC
terials for various building constructions as well. These files ready for BEPS input data generation. The proposed
properties, depending on the material type, are: for opaque workflow consists of four main stages: a) data preparation
materials Conductivity, Density and Specific Heat; for and exportation; b) geometric data correctness check; c)
transparent materials U-factor and Solar heat gain coeffi- data enrichment; and d) data completeness check.
cient (SHGC). Two new MVD concepts depending on the Describing these stages concisely, in the first stage, de-
building element type are applied: a) Opaque materials sign guidelines along with a modified IFC exporter are
(e.g. walls, floors, roofs); and b) Transparent materials provided to enable exportation of information about mate-
(e.g. windows, plates, doors). rials thermal properties and internal gains. In the second
The properties under the opaque and the transparent mate- stage, geometric errors that affect the correct 2nd -level
rials classes are checked against the property name and the space boundaries population are automatically detected
entity type as listed in Table 4 and Table 5, respectively. and reported in a BCF file. These errors are corrected
 manually by the designer and an geometric error-free IFC Hitchcock, R. & Wong, J. (2011), Transforming IFC
that includes information about materials thermal prop- Architectural View BIMs for Energy Simulation, in
erties and internal gains is exported to be processed by ‘Proceedings of the 12th IBPSA Building Simulation
the third stage; here, the IFC data are enriched with the Conference’, Sydney, Australia, 14-16 November 2011,
2nd -level space boundaries data. In the final stage, a data pp. 1089–1095.
completeness check is performed for the enriched IFC file
to ensure that the BEPS information that the IFC schema ISO 16739-1 (2018), ‘Industry Foundation Classes (IFC)
could capture is present. for data sharing in the construction and facility manage-
ment industries – Part 1: Data schema’, International
The presented workflow has been focused strictly on data
Organisation for Standardisation .
preparation for building energy demands estimation as-
pects, since HVAC data inclusion have not been thoroughly Lilis, G. N., Giannakis, G. I. & Rovas, D. V. (2017), ‘Auto-
investigated. A concrete implementation of such data in matic generation of second-level space boundary topol-
the proposed workflow remains the main subject of future ogy from IFC geometry inputs’, Automation in Con-
work. struction 76, 108–124.

Acknowledgments Lilis, G. N., Giannakis, G. & Rovas, D. (2015), Detection

and semi-automatic correction of geometric inaccura-
This research has been partially funded by the European cies in IFC files, in ‘Proceedings of the 14th IBPSA
Commission H2020-EeB5-2015 project “Optimised En- Building Simulation Conference’, Hyderabad, India, 7-
ergy Efficient Design Platform for Refurbishment at Dis- 9 December 2015, pp. 2182–2189.
trict Level" under contract #680676 (OptEEmAL).
Maile, T., Fischer, M. & Bazjanac, V. (2007), ‘Building
References energy performance simulation tools-a life-cycle and in-
Andriamamonjy, A., Saelens, D. & Klein, R. (2018), teroperable perspective’, Center for Integrated Facility
‘An automated IFC-based workflow for building energy Engineering (CIFE) Working Paper 107, 1–49.
performance simulation with Modelica’, Automation in
Pinheiro, S., Wimmer, R., ODonnell, J., Muhic, S., Baz-
Construction 91, 166–181.
janac, V., Maile, T., Frisch, J. & van Treeck, C. (2018),
ASHRAE 90.1 (2010), ‘Energy standard for buildings ‘MVD based information exchange between BIM and
except low-rise residential buildings’, ASHRAE/IESNA building energy performance simulation’, Automation
Standard . in Construction 90, 91–103.
Rose, C. M. & Bazjanac, V. (2015), ‘An algorithm to gen-
Bazjanac, V. (2010), Space boundary requirements for
erate space boundaries for building energy simulation’,
modeling of building geometry for energy and other
Engineering with Computers 31(2), 271–280.
performance simulation, in ‘Proceedings of the 27th
CIB W78 Conference’, Cairo, Egypt, 16-19 November Thorade, M., Rädler, J., Remmen, P., Maile, T., Wimmer,
2010. R., Cao, J., Lauster, M., Nytsch-Geusen, C., Müller, D.
& van Treeck, C. (2015), An open toolchain for gen-
Berkeley, P., Haves, P. & Kolderup, E. (2014), Impact of erating modelica code from building information mod-
modeler decisions on simulation results, in ‘Proceed- els, in ‘Proceedings of the 11th International Modelica
ings of the ASHRAE/IBPSA-USA Building Simulation Conference’, pp. 383–391. Versailles, France, 21-23
Conference’. September 2015.
Chipman, T., Liebich, T. & Weise, M. (2014), ‘Specifi- Wimmer, R., Cao, J., Remmen, P., Maile, T., ODonnel,
cation of a standardized format to define and exchange J., Frisch, J., Streblow, R., Müller, D. & van Treeck, C.
model view definitions with exchange requirements and (2015), Implementation of advanced bim-based map-
validation rules (b1. 0)’. ping rules for automated conversions to modelica, in
‘Proceedings of the 14th IBPSA Building Simulation
Giannakis, G., Lilis, G. N., Garcia, M., Kontes, G., Val- Conference’, pp. 419–426. Hyderabad, India, 7-9 De-
maseda, C. & Rovas, D. (2015), A methodology to au- cember 2015.
tomatically generate geometry inputs for energy perfor-
mance simulation from IFC BIM models, in ‘Proceed- Zhang, C., Beetz, J. & Weise, M. (2015), ‘Interoperable
ings of the 14th IBPSA Building Simulation Confer- validation for ifc building models using open standards’,
ence’, Hyderabad, India, 7-9 December 2015, pp. 504– Journal of Information Technology in Construction (IT-
511. con) 20(2), 24–39.

View publication stats