DANISH BUILDING RESEARCH INSTITUTE

AALBORG UNIVERSITY COPENHAGEN

Daylight calculations in practice

An investigation of the ability of nine daylight simulation
programs to calculate the daylight factor in five typical rooms
SBI 2013:26

SIMPLE ROOM

ROOM WITH OBSTRUCTION

DEEP ROOM

ROOM WITH LIGHT SHELF

ROOM WITH BORROWED LIGHT

Daylight calculations in practice

An investigation of the ability of nine daylight simulation programs to
calculate the daylight factor in five typical rooms

Anne Iversen
Nicolas Roy
Mette Hvass
Michael Jrgensen
Jens Christoffersen
Werner Osterhaus
Kjeld Johnsen

The project is supported by

SBi 2013:26
Danish Building Research Institute, Aalborg University 2013

Title
Subtitle
Serial title
Edition
Year of
publication
Authors
Language
Page numbers
References
Keywords

Daylight calculations in practice

An investigation of the ability of nine daylight simulation programs to calculate the daylight
factor in five typical rooms
SBi 2013:26
1st edition
2013
Anne Iversen, Nicolas Roy, Mette Hvass, Michael Jrgensen, Jens Christoffersen, Werner
Osterhaus, Kjeld Johnsen
English
54
Page 54
Daylight, Danish Building Regulations 2010, Danish Working Environment Authority,
daylight factors, daylight calculations, calculation programs

ISBN

978-87-92739-49-0

Cover

Authors

Publisher

Danish Building Research Institute, Aalborg University,

A. C. Meyers Vnge 15, DK-2450 Copenhagen SV
E-mail [email protected]
www.sbi.dk

This publication is covered by the Danish Copyright Act.

Content

Preface ............................................................................................................ 4
Main findings ................................................................................................... 5
Introduction ...................................................................................................... 8
Aim .............................................................................................................. 9
Workshop ..................................................................................................10
Content .................................................................................................10
Participants...........................................................................................10
The work of the core group .......................................................................11
Daylight demands ..........................................................................................12
Danish Building Regulations 2010 ............................................................12
Danish Working Environment Authority ....................................................13
Definition of daylight factor ........................................................................13
Methods for analysing the daylight factor .................................................14
Calculation methods ......................................................................................17
Accuracy of daylight calculation................................................................17
Split flux formula ...................................................................................17
Radiosity...............................................................................................18
Ray tracing ...........................................................................................18
Photon mapping ...................................................................................19
Sky types ..............................................................................................19
The model ............................................................................................20
Surface properties ................................................................................21
User expertise ......................................................................................22
Description of simulation programs ...............................................................23
Radiance ..............................................................................................23
Daysim .................................................................................................23
Desktop Radiance ................................................................................24
IESve ....................................................................................................24
DIALux..................................................................................................24
Relux ....................................................................................................24
Ecotect .................................................................................................25
VELUX Daylight Visualizer ...................................................................25
LightCalc ..............................................................................................25
Graphical presentation of the results.............................................................26
Simulations and analysis ...............................................................................27
Room types ...............................................................................................27
Simulation parameters ..............................................................................29
Guidance to result sheet simulations ........................................................30
ROOM 1 SIMPEL ROOM Simulation sheet ...................................32
ROOM 1 - SIMPEL ROOM Analysis sheet .......................................34
ROOM 2 DEEP ROOM Simulation sheet ......................................35
ROOM 2 DEEP ROOM Analysis sheet ..........................................38
ROOM 3 ROOM WITH OBSTRUCTIONS Simulation sheet .........40
ROOM 3 ROOM WITH OBSTRUCTION Analysis sheet ...............42
ROOM 4 ROOM WITH LIGHT SHELF Simulation sheet ..............45
ROOM 4 ROOM WITH LIGHT SHELF Analysis sheet ...................47
ROOM 5 ROOM WITH BORROWED LIGHT Simulation sheet ....49
ROOM 5 ROOM WITH BORROWED LIGHT Analysis sheet ........51
Assessment of the nine daylight simulation programs ..................................53
References ....................................................................................................54

Preface

This report presents the results of a project initiated by COWI and the Danish Building Research institute (SBi), Aalborg University. The project focused
on daylight calculations in practice and was conducted by researchers from
SBi, Aarhus University (AU-ASE) and the Technical University of Denmark
(DTU) as well as consultants and architects from COWI, VELUX, Alectia and
Arkitema. The following persons formed the core group:
Kjeld Johnsen, SBi Aalborg University
Anne Iversen, SBi Aalborg University
Mette Hvass, COWI
Michael Jrgensen, COWI
Nicolas Roy, VELUX
Jens Christoffersen, VELUX
Glenn Elmbk, Arkitema
Werner Osterhaus, Aarhus University
Steffen Petersen, Aarhus University/Alectia
Christian Anker Hviid, DTU/Alectia
The COWI foundation and VELUX supported the project. The aim of the project was to obtain a better understanding of what daylight calculations show
and also to gain knowledge of how the different daylight simulation programs
perform compared with each other. Furthermore the aim was to provide
knowledge of how to build up the 3D models that were to be daylightanalysed.
The report describes an investigation of the ability of nine simulation programs to calculate the daylight factor on a working plane in five typical
rooms.
The core group would like to thank the participants in the two workshops
held in relation to the project. The participants came from ARUP London,
the School of Architecture in Copenhagen, Rambll, Grontmij, Esbensen,
DTU, Niras, Alectia, Danske Ark, CCO Architects, Schmidt Hammer Lassen
Architects, Aarhus University, VHR Holding and the Danish Centre for Light.
A special thanks to the speakers; Nanet Krogsbk Mathiasen from the
School of Architecture Copenhagen, Francesco Anselmo from ARUP London, Glenn Elmbk from Arkitema, Steffen Petersen from Alectia/AU-ASE
and Mette Hvass from COWI.
Danish Building Research Institute, Aalborg University
Energy and Environment
September 2013
Sren Aggerholm
Research Director

Main findings

This report investigates the performance and accuracy of nine different

simulation programs commonly used by professionals and researchers in
the field of architecture and engineering. The programs are compared based
on their ability to calculate the daylight factor on a working plane and the
depth of daylight penetration.
The programs investigated are: Radiance, Daysim, VELUX Daylight Visualizer, DIAlux, Ecotect, Ecotect/Radiance, IESve, LightCalc and Relux. The
results from Relux are based on two approaches; radiosity and raytracing,
and therefore ten and not nine simulation programs are listed in the simulation and analysis sheets.
For the simulations five different types of rooms were investigated:
Room 1 Simple room
Room 2 Deep room
Room 3 Room with obstruction *)
Room 4 Room with light shelf
Room 5 Room with borrowed light
*) (meaning obstructed daylight access from buildings opposite)
The table below gives an overview of the capability for each simulation program to calculate the daylight factor for the five room types investigated.
Table 1: Capability of each program to calculate daylight factor levels in the five room types. Yes=capable, No=not capable.
Radiance

Desktop
Radiance

Daysim

1. Simple room

yes

yes

yes

2. Deep room

yes

yes

3. Room with obstruction

yes

4. Room with light shelve

5. Room with borrowed light

Velux
Daylight
Visualizer

DIALux

Ecotect

IESVE

LightCalc

Relux
Radiosity

Relux
Raytracing

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

no

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

no

yes

yes

yes

yes

yes

yes

yes

no

yes

no

no

yes

The programs compared in the analysis of the different room types corresponded to the ones with a yes in Table 1. For these programs, it was
found that the different simulation programs gave similar results, see Table
2.
The standard deviation of the average daylight factor for the different simulation programs ranged from 5.1 % to 13.6 %. The smallest standard deviations of 5.1 %; 6.6 % and 7.4 % were obtained for the most simple
rooms; Room 1, Room 2 and Room 4. Whereas the more complex rooms,
Room 3 and Room 5 had standard deviations of 10.3 % and 13.6 % respectively.

Table 2: Variation in average daylight factor [%] for the different simulation programs for each of the simulated
rooms.
Room 1

Room 2

Room 3

Room 4

Room 5

Light shelve**

Borrowed light***

Simple room

Deep room

Obstruction*

Max

3.5

2.1

0.9

2.6

2.4

Min

3.0

1.6

0.6

2.1

1.7

Average value

3.2

1.8

0.8

2.4

2.0

Standard deviation [DF %]

0.2

0.1

0.1

0.2

0.3

Standard deviation (+/-) [%]

5.1

7.4

10.3

6.6

13.6

Relative difference, max vs. min [%]

15.0

22.6

28.1

18.3

31.2

*Ecotect excluded
**LightCalc excluded
***Ecotect, LightCalc and Relux Radiosity excluded

In this report, the daylight penetration depth was defined as the distance
from the faade where a daylight factor of 2% is reached. Looking at the
daylight penetration depth simulated with the different simulation programs,
standard deviations in the range from 4.9% to 8.7% were found, see Table 3. The highest deviation was seen for the room with obstruction, which
can be attributed to the different ways of the simulation programs to deal
with obstructions. For the room with borrowed light, the daylight penetration
depth was not considered as this room describes a room located within the
building with no faade facing the exterior.
Table 3: Variation in distance [m] from facade, where 2% DF is obtained, for the investigated rooms simulated
with the applied simulation programs.
Room 1

Room 2

Room 3

Room 4

Room 5

Simple room

Deep room

Obstruction*

Light shelf**

Borrowed light***

Max

2.5

2.4

0.7

2.2

Min

2.1

2.1

0.6

1.8

Average value

2.3

2.2

0.7

2.0

Standard deviation [DF]

0.1

0.1

0.1

0.1

Standard deviation (+/-) [%]

6.4

4.9

8.7

6.7

Relative difference, max vs. min [%]

18.1

15.9

24.9

20.6

*Ecotect excluded
**LightCalc excluded
***Ecotect, LightCalc and Relux Radiosity excluded

Experience from the work reported in this report showed that the programs
Radiance, Desktop Radiance, Daysim, Velux Daylight Visualizer, DIALux,
IESve and Relux Raytracing were all capable of calculating the daylight factor levels in the five different room types within acceptable agreement.
Ecotect was capable of calculating the simple room, the deep room and the
room with light shelf. The program was not capable of calculating the room
with obstruction and the room with borrowed light. This fact was ascribed to
the simplified calculation engine within Ecotect (split flux method) which the
program applies as default for its calculations.

LightCalc was capable of calculating the daylight factor in the simple room,
the deep room and the room with obstruction within good agreement. LightCalc was not capable of calculating rooms with light shelf and borrowed
light. This is due to the fact that the calculation engine within LightCalc was
not developed to include light shelves, and is only capable of calculating one
room, and therefore cannot consider internally located rooms.

Relux radiosity was capable of calculating the four rooms; simple room,
deep room, room with obstruction and room with light shelf. Relux radiosity
could not calculate a room with borrowed light. This is due to the fact that it
is only possible to place window openings in the facades facing the exterior.
Therefore it was not possible to create a room geometry with a window
placed internally in the building.
Even though Table 1 states that IESve was capable of calculating all room
types, one should be aware that simulations made with the highest accuracy
in IESve, have the amount of ambient bounces set to 3. For simulation programs such as Daysim and Radiance, the amount of ambient bounces was
set to 7. This means that calculations made with IESve underestimates the
daylight levels in more complex scenes, such as the room with obstruction,
where the shadow effect from the obstruction is considered in the calculations. If rooms with obstruction are to be included in IESve and simulated
correctly, the designer should therefore be aware of the need to change the
default settings in the program.
The report describes a number of possible sources for these deviations.
These sources can be due to the structure of the calculation programs, their
capacity to simulate complex rooms/surroundings, as well as the expertise of
the user.

Introduction

During the last years, there has been a growing interest in buildings lit by
daylight. Apart from the aesthetic, visual and spatial qualities obtained from
daylight, daylight also plays an important role in the efforts to minimise energy consumption for artificial lights.
Therefore, daylight calculations are included in the design process as a tool
to predict daylight levels in buildings. On the marked today, different daylight
simulation programs exist, and these programs are being used by both architects and engineers. Experience has shown that results for the same
room, obtained from two daylight simulation programs can give different results. This can be due to restrictions in the program itself and/or be due to
the skills of the persons setting up the models. This is crucial as daylight calculations are used to document that the demands and recommendations to
daylight levels outlined by building authorities are being met, as well as to
document the need for additional artificial light. A misleading daylight calculation can have consequences for the design layout of the building and for
considerations on how to make use of the space in question. It is therefore
important that the results of the daylight simulations are acceptable and that
you as a user know the limitations of the tool you are applying.
The project stems from a specific example involving COWI and SBi. Daylight
simulations made in two different programs gave very different results.
Therefore, collaboration was initiated enlightening the reasons for these differing results. Focus of the collaboration was daylight simulations in practice
and entailed knowledge of program-related limitations as well as knowledge
of model-specific conditions The work was launched through a workshop
with participants from the building industry and research institutions, who all
had experience of daylight simulations and measurements in rooms and
buildings. A core group was formed at the workshop, and this core group
made further work based on the experience gained at the workshop. The
core group made the calculations for this report.

Aim
The aim of the project was to obtain a better understanding of what daylight
simulations show and also to gain knowledge of how the different daylight
simulation programs perform compared with each other. Furthermore, the
aim was to provide knowledge of how to build the 3D models that were to be
daylight-analysed.
It is an overall objective to enhance the quality of consultancy within the field
of daylighting with a greater knowledge of the available daylight simulation
programs and the accuracy of the simulated,.
The project is divided into three phases:
Part 1 - Identification
Conducting the workshop. Identification of the problem. Choosing room
types and calculation programs
Part 2 Calculation and analysis
Analysis of the simulation results from the first part of the work. Description
of the different daylight simulation programs
Part 3 - Conclusion
Evaluation of the different programs
Conclusion

Part 1 - Identification

Part 2 Calculation and analysis

Part 3 - Conclusion

Figure 1: The three phases of the project

Workshop
In order to highlight and clarify the projects, a workshop was held that included engineers, architects, program developers and others with experience of simulation and measurement of daylight in buildings. At the workshop, the participants were given typical examples of daylight simulations as
well as presentations about the calculated and experienced light. With basis
in this common denominator, the participants were divided into smaller
groups where they discussed the daylight challenges they experienced in
everyday life.

Content
The workshop was divided into two parts; 1) a morning with presentations
and 2) an afternoon with discussions in groups. Nanet Krogsbk Mathiasen
from the School of Architecture gave a presentation about the experienced
light and Francesco Anselmo fom ARUP London gave a presentation about
the simulated light. There were three real life cases where problems with
daylight simulations had had a consequence on the layout of a building. The
cases were presented by Glenn Elmbk from Arkitema, Steffen Petersen
from Alectia/AU-ASE and Mette Hvass from COWI.

Participants
SBi and COWI organised the workshop and arranged it in collaboration with
VELUX. Other participants in the workshop were: ARUP London, School of
Architecture Copenhagen, Rambll, Grontmij, Esbensen, DTU, Niras, Alectia, Danske Ark, CCO Architects, Schmidt Hammer Lassen architects, Aarhus University, VKR Holding and the Danish Centre for Light.
Some of the key points from the presentations were:
Daylight quality, What is good daylight quality?
The effect of daylight on our physical and mental well-being
Developers and architects should present their wishes to daylight in a
building in writing
Choice of building details, materials and surfaces and their influence on
daylight
The experienced and calculated light, a holistic approach to daylight calculations
How do we communicate daylight as well as the experienced light and the
simulation results?
At what stage in the design process do we employ daylight simulations?
The influence of the parameters used in the daylight simulation programs
What does the different daylight simulation results tell us about the daylight condition within the room; daylight factor, maximum, minimum and average values, glare, uniformity, light on vertical surfaces

10

The work of the core group

The core group was formed at the workshop and consisted of participants
from universities, engineering, and architectural firms and program developers.
The core group selected room types for daylight simulations and simulation
programs. The group performed the daylight simulations and documented
the results in this report. The simulations were based on consistent input parameters as described in this report.
The working process of the core group included joint meetings and continuous communication to consider the results.

The core group consisted of:

Kjeld Johnsen
Anne Iversen
Mette Hvass
Michael Jrgensen
Nicolas Roy
Jens Christoffersen
Glenn Elmbk
Werner Osterhaus
Steffen Petersen
Christian Anker Hviid

The simulations were made by:

Nicolas Roy: Radiance, Daysim, Velux Daylight Visualizer
Anne Iversen: DIALux
Michael Jrgensen: Ecotect, Ecotect/Radiance, IESVE
Steffen Petersen: LightCalc
Werner Osterhaus: Relux

11

Daylight demands

Daylight simulations are applied to demonstrate that the requirements or

recommendation on the daylight factor given by the Danish Building Regulations is maintained. The following text contains the legal requirements stipulated in the Building Regulations from 2010; the recommendations given by
the Danish Working Environment Authority; and a definition of the daylight
factor.

Danish Building Regulations 2010

The quoted sections are from the English version of the Danish Building
Regulations, from December 2010 (Danish Building Act, 2010). The additional text is from the updated Danish Building Regulations from January
2013.

6.5.2 Daylight (1)

Workrooms, occupiable rooms in institutions, teaching rooms, dining areas,
hereinafter called workrooms etc., and habitable rooms and kitchen must
have sufficient daylight for the rooms to be well lit. Windows must be made,
located and, where appropriate, screened such that sunlight through them
does not cause overheating in the rooms, and such that nuisance from direct
solar heat gain is avoided.

Guideline: (6.5.2 (1))

In workrooms etc., habitable rooms and kitchen the daylight can usually be
taken to be sufficient if the glazed area of side lights corresponds to a minimum of 10% of the room floor area or, in the case of rooflights, no less than
7% of the room floor area, assuming that the light transmittance of the glazing is no less than 0.75. The 10% and 7% are guidelines assuming a normal
location of the building and a normal layout and fitting out of the rooms. If the
type of window is not known at the time of design, the frame clear area can
be converted to the glazed area by multiplying the clear frame area by a factor of 0.7. The glazed area must be increased in proportion to any reduction
in light transmittance (for example solar control glazing) or reduced light ingress to the windows (for example nearby buildings).
Daylight may similarly be deemed to be adequate in habitable rooms and
kitchen when calculation can demonstrate that there is a daylight factor of
2% in half of the room area.
In workrooms, daylight may also be deemed to be adequate when calculation can demonstrate that there is a daylight factor of 2% in the work zone.
This can be calculated by means of a calculation grid that covers the room
or the work zone. The grid has an off set of 0.5m from the walls and contains
evenly distributed grid points with a maximum distance of 0.5m.

12

Daylight may similarly be deemed to be adequate when calculation or

measurements can demonstrate that there is a daylight factor of 2% at the
workplaces. When determining the daylight factor, account must be taken of
actual conditions, including the design of the windows, the light transmittance of the pane and the nature of the room and of the surroundings. See
By og Byg (SBi) Guidelines 203, Beregning af dagslys i bygninger [Calcula-

tion of daylight in buildings] and SBi Guidelines 219, Dagslys i rum og

bygninger [Daylight in rooms and buildings].

Danish Working Environment Authority

In their guideline A1.11, the Danish Working Environment Authority outline
requirements to daylight in workrooms and permanent workplaces (June
2007):
Daylight levels in workrooms should be adequate. Daylight can usually be
taken to be sufficient if the glazed area of side lights corresponds to a minimum of 10% of the room floor area or, in the case of roof lights, no less than
7% of the room floor area.
Another configuration may similarly be deemed to be adequate. The 10%
and 7% is a guideline that will usually give adequate daylight with a normal
layout of the rooms. However, situations may occur, where daylight access
is not adequate. Daylight may similarly be deemed to be adequate when
calculation or measurements can demonstrate that there is a daylight factor
of 2% at the workplaces

Definition of daylight factor

As a simple means to describe the amount of daylight in a specific point in a
room, the term daylight factor can be used.
A daylight factor (DF) is the ratio of the internal light level to the external light
level and is defined as follows, see Figure 2. The daylight factor is normally
given in per cent:
Eindoor
100 %
Eoutdoor
is the illuminance due to daylight at a point on a given plane indoors
=

Eindoor
(lux)
Eoutdoor is the simultaneous outdoor illuminance on a horizontal plane from
an unobstructed hemisphere of overcast sky (lux)

Figure 2: Definition of the daylight factor (DF) when the unobstructed outdoor illuminance level is 10,000
lux and the daylight level is 200 lux indoors, the daylight factor is 2%. The guideline to the Building Regulations 6.5.2 (1) states that in workrooms, the daylight can be seen as adequate when there is a daylight factor of 2% at the workspaces. The figure is taken from SBi Guidelines 219 (Johnsen &
Christoffersen 2008)

For a given point in a room, the daylight factor is a permanent factor, which
occurs on days with overcast skies. The daylight factor is calculated under a
standard overcast sky, which means that the calculation is per definition independent of window orientation and does not express anything about how

13

much light there will be in the point under real sky conditions for a given orientation.
A topical profile for the daylight factor through the room is seen in Figure 3.
The daylight factor is the factor closest to the window and the ratio between
the highest and lowest daylight factors will typically be 20:1 and most often
50:1.

Figure 3: Typical daylight factor through a room with a glazing area of the facade of 40% and a mean
room reflectance of 0.4. The figure is taken from SBi Guidelines 219 (Johnsen & Christoffersen 2008).

Methods for analysing the daylight factor

The daylight factor (DF) can be assessed in different ways. It can be measured at a specific point in the room to ensure DF levels for specific functions,
e.g. at work places, and it can be measured on a working plane or floor area,
dependent on the function of the room. If the area approach is applied, it is recommended to make use of a measurement grid (EN 12464-1:2011).

DFmean
The DFmean describes the average daylight factor for the investigated grid.

DFmedian
The DFmedian is the middle daylight factor when the calculated daylight factors are sorted in an ordered list. The median is thereby a term less dependent on high and low daylight factors in the room, e.g. very high daylight factors close to the window and very low daylight factors farthest away from the
window.

Uniformity
The uniformity of the daylight factors on a work plane can be defined in two
ways:
1) Ratio between DFmin / DFmean and
2) Ratio between DFmin / DFmax
According to the assessment method for sustainable buildings BREEAM
2.08 the ratio between DFmin / DFmean should at least be 0.4 or the minimum point daylight factor should be at least 0.8%.

Daylight zone
A daylight zone in a room can be defined as the distance from the facade
where the recommended daylight factor is achieved.

14

Measurement grid
Grid systems should be created to indicate the points at which the illuminance values are calculated and verified for the task area(s), immediate surrounding area(s) and background area(s). The following description is in accordance with the description given in EN 12464-1:2011 and it is recommended to follow this instruction.
Grid cells approximating to a square are preferred, the ratio of length to
width of a grid cell shall be kept between 0.5 and 2 (EN 12193:2007 and EN
12464-2:2007). The maximum grid size shall be:
= 0.2 510()

(1)

Where
p is the maximum grid cell size (m) 10 m
d is the longer dimension of the calculation area (m),. However, if the ratio of
the longer to the shorter side is 2 or more, then d becomes the shorter dimension of the area
The number of points in the relevant dimension is given by the nearest
whole number of d/p.
The resulting spacing between the grid points is used to calculate the nearest whole number of grid points in the other dimension. This will give a ratio
of length to width of a grid cell close to 1.
A border of 0.5 m from the walls is excluded from the calculation area except
when a task area is in or extends into this border area.
An appropriate grid size shall be applied to walls and ceiling and a border of
0.5 m may be applied also.
NOTE 1 The grid point spacing should not coincide with the luminaire spacing.
NOTE 2 Formula (1) (coming from CIE x005-1992) has been derived under
the assumption that p is proportional to log (d), where:
p = 0.2 m for d = 1 m;
p = 1 m for d = 10 m;
p = 5 m for d = 100 m.
NOTE 3 Typical values of grid point spacing are given in Table 4.
Table 4: Recommended number of grid points

Length of the area

(d)
[m]
0.4
0.6
1.0
2.0
5.0
10.0
25.0
50.0
100.0

Maximum distance between grid point (p)

[m]
0.15
0.20
0.20
0.30
0.60
1.00
2.00
3.00
5.00

Minimum number of grid

points
3
3
5
6
8
10
12
17
30

Example
The examples below show DF measurements over a work plane in two
rooms with identical dimensions (3m x 6m x 2.5m) and material properties
(floor 0.2, wall 0.5, ceiling 0.8), but different window placement and glazing

15

area. Room 1 is solely illuminated by one large faade window with 5.9 m
glazing area, and Room 2 is illuminated by a combination of faade and roof
2
window with 2.4m glazing area. Its possible to observe that the DFmedian
is a more sensitive metric than the DFmean in the room evaluation.

Room 1 (5.9 m2 glazing area)

DFmean

4.0%

DFmedian

2.3%

Uniformity (min/mean)

0.19

Uniformity (min/max)

0.04

Share of room with DF above 2%

54%

Figure 4: DF results in the example room with a window in the facade

Room 2 (2.4 m2 glazing area)

DFmean

4.2%

DFmedian

3.9%

Uniformity (min/mean)

0.41

Uniformity (min/max)

0.7

Share of room with DF above 2%

98%

Figure 5: DF results in a room with two different window configurations, facade window and roof window

From the figures it can be seen that the DFmean is of similar magnitude; 4.0
% and 4.2% within the two rooms. Looking at the room area with a daylight
factor above 2%, a difference between 54% and 98% is observed in spite of
the average daylight factor being almost equal. Comparing DFmean and
DFmedian shows higher difference for the room with large contrasts (Room
1) compared with Room 2. This example demonstrates that the DFmedian
and the area of the room with a daylight factor above 2% are more sensitive
ways of assessing than the DFmean for assessing the daylight performance
of the rooms.

16

Calculation methods

Accuracy of daylight calculation

Lighting simulation software has experienced rapid growth in recent years.
This is both due to requirements stipulated in the Danish Building Regulations and guidelines given by the Danish Working Environment Authority and
to a wish from developers and architects to document the lighting qualities of
the building. The daylighting software is capable of estimating daylight levels
in a room. However, it it is important not to forget that the quality and accuracy of simulations can be influenced by a number of different factors including:
- calculation method
- sky model
- building model
- surface properties
- user expertise
Studies comparing real-life daylight measurements with daylight simulations
show deviations around 10% (Mardaljevic 1995; Reinhart & Walkenhorst
2001; Reinhart & Andersen 2006; Reinhart 2010).

Split flux formula

The split flux formula is a simple algorithm derived from a manual calculation
method established by BRE (The English Building Research Establishment).
This method is based on the principle that the global illumination at a certain
point in a room is the result of three distinctive components of daylight:
1. the direct sky component (SC)
2. the reflections from exterior surfaces (ERC) and
3. the reflections from internal surfaces (IRC).
Each component is calculated separately and then added up to obtain the
global illumination in each sensor point. The internally reflected component
is determined by an equation using the average reflectance of interior surfaces, the total glazing area and a correction factor for the external obstruction. Given these approximations, this method is likely to overestimate or
underestimate the amount of daylight. It is only recommended to use this
method for spaces in which the window openings are parallel to the walls.
The simulation program Ecotect applies this method as default in its calculations.

17

Figure 6: Components of the split flux formula

Radiosity
Radiosity is an algorithm capable of rendering realistic scenes with shadows
and diffuse light (Watt 2000). It is an application of the finite element method
to solve the rendering equation for scenes with purely diffuse surfaces, a
method initially developed to study thermal transfer. This method requires
the surfaces of the scene to be subdivided into a mesh of smaller patches.
View factors between each pair of patches are computed, and the illumination of a patch is determined by adding the contribution of all visible surrounding patches and light sources. This method has constraints that limit its
use for daylighting simulations (diffuse surface, complex description of the
sky) and it should only be used to evaluate a relatively simple space.
The simulation programs Relux, DIALux and LightCalc apply radiosity as
their simulation engine.

Figure 7: Illustration of the radiosity calculation method.

Ray tracing

18

Ray tracing is a rendering technique based on the calculation of the distribution of a large number of rays emitted in a scene either from light sources
(forward ray tracing), or a view point (backward ray tracing) (Larson &
Shakespeare 1998). Backward ray tracing is a faster method than forward
ray tracing because it only calculates rays reaching the view point. On the
other hand it is less or not suitable for use in cases where light sources are
hard to find in the scenes, i.e. narrow light well, light pipe. Ray tracing algorithms support reflection, transmission and refraction properties of surfaces,
which permits the use of complex materials in simulations (Watt 2000).

The simulation program Radiance applies ray tracing in its calculation engine.

Figure 8: Illustration of the backward ray tracing calculation method

Photon mapping
Photon mapping is a rendering technique that uses bi-directional ray tracing.
A first pass combines forward ray tracing and photon mapping to distribute
the light in the room and cache the luminous flux on surfaces. A second
pass consisting of backward ray tracing from the view point is then used to
compute the final image. This optimised technique permits the simulation of
more complex lighting scenes with accuracy, i.e. light pipe, and is faster than
the traditional backward ray tracing.

Figure 9: Illustration of the photon mapping calculation method

VELUX Daylight Visualizer applies photon mapping as its calculation engine.

Sky types
The nature of daylight is extremely varied and yields a wide range of occurring sky conditions. To this effect, CIE (Commission international de
lclairage) has defined 15 normalised skies for use in lighting simulation
(CIE 2003). In addition to these 15 skies, some older sky models remain
such as the Traditional Overcast Sky which is now referred to as sky type
16.
The figure below shows the relative difference in luminance levels at different angular distance between a sky element and the zenith for sky types 1
and 16. The comparison shows a relative difference above 25% for the sky

19

luminance of the sky element located close to the horizon (80 degrees),
which means that sky type 16 will be brighter, or more luminous, closer to
the horizon.
Relative Luminance Difference
30%

1
0,9
Relative Luminace

0,7

10%

0,6

0%

0,5
0,4

-10%

0,3
0,2

-20%

0,1
0

Relative Difference [%]

20%

0,8

10

20

30

40

50

60

70

80

90

-30%

Angular distance between a sky element and the zenith (deg)

CIE_General Sky Type 1

CIE_Standard Overcast (Type 16)

Relative Difference

Figure 10: Relative difference between the luminance of overcast skies types 1 and 16

This is important to consider when comparing the calculations of different

simulation programs, as it can explain some of the differences obtained between programs. For example, when we compare the results obtained between VELUX Daylight Visualizer and Radiance, we can see slightly higher
results in Radiance for rooms 1, 2, 3 and 4, whereas the results are slightly
higher in VELUX Daylight Visualizer for room 5. This can be explained by
the fact that Radiance uses sky type 16 and VELUX Daylight Visualizer uses
sky type 1, and the fact that only room 5 had a roof opening (all other rooms
were lit solely by facade openings) and thus it was the room that would encounter the largest effect of different sky luminances at zenith.

The model
The quality of the building model used in simulation can significantly influence the quality and accuracy of daylighting simulations.

Geometry
It is important to ensure that the geometry of rooms and daylight openings is
correctly modelled and that the correct simulation parameters are used. For
example, in most programs when you model window glass you should make
sure that you only have one layer of polygon to represent the glass, since
that polygons material is assigned with the correct transmittance properties
of the window pane.

20

Figure 11: Illustration of a window model with 2 layers of polygon to model the glass (left) and one layer
(right).

External obstruction
It is important to properly represent elements influencing daylight on the site,
such as external obstruction from surrounding buildings, landscape and
vegetation. The effect of external obstruction can be shown by comparing
results for the simple room (without obstruction), and the room with obstruction.

Surface properties
Surface properties such as the material type, reflectance and transmittance
values will influence the way daylight is distributed in the rooms. It is important to ensure that realistic values are used for the floor, wall and ceiling
surfaces, as well as any other surfaces included in the model. It is also important to ensure that the glass transmittance is properly defined, and differentiated between openings using different type of glazing.

Example
In the example below, a scene with realistic surface properties (floor 0.2,
wall 0.5, ceiling 0.7) is compared with a scene with unrealistic surface properties (floor 0.9, wall 0.9, ceiling 0.9). The results show a difference of 22%
between the average DF values obtained.

21

Figure 12: Daylight factor results for the rooms with realistic surface properties (left) and unrealistic surface properties (right).

User expertise
In order to ensure valid simulations, it is important for users to have a good
control of the interface and data entry of the simulation software. It is also
important to know the capabilities and limitations of the tools, and to select
an appropriate method based on the complexity of the scene to be evaluated.
A study from 2009 made by researchers at Havard University investigated how
correctly daylight simulations were performed by student/new users (Ibarra &
Reinhart 2009). The study showed that typical sources of errors were geometrical errors, such as omitting the thickness of the wall from the model or
that the material properties were wrong.

22

Description of simulation programs

This section gives a short description of the nine daylight simulation programs applied for this report. The description is based on information available on the respective homepages and information given by the software developers. For each simulation program, links are provided indicating where
to find further reading.

Radiance
Radiance is a suite of programs for the analysis and visualisation of lighting
in design. Input files specify the scene geometry, materials, luminaires, time,
date and sky conditions. Calculated values include spectral radiance (ie. luminance + colour), irradiance (illuminance + colour) and glare indices. Simulation results may be displayed as colour images, numerical values and contour plots. The primary advantage of Radiance over simpler lighting calculation and rendering tools is that there are no limitations on the geometry or
the materials that may be simulated. Radiance is used by architects and engineers to predict illumination, visual quality and appearance of innovative
design spaces, and by researchers to evaluate new lighting and daylighting
technologies.
The program is continuously being updated.
Useful links:
Alex Jacobs has written tutorials that can help you get started on Radiance:
http://www.jaloxa.eu/resources/radiance/documentation/index.shtml
The official link to the Radiance homepage is provided below. At this
homepage, there is access to a Radiance mailing list and a possibility to look
through Radiance-related questions asked. It is also from this homepage
that the program can be downloaded for free.
http://www.radiance-online.org/
Furthermore, versions of Radiance that can be run on both UNIX and PC
systems can be downloaded for free from this link:
https://openstudio.nrel.gov/getting-started-developer/getting-started-radiance

Daysim
DAYSIM is a validated, RADIANCE-based daylighting analysis software
that models the annual amount of daylight in and around buildings. DAYSIM
allows users to model dynamic facades systems ranging from standard venetian blinds to state-of-the-art light redirecting elements, switchable glazings and combinations thereof. Users may further specify complex electric
lighting systems and controls including manual light switches, occupancy
sensors and photocell controlled dimming.
Simulation outputs range from climate-based daylighting metrics such as
daylight autonomy and useful daylight illuminance to annual glare and electric lighting energy use. DAYSIM also generates hourly schedules for occupancy, electric lighting loads and shading device status which can be directly
coupled with thermal simulation engines such as EnergyPlus, eQuest and
TRNSYS. Daysim is free.
Useful link:

23

http://www.daysim.com/

Desktop Radiance
Desktop Radiance is a Windows 95/98/NT software package that integrates
the Radiance Synthetic Imaging System with AutoCAD Release 14. Desktop
Radiance includes libraries of materials, glazings, luminaires and furnishings
so you can quickly create realistic lighting models. The goal is to provide a
design tool integrated with popular CAD packages to facilitate the consideration of energy efficient lighting and daylighting strategies in building design.
The development of Desktop Radiance stopped in 2002. Desktop Radiance
is included in this report since reference is made to Desktop Radiance for
the users of Autodesk Ecotect. It is recommended to download the newest
version of Radiance to Windows, as the calculation engine is updated and
will be updated continuously.
Useful link:
http://radsite.lbl.gov/deskrad/

IESve
In IESve, it is possible to build a geometric model and simulate daylight conditions within this model.
IES have been using Radiance over a wide range of commercial projects
for many years and to facilitate this work IES have integrated Radiance into
the <Virtual Environment>. The latest version of this integration RadianceIES provides users with even greater ease of use through its customised
Graphical User Interface (known as the <Virtual Environment> Framework),
which integrates with all the IES software packages (SunCast, Apache,
etc.).
Useful link:
http://www.iesve.com/

DIALux
DIALux can calculate electric light, daylight and the energy performance of
electric light. The program is oriented towards the European market, and is
widely used for calculation of indoor and outdoor electric lighting systems. It
follows different national standard lighting calculations, and can import photometric databases directly from manufacturers. The daylight calculation capabilities within DIALux make use of German standard DIN 5043 and CIE
Publication 110. Geometric input is limited to certain shapes. Sky choices
are somewhat limited but acceptable for diverse ranges of weather conditions. There is an external radiosity and ray-tracing model, POV-Ray (Persistence of Vision 2010). It is used to produce images from calculation results
and for presentation renderings. DIALux is available free of charge but is not
open source.
Useful link:
http://dial.de/

Relux
Relux is oriented towards the European market and the program can calculate electric light, daylight, and the energy performance of electric light. The
program can import photometrical databases from manufacturers.
Relux applies both radiosity and ray tracing in its simulation engine. It is up
to the user to decide what calculation engine to apply for simulation.
Relux is free of charge but not open source.

24

Useful link:
http://www.relux.biz/

Ecotect
Autodesk Ecotect Analysis sustainable design analysis software is a
comprehensive concept-to-detail sustainable building design tool. Ecotect
Analysis offers a wide range of simulation and building energy analysis functionality that can improve performance of existing buildings and new building
designs. These simulations and analysis include energy performance of
buildings, thermal simulations, solar radiation, daylight levels and shadow
diagrams.
Useful link:
http://usa.autodesk.com/ecotect-analysis/

VELUX Daylight Visualizer

VELUX Daylight Visualizer is a simple, validated (Labayrade et al. 2009)
daylight simulation and visualisation tool for daylighting design and analysis
in buildings. It is intended to promote the use of daylight in buildings and to
aid professionals by predicting and documenting daylight levels and appearance of a space prior to realization of the building design.
The Daylight Visualizer intuitive modelling tool permits quick generation of
3D models in which roof and facade windows are freely inserted. The program also permits users to import 3D models generated by CAD programs in
order to facilitate a good workflow and provide flexibility to the model geometry.
Daylight Visualizer can be used to calculate luminance, illuminance and daylight factors for all 15 sky types defined in CIE Standard General Sky. Outputs from Daylight Visualizer are image based and post-processed in a GUI,
where pseudo colours are applied and pixel values can be exported for further analysis in other software. VELUX Daylight Visualizer has an efficient
workflow and intuitive design, which makes it accessible and easy to use.
The program is freely available and can be accessed from the link below.
Useful link:
http://viz.velux.com

LightCalc
LightCalc is a freely available open-source tool for electric and artificial light
simulations. LightCalc has its own calculation engine, based on forward ray
tracing and radiosity. The light from the sky and other light sources are included in the simulations through forward ray tracing. Radiosity is applied to
calculate the internal inter-reflected contribution.
The program relies on input from the user and has a short calculation time.
Through iDBuild, it is possible to couple light simulations from LightCalc directly to hourly thermal calculations. LightCalc is limited to calculation of one
single room.
Useful link: http://idbuild.dk/

25

Graphical presentation of the results

The graphical output from the various simulation programs varies in their expression. Figure 13 shows examples of the graphical presentation of daylight
factor levels in the simple room. In the following sections, the results are
post-processed in Excel in order to be able to present them in a comparable
graphical way.
Radiance

Relux raytracing

Daysim

Relux radiosity

Desktop Radiance

Ecotect

IESVE

Velux Daylight Visualizer

DIALux

LightCalc

Figure 13: Simulation results (DF) on the working plane 0.85 m above floor level for the different simulation programs

26

Simulations and analysis

Room types
For this investigation, the room types are chosen based on a wish to compare simulations in rooms with a simple geometry with rooms with a slightly
more complex geometry. Some of the chosen simulation programs are capable of calculating all room types whereas others can calculate only the
simpler room types. The following room types were chosen for this investigation:

Room 1 Simple room

A simple room with one window. Light entering the space is reflected both
from the side walls and the end walls. The room dimensions are: Width of
3.5 m, height of 2.8 m, depth of 4 m. The thickness of the wall is 0.4 m and
the thickness of the window frame is 0.06 m. The dimensions of the window
opening in the facade are; 2 m in width and 1.5 m in height. The height of
the spandrel is 0.9 m.

Room 2 Deep room

The room depth is increased from 4 m to 7 m. Otherwise, the room and the
window is the same size as the simple room. The light will not be reflected
from the end walls.

Room 3 Room with obstruction

The dimensions of the room and the window are the same as those of the
simple room. An obstruction is placed outside the window. In the simulation
results, the effect of external obstructions on the daylight levels can be seen.

Room 4 Room with light shelf

The geometry of this room is similar to that of the simple room, with the only
exception that a light shelf is included in the window. The room type is included in these investigations in order to be able to calculate a detail in the
facade/window.

Room 5 Room with borrowed light

Daylight will pass through two windows before it enters the room that requires daylight. Window openings are placed in the facade and as a sky
light. The light has to pass through these openings and then through an internal window before it enters the calculated room.

27

Simple room

Deep room

Room with light

Room with borrowed light
shelf
Figure 14: Chosen room types

28

Room with obstruction

Simulation parameters
The simulation parameters for the rooms are similar. Surface reflectances
applied correspond to the ones shown in the figure below.

Table 5: Reflectances

Ground plane

Table 6: Light transmittance

Reflectance
(r)
0.1

External walls and obstruction

Floor

0.3
0.1

Wall
Ceiling

0.4
0.7

Window frame
Side of window opening, internal

0.8
0.7

Side of window opening, external

Light shelf

0.3
0.5

Window in facade and sky light

Internal glass

Light transmittance(LT)
0.76
0.85

Table 7: Calculation points, calculation height and ground plane

Height of work plane

0.85m

Work zone
Grid

Distance from wall to calculated area is 0.5m

0.25 m horizontal and vertical

Calculation level
Ground plane

High Quality
Rooms are located in the middle of ground plane
of dimensions 50x50 m

29

Guidance to result sheet simulations

The following pages will show the simulation results of the different room
types. A result sheet as shown below is presented for each room. The result
sheet for the simulations includes:
1) Illustration of room type,
2) Figure that shows the daylight factor in the centre line of the room as
a function of the room depth. The results for this case are postprocessed in Excel in order to obtain the same graphical expression
for all the simulation programs.
3) Caption explaining the figure.
4) Simulation results shown for a calculation grid in working plane
height.

3
4

30

Guidance to result sheet analysis

The following pages will show the analysis results of the different room
types. A result sheet as shown below is presented for each room. The result
sheet for the analysis includes:
1) Illustration of room type
2) Explanation of figures
3) Figure showing the daylight factor of the different programs given as
the DFmean, DFmedian, DFmin and DFmax, as well as the uniformity: DFmin/DFmean and DFmin/DFmax
4) Figure showing the distribution of the daylight factor simulations in a
box plot. DFmax is the highest value, followed by the 75% percentile, DFmean, the 25% percentile and DFmin. Percentiles of 75%
and 25% describe the daylight factor for 75% and 25% of the calculation points
5) Figure showing the distance from the facade with a daylight factor of
2% for the different simulation programs

31

ROOM 1 SIMPEL ROOM Simulation sheet

Daylight factor [%]

2
5

14,0
13,0
12,0
11,0
10,0
9,0
8,0
7,0
6,0
5,0
4,0
3,0
2,0
1,0
0,0
0,5

1,0

1,5

2,0

2,5

Distance from window[m]

Figure 15: Daylight factor levels through the room

32

3,0

3,5

Figure 16: Simulation sheet

33

ROOM 1 - SIMPEL ROOM Analysis sheet

Table 8 lists the daylight factors for the simulated simple room given as
DFmean, DFmedian, DFmin and DFmax. Similar results are visualised in
Figure 17 as box plots.
The figure and the table show that the spread of the average daylight factor
for the 10 calculation programs ranges between 3.0% (Visualizer and IESve)
and 3.5% (Relux Raytracing). This means that there is a deviation of 15%
between the maximum and the minimum values of the simulated average
daylight factor for the 10 programs. Furthermore, the standard deviation has
been calculated. The standard deviation is 0.17 which corresponds to a difference of 5.1% between the different programs.
Figure 18 shows the distance from the faade where a daylight factor of 2%
is achieved. The distance from the facade with a DF of 2% varies from 2.1m
to 2.5m leading to differences of 18% in daylight penetration depth between
the maximum and the minimum distances. The standard deviation is 0.15m,
which gives a variation of 6.4% for the different simulation programs.
This type of room represents a standard scene with no spectacular daylighting parameters. It is therefore expected that this simulation will give the
lowest difference between the different simulation programs.
All simulation programs are capable of simulating the standard room. This
suggests that one program is just as good as another for simulating a standard room. The standard deviation resulted in differences between the different simulation programs of 5.5%. And the difference in the average daylight factor between the maximum and the minimum values was 15%. When
looking at the daylight penetration depth, the standard deviation resulted in a
difference of 6.4% between the different simulation programs, and the difference between the maximum and minimum daylight penetration depth was
18%.

34

Table 8: Daylight factor values given as mean, median, min and max values, as well as the uniformity
given as DFmin/DFmean og DFmin/DFmax
VELUX Daylight Visualizer

LightCalc

IESve

Ecotect

Desktop Radiance

Relux Raytracing

Relux Radiosity

Daysim

Radiance

DIALux

DF [%]
DFmean

3.2

3.3

3.4

3.4

3.5

3.3

3.2

3.0

3.3

3.0

DFmedian

2.1

2.2

2.3

2.2

2.6

2.2

2.2

1.9

2.3

1.9

DFmin

0.9

0.9

0.9

0.8

1.1

0.8

1.1

0.7

0.9

0.7

DFmax

12.0

12.9

12.9

13.6

12.3

12.7

12.2

12.2

13.5

12.6

DFmin/DFmean

0.3

0.3

0.3

0.2

0.3

0.3

0.3

0.2

0.3

0.2

DFmin/DFmax

0.1

0.1

0.1

0.1

0.1

0.1

0.1

0.1

0.1

0.1

Uniformity

Figure 17: Box plot of the daylight factor for the simple room simulated using the different simulation programs.

Figure 18: Distance from the facade with a daylight factor of 2%, simulated using the different simulation programs.

35

Daylight factor [%]

ROOM 2 DEEP ROOM Simulation sheet

14,0
13,0
12,0
11,0
10,0
9,0
8,0
7,0
6,0
5,0
4,0
3,0
2,0
1,0
0,0
0,5

1,0

1,5

2,0

2,5

Distance from window[m]

Figure 19: Daylight factor level throughout the room

36

3,0

3,5

Figure 20: Deep room simulation sheet

37

ROOM 2 DEEP ROOM Analysis sheet

Table 9 and Figure 21 show daylight factor values of the simulated deep
room. For the average daylight factor the maximum value is 2.1% (Relux
Raytracing), and the minimum value is 1.6% (IESve). The difference is 23%
between the maximum and the minimum values of the simulated average
DF, and the standard deviation is 0.13 resulting in differences of 7.4%.
Figure 22 shows the distance from the faade with a daylight factor of 2%.
The simulated maximum distance from a facade with a DF of 2% is 2.4m
(Relux Raytracing) and the simulated minimum distance from the facade is
2.1m (DIALux, IESve and VELUX Daylight Visualizer). This is a difference of
0.3m, or 16%. The standard deviation is 0.11m which corresponds to differences between the different simulation programs of 4.9%. The daylight
penetration depth is slightly shorter in the deep room than in the simple room
1. This is due to the reflected, bouncing off from the back wall, contributing
to an increment in DF in the standard room, whereas this is not the case for
the deep room.
All simulation programs are capable of simulating the standard room.

38

Table 9: Daylight factor values given as mean, median, min and max values, as well as the uniformity
given as DFmin/DFmean and DFmin/DFmax

Ecotect

IESve

LightCalc

VELUX Daylight Visualizer

1.9

1.7

1.6

1.8

1.7

DFmedian

0.7

0.8

0.8

0.8

1.0

0.8

0.8

0.6

0.7

0.6

DFmin

0.2

0.2

0.2

0.2

0.3

0.2

0.4

0.1

0.2

0.2

DFmax

12.0 12.9 12.9 13.6 12.4 12.9 10.5 11.7 13.4 12.6

Desktop Radiance

2.1

Relux Raytracing

1.9

Relux Radiosity

1.9

Daysim

1.9

Radiance

1.7

DIALux

DFmean

DF [%]

Uniformity
DFmin/DFmean 0.1

0.1

0.1

0.1

0.2

0.1

0.3

0.1

0.1

0.1

DFmin/DFmax

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

Figure 21: Box plot of daylight factor levels in the deep room simulated using the different simulation programs

Figure 22: Distance from the facade with a daylight factor of 2%

39

Daylight fcator[%]

ROOM 3 ROOM WITH OBSTRUCTIONS Simulation sheet

14,0
13,0
12,0
11,0
10,0
9,0
8,0
7,0
6,0
5,0
4,0
3,0
2,0
1,0
0,0
0,5

1,0

1,5

2,0

2,5

Distance from window[m]

Figure 23: Daylight factor level throughout the room

40

3,0

3,5

Figure 24: Simulation sheet Room with obstruction

41

ROOM 3 ROOM WITH OBSTRUCTION Analysis sheet

Table 10 and Figure 25 show daylight factor values for the simulated room
with obstruction.
From the table and figure it can be seen that the maximum average daylight
factor is 1.2% (Ecotect), and the minimum average daylight factor is 0.7%
(VELUX Daylight Visualizer, Desktop Radiance, DIAlux). However, it shows
clearly that the Ecotect results differ greatly from the other results with generally higher simulation results. If the Ecotect results are excluded, the maximum average daylight factor is 0.9%. This means that the range between
the maximum and the minimum average daylight factor is 25%. The standard deviation for the average daylight factor in Room 3 is a DF of 0.08% corresponding to simulated differences of 10.4%.
It is furthermore worth noting the IESve simulation results. The maximum
value for this simulation program is noticeably lower than those of the other
simulation programs. The difference might be due to the physical composition of the model in the simulation program, or that the default simulation parameters when running highest accuracy in IESve are not set high enough
to consider obstructions correctly. In Radiance, you have a simulation setting
called ambient bounces. This setting tells you the number of inter-reflections
applied in the simulations. For the Daysim simulations and Radiance simulations, the ambient bounces were set to 7, whereas the default IESve setting
was 3.
When a practitioner applies IESve and uses the highest accuracy for his/her
simulation, he/she would of course expect that the program produces reliable results. However, when dealing with obstructions, this does not seem to
be the case.
Figure 26 shows the distance from the faade with a daylight factor of 2%.
From the figure it can be seen that Ecotect is an outlier. Ecotect has therefore been excluded from the analysis. The IESve program results in the lowest daylight penetration depth (0.55m); the maximum daylight penetration
depth of 0.73m was obtained with the Relux Raytracing program. This results in a difference in daylight penetration depths of 25% between the maximum and the minimum distances. The standard deviation is 0.06m resulting
in differences of 8.7%.

42

Except for simulations with Ecotect, the investigated simulation programs

were capable of including obstruction in its simulations. It should be noted
that for the simulation program of IESve, there is a tendency that this pro-

gram underestimates the daylight availability, when you apply the highest
accuracy in its simulation.
Ecotect is the only program that applies the BRE Split Flux method for its
calculation. This calculation method is a simplified method and is limited in
terms of considering multiple reflections, and this might be the reason why
we see this high discrepancy between Ecotect and the other simulation programs.

43

Table 10: Daylight factor values given as mean, median, min and max values, as well as the
uniformity given as DFmin/DFmean og DFmin/DFmax

Radiance

Daysim

Relux Radiosity

Desktop Radiance

Ecotect

IESve

LightCalc

VELUX Daylight Visualizer

0.7

0.8

0.8

0.9

DF [%]
0.9 0.7

1.2

0.6

0.8

0.7

DFmedian

0.4

0.5

0.5

0.6

0.7

0.5

0.9

0.5

0.6

0.4

DFmin

0.3

0.2

0.2

0.3

0.3

0.2

0.6

0.2

0.3

0.2

DFmax

2.7

3.1

3.1

3.4

3.1

2.8

4.4

2.1

2.7

3.0

DFmin/DFmean

0.4

0.3

0.3

0.4

0.3

0.3

0.5

0.3

0.3

0.2

DFmin/DFmax

0.1

0.1

0.1

0.1

0.1

0.1

0.1

0.1

0.1

0.1

Relux Raytracing

DIALux

DFmean

Uniformity

Figure 25: Box plot of daylight factor levels in the room with obstruction simulated using the different simulation programs

Figure 26: Distance from the facade with a daylight factor of 2% obtained using the different simulation programs

44

Daylight factor[%]

ROOM 4 ROOM WITH LIGHT SHELF Simulation sheet

14,0
13,0
12,0
11,0
10,0
9,0
8,0
7,0
6,0
5,0
4,0
3,0
2,0
1,0
0,0
0,5

1,0

1,5

2,0

2,5

3,0

3,5

Distance from window[m]

Figure 27: Daylight factor level through the room

45

Figure 28: Simulation sheet Room with light shelf

46

ROOM 4 ROOM WITH LIGHT SHELF Analysis sheet

Table 11 and Figure 29 show the daylight factor levels for the simulated
room with a shelf. The maximum average DF is 2.6% (Relux Raytracing and
Radiance 2.0 BETA), and the minimum average DF is 2.1% (IESve). The
relative difference in average DF between the program that results in the
maximum average DF and the minimum DF is 18%; the standard deviation
is 6.6%.
Compared with the simple room (Room 1), the average DF for the room with
light shelf is decreased due to less light entering the room, because the light
shelf has the same effect as an overhang. The redirecting effect of light from
a light shelf cannot be seen in the daylight factor simulations since this simulation is made under diffuse overcast skies and the reflections are ideal diffuse lambertian with an even distribution of the reflected light.
Figure 30 shows the distance from the faade where a daylight factor of 2%
is obtained. Looking at the daylight penetration depth, the minimum daylight
penetration depth obtained is 1.8 m from the faade (IESve) and the maximum daylight penetration depth is 2.2 m from the faade (Relux Raytracing
and Ecotect). This means that there is a difference of 0.4m between the
maximum and the minimum distances from the faade where 2% DF is
reached for all the applied simulation programs. This corresponds to a difference between the maximum and the minimum daylight penetration depth
of 21%. The standard deviation is 0.14m corresponding to a simulated difference of 6.7%.
LightCalc was not capable of simulating the light shelf. However, the other
programs were capable of simulating a light shelf as a shelf with a reflectance of 0.5.

47

Table 11: Daylight factor values given as mean, median, min and max values, as well as the uniformity
given as DFmin/DFmean og DFmin/DFmax
Velux Daylight Visualizer

IESve

LightCalc

Ecotect

Desktop Radiance

Relux Raytracing

Relux Radiosity

Daysim

Radiance

2.2

2.5

2.5

2.4

2.6

2.6

2.5

2.1

2.2

DFmedian

1.6

1.8

1.9

1.6

2.0

1.9

1.8

1.5

1.6

DFmin

0.7

0.7

0.8

0.6

1.0

0.8

0.9

0.5

0.6

DFmax

7.5

8.4

8.4

8.5

8.6

8.7

8.8

7.4

7.9

DFmin/DFmean

0.3

0.3

0.3

0.2

0.4

0.3

0.4

0.2

0.3

DFmin/DFmax

0.1

0.1

0.1

0.1

0.1

0.1

0.1

0.1

0.1

DIALux

DFmean

DF [%]

Uniformity

Figure 29: Box plot of daylight factor levels in the room with light shelf simulated with the different simulation programs

Figure 30: Distance from the facade with a daylight factor of 2% obtained with the different simulation programs

48

Daylight factor[%]

ROOM 5 ROOM WITH BORROWED LIGHT Simulation sheet

14,0
13,0
12,0
11,0
10,0
9,0
8,0
7,0
6,0
5,0
4,0
3,0
2,0
1,0
0,0
5,8

6,0

6,2

6,4

6,6

6,8

7,0

7,2

7,4

7,6

7,8

8,0

Distance from window [m]

Figure 31: Daylight factor level throughout the room

49

Figure 32: Simulation sheet Room with borrowed light

50

ROOM 5 ROOM WITH BORROWED LIGHT Analysis sheet

Table 12 and Figure 33 show the daylight factor values for the simulated
room with borrowed light. It can clearly be seen that the results from Ecotect
differ greatly from those of the other programs. With Ecotect, the DFmean is
6%, whereas the mean values for the other programs range from 1.7% to
2.4%. This large difference is caused by the limitations of the split flux method applied in Ecotect.
When excluding Ecotect from the results, the maximum average daylight
factor is 2.4% (DIALux) and the minimum average daylight factor is 1.7%
(Relux Raytracing, Desktop Radiance, IESve) and the difference between
max and mean is 31%. The standard deviation in average daylight factor for
this room is 0.24 corresponding to differences of 9.5%.
No analysis of the daylight penetration depth has been made in this room as
it is an internal room with borrowed light.

51

Table 12: Daylight factor values given as mean, median, minimum and maximum values. As well as the
uniformity given as DFmin/DFmean og DFmin/DFmax
VELUX Daylight Visualizer

IESve

LightCalc

Ecotect

Desktop Radiance

Relux Raytracing

Relux Radiosity

Daysim

Radiance

2.4 2.1 2.1

1.7 1.7 6.0

1.7

2.2

DFmedian

2.0 1.8 1.8

1.5 1.4 4.8

1.5

2.0

DFmin

0.9 0.7 0.7

0.6 0.5 2.9

0.4

0.6

DFmax

5.7 5.0 5.0

3.5 4.2 12.7 4.6

5.0

DFmin/DFmean 0.4 0.3 0.3

0.4 0.3 0.5

0.2

0.3

DFmin/DFmax

0.2 0.1 0.2

0.1

0.1

DIALux

DFmean

DF [%]

Uniformity

0.2 0.1 0.1

Figure 33: Box plot of daylight factor levels in the room with borrowed light simulated using the different simulation programs

52

Assessment of the nine daylight simulation

programs

An overall assessment of the nine simulations programs based on their simulation engine, ability to perform calculations of the five room types, their
output, their user friendliness and price is given in the table below.
Radiance

Desktop
Radiance

Daysim

VELUX
Daylight
Visualizer

DIAlux

Ecotect

IESVE

LightCalc

Relux*

Backward
raytracing

Backward
raytracing

Backward
raytracing

Bidirectional
raytracing with
photon mapping and irradiance caching

Radiosity

BRE Split
Flux

Backward
raytracing

Forward
raytracing

Radiosity
og
backward

1. Simple room

Yes

yes

Yes

Yes

yes

yes

Yes

yes

yes

2.Deep room

Yes

yes

yes

Yes

yes

yes

Yes

yes

yes

3. Room with obstruction

Yes

yes

yes

Yes

yes

no

Yes

yes

yes

4. Room with light shelf

Yes

yes

yes

Yes

yes

yes

Yes

no

yes

5. Room with borrowed light

Yes

yes

yes

Yes

yes

no

Yes

no

Yes/No**

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

yes

yes

Yes

no

yes

Yes

yes

yes

Generel interface

-2

-2

-1

Geometric modelling facility within

the program

-2

-2

Ikke
muligt

Import of geometric model in

the program

Yes

yes

yes

Yes

yes

yes

Yes

yes

yes

File types***

.obj
.skp
.disxml
.dxf
.mgf

.obj

.eco
.skp
.3ds
.rad

.obj
.skp
.dwg
.dxxf

.skp
.sat
.m3d
.3ds

.obj
.dwg
.dxf

.skp
.3ds
.rvt

.skp

.skp
.3ds
.dxf
.wrl

Graphical treatment
of results

-1

-2

-1

-2

No

No

No

No

No

Yes

Yes

No

No

Calculation
method

The programs ability to

calculate on the five
room types

Output
Daylight factor and
illuminance
Mean daylight factor
User friendlyness

User license

* Relux applied two simulation approaches, radiosity and raytracing

**Yes: Relux Raytracing can simulate internal rooms. No: Relux Radiosity cannot simulate internal rooms.
***file types that the authors of this report have knowledge about that can be imported. Other
file types might be imported as well.

-2 = very bad / very difficult. -1 = slightly good / slightly difficult. 0 = average.

1= easy. 2 = very easy

53

References

Christoffersen, J., Johnsen, K. & Petersen, E. (2002). Beregning af dagslys i

bygninger (SBi-Anvisning 203). Statens Byggeforskningsinstitut.
Hrsholm.
CIE. (2003) Spatial Distribution of Daylight - CIE Standard General Sky. CIE
Standard, CIE S 011/E:2003.
CIEx005. (1992) Proceedings of the CIE Seminar 1992 on Computer
Programs for Light and Lighting. CIE.
EN12193. (2007) Light and lighting - Sports lighting. Brussels: European
committee for standardization.
EN12464-1. (2011) Light and lighting - lighting of work places - part 1: indoor
work places. Brussels: European committee for standardization.
EN12464-2. (2007) Lighting of work places - Part 2: Outdoor work places.
Brussels: European committee for standardization.
Ibarra, I.D. & Reinhart, C.F. (2009) DAYLIGHT FACTOR SIMULATIONS
HOW CLOSE DO SIMULATION BEGINNERS REALLY GET?
Building Simulation 2009, Eleventh International IBPSA Conference pp.
196203. Glasgow, Scotland,July 27-30.
Johnsen, K. & Christoffersen, J. (2008). Dagslys i rum og bygninger. (SBianvisning 219) Statens byggeforskningsinstitut, Aalborg Universitet.
Kbenhavn.
Labayrade, R., Jensen, H.W., Jensen, C., Lyon, U. De, Ecole, F., Publics, T.,
Etat, D., Gnie, D. & Audin, M. (2009) VALIDATION OF VELUX
DAYLIGHT VISUALIZER 2 AGAINST CIE 171: 2006 TEST CASES.
Building Simulation 2009, Eleventh International IBPSA Conference pp.
15061513.
Larson, G.W. & Shakespeare, R. (1998) Rendering with Radiance - The Art
and Science of Lighting Visualization. Morgan Kaufmann, San
Francisco, CA.
Mardaljevic, J. (1995) Validation of a lighting simulation program under real
sky conditions. Lighting Research and Technology, 27.
Reinhart, C.F. (2010) Tutorial on the Use of Daysim Simulations for
Sustainable Design.
Reinhart, C.F. & Andersen, M. (2006) Development and validation of a
Radiance model for a translucent panel. Energy and Buildings, 38,
890904.
Reinhart, C.F. & Walkenhorst, O. (2001) Validation of dynamic RADIANCEbased daylight simulations for a test office with external blinds. Energy
and Buildings, 33, 683697.
Watt, A. (2000) 3D Computer Graphics. Pearson Education Limited.

54

Experience has shown that results for the same

room, obtained from two daylight simulation programs can give different results. This can be due
to restrictions in the program itself and/or be due
to the skills of the persons setting up the models.
This is crucial as daylight calculations are used to
document that the demands and recommendations
to daylight levels outlined by building authorities.
The aim of the project was to obtain a better understanding of what daylight calculations show and
also to gain knowledge of how the different daylight simulation programs perform compared with
each other. Furthermore the aim was to provide
knowledge of how to build up the 3D models that
were to be daylight-analysed. It is an overall objective to enhance the quality of consultancy within
the field of daylighting with a greater knowledge of
the available daylight simulation programs and the
accuracy of the simulated.
The programs investigated are: Radiance, Daysim,
VELUX Daylight Visualizer, DIAlux, Ecotect, Ecotect/
Radiance, IESve, LightCalc and Relux.
The project is supported by

1st edition, 2013

ISBN 978-87-92739-49-0